Advances in Cancer Research, Volume 97

  • 43 54 1
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up

Advances in Cancer Research, Volume 97

Advances in CANCER RESEARCH Volume 97 This page intentionally left blank Advances in CANCER RESEARCH Volume 97

1,174 291 6MB

Pages 364 Page size 336 x 530.4 pts Year 2007

Report DMCA / Copyright

DOWNLOAD FILE

Recommend Papers

File loading please wait...
Citation preview

Advances in

CANCER RESEARCH

Volume 97

This page intentionally left blank

Advances in

CANCER RESEARCH

Volume 97

Edited by

George F. Vande Woude Van Andel Research Institute Grand Rapids, Michigan

George Klein Microbiology and Tumor Biology Center Karolinska Institute Stockholm, Sweden

AMSTERDAM • BOSTON • HEIDELBERG • LONDON NEW YORK • OXFORD • PARIS • SAN DIEGO SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO Academic Press is an imprint of Elsevier

Academic Press is an imprint of Elsevier 525 B Street, Suite 1900, San Diego, California 92101-4495, USA 84 Theobald’s Road, London WC1X 8RR, UK

This book is printed on acid-free paper. Copyright # 2007, Elsevier Inc. All Rights Reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the Publisher. The appearance of the code at the bottom of the first page of a chapter in this book indicates the Publisher’s consent that copies of the chapter may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc. (www.copyright.com), for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Copy fees for pre-2007 chapters are as shown on the title pages. If no fee code appears on the title page, the copy fee is the same as for current chapters. 0065-230X/2007 $35.00 Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone: (þ44) 1865 843830, fax: (þ44) 1865 853333, E-mail: [email protected]. You may also complete your request on-line via the Elsevier homepage (http://elsevier.com), by selecting ‘‘Support & Contact’’ then ‘‘Copyright and Permission’’ and then ‘‘Obtaining Permissions.’’ For information on all Elsevier Academic Press publications visit our Web site at www.books.elsevier.com ISBN-13: 978-0-12-006697-1 ISBN-10: 0-12-006697-1 PRINTED IN THE UNITED STATES OF AMERICA 07 08 09 10 9 8 7 6 5 4 3 2 1

Contents

Contributors to Volume 97 ix Dedication xiii

Structural Biology of the Tumor Suppressor p53 and Cancer-Associated Mutants Andreas C. Joerger and Alan R. Fersht I. II. III. IV. V. VI.

Introduction 2 The Domain Structure of Human p53 2 The Structure of the DNA-Binding Domain 4 Effects of Common Cancer Mutations 8 Rescuing Mutant p53 15 Concluding Remarks 19 References 19

Immunotherapy by Allogeneic Stem Cell Transplantation Olle Ringde´n I. II. III. IV. V. VI. VII. VIII.

Introduction 26 Graft-Versus-Host Disease 27 The Graft-Versus-Leukemia Effect 30 NK Cells 34 Early Detection of Relapse 35 The Graft-Versus-Cancer Effect 37 Mesenchymal Stem Cells 41 Future Directions 44 References 47

Mnt Takes Control as Key Regulator of the Myc/Max/Mxd Network Therese Wahlstro¨m and Marie Henriksson I. Myc: The Most Frequently Deregulated Oncogene in Human Tumors 62 II. Mnt: The Key Transcriptional Regulator of the Myc/Max/Mxd Network 66

v

vi

Contents

III. Concluding Remarks 75 References 76

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses George Miller, Ayman El-Guindy, Jill Countryman, Jianjiang Ye, and Lyn Gradoville I. Two Life Cycles of EBV and KSHV: Latency and Lytic Replication 82 II. Virally Encoded Lytic Cycle Activator Genes 84 III. Conclusions: Some Unsolved Mysteries About Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses 103 References 105

No Life Without Death Peter H. Krammer, Marcin Kamin´ski, Michael Kießling, and Karsten Gu¨low I. II. III. IV. V.

Introduction 111 Apoptosis in Life and Disease 112 The Apoptotic Machinery 113 The CD95/CD95L System 116 HIV and Apoptosis 126 References 129

Control of Apoptosis in Human Multiple Myeloma by Insulin-like Growth Factor I (IGF-I) Helena Jernberg-Wiklund and Kenneth Nilsson I. II. III. IV.

Selected Biological Properties of Human Multiple Myeloma 140 Human MM Models In Vitro and In Vivo 142 Targeting Anti-apoptosis and Proliferative Signals in Human MM 143 The Effect of Combinational Treatment with PPP on Human MM Cells Is Additive and Synergistic 154 References 159

c-MYC Impairs Immunogenicity of Human B Cells Martin Schlee, Marino Schuhmacher, Michael Ho¨lzel, Gerhard Laux, and Georg W. Bornkamm I. II. III. IV.

Introduction 168 Personal Perspective by G.W.B. 168 Taking Over the Work from Eva and George    Discussions 180 References 184

172

vii

Contents

Cancer Dormancy: Lessons from a B Cell Lymphoma and Adenocarcinoma of the Prostate Rosalia Rabinovsky, Jonathan W. Uhr, Ellen S. Vitetta, and Eitan Yefenof I. II. III. IV. V. VI.

Introduction 190 Scope of the Present Discussion 190 Clinical Studies 192 Experimental Dormancy of B Cell Lymphoma The Prostate Adenocarcinoma Model 195 Concluding Remarks 197 References 198

193

Therapeutic Targets of Multiple Angiogenic Factors for the Treatment of Cancer and Metastasis Yihai Cao and Qi Liu I. II. III. IV. V. VI.

Introduction 204 Tumor Angiogenesis 205 Angiogenesis Inhibitors 210 Lymphangiogenesis and Lymphatic Metastasis 215 Clinical Development of Antiangiogenic Drugs 216 Conclusions and Perspectives 217 References 219

Novel Three-Dimensional Organotypic Liver Bioreactor to Directly Visualize Early Events in Metastatic Progression Clayton Yates, Chistopher R. Shepard, Glenn Papworth, Ajit Dash, Donna Beer Stolz, Steven Tannenbaum, Linda Griffith, and Alan Wells I. II. III. IV. V.

Introduction 226 Bioreactors 230 Tumor Growth in the Bioreactor 233 Tumor–Hepatocyte Juxtapositioning 236 Future Studies 239 References 242

PDGF Receptors as Targets in Tumor Treatment ¨ stman and Carl-Henrik Heldin Arne O I. Molecular Biology of PDGF 248 II. Physiological Roles of PDGF 251 III. Roles of PDGF Receptors in Tumors

252

viii

Contents

IV. Clinical Studies 263 V. Future Perspectives 266 References 267

Extracellular Matrix, Nuclear and Chromatin Structure, and Gene Expression in Normal Tissues and Malignant Tumors: A Work in Progress Virginia A. Spencer, Ren Xu, and Mina J. Bissell I. II. III. IV.

Introduction 276 The ECM 277 ECM-Response DNA Elements 278 Potential Mechanisms for the Transcriptional Activation of ECM-Response DNA Elements 280 V. Potential Mechanisms Through Which ECM Influences the General Organization of Nuclear Factors and Overall Transcriptional Activity 284 VI. Advancing Toward a Deeper Understanding of the Malignant Phenotype VII. A 3D Reconstruction for the Future of Cancer Research 289 References 289

287

Targeted Cancer Therapy: Promise and Reality Shoshana Klein and Alexander Levitzki I. II. III. IV. V. VI. VII.

What Is Signal Transduction Therapy? 295 Types of Signaling Inhibitors 296 Signaling Networks 299 Target and Drug Evaluation Using Preclinical Tumor Models 300 How Successful Is Signal Transduction Therapy in the Clinic? 303 Using PK Receptors as Homing Molecules for Cancer Therapy 310 Conclusions 311 References 312

Restoration of Wild-Type p53 Function in Human Tumors: Strategies for Efficient Cancer Therapy Klas G. Wiman I. II. III. IV. V.

The Emergence of p53 as a Key Tumor Suppressor 321 p53 Biological Activity and Binding to DNA 322 Reactivation of Mutant p53 326 Virus-Based Therapeutic Strategies for Mutant p53-Carrying Tumors Concluding Remarks 331 References 333 Index

339

330

Contributors

Numbers in parentheses indicate the pages on which the authors’ contributions begin.

Mina J. Bissell, Life Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720 (275) Georg W. Bornkamm, Institute of Clinical Molecular Biology and Tumor Genetics, GSF-National Research Center for Environment and Health, D-81377 Mu¨nchen, Germany (167) Yihai Cao, Laboratory of Angiogenesis Research, Microbiology and Tumor Biology Center, Karolinska Institutet, 171 77 Stockholm, Sweden (203) Jill Countryman, Department of Molecular Biophysics and Biochemistry, Yale University School of Medicine, New Haven, Connecticut 06520 (81) Ajit Dash, Biological Engineering Division, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 (225) Ayman El-Guindy, Department of Molecular Biophysics and Biochemistry, Yale University School of Medicine, New Haven, Connecticut 06520 (81) Alan R. Fersht, MRC Centre for Protein Engineering, Cambridge CB2 2QH, United Kingdom (1) Karsten Gu¨low, Tumor Immunology Program D030, German Cancer Research Center (DKFZ), Im Neuenheimer Feld 280, 69120 Heidelberg, Germany (111) Lyn Gradoville, Department of Pediatrics, Yale University School of Medicine, New Haven, Connecticut 06520 (81) Linda Griffith, Biological Engineering Division, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 (225) Carl-Henrik Heldin, Ludwig Institute for Cancer Research, Uppsala University, SE-751 24 Uppsala, Sweden (247) Marie Henriksson, Department of Microbiology, Tumor, and Cell Biology (MTC), Karolinska Institutet, SE-171 77 Stockholm, Sweden (61) Michael Ho¨lzel, Institute of Clinical Molecular Biology and Tumor Genetics, GSF-National Research Center for Environment and Health, D-81377 Mu¨nchen, Germany (167) Helena Jernberg-Wiklund, Department of Genetics and Pathology, Rudbeck Laboratory, Uppsala University, SE-751 85 Uppsala, Sweden (139)

ix

x

Contributors

Andreas C. Joerger, MRC Centre for Protein Engineering, Cambridge CB2 2QH, United Kingdom (1) Marcin Kamin´ski, Tumor Immunology Program D030, German Cancer Research Center (DKFZ), Im Neuenheimer Feld 280, 69120 Heidelberg, Germany (111) Michael Kießling, Tumor Immunology Program D030, German Cancer Research Center (DKFZ), Im Neuenheimer Feld 280, 69120 Heidelberg, Germany (111) Shoshana Klein, Department of Biological Chemistry, The Alexander Silberman Institute of Life Sciences, The Hebrew University of Jerusalem, Jerusalem, Israel (295) Peter H. Krammer, Tumor Immunology Program D030, German Cancer Research Center (DKFZ), Im Neuenheimer Feld 280, 69120 Heidelberg, Germany (111) Gerhard Laux, Institute of Clinical Molecular Biology and Tumor Genetics, GSF-National Research Center for Environment and Health, D-81377 Mu¨nchen, Germany (167) Alexander Levitzki, Department of Biological Chemistry, The Alexander Silberman Institute of Life Sciences, The Hebrew University of Jerusalem, Jerusalem, Israel (295) Qi Liu, Shandong Provincial Hospital, Jinan, Shandong Province, People’s Republic of China (203) George Miller, Department of Pediatrics, Yale University School of Medicine, New Haven, Connecticut 06520; Department of Epidemiology and Public Health, Yale University School of Medicine, New Haven, Connecticut 06520 (81) Kenneth Nilsson, Department of Genetics and Pathology, Rudbeck Laboratory, Uppsala University, SE-751 85 Uppsala, Sweden (139) ¨ stman, Department of Pathology-Oncology, Cancer Center KaroArne O linska, Karolinska Institutet, R8:03, SE-171 76 Stockholm, Sweden (247) Glenn Papworth, Center for Biologic Imaging, Cell Biology, and Physiology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261 (225) Rosalia Rabinovsky, Lautenberg Center for General and Tumor Immunology, Hebrew University of Jerusalem, Jerusalem 91120, Israel (189) Olle Ringde´n, Division of Clinical Immunology, Karolinska Institutet, Karolinska University Hospital, Huddinge, SE-141 86 Stockholm, Sweden; and Center for Allogeneic Stem Cell Transplantation, Karolinska Institutet, Karolinska University Hospital, Huddinge, SE-141 86 Stockholm, Sweden (25) Martin Schlee, Institute of Clinical Molecular Biology and Tumor Genetics, GSF-National Research Center for Environment and Health, D-81377 Mu¨nchen, Germany (167)

Contributors

xi

Marino Schuhmacher, Institute of Clinical Molecular Biology and Tumor Genetics, GSF-National Research Center for Environment and Health, D-81377 Mu¨nchen, Germany; GPC Biotech-AG, D-82152 Martinsried, Germany (167) Chistopher R. Shepard, Department of Pathology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261 (225) Virginia A. Spencer, Life Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720 (275) Donna Beer Stolz, Department of Pathology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261; Center for Biologic Imaging, Cell Biology, and Physiology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261 (225) Steven Tannenbaum, Biological Engineering Division, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 (225) Jonathan W. Uhr, Cancer Immunobiology Center, University of Texas, Southwestern Medical Center at Dallas, Texas 75235 (189) Ellen S. Vitetta, Cancer Immunobiology Center, University of Texas, Southwestern Medical Center at Dallas, Texas 75235 (189) Therese Wahlstro¨m, Department of Microbiology, Tumor, and Cell Biology (MTC), Karolinska Institutet, SE-171 77 Stockholm, Sweden (61) Alan Wells, Department of Pathology, University of Pittsburgh and Pittsburgh VAMC, Pittsburgh, Pennsylvania 15261 (225) Klas G. Wiman, Department of Oncology-Pathology, Cancer Center Karolinska (CCK), Karolinska Institutet, SE-171 76 Stockholm, Sweden (321) Ren Xu, Life Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720 (275) Clayton Yates, Department of Pathology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261 (225) Jianjiang Ye, Department of Molecular Biophysics and Biochemistry, Yale University School of Medicine, New Haven, Connecticut 06520 (81) Eitan Yefenof, Lautenberg Center for General and Tumor Immunology, Hebrew University of Jerusalem, Jerusalem 91120, Israel (189)

This page intentionally left blank

Dedication

GEORGE VANDE WOUDE It is with great honor and pleasure that we dedicate this 97th volume of Advances in Cancer Research to celebrate the 80th birthdays of George and Eva Klein. Cancer biology and tumor immunology have flourished because of their many important discoveries and contributions and generations of students have grown to become leading members of the community of scientists because of their mentorship and tutelage. I remember my first visit

xiii

xiv

Dedication

to the Karolinska in 1972. The Klein’s laboratory was bustling with many bright students and colleagues signing up around the clock for time on their precious microscope to view EBNA nuclear fluorescence. George and Eva were away, but, everywhere in the laboratory, the excitement among its members and their enthusiasm for discovery was obvious. A decade later at meetings organized at Engedi, George and Eva brought international science and scientists to Israel that stimulated collaborations and helped establish contacts of new generations of scientists with colleagues around the world. On reflection, we realize that there are few areas of modern cancer biology that the contributions of George and Eva did not help to seed. We are all humbled by their dedication, vision, and compassion. Their symposium, “Molecular Oncology: From Bench to Bedside,” held at the Karolinska in June 2005, paid tribute to their many accomplishments and provided a wonderful opportunity for their friends, colleagues, and former students to honor them. In this commemorative volume of Advances in Cancer Research, we have all joined forces to dedicate this special volume to George and Eva as two very special scientists and two very special people of the world.

KLAS WIMAN As a post doc in the US in the early 1980s, I got a reprint request from George Klein. This was the beginning of a long‐lasting correspondence and friendship. I joined the Department of Tumor Biology a few years later and immediately felt at home in the multicultural and rather chaotic atmosphere. I am very grateful to the Kleins for giving me the opportunity to work and develop in their outstanding scientific environment and for many fruitful interactions and collaborations over the years. It is a great honor to contribute to this book dedicated to George and Eva.

OLLE RINGDE´N George and Eva Klein have contributed so much in research and, being a scientific grandchild, I feel much indebted to the atmosphere and spirit that they have created. I do feel honored to have the opportunity to give a lecture and to write a paper in tribute to George and Eva.

CARL‐HENRIK HELDIN AND ARNE O¨STMAN It is an honor to have the opportunity to contribute to this volume in honor of Eva and George Klein, who have made outstanding contributions to the field of cancer research. Their presence in Sweden has been of

Dedication

xv

immense importance for the development of cancer research in our country, and their pupils now have key roles in Swedish cancer research. Most impressive has been the intellectual curiosity of the Kleins and their ability to contribute to and stay updated with novel developments in the wide area of cancer research.

ALAN R. FERSHT One of us (A.R.F.) has had the personal acquaintance of Eva and George Klein only in very recent years. It has been my greatest of fortunes to have the close mentorship and friendship of giants of science such as Max Perutz, Aaron Klug, and Cesar Milstein. Their great characteristics of enthusiasm, love of science, incredible memories, curiosity and honesty, all combined with humanity and culture are clearly shared by Eva and George. Fortunately, such individuals inspire others with their fine qualities and so their virtues as well as their scientific legacies pass down the generations.

PETER H. KRAMMER We dedicate our chapter to Eva and George Klein on the occasion of their 80th birthday. Ad multos annos.

KENNETH NILSSON For almost 60 years, Eva and George Klein have been central figures in the cancer research community of Sweden and internationally. Generations of students and young guest researchers have been trained by Eva and George in their supervisory role first at the Department of Tumor Biology and later at the Microbiology and Tumor Biology Center of the Karolinska Institute. I was one of those students in the early 1970s and was privileged to collaborate with George during my years as a Ph.D. student in Uppsala, and to have Eva as the opponent of my Ph.D. thesis. A close collaboration continued during my postdoc years with both of them. Later in life, as a professor at Uppsala University, my relationship to the Kleins naturally changed. I have enjoyed having them as colleagues at a short distance, always available for interesting, stimulating discussions on scientific problems, philosophical questions, and more personal matters. During my years as chairman of the Scientific Committee of the Swedish Cancer Society, Eva and George by their wisdom and experience were extremely

xvi

Dedication

important as a sort of personal advisor coaching me gently when I needed guidance in difficult matters. I thank them for all this and wish them all the best for the future.

MARIE HENRIKSSON In 1984, I was finishing my studies for a Bachelor of Sciences at Uppsala University and our final course was in Immunology. By then, the T cell receptor was not yet cloned and our lecturer, Hans Wigzell, was transmitting his fascination for the immune system to us. He also discussed the work of George and Eva Klein at the Karolinska Institutet and these lectures made a strong impression on me. This was the first time I heard about the Myc oncogene and the Myc‐immunoglobulin translocation in Burkitt’s lymphoma. These seminars inspired me to get in contact with George and suddenly I found myself at the Department of Tumor Biology for the meeting with him. In the elevator, I met Gunnar Klein and I was wondering whether he was the son of George or not (I later found out that he was not). During the meeting George asked me when I could start and I heard myself saying in June, after the University semester has finished. Little did I know how this first meeting would change the path of my life. The former Department of Tumor Biology at the Karolinska Institutet was located in an ordinary redbrick building with the name in golden letters above the door as for all other departments at the campus, but I guess that the similarity stopped there. Once entering the door there was a staircase, the elevator, and the first floor corridor with dark blue doors that at first gave you an impression of being in a factory. People were running in and out of the doors with laboratory coats flapping, bottles of cells, pipettes, or manuscripts in their hands, laughing, arguing, and discussing in English, Spanish, and also in Swedish. The place was vibrating with energy. The department of George and Eva had many particular routines; individual meetings with front line scientists, participation in writing grants including the ones to NIH, student duty schemes, and a buzzing system with local telephones in each room and an individual morse signal for every person (this was long before the cellular phones). The student duty schemes meant that you during one or two weeks were responsible for either showing the slides during the seminars or to make the program for visiting scientists. Last but not least, the famous Friday letters where all students, and also many scientists, were summarizing their results, ideas, and reading of scientific literature. As with many important things in life, you either find too many words to express your feelings or you do not find any words at all. I do not have

Dedication

xvii

words for what George and Eva have meant for me both on a professional as well as personal level. I best can summarize it in the word inspiration. Inspiration that you can and should do all what you can do and also what you think that you cannot do. Professionally, it meant that I got the best research education one can imagine that led to a Ph.D. from the Karolinska Institutet. The core of the department was the belief that you should be able to explain complicated biological processes so that a 12‐year old (the very age of my oldest daughter now) can understand, that your knowledge rather than your title counts, always to look for solutions rather than for problems, to believe in oneself. Personally it meant that I met the people that have played and still play major roles in my life, my husband and my closest friends. I do hope that I can convey at least some of the atmosphere that shaped my student years to the next generation of scientists as represented by my Ph.D. students.

Structural Biology of the Tumor Suppressor p53 and Cancer‐Associated Mutants Andreas C. Joerger and Alan R. Fersht MRC Centre for Protein Engineering, Cambridge CB2 2QH, United Kingdom

I. Introduction II. The Domain Structure of Human p53 III. The Structure of the DNA‐Binding Domain A. p53C Has Evolved to Be Unstable B. Design of a Superstable Variant of p53C C. From Core Domain to the Full‐Length Protein IV. Effects of Common Cancer Mutations A. Structural Effects of Oncogenic Mutations V. Rescuing Mutant p53 A. Lessons from Second‐Site Suppressor Mutations B. p53C as a Drug Target VI. Concluding Remarks References

The tumor suppressor protein p53 is a transcription factor that plays a key role in the prevention of cancer development. In response to oncogenic or other stresses, the p53 protein is activated and regulates the expression of a variety of target genes, resulting in cell cycle arrest, senescence, or apoptosis. Mutation of the p53 gene is the most common genetic alteration in human cancer, affecting more than 50% of human tumors. Most of these mutations inactivate the DNA‐binding domain of the protein. In this chapter, we describe the structure of the wild‐type p53 protein and present structural and functional data that provide the molecular basis for understanding the effects of common cancer mutations. Further, we assess novel therapeutic strategies that aim to rescue the function of p53 cancer mutants. # 2007 Elsevier Inc.

ABBREVIATIONS p53C, p53 core domain; T‐p53C, p53 core domain containing the four point mutations M133L, V203A, N239Y, and N268D. Advances in CANCER RESEARCH Copyright 2007, Elsevier Inc. All rights reserved.

1

0065-230X/07 $35.00 DOI: 10.1016/S0065-230X(06)97001-8

2

Andreas C. Joerger and Alan R. Fersht

I. INTRODUCTION The tumor suppressor protein p53, “the guardian of the genome,” is a transcription factor that is at the center of several complex cellular networks and plays a key role in the prevention of cancer development. In response to oncogenic and other stresses, p53 is activated and causes up‐ or down‐regulation of a variety of target genes, resulting in cell‐cycle arrest, senescence, or apoptosis (Prives and Hall, 1999; Vogelstein et al., 2000; Vousden and Lu, 2002). Its activity is highly regulated by posttranslational modifications (Lavin and Gueven, 2006) and by a large number of signaling proteins (Braithwaite et al., 2006). A knowledge of the structural biology of p53 is important to understanding how the protein is inactivated by common cancer mutations. Such knowledge provides the framework for devising therapeutic strategies to develop pharmacologically active anticancer drugs. Here we review the structure of p53, in particular of the DNA‐binding core domain, the structural and functional consequences of mutation, and therapeutic strategies for mutant p53 rescue.

II. THE DOMAIN STRUCTURE OF HUMAN p53 The 393‐residue p53 tumor suppressor protein exists in a dynamic equilibrium to form homotetramers (Sakaguchi et al., 1997). Each chain comprises several functional domains (Courtois et al., 2004; Vousden and Lu, 2002) (Fig. 1A). The N‐terminal part of the protein consists of the transactivation domain (residues 1–63) followed by a proline‐rich region (64–92). The transactivation domain is involved in protein–protein interactions with regulatory proteins, such as MDM2 (residues 19–26 of p53), which causes ubiquitination of p53 resulting in proteasomal digestion (Michael and Oren, 2002, 2003; Momand et al., 2000), and p300/CBP, which regulates p53 function by causing acetylation of residues in the C‐terminal region of p53 (Grossman, 2001; Gu and Roeder, 1997). The proline‐rich region (residues 64–92) is thought to have a regulatory role (Mu¨ller‐Tiemann et al., 1998; Walker and Levine, 1996). The central (core) domain (p53 core domain, p53C) is responsible for binding to target DNA (residues 94–292). It binds specifically to a double‐stranded DNA consensus site containing two copies of the 10‐base pair “half‐site” motif 50 ‐PuPuPuC(A/T)(T/A) GPyPyPy‐30 (Pu ¼ A/G, Py ¼ T/C) separated by up to 13 base pairs (el‐Deiry et al., 1992). p53C monomers bind this target DNA to give a 4:1 complex (Balagurumoorthy et al., 1995; Weinberg et al., 2004b). Most p53 cancer mutations are located in the DNA‐binding domain (Olivier et al., 2002) (cf. TP53 mutation database of the International Agency for Research on Cancer at www‐p53.iarc.fr). The C‐terminal part of p53 contains the tetramerization domain (residues 326–355) (Clore et al., 1994) and the

3

Structural Biology of the Tumor Suppressor p53 and Cancer‐Associated Mutants

A Trans- Proline-rich region activation 1

Tet C-terminal domain domain

DNA-binding (core) domain

292

94

64

B

326

355

393

C

R248 Zn G245

H179

R280 R273

K120

Zn C242

C C176

H168

R249

R248

R175

G245 R282 R249

N

F270 V143

H168 E171

Y220

Fig. 1 Structure of human p53. (A) Schematic view of the domain structure of human p53. The main functional domains are the transactivation domain at the N‐terminus, followed by a proline‐rich region, the central DNA‐binding core domain, the tetramerization domain (Tet domain), and the C‐terminal negative regulatory domain (see text for further details). (B) Cartoon representation of the structure of the DNA‐binding (core) domain in complex with consensus DNA (PDB ID code 1TSR, chain B) (Cho et al., 1994). The six residues that are most frequently mutated in human cancer are shown in orange. Blue spheres indicate the mutation sites in the superstable quadruple mutant T‐p53C (Joerger et al., 2004; Nikolova et al., 1998). (C) Close‐up view of the DNA‐binding domain of p53, showing the zinc‐binding site, and the L3 loop region, which is anchored to the minor groove of target DNA via Arg248. The conformation of the L3 loop is stabilized by the side chain of Arg249 (stabilizing interactions are highlighted by dotted lines). The structural figures were generated using MOLSCRIPT (Kraulis, 1991) and RASTER3D (Merritt and Bacon, 1997).

negative regulatory domain at the extreme C‐terminus (363–393), which contains phosphorylation and acetylation sites and regulates the DNA‐ binding activity of p53 (Friedler et al., 2005a; Prives and Manley, 2001; Weinberg et al., 2004a). In addition, a nuclear localization signal is located between the core domain and tetramerization domain (residues 302–322)

4

Andreas C. Joerger and Alan R. Fersht

(Shaulsky et al., 1990), and a highly conserved leucine‐rich nuclear export signal is found within the tetramerization domain (residues 340–350) (Stommel et al., 1999). Several crystal structures of the DNA‐binding core domain in complex with target DNA (Cho et al., 1994; Kitayner et al., 2006) and the structures of various mutants in their DNA‐free form have been solved (Joerger et al., 2004, 2005, 2006). In addition, the solution structure of the core domain has been determined by NMR spectroscopy (Canadillas et al., 2006). The structure of the tetramerization domain has also been solved by X‐ray crystallography (Jeffrey et al., 1995; Mittl et al., 1998) and NMR spectroscopy (Clore et al., 1994, 1995). Four chains of this domain, consisting of a short ‐strand followed by an ‐helix, are assembled in such a way that they form a tetramer with D2 symmetry. In contrast, the N‐terminal domain is natively unfolded in isolation (Dawson et al., 2003; Lee et al., 2000), apart from regions that show nascent helix or turn formation (Lee et al., 2000). Interestingly, a peptide derived from one of these regions (residues 15–29) forms a full amphipathic ‐helix on binding to a hydrophobic cleft in the N‐terminal domain of MDM2, as revealed by X‐ray crystallography (Kussie et al., 1996). The C‐terminal negative regulatory domain is also unstructured, but as in the case of the N‐terminal domain, binding to other proteins appears to induce helix formation. NMR spectroscopy has revealed that a peptide derived from the C‐terminal region of p53 (residues 367–388), which has no regular structure in its native form, becomes helical on binding to S100B (Rustandi et al., 2000). Relatively little is known about the assembly of the different domains of p53 in the full‐length tetramer.

III. THE STRUCTURE OF THE DNA‐BINDING DOMAIN ˚ In 1994, the structure of p53C in complex with consensus DNA at 2.2‐A resolution was published (Cho et al., 1994). This structure contained three molecules in the asymmetric unit, with two molecules bound to DNA, one of which contacted the consensus sequence directly, and the third molecule effectively representing the DNA‐free form of the core domain. Overall, the structure of the three molecules was very similar, apart from some loop regions, indicating that only small conformational changes occur on DNA binding. The core domain consists of a central ‐sandwich that provides the basic scaffold for the DNA‐binding surface (Fig. 1B). Two antiparallel ‐sheets of four and five twisted ‐strands form a compact barrel‐like structure reminiscent of the immunoglobulin fold. The DNA‐binding surface at one end of the ‐sandwich consists of two large loops, L2 (residues 164–194, interrupted by the short ‐helix H1) and L3 (residues 237–250),

Structural Biology of the Tumor Suppressor p53 and Cancer‐Associated Mutants

5

and a loop‐sheet‐helix motif. The conformation of the two large loops is in part stabilized by a zinc ion, which is tetrahedrally coordinated by a histidine and three cysteine side chains (Cys176, His179, Cys238, and Cys242). In the complex with consensus DNA, residues of these structural elements make extensive contacts with the DNA. The L3 loop is anchored to the minor groove of the target DNA via Arg248, which interacts with the negatively charged DNA backbone (Fig. 1C). Parts of the loop‐sheet‐helix motif make contact with the major groove, which involves interactions of Lys120 and Arg280 with bases of the pentameric consensus sequence. The crystal structure of the core domain provided a framework for understanding how common cancer‐associated mutations inactivate p53. The six most frequently mutated residues (Arg175, Gly245, Arg248, Arg249, Arg273, and Arg282), so‐called mutation hotspots, are located in the DNA‐ binding surface (www‐p53.iarc.fr). On the basis of their exact location in the structure, these mutations were divided into two classes. The first class, the “contact” mutations, such as R248Q and R273H, directly remove a residue that contacts DNA. The second class, the “structural mutations,” such as R175H, G245S, R249S, and R282W, were thought to compromise the structural integrity of the DNA‐binding surface and thus affect DNA binding indirectly (Fig. 1B). More recently, high‐resolution crystal structures of human p53C in complex with double‐stranded DNA dodecamers comprising different half‐site motifs have been solved (Kitayner et al., 2006). These structures provided important novel insights into DNA recognition by p53 tetramers. They elucidate the core–core domain interactions on binding as a tetramer and the structural basis of how p53 discriminates between different response elements. In all four structures reported—for one DNA construct, the structure was solved in two different crystal forms—two core domains are bound to a DNA half‐site, and two of these dimers form a tetramer via both protein–protein interactions and base‐pair stacking. Thus, they mimic the scenario of four core domains binding to two half‐sites on a continuous strand of DNA, separated by a two base‐pair linker. The two core domains that bind to a DNA half‐site are connected via a protein–protein interface with twofold symmetry (Fig. 2). This symmetrical core–core domain interface comprises a network of hydrophobic and partly water‐mediated polar interactions of conserved residues from the L3 loop and helix H1 at the periphery of the DNA‐binding surface, which is consistent with earlier studies in solution by NMR (Klein et al., 2001; Rippin et al., 2002b). With a total buried surface of about 600 A˚2, the interface is relatively small. The core domain dimer, however, is substantially stabilized by contacts with the DNA half‐site. A similar core domain dimer on binding to a DNA half‐site has been observed for p53C from mouse cross‐linked to decameric target DNA (Ho et al., 2006).

6

Andreas C. Joerger and Alan R. Fersht

H2

H2

L3 L3 Zn

Zn H1 H1

Fig. 2 Symmetrical dimer of two p53 core domains bound to a DNA half‐site (PDB ID code 2AC0, molecules C and D) (Kitayner et al., 2006). The core–core domain interface is made by residues from helix H1 and the L3 loop and is stabilized by contacts with the DNA half‐site. The figure was generated using MOLSCRIPT (Kraulis, 1991) and RASTER3D (Merritt and Bacon, 1997).

The DNA‐core domain contact geometry shows small but distinct differences depending on the sequence, concomitant with altered affinities (Kitayner et al., 2006). In all structures, Arg248 anchors the L3 loop to the minor groove, although the conformation of the arginine varies substantially. It ranges from fully extended to folded, making either direct or water‐mediated contact with the DNA backbone. Arg273 binds to the phosphate backbone in the central region of the half‐site. The contacts with the major groove reveal invariant contacts of Arg280 with the conserved G base. In contrast, Lys120, Ala276, and Cys277 show changing interaction patterns when the central base pair of the pentameric quarter‐site is changed, thus accounting for differential binding affinities. Interestingly, parts of the DNA‐binding surface and the core–core domain interface in the DNA‐bound form overlap with the docking sites for proteins of various functions, such as ASPP2, which stimulates p53‐induced apoptosis. The L3 loop forms the central part of the interface with the 53BP2 region of the ASPP2 protein, as revealed by the crystal structure of the p53C‐53BP2 complex (Gorina and Pavletich, 1996). Similarly, structural studies have shown that the L3 loop is part of the interface with 53BP1 (Derbyshire et al., 2002; Joo et al., 2002), a protein that is implicated in the response to DNA damage (DiTullio et al., 2002). It appears that the DNA‐ binding surface is a highly promiscuous binding site that interacts with a number of different proteins, another example being Rad51, which plays a role in homologous recombination (Friedler et al., 2005b).

Structural Biology of the Tumor Suppressor p53 and Cancer‐Associated Mutants

7

A. p53C Has Evolved to Be Unstable Equilibrium‐unfolding studies by urea denaturation have shown that p53C has a relatively low thermodynamic stability and is only marginally stable at body temperature (Bullock et al., 2000). The stability of full‐length p53 is dictated by its core domain, as measured by differential scanning calorimetry (Ang et al., 2006). The solution structure of p53C, which has been solved by NMR spectroscopy using state‐of‐the‐art‐labeling techniques, has provided insights into the molecular basis for its instability (Canadillas et al., 2006). While the overall structure in solution is similar to the crystal structure, the solution structure reveals a high degree of flexibility, not only for surface loops but also for particular buried residues. Most importantly, the analysis of the hydrogen‐bond patterns of polar groups revealed several buried hydroxyl and sulfhydryl groups that form suboptimal hydrogen‐bond networks. One such buried pair is formed by Tyr236 and Thr253. In the thermodynamically more stable orthologues p63 and p73, the equivalent residues are replaced by phenylalanine and isoleucine. Intriguingly, introduction of the corresponding mutations into the p53 scaffold does indeed increase the stability of the core domain, and the double mutant Y236F/T253I is stabilized by 1.6 kcal/mol (Canadillas et al., 2006). These observations strongly suggest that p53C has evolved to be dynamic and only marginally stable at physiological conditions, presumably to facilitate regulation of its cellular levels and to allow for enough conformational flexibility to fulfill its multiple cellular functions. This assumption is also supported by the structure of the p53 homologue from Caenorhabditis elegans, which shows that there have been considerable changes in the DNA‐binding domain during evolution, for example loss of secondary structure in functionally important loops (Huyen et al., 2004). Molecular dynamics simulations also suggest an elevated degree of structural fluctuations in the human variant (Pan et al., 2006).

B. Design of a Superstable Variant of p53C Accurate biophysical measurements on p53 are in many cases very difficult to achieve because of the low intrinsic stability of the protein and associated aggregation tendencies. A pseudo wild type with improved stability, which retains the structural and functional characteristics of the wild type, would solve this problem. By semirational design, Nikolova et al. (1998) developed a superstable variant of p53C (T‐p53C), which contains four point mutations (M133L, V203A, N239Y, and N268D) that stabilize the core domain by 2.6 kcal/mol. These mutations are either naturally occurring in p53 from different species or second‐site suppressor mutations that have been reported

8

Andreas C. Joerger and Alan R. Fersht

to reverse the effects of common cancer mutations (Brachmann et al., 1998). Interestingly, the two mutations that contribute most to the stability increase (N239Y and N268D) were also found in a p53 variant selected by in vitro evolution of thermostable p53 (Matsumura and Ellington, 1999). Crystallographic studies have shown that T‐p53C has virtually the same structure as the wild type, apart from the mutated side chains, which confer additional stability (Joerger et al., 2004); for example by creating an energetically more favorable hydrogen‐bond pattern in the case of the N268D mutation (Fig. 3A). T‐p53C has wild‐type‐like functional properties (Joerger et al., 2005; Nikolova et al., 1998), and its full‐length version is active in human cell lines (Sebastian Mayer and A.R.F., unpublished data). The mutations increase the melting temperature of full‐length p53 and its mutants by about 6  C, making them more amenable to extended study (Ang et al., 2006).

C. From Core Domain to the Full‐Length Protein The structural organization of full‐length p53 is only poorly understood because of the lack of detailed structural information. Full‐length p53 is aggregation‐prone and has not been successfully crystallized as a consequence of its extended natively unfolded regions. With a molecular weight of about 170 kDa, the p53 tetramer is also well above the limit of conventional NMR spectroscopy. Using the stabilized quadruple mutant T‐p53, it was possible to acquire NMR spectra on the full‐length protein and shorter domain constructs (Veprintsev et al., 2006). Overall, the NMR spectrum (15N,1H transverse relaxation‐optimized spectroscopy) of full‐length p53 was similar to the sum of the spectra of the individual domains in isolation. Comparison of these spectra, however, revealed a defined pattern of differences in chemical shifts in surface regions of the core domain, indicating a possible role of these regions in domain interactions. In future, the strategy of combining NMR spectra of various domain constructs with modeling and mutagenesis experiments could be further extended in order to systematically map the interactions of individual domains in both the DNA‐free and DNA‐bound forms of p53. Such an approach may eventually lead to a more comprehensive structural model of tetrameric full‐length p53 and the domain motions that occur on binding to DNA or signaling proteins.

IV. EFFECTS OF COMMON CANCER MUTATIONS p53 mutations often change the spectrum and expression of p53 target genes, resulting in a large number of distinct phenotypes (Menendez et al., 2006; Resnick and Inga, 2003). More importantly, a clinical study of breast

Structural Biology of the Tumor Suppressor p53 and Cancer‐Associated Mutants

A

9

B

DNA F270 N/D268

V143 F109

D281

*273 S240 L111

K132 F113

F134

V274 V272

C

L3 loop

L3 loop

L2 loop

L2 loop Zn

168 249

Zn

168 249

Fig. 3 Crystal structures of p53 core domain mutants. (A) View of the N268D mutation site in the structure of T‐p53C (PDB ID code 1UOL, chain A; yellow) superimposed on wild‐type core domain (PDB ID code 1TSR, chain A; light gray). (B) Structure of T‐p53C‐R273H (PDB ID code 2BIM, chain A; yellow) superimposed on the structures of T‐p53C (PDB ID code 1UOL, chain A; green), and DNA‐bound wild type (PDB ID code 1TSR, chain B; light gray). (C) Stereo view of a superposition of the C traces of the structures of T‐p53C (PDB ID code 1UOL, chain A; cyan), T‐p53C‐H168R (PDB ID code 2BIN; red), and T‐p53C‐R249S (PDB ID code 2BIO; yellow). Small spheres indicate the start and end of missing regions in the model of the polypeptide chain. The figures were generated using MOLSCRIPT (Kraulis, 1991) and RASTER3D (Merritt and Bacon, 1997).

cancer patients indicates that the type and location of a cancer mutation dictate the prognosis and the response to drug treatment (Olivier et al., 2006). In order to understand the effects of mutation on the cellular function of p53 and their role in cancer development, it is paramount to elucidate their effects on the global folding state of the protein as well as the finer

10

Andreas C. Joerger and Alan R. Fersht

details of mutation‐induced structural changes. The overall thermodynamic stability will determine whether a mutant is folded under physiological conditions and hence functional. The functional properties in the folded state, for example the affinity for various p53 target genes, are governed by the exact nature of the structural changes. Systematic urea denaturation and DNA‐binding studies show that common cancer mutations can be subdivided into different classes, depending on their effects on stability and DNA binding (Bullock et al., 1997, 2000) (Fig. 4). The DNA‐contact mutation R273H, for instance, has virtually no effect on the stability of the protein but dramatically impairs DNA binding due to the loss of a crucial DNA contact. Other hotspot mutations, such as G245S and R249S in the L3 loop, destabilize the core domain by 2 kcal/mol, with varying effects on DNA binding, whereas many cancer mutations substantially destabilize the core domain by 3 kcal/mol. Because of the low instrinsic thermodynamic stability of the wild‐type core domain, such a degree of destabilization has dramatic consequences for the folding state of these mutants within the cell, and it is estimated that they are largely unfolded under physiological

Wild-type DNA-binding affinity at 20 ⬚C (%)

100

Wild type DNA contact DNA region Zinc region b-Sandwich

R282W

80

Wild type

V157F V143A

b-Sandwich

60

F270C

40

I232T

P151S

Y220C

G245S

I255F L145Q

I195T

F134L

20

Zinc region M237I R175H C242S

0 −2

−1

DNA contact R249S

R248Q

0 1 H2 O ∆G D-N at 37 ⬚C (kcal/mol)

R273H

2

3

Fig. 4 The different classes of p53 cancer mutants grouped according to thermodynamic stability and DNA‐binding properties. Mutant phenotypes correlate with the site of mutation as shown by a plot of stability (estimated at 37  C) against DNA‐binding affinity at 20  C, a temperature at which all mutants are folded. A free energy of unfolding in water, G, of 0 kcal/mol (shown by a dashed line) corresponds to 50% denatured protein. Data taken from Bullock et al. (2000).

Structural Biology of the Tumor Suppressor p53 and Cancer‐Associated Mutants

11

conditions. One class of these highly destabilizing mutations either directly affects residues coordinating the zinc ion (C242S) or perturbs the zinc‐ binding site (R175H), resulting in dramatic loss of DNA‐binding activity. All tested cancer mutations in the ‐sandwich region of the protein have a highly destabilizing effect, ranging from 3.0 to 4.5 kcal/mol (e.g., V143A on ‐strand S3, Y220C at the beginning of the S7‐S8 loop, and F270C on ‐strand S10). Most interestingly, these mutants retain DNA‐binding activity at subphysiological temperature, at which they are largely folded. Similar effects on stability and DNA binding have been found for structural mutations in the loop‐sheet‐helix motif (F134L and R282W) (Bullock et al., 2000). These differences in the thermodynamic stability of p53C (i.e., the G for the denaturation of wild type compared with that of a mutant) are not to be confused with the cellular levels of p53, an increase in which is often referred to as “stabilization” in p53‐related literature. The thermodynamic stability of a protein determines the folding–unfolding equilibrium and hence the amount of functional, folded protein under given conditions. The cellular levels of the protein, however, also depend on the rates of synthesis and degradation. In the case of p53, mutation often results in an accumulation of unfolded protein as a result of impaired MDM2‐dependent degradation.

A. Structural Effects of Oncogenic Mutations NMR studies, analyzing changes in the chemical shifts in the spectra of mutants compared to wild type, provided initial, albeit qualitative, information that common p53 cancer mutants exhibit characteristic structural changes (Wong et al., 1999). Rational drug design strategies aimed at rescuing the function of mutant p53 require structural information of high resolution. Structural studies of highly destabilized p53 mutants, however, are hampered because of the low thermodynamic stability of the protein, making them difficult to handle. We circumvented this problem by introducing common cancer mutations into the superstable quadruple mutant T‐p53C. The validity of this approach is reinforced by the fact that all structurally studied cancer mutations have the same relative effect on the thermodynamic stability of T‐p53C and p53C (Ang et al., 2006; Joerger et al., 2005). Using this stabilized variant, we were able to solve high‐ resolution crystal structures of a number of cancer mutants at a resolution ranging from 1.6 to 1.9 A˚. These structures cover a broad spectrum of mutants and comprise DNA‐contact mutants, diverse structural mutants of the DNA‐binding region, as well as different types of ‐sandwich mutants. An intriguing picture emerged from these studies, showing that common cancer mutations cause distinct local structural changes, while the

12

Andreas C. Joerger and Alan R. Fersht

overall structural scaffold is conserved. A variety of structural responses to mutation were found: merely the removal of a critical DNA‐contact residue, distortions to the DNA‐binding surface, creation of internal hydrophobic cavities, creation of solvent‐accessible crevices at the surface, and conformational changes in regions that are essential for core–core domain interactions on DNA binding (Joerger et al., 2005, 2006).

1. DNA‐CONTACT MUTATIONS Crystallographic studies confirm that R273H and R273C are classic contact mutations (Joerger et al., 2005, 2006). These mutations simply replace Arg273, which makes major contacts with the backbone of target DNA, by a histidine or cysteine without inducing significant structural perturbations. Overall, the high‐resolution structures of T‐p53C, T‐p53C‐ R273H, and T‐p53C‐R273C are virtually identical. Even residues that are in direct contact with the side chain of Arg273 in the wild type, such as Asp281 and Phe134, show only minor structural shifts in T‐p53C‐R273C and T‐p53C‐R273H (Fig. 3B). This preservation of the overall architecture of the DNA‐binding surface is consistent with mammalian cell‐based studies in which residual transactivation function is observed for the mutant R273H (Chen et al., 1993; Chumakov et al., 1993; Pietenpol et al., 1994). In vitro, the R273H mutation reduces the binding of full‐length p53 to gadd45 DNA by about 700–1000 times, although the residual binding can be largely attributed to nonspecific binding to the C‐terminus (Ang et al., 2006). The residual DNA‐binding activity, however, appears to be too weak for normal transactivation of p53 target genes (Ang et al., 2006). The mutant structures also rationalize observations that mixed tetramers of the R273H mutant with two or more wild‐type monomers are transcriptionally active (Chan et al., 2004).

2. STRUCTURAL MUTATIONS IN THE L3 LOOP The G245S and R249S mutations are located in the L3 loop, which is essential for DNA binding. This loop directly anchors the core domain to the minor groove of target DNA via Arg248 and forms an integral part of the protein–protein interface between the two core domains binding a DNA half‐site (Kitayner et al., 2006). Arg249 is involved in a number of interactions that stabilize the conformation of the L3 loop (Fig. 1C). In T‐p53C‐ R249S, these interactions are lost, and the L3 loop undergoes a substantial conformational change involving a peculiar rearrangement of two methionines (Joerger et al., 2005). It shows a high degree of flexibility and adopts a nonnative conformation, which is no longer compatible with effective DNA binding as both the DNA‐contact residue Arg248 and residues forming the

Structural Biology of the Tumor Suppressor p53 and Cancer‐Associated Mutants

13

core–core domain interface on DNA binding are affected (Fig. 3C). Compared to R249S, the G245S hotspot mutation has only a moderate effect on both the thermodynamic stability of the protein and its DNA‐binding activity (Ang et al., 2006; Bullock et al., 2000). In accordance with the biophysical data, the structural perturbations induced by G245S are smaller than those for R249S. In T‐p53C‐G245S, the hydroxyl group of Ser245 points toward the zinc‐coordination sphere, displaces a structural water molecule, and induces small structural changes in its immediate environment to accommodate the larger side chain (Joerger et al., 2006). Most notably, there is a flip of the peptide bond between Met243 and Gly244, and the adjacent Pro177 has shifted by about 1 A˚. These are key residues forming the core–core domain interface in the DNA‐bound form. Hence, it is not surprising that the affinity of full‐length T‐p53‐G245S for gadd45 DNA is about 15‐fold reduced (Ang et al., 2006). The structures of T‐p53C‐ G245S and T‐p53C‐R249S also rationalize observations that these mutants do not bind to the 53BP2 region of the apoptosis‐stimulating protein ASPP2 in their folded state, unlike wild‐type core domain and some of its mutants (Tidow et al., 2006). In both cases, the mutation‐induced structural changes are located in the L3 loop, which has been shown to form an essential part of the p53C‐53BP2 interface (Gorina and Pavletich, 1996).

3. HIGHLY DESTABILIZING MUTATIONS IN THE DNA‐BINDING SURFACE The highly destabilizing R282W hotspot mutation is located in the C‐terminal helix of the core domain, a part of the loop‐sheet‐helix motif, which docks to the major groove of target DNA. In the wild type, the side chain of Arg282 forms several stabilizing interactions at the interface of the different structural elements that form the loop‐sheet‐helix motif (cf. Fig. 1B). In T‐p53C‐R282W, introduction of the tryptophan compromises the structural integrity of this motif (Joerger et al., 2006). Stabilizing wild‐type interactions are lost, and the benzene moiety of the tryptophan side chain is solvent exposed, accounting for the observed stability loss. Because of steric hindrance, the L1 loop, which contains the DNA‐contact residue Lys120, is no longer packed against the core of the protein but is highly disordered, as evidenced by the lack of conclusive electron density for parts of this loop. Apart from the L1 loop, the structural integrity of the DNA‐binding surface is not significantly compromised, which explains why the R282W mutant shows binding to gadd45 DNA in its folded state at below body temperature (Bullock et al., 2000). The H168R mutation is much less frequent than R282W but has a similar effect on the thermodynamic stability of the protein, that is, destabilization

14

Andreas C. Joerger and Alan R. Fersht

of about 3 kcal/mol (Bullock et al., 2000; Joerger et al., 2005; Nikolova et al., 2000). It is located in the ‐turn region of the L2 loop at the periphery of the DNA‐binding surface. Replacing His168 by an arginine induces a substantial structural distortion at the mutation site, as revealed by the structure of T‐p53C‐H168R. Several residues, including the mutated residue, were not detected in the crystal structure because they are highly disordered (Joerger et al., 2005). As found for the other structural cancer mutants, the overall fold of the core domain is conserved (Fig. 3C).

4. ‐SANDWICH MUTATIONS As mentioned above, cancer mutations in the ‐sandwich region of the core domain generally result in a high stability loss, causing the protein to unfold at body temperature. At subphysiological temperature, however, they seem to retain wild‐type‐like functional properties. The classic example of a temperature‐sensitive ‐sandwich mutant is V143A. The mutant is inactive at body temperature in both yeast and mammalian cell lines but retains specific transactivating ability at subphysiological temperatures (Di Como and Prives, 1998; Zhan et al., 2001; Zhang et al., 1994). A similar temperature‐sensitive behavior was observed for the Y220C mutant (Di Como and Prives, 1998). A study has identified a large number of temperature‐sensitive p53 mutants from a comprehensive missense‐mutation library (Shiraishi et al., 2004). According to this study, the majority of p53 mutations with associated temperature‐sensitive phenotypes are located in the ‐sandwich region of the protein and are predominantly “large‐to‐small” substitutions. The side chains of Val143 and Phe270 are facing each other and form an integral part of the hydrophobic core of the ‐sandwich (Fig. 1B). Most interestingly, the V143A and F270L cancer mutations leave the structure of the core domain virtually unchanged, apart from the mutated side chain, as revealed by the structures of T‐p53C‐V143A and T‐p53C‐F270L (Joerger et al., 2006). Truncation of both hydrophobic side chains simply creates internal cavities, while the overall structure of the core domain is perfectly conserved. There is no collapse of the structure, but a high loss of thermodynamic stability is observed as a result of mutation. Even though the created cavities at the mutation site are large enough to accommodate a water molecule, no buried water was detected in the crystal structure. This is to be expected, given the hydrophobic nature of these cavities. The Y220C mutation is located at the far end of the ‐sandwich, at the beginning of the S7‐S8 loop (Fig. 1B). According to the latest version of the TP53 mutation database of the International Agency for Research on Cancer (release R10 at www‐p53.iarc.fr), this mutation is the most common cancer mutation outside the DNA‐binding surface and occurs almost as frequently in human tumors as the hotspot mutation G245S (263 and 343 cases,

Structural Biology of the Tumor Suppressor p53 and Cancer‐Associated Mutants

15

respectively). The 1.65‐A˚ crystal structure of T‐p53C‐Y220C shows that the mutation creates a solvent accessible crevice, connecting two rather shallow preexisting surface clefts to form a long extended crevice (Joerger et al., 2006). The overall structure of the core domain is conserved, and, more importantly, there are no structural perturbations in functionally important surface regions, be it the DNA‐binding surface, docking sites for signaling proteins, or interfaces known to be involved in domain interactions in the p53 tetramer on DNA binding. These findings, in combination with the observed stability loss on mutation, provide the molecular basis for the reported temperature‐sensitive phenotypes.

V. RESCUING MUTANT p53 Novel therapeutic strategies aim at rescuing the function of mutant p53. Most p53 cancer mutants are inactivated as a result of a loss of thermodynamic stability of the core domain (Bullock et al., 2000). In addition, they are also kinetically unstable and have a much shorter half‐life (t1/2) of unfolding than the wild type (Friedler et al., 2003). Stabilization of p53C is, therefore, an attractive target of novel therapeutic anticancer strategies. In principle, stabilization can be achieved with small molecule drugs that stabilize p53 by binding to it. Any ligand that binds to the folded state of p53 but not the denatured state will stabilize the protein by shifting the folding equilibrium toward the folded state. To be effective in the cellular context, the drug has to act as a chemical chaperone, that is, it has to bind to p53 immediately after biosynthesis and keep it in its folded state until it is translocated to the nucleus where it binds target DNA to trigger the cellular response. A number of p53‐activating or, more generally, apoptosis‐stimulating compounds were identified from chemical libraries using protein‐ and cell‐ based screening assays (reviewed by Bykov et al., 2003; Wiman, 2006). In many cases, the exact mechanism of activation is not clear, and most of these compounds do not directly interact with p53 (Krajewski et al., 2005; Rippin et al., 2002a; Wang et al., 2003). The putative p53‐stabilizing compound CP‐31398, for example, was reported to reactivate the tumor suppressor function of mutant p53 (Foster et al., 1999). It does not bind to the core domain in vitro but intercalates with DNA (Rippin et al., 2002a). Subsequent studies in vivo reported that CP‐31398 blocks the p53‐degradation pathway by inhibiting ubiquitination of p53 without blocking the association between p53 and MDM2 (Wang et al., 2003). A study controversially found that the CP‐31398 compound is highly toxic to some human cell lines at concentrations where the original experiments were performed, and no reactivation of mutant p53 could be detected in vivo (Tanner and Barberis, 2004).

16

Andreas C. Joerger and Alan R. Fersht

A. Lessons from Second‐Site Suppressor Mutations Many valuable lessons have been learned from studies on second‐site suppressor mutations, which prove the principle that function can be restored in mutant p53 (Baroni et al., 2004; Brachmann et al., 1998). Interestingly, some of these second‐site suppressor mutations rescue a whole subset of cancer mutants, whereas others are rather specific. Two of the second‐site suppressor mutations, N239Y and N268D, have been shown to act as “global stability” suppressors (Nikolova et al., 2000). They stabilize the core domain by 1.4 and 1.2 kcal/mol, respectively, and rescue cancer mutants by compensating for the stability loss caused by the oncogenic mutation. The structural basis of these stabilizing interactions has been revealed by the 1.9‐A˚ crystal structure of the superstable quadruple mutant T‐p53C (Joerger et al., 2004), which contains both second‐site suppressor mutations (see above). Some cancer mutants are not rescued by these “global stability” suppressors but require specific second‐site suppressor mutations. The R249S hotspot mutation is such an example. This mutation not only reduces the thermodynamic stability of the protein by 2 kcal/mol but also induces distinct structural changes in the DNA‐binding surface, as described above. Various rescue combinations were found, but all of them had the H168R mutation in common (Baroni et al., 2004; Brachmann et al., 1998). Interestingly, this mutation has also been found in tumors. Individually, the R249S and H168R mutations are oncogenic, reduce the thermodynamic stability of the protein, and induce structural perturbations in or near the DNA‐binding surface. If they are combined, however, the structural perturbations are largely reversed, and Arg168 mimics the structural role of Arg249 in the wild type (Joerger et al., 2005) (Fig. 5A). As a consequence, the DNA‐binding activity of the core domain is restored (Joerger et al., 2005; Nikolova et al., 2000). Similarly, DNA‐contact mutants also require specific suppressor mutations. Two second‐site suppressor mutations have been reported for the R273H mutant that potentially create novel DNA contacts and thus compensate for the loss of the Arg273‐mediated contacts (Baroni et al., 2004; Wieczorek et al., 1996).

B. p53C as a Drug Target Screening methods have yet to deliver a small molecule compound that rescues oncogenic p53 mutants by directly binding to the core domain. A possible explanation is the lack of well‐defined binding pockets for small molecules on the surface of the wild‐type core domain. By contrast,

Structural Biology of the Tumor Suppressor p53 and Cancer‐Associated Mutants

A

179

179 242

242

239

176

239

176 248

245

248

245

249

171

17

163 168

249

171

163 168

B

H2 CDB3

L1 Zn S8

p53

53BP2

Fig. 5 Rescuing mutant p53. (A) Structural basis for the rescue of the cancer hotspot mutant R249S by the intragenic suppressor mutation H168R. Stereo view of the structure of T‐p53C‐ T123A/H168R/R249S (PDB code 1BIQ; yellow) superimposed on the structure of wild type (PDB code 1TSR, molecule A; light gray). Arg249‐mediated stabilizing interactions in the wild type are indicated by dotted lines. In the rescued mutant, Arg168 mimics the structural role of Arg249 in the wild type (highlighted in magenta). The picture was adapted from Joerger et al. (2005). (B) Design of the chemical chaperone CDB3 by Friedler et al. (2002). Ribbon representation of the structure of the p53 core domain bound the 53BP2 region of the apoptosis‐ stimulating protein ASPP2 (PDB code 1YCS) (Gorina and Pavletich, 1996). The p53‐binding loop in 53BP2 from which the sequence of CDB3 was derived is shown in magenta (residues 490–498 of 53BP2). The figure was generated using MOLSCRIPT (Kraulis, 1991) and RASTER3D (Merritt and Bacon, 1997).

18

Andreas C. Joerger and Alan R. Fersht

a compound that binds to and stabilizes p53 was found by rational design, based on the structure of the 53BP2 region of the apoptosis‐stimulation protein ASPP2 in complex with p53C (Friedler et al., 2002). Core domain binding peptide 3 (CDB3), a nine‐residue peptide derived from one of the p53‐binding loops of 53BP2 (Fig. 5B), binds to p53 with a strong electrostatic component and raises the apparent melting temperature of the wild type and oncogenic mutants of the core domain. Its fluorescein‐labeled form, FL‐CDB3, is an even more potent binder. Moreover, FL‐CDB3 slows down the unfolding rate of p53C (Friedler et al., 2003), and NMR studies on the R249S mutant suggest that some of the mutation‐induced structural perturbations may be reversed in the presence of FL‐CDB3 (Friedler et al., 2004). FL‐CDB3 was also found to up‐regulate wild‐type and mutant p53 in human cell lines (Issaeva et al., 2003). Heteronuclear single quantum correlation (HSQC)‐NMR spectroscopy showed that CDB3 binds to p53C directly. The binding site partly overlaps with the DNA‐binding surface and includes the H2 helix, L1 loop, and ‐strand S8 (Friedler et al., 2002) but differs from that in the p53C‐53BP2 complex (Gorina and Pavletich, 1996). The stabilizing effect and the binding properties of FL‐CDB3 show that reactivating common p53 cancer mutants by compounds that bind to the core domain is a feasible strategy. Future strategies could combine screening and structure‐based design strategies to identify and optimize nonpeptide compounds that mimic the effects of CDB3. To be suitable drugs for cancer treatment, such compounds have to be chemically stable and display physicochemical properties that facilitate bioavailability and drug delivery, for example, they must be able to permeate the phospholipid bilayers of biological membranes. The emerging structural information on the effects of p53 cancer mutations, in combination with functional and biophysical data, provides a framework to assess suitable rescue strategies for individual mutants. Many ‐sandwich mutants, such as V143A, are ideal candidates for rescue by generic small molecule drugs. Since this mutant is destabilized but retains wild‐type conformation in the folded state, simple stabilization should be enough to restore function under physiological conditions. In the case of other structural mutations, such as in the hotspot mutants G245S and R249S, for which crystal structures show distinct conformational changes in functionally important regions, simple stabilization may only result in partial functional rescue. Altered surface properties, as for example observed for the Y220C mutation, which creates a solvent‐accessible crevice, may be exploited to design mutant‐ selective drugs. For the classic contact mutants, however, it is difficult to see how DNA‐binding activity could be restored by a generic small molecule compound. Rescue of these mutants would require a compound that creates an additional DNA contact, which compensates for the affinity loss on mutation of an essential DNA‐contact residue.

Structural Biology of the Tumor Suppressor p53 and Cancer‐Associated Mutants

19

VI. CONCLUDING REMARKS Since its discovery in 1979, a wealth of information about the structure and function of p53 has been gathered. While many aspects of the structure and function of p53 have been elucidated, even more unanswered questions remain. One of the great challenges will be to fully understand the complex interplay of the individual domains in the full‐length protein and the motions involved in binding to DNA and regulatory proteins; or to elucidate the exact role of the various posttranslational modifications. It has become clear that these questions can only be answered by using an interdisciplinary approach that combines biophysical and structural studies with experiments in the cellular context. Ultimately, biophysical data and functional data may be incorporated into system genomics projects to provide a quantitative mathematical model of the transcriptional network of p53 in the cell. Evidently, it is necessary to thoroughly understand the mutation‐structure and mutation‐ function relationship, in order to predict associated phenotypes and to evaluate possible therapeutic strategies. The emerging structural information on p53 cancer mutants opens novel avenues for the rescue of these mutants. In the future, these structures will provide a platform for a more rational design of pharmacologically active drugs. While many cancer mutants can be rescued by ligands directed toward the wild‐type conformation, others, such as the ‐sandwich mutant Y220C with its specific surface properties, can also be specifically targeted by a mutant‐selective drug.

REFERENCES Ang, H. C., Joerger, A. C., Mayer, S., and Fersht, A. R. (2006). Effects of common cancer mutations on stability and DNA binding of full‐length p53 compared with isolated core domains. J. Biol. Chem. 281, 21934–21941. Balagurumoorthy, P., Sakamoto, H., Lewis, M. S., Zambrano, N., Clore, G. M., Gronenborn, A. M., Appella, E., and Harrington, R. E. (1995). Four p53 DNA‐binding domain peptides bind natural p53‐response elements and bend the DNA. Proc. Natl. Acad. Sci. USA 92, 8591–8595. Baroni, T. E., Wang, T., Qian, H., Dearth, L. R., Truong, L. N., Zeng, J., Denes, A. E., Chen, S. W., and Brachmann, R. K. (2004). A global suppressor motif for p53 cancer mutants. Proc. Natl. Acad. Sci. USA 101, 4930–4935. Brachmann, R. K., Yu, K., Eby, Y., Pavletich, N. P., and Boeke, J. D. (1998). Genetic selection of intragenic suppressor mutations that reverse the effect of common p53 cancer mutations. EMBO J. 17, 1847–1859. Braithwaite, A. W., Del Sal, G., and Lu, X. (2006). Some p53‐binding proteins that can function as arbiters of life and death. Cell Death Differ. 13, 984–993. Bullock, A. N., Henckel, J., DeDecker, B. S., Johnson, C. M., Nikolova, P. V., Proctor, M. R., Lane, D. P., and Fersht, A. R. (1997). Thermodynamic stability of wild‐type and mutant p53 core domain. Proc. Natl. Acad. Sci. USA 94, 14338–14342.

20

Andreas C. Joerger and Alan R. Fersht

Bullock, A. N., Henckel, J., and Fersht, A. R. (2000). Quantitative analysis of residual folding and DNA binding in mutant p53 core domain: Definition of mutant states for rescue in cancer therapy. Oncogene 19, 1245–1256. Bykov, V. J., Selivanova, G., and Wiman, K. G. (2003). Small molecules that reactivate mutant p53. Eur. J. Cancer 39, 1828–1834. Canadillas, J. M., Tidow, H., Freund, S. M., Rutherford, T. J., Ang, H. C., and Fersht, A. R. (2006). Solution structure of p53 core domain: Structural basis for its instability. Proc. Natl. Acad. Sci. USA 103, 2109–2114. Chan, W. M., Siu, W. Y., Lau, A., and Poon, R. Y. (2004). How many mutant p53 molecules are needed to inactivate a tetramer? Mol. Cell. Biol. 24, 3536–3551. Chen, J. Y., Funk, W. D., Wright, W. E., Shay, J. W., and Minna, J. D. (1993). Heterogeneity of transcriptional activity of mutant p53 proteins and p53 DNA target sequences. Oncogene 8, 2159–2166. Cho, Y., Gorina, S., Jeffrey, P. D., and Pavletich, N. P. (1994). Crystal structure of a p53 tumor suppressor‐DNA complex: Understanding tumorigenic mutations. Science 265, 346–355. Chumakov, A. M., Miller, C. W., Chen, D. L., and Koeffler, H. P. (1993). Analysis of p53 transactivation through high‐affinity binding sites. Oncogene 8, 3005–3011. Clore, G. M., Omichinski, J. G., Sakaguchi, K., Zambrano, N., Sakamoto, H., Appella, E., and Gronenborn, A. M. (1994). High‐resolution structure of the oligomerization domain of p53 by multidimensional NMR. Science 265, 386–391. Clore, G. M., Ernst, J., Clubb, R., Omichinski, J. G., Kennedy, W. M., Sakaguchi, K., Appella, E., and Gronenborn, A. M. (1995). Refined solution structure of the oligomerization domain of the tumour suppressor p53. Nat. Struct. Biol. 2, 321–333. Courtois, S., de Fromentel, C. C., and Hainaut, P. (2004). p53 protein variants: Structural and functional similarities with p63 and p73 isoforms. Oncogene 23, 631–638. Dawson, R., Muller, L., Dehner, A., Klein, C., Kessler, H., and Buchner, J. (2003). The N‐terminal domain of p53 is natively unfolded. J. Mol. Biol. 332, 1131–1141. Derbyshire, D. J., Basu, B. P., Serpell, L. C., Joo, W. S., Date, T., Iwabuchi, K., and Doherty, A. J. (2002). Crystal structure of human 53BP1 BRCT domains bound to p53 tumour suppressor. EMBO J. 21, 3863–3872. Di Como, C. J., and Prives, C. (1998). Human tumor‐derived p53 proteins exhibit binding site selectivity and temperature sensitivity for transactivation in a yeast‐based assay. Oncogene 16, 2527–2539. DiTullio, R. A., Jr., Mochan, T. A., Venere, M., Bartkova, J., Sehested, M., Bartek, J., and Halazonetis, T. D. (2002). 53BP1 functions in an ATM‐dependent checkpoint pathway that is constitutively activated in human cancer. Nat. Cell Biol. 4, 998–1002. el‐Deiry, W. S., Kern, S. E., Pietenpol, J. A., Kinzler, K. W., and Vogelstein, B. (1992). Definition of a consensus binding site for p53. Nat. Genet. 1, 45–49. Foster, B. A., Coffey, H. A., Morin, M. J., and Rastinejad, F. (1999). Pharmacological rescue of mutant p53 conformation and function. Science 286, 2507–2510. Friedler, A., Hansson, L. O., Veprintsev, D. B., Freund, S. M., Rippin, T. M., Nikolova, P. V., Proctor, M. R., Ru¨diger, S., and Fersht, A. R. (2002). A peptide that binds and stabilizes p53 core domain: Chaperone strategy for rescue of oncogenic mutants. Proc. Natl. Acad. Sci. USA 99, 937–942. Friedler, A., Veprintsev, D. B., Hansson, L. O., and Fersht, A. R. (2003). Kinetic instability of p53 core domain mutants: Implications for rescue by small molecules. J. Biol. Chem. 278, 24108–24112. Friedler, A., DeDecker, B. S., Freund, S. M., Blair, C., Ru¨diger, S., and Fersht, A. R. (2004). Structural distortion of p53 by the mutation R249S and its rescue by a designed peptide: Implications for “mutant conformation.” J. Mol. Biol. 336, 187–196.

Structural Biology of the Tumor Suppressor p53 and Cancer‐Associated Mutants

21

Friedler, A., Veprintsev, D. B., Freund, S. M., von Glos, K. I., and Fersht, A. R. (2005a). Modulation of binding of DNA to the C‐terminal domain of p53 by acetylation. Structure 13, 629–636. Friedler, A., Veprintsev, D. B., Rutherford, T., von Glos, K. I., and Fersht, A. R. (2005b). Binding of Rad51 and other peptide sequences to a promiscuous, highly electrostatic binding site in p53. J. Biol. Chem. 280, 8051–8059. Gorina, S., and Pavletich, N. P. (1996). Structure of the p53 tumor suppressor bound to the ankyrin and SH3 domains of 53BP2. Science 274, 1001–1005. Grossman, S. R. (2001). p300/CBP/p53 interaction and regulation of the p53 response. Eur. J. Biochem. 268, 2773–2778. Gu, W., and Roeder, R. G. (1997). Activation of p53 sequence‐specific DNA binding by acetylation of the p53 C‐terminal domain. Cell 90, 595–606. Ho, W. C., Fitzgerald, M. X., and Marmorstein, R. (2006). Structure of the p53 core domain dimer bound to DNA. J. Biol. Chem. 281, 20494–20502. Huyen, Y., Jeffrey, P. D., Derry, W. B., Rothman, J. H., Pavletich, N. P., Stavridi, E. S., and Halazonetis, T. D. (2004). Structural differences in the DNA binding domains of human p53 and its C. elegans ortholog Cep‐1. Structure 12, 1237–1243. Issaeva, N., Friedler, A., Bozko, P., Wiman, K. G., Fersht, A. R., and Selivanova, G. (2003). Rescue of mutants of the tumor suppressor p53 in cancer cells by a designed peptide. Proc. Natl. Acad. Sci. USA 100, 13303–13307. Jeffrey, P. D., Gorina, S., and Pavletich, N. P. (1995). Crystal structure of the tetramerization domain of the p53 tumor suppressor at 1.7 angstroms. Science 267, 1498–1502. Joerger, A. C., Allen, M. D., and Fersht, A. R. (2004). Crystal structure of a superstable mutant of human p53 core domain. Insights into the mechanism of rescuing oncogenic mutations. J. Biol. Chem. 279, 1291–1296. Joerger, A. C., Ang, H. C., Veprintsev, D. B., Blair, C. M., and Fersht, A. R. (2005). Crystal structures of p53 cancer mutants and mechanism of rescue by second‐site suppressor mutations. J. Biol. Chem. 280, 16030–16037. Joerger, A. C., Ang, H. C., and Fersht, A. R. (2006). Structural basis for understanding oncogenic p53 mutations and designing rescue drugs. Proc. Natl. Acad. Sci. USA 103, 15056–15061. Joo, W. S., Jeffrey, P. D., Cantor, S. B., Finnin, M. S., Livingston, D. M., and Pavletich, N. P. (2002). Structure of the 53BP1 BRCT region bound to p53 and its comparison to the Brca1 BRCT structure. Genes Dev. 16, 583–593. Kitayner, M., Rozenberg, H., Kessler, N., Rabinovich, D., Shaulov, L., Haran, T. E., and Shakked, Z. (2006). Structural basis of DNA recognition by p53 tetramers. Mol. Cell 22, 741–753. Klein, C., Planker, E., Diercks, T., Kessler, H., Kunkele, K. P., Lang, K., Hansen, S., and Schwaiger, M. (2001). NMR spectroscopy reveals the solution dimerization interface of p53 core domains bound to their consensus DNA. J. Biol. Chem. 276, 49020–49027. Krajewski, M., Ozdowy, P., D’Silva, L., Rothweiler, U., and Holak, T. A. (2005). NMR indicates that the small molecule RITA does not block p53‐MDM2 binding in vitro. Nat. Med. 11, 1135–1136; author reply, 1136–1137. Kraulis, P. J. (1991). MOLSCRIPT: A program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr. 24, 946–950. Kussie, P. H., Gorina, S., Marechal, V., Elenbaas, B., Moreau, J., Levine, A. J., and Pavletich, N. P. (1996). Structure of the MDM2 oncoprotein bound to the p53 tumor suppressor transactivation domain. Science 274, 948–953. Lavin, M. F., and Gueven, N. (2006). The complexity of p53 stabilization and activation. Cell Death Differ. 13, 941–950. Lee, H., Mok, K. H., Muhandiram, R., Park, K. H., Suk, J. E., Kim, D. H., Chang, J., Sung, Y. C., Choi, K. Y., and Han, K. H. (2000). Local structural elements in the mostly unstructured transcriptional activation domain of human p53. J. Biol. Chem. 275, 29426–29432.

22

Andreas C. Joerger and Alan R. Fersht

Matsumura, I., and Ellington, A. D. (1999). In vitro evolution of thermostable p53 variants. Protein Sci. 8, 731–740. Menendez, D., Inga, A., and Resnick, M. A. (2006). The biological impact of the human master regulator p53 can be altered by mutations that change the spectrum and expression of its target genes. Mol. Cell. Biol. 26, 2297–2308. Merritt, E. A., and Bacon, D. J. (1997). Raster3D: Photorealistic molecular graphics. Methods Enzymol. 277, 505–524. Michael, D., and Oren, M. (2002). The p53 and Mdm2 families in cancer. Curr. Opin. Genet. Dev. 12, 53–59. Michael, D., and Oren, M. (2003). The p53‐Mdm2 module and the ubiquitin system. Semin. Cancer. Biol. 13, 49–58. Mittl, P. R., Chene, P., and Grutter, M. G. (1998). Crystallization and structure solution of p53 (residues 326–356) by molecular replacement using an NMR model as template. Acta Crystallogr. D 54, 86–89. Momand, J., Wu, H. H., and Dasgupta, G. (2000). MDM2—master regulator of the p53 tumor suppressor protein. Gene 242, 15–29. Mu¨ller‐Tiemann, B. F., Halazonetis, T. D., and Elting, J. J. (1998). Identification of an additional negative regulatory region for p53 sequence‐specific DNA binding. Proc. Natl. Acad. Sci. USA 95, 6079–6084. Nikolova, P. V., Henckel, J., Lane, D. P., and Fersht, A. R. (1998). Semirational design of active tumor suppressor p53 DNA binding domain with enhanced stability. Proc. Natl. Acad. Sci. USA 95, 14675–14680. Nikolova, P. V., Wong, K. B., DeDecker, B., Henckel, J., and Fersht, A. R. (2000). Mechanism of rescue of common p53 cancer mutations by second‐site suppressor mutations. EMBO J. 19, 370–378. Olivier, M., Eeles, R., Hollstein, M., Khan, M. A., Harris, C. C., and Hainaut, P. (2002). The IARC TP53 database: New online mutation analysis and recommendations to users. Hum. Mutat. 19, 607–614. Olivier, M., Langerod, A., Carrieri, P., Bergh, J., Klaar, S., Eyfjord, J., Theillet, C., Rodriguez, C., Lidereau, R., Bieche, I., Varley, J., Bignon, Y., et al. (2006). The clinical value of somatic TP53 gene mutations in 1,794 patients with breast cancer. Clin. Cancer Res. 12, 1157–1167. Pan, Y., Ma, B., Levine, A. J., and Nussinov, R. (2006). Comparison of the human and worm p53 structures suggests a way for enhancing stability. Biochemistry 45, 3925–3933. Pietenpol, J. A., Tokino, T., Thiagalingam, S., el‐Deiry, W. S., Kinzler, K. W., and Vogelstein, B. (1994). Sequence‐specific transcriptional activation is essential for growth suppression by p53. Proc. Natl. Acad. Sci. USA 91, 1998–2002. Prives, C., and Hall, P. A. (1999). The p53 pathway. J. Pathol. 187, 112–126. Prives, C., and Manley, J. L. (2001). Why is p53 acetylated? Cell 107, 815–818. Resnick, M. A., and Inga, A. (2003). Functional mutants of the sequence‐specific transcription factor p53 and implications for master genes of diversity. Proc. Natl. Acad. Sci. USA 100, 9934–9939. Rippin, T. M., Bykov, V. J., Freund, S. M., Selivanova, G., Wiman, K. G., and Fersht, A. R. (2002a). Characterization of the p53‐rescue drug CP‐31398 in vitro and in living cells. Oncogene 21, 2119–2129. Rippin, T. M., Freund, S. M., Veprintsev, D. B., and Fersht, A. R. (2002b). Recognition of DNA by p53 core domain and location of intermolecular contacts of cooperative binding. J. Mol. Biol. 319, 351–358. Rustandi, R. R., Baldisseri, D. M., and Weber, D. J. (2000). Structure of the negative regulatory domain of p53 bound to S100B(betabeta). Nat. Struct. Biol. 7, 570–574. Sakaguchi, K., Sakamoto, H., Xie, D., Erickson, J. W., Lewis, M. S., Anderson, C. W., and Appella, E. (1997). Effect of phosphorylation on tetramerization of the tumor suppressor protein p53. J. Protein Chem. 16, 553–556.

Structural Biology of the Tumor Suppressor p53 and Cancer‐Associated Mutants

23

Shaulsky, G., Goldfinger, N., Ben‐Zeev, A., and Rotter, V. (1990). Nuclear accumulation of p53 protein is mediated by several nuclear localization signals and plays a role in tumorigenesis. Mol. Cell. Biol. 10, 6565–6577. Shiraishi, K., Kato, S., Han, S. Y., Liu, W., Otsuka, K., Sakayori, M., Ishida, T., Takeda, M., Kanamaru, R., Ohuchi, N., and Ishioka, C. (2004). Isolation of temperature‐sensitive p53 mutations from a comprehensive missense mutation library. J. Biol. Chem. 279, 348–355. Stommel, J. M., Marchenko, N. D., Jimenez, G. S., Moll, U. M., Hope, T. J., and Wahl, G. M. (1999). A leucine‐rich nuclear export signal in the p53 tetramerization domain: Regulation of subcellular localization and p53 activity by NES masking. EMBO J. 18, 1660–1672. Tanner, S., and Barberis, A. (2004). CP‐31398, a putative p53‐stabilizing molecule tested in mammalian cells and in yeast for its effects on p53 transcriptional activity. J. Negat. Results Biomed. 3, 5. Tidow, H., Veprintsev, D. B., Freund, S. M. V., and Fersht, A. R. (2006). Effects of oncogenic mutations and DNA response elements on the binding of p53 and p53‐binding protein 2 (53BP2). J. Biol. Chem. 281, 32526–32533. Veprintsev, D. B., Freund, S. M., Andreeva, A., Rutledge, S. E., Tidow, H., Canadillas, J. M., Blair, C. M., and Fersht, A. R. (2006). Core domain interactions in full‐length p53 in solution. Proc. Natl. Acad. Sci. USA 103, 2115–2119. Vogelstein, B., Lane, D., and Levine, A. J. (2000). Surfing the p53 network. Nature 408, 307–310. Vousden, K. H., and Lu, X. (2002). Live or let die: The cell’s response to p53. Nat. Rev. Cancer 2, 594–604. Walker, K. K., and Levine, A. J. (1996). Identification of a novel p53 functional domain that is necessary for efficient growth suppression. Proc. Natl. Acad. Sci. USA 93, 15335–15340. Wang, W., Takimoto, R., Rastinejad, F., and El‐Deiry, W. S. (2003). Stabilization of p53 by CP‐ 31398 inhibits ubiquitination without altering phosphorylation at serine 15 or 20 or MDM2 binding. Mol. Cell. Biol. 23, 2171–2181. Weinberg, R. L., Freund, S. M., Veprintsev, D. B., Bycroft, M., and Fersht, A. R. (2004a). Regulation of DNA binding of p53 by its C‐terminal domain. J. Mol. Biol. 342, 801–811. Weinberg, R. L., Veprintsev, D. B., and Fersht, A. R. (2004b). Cooperative binding of tetrameric p53 to DNA. J. Mol. Biol. 341, 1145–1159. Wieczorek, A. M., Waterman, J. L., Waterman, M. J., and Halazonetis, T. D. (1996). Structure‐ based rescue of common tumor‐derived p53 mutants. Nat. Med. 2, 1143–1146. Wiman, K. G. (2006). Strategies for therapeutic targeting of the p53 pathway in cancer. Cell Death Differ. 13, 921–926. Wong, K. B., DeDecker, B. S., Freund, S. M., Proctor, M. R., Bycroft, M., and Fersht, A. R. (1999). Hot‐spot mutants of p53 core domain evince characteristic local structural changes. Proc. Natl. Acad. Sci. USA 96, 8438–8442. Zhan, M., Yu, D., Lang, A., Li, L., and Pollock, R. E. (2001). Wild type p53 sensitizes soft tissue sarcoma cells to doxorubicin by down‐regulating multidrug resistance‐1 expression. Cancer 92, 1556–1566. Zhang, W., Guo, X. Y., Hu, G. Y., Liu, W. B., Shay, J. W., and Deisseroth, A. B. (1994). A temperature‐sensitive mutant of human p53. EMBO J. 13, 2535–2544.

This page intentionally left blank

Immunotherapy by Allogeneic Stem Cell Transplantation Olle Ringde´n Division of Clinical Immunology, Karolinska Institutet, Karolinska University Hospital, Huddinge, SE‐141 86 Stockholm, Sweden; and Center for Allogeneic Stem Cell Transplantation, Karolinska Institutet, Karolinska University Hospital, Huddinge, SE‐141 86 Stockholm, Sweden

I. Introduction II. Graft‐Versus‐Host Disease A. Mechanisms B. Acute GVHD C. Chronic GVHD D. Prevention and Treatment of GVHD III. The Graft‐Versus‐Leukemia Effect A. Preliminary Studies B. Tumor Burden C. Enhancement of Graft‐Versus‐Leukemia D. Cytotoxic T Cells E. Pathophysiology of Graft‐Versus‐Leukemia IV. NK Cells V. Early Detection of Relapse A. Minimal Residual Disease B. Mixed Chimerism C. Molecular Detection of CML D. Immunoglobulin and T Cell Receptor Gene Rearrangement VI. The Graft‐Versus‐Cancer Effect A. Immunotherapy Against Cancer B. Reduced Intensity Conditioning and Allogeneic Stem Cell Transplantation C. Reduced Intensity Conditioning and Stem Cell Transplantation for Renal Carcinoma D. Reduced Intensity Conditioning and Stem Cell Transplantation for Various Solid Tumors E. Combined Liver Transplantation and Stem Cell Transplantation for Liver Cancer F. Future of Stem Cell Transplantation for Solid Cancers VII. Mesenchymal Stem Cells A. Surface Markers and Homing B. Immunity and Safety of MSCs C. Immunomodulation by MSC D. Immunosuppressive Mechanisms by MSCs VIII. Future Directions References

Advances in CANCER RESEARCH Copyright 2007, Elsevier Inc. All rights reserved.

25

0065-230X/07 $35.00 DOI: 10.1016/S0065-230X(06)97002-X

26

Olle Ringde´n During the past three decades, allogeneic stem cell transplantation (ASCT) has developed from being an experimental therapy in patients with endstage leukemia into a well‐established therapy in patients with a range of disorders of the immunohematopoietic system. Graft‐versus‐host disease (GVHD), acute or chronic, attacking host tissue is a major threat. However, donor immunocompetent T cells have a potent graft‐versus‐leukemia effect. A combination of calcineurin inhibitors and methotrexate is the standard therapy to prevent GVHD. Modulation of the immunosuppressive regimen may induce mild acute and mild chronic GVHD, reduce the risk of relapse, and improve long‐term survival. Natural killer cells also play a role in this context. Killer cell immunoglobulin‐like receptor incompatibility between recipient and donor may reduce the risk of relapse in patients with myeloid leukemia. Relapse of leukemia is a major cause of death after ASCT. Minimal residual disease and recipient leukemia lineage‐specific chimerism are sensitive techniques for early detection of leukemic relapse. Donor lymphocyte infusions can enhance the antitumor effect, especially for patients with molecular relapse. The allogeneic graft‐versus‐cancer effect has been demonstrated in patients with metastatic breast, renal, colorectal, ovarian, prostatic, and pancreatic carcinoma. Mesenchymal stem cells have immunomodulatory properties and may be used for immunomodulation of GVHD and tissue repair. All things considered, the future looks promising for ASCT. # 2007 Elsevier Inc.

I. INTRODUCTION The immune system may control cancer, which is evident from both experimental and clinical studies (Horowitz et al., 1990; Ringden and Horowitz, 1989; Truitt et al., 1996; Weiden et al., 1981). High‐dose chemoradiotherapy followed by allogeneic hematopoietic stem cell transplantation (ASCT) is an effective and well‐established therapy for many patients with otherwise fatal hematological malignancies (Appelbaum et al., 1987; Barrett et al., 1989; Gajewski et al., 1996; Goldman et al., 1988; Thomas et al., 1977). In patients with acute leukemia in first remission or chronic myeloid leukemia (CML) in the first chronic phase, long‐term survival or cure rate may be in the range of 60–80%, with a relapse probability of 10–30% (Hansen et al., 1998). In patients undergoing transplantation in second or later remission, the relapse probability may be up to 50% with long‐term survival of around 40% (Barrett et al., 1989). In patients undergoing ASCT for advanced or refractory disease, the risk of relapse may be 70% or higher with a long‐term survival of only 10% (Ringden et al., 1987; Thomas et al., 1977). The aim of using high‐dose chemoradiotherapy is to kill as many malignant cells as possible and rescue the patient with the hematopoietic stem cell graft from pancytopenia and toxic side effects. By transplanting pluripotent hematopoietic stem cells contained in bone marrow or spleen cells, it became possible to administer far higher doses of chemotherapy than was otherwise possible. Furthermore, the conditioning immunosuppresses the patient to eradicate host T cells, prevents graft rejection, and paves the way for the donor immune hematopoietic system.

Immunotherapy by Allogeneic Stem Cell Transplantation

27

In the last two decades, it became apparent that donor immunocompetent T cells have a potent graft‐versus‐leukemia effect (Horowitz et al., 1990; Ringden and Horowitz 1989; Weiden et al., 1981). Ther are several immunological consequences of ASCT; for example, residual immunocompetent recipient T cells may cause graft rejection. A recovering immature immunohematopoietic system suffers from pancytopenia and immune incompetence, which may lead to infections by bacteria, fungi, and viruses. Furthermore, immune donor T cells induce acute and/or chronic graft‐versus‐host disease (GVHD). Due to immune incompetence, not only infections but also secondary malignancies may appear (Curtis et al., 1997; Socie et al., 1993; Witherspoon et al., 1989). This chapter will focus on GVHD, the graft‐versus‐leukemia effect, early detection and treatment of leukemic relapse, the graft‐versus‐solid‐ tumor effect, mesenchymal stem cells (MSCs), and future directions in immunotherapy by ASCT.

II. GRAFT‐VERSUS‐HOST DISEASE A. Mechanisms Acute GVHD was first detected in experimental animals and is one of the major hazards in clinical ASCT (Ringden and Deeg, 1996; Storb and Thomas, 1985; van Bekkum, 1985). Donor T cells are responsible for triggering GVHD and proliferate after activation by recipient antigens, which are expressed on host cells in the form of major histocompatibility complex (MHC), class I or class II antigens, viral antigens, or minor antigenic peptides, including epithelial cell‐associated antigens (de Gast et al., 1987; Ferrara et al., 1999). Antigen‐presenting cells, such as dendritic cells or macrophages, present the antigens to T cells. CD4þ T cells recognize antigens in association with human leukocyte antigen system (HLA) class II molecules. IL‐1 and other factors, such as epidermal‐derived thymocyte‐ activating factor produced by monocytes, stimulate the CD4þ T cells, which release IL‐2. IL‐2 activates CD8þ T cells, which in turn react with MHC class I‐positive targets. In addition, natural killer (NK) cells and macrophages participate in the development of GVHD. A subset of activated CD4þ T cells produces interferon (IFN) , which enhances the expression of MHC class II on epithelial cells and macrophages, further stimulating T cell and NK cell activation (de Gast et al., 1987; Ferrara et al., 1999).

28

Olle Ringde´n

B. Acute GVHD In humans, acute GVHD generally appears during the first 3 months after ASCT, but may also appear later. The main target organs for acute GVHD are the skin, gut, and liver (Ringden and Deeg, 1996; Storb and Thomas, 1985; van Bekkum, 1985). Depending on severity, acute GVHD is graded on a five‐point scale from 0 to IV (Glucksberg et al., 1974). Grade 0 is no GVHD and grade I (mild GVHD) is a local skin rash. Grade II (moderate GVHD) can be a skin rash alone affecting most of the body, or a skin rash in association with gut or liver symptoms. Grade III involves more severe symptoms from the skin, gut, and/or liver, and grade IV is life‐threatening. Skin GVHD often presents as a macropapular skin rash, which may start in the palms, the soles of the feet, or the face. In more severe forms, the entire body may be affected with necrosis. Liver GVHD involves increased bilirubin and liver enzymes, with ascites and liver enlargement in more severe forms. Patients with gastrointestinal GVHD have diarrhea, abdominal pain and, in severe cases, hemorrhaging. GVHD and its treatment are also immunosuppressive, and infections by bacteria, fungi, and viruses occur frequently. In severe cases these can be fatal. The major risk factor in acute GVHD is HLA disparity between recipient and donor. Other risk factors are a female donor to a male recipient, seropositivity for several herpes viruses in the recipient and donor, certain HLA alleles, and the host environment (Bostrom et al., 1990a; Gale et al., 1987; Renkonen et al., 1986; Ringden and Deeg, 1996; Storb et al., 1983a). Granulocyte colony‐ stimulating factor (G‐CSF) was used posttransplant to accelerate neutrophil recovery and hopefully reduce the risk of infections (Appelbaum, 1995). G‐CSF treatment after transplant was found to be associated with an increased risk of GVHD (Remberger et al., 2003; Ringden et al., 2004a). A large retrospective European study showed that bone marrow transplant recipients with acute leukemia who were treated with G‐CSF ran a significantly increased risk of acute and chronic GVHD, transplant‐related mortality and death, with reduced survival and leukemia‐free survival (Ringden et al., 2004a). G‐CSF treatment, while shortening the neutropenic phase, was found to be associated with slower platelet engraftment, probably due to an increase in platelet aggregation. However, there is some controversy regarding these findings, and several small studies have not found an increased risk of GVHD using G‐CSF (Ringden et al., 2005). One proposed mechanism for the increased risk of acute GVHD follows from the observation that when G‐CSF is given to recipients as prophylaxis following transplantation, the levels of soluble IL‐2 receptor‐ increase. This was found to be associated with aggravated GVHD (Kobayashi et al., 1999; Miyamoto et al., 1996; Remberger and Sundberg, 2005).

Immunotherapy by Allogeneic Stem Cell Transplantation

29

C. Chronic GVHD GVHD may also appear in a chronic form, which may develop 3 months or later after ASCT (Ringden and Deeg, 1996; Sullivan et al., 1981). Chronic GVHD manifestations include skin disease, keratoconjunctivitis, generalized sicca syndrome, oral mucositis, esophageal and vaginal strictures, malabsorption, wasting, liver disease, obstructive bronchiolitis, myositis, and neuropathy. Chronic GVHD is also associated with immunodeficiency and frequent infections with gram‐positive bacteria, which may cause septicemia, sinusitis, and bronchopneumonia (Ringden and Deeg, 1996). Chronic GVHD may be graded as limited, localized skin disease or hepatic dysfunction (Shulman et al., 1978). Manifestations of extensive chronic GVHD are generalized skin involvement and/or involvement of other organs. Acute GVHD may pave the way for chronic GVHD. Other risk factors for development of chronic GVHD include high recipient or donor age, an alloimmune female donor to a male recipient, treatment with donor buffy‐coat cells, peripheral blood stem cell grafts (PBSC), compared with bone marrow, CML, and seropositivity for several herpes viruses in the recipient and donor (Bostrom et al., 1990b; Carlens et al., 1998; Ringden and Deeg, 1996; Storb et al., 1983b).

D. Prevention and Treatment of GVHD GVHD may be completely abolished if a T cell‐depleted graft is transplanted (Marmont et al., 1991; Prentice et al., 1984; Ringden and Deeg, 1996). However, T cell depletion of the graft is associated with an increased risk of rejection, leukemic relapse, and infectious complications. GVHD appears even when there is genotypic HLA identity between patient and donor. Thus, all patients except those receiving grafts from syngeneic donors need immunosuppression to prevent GVHD. Initially, methotrexate and cyclosporine were used as single agents (Ringden et al., 1986; Storb and Thomas, 1985; Thomas et al., 1977). These drugs resulted in comparable incidences of acute and chronic GVHD (Ringden et al., 1986). When methotrexate and cyclosporine were combined, this resulted in a decrease in acute and chronic GVHD and improved leukemia‐free survival (Ringden et al., 1993b; Storb et al., 1989). However, improved immunosuppression by combining these two drugs was achieved at the price of an increased risk of leukemic relapse (Aschan et al., 1991). Combining cyclosporine and methotrexate resulted in a similar incidence of GVHD and a similar incidence of relapse, compared to T cell depletion of the graft, using HLA‐identical sibling donors (Ringden et al., 1991). Cyclosporine may be

30

Olle Ringde´n

replaced by tacrolimus, with a similar outcome (Nash et al., 1996). Other drugs used for the prevention of GVHD include mycophenolate mofetil and rapamycin. However, their role in the prevention of GVHD must still be established in controlled trials. The first‐line standard treatment of acute GVHD is high‐dose steroids (Groth et al., 1979). If this treatment fails, a variety of other agents have been tried, including antithymocyte globulin, anti‐IL‐2 receptor antibodies, monoclonal antibodies against the CD3 receptor, antibody against tumor necrosis factor‐a, recombinant human IL‐1 receptor antibodies, methotrexate, and rapamycin (Anasetti et al., 1994; Benito et al., 2001; Deeg et al., 2001; Herve et al., 1990; Kobbe et al., 2001; McCarthy et al., 1996; Ringden and Deeg, 1996). Patients with steroid‐refractory acute GVHD have a very poor outcome with a high mortality of hemorrhages, multiorgan failure, and infectious complications (Remberger et al., 2001). A new experimental approach to steroid‐resistant acute GVHD is the use of MSCs, which have been found to repair the gut and reverse life‐threatening grade‐IV acute GVHD (Le Blanc et al., 2004b; Ringden et al., 2006). Chronic GVHD may be treated with steroids, cyclosporine, tacrolimus, psoralene and ultraviolet light, azathioprin, 1 Gy of total body irradiation, thalidomide, mycophenolate mofetil, sirolimus, and anti‐B cell antibodies (Canninga‐van Dijk et al., 2004; Ringden and Deeg, 1996). Due to the very high mortality associated with moderate‐to‐severe acute GVHD, this complication is of utmost importance for the outcome and survival after ASCT. However, acute and especially chronic GVHD have a pronounced antileukemic effect. Thus, the antileukemic effect has to be balanced against morbidity and mortality from these complications.

III. THE GRAFT‐VERSUS‐LEUKEMIA EFFECT A. Preliminary Studies It was noted some 50 years ago that leukemia could be eradicated in irradiated mice receiving allogeneic bone marrow grafts, but not syngeneic grafts (Barnes et al., 1956). In clinical ASCT, it was first reported by Weiden et al. (1979) from Seattle that leukemic patients with GVHD had an increased probability of being in remission. In a subsequent study, it was shown that chronic GVHD has a stronger antileukemic effect than acute GVHD (Weiden et al., 1981). As had been reported previously in mice, twins who underwent ASCT and who did not develop GVHD ran a higher risk of relapse than recipients of grafts from HLA‐identical siblings (Fefer et al., 1987; Gale et al., 1994; Goldman et al., 1988; Horowitz et al., 1990; Ringden et al., 1987). As mentioned above, T cell depletion of the stem cell

Immunotherapy by Allogeneic Stem Cell Transplantation

31

graft, which prevents GVHD, also increases the risk of relapse, especially in patients with CML (Aschan et al., 1993; Goldman et al., 1988; Horowitz et al., 1990; Marmont et al., 1991). A study from the International Bone Marrow Transplant Registry showed that among 2000 patients with leukemia in first remission or first chronic phase, for those with acute lymphoblastic leukemia, there was a high correlation between acute GVHD and antileukemic activity, whereas in patients with acute myeloid leukemia and CML, the association was stronger with chronic GVHD (Horowitz et al., 1990). A large European study showed that chronic GVHD had a stronger antileukemic effect than acute GVHD in patients with acute leukemia (Ringden et al., 1996b). The higher the grade of acute GVHD, the stronger was the antileukemic effect (Gratwohl et al., 1995; Horowitz et al., 1990; Ringden and Horowitz, 1989; Ringden et al., 1996a). With more severe acute GVHD, transplant‐related mortality was increased and thus the highest leukemia‐free survival in patients with leukemia was seen in patients with a mild, grade‐I, acute GVHD (Gratwohl et al., 1995; Ringden et al., 1996a). However, since chronic GVHD had an even better antileukemic effect, the best leukemia‐free survival was seen in patients with mild acute and mild chronic GVHD (Ringden and Horowitz, 1989).

B. Tumor Burden Tumor burden may be important for the antitumor effect of immunotherapy. In experimental animals, adoptive immunotherapy was most effective when the tumor burden was low (Fefer et al., 1976). In contrast, a study in humans showed that GVHD patients with advanced leukemia had a reduced rate of relapse, but this did not apply to patients in first remission (Sullivan et al., 1989). A European study showed that the best antileukemic effect in patients with acute leukemia was seen in those in first remission, and that the effect was less pronounced in patients with second or later remission, or in relapse (Ringden et al., 1996b). This study also showed that in patients with early leukemia, the antileukemic effect was the same in patients with limited or extensive chronic GVHD. However, in patients with more advanced leukemia, extensive chronic GVHD had a stronger antileukemic effect than more limited disease. With more therapy‐resistant leukemia, a stronger antileukemic effect may be required to reduce the risk of leukemic relapse.

C. Enhancement of Graft‐Versus‐Leukemia Is it possible to use the graft‐versus‐leukemia effect to improve survival in patients undergoing ASCT for leukemia? GVHD can be prevented by T cell depletion of the graft, or by effective immunosuppression (Marmont

32

Olle Ringde´n

et al., 1991; Ringden et al., 1993b; Storb et al., 1989). However, better prevention of GVHD may be at the price of an increased risk of relapse and, subsequently, unchanged long‐term survival. Bacigalupo et al. (1991) showed that in patients with acute myeloid leukemia, a high dose of cyclosporine was associated with an increased risk of leukemic relapse, compared to a low dose. Thus, we attempted to take advantage of the graft‐versus‐ leukemia effect by reducing the immunosuppressive therapy to allow mild acute GVHD to develop (Carlens et al., 1999). By giving low doses of cyclosporine combined with methotrexate and discontinuing immunosuppression early, we were able to increase the risk of mild and moderate acute GVHD, and chronic GVHD, and subsequently reduce the risk of leukemic relapse (Carlens et al., 1999). By this strategy, we were able to optimize the graft‐versus‐leukemia effect without having an increased risk of severe acute GVHD and transplant‐related mortality. It is possible to enhance the graft‐versus‐leukemia effect further by treatment with immunocompetent donor cells, so‐called donor lymphocyte infusions (DLI) (Kolb et al., 1990). The introduction of DLI has been one of the most important advances in the field of ASCT in recent years. DLI was especially effective in patients with CML who had a relapse after ASCT, with a complete remission rate of 73% (Kolb et al., 1995). Leukemic burden is also important for the effect of DLI. For instance, patients receiving DLI at an early disease stage, for example molecular or cytogeneic relapse, had a better response to this therapy than patients treated for more advanced hematological relapse (Carlens et al., 2001). There are also side effects to DLI, including pancytopenia and GVHD. However, it seems important to develop chronic GVHD to have an optimal antileukemic response to DLI (Kolb et al., 1995). For patients with CML, to achieve an antileukemic response, a dose of 107 T cells/kg appears to be required if the donor is an HLA‐identical sibling (Mackinnon et al., 1995). If the donor is unrelated, a similar effect seems to be possible using a T cell dose that is one log lower. To avoid inducing severe GVHD by DLI, it may be helpful to start with a low dose and to escalate the T cell dose by one log per month, instead of giving one high dose of bulk DLI (Dazzi et al., 2000a). Using this protocol, it was possible to achieve good antileukemic control with an acceptable risk of severe GVHD. There is also a graft‐versus‐leukemia effect in the absence of GVHD (Horowitz et al., 1990; Ringden and Horowitz, 1989; Ringden et al., 2000a). As an example of this, patients receiving ASCT from identical twins had an increased risk of relapse compared to recipients of allografts without any GVHD. There also appears to be an antileukemic cell‐dose effect. For example, in recipients of syngeneic grafts, those who received a higher than median donor stem cell dose (3  108 nucleated cells/kg) had a reduced risk of

Immunotherapy by Allogeneic Stem Cell Transplantation

33

relapse compared to those receiving a lower cell dose (Barrett et al., 2000). The cell‐dose effect was also seen in recipients of peripheral blood stem cell transplants from HLA‐identical sibling grafts. In multivariate analysis, those receiving more than the median (>6  106 CD34þ cells/kg) had a reduced risk of relapse and treatment failure compared to those receiving a lower stem cell dose, taking other significant covariates into consideration (Ringden et al., 2003).

D. Cytotoxic T Cells It is obvious that cytotoxic T‐lymphocytes (CTLs) have a profound antileukemic effect after ASCT. This may raise the possibility of differentiating the graft‐versus‐leukemia effect from GVHD. However, is it possible to sort out whether different T cell clones induce GVHD and a graft‐versus‐ leukemia effect (Truitt et al., 1996)? One possible way of exploring this in the clinic may be to identify antigens which are solely expressed on the patient’s leukemic cells, and to separate them from other antihost reactions that may induce GVHD. One strategy for differentiation of the graft‐versus‐ leukemia effect has been to study polymorphic minor histocompatibility antigens. Another way to discriminate leukemic cells from other host tissues may be to identify antigens associated with a malignant phenotype. One approach was to identify minor histocompatibility antigens of importance for either graft‐versus‐leukemia or GVHD (Goulmy, 1997). Five human minor histocompatibility antigens were identified, three of which were broadly expressed and may serve as targets for both GVHD and for graft‐ versus‐leukemia. These were expressed selectively by hematopoietic cells and could thus serve as targets for a specific antileukemic response (den Haan et al., 1998; Mutis et al., 1999). The other two were only expressed on hematopoietic cells and may therefore be targets for a specific antileukemic cytotoxic response. By cell culture techniques, it has been possible to isolate cytotoxic T cell clones specific for minor histocompatibility antigens from patients who have undergone ASCT (Warren et al., 1998a,b). Using this technique, it was possible to characterize several minor histocompatibility antigens. Many of these antigens were only expressed on hematopoietic tissue. T cell clones were shown to be capable of lysing hematopoietic cells, but not fibroblasts from the ASCT recipients. Furthermore, in patients with B cell acute lymphoblastic leukemia, specific cytotoxic T‐cells against B cell leukemia‐associated minor histocompatibility antigens were detected (Dolstra et al., 1997). An alternative approach to induction of an antileukemic effect may be to use G‐CSF, which has been reported to induce complete remission of leukemic relapse in 43% of patients (Bishop et al., 2000).

34

Olle Ringde´n

E. Pathophysiology of Graft‐Versus‐Leukemia Although the exact pathophysiological mechanisms involved in graft‐versus‐ leukemia effects are not known, minor and major histocompatibility‐restricted cytotoxic T‐cells with antileukemic activity probably play a significant role. Several mechanisms of action have been proposed (Truitt et al., 1996). Antigens mismatched at MHC class I and class II are probably involved. As discussed above, minor histocompatibility antigens expressed on the leukemic cells may also be target antigens. Not only T cells, but also other cell populations have an antitumor effect—including lymphokine‐activated killer cells and NK cells. Cytokines are involved in the pathogenesis of GVHD and may also be involved in the allogeneic graft‐versus‐leukemia phenomenon. Cytokines secreted from T cells, such as IL‐2, IL‐3, IFN , and tumor necrosis factor‐, may have a direct antitumor effect by recruiting accessory cells or enhancing the cytotoxic T‐cells involved in this reaction.

IV. NK CELLS NK cells interact with class I MHC molecules on target cells and are inhibited in line with the original “missing‐self” hypothesis (Karre et al., 1986; Ljunggren and Karre, 1990). ASCT across HLA barriers may trigger donor NK cell alloreactivity if the recipient lacks killer cell immunoglobulin‐ like receptors (KIRs) that should recognize groups of human HLA class I alleles that are present in the donor (Farag et al., 2002). In ASCT between HLA‐haploidentical donors and recipients, it has been observed that donor NK cells encountering recipient target cells lacking an HLA class I allele present in the donor HLA genotype can mediate antileukemic effects in patients with acute myeloid leukemia—if the HLA class I mismatch predicts lack of ligand for donor‐inhibitory KIR (Ruggeri et al., 2002). These antileukemic effects were found to induce a lower rate of relapse, graft failure, and GVHD, and resulted in improved survival. Conflicting results have been found in studying the importance of KIR‐ligand incompatibility between the recipient of ASCT and an unrelated donor. For example, Davies et al. (2002) could not demonstrate an improved survival rate for patients receiving KIR‐ligand mismatched transplants. In contrast, Giebel et al. (2003) reported that overall survival was significantly better for patients with KIR‐ ligand‐incompatible donors. We have found an increased mortality from infections using unrelated KIR‐ligand mismatched donors (Schaffer et al., 2004). Our results suggest that the presence of donor‐derived alloreactive NK cells may interfere with immunity to infection. There is evidence that NK cells can kill dendritic cells in vitro and in vivo (Cooper et al., 2004). It is

Immunotherapy by Allogeneic Stem Cell Transplantation

35

possible that dendritic cells may play a role in the patient’s immune defense against infections. Bornhauser et al. (2004) reported an increased risk of relapse in patients receiving stem cell grafts from KIR‐ligand mismatched donors. However, a report on leukemic patients undergoing ASCT using unrelated donors showed that lack of HLA‐ligand for donor‐inhibitory KIR had no effect on relapse or survival in patients undergoing ASCT for acute lymphoblastic leukemia or CML (Hsu et al., 2005). However, in patients with acute myeloid leukemia, there was a significant missing KIR‐ligand effect with a lower incidence of relapse, and improved survival and leukemia‐free survival in HLA‐identical sibling transplants for patients who lacked the HLA‐ligand for donor‐inhibitory KIR. Thus, it seems that the absence of HLA class I ligand in the recipient for donor‐inhibitory KIR is only a good prognostic factor for ASCT in patients with acute myeloid leukemia.

V. EARLY DETECTION OF RELAPSE A. Minimal Residual Disease At the time of diagnosis, patients with leukemia may have as many as 1012 leukemic cells (Ryan and van Dongen, 1988). After intensive chemotherapy, the patient may hopefully enter remission, but 1010 leukemic cells may still be present. Minimal residual disease (MRD) refers to detection of leukemic cells below the threshold of standard morphological assessment. ASCT aims to induce complete eradication of leukemic cells, as well as complete donor hematopoietic chimerism. However, so‐called mixed chimerism with both donor and persistent host hematopoietic cells is compatible with prolonged survival. Using molecular techniques, it is possible to detect even low numbers of residual host cells after ASCT. Polymerase chain reaction (PCR) for characterization of molecular lesions can monitor residual disease at a sensitivity of 1 malignant cell in 105 healthy hematopoietic cells (Bader et al., 2005; Potter et al., 1993). It is also possible to use fluorescence in situ hybridization (FISH) for detection of leukemic cells.

B. Mixed Chimerism Another approach for detection of leukemic relapse may be to perform chimerism analysis, using PCR amplification of variable‐number tandem repeats to detect leukemia lineage‐specific chimerism. If mixed chimerism analysis, that is persistence of recipient‐derived hematopoietic cells, can be

36

Olle Ringde´n

used as a tool for early detection of leukemia relapse after ASCT is controversial (Socie et al., 1995). Some studies have shown no correlation between mixed chimerism and relapse in patients with acute leukemia (Schattenberg et al., 1989; van Leeuwen et al., 1994). However, other investigators have found an increased amount of recipient cells after ASCT, and this has been found to be correlated with an increased risk of relapse (Bader et al., 1998; Bertheas et al., 1991; Lawler et al., 1991; Mackinnon et al., 1992). In patients with CML, MRD was commoner in patients who had mixed T cell chimerism after ASCT (Mackinnon et al., 1994). In a study of patients with acute myeloid leukemia, mixed chimerism in the leukemia‐affected cell lineage was detected in 14 patients. Ten of these patients underwent relapse, compared to only 2 relapses in 16 patients who had complete donor chimerism (p < 0.01) (Mattsson et al., 2001b). Mixed chimerism of the leukemia‐affected cell lineage was detected at a median of 66 days before hematological relapse.

C. Molecular Detection of CML In patients with CML, reverse transcriptase (RT)‐PCR of the BCR‐ABL region was used extensively to study MRD after ASCT (Gabert et al., 1989; Mackinnon et al., 1994; Morgan et al., 1989). Using quantitative PCR methods, the kinetics of BCR‐ABL transcripts could be followed in detail (Cross et al., 1993; Preudhomme et al., 1999). Serial quantitative RT‐PCR analysis can distinguish patients who will most probably relapse, that is, those with high or increasing BCR‐ABL levels from those who will remain in clinical remission, those with low or decreasing BCR‐ABL levels (Lin et al., 1996). Patients with CML undergoing ASCT and treated with DLI at the time of molecular relapse had a better response rate than patients given DLI at the time of hematological relapse (Carlens et al., 2001; Dazzi et al., 2000b).

D. Immunoglobulin and T Cell Receptor Gene Rearrangement In patients with acute lymphoblastic leukemia, immunoglobulin‐gene and T cell receptor gene rearrangement may be used as clonal markers for MRD analysis. In patients with acute lymphoblastic leukemia who were treated with T cell‐depleted marrow, MRD before and after ASCT was found to be a significant predictor of leukemic relapse (Knechtli et al., 1998a,b). Also, in patients with acute lymphoblastic leukemia, several studies have been performed to evaluate the effect of mixed chimerism after enrichment of the

Immunotherapy by Allogeneic Stem Cell Transplantation

37

cell population carrying the leukemic phenotype. Using unmanipulated grafts, detection of MRD before ASCT has also been found to be associated with an increased risk of relapse (Uzunel et al., 2001). In patients with high levels of MRD, acute and chronic GVHD protected against relapse. Appearance of MRD after ASCT in patients with acute lymphoblastic leukemia was also associated with an increased risk of relapse. Relapse occurred in 8/9 MRD‐positive patients with acute lymphoblastic leukemia, as compared to 6/23 MRD‐negative patients after ASCT (p < 0.01) (Preudhomme et al., 1999). MRD was detected at a median of 6 months before hematological relapse in these patients. In patients with acute lymphoblastic leukemia, chimerism analysis can be made more sensitive using real‐time PCR single nucleotide polymorphism (Uzunel et al., manuscript in preparation). With this technique and with a detection level of more than 0.5% in blood, leukemia‐free survival was found to be only 21% five years after ASCT, compared to 80% in patients with acute lymphoblastic leukemia and less than 0.1% chimerism (p < 0.001). Chimerism and MRD analysis in leukemic patients provide the possibility of identifying patients at high risk of relapse after ASCT. Furthermore, monitoring of these sensitive tests after transplant may detect relapse several months before hematological relapse. Thus, based on such strategies, DLI may be given to prevent relapse and improve leukemia‐free survival.

VI. THE GRAFT‐VERSUS‐CANCER EFFECT A. Immunotherapy Against Cancer Because of the graft‐versus‐leukemia effect, ASCT has also been tried in patients with metastatic solid tumors and a poor prognosis. Metastatic melanoma and renal cell carcinoma occasionally show spontaneous regression or prolonged periods of stable disease. This leads to the concept that the immune system may have a role in the control of these malignancies. The first spontaneous regression of metastatic renal cell carcinoma was described after nephrectomy, and was thought to be due to an antibody‐mediated immune response (Bumpus, 1928). Later immune‐based therapies were tried in metastatic solid tumors. Therapies included treatment with IL‐2 and adoptive transfer of lymphokine‐activated killer cells (Rosenberg et al., 1993). In mice, it was shown that the incidence of spontaneous lymphosarcomas in a hybrid strain could be significantly diminished by ASCT from another strain (Moscovitch and Slavin, 1984). Furthermore, challenge with mouse mammary adenocarcinoma tumor was found to be protected by histocompatibility antigen mismatch ASCT (Morecki et al., 1997).

38

Olle Ringde´n

These animal studies have formed the basis for ASCT in humans. The first reports on an allogeneic graft‐versus‐tumor effect were reported in some anecdotal patients with metastatic breast cancer (Ben‐Yosef et al., 1996; Eibl et al., 1996; Ueno et al., 1998).

B. Reduced Intensity Conditioning and Allogeneic Stem Cell Transplantation Reduced intensity conditioning (RIC) was introduced to enable transplantation in older patients or those with organ impairment, who could not tolerate high‐dose myeloablative therapy (Giralt et al., 1997; McSweeney and Storb, 1999; Slavin et al., 1998). Patients above the age of 60 years have an increased transplant‐related mortality and a reduced survival rate compared to younger patients who have undergone myeloablative conditioning (Ringden et al., 1993a). Low‐intensity conditioning regimens have a low toxicity and low transplant‐related mortality, making ASCT safer and possible to use on older or debilitated patients. Furthermore, such conditioning regimens should be sufficiently immunosuppressive to achieve donor engraftment. Also, they should allow hematological recovery in the recipient in the event of donor graft failure. To achieve donor engraftment, it is important to have a high dose of donor cells. Thus, PBSC are preferentially used because they have a significantly higher CD34 cell dose, and a one‐log higher T cell and NK cell dose, and therefore enable a faster engraftment of neutrophils and platelets compared to bone marrow as the source of stem cells (Ringden et al., 2002; Schmitz et al., 2002).

C. Reduced Intensity Conditioning and Stem Cell Transplantation for Renal Carcinoma Donor engraftment and GVHD can be monitored by analyzing lineage‐ specific chimerism for T cells, B cells, and myeloid cells using PCR techniques for variable‐number tandem repeats (Mattsson et al., 2000; Potter et al., 1993; Zetterquist et al., 2000). If there is mixed chimerism after transplantation, DLI can be given in escalating doses to enhance donor cell engraftment (Carlens et al., 2001; Kolb et al., 1990, 1995; Mackinnon, et al., 1995). For several reasons, RIC has been an attractive conditioning for patients with metastatic solid tumors that were not responsive to chemoradiotherapy. These tumors do not respond well to chemoradiotherapy, and therefore conditioning is only used to immunosuppress the patient to accept the transplanted immunocompetent donor cells. Childs et al. (1999) were

Immunotherapy by Allogeneic Stem Cell Transplantation

39

the first to report an ASCT‐associated antitumor effect in metastatic renal carcinoma. Subsequently, this group reported an overall response rate of 44% in 50 patients who underwent ASCT for metastatic renal carcinoma (Childs and Srinivasan, 2004). Four of these patients had complete responses. This antitumor effect in renal carcinoma has been confirmed in several subsequent studies (Barkholt et al., 2006; Blaise et al., 2004; Bregni et al., 2002; Hentschke et al., 2003; Pedrazzoli et al., 2002; Rini et al., 2001). The vast majority of these patients had advanced disease, with several metastases and also bulky tumors. For patients with renal cell carcinoma, good prognostic factors included a Karnofsky score of 70% or above, and no more than two metastatic sites at the time of ASCT (Barkholt et al., 2006). To induce a tumor response, the patients who were treated with DLI following transplantation and also developed chronic GVHD had an improved survival. These patients had a 70% survival rate 3 years after ASCT.

D. Reduced Intensity Conditioning and Stem Cell Transplantation for Various Solid Tumors Other solid tumors to which ASCT has induced an antitumor response include metastatic colon carcinoma, ovarian carcinoma, and pancreatic carcinoma (Bay et al., 2002; Omuro et al., 2003; Takahashi et al., 2004; Zetterquist et al., 2001). Sixteen patients with metastatic breast cancer that had progressed after treatment with conventional therapy received ASCT (Bishop et al., 2004). The patients received T cell‐depleted grafts and were given DLI in escalating doses posttransplant. Objective tumor regression occurred 1 month after transplant in six patients, and was attributed to DLI. Regression of tumors occurred concomitantly with the establishment of complete donor T cell engraftment, was associated with the development of GVHD, and was abrogated by a subsequent systemic immunosuppression for GVHD. Ueno et al. (2003) described eight patients with metastatic breast cancer who underwent ASCT after reduced intensity conditioning. After transplant, there was complete response in two patients, mixed response in one and stable disease in the remaining five. Tumor regression was associated with GVHD. In a Japanese study performed on patients with locally advanced or metastatic pancreatic cancer, where curative resection was not possible, seven patients underwent reduced intensity conditioning and ASCT (Kanda et al., 2005). An objective tumor response on computer tomographic scan was observed in two patients and another had a tumor marker response. Marked tumor shrinkage was observed in one of the remaining patients after DLI. Although most patients have died of progressive disease,

40

Olle Ringde´n

significant antitumor effects, often associated with GVHD, were reported. Pancreatic cancer appeared to be sensitive to the graft‐versus‐tumor effect, but ASCT should probably be introduced at an earlier stage. To be more successful in using ASCT in patients with metastatic solid tumors, early treatment—before the disease becomes too widespread—may be important. Furthermore, removal of bulky tumors by surgery, high‐dose‐focused irradiation therapy, and/or radiofrequency ablation may be added. Because the liver seems to be a privileged site for tumor survival and spread of metastases, targeted DLI given via Arteria hepatica may be used to give a better graft‐versus‐tumor effect at this site (Barkholt et al., 2004). In patients with breast cancer, one way of debulking was to perform an autologous stem cell transplant, followed by RIC and ASCT (Carella et al., 2005). With this approach, an overall response rate of 24% and a survival rate of 29% were seen in 17 patients with metastatic breast cancer.

E. Combined Liver Transplantation and Stem Cell Transplantation for Liver Cancer Patients with extensive hepatocellular cancer have a poor life expectancy. In such patients, we have combined liver transplantation with ASCT. In histocompatibility mismatched experimental animals, a combination of T cell‐depleted autologous and allogeneic marrow was found to induce mixed chimerism and tolerance (Ildstad and Sachs, 1984). We used high‐dose chemoradiotherapy and T cell‐depleted autologous and cadaveric major HLA‐ mismatched bone marrow at a ratio of CD34 cells of 1:6 (Ringden et al., 2000b). This resulted in initial chimerism in two patients (Mattsson et al., 2001a; Ringden et al., 2000b). To treat rejection, a low dose of DLI was given in the first patient, who then achieved 100% donor engraftment. Chimera cells did not respond to recipient or donor cells in mixed lymphocyte culture, but reacted vigorously to third‐party cells, suggesting tolerance (Ringden et al., 2000b). The second patient rejected the graft and had autologous recovery (Mattsson et al., 2001a). Both patients were severely immunocompromised, however, and died from fulminant widespread infections. In the first patient, an antitumor effect was obtained, and a‐fetoprotein levels dropped from 440 to normal. At autopsy, no signs of the tumor were detected. In contrast to experimental animals, major histocompatibility mismatched ASCT in humans is not possible and results in profound immune incompetence. We then switched to another approach, performing primary liver transplantation followed by RIC and ASCT, using an HLA‐identical sibling donor or an HLA‐identical unrelated donor (Soderdahl et al., 2003). In the first four patients, conditioning before ASCT was done with fludarabine

Immunotherapy by Allogeneic Stem Cell Transplantation

41

combined with 2 Gy of total body irradiation. Two of these patients rejected the hematopoietic stem cell graft. Subsequently, we switched to conditioning using fludarabine and cyclophosphamide (Ringden et al., 2004c). Of nine patients who underwent combined liver transplantation and ASCT, three died of progressive disease, one died from hemorrhages, and five are still alive between 4 and 48 months after ASCT (Ringden et al., 2004c). Antitumor effects can be seen by the disappearance of tumor markers such as ‐fetoprotein (Soderdahl et al., 2003). Patients who underwent combined liver and ASCT for advanced primary liver cancer had significantly delayed tumor progression compared to patients who underwent liver transplantation alone.

F. Future of Stem Cell Transplantation for Solid Cancers The use of ASCT to treat solid tumors is still in its infancy. To date, progressive cancer has been the main problem. Thus, it seems probable that most solid tumors will not be cured by ASCT. To have a sustained antitumor effect, not only ASCT but also several other modalities will have to be used because tumors have several mechanisms to escape immunologically effective cells. However, it is possible that ASCT with its antitumor effect may prolong survival, compared to conventional anticancer therapy. If so, ASCT may have a place in the treatment repertoire. So far, ASCT for metastatic solid tumors has been used in investigational pilot studies. Now the time has come to do comparative studies between conventional antitumor therapy and ASCT.

VII. MESENCHYMAL STEM CELLS A. Surface Markers and Homing Friedenstein et al. (1968) were the first to identify an adherent fibroblast‐ like population in the adult bone marrow that could regenerate rudiments of bone in vivo. Although these cells may not be true stem cells, because there seems to be an expansion limit, these cells are called MSCs. Apart from bone marrow, MSCs have been isolated from fat, cord blood, fetal liver, blood, and lung (Campagnoli et al., 2001). MSCs are very rare in the human bone marrow and have been proposed to represent 1 in 10,000 nucleated cells (Friedenstein and Kuralesova, 1971). MSCs have the potential to differentiate into several mesenchymal tissues, including bone, cartilage, and fat. It has

42

Olle Ringde´n

been difficult to identify MSCs because there is no specific marker. The cells stain positive for CD73, CD105, CD166, CD90, and CD29 by flow cytometry, and are negative for CD34, CD45, and CD14 (Campagnoli et al., 2001; Haynesworth et al., 1992; Pittenger et al., 1999). MSCs express HLA class I but not HLA class II, unless stimulated with IFN (Haynesworth et al., 1992; Le Blanc et al., 2003a). Human MSCs have high expansion potential, genetic stability, reproducible characteristics, and raised expectations because they show promise as a potential source of therapeutic cells. Following intravenous infusion, MSCs can be detected at low levels and appear to preferentially home in sites of injury (Caplan, 1991; Devine et al., 2003; Gao et al., 2001). Furthermore, MSCs have immunological properties that may be suitable for allogeneic transplantation. MSCs evoke little, if any, immune response (Le Blanc and Ringden, 2005).

B. Immunity and Safety of MSCs MSCs give rise to little lymphocyte proliferation and seem to escape cytotoxic lysis by cytotoxic T‐cells and NK cells (Rasmusson et al., 2003). Although experience is limited, in vivo data also suggest that MSCs are transplantable across major histocompatibility barriers. In baboons, both allogeneic and autologous MSCs were found to engraft in a wide range of tissues (Devine et al., 2003). However, in a xenogenic model, human MSCs were rejected in immunocompetent rats after acute myocardial infarction (Grinnemo et al., 2004). Infusion of autologous and allogeneic MSCs in humans appears to be safe with no acute side effects, and no development of ectopic tissue (Koc et al., 2000; Lazarus et al., 2005).

C. Immunomodulation by MSC It was found that MSCs had immunomodulatory effects in vitro and could inhibit T cell proliferation in mixed lymphocyte cultures, and after lymphocyte stimulation by a wide range of T cell mitogens (Di Nicola et al., 2002; Le Blanc et al., 2003b; McIntosh and Bartholomew, 2000; Tse et al., 2003). In vivo, MSCs have been shown to prolong skin allografts in baboons (Bartholomew et al., 2002). We have shown that allogeneic MSCs from HLA‐identical sibling donors, haploidentical donors, or completely HLA mismatched donors can reverse therapy‐resistant life‐threatening acute GVHD in humans after ASCT (Le Blanc et al., 2004b; Ringden et al., 2006). Acute GVHD disappeared completely in six of eight patients treated with MSCs for therapy‐resistant acute GVHD. Response (by normalization of stool) was seen within 1 or 2 weeks after MSC infusion. However, transplantation tolerance was not

Immunotherapy by Allogeneic Stem Cell Transplantation

43

induced, and discontinuation of immunosuppression showed a rebound phenomenon of acute GVHD (Le Blanc et al., 2004b). Of eight patients treated for life‐threatening acute GVHD, five are still alive between 2 months and 3 years after transplantation (Ringden et al., 2006). Their rate of survival was significantly better than that of 16 patients with steroid‐resistant biopsy‐proven gastrointestinal GVHD who were not treated with MSCs during the same time period. A patient treated for chronic GVHD had only transient response in the liver, but died of Epstein–Barr virus lymphoma in the gut (Ringden et al., 2006). One concern is that by reducing T cell responses, MSCs may pave the way for viral infections, EBV PTLD, and also increase the risk of leukemic relapse. MSCs inhibit the initial phase of alloreactivity in mixed lymphocyte cultures and the formation of cytotoxic T‐lymphocytes (Rasmusson et al., 2003). However, if cytotoxic T‐lymphocytes have already developed, they exert their cytotoxic effect in vitro despite the presence of MSCs. The exact mechanism whereby MSCs induce immunosuppression is not known, although several mechanisms of action have been put forward (Le Blanc and Ringden, 2005).

D. Immunosuppressive Mechanisms by MSCs Among other effects, MSCs reduced the expression of CD4 activation markers, CD25, CD38, and CD69, on phytohemagglutinin‐stimulated lymphocytes (Le Blanc et al., 2004a). Numbers of regulatory T cells are increased in the presence of MSCs (Aggarwal and Pittenger, 2005; Maccario et al., 2005). Activated dendritic cells have been shown to have decreased tumor necrosis factor‐a and IL‐12 secretion and increased IL‐10 secrection. T‐helper cell secretion by IFN and IL‐5 was reduced, and IL‐4 was increased in the presence of MSCs. Suppression by MSCs appears to be mediated by a soluble factor because suppression still occurs if MSCs and lymphocytes are separated in a transwell system (Di Nicola et al., 2002; Rasmusson et al., 2003; Tse et al., 2003). Several factors have been suggested to have this role, including transforming growth factor‐a, hepatocyte growth factor, prostaglandin E2, indoleamine 2,3‐dioxygenase, IL‐10, IL‐2, IL‐6, IL‐8, stem‐cell‐derived factor 1, and vascular endothelial growth factor (Le Blanc and Ringden, 2005). However, several investigators have failed to confirm these data, and there have been contradictory findings (Le Blanc et al., 2004a; Tse, et al., 2003). The controversy may be due to the use of different types of MSCs, culture conditions, MSC doses, kinetics, different lymphocyte subpopulations, different species, and different stimuli. For instance, MSCs increase the levels of IL‐2 and soluble IL‐2 receptor in mixed lymphocyte cultures, but decrease the levels in lymphocytes

44

Olle Ringde´n

stimulated with phytohemagglutinin (Rasmusson et al., 2005). Furthermore, IL‐10 levels increased in MLCs cocultured with MSCs, but not in lymphocytes stimulated with PHA. The addition of indomethacin, which inhibits prostaglandin E2 synthesis, restored part of the inhibition by MSCs in lymphocytes stimulated with PHA, but not in MLCs. Furthermore, the dose of MSCs has a major effect on lymphocyte proliferation in different systems. Some investigators have had to use 10 times as many MSCs as lymphocytes, and some have used the same amount, to demonstrate effects (Di Nicola et al., 2002; Maccario et al., 2005; Maitra et al., 2004). We and others have found that high doses of MSCs are immunosuppressive, while low numbers sometimes enhance proliferation (Le Blanc et al., 2003b; Liu et al., 2004). The vast majority of studies using MSCs have been from in vitro experiments and more in vivo studies will be required before the mechanism and role of MSCs have been thoroughly investigated.

VIII. FUTURE DIRECTIONS GVHD is still a major threat to successful ASCT. Regulatory T cells (CD4þ CD25þ cells) may suppress GVHD (Blazar and Taylor, 2005). Clinically, regulatory T cells may be used to prevent GVHD, to promote alloengraftment, and to restore peripheral tolerance following ASCT. Greater understanding is required before this can be established in the clinic. In animals, tolerance can be achieved even in a xenograft model (Ildstad and Sachs, 1984), but fails in humans mismatched for major HLA antigens (Ringden et al., 2000b). ASCT beyond haploidentical transplants can probably not be achieved and is not necessary, as haploidentical transplants can be identified for most patients. One way of tackling the problem of induction of GVHD may be to introduce a suicide gene into the T cells of the graft (Gallot et al., 1996; Tiberghien et al., 1994). This approach has been used directly by giving a T cell‐depleted graft and adding thymidine kinase gene‐manipulated T cells. When GVHD appeared, the patients were treated with ganciclovir and the GVHD‐reactive T cells were killed. One concern has been that these culture‐expanded, gene‐ manipulated cells do not react against EBV and will therefore not protect the patient against EBV disease or EBV PTLD (Carlens et al., 2002). The method needs to be refined or combined with cytotoxic T cell (CTL) therapy against EBV. Because MSCs have immunomodulatory and tissue‐repairing effects, they may be used for both prevention and treatment of GVHD. Prospective randomized multicenter studies using MSCs have recently started. There has been a general trend toward reduced transplant‐related mortality following ASCT in the last two decades due to better prevention of GVHD and better diagnosis and treatment of infectious complications. However,

Immunotherapy by Allogeneic Stem Cell Transplantation

45

there has been no clear reduction in disease relapse. Thus, harnessing the graft‐versus‐leukemia effect and separating it from GVHD reactions remain an important goal for improving the success of ASCT. DLI can induce remission, especially with molecular relapse in patients with CML, but it can also induce GVHD (Kolb et al., 1995). Thus, it is important to drive the donor immune system in a specific fashion, against antigens exclusively or preferentially presented by tumor cells, without damaging normal host cells. Antigens expressed on leukemia cells or other tumor cells may provide targets for CTL therapy and possibly induce an antitumor effect without GVHD. Such antigens may be tumor‐specific, for example, BCR‐ABL in CML. An alternative to tumor‐specific antigens may be CTLs generated against tissues such as the prostate and pancreas for patients with prostate carcinoma and pancreas carcinoma. It seems much more difficult to develop tumor‐ specific CTL therapy than CTL therapy against EBV, for example, which has already been used successfully for the treatment and prevention of EBV‐associated lymphoma (Gustafsson et al., 2000; Rooney et al., 1995). The adoptive transfer of EBV‐specific CTLs has established the safety of this approach, showing that expanded T cells can survive and perform effectively in vivo. An alternative approach, which has been used for cytomegalovirus (CMV), is to isolate virus‐specific CTLs from an ASCT donor before transplant, to expand the cells in vitro, and to give them intravenously to the transplant recipient to institute protection against the virus (in this case CMV) after transplantation (Riddell et al., 1992). In a similar way, adoptive transfer of ex vivo‐expanded tumor‐specific CTLs that are generated from the donor could be used to enhance a graft‐versus‐leukemia or graft‐versus‐tumor effect in the recipient after ASCT. This approach has been used successfully in mice, where it was demonstrated that tumor‐specific immunity could be boosted following posttransplant tumor immunization. In a minor histocompatibility antigen disparate mouse allotransplant model, immunization of the recipient (after ASCT) against either leukemia or fibrosarcoma resulted in enhanced antitumor activity without exacerbating GVHD (Anderson et al., 2000). In mice, it has also been demonstrated that NK cells have the capacity to reach sites of tumor growth, to infiltrate, interact with, and then selectively kill malignant cells in solid tissues without affecting normal recipient cells. As discussed above, alloreactive NK cells have been found to exert a potent antileukemia effect against acute myeloid leukemia (Ruggeri et al., 2002). In vitro, it has been demonstrated that human KIR‐ligand mismatched, alloreactive NK cells are capable of lysing melanoma and renal cell carcinoma cell lines (Igarashi et al., 2004). KIR‐incompatible allogeneic NK cells may have superior antitumor effects against solid tumors, compared to autologous NK cells. It is probable that alloreactive NK cells may be used to boost an antitumor effect after ASCT when there is a KIR‐ligand mismatched tumor cell. We require a better understanding of how to use

46

Olle Ringde´n

antitumor CTLs in an optimal way. In a mouse adoptive cell transfer model, IL‐7 and IL‐15 were found to be required for T cell antitumor activity (Gattinoni et al., 2005). During the pancytopenic phase after ASCT, patients are isolated in the hospital by reversed isolation or in laminar airflow rooms to prevent infectious complications. We challenged this approach by letting the patients go home after conditioning (Svahn et al., 2002). Compared to controls who were protected in the hospital, the patients treated at home had fewer days on total parenteral nutrition, less acute GVHD, lower transplant‐related mortality, and lower costs. Despite a decrease in acute GVHD, home care did not abrogate the graft‐versus‐leukemia effect, and rate of relapse was the same as in patients treated in the hospital. Also, there was improved long‐term survival in patients treated at home (Svahn et al., 2005). There may be several reasons for the reduced risk of GVHD in patients treated at home such as better nutrition, less microorganisms, and less stress. The latter may stimulate cytokine release, which could in turn trigger acute GVHD. Home care will be adopted in more centers, and this will enable a prospective randomized study and permit further studies to evaluate the mechanism behind the positive effects of home care. It is expected that home care will successfully replace isolation after ASCT in the future. Because of the advent of genomic tissue typing, unrelated donors can be better matched to the patient—it reduces the risk of GVHD and infections and improves survival (Ringden et al., 2004b). The pool of unrelated donors has increased and is now around 10 million donors worldwide. Furthermore, cord blood transplants are being increasingly used. Tissue typing is becoming less stringent, and 2 HLA antigen mismatches are acceptable (Gluckman et al., 1997; Wagner et al., 1996). The cell dose is more important, and should be above 2  107 nucleated cells/kg recipient weight. When the cell dose is too low, dual cord blood transplants may overcome this problem (Barker et al., 2005). If an unrelated donor or a cord blood graft is not available, there is always the possibility of a haploidentical transplant (Ruggeri et al., 2002). Thus, today it is possible to transplant almost all patients who need an ASCT. In the future, with an ever‐increasing pool of unrelated donors, it will be possible to match donors and recipients not only for HLA class I and class II antigens, but also for viral immunity in recipient and donor, and for KIR‐ligand incompatibilities when desired. Because many disorders which can be cured by ASCT occur in elderly and disabled patients, RIC conditioning enables ASCT in these patient cohorts. Prospective randomized trials are required to evaluate the role of RIC in ASCT, as compared to heavier chemoirradiotherapy. Immunosuppression and repeated infections are major problems after ASCT. One alternative to development of EBV‐, CMV‐, or adenovirus‐specific CTL therapy may be to deplete alloreactive donor T cells (Cavazzana‐Calvo

Immunotherapy by Allogeneic Stem Cell Transplantation

47

et al., 1990; Mavroudis et al., 1996; Solomon et al., 2005). Using MLCs, it is possible to deplete alloreactivity by treatment with anti‐CD25 in vitro. This works both in an HLA‐mismatched and HLA‐matched setting, while maintaining third‐party alloresponses and reactivity against viruses and other microorganisms. Preliminary clinical trials have demonstrated successful depletion of host‐reactive donor T cells with conservation of third‐party responses, following ASCT (Solomon et al., 2005). Despite the progress and success in using ASCT for immunotherapy against cancer, the major challenges—such as severe GVHD, leukemia, or tumor recurrence, and immunoincompetence—still remain. Several promising experimental approaches to these problems are being developed and explored today. Exciting developments in the field can therefore be expected in the near future.

ACKNOWLEDGMENTS I wish to thank Inger Hammarberg for skillful preparation of this chapter. The author is grateful for financial support from the Swedish Cancer Society (0070‐B02–16XAC), the Children’s Cancer Foundation (03/039), the Swedish Research Council (K2003–32X‐05971– 23A), the Cancer Society of Stockholm, the Swedish Cancer and Allergy Foundation, and the Karolinska Institute.

REFERENCES Aggarwal, S., and Pittenger, M. F. (2005). Human mesenchymal stem cells modulate allogeneic immune cell responses. Blood 105, 1815–1822. Anasetti, C., Hansen, J. A., Waldmann, T. A., Appelbaum, F. R., Davis, J., Deeg, H. J., Doney, K., Martin, P. J., Nash, R., Storb, R., Sullivan, K. M., Witherspoon, R. P., et al. (1994). Treatment of acute graft‐versus‐host disease with humanized anti‐Tac: An antibody that binds to the interleukin‐2 receptor. Blood 84, 1320–1327. Anderson, L. D., Jr., Savary, C. A., and Mullen, C. A. (2000). Immunization of allogeneic bone marrow transplant recipients with tumor cell vaccines enhances graft‐versus‐tumor activity without exacerbating graft‐versus‐host disease. Blood 95, 2426–2433. Appelbaum, F. R. (1995). Allogeneic marrow transplantation and the use of hematopoietic growth factors. Stem Cells 13, 344–350. Appelbaum, F. R., Sullivan, K. M., Buckner, C. D., Clift, R. A., Deeg, H. J., Fefer, A., Hill, R., Mortimer, J., Neiman, P. E., Sanders, J. E., et al. (1987). Treatment of malignant lymphoma in 100 patients with chemotherapy, total body irradiation, and marrow transplantation. J. Clin. Oncol. 5, 1340–1347. Aschan, J., Ringden, O., Sundberg, B., Gahrton, G., Ljungman, P., and Winiarski, J. (1991). Methotrexate combined with cyclosporin A decreases graft‐versus‐host disease, but increases leukemic relapse compared to monotherapy. Bone Marrow Transplant. 7, 113–119.

48

Olle Ringde´n

Aschan, J., Ringden, O., Sundberg, B., Klaesson, S., Ljungman, P., and Lonnqvist, B. (1993). Increased risk of relapse in patients with chronic myelogenous leukemia given T‐cell depleted marrow compared to methotrexate combined with cyclosporin or monotherapy for the prevention of graft‐versus‐host disease. Eur. J. Haematol. 50, 269–274. Bacigalupo, A., Van Lint, M. T., Occhini, D., Gualandi, F., Lamparelli, T., Sogno, G., Tedone, E., Frassoni, F., Tong, J., and Marmont, A. M. (1991). Increased risk of leukemia relapse with high‐dose cyclosporine A after allogeneic marrow transplantation for acute leukemia. Blood 77, 1423–1428. Bader, P., Beck, J., Frey, A., Schlegel, P. G., Hebarth, H., Handgretinger, R., Einsele, H., Niemeyer, C., Benda, N., Faul, C., Kanz, L., Niethammer, D., et al. (1998). Serial and quantitative analysis of mixed hematopoietic chimerism by PCR in patients with acute leukemias allows the prediction of relapse after allogeneic BMT. Bone Marrow Transplant. 21, 487–495. Bader, P., Niethammer, D., Willasch, A., Kreyenberg, H., and Klingebiel, T. (2005). How and when should we monitor chimerism after allogeneic stem cell transplantation? Bone Marrow Transplant. 35, 107–119. Barker, J. N., Weisdorf, D. J., DeFor, T. E., Blazar, B. R., McGlave, P. B., Miller, J. S., Verfaillie, C. M., and Wagner, J. E. (2005). Transplantation of 2 partially HLA‐matched umbilical cord blood units to enhance engraftment in adults with hematologic malignancy. Blood 105, 1343–1347. Barkholt, L., Danielsson, R., Calissendorff, B., Svensson, L., Malihi, R., Remberger, M., Uzunel, M., Thorne, A., and Ringden, O. (2004). Indium‐111‐labelled donor‐lymphocyte infusion by way of hepatic artery and radio‐frequency ablation against liver metastases of renal and colon carcinoma after allogeneic hematopoietic stem‐cell transplantation. Transplantation 78, 697–703. Barkholt, L., Bregni, M., Remberger, M., Blaise, D., Peccatori, J., Massenkeil, G., Pedrazzoli, P., Zambelli, A., Bay, J. O., Francois, S., Martino, R., Bengala, C., et al. (2006). Allogeneic hematopoetic stem cell transplantation for metastatic round carcinoma in Europe. Annals of Oncology 17, 1134–1140. Barnes, D. W., Corp, M. J., Loutit, J. F., and Neal, F. E. (1956). Treatment of murine leukaemia with X rays and homologous bone marrow; preliminary communication. Br. Med. J. 32, 626–627. Barrett, A. J., Horowitz, M. M., Gale, R. P., Biggs, J. C., Camitta, B. M., Dicke, K. A., Gluckman, E., Good, R. A., Herzig, R. H., Lee, M. B., et al. (1989). Marrow transplantation for acute lymphoblastic leukemia: Factors affecting relapse and survival. Blood 74, 862–871. Barrett, A. J., Ringden, O., Zhang, M. J., Bashey, A., Cahn, J. Y., Cairo, M. S., Gale, R. P., Gratwohl, A., Locatelli, F., Martino, R., Schultz, K. R., and Tiberghien, P. (2000). Effect of nucleated marrow cell dose on relapse and survival in identical twin bone marrow transplants for leukemia. Blood 95, 3323–3327. Bartholomew, A., Sturgeon, C., Siatskas, M., Ferrer, K., McIntosh, K., Patil, S., Hardy, W., Devine, S., Ucker, D., Deans, R., Moseley, A., and Hoffman, R. (2002). Mesenchymal stem cells suppress lymphocyte proliferation in vitro and prolong skin graft survival in vivo. Exp. Hematol. 30, 42–48. Bay, J. O., Fleury, J., Choufi, B., Tournilhac, O., Vincent, C., Bailly, C., Dauplat, J., Viens, P., Faucher, C., and Blaise, D. (2002). Allogeneic hematopoietic stem cell transplantation in ovarian carcinoma: Results of five patients. Bone Marrow Transplant. 30, 95–102. Benito, A. I., Furlong, T., Martin, P. J., Anasetti, C., Appelbaum, F. R., Doney, K., Nash, R. A., Papayannopoulou, T., Storb, R., Sullivan, K. M., Witherspoon, R., and Deeg, H. J. (2001). Sirolimus (rapamycin) for the treatment of steroid‐refractory acute graft‐versus‐host disease. Transplantation 72, 1924–1929. Ben‐Yosef, R., Or, R., Nagler, A., and Slavin, S. (1996). Graft‐versus‐tumour and graft‐versus‐ leukaemia effect in patient with concurrent breast cancer and acute myelocytic leukaemia. Lancet 348, 1242–1243.

Immunotherapy by Allogeneic Stem Cell Transplantation

49

Bertheas, M. F., Lafage, M., Levy, P., Blaise, D., Stoppa, A. M., Viens, P., Mannoni, P., and Maraninchi, D. (1991). Influence of mixed chimerism on the results of allogeneic bone marrow transplantation for leukemia. Blood 78, 3103–3106. Bishop, M. R., Tarantolo, S. R., Pavletic, Z. S., Lynch, J. C., Morris, M. E., Zacharias, D., Armitage, J. O., and Kessinger, A. (2000). Filgrastim as an alternative to donor leukocyte infusion for relapse after allogeneic stem‐cell transplantation. J. Clin. Oncol. 18, 2269–2272. Bishop, M. R., Fowler, D. H., Marchigiani, D., Castro, K., Kasten‐Sportes, C., Steinberg, S. M., Gea‐Banacloche, J. C., Dean, R., Chow, C. K., Carter, C., Read, E. J., Leitman, S., et al. (2004). Allogeneic lymphocytes induce tumor regression of advanced metastatic breast cancer. J. Clin. Oncol. 22, 3886–3892. Blaise, D., Bay, J. O., Faucher, C., Michallet, M., Boiron, J. M., Choufi, B., Cahn, J. Y., Gratecos, N., Sotto, J. J., Francois, S., Fleury, J., Mohty, M., et al. (2004). Reduced‐intensity preparative regimen and allogeneic stem cell transplantation for advanced solid tumors. Blood 103, 435–441. Blazar, B. R., and Taylor, P. A. (2005). Regulatory T cells. Biol. Blood Marrow Transplant. 11, 46–49. Bornhauser, M., Schwerdtfeger, R., Martin, H., Frank, K. H., Theuser, C., and Ehninger, G. (2004). Role of KIR ligand incompatibility in hematopoietic stem cell transplantation using unrelated donors. Blood 103, 2860–2861; author reply 2862. Bostrom, L., Ringden, O., Gratama, J. W., Jacobsen, N., Prentice, H. G., Zwaan, F. E., and Nilsson, B. (1990a). A role of herpes virus serology for the development of acute graft‐ versus‐host disease. Leukaemia Working Party of the European Group for Bone Marrow Transplantation. Bone Marrow Transplant. 5, 321–326. Bostrom, L., Ringden, O., Jacobsen, N., Zwaan, F., and Nilsson, B. (1990b). A European multicenter study of chronic graft‐versus‐host disease. The role of cytomegalovirus serology in recipients and donors: Acute graft‐versus‐host disease, and splenectomy. Transplantation 49, 1100–1105. Bregni, M., Dodero, A., Peccatori, J., Pescarollo, A., Bernardi, M., Sassi, I., Voena, C., Zaniboni, A., Bordignon, C., and Corradini, P. (2002). Nonmyeloablative conditioning followed by hematopoietic cell allografting and donor lymphocyte infusions for patients with metastatic renal and breast cancer. Blood 99, 4234–4236. Bumpus, H. C. (1928). The apparent disappearance of pulmonary metastasis in a case of hypernephroma following nephrectomy. J. Urol. 20, 185. Campagnoli, C., Roberts, I. A., Kumar, S., Bennett, P. R., Bellantuono, I., and Fisk, N. M. (2001). Identification of mesenchymal stem/progenitor cells in human first‐trimester fetal blood, liver, and bone marrow. Blood 98, 2396–2402. Canninga‐van Dijk, M. R., van der Straaten, H. M., Fijnheer, R., Sanders, C. J., van den Tweel, J. G., and Verdonck, L. F. (2004). Anti‐CD20 monoclonal antibody treatment in 6 patients with therapy‐refractory chronic graft‐versus‐host disease. Blood 104, 2603–2606. Caplan, A. I. (1991). Mesenchymal stem cells. J. Orthop. Res. 9, 641–650. Carella, A. M., Beltrami, G., Corsetti, M. T., Nati, S., Musto, P., Scalzulli, P., Gonella, R., Ballestrero, A., and Patrone, F. (2005). Reduced intensity conditioning for allograft after cytoreductive autograft in metastatic breast cancer. Lancet 366, 318–320. Carlens, S., Ringden, O., Remberger, M., Lonnqvist, B., Hagglund, H., Klaesson, S., Mattsson, J., Svahn, B. M., Winiarski, J., Ljungman, P., and Aschan, J. (1998). Risk factors for chronic graft‐ versus‐host disease after bone marrow transplantation: A retrospective single centre analysis. Bone Marrow Transplant. 22, 755–761. Carlens, S., Aschan, J., Remberger, M., Dilber, M., and Ringden, O. (1999). Low‐dose cyclosporine of short duration increases the risk of mild and moderate GVHD and reduces the risk of relapse in HLA‐identical sibling marrow transplant recipients with leukaemia. Bone Marrow Transplant. 24, 629–635.

50

Olle Ringde´n

Carlens, S., Remberger, M., Aschan, J., and Ringden, O. (2001). The role of disease stage in the response to donor lymphocyte infusions as treatment for leukemic relapse. Biol. Blood Marrow Transplant. 7, 31–38. Carlens, S., Liu, D., Ringden, O., Aschan, J., Christensson, B., Levitsky, V., and Dilber, M. S. (2002). Cytolytic T cell reactivity to Epstein‐Barr virus is lost during in vitro T cell expansion. J. Hematother. Stem Cell Res. 11, 669–674. Cavazzana‐Calvo, M., Fromont, C., Le Deist, F., Lusardi, M., Coulombel, L., Derocq, J. M., Gerota, I., Griscelli, C., and Fischer, A. (1990). Specific elimination of alloreactive T cells by an anti‐interleukin‐2 receptor B chain‐specific immunotoxin. Transplantation 50, 1–7. Childs, R. W., and Srinivasan, R. (2004). Allogeneic hematopoetic cell transplantation for solid tumors. In “Thomas’ Hematopoietic Cell Transplantation” (K. G. Blume, S. J. Forman, and F. R. Appelbaum, Eds.), pp. 1177–1187. Blackwell Publishing Ltd., Oxford. Childs, R. W., Clave, E., Tisdale, J., Plante, M., Hensel, N., and Barrett, J. (1999). Successful treatment of metastatic renal cell carcinoma with a nonmyeloablative allogeneic peripheral‐ blood progenitor‐cell transplant: Evidence for a graft‐versus‐tumor effect. J. Clin. Oncol. 17, 2044–2049. Cooper, M. A., Fehniger, T. A., Fuchs, A., Colonna, M., and Caligiuri, M. A. (2004). NK cell and DC interactions. Trends Immunol. 25, 47–52. Cross, N. C., Feng, L., Chase, A., Bungey, J., Hughes, T. P., and Goldman, J. M. (1993). Competitive polymerase chain reaction to estimate the number of BCR‐ABL transcripts in chronic myeloid leukemia patients after bone marrow transplantation. Blood 82, 1929–1936. Curtis, R. E., Rowlings, P. A., Deeg, H. J., Shriner, D. A., Socie, G., Travis, L. B., Horowitz, M. M., Witherspoon, R. P., Hoover, R. N., Sobocinski, K. A., Fraumeni, J. F., Jr., and Boice, J. D., Jr. (1997). Solid cancers after bone marrow transplantation. N. Engl. J. Med. 336, 897–904. Davies, S. M., Ruggieri, L., DeFor, T., Wagner, J. E., Weisdorf, D. J., Miller, J. S., Velardi, A., and Blazar, B. R. (2002). Evaluation of KIR ligand incompatibility in mismatched unrelated donor hematopoietic transplants. Killer immunoglobulin‐like receptor. Blood 100, 3825–3827. Dazzi, F., Szydlo, R. M., Craddock, C., Cross, N. C., Kaeda, J., Chase, A., Olavarria, E., van Rhee, F., Kanfer, E., Apperley, J. F., and Goldman, J. M. (2000a). Comparison of single‐dose and escalating‐dose regimens of donor lymphocyte infusion for relapse after allografting for chronic myeloid leukemia. Blood 95, 67–71. Dazzi, F., Szydlo, R. M., Cross, N. C., Craddock, C., Kaeda, J., Kanfer, E., Cwynarski, K., Olavarria, E., Yong, A., Apperley, J. F., and Goldman, J. M. (2000b). Durability of responses following donor lymphocyte infusions for patients who relapse after allogeneic stem cell transplantation for chronic myeloid leukemia. Blood 96, 2712–2716. de Gast, G. C., Gratama, J. W., Ringden, O., and Gluckman, E. (1987). The multifactorial etiology of graft‐versus‐host disease. Immunol. Today 8, 209–212. Deeg, H. J., Blazar, B. R., Bolwell, B. J., Long, G. D., Schuening, F., Cunningham, J., Rifkin, R. M., Abhyankar, S., Briggs, A. D., Burt, R., Lipani, J., Roskos, L. K., et al. (2001). Treatment of steroid‐refractory acute graft‐versus‐host disease with anti‐CD147 monoclonal antibody ABX‐CBL. Blood 98, 2052–2058. den Haan, J. M., Meadows, L. M., Wang, W., Pool, J., Blokland, E., Bishop, T. L., Reinhardus, C., Shabanowitz, J., Offringa, R., Hunt, D. F., Engelhard, V. H., and Goulmy, E. (1998). The minor histocompatibility antigen HA‐1: A diallelic gene with a single amino acid polymorphism. Science 279, 1054–1057. Devine, S. M., Cobbs, C., Jennings, M., Bartholomew, A., and Hoffman, R. (2003). Mesenchymal stem cells distribute to a wide range of tissues following systemic infusion into nonhuman primates. Blood 101, 2999–3001.

Immunotherapy by Allogeneic Stem Cell Transplantation

51

Di Nicola, M., Carlo‐Stella, C., Magni, M., Milanesi, M., Longoni, P. D., Matteucci, P., Grisanti, S., and Gianni, A. M. (2002). Human bone marrow stromal cells suppress T‐lymphocyte proliferation induced by cellular or nonspecific mitogenic stimuli. Blood 99, 3838–3843. Dolstra, H., Fredrix, H., Preijers, F., Goulmy, E., Figdor, C. G., de Witte, T. M., and van de Wiel‐van Kemenade, E. (1997). Recognition of a B cell leukemia‐associated minor histocompatibility antigen by CTL. J. Immunol. 158, 560–565. Eibl, B., Schwaighofer, H., Nachbaur, D., Marth, C., Gachter, A., Knapp, R., Bock, G., Gassner, C., Schiller, L., Petersen, F., and Niederwieser, D. (1996). Evidence for a graft‐versus‐tumor effect in a patient treated with marrow ablative chemotherapy and allogeneic bone marrow transplantation for breast cancer. Blood 88, 1501–1508. Farag, S. S., Fehniger, T. A., Ruggeri, L., Velardi, A., and Caligiuri, M. A. (2002). Natural killer cell receptors: New biology and insights into the graft‐versus‐leukemia effect. Blood 100, 1935–1947. Fefer, A., Einstein, A. B., and Cheever, M. A. (1976). Adoptive chemoimmunotherapy of cancer in animals: A review of results, principles, and problems. Ann. NY Acad. Sci. 277, 492–504. Fefer, A., Sullivan, K. M., Weiden, P. L., Buckner, C. D., Schoch, G., Storb, R., and Thomas, E. D. (1987). In “Cellular Immunotherapy of Cancer” (R. L. Truitt, R. P. Gale, and M. M. Bortin, Eds.), pp. 401–408. AR Liss, New York. Ferrara, J. L., Levy, R., and Chao, N. J. (1999). Pathophysiologic mechanisms of acute graft‐ vs.‐host disease. Biol. Blood Marrow Transplant. 5, 347–356. Friedenstein, A. J., Petrakova, K. V., Kurolesova, A. I., and Frolova, G. P. (1968). Heterotopic of bone marrow. Analysis of precursor cells for osteogenic and hematopoietic tissues. Transplantation 6, 230–247. Friedenstein, A., and Kuralesova, A. I. (1971). Osteogenic precursor cells of bone marrow in radiation chimeras. Transplantation 12, 99–108. Gabert, J., Thuret, I., Lafage, M., Carcassonne, Y., Maraninchi, D., and Mannoni, P. (1989). Detection of residual bcr/abl translocation by polymerase chain reaction in chronic myeloid leukaemia patients after bone‐marrow transplantation. Lancet 2, 1125–1128. Gajewski, J. L., Phillips, G. L., Sobocinski, K. A., Armitage, J. O., Gale, R. P., Champlin, R. E., Herzig, R. H., Hurd, D. D., Jagannath, S., Klein, J. P., Lazarus, H. M., McCarthy, P. L., Jr., et al. (1996). Bone marrow transplants from HLA‐identical siblings in advanced Hodgkin’s disease. J. Clin. Oncol. 14, 572–578. Gale, R. P., Bortin, M. M., van Bekkum, D. W., Biggs, J. C., Dicke, K. A., Gluckman, E., Good, R. A., Hoffmann, R. G., Kay, H. E., Kersey, J. H., et al. (1987). Risk factors for acute graft‐ versus‐host disease. Br. J. Haematol. 67, 397–406. Gale, R. P., Horowitz, M. M., Ash, R. C., Champlin, R. E., Goldman, J. M., Rimm, A. A., Ringden, O., Stone, J. A., and Bortin, M. M. (1994). Identical‐twin bone marrow transplants for leukemia. Ann. Intern. Med. 120, 646–652. Gallot, G., Hallet, M. M., Gaschet, J., Moreau, J. F., Vivien, R., Bonneville, M., Milpied, N., and Vie, H. (1996). Human HLA‐specific T‐cell clones with stable expression of a suicide gene: A possible tool to drive and control a graft‐versus‐host‐graft‐versus‐leukemia reaction? Blood 88, 1098–1103. Gao, J., Dennis, J. E., Muzic, R. F., Lundberg, M., and Caplan, A. I. (2001). The dynamic in vivo distribution of bone marrow‐derived mesenchymal stem cells after infusion. Cells Tissues Organs. 169, 12–20. Gattinoni, L., Finkelstein, S. E., Klebanoff, C. A., Antony, P. A., Palmer, D. C., Spiess, P. J., Hwang, L. N., Yu, Z., Wrzesinski, C., Heimann, D. M., Surh, C. D., Rosenberg, S. A., et al. (2005). Removal of homeostatic cytokine sinks by lymphodepletion enhances the efficacy of adoptively transferred tumor‐specific CD8þ T cells. J. Exp. Med. 202, 907–912.

52

Olle Ringde´n

Giebel, S., Locatelli, F., Lamparelli, T., Velardi, A., Davies, S., Frumento, G., Maccario, R., Bonetti, F., Wojnar, J., Martinetti, M., Frassoni, F., Giorgiani, G., et al. (2003). Survival advantage with KIR ligand incompatibility in hematopoietic stem cell transplantation from unrelated donors. Blood 102, 814–819. Giralt, S., Estey, E., Albitar, M., van Besien, K., Rondon, G., Anderlini, P., O’Brien, S., Khouri, I., Gajewski, J., Mehra, R., Claxton, D., Andersson, B., et al. (1997). Engraftment of allogeneic hematopoietic progenitor cells with purine analog‐containing chemotherapy: Harnessing graft‐versus‐leukemia without myeloablative therapy. Blood 89, 4531–4536. Gluckman, E., Rocha, V., Boyer‐Chammard, A., Locatelli, F., Arcese, W., Pasquini, R., Ortega, J., Souillet, G., Ferreira, E., Laporte, J. P., Fernandez, M., and Chastang, C. (1997). Outcome of cord‐blood transplantation from related and unrelated donors. Eurocord Transplant Group and the European Blood and Marrow Transplantation Group. N. Engl. J. Med. 337, 373–381. Glucksberg, H., Storb, R., Fefer, A., Buckner, C. D., Neiman, P. E., Clift, R. A., Lerner, K. G., and Thomas, E. D. (1974). Clinical manifestations of graft‐versus‐host disease in human recipients of marrow from HL‐A‐matched sibling donors. Transplantation 18, 295–304. Goldman, J. M., Gale, R. P., Horowitz, M. M., Biggs, J. C., Champlin, R. E., Gluckman, E., Hoffmann, R. G., Jacobsen, S. J., Marmont, A. M., McGlave, P. B., et al. (1988). Bone marrow transplantation for chronic myelogenous leukemia in chronic phase. Increased risk for relapse associated with T‐cell depletion. Ann. Intern. Med. 108, 806–814. Goulmy, E. (1997). Human minor histocompatibility antigens: New concepts for marrow transplantation and adoptive immunotherapy. Immunol. Rev. 157, 125–140. Gratwohl, A., Hermans, J., Apperley, J., Arcese, W., Bacigalupo, A., Bandini, G., di Bartolomeo, P., Boogaerts, M., Bosi, A., Carreras, E., et al. (1995). Acute graft‐versus‐host disease: Grade and outcome in patients with chronic myelogenous leukemia. Working Party Chronic Leukemia of the European Group for Blood and Marrow Transplantation. Blood 86, 813–818. Grinnemo, K. H., Mansson, A., Dellgren, G., Klingberg, D., Wardell, E., Drvota, V., Tammik, C., Holgersson, J., Ringden, O., Sylven, C., and Le Blanc, K. (2004). Xenoreactivity and engraftment of human mesenchymal stem cells transplanted into infarcted rat myocardium. J. Thorac. Cardiovasc. Surg. 127, 1293–1300. Groth, C. G., Gahrton, G., Lundgren, G., Moller, E., Pihlstedt, P., Ringden, O., and Sundelin, P. (1979). Successful treatment with prednisone and graft‐versus‐host disease in an allogeneic bone‐marrow transplant recipient. Scand. J. Haematol. 22, 333–338. Gustafsson, A., Levitsky, V., Zou, J. Z., Frisan, T., Dalianis, T., Ljungman, P., Ringden, O., Winiarski, J., Ernberg, I., and Masucci, M. G. (2000). Epstein‐Barr virus (EBV) load in bone marrow transplant recipients at risk to develop posttransplant lymphoproliferative disease: Prophylactic infusion of EBV‐specific cytotoxic T cells. Blood 95, 807–814. Hansen, J. A., Gooley, T. A., Martin, P. J., Appelbaum, F., Chauncey, T. R., Clift, R. A., Petersdorf, E. W., Radich, J., Sanders, J. E., Storb, R. F., Sullivan, K. M., and Anasetti, C. (1998). Bone marrow transplants from unrelated donors for patients with chronic myeloid leukemia. N. Engl. J. Med. 338, 962–968. Haynesworth, S. E., Baber, M. A., and Caplan, A. I. (1992). Cell surface antigens on human marrow‐derived mesenchymal cells are detected by monoclonal antibodies. Bone 13, 69–80. Hentschke, P., Barkholt, L., Uzunel, M., Mattsson, J., Wersall, P., Pisa, P., Martola, J., Albiin, N., Wernerson, A., Soderberg, M., Remberger, M., Thorne, A., et al. (2003). Low‐intensity conditioning and hematopoietic stem cell transplantation in patients with renal and colon carcinoma. Bone Marrow Transplant. 31, 253–261. Herve, P., Wijdenes, J., Bergerat, J. P., Bordigoni, P., Milpied, N., Cahn, J. Y., Clement, C., Beliard, R., Morel‐Fourrier, B., Racadot, E., et al. (1990). Treatment of corticosteroid resistant acute graft‐versus‐host disease by in vivo administration of anti‐interleukin‐2 receptor monoclonal antibody (B‐B10). Blood 75, 1017–1023.

Immunotherapy by Allogeneic Stem Cell Transplantation

53

Horowitz, M. M., Gale, R. P., Sondel, P. M., Goldman, J. M., Kersey, J., Kolb, H. J., Rimm, A. A., Ringden, O., Rozman, C., Speck, B., et al. (1990). Graft‐versus‐leukemia reactions after bone marrow transplantation. Blood 75, 555–562. Hsu, K. C., Keever‐Taylor, C. A., Wilton, A., Pinto, C., Heller, G., Arkun, K., O’Reilly, R. J., Horowitz, M. M., and Dupont, B. (2005). Improved outcome in HLA‐identical sibling hematopoietic stem‐cell transplantation for acute myelogenous leukemia predicted by KIR and HLA genotypes. Blood 105, 4878–4884. Igarashi, T., Wynberg, J., Srinivasan, R., Becknell, B., McCoy, J. P., Jr., Takahashi, Y., Suffredini, D. A., Linehan, W. M., Caligiuri, M. A., and Childs, R. W. (2004). Enhanced cytotoxicity of allogeneic NK cells with killer immunoglobulin‐like receptor ligand incompatibility against melanoma and renal cell carcinoma cells. Blood 104, 170–177. Ildstad, S. T., and Sachs, D. H. (1984). Reconstitution with syngeneic plus allogeneic or xenogeneic bone marrow leads to specific acceptance of allografts or xenografts. Nature 307, 168–170. Kanda, Y., Komatsu, Y., Akahane, M., Kojima, S., Asano‐Mori, Y., Tada, M., Oshima, K., Isayama, H., Ogawa, S., Motokura, T., Chiba, S., Ohtomo, K., et al. (2005). Graft‐versus‐ tumor effect against advanced pancreatic cancer after allogeneic reduced‐intensity stem cell transplantation. Transplantation 79, 821–827. Karre, K., Ljunggren, H. G., Piontek, G., and Kiessling, R. (1986). Selective rejection of H‐2‐deficient lymphoma variants suggests alternative immune defence strategy. Nature 319, 675–678. Knechtli, C. J., Goulden, N. J., Hancock, J. P., Grandage, V. L., Harris, E. L., Garland, R. J., Jones, C. G., Rowbottom, A. W., Hunt, L. P., Green, A. F., Clarke, E., Lankester, A. W., et al. (1998a). Minimal residual disease status before allogeneic bone marrow transplantation is an important determinant of successful outcome for children and adolescents with acute lymphoblastic leukemia. Blood 92, 4072–4079. Knechtli, C. J., Goulden, N. J., Hancock, J. P., Harris, E. L., Garland, R. J., Jones, C. G., Grandage, V. L., Rowbottom, A. W., Green, A. F., Clarke, E., Lankester, A. W., Potter, M. N., et al. (1998b). Minimal residual disease status as a predictor of relapse after allogeneic bone marrow transplantation for children with acute lymphoblastic leukaemia. Br. J. Haematol. 102, 860–871. Kobayashi, S., Imamura, M., Hashino, S., Noto, S., Mori, A., Tanaka, J., Naohara, T., Kasai, M., and Asaka, M. (1999). Possible role of granulocyte colony‐stimulating factor in increased serum soluble interleukin‐2 receptor‐alpha levels after allogeneic bone marrow transplantation. Leuk. Lymphoma 33, 559–566. Kobbe, G., Schneider, P., Rohr, U., Fenk, R., Neumann, F., Aivado, M., Dietze, L., Kronenwett, R., Hunerliturkoglu, A., and Haas, R. (2001). Treatment of severe steroid refractory acute graft‐ versus‐host disease with infliximab, a chimeric human/mouse antiTNFalpha antibody. Bone Marrow Transplant. 28, 47–49. Koc, O. N., Gerson, S. L., Cooper, B. W., Dyhouse, S. M., Haynesworth, S. E., Caplan, A. I., and Lazarus, H. M. (2000). Rapid hematopoietic recovery after coinfusion of autologous‐ blood stem cells and culture‐expanded marrow mesenchymal stem cells in advanced breast cancer patients receiving high‐dose chemotherapy. J. Clin. Oncol. 18, 307–316. Kolb, H. J., Mittermuller, J., Clemm, C., Holler, E., Ledderose, G., Brehm, G., Heim, M., and Wilmanns, W. (1990). Donor leukocyte transfusions for treatment of recurrent chronic myelogenous leukemia in marrow transplant patients. Blood 76, 2462–2465. Kolb, H. J., Schattenberg, A., Goldman, J. M., Hertenstein, B., Jacobsen, N., Arcese, W., Ljungman, P., Ferrant, A., Verdonck, L., Niederwieser, D., van Rhee, F., Mittermueller, J., et al. (1995). Graft‐versus‐leukemia effect of donor lymphocyte transfusions in marrow grafted patients. Blood 86, 2041–2050. Lawler, M., Humphries, P., and McCann, S. R. (1991). Evaluation of mixed chimerism by in vitro amplification of dinucleotide repeat sequences using the polymerase chain reaction. Blood 77, 2504–2514.

54

Olle Ringde´n

Lazarus, H. M., Koc, O. N., Devine, S. M., Curtin, P., Maziarz, R. T., Holland, H. K., Shpall, E. J., McCarthy, P., Atkinson, K., Cooper, B. W., Gerson, S. L., Laughlin, M. J., et al. (2005). Cotransplantation of HLA‐identical sibling culture‐expanded mesenchymal stem cells and hematopoietic stem cells in hematologic malignancy patients. Biol. Blood Marrow Transplant. 11, 389–398. Le Blanc, K., and Ringden, O. (2005). Immunobiology of human mesenchymal stem cells and future use in hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 11, 321–334. Le Blanc, K., Tammik, C., Rosendahl, K., Zetterberg, E., and Ringden, O. (2003a). HLA expression and immunologic properties of differentiated and undifferentiated mesenchymal stem cells. Exp. Hematol. 31, 890–896. Le Blanc, K., Tammik, L., Sundberg, B., Haynesworth, S. E., and Ringden, O. (2003b). Mesenchymal stem cells inhibit and stimulate mixed lymphocyte cultures and mitogenic responses independently of the major histocompatibility complex. Scand. J. Immunol. 57, 11–20. Le Blanc, K., Rasmusson, I., Gotherstrom, C., Seidel, C., Sundberg, B., Sundin, M., Rosendahl, K., Tammik, C., and Ringden, O. (2004a). Mesenchymal stem cells inhibit the expression of CD25 (interleukin‐2 receptor) and CD38 on phytohaemagglutinin‐activated lymphocytes. Scand. J. Immunol. 60, 307–315. Le Blanc, K., Rasmusson, I., Sundberg, B., Gotherstrom, C., Hassan, M., Uzunel, M., and Ringden, O. (2004b). Treatment of severe acute graft‐versus‐host disease with third party haploidentical mesenchymal stem cells. Lancet 363, 1439–1441. Lin, F., van Rhee, F., Goldman, J. M., and Cross, N. C. (1996). Kinetics of increasing BCR‐ABL transcript numbers in chronic myeloid leukemia patients who relapse after bone marrow transplantation. Blood 87, 4473–4478. Liu, J., Lu, X. F., Wan, L., Li, Y. P., Li, S. F., Zeng, L. Y., Zeng, Y. Z., Cheng, L. H., Lu, Y. R., and Cheng, J. Q. (2004). Suppression of human peripheral blood lymphocyte proliferation by immortalized mesenchymal stem cells derived from bone marrow of Banna Minipig inbred‐line. Transplant. Proc. 36, 3272–3275. Ljunggren, H. G., and Karre, K. (1990). In search of the “missing self”: MHC molecules and NK cell recognition. Immunol. Today. 11, 237–244. Maccario, R., Podesta, M., Moretta, A., Cometa, A., Comoli, P., Montagna, D., Daudt, L., Ibatici, A., Piaggio, G., Pozzi, S., Frassoni, F., and Locatelli, F. (2005). Interaction of human mesenchymal stem cells with cells involved in alloantigen‐specific immune response favors the differentiation of CD4þ T‐cell subsets expressing a regulatory/suppressive phenotype. Haematologica 90, 516–525. Mackinnon, S., Barnett, L., Bourhis, J. H., Black, P., Heller, G., and O’Reilly, R. J. (1992). Myeloid and lymphoid chimerism after T‐cell‐depleted bone marrow transplantation: Evaluation of conditioning regimens using the polymerase chain reaction to amplify human minisatellite regions of genomic DNA. Blood 80, 3235–3241. Mackinnon, S., Barnett, L., Heller, G., and O’Reilly, R. J. (1994). Minimal residual disease is more common in patients who have mixed T‐cell chimerism after bone marrow transplantation for chronic myelogenous leukemia. Blood 83, 3409–3416. Mackinnon, S., Papadopoulos, E. B., Carabasi, M. H., Reich, L., Collins, N. H., Boulad, F., Castro‐Malaspina, H., Childs, B. H., Gillio, A. P., Kernan, N. A., et al. (1995). Adoptive immunotherapy evaluating escalating doses of donor leukocytes for relapse of chronic myeloid leukemia after bone marrow transplantation: Separation of graft‐versus‐leukemia responses from graft‐versus‐host disease. Blood 86, 1261–1268. Maitra, B., Szekely, E., Gjini, K., Laughlin, M. J., Dennis, J., Haynesworth, S. E., and Koc, O. N. (2004). Human mesenchymal stem cells support unrelated donor hematopoietic stem cells and suppress T‐cell activation. Bone Marrow Transplant. 33, 597–604.

Immunotherapy by Allogeneic Stem Cell Transplantation

55

Marmont, A. M., Horowitz, M. M., Gale, R. P., Sobocinski, K., Ash, R. C., van Bekkum, D. W., Champlin, R. E., Dicke, K. A., Goldman, J. M., Good, R. A., et al. (1991). T‐cell depletion of HLA‐identical transplants in leukemia. Blood 78, 2120–2130. Mattsson, J., Uzunel, M., Remberger, M., Ljungman, P., Kimby, E., Ringden, O., and Zetterquist, H. (2000). Minimal residual disease is common after allogeneic stem cell transplantation in patients with B cell chronic lymphocytic leukemia and may be controlled by graft‐versus‐host disease. Leukemia 14, 247–254. Mattsson, J., Uzunel, M., Remberger, M., Tammik, L., Omazic, B., Levitsky, V., Zou, J. Z., Hentschke, P., and Ringden, O. (2001a). Poor immune reconstitution after four or five major HLA antigens mismatched T cell‐depleted allogeneic and autologous stem cell transplantation. Clin. Exp. Immunol. 123, 162–169. Mattsson, J., Uzunel, M., Tammik, L., Aschan, J., and Ringden, O. (2001b). Leukemia lineage‐ specific chimerism analysis is a sensitive predictor of relapse in patients with acute myeloid leukemia and myelodysplastic syndrome after allogeneic stem cell transplantation. Leukemia 15, 1976–1985. Mavroudis, D. A., Jiang, Y. Z., Hensel, N., Lewalle, P., Couriel, D., Kreitman, R. J., Pastan, I., and Barrett, A. J. (1996). Specific depletion of alloreactivity against haplotype mismatched related individuals by a recombinant immunotoxin: A new approach to graft‐versus‐ host disease prophylaxis in haploidentical bone marrow transplantation. Bone Marrow Transplant. 17, 793–799. McCarthy, P. L., Jr., Williams, L., Harris‐Bacile, M., Yen, J., Przepiorka, D., Ippoliti, C., Champlin, R., Fay, J., Blosch, C., Jacobs, C., and Anasetti, C. (1996). A clinical phase I/II study of recombinant human interleukin‐1 receptor in glucocorticoid‐resistant graft‐versus‐ host disease. Transplantation 62, 626–631. McIntosh, K., and Bartholomew, A. (2000). Stromal cell modulation of the immune system. Graft 3, 324–328. McSweeney, P. A., and Storb, R. (1999). Mixed chimerism: Preclinical studies and clinical applications. Biol. Blood Marrow Transplant. 5, 192–203. Miyamoto, T., Akashi, K., Hayashi, S., Gondo, H., Murakawa, M., Tanimoto, K., Harada, M., and Niho, Y. (1996). Serum concentration of the soluble interleukin‐2 receptor for monitoring acute graft‐versus‐host disease. Bone Marrow Transplant. 17, 185–190. Morecki, S., Moshel, Y., Gelfend, Y., Pugatsch, T., and Slavin, S. (1997). Induction of graft vs. tumor effect in a murine model of mammary adenocarcinoma. Int. J. Cancer 71, 59–63. Morgan, G. J., Hughes, T., Janssen, J. W., Gow, J., Guo, A. P., Goldman, J. M., Wiedemann, L. M., and Bartram, C. R. (1989). Polymerase chain reaction for detection of residual leukaemia. Lancet 1, 928–929. Moscovitch, M., and Slavin, S. (1984). Anti‐tumor effects of allogeneic bone marrow transplantation in (NZB  NZW)F1 hybrids with spontaneous lymphosarcoma. J. Immunol. 132, 997–1000. Mutis, T., Verdijk, R., Schrama, E., Esendam, B., Brand, A., and Goulmy, E. (1999). Feasibility of immunotherapy of relapsed leukemia with ex vivo‐generated cytotoxic T lymphocytes specific for hematopoietic system‐restricted minor histocompatibility antigens. Blood 93, 2336–2341. Nash, R. A., Pineiro, L. A., Storb, R., Deeg, H. J., Fitzsimmons, W. E., Furlong, T., Hansen, J. A., Gooley, T., Maher, R. M., Martin, P., McSweeney, P. A., Sullivan, K. M., et al. (1996). FK506 in combination with methotrexate for the prevention of graft‐versus‐host disease after marrow transplantation from matched unrelated donors. Blood 88, 3634–3641. Omuro, Y., Matsumoto, G., Sasaki, T., Tanaka, Y., Maeda, Y., Sakamaki, H., Hiruma, K., Tsuruta, K., and Takahashi, T. (2003). Regression of an unresectable pancreatic tumor following nonmyeloablative allogeneic peripheral‐blood stem‐cell transplantation. Bone Marrow Transplant. 31, 943–945.

56

Olle Ringde´n

Pedrazzoli, P., Da Prada, G. A., Giorgiani, G., Schiavo, R., Zambelli, A., Giraldi, E., Landonio, G., Locatelli, F., Siena, S., and Della Cuna, G. R. (2002). Allogeneic blood stem cell transplantation after a reduced‐intensity, preparative regimen: A pilot study in patients with refractory malignancies. Cancer 94, 2409–2415. Pittenger, M. F., Mackay, A. M., Beck, S. C., Jaiswal, R. K., Douglas, R., Mosca, J. D., Moorman, M. A., Simonetti, D. W., Craig, S., and Marshak, D. R. (1999). Multilineage potential of adult human mesenchymal stem cells. Science 284, 143–147. Potter, M. N., Cross, N. C., van Dongen, J. J., Saglio, G., Oakhill, A., Bartram, C. R., and Goldman, J. M. (1993). Molecular evidence of minimal residual disease after treatment for leukaemia and lymphoma: An updated meeting report and review. Leukemia 7, 1302–1314. Prentice, H. G., Blacklock, H. A., Janossy, G., Gilmore, M. J., Price‐Jones, L., Tidman, N., Trejdosiewicz, L. K., Skeggs, D. B., Panjwani, D., Ball, S., et al. (1984). Depletion of T lymphocytes in donor marrow prevents significant graft‐versus‐host disease in matched allogeneic leukaemic marrow transplant recipients. Lancet 1, 472–476. Preudhomme, C., Revillion, F., Merlat, A., Hornez, L., Roumier, C., Duflos‐Grardel, N., Jouet, J. P., Cosson, A., Peyrat, J. P., and Fenaux, P. (1999). Detection of BCR‐ABL transcripts in chronic myeloid leukemia (CML) using a “real time” quantitative RT‐PCR assay. Leukemia 13, 957–964. Rasmusson, I., Ringden, O., Sundberg, B., and Le Blanc, K. (2003). Mesenchymal stem cells inhibit the formation of cytotoxic T lymphocytes, but not activated cytotoxic T lymphocytes or natural killer cells. Transplantation 76, 1208–1213. Rasmusson, I., Ringden, O., Sundberg, B., and Le Blanc, K. (2005). Mesenchymal stem cells inhibit lymphocyte proliferation by mitogens and alloantigens by different mechanisms. Exp. Cell Res. 305, 33–41. Remberger, M., and Sundberg, B. (2005). Granulocyte colony‐stimulating factor affects serum levels of soluble interleukin‐2 receptors after allogeneic stem cell transplantation. Haematologica 90, 427–429. Remberger, M., Aschan, J., Barkholt, L., Tollemar, J., and Ringden, O. (2001). Treatment of severe acute graft‐versus‐host disease with anti‐thymocyte globulin. Clin. Transplant. 15, 147–153. Remberger, M., Naseh, N., Aschan, J., Barkholt, L., LeBlanc, K., Svennberg, P., and Ringden, O. (2003). G‐CSF given after haematopoietic stem cell transplantation using HLA‐identical sibling donors is associated to a higher incidence of acute GVHD II‐IV. Bone Marrow Transplant. 32, 217–223. Renkonen, R., Wangel, A., and Hayry, P. (1986). Bone marrow transplantation in the rat. B lymphocyte activation in acute graft‐versus‐host disease. Transplantation 41, 290–296. Riddell, S. R., Watanabe, K. S., Goodrich, J. M., Li, C. R., Agha, M. E., and Greenberg, P. D. (1992). Restoration of viral immunity in immunodeficient humans by the adoptive transfer of T cell clones. Science 257, 238–241. Ringden, O., and Deeg, H. J. (1996). Clinical spectrum of graft‐versus‐host disease. In “Graft Vs Host Disease” (J. L. M. Ferrara, H. J. Deeg, and S. Burakoff, Eds.), pp. 525–559. Marcel Dekker, Inc., New York. Ringden, O., and Horowitz, M. M. (1989). Graft‐versus‐leukemia reactions in humans. The Advisory Committee of the International Bone Marrow Transplant Registry. Transplant. Proc. 21, 2989–2992. Ringden, O., Backman, L., Lonnqvist, B., Heimdahl, A., Lindholm, A., Bolme, P., and Gahrton, G. (1986). A randomized trial comparing use of cyclosporin and methotrexate for graft‐versus‐host disease prophylaxis in bone marrow transplant recipients with haematological malignancies. Bone Marrow Transplant. 1, 41–51. Ringden, O., Zwaan, F., Hermans, J., and Gratwohl, A. (1987). European experience of bone marrow transplantation for leukemia. Transplant. Proc. 19, 2600–2604. Ringden, O., Pihlstedt, P., Markling, L., Aschan, J., Baryd, I., Ljungman, P., Lonnqvist, B., Tollemar, J., Janossy, G., and Sundberg, B. (1991). Prevention of graft‐versus‐host disease

Immunotherapy by Allogeneic Stem Cell Transplantation

57

with T cell depletion or cyclosporin and methotrexate. A randomized trial in adult leukemic marrow recipients. Bone Marrow Transplant. 7, 221–226. Ringden, O., Horowitz, M. M., Gale, R. P., Biggs, J. C., Gajewski, J., Rimm, A. A., Speck, B., Veum‐Stone, J. A., de Witte, T., and Bortin, M. M. (1993a). Outcome after allogeneic bone marrow transplant for leukemia in older adults. JAMA 270, 57–60. Ringden, O., Horowitz, M. M., Sondel, P., Gale, R. P., Biggs, J. C., Champlin, R. E., Deeg, H. J., Dicke, K., Masaoka, T., Powles, R. L., et al. (1993b). Methotrexate, cyclosporine, or both to prevent graft‐versus‐host disease after HLA‐identical sibling bone marrow transplants for early leukemia? Blood 81, 1094–1101. Ringden, O., Hermans, J., Labopin, M., Apperley, J., Gorin, N. C., and Gratwohl, A. (1996a). The highest leukaemia‐free survival after allogeneic bone marrow transplantation is seen in patients with grade I acute graft‐versus‐host disease. Acute and Chronic Leukaemia Working Parties of the European Group for Blood and Marrow Transplantation (EBMT). Leuk. Lymphoma 24, 71–79. Ringden, O., Labopin, M., Gluckman, E., Reiffers, J., Vernant, J. P., Jouet, J. P., Harrousseau, J. L., Fiere, D., Bacigalupo, A., Frassoni, F., and Gorin, N. C. (1996b). Graft‐versus‐leukemia effect in allogeneic marrow transplant recipients with acute leukemia is maintained using cyclosporin A combined with methotrexate as prophylaxis. Acute Leukemia Working Party of the European Group for Blood and Marrow Transplantation. Bone Marrow Transplant. 18, 921–929. Ringden, O., Labopin, M., Gorin, N. C., Schmitz, N., Schaefer, U. W., Prentice, H. G., Bergmann, L., Jouet, J. P., Mandelli, F., Blaise, D., Fouillard, L., and Frassoni, F. (2000a). Is there a graft‐versus‐leukaemia effect in the absence of graft‐versus‐host disease in patients undergoing bone marrow transplantation for acute leukaemia? Br. J. Haematol. 111, 1130–1137. Ringden, O., Soderdahl, G., Mattsson, J., Uzunel, M., Remberger, M., Hentschke, P., Hagglund, H., Sparrelid, E., Elmhorn‐Rosenborg, A., Duraj, F., Zetterquist, H., and Ericzon, B. G. (2000b). Transplantation of autologous and allogeneic bone marrow with liver from a cadaveric donor for primary liver cancer. Transplantation 69, 2043–2048. Ringden, O., Labopin, M., Bacigalupo, A., Arcese, W., Schaefer, U. W., Willemze, R., Koc, H., Bunjes, D., Gluckman, E., Rocha, V., Schattenberg, A., and Frassoni, F. (2002). Transplantation of peripheral blood stem cells as compared with bone marrow from HLA‐identical siblings in adult patients with acute myeloid leukemia and acute lymphoblastic leukemia. J. Clin. Oncol. 20, 4655–4664. Ringden, O., Barrett, A. J., Zhang, M. J., Loberiza, F. R., Bolwell, B. J., Cairo, M. S., Gale, R. P., Hale, G. A., Litzow, M. R., Martino, R., Russell, J. A., Tiberghien, P., et al. (2003). Decreased treatment failure in recipients of HLA‐identical bone marrow or peripheral blood stem cell transplants with high CD34 cell doses. Br. J. Haematol. 121, 874–885. Ringden, O., Labopin, M., Gorin, N. C., Le Blanc, K., Rocha, V., Gluckman, E., Reiffers, J., Arcese, W., Vossen, J. M., Jouet, J. P., Cordonnier, C., and Frassoni, F. (2004a). Treatment with granulocyte colony‐stimulating factor after allogeneic bone marrow transplantation for acute leukemia increases the risk of graft‐versus‐host disease and death: A study from the Acute Leukemia Working Party of the European Group for Blood and Marrow Transplantation. J. Clin. Oncol. 22, 416–423. Ringden, O., Schaffer, M., Le Blanc, K., Persson, U., Hauzenberger, D., Abedi, M. R., Olerup, O., Ljungman, P., and Remberger, M. (2004b). Which donor should be chosen for hematopoietic stem cell transplantation among unrelated HLA‐A, ‐B, and ‐DRB1 genomically identical volunteers? Biol. Blood Marrow Transplant. 10, 128–134. Ringden, O., Soderdahl, G., Barkholt, L., and Ericzon, B. G. (2004c). Solid oxygen and marrow transplants: A combined strategy for tolerance and anti‐tumor activity. In “New Perspectives in Allogeneic Transplantation of Hematopoietic Cells” (A. Bacigalupo and P. Corradini, Eds.), pp. 129–134. GITMO, Universita Milano, Instituto Nazionale per lo Studio e la Cura dei Tumori, Milano.

58

Olle Ringde´n

Ringden, O., Uzunel, M., Rasmusson, I., Remberger, M., Sundberg, B., Lonnies, H., Marschall, H. U., Dlugosz, A., Szakos, A., Hassan, Z., Omazic, B., Sundin, M., et al. (2006). Mesenchymal stem cells for treatment of therapy‐resistant graft‐versus‐host disease. Transplantation 81, 1390–1397. Ringden, O. T., Le Blanc, K., and Remberger, M. (2005). Granulocyte and granulocyte‐ macrophage colony‐stimulating factors in allografts: Uses, misuses, misconceptions, and future applications. Exp. Hematol. 33, 505–512. Rini, B. I., Zimmerman, T. M., Gajewski, T. F., Stadler, W. M., and Vogelzang, N. J. (2001). Allogeneic peripheral blood stem cell transplantation for metastatic renal cell carcinoma. J. Urol. 165, 1208–1209. Rooney, C. M., Smith, C. A., Ng, C. Y., Loftin, S., Li, C., Krance, R. A., Brenner, M. K., and Heslop, H. E. (1995). Use of gene‐modified virus‐specific T lymphocytes to control Epstein‐ Barr‐virus‐related lymphoproliferation. Lancet 345, 9–13. Rosenberg, S. A., Lotze, M. T., Yang, J. C., Topalian, S. L., Chang, A. E., Schwartzentruber, D. J., Aebersold, P., Leitman, S., Linehan, W. M., Seipp, C. A., et al. (1993). Prospective randomized trial of high‐dose interleukin‐2 alone or in conjunction with lymphokine‐ activated killer cells for the treatment of patients with advanced cancer. J. Natl. Cancer Inst. 85, 622–632. Ruggeri, L., Capanni, M., Urbani, E., Perruccio, K., Shlomchik, W. D., Tosti, A., Posati, S., Rogaia, D., Frassoni, F., Aversa, F., Martelli, M. F., and Velardi, A. (2002). Effectiveness of donor natural killer cell alloreactivity in mismatched hematopoietic transplants. Science 295, 2097–2100. Ryan, D. H., and van Dongen, J. J. (1988). Detection of residual disease in acute leukemia using immunological markers. Cancer Treat. Res. 38, 173–207. Schaffer, M., Malmberg, K. J., Ringden, O., Ljunggren, H. G., and Remberger, M. (2004). Increased infection‐related mortality in KIR‐ligand‐mismatched unrelated allogeneic hematopoietic stem‐cell transplantation. Transplantation 78, 1081–1085. Schattenberg, A., De Witte, T., Salden, M., Vet, J., Van Dijk, B., Smeets, D., Hoogenhout, J., and Haanen, C. (1989). Mixed hematopoietic chimerism after allogeneic transplantation with lymphocyte‐depleted bone marrow is not associated with a higher incidence of relapse. Blood 73, 1367–1372. Schmitz, N., Beksac, M., Hasenclever, D., Bacigalupo, A., Ruutu, T., Nagler, A., Gluckman, E., Russell, N., Apperley, J. F., Gorin, N. C., Szer, J., Bradstock, K., et al. (2002). Transplantation of mobilized peripheral blood cells to HLA‐identical siblings with standard‐risk leukemia. Blood 100, 761–767. Shulman, H. M., Sale, G. E., Lerner, K. G., Barker, E. A., Weiden, P. L., Sullivan, K., Gallucci, B., Thomas, E. D., and Storb, R. (1978). Chronic cutaneous graft‐versus‐host disease in man. Am. J. Pathol. 91, 545–570. Slavin, S., Nagler, A., Naparstek, E., Kapelushnik, Y., Aker, M., Cividalli, G., Varadi, G., Kirschbaum, M., Ackerstein, A., Samuel, S., Amar, A., Brautbar, C., et al. (1998). Nonmyeloablative stem cell transplantation and cell therapy as an alternative to conventional bone marrow transplantation with lethal cytoreduction for the treatment of malignant and nonmalignant hematologic diseases. Blood 91, 756–763. Socie, G., Henry‐Amar, M., Bacigalupo, A., Hows, J., Tichelli, A., Ljungman, P., McCann, S. R., Frickhofen, N., Van’t Veer‐Korthof, E., and Gluckman, E. (1993). Malignant tumors occurring after treatment of aplastic anemia. European Bone Marrow Transplantation‐ Severe Aplastic Anaemia Working Party. N. Engl. J. Med. 329, 1152–1157. Socie, G., Lawler, M., Gluckman, E., McCann, S. R., and Brison, O. (1995). Studies on hemopoietic chimerism following allogeneic bone marrow transplantation in the molecular biology era. Leuk. Res. 19, 497–504.

Immunotherapy by Allogeneic Stem Cell Transplantation

59

Soderdahl, G., Barkholt, L., Hentschke, P., Mattsson, J., Uzunel, M., Ericzon, B. G., and Ringden, O. (2003). Liver transplantation followed by adjuvant nonmyeloablative hemopoietic stem cell transplantation for advanced primary liver cancer in humans. Transplantation 75, 1061–1066. Solomon, S. R., Mielke, S., Savani, B. N., Montero, A., Wisch, L., Childs, R., Hensel, N., Schindler, J., Ghetie, V., Leitman, S. F., Mai, T., Carter, C. S., et al. (2005). Selective depletion of alloreactive donor lymphocytes: A novel method to reduce the severity of graft‐versus‐host disease in older patients undergoing matched sibling donor stem cell transplantation. Blood 106, 1123–1129. Storb, R., and Thomas, E. D. (1985). Graft‐versus‐host disease in dog and man: The Seattle experience. Immunol. Rev. 88, 215–238. Storb, R., Prentice, R. L., Hansen, J. A., and Thomas, E. D. (1983a). Association between HLA‐B antigens and acute graft‐versus‐host disease. Lancet 2, 816–819. Storb, R., Prentice, R. L., Sullivan, K. M., Shulman, H. M., Deeg, H. J., Doney, K. C., Buckner, C. D., Clift, R. A., Witherspoon, R. P., Appelbaum, F. A., Sanders, J. E., Stewart, P. S., et al. (1983b). Predictive factors in chronic graft‐versus‐host disease in patients with aplastic anemia treated by marrow transplantation from HLA‐identical siblings. Ann. Intern. Med. 98, 461–466. Storb, R., Deeg, H. J., Pepe, M., Appelbaum, F., Anasetti, C., Beatty, P., Bensinger, W., Berenson, R., Buckner, C. D., Clift, R., et al. (1989). Methotrexate and cyclosporine versus cyclosporine alone for prophylaxis of graft‐versus‐host disease in patients given HLA‐ identical marrow grafts for leukemia: Long‐term follow‐up of a controlled trial. Blood 73, 1729–1734. Sullivan, K. M., Shulman, H. M., Storb, R., Weiden, P. L., Witherspoon, R. P., McDonald, G. B., Schubert, M. M., Atkinson, K., and Thomas, E. D. (1981). Chronic graft‐versus‐host disease in 52 patients: Adverse natural course and successful treatment with combination immunosuppression. Blood 57, 267–276. Sullivan, K. M., Weiden, P. L., Storb, R., Witherspoon, R. P., Fefer, A., Fisher, L., Buckner, C. D., Anasetti, C., Appelbaum, F. R., Badger, C., et al. (1989). Influence of acute and chronic graft‐versus‐host disease on relapse and survival after bone marrow transplantation from HLA‐identical siblings as treatment of acute and chronic leukemia. Blood 73, 1720–1728. Svahn, B. M., Remberger, M., Myrback, K. E., Holmberg, K., Eriksson, B., Hentschke, P., Aschan, J., Barkholt, L., and Ringden, O. (2002). Home care during the pancytopenic phase after allogeneic hematopoietic stem cell transplantation is advantageous compared with hospital care. Blood 100, 4317–4324. Svahn, B. M., Ringden, O., and Remberger, M. (2005). Long‐term follow‐up of patients treated at home during the pancytopenic phase after allogeneic haematopoietic stem cell transplantation. Bone Marrow Transplant. 36, 511–516. Takahashi, T., Omuro, Y., Matsumoto, G., Sakamaki, H., Maeda, Y., Hiruma, K., Tsuruta, K., and Sasaki, T. (2004). Nonmyeloablative allogeneic stem cell transplantation for patients with unresectable pancreatic cancer. Pancreas 28, e65–e69. Thomas, E. D., Buckner, C. D., Banaji, M., Clift, R. A., Fefer, A., Flournoy, N., Goodell, B. W., Hickman, R. O., Lerner, K. G., Neiman, P. E., Sale, G. E., Sanders, J. E., et al. (1977). One hundred patients with acute leukemia treated by chemotherapy, total body irradiation, and allogeneic marrow transplantation. Blood 49, 511–533. Tiberghien, P., Reynolds, C. W., Keller, J., Spence, S., Deschaseaux, M., Certoux, J. M., Contassot, E., Murphy, W. J., Lyons, R., Chiang, Y., et al. (1994). Ganciclovir treatment of herpes simplex thymidine kinase‐transduced primary T lymphocytes: An approach for specific in vivo donor T‐cell depletion after bone marrow transplantation? Blood 84, 1333–1341.

60

Olle Ringde´n

Truitt, R. L., Johnson, B. D., McCabe, C. M., and Weiler, M. B. (1996). In “Graft Vs Host Disease” (J. L. M. Ferrara, H. J. Deeg, and S. Burakoff, Eds.), 2nd edn., pp. 385–423. Marcel Dekker, Inc. New York. Tse, W. T., Pendleton, J. D., Beyer, W. M., Egalka, M. C., and Guinan, E. C. (2003). Suppression of allogeneic T‐cell proliferation by human marrow stromal cells: Implications in transplantation. Transplantation 75, 389–397. Ueno, N. T., Rondon, G., Mirza, N. Q., Geisler, D. K., Anderlini, P., Giralt, S. A., Andersson, B. S., Claxton, D. F., Gajewski, J. L., Khouri, I. F., Korbling, M., Mehra, R. C., et al. (1998). Allogeneic peripheral‐blood progenitor‐cell transplantation for poor‐risk patients with metastatic breast cancer. J. Clin. Oncol. 16, 986–993. Ueno, N. T., Cheng, Y. C., Rondon, G., Tannir, N. M., Gajewski, J. L., Couriel, D. R., Hosing, C., de Lima, M. J., Anderlini, P., Khouri, I. F., Booser, D. J., Hortobagyi, G. N., et al. (2003). Rapid induction of complete donor chimerism by the use of a reduced‐intensity conditioning regimen composed of fludarabine and melphalan in allogeneic stem cell transplantation for metastatic solid tumors. Blood 102, 3829–3836. Uzunel, M., Mattsson, J., Jaksch, M., Remberger, M., and Ringden, O. (2001). The significance of graft‐versus‐host disease and pretransplantation minimal residual disease status to outcome after allogeneic stem cell transplantation in patients with acute lymphoblastic leukemia. Blood 98, 1982–1984. van Bekkum, D. W. (1985). “Graft‐Versus‐Host Disease.” Marcel Dekker, Inc., New York. van Leeuwen, J. E., van Tol, M. J., Joosten, A. M., Wijnen, J. T., Verweij, P. J., Khan, P. M., and Vossen, J. M. (1994). Persistence of host‐type hematopoiesis after allogeneic bone marrow transplantation for leukemia is significantly related to the recipient’s age and/or the conditioning regimen, but it is not associated with an increased risk of relapse. Blood 83, 3059–3067. Wagner, J. E., Rosenthal, J., Sweetman, R., Shu, X. O., Davies, S. M., Ramsay, N. K., McGlave, P. B., Sender, L., and Cairo, M. S. (1996). Successful transplantation of HLA‐matched and HLA‐mismatched umbilical cord blood from unrelated donors: Analysis of engraftment and acute graft‐versus‐host disease. Blood 88, 795–802. Warren, E. H., Gavin, M., Greenberg, P. D., and Riddell, S. R. (1998a). Minor histocompatibility antigens as targets for T‐cell therapy after bone marrow transplantation. Curr. Opin. Hematol. 5, 429–433. Warren, E. H., Greenberg, P. D., and Riddell, S. R. (1998b). Cytotoxic T‐lymphocyte‐defined human minor histocompatibility antigens with a restricted tissue distribution. Blood 91, 2197–2207. Weiden, P. L., Flournoy, N., Thomas, E. D., Prentice, R., Fefer, A., Buckner, C. D., and Storb, R. (1979). Antileukemic effect of graft‐versus‐host disease in human recipients of allogeneic‐ marrow grafts. N. Engl. J. Med. 300, 1068–1073. Weiden, P. L., Sullivan, K. M., Flournoy, N., Storb, R., and Thomas, E. D. (1981). Antileukemic effect of chronic graft‐versus‐host disease: Contribution to improved survival after allogeneic marrow transplantation. N. Engl. J. Med. 304, 1529–1533. Witherspoon, R. P., Fisher, L. D., Schoch, G., Martin, P., Sullivan, K. M., Sanders, J., Deeg, H. J., Doney, K., Thomas, D., Storb, R., et al. (1989). Secondary cancers after bone marrow transplantation for leukemia or aplastic anemia. N. Engl. J. Med. 321, 784–789. Zetterquist, H., Mattsson, J., Uzunel, M., Nasman‐Bjork, I., Svenberg, P., Tammik, L., Bayat, G., Winiarski, J., and Ringden, O. (2000). Mixed chimerism in the B cell lineage is a rapid and sensitive indicator of minimal residual disease in bone marrow transplant recipients with pre‐B cell acute lymphoblastic leukemia. Bone Marrow Transplant. 25, 843–851. Zetterquist, H., Hentschke, P., Thorne, A., Wernerson, A., Mattsson, J., Uzunel, M., Martola, J., Albiin, N., Aschan, J., Papadogiannakis, N., and Ringden, O. (2001). A graft‐versus‐colonic cancer effect of allogeneic stem cell transplantation. Bone Marrow Transplant. 28, 1161–1166.

Mnt Takes Control as Key Regulator of the Myc/Max/Mxd Network Therese Wahlstro¨m and Marie Henriksson Department of Microbiology, Tumor, and Cell Biology (MTC), Karolinska Institutet, SE‐171 77 Stockholm, Sweden

I. Myc: The Most Frequently Deregulated Oncogene in Human Tumors A. Myc in Control of Cell Fate B. Deregulation of Myc in Tumor Development C. Myc as a Therapeutic Target for Human Cancer II. Mnt: The Key Transcriptional Regulator of the Myc/Max/Mxd Network A. Discovery and Characterization of Mnt B. Mnt, the Major Myc Antagonist C. Effects of Mnt Deficiency D. Relief of Mnt‐Mediated Repression: The Critical Event for Myc Target Gene Activation E. Mnt, to Be or Not to Be a Tumor Suppressor III. Concluding Remarks References

Myc is the most frequently deregulated oncogene in human tumors. The protein belongs to the Myc/Max/Mxd network of transcriptional regulators important for cell growth, proliferation, differentiation, and apoptosis. The ratio between Mnt/ Max and c‐Myc/Max on the 50 ‐CACGTG‐30 E‐box sequence at shared target genes is of great importance for cell cycle progression and arrest. Serum stimulation of quiescent cells results in phosphorylation of Mnt and disruption of the critical Mnt‐mSin3‐HDAC1 interaction. This in turn leads to increased expression of the Myc/Mnt target gene cyclin D2. It is therefore possible that Myc function relies on its ability to overcome transcriptional repression by Mnt and that relief of Mnt‐mediated transcriptional repression is of greater importance for regulation of target genes than the sole activation by Myc. In addition, Mnt has many features of a tumor suppressor and may thus be nonfunctional or inactivated in human tumors. In summary, accumulating evidence supports the model of Mnt as the key regulator of the network in vivo. # 2007 Elsevier Inc.

Advances in CANCER RESEARCH Copyright 2007, Elsevier Inc. All rights reserved.

61

0065-230X/07 $35.00 DOI: 10.1016/S0065-230X(06)97003-1

62

Therese Wahlstro¨m and Marie Henriksson

I. Myc: THE MOST FREQUENTLY DEREGULATED ONCOGENE IN HUMAN TUMORS A. Myc in Control of Cell Fate The Myc/Max/Mxd network of transcription factors plays a key role in the regulation of cell growth, proliferation, apoptosis, and differentiation (Grandori et al., 2000; Henriksson and Lu¨scher, 1996). Ectopic expression of c‐Myc triggers quiescent cells to enter the cell cycle and once proliferating, cells express c‐Myc during all stages of the cell cycle (Eilers et al., 1991). Conversely, Myc transcription is turned off in quiescent cells, but on mitogenic stimulation, c‐Myc mRNA and protein are rapidly up‐regulated. The high c‐Myc protein level is sustained for a short period of time and subsequently decline to a low, but detectable levels once cells are proliferating (Henriksson and Lu¨scher, 1996; Oster et al., 2002). Both the up‐regulation and subsequent down‐regulation of c‐Myc are critical in the normal transition from quiescent to proliferating cells. c‐Myc is a transcription factor containing a basic‐region‐helix‐loop‐helix‐ leucine‐zipper (bHLHZip) and a transactivation domain (Oster et al., 2002). Other members of the Myc family include MYCN and L‐Myc. Myc forms a complex with Max through its bHLHZip domain, and these heterodimers bind specifically to 50 ‐CACGTG‐30 E‐box sequences to activate transcription (Fig. 1). The idea that Myc exerts its functions through gene regulation is supported by the findings that Myc interacts with proteins essential for transcriptional regulation, for example TRRAP and histone acetyltransferases (HATs), via a conserved region called Myc BoxII (MBII) (Fig. 1) (Oster et al., 2002). The Myc/Max/Mxd network also includes the Mxd (formerly Mad) family, Mnt, and Mga (Hooker and Hurlin, 2006). In similarity with Myc, all these proteins form heterodimers with Max and bind to the consensus E‐box element. Transcriptional repression by Mxd, Mnt, or Mga is mediated through interaction with mSin3, which in turn results in recruitment of histone decaetylases (HDACs) and corepressor molecules. These complexes cause condensation of the chromatin structure and thereby silencing of the expression of specific genes (Fig. 1) (Grandori et al., 2000; Hooker and Hurlin, 2006). Concomitant with transcriptional repression, there is a switch from c‐Myc/Max to Mxd1/Max at the E‐boxes of the hTERT and cyclin D2 genes when proliferating HL60 cells are induced to differentiate (Bouchard et al., 2001; Xu et al., 2001). As in HL60 cells, Myc expression is normally down‐regulated when cells are induced to differentiate, whereas enforced Myc expression inhibits differentiation (Larsson et al., 1994; Oster et al., 2002). The ability of Myc to block differentiation can

Mnt, the Key Regulator of Myc Function

63

Fig. 1 Model for transcriptional regulation of E‐box containing genes by the Myc/Max/Mxd network. The Myc, Mxd, and Mnt transcription factors form heterodimers with Max that bind specifically to the 50 ‐CACGTG‐30 E‐box sequence in target genes. In quiescent and differentiated cells, the Mxd proteins cause transcriptional repression mediated through interaction with mSin3 via the Sin3‐interacting domain (SID). This binding results in chromatin remodeling due to recruitment of histone decaetylases (HDACs) and corepressor molecules to the promoter. During cell cycle entry and in proliferating cells, transcription is activated in response to binding of Myc/Max to the E‐box. Myc interacts with proteins essential for transcriptional regulation, for example TRRAP and histone acetyltransferases (HATs), through the conserved Myc Box II (MBII) domain, resulting in chromatin modification. The Mnt protein is expressed both during quiescence and in proliferating cells and represses transcription, like the Mxd proteins, via interaction with mSin3 proteins. Ac, acetyl group.

be uncoupled from its proliferative function, making Myc an important mediator in the process between these events (Pelengaris et al., 2002a). In addition, c‐Myc has been shown to have a novel role during the first steps of hematopoietic stem cell (HSC) differentiation within bone marrow niches. c‐Myc represses the expression of specific integrins and the HSC‐anchoring protein N‐cadherin, suggesting a model in which c‐Myc controls the balance between self‐renewal and differentiation by modulating migration and/or adhesion of HSCs to the niche (Wilson et al., 2004).

64

Therese Wahlstro¨m and Marie Henriksson

The central player in the network, Max, is crucial for development since Max deficiency results in a severely altered phenotype and early embryonic lethality in mice (Shen‐Li et al., 2000). Similarly, genetic deletion of c‐Myc or N‐Myc results in cell cycle arrest, impaired development, and embryonic lethality (Charron et al., 1992; Davis et al., 1993; de Alboran et al., 2001; Giroux and Charron, 1998; Knoepfler et al., 2002; Moens et al., 1992; Sawai et al., 1991; Stanton et al., 1990; Trumpp et al., 2001). Mouse embryo fibroblasts (MEFs) lacking c‐Myc have been shown to arrest in G0 and to be unable to enter the cell cycle and proliferate (de Alboran et al., 2001; Trumpp et al., 2001). However, immortalized Rat1 fibroblasts with both c‐Myc alleles homozygously deleted proliferate, albeit at very slow rates (Mateyak et al., 1997). Myc is also required for cell growth in vivo since mice with c‐Myc deficiency, as well as dMyc null fly embryos (Pierce et al., 2004), die during development. Moreover, both flies (Johnston et al., 1999; Maines et al., 2004; Moberg et al., 2004; Pierce et al., 2004) and mice (de Alboran et al., 2001; Trumpp et al., 2001) carrying Myc mutations are small in size. Conversely, overexpression of dMyc in the eye of the fly resulted in increased size by accelerating cell growth, and when overexpressed throughout the entire animal, the size of the fly increased by nearly 30% (de la Cova et al., 2004; Johnston et al., 1999). The most likely mechanism for growth regulation by Myc is through transcriptional control of key regulators of ribosome biogenesis (Arabi et al., 2005; Grandori et al., 2005; Grewal et al., 2005). Cells respond to high c‐Myc levels by undergoing apoptosis, in particular when growth and survival factors become limiting (Askew et al., 1991; Evan et al., 1992). Myc sensitizes cells to apoptosis induced by different cellular insults such as ligation of the Fas death receptor and cytotoxic drugs (Oster et al., 2002). The ability of Myc to enhance the effect of many mechanistically distinct inducers suggests that Myc acts in a common control and/or execution pathway of apoptosis. The paradox that Myc expression can induce apoptosis during serum deprivation is probably attributed to simultaneous activation of both proliferation and apoptosis as normal physiological functions of Myc (Hueber and Evan, 1998). Two signals are needed for cell survival, one to stimulate cell proliferation and one to inhibit apoptosis. The availability of apoptosis inhibitors determines whether the cell divides or dies.

B. Deregulation of Myc in Tumor Development Inactivation of the strict regulation of Myc expression results in uncontrolled cell division and an increased risk for secondary mutations in the accumulating pool of proliferating cells, which may contribute to tumor

Mnt, the Key Regulator of Myc Function

65

development (Grandori et al., 2000). Mutations in the apoptotic program result in an imbalance between proliferation and cell death with a predominance for the former, thereby facilitating Myc‐driven tumorigenesis. Myc is activated in many different tumors such as small cell lung cancer, breast carcinoma, osteosarcoma, glioblastoma, cervix carcinoma, myeloid leukemia, plasma cell leukemia, Burkitt’s lymphoma, and neuroblastoma (Vita and Henriksson, 2006). The frequent activation of Myc in a broad spectrum of human tumors emphasizes the importance to study regulation of Myc, as well as the role of Mnt, the potential key regulator of the Myc/Max/Mxd network.

C. Myc as a Therapeutic Target for Human Cancer There are several reasons for considering Myc as an interesting target for therapy. First, the gene is often overexpressed or activated in human tumors. Second, Myc is expressed only in cycling cells and a targeted therapy would therefore mainly affect the tumor cells. However, proliferating cells found in the gastrointestinal tract and the hematopoietic system may also be at risk. Third, Myc activates both cell proliferation and apoptosis and thus strategies to target Myc could either be to block its proliferative function and/or to activate its pro‐apoptotic effect. The fourth reason is that Myc‐ induced tumorigenesis has been shown to be reversible in experimental models. Constitutive Myc expression in hematopoietic cells has been demonstrated to cause leukemias and lymphomas. Importantly, inactivation of the Myc gene resulted in regression of the tumors and restoration of normal hematopoiesis (Felsher and Bishop, 1999). It has also been reported that Myc‐induced tumor formation is reversible in pancreatic beta cells (Pelengaris et al., 2002b). Taken together, pharmacological inactivation of Myc on the gene or protein level could therefore be an efficient therapy for treating malignancies with Myc overexpression. Since Myc is activated in several cancer forms, other tumor types could also be affected by a Myc‐targeted therapeutic intervention (Henriksson et al., 2001; Vita and Henriksson, 2006). One drawback in targeting Myc for cancer therapy would be that cessation of drug administration could result in reactivation of the oncoprotein and tumor regrowth. This issue was addressed in a very elegant study using a conditional transgenic mouse model for Myc‐induced tumorigenesis where even brief inactivation of Myc resulted in regression of the tumor and differentiation of the cancer cells (Jain et al., 2002). More importantly, subsequent reactivation of Myc did not restore the malignant properties of the cells, but instead induced apoptosis. Thus, brief inactivation of Myc appears to cause changes in tumor cells that render them insensitive to Myc‐induced tumorigenesis. These results raise the possibility that transient inactivation

66

Therese Wahlstro¨m and Marie Henriksson

of Myc may be an effective therapy for certain cancers. However, it has yet to be proven that inactivation of Myc can induce regression of the more genomic complex human tumors.

II. Mnt: THE KEY TRANSCRIPTIONAL REGULATOR OF THE Myc/Max/Mxd NETWORK A. Discovery and Characterization of Mnt Mnt was identified in 1997 as a member of the Myc/Max/Mxd network. Hurlin et al. (1997a) described the murine Mnt in a yeast two‐hybrid screen using cDNA libraries from day 9.5 and 10.5 mouse embryos while Meroni et al. (1997) identified the human homologue of Mnt (originally termed Rox) in a search for transcribed sequences from the human chromosome 17p13.3 locus. The protein is localized in the cell nucleus, and it migrates as a doublet of 72 and 74 kDa in SDS‐PAGE (Hurlin et al., 1997a; Meroni et al., 1997; Popov et al., 2005). Phosphorylation of Mnt converts the faster migrating 72 kDa form into the 74‐kDa protein (Hurlin et al., 1997a; Popov et al., 2005). In similarity with the other Myc/Max/Mxd network proteins, Mnt forms heterodimers with Max to bind CACGTG E‐box sites. However, Mnt/Max dimers have been shown to have higher affinity to CACGCG sites compared to CACGTG E‐box sequences (Hurlin et al., 1997a; Meroni et al., 1997). In addition, Mnt is the most abundant Max‐binding partner in several human tumor cell lines (Popov et al., 2005; Smith et al., 2004; Sommer et al., 1998, 1999). The Mnt protein is expressed at a constant level during quiescence, throughout cell cycle progression and during differentiation (Hurlin et al., 1997a, 2003, 2004; Meroni et al., 1997; Popov et al., 2005). Despite its ubiquitous expression pattern, Mnt has been classified as a member of the Mxd family since Mnt/Max complexes act as transcriptional repressors and suppress Myc‐dependent activation of E‐box containing promoters (Fig. 1) (Hurlin et al., 1997a; Meroni et al., 1997). In similarity with the Mxd proteins, Mnt contains a Sin3‐interacting domain (SID), which enables recruitment of HDACs through interaction with the paired amphipathic helix 2 (PAH2) domain in Sin3 proteins (Fig. 1) (Ayer et al., 1995; Hurlin et al., 1997a; Meroni et al., 1997; Schreiber‐Agus et al., 1995). Deletion of the SID has been shown to convert Mnt into a transcriptional activator, probably due to the presence of proline‐rich residues, which resemble the activation domain of several transcription factors including Myc Box I (Hurlin et al., 1997a). Another possibility is that Myc can function as an activator when Mnt is lacking the SID due to relief of transcriptional repression.

Mnt, the Key Regulator of Myc Function

67

B. Mnt, the Major Myc Antagonist The conventional model, in which the Mnt and Mxd proteins function as antagonists of Myc activities, has mainly been supported by studies demonstrating that overexpression of the Mxd (Ayer et al., 1993; Lahoz et al., 1994; Roussel et al., 1996), Mnt (Hurlin et al., 1997a), or Mga genes (Hurlin et al., 1999; Ogawa et al., 2002) led to growth arrest and suppressed Myc/Ras‐ induced transformation of fibroblasts. It was therefore expected that deletion of the Mxd family genes would release Myc activity and that mice harboring these deletions would become tumor prone. However, Mxd1, Mxd2, or Mxd3 null mice were fertile and viable and did not develop tumors (Foley et al., 1998; Queva et al., 2001; Schreiber‐Agus et al., 1998). This is probably due to the fact that the other Mxd family genes compensate for the loss of the Mxd1, Mxd2, or Mxd3 genes. It will therefore be of great importance to study the effects in mice where all the four Mxd genes have been deleted. On the other hand, Mnt transgenic mouse embryos exhibited a delay in development and died in midgestation when Myc function is critical (Hurlin et al., 1997a). This highlights the importance of Mnt in the restriction of Myc activities associated with cell proliferation. However, the embryonic lethality observed in c‐Myc‐ or N‐Myc‐deficient mice was not rescued by simultaneous deletion of Mnt (Dezfouli et al., 2006). Thus, it has been suggested that Mnt can antagonize Myc activities and that Mxd family proteins may compensate, at least partially, for the loss of Mnt in certain tissues. Alternatively, Mnt and Myc may not function solely by controlling the same set of genes and molecular pathways, but may possess distinct and perhaps cell type‐specific biological activities (Dezfouli et al., 2006). In this respect, it is interesting that Mnt and Mnt/Max complexes have been detected in P19 mouse embryonic carcinoma cells, which lack Myc expression. This could imply that Myc may not be required for active proliferation in certain cell types. It is possible that pluripotent cells like P19 rely on Mnt to ensure that Myc target genes are in a repressed state even in the absence of Myc (Hurlin et al., 1997b). In addition, Drosophila Mnt (dMnt) is important for cell cycle progression, proliferation, and cell growth in the fly and, like in mammalian cells, binding of dMnt/dMax complexes at E‐box sequences and interaction with the dSin3 protein is required for transcriptional repression. Since both dMyc and dMax are highly conserved but no dMxd proteins have been identified, dMnt has been proposed to act as the sole repressor within this network in Drosophila (Orian et al., 2003). Furthermore, overexpression of dMnt resulted in smaller flies whereas dMnt null flies had larger cells, increased weight, and decreased life span compared to wild‐type flies, suggesting that dMnt is the regulator of Drosophila body size (Loo et al., 2005). However, in contrast to the mammalian protein, dMnt is not essential for the normal developmental program since no significant effects on development

68

Therese Wahlstro¨m and Marie Henriksson

during embryogenesis or delays in larval molting pupation or eclosion were observed following loss of dMnt (Loo et al., 2005). Furthermore, recent data has shown that deletion of both dMnt and dMyc generates a partial rescue of the lethality observed in dMyc null flies. The development continues nearly to the pupal stage, where the flies eventually die. During this rescue period, there is a huge amount of proliferation, DNA synthesis, and growth occurring in the absence of both dMyc and dMnt, probably due to that other factors can regulate the dMyc target genes involved. Pierce and Eisenman suggested that the growth inhibition observed in dMyc null flies is due to that the remaining dMnt represses dMyc target genes and that the repressional function of dMnt is normally overridden by dMyc during normal larval development (Pierce and Eisenman, personal communication). In Xenopus, expression of the different XMxd/XMnt genes is specific and mostly nonoverlapping, suggesting distinct roles during frog embryogenesis (Juergens et al., 2005). Interestingly, in the worm Caenorhabditis elegans, functional Max and Mnt but no Myc orthologues have been found, indicating that these genes are more evolutionary conserved than Myc (Yuan et al., 1998). In conclusion, the findings that dMnt plays a similar role in Drosophila as the Mxd proteins in vertebrates together with the fact that Mnt is ubiquitously expressed, in contrast to the more restricted expression of Mxd proteins (Hooker and Hurlin, 2006), support the indications of Mnt as the key Myc antagonist.

C. Effects of Mnt Deficiency Mice lacking Mnt die within 24 h after birth (Toyo‐oka et al., 2004), and spontaneous tumors develop in mice with conditional deletion of Mnt in breast epithelium (Hurlin et al., 2003, 2004). Characterization of the mammary gland tissue after Mnt deletion showed that loss of Mnt severely disrupts involution and leads to hyperplastic duct formation associated with reduced numbers of apoptotic cells (Toyo‐Oka et al., 2006). In addition, conditional deletion of Mnt in T cells has been shown to cause inflammatory disease, increased apoptosis in thymic T cells, interference with T cell development and T cell lymphomas (Dezfouli et al., 2006). These data reveal a critical role for Mnt in the regulation of T cell development, T cell‐dependent immune homeostasis, as well as a possible tumor suppressor function. Furthermore, Mnt null mice were not viable (Hurlin et al., 2003) in contrast to mice with deletion of Mxd1, Mxd2, or Mxd3 (Foley et al., 1998; Queva et al., 2001; Schreiber‐Agus et al., 1998). The Mnt‐ deficient embryos exhibited a small size throughout development with reduced levels of c‐Myc and N‐Myc, suggesting an important role for Mnt during embryonic development and survival (Toyo‐Oka et al., 2004). In this respect, it is interesting to note that XMnt is expressed in migrating cranial neuronal crest cells in the frog (Juergens et al., 2005), whereas Mnt‐deficient mice have cleft palates and craniofacial deformities (Toyo‐Oka et al., 2004), indicating a function for Mnt in cranial neuronal cells.

Mnt, the Key Regulator of Myc Function

69

Since deletion of Mnt in mice was lethal and the litters died early after birth, Mnt‐deficient MEFs were generated and analyzed (Hurlin et al., 2003). These cells entered S phase prematurely and proliferated more rapidly than wild type (wt) MEFs. The defective cell cycle control in the Mnt/ MEFs was linked to up‐regulation of cdk4 and cyclin E. In addition, other hallmarks of Myc overexpression, such as sensitivity to apoptosis, more efficient escape from senescence, and the ability for transformation induced by Ras were observed (Hurlin et al., 2003, 2004). Similarly, cells in which Mnt was down‐regulated by RNA interference (RNAi) were also shown to mimic Myc‐overexpressing cells (Nilsson et al., 2004). These data further demonstrate the unique role of Mnt in antagonizing Myc function. However, different results have been obtained depending on the techniques used to down‐regulate/delete Mnt. Mnt RNAi knockdown cells had a dramatic increase in proliferation, whereas the Mnt/ MEFs showed hyper‐proliferation at early passages and then displayed a sequential loss in c‐Myc expression and proliferated at a more reduced rate with increasing passage number (Hurlin et al., 2003; Nilsson and Cleveland, 2004; Nilsson et al., 2004). Furthermore, Mnt null MEFs showed a relatively modest increase in apoptosis in response to serum withdrawal, whereas Mnt RNAi knockdown generated a robust apoptotic response, also in fibroblasts lacking c‐Myc. In addition, the spectrum of Myc target genes induced by Mnt deficiency in the Mnt/ MEFs differed from that in the Mnt RNAi knockdown cells (Hurlin et al., 2003; Nilsson and Cleveland, 2004; Nilsson et al., 2004; Walker et al., 2005). The explanation for these discrepancies could be the use of different cell types and/or be attributed to the fact that transient down‐regulation of Mnt expression may have different effects compared to the sustained loss of Mnt resulting from gene deletion. It has been suggested that changes in the expression of the Mxd, Mga, or even Myc genes would compensate for the effects of Mnt loss (Hurlin et al., 2003). Such compensatory mechanisms have previously been observed in retinoblastoma (Rb) null cells, where an increase in the pRb‐related protein p107 prevented quiescent Rb/ fibroblasts from entering the cell cycle. In contrast, cells from the conditional Rb null mice cycled continuously (Sage et al., 2003). In conclusion, deletion of Mnt in mice and down‐regulation of Mnt in cells support the idea that Mnt may be of greater importance for the regulation of the Myc/Max/Mxd network than what was first anticipated.

D. Relief of Mnt‐Mediated Repression: The Critical Event for Myc Target Gene Activation In contrast to the rapid up‐regulation of c‐Myc upon mitogenic stimulation of quiescent cells, Mnt is expressed throughout the cell cycle and c‐Myc/Max and Mnt/Max coexist in a number of proliferating cell types (Hurlin et al., 1997a,b; Popov et al., 2005; Pulverer et al., 2000; Sommer et al., 1998, 1999). Mnt has been suggested to have an important role in regulation of cell cycle

70

Therese Wahlstro¨m and Marie Henriksson

progression because it suppresses cell cycle entry and proliferation when overexpressed and the G0 to S transition is accelerated in Mnt‐deficient MEFs (Hurlin et al., 2003). In addition, down‐regulation of Mnt by RNAi has been shown to be sufficient to provoke typical Myc responses, such as increased cell proliferation and transformation by Ras, also in c‐Myc null cells. However, even though cells lacking Mnt seem to be less dependent on Myc for proliferation, Myc up‐regulation is needed to force Mnt/ MEFs from quiescence to proliferation (Hurlin et al., 2003). The Myc target genes cdk4, cyclin D2, and ornithine decarboxylase (Odc) have been shown to be direct targets also for the Mnt protein (Hurlin et al., 2003; Nilsson et al., 2004; Popov et al., 2005; Walker et al., 2005). Thus, Mnt/Max and Myc/Max complexes compete for the same site at target gene promoters and these are derepressed in the absence of Mnt (Hurlin et al., 2003). The ability of Mnt to antagonize Myc has been suggested to be due to competition for interaction with Max and between Mnt/Max and Myc/Max for binding to the E‐box element at shared target genes (Hurlin et al., 2003). In addition, Mnt and c‐Myc were shown to have a significant overlap in promoter binding in tumor tissues of MMTV‐c‐Myc transgenic mice and similar mRNA expression patterns in mammary tumors from MMTV‐Cre/MntKO/CKO and MMTV‐c‐Myc transgenic mice (Toyo‐Oka et al., 2006). Furthermore, the level of Mnt/Max complexes has been shown to decrease concomitantly with a transient switch from Mnt/Max to c‐Myc/Max at E‐boxes on shared target genes when quiescent cells are reentering the cell cycle even though the total levels of Mnt protein remained unchanged (Fig. 2) (Hooker and Hurlin, 2006; Walker et al., 2005). Like Myc, the Mnt protein has a short half‐life of 25–30 min (Hurlin et al., 1997b), whereas Max has a half‐life of 24 h (Blackwood et al., 1992). Similarly, dMax has been shown to be the nonlimiting factor for genomic binding of the Myc network in Drosophila (Orian et al., 2003). The formation of novel Mnt/Max and c‐Myc/Max complexes is thus dependent on the stability of the Mnt/Max complex as well as of the ratio of newly synthesized Mnt and c‐Myc proteins at a certain time‐point since both proteins have similar half‐lives (Hooker and Hurlin, 2006; Walker et al., 2005). Mnt‐mediated repression may therefore be of greater importance for the impact on target genes than the sole activation by Myc as previously thought. We have presented data suggesting a model in which phosphorylation of Mnt at cell cycle entry results in disruption of the Mnt‐mSin3‐HDAC1 interaction, allowing induction of Myc target genes by relief of Mnt‐mediated transcriptional repression (Fig. 3) (Popov et al., 2005). Walker et al. (2005) have also demonstrated binding of the Mnt‐mSin3‐HDAC complex to the E‐box sequence, but did not observe regulation of the Mnt‐mSin3 interaction during cell cycle progression. In contrast, we found that phosphorylation of Mnt generated more of the slower migrating 74 kDa form of Mnt which,

Mnt, the Key Regulator of Myc Function

71

Fig. 2 Expression profiles for c‐Myc/Max and Mnt/Max complexes. The graph shows the relative expression levels of c‐Myc/Max and Mnt/Max dimers during cell cycle entry and during quiescence/differentiation (upper panel). A model for the relationship of c‐Myc/Max and Mnt/ Max complexes binding at target gene promoters at cell cycle entry (serum stimulation) during proliferation and quiescence is shown (lower panel). Mnt is expressed ubiquitously but the binding of Mnt/Max dimers at E‐box sequences has been shown to decrease slightly concomitant with the rapid up‐regulation of c‐Myc and binding of c‐Myc/Max to target gene promoters as cells reenter the cell cycle. Ac, acetyl group. The graph is modified from Fig. 2 in Hooker and Hurlin, of Myc and Mnt, J. Cell Sci. 119, 208–216, 2006 with permission from The Company of Biologists Ltd.

already in 1997, was suggested to be derived from the 72‐kDa protein (Hurlin et al., 1997a; Popov et al., 2005). We showed that the nonphosphorylated, faster migrating form of Mnt is the interaction partner of mSin3, and that Mnt‐ mediated transcriptional repression is reduced due to less efficient recruitment

72

Therese Wahlstro¨m and Marie Henriksson

Fig. 3 Model for regulation of Mnt‐mediated repression at cell cycle entry. The upper graph shows the relative expression levels of c‐Myc/Max and Mnt/Max dimers during cell cycle entry and during quiescence/differentiation. The cyclin D2 gene is a target for both the c‐Myc and Mnt proteins. In resting or differentiated cells, the Mnt‐mSin3‐HDAC1 interaction mediates repression of cyclin D2 transcription (quiescence). Phosphorylation of Mnt after serum stimulation disrupts the interaction with mSin3 proteins resulting in relief of transcriptional repression (Alt. 1). Serum stimulation also results in up‐regulated levels of c‐Myc and in binding of c‐Myc/Max to one of the E‐boxes in combination with Mnt/Max (Alt. 2), or alternatively, to both E‐boxes (Alt. 3) at the cyclin D2 promoter. This in turn results in cyclin D2 transcription. Ac, acetyl group. The graph is modified from Fig. 2 in Hooker and Hurlin, of Myc and Mnt, J. Cell Sci. 119, 208–216, 2006 with permission from The Company of Biologists Ltd.

of HDACs upon phosphorylation of Mnt at cell cycle entry (Popov et al., 2005). The consequences of this regulation have been shown on the cyclin D2 gene where we observed that the disruption between Mnt and mSin3 results in augmented cyclin D2 levels (Popov et al., 2005). The combination of a

Mnt, the Key Regulator of Myc Function

73

decreased Mnt‐mediated repression with an increased binding of newly synthesized c‐Myc/Max in relation to Mnt/Max heterodimers at the E‐box(es) of the cyclin D2 promoter are the likely reason for the increased levels of cyclin D2 after serum stimulation (Fig. 3). Conversely, there was an increase in the nonphosphorylated form of Mnt on DMSO‐induced differentiation of HL60 cells (Popov et al., 2005).

E. Mnt, to Be or Not to Be a Tumor Suppressor In addition to its function as an important antagonist and regulator of Myc activities, Mnt has been proposed as a potential tumor suppressor since the gene is located in a region frequently undergoing loss of heterozygosity (LOH) in a number of malignancies, including breast cancer (Devilee et al., 1989; Lindblom et al., 1993; Mackay et al., 1988; Stack et al., 1995), ovarian cancer (Phillips et al., 1993), astrocytoma (Saxena et al., 1992), bladder cancer (Williamson et al., 1994), medulloblastoma (McDonald et al., 1994), neuroectodermal cancer (Biegel et al., 1992), and ostesarcoma (Andreassen et al., 1993) (Table I). The fact that evidence for inactivation of Mnt in human tumors is still missing as neither mutations nor deletions in Mnt have been found in breast cancer (Nigro et al., 1998), lung cancer (Takahashi et al., 1998), or medulloblastoma (Sommer et al., 1999) raises the question whether Mnt is a bona fide tumor suppressor or not. One reason that Mnt inactivation has not been detected in human tumors could be that loss of Mnt results in increased sensitivity toward apoptosis inducers, and thus that such cells are most probably lost (Nilsson et al., 2004). On the other hand, expression of Mnt was reported to be reduced in 6 out of 14 medulloblastoma tumors analyzed, indicating that Mnt is indeed a potential tumor suppressor gene (Cvekl et al., 2004) (Table I). The notion that Mnt can suppress tumor growth is also supported by the gene deletion and RNAi studies, as discussed above (Dezfouli et al., 2006; Hurlin et al., 2003; Nilsson et al., 2004; Toyo‐Oka et al., 2006). In addition, mice injected with MEFs expressing Mnt RNAi and Ras developed rapidly growing fibrosarcomas and Ras alone transformed primary fibroblasts lacking c‐Myc when Mnt was down‐regulated by RNAi. Similarly, Balb/c‐fibroblasts expressing Mnt RNAi were capable of forming tumors, indicating that removal of Mnt is sufficient to provoke tumorigenesis in cells that have undergone immortalization (Nilsson et al., 2004) (Table I). LOH at chromosome 17p13.3 has been shown to involve a region telomeric to p53 (Hirano et al., 2001) and data clearly indicate the presence of another tumor suppressor gene(s) located at this chromosome. One candidate has been the hypermethylated in cancer (HIC1), but since heterozygous HIC1 mutant mice do not display mammary gland tumors, there are indications of another candidate tumor suppressor gene within 17p13.3 that remains to be identified.

74

Therese Wahlstro¨m and Marie Henriksson

Table I

a

Mnt, to Be or Not to Be a Tumor Suppressor Evidence in the literature

þ

Mnt is located at human chromosome 17p13.3, a region that frequently undergoes loss of heterozygosity in a number of malignancies

þ

Expression of Mnt is reduced in 6 out of 14 medulloblastoma tumors analyzed Mnt is ubiquitously expressed

þ þ

Cells with loss of Mnt show many hallmarks of Myc overexpression such as increased proliferation and sensitivity to apoptosis, more efficient escape from senescence, and the ability for in vitro and in vivo transformation by Ras Myc null primary fibroblasts are transformed by Ras alone on RNAi‐down‐regulation of Mnt Mnt loss is sufficient to provoke tumorigenesis in immortalized cells since Balb/c fibroblasts with RNAi‐down‐regulation of Mnt form tumors in injected mice Mnt deficiency or c‐Myc overexpression in mammary glands result in tumor development Relief of transcriptional repression by regulation of Mnt at cell cycle entry emphasizes the functional importance of the Mnt‐mSin3 interaction for suppression of tumor formation No mutation or deletion of Mnt has been found in medulloblastoma, breast, or lung cancer

þ þ

þ þ



a

References Andreassen et al., 1993; Biegel et al., 1992; Devilee et al., 1989; Hurlin et al., 1997a; Lindblom et al., 1993; Mackay et al., 1988; McDonald et al., 1994; Meroni et al., 1997; Phillips et al., 1993; Saxena et al., 1992; Stack et al., 1995; Williamson et al., 1994 Cvekl et al., 2004

Hurlin et al., 1997a, 2003, 2004; Meroni et al., 1997; Popov et al., 2005 Dezfouli et al., 2006; Hurlin et al., 2003; Nilsson et al., 2004; Toyo‐Oka et al., 2004, 2006

Nilsson et al., 2004

Nilsson et al., 2004

Hurlin et al., 2003; Hutchinson and Muller, 2000 Popov et al., 2005

Nigro et al., 1998; Sommer et al., 1999; Takahashi et al., 1998

Summary of findings supporting (þ) or discarding () the view that Mnt is a tumor suppressor gene. Accumulating data suggest that Mnt may have tumor suppressor function, but there is no evidence that the gene is nonfunctional in human tumors.

Mnt, the Key Regulator of Myc Function

75

Toyo‐Oka et al. (2006) have suggested Mnt as a possible candidate since it is well established that c‐Myc promotes mammary gland tumorigenesis in transgenic mice (Hutchinson and Muller, 2000) and that deletion of Mnt in mammary glands results in adenocarcinoma (Hurlin et al., 2003) (Table I). In addition, Mnt probably has an important role in mammary gland development preceding tumor development, as loss of Mnt impairs mammary gland involution. Further characterization of the tumor formation caused by loss of Mnt has also shown that the gene expression profile of tumors lacking Mnt is similar to that caused by c‐Myc overexpression (Toyo‐Oka et al., 2006). The finding that the Mnt‐ mSin3 binding is regulated by phosphorylation of Mnt at cell cycle entry (Fig. 3) emphasizes the importance of a functional interaction for suppression of tumor formation and thus of Mnt as a tumor suppressor gene (Table I). It will therefore be of great importance to analyze whether the interaction between Mnt and mSin3 is dysfunctional in tumors and to investigate whether a Mnt protein incapable of binding mSin3 contributes to tumor development due to lack of Mnt tumor suppressor function.

III. CONCLUDING REMARKS Mnt has a critical role in entry and regulation of the cell cycle and the ratio between Mnt/Max and c‐Myc/Max on the 50 ‐CACGTG‐30 sequence at shared target genes is of importance for proliferation. Phosphorylation of Mnt at cell cycle entry results in disruption of the Mnt‐mSin3‐HDAC1 interaction, which in turn causes concomitant up‐regulation of the Myc/Mnt target gene cyclin D2. Similarly, cyclin D2 protein levels increase upon expression of Mnt‐ targeting RNAi in quiescent cells. These data support the model of Mnt as the key transcriptional regulator of the Myc/Max/Mxd network. Thus, Myc may function through relief of Mnt‐mediated repression, and its oncogenic activity may rely on its ability to block transcriptional inhibition by Mnt. In addition, Mnt may act as a tumor suppressor and as such be inactivated or nonfunctional in human cancers. It will therefore be of great interest to further explore the putative tumor suppressor function of Mnt.

ACKNOWLEDGMENTS We apologize to colleagues whose work we were not able to cite due to the specific focus and to space limitations. We are indebted to A. Albihn for creating the figures and for critical reading and to J. Love´n for valuable comments. Research from the Henriksson Laboratory is supported by grants from the Swedish Cancer Society, the Swedish Children’s Cancer Foundation, the Hedlund’s Foundation, the Karolinska Institutet, and the Swedish Research Council.

76

Therese Wahlstro¨m and Marie Henriksson

REFERENCES Andreassen, A., Oyjord, T., Hovig, E., Holm, R., Florenes, V. A., Nesland, J. M., Myklebost, O., Hoie, J., Bruland, O. S., and Borresen, A. L. (1993). p53 abnormalities in different subtypes of human sarcomas. Cancer Res. 53, 468–471. Arabi, A., Wu, S., Ridderstrale, K., Bierhoff, H., Shiue, C., Fatyol, K., Fahlen, S., Hybring, P., Soderberg, O., Grummt, I., Larsson, L. G., and Wright, A. P. (2005). C‐Myc associates with ribosomal DNA and activates RNA polymerase I transcription. Nat. Cell Biol. 7, 303–310. Askew, D. S., Ashmun, R. A., Simmons, B. C., and Cleveland, J. L. (1991). Constitutive c‐myc expression in an IL‐3‐dependent myeloid cell line suppresses cell cycle arrest and accelerates apoptosis. Oncogene 6, 1915–1922. Ayer, D. E., Kretzner, L., and Eisenman, R. N. (1993). Mad: A heterodimeric partner for Max that antagonizes Myc transcriptional activity. Cell 72, 211–222. Ayer, D. E., Lawrence, Q. A., and Eisenman, R. N. (1995). Mad‐Max transcriptional repression is mediated by ternary complex formation with mammalian homologs of yeast repressor Sin3. Cell 80, 767–776. Biegel, J. A., Burk, C. D., Barr, F. G., and Emanuel, B. S. (1992). Evidence for a 17p tumor related locus distinct from p53 in pediatric primitive neuroectodermal tumors. Cancer Res. 52, 3391–3395. Blackwood, E. M., Luscher, B., and Eisenman, R. N. (1992). Myc and Max associate in vivo. Genes Dev. 6, 71–80. Bouchard, C., Dittrich, O., Kiermaier, A., Dohmann, K., Menkel, A., Eilers, M., and Luscher, B. (2001). Regulation of cyclin D2 gene expression by the Myc/Max/Mad network: Myc‐ dependent TRRAP recruitment and histone acetylation at the cyclin D2 promoter. Genes Dev. 15, 2042–2047. Charron, J., Malynn, B. A., Fisher, P., Stewart, V., Jeannotte, L., Goff, S. P., Robertson, E. J., and Alt, F. W. (1992). Embryonic lethality in mice homozygous for a targeted disruption of the N‐myc gene. Genes Dev. 6, 2248–2257. Cvekl, A., Jr., Zavadil, J., Birshtein, B. K., Grotzer, M. A., and Cvekl, A. (2004). Analysis of transcripts from 17p13.3 in medulloblastoma suggests ROX/MNT as a potential tumour suppressor gene. Eur. J. Cancer 40, 2525–2532. Davis, A. C., Wims, M., Spotts, G. D., Hann, S. R., and Bradley, A. (1993). A null c‐myc mutation causes lethality before 10.5 days of gestation in homozygotes and reduced fertility in heterozygous female mice. Genes Dev. 7, 671–682. de Alboran, I. M., O’Hagan, R. C., Gartner, F., Malynn, B., Davidson, L., Rickert, R., Rajewsky, K., DePinho, R. A., and Alt, F. W. (2001). Analysis of C‐MYC function in normal cells via conditional gene‐targeted mutation. Immunity 14, 45–55. de la Cova, C., Abril, M., Bellosta, P., Gallant, P., and Johnston, L. A. (2004). Drosophila myc regualtes organ size by inducing cell competition. Cell 117, 107–116. Devilee, P., van den Broek, M., Kuipers‐Dijkshoorn, N., Kolluri, R., Khan, P. M., Pearson, P. L., and Cornelisse, C. J. (1989). At least four different chromosomal regions are involved in loss of heterozygosity in human breast carcinoma. Genomics 5, 554–560. Dezfouli, S., Bakke, A., Huang, J., Wynshaw‐Boris, A., and Hurlin, P. J. (2006). Inflammatory disease and lymphomagenesis caused by deletion of the Myc antagonist Mnt in T cells. Mol. Cell Biol. 26, 2080–2092. Eilers, M., Schirm, S., and Bishop, J. M. (1991). The MYC protein activates transcription of the alpha‐prothymosin gene. EMBO J. 10, 133–141. Evan, G. I., Wyllie, A. H., Gilbert, C. S., Littlewood, T. D., Land, H., Brooks, M., Waters, C. M., Penn, L. Z., and Hancock, D. C. (1992). Induction of apoptosis in fibroblasts by c‐myc protein. Cell 69, 119–128.

Mnt, the Key Regulator of Myc Function

77

Felsher, D. W., and Bishop, J. M. (1999). Reversible tumorigenesis by MYC in hematopoietic lineages. Mol. Cell 4, 199–207. Foley, K. P., McArthur, G. A., Queva, C., Hurlin, P. J., Soriano, P., and Eisenman, R. N. (1998). Targeted disruption of the MYC antagonist MAD1 inhibits cell cycle exit during granulocyte differentiation. EMBO J. 17, 774–785. Giroux, S., and Charron, J. (1998). Defective development of the embryonic liver in N‐myc‐ deficient mice. Dev. Biol. 195, 16–28. Grewal, S. S., Li, L., Orian, A., Eisenman, R. N., and Edgar, B. A. (2005). Myc‐dependent regulation of ribosomal RNA sythesis during Drosophila development. Nat. Cell Biol. 7, 295–302. Grandori, C., Cowley, A. M., James, L. P., and Eisenman, R. N. (2000). The Myc/Max/Mad network and the transcriptional control of cell behavior. Annu. Rev. Dev. Biol. 16, 653–699. Grandori, C., Gomez‐Roman, N., Felton‐Edkins, Z. A., Ngouenet, C., Galloway, D. A., Eisenman, R. N., and White, R. J. (2005). C‐Myc binds to human ribosomal DNA and stimulates transcription of rRNA genes by RNA polymerase I. Nat. Cell Biol. 7, 311–318. Henriksson, M., and Lu¨scher, B. (1996). Proteins of the Myc network: Essential regulators of cell growth and differentiation. Adv. Cancer Res. 68, 109–182. Henriksson, M., Selivanova, G., Lindstrom, M., and Wiman, K. G. (2001). Inactivation of Myc‐induced p53‐dependent apoptosis in human tumors. Apoptosis 6, 133–137. Hirano, A., Emi, M., Tsuneizumi, M., Utada, Y., Yoshimoto, M., Kasumi, F., Akiyama, F., Sakamoto, G., Haga, S., Kajiwara, T., and Nakamura, Y. (2001). Allelic losses of loci at 3p25.1, 8p22, 13q12, 17p13.3, and 22q13 correlates with postoperative recurrence in breast cancer. Clin. Cancer Res. 7, 876–882. Hooker, C. W., and Hurlin, P. J. (2006). Of Myc and Mnt. J. Cell Sci. 119, 208–216. Hueber, A. O., and Evan, G. I. (1998). Traps to catch unwary oncogenes. Trends Genet. 14, 364–367. Hurlin, P. J., Queva, C., and Eisenman, R. N. (1997a). Mnt, a novel Max‐interacting protein is coexpressed with Myc in proliferating cells and mediates repression at Myc binding sites. Genes Dev. 11, 44–58. Hurlin, P. J., Queva, C., and Eisenman, R. N. (1997b). Mnt: A novel Max‐interacting protein and Myc antagonist. Curr. Top. Microbiol. Immunol. 224, 115–121. Hurlin, P. J., Steingrimsson, E., Copeland, N. G., Jenkins, N. A., and Eisenman, R. N. (1999). Mga, a dual‐specificity transcription factor that interacts with Max and contains a T‐domain DNA‐binding motif. EMBO J. 18, 7019–7028. Hurlin, P. J., Zhou, Z. Q., Toyo‐Oka, K., Ota, S., Walker, W. L., Hirotsune, S., and Wynshaw‐ Boris, A. (2003). Deletion of Mnt leads to disrupted cell cycle control and tumorigenesis. EMBO J. 22, 4584–4596. Hurlin, P. J., Zhou, Z. Q., Toyo‐Oka, K., Ota, S., Walker, W. L., Hirotsune, S., and Wynshaw‐ Boris, A. (2004). Evidence of mnt‐myc antagonism revealed by mnt gene deletion. Cell Cycle 3, 97–99. Hutchinson, J. N., and Muller, W. J. (2000). Transgenic mouse models of human breast cancer. Oncogene 19, 6130–6137. Jain, M., Arvanitis, C., Chu, K., Dewey, W., Leonhardt, E., Trinh, M., Sundberg, C. D., Bishop, J. M., and Felsher, D. W. (2002). Sustained loss of a neoplastic phenotype by brief inactivation of MYC. Science 297, 102–104. Johnston, L. A., Prober, D. A., Edgar, B. A., Eisenman, R. N., and Gallant, P. (1999). Drosophila myc regulates cellular growth during development. Cell 98, 779–790. Juergens, K., Rust, B., Pieler, T., and Henningfeld, K. A. (2005). Isolation and comparative expression analysis of the Myc‐regulatory proteins Mad1, Mad3, and Mnt during Xenopus development. Dev. Dyn. 233, 1554–1559.

78

Therese Wahlstro¨m and Marie Henriksson

Knoepfler, P. S., Cheng, P. F., and Eisenman, R. N. (2002). N‐myc is essential during neurogenesis for the rapid expansion of progenitor cell populations and the inhibition of neuronal differentiation. Genes Dev. 16, 2699–2712. Lahoz, E. G., Xu, L., Schreiber‐Agus, N., and DePinho, R. A. (1994). Suppression of Myc, but not E1a, transformation activity by Max‐associated proteins, Mad and Mxi1. Proc. Natl. Acad. Sci. USA 91, 5503–5507. Larsson, L. G., Pettersson, M., Oberg, F., Nilsson, K., and Luscher, B. (1994). Expression of Mad, mxi1, max and c‐myc during induced differentiation of hematopoietic cells: Opposite regulation of mad and c‐myc. Oncogene 9, 1247–1252. Lindblom, A., Skoog, L., Andersen, T. I., Rotstein, S., Nordenskjold, M., and Larsson, C. (1993). Four separate regions on chromosome 17 show loss of heterozygosity in familial breast carcinomas. Hum. Genet. 91, 6–12. Loo, L. W., Secombe, J., Little, J. T., Carlos, L. S., Yost, C., Cheng, P. F., Flynn, E. M., Edgar, B. A., and Eisenman, R. N. (2005). The transcriptional repressor dMnt is a regulator of growth in Drosophila melanogaster. Mol. Cell. Biol. 25, 7078–7091. Mackay, J., Steel, C. M., Elder, P. A., Forrest, A. P., and Evans, H. J. (1988). Allele loss on short arm of chromosome 17 in breast cancers. Lancet 2, 1384–1385. Maines, J. Z., Stevens, L. M., Tong, X., and Stein, D. (2004). Drosophila dMyc is required for ovary cell growth endoreplication. Development 131, 775–786. Mateyak, M. K., Obaya, A. J., Adachi, S., and Sedivy, J. M. (1997). Phenotypes of c‐Myc‐ deficient rat fibroblasts isolated by targeted homologous recombination. Cell Growth Differ. 8, 1039–1048. McDonald, J. D., Daneshvar, L., Willert, J. R., Matsumura, K., Waldman, F., and Cogen, P. H. (1994). Physical mapping of chromosome 17p13.3 in the region of a putative tumor suppressor gene important in medulloblastoma. Genomics 23, 229–232. Meroni, G., Reymond, A., Alcalay, M., Borsani, G., Tanigami, A., Tonlorenzi, R., Nigro, C. L., Messali, S., Zollo, M., Ledbetter, D. H., Brent, R., Ballabio, A., et al. (1997). Rox, a novel bHLHZip protein expressed in quiescent cells that heterodimerizes with Max, binds a non‐ canonical E box and acts as a transcriptional repressor. EMBO J. 16, 2892–2906. Moberg, K. H., Mukherjee, A., Veraksa, A., Artavanis‐Tsakonas, S., and Hariharan, I. K. (2004). The Drosophila F box protein archipelago regulates dMyc protein levels in vivo. Curr. Biol. 14, 965–974. Moens, C. B., Auerbach, A. B., Conlon, R. A., Joyner, A. L., and Rossant, J. (1992). A targeted mutation reveals a role for N‐myc in branching morphogenesis in the embryonic mouse lung. Genes Dev. 6, 691–704. Nigro, C. L., Venesio, T., Reymond, A., Meroni, G., Alberici, P., Cainarca, S., Enrico, F., Stack, M., Ledbetter, D. H., Liscia, D. S., Ballabio, A., and Carrozzo, R. (1998). The human ROX gene: Genomic structure and mutation analysis in human breast tumors. Genomics 49, 275–282. Nilsson, J. A., and Cleveland, J. L. (2004). Mnt: Master regulator of the Max network. Cell Cycle 3, 588–590. Nilsson, J. A., Maclean, K. H., Keller, U. B., Pendeville, H., Baudino, T. A., and Cleveland, J. L. (2004). Mnt loss triggers Myc transcription targets, proliferation, apoptosis, and transformation. Mol. Cell. Biol. 24, 1560–1569. Ogawa, H., Ishiguro, K., Gaubatz, S., Livingston, D. M., and Nakatani, Y. (2002). A complex with chromatin modifiers that occupies E2F‐ and Myc‐responsive genes in G0 cells. Science 296, 1132–1136. Orian, A., van Steensel, B., Delrow, J., Bussemaker, H. J., Li, L., Sawado, T., Williams, E., Loo, L. W., Cowley, S. M., Yost, C., Pierce, S., Edgar, B. A., et al. (2003). Genomic binding by the Drosophila Myc, Max, Mad/Mnt transcription factor network. Genes Dev. 17, 1101–1114.

Mnt, the Key Regulator of Myc Function

79

Oster, S. K., Ho, C. S., Soucie, E. L., and Penn, L. Z. (2002). The myc oncogene: Marvelously complex. Adv. Cancer Res. 84, 81–154. Pelengaris, S., Khan, M., and Evan, G. (2002a). c‐MYC: More than just a matter of life and death. Nat. Rev. Cancer 2, 764–776. Pelengaris, S., Khan, M., and Evan, G. (2002b). Suppression of Myc‐induced apoptosis in beta cells exposes multiple oncogenic properties of Myc and triggers carcinogenic progression. Cell 109, 321–334. Phillips, N., Ziegler, M., Saha, B., and Xynos, F. (1993). Allelic loss on chromosome 17 in human ovarian cancer. Int. J. Cancer 54, 85–91. Pierce, S., Yost, C., Britton, J. S., Loo, L. W., Flynn, E. M., Edgar, B. A., and Eisenman, R. N. (2004). dMyc is required for larval growth and endoreplication in Drosophila. Development 131, 2317–2327. Popov, N., Wahlstrom, T., Hurlin, P. J., and Henriksson, M. (2005). Mnt transcriptional repressor is functionally regulated during cell cycle progression. Oncogene 24, 8326–8337. Pulverer, B., Sommer, A., McArthur, G. A., Eisenman, R. N., and Luscher, B. (2000). Analysis of Myc/Max/Mad network members in adipogenesis: Inhibition of the proliferative burst and differentiation by ectopically expressed Mad1. J. Cell. Physiol. 183, 399–410. Queva, C., McArthur, G. A., Iritani, B. M., and Eisenman, R. N. (2001). Targeted deletion of the S‐phase‐specific Myc antagonist Mad3 sensitizes neuronal and lymphoid cells to radiation‐induced apoptosis. Mol. Cell. Biol. 21, 703–712. Roussel, M. F., Ashmun, R. A., Sherr, C. J., Eisenman, R. N., and Ayer, D. E. (1996). Inhibition of cell proliferation by the Mad1 transcriptional repressor. Mol. Cell. Biol. 16, 2796–2801. Sage, J., Miller, A. L., Perez‐Mancera, P. A., Wysocki, J. M., and Jacks, T. (2003). Acute mutation of retinoblastoma gene function is sufficient for cell cycle re‐entry. Nature 424, 223–228. Sawai, S., Shimono, A., Hanaoka, K., and Kondoh, H. (1991). Embryonic lethality resulting from disruption of both N‐myc alleles in mouse zygotes. New Biol. 3, 861–869. Saxena, A., Clark, W. C., Robertson, J. T., Ikejiri, B., Oldfield, E. H., and Ali, I. U. (1992). Evidence for the involvement of a potential second tumor suppressor gene on chromosome 17 distinct from p53 in malignant astrocytomas. Cancer Res. 52, 6716–6721. Schreiber‐Agus, N., Chin, L., Chen, K., Torres, R., Rao, G., Guida, P., Skoultchi, A. I., and DePinho, R. A. (1995). An amino‐terminal domain of Mxi1 mediates anti‐Myc oncogenic activity and interacts with a homolog of the yeast transcriptional repressor SIN3. Cell 80, 777–786. Schreiber‐Agus, N., Meng, Y., Hoang, T., Hou, H., Jr., Chen, K., Greenberg, R., Cordon‐ Cardo, C., Lee, H. W., and DePinho, R. A. (1998). Role of Mxi1 in ageing organ systems and the regulation of normal and neoplastic growth. Nature 393, 483–487. Shen‐Li, H., O’Hagan, R. C., Hou, H., Jr., Horner, J. W., III, Lee, H. W., and DePinho, R. A. (2000). Essential role for Max in early embryonic growth and development. Genes Dev. 14, 17–22. Smith, A. G., Popov, N., Imreh, M., Axelson, H., and Henriksson, M. (2004). Expression and DNA‐binding activity of MYCN/Max and Mnt/Max during induced differentiation of human neuroblastoma cells. J. Cell. Biochem. 92, 1282–1295. Sommer, A., Bousset, K., Kremmer, E., Austen, M., and Luscher, B. (1998). Identification and characterization of specific DNA‐binding complexes containing members of the Myc/Max/ Mad network of transcriptional regulators. J. Biol. Chem. 273, 6632–6642. Sommer, A., Waha, A., Tonn, J., Sorensen, N., Hurlin, P. J., Eisenman, R. N., Luscher, B., and Pietsch, T. (1999). Analysis of the Max‐binding protein MNT in human medulloblastomas. Int. J. Cancer 82, 810–816. Stack, M., Jones, D., White, G., Liscia, D. S., Venesio, T., Casey, G., Crichton, D., Varley, J., Mitchell, E., Heighway, J., and Santibanez‐Koref, M. (1995). Detailed mapping and loss of

80

Therese Wahlstro¨m and Marie Henriksson

heterozygosity analysis suggests a suppressor locus involved in sporadic breast cancer within a distal region of chromosome band 17p13.3. Hum. Mol. Genet. 4, 2047–2055. Stanton, B. R., Reid, S. W., and Parada, L. F. (1990). Germ line transmission of an inactive N‐myc allele generated by homologous recombination in mouse embryonic stem cells. Mol. Cell. Biol. 10, 6755–6758. Takahashi, T., Konishi, H., Kozaki, K., Osada, H., Saji, S., Takahashi, T., and Takahashi, T. (1998). Molecular analysis of a Myc antagonist, ROX/Mnt, at 17p13.3 in human lung cancers. Jpn. J. Cancer Res. 89, 347–351. Toyo‐Oka, K., Hirotsune, S., Gambello, M. J., Zhou, Z. Q., Olson, L., Rosenfeld, M. G., Eisenman, R., Hurlin, P., and Wynshaw‐Boris, A. (2004). Loss of the Max‐interacting protein Mnt in mice results in decreased viability, defective embryonic growth and craniofacial defects: Relevance to Miller‐Dieker syndrome. Hum. Mol. Genet. 13, 1057–1067. Toyo‐Oka, K., Bowen, T. J., Hirotsune, S., Li, Z., Jain, S., Ota, S., Escoubet Lozach, L., Garcia Basset, I., Lozach, J., Rosenfeldt, M. G., Glass, C. K., Eisenman, R., et al. (2006). Mnt‐ deficient mammary glands exhibit impaired involution and tumors with characteristics of Myc overexpression. Cancer Res. 66, 5565–5573. Trumpp, A., Refaeli, Y., Oskarsson, T., Gasser, S., Murphy, M., Martin, G. R., and Bishop, J. M. (2001). c‐Myc regulates mammalian body size by controlling cell number but not cell size. Nature 414, 768–773. Vita, M., and Henriksson, M. (2006). The Myc oncoprotein as a therapeutic target for human cancer. Semin. Cancer Biol. 16, 318–330. Walker, W., Zhou, Z. Q., Ota, S., Wynshaw‐Boris, A., and Hurlin, P. J. (2005). Mnt‐Max to Myc‐Max complex switching regulates cell cycle entry. J. Cell Biol. 169, 405–413. Williamson, M. P., Elder, P. A., and Knowles, M. A. (1994). The spectrum of TP53 mutations in bladder carcinoma. Genes Chromosomes Cancer 9, 108–118. Wilson, A., Murphy, M. J., Oskarsson, T., Kaloulis, K., Bettess, M. D., Oser, G. M., Pasche, A. C., Knabenhans, C., Macdonald, H. R., and Trumpp, A. (2004). c‐Myc controls the balance between hematopoietic stem cell self‐renewal and differentiation. Genes Dev. 18, 2747–2763. Xu, D., Popov, N., Hou, M., Wang, Q., Bjorkholm, M., Gruber, A., Menkel, A. R., and Henriksson, M. (2001). Switch from Myc/Max to Mad1/Max binding and decrease in histone acetylation at the telomerase reverse transcriptase promoter during differentiation of HL60 cells. Proc. Natl. Acad. Sci. USA 98, 3826–3831. Yuan, J., Tirabassi, R. S., Bush, A. B., and Cole, M. D. (1998). The C. elegans MDL‐1 and MXL‐1 proteins can functionally substitute for vertebrate MAD and MAX. Oncogene 17, 1109–1118.

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses1 George Miller,*,{ Ayman El‐Guindy,z Jill Countryman,z Jianjiang Ye,z and Lyn Gradoville* *Department of Pediatrics, Yale University School of Medicine, New Haven, Connecticut 06520; { Department of Epidemiology and Public Health, Yale University School of Medicine, New Haven, Connecticut 06520; z Department of Molecular Biophysics and Biochemistry, Yale University School of Medicine, New Haven, Connecticut 06520

I. Two Life Cycles of EBV and KSHV: Latency and Lytic Replication II. Virally Encoded Lytic Cycle Activator Genes A. Upstream and Downstream Events in Lytic Cycle Activation B. Agents That Induce the Lytic Cycle C. Role of Phosphorylation in the Downstream Functions of the EBV ZEBRA Protein III. Conclusions: Some Unsolved Mysteries About Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses References

The seminal experiments of George and Eva Klein helped to define the two life cycles of Epstein–Barr Virus (EBV), namely latency and lytic or productive infection. Their laboratories described latent nuclear antigens expressed during latency and discovered several chemicals that activated the viral lytic cycle. The mechanism of the switch between latency and the lytic cycle of EBV and Kaposi’s sarcoma‐associated herpesvirus (KSHV) can be studied in cultured B cell lines. Lytic cycle activation of EBV is controlled by two viral transcription factors, ZEBRA and Rta. The homologue of Rta encoded in ORF50 is the lytic cycle activator of KSHV. Control of the lytic cycle can be divided into two distinct phases. Upstream events control expression of the virally encoded lytic cycle activator genes. Downstream events represent tasks carried out by the viral proteins in driving expression of lytic cycle genes and lytic viral DNA replication. In this chapter, we report three recent groups of experiments relating to upstream and downstream events. Azacytidine (AzaC) is a DNA methyltransferase inhibitor whose lytic cycle activation capacity was discovered by G. Klein and coworkers. We find that AzaC rapidly activates the EBV lytic cycle but does not detectably alter DNA methylation or histone acetylation on the promoters of the EBV lytic cycle activator genes. AzaC probably acts via a novel, yet to be elucidated, mechanism. The lytic cycle of both

1

Nobel Symposium on Cancer Research, “Molecular Oncology‐from Bench to Bedside,” in Honor of George and Eva Klein. Advances in CANCER RESEARCH Copyright 2007, Elsevier Inc. All rights reserved.

81

0065-230X/07 $35.00 DOI: 10.1016/S0065-230X(06)97004-3

George Miller et al.

82

EBV and KSHV can be activated by sodium butyrate (NaB), a histone deacetylase inhibitor whose activity in disrupting latency was also discovered by G. Klein and coworkers. Activation of EBV by NaB requires protein synthesis; activation of KSHV is independent of protein synthesis. Thus, NaB works by a different pathway on the two closely related viruses. ZEBRA, the major downstream mediator of EBV lytic cycle activation is both a transcription activator and an essential replication protein. We show that phosphorylation of ZEBRA at its casein kinase 2 (CK2) site separates these two functions. Phosphorylation by CK2 is required for ZEBRA to activate lytic replication but not to induce expression of early lytic cycle genes. We discuss a number of unsolved mysteries about lytic cycle activation which should provide fertile territory for future research. # 2007 Elsevier Inc.

I. TWO LIFE CYCLES OF EBV AND KSHV: LATENCY AND LYTIC REPLICATION From the earliest days of research on Epstein–Barr virus (EBV) it was clear that not all cells that were infected with the virus produced mature viral progeny. When the Henles originally described the “viral capsid antigen,” using indirect immunofluorescence techniques, they observed the antigen only in a subpopulation of cells from Burkitt’s lymphoma (Henle and Henle, 1966). Single cell cloning experiments showed that the progeny of single cells invariably gave rise to populations of cells in which only a subpopulation expressed the viral capsid antigen (Miller et al., 1970). This result suggested that all cells contained latent viral genomes which were periodically reactivated to produce virus in a subpopulation. Moreover, some cell lines, such as the famous Raji Burkitt’s lymphoma cells, never spontaneously expressed viral capsid antigen (Epstein et al., 1966). However, Raji cells could be demonstrated to contain an antigen that was recognized by sera from EBV‐infected individuals. Complement fixation was required to detect this antigen (Armstrong et al., 1966; Pope et al., 1969). In a classic paper in 1973, Beverly Reedman and George Klein used immunofluorescence to detect the complement‐fixing antigen in the nucleus (Reedman and Klein, 1973) (Fig. 1A). This nuclear antigen was found in all cells in cell lines that spontaneously produced virus in a few cells and in those, such as Raji, that did not produce virus. Further work showed that Raji did not express capsid proteins because EBV DNA in Raji is unable to replicate in a lytic fashion. The EBV genome in Raji cells contains a deletion of BALF2, the gene encoding single‐stranded DNA‐binding protein, required for lytic EBV DNA replication (Decaussin et al., 1995; Zhang et al., 1988). The complement‐fixing nuclear antigen discovered by Reedman and Klein, which they called “EBNA” (EB nuclear antigen), is now known to consist of the products of at least six different EBV genes. The component of EBNA most commonly detected by infected individuals is EBNA1, the product of the EBV BKRF1 gene (Fischer et al., 1984; Summers et al., 1982). EBNA1 is responsible for partitioning of EBV DNA during latency from parental to daughter cells and for mediating

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

83

Fig. 1 Detection of latent and lytic EBV gene expression in B95–8 cells (Miller and Lipman, 1973). (A) Anticomplement immunofluorescence according to the method of Reedman and Klein (Reedman and Klein, 1973). The majority of cells show nuclear staining. Two bright cells (arrows) contain “viral capsid antigen.” (B) Fluorescent in situ hybridization (FISH) with EBV Bam H1 W as a probe. Latently infected cells contain approximately 20 genomes per cell. Following treatment with TPA approximately 25% of the cell population contain brightly staining viral replication factories indicative of lytic cycle activation.

attachment of the viral genome to cell chromosomes (Kapoor et al., 2001; Yates et al., 1984). During latency, the EBV genome exists as an extrachromosomal element (Nonoyama and Pagano, 1972). Most latently infected cells contain approximately 20 copies of viral DNA that can be visualized by fluorescence in situ hybridization (Fig 1B). When the virus is induced to replicate, either spontaneously or after the addition of a chemical‐inducing agent, the viral DNA in a subpopulation of cells aggregates into replication factories (Gradoville

84

George Miller et al.

et al., 2002). Thus, there are two distinct life cycles of EBV, latency and lytic replication, which can be studied in cultured lymphoid cells. The relationship between EBVand lymphoid cells offers a unique experimental system in which to study control of the latency to lytic cycle switch. A very similar system exists for study of the related gammaherpesvirus, Kaposi’s sarcoma‐associated herpesvirus (KSHV), which remains latent in cultured B cells derived from primary effusion lymphoma (Cesarman et al., 1995). A number of laboratories, including those of George and Eva Klein, have spent considerable effort over the past 40 years attempting to unravel the complex biology of the latency to lytic cycle switch of the oncogenic human gammaherpesviruses. Aside from the obvious advantage, by comparison to other herpesviruses, that the latency to lytic cycle switch of gammaherpesviruses can be experimentally manipulated in cultured cells, why all this effort? It is a biologically interesting example of the combinatorial regulation of eukaryotic gene expression. Control is exerted by groups of virally and cellular‐encoded transcription factors, acting both as activators and repressors. Epigenetic controls of promoter chromatinization and methylation come into play. The latency to lytic switch has obvious implications for the pathogenesis of malignant and nonmalignant diseases associated with EBV and KSHV. While latency may be the predominant life cycle form in virus‐associated cancers, the virus must replicate and assume a lytic state in order to be transmitted between cells and among individuals. Being able to manipulate the transition between latency and the lytic cycle offers the promise of translational application to virus‐associated cancers. For example, tumor lysis might be achieved by activating lytic viral replication. Thereafter, spread of virus to new cells could be inhibited by antiviral drugs directed at viral DNA replication (Westphal et al., 2000). Although the latency to lytic cycle switch can be manipulated in cell culture, particularly by the addition of one or more chemical‐inducing agents, little is understood about the physiologic stimulus that triggers the event in vivo. The stimulus may be external, for example receipt of an activating cytokine signal or removal of an inhibitory signal. Or the physiologic stimulus may be internal, reflecting the metabolic state of the cell, or its position in the cell cycle.

II. VIRALLY ENCODED LYTIC CYCLE ACTIVATOR GENES Through study of a defective EBV genome in a Burkitt’s lymphoma cell line, P3J HR‐1, Jill Countryman and colleagues discovered that the lytic cycle of EBV was controlled by a transcription factor protein encoded in the EBV BZLF1 gene (Countryman and Miller, 1985). When expressed in latently

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

85

infected cells, this protein, variably named ZEBRA, Zta, or EB‐1, is sufficient to activate the entire EBV lytic cycle cascade, leading to production of infectious virus (Grogan et al., 1987). ZEBRA protein, in some ways analogous to the prokaryotic lambda phage repressor, was the first example of latency to lytic cycle switch protein encoded for by a eukaryotic virus. ZEBRA plays many roles in the lytic phase of the viral life cycle. ZEBRA is essential to activate lytic cycle genes (Feederle et al., 2000). One of its earliest and most critical functions is to activate expression of another transcription factor, Rta, encoded by the viral BRLF1 gene (Kolman et al., 1996). ZEBRA and Rta then act in synergy to activate transcription of a subset of early lytic cycle genes many of which encode proteins required for lytic viral DNA replication (Quinlivan et al., 1993; Ragoczy and Miller, 1999). Rta appears to be capable of activating a distinct set of viral genes independent of ZEBRA (Chen et al., 2005; Ragoczy and Miller, 1999). ZEBRA represses the action of Rta on some promoters in a temporally controlled manner (El‐Guindy and Miller, 2004; Ragoczy and Miller, 1999). ZEBRA plays a distinct role in activating viral lytic DNA replication by binding to the origin of lytic cycle replication (oriLyt) and by interacting with and recruiting viral proteins that are essential for lytic replication (Baumann et al., 1999; Fixman et al., 1992; Gao et al., 1998; Lieberman et al., 1990; Schepers et al., 1993, 1996) (Fig. 10B). Experiments using EBV bacmids in which the BZLF1 and BRLF1 have been insertionally inactivated show that both genes are required for lytic cycle activation (Feederle et al., 2000). An unanswered question is why EBV employs two different viral transcription factors to control the lytic cycle. KSHV or Human herpesvirus 8 encodes homologues of ZEBRA and Rta (Fig. 2A). K8‐bZIP, the ZEBRA homologue, exists in several spliced isoforms (Lin et al., 1999). K8‐bZIP is essential for viral DNA replication, but unlike ZEBRA, it does not seem to play a role as a transcriptional activator of early genes (Fig 2B). The job of activation of KSHV viral early lytic cycle genes is performed by KSHV ORF50 protein, the homologue of EBV Rta (Sun et al., 1998).

A. Upstream and Downstream Events in Lytic Cycle Activation Early experiments with EBV showed that expression of ZEBRA strictly correlated with lytic cycle activation (Fig. 3A). When ZEBRA was constitutively expressed, as the result of genome rearrangements in the defective EBV genome (so‐called “het” DNA), or induced to be expressed, as the result of treatment of cells with inducing agents such as phorbol esters or sodium butyrate, the virus was forced to enter the lytic cycle. When neither ZEBRA protein nor BZLF1 mRNA were expressed, the virus was in latency. Expression of ZEBRA obligated the virus to enter the full lytic cascade.

86

George Miller et al.

Fig. 2 Lytic cycle activator genes encoded by EBV and KSHV. (A) Homologous genes and gene products of EBV and KSHV. (B) Functions of the activator genes in lytic cycle expression. Notably K8‐bZIP, the ZEBRA homologue in KSHV, is a replication protein but not an activator of early gene transcription.

Thus, the latency to lytic cycle switch might be envisioned to be composed of two distinct sets of events: (1) upstream events leading to expression of the lytic cycle activator genes (ZEBRA and Rta in the case of EBV; ORF50 in the case of KSHV) and (2) downstream events by which the products of the lytic cycle activator genes control viral and cellular gene expression and viral and cellular DNA replication. A simple model for the entry pathway into the EBV lytic cascade (Fig. 3B) suggests that during latency the BZLF1 and BRLF1 genes are repressed. The mechanism of repression is likely to involve both epigenetic mechanisms such as chromatization and DNA methylation of the BZLF1 promoter, Zp, and BRLF1 promoter, Rp, as well as occupancy of these promoters by specific repressors. One candidate repressor for Zp, called ZEB1, has been identified (Kraus et al., 2001, 2003). Following application of an inducing stimulus, there is likely to be a signaling pathway that allows cell‐encoded activators to bind to Zp and Rp. Candidates for this phase include cellular proteins such as Sp1, c/EBP , AP‐1, and CREB. These cellular factors lead to low‐level expression of ZEBRA and Rta, which then amplify the signal by autostimulating and cross‐stimulating each other’s expression. ZEBRA specifically binds to several ZEBRA response sites in Rp. A methylated CpG is embedded in one of those sites. ZEBRA binds preferentially to methylated DNA (Bhende et al., 2004, 2005). Rta autostimulates Rp by an indirect mechanism that involves interaction with other proteins such as those of the Sp1 family (Ragoczy and Miller, 2001). ZEBRA appears to mediate all of its downstream activities by a direct DNA‐binding mechanism. We have completed saturation mutagenesis of

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

87

Fig. 3 Models for activation of the EBV lytic cycle cascade. (A) An early “ZEBRA‐centric” model in which all inducing stimuli lead to activation of ZEBRA expression and thereafter downstream events. (B) A later model incorporating control of the EBV lytic cycle by BRLF1 and BZLF1 genes. During latency, Rp and Zp, the promoters of these genes are repressed by histones (H). These promoters contain TPA response elements (TRE) and ZEBRA (Z) response elements (ZRE). Following induction, cell encoded activators (A) including Ap‐1 factors bind to the promoters. In the autostimulation phase both ZEBRA (Z) and Rta (R) activate Rp. Autoactivation of the Rp by Rta is indirectly mediated by X, probably a member of the Sp1 family.

88

George Miller et al.

the DNA recognition domain of ZEBRA and found that any mutation that eliminated ZEBRA binding to its highest affinity ZEBRA response elements (ZRE) abolishes the biologic activity of the protein (Heston et al., 2006). EBV Rta and its homologue KSHV ORF50 activate gene expression by more diverse mechanisms than does ZEBRA (Fig. 4A and B). EBV Rta has three types of target genes. One group consists of those genes it activates independently of ZEBRA by binding to Rta response elements in the promoter (Chen et al., 2005). A second group of genes is activated as the result of Rta and ZEBRA both binding to the promoter of responsive genes (Francis et al., 1999; Quinlivan et al., 1993). A third group of target genes is activated indirectly by Rta operating through binding sites for cellular transcription factors (Furnari et al., 1992). Rta may influence the activity of cellular

Fig. 4 Mechanisms of action of the EBV Rta and KSHV ORF50 proteins. (A) EBV/Rta activates one group of genes autonomously by a direct DNA‐binding mechanism. A second group of genes is activated in synergy with ZEBRA. A third group of genes is activated via cellular transcription factors. (B) KSHV/ORF50 activates one group of genes by direct DNA binding and two other groups by interaction with cellular transcription factors.

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

89

transcription factors by protein–protein interactions and by activating signaling cascades that lead to posttranslational modification of the cellular transcription factors (Adamson et al., 2000). KSHV ORF50 also activates downstream targets by a variety of mechanisms. These mechanisms can be distinguished by ORF50 mutants that do or do not bind to DNA (Chang et al., 2005). The promoters of direct targets, such as PAN and K12, share response elements that bind ORF50 protein (Chang et al., 2002). One group of indirect target genes has been shown to be activated by ORF50 mutants that do not bind DNA. These mutants interact with the RBP‐J protein (Liang et al., 2002).

B. Agents That Induce the Lytic Cycle A number of different chemicals have been found to induce the EBV and KSHV lytic cycle (Fig. 5 presents a partial list). All the chemicals appear to operate by activating expression of the virally encoded lytic cycle activator genes, that is, they impinge on crucial upstream events. George Klein was a leader in investigating chemicals that could activate EBV lytic gene expression. For example, Klein and his colleagues, Janos Luka and Bengt Kallin, were the first to show that sodium butyrate, an inhibitor of histone deacetylase (HDAC), is a potent lytic cycle activator in some cell backgrounds (Luka et al., 1979). In our laboratory, we often use two prototype cancer cell lines to study the mechanism of lytic cycle activation—HH514‐16, an EBV‐infected cell line from Burkitt’s lymphoma, and HH‐B2, a KSHV‐ infected cell line from primary effusion lymphoma (Gradoville et al., 2000). The advantage of these cell lines as model systems is that they manifest a very low background level of spontaneous lytic cycle activation; on application of the inducing stimulus a high proportion (up to 50%) of the cells undergo lytic

Fig. 5 A partial list of agents commonly used to induce the EBV and KSHV lytic cycle in cultured lymphoid cell lines. Prototype lymphoma cell lines used to study gammaherpesvirus lytic cycle activation in our laboratory are HH514–16 from Burkitt’s lymphoma and HH‐B2 from primary effusion lymphoma.

90

George Miller et al.

cycle activation. In both of these cell lines, sodium butyrate reproducibly activates the viral lytic cycle. There are at least three major unanswered questions about the mechanism by which lytic cycle activation stimuli work. First, the list of chemicals in Fig. 5 shows that known inducing stimuli operate via many different modes of action. The inducing chemicals include activators of protein kinase C (PKC), inhibitors of HDACs, and DNA methyltransferase inhibitors. Is there a final common pathway to activation of the promoters of the lytic cycle activator genes or can several distinct unrelated events activate these genes? Second, even with the most potent activation stimulus in the most susceptible cell line, only a subpopulation of the cells respond. What accounts for this refractory subpopulation? Is it due to permanent or transient changes in the viral or cellular genome? Finally, the same inducing stimulus does not activate lytic cycle gene expression in all cells backgrounds. What accounts for the differences among cell lines in their response to inducing stimuli?

1. THE SAME STIMULUS DOES NOT ACTIVATE EBV LYTIC CYCLE GENE EXPRESSION IN ALL CELL BACKGROUNDS At the symposium George Klein pointed out that the differences in response of EBV‐containing cell lines to various inducing stimuli that work by many diverse mechanisms represents a major unsolved puzzle about the gammaherpes lytic cycle switches. For example, our laboratory stocks of B95‐8 cells, a marmoset lymphoblastoid cell line derived by in vitro immortalization with EBV, respond strongly to phorbol ester (Fig. 6A). The cell line HH514‐16, a subclone of the Burkitt’s lymphoma Jijoye line, fails to express lytic cycle genes after treatment with phorbol ester (Gradoville et al., 2002). TPA is a potent agonist for PKC. The levels of PKC activity increase in both responsive and unresponsive cell lines following treatment with TPA. Therefore, it is likely that some additional event, downstream of PKC activation, that is required for EBV lytic activation does not take place in the HH514‐16 Burkitt’s lymphoma‐ derived cell line. Conversely, EBV lytic cycle gene expression is reproducibly activated to a high level in HH514‐16 cells by HDAC inhibitors such as sodium butyrate and Trichostatin A (TSA) (Fig. 6B). Activation of the lytic cycle in HH514‐16 cells by HDAC inhibitors is not accompanied by detectable activation of PKC nor is lytic cycle activation by HDAC inhibitors blocked by bisindolymaleimide, a potent inhibitor of PKC (Gradoville et al., 2002). These results lead to the conclusion that activation of PKC is not universally required for EBV lytic cycle activation. B95‐8 cells do not increase lytic cycle gene expression following treatment with sodium butyrate or TSA. However, as expected, these agents induce hyperacetylation of histone tails, H3 and H4, in B95‐8 cells. The hyperacetylation is both “global” (i.e., detectable by increased levels of H3 and H4 hyperacetylation on immunoblots) and specific to the promoters of the BZLF1 (ZEBRA) and BRLF1 (Rta) genes (i.e., detectable by

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

A

91

TPA

Cell line A (e.g., B95-8) PKC TPA Cell line B (e.g., HH514-16) PKC B

TSA

Cell line A (e.g., B95-8) acH3 and acH4 TSA Cell line B (e.g., HH514-16) acH3 and acH4

Fig. 6 Inducing agents exhibit cell line‐specific behavior. (A) Effect of phorbol ester (TPA). TPA activates PKC in two EBV‐containing cell lines, but in only one cell line, A, is this stimulus adequate to induce the lytic cycle. Latent genomes: open black circles; replicating genomes: black‐shaded circles; virions: open gray circles. (B) Effect of an HDAC inhibitor, Trichostatin A (TSA). In both cell lines, TSA causes hyperacetylation of histone tails acH3 and acH4. Only in cell line B does this stimulus lead to EBV lytic cycle reactivation.

chromatin immunoprecipitation with antibody to acetylated H3 and histone tails H4) (Countryman et al., 2006). These results show that histone hyperacetylation is neither a sufficient nor universal stimulus to activate EBV lytic gene expression. There appears to be an activity, in addition to hyperacetylation of Rp and Zp, that is essential for HDAC inhibitors to disrupt EBV latency. This activity may involve acetylation of nonhistone proteins or other properties of the HDAC inhibitors that are not related to protein acetylation.

2. MYSTERIES ABOUT THE MECHANISMS BY WHICH AZACYTIDINE INDUCES EBV LYTIC GENE EXPRESSIONS Samuel Ben‐Sasson and George Klein demonstrated that 5‐Azacytidine activated EBV early lytic gene expression in human lymphoid lines latently infected with EBV (Ben‐Sasson and Klein, 1981). We have found that in some B cell backgrounds, such as HH514‐16 cells, the DNA methyltransferase

George Miller et al.

92

inhibitor 5‐Aza‐20 ‐deoxcytidine (5Aza20 dC) very rapidly and potently induces expression of the EBV BRLF1 and BZLF1 mRNAs. An increase in the levels of lytic cycle activator mRNAs can be detected within 4 h after addition of 5Aza20 dC (Fig. 7A). The effects are more rapid than those of A

4

6

8 NaB 5Aza2⬘dC

Induction:

2

NaB 5Aza2⬘dC

0

NaB 5Aza2⬘dC

HRS.

NaB 5Aza2⬘dC

HH514-16

Rp (4.0) -

Markers (Kb) − 4.4

Rp (3.0) − 2.4 − 1.35 Zp (1.0) -

RNaseP (0.35) − 0.24 1

2

3 4

5

6 7

8 9 10

11 12 13

Probe: Z(300) B

4.5 4

Fold stimulation

3.5 3 2.5 2 1.5 1 0.5 0 0

4 Hours HH514-16

6 NaB 5Aza2⬘dC

93

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

C (a)

Zp

100.0 90.0

% Methylation

80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0

Uninduced

(b)

NaB

+188/+189

+56/+57

+13/+14

−79/−78

−274/−273

−337/−336

−351/−350

−359/−358

0.0

5Aza2⬘dC

Rp

100.0 90.0

% Methylation

80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 −5/−4

−49/−48

−58/−57

−96/−95

−113/−112

−144/−143

−196/−195

−254/−253

−256/−255

0.0

HH514-16 cells Uninduced

NaB

5Aza2⬘dC

Fig. 7 Activation of the EBV lytic cycle by 5‐Aza‐20 ‐deoxycytidine (5Aza20 dC) and sodium butyrate (NaB). Experiments were conducted in HH514‐16 cells (cell line B in Fig. 6). (A) Kinetics of activation of BRLF1 (Rp 3.0) and BZLF1 (Zp 1.0) mRNAs by the two stimuli. (B) Hyperacetylation of Rp determined by chromatin immunoprecipitation with antibodies to acetylated H3. (C) Lack of effect of the two inducing stimuli on CpG methylation of Zp or Rp determined by bisulfite sequencing 8 h after induction.

94

George Miller et al.

sodium butyrate and, at early times, relatively more pronounced on the BLZF1 mRNA than on the BRLF1 mRNA. 5Aza20 dC does not detectably effect the acetylation state of histone H3 on Rp at 4 or 6 h after addition (Fig. 7B). Moreover, at 8 h after addition neither Azacytidine (AzaC) nor sodium butyrate affects the methylation state of the promoters of the BZLF1 and BRLF1 genes (Fig. 7C). Thus, at times when 5Aza20 dC is inducing the lytic cycle, there is no detectable change either in DNA methylation of the promoters of the lytic cycle activator genes or in hyperacetylation of H3 on these promoters. It is unlikely, therefore, that EBV lytic cycle induction by 5Aza20 dC is operating directly by epigenetic alterations on the promoters of BZLF1 and BRLF1. There appears to be another yet to be discovered mechanism of action of AzaC.

3. DIFFERING EFFECTS OF CYCLOHEXIMIDE ON LYTIC CYCLE REACTIVATION OF EBV AND KSHV BY SODIUM BUTYRATE An important unanswered question is whether stimuli that induce the EBV and KSHV lytic cycle work on transcription factors that are already expressed in the cell at the time when the cell receives the lytic cycle‐ inducing stimulus. The transcription factors may be poised, ready to work after they are activated, for example, by posttranslational modification. Alternatively, lytic cycle activation stimuli might change the epigenetic state of the promoters of the lytic cycle activation genes or remove repressors, thus allowing preexisting competent transcription factors access to the promoters. If either of these two scenarios is operative, a lytic cycle activation stimulus, such as sodium butyrate, might not require de novo protein synthesis. Rather, sodium butyrate might work by activating a signal transduction cascade that modifies a repressor or by “opening chromatin” via the hyperacetylation of histone tails. As originally defined by Roizman and colleagues, a herpesvirus “immediate‐ early gene” can be transcribed in the presence of an inhibitor of protein synthesis such as cycloheximide (CHX) (Frenkel et al., 1973; Honess and Roizman, 1974). An explanation for immediate‐early kinetics in herpes simplex virus is that the virus carries a preformed transcription factor (VP16) within the virion. This transcription factor activates immediate‐early downstream genes following de novo infection of cells with the virus. The question whether EBV BZLF1 and BRLF1 and KSHV ORF50 are immediate‐early genes has not yet been carefully investigated. It is unknown whether EBV BZLF1 and BRLF1 and KSHV ORF50, the lytic cycle activator genes, behave with “immediate‐early” kinetics on de novo infection or on lytic cycle reactivation after addition of a potent lytic cycle‐inducing stimulus such as sodium butyrate.

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

95

We recently investigated the question of immediate‐early behavior on lytic cycle reactivation in two prototypic cell lines infected with EBV or KSHV in which sodium butyrate activates the lytic cycle (Fig. 5). CHX was added at intervals after the inducing stimulus. In HH514‐16 cells (Fig. 8), appearance of the EBV BRLF1 and BZLF1 mRNAs was inhibited by addition of CHX added up to 4 h after sodium butyrate. By contrast, induction of ORF50 mRNA by sodium butyrate in HH‐B2 cells was resistant to the action of CHX at all times. In fact, if CHX was added simultaneously with sodium butyrate, there was enhanced expression of ORF50 mRNA, a finding consistent with the known activity of CHX in stabilizing mRNA. These results indicated that for lytic cycle reactivation, de novo protein synthesis appeared to be required for expression of the EBV BRLF1 and BZLF1 mRNAs but not for KSHV ORF50 mRNA. One limitation of this initial experiment was that the effects of CHX on lytic cycle reactivation were studied in two different cell backgrounds. Many primary effusion lymphoma cell lines are dually infected with EBV and KSHV (Cesarman et al., 1995). In one such dually infected cell line, BC‐1, sodium butyrate induces lytic cycle expression of both EBV and KSHV. Five hours after CHX was added together with sodium butyrate (NaB),

Fig. 8 Differential effects of CHX on lytic cycle induction of EBV and KSHV by sodium butyrate (NaB). Each cell line was untreated or treated with sodium butyrate (NaB). CHX was added simultaneously with NaB or at 2‐h intervals thereafter. Lytic cycle activator mRNAs were analyzed on Northern blots 8 h after addition of NaB. (A) Probed for EBV BRLF1 (Rp) and BZLF1 (Zp) mRNAs. (B) Probed for KSHV ORF50 mRNA.

96

George Miller et al.

the abundance of EBV BRLF1 mRNA was 33% of the level measured after NaB in the absence of CHX (Fig. 9A). At the same time, in the same cell line, the expression of KSHV ORF50 mRNA was not inhibited; in fact, it was slightly stabilized (Fig. 9B). These experiments indicate that KSHV ORF50 behaves with immediate‐ early kinetics on lytic cycle reactivation. Lytic cycle reactivation of KSHV, induced by Na butyrate, is likely to require modification of existing transcription factors and/or chromatin remodeling of the ORF50 promoter. Studies show that sodium butyrate treatment enhances the interaction of Sp1 protein with the ORF50 promoter (Lu et al., 2003; Ye et al., 2005).

Fig. 9 Effect of CHX on EBV and KSHV lytic cycle induction by sodium butyrate (NaB) in dually infected BC‐1 cells. BC‐1 cells were untreated or treated with NaB. One set of cultures received CHX. RNA, harvested at intervals after addition of NaB, was analyzed for EBV BRLF1 or KSHV ORF50 by quantitative real‐time reverse transcriptase PCR. At 3, 5, and 8 h, EBV BRLF1 was inhibited by CHX. At 3 and 5 h, KSHV ORF50 was resistant to CHX. S.I., stimulation index.

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

97

In striking contrast, EBV BRLF1 (Rta) and BZLF1 (ZEBRA) require de novo protein synthesis before reactivation and are therefore not “immediate‐early” genes. One hypothesis is that EBV reactivation requires de novo protein synthesis of cellular transcription factor(s) that regulate the promoters of the BZLF1 and BRLF1 genes.

C. Role of Phosphorylation in the Downstream Functions of the EBV ZEBRA Protein The preliminary unpublished experiments presented in Figs. 7–9 relate to upstream events leading to lytic cycle activation. As an example of downstream studies, we will describe how phosphorylation influences the function of the ZEBRA protein. ZEBRA is a member of the basic zipper (bZIP) family of transcription factor proteins (Flemington and Speck, 1990; Kouzarides et al., 1991) (Fig. 10A). It shares homology with cellular bZIP proteins such as c‐Jun, c‐Fos, and GCN4, especially in the basic DNA recognition domain. ZEBRA forms a homodimer in the absence of DNA. It binds specifically to AP‐1 sites and to a large family of heterogenous ZRE (Lehman et al., 1998). ZEBRA can be considered to consist of five modular elements: a transcriptional activation domain (aa 1–167), a regulatory domain (aa 168– 177), a basic DNA recognition domain (aa 178–194), a dimerization domain (aa 195–227), and an accessory tail region that stabilizes the homodimer (aa 228–245) (Petosa et al., 2006). ZEBRA is a phosphoprotein: it is constitutively phosphorylated in vivo at S167 and S173 [casein kinase 2 (CK2) sites], at T14 (a putative JNK site), and at S6, T7, S8 (El‐Guindy et al., 2006). It can be phosphorylated at T159 and S186 by PKC in vitro and in vivo when PKC is activated (Baumann et al., 1998). A report indicates that ZEBRA can also be phosphorylated by EBV protein kinase encoded in BGLF4 (Asai et al., 2006). ZEBRA performs many functions in the lytic viral life cycle (Fig. 10B): (1) it directly activates the BRLF1 (Rta) gene, (2) it synergizes with Rta in activating some downstream early genes, (3) it represses the activation of late genes by Rta at early times in the viral lytic cycle, and (4) it acts as an essential lytic viral replication protein (Fixman et al., 1992). Furthermore, ZEBRA impinges on many cellular pathways: (1) it activates MAP kinases such as p38 and JNK (Adamson et al., 2000), (2) it mediates cell cycle arrest (Cayrol and Flemington, 1996a,b), and (3) it exerts immunomodulatory effects by inhibiting the action of interferon, TNF‐ and NF‐B (Morrison and Kenney, 2004; Morrison et al., 2001, 2004). It is of obvious interest to know which of these many functions of the ZEBRA protein are influenced by its phosphorylation state.

98

George Miller et al.

Fig. 10 Functions of EBV ZEBRA protein. (A) Functional domains of the ZEBRA protein with identified phosphorylation sites. Constitutive phosphorylation by CK2 and JNK is shown in bold. Phosphorylation of ZEBRA by PKC is inducible (Baumann et al., 1998). Residues that are strongly phosphorylated in vitro (T14, S173, and S186) are shown in bold (El‐Guindy et al., 2006). (B) Four different functions of ZEBRA in EBV lytic cycle activation: (1) activation of transcription of the EBV BRLF1 promoter, (2) synergy with Rta in activation of the EBV BMRF1 promoter, (3) repression of Rta activation of the BLRF2 promoter at early times, and (4) as a lytic origin of DNA replication‐binding protein.

Phosphorylation of ZEBRA at its CK2 sites, especially S173, influences its capacity to act as a repressor (El‐Guindy and Miller, 2004). A ZEBRA mutant Z(S173A) that cannot be phosphorylated by CK2 loses its capacity to repress the ability of Rta to activate a viral late gene BRLF2 at early times during the lytic cycle. An inhibitor of CK2, TBB, also blocks the capacity of ZEBRA to act as a repressor of Rta. Neither the mechanism of repression by ZEBRA, when phosphorylated at S173, nor its temporal control is well understood. At early times in the lytic cycle, when ZEBRA is phosphorylated at its major CK2 site, it may inhibit the activity of Rta or block the access of Rta to DNA. At late times in the lytic cycle, after DNA replicates,

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

99

ZEBRA no longer acts as a repressor of late gene activation by Rta. It is likely that the phosphorylation state of ZEBRA may be temporally regulated, with ZEBRA becoming dephosphorylated at its CK2 site after, or concomitant with, lytic viral DNA replication (Fig. 12A). Many other questions about the role of ZEBRA as a repressor remain to be investigated. Why is the repressive activity of ZEBRA, when phosphorylated by CK2, promoter specific? What differentiates promoters that are activated from those that are repressed by ZEBRA? Do ZEBRA proteins in different phosphorylated states act as activators and repressors? How does a constitutively expressed kinase, such as CK2, control the activity of a transcription factor? If the phosphorylation state of ZEBRA changes at different time‐points during the viral life cycle, as we postulate, is ZEBRA phosphorylation somehow related to lytic viral DNA replication?

1. CK2 SITE MUTANTS ARE SELECTIVELY DEFECTIVE AT ACTIVATING EBV LATE GENE EXPRESSION Ayman El‐Guindy recently investigated the role of the CK2 phosphorylation site in temporal control of EBV lytic gene expression by comparing viral early and late gene expression induced by wild‐type ZEBRA and a mutant ZEBRA, Z(S167A/S173A) in which both potential CK2 sites were changed to alanine. The experiments were conducted in BZKO cells which harbor an EBV genome containing an insertionally inactivated BZLF1 gene (Delecluse et al., 1999; Feederle et al., 2000). ZEBRA with mutant CK2 sites was equivalent to wild‐type ZEBRA protein in its ability to activate expression of viral early products, the transcription factor Rta, and the EA‐D (BMRF1) DNA polymerase processivity factor. However, the Z(S167A/ S173A) mutant was markedly defective at activating expression of a late small viral capsid protein encoded by EBV BFRF3 (Fig. 11A). We considered a number of possible explanations to account for this remarkable phenotype. (1) Since the CK2 site mutant fails to act as a repressor, it could permit or facilitate the accumulation of a product that inhibits late gene expression. (2) Since Z(S167A/S173A) activates the EBV early genes BRLF1 and BMRF1, a ZEBRA protein mutated at the CK2 site does not possess a general deficit in transcriptional activation. However, such a mutant might have a selective defect that renders it unable to activate specific viral early genes or cellular genes whose products are required for lytic viral DNA replication. (3) The CK2 site mutant ZEBRA protein might be unable to activate transcription of late genes. (4) Since transcription of late genes is dependent on lytic DNA replication, the CK2 site mutant might be deficient in promoting lytic viral DNA replication. El‐Guindy has obtained convincing evidence favoring the last hypothesis.

100

George Miller et al.

Fig. 11 An alanine substitution mutant at ZEBRA’s principal CK2 phosphorylation site is defective at inducing late gene expression and lytic viral DNA replication. (A) BZKO cells, which contain an insertionally inactivated BZLF1 gene, were transfected with CMV vector, ZEBRA, or Z(S167A/S173A) mutant. At 24‐h intervals after transfection the cells were analyzed for latent, immediate‐early, early, or late EBV protein expression by immunoblotting with specific antibodies. (B) BZKO cells transfected with wild‐type or CK2 site mutant ZEBRA expression vectors were analyzed for EBV lytic DNA replication by Southern blotting with a probe adjacent to the viral terminal repeats (Raab‐Traub and Flynn, 1986).

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

101

2. CK2 SITE MUTANTS OF ZEBRA ARE IMPAIRED IN THEIR ABILITY TO ACTIVATE LYTIC VIRAL DNA REPLICATION Two complementary approaches were used to investigate the importance of the CK2 site of ZEBRA in activating lytic EB viral DNA replication. Using quantitative real‐time PCR, the total amount of EBV DNA in BZKO cells that had been transfected with an empty vector, or expression vectors for ZEBRA or the CK2 site mutant was measured. Wild‐type ZEBRA stimulated a 62‐fold increase in viral DNA; the Z(S167A/S173A) mutant activated a sixfold increase. Southern blots were used to detect lytic DNA replication by probing for the ladder of DNA restriction fragments containing variable number of EBV terminal repeats (Fig. 11B). The ladder that reflects lytic replication was observed when BZKO cells were transfected with wild‐type ZEBRA or the Z(S167A) mutant. However, only a faint ladder was detected after transfection of the mutant Z(S173A) or the double mutant Z(S167A/S173A). These experiments showed that the CK2 site mutant was impaired in activation of lytic viral DNA replication and that residue S173 in the regulatory domain was principally involved in regulating lytic viral DNA synthesis. The question remains, what is the mechanism by which phosphorylation of ZEBRA at residue S173 licenses the protein to function in lytic DNA replication? ZEBRA is an origin‐binding protein. The EBV lytic origin of replication contains seven binding sites for ZEBRA, at least four of which are essential for the origin to function (Schepers et al., 1996). El‐Guindy has shown in three consecutive replicate chromatin immunoprecipitation experiments that the ZEBRA S173A mutant is approximately one‐third as active as wild‐type protein in binding to oriLyt in vivo. In vitro electrophoretic mobility shift DNA‐binding experiments also demonstrate that the Z(S173A) mutant is markedly deficient by comparison to wild type at binding all seven of the ZRE in oriLyt. Therefore, phosphorylation of ZEBRA at S173 achieves maximal EBV lytic DNA replication by promoting high DNA‐binding affinity. However, DNA‐binding affinity may not be the only property of ZEBRA that is regulated by phosphorylation at S173. In order to function as an essential EBV, lytic replication protein ZEBRA must possess at least two other functions. It must localize to the subcompartment of the nucleus, so called “factories” (Fig. 1B) where lytic viral DNA replication occurs. ZEBRA must also recruit and form a complex with other virally encoded replication proteins (Fig. 10B). Phosphorylation of ZEBRA at S173 may be required for the protein to fulfill these functions. The ZEBRA mutant S173A is fully competent to activate expression of Rta or diffuse early antigen (EA‐D) (Fig. 11A). Although it is possible that this mutant is specifically defective at activating expression of certain genes required for DNA replication, the current data support the view that

102

George Miller et al.

phosphorylation at S173 is not essential for maximal transcriptional activation of viral early genes. El‐Guindy has done ChIP experiments to measure the association of wild‐type ZEBRA and the Z(S173A) mutant to Rp, the promoter of the BRLF1 (Rta) gene. Again the Z(S173A) mutant is about one‐third as active as wild type in associating with Rp in vivo. This interesting finding has two implications: first, phosphorylation of S173 has a general enhancing effect on DNA binding in vivo rather than a site‐specific effect. Second, since the CK2 site mutant of ZEBRA is impaired in replication but not in transcriptional activation, replication may be more sensitive than transcription to decreases in DNA‐binding affinity. The EBV origin of lytic replication contains seven contiguous ZEBRA‐ binding sites, while most ZEBRA responsive promoters contain 1, 2, or 3 sites. Thus, cooperative high affinity binding by ZEBRA may be needed to initiate DNA replication. The DNA‐binding requirement for recruitment of the transcription machinery may not be so stringent.

3. SUMMARY OF FUNCTIONAL ROLE OF CK2 PHOSPHORYLATION OF ZEBRA ON LYTIC CYCLE ACTIVATION Lytic cycle‐inducing stimuli that cause EBV to leave the latent state lead to expression of the viral ZEBRA and Rta proteins. These two transcription factors synergize to activate most early lytic genes. Early genes encode

Fig. 12 Models for the role of phosphorylation of ZEBRA at its CK2 site in the temporal control of the EBV lytic cycle. (A) At early times in the lytic cycle, when ZEBRA is phosphorylated at S173, it acts as a repressor of Rta’s capacity to activate an EBV late gene. At late times, ZEBRA may become dephosphorylated and no longer acts as a repressor. (B) At early times in the lytic cycle when ZEBRA is phosphorylated at S173, it can function optimally as an origin‐binding protein and promote lytic viral DNA replication.

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

103

products required for DNA replication. Lytic replication is needed for late gene expression. Rta has the capacity to activate expression of some late genes at early times in the viral lytic cycle. Such aberrant late gene expression may inhibit DNA replication by a feedback pathway. Phosphorylation of ZEBRA at its CK2 site in the regulatory domain maximizes the DNA‐binding affinity of ZEBRA. ZEBRA in its phosphorylated high‐affinity DNA‐binding state performs two functions that are critical for proper temporal control of the EBV life cycle. Phosphorylated ZEBRA represses the activation of Rta of late genes at early times. Phosphorylated ZEBRA acts as a lytic origin‐binding protein, thus facilitating lytic viral DNA replication (Fig. 12B).

III. CONCLUSIONS: SOME UNSOLVED MYSTERIES ABOUT LYTIC CYCLE SWITCHES OF ONCOGENIC HUMAN GAMMAHERPESVIRUSES In this essay, we classify the events leading to lytic cycle activation either as upstream events, resulting in repression or activation of the promoters of the lytic cycle activator genes of EBV and KSHV, and downstream events, resulting from activation or repression of viral lytic gene promoters and viral lytic origins of replication by the lytic cycle activator proteins, the ZEBRA and Rta proteins of EBV and ORF50 protein of KSHV. We have pointed out many, yet to be resolved, mysteries about this deceptively simple biologic system. Among the mysteries about upstream events, we would mention the following: why do EBV and KSHV genomes exist as multicopy plasmids? When lytic induction occurs, does each genome respond or are the genomes heterogeneous in their response? Although the lytic cycle can be manipulated in cultured B cells, we know very little about the physiologic stimuli that repress or activate the lytic cycle in vivo. Is the physiologic signal external, mediated by cytokines, hormones, or cells of the immune system, or does lytic cycle activation occur as a cellular response to stress such as hypoxia or change in energy charge of the cell? Not all cultured cell lines respond to the same inducing stimulus. Some cell lines respond to inducing stimuli, such as PKC agonists, HDAC inhibitors, and DNA methyltransferase inhibitors, which act via quite distinct mechanisms. These observations raise several questions about lytic cycle activation in cultured cells. What accounts for differences in response of different cell lines to inducing stimuli? Do lytic cycle‐inducing agents that act by different biochemical mechanisms operate via a common pathway or are there several distinct routes to lytic cycle activation?

104

George Miller et al.

There are many unresolved issues about the mechanism of action of inducing stimuli. For example, is the action of HDAC inhibitors, such as sodium butyrate or Trichostatin A, mediated solely by opening chromatin on the promoters of the lytic cycle activator genes or are additional events required? In preliminary data, we show that the DNA methyltransferase inhibitor 5Aza20 dC rapidly induces lytic cycle gene expression without detectably altering the methylation state or acetylation of histone tails on the promoters of the lytic cycle activator genes. What additional property of 5Aza20 dC is responsible for this rapid effect? Even in the most responsive cell line exposed to the most potent inducing stimulus only a subpopulation of the cells respond. What accounts for the refractory subpopulation? What mediates the refractory state? Are some cells permanently marked for lack of response or is the refractory state transient? In one set of experiments (Figs. 8 and 9), we ask whether inducing stimuli activate transcription factors that are already present in the cell, or whether these transcription factors need to be synthesized before lytic cycle activation can occur. We find, surprisingly, that CHX, an inhibitor of protein synthesis, blocks lytic cycle induction of EBV but not KSHV. Thus this question has two different answers for the two gammaherpesviruses. This result poses another important question: What protein(s) need to be synthesized in order for EBV ZEBRA and Rta to be made? A central unanswered question about the downstream events is why EBV employs two different transcription factors, ZEBRA and Rta, to control expression of lytic cycle genes, whereas KSHV employs only one, ORF50 the homologue of Rta? What is the mechanism of synergy between ZEBRA and Rta? The EBV ZEBRA protein plays two distinct roles: as a transcription factor and as a DNA replication protein. The ZEBRA homologue in KSHV, that is K8‐bZIP, is mainly a DNA replication protein. What features of the two bZIP proteins account for these differences in biologic activity? ZEBRA acts both as an activator and as a repressor of the promoters of lytic cycle genes. The capacity of ZEBRA to act as a repressor of late genes is mediated by phosphorylation of residue S173. How does phosphorylation mediate this activity? How is phosphorylation temporally regulated? What mechanism accounts for the promoter specificity of repression? Finally, how does phosphorylation of ZEBRA license its role as an origin‐binding protein? Does enhanced DNA‐binding activity suffice to explain the role of phosphorylated ZEBRA in DNA replication? Does phosphorylation of ZEBRA affect interactions with cellular and viral replication proteins? Does phosphorylation of ZEBRA alter its subnuclear localization? The search for answers to these and many other questions makes study of the lytic cycle switches of the oncogenic human herpesviruses a fascinating and fertile field.

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

105

ACKNOWLEDGMENTS Work in G.M.’s laboratory was supported by grants from the United States National Cancer Institute: CA12055, CA16038, CA70036, and CA52228. This essay emphasizes work in the authors’ laboratories and is not intended as a formal review of the published literature. We thank Susan Prisley and Karen Lavery for assistance in the preparation of the figures and chapter. We thank H. J. Delecluse for the BZKO cells.

REFERENCES Adamson, A. L., Darr, D., Holley‐Guthrie, E., Johnson, R. A., Mauser, A., Swenson, J., and Kenney, S. (2000). Epstein‐Barr virus immediate‐early proteins BZLF1 and BRLF1 activate the ATF2 transcription factor by increasing the levels of phosphorylated p38 and c‐Jun N‐terminal kinases. J. Virol. 74(3), 1224–1233. Armstrong, D., Henle, G., and Henle, W. (1966). Complement‐fixation tests with cell lines derived from Burkitt’s lymphoma and acute leukemias. J. Bacteriol. 91(3), 1257–1262. Asai, R., Kato, A., Kato, K., Kanamori‐Koyama, M., Sugimoto, K., Sairenji, T., Nishiyama, Y., and Kawaguchi, Y. (2006). Epstein‐Barr virus protein kinase BGLF4 is a virion tegument protein that dissociates from virions in a phosphorylation‐dependent process and phosphorylates the viral immediate‐early protein BZLF1. J. Virol. 80(11), 5125–5134. Baumann, M., Mischak, H., Dammeier, S., Kolch, W., Gires, O., Pich, D., Zeidler, R., Delecluse, H. J., and Hammerschmidt, W. (1998). Activation of the Epstein‐Barr virus transcription factor BZLF1 by 12‐O‐tetradecanoylphorbol‐13‐acetate‐induced phosphorylation. J. Virol. 72(10), 8105–8114. Baumann, M., Feederle, R., Kremmer, E., and Hammerschmidt, W. (1999). Cellular transcription factors recruit viral replication proteins to activate the Epstein‐Barr virus origin of lytic DNA replication, oriLyt. [published erratum appears in EMBO J. 2000 Jan 17; 19(2), 315]. EMBO J. 18(21), 6095–6105. Ben‐Sasson, S. A., and Klein, G. (1981). Activation of the Epstein‐Barr virus genome by 5‐aza‐ cytidine in latently infected human lymphoid lines. Int. J. Cancer 28(2), 131–135. Bhende, P. M., Seaman, W. T., Delecluse, H. J., and Kenney, S. C. (2004). The EBV lytic switch protein, Z, preferentially binds to and activates the methylated viral genome. Nat. Genet. 36(10), 1099–1104. Bhende, P. M., Seaman, W. T., Delecluse, H. J., and Kenney, S. C. (2005). BZLF1 activation of the methylated form of the BRLF1 immediate‐early promoter is regulated by BZLF1 residue 186. J. Virol. 79(12), 7338–7348. Cayrol, C., and Flemington, E. (1996a). G0/G1 growth arrest mediated by a region encompassing the basic leucine zipper (bZIP) domain of the Epstein‐Barr virus transactivator Zta. J. Biol. Chem. 271(50), 31799–31802. Cayrol, C., and Flemington, E. K. (1996b). The Epstein‐Barr virus bZIP transcription factor Zta causes G0/G1 cell cycle arrest through induction of cyclin‐dependent kinase inhibitors. EMBO J. 15(11), 2748–2759. Cesarman, E., Chang, Y., Moore, P. S., Said, J. W., and Knowles, D. M. (1995). Kaposi’s sarcoma‐associated herpesvirus‐like DNA sequences in AIDS‐related body‐cavity‐based lymphomas. N. Engl. J. Med. 332(18), 1186–1191. Chang, P. J., Shedd, D., Gradoville, L., Cho, M. S., Chen, L. W., Chang, J., and Miller, G. (2002). Open reading frame 50 protein of Kaposi’s sarcoma‐associated herpesvirus directly

106

George Miller et al.

activates the viral PAN and K12 genes by binding to related response elements. J. Virol. 76(7), 3168–3178. Chang, P. J., Shedd, D., and Miller, G. (2005). Two subclasses of Kaposi’s sarcoma‐associated herpesvirus lytic cycle promoters distinguished by open reading frame 50 mutant proteins that are deficient in binding to DNA. J. Virol. 79(14), 8750–8763. Chen, L. W., Chang, P. J., Delecluse, H. J., and Miller, G. (2005). Marked variation in response of consensus binding elements for the Rta protein of Epstein‐Barr virus. J. Virol. 79(15), 9635–9650. Countryman, J., and Miller, G. (1985). Activation of expression of latent Epstein‐Barr herpesvirus after gene transfer with a small cloned subfragment of heterogeneous viral DNA. Proc. Natl. Acad. Sci. USA 82, 4085–4089. Countryman, J., Gradoville, L., Heston, L., Ye, J., and Miller, G. (2006). Evidence for a second activity of HDAC inhibitors required for activation of EBV lytic cycle (in preparation). Decaussin, G., Leclerc, V., and Ooka, T. (1995). The lytic cycle of Epstein‐Barr virus in the nonproducer Raji line can be rescued by the expression of a 135‐kilodalton protein encoded by the BALF2 open reading frame. J. Virol. 69(11), 7309–7314. Delecluse, H. J., Pich, D., Hilsendegen, T., Baum, C., and Hammerschmidt, W. (1999). A first‐ generation packaging cell line for Epstein‐Barr virus‐derived vectors. Proc. Natl. Acad. Sci. USA 96(9), 5188–5193. El‐Guindy, A. S., and Miller, G. (2004). Phosphorylation of Epstein‐Barr virus ZEBRA protein at its casein kinase 2 sites mediates its ability to repress activation of a viral lytic cycle late gene by Rta. J. Virol. 78(14), 7634–7644. El‐Guindy, A. S., Paek, S. Y., Countryman, J., and Miller, G. (2006). Identification of constitutive phosphorylation sites on the Epstein‐Barr virus ZEBRA protein. J. Biol. Chem. 281(6), 3085–3095. Epstein, M. A., Achong, B. G., Barr, Y. M., Zajac, B., Henle, G., and Henle, W. (1966). Morphological and virological investigations on cultured Burkitt tumor lymphoblasts (strain Raji). J. Natl. Cancer Inst. 37(4), 547–559. Feederle, R., Kost, M., Baumann, M., Janz, A., Drouet, E., Hammerschmidt, W., and Delecluse, H. J. (2000). The Epstein‐Barr virus lytic program is controlled by the co‐operative functions of two transactivators. EMBO J. 19(12), 3080–3089. Fischer, D. K., Robert, M. F., Shedd, D., Summers, W. P., Robinson, J. E., Wolak, J., Stefano, J. E., and Miller, G. (1984). Identification of Epstein‐Barr nuclear antigen polypeptide in mouse and monkey cells after gene transfer with a cloned 2.9‐kilobase‐pair subfragment of the genome. Proc. Natl. Acad. Sci. USA 81(1), 43–47. Fixman, E. D., Hayward, G. S., and Hayward, S. D. (1992). Trans‐acting requirements for replication of Epstein‐Barr virus ori‐Lyt. J. Virol. 66(8), 5030–5039. Flemington, E., and Speck, S. H. (1990). Evidence for coiled‐coil dimer formation by an Epstein‐Barr virus transactivator that lacks a heptad repeat of leucine residues. Proc. Natl. Acad. Sci. USA 87(23), 9459–9463. Francis, A., Ragoczy, T., Gradoville, L., El‐Guindy, A., and Miller, G. (1999). Amino acid substitutions reveal distinct functions of serine 186 of the ZEBRA protein in activation of lytic cycle genes and synergy with the EBV Rta transactivator. J. Virol. 73(6), 4543–4551. Frenkel, N., Silverstein, S., Cassai, E., and Roizman, B. (1973). RNA synthesis in cells infected with herpes simplex virus. VII. Control of transcription and of transcript abundancies of unique and common sequences of herpes simplex virus 1 and 2. J. Virol. 11(6), 886–892. Furnari, F. B., Adams, M. D., and Pagano, J. S. (1992). Regulation of the Epstein‐Barr virus DNA polymerase gene. J. Virol. 66(5), 2837–2845. Gao, Z., Krithivas, A., Finan, J. E., Semmes, O. J., Zhou, S., Wang, Y., and Hayward, S. D. (1998). The Epstein‐Barr virus lytic transactivator Zta interacts with the helicase‐primase replication proteins. J. Virol. 72(11), 8559–8567.

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

107

Gradoville, L., Gerlach, J., Grogan, E., Shedd, D., Nikiforow, S., Metroka, C., and Miller, G. (2000). Kaposi’s sarcoma‐associated herpesvirus open reading frame 50/Rta protein activates the entire viral lytic cycle in the HH‐B2 primary effusion lymphoma cell line. J. Virol. 74(13), 6207–6212. Gradoville, L., Kwa, D., El‐Guindy, A., and Miller, G. (2002). Protein kinase C‐independent activation of the Epstein‐Barr virus lytic cycle. J. Virol. 76(11), 5612–5626. Grogan, E. J., Jenson, J., Countryman, J., Heston, L., Gradoville, L., and Miller, G. (1987). Transfection of a rearranged viral DNA fragment WZhet, stably converts latent Epstein‐Barr virus infection to productive infection in lymphoid cells. Proc. Natl. Acad. Sci. USA 84, 1332–1336. Henle, G., and Henle, W. (1966). Immunofluorescence in cells derived from Burkitt’s lymphoma. J. Bacteriol. 91(3), 1248–1256. Heston, L., El‐Guindy, A., Countryman, J., DelaCruz, C., Delecluse, H. J., and Miller, G. (2006). Amino acids in the basic domain of Epstein‐Barr virus ZEBRA protein play distinct roles in DNA binding, activation of early lytic gene expression and promotion of viral DNA replication. J. Virol. 80, 9115–9133. Honess, R. W., and Roizman, B. (1974). Regulation of herpesvirus macromolecular synthesis. I. Cascade regulation of the synthesis of three groups of viral proteins. J. Virol. 14(1), 8–19. Kapoor, P., Shire, K., and Frappier, L. (2001). Reconstitution of Epstein‐Barr virus‐based plasmid partitioning in budding yeast. EMBO J. 20(1–2), 222–230. Kolman, J. L., Taylor, N., Gradoville, L., Countryman, J., and Miller, G. (1996). Comparing transcriptional activation and autostimulation by ZEBRA and ZEBRA/c‐Fos chimeras. J. Virol. 70, 1493–1504. Kouzarides, T., Packham, G., Cook, A., and Farrell, P. J. (1991). The BZLF1 protein of EBV has a coiled coil dimerisation domain without a heptad leucine repeat but with homology to the C/EBP leucine zipper. Oncogene 6(2), 195–204. Kraus, R. J., Mirocha, S. J., Stephany, H. M., Puchalski, J. R., and Mertz, J. E. (2001). Identification of a novel element involved in regulation of the lytic switch BZLF1 gene promoter of Epstein‐Barr virus. J. Virol. 75(2), 867–877. Kraus, R. J., Perrigoue, J. G., and Mertz, J. E. (2003). ZEB negatively regulates the lytic‐switch BZLF1 gene promoter of Epstein‐Barr virus. J. Virol. 77(1), 199–207. Lehman, A. M., Ellwood, K. B., Middleton, B. E., and Carey, M. (1998). Compensatory energetic relationships between upstream activators and the RNA polymerase II general transcription machinery. J. Biol. Chem. 273(2), 932–939. Liang, Y., Chang, J., Lynch, S. J., Lukac, D. M., and Ganem, D. (2002). The lytic switch protein of KSHV activates gene expression via functional interaction with RBP‐Jkappa (CSL), the target of the Notch signaling pathway. Genes Dev. 16(15), 1977–1989. Lieberman, P. M., Hardwick, J. M., Sample, J., Hayward, G. S., and Hayward, S. D. (1990). The zta transactivator involved in induction of lytic cycle gene expression in Epstein‐Barr virus‐infected lymphocytes binds to both AP‐1 and ZRE sites in target promoter and enhancer regions. J. Virol. 64(3), 1143–1155. Lin, S. F., Robinson, D. R., Miller, G., and Kung, H. J. (1999). Kaposi’s sarcoma‐associated herpesvirus encodes a bZIP protein with homology to BZLF1 of Epstein‐Barr virus. J. Virol. 73(3), 1909–1917. Lu, F., Zhou, J., Wiedmer, A., Madden, K., Yuan, Y., and Lieberman, P. M. (2003). Chromatin remodeling of the Kaposi’s sarcoma‐associated herpesvirus ORF50 promoter correlates with reactivation from latency. J. Virol. 77(21), 11425–11435. Luka, J., Kallin, B., and Klein, G. (1979). Induction of the Epstein‐Barr virus (EBV) cycle in latently infected cells by n‐butyrate. Virology 94(1), 228–231.

108

George Miller et al.

Miller, G., and Lipman, M. (1973). Release of infectious Epstein‐Barr virus by transformed marmoset leukocytes. Proc. Natl. Acad. Sci. USA 70(1), 190–194. Miller, M. H., Stitt, D., and Miller, G. (1970). Epstein‐Barr viral antigen in single cell clones of two human leukocytic lines. J. Virol. 6(5), 699–701. Morrison, T. E., and Kenney, S. C. (2004). BZLF1, an Epstein‐Barr virus immediate‐early protein, induces p65 nuclear translocation while inhibiting p65 transcriptional function. Virology 328(2), 219–232. Morrison, T. E., Mauser, A., Wong, A., Ting, J. P., and Kenney, S. C. (2001). Inhibition of IFN‐ gamma signaling by an Epstein‐Barr virus immediate‐early protein. Immunity 15(5), 787–799. Morrison, T. E., Mauser, A., Klingelhutz, A., and Kenney, S. C. (2004). Epstein‐Barr virus immediate‐early protein BZLF1 inhibits tumor necrosis factor alpha‐induced signaling and apoptosis by downregulating tumor necrosis factor receptor 1. J. Virol. 78(1), 544–549. Nonoyama, M., and Pagano, J. S. (1972). Separation of Epstein‐Barr virus DNA from large chromosomal DNA in non‐virus‐producing cells. Nat. New Biol. 238(84), 169–171. Petosa, C., Morand, P., Baudin, F., Moulin, M., Artero, J. B., and Muller, C. W. (2006). Structural basis of lytic cycle activation by the Epstein‐Barr virus ZEBRA protein. Mol. Cell 21(4), 565–572. Pope, J. H., Horne, M. K., and Wetters, E. J. (1969). Significance of a complement‐fixing antigen associated with herpes‐like virus and detected in the Raji cell line. Nature 222(189), 186–187. Quinlivan, E. B., Holley‐Guthrie, E. A., Norris, M., Gutsch, D., Bachenheimer, S. L., and Kenney, S. C. (1993). Direct BRLF1 binding is required for cooperative BZLF1/BRLF1 activation of the Epstein‐Barr virus early promoter, BMRF1. [corrected and republished with original paging, article originally printed in Nucleic Acids Res. 1993 April 25; 21(8), 1999–2007]. Nucleic Acids Res. 21(14), 1999–2007. Raab‐Traub, N., and Flynn, K. (1986). The structure of the termini of the Epstein‐Barr virus as a marker of clonal cellular proliferation. Cell 47(6), 883–889. Ragoczy, T., and Miller, G. (1999). Role of the Epstein‐Barr virus Rta protein in activation of distinct classes of viral lytic cycle genes. J. Virol. 73(12), 9858–9866. Ragoczy, T., and Miller, G. (2001). Autostimulation of the Epstein‐Barr virus BRLF1 promoter is mediated through consensus Sp1 and Sp3 binding sites. J. Virol. 75(11), 5240–5251. Reedman, B. M., and Klein, G. (1973). Cellular localization of an Epstein‐Barr virus (EBV)‐ associated complement‐fixing antigen in producer and non‐producer lymphoblastoid cell lines. Int. J. Cancer 11(3), 499–520. Schepers, A., Pich, D., and Hammerschmidt, W. (1993). A transcription factor with homology to the AP‐1 family links RNA transcription and DNA replication in the lytic cycle of Epstein‐ Barr virus. EMBO J. 12, 3921–3929. Schepers, A., Pich, D., and Hammerschmidt, W. (1996). Activation of oriLyt, the lytic origin of DNA replication of Epstein‐Barr virus, by BZLF1. Virology 220(2), 367–376. Summers, W. P., Grogan, E. A., Shedd, D., Robert, M., Liu, C. R., and Miller, G. (1982). Stable expression in mouse cells of nuclear neoantigen after transfer of a 3.4‐megadalton cloned fragment of Epstein‐Barr virus DNA. Proc. Natl. Acad. Sci. USA 79(18), 5688–5692. Sun, R., Lin, S. F., Gradoville, L., Yuan, Y., Zhu, F., and Miller, G. (1998). A viral gene that activates lytic cycle expression of Kaposi’s sarcoma‐associated herpesvirus. Proc. Natl. Acad. Sci. USA 95(18), 10866–10871. Westphal, E. M., Blackstock, W., Feng, W., Israel, B., and Kenney, S. C. (2000). Activation of lytic Epstein‐Barr virus (EBV) infection by radiation and sodium butyrate in vitro and in vivo: A potential method for treating EBV‐positive malignancies. Cancer Res. 60(20), 5781–5788.

Lytic Cycle Switches of Oncogenic Human Gammaherpesviruses

109

Yates, J., Warren, N., Reisman, D., and Sugden, B. (1984). A cis‐acting element from the Epstein‐Barr viral genome that permits stable replication of recombinant plasmids in latently infected cells. Proc. Natl. Acad. Sci. USA 81(12), 3806–3810. Ye, J., Shedd, D., and Miller, G. (2005). An Sp1 response element in the Kaposi’s sarcoma‐ associated herpesvirus open reading frame 50 promoter mediates lytic cycle induction by butyrate. J. Virol. 79(3), 1397–1408. Zhang, C. X., Decaussin, G., Daillie, J., and Ooka, T. (1988). Altered expression of two Epstein‐Barr virus early genes localized in BamHI‐A in nonproducer Raji cells. J. Virol. 62(6), 1862–1869.

This page intentionally left blank

No Life Without Death Peter H. Krammer, Marcin Kamin´ski, Michael Kießling, and Karsten Gu¨low Tumor Immunology Program D030, German Cancer Research Center (DKFZ), Im Neuenheimer Feld 280, 69120 Heidelberg, Germany

I. Introduction II. Apoptosis in Life and Disease III. The Apoptotic Machinery A. Mitochondria and Cell Death: The Intrinsic Pathway B. DR‐Induced Apoptosis: The Extrinsic Pathway IV. The CD95/CD95L System A. Regulation of CD95L Expression in Activation‐Induced Cell Death B. Transcriptional Regulation of CD95L Expression in T Cells C. Regulation of CD95L Expression by Oxidative Signals V. HIV and Apoptosis A. The Genetic Structure of HIV B. HIV Proteins and Apoptosis References

Apoptosis—programed cell death—is the most common form of death in the body. Once apoptosis is induced, proper execution of the cell death program requires the coordinated activation and execution of multiple molecular processes. Here, we describe the pathways and the basic components of the death‐inducing machinery. Since apoptosis is a key regulator of tissue homeostasis, an imbalance of apoptosis results in severe diseases like cancer, autoimmunity, and AIDS. # 2007 Elsevier Inc.

I. INTRODUCTION Cell death is crucial for living. At least two modes of cell death can be distinguished: necrosis and apoptosis (programed cell death) (Kerr et al., 1972). Apoptosis is a strictly regulated process which is essential for development, establishment, and maintenance of tissue architecture. In contrast to necrotic cells, which can elicit an inflammatory reaction, apoptotic cells are removed in an inconspicuous fashion, mainly by phagocytosis by neighboring cells or by specialized macrophage‐like cells. Apoptosis is distinguished by typical morphological and biochemical traits including cell shrinkage, nuclear DNA fragmentation, and membrane blebbing (Hengartner, 2000). A characteristic feature of the classical apoptotic process is the activation of proteases called caspases, which act as key Advances in CANCER RESEARCH Copyright 2007, Elsevier Inc. All rights reserved.

111

0065-230X/07 $35.00 DOI: 10.1016/S0065-230X(06)97005-5

112

Peter H. Krammer et al.

effector molecules (Thornberry and Lazebnik, 1998). The activity of caspases is largely responsible for the observed morphological changes seen in apoptotic cells. The caspase family members are synthesized as zymogens and are commonly activated by proteolytic cleavage. Caspases are characterized by the presence of a cysteine residue in their active center and by the ability to cleave their substrates at aspartate residues. The majority of known caspases has a function in apoptosis and can be subdivided into two classes: initiator caspases (e.g., caspase‐8, ‐9, ‐10) and effector caspases (e.g., caspase‐3, ‐6, ‐7). Activation of initiator caspases leads to proteolytic activation of effector caspases, which in turn results in dismantling of the vital cellular apparatus.

II. APOPTOSIS IN LIFE AND DISEASE In invertebrates, such as Drosophila melanogaster or Caenorhabditis elegans, apoptosis is mainly confined to early live and ends at birth or metamorphosis, whereas vertebrates show remodeling and regeneration of different tissues over their entire lifetime. In humans, several hundred thousand cells are produced every second by mitosis, and an equal number has to die by apoptosis. The physiological mechanism of cell removal is vital for tissue repair in response to injury or insult, to control cell number, and maintain cell homeostasis, for example in the liver or the immune system, and as a strategy of defense to eliminate infected, mutated, or damaged cells (Vaux and Korsmeyer, 1999). The immune system is the place where B and T cells meet their fate. The cells whose receptors do not match the exposed antigen and therefore prove to be insufficient for an immune response, or worse, could attack the own organism, are deleted. In mice for instance, about 5  107 new T cell precursors are generated everyday, but only 2  106 mature T cells will leave the thymus; approximately 95% of developing thymocytes die in the thymus due to negative selection (Surh and Sprent, 1994; Thompson, 1995). Apoptosis also plays an important role as mechanism of cellular homeostasis in the immune system after an immune response. In the periphery, the B and T cells that were produced in high numbers to cope with pathogens, for example microbes and viruses, have to be cleared from the blood system to reinstall the original level of immune cells (Krammer, 2000). The critical role of apoptosis throughout the extensive mammalian live span manifests itself in apoptotic dysfunctions which may cause abnormalities not only in development but also in the pathogenesis of various diseases in adults. Many diseases are associated with either too much or too little cell death. Cancer, for example, is a disease in which apoptosis is impaired.

No Life Without Death

113

Apoptosis represents a line of defense for the organism against cells that have been transformed and that are predisposed to cell proliferation. Their inability to undergo cell death is due to nonfunctional proteins of the intracellular cell death system or to overexpression of proteins which counteract the death signal and promote cell survival. Insufficient cell death shifts the equilibrium of cell homeostasis toward cell proliferation and can therefore lead to cancer (Thompson, 1995). Autoimmune diseases are another case where autoreactive cells fail to commit suicide and pose a potential threat to the organism as seen, for example in Hashimoto’s thyroiditis (Stassi and De Maria, 2002) or Grave’s disease (Feldmann et al., 1992). In contrast, many diseases, as AIDS, neurodegenerative disorders such as Parkinson’s, Alzheimer’s, and retinitis pigmentosa involve an excess of apoptosis (Thompson, 1995). In human immunodeficiency virus (HIV)‐ infected patients, AIDS is characterized by a depletion of T cells that is partly due to massive apoptosis (Badley et al., 2000; Finkel et al., 1995; Gougeon and Montagnier, 1999). Therefore, infected individuals ultimately die of opportunistic infections or cancer due to massive depletion of potential effector T cells. The need for therapeutic agents to either induce or prevent apoptosis is obvious. In the case of AIDS for example, approximately 42 million people worldwide are HIV‐infected. To fight AIDS and the other diseases and disorders mentioned above, an exact understanding and profound knowledge of apoptotic signaling pathways composes an indispensable prerequisite.

III. THE APOPTOTIC MACHINERY Pathways transducing apoptotic signals are distinct. Many environmental stressors are detected within the cell and initiate the signaling cascade leading to apoptosis. Other stimuli act via cell surface receptors. Apoptotic signaling events can thus be roughly divided into two pathways, depending on the mechanism of initiation: the intrinsic pathway, which mainly depends on mitochondrial changes and the extrinsic pathway, which is activated by extracellular signals that act via death receptors (DRs). Although different molecules participate in the core machinery of both apoptosis signaling pathways, a cross talk exists at multiple levels.

A. Mitochondria and Cell Death: The Intrinsic Pathway Precise mechanisms for mitochondria‐induced apoptosis remain to be elucidated. The major role of mitochondria is the release of certain pro‐ apoptotic proteins that were previously localized to the mitochondrial

114

Peter H. Krammer et al.

matrix and intermembrane space. Key regulators of mitochondrial events during initiation of the apoptotic process are proteins of the Bcl‐2 family (Cory and Adams, 2002), cytochrome c (Cyt C) (Liu et al., 1996), and SMAC/DIABLO (Du et al., 2000). In addition, several other less well‐defined factors were reported, among them are AIF (Joza et al., 2001), Endonuclease G (EndoG) (Li et al., 2001; Parrish et al., 2001), and HtrA/Omi (Suzuki et al., 2001). The Bcl‐2 family plays a central role in controlling the mitochondrial pathway. The C. elegans Bcl‐2 homologue CED‐9 is expressed from a bicistronic mRNA that encodes both CED‐9 and cytochrome b, suggesting that Bcl‐2 family proteins may have originated from mitochondria but were transferred to the nuclear genome along with other mitochondrial genes (Hengartner and Horvitz, 1994). The first link between apoptosis and Bcl‐2 was drawn by the discovery that the product of the bcl‐2 gene can suppress apoptosis (Vaux et al., 1988). Further, cell death is promoted in C. elegans when the ced‐9 gene is mutated (Hengartner and Horvitz, 1994). In humans, more than 20 members of the Bcl‐2 family have been identified, including anti‐apoptotic (Bcl‐2, Bcl‐XL, Mcl‐1, Bfl‐1/A1, Bcl‐W, Bcl‐G) and pro‐apoptotic proteins (Bax, Bak, Bok/Mtd, Bad, Bid, Bik, Bim, Hrk, Puma/Bbc3, Noxa, EGL‐1, Bcl‐B) (Cory and Adams, 2002). Bcl‐2 proteins are localized at, or translocated to the mitochondrial membrane. They are thought to modulate apoptosis by permeabilization of the inner and/or outer membrane, leading to release of Cyt C and other pro‐apoptotic proteins, or by stabilizing mitochondrial barrier function, respectively. Cyt C, an important component of the mitochondrial respiratory chain, is released into the cytosol where it causes ATP‐dependent oligomerization of the adaptor Apaf‐1 and recruitment of procaspase‐9. This complex is called the apoptosome and activates further downstream caspases such as caspase‐3 (Lorenzo and Susin, 2004). Smac/DIABLO (second mitochondria‐derived activator of caspases/direct IAP‐binding protein with low isoelectric point), is another protein released from the mitochondria on induction of apoptosis that antagonizes IAPs (caspase inhibiting proteins) and therefore, promotes apoptosis (Du et al., 2000). In addition, mitochondria release a number of other pro‐apoptotic proteins, whose function is not yet clearly defined. One of these factors is apoptosis‐inducing factor (AIF), whose role is controversial. AIF was reported to translocate into the nucleus, triggering chromatin condensation and large‐scale DNA fragmentation (Klein et al., 2002; Susin et al., 1999). It is also proposed to participate in the regulation of apoptotic mitochondrial membrane permeabilization (Joza et al., 2001) and may trigger cell death through reactive oxygen species production

No Life Without Death

115

(Lipton and Bossy‐Wetzel, 2002). Another player in the orchestra of mitochondrial effectors in programed cell death is EndoG. This sequence‐ unspecific DNase normally functions in replication of mitochondrial DNA, but on apoptotic stimuli, it translocates from mitochondria to the nucleus, where it degrades nuclear DNA into oligosomal fragments, similar to caspase‐activated DNase (CAD) (Li et al., 2001; Parrish et al., 2001). In contrast to CAD, activation of EndoG is caspase independent. Omi/HtrA2 is a mammalian serine protease that shares homology with the bacterial endoprotease HtrA (Faccio et al., 2000). In mammals, Omi/HtrA2 is an ubiquitously expressed protein with dual function. On release from mitochondria, Omi/HtrA2 is able to bind to IAPs and disrupt their binding to caspases. In addition, it can also induce apoptosis in a caspase‐independent manner that seems to depend on its own serine protease activity (Lorenzo and Susin, 2004; Suzuki et al., 2001).

B. DR‐Induced Apoptosis: The Extrinsic Pathway A well‐elaborated mechanism to regulate life and death of cells is the interaction of surface receptors with their cognate ligands, thereby transmitting extracellular signals into the cytosol. Several DRs have been described that all belong to the tumor necrosis factor (TNF) receptor superfamily, defined by cysteine‐rich extracellular domains (Schulze‐Osthoff et al., 1998; Smith et al., 1994). In addition, DRs contain a so‐called “death domain” (DD), a cytosolic 70‐amino acid sequence, both necessary and sufficient for induction of apoptosis. DRs reported so far are TNF‐R1 (CD120a), CD95 (APO‐1, Fas), DR3 (APO‐3, LARD, TRAMP, WSL1), TRAIL‐R1 (APO‐2, DR4), TRAIL‐R2 (DR5, KILLER, TRICK2), and DR6 (Ashkenazi and Dixit, 1998; Krueger et al., 2003; Schulze‐Osthoff et al., 1998). DRs are identified as type I transmembrane proteins whereas natural ligands for these receptors are type II transmembrane proteins. Ligands belong to the TNF gene superfamily (Smith et al., 1994). Several ligands have been identified; for example CD95 ligand (CD95L, FasL) binds to CD95, TNF binds to TNFR1, APO‐3 ligand (APO‐3L or TWEAK) binds to DR3, and APO‐2 ligand (APO‐2L or TRAIL) binds to DR4 and DR5 (Ashkenazi and Dixit, 1998; Krueger et al., 2003) (Fig. 1). Soluble forms of these ligands have been reported to be generated by cleavage of the C‐terminal portion by specific metalloproteinases (Krueger et al., 2003). The crucial point of DR signaling is the formation of a multimolecular complex of proteins triggered by receptor cross‐linking either with agonistic antibodies or with death ligands. The complex that is formed is called the death‐ inducing signaling complex (DISC) (Walczak and Sprick, 2001).

Peter H. Krammer et al.

116 CD95L

CD95

DcR3

TNF

APO-3L/TWEAK

TNFR1

DR3

APO-2L/TRAIL

TRAILR1 TRAILR2 DcR1 Death receptor

DcR2

OPG

Cystein-rich motif Death domain Transmembrane domain

Fig. 1 Death receptors are type I transmembrane proteins and harbor cysteine‐rich extracellular repeats. Their common feature is a cytoplasmic protein interaction domain, the death domain (DD), of approximately 70 amino acids. The DD is necessary for induction of apoptosis. The natural ligands of death receptors belong to the TNF‐family. TNF , CD95L, and TRAIL are most frequently involved in apoptotic signaling and are synthesized as type II transmembrane proteins. Yellow squares: cysteine‐rich motifs and red: death domains.

IV. THE CD95/CD95L SYSTEM CD95 is a glycosylated type I transmembrane cell surface receptor of a relative molecular mass of approximately 45–52 kDa (consisting of 335‐amino acid residues). CD95 expression can be enhanced in T cells by cytokines such as interferon‐ (IFN ) and TNF (Krammer, 2000). Resting B cells express low levels of CD95 on induction by CD40L and endotoxins (Watanabe et al., 1995). The CD95L, a 40‐kDa glycoprotein is only present on few cell types, such as activated T cells (Suda et al., 1993) or natural killer (NK) cells (Oshimi et al., 1996), where it participates in cell‐mediated cytotoxicity. Its presence on cells of immune‐privileged sites, such as the testes, placenta, anterior chamber of the eye, and brain, has also been reported (French et al., 1996; Lee et al., 1997; Mitsiades et al., 2003). Presence of CD95L in these tissues may contribute to elimination of infiltrating lymphocytes. The physiological importance of CD95 and CD95L is underlined by mutations found in these genes leading to diseases in mice and humans. Mice carrying the homozygous lpr/lpr (lymphoproliferation) allele

No Life Without Death

117

(a mutation in the CD95 gene) show symptoms that are highly similar to the generalized lymphoproliferative disease (gld) mutation (a mutation in the CD95L gene). These symptoms include autoimmune disorder syndromes such as accumulation of peripheral lymphocytes, massive enlargement of lymph nodes (lymphadenopathy) and spleen (splenomegaly) (Takahashi et al., 1994; Watanabe‐Fukunaga et al., 1992). In humans, mutations in CD95 are relevant for an autoimmune disorder termed ALPS Ia (autoimmune lymphopoliferative syndrome). ALPS Ib is associated with mutations in the CD95L (Rieux‐Laucat et al., 2003). ALPS patients typically present childhood autoimmunity (Canal–Smith syndrome), accumulation of lymphocytes, and an increased risk of developing lymphomas. Most TNF receptor‐like molecules require trimerization for the recruitment of signaling molecules and activation of the signaling cascade. In addition, the CD95 receptor was also suggested to exist in a trimeric (inactive) form preassembled via preligand‐binding assembly domains (PLAD) (Siegel et al., 2000). In this case, signaling is induced either by conformational changes of preformed DR trimers or alternatively, by the formation of multimeric complexes on ligand binding (Krueger et al., 2003). Furthermore, a refined but still intensively discussed model of proximal steps of CD95 signaling was proposed, involving (1) formation of CD95 microaggregates, (2) DISC formation, (3) formation of large CD95 surface clusters, and (4) internalization of activated CD95 (Algeciras‐Schimnich et al., 2002). Transduction of the apoptotic signal starts with the formation of the DISC within seconds after receptor engagement. The CD95 DISC consists of oligomerized CD95 receptors, the adapter protein Fas‐associated DD (FADD, also known as MORT‐1), two isoforms of procaspase‐8 [procaspase‐8/a (FADD‐like interleukin‐1 (IL‐1 )‐converting enzyme (FLICE), Mach‐ 1, Mch5 ) and procaspase‐8/b (Mach‐ 2)] (Muzio et al., 1996), procaspase‐10, and c‐FLIPL/S/R. The crucial factor for DISC formation is FADD, which bridges the receptors with caspases by homotypic interactions of the DD and death effector domain (DED). FADD binds via its DD to the DD of the receptor, and via its DED it recruits DED‐containing procaspases‐8 and ‐10. Thereafter, procaspase‐8 is cleaved at the DISC, which leads to formation of active caspase‐8, a major constituent in the pathway. The active form of caspase‐8 is a heterotetramer of two small (p10) subunits and two large (p18) subunits (Medema et al., 1997; Muzio et al., 1998). The prodomain of caspase‐8 remains bound to the DISC, while activated caspase‐8 dissociates from the DISC to initiate further caspase activation and the execution of apoptosis (Medema et al., 1997). The major regulator of CD95‐mediated apoptosis at the DISC level is the cellular FLICE inhibitory protein (c‐FLIP, also called FLAME‐1, I‐FLICE, Casper, CASH, MRIT, CLARP, and Ursupin) (Goltsev et al., 1997; Han et al., 1997; Hu et al., 1997; Inohara et al., 1997; Irmler et al., 1997;

118

Peter H. Krammer et al.

Krueger et al., 2001; Rasper et al., 1998; Shu et al., 1997; Srinivasula et al., 1997). Multiple human splice isoforms exist of which three were identified at the protein level: c‐FLIPS (short), c‐FLIPL (long), and c‐FLIPR (Raji) (Golks et al., 2005). c‐FLIPS is anti‐apoptotic and blocks caspase‐8 processing and activation at the DISC and CD95‐induced cell death. c‐FLIPR was reported to be similar to c‐FLIPS in being recruited to the DISC after CD95 stimulation where it reveals anti‐apoptotic effects (Golks et al., 2005). The role of c‐FLIPL is ambiguous, described as pro‐ or anti‐apoptotic, depending on its expression levels and experimental conditions (Chang et al., 2002; Krueger et al., 2001). Two pathways of CD95 apoptosis signaling, based on the quantity of production of active caspase‐8 at the DISC were described (Scaffidi et al., 1998). In type I cells, a high production of caspase‐8 at the DISC leads to direct processing and activation of the effector caspase, caspase‐3, and thus leads to apoptosis. In type II cells, however, only a small amount of caspase‐8 is produced at the DISC. The DISC in type II cells is formed quite poorly and subsequently active caspase‐8 is generated in lower amounts. Apoptosis in these cells is dependent, at least in part, on the cleavage of the Bcl‐2 family member Bid (Li et al., 1998; Luo et al., 1998). This cleavage results in a pro‐ apoptotic fragment termed truncated Bid (tBid). This fragment induces pro‐apoptotic functions of mitochondria by causing aggregation of Bax and Bak (Korsmeyer et al., 2000) and subsequent release of Cyt C from the mitochondrial intermembrane space. Apaf‐1 and Cyt C form a large protein complex, the apoptosome, at which caspase‐9 is activated as initiator caspase (Li et al., 1997). Induction of apoptosis is accomplished by activation of effector caspases like caspase‐3. CD95‐mediated apoptosis in type II cells is further affected by the expression of antiapoptotic members of the Bcl‐2 family. Expression of either Bcl‐2 or Bcl‐xL renders type II cells resistant to CD95‐mediated apoptosis (Scaffidi et al., 1998). In contrast, type I cells are not protected from CD95‐mediated apoptosis even by the expression of very high levels of Bcl‐2 or Bcl‐xL (Scaffidi et al., 1998).

A. Regulation of CD95L Expression in Activation‐Induced Cell Death Apoptosis, which occurs after activation of peripheral T cells, is of crucial importance for termination of the acquired immune response, as well as for maintenance of T cell homeostasis (Krueger et al., 2003). A typical immune response to foreign antigen is characterized by a multistep process. First, T cells are activated on encounter of antigen presented by an antigen‐presenting cell (APC), for example in the lymph nodes or the spleen.

No Life Without Death

119

The activation is mediated by T cell receptor (TCR) triggering on cross‐ linking with antigenic peptides bound to major histocompatibility complex (MHC) molecules. Then, clonal expansion proceeds, during which activated T cells rapidly proliferate and differentiate into effector cells. Subsequently, the majority of activated T cells enter the so‐called deletion phase and are eliminated. However, certain antigen‐specific T cells may survive, forming a pool of memory T cells (Krueger et al., 2003). The apoptotic process of elimination of activated T cells during the termination phase of an immune response is called activation‐induced cell death (AICD). The pro‐apoptotic stimuli playing a role in AICD are engagement of DRs and their ligands, mainly CD95/CD95L (Krammer, 2000). CD95L is not expressed by naı¨ve, resting T cells. Its expression is triggered by signaling events occurring after T cell receptor triggering (Krueger et al., 2003). The TCR is composed out of six different polypeptide chains. The specificity of peptide binding is determined by TCR and TCR chains (or and , alternatively), which arise from a process of genetic rearrangements. Communication between TCR , bound to the peptide–MHC complex, and the intracellular signaling machinery occurs via the TCR‐associated CD3 chains. Each CD3 chain contains immunoreceptor tyrosine‐based activation motives (ITAMS), which become rapidly phosphorylated by the Src‐family protein tyrosine kinase (PTK) Lck after TCR stimulation (Malissen, 2003). Once fully phosphorylated, these motives serve as binding sites for the SYC‐family PTK ZAP70. ZAP70 phosphorylates a restricted set of substrates, which includes Shc, SLP76, and a main mediator of signaling—LAT (Clements, 2003; Finco et al., 1998; Jackman et al., 1995; Zhang et al., 1998). On phosphorylation, both LAT and SLP76 act as adapter proteins and recruit several Src‐homology 2 (SH2) domain‐containing proteins, including phospholipase C 1 (PLC 1) into lipid rafts (Clements, 2003). Activation of PLC 1 leads to hydrolysis of phosphatidylinositol‐ 4,5‐diphosphate [(PI)‐4,5‐P2], which results in generation of diacylglycerol (DAG) and inositol‐1,4,5‐triphosphate (IP3). IP3 production leads to increase of cytosolic calcium (Ca2þ), whereas DAG can activate both protein kinase C (PKC) and Ras guanyl nucleotide‐releasing protein (RasGRP) (Wange and Huang, 2004). The increase in cytosolic Ca2þ causes activation of calcineurin (Williamson, 1986), which dephosphorylates the nuclear factor of activated T cells (NF‐AT). On translocation into the nucleus, activated NF‐AT initiates gene transcription and therefore, it is regarded as one of the key participants in CD95L regulation (Li‐Weber et al., 1999). The  isoform of PKC has been shown to be essential for activation‐induced CD95L expression (Villalba et al., 1999; Villunger et al., 1999). In addition, activated RasGRP leads to Ras‐activated cascade of kinase activity including Raf, Mek, Erk, and p38 mitogen‐activated thyrosine kinase (MAPK),

Peter H. Krammer et al.

120

an important pathway for optimal activation‐induced CD95L expression (Hsu et al., 1999; Latinis et al., 1997).

B. Transcriptional Regulation of CD95L Expression in T Cells Like many other genes, the CD95L gene is under control of a large array of cis‐acting promoter elements which act in concert to achieve a fine degree of control over transcriptional activity of the gene. A series of promoter‐ regulated elements located within a 960 bp fragment extending from position 860 to þ100 of the human CD95L promoter have been identified (Li‐Weber and Krammer, 2003) (Fig. 2).

1. NUCLEAR FACTOR OF ACTIVATED T CELLS TCR‐stimulated increase in Ca2þ leads to activation of the Ca2þ‐calmodulin‐ regulated phosphatase, calcineurin, which in turn dephosphorylates the transcription factor NF‐AT. Removal of the phosphate groups reveals a nuclear localization signal and results in translocation of transcriptionally active NF‐ AT to the nucleus (Hogan et al., 2003). Four NF‐AT‐binding sites have been found in the CD95L promoter, three major (680, 180, and 120) and one site of minor effect for promoter induction (–30; located next to the TATA‐box) (Li‐Weber and Krammer, 2003).

2. IMMEDIATE EARLY RESPONSE PROTEIN (Egr) The Egr proteins, Egr‐1, ‐2, ‐3, and ‐4, are closely related members of a subclass of immediate early gene‐encoded zinc finger transcription factors and are inducibly expressed in response to diverse stimuli. The first

−180

−120

−65 −50

−30 +1 TATA

+65

+90

2 r/S NF P-1 Eg -AT r/S IR NF P-1 F- -A 1/ T vI RF IR F1/ NF vIR F -κ B/ SP -1 cM N F yc -A T IR F1/ vI AP RF -1 (J un /F os )

−230

Eg

n/

AT F

−530

Ju

−630

N Eg F-A r/S T P1 NF -κ B/ SP -1

N FκB

−980

Fig. 2 The CD95L gene is under control of a large array of other cis‐acting promoter elements (AP‐1, NF‐AT, NF‐B, ATF2, c‐Myc, INF‐regulatory factors [IRF], SP‐1, and Egr), which act in concert to achieve a fine degree of control over the transcriptional activity of the gene.

No Life Without Death

121

Egr‐interacting site was identified at the 120 region of the CD95L promoter. Later, it was found that all three major NF‐AT sites (680, 180, and 120) are combined Egr and NF‐AT sites. NF‐AT and Egr at these three sites were shown to display a cooperative and synergistic activation of the CD95L promoter. Among the Egr factors, Egr‐2 and Egr‐3 were proposed to confer most to activation‐induced CD95L expression (Li‐Weber and Krammer, 2003).

3. NUCLEAR FACTOR KAPPA B Nuclear factor kappa B (NF‐B) is a collective name given to dimeric transcription factors of the Rel family (Rothwarf and Karin, 1999). In absence of activating signals, tight binding to the inhibitory IB protein retains NF‐B in the cytoplasm. Stimulation activates the IB kinase (IKK) complex, which phosphorylates IB and thereby targets it for ubiquitinilation and degradation. The degradation of IB unmasks the nuclear localization sequence and allows the translocation of NF‐B into the nucleus (Rothwarf and Karin, 1999; Wange and Huang, 2004). Several studies have shown that high expression of CD95L on TCR stimulation depends on NF‐B. Two NF‐B‐binding sites are localized at the 530 and 50 regions of the human CD95L promoter. Mutation of these sites results in 30–70% reduction of activation‐induced CD95L expression (Li‐Weber and Krammer, 2003). An additional NF‐B‐binding site was described at 980 region of the CD95L promoter. However, conflicting results have been obtained from mutation and deletion studies of the 980 binding site (Li‐Weber and Krammer, 2003).

4. ACTIVATOR PROTEIN‐1 The activator protein‐1 (AP‐1) transcription factor is composed of dimers of c‐Jun and c‐Fos family proteins and can be activated both by phosphorylation of c‐Jun by JNK or by up‐regulation of c‐Jun and c‐Fos expression (Wange and Huang, 2004). Pathways that enhance activation‐induced CD95L expression including Ras/Raf/Mek/Erk have also been described (Wange and Huang, 2004). An important aspect of AP‐1 function is the ability to form complexes with NF‐AT and NF‐B transcription factors. In the promoter region of the CD95L gene, a typical AP‐1 site was identified at position þ90. A second AP‐1 responsive element was identified at position 230. However, this element is not homologous to a consensus AP‐1 site and binds to c‐Jun and ATF‐2 but not to c‐Fos. A distal AP‐1‐binding site at the 950 region was shown to respond to DNA‐damaging‐agent‐induced CD95L expression. However, it is not required for activation‐induced CD95L expression (Li‐Weber and Krammer, 2003).

122

Peter H. Krammer et al.

5. c‐Myc AND Max The transcription factor, c‐Myc, a helix‐loop‐helix basic leucine zipper protein, is induced by growth stimulatory signals and other stimuli and has been shown to play an important role in regulation of proliferation, differentiation, and apoptosis (Grandori et al., 2000). Binding of c‐Myc to Max switches on c‐Myc’s DNA‐binding activity. Several studies show c‐Myc is involved in AICD via up‐regulation of CD95L expression. A putative noncanonical‐binding site was detected at position 27. However, this site is exactly located in the TATA‐box region normally required for binding of TBP/TFIID and RNA polymerase II (Featherstone, 2002). Thus, whether it is a physiological c‐Myc‐binding site needs to be clarified (Li‐Weber and Krammer, 2003).

6. Nur77 Nur77 (also called NGFI‐B, N10, Nak1, TR3) is a member of the nuclear orphan steroid receptor superfamily. It was identified as an immediate early serum‐induced gene and subsequently shown to be up‐regulated by nerve growth factor, ionomycin, and TCR/CD3 signaling. It has been shown that overexpression of a dominant‐negative Nur77 protein can inhibit TCR‐ mediated AICD (Liu et al., 1994; Woronicz et al., 1994). In contrast, peripheral T cells and thymocytes of Nur77 knockout mice still show AICD after TCR/CD3 ligation (Lee et al., 1995). However, so far, a recognizable Nur77‐binding site within the CD95L promoter has not been found (Li‐Weber and Krammer, 2003).

7. IFN‐REGULATORY FACTORS (IRFS) AND VIRAL IRFS IRF‐1 is expressed on IFN stimulation in T cells. IRF‐1 was shown to strongly reduce PMA/ionomycin‐induced CD95L expression (Chow et al., 2000). Three putative IRF‐1‐binding sites (at position 120, 65, and þ65) have been found in the CD95L promoter (Chow et al., 2000; Kirchhoff et al., 2002; Li‐Weber and Krammer, 2003).

8. SP‐1 SP‐1 is a ubiquitously expressed transcription factor which preferentially interacts with GC‐rich DNA sequences. All NF‐B and Egr/NF‐AT composite sites contain GC‐rich sequences and were found to interact with SP‐1. SP‐1 has been implicated in maintaining the basal level of CD95L in T cells and in constitutive expression of CD95L in other tissues (Kavurma et al., 2001; McClure et al., 1999).

No Life Without Death

123

9. ADDITIONAL TRANSCRIPTION FACTORS Other transcription factors like apoptosis‐linked gene 4 (ALG‐4) and ROR t were also reported to regulate CD95L expression despite the absence of binding sites within the CD95L promoter (He et al., 1998; Lacana and D’Adamio, 1999). The class II transactivator (CIITA) is a transcription factor regulating multiple genes in antigen presentation. It was shown that expression of CIITA in T cells activates MHC class II but inhibits CD95L expression and prevents TCR/CD3‐induced AICD. Repression of CD95L transcription by CIITA was proposed to be due to competition between CIITA and NF‐AT for binding to the common cofactor CBP/p300 (Gourley and Chang, 2001).

C. Regulation of CD95L Expression by Oxidative Signals Besides the previously assumed cell‐damaging role of reactive oxygen species (ROS), more and more data indicate a role for ROS as a second messenger for signal transduction and amplification (Reth, 2002). There are several potential sources of ROS production inside the cell. Nevertheless, the reaction always starts with the transfer of a single electron to molecular oxygen (O2). This one electron reduction results in generation of superoxide anion (O2) that, in contact with protons in water, is rapidly converted to H2O2 and O2 (2O2 þ 2Hþ $ H2O2 þ O2). The conversion takes place either spontaneously or is catalyzed by superoxide dismutases (SOD). The half‐life of H2O2 is critically dependent on the redox equilibrium inside the cell. The cytosol has a strong reducing capacity. This is due to redox regulators, for example catalase and cellular reductants, for example glutathione (GSH). H2O2 oxidizes GSH into glutathione disulfide (GSSG). In addition, catalase can convert H2O2 into water (2H2O2 $ 2H2O þ O2) (Reth, 2002).

1. H2O2 AS SECOND MESSENGER H2O2 shares several features of a second messenger. It is a small electrically neutral molecule that can diffuse locally inside the cell. It is rapidly generated on extracellular stimuli and can be easily removed by numerous mechanisms. Compared to other short‐lived ROS molecules—for example O2 (half‐life 1 ms)—H2O2 is more stable (half‐live 1 ms) (Reth, 2002). H2O2 oxidizes the thiol group (–SH) of cysteines to sulfenic acid (–SO2H), which is readily reduced to cysteine by cellular‐reducing agents, GSH and thioredoxin (Trx). However, not each cysteine is a good target for the oxidation by H2O2. The cysteine must exist in a deprotonated form which

124

Peter H. Krammer et al.

is the case when it is located in the vicinity of positively charged amino acids (Reth, 2002).

2. GENERATION OF ROS An inducible source of ROS production in leukocytes is the NADPH oxidase (respiratory burst oxidase) (Segal and Shatwell, 1997). This plasma membrane‐associated enzyme is best studied in phagocytes, but it is also found in other cell lineages. NADPH oxidase, a membrane‐associated enzyme, is constituted of several components: the membrane‐bound flavocytochrome b558 (a heterodimer containing gp91phox and p22phox) and the four cytosolic proteins p47phox, p67phox, p40phox, and the small G protein Rac (Reth, 2002). The NADPH oxidase transfers electrons across the plasma membrane to extracellular O2. Thus, O2 is generated in the extracellular space. The O2 anion is membrane impermeable but it can be converted rapidly into H2O2. H2O2 can cross membranes and can therefore reenter the cell (Reth, 2002). However, most estimations suggest that the majority of intracellular ROS is derived from mitochondria as a by‐product of oxidative phosphorylation, so‐called mitochondrial leakage (Turrens et al., 1982). The production of mitochondrial superoxide radicals mainly takes place at two different points in the electron transport chain, at complex I (NADH dehydrogenase) and at complex III (ubiquinone‐Cyt C reductase). Recently, an alternative ROS‐generating pathway involving the respiratory chain has been described. Electrons are transferred from Cyt C to p66shc a redox enzyme, located in the mitochondrial intramembrane space that generates H2O2 directly (Giorgio et al., 2005). Other ROS producers in aerobic organisms are cytochrome P450 at ER membranes (Rashba‐Step et al., 1993), the xanthine oxidase enzymes in the cytoplasm (Fridovich, 1970), phospholipases (Robinson et al., 1998), and peroxisomes (Sies, 1977).

3. TARGETS OF ROS AND CD95L EXPRESSION NF‐B is activated by many stimuli and it was the first eukaryotic transcription factor shown to respond directly to oxidative stress in certain cell types (Schreck et al., 1991). H2O2 activates NF‐B at micromolar concentrations in human breast MCF‐7 and 70 Z/3 pre‐B cells by causing the release of the inhibitory subunit IB from the NF‐B complex (Schreck et al., 1991). However, NF‐B is not strictly dependent on ROS, and H2O2 does not induce NF‐B activation in all cell types. Nevertheless, in T cells, NF‐B activation is suppressed by antioxidants (Israel et al., 1992), suggesting the involvement of ROS in NF‐B activation. In addition,

No Life Without Death

125

a connection between H2O2‐mediated NF‐B activation and increased CD95L expression could be established. A study showed that antioxidants reduce the NF‐B DNA‐binding activity and therefore decrease CD95L transcription (Bauer et al., 1998). Many different oxidative stress‐inducing stimuli, for example low concentrations of hydrogen peroxide, UV light, ‐irradiation, and interleukin‐1, result in AP‐1 activation by two different mechanisms, JNK activation or increased c‐Jun and c‐Fos transcription. In Jurkat T cells, mRNA transcription of c‐Jun and c‐Fos are enhanced after treatment with H2O2, in the absence of any mitogenic stimulus (Beiqing et al., 1996). A mild oxidative shift of the redox state by different oxidants, such as H2O2 or 1,3‐bis(2‐ choloethyl)‐1‐nitrosourea (BCNU), leads to a strong increase in activation of the family of the mitogen‐activated protein kinases JNK that were found to be important conductors in oxidative stress signaling; in addition, other members of the MAPK superfamily such as ERK1 and ERK2 were found to be activated by high hydrogen peroxide concentrations (millimolar range) in Jurkat T cells (Griffith et al., 1998). Since AP‐1 is involved in CD95L expression, both NF‐B and AP‐1 may be targets of ROS‐regulated CD95L expression.

4. T CELL STIMULATION, ROS, AND AICD First suggestions for a role of ROS in T cell activation came from an analysis investigating the role of antioxidants on primary T cell activation by mitogens, antibodies to the TCR, or antigens (Chaudhri et al., 1988; Novogrodsky et al., 1982). Extending these observations to allo‐antigen‐ stimulated cultures suggested that antigen‐mediated T cell activation requires ROS production (Chaudhri et al., 1986). Later, ROS production was shown directly by application of oxidation‐sensitive fluorescent molecules. Nowadays, it is a widely accepted fact that T cells produce ROS on treatment with mitogens (Orie et al., 1999; Sekkat et al., 1988; Tatla et al., 1999; Williams and Henkart, 1996), cognate antigen (Matsue et al., 2003), superantigens (Hildeman et al., 1999), and antibodies to the TCR complex (Devadas et al., 2002; Gulow et al., 2005; Jackson et al., 2004; Kwon et al., 2003; Lahdenpohja and Hurme, 1998; Los et al., 1995; Williams and Henkart, 1996). Restimulation of activated T cells induces AICD. It has been demonstrated that multiple chemical antioxidants inhibit not only proliferation but also AICD when added to restimulated T cells (Bauer et al., 1998; Sandstrom et al., 1993, 1994). Studies have extended these observations and have suggested that control of AICD by ROS occurs via regulation of CD95L expression (Bauer et al., 1998; Devadas et al., 2002; Gulow et al.,

Peter H. Krammer et al.

126

A

B

C

TCR

TCR

CD4

Tat

Tat Ca2+

H2O2

Ca2+

CD95L

H2O2

CD95L

Ca2+

H2O2

CD95L

Fig. 3 H2O2 and Ca2þ signaling. (A) TCR triggering induces H2O2 and Ca2þ signals and CD95L expression. (B) HIV‐Tat enhances the oxidative signal induced by TCR triggering. (C) CD4 triggering induces a Ca2þ signal and Tat an H2O2 signal. Both signals together are sufficient to induce CD95L expression.

2005; Kwon et al., 2003). We have shown that H2O2 functions as an essential second messenger after TCR stimulation. It is an H2O2 signal combined with a simultaneous calcium influx into the cytosol that constitutes the minimal requirement for induction of CD95L expression. Either signal alone has shown to be insufficient to induce CD95L expression (Gulow et al., 2005). Thus, H2O2 acts as a second messenger regulating the induction of AICD (Fig. 3A). Interestingly, the transcriptional transactivator (Tat) of the HIV interferes with activation‐induced ROS production and, therefore, increases CD95L expression and AICD (Gulow et al., 2005). This observation underlines the fact that in HIV‐infected individuals, T cell depletion occurs at least partially due to massive apoptosis (Badley et al., 2000; Finkel et al., 1995; Gougeon and Montagnier, 1999).

V. HIV AND APOPTOSIS Infection by HIV leads to a complex disease characterized by a variety of clinical symptoms such as deregulation of the immune system ranging from profound T cell depletion to autoimmunity, opportunistic infections, systemic inflammatory responses, energy deficit, dementia, and increased incidence of cancers (Rosenberg and Fauci, 1989).

No Life Without Death

127

A. The Genetic Structure of HIV HIV encodes three structural genes, gag, pol, and env, which are common to all replication competent retroviruses. The product of gag is translated from unspliced mRNA as a precursor polyprotein. This precursor is cleaved by the viral protease (PRO) into the subunits: matrix (MAp21), capsid (CAp24), nucleocapsid (NCp7), and several additional polypeptides of small size and unknown function (Roshal et al., 2001). The pol gene also encodes a polyprotein and is expressed as a fusion protein together with Gag on ribosomal frameshifting. The Gag‐Pol protein is also processed by PRO. The cleavage of Pol gives rise to three proteins: PRO, reverse transcriptase (RT), and integrase (IN) (Roshal et al., 2001). The env gene encodes the precursor glycoprotein gp160, which is cleaved into the surface glycoprotein gp120, and a transmembrane glycoprotein, gp41 (Roshal et al., 2001). In addition to these structural genes, HIV encodes six open reading frames. Two of those are regulatory genes, tat and rev, which encode transactivator proteins essential for viral replication. Tat is a transcriptional transactivator and Rev is a posttranscriptional activator allowing nuclear export of unspliced mRNAs encoding viral proteins (Roshal et al., 2001). The remaining four open reading frames are also known as “accessory” genes not essential for viral replication including vif, vpu, and nef (Roshal et al., 2001).

B. HIV Proteins and Apoptosis Many HIV genes are known to be involved in apoptosis, for instance HIV‐Nef. Nef was shown to induce activation of lymphocytes leading to up‐ regulation of CD95L which, in turn, is able to induce apoptosis in CD4þ and CD8þ positive cells (Hodge et al., 1998; Xu et al., 1997, 1999; Zauli et al., 1999). In addition, it has been reported that Nef is also responsible for CD95 up‐regulation. On the contrary, Nef also down‐regulates cell surface expression of MHC class I to protect infected cells from being killed by cytotoxic T‐lymphocytes (Collins et al., 1998; Piguet et al., 1998; Schwartz et al., 1996). HIV Vpr contributes to multiple cytopathic effects induced by HIV by inducing cell cycle arrest in G2 (He et al., 1995; Jowett et al., 1995; Planelles et al., 1996; Re et al., 1995; Rogel et al., 1995) and apoptosis (Chang et al., 2000; Shostak et al., 1999; Stewart et al., 1997; Watanabe et al., 2000). The apoptotic properties have been mapped to the C‐terminal domain of Vpr (Chen et al., 1999; Macreadie et al., 1995). The pathways of Vpr‐induced apoptosis are so far not fully clear. One hypothesis is that cells

128

Peter H. Krammer et al.

die due to abnormal multipolar mitosis induced by aberrant centrosome duplication (Minemoto et al., 1999; Watanabe et al., 2000). Furthermore, it has been reported that in several cell lines Vpr expression leads to a genomic instability including chromosome brakes, micronuclei formation, aneuploidy, and gene amplification (Shimura et al., 1999a,b). Other groups report that the apoptotic action of Vpr is due to interference with mitochondria. It has been shown that Vpr can cause loss of mitochondrial membrane potential and that Vpr is able to interact directly with the PTP complex to increase ion permeability swelling of mitochondria and release of Cyt C (Jacotot et al., 2000). However, it has also been described that Vpr acts as a prosurvival protein up‐regulating Bcl‐2 and down‐regulating Bax (Conti et al., 1998). In order to enter a target cell, HIV has to bind to molecules on the cell surface. Depending on the cell type, the envelope glycoprotein gp120 interacts with its primary receptor CD4 and an accessory protein, typically but not exclusively CXCR4 or CCR5 (Dittmar et al., 1997; Moore, 1997). The binding of gp120 to cell surface receptors can induce signaling events that have been implicated to induce apoptosis in both lymphoid and nonlymphoid cells (Banda et al., 1992; Davis et al., 1997; Neudorf et al., 1990; Weissman et al., 1997). Cross‐linking of bound gp120 on human CD4þ T cells leads to up‐regulation of CD95 (Desbarats et al., 1996; Oyaizu et al., 1994) and CD95L (Oyaizu et al., 1997; Tateyama et al., 2000). In addition, FLIP and Bcl‐2 levels are down‐regulated whereas Bax expression is increased (Hashimoto et al., 1997; Somma et al., 2000). gp120 interaction with CXCR4 has also been implicated in case of increased apoptosis of CD8þ T cells in the presence of macrophages (Herbein et al., 1998). The viral protein Tat up‐regulates viral transcription on the level of elongation via interaction with the Tat activation region (TAR) located at the 50 end of all viral mRNAs. In addition, HIV‐Tat can be secreted by virus‐ infected cells (Ensoli et al., 1990) and can be taken up by bystander cells via endocytosis (Mann and Frankel, 1991). Due to these properties the biological activities of Tat can be exerted in uninfected cells. We and others have previously shown that CD95L mRNA expression in activated T cells was strongly increased in the presence of HIV‐Tat (Katsikis et al., 1995; Li‐Weber et al., 2000; Westendorp et al., 1995a). Enhancement of activation‐induced CD95L expression was shown to be closely linked to a disturbance of the redox equilibrium (Ehret et al., 1996; Westendorp et al., 1995b). Recently, we have shown that Tat induces generation of H2O2. The increase in H2O2 induced by Tat interferes with TCR signaling. Thus, cells preincubated with Tat and stimulated via the TCR reveal a strong increase in the H2O2 signal in comparison to untreated cells. Such an enhanced oxidative signal combined with Ca2þ influx into the cytosol causes a significant increase in AICD (Gulow et al., 2005) (Fig. 3B). Since CD4 stimulation

No Life Without Death

129

induces Ca2þ influx into the cytosol, H2O2 generation due to Tat treatment is sufficient to induce CD95L expression (Gulow et al., 2005) (Fig. 3C). Thus, understanding determinants of apoptosis in HIV infection will have profound implications for design of an anti‐HIV vaccine.

ACKNOWLEDGMENTS This chapter is dedicated to Eva and George Klein on behalf of their 80th birthday. The work is supported by the “Wilhelm Sander Stiftung,” the “Deutsche Forschungsgemeinschaft” (DFG), the European Community, and the Tumor Center Heidelberg/Mannheim.

REFERENCES Algeciras‐Schimnich, A., Shen, L., Barnhart, B. C., Murmann, A. E., Burkhardt, J. K., and Peter, M. E. (2002). Molecular ordering of the initial signaling events of CD95. Mol. Cell. Biol. 22, 207–220. Ashkenazi, A., and Dixit, V. M. (1998). Death receptors: Signaling and modulation. Science 281, 1305–1308. Badley, A. D., Pilon, A. A., Landay, A., and Lynch, D. H. (2000). Mechanisms of HIV‐ associated lymphocyte apoptosis. Blood 96, 2951–2964. Banda, N. K., Bernier, J., Kurahara, D. K., Kurrle, R., Haigwood, N., Sekaly, R. P., and Finkel, T. H. (1992). Crosslinking CD4 by human immunodeficiency virus gp120 primes T cells for activation‐induced apoptosis. J. Exp. Med. 176, 1099–1106. Bauer, M. K., Vogt, M., Los, M., Siegel, J., Wesselborg, S., and Schulze‐Osthoff, K. (1998). Role of reactive oxygen intermediates in activation‐induced CD95 (APO‐1/Fas) ligand expression. J. Biol. Chem. 273, 8048–8055. Beiqing, L., Chen, M., and Whisler, R. L. (1996). Sublethal levels of oxidative stress stimulate transcriptional activation of c‐jun and suppress IL‐2 promoter activation in Jurkat T cells. J. Immunol. 157, 160–169. Chang, D. W., Xing, Z., Pan, Y., Algeciras‐Schimnich, A., Barnhart, B. C., Yaish‐Ohad, S., Peter, M. E., and Yang, X. (2002). c‐FLIP(L) is a dual function regulator for caspase‐8 activation and CD95‐mediated apoptosis. EMBO J. 21, 3704–3714. Chang, L. J., Chen, C. H., Urlacher, V., and Lee, T. Z. (2000). Differential apoptosis effects of primate lentiviral Vpr and Vpx in mammalian cells. J. Biomed. Sci. 7, 322–333. Chaudhri, G., Clark, I. A., Hunt, N. H., Cowden, W. B., and Ceredig, R. (1986). Effect of antioxidants on primary alloantigen‐induced T cell activation and proliferation. J. Immunol. 137, 2646–2652. Chaudhri, G., Hunt, N. H., Clark, I. A., and Ceredig, R. (1988). Antioxidants inhibit proliferation and cell surface expression of receptors for interleukin‐2 and transferrin in T lymphocytes stimulated with phorbol myristate acetate and ionomycin. Cell. Immunol. 115, 204–213. Chen, M., Elder, R. T., Yu, M., O’Gorman, M. G., Selig, L., Benarous, R., Yamamoto, A., and Zhao, Y. (1999). Mutational analysis of Vpr‐induced G2 arrest, nuclear localization, and cell death in fission yeast. J. Virol. 73, 3236–3245.

130

Peter H. Krammer et al.

Chow, W. A., Fang, J. J., and Yee, J. K. (2000). The IFN regulatory factor family participates in regulation of Fas ligand gene expression in T cells. J. Immunol. 164, 3512–3518. Clements, J. L. (2003). Known and potential functions for the SLP‐76 adapter protein in regulating T‐cell activation and development. Immunol. Rev. 191, 211–219. Collins, K. L., Chen, B. K., Kalams, S. A., Walker, B. D., and Baltimore, D. (1998). HIV‐1 Nef protein protects infected primary cells against killing by cytotoxic T lymphocytes. Nature 391, 397–401. Conti, L., Rainaldi, G., Matarrese, P., Varano, B., Rivabene, R., Columba, S., Sato, A., Belardelli, F., Malorni, W., and Gessani, S. (1998). The HIV‐1 vpr protein acts as a negative regulator of apoptosis in a human lymphoblastoid T cell line: Possible implications for the pathogenesis of AIDS. J. Exp. Med. 187, 403–413. Cory, S., and Adams, J. M. (2002). The Bcl2 family: Regulators of the cellular life‐or‐death switch. Nat. Rev. Cancer 2, 647–656. Davis, C. B., Dikic, I., Unutmaz, D., Hill, C. M., Arthos, J., Siani, M. A., Thompson, D. A., Schlessinger, J., and Littman, D. R. (1997). Signal transduction due to HIV‐1 envelope interactions with chemokine receptors CXCR4 or CCR5. J. Exp. Med. 186, 1793–1798. Desbarats, J., Freed, J. H., Campbell, P. A., and Newell, M. K. (1996). Fas (CD95) expression and death‐mediating function are induced by CD4 cross‐linking on CD4þ T cells. Proc. Natl. Acad. Sci. USA 93, 11014–11018. Devadas, S., Zaritskaya, L., Rhee, S. G., Oberley, L., and Williams, M. S. (2002). Discrete generation of superoxide and hydrogen peroxide by T cell receptor stimulation: Selective regulation of mitogen‐activated protein kinase activation and fas ligand expression. J. Exp. Med. 195, 59–70. Dittmar, M. T., McKnight, A., Simmons, G., Clapham, P. R., Weiss, R. A., and Simmonds, P. (1997). HIV‐1 tropism and co‐receptor use. Nature 385, 495–496. Du, C., Fang, M., Li, Y., Li, L., and Wang, X. (2000). Smac, a mitochondrial protein that promotes cytochrome c‐dependent caspase activation by eliminating IAP inhibition. Cell 102, 33–42. Ehret, A., Westendorp, M. O., Herr, I., Debatin, K. M., Heeney, J. L., Frank, R., and Krammer, P. H. (1996). Resistance of chimpanzee T cells to human immunodeficiency virus type 1 Tat‐enhanced oxidative stress and apoptosis. J. Virol. 70, 6502–6507. Ensoli, B., Barillari, G., Salahuddin, S. Z., Gallo, R. C., and Wong‐Staal, F. (1990). Tat protein of HIV‐1 stimulates growth of cells derived from Kaposi’s sarcoma lesions of AIDS patients. Nature 345, 84–86. Faccio, L., Fusco, C., Chen, A., Martinotti, S., Bonventre, J. V., and Zervos, A. S. (2000). Characterization of a novel human serine protease that has extensive homology to bacterial heat shock endoprotease HtrA and is regulated by kidney ischemia. J. Biol. Chem. 275, 2581–2588. Featherstone, M. (2002). Coactivators in transcription initiation: Here are your orders. Curr. Opin. Genet. Dev. 12, 149–155. Feldmann, M., Dayan, C., Grubeck‐Loebenstein, B., Rapoport, B., and Londei, M. (1992). Mechanism of Graves thyroiditis: Implications for concepts and therapy of autoimmunity. Int. Rev. Immunol. 9, 91–106. Finco, T. S., Kadlecek, T., Zhang, W., Samelson, L. E., and Weiss, A. (1998). LAT is required for TCR‐mediated activation of PLCgamma1 and the Ras pathway. Immunity 9, 617–626. Finkel, T. H., Tudor‐Williams, G., Banda, N. K., Cotton, M. F., Curiel, T., Monks, C., Baba, T. W., Ruprecht, R. M., and Kupfer, A. (1995). Apoptosis occurs predominantly in bystander cells and not in productively infected cells of HIV‐ and SIV‐infected lymph nodes. Nat. Med. 1, 129–134. French, L. E., Hahne, M., Viard, I., Radlgruber, G., Zanone, R., Becker, K., Muller, C., and Tschopp, J. (1996). Fas and Fas ligand in embryos and adult mice: Ligand expression in

No Life Without Death

131

several immune‐privileged tissues and coexpression in adult tissues characterized by apoptotic cell turnover. J. Cell Biol. 133, 335–343. Fridovich, I. (1970). Quantitative aspects of the production of superoxide anion radical by milk xanthine oxidase. J. Biol. Chem. 245, 4053–4057. Giorgio, M., Migliaccio, E., Orsini, F., Paolucci, D., Moroni, M., Contursi, C., Pelliccia, G., Luzi, L., Minucci, S., and Marcaccio, M. (2005). Electron transfer between cytochrome c and p66Shc generates reactive oxygen species that trigger mitochondrial apoptosis. Cell 122, 221–233. Golks, A., Brenner, D., Fritsch, C., Krammer, P. H., and Lavrik, I. N. (2005). C‐FLIPR‐a new regulator of death receptor‐induced apoptosis. J. Biol. Chem. 280, 14507–14513. Goltsev, Y. V., Kovalenko, A. V., Arnold, E., Varfolomeev, E. E., Brodianskii, V. M., and Wallach, D. (1997). CASH, a novel caspase homologue with death effector domains. J. Biol. Chem. 272, 19641–19644. Gougeon, M. L., and Montagnier, L. (1999). Programmed cell death as a mechanism of CD4 and CD8 T cell deletion in AIDS. Molecular control and effect of highly active anti‐retroviral therapy. Ann. NY Acad. Sci. 887, 199–212. Gourley, T. S., and Chang, C. H. (2001). Cutting edge: The class II transactivator prevents activation‐induced cell death by inhibiting Fas ligand gene expression. J. Immunol. 166, 2917–2921. Grandori, C., Cowley, S. M., James, L. P., and Eisenman, R. N. (2000). The Myc/Max/Mad network and the transcriptional control of cell behavior. Annu. Rev. Cell Dev. Biol. 16, 653–699. Griffith, C. E., Zhang, W., and Wange, R. L. (1998). ZAP‐70‐dependent and ‐independent activation of Erk in Jurkat T cells. Differences in signaling induced by H2o2 and Cd3 cross‐ linking. J. Biol. Chem. 273, 10771–10776. Gulow, K., Kaminski, M., Darvas, K., Suss, D., Li‐Weber, M., and Krammer, P. H. (2005). HIV‐1 trans‐activator of transcription substitutes for oxidative signaling in activation‐ induced T cell death. J. Immunol. 174, 5249–5260. Han, D. K., Chaudhary, P. M., Wright, M. E., Friedman, C., Trask, B. J., Riedel, R. T., Baskin, D. G., Schwartz, S. M., and Hood, L. (1997). MRIT, a novel death‐effector domain‐ containing protein, interacts with caspases and BclXL and initiates cell death. Proc. Natl. Acad. Sci. USA 94, 11333–11338. Hashimoto, F., Oyaizu, N., Kalyanaraman, V. S., and Pahwa, S. (1997). Modulation of Bcl‐2 protein by CD4 cross‐linking: A possible mechanism for lymphocyte apoptosis in human immunodeficiency virus infection and for rescue of apoptosis by interleukin‐2. Blood 90, 745–753. He, J., Choe, S., Walker, R., Di Marzio, P., Morgan, D. O., and Landau, N. R. (1995). Human immunodeficiency virus type 1 viral protein R (Vpr) arrests cells in the G2 phase of the cell cycle by inhibiting p34cdc2 activity. J. Virol. 69, 6705–6711. He, Y. W., Deftos, M. L., Ojala, E. W., and Bevan, M. J. (1998). RORgamma t, a novel isoform of an orphan receptor, negatively regulates Fas ligand expression and IL‐2 production in T cells. Immunity 9, 797–806. Hengartner, M. O. (2000). The biochemistry of apoptosis. Nature 407, 770–776. Hengartner, M. O., and Horvitz, H. R. (1994). C. elegans cell survival gene ced‐9 encodes a functional homolog of the mammalian proto‐oncogene bcl‐2. Cell 76, 665–676. Herbein, G., Mahlknecht, U., Batliwalla, F., Gregersen, P., Pappas, T., Butler, J., O’rien, W. A., and Verdin, E. (1998). Apoptosis of CD8þ T cells is mediated by macrophages through interaction of HIV gp120 with chemokine receptor CXCR4. Nature 395, 189–194. Hildeman, D. A., Mitchell, T., Teague, T. K., Henson, P., Day, B. J., Kappler, J., and Marrack, P. C. (1999). Reactive oxygen species regulate activation‐induced T cell apoptosis. Immunity 10, 735–744.

132

Peter H. Krammer et al.

Hodge, S., Novembre, F. J., Whetter, L., Gelbard, H. A., and Dewhurst, S. (1998). Induction of fas ligand expression by an acutely lethal simian immunodeficiency virus, SIVsmmPBj14. Virology 252, 354–363. Hogan, P. G., Chen, L., Nardone, J., and Rao, A. (2003). Transcriptional regulation by calcium, calcineurin, and NFAT. Genes Dev. 17, 2205–2232. Hsu, S. C., Gavrilin, M. A., Lee, H. H., Wu, C. C., Han, S. H., and Lai, M. Z. (1999). NF‐ kappa B‐dependent Fas ligand expression. Eur. J. Immunol. 29, 2948–2956. Hu, S., Vincenz, C., Ni, J., Gentz, R., and Dixit, V. M. (1997). I‐FLICE, a novel inhibitor of tumor necrosis factor receptor‐1‐ and CD‐95‐induced apoptosis. J. Biol. Chem. 272, 17255–17257. Inohara, N., Koseki, T., Hu, Y., Chen, S., and Nunez, G. (1997). CLARP, a death effector domain‐containing protein interacts with caspase‐8 and regulates apoptosis. Proc. Natl. Acad. Sci. USA 94, 10717–10722. Irmler, M., Thome, M., Hahne, M., Schneider, P., Hofmann, K., Steiner, V., Bodmer, J. L., Schroter, M., Burns, K., Mattmann, C., et al. (1997). Inhibition of death receptor signals by cellular FLIP. Nature 388, 190–195. Israel, N., Gougerot‐Pocidalo, M. A., Aillet, F., and Virelizier, J. L. (1992). Redox status of cells influences constitutive or induced NF‐kappa B translocation and HIV long terminal repeat activity in human T and monocytic cell lines. J. Immunol. 149, 3386–3393. Jackman, J. K., Motto, D. G., Sun, Q., Tanemoto, M., Turck, C. W., Peltz, G. A., Koretzky, G. A., and Findell, P. R. (1995). Molecular cloning of SLP‐76, a 76‐kDa tyrosine phosphoprotein associated with Grb2 in T cells. J. Biol. Chem. 270, 7029–7032. Jackson, S. H., Devadas, S., Kwon, J., Pinto, L. A., and Williams, M. S. (2004). T cells express a phagocyte‐type NADPH oxidase that is activated after T cell receptor stimulation. Nat. Immunol. 5, 818–827. Jacotot, E., Ravagnan, L., Loeffler, M., Ferri, K. F., Vieira, H. L., Zamzami, N., Costantini, P., Druillennec, S., Hoebeke, J., Briand, J. P., et al. (2000). The HIV‐1 viral protein R induces apoptosis via a direct effect on the mitochondrial permeability transition pore. J. Exp. Med. 191, 33–46. Jowett, J. B., Planelles, V., Poon, B., Shah, N. P., Chen, M. L., and Chen, I. S. (1995). The human immunodeficiency virus type 1 vpr gene arrests infected T cells in the G2 þ M phase of the cell cycle. J. Virol. 69, 6304–6313. Joza, N., Susin, S. A., Daugas, E., Stanford, W. L., Cho, S. K., Li, C. Y., Sasaki, T., Elia, A. J., Cheng, H. Y., Ravagnan, L., et al. (2001). Essential role of the mitochondrial apoptosis‐ inducing factor in programmed cell death. Nature 410, 549–554. Katsikis, P. D., Wunderlich, E. S., Smith, C. A., and Herzenberg, L. A. (1995). Fas antigen stimulation induces marked apoptosis of T lymphocytes in human immunodeficiency virus‐ infected individuals. J. Exp. Med. 181, 2029–2036. Kavurma, M. M., Santiago, F. S., Bonfoco, E., and Khachigian, L. M. (2001). Sp1 phosphorylation regulates apoptosis via extracellular FasL‐Fas engagement. J. Biol. Chem. 276, 4964–4971. Kerr, J. F., Wyllie, A. H., and Currie, A. R. (1972). Apoptosis: A basic biological phenomenon with wide‐ranging implications in tissue kinetics. Br. J. Cancer 26, 239–257. Kirchhoff, S., Sebens, T., Baumann, S., Krueger, A., Zawatzky, R., Li‐Weber, M., Meinl, E., Neipel, F., Fleckenstein, B., and Krammer, P. H. (2002). Viral IFN‐regulatory factors inhibit activation‐induced cell death via two positive regulatory IFN‐regulatory factor 1‐dependent domains in the CD95 ligand promoter. J. Immunol. 168, 1226–1234. Klein, J. A., Longo‐Guess, C. M., Rossmann, M. P., Seburn, K. L., Hurd, R. E., Frankel, W. N., Bronson, R. T., and Ackerman, S. L. (2002). The harlequin mouse mutation downregulates apoptosis‐inducing factor. Nature 419, 367–374.

No Life Without Death

133

Korsmeyer, S. J., Wei, M. C., Saito, M., Weiler, S., Oh, K. J., and Schlesinger, P. H. (2000). Pro‐ apoptotic cascade activates BID, which oligomerizes BAK or BAX into pores that result in the release of cytochrome c. Cell Death Differ. 7, 1166–1173. Krammer, P. H. (2000). CD95’s deadly mission in the immune system. Nature 407, 789–795. Krueger, A., Baumann, S., Krammer, P. H., and Kirchhoff, S. (2001). FLICE‐inhibitory proteins: Regulators of death receptor‐mediated apoptosis. Mol. Cell. Biol. 21, 8247–8254. Krueger, A., Fas, S. C., Baumann, S., and Krammer, P. H. (2003). The role of CD95 in the regulation of peripheral T‐cell apoptosis. Immunol. Rev. 193, 58–69. Kwon, J., Devadas, S., and Williams, M. S. (2003). T cell receptor‐stimulated generation of hydrogen peroxide inhibits MEK‐ERK activation and lck serine phosphorylation. Free Radic. Biol. Med. 35, 406–417. Lacana, E., and D’Adamio, L. (1999). Regulation of Fas ligand expression and cell death by apoptosis‐linked gene 4. Nat. Med. 5, 542–547. Lahdenpohja, N., and Hurme, M. (1998). CD28‐mediated activation in CD45RAþ and CD45ROþ T cells: Enhanced levels of reactive oxygen intermediates and c‐Rel nuclear translocation in CD45RAþ cells. J. Leukoc. Biol. 63, 775–780. Latinis, K. M., Carr, L. L., Peterson, E. J., Norian, L. A., Eliason, S. L., and Koretzky, G. A. (1997). Regulation of CD95 (Fas) ligand expression by TCR‐mediated signaling events. J. Immunol. 158, 4602–4611. Lee, J., Richburg, J. H., Younkin, S. C., and Boekelheide, K. (1997). The Fas system is a key regulator of germ cell apoptosis in the testis. Endocrinology 138, 2081–2088. Lee, S. L., Wesselschmidt, R. L., Linette, G. P., Kanagawa, O., Russell, J. H., and Milbrandt, J. (1995). Unimpaired thymic and peripheral T cell death in mice lacking the nuclear receptor NGFI‐B (Nur77). Science 269, 532–535. Li, H., Zhu, H., Xu, C. J., and Yuan, J. (1998). Cleavage of BID by caspase 8 mediates the mitochondrial damage in the Fas pathway of apoptosis. Cell 94, 491–501. Li, L. Y., Luo, X., and Wang, X. (2001). Endonuclease G is an apoptotic DNase when released from mitochondria. Nature 412, 95–99. Li, P., Nijhawan, D., Budihardjo, I., Srinivasula, S. M., Ahmad, M., Alnemri, E. S., and Wang, X. (1997). Cytochrome c and dATP‐dependent formation of Apaf‐1/caspase‐9 complex initiates an apoptotic protease cascade. Cell 91, 479–489. Lipton, S. A., and Bossy‐Wetzel, E. (2002). Dueling activities of AIF in cell death versus survival: DNA binding and redox activity. Cell 111, 147–150. Liu, X., Kim, C. N., Yang, J., Jemmerson, R., and Wang, X. (1996). Induction of apoptotic program in cell‐free extracts: Requirement for dATP and cytochrome c. Cell 86, 147–157. Liu, Z. G., Smith, S. W., McLaughlin, K. A., Schwartz, L. M., and Osborne, B. A. (1994). Apoptotic signals delivered through the T‐cell receptor of a T‐cell hybrid require the immediate‐early gene nur77. Nature 367, 281–284. Li‐Weber, M., and Krammer, P. H. (2003). Function and regulation of the CD95 (APO‐1/Fas) ligand in the immune system. Semin. Immunol. 15, 145–157. Li‐Weber, M., Laur, O., and Krammer, P. H. (1999). Novel Egr/NF‐AT composite sites mediate activation of the CD95 (APO‐1/Fas) ligand promoter in response to T cell stimulation. Eur. J. Immunol. 29, 3017–3027. Li‐Weber, M., Laur, O., Dern, K., and Krammer, P. H. (2000). T cell activation‐induced and HIV tat‐enhanced CD95(APO‐1/Fas) ligand transcription involves NF‐kappaB. Eur. J. Immunol. 30, 661–670. Lorenzo, H. K., and Susin, S. A. (2004). Mitochondrial effectors in caspase‐independent cell death. FEBS Lett. 557, 14–20. Los, M., Schenk, H., Hexel, K., Baeuerle, P. A., Droge, W., and Schulze‐Osthoff, K. (1995). IL‐2 gene expression and NF‐kappa B activation through CD28 requires reactive oxygen production by 5‐lipoxygenase. EMBO J. 14, 3731–3740.

134

Peter H. Krammer et al.

Luo, X., Budihardjo, I., Zou, H., Slaughter, C., and Wang, X. (1998). Bid, a Bcl2 interacting protein, mediates cytochrome c release from mitochondria in response to activation of cell surface death receptors. Cell 94, 481–490. Macreadie, I. G., Castelli, L. A., Hewish, D. R., Kirkpatrick, A., Ward, A. C., and Azad, A. A. (1995). A domain of human immunodeficiency virus type 1 Vpr containing repeated H(S/F) RIG amino acid motifs causes cell growth arrest and structural defects. Proc. Natl. Acad. Sci. USA 92, 2770–2774. Malissen, B. (2003). An evolutionary and structural perspective on T cell antigen receptor function. Immunol. Rev. 191, 7–27. Mann, D. A., and Frankel, A. D. (1991). Endocytosis and targeting of exogenous HIV‐1 Tat protein. EMBO J. 10, 1733–1739. Matsue, H., Edelbaum, D., Shalhevet, D., Mizumoto, N., Yang, C., Mummert, M. E., Oeda, J., Masayasu, H., and Takashima, A. (2003). Generation and function of reactive oxygen species in dendritic cells during antigen presentation. J. Immunol. 171, 3010–3018. McClure, R. F., Heppelmann, C. J., and Paya, C. V. (1999). Constitutive Fas ligand gene transcription in Sertoli cells is regulated by Sp1. J. Biol. Chem. 274, 7756–7762. Medema, J. P., Scaffidi, C., Kischkel, F. C., Shevchenko, A., Mann, M., Krammer, P. H., and Peter, M. E. (1997). FLICE is activated by association with the CD95 death‐inducing signaling complex (DISC). EMBO J. 16, 2794–2804. Minemoto, Y., Shimura, M., Ishizaka, Y., Masamune, Y., and Yamashita, K. (1999). Multiple centrosome formation induced by the expression of Vpr gene of human immunodeficiency virus. Biochem. Biophys. Res. Commun. 258, 379–384. Mitsiades, C. S., Poulaki, V., and Mitsiades, N. (2003). The role of apoptosis‐inducing receptors of the tumor necrosis factor family in thyroid cancer. J. Endocrinol. 178, 205–216. Moore, J. P. (1997). Coreceptors: Implications for HIV pathogenesis and therapy. Science 276, 51–52. Muzio, M., Chinnaiyan, A. M., Kischkel, F. C., O’Rourke, K., Shevchenko, A., Ni, J., Scaffidi, C., Bretz, J. D., Zhang, M., Gentz, R., et al. (1996). FLICE, a novel FADD‐homologous ICE/ CED‐3‐like protease, is recruited to the CD95 (Fas/APO‐1) death—inducing signaling complex. Cell 85, 817–827. Muzio, M., Stockwell, B. R., Stennicke, H. R., Salvesen, G. S., and Dixit, V. M. (1998). An induced proximity model for caspase‐8 activation. J. Biol. Chem. 273, 2926–2930. Neudorf, S. M., Jones, M. M., McCarthy, B. M., Harmony, J. A., and Choi, E. M. (1990). The CD4 molecule transmits biochemical information important in the regulation of T lymphocyte activity. Cell. Immunol. 125, 301–314. Novogrodsky, A., Ravid, A., Rubin, A. L., and Stenzel, K. H. (1982). Hydroxyl radical scavengers inhibit lymphocyte mitogenesis. Proc. Natl. Acad. Sci. USA 79, 1171–1174. Orie, N. N., Zidek, W., and Tepel, M. (1999). Chemoattractant‐ and mitogen‐induced generation of reactive oxygen species in human lymphocytes: The role of calcium. Exp. Physiol. 84, 515–520. Oshimi, Y., Oda, S., Honda, Y., Nagata, S., and Miyazaki, S. (1996). Involvement of Fas ligand and Fas‐mediated pathway in the cytotoxicity of human natural killer cells. J. Immunol. 157, 2909–2915. Oyaizu, N., McCloskey, T. W., Than, S., Hu, R., Kalyanaraman, V. S., and Pahwa, S. (1994). Cross‐linking of CD4 molecules upregulates Fas antigen expression in lymphocytes by inducing interferon‐gamma and tumor necrosis factor‐alpha secretion. Blood 84, 2622–2631. Oyaizu, N., Adachi, Y., Hashimoto, F., McCloskey, T. W., Hosaka, N., Kayagaki, N., Yagita, H., and Pahwa, S. (1997). Monocytes express Fas ligand upon CD4 cross‐linking and induce CD4þ T cells apoptosis: A possible mechanism of bystander cell death in HIV infection. J. Immunol. 158, 2456–2463.

No Life Without Death

135

Parrish, J., Li, L., Klotz, K., Ledwich, D., Wang, X., and Xue, D. (2001). Mitochondrial endonuclease G is important for apoptosis in C. elegans. Nature 412, 90–94. Piguet, V., Chen, Y. L., Mangasarian, A., Foti, M., Carpentier, J. L., and Trono, D. (1998). Mechanism of Nef‐induced CD4 endocytosis: Nef connects CD4 with the mu chain of adaptor complexes. EMBO J. 17, 2472–2481. Planelles, V., Jowett, J. B., Li, Q. X., Xie, Y., Hahn, B., and Chen, I. S. (1996). Vpr‐induced cell cycle arrest is conserved among primate lentiviruses. J. Virol. 70, 2516–2524. Rashba‐Step, J., Turro, N. J., and Cederbaum, A. I. (1993). Increased NADPH‐ and NADH‐ dependent production of superoxide and hydroxyl radical by microsomes after chronic ethanol treatment. Arch. Biochem. Biophys. 300, 401–408. Rasper, D. M., Vaillancourt, J. P., Hadano, S., Houtzager, V. M., Seiden, I., Keen, S. L., Tawa, P., Xanthoudakis, S., Nasir, J., Martindale, D., et al. (1998). Cell death attenuation by ‘Usurpin’, a mammalian DED‐caspase homologue that precludes caspase‐8 recruitment and activation by the CD‐95 (Fas, APO‐1) receptor complex. Cell Death Differ. 5, 271–288. Re, F., Braaten, D., Franke, E. K., and Luban, J. (1995). Human immunodeficiency virus type 1 Vpr arrests the cell cycle in G2 by inhibiting the activation of p34cdc2‐cyclin B. J. Virol. 69, 6859–6864. Reth, M. (2002). Hydrogen peroxide as second messenger in lymphocyte activation. Nat. Immunol. 3, 1129–1134. Rieux‐Laucat, F., Fischer, A., and Deist, F. L. (2003). Cell‐death signaling and human disease. Curr. Opin. Immunol. 15, 325–331. Robinson, B. S., Hii, C. S., and Ferrante, A. (1998). Activation of phospholipase A2 in human neutrophils by polyunsaturated fatty acids and its role in stimulation of superoxide production. Biochem. J. 336(Pt. 3), 611–617. Rogel, M. E., Wu, L. I., and Emerman, M. (1995). The human immunodeficiency virus type 1 vpr gene prevents cell proliferation during chronic infection. J. Virol. 69, 882–888. Rosenberg, Z. F., and Fauci, A. S. (1989). The immunopathogenesis of HIV infection. Adv. Immunol. 47, 377–431. Roshal, M., Zhu, Y., and Planelles, V. (2001). Apoptosis in AIDS. Apoptosis 6, 103–116. Rothwarf, D. M., and Karin, M. (1999). The NF‐kappa B activation pathway: A paradigm in information transfer from membrane to nucleus. Sci STKE 1999, RE1. Sandstrom, P. A., Roberts, B., Folks, T. M., and Buttke, T. M. (1993). HIV gene expression enhances T cell susceptibility to hydrogen peroxide‐induced apoptosis. AIDS Res. Hum. Retroviruses 9, 1107–1113. Sandstrom, P. A., Mannie, M. D., and Buttke, T. M. (1994). Inhibition of activation‐induced death in T cell hybridomas by thiol antioxidants: Oxidative stress as a mediator of apoptosis. J. Leukoc. Biol. 55, 221–226. Scaffidi, C., Fulda, S., Srinivasan, A., Friesen, C., Li, F., Tomaselli, K. J., Debatin, K. M., Krammer, P. H., and Peter, M. E. (1998). Two CD95 (APO‐1/Fas) signaling pathways. EMBO J. 16, 1675–1678. Schreck, R., Rieber, P., and Baeuerle, P. A. (1991). Reactive oxygen intermediates as apparently widely used messengers in the activation of the NF‐kappa B transcription factor and HIV‐1. EMBO J. 10, 1675–2258. Schulze‐Osthoff, K., Ferrari, D., Los, M., Wesselborg, S., and Peter, M. E. (1998). Apoptosis signaling by death receptors. Eur. J. Biochem. 254, 439–459. Schwartz, O., Marechal, V., Le Gall, S., Lemonnier, F., and Heard, J. M. (1996). Endocytosis of major histocompatibility complex class I molecules is induced by the HIV‐1 Nef protein. Nat. Med. 2, 338–342. Segal, A. W., and Shatwell, K. P. (1997). The NADPH oxidase of phagocytic leukocytes. Ann. NY Acad. Sci. 832, 215–222.

136

Peter H. Krammer et al.

Sekkat, C., Dornand, J., and Gerber, M. (1988). Oxidative phenomena are implicated in human T‐cell stimulation. Immunology 63, 431–437. Shimura, M., Onozuka, Y., Yamaguchi, T., Hatake, K., Takaku, F., and Ishizaka, Y. (1999a). Micronuclei formation with chromosome breaks and gene amplification caused by Vpr, an accessory gene of human immunodeficiency virus. Cancer Res. 59, 2259–2264. Shimura, M., Tanaka, Y., Nakamura, S., Minemoto, Y., Yamashita, K., Hatake, K., Takaku, F., and Ishizaka, Y. (1999b). Micronuclei formation and aneuploidy induced by Vpr, an accessory gene of human immunodeficiency virus type 1. FASEB J. 13, 621–637. Shostak, L. D., Ludlow, J., Fisk, J., Pursell, S., Rimel, B. J., Nguyen, D., Rosenblatt, J. D., and Planelles, V. (1999). Roles of p53 and caspases in the induction of cell cycle arrest and apoptosis by HIV‐1 vpr. Exp. Cell Res. 251, 156–165. Shu, H. B., Halpin, D. R., and Goeddel, D. V. (1997). Casper is a FADD‐ and caspase‐related inducer of apoptosis. Immunity 6, 751–763. Siegel, R. M., Frederiksen, J. K., Zacharias, D. A., Chan, F. K., Johnson, M., Lynch, D., Tsien, R. Y., and Lenardo, M. J. (2000). Fas preassociation required for apoptosis signaling and dominant inhibition by pathogenic mutations. Science 288, 2354–2357. Sies, H. (1977). Peroxisomal enzymes and oxygen metabolism in liver. Adv. Exp. Med. Biol. 78, 51–60. Smith, C. A., Farrah, T., and Goodwin, R. G. (1994). The TNF receptor superfamily of cellular and viral proteins: Activation, costimulation, and death. Cell 76, 959–962. Somma, F., Tuosto, L., Gilardini Montani, M. S., Di Somma, , M. S., Cundari, , M. M., and Piccolella, E. (2000). Engagement of CD4 before TCR triggering regulates both Bax‐ and Fas (CD95)‐mediated apoptosis. J. Immunol. 164, 5078–5087. Srinivasula, S. M., Ahmad, M., Ottilie, S., Bullrich, F., Banks, S., Wang, Y., Fernandes‐Alnemri, T., Croce, C. M., Litwack, G., Tomaselli, K. J., et al. (1997). FLAME‐1, a novel FADD‐like anti‐apoptotic molecule that regulates Fas/TNFR1‐induced apoptosis. J. Biol. Chem. 272, 18542–18545. Stassi, G., and De Maria, R. (2002). Autoimmune thyroid disease: New models of cell death in autoimmunity. Nat. Rev. Immunol. 2, 195–204. Stewart, S. A., Poon, B., Jowett, J. B., and Chen, I. S. (1997). Human immunodeficiency virus type 1 Vpr induces apoptosis following cell cycle arrest. J. Virol. 71, 5579–5592. Suda, T., Takahashi, T., Golstein, P., and Nagata, S. (1993). Molecular cloning and expression of the Fas ligand, a novel member of the tumor necrosis factor family. Cell 75, 1169–1178. Surh, C. D., and Sprent, J. (1994). T‐cell apoptosis detected in situ during positive and negative selection in the thymus [see comments]. Nature 372, 100–103. Susin, S. A., Lorenzo, H. K., Zamzami, N., Marzo, I., Snow, B. E., Brothers, G. M., Mangion, J., Jacotot, E., Costantini, P., Loeffler, M., et al. (1999). Molecular characterization of mitochondrial apoptosis‐inducing factor. Nature 397, 441–446. Suzuki, Y., Imai, Y., Nakayama, H., Takahashi, K., Takio, K., and Takahashi, R. (2001). A serine protease, HtrA2, is released from the mitochondria and interacts with XIAP, inducing cell death. Mol. Cell 8, 613–621. Takahashi, T., Tanaka, M., Brannan, C. I., Jenkins, N. A., Copeland, N. G., Suda, T., and Nagata, S. (1994). Generalized lymphoproliferative disease in mice, caused by a point mutation in the Fas ligand. Cell 76, 969–976. Tateyama, M., Oyaizu, N., McCloskey, T. W., Than, S., and Pahwa, S. (2000). CD4 T lymphocytes are primed to express Fas ligand by CD4 cross‐linking and to contribute to CD8 T‐cell apoptosis via Fas/FasL death signaling pathway. Blood 96, 195–202. Tatla, S., Woodhead, V., Foreman, J. C., and Chain, B. M. (1999). The role of reactive oxygen species in triggering proliferation and IL‐2 secretion in T cells. Free Radic. Biol. Med. 26, 14–24.

No Life Without Death

137

Thompson, C. B. (1995). Apoptosis in the pathogenesis and treatment of disease. Science 267, 1456–1462. Thornberry, N. A., and Lazebnik, Y. (1998). Caspases: Enemies within. Science 281, 1312–1316. Turrens, J. F., Freeman, B. A., Levitt, J. G., and Crapo, J. D. (1982). The effect of hyperoxia on superoxide production by lung submitochondrial particles. Arch. Biochem. Biophys. 217, 401–410. Vaux, D. L., and Korsmeyer, S. J. (1999). Cell death in development. Cell 96, 245–254. Vaux, D. L., Cory, S., and Adams, J. M. (1988). Bcl‐2 gene promotes haemopoietic cell survival and cooperates with c‐myc to immortalize pre‐B cells. Nature 335, 440–442. Villalba, M., Kasibhatla, S., Genestier, L., Mahboubi, A., Green, D. R., and Altman, A. (1999). Protein kinase ctheta cooperates with calcineurin to induce Fas ligand expression during activation‐induced T cell death. J. Immunol. 163, 5813–5819. Villunger, A., Ghaffari‐Tabrizi, N., Tinhofer, I., Krumbock, N., Bauer, B., Schneider, T., Kasibhatla, S., Greil, R., Baier‐Bitterlich, G., Uberall, F., et al. (1999). Synergistic action of protein kinase C theta and calcineurin is sufficient for Fas ligand expression and induction of a crmA‐sensitive apoptosis pathway in Jurkat T cells. Eur. J. Immunol. 29, 3549–3561. Walczak, H., and Sprick, M. R. (2001). Biochemistry and function of the DISC. Trends Biochem. Sci. 26, 452–453. Wange, R. L., and Huang, Y. (2004). T cell receptor signaling: Beyond complex complexes. J. Biol. Chem. 279, 28827–28830. Watanabe, D., Suda, T., and Nagata, S. (1995). Expression of Fas in B cells of the mouse germinal center and Fas‐dependent killing of activated B cells. Int. Immunol. 7, 1949–1956. Watanabe, N., Yamaguchi, T., Akimoto, Y., Rattner, J. B., Hirano, H., and Nakauchi, H. (2000). Induction of M‐phase arrest and apoptosis after HIV‐1 Vpr expression through uncoupling of nuclear and centrosomal cycle in HeLa cells. Exp. Cell Res. 258, 261–269. Watanabe‐Fukunaga, R., Brannan, C. I., Copeland, N. G., Jenkins, N. A., and Nagata, S. (1992). Lymphoproliferation disorder in mice explained by defects in Fas antigen that mediates apoptosis. Nature 356, 314–317. Weissman, D., Rabin, R. L., Arthos, J., Rubbert, A., Dybul, M., Swofford, R., Venkatesan, S., Farber, J. M., and Fauci, A. S. (1997). Macrophage‐tropic HIV and SIV envelope proteins induce a signal through the CCR5 chemokine receptor. Nature 389, 981–985. Westendorp, M. O., Frank, R., Ochsenbauer, C., Stricker, K., Dhein, J., Walczak, H., Debatin, K. M., and Krammer, P. H. (1995a). Sensitization of T cells to CD95‐mediated apoptosis by HIV‐1 Tat and gp120. Nature 375, 497–500. Westendorp, M. O., Shatrov, V. A., Schulze‐Osthoff, K., Frank, R., Kraft, M., Los, M., Krammer, P. H., Droge, W., and Lehmann, V. (1995b). HIV‐1 Tat potentiates TNF‐induced NF‐kappa B activation and cytotoxicity by altering the cellular redox state. EMBO J. 14, 546–554. Williams, M. S., and Henkart, P. A. (1996). Role of reactive oxygen intermediates in TCR‐ induced death of T cell blasts and hybridomas. J. Immunol. 157, 2395–2402. Williamson, J. R. (1986). Role of inositol lipid breakdown in the generation of intracellular signals. State of the art lecture. Hypertension 8, II140–II156. Woronicz, J. D., Calnan, B., Ngo, V., and Winoto, A. (1994). Requirement for the orphan steroid receptor Nur77 in apoptosis of T‐cell hybridomas. Nature 367, 277–281. Xu, X. N., Screaton, G. R., Gotch, F. M., Dong, T., Tan, R., Almond, N., Walker, B., Stebbings, R., Kent, K., Nagata, S., et al. (1997). Evasion of cytotoxic T lymphocyte (CTL) responses by nef‐dependent induction of Fas ligand (CD95L) expression on simian immunodeficiency virus‐infected cells. J. Exp. Med. 186, 7–16.

138

Peter H. Krammer et al.

Xu, X. N., Laffert, B., Screaton, G. R., Kraft, M., Wolf, D., Kolanus, W., Mongkolsapay, J., McMichael, A. J., and Baur, A. S. (1999). Induction of Fas ligand expression by HIV involves the interaction of Nef with the T cell receptor zeta chain. J. Exp. Med. 189, 1489–1496. Zauli, G., Gibellini, D., Secchiero, P., Dutartre, H., Olive, D., Capitani, S., and Collette, Y. (1999). Human immunodeficiency virus type 1 Nef protein sensitizes CD4(þ) T lymphoid cells to apoptosis via functional upregulation of the CD95/CD95 ligand pathway. Blood 93, 1000–1010. Zhang, W., Sloan‐Lancaster, J., Kitchen, J., Trible, R. P., and Samelson, L. E. (1998). LAT: The ZAP‐70 tyrosine kinase substrate that links T cell receptor to cellular activation. Cell 92, 83–92.

Control of Apoptosis in Human Multiple Myeloma by Insulin‐like Growth Factor I (IGF‐I) Helena Jernberg‐Wiklund and Kenneth Nilsson Department of Genetics and Pathology, Rudbeck Laboratory, Uppsala University, SE‐751 85 Uppsala, Sweden

I. Selected Biological Properties of Human Multiple Myeloma II. Human MM Models In Vitro and In Vivo III. Targeting Anti‐apoptosis and Proliferative Signals in Human MM A. Anti‐apoptotic Events in Human MM B. The IGF‐I Signaling Pathways C. IGF‐I as a Growth and Survival Factor in MM D. IGF‐I as a Target for Therapy IV. The Effect of Combinational Treatment with PPP on Human MM Cells Is Additive and Synergistic References

Human multiple myeloma (MM) is characterized by the expansion of neoplastic plasmablasts/plasma cells with complex genetic aberrations and high dependence for survival and growth on cytokines produced in the bone marrow microenvironment. As tools in the study of MM about 80 authentic MM cell lines and a few relevant in vivo mouse models are available. The dependence on insulin‐like growth factor receptor (IGF‐IR) signaling in the development and maintenance of the malignant phenotype in a variety of cancers is a rationale for attempts to improve tumor treatment by selectively inhibiting the IGF‐IR in malignant cells by neutralizing antibodies, dominant negative IGF‐IR, and IGF‐IR siRNA. Testing the hypothesis that abrogating IGF‐IR‐mediated signaling of survival should make MM cells more susceptible to apoptosis, our studies have so far provided proof‐of‐principle by the demonstration that inhibition of a signaling pathway stimulating survival renders cells susceptible to drug‐induced apoptosis when the drug (dexamethasone) and inhibitor (rapamycin) converge on the same target, that is p70S6K. The recent publication of the three‐dimensional structure of the IGF‐IR kinase domain has facilitated the development of IGF‐IR inhibitors of the cyclolignan family, that is picropodophyllin, with capacity to distinguish also in vivo between the IGF‐IR and the insulin receptor. Studies in vitro and in vivo with picropodophyllin show promising effects, that is apoptosis induction and growth arrest, and have made it possible to evaluate the biological and therapeutic effects of inhibition of the IGF‐IR signaling in MM. # 2007 Elsevier Inc.

Advances in CANCER RESEARCH Copyright 2007, Elsevier Inc. All rights reserved.

139

0065-230X/07 $35.00 DOI: 10.1016/S0065-230X(06)97006-7

140

Helena Jernberg‐Wiklund and Kenneth Nilsson

I. SELECTED BIOLOGICAL PROPERTIES OF HUMAN MULTIPLE MYELOMA Important characteristic features of human multiple myeloma (MM) include that the slow‐growing and relatively apoptosis‐resistant malignant plasmablasts/plasma cells almost exclusively home in the bone marrow, that complex genetic aberrations are present, and that a high frequency of illegitimate switch recombinations in the tumor cells result in translocations to the immunoglobulin heavy chain locus (Bergsagel et al., 2005; Kuehl and Bergsagel, 2002). Extensive studies on the interaction of MM cells with the bone marrow environment have documented that the growth, survival, and apoptosis of the tumor cells are regulated by paracrine cytokine signaling by normal bone marrow cells and by complex interactions with the extracellular matrix. Studies using fluorescent in situ hybridization (FISH) indicate that a translocation to IgH locus is present in 65–75% of biopsy specimens of MM. A few partner genes have been identified to be involved in these immunoglobulin translocations [i.e., 11q13 (cyclin D1/CCND1), 6p21 (cyclin D3/CCND3), 4p16 (FGFR3/MMSET), 16q23 (c‐maf)] (Bergsagel et al., 2005). Such translocations are found in approximately 40% of MM biopsy cells and in established cell MM cell lines (Bergsagel et al., 1996; Jernberg‐ Wiklund and Nilsson, 2000). Secondary, non‐B cell‐specific translocations are found in approximately 5% of MM patients and involve 8q24 (c‐MYC), 6p25 (MUM1/IRF4), and 20q11 (MAFB). Other potentially oncogenic genetic alterations in MM are activating mutations of K‐ and N‐Ras in early progression of MM, deregulation of c‐myc, and inactivation of the p53 and Rb pathways occurring late during progression (Corradini et al., 1994; Hideshima et al., 2004; Jernberg‐Wiklund et al., 1992b; Kuehl and Bergsagel, 2002; Nilsson, 1994). In addition, bcl‐2 has been reported by us and others to be expressed in high but variable levels in MM (Hamilton et al., 1991; Nilsson et al., 1990; Pettersson et al., 1992). Although the bcl‐2 gene has not been identified as a partner in IgH translocations of MM, we have previously reported that bcl‐2 gene amplification may be a mechanism explaining overexpression of this gene (Pettersson et al., 1992). In addition, the increased expression of members of the Bcl‐2 family of anti‐apoptotic proteins, that is Bcl‐XL and Mcl‐1, has frequently been associated with the malignant phenotype and resistance to cytotoxic drugs (Gazitt et al., 1998; Jourdan et al., 2003; Spets et al., 2002; Tu et al., 1998). Several non‐neoplastic cell types in the bone marrow seem to influence the growth of MM cells, for example stromal cells, osteoclasts, ostoblasts, vascular endothelial cells, lymphocytes, and monocytes. Indirect and direct reciprocal tumor–bone marrow cell interactions are mediated by cytokines, receptors, and adhesion molecules and guide the generation of paracrine cytokines regulating survival, proliferation, and homing of tumor cells to

141

IGF‐IR Signaling in Multiple Myeloma

this site. Among these, IL‐6 and IGF‐I represent important paracrine growth and survival factors (Fig. 1). It is plausible that malignant cells during transformation and progression acquire independence of these paracrine interactions and develop autocrine production of cytokines (Jernberg‐ Wiklund et al., 1992a). Several studies have confirmed IL‐6 to be a major paracrine and autocrine growth factor in MM cells in vitro and in vivo (Hideshima et al., 2004; Jernberg et al., 1991; Kawano et al., 1988; Klein et al., 1989; Nilsson et al., 1990). Additional growth factors, for example

FasL

CD40L

Tc

TH IFN-g

IL-1a/b TNFa/b

MM

IL-10

Adh., TGFb SC

IL-6

IFN-a M

IL-6 IGF-I

OB

Stroma

OC Bone

Stimulatory signal Inhibitory signal

Stimulatory and/or inhibitory signal

Fig. 1 Possible interactions of MM with nonneoplastic cells of the bone marrow microenvironment. IL‐1/, TNF/, IL‐6, and some other cytokines and/or cell–cell contacts mediate the activation of cells in the bone marrow. Data from our group suggest that MM cells may produce TGF1 that may activate cells of the bone marrow microenvironment. The activation leads to an increased production of factors, for example IL‐6, IGF‐I, IFN, IFN, FasL, CD40L, and IL‐10, that regulate the growth and survival of MM cells. IFN and IFN are mainly growth inhibitory to MM cells, while IL‐6 and IGF‐I stimulate paracrine and autocrine growth and survival (Jernberg‐Wiklund and Nilsson, 2000). TH/T c, T‐helper/T‐cytotoxic lymphocytes; M, macrophages; SC, other stromal cell; OB, osteoblast; OC, osteoclast. Originally published in Nilsson et al. (1999) with kind permission of Springer Science and Business Media.

142

Helena Jernberg‐Wiklund and Kenneth Nilsson

HB‐EGF and IL‐21, have been attributed a role in the control of MM growth and survival (Mahtouk et al., 2005; Wang et al., 2002) as have BAFF and APRIL that seem to promote survival in MM cells deprived of IL‐6 and serum (van de Donk et al., 2005). Adhesion of MM cells to stromal cells and osteoblasts, as well as classical bone‐resorbing cytokines (IL‐1, IL‐1, TNF‐, TNF‐, and M‐CSF) may trigger the production of IL‐6 in these cell types (Barille et al., 1995; Carter et al., 1990; Nemunaitis et al., 1989; Uchiyama et al., 1993). Also processes indirectly affecting tumor growth, that is, angiogenesis and the function of osteoblasts/osteoclasts are regulated by reciprocal interactions between MM and the local environment. Increased levels of VEGF, bFGF, and HGF have been detected in peripheral blood of MM patients possibly attributed to the up‐regulated production by a tumor– host interactions (Bisping et al., 2003; Borset et al., 1996). While both VEGF and HGF may directly stimulate IL‐6 production, HGF seems to indirectly stimulate IL‐6 production via IL‐11 synthesis in osteoblasts.

II. HUMAN MM MODELS IN VITRO AND IN VIVO Studies on MM were for long hampered by the lack of relevant model systems in vitro and in vivo (Nilsson, 1977). With the introduction of organ cultures on monolayers of stromal cells from bone marrow, the success rate of establishment of MM cell lines increased (Nilsson, 1977; Nilsson et al., 1970). However, only by the better characterization of the survival requirements of MM cells in vitro, that is the presence of stromal cells, IL‐6, and other cytokines, the establishment of authentic MM cell lines in ordinary nonstirred suspensions cultures was possible at a relatively high success rate from patients with advanced disease (Jernberg‐Wiklund and Nilsson, 2000). To date, approximately 80 cell lines have been characterized as authentic and EBV negative, originating from MM and plasma cell leukemia. In a majority of the cases, these have been established from blood, pleural effusions, or ascites as suspension cultures using IL‐6 as an exogenous growth factor (Jernberg‐Wiklund and Nilsson, 2000). Several in vivo models for MM have also been described. For long, the pristane‐induced plasmacytoma in Balb/c mice (Potter and Wiener, 1992) was the only available in vivo model of MM. However, although the oil‐induced granuloma resembles the bone marrow in its local production of paracrine IL‐6 and IGF‐I for growth, tumorigenesis, and survival of the tumor, the genetic alterations (mainly B cell‐specific translocations of c‐myc) differed from those found in the human MM. Human MM cell lines may also be grown in human fetal bone marrow or in other lymphoid tissues inoculated into severe combined immunodeficiency (SCID) mice (Yaccoby et al., 1998). However, with this

IGF‐IR Signaling in Multiple Myeloma

143

method, the success is only limited with primary MM cells or with MM cell lines dependent on interactions with host stromal cell. However, Tassone et al. (2005) published that the IL‐6‐dependent INA‐6 cell line could be grown in a human fetal bone chip in SCID‐hu mice, suggesting that this model may allow growth of MM clones that are highly dependent on the microenvironment for survival and growth. As a model for in vivo studies on targeted drugs and survival, the 5TMM immunocompetent murine “Radl model” is a highly relevant in vivo model for the human disease (Radl, 1994; Radl et al., 1979). Two models of the 5TMM in vivo model are available, the 5T2 transplantable tumor, representing a model of the most common forms of human MM, with moderate growth and osteolytic lesions, and the 5T33 tumor, a more aggressive form with rapid growth of tumor cells and with involvement of the liver apart from the bone marrow (Vanderkerken et al., 2003). The 5TMM model originates from spontaneously developed monoclonal immunoglobulin‐secreting plasmablasts/plasma cells in aged C57BL KalwRij mice. 5TMM cells are propagated by intravenous injection of BM cells into young naive syngeneic recipients. Clinical, biological, and genetic characterization of 5TMM cells suggest that this model is similar to human MM in several major aspects, for example phenotypic properties and localization to the bone marrow. In fact, the first evidence of IGF‐I acting as a chemoattractant in the homing process of MM originated from studies using this model (Vanderkerken et al., 1999).

III. TARGETING ANTI‐APOPTOSIS AND PROLIFERATIVE SIGNALS IN HUMAN MM A. Anti‐apoptotic Events in Human MM As mentioned, resistance to apoptosis is an important aspect of the MM phenotype and increases as the disease progresses. A tentative and attractive model on how a genetic aberration early in the tumorigenesis, that is c‐myc, may provide an initial sensitization to apoptosis in tumors came from the studies of Evan and Littlewood (1998) and Evan et al. (1992). Applied to the MM model, a genetic alteration leading to overexpression of cyclin D1, D2, or D3 may constitute an early event in MM tumorigenesis (Bergsagel et al., 2005; Kuehl and Bergsagel, 2002) (Fig. 2). Reciprocal interactions between neoplastic cells and their microenvironment will then provide paracrine production of IGF‐I and IL‐6 and other less well‐defined cytokines governing survival of at least the classical intramedullary MM. Later during tumor development, mutations that disable normal cellular control of

144

Helena Jernberg‐Wiklund and Kenneth Nilsson

IFN

FasL

DNA damage MDM2 IL-6 IGF-I Bcl-2

IFNR Fas

p53

c-Myc Cyclin D

GC

GR

IGF-I

Sensitization to apoptosis

Apoptosis

Proliferation

Progression

IL-6 IGF-I

Rb del. p53 mut. Ras mut.

Factors that inhibit apoptosis Pro-apoptotic factors

Fig. 2 A tentative model of the regulation of growth and survival in MM. A genetic alteration and/or overexpression of cyclin D and alternative genetic aberrations are early event that leads to increased proliferation and increased sensitivity to apoptotic signals. The MM cells are dependent on survival factors, such as IL‐6, IGF‐I, and Bcl‐2, which inhibit apoptosis. The transformed cells are initially more sensitive to apoptotic stimuli than nonneoplastic cells, providing a therapeutic window. During the progression, the tumor accumulates additional genetic aberrations, for example Rb deletions, p53 mutations, and N and K‐Ras mutations, which may make the MM cells gradually less dependent on paracrine signals for growth and survival. The sensitivity to apoptosis is also decreased, probably contributing to the development of resistance to cytotoxic drugs [adopted from a general model by Evan and Littlewood (1998)]. GC, glucocorticoid; GR, glucocorticoid receptor. Originally published in Nilsson et al. (1999) with kind permission of Springer Science and Business Media.

proliferation and apoptosis, for example p53 mutations, development of autonomous growth by autocrine loops of IGF‐I and IL‐6, and/or overexpression of downstream anti‐apoptosis gene targets such as Bcl‐XL and Mcl‐1 seem to be required for progression of MM. The IGF‐IR promoter is regulated by several genes associated with the tumor progression in MM, including p53. Ultimately, mutant p53 or overexpression of mdm2 as an alternate mechanism for p53 inactivation may up‐regulate the expression of IGF‐IR (Larsson et al., 2005). However, bearing in mind the function of the MDM2 in targeting the IGF‐IR for proteasome degradation, it is plausible that a delicate balance of both the expression and nuclear versus cytoplasmic location of these regulatory proteins will be decisive for the biological effect (Girnita et al., 2003).

IGF‐IR Signaling in Multiple Myeloma

145

The strategy of targeting anti‐apoptotic signals in MM cells is based on the hypothesis that blocking the survival signaling circuits will reset their level of sensitivity for apoptosis. One example of decreased sensitivity to pro‐apoptotic signals in MM is the resistance to Fas‐mediated apoptosis, in some cases also associated with Fas expression, as reported by us and others (Egle et al., 1997; Landowski et al., 1999; Shain et al., 2000; Spets et al., 1998; Villunger et al., 1997). However, others have contradicted this finding and were unable to establish a correlation between Fas expression and susceptibility to apoptosis (Hata et al., 1995; Westendorf et al., 1995). A decreased sensitivity to Fas‐induced apoptosis has been ascribed to multiple mechanisms, including IL‐6‐mediated down‐regulation of the Jun kinase (Xu et al., 1998), a decreased Bax expression (Egle et al., 1997), and to the induction of signal transducer and activator of transcription 3 (STAT3)‐ dependent up‐regulation of Bcl‐XL (Catlett‐Falcone et al., 1999). The death receptor Fas/Apo‐1/CD95 is a homotrimer promiscuously expressed, particularly high in thymus, liver, heart, and kidney (Scaffidi et al., 2000). The natural ligand, FasL, induces clustering of Fas subsequently leading to the recruitment of pro‐caspase 8, which together with the receptor and Fas‐associated death domain protein (FADD) forms the death‐inducing signaling complex (DISC). In analogy with the apoptosome of the mitochondrial pathway of apoptosis, this complex facilitates the activation of the effector caspase 3. We previously demonstrated that the sensitivity to Fas‐induced apoptosis in MM can be restored by IFN type I and type II stimulation prior to the exposure to Fas‐agonistic antibodies, mimicing the action of FasL (Spets et al., 1998). Several oncogenic signaling pathways converge on a limited set of nuclear transcription factors. Shifting the balance of gene expression of those genes favoring survival to genes thereby sensitizing cells to apoptosis may therefore represent a new therapeutic possibility. Thus, targeting transcription factor(s) by selective inhibitors may result in blockade of the effects of a multitude of upstream genetic aberrations and exogenous factors from the microenvironment. It is well known that IFNs induce and activate STAT1, whereas the MM growth factor IL‐6 induces STAT3 activity (Calo et al., 2003). In light of our previous findings, we hypothesized that the STAT1‐ activating action of IFNs may counteract STAT3 thereby shifting the balance from survival to apoptosis by influencing the gene expression of apoptosis‐ related genes in MM (Dimberg et al., 2005). Supporting this line of thinking, we found that IFNs attenuated STAT3 phosphorylation and activation in addition to the induction of STAT1 activation. Neither Bcl‐XL, previously identified (Catlett‐Falcone et al., 1999; Stephanou et al., 1999) as a mediator of STAT3‐induced survival, nor Bcl‐2 was, however, affected by IFNs. Identified potential IFN targets possibly participating in the apoptosis susceptibility process were, however, Fas and Apo2L/TRAIL. TRAIL is

146

Helena Jernberg‐Wiklund and Kenneth Nilsson

structurally similar to FasL, and the pathways employed by these inducers share multiple downstream components (Nagata, 1997; Scaffidi et al., 1998). Furthermore, both Fas and TRAIL are likely to be potentially regulated by the described shifted balance. In addition, we could demonstrate that MM cells with endogenous or ectopically high expression of Fas were more prone to undergo apoptosis. We could, however, not find support for TRAIL being a major determinant of IFN‐mediated sensitization. These results indicate that IFN increases the susceptibility to Fas‐induced apoptosis in MM cells by inducing a shift in STAT protein activation from STAT3 to STAT1. The subsequent transcriptional up‐regulation of selected target genes may thus define a pro‐apoptotic MM phenotype.

B. The IGF‐I Signaling Pathways During the last decade a role of IGF‐I signaling in development and progression of cancer has been well established (Baserga, 2000; LeRoith and Roberts, 2003; Pollak et al., 2004). The IGF axis is composed of at least two IGF‐ligands (IGF‐I and IGF‐II) and several receptors (IGF‐IR, IR, and IGF‐IIR) mediating the biological effect of the ligands. The complexity of IGF signaling is further increased by the interactions of IGFs with six IGF‐ binding proteins (IGFBPs) that exhibit higher affinity for IGFs than does IGF‐IR. IGF‐BP‐3 is the most abundant in serum and binds approximately 90% of the circulating IGF‐I in complex with the acid‐labile subunit (ALS). Thus, IGF bioactivity in the tissues is not merely a function of circulating levels of IGF. Rather, both IGF‐I stability and ligand bioavailability are under the influence of several variables of the complex IGF network, that is IGF expression, receptor surface exposure, IGFBPs, and proteases‐digesting IGFBPs (Jones and Clemmons, 1995). The IGF‐I and IGF‐II are single polypeptide molecules that show 62% homology to insulin. In contrast to the level of IGF‐I in mice, increasing from birth to puberty and slowly declining through adulthood, the level of IGF‐I and IGF‐II in human tissue and serum, albeit at different levels, seems to remain stable throughout life (LeRoith and Roberts, 2003). These ligands may interact with several structurally distinct receptors in homo‐ or heterocombinations (Pollak et al., 2004). The IGF‐IR is structurally and functionally closely related to the insulin receptor (IR). Both IGF‐IR and IR are preformed heterotetrameric tyrosine kinase (TK) receptors consisting of two extracellular alpha‐subunits and two membrane‐ spanning beta‐subunits, including the TK domain and the C domain. The TK domains of the IGF‐IR and IR are highly homologous and share 84% sequence identity (Adams et al., 2000). In contrast, the IGF‐IIR lacks the intracellular TK domain and is identical to the monomeric mannose‐6

IGF‐IR Signaling in Multiple Myeloma

147

phosphate receptor that exerts its function mainly in endocytosis and intracellular protein trafficking. Most, if not all, effects of the IGF‐I result from its binding to the IGF‐IR and the subsequent activation of the receptor. Although IGF‐I may also interact with the IR, IGF‐I has predominant affinity for the IGF‐IR over the IR and only binds the IR at pharmacological doses. Interestingly, two splice variants of IR have been documented, where one of these, the IR‐A fetal splice variant, is frequently found in cancer (Frasca et al., 1999). Although IGF‐II may interact with the IGF‐IR, the IGF‐II is likely to exert its proliferative effect via this isoform (IR‐A) (Frasca et al., 1999) as demonstrated in breast cancer cell lines (Sciacca et al., 1999). Some reports have also indicated the formation of hybrid receptors, that is dimerization of IGF‐IR and IR hemireceptors, that retain a high affinity for IGF‐I, but the functional significance of these receptors are as yet unknown in vivo (Federici et al., 1997). The TK domain of the IGF‐IR spans the region of 973–1229 amino acids of the beta‐subunit of the IGF‐IR. High‐affinity ligand‐induced activity of the IGF‐IR kinase requires the sequential phosphorylation of tyrosines 1131, 1135, and 1136 within the activation loop of the catalytic domain (Baserga, 2000; Favelyukis et al., 2001). These changes allow access of substrates and ATP to the loop, and subsequent phosphorylation of adjacent tyrosines in the juxtamembrane and C‐terminal regions, flanking the TK domain. Phosphorylation within this cytoplasmic domain provides sites for the recruitment of signaling intermediates. The two best characterized pathways associated with IGF‐IR activation are mediated via distinct adaptor proteins, among these are the Shc and the IR substrate (IRS)‐1 and ‐2. This in turn leads to activation of parallel as well as cross‐talking signaling pathways including the mitogen‐activated protein kinase (MAPK) cascade that culminates in the activation of extracellular signal‐related kinase (ERK)1 and ERK2, and the phosphatidylinositol‐3‐kinase (PI3K)‐Akt pathway (Adams et al., 2000). The perhaps best characterized substrates of PI 3‐kinase signaling are the Akt/PKB serine/threonine kinases, constituting targets of both IGF‐I and insulin signaling. Additional activation pathways and adaptor proteins have been reported including phospholipase C and protein kinase C (PKC), the STATs, and the suppressors of cytokine signaling (SOCS) (Kayali et al., 1998; Siddle et al., 2001; Zhang et al., 2006; Zong et al., 2000). The rational for targeting the IGF‐IR in anticancer treatment is based on two findings: (1) the IGF‐IR mediates fundamental aspects of the malignant phenotype, that is growth, motility, and protection from apoptosis and (2) IGF‐IR does not appear to be an absolute requirement for maintenance of normal cell homeostasis in adult individuals (LeRoith et al., 1995; Valentinis and Baserga, 2001). In contrast to the, for example, EGF family of receptors, activation of IGF‐IR, IR, and hybrid receptors by overexpression

148

Helena Jernberg‐Wiklund and Kenneth Nilsson

alone is not seen. With a few exceptions of primary tumor cells originating from breast, melanoma, and colon (Surmacz, 2003), there are currently no reports on a persistent pattern of IGF‐IR overexpression in malignant cells in comparison to normal counterpart cells. With the exception for what has been demonstrated in cell models (Kozma and Weber, 1990), activation of IGF‐IR thus requires ligand binding. Therefore, and since an increased IGF‐IR expression per se has not been found to correlate to tumor progression (Mitsiades et al., 2004), it is likely that the availability of the ligand in a paracrine or autocrine setting would be the limiting factor guarding the growth and survival in the malignant cells. Since inhibition of IR will have profound and irreversible effects on glucose homeostasis, targeting of only IGF‐IR would be preferred to blocking of tumor growth regulated by the IGFs and insulin. This aim has encouraged the development of IGF‐IR inhibitors for clinical use, possibly without the severe side effects associated with traditional treatment protocols using cytotoxic drugs.

C. IGF‐I as a Growth and Survival Factor in MM Genetic aberrations of the IGF axis, with the exception of alterations in downstream substrates of the IGF signaling pathway, for example PTEN mutations (Ge and Rudikoff, 2000; Maehama and Dixon, 1999) leading to an indirect perturbed activation of the receptor, have so far not been reported. However, during tumor development, IGF‐I is accessible and can be produced locally or at remote sites. Notably, IGF‐I is present in serum but is also abundantly residing in bone matrix where it is produced by the bone marrow stromal cells and osteoblasts (Chenu et al., 1990; Govoni et al., 2005) adjacent to the tumor cells. Likely, this microenvironment in MM provides the bioavailability of the ligand to the tumor cells, and as a result of autocrine or paracrine IGF‐I stimulation the IGF‐IR may become hyperactive and stimulate to increased survival and growth of the MM cells (Nilsson et al., 1999). IGF‐I may also indirectly facilitate the expansion of the MM clone in the bone marrow. Thus, the increased angiogenesis in the bone marrow of MM patients, that is positively correlated to increased tumor growth, seems to be attributed to the secretion of pro‐angiogenic cytokines among which VEGF is produced not only by the tumor cells but also by osteoblasts on IGF‐I stimulation (Dunn, 2000; Menu et al., 2004). Direct and indirect actions of IGF‐I make the survival signaling molecules downstream of the IGF‐IR in MM equally important targets for therapy as mutated oncogenes. Although the data concerning the prognostic impact of IGF‐I and IGF‐IR serum levels are conflicting (Bataille et al., 2005; Mitsiades and Mitsiades, 2005; Standal et al., 2002), an increased IGF‐IR

IGF‐IR Signaling in Multiple Myeloma

149

expression has been reported in highly aggressive variants of MM (Bataille et al., 2005; Chng et al., 2005). We and others have previously ascribed a pivotal role of IGF‐I in MM cells as an important component of the bone marrow microenvironment in promoting proliferation, survival, and drug‐ resistance (Freund et al., 1993; Georgii‐Hemming et al., 1996; Jelinek et al., 1997; Nilsson et al., 1999). Our studies indicate that IGFs (in serum or from autocrine or paracrine sources) via IGF‐IR signaling can stimulate growth in MM cell lines and also protect MM cells from apoptosis induced by serum starvation (Georgii‐Hemming et al., 1996; Nilsson et al., 1999). IGF‐I has been suggested both to modulate IL‐6‐induced growth (Jelinek et al., 1997) and to act independently of IL‐6 (Ferlin et al., 2000). A synergistic enhancement of apoptosis induced by IGF‐IR inhibition and activation of conceptually different signaling pathways such as dexamethasone and FasL has been shown by us and a few other groups (Nilsson et al., 1999; Ogawa et al., 2000; Xu et al., 1997). IGF‐IR inhibitory molecules interfering with the IGF‐I autocrine loop (anti‐IGF‐IR/IR3), or with the IGF‐I axis (somatostatin) (Georgii‐Hemming et al., 1999), had comparable effects in profoundly suppressing serum‐induced survival. In addition, the abrogation of this survival pathway potentiated apoptosis induced by dexamethasone and the death receptor Fas in MM cells (Fig. 2). Also taken into account the importance of IGF‐I in plasma cell tumor development and homing in the murine models of MM (Asosingh et al., 2000; Li et al., 2000; Vanderkerken et al., 1999, 2000), it has become increasingly evident that IGF‐IR is a promising target for MM therapy.

D. IGF‐I as a Target for Therapy In the light of the general importance of the bone marrow microenvironment for the expansion of MM tumor cell clones, our experimental plan was based on the hypothesis that a local paracrine or autocrine hyperactivation of the IGF‐IR, and selected downstream targets, will result in survival advantage of the malignant cells versus normal cells. Various attempts have been made to inhibit the IGF‐IR function and expression (Surmacz, 2003). Only the development of inhibitors of the receptor TK domain (Girnita et al., 2004; Mitsiades and Mitsiades, 2005), with capacity to distinguish also in vivo between the IGF‐IR and IR (Girnita et al., 2004; Menu et al., 2006), has made it possible to approach a mechanistic understanding of using the IGF‐IR as a target for therapy in MM. Studies on the biological effects and molecular mechanisms of these inhibitors in the MM model constitute an important experimental platform for future therapeutic implementation. Information from crystallographic studies revealed the conformational differences of IGF‐IR and IR kinases and thus provided the basis

150

Helena Jernberg‐Wiklund and Kenneth Nilsson

for the development of selective inhibitors of the cyclolignan family (Fig. 3). The inhibitory effect of one of these members, picropodophyllin (PPP), proved promising, since it displayed selectivity for the IGF‐IR and did not coinhibit tyrosine phosphorylation of the IR, or of a selected panel of receptors less related to the IGF‐IR (FGF‐R, PDGF‐R, or EGF‐R) (Girnita et al., 2004). As demonstrated, PPP down‐regulates autophosphorylation of the IGF‐IR in both human (Fig. 4) and murine MM cells without inhibiting the autophosphorylation of the IR (Menu et al., 2006; Stromberg et al., 2006) (Fig. 4 and data not shown). Importantly, inhibition of the IGF‐I RTK with PPP was shown to be noncompetitive with ATP, suggesting interference with the IGF‐IR at the substrate level (Girnita et al., 2004). A study from the same laboratory has shown that PPP initially blocks phosphorylation of the residue Tyr1136 in the activation loop of the IGF‐IR kinase (Vasilcanu et al.,

0.93 nm

PPT

1131 1135

0.95 nm

0.91 nm

1136

PPP

IGF-1R

Fig. 3 The three‐dimensional structure of the insulin‐like growth factor (IGF)‐IR a‐loop. The three‐ dimensional structures of a peptide constructed by a computer from the amino acid sequence 1127–1138 of the IGF‐IR (activation loop of the tyrosine kinase domain) using the Internal Coordinate Mechanics software compared with the structures of the cyclolignans podophyllotoxin (PPT) and picropodophyllin (PPP). The tyrosines 1131, 1135, and 1136 are indicated. Originally published in Girnita et al. (2004). Copyright American Association for Cancer Research, Inc., used with permission.

151

IGF‐IR Signaling in Multiple Myeloma

A

Relative IGF-1R kinase activity (%)

125

100

75 RPMI 8226 Karpas 707 50

25

0 0

0.5

1

1.5

2

2.5

3

PPP (µM) B

Relative insulin R kinase activity (%)

125

100

75 RPMI 8226 Karpas 707 50

25

0 0

0.5

1

1.5

2

2.5

3

PPP (µM)

Fig. 4 The effects of PPP on IGF‐I RTK autophosphorylation. (A) IGF‐IR was extracted from total cell lysates of the RPMI 8226 and the Karpas 707 MM cell lines and incubated with PPP at indicated concentrations before the kinase reaction was initiated by the addition of ATP. The autophosphorylation of the IGF‐I RTK was quantified spectrophotometrically. (B) Insulin R was extracted from total cell lysates of the RPMI 8226 and Karpas 707 cell lines and incubated with PPP at indicated concentrations before kinase reaction was initiated by the addition of ATP. The autophosphorylation of the insulin R was quantified spectrophotometrically. Originally published in Stromberg et al. (2006). Copyright American Society of Hematology, used with permission.

152

Helena Jernberg‐Wiklund and Kenneth Nilsson

2004). Phosphorylation of Tyr1136 seems to stabilize the catalytically optimized conformation of the receptor and as such it has been suggested to be of paramount importance for IGF‐IR activity (Favelyukis et al., 2001). This compound is now available in oral preparation for gavage and in crystalline form for preclinical studies and phase I studies. Importantly, no substantial toxicity by PPP in animals (mice to primates) has been documented (LD50 > 500 mg/kg in rodents). All MM cell lines of the panel, including IL‐6‐dependent and ‐independent as well as the in vitro selected drug‐resistant subclones of the RPMI 8226 cell line, responded to PPP with growth inhibition (Fig. 5). Confirming the results from the cell lines, PPP potently inhibited DNA synthesis in patient samples of MM cells, notably, also in the presence of bone marrow stromal cells. However, in comparison with the IR3 antibody, previously shown to induce an increased susceptibility to apoptosis, PPP more potently inhibited growth in MM, probably reflecting the direct effect on the kinase activity of the IGF‐IR. Some reports have indicated that an efficient blockade of the IGF‐IR may induce a compensatory response in vivo, that is up‐regulation of IGF‐I in serum (Mitsiades et al., 2004). However, even considering such a scenario, our studies rather emphasize the potency of PPP since pretreatment of the MM cell lines with high concentrations of IGF‐I, IGF‐2, insulin, or IL‐6 did not reduce the inhibitory effects of PPP (Fig. 6). In line with this notion, PPP also exerted antitumor activity in vivo and significantly reduced serum paraprotein concentrations and percentage of idiotype‐positive tumor cells in the 5TMM mouse model. Notably, PPP treatment did not negatively affect the levels of insulin‐dependent metabolic parameters (glucose and albumin) (Fig. 7 and data not shown). In the 5T33MM model, MM cells also accumulate in spleen and liver, in Fig. 7 this is obvious as a reduced splenomegaly and hepatomegaly can be noted. Intriguingly, the mice treated with PPP had clearly a prolonged survival according to the Kaplan–Meier analysis. The mice, however, ultimately become nonresponsive to PPP as a consequence of treatment‐induced fibrosis (Menu et al., 2006) (Fig. 8). Taken together, these data show a preferable scenario in vivo where therapeutically achievable concentrations of PPP does not negatively influence IR signaling and metabolism but has a profound effect on survival of the mice (Menu et al., 2006). The molecular mechanisms so far revealed show that the inhibition of the IGF‐IR by PPP is associated with growth arrest, caspase‐dependent, and caspase‐independent apoptosis in human MM cells and the reduced expression of a few anti‐apoptotic genes, including Mcl‐1. The IGF‐I‐induced phosphorylation of the downstream substrates ERK1 and ERK2 was down‐regulated by treatment with PPP, which was in concordance to previous findings in this model, and to the effect of PPP in MM cells of the 5TMM mouse model.

A

EJM

100

JJN3 Karpas 707

75

L363 50

LP-1

25

OPM-2 U-1996

0 0

0.1

0.2

0.3 0.4 PPP (µM)

0.5

1

Relative number of viable cells (%)

Relative number of viable cells (%)

125

125

U-1958 50

0 0

RPMI 8226 75

RPMI 8226/ Dox40

50

RPMI 8226/LR5

25 0 0.2

0.3 0.4 PPP (µM)

0.5

1

Relative number of viable cells (%)

Relative number of viable cells (%)

100

0.1

U-266−1970

25

2

C

0

U-1957

75

125 125

B

100

2

0.1

0.2

0.3 0.4 PPP (µM)

D Control PPP 0.1 µM

75

PPP 0.2 µM 50

PPP 0.5 µM PPP 1 µM

25

1

Patient 5 Patient 8

75 Patient 9 Patient 10

0 Control

PPP 1 µM

Relative number of viable cells (%)

Relative 3H-TdR incorporation (%)

125

125

25

2

0

E

50

1

100

2

3

4

150

100

0.5

5 6 Patient

7

8

9

10

F

100

75

BMSC SK1069

50

25

0 0 0.01 0.02 0.05 0.1 0.2 0.5

1

2

PPP (µM)

Fig. 5 The effect of cyclolignan PPP on growth in MM cell lines and primary MM cells. A panel of 13 MM cell lines was treated with PPP at indicated concentrations for 48 h followed by analysis using the resazurin assay. The cell lines were divided into three groups showing (A) IL‐6‐independent EJM, JJN3, Karpas 707, L363, LP‐1, OPM‐2, and U‐1996 cell lines, (B) IL‐6‐dependent U‐1957, U‐1958, and U‐266–1970 cell lines, and (C) the IL‐6‐independent RPMI 8226 cell line and its subclones RPMI 8226/Dox40 and RPMI 8226/LR5 with increased resistance to doxorubicin and melphalan, respectively. IL‐6 was added to

154

Helena Jernberg‐Wiklund and Kenneth Nilsson

IV. THE EFFECT OF COMBINATIONAL TREATMENT WITH PPP ON HUMAN MM CELLS IS ADDITIVE AND SYNERGISTIC Apart from the potential use of PPP as a single agent, targeting of the IGF‐ IR might increase the efficacy of other drugs designed to induce apoptosis. A mechanistic understanding of the substrates and gene expression profile involved in IGF‐I‐activated survival may shed light on molecular candidates to be targeted in rational combinatorial MM therapy. Several substrates regulating caspase activity and apoptosis downstream of PI3K/Akt activation have been suggested. Some of these, Bad, caspase 9, the FKHR family of transcription factors, and the IKK/inhibitor of NF‐B may not all constitute ideal candidate targets for therapy. However, the mammalian target of rapamycin (mTOR) regulating translation via the downstream target p70S6K and transcription of the c‐myc gene may well be a candidate target if proven functional in IGF‐IR‐induced survival in MM. We have previously shown that targeting mTOR in combinatorial regimen with dexamethasone may increase the susceptibility of MM cells to undergo apoptosis (Stromberg et al., 2004). Our studies provide a proof‐of‐principle that abrogating the survival factor pathway by a specific inhibitor renders cells susceptible to drug‐induced apoptosis if the drug and inhibitor converge on the same target, the p70S6K. The selective inhibitor of mTOR, rapamycin, can inhibit cytokine‐dependent proliferation, although the effect on survival remains unclear when used as a single agent. In our studies, rapamycin mediated a significant potentiation of dexamethasone‐induced apoptosis and suppressed both constitutive, serum‐, IGF‐I‐, and IL‐6‐induced survival in MM cell lines (Stromberg et al., 2004). This was also demonstrated in purified cells from MM patients, one of these exhibiting resistance to first‐line treatment, indicating that this drug combination may have potential also in vivo. The studies on the molecular mechanisms of apoptosis sensitization via rapamycin and

U‐1957, U‐1958, U‐1996, and U‐266–1970. Four experiments were performed and one representative is shown. (D) Plasma cells were purified from bone marrow samples of 10 patients with MM and treated for 72 h with indicated concentrations of PPP. At harvest the relative number of viable cells was analyzed using the resazurin assay. (E) Alternatively, primary MM cells from four patients were allowed to adhere to BMSCs before treatment with PPP for 72 h. By using 3H‐TdR, the amount of DNA synthesis was quantified where 3H‐TdR incorporation of the MM cells was calculated as described in Materials and Methods. The experiments were performed in triplicate and data presented as mean SD. Error bars not visible are included within the symbols. (F) BMSC and SK1069 fibroblasts were treated with PPP at 1 mM for 72 h. At harvest, the relative number of viable cells was analyzed using the resazurin assay. Three experiments were performed and one representative is shown where data is presented as mean SD. Originally published in Stromberg et al. (2006). Copyright American Society of Hematology, used with permission.

155

IGF‐IR Signaling in Multiple Myeloma 150

A

Relative number of viable cells (%)

125 100 75 50 25 0 0.1

0.2 0.3 PPP (µM)

0.4

25

0

C

Relative number of viable cells (%)

0.1

0.2 0.3 PPP (µM)

0.4

0.5

125

100

75

D 125

100

75

In su lin

C on tro l

In su lin

IG FII Lo ng R 3IG FI

IG FI

50 IG FII Lo ng R 3IG FI

Relative number of viable cells (%)

50

150

C on tro l

125

E

Relative number of viable cells (%)

Relative number of viable cells (%)

75

0.5

50

100

75

50

25

0

F 100

75

50

25

0 0

0.1

0.2 0.3 PPP (µM)

0.4

0.5

0

125

0.1

0.2 0.3 PPP (µM)

0.4

0.5

0.1

0.2 0.3 PPP (µM)

0.4

0.5

125

G

Relative number of viable cells (%)

Relative number of viable cells (%)

100

0 0

150

125

B

125

IG FI

Relative number of viable cells (%)

150

100

75

50

25

0

H 100

75

50

25

0 0.1

0.2 0.3 PPP (µM) 200 Relative number of viable cells (%)

0

0.4

0.5

0

I

175 150 125 100 75 50 0

0.01 0.02 0.1 0.2 1 2 IL-6 (ng/mL)

10

20 100

Fig. 6 Effects of IGFs, insulin, and IL‐6 on growth inhibition induced by PPP. (A) Karpas 707 and (B) RPMI 8226 cells were treated for 48 h with PPP alone (open squares) or in the presence of IGF‐I

156

Helena Jernberg‐Wiklund and Kenneth Nilsson Paraprotein **

*

4

(g/dl)

3 2 1 0 Naive

C

Vehicle

Low PPP

*

60 40 20 0 Naive

Vehicle

Low PPP

High PPP

Liver weight 2.5

*

**

2 Weight (g)

1 Weight (g)

80

D **

**

*

100

High PPP

Spleen weight

1.2

Idiotype staining

B Myeloma cells in BM (%)

A

0.8 0.6 0.4

1.5 1 0.5

0.2

0

0 Naive

Vehicle

Low PPP

MVD

E

Naive

High PPP

Vehicle

Low PPP

High PPP

Microvessel density

35 30 25 20 15 10 5 0

***

Naive

Vehicle

****

Low PPP

High PPP

Fig. 7 Effect of PPP in vivo. (A) Serum paraprotein concentrations as determined by serum electrophoresis. (B) Tumor load as determined by flow cytometric analysis. Data are expressed as percentage 5T33MM cells of total cell number. (C–D) Weight of spleen and liver in grams of naive and treated or untreated 5T33MM‐bearing mice. (E) Microvessel density. The number of microvessels in the tibiae and femora of the mice counted by CD31 staining. Mean values SD for groups of 10 mice are shown.*p > 0.05; **p < 0.001; ***p < 0.05; ****p < 0.02. Originally published in Menu et al. (2006). Copyright American Society of Hematology, used with permission.

(closed circles), IGF‐2 (open circles), long R3‐IGF‐I (closed triangles), or insulin (open triangles). In control cultures (C) Karpas 707 and (D) RPMI 8226 cells were treated for 72 h with the indicated ligands alone (open bars) or with the addition of 1 mM dexamethasone (filled bars). Each ligand was added at the concentration 100 nM and always 1 h before treatment with PPP/dexamethasone was initiated. The four IL‐6‐dependent/responsive MM cell lines (A) U‐1957, (B) U‐1958, (C) U‐1996, and (D) U‐266–1970 were treated for 48 h with PPP alone (open squares) or in the presence of IL‐6 at 10 ng/mL (open circles) or 100 ng/mL (closed circles). IL‐6 was added 1 h before treatment with PPP, at indicated concentrations, was initiated. (E) As a control experiment the IL‐6‐ dependent U‐1958 cell line was treated with increasing concentrations of IL‐6. Three independent experiments were performed in triplicate and one representative is shown. Data are presented as mean SD. Error bars not visible are included within the symbols. Originally published in Stromberg et al. (2006). Copyright American Society of Hematology, used with permission.

157

IGF‐IR Signaling in Multiple Myeloma

1

Cum. survlval

0.8 : Vehicle 0.6

: Untreated control : PPP 50 mM

0.4

: Naive

0.2 0 0

5

10

15 20 Time (s)

25

30

35

Fig. 8 Effect of PPP on disease‐free survival. Mice were either untreated or treated with 50 mM PPP or the vehicle (DMSO/oil, 9:1) from the day of injection with 5T33MM onward. The first day of onset of morbidity was in the vehicle group on the 17th day. All naive mice were killed on the last day. The PPP group lived 10 days longer (p < 0.001 between vehicle group and PPP group). Originally published in Menu et al. (2006). Copyright American Society of Hematology, used with permission.

dexamethasone revealed the inhibition of mTOR‐ or MEK/ERK‐specific phosphorylation of the kinase p70S6K at a unique and common activating sites as well as down‐regulation of the cell cycle‐regulated protein cyclin D2 and D3 (Stromberg et al., 2004). In line with the findings from this rational combinatorial treatment targeting multiple substrates in MM survival pathways, the IGF‐IR inhibitor PPP was proven to sensitize MM cells to both mTOR inhibitor rapamycin and the p38 inhibitor SB205380 (Fig. 9). Apart from a rational combinatorial treatment regimen, IGF‐IR blockade might also prove to potentiate various conventional or novel therapies. So far, we have shown that pretreatment with PPP significantly sensitized MM cells to dexamethasone, doxorubicin, and melphalan (Fig. 9). However, as concluded from our data, caution must be taken in the sequential addition using the IGF‐IRTK inhibitor. A schedule‐dependent sensitization effect of PPP pretreatment was observed in vitro, while the effect was diminished when adding doxorubicin and melphalan prior or simultaneously with PPP. Studies on whether PPP will be useful as single agent or in combinatorial regimens are currently underway using extensive drug libraries and drugs used in novel and conventional therapy. In addition to the drugs listed herein, there is ample evidence that increased IGF‐IR signaling may indeed be the underlying mechanism to resistance to compounds targeting other receptors and associated kinases, for example ErbB2 (Camirand et al., 2002; Pollak et al., 2004). Therefore, PPP should also be regarded as potentially useful in sensitizing MM cells to other selective kinase inhibitors. In summary, PPP constitutes an attractive substance with selectivity for the IGF‐IR over the IR at the cellular level. It exerts antitumor activity in MM on established cell lines and primary MM cells in vitro and in vivo as determined

158

Helena Jernberg‐Wiklund and Kenneth Nilsson

125 Relative number of viable cells (%)

A 100 -

75

Dexamethasone 1 µM Doxorubicin 1.5 µg/ml

50

Melphalan 15 µg/ml 25

0 Control

PPP 0.3 µM

125 Relative number of viable cells (%)

B 100

-

75

Rapamycin 20 nM 50

SB203580 20 µM

25

0 Control

PPP 0.3 µM

Fig. 9 Effects of PPP in combination with cytotoxic drugs and pharmacological inhibitors. (A) RPMI 8226 cells were incubated for 24 h with 0.3 mM PPP before treatment with dexamethasone, doxorubicin, and melphalan for another 24 h followed by analysis using the resazurin assay. (B) RPMI 8226 cells were treated simultaneously with PPP and rapamycin or SB203580 for 48 h followed by analysis using the resazurin assay. Three independent experiments were performed in triplicate and one representative is shown. Data are presented as mean SD. Error bars not visible are included within the symbols (p < 0.001). Originally published in Stromberg et al. (2006). Copyright American Society of Hematology, used with permission.

by tumor burden and increased overall survival in the 5TMM mouse model. PPP is therapeutically potent and lacks major cytotoxic side zeffects. PPP may be proven useful for single drug treatment and, if synergistically activity with other kinase inhibitors will be found, in rational combinatorial drug regimens.

IGF‐IR Signaling in Multiple Myeloma

159

ACKNOWLEDGMENTS This work was supported by grants from the Swedish Cancer Society, Go¨ran Gustafssons Stiftelse, the Multiple MM Research Foundation (MMRF), Hans von Kantzows and Magnus Bergwalls Stiftelse, Selanders Stiftelse, and the Swedish Research Council.

REFERENCES Adams, T. E., Epa, V. C., Garrett, T. P. J., and Ward, C. W. (2000). Structure and function of the type 1 insulin‐like growth factor receptor. Cell. Mol. Life Sci. 57, 1050–1093. Asosingh, K., Gunthert, U., Bakkus, M. H., De Raeve, H., Goes, E., Van Riet, I., Van Camp, B., and Vanderkerken, K. (2000). In vivo induction of insulin‐like growth factor‐I receptor and CD44v6 confers homing and adhesion to murine multiple myeloma cells. Cancer Res. 60, 3096–3104. Barille, S., Collette, M., Bataille, R., and Amiot, M. (1995). Myeloma cells upregulate interleukin‐6 secretion in osteoblastic cells through cell‐to‐cell contact but downregulate osteocalcin. Blood 86, 3151–3159. Baserga, R. (2000). The contradictions of the insulin‐like growth factor 1 receptor. Oncogene 19, 5574–5581. Bataille, R., Robillard, N., Avet‐Loiseau, H., Harousseau, J. L., and Moreau, P. (2005). CD221 (IGF‐1R) is aberrantly expressed in multiple myeloma, in relation to disease severity. Haematologica 90, 706–707. Bergsagel, P. L., Chesi, M., Nardini, E., Brents, L. A., Kirby, S. L., and Kuehl, W. M. (1996). Promiscuous translocations into immunoglobulin heavy chain switch regions in multiple myeloma. Proc. Natl. Acad. Sci. USA 93, 13931–13936. Bergsagel, P. L., Kuehl, W. M., Zhan, F., Sawyer, J., Barlogie, B., and Shaughnessy, J., Jr. (2005). Cyclin D dysregulation: An early and unifying pathogenic event in multiple myeloma. Blood 106, 296–303. Bisping, G., Leo, R., Wenning, D., Dankbar, B., Padro, T., Kropff, M., Scheffold, C., Kroger, M., Mesters, R. M., Berdel, W. E., and Kienast, J. (2003). Paracrine interactions of basic fibroblast growth factor and interleukin‐6 in multiple myeloma. Blood 101, 2775–2783. Borset, M., Hjorth‐Hansen, H., Seidel, C., Sundan, A., and Waage, A. (1996). Hepatocyte growth factor and its receptor c‐met in multiple myeloma. Blood 88, 3998–4004. Calo, V., Migliavacca, M., Bazan, V., Macaluso, M., Buscemi, M., Gebbia, N., and Russo, A. (2003). STAT proteins: From normal control of cellular events to tumorigenesis. J. Cell. Physiol. 197, 157–168. Camirand, A., Lu, Y., and Pollak, M. (2002). Co‐targeting HER2/ErbB2 and insulin‐like growth factor‐1 receptors causes synergistic inhibition of growth in HER2‐overexpressing breast cancer cells. Med. Sci. Monit. 8, BR521–BR526. Carter, A., Merchav, S., Silvian‐Draxler, I., and Tatarsky, I. (1990). The role of interleukin‐1 and tumour necrosis factor‐alpha in human multiple myeloma. Br. J. Haematol. 74, 424–431. Catlett‐Falcone, R., Landowski, T. H., Oshiro, M. M., Turkson, J., Levitzki, A., Savino, R., Ciliberto, G., Moscinski, L., Fernandez‐Luna, J. L., Nunez, G., Dalton, W. S., and Jove, R. (1999). Constitutive activation of Stat3 signaling confers resistance to apoptosis in human U266 myeloma cells. Immunity 10, 105–115. Chenu, C., Valentin‐Opran, A., Chavassieux, P., Saez, S., Meunier, P. J., and Delmas, P. D. (1990). Insulin like growth factor I hormonal regulation by growth hormone and by 1,25

160

Helena Jernberg‐Wiklund and Kenneth Nilsson

(OH)2D3 and activity on human osteoblast‐like cells in short‐term cultures. Bone 11, 81–86. Chng, W. J., Gualberto, A., and Fonseca, R. (2005). IGF‐1R is overexpressed in poor‐ prognostic subtypes of multiple myeloma. Leukemia 20, 174–176. Corradini, P., Inghirami, G., Astolfi, M., Ladetto, M., Voena, C., Ballerini, P., Gu, W., Nilsson, K., Knowles, D. M., Boccadoro, M., et al. (1994). Inactivation of tumor suppressor genes, p53 and Rb1, in plasma cell dyscrasias. Leukemia 8, 758–767. Dimberg, L. Y., Dimberg, A. I., Ivarsson, K., Stromberg, T., Osterborg, A., Nilsson, K., Oberg, F., and Jernberg‐Wiklund, H. (2005). Ectopic and IFN‐induced expression of Fas overcomes resistance to Fas‐mediated apoptosis in multiple myeloma cells. Blood 106, 1346–1354. Dunn, S. E. (2000). Insulin‐like growth factor I stimulates angiogenesis and the production of vascular endothelial growth factor. Growth Horm. IGF Res. 10(Suppl. A), S41–S42. Egle, A., Villunger, A., Marschitz, I., Kos, M., Hittmair, A., Lukas, P., Grunewald, K., and Greil, R. (1997). Expression of Apo‐1/Fas (CD95), Bcl‐2, Bax and Bcl‐x in myeloma cell lines: Relationship between responsiveness to anti‐Fas mab and p53 functional status [see comments]. Br. J. Haematol. 97, 418–428. Evan, G., and Littlewood, T. (1998). A matter of life and cell death. Science 281, 1317–1322. Evan, G. I., Wyllie, A. H., Gilbert, C. S., Littlewood, T. D., Land, H., Brooks, M., Waters, C. M., Penn, L. Z., and Hancock, D. C. (1992). Induction of apoptosis in fibroblasts by c‐myc protein. Cell 69, 119–128. Favelyukis, S., Till, J. H., Hubbard, S. R., and Miller, W. T. (2001). Structure and autoregulation of the insulin‐like growth factor 1 receptor kinase. Nat. Struct. Biol. 8, 1058–1063. Federici, M., Porzio, O., Zucaro, L., Fusco, A., Borboni, P., Lauro, D., and Sesti, G. (1997). Distribution of insulin/insulin‐like growth factor‐I hybrid receptors in human tissues. Mol. Cell. Endocrinol. 129, 121–126. Ferlin, M., Noraz, N., Hertogh, C., Brochier, J., Taylor, N., and Klein, B. (2000). Insulin‐like growth factor induces the survival and proliferation of myeloma cells through an interleukin‐ 6‐independent transduction pathway. Br. J. Haematol. 111, 626–634. Frasca, F., Pandini, G., Scalia, P., Sciacca, L., Mineo, R., Costantino, A., Goldfine, I. D., Belfiore, A., and Vigneri, R. (1999). Insulin receptor isoform A, a newly recognized, high‐affinity insulin‐like growth factor II receptor in fetal and cancer cells. Mol. Cell. Biol. 19, 3278–3288. Freund, G. G., Kulas, D. T., and Mooney, R. A. (1993). Insulin and IGF‐1 increase mitogenesis and glucose metabolism in the multiple myeloma cell line, RPMI 8226. J. Immunol. 151, 1811–1820. Gazitt, Y., Fey, V., Thomas, C., and Alvarez, R. (1998). Bcl‐2 overexpression is associated with resistance to dexamethasone, but not melphalan, in multiple myeloma cells. Int. J. Oncol. 13, 397–405. Ge, N. L., and Rudikoff, S. (2000). Expression of PTEN in PTEN‐deficient multiple myeloma cells abolishes tumor growth in vivo. Oncogene 19, 4091–4095. Georgii‐Hemming, P., Wiklund, H. J., Ljunggren, O., and Nilsson, K. (1996). Insulin‐like growth factor I is a growth and survival factor in human multiple myeloma cell lines. Blood 88, 2250–2258. Georgii‐Hemming, P., Stromberg, T., Janson, E. T., Stridsberg, M., Wiklund, H. J., and Nilsson, K. (1999). The somatostatin analog octreotide inhibits growth of interleukin‐6 (IL‐6)‐dependent and IL‐6‐independent human multiple myeloma cell lines. Blood 93, 1724–1731. Girnita, A., Girnita, L., del Prete, F., Bartolazzi, A., Larsson, O., and Axelson, M. (2004). Cyclolignans as inhibitors of the insulin‐like growth factor‐1 receptor and malignant cell growth. Cancer Res. 64, 236–242.

IGF‐IR Signaling in Multiple Myeloma

161

Girnita, L., Girnita, A., and Larsson, O. (2003). Mdm2‐dependent ubiquitination and degradation of the insulin‐like growth factor 1 receptor. Proc. Natl. Acad. Sci. USA 100, 8247–8252. Govoni, K. E., Baylink, D. J., and Mohan, S. (2005). The multi‐functional role of insulin‐like growth factor binding proteins in bone. Pediatr. Nephrol. 20, 261–268. Hamilton, M. S., Barker, H. F., Ball, J., Drew, M., Abbot, S. D., and Franklin, I. M. (1991). Normal and neoplastic human plasma cells express bcl‐2 antigen. Leukemia 5, 768–771. Hata, H., Matsuzaki, H., Takeya, M., Yoshida, M., Sonoki, T., Nagasaki, A., Kuribayashi, N., Kawano, F., and Takatsuki, K. (1995). Expression of Fas/Apo‐1 (CD95) and apoptosis in tumor cells from patients with plasma cell disorders. Blood 86, 1939–1945. Hideshima, T., Bergsagel, P. L., Kuehl, W. M., and Anderson, K. C. (2004). Advances in biology of multiple myeloma: Clinical applications. Blood 104, 607–618. Jelinek, D. F., Witzig, T. E., and Arendt, B. K. (1997). A role for insulin‐like growth factor in the regulation of IL‐6‐responsive human myeloma cell line growth. J. Immunol. 159, 487–496; ISSN: 0022–1767. Jernberg, H., Pettersson, M., Kishimoto, T., and Nilsson, K. (1991). Heterogeneity in response to interleukin 6 (IL‐6), expression of IL‐6 and IL‐6 receptor mRNA in a panel of established human multiple myeloma cell lines [published erratum appears in Leukemia 1991 Jun;5(6): following 530]. Leukemia 5, 255–265. Jernberg‐Wiklund, H., and Nilsson, K. (2000). Plasma cell and myeloma cell lines. In “Human Cell Culture” (J. R. W. Masters and B. Palsson, Eds.), Vol. III, 3, pp. 81–155. Kluwer Academic Publishers, Dordrecht/Boston/London. Jernberg‐Wiklund, H., Pettersson, M., Carlsson, M., and Nilsson, K. (1992a). Increase in interleukin‐6 (IL‐6) and IL‐6 receptor expression in a human multiple myeloma cell line, U‐266, during long‐term in vitro culture and the development of a possible autocrine IL‐6 loop. Leukemia 6, 310–318. Jernberg‐Wiklund, H., Pettersson, M., Larsson, L. G., Anton, R., and Nilsson, K. (1992b). Expression of myc‐family genes in established human multiple myeloma cell lines: L‐myc but not c‐myc gene expression in the U‐266 myeloma cell line. Int. J. Cancer 51, 116–123. Jones, J. I., and Clemmons, D. R. (1995). Insulin‐like growth factors and their binding proteins: Biological actions. Endocr. Rev. 16, 3–34. Jourdan, M., Veyrune, J. L., Vos, J. D., Redal, N., Couderc, G., and Klein, B. (2003). A major role for Mcl‐1 antiapoptotic protein in the IL‐6‐induced survival of human myeloma cells. Oncogene 22, 2950–2959. Kawano, M., Hirano, T., Matsuda, T., Taga, T., Horii, Y., Iwato, K., Asaoku, H., Tang, B., Tanabe, O., Tanaka, H., et al. (1988). Autocrine generation and requirement of BSF‐2/IL‐6 for human multiple myelomas. Nature 332, 83–85. Kayali, A. G., Eichhorn, J., Haruta, T., Morris, A. J., Nelson, J. G., Vollenweider, P., Olefsky, J. M., and Webster, N. J. G. (1998). Association of the insulin receptor with phospholipase C‐gamma (PLCgamma) in 3T3‐L1 adipocytes suggests a role for PLCgamma in metabolic signaling by insulin. J. Biol. Chem. 273, 13808–13818. Klein, B., Zhang, X. G., Jourdan, M., Content, J., Houssiau, F., Aarden, L., Piechaczyk, M., and Bataille, R. (1989). Paracrine rather than autocrine regulation of myeloma‐cell growth and differentiation by interleukin‐6. Blood 73, 517–526. Kozma, L. M., and Weber, M. J. (1990). Constitutive Phosphorylation of the receptor for insulin‐like growth factor‐I in cells transformed by the Src oncogene. Mol. Cell. Biol. 10, 3626–3634. Kuehl, W. M., and Bergsagel, P. L. (2002). Multiple myeloma: Evolving genetic events and host interactions. Nat. Rev. Cancer 2, 175–187.

162

Helena Jernberg‐Wiklund and Kenneth Nilsson

Landowski, T. H., Shain, K. H., Oshiro, M. M., Buyuksal, I., Painter, J. S., and Dalton, W. S. (1999). Myeloma cells selected for resistance to CD95‐mediated apoptosis are not cross‐ resistant to cytotoxic drugs: Evidence for independent mechanisms of caspase activation. Blood 94, 265–274. Larsson, O., Girnita, A., and Girnita, L. (2005). Role of insulin‐like growth factor 1 receptor signalling in cancer. Br. J. Cancer 92, 2097–2101. LeRoith, D., and Roberts, J. C. T. (2003). The insulin‐like growth factor system and cancer. Cancer Lett. 195, 127–137. LeRoith, D., Werner, H., Beitner‐Johnson, D., and Roberts, C. T., Jr. (1995). Molecular and cellular aspects of the insulin‐like growth factor I receptor. Endocr. Rev. 16, 143–163. Li, W., Hyun, T., Heller, M., Yam, A., Flechner, L., Pierce, J. H., and Rudikoff, S. (2000). Activation of insulin‐like growth factor I receptor signaling pathway is critical for mouse plasma cell tumor growth. Cancer Res. 60, 3909–3915. Maehama, T., and Dixon, J. E. (1999). PTEN: A tumour suppressor that functions as a phospholipid phosphatase. Trends Cell Biol. 9, 125–128. Mahtouk, K., Hose, D., Reme, T., De Vos, J., Jourdan, M., Moreaux, J., Fiol, G., Raab, M., Jourdan, E., Grau, V., Moos, M., Goldschmidt, H., et al. (2005). Expression of EGF‐family receptors and amphiregulin in multiple myeloma. Amphiregulin is a growth factor for myeloma cells. . Oncogene 24, 3512–3524. Menu, E., Kooijman, R., Van Valckenborgh, E., Asosingh, K., Bakkus, M., Van Camp, B., and Vanderkerken, K. (2004). Specific roles for the PI3K and the MEK‐ERK pathway in IGF‐1‐ stimulated chemotaxis, VEGF secretion and proliferation of multiple myeloma cells: Study in the 5T33MM model. Br. J. Cancer 90, 1076–1083. Menu, E., Jernberg‐Wiklund, H., Stromberg, T., De Raeve, H., Girnita, L., Larsson, O., Axelson, M., Asosingh, K., Nilsson, K., Van Camp, B., and Vanderkerken, K. (2006). Inhibiting the IGF‐1 receptor tyrosine kinase with the cyclolignan PPP: An in vitro and in vivo study in the 5T33MM mouse model. Blood 107, 655–660. Mitsiades, C. S., and Mitsiades, N. (2005). Treatment of hematologic malignancies and solid tumors by inhibiting IGF receptor signaling. Expert Rev. Anticancer Ther. 5, 487–499. Mitsiades, C. S., Mitsiades, N. S., McMullan, C. J., Poulaki, V., Shringarpure, R., Akiyama, M., Hideshima, T., Chauhan, D., Joseph, M., Libermann, T. A., Garcia‐Echeverria, C., Pearson, M. A., et al. (2004). Inhibition of the insulin‐like growth factor receptor‐1 tyrosine kinase activity as a therapeutic strategy for multiple myeloma, other hematologic malignancies, and solid tumors. Cancer Cell 5, 221–230. Nagata, S. (1997). Apoptosis by death factor. Cell 88, 355–365. Nemunaitis, J., Andrews, D. F., Mochizuki, D. Y., Lilly, M. B., and Singer, J. W. (1989). Human marrow stromal cells: Response to interleukin‐6 (IL‐6) and control of IL‐6 expression. Blood 74, 1929–1935. Nilsson, K. (1977). Established cell lines as tools in the study of human lymphoma and myeloma cell characteristics. Hamatol. Bluttransfus. 20, 253–264. Nilsson, K. (1994). The control of growth in human multiple myeloma cell lines. In “Multiple Myeloma and Related Disorders” (B. Barlogie and F. Dammacco, Eds.), pp. 83–92. Ares‐ Serano Symposia Publications, Rome, Italy. Nilsson, K., Bennich, H., Johansson, S., and Ponte´n, J. (1970). Established immunoglobulin producing myeloma (IgE) and lymphoblastoid (IgG) cell line from an IgE myeloma patient. Clin. Exp. Immunol. 7, 477–489. Nilsson, K., Jernberg, H., and Pettersson, M. (1990). IL‐6 as a growth factor for human multiple myeloma cells—a short overview. Curr. Top. Microbiol. Immunol. 166, 3–12.

IGF‐IR Signaling in Multiple Myeloma

163

Nilsson, K., Georgii‐Hemming, P., Spets, H., and Jernberg‐Wiklund, H. (1999). The control of proliferation, survival and apoptosis in human multiple myeloma cells in vitro. Curr. Top. Microbiol. Immunol. 246, 325–332; discussion 333. Ogawa, M., Nishiura, T., Oritani, K., Yoshida, H., Yoshimura, M., Okajima, Y., Ishikawa, J., Hashimoto, K., Matsumura, I., Tomiyama, Y., and Matsuzawa, Y. (2000). Cytokines prevent dexamethasone‐induced apoptosis via the activation of mitogen‐activated protein kinase and phosphatidylinositol 3‐kinase pathways in a new multiple myeloma cell line. Cancer Res. 60, 4262–4269. Pettersson, M., Jernberg‐Wiklund, H., Larsson, L. G., Sundstrom, C., Givol, I., Tsujimoto, Y., and Nilsson, K. (1992). Expression of the bcl‐2 gene in human multiple myeloma cell lines and normal plasma cells. Blood 79, 495–502. Pollak, M. N., Schernhammer, E. S., and Hankinson, S. E. (2004). Insulin‐like growth factors and neoplasia. Nat. Rev. Cancer 4, 505–518. Potter, M., and Wiener, F. (1992). Plasmacytomagenesis in mice: Model of neoplastic development dependent upon chromosomal translocations. Carcinogenesis 13, 1681–1697. Radl, J. (1994). The C57BL/KaLwRij mouse model in the studies of monoclonal gammapathies. In “Multiple Myeloma and Related Disorders” (F. Dammacco and B. Barlogie, Eds.), pp. 11–17. Ares‐Serono Symposia, Rome. Radl, J., De Glopper, E. D., Schuit, H. R., and Zurcher, C. (1979). Idiopathic paraproteinemia. II. Transplantation of the paraprotein‐producing clone from old to young C57BL/KaLwRij mice. J. Immunol. 122, 609–613. Scaffidi, C., Fulda, S., Srinivasan, A., Friesen, C., Li, F., Tomaselli, K. J., Debatin, K. M., Krammer, P. H., and Peter, M. E. (1998). Two CD95 (APO‐1/Fas) signaling pathways. EMBO J. 17, 1675–1687. Scaffidi, C., Kischkel, F. C., Krammer, P. H., and Peter, M. E. (2000). Analysis of the CD95 (APO‐1/Fas) death‐inducing signaling complex by high‐resolution two‐dimensional gel electrophoresis. Methods Enzymol. 322, 363–373. Sciacca, L., Costantino, A., Pandini, G., Mineo, R., Frasca, F., Scalia, P., Sbraccia, P., Goldfine, I. D., Vigneri, R., and Belfiore, A. (1999). Insulin receptor activation by IGF‐II in breast cancers: Evidence for a new autocrine/paracrine mechanism. Oncogene 18, 2471–2479. Shain, K. H., Landowski, T. H., Buyuksal, I., Cantor, A. B., and Dalton, W. S. (2000). Clonal variability in CD95 expression is the major determinant in Fas‐medicated, but not chemotherapy‐medicated apoptosis in the RPMI 8226 multiple myeloma cell line. Leukemia 14, 830–840. Siddle, K., Urso, B., Niesler, C. A., Cope, D. L., Molina, L., Surinya, K. H., and Soos, M. A. (2001). Specificity in ligand binding and intracellular signalling by insulin and insulin‐like growth factor receptors. Biochem. Soc. Trans. 29, 513–525. Spets, H., Georgii‐Hemming, P., Siljason, J., Nilsson, K., and Jernberg‐Wiklund, H. (1998). Fas/APO‐1 (CD95)‐mediated apoptosis is activated by interferon‐gamma and interferon‐ in interleukin‐6 (IL‐6)‐dependent and IL‐6‐independent multiple myeloma cell lines. Blood 92, 2914–2923. Spets, H., Stromberg, T., Georgii‐Hemming, P., Siljason, J., Nilsson, K., and Jernberg‐Wiklund, H. (2002). Expression of the bcl‐2 family of pro‐ and anti‐apoptotic genes in multiple myeloma and normal plasma cells: Regulation during interleukin‐6(IL‐6)‐induced growth and survival. Eur. J. Haematol. 69, 76–89. Standal, T., Borset, M., Lenhoff, S., Wisloff, F., Stordal, B., Sundan, A., Waage, A., and Seidel, C. (2002). Serum insulinlike growth factor is not elevated in patients with multiple myeloma but is still a prognostic factor. Blood 100, 3925–3929.

164

Helena Jernberg‐Wiklund and Kenneth Nilsson

Stephanou, A., Isenberg, D. A., Nakajima, K., and Latchman, D. S. (1999). Signal transducer and activator of transcription‐1 and heat shock factor‐1 interact and activate the transcription of the Hsp‐70 and Hsp‐90beta gene promoters. J. Biol. Chem. 274, 1723–1728. Stromberg, T., Dimberg, A., Hammarberg, A., Carlson, K., Osterborg, A., Nilsson, K., and Jernberg‐Wiklund, H. (2004). Rapamycin sensitizes multiple myeloma cells to apoptosis induced by dexamethasone. Blood 103, 3138–3147. Stromberg, T., Ekman, S., Girnita, L., Dimberg, L. Y., Larsson, O., Axelson, M., Lennartsson, J., Hellman, U., Carlson, K., Osterborg, A., Vanderkerken, K., Nilsson, K., et al. (2006). IGF‐1 receptor tyrosine kinase inhibition by the cyclolignan PPP induces G2/M‐phase accumulation and apoptosis in multiple myeloma cells. Blood 107, 669–678. Surmacz, E. (2003). Growth factor receptors as therapeutic targets: Strategies to inhibit the insulin‐like growth factor I receptor. Oncogene 22, 6589–6597. Tassone, P., Neri, P., Carrasco, D. R., Burger, R., Goldmacher, V. S., Fram, R., Munshi, V., Shammas, M. A., Catley, L., Jacob, G. S., Venuta, S., Anderson, K. C., et al. (2005). A clinically relevant SCID‐hu in vivo model of human multiple myeloma. Blood 106, 713–716. Tu, Y., Renner, S., Xu, F., Fleishman, A., Taylor, J., Weisz, J., Vescio, R., Rettig, M., Berenson, J., Krajewski, S., Reed, J. C., and Lichtenstein, A. (1998). BCL‐X expression in multiple myeloma: Possible indicator of chemoresistance. Cancer Res. 58, 256–262. Uchiyama, H., Barut, B. A., Mohrbacher, A. F., Chauhan, D., and Anderson, K. C. (1993). Adhesion of human myeloma‐derived cell lines to bone marrow stromal cells stimulates interleukin‐6 secretion. Blood 82, 3712–3720. Valentinis, B., and Baserga, R. (2001). IGF‐I receptor signalling in transformation and differentiation. Mol. Pathol. 54, 133–137. van de Donk, N. W. C. J., Lokhorst, H. M., and Bloem, A. C. (2005). Growth factors and antiapoptotic signaling pathways in multiple myeloma. Leukemia 19, 2177–2185. Vanderkerken, K., Asosingh, K., Braet, F., Van Riet, I., and Van Camp, B. (1999). Insulin‐like growth factor‐1 acts as a chemoattractant factor for 5T2 multiple myeloma cells. Blood 93, 235–241. Vanderkerken, K., Van Camp, B., De Greef, C., VandeBroek, I., Asosingh, K., and Van Riet, I. (2000). Homing of the myeloma cell clone. Acta Oncol. 39, 771–776. Vanderkerken, K., Asosingh, K., Croucher, P., and Van Camp, B. (2003). Multiple myeloma biology: Lessons from the 5TMM models. Immunol. Rev. 194, 196–206. Vasilcanu, D., Girnita, A., Girnita, L., Vasilcanu, R., Axelson, M., and Larsson, O. (2004). The cyclolignan PPP induces activation loop‐specific inhibition of tyrosine phosphorylation of the insulin‐like growth factor‐1 receptor. Link to the phosphatidyl inositol‐3 kinase/Akt apoptotic pathway. Oncogene 23, 7854–7862. Villunger, A., Egle, A., Marschitz, I., Kos, M., Bock, G., Ludwig, H., Geley, S., Kofler, R., and Greil, R. (1997). Constitutive expression of Fas (Apo‐1/CD95) ligand on multiple myeloma cells: A potential mechanism of tumor‐induced suppression of immune surveillance. Blood 90, 12–20. Wang, Y. D., De Vos, J., Jourdan, M., Couderc, G., Lu, Z. Y., Rossi, J. F., and Klein, B. (2002). Cooperation between heparin‐binding EGF‐like growth factor and interleukin‐6 in promoting the growth of human myeloma cells. Oncogene 21, 2584–2592. Westendorf, J. J., Lammert, J. M., and Jelinek, D. F. (1995). Expression and function of Fas (APO‐1/CD95) in patient myeloma cells and myeloma cell lines. Blood 85, 3566–3576. Xu, F., Gardner, A., Tu, Y., Michl, P., Prager, D., and Lichtenstein, A. (1997). Multiple myeloma cells are protected against dexamethasone‐induced apoptosis by insulin‐like growth factors. Br. J. Haematol. 97, 429–440.

IGF‐IR Signaling in Multiple Myeloma

165

Xu, F. H., Sharma, S., Gardner, A., Tu, Y., Raitano, A., Sawyers, C., and Lichtenstein, A. (1998). Interleukin‐6‐induced inhibition of multiple myeloma cell apoptosis: Support for the hypothesis that protection is mediated via inhibition of the JNK/SAPK pathway. Blood 92, 241–251. Yaccoby, S., Barlogie, B., and Epstein, J. (1998). Primary myeloma cells growing in SCID‐hu mice: A model for studying the biology and treatment of myeloma and its manifestations. Blood 92, 2908–2913. Zhang, W., Zong, C. S., Hermanto, U., Lopez‐Bergami, P., Ronai, Z. E., and Wang, L.‐H. (2006). RACK1 recruits STAT3 specifically to insulin and insulin‐like growth factor 1 receptors for activation, which is important for regulating anchorage‐independent growth. Mol. Cell. Biol. 26, 413–424. Zong, C. S., Chan, J., Levy, D. E., Horvath, C., Sadowski, H. B., and Wang, L.‐H. (2000). Mechanism of STAT3 activation by insulin‐like growth factor I receptor. J. Biol. Chem. 275, 15099–15105.

This page intentionally left blank

c‐MYC Impairs Immunogenicity of Human B Cells Martin Schlee,*,1 Marino Schuhmacher,*,{ Michael Ho¨lzel,*,2 Gerhard Laux,* and Georg W. Bornkamm* *Institute of Clinical Molecular Biology and Tumor Genetics, GSF‐National Research Center for Environment and Health, D‐81377 Mu¨nchen, Germany; { GPC Biotech‐AG, D‐82152 Martinsried, Germany

I. Introduction II. Personal Perspective by G.W.B. III. Taking Over the Work from Eva and George . . . A. Growth Pattern and Cell Surface Phenotype of Burkitt’s Lymphoma Cells Can Be Recapitulated by c‐myc Overexpression in Primary Human B Cells Conditionally Immortalized by EBV B. Conditional B Cells Driven into Proliferation by c‐myc Overexpression Loose Their Ability to Stimulate Allogeneic T Cells and Become Invisible to Cytotoxic T Cells C. c‐MYC Down‐Regulates NF‐B and Interferon Response Genes D. c‐MYC Impairs the Interferon Response at Different Levels: At the Level of Induction as Well as at the Level of Action of Type I Interferons IV. Discussions References

Deregulation of c‐myc expression through chromosomal translocation is essential in the pathogenesis of Burkitt’s lymphoma (BL). A characteristic feature of BL cells, compared to Epstein–Barr Virus (EBV)‐immortalized B cells, is their lack of immunogenicity. To study the contribution of EBV genes and of the c‐MYC protein to this phenotype, we have generated a conditional B cell system in which the viral proliferation program and expression of c‐myc can be regulated independently of each other. In cells proliferating due to exogenous c‐myc overexpression, the cell surface phenotype, the pattern of proliferation in single cell suspension, and the immunological characteristics of BL cells could be completely recapitulated. Yet, it had remained open whether nonimmunogenicity is the default phenotype when EBNA2 and LMP1 are switched off, or whether c‐MYC actively contributes to immunosuppression. We provide evidence also for the latter by showing that c‐MYC down‐regulates genes of the NF‐B and interferon pathway in a dose‐dependent fashion. c‐MYC acts at at least two different levels, the level of interferon induction as well as at the level of action of type I and type II 1

Present address: Division of Clinical Pharmacology, Department of Internal Medicine I, University of Bonn, D‐53127 Bonn, Germany. 2 Present address: Division of Molecular Carcinogenesis and Centre for Biomedical Genetics, The Netherlands Cancer Institute, NL‐1066 CX Amsterdam, The Netherlands. Advances in CANCER RESEARCH Copyright 2007, Elsevier Inc. All rights reserved.

167

0065-230X/07 $35.00 DOI: 10.1016/S0065-230X(06)97007-9

Martin Schlee et al.

168

interferons on their respective target promoters. c‐MYC does not block the interferon pathway completely, it shifts the balance and increases the threshold of interferon induction and action. # 2007 Elsevier Inc.

I. INTRODUCTION Burnet’s immunosurveillance hypothesis of cancer is still strongly debated (Dunn et al., 2004; Qin and Blankenstein, 2004; Willimsky and Blankenstein, 2005). It assumes that the immune system has an inherent ability to recognize evolving mutations in tumor cells and would thereby destroy them in a premalignant state. Eva and George Klein have pointed out that immunosuppression leads to a strong increase in the incidence of virus‐associated tumors, yet, the frequency of most common cancers, for example lung, breast, colon, and prostate cancers, is not affected (Klein and Klein, 1977). This suggests that at least for most virus‐associated cancers, immunosurveillance plays indeed an important role whereas its role remains elusive for the majority of human tumors. Tumor immune escape is generally believed to be part of Darwinian selection taking place during evolution of a tumor in vivo. But this may not be the only scenario. It is also conceivable that at least in some instances immune escape might be coupled to activation of a cellular protooncogene. We have addressed this issue taking human EBV‐associated endemic Burkitt’s lymphoma (BL) as a model and provide evidence that c‐MYC directly impairs key regulators of cell immunogenicity. The three topics, EBV, the c‐myc protooncogene activated by chromosomal translocation into one of the immunoglobulin gene loci, and the lack of immunogenicity of BL cells (or in more general terms, the rejectability vs nonrejectability of tumors), have been central issues in the work of Eva and George over the last decades, and our work is based to a large extent on the ground that they have laid. Before we go in medias res, one of us (G.W.B.) is giving a personal perspective of the field and how his scientific career has been connected to and coined by Eva and George.

II. PERSONAL PERSPECTIVE BY G.W.B. After finishing my medical education I moved into the field of tumor virology in 1972, fascinated by the advent of modern molecular biology and the beauty and power of bacterial and phage genetics. As the logical extension of phage genetics, tumor viruses with their limited number of genes promised to be the gate opener for cancer research. I joined the laboratory of Harald zur Hausen at Erlangen who had worked as a postdoc with Gertrud and Werner Henle in Philadelphia and, back in Germany, had purified Epstein–Barr virus (EBV) and the viral DNA from virus‐producing BL cell

c‐MYC Impairs B Cell Immunogenicity

169

lines. Using in vivo 3H‐labeled viral DNA, Harald succeeded to detect EBV DNA in specimens from African BL and nasopharyngeal carcinoma (NPC) in collaboration with George Klein, Gertrud and Werner Henle, and Peter Clifford (zur Hausen et al., 1970). During my early years at Erlangen, tumor specimens from Stockholm arrived once a week. They had been shipped to Stockholm from Africa by Peter Clifford the day before, were portionated at Stockholm, and sent out again on the same evening to the collaborating partners all over the world. George and Eva have thus accompanied me from the very beginning of my scientific career. The question addressed at this time was whether EBV is present in the epithelial or lymphoid cells of nasopharyngeal carcinoma (also called Schmincke’s lymphoepithelioma). EBV was known to infect and immortalize B cells and the first guess was that EBV as a lymphotropic virus is present in the lymphoid cells, which represent a large part of this tumor. Three different experimental approaches were taken and all three converged in the finding that EBV is present in the malignant epithelial cells. Hans Wolf working with Harald as a student detected EBV DNA in epithelial cells but not in lymphoid cells of NPC by DNA in situ hybridization (Wolf et al., 1973). Claude Desgranges from Guy de The´’s laboratory at the International Agency for Research on Cancer at Lyon separated epithelial cells from lymphoid cells by cell separation techniques, came to our laboratory and confirmed Hans’ and Harald’s finding (Desgranges et al., 1975). George, not totally satisfied with these approaches, had started a collaboration with Giovanella who succeeded to grow the epithelial tumor cells in nude mice without contaminating lymphoid cells and provided definitive evidence that EBV is indeed confined to the epithelial cell compartment in this tumor (Klein et al., 1974). In 1975, I had the chance to work at the Karolinska Institute on the episomal nature of EBV DNA in BL cell lines and in the NPC tumor cells grown in nude mice by Giovanella together with Tomas Lindahl, Alice Adams, Christine Kaschka‐Dierich, Gunnar Bjursell, and George Klein. I remember very well not only how efficient this collaboration was (Kaschka‐Dierich et al., 1976; Lindahl et al., 1976), but also that we had a wonderful celebration of Eva’s and George’s 50th birthday during my stay in Stockholm. In the early 1970s, only the endemic form of BL was known as a high‐ grade B cell malignancy that occurs with high frequency in the tropical areas of Africa and New Guinea and almost invariably carries viral DNA and expresses the nuclear antigen EBNA1 discovered in 1973 by Beverly Reedman and George Klein (Reedman and Klein, 1973; Reedman et al., 1974). The differences between EBV‐immortalized cells and BL cells only started to emerge from the work done in George’s and Eva’s as well as Kenneth Nilsson’s laboratory (Bechet et al., 1973, 1974). BL attracted the interest of people from many different areas: clinicians, pathologists, epidemiologists, virologists, immunologists, cell biologists, and cytogeneticists.

170

Martin Schlee et al.

Eva and George together combined virtually all this expertise. The most important discovery in the field, the description of the reciprocal balanced t(8;14) translocation in BL cells, was made by Manolov and Manolova (1972) who were working as guests of the Swedish Cancer Society in the laboratory of Albert Levan in Lund. Eva and George had started to collaborate with the Manolovs already a couple of years earlier and it is obvious from their joint publications that they have been deliberately searching for specific chromosomal aberrations in BL (Manolov et al., 1971a,b) until Manolov and Manolova were finally successful. Eva and George had the prepared mind to acknowledge the importance of cytogenetic aberrations in human cancer not at least due to the specific education that George had experienced in Sweden after the war. George had joined the laboratory of Torbjo¨rn O. Caspersson at Karolinska as a student in 1947 and closely followed the work of his former mentor over the years. Caspersson was the great pioneer of quantitative cytometry and developed chromosomal banding techniques in the 1960s, paving the way for modern (tumor) cytogenetics (Caspersson et al., 1968, 1970). Eva and George had noticed the specific dedication of the Manolovs, gave them the support they needed, and did a great job bringing the importance of their discovery to general attention. Nowadays, the role of chromosomal translocations in leukemias and lymphomas is text book knowledge. But this was quite different at the end of the 1960s and the beginning of the 1970s. Pathologists were well aware of the fact that most tumor cells carry cytogenetic aberrations; yet, most people believed that they arise as a consequence of malignant transformation during tumor progression rather than as initiating events. The Philadelphia (Ph) chromosome in chronic myeloid leukemia (CML) had been described by Nowell and Hungerford already in 1960 (Nowell and Hungerford, 1960), but the importance of the Ph chromosome has only come to general perception with the description of the recurrent t(8;14) chromosomal translocations in BL that provided a second independent example for the involvement of specific chromosomal translocations in the development of leukemias and lymphomas as a new paradigm in cancer research. Eva’s and George’s role in the discovery of the chromosomal translocations in BL sheds a characteristic light on their personalities: in addition to their important personal contributions to science, they have played as an important role in communicating, discussing, and promoting important developments in science and moving the field of cancer research forward. This is illustrated by George’s immense load of work as a restless letter writer (this is a unique feature of George) and communicator of ideas, by his conceptual reviews and commentaries, and last but not least, by his not ending work as an editor identifying the most important topics and concepts, and attracting the best in the field as contributors, for example, for Advances in Cancer Research.

c‐MYC Impairs B Cell Immunogenicity

171

Back to the late 1970s and early 1980s: having a cytogenetic marker at hand, a number of groups including Gilbert Lenoir’s, Ian Magrath’s, Miyoshi’s, George’s, and our own group have shown that BL is not restricted to Africa and New Guinea, but occurs with about 20‐fold lower incidence all over the world (Bornkamm et al., 1980; Douglass et al., 1980; Lenoir and Philip, 1979; Lenoir et al., 1979; Magrath et al., 1980; Miyoshi et al., 1978). The incidence of BL additionally increased in the Caucasian population with the AIDS epidemic (Kalter et al., 1985). Only 10–20% of the sporadic cases and about half of those in HIV‐infected patients turned out to be associated with EBV. Cytogenetic examination of these cases and reevaluation of the endemic cases revealed that the t(8;14) translocation is not the only translocation in BL (Bernheim et al., 1980, 1981; Bornkamm et al., 1980; Miyoshi et al., 1979, 1981). Chromosome 8 (band 8q24) is invariably involved and is fused to the long or short arm of chromosomes 14, 2, and 22 that carry the heavy and light chain immunoglobulin loci, respectively. The importance of specific chromosomal translocations in B cell malignancies was further underlined by the collaborative work of Michael Potter’s and George Klein’s laboratories showing that mouse plasmacytoma (MPC) is characterized by similar recurrent chromosomal translocations involving the immunoglobulin gene loci (Ohno et al., 1979). The identification of the c‐myc protooncogene at the breakpoint of the chromosomal translocations in BL and MPC juxtaposed to one of the Ig loci by Riccardo Dalla‐Favera et al. and Grace Shen‐Ong et al. in 1982 (Dalla‐Favera et al., 1982; Shen‐Ong et al., 1982) and of the c‐abl protooncogene at the breakpoint of the Ph chromosome by de Klein et al. (de Klein et al., 1982) have bridged the gap between classical cytogenetics, molecular biology, and tumor virology for the first time and has also provided a molecular explanation for the mechanism of oncogene activation either by gene dysregulation (BL) or by gene to gene fusion (CML). During the last decades, the concept of oncogene activation by chromosomal translocation has been verified for a large number of malignancies and has gained general acceptance. Another issue that Eva and George have pursued throughout their scientific career is the recognition and rejectability of virus‐induced and spontaneously arising tumors. A particular highlight of this part of their work has been the first description of NK cells by Rolf Kiessling, Eva Klein, and Hans Wigzell in 1975 (Kiessling et al., 1975a,b). Pursuing the issue of cellular recognition of BL cells and EBV‐immortalized cells, Eva and George together with their collaborators Maria Masucci and Sibba Torsteindottir have shown that BL cells, in contrast to their EBV‐immortalized counterparts, express HLA class I antigens at low level, have poor antigen presenting capacity, and are unable to stimulate syngeneic or allogeneic T cells (Avila‐Carino et al., 1987; Frisan et al., 1996; Klein et al., 1986; Masucci et al., 1987; Torsteinsdottir et al., 1986). Similar observations have also been made

172

Martin Schlee et al.

by the groups of Alan Rickinson and Dennis Moss (Khanna et al., 1994; Rowe et al., 1995).

III. TAKING OVER THE WORK FROM EVA AND GEORGE . . . This is the point where we have taken over and continued the work in a wonderful collaboration with Maria Masucci, Riccardo Gavioli, and Teresa Frisan at Stockholm and Alan Rickinson and Steve Lee at Birmingham (Gavioli et al., 2001; Staege et al., 2002). We have addressed the question whether there is a link between c‐myc activation by chromosomal translocation and the poor immunogenicity of BL cells. Proliferation of EBV‐immortalized B cells is driven by the genes of the viral latency growth program (Bornkamm and Hammerschmidt, 2001), which is initiated and controlled by the viral transcription factor EBNA2 that induces viral and cellular target genes. The main players are two direct EBNA2 targets, the viral LMP1 and the cellular c‐myc gene, that, in contrast to its high expression in BL cells, is only moderately induced by EBNA2 in LCLs. LMP1 is a transmembrane protein mimicking signaling of constitutively cross‐linked CD40 and activates the NF‐B, AP‐1, MAP‐kinase, and STAT pathways. To gain insight into the changes imposed by different EBV genes and the cellular oncogene c‐myc, we have established in vitro model systems that recapitulate important features of EBV‐driven B cell immortalization and BL cells. Exploiting the possibility to modulate the c‐MYC expression level in a conditional cell line, we have provided evidence that c‐MYC is a negative regulator of NF‐B and interferon response genes, the genes playing a pivotal role determining the cell’s surface phenotype and antigen‐presenting capacity.

A. Growth Pattern and Cell Surface Phenotype of Burkitt’s Lymphoma Cells Can Be Recapitulated by c‐myc Overexpression in Primary Human B Cells Conditionally Immortalized by EBV To study the contribution of EBV genes and of the c‐myc protooncogene to growth pattern, cell surface phenotype, and immunogenicity of human B cells, we have established a conditional in vitro system in which the viral proliferation program and an exogenous c‐myc gene can be expressed independently of each other. To generate a conditionally EBV‐immortalized

173

c‐MYC Impairs B Cell Immunogenicity

cell line, we have coinfected primary human B cells with the nontransforming, EBNA2‐deficient P3HR1 virus and recombinant EBV encoding an estrogen receptor‐EBNA2 fusion protein. The cell line thus established (EREB2, estrogen‐regulated EBNA2) is a typical lymphoblastoid cell line, except that for proliferation and survival the cells are strictly dependent on estrogen (Kempkes et al., 1995). Withdrawal of estrogen leads to inactivation of EBNA2, proliferation arrest, and cell death within 4 days. Starting from EREB2 cells as recipient cells, two additional cell lines were generated by transfection with a constitutively active c‐myc gene (cell line A1) and a tetracycline‐regulatable c‐myc gene (P493), respectively (Fig. 1). The c‐myc‐transfected cells can proliferate in the absence of estrogen and adopt the growth pattern and cell

Resting B cell from umbilical cord blood

Hsp90

Non functional ER-EBNA2 (cytoplasmic) Estrogen

Virus particle ER-EBNA2

EBNA2 deficient EBV (P3HR1)

Functional ER-EBNA2 (nuclear)

c-myc-Igk

tet(off) c-myc Lymphoblastoid cell line EREB2 +Estrogen

−EBNA2

+EBNA2

P493 A1 "Burkitt like growth pattern and phenotype" Conditional Constitutive

+Tetracycline

−MYC

+MYC

Fig. 1 The cellular system described here is mimicking important features of the pathogenesis of Bukritt’s lymphoma in vitro. Primary B cells were infected with an EBNA2‐deficient virus harvested from HH514 cells and the EBNA2 defect complemented by recombinant EBV expressing an EBNA2‐estrogen receptor fusion protein (inset). This double infection resulted in outgrowth of an LCL (EREB2) that is dependent in its proliferation and survival on estrogen. EREB2 cells were transfected with an episomal vector expressing c‐myc at a very high constitutive level (A1 cells) or a tetracycline‐regulatable c‐myc gene (P493 cells). The construct driving c‐myc expression in A1 cells is shown in Fig. 2A. Constitutive or tetracycline‐regulated expression of c‐myc rendered A1 and P493 cells independent of estrogen.

174

Martin Schlee et al.

surface phenotype of BL cells (Fig. 2). Genes involved in adhesion and activation were expressed at much higher level in the parental EREB2 cells as compared to the c‐myc‐driven conditional cell lines A1 and P493 (Pajic et al., 2001). The same holds true for HLA class I and class II expression (Staege et al., 2002). A notable difference between BL cells on one hand and A1 and P493 cells on the other hand is the expression of BCL‐6. BCL‐6 is highly expressed in BL cells but is absent from A1 and EREB2 cells (Fig. 2B). This is compatible with the notion that BL cells are derived from germinal center B cells, whereas A1 and P493 cells originate from normal naı¨ve peripheral B cells that have neither encountered antigen nor passed through a germinal center.

B. Conditional B Cells Driven into Proliferation by c‐myc Overexpression Loose Their Ability to Stimulate Allogeneic T Cells and Become Invisible to Cytotoxic T Cells It has been shown by Eva and George Klein and their collaborators that EBV converted BL lines are potent stimulators of allogeneic T cell proliferation in contrast to their parental EBV‐negative counterparts and that 5AzaC treatment of EBV‐positive group I BL lines, but not of EBV‐negative BL lines, increases their allostimulatory potential (Cuomo et al., 1993). As shown in Fig. 3A, conditional B cells proliferating on a c‐myc‐driven proliferation program have completely lost their ability to stimulate allogeneic T cells. Switching P493 cells back to the EBV‐mediated proliferation program by the addition of tetracycline (switching off the exogenous c‐myc gene) and estrogen (activation of EBNA2) restored the capacity of the cells to stimulate proliferation of allogeneic T cells. BL cells are characterized not only by the lack of stimulatory capacity in a mixed lymphocyte reaction (MLR), they are also not recognized by antigen‐ specific cytotoxic T cells if the respective antigen is expressed in the cells. This is a very important feature of BL cells as the viral antigens expressed in EBV‐ positive BLs are foreign antigens and should therefore be easily recognized. EBNA1 is expressed in all EBV‐positive BLs. EBNA1 has long been regarded as a nonimmunogenic protein due to the presence of a gly‐ala repeat that impairs proteasomal degradation (Levitskaya et al., 1995). There is, however, recent evidence which indicates that this effect is not complete and that EBNA1 can indeed be recognized by cytotoxic T cells in the context of LCLs but not of BL cells. Most importantly, a minority of endemic BLs was shown to express also the highly immunogenic proteins EBNA3A, ‐3B, and ‐3C, and yet, the cells cannot be recognized by HLA class I‐restricted cytotoxic T cells

A

B Ig-kappa locus MAR Ei V

J

E3⬘ CD10 CD38 CD77 IgM

C

BL-myc c-myc

Plus

Estrogen

0

BCL-6

Western blot

Minus

Time after-induction P

133 93 67 56 42

MHC class I low MHC class II low

MHC class I high MHC class II high

C

kD

BL 2 BL 41 BL 70 BL 29 Ak at EL a I A1 ER EB Ra 2 ji

CD21 CD23 CD39 CD40 CD54 CD58 CD80 CD86 CD95

A1 cells

20⬘ 40⬘ 1 h 2 h 3 h 4 h 5 h 6 h 8 h 24 h 48 h

Western blot

D Cell number × 105 cells

EREB2 cells

20

+c-MYC

15 10 -c-MYC

5 0

1

2

3 Days

4

5

Fig. 2 Phenotype and growth, pattern of A1 and the parental cell line EREB2 (A). Switching off c‐myc expression by addition of tetracycline induces cell cycle arrest in P493 cells, and withdrawal of tetracycline allows the cells to reenter the cell cycle and to resume proliferation (C, D). BCL‐6 is not expressed in A1 and EREB2 cells (B).

T cell stimulation index

120

ce ER lls EB 2

1

c-MYC EBNA2

80

b -2m

LMP1 Lmp2

40

B49 A11 B7

Lmp7 PA28a

0 EBNA2 c-MYC

EREB2

A1

P493

+ −

− +

− +

PA28b

P493 P493 + −

TAP1

− −

TAP2 C

B 40

40

% specific lysis

% specific lysis

E

A

ER E A B2 1 ce lls

D

A

30 20 10 0

30 20 10

+ − EREB2 cells

+ − A1 cells

0

EBNA3A Vaccinia virus

+ − EREB2 cells

+ − A1 cells

Peptide

Fig. 3 Immune recogition of A1, P493, and EREB2 cells. A1 and P493 cells have lost the stimulatory capacity for allogeneic T cells in an MLR when they are proliferating as a consequence of c‐myc overexpression. Activating EBNA2 and switching off c‐myc expression in P493 cells by addition of

c‐MYC Impairs B Cell Immunogenicity

177

(Kelly et al., 2002; Lee et al., 2004). To address the question whether this important property of BL cells is also reconstructed in the conditional B cell system, the parental EREB2 and A1 cells were infected with a recombinant vaccinia virus expressing the EBNA3A protein of EBV type 1. The use of an EBV antigen as a target antigen is possible because the virus used for generating the parental EREB2 cell line is of type 2 origin (P3HR1 virus) (Zimber et al., 1986) (except for ER‐EBNA2 that complemented the EBNA2 deficiency of P3HR1 virus), and cannot be recognized by a cytotoxic T cell clone specific for an epitope of EBNA3A of EBV type 1. Ectopic expression of EBV type 1 EBNA3A in the parental EREB2 cell line by a recombinant vaccinia virus led to efficient T cell killing, whereas recombinant vaccinia virus‐ infected A1 cells were completely resistant (Fig. 3B) despite equal vaccinia virus‐mediated EBNA3A expression in EREB2 and A1 cells (data not shown). An obvious possibility for the lack of T cell recognition in A1 cells was the low HLA class I expression level. This could be excluded as a cause for nonrecognition by a peptide rescue experiment. Addition of the peptide restored T cell killing to about 70% indicating that HLA class I expression is not the limiting factor for T cell recognition and killing (Fig. 3C). If the antigen is expressed but not recognized, there must be a defect in antigen degradation, peptide transport into the endoplasmatic reticulum, and/or processing and loading onto the respective HLA class I molecule. Western blot analysis revealed that the inducible components of the immune proteasome and the peptide transporters TAP1 and TAP2 are expressed at much lower level in A1 cells as compared to parental EREB2 cells, providing an explanation for the poor antigen‐presenting capacity of A1 cells. Most likely, down‐regulation of HLA class I and of adhesion molecules has additionally contributed to the lack of immunogenicity of c‐ myc‐overexpressing cells by lowering the threshold of T cell recognition.

C. c‐MYC Down‐Regulates NF‐kB and Interferon Response Genes The comparison of EREB2 cells on one hand and A1 and P493 on the other hand has revealed clear differences in the cell surface phenotype and tetracycline restores the allostimulatory capacity (A). A1 cells expressing vaccinia virus encoded EBNA3A of EBV type 1 are not recognized by an antigen‐specific cytotoxic T cell clone (B). Recognition can be partly restored by addition of the respective peptide (C). The components of the immune proteasome Lmp2, Lmp7, PA28 , and PA28 and peptide transporters TAP1 and TAP2 are down‐regulated in A1 cells (D) as well as HLA class I molecules as shown by isoelectric focusing (E). LMP1 denotes viral membrane protein 1, whereas Lmp2 and Lmp7 denote components of the immune proteasome. The data of this figure are derived from the paper by Staege et al. (2002).

178

Martin Schlee et al.

antigen‐presenting capacity of cells driven into proliferation by either EBV or c‐MYC. Yet, it has left an important question open, that is, whether nonimmunogenicity is the default pathway when the viral proliferation program is switched off or whether c‐MYC is actively contributing to the cell’s low or nonimmunogenicity. The finding that c‐MYC down‐regulates expression of HLA class I genes (Schrier and Peltenburg, 1993) and a number of adhesion molecules supports the latter assumption. We have addressed this issue by studying changes in gene expression in P493 cells manipulating c‐myc expression by addition or withdrawal of tetracycline as the only variable. Two approaches were used to modulate c‐myc expression: switching c‐myc expression on by washing out tetracycline from tetracycline‐arrested P493 cells, and lowering or switching c‐myc expression off by adding tetracycline to proliferating P493 cells. The latter approach was particularly useful, as it did not require excessive washing steps and allowed to add tetracycline in varying concentrations. We focused our attention on genes that were (1) differentially expressed and (2) related to cell adhesion and immune functions by gene ontology classification. One hundred twenty‐two genes followed these criteria, of which 97 were down‐regulated and 25 up‐regulated by c‐MYC. Figure 4 shows the expression of the negatively regulated genes in a tetracycline titration experiment. Many of the genes down‐regulated by c‐MYC were found to be NF‐B‐ and interferon‐regulated genes. Three typical NF‐B‐regulated genes (Lymphotoxin‐ and ‐ , TNF‐ ) (Pahl, 1999) and three interferon‐regulated genes (IFI‐6‐16, IFP‐35, and OAS‐1) (de Veer et al., 2001; Der et al., 1998) were selected and their expression presented as a function of c‐myc expression. As shown in the middle and right panel of Fig. 4, expression of these genes is inversely correlated to that of c‐myc. This supports the notion that the phenotype of A1 and P493 cells (Fig. 2A) is not a default phenotype solely brought about by switching off the expression of EBNA2 and LMP1, but is additionally actively imposed by c‐MYC. This holds true not only for the expression pattern of cell surface markers but also for other important regulators of immune recognition like the components of the immune proteasome and the peptide transporters. Many of the genes involved in antigen presentation and cell‐to‐cell communication are known to be coregulated by NF‐B and interferons. Among the genes negatively regulated by c‐MYC is also the IFN gene precursor that is moderately down‐regulated by c‐MYC. The fact that the IFN precursor is moderately down‐regulated and many interferon‐regulated genes are strongly down‐regulated in P493 cells upon addition of tetracycline suggests that c‐MYC interrupts the positive feedback loop that involves secretion of interferons (Schlee et al., in press).

Interferon regulated genes

NF-kB regulated genes 1.2

1

Lymphotoxin-a

0.8 0.6 0.4 0.2

Fold change

Gene symbol

Fold change

1.2

0 6

5

4

3

0.4 0.2 100 8

6

5

4

3

2 1.5 1 0.5 0

1.2

TNF-a

1 0.8 0.6 0.4 0.2

Fold change

Fold change

0.6

2 1.5 1 0.5 0

0

1

IFP 35

0.8 0.6 0.4 0.2 0

100 8

6

5

4

3

2 1.5 1 0.5 0

100 8

6

5

4

3

2 1.5 1 0.5 0

1.2

1

Lymphotoxin-b

0.8 0.6 0.4 0.2

Fold change

1.2

Fold change

IFI-6-16

0 100 8

1.2

0

1

OAS1

0.8 0.6 0.4 0.2 0

6

5

100 90 80 70 60 50 40 30 20 10 0

4

3

2 1.5 1 0.5 0

Fold change

c-MYC

100 8

6

5

4

3

2 1.5 1 0.5 0

Tetracycline (ng/ml)

100 8

6

5

4

100 90 80 70 60 50 40 30 20 10 0

3

2 1.5 1 0.5 0

c-MYC

Fold change

100 8

c-MYC Tetracycline

1 0.8

100 8

6

5

4

3

2 1.5 1 0.5 0

Tetracycline (ng/ml)

Fig. 4 Genes differentially regulated by c‐MYC were selected as related to immune function and cell adhesion by gene ontology from a filter array of 5000 genes. Of 122 genes differentially regulated by a factor of 2 or more, 25 were positively and 97 negatively regulated. The genes negatively regulated by c‐MYC are presented in TreeView in the left panel. The expression of three selected NF‐B and interferon target genes is presented in the middle and right panel as a function of tetracycline and c‐MYC concentration.

180

Martin Schlee et al.

D. c‐MYC Impairs the Interferon Response at Different Levels: At the Level of Induction as Well as at the Level of Action of Type I Interferons On the basis of the observation that IFN is up‐regulated when c‐myc expression is switched off in P493 cells (Schlee et al., in press; Fig. 4B), we have studied the response of the IFN promoter to the interferon inducer poly(IC) in HeLa cells in the absence and presence of c‐MYC. Activation of the IFN promoter is mediated by the transcription factor interferon regulatory factor 3 (IRF3). IRF3 is constitutively expressed in an inactive form and is activated by phosphorylation in response to an interferon‐inducing stimulus like poly(IC) or viral infection. As shown in Fig. 5A, coexpression of c‐MYC abolished activation of the IFN promoter by poly(IC) and also reduced the basal promoter activity. Virtually identical results were obtained when a multimerized positive regulatory domain I (PRDI) interferon response element representing the binding site for IRF3 was used for transfection (lower panel of Fig. 5A). These results emphasize a general, cell system‐independent effect of c‐MYC on the IRF3 pathway. To ask whether c‐MYC interferes with the action of activated IRF3, we have studied activation of the IFN promoter by a phosphomimetic constitutively active mutant of IRF3, IRF3‐5D (Lin et al., 1998), in the absence and presence of c‐MYC. As shown in Fig. 5B, coexpression of c‐MYC strongly abolished activation of the IFN promoter by IRF3‐5D. We have finally addressed the question whether c‐MYC acts not only at the level of interferon induction but also at the level of interferon action. To this end, we have studied the activation of type I or type II interferon promoter reporter constructs by IFN and IFN , respectively, in the absence and presence of c‐MYC. As shown in Fig. 5C and D, c‐MYC abolished induction of the MxA‐promoter by IFN as well as induction of a multimerized GAS element by IFN . We thus propose that c‐MYC interferes with the interferon response at at least two different levels, the level of interferon induction as well as at the level of interferon‐induced target genes.

IV. DISCUSSIONS The EBV‐ and c‐MYC‐driven conditional cell lines that we have developed mimic the dramatic differences between EBV‐immortalized cells and BL cells in antigen presentation, T cell stimulation, and the ability to be recognized by cytotoxic T‐lymphocytes (CTL). EBV‐immortalized cells are excellent antigen‐presenting cells, whereas their c‐MYC‐driven counterparts have lost their antigen‐presenting capacity and cannot be recognized

181

c‐MYC Impairs B Cell Immunogenicity

A

B

ATF2 c-jun IRF3/7 IRF3 NFkB −125 PRDIV PRDIII PRDI PRDII

IFNb

RLU

10,000

150,000

IFNb

0 50,000 40,000 30,000 20,000 10,000 0 c-MYC − pIC −

100,000

0 IRF3-5D IRF7-2D Con

− +

+ −

+ +

D 20,000 12,000 8000 4000 − c-MYC −

900 800 700 600 500 400 300 200 100 0 IFNg c-MYC

RLU

16,000

0 IFNb

IFNb

50,000

PRDI(8x)

C

RLU

c-MYC GFP

5000

RLU

RLU

15,000

+ −

− +

+ +

− −

+ −

− +

+ +

Fig. 5 c‐MYC impairs induction of the IFN promoter (A, B) and of the promoters of interferon‐induced genes (C, D) in HeLa cells. Firefly luciferase was cloned behind the respective promoters and promoter activity measured as luciferase activity (RLU: relative light units). The induction of the IFN promoter (upper panel in A) and of a multimerized PRDI element (lower panel in A) by poly(IC) (A), the induction of the IFN promoter by constitutively activated IRF3‐5D, but not by IRF7‐2D (B), the induction of the MxA promoter by IFN (C), and of a multimerized GAS element by IFN (D) were all impaired by c‐MYC. In A, luciferase promoter constructs (5 mg) and 5 mg of a c‐myc expression plasmid or empty pCDNA3 vector were cotransfected together with or without poly(IC) as inducer (A). In B, the IFN promoter luciferase reporter plasmid was cotransfected with either a c‐myc (black bars) or GFP expression plasmid (gray bars) together with expression constructs of either IRF3‐5D, IRF7‐2D or pcDNA3 vector control (con) as inducers, respectively (B). In C and D, an MxA promoter and a multimerized GAS‐luciferase construct were selected as IFN / ‐ and IFN ‐responsive promoters, respectively, and were cotransfected with a c‐myc expression plasmid or empty vector. The transfected cells were treated with IFN (C) and IFN (D), respectively, or left untreated.

by CTLs. Here we have shown that c‐MYC‐driven proliferation and immunogenicity are mechanistically coupled in a reciprocal fashion through the NF‐B and interferon system. The interferon system is well known as an antiviral defense system, it has, however, also strong antiproliferative activity. By antagonizing the interferon response, c‐MYC thus relieves the brake

182

Martin Schlee et al.

on cell proliferation. Importantly, c‐MYC not only abolishes induction of type I interferons, but also impairs the response to type I as well as type II interferons. Many steps in the processing and presentation of antigens are known to be regulated by the combined action of NF‐B and interferons, including the composition and activity of the immune proteasome, the peptide transporter (TAP1 and TAP2), the ER‐associated aminopeptidase ERAP1, as well as expression of MHC class I and class II and adhesion molecules (Kaplan et al., 1998; Kloetzel, 2001; Rock et al., 2004). Impaired immunogenicity of a tumor is thus not necessarily exclusively the end‐ product of a selection process in vivo. The initial oncogenic event leading to c‐myc activation may additionally contribute. As the c‐myc gene is activated in a large set of rapidly proliferating tumors of hematopoietic and epithelial origin including, for example, lung, mammary, and colon carcinoma, it is important to answer the question whether the reciprocal link between c‐myc activation and immunogenicity holds true for other types of human cancers. Whether the N‐myc gene activated in childhood neuroblastomas has a similar anti‐immunogenic activity remains also to be determined. It is noteworthy that c‐myc overexpression increases the threshold but does not block induction of the NF‐B and interferon pathway completely. Resistance to interferon treatment of uveal melanoma cells cultured ex vivo has indeed been correlated with the level of c‐MYC expression in the original formalin‐fixed tumor indicating that the quantitative aspect is important (Tulley et al., 2004). MYC’s inhibitory action on the interferon pathway should also be considered in the context of clinical tumor vaccination approaches (Leitner et al., 2003). Genetic as well as biochemical evidence has revealed a number of properties shared by the c‐MYC and adenovirus E1A oncoproteins (Deleu et al., 2001; McMahon et al., 1998; Ralston, 1991). Both induce proliferation as well as apoptosis, require interaction with ATM‐related transformation/ transcription‐domain‐associated protein (TRRAP) for transformation, interact with CBP/p300 (Eckner et al., 1994; Vervoorts et al., 2003), and down‐regulate HLA class I genes (Vasavada et al., 1986; Versteeg et al., 1988). We extend this functional homology by showing that MYC also shares with E1A the ability to down‐regulate IFN transcription and to inhibit IFN type I and type II signaling (Ackrill et al., 1991; Kalvakolanu et al., 1991). It will be important to find out how c‐MYC impairs the interferon response, which of the interacting proteins contribute to this action, and whether E1A and c‐MYC share the basic principle of action. The antagonism of c‐MYC and the interferon system is not restricted to the finding reported here that c‐MYC inhibits induction of IFN and STAT1, and impairs the interferon response. STAT1 is not only regulated by c‐MYC, it is also a negative regulator of c‐myc expression (Ramana et al., 2000). c‐MYC and STAT1 thus regulate each other at a transcriptional level in a negative feedback loop favoring that either the c‐MYC‐ or

183

c‐MYC Impairs B Cell Immunogenicity

c-MYC

Targets

STAT1

Targets

Fig. 6 Proposed c‐MYC–STAT1 network regulating proliferation versus cell cycle arrest. c‐MYC and STAT1 regulate each other negatively not only at the level of expression but also at the level of action on target genes. Downstream target genes of c‐MYC and STAT1 may additionally participate in this network.

the STAT1‐program is active (Fig. 6). Remarkably, besides the reciprocal negative transcriptional regulation of c‐MYC and STAT1, there is an antagonism at still another level. IFN has been reported to relieve the v‐myc‐ mediated block in TPA‐induced differentiation of U937 cells without affecting the level of v‐myc protein expression (Oberg et al., 1991). This indicates that IFN is able to antagonize and overcome the action of the oncogene at a level downstream of v‐myc. Likewise, the STAT1 target gene IRF1 was shown to overcome c‐myc‐ and FosB‐induced transformation of mouse embryonic fibroblasts (Tanaka et al., 1994a) and to revert the phenotype and tumorigenicity of c‐myc þ v‐ras transformed cells without affecting the level of c‐myc expression in these cells (Kroger et al., 2003). This pronounced antagonism of c‐myc and the interferon system at more than one level of regulation suggests that the interferon system is involved in physiological growth control. The role of IRF1 as a negative growth regulator in vitro is well documented (Kirchhoff et al., 1993; Kroger et al., 2003; Tanaka et al., 1994b). IRF1/ cells, but not IRF1þ/þ cells, can be transformed in vitro by a mutated ras oncogene (Tanaka et al., 1994a), and loss of IRF1 increases the susceptibility to tumor development of mice with a Ha‐ras transgene or p53 nullizygosity in vivo (Nozawa et al., 1999). There is increasing evidence that the p53 and the interferon system are interconnected (Baran‐ Marszak et al., 2004; Takaoka et al., 2003; Townsend et al., 2004). Last but not least, ICSBP/ (IRF8/) mice develop a CML‐like syndrome (Holtschke et al., 1996). It is thus important to readdress the role of STAT1 and IRFs as tumor suppressor genes and regulators of cellular senescence in vivo.

ACKNOWLEDGMENTS We are grateful to J. Darnell, J. Hiscott, J. Pagano, T. Fujita, C. Horvath, T. Taniguchi, and P. Staeheli for IRF and STAT expression and reporter constructs, and K. Conzelmann for antibodies. We thank Hans‐Jo¨rg Hauser, Thomas Decker, and Klaus Conzelmann for reagents

184

Martin Schlee et al.

and many helpful discussions. The work was supported by a grant to G.W.B. from the Wilhelm Sander‐Stiftung, DFG (SFB455), and Deutsche Krebshilfe. M.S. received a fellowship from the Boehringer Ingelheim‐Stiftung.

REFERENCES Ackrill, A. M., Foster, G. R., Laxton, C. D., Flavell, D. M., Stark, G. R., and Kerr, I. M. (1991). Inhibition of the cellular response to interferons by products of the adenovirus type 5 E1A oncogene. Nucleic Acids Res. 19, 4387–4393. Avila‐Carino, J., Torsteinsdottir, S., Ehlin‐Henriksson, B., Lenoir, G., Klein, G., Klein, E., and Masucci, M. G. (1987). Paired Epstein‐Barr virus (EBV)‐negative and EBV‐converted Burkitt lymphoma lines: Stimulatory capacity in allogeneic mixed lymphocyte cultures. Int. J. Cancer 40, 691–697. Baran‐Marszak, F., Feuillard, J., Najjar, I., Le Clorennec, C., Bechet, J. M., Dusanter‐Fourt, I., Bornkamm, G. W., Raphael, M., and Fagard, R. (2004). Differential roles of STAT1alpha and STAT1beta in fludarabine‐induced cell cycle arrest and apoptosis in human B cells. Blood 104, 2475–2483. Bechet, J. M., Nilsson, K., and Ponten, J. (1973). Establishment of multiple lymphoblastoid lines from single lymph nodes and study of their immunoglobulin synthesis. Int. J. Cancer 11, 58–63. Bechet, J. M., Fialkow, P. J., Nilsson, K., Klein, G., and Singh, S. (1974). Immunoglobulin synthesis and glucose‐6‐phosphate dehydrogenase as cell markers in human lymphoblastoid cell lines. Exp. Cell Res. 89, 275–282. Bernheim, A., Berger, R., and Lenoir, G. (1980). Translocations t(2;8) and t(8;22) in continuous cell lines of African Burkitt’s lymphoma. C. R. Seances Acad. Sci. D 291, 237–239. Bernheim, A., Berger, R., and Lenoir, G. (1981). Cytogenetic studies on African Burkitt’s lymphoma cell lines: t(8;14), t(2;8) and t(8;22) translocations. Cancer Genet. Cytogenet. 3, 307–315. Bornkamm, G. W., and Hammerschmidt, W. (2001). Molecular virology of Epstein‐Barr virus. Philos. Trans. R. Soc. Lond. B Biol. Sci. 356, 437–459. Bornkamm, G. W., Kaduk, B., Kachel, G., Schneider, U., Fresen, K. O., Schwanitz, G., and Hermanek, P. (1980). Epstein‐Barr virus‐positive Burkitt’s lymphoma in a German woman during pregnancy. Blut 40, 167–177. Caspersson, T., Farber, S., Foley, G. E., Kudynowski, J., Modest, E. J., Simonsson, E., Wagh, U., and Zech, L. (1968). Chemical differentiation along metaphase chromosomes. Exp. Cell Res. 49, 219–222. Caspersson, T., Zech, L., and Johansson, C. (1970). Differential binding of alkylating fluorochromes in human chromosomes. Exp. Cell Res. 60, 315–319. Cuomo, L., Trivedi, P., de Campos‐Lima, P. O., Zhang, Q. J., Ragnar, E., Klein, G., and Masucci, M. G. (1993). Selective induction of allostimulatory capacity after 5‐azaC treatment of EBV carrying but not EBV negative Burkitt lymphoma cell lines. Mol. Immunol. 30, 441–450. Dalla‐Favera, R., Bregni, M., Erikson, J., Patterson, D., Gallo, R. C., and Croce, C. M. (1982). Human c‐myc onc gene is located on the region of chromosome 8 that is translocated in Burkitt lymphoma cells. Proc. Natl. Acad. Sci. USA 79, 7824–7827. de Klein, A., van Kessel, A. G., Grosveld, G., Bartram, C. R., Hagemeijer, A., Bootsma, D., Spurr, N. K., Heisterkamp, N., Groffen, J., and Stephenson, J. R. (1982). A cellular oncogene is translocated to the Philadelphia chromosome in chronic myelocytic leukaemia. Nature 300, 765–767.

c‐MYC Impairs B Cell Immunogenicity

185

de Veer, M. J., Holko, M., Frevel, M., Walker, E., Der, S., Paranjape, J. M., Silverman, R. H., and Williams, B. R. (2001). Functional classification of interferon‐stimulated genes identified using microarrays. J. Leukoc. Biol. 69, 912–920. Deleu, L., Shellard, S., Alevizopoulos, K., Amati, B., and Land, H. (2001). Recruitment of TRRAP required for oncogenic transformation by E1A. Oncogene 20, 8270–8275. Der, S. D., Zhou, A., Williams, B. R., and Silverman, R. H. (1998). Identification of genes differentially regulated by interferon alpha, beta, or gamma using oligonucleotide arrays. Proc. Natl. Acad. Sci. USA 95, 15623–15628. Desgranges, C., Wolf, H., De‐The, G., Shanmugaratnam, K., Cammoun, N., Ellouz, R., Klein, G., Lennert, K., Munoz, N., and Zur Hausen, H. (1975). Nasopharyngeal carcinoma. X. Presence of Epstein‐Barr genomes in separated epithelial cells of tumours in patients from Singapore, Tunisia and Kenya. Int. J. Cancer 16, 7–15. Douglass, E. C., Magrath, I. T., Lee, E. C., and Whang‐Peng, J. (1980). Cytogenetic studies in non‐African Burkitt lymphoma. Blood 55, 148–155. Dunn, G. P., Old, L. J., and Schreiber, R. D. (2004). The immunobiology of cancer immunosurveillance and immunoediting. Immunity 21, 137–148. Eckner, R., Ewen, M. E., Newsome, D., Gerdes, M., DeCaprio, J. A., Lawrence, J. B., and Livingston, D. M. (1994). Molecular cloning and functional analysis of the adenovirus E1A‐ associated 300‐kD protein (p300) reveals a protein with properties of a transcriptional adaptor. Genes Dev. 8, 869–884. Frisan, T., Zhang, Q. J., Levitskaya, J., Coram, M., Kurilla, M. G., and Masucci, M. G. (1996). Defective presentation of MHC class I‐restricted cytotoxic T‐cell epitopes in Burkitt’s lymphoma cells. Int. J. Cancer 68, 251–258. Gavioli, R., Frisan, T., Vertuani, S., Bornkamm, G. W., and Masucci, M. G. (2001). c‐Myc overexpression activates alternative pathways for intracellular proteolysis in lymphoma cells. Nat. Cell Biol. 3, 283–288. Holtschke, T., Lohler, J., Kanno, Y., Fehr, T., Giese, N., Rosenbauer, F., Lou, J., Knobeloch, K. P., Gabriele, L., Waring, J. F., Bachmann, M. F., Zinkernagel, R. M., et al. (1996). Immunodeficiency and chronic myelogenous leukemia‐like syndrome in mice with a targeted mutation of the ICSBP gene. Cell 87, 307–317. Kalter, S. P., Riggs, S. A., Cabanillas, F., Butler, J. J., Hagemeister, F. B., Mansell, P. W., Newell, G. R., Velasquez, W. S., Salvador, P., Barlogie, B., Rios, A., and Hersh, E. M. (1985). Aggressive non‐Hodgkin’s lymphomas in immunocompromised homosexual males. Blood 66, 655–659. Kalvakolanu, D. V., Bandyopadhyay, S. K., Harter, M. L., and Sen, G. C. (1991). Inhibition of interferon‐inducible gene expression by adenovirus E1A proteins: Block in transcriptional complex formation. Proc. Natl. Acad. Sci. USA 88, 7459–7463. Kaplan, D. H., Shankaran, V., Dighe, A. S., Stockert, E., Aguet, M., Old, L. J., and Schreiber, R. D. (1998). Demonstration of an interferon gamma‐dependent tumor surveillance system in immunocompetent mice. Proc. Natl. Acad. Sci. USA 95, 7556–7561. Kaschka‐Dierich, C., Adams, A., Lindahl, T., Bornkamm, G. W., Bjursell, G., Klein, G., Giovanella, B. C., and Singh, S. (1976). Intracellular forms of Epstein‐Barr virus DNA in human tumour cells in vivo. Nature 260, 302–306. Kelly, G., Bell, A., and Rickinson, A. (2002). Epstein‐Barr virus‐associated Burkitt lymphomagenesis selects for downregulation of the nuclear antigen EBNA2. Nat. Med. 8, 1098–1104. Kempkes, B., Spitkovsky, D., Jansen‐Durr, P., Ellwart, J. W., Kremmer, E., Delecluse, H. J., Rottenberger, C., Bornkamm, G. W., and Hammerschmidt, W. (1995). B‐cell proliferation and induction of early G1‐regulating proteins by Epstein‐Barr virus mutants conditional for EBNA2. EMBO J. 14, 88–96. Khanna, R., Burrows, S. R., Argaet, V., and Moss, D. J. (1994). Endoplasmic reticulum signal sequence facilitated transport of peptide epitopes restores immunogenicity of an antigen processing defective tumour cell line. Int. Immunol. 6, 639–645.

186

Martin Schlee et al.

Kiessling, R., Klein, E., Pross, H., and Wigzell, H. (1975a). “Natural” killer cells in the mouse. II. Cytotoxic cells with specificity for mouse Moloney leukemia cells. Characteristics of the killer cell. Eur. J. Immunol. 5, 117–121. Kiessling, R., Klein, E., and Wigzell, H. (1975b). “Natural” killer cells in the mouse. I. Cytotoxic cells with specificity for mouse Moloney leukemia cells. Specificity and distribution according to genotype. Eur. J. Immunol. 5, 112–117. Kirchhoff, S., Schaper, F., and Hauser, H. (1993). Interferon regulatory factor 1 (IRF‐1) mediates cell growth inhibition by transactivation of downstream target genes. Nucleic Acids Res. 21, 2881–2889. Klein, E., Torsteinsdottir, S., Masucci, M. G., and Klein, G. (1986). Different sensitivity of B cell lines derived from Burkitt lymphoma (BL) and normal cells (LCL) to cytotoxic T cell clones generated by autologous and allogeneic stimulation. J. Immunogenet. 13, 269–274. Klein, G., and Klein, E. (1977). Immune surveillance against virus‐induced tumors and nonrejectability of spontaneous tumors: Contrasting consequences of host versus tumor evolution. Proc. Natl. Acad. Sci. USA 74, 2121–2125. Klein, G., Giovanella, B. C., Lindahl, T., Fialkow, P. J., Singh, S., and Stehlin, J. S. (1974). Direct evidence for the presence of Epstein‐Barr virus DNA and nuclear antigen in malignant epithelial cells from patients with poorly differentiated carcinoma of the nasopharynx. Proc. Natl. Acad. Sci. USA 71, 4737–4741. Kloetzel, P. M. (2001). Antigen processing by the proteasome. Nat. Rev. Mol. Cell. Biol. 2, 179–187. Kroger, A., Dallugge, A., Kirchhoff, S., and Hauser, H. (2003). IRF‐1 reverts the transformed phenotype of oncogenically transformed cells in vitro and in vivo. Oncogene 22, 1045–1056. Lee, S. P., Brooks, J. M., Al‐Jarrah, H., Thomas, W. A., Haigh, T. A., Taylor, G. S., Humme, S., Schepers, A., Hammerschmidt, W., Yates, J. L., Rickinson, A. B., and Blake, N. W. (2004). CD8 T cell recognition of endogenously expressed Epstein‐Barr virus nuclear antigen 1. J. Exp. Med. 199, 1409–1420. Leitner, W. W., Hwang, L. N., deVeer, M. J., Zhou, A., Silverman, R. H., Williams, B. R., Dubensky, T. W., Ying, H., and Restifo, N. P. (2003). Alphavirus‐based DNA vaccine breaks immunological tolerance by activating innate antiviral pathways. Nat. Med. 9, 33–39. Lenoir, G., and Philip, T. (1979). Burkitt’s lymphoma and Epstein‐Barr virus. Nouv. Presse Med. 8, 4017. Lenoir, G., Philip, T., Bornkamm, G. W., Gillet, P., Bryon, P. A., Bouvier, R., Dodat, H., Dore, J. F., Brunat‐Mentigny, M., and Hermier, M. (1979). Epstein‐Barr virus associated Burkitt’s lymphoma in a French child. Nouv. Presse Med. 8, 4031–4033. Levitskaya, J., Coram, M., Levitsky, V., Imreh, S., Steigerwald‐Mullen, P. M., Klein, G., Kurilla, M. G., and Masucci, M. G. (1995). Inhibition of antigen processing by the internal repeat region of the Epstein‐Barr virus nuclear antigen‐1. Nature 375, 685–688. Lin, R., Heylbroeck, C., Pitha, P. M., and Hiscott, J. (1998). Virus‐dependent phosphorylation of the IRF‐3 transcription factor regulates nuclear translocation, transactivation potential, and proteasome‐mediated degradation. Mol. Cell. Biol. 18, 2986–2996. Lindahl, T., Adams, A., Bjursell, G., Bornkamm, G. W., Kaschka‐Dierich, C., and Jehn, U. (1976). Covalently closed circular duplex DNA of Epstein‐Barr virus in a human lymphoid cell line. J. Mol. Biol. 102, 511–530. Magrath, I. T., Freeman, C. B., Pizzo, P., Gadek, J., Jaffe, E., Santaella, M., Hammer, C., Frank, M., Reaman, G., and Novikovs, L. (1980). Characterization of lymphoma‐derived cell lines: Comparison of cell lines positive and negative for Epstein‐Barr virus nuclear antigen. II. Surface markers. J. Natl. Cancer Inst. 64, 477–483. Manolov, G., and Manolova, Y. (1972). Marker band in one chromosome 14 from Burkitt lymphomas. Nature 237, 33–34. Manolov, G., Manolova, Y., Levan, A., and Klein, G. (1971a). Experiments with fluorescent chromosome staining in Burkitt tumors. Hereditas 68, 235–244.

c‐MYC Impairs B Cell Immunogenicity

187

Manolov, G., Manolova, Y., Levan, A., and Klein, G. (1971b). Fluorescent pattern of apparently normal chromosomes in Burkitt lymphomas. Hereditas 68, 160–163. Masucci, M. G., Torsteindottir, S., Colombani, J., Brautbar, C., Klein, E., and Klein, G. (1987). Down‐regulation of class I HLA antigens and of the Epstein‐Barr virus‐encoded latent membrane protein in Burkitt lymphoma lines. Proc. Natl. Acad. Sci. USA 84, 4567–4571. McMahon, S. B., Van Buskirk, H. A., Dugan, K. A., Copeland, T. D., and Cole, M. D. (1998). The novel ATM‐related protein TRRAP is an essential cofactor for the c‐Myc and E2F oncoproteins. Cell 94, 363–374. Miyoshi, I., Hiraki, S., Tsubota, T., Uno, J., Nakamura, K., Ota, T., Hikita, T., Hayashi, K., Kataoka, M., Tanaka, T., Kimura, I., Sairenji, T., et al. (1978). Epstein‐Barr virus‐positive Japanese Burkitt lymphoma. Gann 69, 449–450. Miyoshi, I., Hiraki, S., Kimura, I., Miyamoto, K., and Sato, J. (1979). 2/8 translocation in a Japanese Burkitt’s lymphoma. Experientia 35, 742–743. Miyoshi, I., Hamazaki, K., Kubonishi, I., Yoshimoto, S., Kitajima, K., Kimura, I., Miyamoto, K., Sato, J., Yorimitsu, S., Tao, S., Ishibashi, K., and Tokuda, M. (1981). A variant translocation (8;22) in a Japanese patient with Burkitt lymphoma. Gann 72, 176–177. Nowell, P. C., and Hungerford, D. A. (1960). Chromosome studies on normal and leukemic human leukocytes. J. Natl. Cancer Inst. 25, 85–109. Nozawa, H., Oda, E., Nakao, K., Ishihara, M., Ueda, S., Yokochi, T., Ogasawara, K., Nakatsuru, Y., Shimizu, S., Ohira, Y., Hioki, K., Aizawa, S., et al. (1999). Loss of transcription factor IRF‐1 affects tumor susceptibility in mice carrying the Ha‐ras transgene or nullizygosity for p53. Genes Dev. 13, 1240–1245. Oberg, F., Larsson, L. G., Anton, R., and Nilsson, K. (1991). Interferon gamma abrogates the differentiation block in v‐myc‐expressing U‐937 monoblasts. Proc. Natl. Acad. Sci. USA 88, 5567–5571. Ohno, S., Babonits, M., Wiener, F., Spira, J., Klein, G., and Potter, M. (1979). Nonrandom chromosome changes involving the Ig gene‐carrying chromosomes 12 and 6 in pristane‐ induced mouse plasmacytomas. Cell 18, 1001–1007. Pahl, H. L. (1999). Activators and target genes of Rel/NF‐kappaB transcription factors. Oncogene 18, 6853–6866. Pajic, A., Staege, M. S., Dudziak, D., Schuhmacher, M., Spitkovsky, D., Eissner, G., Brielmeier, M., Polack, A., and Bornkamm, G. W. (2001). Antagonistic effects of c‐myc and Epstein‐Barr virus latent genes on the phenotype of human B cells. Int. J. Cancer 93, 810–816. Qin, Z., and Blankenstein, T. (2004). A cancer immunosurveillance controversy. Nat. Immunol. 5, 3–4; author reply, 4–5. Ralston, R. (1991). Complementation of transforming domains in E1a/myc chimaeras. Nature 353, 866–868. Ramana, C. V., Grammatikakis, N., Chernov, M., Nguyen, H., Goh, K. C., Williams, B. R., and Stark, G. R. (2000). Regulation of c‐myc expression by IFN‐gamma through Stat1‐ dependent and ‐independent pathways. EMBO J. 19, 263–272. Reedman, B. M., and Klein, G. (1973). Cellular localization of an Epstein‐Barr virus (EBV)‐ associated complement‐fixing antigen in producer and non‐producer lymphoblastoid cell lines. Int. J. Cancer 11, 499–520. Reedman, B. M., Klein, G., Pope, J. H., Walters, M. K., Hilgers, J., Singh, S., and Johansson, B. (1974). Epstein‐Barr virus‐associated complement‐fixing and nuclear antigens in Burkitt lymphoma biopsies. Int. J. Cancer 13, 755–763. Rock, K. L., York, I. A., and Goldberg, A. L. (2004). Post‐proteasomal antigen processing for major histocompatibility complex class I presentation. Nat. Immunol. 5, 670–677. Rowe, M., Khanna, R., Jacob, C. A., Argaet, V., Kelly, A., Powis, S., Belich, M., Croom‐Carter, D., Lee, S., Burrows, S. R., Trowsdale, J., Moss, D. J., et al. (1995). Restoration of endogenous antigen processing in Burkitt’s lymphoma cells by Epstein‐Barr virus latent membrane protein‐1: Coordinate up‐regulation of peptide transporters and HLA‐class I antigen expression. Eur. J. Immunol. 25, 1374–1384.

188

Martin Schlee et al.

Schlee, M., Ho¨lzel, M., Bernard, S., Mailhammer, R., Schuhmacher, M., Reschke, J., Eick, D, Marinkovic, D., Wirth, T., Rosenwald, A., Staudt, L. M., Eilers, M., et al. (2006). c‐MYC activation impairs the NF-B and the interferon response: Implications for the pathogenesis of Burkitt’s lymphoma. Int. J. Cancer, in press. Schrier, P. I., and Peltenburg, L. T. (1993). Relationship between myc oncogene activation and MHC class I expression. Adv. Cancer Res. 60, 181–246. Shen‐Ong, G. L., Keath, E. J., Piccoli, S. P., and Cole, M. D. (1982). Novel myc oncogene RNA from abortive immunoglobulin‐gene recombination in mouse plasmacytomas. Cell 31, 443–452. Staege, M. S., Lee, S. P., Frisan, T., Mautner, J., Scholz, S., Pajic, A., Rickinson, A. B., Masucci, M. G., Polack, A., and Bornkamm, G. W. (2002). MYC overexpression imposes a nonimmunogenic phenotype on Epstein‐Barr virus‐infected B cells. Proc. Natl. Acad. Sci. USA 99, 4550–4555. Takaoka, A., Hayakawa, S., Yanai, H., Stoiber, D., Negishi, H., Kikuchi, H., Sasaki, S., Imai, K., Shibue, T., Honda, K., and Taniguchi, T. (2003). Integration of interferon‐alpha/beta signalling to p53 responses in tumour suppression and antiviral defence. Nature 424, 516–523. Tanaka, N., Ishihara, M., Kitagawa, M., Harada, H., Kimura, T., Matsuyama, T., Lamphier, M. S., Aizawa, S., Mak, T. W., and Taniguchi, T. (1994a). Cellular commitment to oncogene‐ induced transformation or apoptosis is dependent on the transcription factor IRF‐1. Cell 77, 829–839. Tanaka, N., Ishihara, M., and Taniguchi, T. (1994b). Suppression of c‐myc or fosB‐induced cell transformation by the transcription factor IRF‐1. Cancer Lett. 83, 191–196. Torsteinsdottir, S., Masucci, M. G., Ehlin‐Henriksson, B., Brautbar, C., Ben Bassat, H., Klein, G., and Klein, E. (1986). Differentiation‐dependent sensitivity of human B‐cell‐derived lines to major histocompatibility complex‐restricted T‐cell cytotoxicity. Proc. Natl. Acad. Sci. USA 83, 5620–5624. Townsend, P. A., Scarabelli, T. M., Davidson, S. M., Knight, R. A., Latchman, D. S., and Stephanou, A. (2004). STAT‐1 interacts with p53 to enhance DNA damage‐induced apoptosis. J. Biol. Chem. 279, 5811–5820. Tulley, P. N., Neale, M., Jackson, D., Chana, J. S., Grover, R., Cree, I., Grobbelaar, A. O., and Wilson, G. D. (2004). The relation between c‐myc expression and interferon sensitivity in uveal melanoma. Br. J. Ophthalmol. 88, 1563–1567. Vasavada, R., Eager, K. B., Barbanti‐Brodano, G., Caputo, A., and Ricciardi, R. P. (1986). Adenovirus type 12 early region 1A proteins repress class I HLA expression in transformed human cells. Proc. Natl. Acad. Sci. USA 83, 5257–5261. Versteeg, R., Noordermeer, I. A., Kruse‐Wolters, M., Ruiter, D. J., and Schrier, P. I. (1988). c‐Myc down‐regulates class I HLA expression in human melanomas. EMBO J. 7, 1023–1029. Vervoorts, J., Luscher‐Firzlaff, J. M., Rottmann, S., Lilischkis, R., Walsemann, G., Dohmann, K., Austen, M., and Luscher, B. (2003). Stimulation of c‐MYC transcriptional activity and acetylation by recruitment of the cofactor CBP. EMBO Rep. 4, 484–490. Willimsky, G., and Blankenstein, T. (2005). Sporadic immunogenic tumours avoid destruction by inducing T‐cell tolerance. Nature 437, 141–146. Wolf, H., Hausen, H. Z., and Becker, V. (1973). EB viral genomes in epithelial nasopharyngeal carcinoma cells. Nat. New Biol. 244, 245–247. Zimber, U., Adldinger, H. K., Lenoir, G. M., Vuillaume, M., Knebel‐Doeberitz, M. V., Laux, G., Desgranges, C., Wittmann, P., Freese, U. K., Schneider, U., et al. (1986). Geographical prevalence of two types of Epstein‐Barr virus. Virology 154, 56–66. zur Hausen, H., Schulte‐Holthausen, H., Klein, G., Henle, W., Henle, G., Clifford, P., and Santesson, L. (1970). EBV DNA in biopsies of Burkitt tumours and anaplastic carcinomas of the nasopharynx. Nature 228, 1056–1058.

Cancer Dormancy: Lessons from a B Cell Lymphoma and Adenocarcinoma of the Prostate Rosalia Rabinovsky,* Jonathan W. Uhr,{ Ellen S. Vitetta,{ and Eitan Yefenof* *Lautenberg Center for General and Tumor Immunology, Hebrew University of Jerusalem, Jerusalem 91120, Israel; Cancer Immunobiology Center, University of Texas, Southwestern Medical Center at Dallas, Texas 75235

{

I. II. III. IV. V. VI.

Introduction Scope of the Present Discussion Clinical Studies Experimental Dormancy of B Cell Lymphoma The Prostate Adenocarcinoma Model Concluding Remarks References

Cancer dormancy delineates a situation in which residual tumor cells persist in a patient with no apparent clinical symptoms. Although the precise mechanisms underlying cancer dormancy have not been explained, experimental models have provided some insights into the factors that might be involved in the induction and maintenance of a tumor dormant state. The authors of the present chapter studied a murine B cell lymphoma that can be made dormant when interacting with antibodies directed against the idiotype on its immunoglobulin Ig receptor. This experimental model of antibody‐ induced dormancy enabled the isolation and characterization of dormant lymphoma cells. The results indicated that anti‐Ig antibodies activate growth‐inhibiting signals that induced cycle arrest and apoptosis. This process appeared to be balanced by the growth of the tumor cells such that the tumor did not expand. In contrast, antibodies against HER‐2expressed on prostate adenocarcinoma (PAC) cells were not growth inhibitory. However, an immunotoxin (IT) prepared by conjugating HER‐2 to the A‐chain of ricin (RTA) was internalized by PAC cells, followed by induction of cycle arrest and apoptotic death. Infusion of HER‐2‐specific IT into PAC‐bearing immunodeficient mice did not eradicate the tumor but retained it dormant over an extended period of time. Hence, certain aspects of signaling receptors expressed on cancer can be manipulated by antibodies to induce and maintain a tumor dormant state. # 2007 Elsevier Inc.

Advances in CANCER RESEARCH Copyright 2007, Elsevier Inc. All rights reserved.

189

0065-230X/07 $35.00 DOI: 10.1016/S0065-230X(06)97008-0

Rosalia Rabinovsky et al.

190

I. INTRODUCTION Professor Yigal Yadin, the Israeli archeologist who excavated Massada, recollected in his memoirs his first visit to Stockholm. It was in the mid‐ 1960s, when he was invited to speak at the annual meeting of the Swedish Archeological Society. The King of Sweden, Gustav VII Adolf, attended the opening session and the president of the society gave the opening address in Swedish. Immediately thereafter, Yadin was introduced to the king. “How did you like the opening lecture?” asked the king. “Since I do not understand Swedish I was a little bored,” answered Yadin. “Well,” the king responded, “I can assure you that those who understand Swedish were bored much more than you.” Language fulfills its destiny if there is a speaker and a listener. The last author of this chapter (E.Y.) had the privilege of listening to Eva and George Klein speak in three languages (Fig. 1). It began with English, which was dominant among a dozen languages spoken at the Department of Tumor Biology. After a while it was Swedish due to my desire to practice a newly acquired tongue (I recall many incidents of misunderstanding on my part both as a speaker and as a listener). Then it became Hebrew, when George was determined to master the ancient language at the age of 50. Many Israelis, including myself, were delighted to hear him speak and even lecture in Hebrew on his frequent visits to our faculty. With this experience, I am confident to testify that no matter what language Eva and George spoke, what topic they discussed, what story they told, I never was bored when listening to them. Each time there was a novel insight, a different view, a lesson to follow, or a new idea to embrace and study. One tenet I learned from Eva and George through their teaching and practice is that in scientific conduct, collaboration is superior to competition. Hence, the present chapter highlights a topic that has been studied in collaboration between the authors and their respective laboratories: what is cancer dormancy, how does it come about to exist and persist, what are the properties of dormant cancer cells, and why is tumor dormancy many times terminated with a relapse?

II. SCOPE OF THE PRESENT DISCUSSION Three alternative outcomes of cellular transformation may be discerned. If malignant cells continue to grow with no restraint, a deadly tumor develops. Alternatively, such cells can be eradicated by potent therapeutic interventions such as surgery, chemotherapy, radiotherapy, targeted therapy, or combinations thereof. A third possibility is that cancerous cells remain in the body but at the same time they do not progress to a full‐fledged disease.

B

A

C

Fig. 1 Eva and George Klein in three languages. (A) English, (B) Swedish, and (C) Hebrew.

192

Rosalia Rabinovsky et al.

The tumor is dormant in various organs and tissues with no apparent clinical manifestations. How can cells with a malignant character be quiescent over a long period of time? We tend to think that a tumor is a mass of proliferating cells that continues to grow and expand. However, under most circumstances, most cancerous cells within a tumor do not divide in a progressive manner. Some undergo terminal differentiation. Some are cycle arrested and others die of necrosis or apoptosis. When cell growth comes into balance with growth inhibition and death, the tumor remains dormant. It is however, a delicate balance that may be interrupted at any time if one of the factors controlling it goes awry. There appear to be two categories of tumor dormancy: one involves macrometastatic lesions that fail to induce angiogenesis because antiangiogenic factors are produced by the primary tumor or given externally (Gasparini et al., 2001; O’Reilly et al., 1994). Local hypoxia is insufficient to eradicate the tumor but limits its expansion by inducing apoptotic death (O’Reilly et al., 1997). Another hypothesis that is not mutually exclusive is that micrometastatic foci or solitary cells remain quiescent in the bone marrow, blood, or other tissues because they respond to immunological and nonimmunological surveillance mechanisms (Mejean et al., 2000; Pantel et al., 1999). In the present chapter, we will focus on the second form of tumor dormancy as studied by others and ourselves in clinical observations and experimental approaches.

III. CLINICAL STUDIES Successful therapy of cancer often results in long‐term remission during which no disease is detectable. Recurrence of a tumor may occur many years later indicating that such patients are carriers of a residual population of cancerous cells that does not increase over time (Sagalowsky and Molberg, 1999; Talpaz et al., 1994). This outcome is inconsistent with the inherent property of cancer cells to divide and multiply. It implies that dormant tumor cells display unique properties at the single cell or populational level. For instance, such cells may divide and die at an equal rate or they may be cell cycle‐arrested. It is obvious that experimental models of tumor dormancy are necessary to investigate how cells with a fully malignant character can remain dormant and what factors are involved in the termination of the tumor dormant state. The conclusion that patients with long‐term remission can carry an undetectable disease is based on recurrences at a time when the patient

Cancer Dormancy

193

has no signs of tumor and the risk of recurrence is very low. This contrasts with the demonstration that tumor cells can persist after resection of a primary tumor when the risk of recurrence is high. Thus, several researchers were able to identify occult tumor cells in patients following removal of a primary tumor. Such cells were detected in the blood and bone marrow using sensitive morphologic, immunologic, and molecular techniques (Bilchik et al., 2000; Pantel et al., 1999; Racila et al., 1998; Zippelius and Pantel, 2000), indicating that dissemination of solitary cells from a primary tumor is an early event in the progression of the disease (Braun and Harbeck, 2001; Riethmuller and Klein, 2001). Of major interest is the study by Meng et al. (2004), who demonstrated the presence of circulating tumor cells in breast cancer patients 7–22 years after removal of the primary tumor, when the risk of recurrence is 1% per year. In theses studies, it was estimated that the half‐life of such cells was only a few hours, indicating that dormancy is a steady state in which potentially cancerous cells continuously divide and die. A dynamic persistence of dividing cancerous cells suggests that termination of dormancy results from either the eventual emergence of a cell variant less sensitive to death which gives rise to a secondary tumor or a systemic change possibly induced by an environmental event, for example infection or trauma.

IV. EXPERIMENTAL DORMANCY OF B CELL LYMPHOMA BCL‐1 is an aggressive B cell lymphoma that evolved spontaneously in a Balb/c mouse (Slavin and Strober, 1978). Syngeneic mice inoculated with BCL‐1 cells develop lymphoma involving the spleen, blood, and lymph nodes, and die within weeks. However, when the mice are preimmunized with the Ig idiotype (Id) of BCL‐1, no tumor growth is apparent (Uhr et al., 1991). Such mice, however, continue to carry BCL‐1 cells for life (George et al., 1987; Siu et al., 1986). Thus, if spleen cells of an Id immunized, BCL‐1 inoculated mouse are transferred to a nonimmunized mouse, the recipient will develop a BCL‐1 tumor and die within 2–3 weeks. In addition, every now and again, individual BCL‐1‐inoculated mice that have been protected by immunization with the Id will relapse and develop a BCL‐1 tumor months after the initial inoculation. Regrowth of a BCL‐1 tumor results from a drop in the titers of anti‐Id antibodies or because some of the cells undergo mutations in the Id signaling pathway that confers on wild‐type cells sensitivity to death or cycle arrest (Vitetta et al., 1997). The conclusion from these observations was that anti‐idiotypic immunity is insufficient to eradicate BCL‐1 cells but is effective in maintaining them dormant over an extended period of time.

194

Rosalia Rabinovsky et al.

Since the light chain of BCL‐1 is l3, which is a rare isotype in the mouse, BCL‐1 cells could be identified in the spleen by staining with anti‐l antibodies. Flow cytometry analysis revealed a unique population of lþ small lymphocytes in the dormant mouse that resemble resting B lymphocytes of the normal Balb/c spleen. Such cells were isolated by FACS sorting and analyzed microscopically (Yefenof et al., 1993a). Whereas growing BCL‐1 cells were large lymphoblasts with active nuclei and open chromatin, many of the dormant lymphoma cells were small, having nonactive nuclei and clumped chromatin. It thus appears that dormant BCL‐1 cells retain a malignant genotype but display a phenotype that resembles normal B cells (Yefenof et al., 1993b). Preimmunization with the Id is not the only manipulation that induces a dormant state in BCL‐1 neoplasia. One can achieve the same effect in SCID mice by passive administration of the antibody (Racila et al., 1995). In this case, any antibody to the Ig receptor will maintain the BCL‐1 tumor dormant. The antibody can be monoclonal or a polyclonal, directed against the Id, light chain, or heavy chain of the BCL‐1 Ig receptor. The establishment of tumor dormancy in SCID mice demonstrated that antibody alone is sufficient to induce a BCL‐1 dormant state with no direct involvement of T cell immunity, although T cells might be involved in prolonging dormancy (Farrar et al., 1999). Evidence indicates that this is also the case in Id‐immunized immunocompetent mice, since depletion of T regulatory cells or blockade of CTLA‐4 does not affect a BCL‐1 dormant state (Pop et al., 2005). Dormancy in SCID mice enabled exclusive identification of lþ dormant BCL‐1 cells on a background of undetectable autologous B lymphocytes in the SCID spleen. Flow cytometry analysis on gated lþ cells of such mice indicated that most of them were small (resembling the size of resting B lymphocytes) in comparison to growing BCL‐1 cells. Cell cycle analysis revealed a striking difference between growing and dormant cells. The first were cycling with about 40% of the cells in S and G2, whereas the majority of the dormant cells were arrested in G0/G1. Still, few of the dormant cells were cycling, but their expansion has been limited by apoptotic death following interaction with anti‐Ig antibodies. These data suggested that antibodies interacting with the Ig receptor on the BCL‐1 cells either mediate effector function or act as an agonistic ligand that induce growth inhibitory signals in the cells. This latter mode of interaction is a well‐known feature of Ig receptors expressed on premature B‐lymphocytes that undergo clonal anergy or apoptosis when interacting with self‐antigens prior to their differentiation into immunocompetent cells (Gaur et al., 1993). It is a physiological mechanism that prevents the generation of autoreactive B cells during maturation (Hartley et al., 1993). This response remains as a vestige in B‐lymphoma cells with a premature phenotype (e.g., BCL‐1) and can be mimicked by anti‐Ig antibodies.

Cancer Dormancy

195

V. THE PROSTATE ADENOCARCINOMA MODEL The question ensuing from the above conclusion is whether antibody‐ induced dormancy is a phenomenon restricted to B lymphoid tumors or can it occur in nonlymphoid cancers, for example carcinomas. To answer this question we studied the effect of anti‐HER‐2 antibodies on the growth of prostate adenocarcinoma (PAC) cells. HER‐2 is a transmembrane protein expressed on a variety of carcinomas including prostate cancer (Coussens et al., 1985). It is a tyrosine kinase belonging to the EGF receptor family (Kraus et al., 1989). Whereas HER‐2 does not have a specific ligand, it dimerizes with HER‐3 or HER‐4, which are devoid of an autonomous catalytic unit, and by that virtue enables signal transduction induced by EGF (Yarden, 2001). Once the ligand binds, the heterodimer recruits phosphatidylinositol‐30 ‐kinase (PI3K), which then phosphorylates phosphatidylinositols on the inner membrane. These become docking sites for PDK and AKT kinases, which are activated to phosphorylate IKK, leading to NF‐B nuclear translocation and transcriptional activation. At the same time, these kinases phosphorylate the pro‐apoptotic proteins caspase 9 and BAD by which apoptotic death is inhibited (Schlessinger, 2002). The net outcome of this enzymatic reactivity is continuous growth and proliferation of the affected cells. HER‐2 is overexpressed in approximately 20% of human breast carcinomas. Overexpression correlates with resistance to chemotherapy, more aggressive growth, and poor prognosis of the disease (Slamon et al., 1989; Ware et al., 1991). Numerous studies have demonstrated that antibodies against HER‐2 induce growth inhibitory signals in breast cancer cells overexpressing HER‐2 (Bacus et al., 1992; Ghetie et al., 1997). Apparently, homodimerization of HER‐2 is followed by internalization and degradation of the receptor which is no longer available for interaction with HER‐3 or HER‐4 (Alroy and Yarden, 2000; Bacus et al., 2000; Harari and Yarden, 2000). This is the principle on which the drug Herceptin, a humanized anti‐HER‐2 antibody, has been developed for treatment of breast carcinomas when HER‐2 is overexpressed (Cohen, 1999; Colomer et al., 2001; Kearney and McPhail, 2000). The properties of HER‐2 and the consequences of its interaction with anti‐HER‐2 antibodies prompted us to evaluate whether its engagement under proper conditions might lead to growth inhibition and dormancy of PAC cells. To this end, we tested the effect of seven monoclonal antibodies (MAbs) to HER‐2 (Spiridon et al., 2002) on the growth of HER‐2‐positive LNCAP, DU145, and PC3 PAC cell lines. However, none of these antibodies inhibited the proliferation of the PAC cells in vitro. Since these MAbs bound efficiently to the PAC cells, we conjugated them with the A‐chain of the

196

Rosalia Rabinovsky et al.

plant toxin ricin (RTA) and treated the cells with the resulting immunotoxins (ITs). A large body of work has demonstrated that ITs prepared with antibodies bound to RTA are potent inhibitors of cancer cell growth both in vivo and in vitro (Ghetie et al., 1995; Schnell et al., 2003; Vitetta and Thorpe, 1991). Accordingly, anti‐HER‐2 ITs were assayed in vitro. All three PAC lines were growth inhibited when exposed to the ITs, as determined by 3 H‐thymidine uptake and cell counting. Of the seven ITs tested, the HER66‐ deglycosylated (dg) RTA IT has been most effective. We therefore chose HER66‐dgRTA for further experimentation. What is the reason for growth inhibition of PAC cells by HER66‐dgRTA? Several tests have confirmed that LNCAP cells are readily apoptosed when interacting with HER66‐dgRTA at a concentration as low as 108 M. HER81‐dgRTA, which was not growth inhibitory at that concentration, was also poor as inducer of apoptosis. A different pattern emerged when HER66‐dgRTA was tested on DU145 PAC cells. These cells remained alive for more than 72 h, but were cycle arrested at G2. There is, however, a fundamental difference between the lines. Contrary to LNCAP, DU145 has a mutated p53 and lacks the pro‐apoptotic protein BAX (Carroll et al., 1993; Ouyang et al., 2001; Rampino et al., 1997). The IT does not provoke DNA damage but inhibits protein synthesis (Olsnes and Kozlov, 2001). Consequently, cells that are in G1 continue cycling but are arrested at G2, a checkpoint at which progression to mitosis requires de novo protein synthesis. Eventually, the growth‐arrested DU145 cells die of necrosis, as indicted by negative TUNEL and Annexin‐V staining. Hence, there are two outcomes following HER66‐dgRTA treatment of the PAC lines, depending on the cell type tested: cycle arrest and apoptotic death. It has been elaborated in the past that ricin induces cycle arrest followed by necrotic death (Endo and Tsurugi, 1987; Endo et al., 1987). However, data indicate that many cell types respond to ricin with apoptosis, as observed in our study using LNCAP cells (Gan et al., 2000; Oda et al., 1999; Williams et al., 1997). Evidently, engagement of HER‐2 on PAC cells is not inhibitory per se as in HER‐2 overexpressing breast carcinoma cells. This may explain why Herceptin is not effective in the treatment of prostate carcinoma (Lara et al., 2004). On the other hand, binding of HER66‐dgA is followed by internalization of the HER‐2–IT complex. HER‐2 dissociates in the cytosol and is degraded by the proteasome, while the toxin translocates to the ER and inhibits protein synthesis. What effect has the HER‐2‐specific IT on the progression of prostate carcinoma tumors in vivo? To address this question, we transfected LNCAP cells with a luciferase (luc) carrying retroviral vector and used the luc‐ positive cells for intraprostate injection in SCID/Beige mice (Honigman et al., 2001). This manipulation allowed the monitoring of growth and

Cancer Dormancy

197

spread of an orthotopic prostate tumor under a cooled charged‐coupled device (CCCD) camera when the mouse is injected with luciferin. Follow up of individual mice by repeated determinations revealed a progressive growth of LNCAP tumor that begins in the prostate and metastasizes to the lymph nodes, liver, and lungs in a course of 3 months postinoculation. Injection of HER66‐dgRTA into tumor‐bearing mice once or twice resulted with substantial decline in the total tumor mass. However, the tumor resumed a progressive growth pattern as the injection of the drug has been discontinued. To prolong the inhibitory effect of the IT, HER66‐dgRTA was loaded unto ALZET osmotic pumps which were implanted subcutaneously in tumor‐ bearing mice 20 days after tumor cell inoculation. Such mice presented with gradual regression and complete remission to a state of undetectable tumor. The mice were not, however, cured of the disease as transfer of their lung cells to untreated SCID/Beige recipients resulted with growth of prostate tumors that were luc‐positive. Hence, continuous treatment of mice bearing a PAC tumor with HER66‐dgRTA induces and maintains a tumor dormant state as long as the IT is infused.

VI. CONCLUDING REMARKS We have summarized experimental studies indicating that certain properties of signaling receptors expressed on cancer cells can be manipulated to alter the malignant phenotype of the cells. In a B cell lymphoma, it is sufficient to engage an Ig receptor in order to induce cell cycle arrest and apoptotic death in vivo that results in long‐term dormancy. In PAC, engagement of HER‐2 is insufficient to evoke an antitumor response. However, the dynamic nature of the receptor can be harnessed to internalize a toxin that induces apoptosis or cycle arrest, thus reducing the tumor mass and balancing expansion of residual micrometastases, keeping them dormant over time. In their joint paper Eva and George Klein reassess the traditional concept of immune surveillance against cancer (Klein and Klein, 2005). They conclude that immune rejection of cancerous cells may be effective against virally induced tumors, which are recognized as “nonself” by the host. Nonvirally induced tumors, constituting the majority of human cancers, are considered “self” and therefore ignored by the immune system. There exist, however, other protective mechanisms such as DNA repair (Cleaver, 2005; Eyfjord and Bodvarsdottir, 2005), apoptosis (Igney and Krammer, 2002; Klein, 2004), and tumor–stromal interaction (Rubin, 2003) that suppress the growth of the latter. This nonimmunological surveillance is robust, comprehensive, and effective in combating the emergence of precancerous cells or their progression to a full‐fledged disease.

198

Rosalia Rabinovsky et al.

The above conclusion does not preclude the possibility of manipulating various components of the immune response, for example antibodies, T cells, NK cells, dendritic cells, cytokines, and so on, to affect the growth of nonimmunogenic tumors as well. The advent of MAbs provides an insight to the scope and limits of this approach. Dozens of MAbs have been raised against “tumor‐ specific antigens,” which specifically target cancerous cells while ignoring their normal counterparts. However, none of these MAbs has been proven useful for therapeutic purposes. On the other hand, several MAbs directed against cell surface receptors, for example HER‐2, CD19, CD22, and B cell idiotypes, are in clinical use with significant therapeutic benefit (Bhattacharya‐Chatterjee et al., 2002; Cesano and Gayko, 2003; Coscia et al., 2004; Emens, 2005; Leonard and Link, 2002; Maloney, 2005; Messmann et al., 2000; Nguyen et al., 1999; von Schilling, 2003). It thus appears that the absolute distinction between malignant and normal cells is not a prerequisite when considering targeting of tumors. The data summarized in the present chapter demonstrate that the effectiveness of antireceptor antibodies as antitumor drugs is not necessarily a consequence of their ability to recruit immune effector mechanisms such as complement or ADCC, but depends primarily on their activity as agonistic ligands. Binding of such MAbs may dimerize and excite the receptor to relay growth inhibitory signals or facilitate the internalization of an anticancer drug that activates cell death from within. Hence, even if the immune system per se has no direct role in tumor rejection, its powerful capacity to generate high affinity‐specific antibodies can be recruited to activate effective intracellular surveillance that maintains a tumor dormant state over a long period of time.

ACKNOWLEDGMENTS The authors wish to thank Ms. Daniella Krauthamer for secretarial assistance. Studies summarized in this chapter were supported by the Prostate Cancer Foundation and Concern Foundation.

REFERENCES Alroy, I., and Yarden, Y. (2000). Biochemistry of HER2 oncogenesis in breast cancer. Breast Dis. 11, 31–48. Bacus, S. S., Stancovski, I., Huberman, E., Chin, D., Hurwitz, E., Mills, G. B., Ullrich, A., Sela, M., and Yarden, Y. (1992). Tumor‐inhibitory monoclonal antibodies to the HER‐2/Neu receptor induce differentiation of human breast cancer cells. Cancer Res. 52, 2580–2589.

Cancer Dormancy

199

Bacus, S. S., Gudkov, A. V., Esteva, F. J., and Yarden, Y. (2000). Expression of erbB receptors and their ligands in breast cancer: Implications to biological behavior and therapeutic response. Breast Dis. 11, 63–75. Bhattacharya‐Chatterjee, M., Chatterjee, S. K., and Foon, K. A. (2002). Anti‐idiotype antibody vaccine therapy for cancer. Expert Opin. Biol. Ther. 2, 869–881. Bilchik, A., Miyashiro, M., Kelley, M., Kuo, C., Fujiwara, Y., Nakamori, S., Monden, M., and Hoon, D. S. (2000). Molecular detection of metastatic pancreatic carcinoma cells using a multimarker reverse transcriptase‐polymerase chain reaction assay. Cancer 88, 1037–1044. Braun, S., and Harbeck, N. (2001). Recent advances in technologies for the detection of occult metastatic cells in bone marrow of breast cancer patients. Breast Cancer Res. 3, 285–288. Carroll, A. G., Voeller, H. J., Sugars, L., and Gelmann, E. P. (1993). p53 oncogene mutations in three human prostate cancer cell lines. Prostate 23, 123–134. Cesano, A., and Gayko, U. (2003). CD22 as a target of passive immunotherapy. Semin. Oncol. 30, 253–257. Cleaver, J. E. (2005). Splitting hairs—discovery of a new DNA repair and transcription factor for the human disease trichothiodystrophy. DNA Repair (Amst.) 4, 285–287. Cohen, R. L. (1999). Herceptin: Breaking new ground. Cancer Biother. Radiopharm. 14, 1–4. Colomer, R., Shamon, L. A., Tsai, M. S., and Lupu, R. (2001). Herceptin: From the bench to the clinic. Cancer Invest. 19, 49–56. Coscia, M., Mariani, S., Battaglio, S., Di Bello, C., Fiore, F., Foglietta, M., Pileri, A., Boccadoro, M., and Massaia, M. (2004). Long‐term follow‐up of idiotype vaccination in human myeloma as a maintenance therapy after high‐dose chemotherapy. Leukemia 18, 139–145. Coussens, L., Yang‐Feng, T. L., Liao, Y. C., Chen, E., Gray, A., McGrath, J., Seeburg, P. H., Libermann, T. A., Schlessinger, J., Francke, U., Levinson, A., and Ullrich, A. (1985). Tyrosine kinase receptor with extensive homology to EGF receptor shares chromosomal location with neu oncogene. Science 230, 1132–1139. Emens, L. A. (2005). Trastuzumab: Targeted therapy for the management of HER‐2/neu‐ overexpressing metastatic breast cancer. Am. J. Ther. 12, 243–253. Endo, Y., and Tsurugi, K. (1987). RNA N‐glycosidase activity of ricin A‐chain. Mechanism of action of the toxic lectin ricin on eukaryotic ribosomes. J. Biol. Chem. 262, 8128–8130. Endo, Y., Mitsui, K., Motizuki, M., and Tsurugi, K. (1987). The mechanism of action of ricin and related toxic lectins on eukaryotic ribosomes. The site and the characteristics of the modification in 28 S ribosomal RNA caused by the toxins. J. Biol. Chem. 262, 5908–5912. Eyfjord, J. E., and Bodvarsdottir, S. K. (2005). Genomic instability and cancer: Networks involved in response to DNA damage. Mutat. Res. 592, 18–28. Farrar, J. D., Katz, K. H., Windsor, J., Thrush, G., Scheuermann, R. H., Uhr, J. W., and Street, N. E. (1999). Cancer dormancy VII: A regulatory role for CD8þ T cells and IFN‐ in establishing and maintaining the tumor‐dormant state. J. Immunol. 162, 2842–2849. Gan, Y. H., Peng, S. Q., and Liu, H. Y. (2000). Molecular mechanism of apoptosis induced by ricin in HeLa cells. Acta Pharmacol. Sin. 21, 243–248. Gasparini, G., Biganzoli, E., Bonoldi, E., Morabito, A., Fanelli, M., and Boracchi, P. (2001). Angiogenesis sustains tumor dormancy in patients with breast cancer treated with adjuvant chemotherapy. Breast Cancer Res. Treat. 65, 71–75. Gaur, A., Yao, X. R., and Scott, D. W. (1993). B cell tolerance induction by cross‐linking of membrane IgM, but not IgD, and synergy by cross‐linking of both isotypes. J. Immunol. 150, 1663–1669. George, A. J., Tutt, A. L., and Stevenson, F. K. (1987). Anti‐idiotypic mechanisms involved in suppression of a mouse B cell lymphoma, BCL1. J. Immunol. 138, 628–634. Ghetie, M. A., Podar, E. M., Ilgen, A., Gordon, B. E., Uhr, J. W., and Vitetta, E. S. (1997). Homodimerization of tumor‐reactive monoclonal antibodies markedly increases their ability

200

Rosalia Rabinovsky et al.

to induce growth arrest or apoptosis of tumor cells. Proc. Natl. Acad. Sci. USA 94, 7509–7514. Ghetie, V., Engert, A., Schnell, R., and Vitetta, E. S. (1995). The in vivo anti‐tumor activity of immunotoxins containing two versus one deglycosylated ricin A chains. Cancer Lett. 98, 97–101. Harari, D., and Yarden, Y. (2000). Molecular mechanisms underlying ErbB2/HER2 action in breast cancer. Oncogene 19, 6102–6114. Hartley, S. B., Cooke, M. P., Fulcher, D. A., Harris, A. W., Cory, S., Basten, A., and Goodnow, C. C. (1993). Elimination of self‐reactive B lymphocytes proceeds in two stages: Arrested development and cell death. Cell 72, 325–335. Honigman, A., Zeira, E., Ohana, P., Abramovitz, R., Tavor, E., Bar, I., Zilberman, Y., Rabinovsky, R., Gazit, D., Joseph, A., Panet, A., Shai, E., et al. (2001). Imaging transgene expression in live animals. Mol. Ther. 4, 239–249. Igney, F. H., and Krammer, P. H. (2002). Death and anti‐death: Tumor resistance to apoptosis. Nat. Rev. Cancer 2, 227–288. Kearney, N., and McPhail, G. (2000). Herceptin: Implications for breast cancer management. Eur. J. Oncol. Nurs. 4, 37–41. Klein, G. (2004). Cancer, apoptosis, and nonimmune surveillance. Cell Death Differ. 11, 13–17. Klein, G., and Klein, E. (2005). Surveillance against tumors—is it mainly immunological? Immunol. Lett. 100, 29–33. Kraus, M. H., Issing, W., Miki, T., Popescu, N. C., and Aaronson, S. A. (1989). Isolation and characterization of ERBB3, a third member of the ERBB/epidermal growth factor receptor family: Evidence for overexpression in a subset of human mammary tumors. Proc. Natl. Acad. Sci. USA 86, 9193–9197. Lara, P. N., Jr., Chee, K. G., Longmate, J., Ruel, C., Meyers, F. J., Gray, C. R., Edwards, R. G., Gumerlock, P. H., Twardowski, P., Doroshow, J. H., and Gandara, D. R. (2004). Trastuzumab plus docetaxel in HER‐2/neu‐positive prostate carcinoma: Final results from the California Cancer Consortium Screening and Phase II Trial. Cancer 100, 2125–2131. Leonard, J. P., and Link, B. K. (2002). Immunotherapy of non‐Hodgkin’s lymphoma with hLL2 (epratuzumab, an anti‐CD22 monoclonal antibody) and Hu1D10 (apolizumab). Semin. Oncol. 29, 81–86. Maloney, D. G. (2005). Immunotherapy for non‐Hodgkin‐s lymphoma: Monoclonal antibodies and vaccines. Semin. Oncol. 32, S19–S26. Mejean, A., Vona, G., Nalpas, B., Damotte, D., Brousse, N., Chretien, Y., Dufour, B., Lacour, B., Brechot, C., and Paterlini‐Brechot, P. (2000). Detection of circulating prostate derived cells in patients with prostate adenocarcinoma is an independent risk factor for tumor recurrence. J. Urol. 163, 2022–2029. Meng, S., Tripathy, D., Frenkel, E. P., Shete, S., Naftalis, E. Z., Huth, J. F., Beitsch, P. D., Leitch, M., Hoover, S., Euhus, D., Haley, B., Morrison, L., et al. (2004). Circulating tumor cells in patients with breast cancer dormancy. Clin. Cancer Res. 10, 8152–8162. Messmann, R. A., Vitetta, E. S., Headlee, D., Senderowicz, A. M., Figg, W. D., Schindler, J., Michiel, D. F., Creekmore, S., Steinberg, S. M., Kohler, D., Jaffe, E. S., Stetler‐Stevenson, M., et al. (2000). A phase I study of combination therapy with immunotoxins IgG‐HD37‐ deglycosylated ricin A chain (dgA) and IgG‐RFB4‐dgA (Combotox) in patients with refractory CD19(þ), CD22(þ) B cell lymphoma. Clin. Cancer Res. 6, 1302–1313. Nguyen, D. T., Amess, J. A., Doughty, H., Hendry, L., and Diamond, L. W. (1999). IDEC‐ C2B8 anti‐CD20 (rituximab) immunotherapy in patients with low‐grade non‐Hodgkin’s lymphoma and lymphoproliferative disorders: Evaluation of response on 48 patients. Eur. J. Haematol. 62, 76–82.

Cancer Dormancy

201

O’Reilly, M. S., Holmgren, L., Shing, Y., Chen, C., Rosenthal, R. A., Moses, M., Lane, W. S., Cao, Y., Sage, E. H., and Folkman, J. (1994). Angiostatin: A novel angiogenesis inhibitor that mediates the suppression of metastases by a Lewis lung carcinoma. Cell 79, 315–328. O’Reilly, M. S., Boehm, T., Shing, Y., Fukai, N., Vasios, G., Lane, W. S., Flynn, E., Birkhead, J. R., Olsen, B. R., and Folkman, J. (1997). Endostatin: An endogenous inhibitor of angiogenesis and tumor growth. Cell 88, 277–285. Oda, T., Iwaoka, J., Komatsu, N., and Muramatsu, T. (1999). Involvement of N‐acetylcysteine‐ sensitive pathways in ricin‐induced apoptotic cell death in U937 cells. Biosci. Biotechnol. Biochem. 63, 341–348. Olsnes, S., and Kozlov, J. V. (2001). Ricin. Toxicon 39, 1723–1728. Ouyang, X., Wang, X., Xu, K., Jin, D. Y., Cheung, A. L., Tsao, S. W., and Wong, Y. C. (2001). Effect of p53 on centrosome amplification in prostate cancer cells. Biochim. Biophys. Acta 1541, 212–220. Pantel, K., Cote, R. J., and Fodstad, O. (1999). Clinical Detection and importance of micrometastatic disease. J. Natl. Cancer Inst. 91, 1113–1124. Pop, L. M., Smallshaw, J. E., Tucker, T. F., Stevenson, F. K., and Vitetta, E. S. (2005). Failure of vaccination with idiotypic protein or DNA, (þ/IL‐2), the depletion of regulatory T cells, or the blockade of CTLA‐4 to prolong dormancy in mice with BCL1 lymphoma. J. Immunother. 28, 525–534. Racila, E., Scheuermann, R. H., Picker, L. J., Yefenof, E., Tucker, T., Chang, W., Marches, R., Street, N. E., Vitetta, E. S., and Uhr, J. W. (1995). Tumor dormancy and cell signaling II: Antibody as an agonist in inducing dormancy of a B cell lymphoma in SCID mice. J. Exp. Med. 181, 1539–1550. Racila, E., Euhus, D., Weiss, A. J., Rao, C., McConnell, J., Terstappen, L. W., and Uhr, J. W. (1998). Detection and characterization of carcinoma cells in the blood. Proc. Nat. Acad. Sci. USA 95, 4589–4594. Rampino, N., Yamamoto, H., Ionov, Y., Li, Y., Sawai, H., Reed, J. C., and Perucho, M. (1997). Somatic frameshift mutations in the BAX gene in colon cancers of the microsatellite mutator phenotype. Science 275, 967–969. Riethmuller, G., and Klein, C. A. (2001). Early cancer cell dissemination and late metastatic relapse: Clinical reflections and biological approaches to the dormancy problem in patients. Semin. Cancer Biol. 11, 307–311. Rubin, H. (2003). Microenvironmental regulation of the initiated cell. Adv. Cancer Res. 90, 1–62. Sagalowsky, A. I., and Molberg, K. (1999). Solitary metastasis of renal cell carcinoma to the contralateral adrenal gland 22 years after nephrectomy. Urology 54, 162. Schlessinger, J. (2002). Ligand‐induced, receptor‐mediated dimerization and activation of EGF receptor. Cell 110, 669–672. Schnell, R., Borchmann, P., Staak, J. O., Schindler, J., Ghetie, V., Vitetta, E. S., and Engert, A. (2003). Clinical evaluation of ricin A‐chain immunotoxins in patients with Hodgkin’s lymphoma. Ann. Oncol. 14, 729–736. Siu, H., Vitetta, E. S., May, R. D., and Uhr, J. W. (1986). Tumor dormancy. I. Regression of BCL1 tumor and induction of a dormant tumor state in mice chimeric at the major histocompatibility complex. J. Immunol. 137, 1376–1382. Slamon, D. J., Godolphin, W., Jones, L. A., Holt, J. A., Wong, S. G., Keith, D. E., Levin, W. J., Stuart, S. G., Udove, J., Ullrich, A., et al. (1989). Studies of the HER‐2/neu proto‐oncogene in human breast and ovarian cancer. Science 244, 707–712. Slavin, S., and Strober, S. (1978). Spontaneous murine B‐cell leukaemia. Nature 272, 624–626. Spiridon, C. I., Ghetie, M. A., Uhr, J., Marches, R., Li, J. L., Shen, G. L., and Vitetta, E. S. (2002). Targeting multiple Her‐2 epitopes with monoclonal antibodies results in improved

202

Rosalia Rabinovsky et al.

antigrowth activity of a human breast cancer cell line in vitro and in vivo. Clin. Cancer Res. 8, 1720–1730. Talpaz, M., Estrov, Z., Kantarjian, H., Ku, S., Foteh, A., and Kurzrock, R. (1994). Persistence of dormant leukemic progenitors during interferon‐induced remission in chronic myelogenous leukemia. Analysis by polymerase chain reaction of individual colonies. J. Clin. Invest. 94, 1383–1389. Uhr, J. W., Tucker, T., May, R. D., Siu, H., and Vitetta, E. S. (1991). Cancer dormancy: Studies of the murine BCL1 lymphoma. Cancer Res. 51, 5045s–5053s. Vitetta, E. S., and Thorpe, P. E. (1991). Immunotoxins containing ricin or its A chain. Semin. Cell Biol. 2, 47–58. Vitetta, E. S., Tucker, T. F., Racila, E., Huang, Y. W., Marchers, R., Lane, N., Scheuermann, R. H., Street, N. E., Watanobe, T., and Uhr, J. W. (1997). Tumor dormancy and cell signaling. V: Regrowth of the BCL1 tumor after dormancy is established. Blood 89, 4425–4436. von Schilling, C. (2003). Immunotherapy with anti‐CD20 compounds. Semin. Cancer Biol. 13, 211–222. Ware, J. L., Maygarden, S. J., Koontz, W. W., Jr., and Strom, S. C. (1991). Immunohistochemical detection of c‐erbB‐2 protein in human benign and neoplastic prostate. Hum. Pathol. 22, 254–258. Williams, J. M., Lea, N., Lord, J. M., Roberts, L. M., Milford, D. V., and Taylor, C. M. (1997). Comparison of ribosome‐inactivating proteins in the induction of apoptosis. Toxicol. Lett. 91, 121–127. Yarden, Y. (2001). Biology of HER2 and its importance in breast cancer. Oncology 61(Suppl. 2), 1–13. Yefenof, E., Picker, L. J., Scheuermann, R. H., Tucker, T. F., Vitetta, E. S., and Uhr, J. W. (1993a). Cancer dormancy: Isolation and characterization of dormant lymphoma cells. Proc. Natl. Acad. Sci. USA 90, 1829–1833. Yefenof, E., Picker, L. J., Scheuermann, R. H., Vitetta, E. S., Street, N. E., Tucker, T. F., and Uhr, J. W. (1993b). Induction of B cell tumor dormancy by anti‐idiotypic antibodies. Curr. Opin. Immunol. 5, 740–744. Zippelius, A., and Pantel, K. (2000). RT‐PCR‐based detection of occult disseminated tumor cells in peripheral blood and bone marrow of patients with solid tumors. An overview. Ann. NY Acad. Sci. 906, 110–123.

Therapeutic Targets of Multiple Angiogenic Factors for the Treatment of Cancer and Metastasis Yihai Cao* and Qi Liu{ *Laboratory of Angiogenesis Research, Microbiology and Tumor Biology Center, Karolinska Institutet, 171 77 Stockholm, Sweden; Shandong Provincial Hospital, Jinan, Shandong Province, People’s Republic of China

{

I. Introduction II. Tumor Angiogenesis A. Tumor Blood Vessels B. Tumor‐Produced Angiogenic Factors C. VEGF Family and VEGF Receptors D. VEGF‐A‐Induced Angiogenesis and Permeability E. Non‐VEGF Angiogenic Factors III. Angiogenesis Inhibitors A. Growth Factor Antagonists B. VEGF‐A Antagonists C. Antagonists for Non‐VEGF Factors D. Broad‐Spectrum Endogenous Inhibitors E. Oral Angiogenesis Inhibitors IV. Lymphangiogenesis and Lymphatic Metastasis V. Clinical Development of Antiangiogenic Drugs VI. Conclusions and Perspectives References

Like any growing healthy tissues, tumors build up their blood vessels by three mechanisms: angiogenesis, vasculogenesis, and intersucception. Vascular endothelial growth factor‐A (VEGF‐A) is one of the key factors responsible for stimulation and maintenance of the disorganized, leaky, and torturous tumor vasculature. In addition to VEGF‐A, tumors produce multiple other factors to stimulate blood vessel growth. These include members in the platelet‐derived growth factor (PDGF), fibroblast growth factor (FGF), VEGF‐C, insulin‐like growth factor (IGF), angiopoietin (Ang), and hepatocyte growth factor (HGF) families. Recent studies show that these angiogenic factors can also promote lymphangiogenesis and potentially lymphatic metastasis. Understanding the roles of individual and combined angiogenic factors in promoting tumor angiogenesis is crucial for defining therapeutic targets and antiangiogenic drug development for the treatment of cancer. # 2007 Elsevier Inc.

Advances in CANCER RESEARCH Copyright 2007, Elsevier Inc. All rights reserved.

203

0065-230X/07 $35.00 DOI: 10.1016/S0065-230X(06)97009-2

204

Yihai Cao and Qi Liu

I. INTRODUCTION The idea that suppression of tumor blood vessel growth might be a valid approach for cancer therapy was raised more than 30 years ago (Folkman, 1971). Now we know that tumors produce multiple factors to stimulate the growth of blood vessels, which are essential for supporting tumor growth and metastasis. As the information of human genome becomes available, the identity of these tumor‐derived angiogenic factors has been uncovered. Although individual roles of these angiogenic factors in promoting tumor blood vessel growth have been relatively well studied, little is known about the interplay between different growth factors. It appears that the angiogenic activity of most tumor‐derived growth factors is mediated by their tyrosine kinase receptors expressed on endothelial cells and the blood vessel‐associated cell types such as pericytes and smooth muscle cells (Carmeliet, 2005; Ferrara and Kerbel, 2005). Identification and characterization of tumor‐derived angiogenic factors, their endothelial cell receptors, and signaling components have provided opportunities of therapeutic development of new drugs for the treatment of cancer (Ferrara and Kerbel, 2005). Avastin(bevacizumab), a humanized neutralizing antibody against vascular endothelial growth factor (VEGF), was the first successful antiangiogenic drug approved by the Food and Drug Agency (FDA) of the United States for the treatment of certain types of human cancer (Ellis, 2005). Followed by this success, several other anti‐VEGF protein molecules or small chemical compounds are at different stages of clinical trials for cancer therapy (Carmeliet, 2005; Ferrara and Kerbel, 2005). Although the outcome of these ongoing trials remains to be seen, it is predicted that they might effectively block the function of tumor‐produced VEGF and give rise to therapeutic benefits for cancer patients. Unlike VEGF, tumor‐produced non‐VEGF factors are relatively less specific for endothelial cells. For example, members in the fibroblast growth factor (FGF), insulin‐like growth factor (IGF), hepatocyte growth factor (HGF), and EGF families usually have broad biological functions by targeting different cell types in the body. They have also become attractive targets for therapeutic development (Cao, 2005b; Carmeliet, 2005; Ferrara and Kerbel, 2005). In addition to growth factor antagonists, many endogenous angiogenesis inhibitors have also been identified (Cao, 2001; Folkman, 2004; Nyberg et al., 2005). Although these endogenous angiogenesis inhibitors might be detected at low levels in tumor tissues, they are somewhat produced in association with tumor growth (Cao, 2001; Ferrara and Kerbel, 2005; Folkman, 2004; Nyberg et al., 2005). In principle, these endogenous angiogenesis inhibitors (also called direct angiogenesis inhibitors) might block broad‐spectrum pathways triggered by multiple angiogenic factors (Folkman, 2006). Thus, they should in principle be more effective than single angiogenic factor antagonists in the treatment of

Therapeutic Targets for the Treatment of Cancer and Metastasis

205

cancer. Endostatin is the first approved endogenous antiangiogenic drug available in China for the treatment of human cancers (Sun et al., 2005). Unlike indirect angiogenesis inhibitors (growth factor antagonists), the molecular mechanisms of this class of molecules remain an enigma although several possible mechanisms of different endogenous angiogenesis inhibitors have been proposed. It remains to be seen if these endogenous angiogenesis inhibitors have advantages over growth factor antagonists such as Avastin in cancer patients. Tumors also produce a range of growth factors that are able to induce lymphangiogenesis (Cao, 2005b). Unlike blood vessels, the role of lymphatic vessels in the tumor tissue remains to be characterized. The lymphatic endothelial cells might have cross talks with tumor cells and blood vessel endothelial cells by producing growth factors (Cao, 2005b). The intimate interaction between lymphatic endothelial cells and tumor cells might facilitate lymphatic metastasis (Achen et al., 2005; Alitalo et al., 2005; Cao, 2005b). The fact that lymphatic vessels can grow without blood vessels suggests that separate therapeutic molecules specifically targeting lymphatic vessels should be developed for prevention and treatment of lymphatic metastasis (Bjorndahl et al., 2005b; Cao, 2005b; Chang et al., 2004).

II. TUMOR ANGIOGENESIS A. Tumor Blood Vessels Switching on an angiogenic phenotype is essential for expansion of malignant tissues. Tumors are able to secrete a range of angiogenic factors that stimulate the angiogenic sprouting from preexisting blood vessels in the surrounding healthy tissues (Cao, 2005b). Tumor blood vessels are also able to recruit circulating endothelial precursor cells differentiated from bone marrow stem cells (Lyden et al., 2001). Tumor tissues mimic developing embryonic tissues in the aspects that both tumor cells and early embryonic cells remain at an undifferentiated stage and they share a common mechanism of building up new blood vessels. It has also been reported that tumor tissues have the ability to divide large “mother” vessels into smaller “daughter” vessels by growing interstitial tissues in the lumens of large vessels (Feng et al., 2000; Jain, 2005). Although tumors utilize the similar mechanisms as healthy growing tissues to build up their vasculatures, there are fundamental differences between tumor and healthy vasculatures. Tumor blood vessels are usually disorganized, tortuous, leaky,

206

Yihai Cao and Qi Liu

hemorrhagic, and lacking clear separation between arteries and veins (Carmeliet, 2000; Jain, 2005). As a result of the abnormal architecture, the blood flow in tumor vessels is also chaotic. It has been reported that blood stream within a tumor tissue could move forward and backward within the same vessel. In addition to blood vessel endothelial cells, tumor cells might participate in the formation of their blood vessels. In some extreme cases, the tumor vasculature is exclusively formed by tumor cells (a process named vascular mimicry) (Bissell, 1999; Folberg et al., 2000; Maniotis et al., 1999). It is unclear how important the tumor vascular mimicry in contribution of tumor growth and metastasis. The malformation and leaky feature of the tumor vasculature is also reflected by the fact that interstitial fluid pressure (IFP) in the tumor tissue is relatively high (Jain, 2005). The high interstitial pressure might restrict blood flow within tumor blood vessels. This is perhaps one of the reasons why the tumor tissue is hypoxic. Although the density of microvessels within a tumor is relatively high, many of these vessels might be in a poorly perfused or even occlusive state.

B. Tumor‐Produced Angiogenic Factors The tumor tissue contains heterogenous populations of malignant cells with diversity of genetic alterations (Carmeliet, 2005). In addition to tumor cells, stromal cells of fibroblast origins, inflammatory cells, blood vessel endothelial cells, and smooth muscle cells are also frequently found within the tumor tissue. Various cell types and various genetic alterations within the tumor tissue determine the expression of diversity of growth factors that might mediate cross‐communications between these cell types. The most common angiogenic factors detected in the tumor tissue include members of the VEGF, FGF, platelet‐derived growth factor (PDGF), angiopoietin (Ang), HGF, and IGF families (Fig. 1). These angiogenic factors usually exhibit robust angiogenic activity in in vitro and in vivo angiogenic models (Cao et al., 2003). Several of these angiogenic factors can also induce the growth and survival of tumor cells. Thus, they have overlapping functions on blood vessel endothelial cells and tumor cells. Different angiogenic factors in the tumor tissue might synergistically stimulate angiogenesis. For example, angiogenic synergisms between FGF‐2 and VEGF‐A, FGF‐2 and PDGF‐BB, and PDGF‐BB and VEGF‐A have been reported (Cao et al., 2003; Kano et al., 2005; Nico et al., 2001; Richardson et al., 2001). For therapeutic development of effective angiogenic factor antagonists, it is pivotal to block functions of multiple angiogenic factors produced by tumors (Fig. 1).

Therapeutic Targets for the Treatment of Cancer and Metastasis

207

Tumor

Angiogenic factors i.e., VEGF, FGF, IGF, HGF, PDGF, ANg., etc. Angiogenic factor and receptor antagonists i.e., nAbs, sRs, aptamers, PTKIs, etc.

Broad-spectrum endogeneous inhibitors i.e. endostatin, angiostatin, etc.

Fig. 1 Tumors produce multiple angiogenic factors that individually or synergistically stimulate blood vessel growth. Growth factor antagonists including neutralizing antibodies, soluble receptors, and PTKIs block their respective growth factor‐induced angiogenesis. Broad‐ spectrum angiogenesis inhibitors including endogenous angiogenesis inhibitors such as angiostatin and endostatin that directly inhibit endothelial cell growth independent from the source of various angiogenic stimuli. PTKIs, protein tyrosine kinase inhibitors; sRs, soluble receptors; nAbs, neutralizing antibodies.

C. VEGF Family and VEGF Receptors Among all tumor‐produced angiogenic factors, members of the VEGF family, especially VEGF‐A, are most frequently overexpressed and well‐ studied growth factors (Ferrara, 2005). The VEGF family includes VEGF‐A,

208

Yihai Cao and Qi Liu

VEGF‐B, VEGF‐C, VEGF‐D, and placenta growth factor (PlGF) (Eriksson et al., 2002). They are all secretory proteins that contain eight conserved cysteine residues linking each monomeric molecule into a functional dimmer (Cao et al., 1996). These angiogenic factors exert their biological functions through interaction with three structurally related tyrosine kinase receptors, VEGFR‐1, VEGFR‐2, and VEGFR‐3. While VEGFR‐1 and VEGFR‐2 are relatively specifically expressed on blood vessel endothelial cells, VEGFR‐3 is specifically expressed on lymphatic endothelial cells (Joukov et al., 1996, 1997; Kanno et al., 2000; Sato et al., 2000). According to their biological functions and receptor‐binding patterns, the VEGF family can be divided into three subgroups. (1) VEGF‐A is the prototype of VEGF, which binds to both VEGFR‐1 and VEGFR‐2 (Fig. 2). VEGF‐A can be generated as several alternatively spliced isoforms with different heparin‐ binding affinity (Ferrara, 1995). In addition to stimulation of angiogenesis, VEGF‐A is also a potent vascular permeability factor (Senger et al., 1986, 1993). The expression level of VEGF‐A can be dramatically up‐regulated by hypoxia through activation of hypoxia inducible factor 1 (HIF‐1 ) (Cejudo‐Martin and Johnson, 2005; Gustafsson et al., 2005; Makino et al., 2001; Ryan et al., 2000). Deletion of only one allele of VEGF‐A gene in mice results in lethality of early embryos due to lacking blood islets and hematopoiesis (Carmeliet et al., 1996; Ferrara et al., 1996). It is believed that VEGFR‐2 is the functional receptor that transduces the angiogenic and vascular permeability signals for VEGF, whereas the VEGFR‐1‐ mediated biological function is not obvious (Waltenberger et al., 1994). It has been reported that VEGFR‐1 might act as a decoy receptor that does not transduce active angiogenic signals for VEGF‐A, but competes VEGF‐A for the VEGFR‐2 receptor (Clauss, 1998). (2) PlGF and VEGF‐B bind only to VEGFR‐1 and do not induce obvious angiogenic responses (Eriksson et al., 2002; Olofsson et al., 1999). Although PlGF and VEGF‐B do not induce obvious angiogenic responses in vitro and in vivo, these two factors can form heterodimers with VEGF‐A and modulate VEGF‐A‐induced angiogenesis. Deletion of PlGF and VEGF‐B genes in mice did not result in obvious phenotypes (Aase et al., 2001; Luttun et al., 2002). (3) VEGF‐C and VEGF‐D interact with both VEGFR‐2 and VEGFR‐3 receptors and thus are able to induce hemangiogenesis and lymphangiogenesis. It seems that VEGF‐C preferentially induce lymphangiogenesis when both blood and lymphatic vessels are simultaneously exposed to VEGF‐C (Oh et al., 1997).

D. VEGF‐A‐Induced Angiogenesis and Permeability In the mouse corneal angiogenesis model, VEGF‐A induces a disorganized vasculature that consists of leaky, hemorrhagic, and tortuous vascular

Therapeutic Targets for the Treatment of Cancer and Metastasis

VEGF-A

VEGFR1

VEGFR2

TK

209

- Neutralizing Abs - Soluble receptors

- Neutralizing Abs - Aptamers

TK

- PTKIs - Inhibitors of signaling components

Fig. 2 VEGF‐A‐induced angiogenesis is mainly mediated the VEGFR‐2 receptor. Different VEGF‐A antagonists including neutralizing antibodies, soluble receptors, aptamers, and

plexuses due to fusion of microvessels (Eriksson et al., 2003). This unique vascularization pattern is a fingerprint of VEGF‐A and no other factors could induce a similar vasculature (Cao et al., 2003; Eriksson et al., 2003). Electron microscopy analysis of VEGF‐A‐induced vessels reveals that the endothelium contains high numbers of fenestrations that mediate vascular permeability (Cao et al., 2004b; Eriksson et al., 2003). This type of leaky vasculature mimics those found in tumors (Qu et al., 1995; Roberts and Palade, 1995). Vascular permeability is not a prerequisite for angiogenesis since angiogenic factors such as FGF‐2 lacking an ability to induce vascular

210

Yihai Cao and Qi Liu

permeability can potently induce angiogenesis (Cao et al., 2004b; Eriksson et al., 2003). It seems that VEGF‐A‐stimulated angiogenesis and vascular permeability are mediated by the same VEGFR‐2 (Cao et al., 1998). Blockage of VEGF‐A‐induced vascular permeability might be an important approach for cancer therapy in combinations with existing anticancer drugs. The leaky tumor vasculature could contribute to the high IFP, which might restrict blood flow in the tumor vessels (Jain, 2005). Although tumors have relatively high numbers of blood vessels, the tumor environment is hypoxic. Thus, inhibition of the VEGF‐A‐induced vascular permeability would in principle decrease the tumor IFP and improve chemotherapeutic drug delivery via increasing blood flow. Recent signal transduction studies show that VEGF‐ A‐induced vascular permeability and angiogenesis can be separated. For example, Scr and Rac have been reported to specifically mediate the VEGF‐ induced vascular permeability but not angiogenesis (Eriksson et al., 2003; Paul et al., 2001; Schlessinger, 2000). However, it remains to be seen if Src and Rac antagonists might be able to potentiate the efficacy of chemotherapeutic drugs.

E. Non‐VEGF Angiogenic Factors Although VEGF‐A‐triggered signaling pathways have become a focus of angiogenesis research, tumors produce other angiogenic factors with redundant functions. Particularly, interplays between different factors may synergistically stimulate angiogenesis. For example, FGF‐2/VEGF‐A, FGF‐2/PDGF‐BB, and Ang‐1,2N/VEGF‐A have been reported to synergistically induce angiogenesis (Cao et al., 2003; Kano et al., 2005; Shyu et al., 2003). These findings suggest that these growth factors in combinations even expressed at low levels may induce robust angiogenic responses. The most frequently expressed non‐VEGF angiogenic factors by tumors include members of FGF, PDGF, IGF, HGF, and Ang families (Cao, 2005b). Since they bind to heparin with different affinities, these factors might act locally by interaction with heparan sulfate proteoglycans in the extracellular matrix or have systemic effects on remote tissues or organs. Because tumors produce multiple factors, it is pivotal to develop therapeutic strategies aimed to block these multiple factor‐induced angiogenesis for the treatment of malignant and nonmalignant diseases.

III. ANGIOGENESIS INHIBITORS As tumor angiogenesis is crucial for primary tumor growth and metastasis, angiogenesis inhibitors have therapeutic implications in the treatment of malignancies. As a result of joint efforts, hundreds of angiogenesis

Therapeutic Targets for the Treatment of Cancer and Metastasis

211

inhibitors including endogenous angiogenesis protein molecules, growth factor/receptor antagonists, small chemical compounds, and nucleotides have been identified and many of them have clinically been tested for their efficacies against human cancers. According to their mode of actions, angiogenesis inhibitors can be divided into two classes, growth factor antagonists and broad‐spectrum endogenous inhibitors.

A. Growth Factor Antagonists As a typical growth factor exerts its angiogenic activity through its tyrosine kinase receptors expressed on endothelial cells, antagonists blocking different levels of signaling pathways have been designed. These include neutralizing antibodies against different angiogenic factors, soluble receptor molecules, growth factor aptamers, anti‐growth factor receptor antibodies, growth factor tyrosine kinase inhibitors, antisense nucleotides, and small interference RNA (siRNA) (Fig. 2). Although each antagonist is designed to specifically block the angiogenic activity of a particular angiogenic factor, some inhibitors could exhibit a broad spectrum of inhibitory effects by acting on other signaling systems. For example, a tyrosine kinase inhibitor could affect several signaling pathways triggered by several angiogenic factors. In principle, these individual indirect angiogenesis inhibitors may have disadvantages for anticancer therapy because tumors may produce multiple angiogenic factors. Thus, it is possible that angiogenic factor antagonists might encounter drug resistant problems. Indeed, it has been reported that tumors become resistant toward the antiangiogenic monotherapy over long period of treatment (Casanovas et al., 2005).

B. VEGF‐A Antagonists Designing VEGF‐A antagonists has probably become the most attractive approach for development of antiangiogenic drugs (Ferrara and Kerbel, 2005). Indeed, the first antiangiogenic drug, bevacizumab (Avastin), a humanized neutralizing antibody against VEGF‐A, was approved by the US Food and Drug Administration (FDA) in February of 2004 for the treatment of metastatic colorectal cancer in combination with 5‐fluorouracil (5FU)‐based chemotherapy (Meyerhardt and Mayer, 2005; Sharieff, 2004; Sonpavde, 2004) (Fig. 2). In the same year, the FDA approved the second anti‐VEGF‐A drug pegaptinib (Macugen), anti‐VEGF‐A aptamer, for the treatment of the age‐related macular degeneration (AMD) (Gragoudas et al., 2004; Rakic et al., 2005). In addition to these initial successes, several other anti‐VEGF‐A molecules aimed to block different components of the VEGF‐A signaling

212

Yihai Cao and Qi Liu

pathways are under development or have been tested for their therapeutic efficacies against human cancers and nonmalignant disorders. These include soluble VEGFR‐1 and VEGFR‐2, the formation of biologically inactive heterodimers, gene therapy, neutralizing antibodies against VEGFR‐1 and VEGFR‐2, anti‐VEGFR‐1 and VEGFR‐2 tyrosine kinase chemical molecules, and siRNAs blocking VEGF‐A or anti‐VEGFRs. Similar to Avastin and Macugen, it is expected these anti‐VEGF‐A drugs would also produce some beneficial effects for the treatment of malignant and nonmalignant human diseases.

C. Antagonists for Non‐VEGF Factors In addition to VEGF‐A, other members in the VEGF family including VEGF‐C and VEGF‐D are also important targets. Similar to VEGF‐A, antagonists targeting different levels of the signaling pathways of these two factors are under development. These potential inhibitors might also inhibit lymphangiogenesis via blocking the VEGFR‐3‐activated signaling pathways. PDGF is another interesting angiogenic factor family with pluripotent effects on hemangiogenesis, lymphangiogenesis, and tumor cells. Members in the PDGF family primarily target two cell types of blood vessels, endothelial cells, and pericytes/smooth muscle cells (Cao, 2005a). Thus, inhibition of the PDGF signaling system is an important approach for antiangiogenic therapy. Other angiogenic factor antagonists include targets of FGF‐2, HGF, Ang‐1, Ang‐2, IGF, and TGF‐ . The therapeutic efficacy of these growth factor antagonists needs further evaluation.

D. Broad‐Spectrum Endogenous Inhibitors The vascular system is regulated by both positive and negative factors. Under physiological conditions, the quiescent vasculature may expose to imbalanced levels of angiogenic factors and inhibitors with predominant expression of the latter. Interestingly, several potent angiogenesis inhibitors have been identified in association of tumor growth. Although the issue why these negative factors are produced in association with tumor growth remains a paradoxical mystery, it appears that many inhibitors are generated from large circulating molecules by tumor‐derived proteases. Examples of these fragmental angiogenesis inhibitors include endostatin, angiostatin, tumstatin, and vasostatin (Cao, 2001; Folkman, 2004; Gragoudas et al., 2004; Nyberg et al., 2005; Rakic et al., 2005). Interestingly, the large parental molecules from which these angiostatic fragments are cleaved usually lack inhibitory activity. As expected, these angiogenesis inhibitors

Therapeutic Targets for the Treatment of Cancer and Metastasis

213

exhibit potent antitumor activity in experimental xenographic mouse tumor models (Cao et al., 1999; Maeshima et al., 2002; O’Reilly et al., 1994, 1997). However, the physiological functions of these endogenous angiogenesis inhibitors remain unknown. In addition to protein fragments, several intact protein molecules have also been found to inhibit angiogenesis. Thrombospondin‐1, a large multifunctional glycoprotein released by most epithelial cells in the extracellular matrix, is an example of endogenous inhibitor that might have an important function in maintaining the homeostatic nature of the vasculature in most adult tissues (Crawford et al., 1998; Rastinejad et al., 1989). Some inhibitors are produced at a specific site where the inhibitors may act locally. For example, pigment‐derived epithelial factor (PEDF) might prevent retinal vessel growth under physiological conditions (Doll et al., 2003). Unlike angiogenic factors, the underlying molecular mechanisms of endogenous angiogenesis inhibitors are poorly understood, although it has been suggested that they might interact with endothelial cell‐specific surface sensor molecules and integrins to induce apoptosis (Kumar, 2003; Moser et al., 1999; Veitonmaki et al., 2004). A modified version of endostatin, endostar, in combination with chemotherapy, has been approved in China for the treatment of certain types of human cancer (Sun et al., 2005). Endostatin is the first endogenous angiogenesis inhibitor that confers survival advantage on cancer patients. In principle, endogenous angiogenesis inhibitors have broader molecular targets in endothelial cells than indirect inhibitors (Fig. 1). Thus, they might have less drug resistant problems. It remains to be seen if endostatin have therapeutic advantages over Avastin.

E. Oral Angiogenesis Inhibitors Antiangiogenic small chemical molecules might have advantages over large proteins. Small compounds can be easily manufactured as synthetic pure chemicals at relatively low costs. Most chemical molecules might be delivered orally. More importantly, antiangiogenic chemical compounds in their smaller molecular weights might give rise to better penetration in the tumor tissue when delivered systemically. Indeed, large protein molecules such as antibodies have limited abilities for diffusion within the tumor tissue (Pegram and Reese, 2002). Polyphenols in green tea was one of the first oral angiogenesis inhibitors discovered to counteract VEGF‐A‐induced angiogenic activity (Cao and Cao, 1999). Epigallo‐catechin‐3‐gallate (EGCG) when administrated orally is able to inhibit mouse corneal angiogenesis (Fig. 3). Later studies show that EGCG inhibits VEGF production and phosphorylations of VEGF receptors in response to VEGF‐A (Basini et al., 2005;

214

Yihai Cao and Qi Liu

Green tea

OH OH O

HO

H OH

OH

H O

C

OH OH

O

EGCG

OH

VEGF-A Inhibition of angiogenesis

VEGF-A

Angiogenesis

Fig. 3 Epigallo‐catechin‐3‐gallate (EGCG) present in green tea acting as an angiogenesis inhibitor by blocking both VEGF production and VEGF‐A‐induced receptor phosphorylations.

Cao and Cao, 1999; Cao et al., 2002; Masuda et al., 2002; Rodriguez et al., 2005; Tang et al., 2003). These studies provide molecular mechanistic insights on the cancer preventive effects of EGCG. In addition to catechins, resveratrol rich in peanuts, red wine, and other daily consumable food plants

Therapeutic Targets for the Treatment of Cancer and Metastasis

215

has also been found to inhibit angiogenesis (Cao et al., 2002). To date, more than 20 different polyphenols present in various plants and functional foods have been reported to contain antiangiogenic activities (Cao et al., 2002). An important notion of the antitumor activity of these polyphenols is that they could produce synergistic effects with chemotherapeutic agents in suppression of tumor growth. For example, EGCG and chemotherapy synergistically inhibit the growth of a broad spectrum of tumors (Horie et al., 2005; Isogai et al., 2001; Ohishi et al., 2002; Sadzuka et al., 2000; Sugiyama and Sadzuka, 1998). These findings might imply cancer patients under chemotherapy should drink EGCG‐enriched green tea to enhance the antitumor effects of chemotherapeutic agents. Alternatively, intake of EGCG or green tea could reduce the dosages of chemotherapeutic agents and consequently decrease their toxicity‐related side effects. Thalidomide, a nonspecific and oral angiogenesis inhibitor, has been approved as a new drug for the treatment of human cancers (Caceres and Gonzalez, 2003).

IV. LYMPHANGIOGENESIS AND LYMPHATIC METASTASIS Similar to blood vessel growth, the growth of lymphatic vessels (lymphangiogenesis) is regulated by a range of growth factors, including members of VEGF, PDGF, Ang, IGF, FGF, and HGF families (Achen et al., 1998; Bjorndahl et al., 2005a,b; Cao, 2005b; Cao et al., 2004a,b; Cursiefen et al., 2004; Ferrara, 2005; Gale et al., 2002; Kajiya et al., 2005; Kubo et al., 2002; Morisada et al., 2005; Oh et al., 1997). These angiogenic factors seem to have redundant functions on both blood and lymphatic vessels. VEGF‐C and VEGF‐D are probably the most well‐studied and potent lymphangiogenic factors by interacting with VEGFR‐3 specifically expressed on lymphatic endothelial cells (Alitalo et al., 2005). Deletion of VEGF‐C or VEGFR‐3 in mouse embryos leads to completely absence of the lymphatic system, demonstrating the essential roles of this signaling pathway in lymphatic development (Alitalo et al., 2005). In xenographic mouse tumor models, overexpression of VEGF‐C or VEGF‐D in tumors promotes the growth of peri/intratumoral lymphatics and lymphatic metastasis (Skobe et al., 2001; Stacker et al., 2001). The role of VEGF‐A in lymphangiogenesis has become increasingly appreciated, although it might indirectly induce lymphangiogenesis via inflammatory cytokines and VEGF‐C/ VEGFR‐3 signaling system (Cursiefen et al., 2004). Expression of VEGF‐A at high levels also promotes lymphatic metastasis, probably via induction of peritumoral lymphatic vessel growth (Bjorndahl et al., 2005b; Hirakawa et al., 2005). Similar to VEGF‐A, FGF‐2 might stimulate lymphangiogenesis

216

Yihai Cao and Qi Liu

via an indirect pathway although a study demonstrates that FGF‐2 could also directly induce lymphangiogenesis (Chang et al., 2004; Kubo et al., 2002; Shin et al., 2006). PDGF‐BB seems to directly induce lymphangiogenesis and promotes lymphatic metastasis in mouse tumor models. Newly identified lymphangiogenic factors including HGF, IGF‐1, IGF‐2, Ang‐1, and Ang‐2 are all frequently expressed in various tumors and they are most likely involved in promoting lymphatic metastasis (Bjorndahl et al., 2005a; Gale et al., 2002; Kajiya et al., 2005; Morisada et al., 2005). The list of novel lymphangiogenic factors is still growing, suggesting that lymphangiogenesis is a complex process, which is tightly controlled by multiple positive and negative regulators. To date, no endogenous lymphangiogenesis inhibitors have been identified. Lymphangiogenesis research has stimulated considerable interests in therapeutic development of potential inhibitors for the treatment or prevention of lymphatic metastasis. However, further mechanistic studies at deep levels are pivotal for development of effective drugs. For example, all lymphangiogenic factors identified thus far have overlapping angiogenic activities on blood vessels and lymphatic vessels (Cao, 2005b). Does this mean that suppression of these angiogenic factor‐ induced hemangiogenesis is enough to inhibit lymphangiogenesis? At least two separate studies have provided answers against this hypothesis, indicating that under certain circumstances both FGF‐2 and VEGF‐A could specifically induce lymphangiogenesis but not hemangiogenesis. Lymphangiogenic inhibitors specifically targeting the lymphatic vasculature should be considered for drug development (Bjorndahl et al., 2005b; Chang et al., 2004).

V. CLINICAL DEVELOPMENT OF ANTIANGIOGENIC DRUGS Knowing the broad therapeutic implications of angiogenesis regulators, searching and designing for angiogenesis stimulators and inhibitors have become a competitive race among pharmaceutical companies for drug development. Therapeutic angiogenesis by proangiogenic factors in the treatment of ischemic disorders has been reviewed elsewhere (Simons, 2005). Antiangiogenesis will soon probably become a routine therapeutic method in the treatment of cancer and nonmalignant disorders. In 2004, two anti‐VEGF‐A drugs were proofed by the US FDA for the treatment of human cancers and eye diseases. The success of these novel drugs not only approves the concept of antiangiogenesis as a valid and effective therapeutic method but also inspires academic institutions and pharmaceutical

Therapeutic Targets for the Treatment of Cancer and Metastasis

217

industries to develop more effective antiangiogenic drugs. Indeed, several new antiangiogenic drugs are expected for approval within the next couple of years and hundreds of new chemical compounds, nucleotides, and protein molecules either as monotherapies or in combinations with conventional therapeutic methods have been tested in animal disease models or at early phases of clinical trials for the treatment of various angiogenesis‐dependent disorders. These include direct and indirect angiogenesis inhibitors that target different signaling pathways of angiogenesis. It is expected that more than 500,000,000 people globally will benefit from future proangiogenic and antiangiogenic therapies.

VI. CONCLUSIONS AND PERSPECTIVES Despite some initial setbacks and failures of clinical trials, angiogenesis research is now one of the most exciting areas of biomedical research. Owing that more than 70 human diseases are related to dysfunctions of blood or lymphatic vessels, targeting the vasculature in pathological tissues has become an extremely attractive new approach for drug development. The success of the first antiangiogenic agents for the treatment of human cancers and nonmalignant diseases has set up milestones in this field. However, the complexity of human diseases, in particular the malignancy, implies multiple targets should be considered for intervention. Today, we still know very little about the defined mechanisms of tumor blood or lymphatic vessel growth and the complex regulation of these sophisticated processes. At least we know that the genome of malignant cells contain multiple alternations of genetic information, which encodes production of different aberrant protein molecules. Many cell growth signaling molecules such as growth factors and receptors can be amplified at high levels in cancer cells. The complexity of regulation of tumor angiogenesis suggests that combinatorial therapy by targeting different angiogenic factors would be more effective. Indeed, clinical data obtained from anti‐VEGF‐A trials indicate that tyrosine kinase inhibitors blocking multiple kinases have a survival benefit, whereas monotherapy that only neutralizing VEGF‐A is ineffective (Carmeliet, 2005; Ferrara and Kerbel, 2005). The molecular mechanisms by which the anti‐VEGF‐A agents in combination with chemotherapy have survival advantages remain unclear although different hypotheses have been put forward (Ferrara and Kerbel, 2005). It is critically important to understand the molecular mechanisms of the actions of these inhibitors in order to improve therapeutic efficacy for future drug development.

218

Yihai Cao and Qi Liu

Although anti‐VEGF‐A agents confer a survival advantage, they also cause severe cardiovascular side effects, suggesting that these agents have broader targets in the body (Folkman, 2006). Today, we know nothing about the long‐term side effects of antiangiogenic drugs. To avoid side effects, it is important to develop tumor‐targeting strategies by limiting the action of angiogenesis inhibitors within the tumor vasculature. However, if antiangiogenic drugs act on nonmalignant tissues or organs to improve survivals of cancer patients, limiting the action of these drugs within tumor tissues could eliminate their therapeutic effects. As tumor vessels also consist of endothelial cells and mural cells, it is important to target nonendothelial cell types of tumor blood vessels. Recent studies demonstrate that antismooth muscle cell agents can synergistically suppress tumor growth with antiendothelial agents (Bergers et al., 2003). Another important issue is the selection of patient populations that are likely respond to antiangiogenic therapy. Unfortunately, such markers for patient selection are not available. Similarly, there also lack surrogate markers for us to monitor patient responses to antiangiogenic agents, although circulating endothelial cells derived from bone marrow have been suggested to serve as a surrogate marker for therapy (Park et al., 2004; Shaked et al., 2005). It is not expected that antiangiogenic therapy will result in shrinkage of tumors. Instead, this type of therapeutic agents might act as disease stabilizers by preventing further progression of malignancies. It is unclear for how long period the antiangiogenic drugs should be delivered to patients. Should antiangiogenic drugs be delivered to cancer patients for the rest of their lives? These issues should be warranted for further investigation. As for the treatment or prevention of lymphatic metastasis, it is important to understand the relation between blood vessel growth and lymphatic vessel growth. The fact that lymphatic vessels could grow independently from blood vessels indicates that specific inhibitors should be developed for targeting lymphatic vessels.

ACKNOWLEDGMENTS I thank Lasse Dahl Jensen for drawing the artwork. Y.C.’s laboratory is supported by the Swedish Research Council, the Swedish Heart and Lung Foundation, the Swedish Cancer Foundation, the Karolinska Institute Fund, the EU integrated projects of Angiotargeting (Contract No. 504743), VascuPlug (Contract No. STRP 013811), and the So¨derberg Foundation. Y.C. is supported by The Swedish Research Council. This work is dedicated to Drs. George Klein and Eva Klein’s 160th birthday.

Therapeutic Targets for the Treatment of Cancer and Metastasis

219

REFERENCES Aase, K., von Euler, G., Li, X., Ponten, A., Thoren, P., Cao, R., Cao, Y., Olofsson, B., Gebre‐ Medhin, S., Pekny, M., Alitalo, K., Betsholtz, C., et al. (2001). Vascular endothelial growth factor‐B‐deficient mice display an atrial conduction defect. Circulation 104, 358–364. Achen, M. G., Jeltsch, M., Kukk, E., Makinen, T., Vitali, A., Wilks, A. F., Alitalo, K., and Stacker, S. A. (1998). Vascular endothelial growth factor D (VEGF‐D) is a ligand for the tyrosine kinases VEGF receptor 2 (Flk1) and VEGF receptor 3 (Flt4). Proc. Natl. Acad. Sci. USA 95, 548–553. Achen, M. G., McColl, B. K., and Stacker, S. A. (2005). Focus on lymphangiogenesis in tumor metastasis. Cancer Cell 7, 121–127. Alitalo, K., Tammela, T., and Petrova, T. V. (2005). Lymphangiogenesis in development and human disease. Nature 438, 946–953. Basini, G., Bianco, F., and Grasselli, F. (2005). EGCG, a major component of green tea, inhibits VEGF production by swine granulosa cells. Biofactors 23, 25–33. Bergers, G., Song, S., Meyer‐Morse, N., Bergsland, E., and Hanahan, D. (2003). Benefits of targeting both pericytes and endothelial cells in the tumor vasculature with kinase inhibitors. J. Clin. Invest. 111, 1287–1295. Bissell, M. J. (1999). Tumor plasticity allows vasculogenic mimicry, a novel form of angiogenic switch. A rose by any other name? Am. J. Pathol. 155, 675–679. Bjorndahl, M., Cao, R., Nissen, L. J., Clasper, S., Johnson, L. A., Xue, Y., Zhou, Z., Jackson, D., Hansen, A. J., and Cao, Y. (2005a). Insulin‐like growth factors 1 and 2 induce lymphangiogenesis in vivo. Proc. Natl. Acad. Sci. USA 102, 15593–15598. Bjorndahl, M. A., Cao, R., Burton, J. B., Brakenhielm, E., Religa, P., Galter, D., Wu, L., and Cao, Y. (2005b). Vascular endothelial growth factor‐a promotes peritumoral lymphangiogenesis and lymphatic metastasis. Cancer Res. 65, 9261–9268. Caceres, W., and Gonzalez, S. (2003). Angiogenesis and cancer: Recent advances. P. R. Health Sci. J. 22, 149–151. Cao, R., Wu, H. L., Veitonmaki, N., Linden, P., Farnebo, J., Shi, G. Y., and Cao, Y. (1999). Suppression of angiogenesis and tumor growth by the inhibitor K1–5 generated by plasmin‐ mediated proteolysis. Proc. Natl. Acad. Sci. USA 96, 5728–5733. Cao, R., Brakenhielm, E., Pawliuk, R., Wariaro, D., Post, M. J., Wahlberg, E., Leboulch, P., and Cao, Y. (2003). Angiogenic synergism, vascular stability and improvement of hind‐limb ischemia by a combination of PDGF‐BB and FGF‐2. Nat. Med. 9, 604–613. Cao, R., Bjorndahl, M. A., Religa, P., Clasper, S., Garvin, S., Galter, D., Meister, B., Ikomi, F., Tritsaris, K., Dissing, S., Ohhashi, T., Jackson, D. G., et al. (2004a). PDGF‐BB induces intratumoral lymphangiogenesis and promotes lymphatic metastasis. Cancer Cell 6, 333–345. Cao, R., Eriksson, A., Kubo, H., Alitalo, K., Cao, Y., and Thyberg, J. (2004b). Comparative evaluation of FGF‐2‐, VEGF‐A‐, and VEGF‐C‐induced angiogenesis, lymphangiogenesis, vascular fenestrations, and permeability. Circ. Res. 94, 664–670. Cao, Y. (2001). Endogenous angiogenesis inhibitors and their therapeutic implications. Int. J. Biochem. Cell Biol. 33, 357–369. Cao, Y. (2005a). Direct role of PDGF‐BB in lymphangiogenesis and lymphatic metastasis. Cell Cycle 4, 228–230. Cao, Y. (2005b). Opinion: Emerging mechanisms of tumour lymphangiogenesis and lymphatic metastasis. Nat. Rev. Cancer 5, 735–743. Cao, Y., and Cao, R. (1999). Angiogenesis inhibited by drinking tea. Nature 398, 381. Cao, Y., Chen, H., Zhou, L., Chiang, M. K., Anand‐Apte, B., Weatherbee, J. A., Wang, Y., Fang, F., Flanagan, J. G., and Tsang, M. L. (1996). Heterodimers of placenta growth factor/

220

Yihai Cao and Qi Liu

vascular endothelial growth factor. Endothelial activity, tumor cell expression, and high affinity binding to Flk‐1/KDR. J. Biol. Chem. 271, 3154–3162. Cao, Y., Linden, P., Farnebo, J., Cao, R., Eriksson, A., Kumar, V., Qi, J. H., Claesson‐Welsh, L., and Alitalo, K. (1998). Vascular endothelial growth factor C induces angiogenesis in vivo. Proc. Natl. Acad. Sci. USA 95, 14389–14394. Cao, Y., Cao, R., and Brakenhielm, E. (2002). Antiangiogenic mechanisms of diet‐derived polyphenols. J. Nutr. Biochem. 13, 380–390. Carmeliet, P. (2000). VEGF gene therapy: Stimulating angiogenesis or angioma‐genesis? Nat. Med. 6, 1102–1103. Carmeliet, P. (2005). Angiogenesis in life, disease and medicine. Nature 438, 932–936. Carmeliet, P., Ferreira, V., Breier, G., Pollefeyt, S., Kieckens, L., Gertsenstein, M., Fahrig, M., Vandenhoeck, A., Harpal, K., Eberhardt, C., Declercq, C., Pawling, J., et al. (1996). Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 380, 435–439. Casanovas, O., Hicklin, D. J., Bergers, G., and Hanahan, D. (2005). Drug resistance by evasion of antiangiogenic targeting of VEGF signaling in late‐stage pancreatic islet tumors. Cancer Cell 8, 299–309. Cejudo‐Martin, P., and Johnson, R. S. (2005). A new notch in the HIF belt: How hypoxia impacts differentiation. Dev. Cell 9, 575–576. Chang, L. K., Garcia‐Cardena, G., Farnebo, F., Fannon, M., Chen, E. J., Butterfield, C., Moses, M. A., Mulligan, R. C., Folkman, J., and Kaipainen, A. (2004). Dose‐dependent response of FGF‐2 for lymphangiogenesis. Proc. Natl. Acad. Sci. USA 101, 11658–11663. Clauss, M. (1998). Functions of the VEGF receptor‐1 (FLT‐1) in the vasculature. Trends Cardiovasc. Med. 8, 241–245. Crawford, S. E., Stellmach, V., Murphy‐Ullrich, J. E., Ribeiro, S. M., Lawler, J., Hynes, R. O., Boivin, G. P., and Bouck, N. (1998). Thrombospondin‐1 is a major activator of TGF‐beta1 in vivo. Cell 93, 1159–1170. Cursiefen, C., Chen, L., Borges, L. P., Jackson, D., Cao, J., Radziejewski, C., D’Amore, P. A., Dana, M. R., Wiegand, S. J., and Streilein, J. W. (2004). VEGF‐A stimulates lymphangiogenesis and hemangiogenesis in inflammatory neovascularization via macrophage recruitment. J. Clin. Invest. 113, 1040–1050. Doll, J. A., Stellmach, V. M., Bouck, N. P., Bergh, A. R., Lee, C., Abramson, L. P., Cornwell, M. L., Pins, M. R., Borensztajn, J., and Crawford, S. E. (2003). Pigment epithelium‐derived factor regulates the vasculature and mass of the prostate and pancreas. Nat. Med. 9, 774–780. Ellis, L. M. (2005). Bevacizumab. Nat. Rev. Drug. Discov. Suppl., S8–S9. Eriksson, A., Cao, R., Pawliuk, R., Berg, S. M., Tsang, M., Zhou, D., Fleet, C., Tritsaris, K., Dissing, S., Leboulch, P., and Cao, Y. (2002). Placenta growth factor‐1 antagonizes VEGF‐ induced angiogenesis and tumor growth by the formation of functionally inactive PlGF‐1/ VEGF heterodimers. Cancer Cell 1, 99–108. Eriksson, A., Cao, R., Roy, J., Tritsaris, K., Wahlestedt, C., Dissing, S., Thyberg, J., and Cao, Y. (2003). Small GTP‐binding protein Rac is an essential mediator of vascular endothelial growth factor‐induced endothelial fenestrations and vascular permeability. Circulation 107, 1532–1538. Feng, D., Nagy, J. A., Dvorak, A. M., and Dvorak, H. F. (2000). Different pathways of macromolecule extravasation from hyperpermeable tumor vessels. Microvasc. Res. 59, 24–37. Ferrara, N. (1995). The role of vascular endothelial growth factor in pathological angiogenesis. Breast Cancer Res. Treat. 36, 127–137. Ferrara, N. (2005). The role of VEGF in the regulation of physiological and pathological angiogenesis. EXS. 209–231.

Therapeutic Targets for the Treatment of Cancer and Metastasis

221

Ferrara, N., and Kerbel, R. S. (2005). Angiogenesis as a therapeutic target. Nature 438, 967–974. Ferrara, N., Carver‐Moore, K., Chen, H., Dowd, M., Lu, L., O’Shea, K. S., Powell‐Braxton, L., Hillan, K. J., and Moore, M. W. (1996). Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 380, 439–442. Folberg, R., Hendrix, M. J., and Maniotis, A. J. (2000). Vasculogenic mimicry and tumor angiogenesis. Am. J. Pathol. 156, 361–381. Folkman, J. (1971). Tumor angiogenesis: Therapeutic implications. N. Engl. J. Med. 285, 1182–1186. Folkman, J. (2004). Endogenous angiogenesis inhibitors. APMIS 112, 496–507. Folkman, J. (2006). Antiangiogenesis in cancer therapy—endostatin and its mechanisms of action. Exp. Cell Res. 312, 594–607. Gale, N. W., Thurston, G., Hackett, S. F., Renard, R., Wang, Q., McClain, J., Martin, C., Witte, C., Witte, M. H., Jackson, D., Suri, C., Campochiaro, P. A., et al. (2002). Angiopoietin‐2 is required for postnatal angiogenesis and lymphatic patterning, and only the latter role is rescued by Angiopoietin‐1. Dev. Cell 3, 411–423. Gragoudas, E. S., Adamis, A. P., Cunningham, E. T., Jr., Feinsod, M., and Guyer, D. R. (2004). Pegaptanib for neovascular age‐related macular degeneration. N. Engl. J. Med. 351, 2805–2816. Gustafsson, M. V., Zheng, X., Pereira, T., Gradin, K., Jin, S., Lundkvist, J., Ruas, J. L., Poellinger, L., Lendahl, U., and Bondesson, M. (2005). Hypoxia requires notch signaling to maintain the undifferentiated cell state. Dev. Cell 9, 617–628. Hirakawa, S., Kodama, S., Kunstfeld, R., Kajiya, K., Brown, L. F., and Detmar, M. (2005). VEGF‐A induces tumor and sentinel lymph node lymphangiogenesis and promotes lymphatic metastasis. J. Exp. Med. 201, 1089–1099. Horie, N., Hirabayashi, N., Takahashi, Y., Miyauchi, Y., Taguchi, H., and Takeishi, K. (2005). Synergistic effect of green tea catechins on cell growth and apoptosis induction in gastric carcinoma cells. Biol. Pharm. Bull. 28, 574–579. Isogai, E., Isogai, H., Hirose, K., Hayashi, S., and Oguma, K. (2001). In vivo synergy between green tea extract and levofloxacin against enterohemorrhagic Escherichia coli O157 infection. Curr. Microbiol. 42, 248–251. Jain, R. K. (2005). Normalization of tumor vasculature: An emerging concept in antiangiogenic therapy. Science 307, 58–62. Joukov, V., Pajusola, K., Kaipainen, A., Chilov, D., Lahtinen, I., Kukk, E., Saksela, O., Kalkkinen, N., and Alitalo, K. (1996). A novel vascular endothelial growth factor, VEGF‐ C, is a ligand for the Flt4 (VEGFR‐3) and KDR (VEGFR‐2) receptor tyrosine kinases. EMBO J. 15, 1751. Joukov, V., Sorsa, T., Kumar, V., Jeltsch, M., Claesson‐Welsh, L., Cao, Y., Saksela, O., Kalkkinen, N., and Alitalo, K. (1997). Proteolytic processing regulates receptor specificity and activity of VEGF‐C. EMBO J. 16, 3898–3911. Kajiya, K., Hirakawa, S., Ma, B., Drinnenberg, I., and Detmar, M. (2005). Hepatocyte growth factor promotes lymphatic vessel formation and function. EMBO J. 24, 2885–2895. Kanno, S., Oda, N., Abe, M., Terai, Y., Ito, M., Shitara, K., Tabayashi, K., Shibuya, M., and Sato, Y. (2000). Roles of two VEGF receptors, Flt‐1 and KDR, in the signal transduction of VEGF effects in human vascular endothelial cells. Oncogene 19, 2138–2146. Kano, M. R., Morishita, Y., Iwata, C., Iwasaka, S., Watabe, T., Ouchi, Y., Miyazono, K., and Miyazawa, K. (2005). VEGF‐A and FGF‐2 synergistically promote neoangiogenesis through enhancement of endogenous PDGF‐B‐PDGFRbeta signaling. J. Cell Sci. 118, 3759–3768. Kubo, H., Cao, R., Brakenhielm, E., Makinen, T., Cao, Y., and Alitalo, K. (2002). Blockade of vascular endothelial growth factor receptor‐3 signaling inhibits fibroblast growth factor‐ 2‐induced lymphangiogenesis in mouse cornea. Proc. Natl. Acad. Sci. USA 99, 8868–8873.

222

Yihai Cao and Qi Liu

Kumar, C. C. (2003). Integrin alpha v beta 3 as a therapeutic target for blocking tumor‐induced angiogenesis. Curr. Drug Targets 4, 123–131. Luttun, A., Tjwa, M., Moons, L., Wu, Y., Angelillo‐Scherrer, A., Liao, F., Nagy, J. A., Hooper, A., Priller, J., De Klerck, B., Compernolle, V., Daci, E., et al. (2002). Revascularization of ischemic tissues by PlGF treatment, and inhibition of tumor angiogenesis, arthritis and atherosclerosis by anti‐Flt1. Nat. Med. 8, 831–840. Lyden, D., Hattori, K., Dias, S., Costa, C., Blaikie, P., Butros, L., Chadbum, A., Heissig, B., Marks, W., Witte, L., Wu, Y., Hicklin, D., et al. (2001). Impaired recruitment of bone‐ marrow‐derived endothelial and hematopoietic precursor cells blocks tumor angiogenesis and growth. Nat. Med. 7, 1194–1201. Maeshima, Y., Sudhakar, A., Lively, J. C., Ueki, K., Kharbanda, S., Kahn, C. R., Sonenberg, N., Hynes, R. O., and Kalluri, R. (2002). Tumstatin, an endothelial cell‐specific inhibitor of protein synthesis. Science 295, 140–143. Makino, Y., Cao, R., Svensson, K., Bertilsson, G., Asman, M., Tanaka, H., Cao, Y., Berkenstam, A., and Poellinger, L. (2001). Inhibitory PAS domain protein is a negative regulator of hypoxia‐ inducible gene expression. Nature 414, 550–554. Maniotis, A. J., Folberg, R., Hess, A., Seftor, E. A., Gardner, L. M., Pe’er, J., Trent, J. M., Meltzer, P. S., and Hendrix, M. J. (1999). Vascular channel formation by human melanoma cells in vivo and in vitro: Vasculogenic mimicry. Am. J. Pathol. 155, 739–752. Masuda, M., Suzui, M., Lim, J. T., Deguchi, A., Soh, J. W., and Weinstein, I. B. (2002). Epigallocatechin‐3‐gallate decreases VEGF production in head and neck and breast carcinoma cells by inhibiting EGFR‐related pathways of signal transduction. J. Exp. Ther. Oncol. 2, 350–359. Meyerhardt, J. A., and Mayer, R. J. (2005). Systemic therapy for colorectal cancer. N. Engl. J. Med. 352, 476–487. Morisada, T., Oike, Y., Yamada, Y., Urano, T., Akao, M., Kubota, Y., Maekawa, H., Kimura, Y., Ohmura, M., Miyamoto, T., Nozawa, S., Koh, G. Y., et al. (2005). Angiopoietin‐1 promotes LYVE‐1‐positive lymphatic vessel formation. Blood 105, 4649–4656. Moser, T. L., Stack, M. S., Asplin, I., Enghild, J. J., Hojrup, P., Everitt, L., Hubchak, S., Schnaper, H. W., and Pizzo, S. V. (1999). Angiostatin binds ATP synthase on the surface of human endothelial cells. Proc. Natl. Acad. Sci. USA 96, 2811–2816. Nico, B., de Falco, G., Vacca, A., Roncali, L., and Ribatti, D. (2001). In vivo absence of synergism between fibroblast growth factor‐2 and vascular endothelial growth factor. J. Hematother. Stem Cell Res. 10, 905–912. Nyberg, P., Xie, L., and Kalluri, R. (2005). Endogenous inhibitors of angiogenesis. Cancer Res. 65, 3967–3979. O’Reilly, M. S., Holmgren, L., Shing, Y., Chen, C., Rosenthal, R. A., Moses, M., Lane, W. S., Cao, Y., Sage, E. H., and Folkman, J. (1994). Angiostatin: A novel angiogenesis inhibitor that mediates the suppression of metastases by a Lewis lung carcinoma. Cell 79, 315–328. O’Reilly, M. S., Boehm, T., Shing, Y., Fukai, N., Vasios, G., Lane, W. S., Flynn, E., Birkhead, J. R., Olsen, B. R., and Folkman, J. (1997). Endostatin: An endogenous inhibitor of angiogenesis and tumor growth. Cell 88, 277–285. Oh, S. J., Jeltsch, M. M., Birkenhager, R., McCarthy, J. E., Weich, H. A., Christ, B., Alitalo, K., and Wilting, J. (1997). VEGF and VEGF‐C: Specific induction of angiogenesis and lymphangiogenesis in the differentiated avian chorioallantoic membrane. Dev. Biol. 188, 96–109. Ohishi, T., Kishimoto, Y., Miura, N., Shiota, G., Kohri, T., Hara, Y., Hasegawa, J., and Isemura, M. (2002). Synergistic effects of (–)‐epigallocatechin gallate with sulindac against colon carcinogenesis of rats treated with azoxymethane. Cancer Lett. 177, 49–56. Olofsson, B., Jeltsch, M., Eriksson, U., and Alitalo, K. (1999). Current biology of VEGF‐B and VEGF‐C. Curr. Opin. Biotechnol. 10, 528–535.

Therapeutic Targets for the Treatment of Cancer and Metastasis

223

Park, J. W., Kerbel, R. S., Kelloff, G. J., Barrett, J. C., Chabner, B. A., Parkinson, D. R., Peck, J., Ruddon, R. W., Sigman, C. C., and Slamon, D. J. (2004). Rationale for biomarkers and surrogate end points in mechanism‐driven oncology drug development. Clin. Cancer Res. 10, 3885–3896. Paul, R., Zhang, Z. G., Eliceiri, B. P., Jiang, Q., Boccia, A. D., Zhang, R. L., Chopp, M., and Cheresh, D. A. (2001). Src deficiency or blockade of Src activity in mice provides cerebral protection following stroke. Nat. Med. 7, 222–227. Pegram, M. D., and Reese, D. M. (2002). Combined biological therapy of breast cancer using monoclonal antibodies directed against HER2/neu protein and vascular endothelial growth factor. Semin. Oncol. 29, 29–37. Qu, H., Nagy, J. A., Senger, D. R., Dvorak, H. F., and Dvorak, A. M. (1995). Ultrastructural localization of vascular permeability factor/vascular endothelial growth factor (VPF/VEGF) to the abluminal plasma membrane and vesiculovacuolar organelles of tumor microvascular endothelium. J. Histochem. Cytochem. 43, 381–389. Rakic, J. M., Blaise, P., and Foidart, J. M. (2005). Pegaptanib and age‐related macular degeneration. N. Engl. J. Med. 352, 1720–1721; author reply 1720–1721. Rastinejad, F., Polverini, P. J., and Bouck, N. P. (1989). Regulation of the activity of a new inhibitor of angiogenesis by a cancer suppressor gene. Cell 56, 345–355. Richardson, T. P., Peters, M. C., Ennett, A. B., and Mooney, D. J. (2001). Polymeric system for dual growth factor delivery. Nat. Biotechnol. 19, 1029–1034. Roberts, W. G., and Palade, G. E. (1995). Increased microvascular permeability and endothelial fenestration induced by vascular endothelial growth factor. J. Cell Sci. 108(Pt. 6), 2369–2379. Rodriguez, S. K., Guo, W., Liu, L., Band, M. A., Paulson, E. K., and Meydani, M. (2005). Green tea catechin, epigallocatechin‐3‐gallate, inhibits vascular endothelial growth factor angiogenic signaling by disrupting the formation of a receptor complex. Int. J. Cancer 118, 1635–1644. Ryan, H. E., Poloni, M., McNulty, W., Elson, D., Gassmann, M., Arbeit, J. M., and Johnson, R. S. (2000). Hypoxia‐inducible factor‐1alpha is a positive factor in solid tumor growth. Cancer Res. 60, 4010–4015. Sadzuka, Y., Sugiyama, T., and Sonobe, T. (2000). Improvement of idarubicin induced antitumor activity and bone marrow suppression by theanine, a component of tea. Cancer Lett. 158, 119–124. Sato, Y., Kanno, S., Oda, N., Abe, M., Ito, M., Shitara, K., and Shibuya, M. (2000). Properties of two VEGF receptors, Flt‐1 and KDR, in signal transduction. Ann. NY Acad. Sci. 902, 201–205; discussion 205–207. Schlessinger, J. (2000). New roles for Src kinases in control of cell survival and angiogenesis. Cell 100, 293–296. Senger, D. R., Perruzzi, C. A., Feder, J., and Dvorak, H. F. (1986). A highly conserved vascular permeability factor secreted by a variety of human and rodent tumor cell lines. Cancer Res. 46, 5629–5632. Senger, D. R., Van de Water, L., Brown, L. F., Nagy, J. A., Yeo, K. T., Yeo, T. K., Berse, B., Jackman, R. W., Dvorak, A. M., and Dvorak, H. F. (1993). Vascular permeability factor (VPF, VEGF) in tumor biology. Cancer Metastasis Rev. 12, 303–324. Shaked, Y., Bocci, G., Munoz, R., Man, S., Ebos, J. M., Hicklin, D. J., Bertolini, F., D’Amato, R., and Kerbel, R. S. (2005). Cellular and molecular surrogate markers to monitor targeted and non‐targeted antiangiogenic drug activity and determine optimal biologic dose. Curr. Cancer Drug Targets 5, 551–559. Sharieff, W. (2004). Bevacizumab in colorectal cancer. N. Engl. J. Med. 351, 1690–1691; author reply 1690–1691.

224

Yihai Cao and Qi Liu

Shin, J. W., Min, M., Larrieu‐Lahargue, F., Canron, X., Kunstfeld, R., Nguyen, L., Henderson, J. E., Bikfalvi, A., Detmar, M., and Hong, Y. K. (2006). Prox1 promotes lineage‐specific expression of FGF receptor‐3 in lymphatic endothelium: A role for FGF signaling in lymphangiogenesis. Mol. Biol. Cell 17, 576–584. Shyu, K. G., Chang, H., and Isner, J. M. (2003). Synergistic effect of angiopoietin‐1 and vascular endothelial growth factor on neoangiogenesis in hypercholesterolemic rabbit model with acute hindlimb ischemia. Life Sci. 73, 563–579. Simons, M. (2005). Angiogenesis: Where do we stand now? Circulation 111, 1556–1566. Skobe, M., Hawighorst, T., Jackson, D. G., Prevo, R., Janes, L., Velasco, P., Riccardi, L., Alitalo, K., Claffey, K., and Detmar, M. (2001). Induction of tumor lymphangiogenesis by VEGF‐C promotes breast cancer metastasis. Nat. Med. 7, 192–198. Sonpavde, G. (2004). Bevacizumab in colorectal cancer. N. Engl. J. Med. 351, 1690–1691; author reply 1690–1691. Stacker, S. A., Caesar, C., Baldwin, M. E., Thornton, G. E., Williams, R. A., Prevo, R., Jackson, D. G., Nishikawa, S., Kubo, H., and Achen, M. G. (2001). VEGF‐D promotes the metastatic spread of tumor cells via the lymphatics. Nat. Med. 7, 186–191. Sugiyama, T., and Sadzuka, Y. (1998). Enhancing effects of green tea components on the antitumor activity of adriamycin against M5076 ovarian sarcoma. Cancer Lett. 133, 19–26. Sun, Y., Wang, J., Liu, Y., Song, X., Zhang, Y., Li, K., Zhu, Y., Zhou, Q., You, L., and Yao, C. (2005). Results of phase III trial of rh‐endostatin (YH‐16) in advanced non‐small cell lung cancer (NSCLC) patients. J. Clin. Oncol. Abstract‐GEOBASE. Tang, F. Y., Nguyen, N., and Meydani, M. (2003). Green tea catechins inhibit VEGF‐induced angiogenesis in vitro through suppression of VE‐cadherin phosphorylation and inactivation of Akt molecule. Int. J. Cancer 106, 871–878. Veitonmaki, N., Cao, R., Wu, L. H., Moser, T. L., Li, B., Pizzo, S. V., Zhivotovsky, B., and Cao, Y. (2004). Endothelial cell surface ATP synthase‐triggered caspase‐apoptotic pathway is essential for k1–5‐induced antiangiogenesis. Cancer Res. 64, 3679–3686. Waltenberger, J., Claesson‐Welsh, L., Siegbahn, A., Shibuya, M., and Heldin, C. H. (1994). Different signal transduction properties of KDR and Flt1, two receptors for vascular endothelial growth factor. J. Biol. Chem. 269, 26988–26995.

Novel Three‐Dimensional Organotypic Liver Bioreactor to Directly Visualize Early Events in Metastatic Progression Clayton Yates,* Chistopher R. Shepard,* Glenn Papworth,{ Ajit Dash,z Donna Beer Stolz,*,{ Steven Tannenbaum,z Linda Griffith,z and Alan Wells*,} *Department of Pathology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261; { Center for Biologic Imaging, Cell Biology, and Physiology, University of Pittsburgh, Pittsburgh, Pennsylvania 15261; ‡ Biologicl Egineering Division, Massachusetts Institute of Technology, Cambrige, Massachusetts 02139; } Department of Pathology, University of Pittsburgh and Pittsburgh VAMC, Pittsburgh, Pennsylvania 15261

I. Introduction A. Metastasis B. Models to Study Metastasis II. Bioreactors A. Liver Bioreactor III. Tumor Growth in the Bioreactor IV. Tumor–Hepatocyte Juxtapositioning V. Future Studies References

Metastatic seeding leads to most of the morbidity from carcinomas. However, little is known of this key event as current methods to study the cellular behaviors utilize nonrepresentative in vitro models or follow indirect subsequent developments in vivo. Therefore, we developed a system to visualize over a multiday to multiweek period the interactions between tumor cells and target organ parenchyma. We employ an ex vivo microscale perfusion culture system that provides a tissue‐relevant environment to assess metastatic seeding behavior. The bioreactor recreates many features of the fluid flow, scale, and biological functionality of a hepatic parenchyma, a common site of metastatic spread for a wide range of carcinomas. As a test of this model, prostate and breast carcinoma cells were introduced. Tumor cell invasion and expansion could be observed by two‐photon microscopy of red fluorescent protein (RFP)‐ and CellTracker‐labeled carcinoma cells against a green fluorescent protein (GFP)‐labeled hepatic tissue bed over a 14‐day period. Tumors visible to the naked eye could be formed by day 25, without evident necrosis in the >0.3‐mm tumor mass. These tumor cells failed to grow in the Advances in CANCER RESEARCH Copyright 2007, Elsevier Inc. All rights reserved.

225

0065-230X/07 $35.00 DOI: 10.1016/S0065-230X(06)97010-9

Clayton Yates et al.

226

absence of the supporting three‐dimensional (3D) hepatic microtissue, suggesting paracrine or stromal support function for the liver structure in tumor progression. Initial ultrastructural studies suggest that early during the tumor–parenchyma interactions, there are extensive interactions between and accommodations of the cancer and host cells, suggesting that the tumor‐related epithelial–mesenchymal transition (EMT) reverts, at least transiently, to promote metastatic seeding. In sum, our 3D ex vivo organotypic liver tissue system presents a critical vehicle to examine tumor–host interactions during cancer metastasis and/or invasion. It also circumvents current limitations in assays to assess early events in metastasis, and provides new approaches to study molecular events during tumor progression. # 2007 Elsevier Inc.

I. INTRODUCTION Metastases cause the major part of mortality and morbidity in cancer patients. Therefore, preventing the early steps in the metastatic cascade would yield outsized benefit for patients, a fact that is driving extensive investigative interest in the underlying mechanisms. This activity comes at the same time as quantal advances in the development of molecularly targeted therapies. Unfortunately, the translation of the current and future findings to patient care are hampered by two aspects. First, extant clinical testing paradigms are not informative as to efficacy in limiting metastatic spread due to the reliance on advanced patients or the need for long‐term follow‐up. Second, and more to the immediate point, current analytical systems are not optimal to evaluate this critical step in tumor progression.

A. Metastasis The biology of carcinoma metastasis is being deciphered slowly. Greater understanding has been derived for the initial stages of escape from the primary mass, increasingly made possible by new imaging advances of tumors in situ (Condeelis et al., 2005). However, the major rate‐limiting steps are at the site of metastatic seeding, which is underserved by current models. Still, the following sequence can be pieced together from existing data. Carcinomas develop from epithelial cells, or their precursors, which escape the growth and differentiation control of their orthotopic microenvironment. These cells acquire properties which enable them to separate from the primary tumor, penetrate surrounding basement membranes and invade local tissue, and gain access to conduits for dissemination. Other cellular changes allow the cells to grow independently of the orthotopic environment; metastases are noted primarily in lung, liver, bone marrow, and local lymph nodes. The altered cell phenotype at the metastatic site provides for the disseminated tumor cells to recognize the target organ as “receptive” and to proceed to form a cohesive tumor mass (Fidler, 2003).

Liver Microtissue Bioreactor to Study Tumor Metastasis

227

Thus, the properties that allow escape from the primary tumor may be opposite those that enable metastatic competency. Separation from the primary carcinoma mass requires loosening of the cell–cell bonds. This is characteristic of the dedifferentiation or epithelial– mesenchymal transition (EMT) as loosely defined (Tarin, 2005; Thompson and Newgreen, 2005), noted in most all metastatic carcinomas. Down‐ regulation of E‐cadherin, which is indicative of EMT (Grunert et al., 2003), usually occurs via epigenetic processes and not the irreversible and stochastic mechanisms of gene mutation or deletion (Jones and Baylin, 2002). While the mechanisms underlying this loss of E‐cadherin function are poorly understood, the result of such a loss of the cell–cell adhesions is that it now allows for autocrine signaling of physiologically separated receptor and ligand pairs (Kim et al., 1999). The cells must then migrate to, transit through, and move from a vascular conduit. Our previous work in prostate (Mamoune et al., 2003, 2004; Turner et al., 1996, 1997; Xie et al., 1995), bladder (Kassis et al., 2002), and breast cancer (Kassis et al., 1999, 2001), and studies from others studying breast carcinomas (Wang et al., 2005) have highlighted the central role of epidermal growth factor receptor (EGFR)‐mediated motility in this process (Wells, 2000). Transit through the vasculatures (both hematogenous and lymphatic) is poorly understood, as to the track, length of time, and survival of the tumor cells. The ready finding of “circulating” tumor cells in peripheral venous blood (Cristofanilli et al., 2004) raised the possibility of tumor cells being able to pass at least two capillary beds during their time in circulation. The signals that allow these large cells to deform and survive this difficult passage are still unknown, and as such do not at present offer opportunities for targeted interventions. Extravasation involves both tumor cell and endothelial adaptations. The initial step of arrest was captured as a physical mismatch between vessel caliber and tumor cell size. The molecular basis of tumor cell–endothelial cell interaction is via cell adhesion molecules (CAM) which lead to intracellular signaling. While this vascular access and transit is critical for dissemination, it appears not to be the major rate‐limiting step under experimental challenges situations (Luzzi et al., 1998). Once the carcinoma cell attains the organ parenchyma, it moves beyond the immediate site of extravasation. Findings have suggested that sites of metastases might be preconditioned by nontumor/nontarget organ cells (Kaplan et al., 2005). The motility processes required to invade into the metastasis organ should be similar mechanistically to those during the initial escape from the primary tumor site. However, the stimulating factors and extracellular matrix components involved in these migrations have yet to be deciphered. During the initial tumor invasion into the vasculature, the cells move through the barrier matrix/basement membrane utilizing integrin‐mediated adhesions. It is likely that integrins are also utilized during

228

Clayton Yates et al.

migration into the parenchyma, but this is yet to be shown. Further, studies must also determine whether this invasion into the parenchyma is required for subsequent metastatic growth. However, regardless of specific migration processes, the carcinoma cells need to establish adhesions, either to the matrix (likely via integrins) or to other cells (likely involving the cadherin systems) to avoid anoikis‐initiated death (Assoian, 1997). The definition of a permissive organ, or “soil,” for metastatic spread likely arises from paracrine factors released from organ stromal and parenchymal cells; though it may also involve some preconditioning from bone marrow‐ derived cells (Kaplan et al., 2005). That only some organs are “permissive” is not only noted by the limited and predictable distribution of carcinoma metastases but also has been shown experimentally by the ability of bone marrow (and to some extent lymph node) fibroblasts but not dermal or lung fibroblasts to support the growth of LNCaP prostate cells (Gleave et al., 1992). Carcinoma cell autocrine EGFR signaling, as well as signals from other tyrosine kinase receptors such as IGF‐1R, may play into this survival advantage. In addition, adhesion to the organ body and parenchyma likely contribute. Long‐term dormancy of tumor cells following the initial seeding is still undeciphered, even as to whether it involves balanced proliferation/ apoptosis or G0‐like arrest (Ghiso et al., 1999), the latter would likely involve some cellular interactions between the carcinoma cells and the parenchymal cells. The presence of such reciprocal cell signaling would imply adaptation to the host environment and the ability to avoid/escape proliferation‐targeted therapies (most current cancer chemotherapy). Thus, the nature of carcinoma cell dormancy following seeding in target metastatic organs must be addressed experimentally.

B. Models to Study Metastasis As novel therapies are being developed that might target molecular steps which contribute to metastasis (Minn et al., 2005), the development of new analytical methods to study development of metastatic lesions at the cellular level are needed. Currently, few modes are available to evaluate this critical progression in the target organ (Table I). Attempts have been made to dissect the individual aspects of the metastatic process. There are established assays for cell proliferation, migration, adhesion, and survival, as well as specific assays for key regulatory molecules. However, these are limited in that metastasis requires a constellation of individual cell properties and molecular activations; no one assay is full predictive either positive or negative. Furthermore, as these cellular and molecular events are interdependent, assays designed to isolate each may provide a panel of responses not indicative of the integrated situation.

229

Liver Microtissue Bioreactor to Study Tumor Metastasis

Table I

Model Properties Current models

1. In vitro models lack tissue complexity 2. In vivo models are short‐term or single snapshots of a process 3. In vivo models present challenges to targeted molecular and pharmacological interventions 4. Relevance to human tissues is uncertain

Proposed models 1. An integrated epithelial/ stromal/endothelial cell architecture representing key target organs 2. Long‐term (weeks) visualization during tumor evolution 3. Reproducible and direct and temporally constrained manipulation of distinct cell types 4. Humanizable for linked metabolism and action

For this reason, investigators seek experimental systems that recapitulate this integrated process (Fig. 1). Endpoint animal models of metastasis, the most common assays, have yielded important information for the size of metastases and the number of cells at the target organ, but have failed to provide information about the cellular processes that occur during the development of metastasis. To overcome this limitation, some investigators have developed real‐time (intravital video microscopy) in vivo systems that allow for short‐term imaging and evaluation of the tumor cell behavior (Chambers et al., 1995; Condeelis et al., 2001; Timmers et al., 2004). Intravital imaging relies on confocal or multiphoton imaging to follow the behavior of individual fluorescently labeled cells within a particular target organ. The depth of focus is less than 0.5 mm and thus to gain images a portion of a target organ is exposed and placed on the microscopy viewing platform. The advent of these new imaging methods offer greater possibilities to follow the behavior of individual cells and have led to new appreciation for the roles played by nontumor cells and matrix (Wyckoff et al., 2004). Still, the window of examination is in hours at most and the systems are not easily manipulated preventing higher throughput investigations. Herein, we report using a microscale bioreactor that fosters three‐ dimensional (3D) liver tissue formation and maintenance in culture. Each of the individual capillary bed‐sized tissue units in the reactor is locally perfused with culture medium in a manner that mimics flow through the liver capillary bed and can be observed by light or two‐photon microscopy (Fig. 1). This system appears to allow maintenance of a high level of liver‐ specific gene expression (Sivaraman et al 2005) and thus affords for the recreation of many key features of the complex in vivo physiological

Clayton Yates et al.

230

Cell injection

Imaging & spectroscopy Reactor

Pump 1

Pump 2

Reservoir

Fig. 1 Schematic of the liver microtissue bioreactor. The schematic of the microtissue assembly demonstrates both forward flow and cross‐flow as well as injection port. An optional sampling port downstream of the microtissue is not shown. The picture to the left shows one version of the machined product shown the scale of a US penny.

environment for in vitro observation (Powers et al., 2002a,b). With the liver being a preferred ectopic site for metastasis of many cancers including prostate cancer (Ewing, 1922; Fidler, 2003; Paget, 1989; Shah et al., 2004), we propose that tumor growth and invasion at the metastatic target organ can be observed at the cellular level in real‐time with the use of fluorescent markers for visualization. The ultimate advantage of this system is the potential to create human‐perfused tissue structures for supporting human tumor growth, allowing for an easily manipulatable procedure for visualizing in real‐time invasion and growth of a target organ capillary bed during metastasis.

II. BIOREACTORS The tumor microenvironment is emerging as a critical factor in development and progression of primary tumors (Proia and Kuperwasser, 2005; Tlsty, 2001) and progression of secondary tumors that have metastasized from a primary site (Bhowmick and Moses, 2005; Bhowmick et al., 2004). The interplay between tumor cells and matrix is also a strong determinant of tumor phenotype. Thus, culture systems that provide an in vivo‐like environment offer increased advantage to observe relevant responses. Most approaches to creating 3D in vitro systems for analysis of heterotypic interactions in cancer employ culturing cells in 3D extracellular matrix gels (such as type I collagen or matrigel) where no external flows are imposed (Goswami et al., 2005; Griffith and Swartz, 2006). To model tumors, stromal

Liver Microtissue Bioreactor to Study Tumor Metastasis

231

cells are typically mixed in a gel or cultured in a 3D scaffold, and the tumor cell might be added directly on top of the gel‐containing fibroblasts or endothelial cells, mixed in another gel layer on top, or various other configurations (Donovan et al., 2001; Nelson and Bissell, 2005; Tsai et al., 2005). The hanging drop method has also been used to foster formation of cell aggregates from defined numbers of tumor cells with endothelial cells, with an advantage that the transport of oxygen to the tissue is facilitated by the culture geometry (Landman and Please, 2001; Timmins et al., 2004). Bioreactor approaches to studies of cancer have primarily been focused on creating relatively large tumor cell aggregates that mimic the early stages of avascular tumors. In this regard, fluid‐filled spinner flasks have been used for decades to create 3D spheroids of tumor cells under controlled environmental conditions (oxygen and pH) in the bulk cell culture medium (Franko and Koch, 1983; Margolis et al., 1999; Santini and Rainaldi, 1999; Schmeichel and Bissell, 2003), and this approach has been extended to formation of 3D differentiated structures from embryonic stem cells (Bauwens et al., 2005). Limitations of this system include lack of mass transfer through large (>0.3‐mm diameter) spheroids, resulting in necrosis of the center due to nutrient and oxygen deprivation (Kunz‐Schughart, 1999; Santini and Rainaldi, 1999). The rotating wall vessel, an analogue of the spinner flask that provides lower fluid shear stresses at the exterior surfaces of spheroids and can foster cell aggregation (Brown et al., 2003), is gaining favor as a method for culture of tumor spheroids and for coculture of heterotypic cells types with tumors (Rhiel et al., 2004). Still, spheroidal cultures lack an important component of tissue physiology—localized, microscale flow through the tissue (Griffith and Swartz, 2006). A variety of bioreactor configurations have been developed to provide flow past the surface of a 3D tissue, an arrangement that enhances mass transfer at the tissue–fluid interface (Fiegel et al., 2004; Jasmund and Bader, 2002; Zhao et al., 2005) and that can provide stimulatory shear stress to cells (Barron et al., 2003; Martin et al., 2004). Fluid flow through large 3D tissue structures in culture has also been described as a means to enhance mass transfer in structures with dimensions in the 1‐ to 10‐mm (or more) size range (Braccini et al., 2005; Cartmell et al., 2003; Navarro et al., 2001), but relatively little attention has been given to microscale control of flow. Elegant studies using in vitro reactors to illustrate the role of interstitial flows on molecular processes governing angiogenesis will likely stimulate additional efforts in this important area (Helm et al., 2005; Semino et al., 2006). Bioreactors have the potential to fulfill an important gap between the well‐defined cultures of single cell types and the complexity of the whole animal. They also provide a greater appreciation of tissue‐specific

232

Clayton Yates et al.

microenvironmental influences contributing to tissue‐specific malignant behavior of epithelial tumors (Atula et al., 1997; Bhowmick and Moses, 2005; Kunz‐Schughart, 1999).

A. Liver Bioreactor As a step toward creating true physiological mimics of human and animal tissues that recapitulate the features of a capillary bed, we have developed a microfabricated bioreactor system that facilitates perfusion of 3D heterotypic cocultures at the length scale of the capillary bed in an arrangement that also allows in situ analysis of cell behavior via microscopy (Powers et al., 2002a,b; Sivaraman et al., 2005; Torisawa et al., 2005). This system circumvents rapid loss of liver‐specific functions that normally occurs when hepatocytes are maintained under standard culture conditions, thus, providing a reasonable model system for the testing of tumor–host interactions in ex vivo environment. That the liver is a major site of metastasis for many carcinomas enhances the value of this model. The model also has the potential for recreating the hepatic parenchymal architecture, with the possibility of physiological distribution of endothelial and other nonhepatocytes cellular elements (Powers and Griffith, 1998; Powers et al., 2002a,b). The self‐assembly from individual cellular components enables one to provide cell types that are engineered to investigate specific aspects. As an example, mixing liver endothelial cells isolated from a rat that expresses enhanced green fluorescent protein (GFP) in all its cells with hepatocytes from a wild‐type animal allows the endothelial network to be visualized in situ. A second example could encompass VE‐CAM‐devoid endothelial cells could be mixed with normal hepatocytes and stromal elements by isolating components from different transgenic rodents. Our cross‐flow perfusion reactor is designed to address several needs for 3D liver tissue culture (Powers et al., 2002a,b). The classical challenges in reactor design for 3D perfusion culture—ensuring a relatively homogeneous distribution of flow and mass transfer throughout the system to meet the metabolic demands of the cells—are augmented in the case of 3D cultures of primary cells by the need to provide a scaffold appropriate for tissue morphogenesis. Varying degrees of histotypic reorganization have been observed in several types of 3D liver cell cultures, particularly those incorporating perfusion through the tissue mass (Gerlach et al., 1995; Griffith et al., 1997; Kaihara et al., 2000; Michalopoulos et al., 1999). Distinguishing features of our design include: an appropriate scaffold for tissue morphogenesis, uniform distribution of fluid flow and nutrients throughout the 3D culture, and an optical window to allow repeated in situ observation of cells via light or two‐photon microscopy during perfusion culture.

Liver Microtissue Bioreactor to Study Tumor Metastasis

233

Herein, we will describe its first use as a host for tumor invasion and growth.

III. TUMOR GROWTH IN THE BIOREACTOR To determine whether human prostate cancer cells could grow in serum‐ free medium in the context of the 3D liver bioreactor, we introduced a single cell suspension of red fluorescent protein (RFP)‐expressing DU‐145 carcinoma cells into a bioreactor containing stable liver microtissue units. These bioreactor cocultures were then imaged at the same sites with two‐photon microscopy until day 14 (Fig. 2). Initial attachment of RFP‐expressing cells is apparent by day 2. 3D reconstruction of day 2 and day 4 images from the same channels demonstrated that expanded tumor cell proliferation derives from single cell origins (Fig. 3). By day 6, there was apparent growth and invasion into the parenchyma of the hepatic tissue, which resulted in overgrowth of individual channels in the bioreactor by cancer tissue by day 14. As a control, bioreactors that had not been seeded with DU‐145 cells were imaged through the 14‐day period (Fig. 2) and beyond to 30 days (data not shown); these bioreactors showed stable persistence of the liver microtissue units. The liver microtissues were assessed by microscopy on day 14. We detected prostate cancer cells invading the hepatic parenchyma (Fig. 4). Viable prostate cancer cells were observed by histology and transmission electron microscopy (TEM) after 14 days, and these were closely juxtaposed to the remaining hepatocytes. The ability of tumor cells to persist in the liver bioreactor and invade the established parenchyma was not limited to prostate tumors or cell lines. Initial passage explants from mammary carcinomas were stained with CellTracker dye, introduced into the liver bioreactor, and followed by two‐ photon microscopy (Fig. 5). These cells survived and entered between hepatocyes. In addition to cancer cell invasion, overt tumor formation was observed in microtissue bioreactors by 25 days (Fig. 6). On closer examination by toluidine blue and TEM, we found that the tumor cell mass was not necrotic from the center to the perimeter, despite exceeding 300 mm in diameter (Fig. 6). The tumor masses consisted of viable prostate cells closely juxtaposed to nonparenchymal cells. This constitutes adequate supply of oxygen and nutrients to avoid necrotic centers in tumor of this mass; this is likely due to the active perfusion through the chambers. An open question remained whether this tumor growth was intrinsic to structural environment provided by our 3D system or simply due to the

Clayton Yates et al.

234 A

Day 2

Day 4

Day 6

Day 9

Day 11

Day 14

B

Fig. 2 DU‐145 cells proliferate within the liver microtissue. RFP‐expressing DU‐145 human prostate cancer cells were introduced to primary hepatocytes obtained from GFP transgenic rats. (A) Two‐photon images were taken of DU‐145 prostate cancer (red) in the presence of hepatic tissue (green). DU‐145 growth was assessed over a 14‐day period. (B) Bioreactors with only liver cells demonstrate hepatocyte structure and function stability over this time period. Shown is one representative channel from experiments repeated at least five times. DU‐145 cells (150,000/bioreactor were entered through the port) stably expressing RFP were introduced on day 5 into a preformed liver microtissue bioreactor. DU‐145 cells were introduced in the mode of forward flow and cross‐flow stopped for 24 h before reinstating crossflow. At the first imaging after attachment, channels that contained only a few (300 mm in diameter. (B) Tumor masses were removed at day 25 from reactor and stained with toluidine blue. (C) Removed tumor masses were also imaged by electron microscopy. Tumor‐parenchymal heterogeneity remained intact throughout tumor. Shown are images from one of three similar experiments.

possibly tumor dormancy are all unknown. However, they do suggest a model of tumor cell plasticity in which carcinomas are not viewed as being increasingly dedifferentiated as they progress toward invasion and

Clayton Yates et al.

238

Fluorescence intensity

A

1200

DMEW 10% FBS DMEW serum free HGM serum free

1000 800 600 400 200 0

TO

Day 1 Day 2 Day 3 Day 4 Day 5 Day 6

4 hr

B

Day 2

Day 4

Day 6

Polystyrene

C

500

Fluorescence intensity

Collagen coated

400

Polystyrene Collagen coated 4 µM transwell

300 200 100 0 Day 1

Day 2

Day 3

Day 4

Day 5

Day 6

Fig. 7

DU‐145 cells fail to expand on 2D surfaces in the presence of hepatocytes and hepatocyte growth medium (HGM). (A) Growth of DU‐145 human prostate cancer cells was assessed by fluorescent intensity in the presence of HGM, 10% FBS‐supplemented DMEM, or serum‐free DMEM over a 6‐day period. (B) Fluorescence images of hepatocytes (green) and DU‐145 cells (red) on polystyrene or collagen‐coated surfaces in the presence of HGM. (C) Growth of DU‐145 prostate cancer cells in HGM on polystyrene, collagen‐coated, or 4‐mM pore transwell plates in coculture systems. All experiments were performed in triplicate and repeated three times; in the graphs the data were normalized to day 1 (set at 100) and are presented as percentage growth SEM.

metastasis, but one in which the EMT is a reversible state (Fig. 9). EMT promotes tumor dissemination from the primary mass by weakening cell– cell bonds. However, to gain critical survival signals in the new ectopic location, the metastasis, this dedifferentiation at least partially reverts, a

Liver Microtissue Bioreactor to Study Tumor Metastasis

239

DU-145

Co culture

Green = hepatocytes Red = DU-145 prostate cells Blue = huE-cadherin staining

Fig. 8 Carcinoma and hepatocyte coculture results in E‐cadherin expression by DU‐145 cells. DU‐145 cells (2000 cells/cm2) were cocultured with primary rat hepatocytes (50,000 cells/cm2) in HGM media. Cells were then stained for E‐cadherin using a human‐specific antibody. In the two left‐hand panels, the cells were cultured for 14 days and imaged by standard immunohistochemical techniques, with the top panel being DU‐145 in the absence of hepatocytes and the bottom in the presence. In the right‐hand picture, immunofluorescence captures E‐cadherins (blue) clustering along the interface between DU‐145 cells (RFP, mainly nuclear) and hepatocytes after 2 day of coculture (eGFP).

“mesenchymal–epithelial reversion transition (MErT),” to link to epithelial cells in the target organ. This model may explain the few reports of E‐cadherin expression in carcinoma metastases (Graff et al., 2000; Kowalski et al., 2003), though these also may be indicative of dissemination of E‐cadherin‐expressing primary tumors. Additionally, one might expect the MErT state itself to be transient as the metastasis grows to the size at which connections to host parenchyma is no longer necessary. Such fundamental questions as to the initial stages of metastases must be approached experimentally through real‐time visualization of a continuous process. Our carcinoma‐bearing microtissue bioreactor provides such an ex vivo model system to tackle this question at the cellular and molecular levels.

V. FUTURE STUDIES We developed a model for cancer cell establishment and growth in the liver, second only to bone for prostate cancer metastases (Ewing et al., 1995; Shah et al., 2004), using a bioreactor that provides an appropriate

Clayton Yates et al.

240 EGF

EGF

EGF Primary carcinogenesis

H

X GF X

X

EMT for dissemination H

H MErT for establishment

Fig. 9 Proposed model of phenotypic plasticity during tumor progression. Shown are the postulated stages of tumor cell differentiation with a subset of cells in the primary tumor losing cell–cell adhesions and polarity, reinforced by various ubiquitous autocrine signaling loops (Kim et al., 1999). This enables the tumor cells to disseminate from the primary mass. On reaching a metastatic target organ, the carcinoma cells reexpress the cell–cell adhesion molecules to interact directly with the parenchymal cells. Heavy bars represent E‐cadherins, Ys are growth factor receptors, lines are matrix, and Hs are parenchymal cells (hepatocytes).

environment for the morphogenesis of primary liver cell isolates into functional 3D microtissue (Powers et al., 2002a,b). Our bioreactor system, on establishment of a functional liver parchencyma including nonhepatocyte support cells, addresses the concerns of an appropriate environment to study molecular events of metastasis. Human prostate cancer cells, stably expressing RFP, were introduced into the established liver bioreactor and attachment was seen by day 2 with subsequent growth noticeable by day 4. This was not observed on 2D culture plates. Obviously, the media and conditions lacked signals that promoted tumor cell growth that the liver microtissue provided. Interestingly, a primary human breast cancer explant demonstrated invasion and survival in the liver bioreactor but no obvious growth. Whether this is related to the vexing issue of tumor dormancy remains to be seen. Still, the bioreactor not only supported cell proliferation but another integrated cell response, the relocation/migration of tumor cells across the tissue mass, was seen in the early days after inoculation. As currently established, the liver bioreactor contains the parenchymal cells of the liver, including hepatocytes, stromal cells, and large and small vessel endothelial cells. Resident immune response‐derived cells (Kupffer cells) are also present (data not shown). Data suggest cells of the innate and acquired

Liver Microtissue Bioreactor to Study Tumor Metastasis

241

immune response (Kaplan et al., 2005) play a role in metastatic seed and tumor progression. Also, bone marrow‐derived endothelial precursors might contribute to tumor angiogenesis (Salven et al., 2003). A future version of the bioreactor could incorporate these cells by infusing leukocyte or bone marrow fractions into the tumor‐bearing bioreactor using the available ports; this is the same mode in which the tumor cells are introduced and can be used for repeated additions. This will also enable the study of immunomodulation/ immunotherapy of metastases, an area of intense investigation. The liver bioreactor also provides an ideal environment with which to study metastasis‐directed therapies. One central issue in chemotherapy is the metabolism of the agents; another is the seeming resistance of metastatic foci. A tumor‐bearing liver bioreactor addresses both points. First, the tumor cells are presented within a metastasis‐related environment. The signals from the ectopic tissue that may contribute to chemoresistance are likely extant and operative. Second, liver is the primary metabolizer of chemotherapeutics; thus, the tumor and the metabolism of the agent are linked in the bioreactor. A major obstacle in translating preclinical data to patients is the differences in metabolism between even the most closely related species. We have demonstrated that the maintenance of liver‐specific metabolic functions is significantly enhanced in 3D bioreactor culture compared to polarized (collagen gel sandwich) 2D culture (Sivaraman et al., 2005). Our bioreactor potentially can be fully humanized, with liver cells obtained from limited resections obtained during colorectal carcinoma procedures or from livers unfit for transplantation. Such a bioreactor has been established in an initial feasibility study (Griffith Stolz et al., unpublished data). In short, successful establishment of this organotypic liver system that supports tumor cell growth opens many avenues for future investigation. Current approved therapies aim at cell proliferation and do not expressly target the stages of metastasis establishment and progression. Thus, an integrative model of tumor progression including the target environment would be a significant advancement to highlight total systemic responses. Initial studies have been able to humanize this microtissue by using cells from liver resections. Lastly, by utilizing a fully functional liver bioreactor, the ability to intimately link drug metabolism in real time to target actions opens up new possibilities for the development and testing of agents.

ACKNOWLEDGMENTS These studies were supported by grants from the US Army and the NCI/NIH. We thank members of the Wells Laboratory and the Griffith Laboratory for technical assistance and helpful discussions.

242

Clayton Yates et al.

REFERENCES Assoian, R. (1997). Anchorage‐dependent cell cycle progression. J. Cell Biol. 136, 1–4. Atula, S., Grenman, R., and Syrjanen, S. (1997). Fibroblasts can modulate the phenotype of malignant epithelial cells in vitro. Exp. Cell Res. 235, 180–187. Barron, V., Lyons, E., Stenson‐Cox, C., McHugh, P. E., and Pandit, A. (2003). Bioreactors for cardiovascular cell and tissue growth: A review. Ann. Biomed. Eng. 31, 1017–1030. Bauwens, C., Yin, T., Dang, S., Peerani, R., and Zandstra, P. W. (2005). Development of a perfusion fed bioreactor for embryonic stem cell‐derived cardiomyocyte generation: Oxygen‐ mediated enhancement of cardiomyocyte output. Biotechnol. Bioeng. 90, 452–461. Bhowmick, N. A., and Moses, H. L. (2005). Tumor‐stroma interactions. Curr. Opin. Genet. Dev. 15, 97–101. Bhowmick, N. A., Chytil, A., Plieth, D., Gorska, A. E., Dumont, N., Shappell, S., Washington, M. K., Neilson, E. G., and Moses, H. L. (2004). TGF‐ signaling in fibroblasts modulate the oncogenic potential of adjacent epithelia. Science 303, 848–851. Braccini, A., Wendt, D., Jaquiery, C., Jakob, M., Heberer, M., Kenins, L., Wodnar‐Filipowicz, A., Quarto, R., and Martin, I. (2005). Three‐dimensional perfusion culture of human bone marrow cells and generation of osteoinductive grafts. Stem Cells 23, 1066–1072. Brown, L. A., Arterburn, L. M., Miller, A. P., Cowger, N. L., Hartley, S. M., Andrews, A., Silber, P. M., and Li, A. P. (2003). Maintenance of liver functions in rat hepatocytes cultured as spheroids in a rotating wall vessel. In vitro Cell. Dev. Biol. Anim. 39, 13–20. Cartmell, S. H., Porter, B. D., Garcia, A. J., and Guldberg, R. E. (2003). Effects of medium perfusion rate on cell‐seeded three‐dimensional bone constructs in vitro. Tissue Eng. 9, 1197–1203. Chambers, A. F., MacDonald, I. C., Schmidt, E. E., Koop, S., Morris, V. L., Khokha, R., and Groom, A. C. (1995). Steps in tumor metastasis: New concepts from intravital videomicroscopy. Cancer Metastasis Rev. 14, 279–301. Condeelis, J., Singer, R., and Segall, J. E. (2005). The great escape: When cancer cells hijack the genes for chemotaxis and motility. Ann Rev. Cell Dev. Biol. 21, 697–718. Condeelis, J. S., Wyckoff, J. B., Bailly, M., Pestell, R., Lawrence, D., Backer, J., and Segall, J. E. (2001). Lamellipodia in invasion. Semin. Cancer Biol. 11, 119–128. Cristofanilli, M., Budd, G. T., Ellis, M. J., Stopeck, A., Matera, J., Miller, M. C., Reuben, J. M., Doyle, G. V., Allard, W. J., Terstappen, L. W., and Hayes, D. F. (2004). Circulating tumor cells, disease progression, and survival in metastatic breast cancer. N. Engl. J. Med. 351, 781–791. Donovan, D., Brown, N. J., Bishop, E. T., and Lewis, C. E. (2001). Comparison of three in vitro human “angiogenesis” assays with capillaries formed in vivo. Angiogenesis 4, 113–121. Ewing, C. M., Ru, N., Morton, R. A., Robinson, J. C., Whellock, M. J., Johnson, K. R., Barrett, J. C., and Isaacs, W. B. (1995). Chromosome 5 suppresses tumorigenicity of PC3 prostate cancer cells: Correlation with re‐expression of a‐catenin and restoration of E‐cadherin function. Cancer Res. 55, 4813–4817. Ewing, J. (1922). In “Neoplastic Diseases,” pp. 784–785. WB Saunders Company, Philadelphia, London. Fidler, I. J. (2003). The pathogenesis of cancer metastasis: “The seed and soil” hypothesis revisited. Nat. Rev. Cancer 3, 453–458. Fiegel, H. C., Havers, J., Kneser, U., Smith, M. K., Moeller, T., Kluth, D., Mooney, D. J., Rogiers, X., and Kaufmann, P. M. (2004). Influence of flow conditions and matrix coatings on growth and differentiation of three‐dimensionally cultured rat hepatocytes. Tissue Eng. 10, 165–174.

Liver Microtissue Bioreactor to Study Tumor Metastasis

243

Franko, A. J., and Koch, C. J. (1983). The radiation response of hypoxic cells in EMT6 spheroids in suspension culture does model data from EMT6 tumors. Radiat. Res. 96, 497–504. Gerlach, J. C., Schnoy, N., Encke, J., Smith, M. D., Muller, C., and Neuhaus, P. (1995). Improved hepatocyte in vitro maintenance in a culture model with woven multicompartment capillary systems: Electron microscopy studies. Hepatology 22, 546–552. Ghiso, J. A. A., Kovalski, K., and Ossowski, L. (1999). Tumor dormancy induced by downregulation of urokinase receptor in human carcinoma involves integrin and MAPK signaling. J. Cell Biol. 147, 89–103. Gleave, M. E., Hsieh, J. T., vonEschenbach, A. C., and Chung, L. W. K. (1992). Prostate and bone fibroblasts induce human prostate cancer growth in vivo: Implications for bidirectional tumor‐stromal cell interaction in prostate carcinoma growth and metastasis. J. Urol. 147, 1151–1159. Goswami, S., Sahai, E., Wyckoff, J. B., Cammer, M., Cox, D., Pixley, F. J., Stanley, E. R., Segall, J. E., and Condeelis, J. S. (2005). Macrophages promote the invasion of breast carcinoma cells via a colony‐stimulating factor‐1/epidermal growth factor paracrine loop. Cancer Res. 65, 5278–5283. Graff, J. R., Gabrielson, E., Fujii, H., Baylin, S. B., and Herman, J. G. (2000). Methylation patterns of the E‐cadherin 50 X island are unstable and reflect the dynamic, heterogeneous loss of E‐cadherin expression during metastatic progression. J. Biol. Chem. 275, 2727–2732. Griffith, L. G., and Swartz, M. A. (2006). Capturing complex 3D tissue physiology in vitro. Nat. Rev. Mol. Cell. Biol. 7, 211–224. Griffith, L. G., Wu, B., Cima, M. J., Powers, M. J., Chaignaud, B., and Vacanti, J. P. (1997). In vitro organogenesis of liver tissue. Ann. NY Acad. Sci. 831, 382–397. Grunert, S., Jechlinger, M., and Beug, H. (2003). Diverse cellular and molecular mechanisms contribute to epithelial plasticity and metastasis. Nat. Rev. Mol. Cell. Biol. 4, 657–665. Helm, C. E., Fleury, M. E., Zisch, A. H., Boschetti, F., and Swartz, M. A. (2005). Synergy between interstitial flow and VEGF directs capillary morphogenesis in vitro through a gradient amplification mechanism. Proc. Natl. Acad. Sci. USA 44, 15779–15784. Jasmund, I., and Bader, A. (2002). Bioreactor developments for tissue engineering applications by the example of the bioartificial liver. Adv. Biochem. Eng. Biotechnol. 74, 99–109. Jones, P. A., and Baylin, S. B. (2002). The fundamental role of epigenetic events in cancer. Nat. Rev. Genet. 3, 415–428. Kaihara, S., Kim, S., Kim, B. S., Mooney, D. J., Tanaka, K., and Vacanti, J. P. (2000). Survival and function of rat hepatocytes cocultured with nonparenchymal cells or sinusoidal endothelial cells on biodegradable polymers under flow conditions. J. Pediatr. Surg. 35, 1287–1290. Kaplan, R. N., Riba, R. D., Zacharoulis, S., Bramley, A. H., Vincent, L., Costa, C., MacDonald, D. D., Jin, D. K., Shido, K., Kerns, S. A., Zhu, Z., Hicklin, D., et al. (2005). VEGFR1‐ positive haematopoietic bone marrow progenitors initiate the pre‐metastatic niche. Nature 438, 820–827. Kassis, J., Moellinger, J., Lo, H., Greenberg, N., Kim, H.‐G., and Wells, A. (1999). A role for phospholipase C‐‐mediated signaling in tumor cell invasion. Clin. Cancer Res. 5, 2251–2260. Kassis, J., Lauffenburger, D. A., Turner, T., and Wells, A. (2001). Tumor invasion as dysregulated cell motility. Semin. Cancer Biol. 11, 105–118. Kassis, J., Radinsky, R., and Wells, A. (2002). Motility is rate‐limiting for invasion of bladder carcinoma cell lines. Int. J. Biochem. Cell Biol. 34, 262–275. Kim, H., Turner, T., Kassis, J., Souto, J., and Wells, A. (1999). EGF receptor signaling in prostate development. Histol. Histopathol. 14, 1175–1182.

244

Clayton Yates et al.

Kowalski, P. J., Rubin, M. A., and Kleer, C. G. (2003). E‐cadherin expression in primary carcinoma of the breast and its distant metastases. Breast Cancer Res. 5, R217–R222. Kunz‐Schughart, L. A. (1999). Multicellular tumor spheroids: Intermediates between monolayer culture and in vivo tumor. Cell Biol. Int. 23, 157–161. Landman, K. A., and Please, C. P. (2001). Tumour dynamics and necrosis: Surface tension and stability. Math. Med. Biol. 18, 131–158. Luzzi, K. J., MacDonald, I. C., Schmidt, E. E., Kerkvliet, N., Morris, V. L., Chambers, A. F., and Groom, A. C. (1998). Multistep nature of metastatic inefficiency: Dormancy of solitary cells after successful extravasation and limited survival of early micrometastases. Am. J. Pathol. 153, 865–873. Mamoune, A., Luo, J.‐H., Lauffenburger, D. A., and Wells, A. (2003). m‐Calpain as a target for limiting prostate cancer invasion. Cancer Res. 63, 4632–4640. Mamoune, A., Kassis, J., Kharait, S., Kloeker, S., Manos, E., Jones, D. A., and Wells, A. (2004). DU145 human prostate carcinoma invasiveness is modulated by urokinase receptor (uPAR) downstream of epidermal growth factor receptor (EGFR) signaling. Exp. Cell Res. 299, 91–100. Margolis, L., Hatfill, S., Chuaqui, R., Vocke, C., Emmert‐Buck, M., Linehan, W. M., and Duray, P. H. (1999). Long term organ culture of human prostate tissue in a NASA‐designed rotating wall bioreactor. J. Urol. 161, 290–297. Martin, I., Wendt, D., and Heberer, M. (2004). The role of bioreactors in tissue engineering. Trends Biotechnol. 22, 80–86. Michalopoulos, G. K., Bowen, W. C., Zajac, V. F., Beer‐Stolz, D., Watkins, S., Kostrubsky, V., and Strom, S. (1999). Morphogenetic events in mixed cultures of rat hepatocytes and non‐ parenchymal cells maintained in biological matrices in the presence of hepatocyte growth factor and epidermal growth factor. Hepatology 29, 90–100. Minn, A. J., Gupta, G. P., Siegel, P. M., Bos, P. D., Shu, W., Girl, D. D., Viale, A., Olshen, A. B., Gerald, W. L., and Massague, J. (2005). Genes that mediate breast cancer metastasis to lung. Nature 436, 518–524. Navarro, F. A., Mizuno, S., Huertas, J. C., Glowacki, J., and Orgill, D. P. (2001). Perfusion of medium improves growth of human oral neomucosal tissue constructs. Wound Repair Regen. 9, 507–512. Nelson, C. M., and Bissell, M. J. (2005). Modeling dynamic reciprocity: Engineering three‐ dimensional culture models of breast architecture, function, and neoplastic transformation. Semin. Cancer Biol. 15, 342–352. Paget, S. (1989). The distribution of secondary growths in cancer of the breast. 1889. Cancer Metastasis Rev. 8, 98–101. Powers, M. J., and Griffith, L. G. (1998). Adhesion‐guided in vitro morphogenesis in pure and mixed cultures. Microsc. Res. Tech. 43, 379–384. Powers, M. J., Domansky, K., Capitano, A., Kalezi, A., Udapadhaya, A., Kaazempur‐Mofrad, M. R., Kurzawski, P., Wack, K. E., Stolz, D. B., Kamm, R., and Griffith, L. G. (2002a). A microarray perfusion bioreactor for 3D liver culture. Biotechnol. Bioeng. 78, 257–269. Powers, M. J., Janigian, J., Wack, K. E., Baker, C., Stolz, D. B., and Griffith, L. G. (2002b). Functional behavior of primary rat liver cells in a 3D perfused microarray bioreactor. Tissue Eng. 8, 499–513. Proia, D. A., and Kuperwasser, C. (2005). Stroma: Tumor agonist or antagonist. Cell Cycle 4, 1022–1025. Rhiel, M. H., Cohen, M. B., Arnold, M. A., and Murhammer, D. W. (2004). On‐line monitoring of human prostate cancer cells in a perfusion rotating wall vessel by near‐infrared spectroscopy. Biotechnol. Bioeng. 86, 852–861.

Liver Microtissue Bioreactor to Study Tumor Metastasis

245

Salven, P., Mustjoki, S., Alitalo, R., Alitalo, K., and Rafii, S. (2003). VEGFR‐3 and CD133 identify a population of CD34þ lymphatic/vascular endothelial precursor cells. Blood 101, 168–172. Santini, M. T., and Rainaldi, G. (1999). Three‐dimensional spheroid model in tumor biology. Pathobiology 67, 148–157. Schmeichel, K. L., and Bissell, M. J. (2003). Modeling tissue‐specific signaling and organ function in three dimensions. J. Cell Sci. 116, 2377–2388. Semino, C. E., Kamm, R. D., and Lauffenburger, D. A. (2006). Autocrine EGF receptor activation mediates endothelial cell migration and vascular morphogenesis induced by VEGF under interstitial flow. Exp. Cell Res. 312, 289–298. Shah, R. B., Mehra, R., Chinnaiyan, A. M., Shen, R., Ghosh, D., Zhou, M., MacVicar, G. R., Varambally, S., Harwood, J., Bismar, T. A., Kim, R., Rubin, M. A., et al. (2004). Androgen‐ independent prostate cancer is a heterogeneous group of diseases. Cancer Res. 64, 9209–9216. Sivaraman, A., Iida, T., Leach, J. K., Townsend, S., Hogan, B. J., Fry, R., Samson, L., Tannenbaum, S. R., and Griffith, L. G. (2005). A microscale in vitro physiological modle of the liver: Preditive screens for drug metabolism and enzyme induction. Curr. Drug Metab. 6, 569–592. Tarin, D. (2005). The fallacy of epithelial mesenchymal transition in neoplasia. Cancer Res. 65, 5996–6000. Thompson, E. W., and Newgreen, D. F. (2005). Carcinoma invasion and metastasis: A role for epithelial‐mesenchymal transition. Cancer Res. 65, 5991–5995. Timmers, M., Vekemans, K., Vermijlen, D., Asosingh, K., Kuppen, P., Bouwens, L., Wisse, E., and Braet, F. (2004). Interactions between rat colon carcinoma cells and kupffer cells during the onset of hepatic metastasis. Int. J. Cancer 112, 793–802. Timmins, N. E., Dietmair, S., and Nielsen, L. K. (2004). Hanging‐drop multicellular spheroids as a model of tumour angiogenesis. Angiogenesis 7, 97–103. Tlsty, T. D. (2001). Stromal cells can contribute oncogenic signals. Semin. Cancer Biol. 11, 97–104. Torisawa, Y. S., Shiku, H., Yasukawa, T., Nishizawa, M., and Matsue, T. (2005). Multi‐channel 3‐D cell culture device integrated on a silicon chip for anticancer drug sensitivity test. Biomaterials 26, 2165–2172. Tsai, K. K., Chuang, E. Y., Little, J. B., and Yuan, Z. M. (2005). Cellular mechanisms for low‐ dose ionizing radiation‐induced perturbation of the breast tissue microenvironment. Cancer Res. 65, 6734–6744. Turner, T., Chen, P., Goodly, L. J., and Wells, A. (1996). EGF receptor signaling enhances in vivo invasiveness of DU‐145 human prostate carcinoma cells. Clin. Exp. Metastasis 14, 409–418. Turner, T., VanEpps‐Fung, M., Kassis, J., and Wells, A. (1997). Molecular inhibition of PLC signaling abrogates DU‐145 prostate tumor cell invasion. Clin. Cancer Res. 3, 2275–2282. Wang, W., Goswami, S., Sahai, E., Wyckoff, J. B., Segall, J. E., and Condeelis, J. S. (2005). Tumor cells caught in the act of invading: Their strategy for enhanced cell motility. Trends Cell Biol. 15, 138–145. Wells, A. (2000). Tumor invasion: Role of growth factor‐induced cell motility. Adv. Cancer Res. 78, 31–101. Wyckoff, J., Wang, W., Lin, E. Y., Wang, Y., Pixley, F., Stanley, E. R., Graf, T., Pollard, J. W., Segall, J., and Condeelis, J. (2004). A paracrine loop between tumor cells and macrophages is required for tumor cell migration in mammary tumors. Cancer Res. 64, 7022–7029.

246

Clayton Yates et al.

Xie, H., Turner, T., Wang, M.‐H., Singh, R. K., Siegal, G. P., and Wells, A. (1995). In vitro invasiveness of DU‐145 human prostate carcinoma cells is modulated by EGF receptor‐ mediated signals. Clin. Exp. Metastasis 13, 407–419. Yates, C., Wells, A., and Turner, T. (2005). Luteinizing hormone releasing hormone (LHRH) analog reverses the cell adhesion profile of DU‐145 human prostate carcinoma. Br. J. Cancer 92, 366–375. Zhao, F., Pathi, P., Grayson, W., Xing, Q., Locke, B. R., and Ma, T. (2005). Effects of oxygen transport on 3‐d human mesenchymal stem cell metabolic activity in perfusion and static cultures: Experiments and mathematical model. Biotechnol. Progress 21, 1269–1280.

PDGF Receptors as Targets in Tumor Treatment Arne O¨stman* and Carl‐Henrik Heldin{ *Department of Pathology‐Oncology, Cancer Center Karolinska, Karolinska Institutet, R8:03, SE‐171 76 Stockholm, Sweden; { Ludwig Institute for Cancer Research, Uppsala University, SE‐751 24 Uppsala, Sweden

I. Molecular Biology of PDGF A. PDGF Isoforms and PDGF Receptors B. Signaling via PDGF Receptors II. Physiological Roles of PDGF III. Roles of PDGF Receptors in Tumors A. PDGF Stimulation of Malignant Cells B. Tumor Angiogenesis and PDGF Receptor Signaling C. PDGF and Recruitment of Tumor Fibroblasts D. Regulation of Tumor Drug Uptake and IFP by PDGF Receptors E. Implications of Roles for PDGF Receptor Signaling in Metastasis IV. Clinical Studies A. PDGF Antagonists B. Clinical Effects Ascribed to PDGF Receptor Inhibition V. Future Perspectives References

Signaling through platelet‐derived growth factor (PDGF) receptors contributes to multiple tumor‐associated processes. The recent introduction of clinically useful PDGF inhibitors have the last years validated PDGF receptors in malignant and stromal cells as relevant cancer drug targets. Mutational activation of PDGF receptor signaling in malignant cells has been described in some rare tumor types such as dermatofibrosarcoma protuberans, a subset of GISTs, and some hematologic malignancies. Furthermore, expression of PDGF receptors on pericytes is a common characteristic of solid tumors. The clinical efficacy of novel multikinase inhibitors, such as sunitinib and sorafenib, most likely involves targeting of PDGF receptor‐dependent pericytes. Preclinical studies suggest that targeting of stromal PDGF receptors might also constitute a novel strategy to enhance tumor drug uptake. Finally, recent studies have implied both pro‐ and antimetastatic effects of PDGF receptors on malignant and stromal cells. The studies on the roles of PDGF receptors in cancer signaling are thus presently in a dynamic phase where collaborations between oncologists, pathologists, and tumor biologists are predicted to be highly productive. # 2007 Elsevier Inc.

Advances in CANCER RESEARCH Copyright 2007, Elsevier Inc. All rights reserved.

247

0065-230X/07 $35.00 DOI: 10.1016/S0065-230X(06)97011-0

Arne O¨stman and Carl‐Henrik Heldin

248

I. MOLECULAR BIOLOGY OF PDGF Members of the platelet‐derived growth factor (PDGF) family stimulate the proliferation, survival, and motility of connective tissue cells and certain other cell types (Heldin and Westermark, 1999). PDGF isoforms have important roles during the embryonal development, particularly to promote the development of various mesenchymal cell types in different organs (Betsholtz, 2004). In the adult, PDGF stimulates normal wound healing (Robson et al., 1992) and regulates the interstitial fluid pressure (IFP) of tissues (Rodt et al., 1996). Overactivity of PDGF has been linked to tumorigenesis, as well as to the development of other diseases involving excessive cell proliferation, such as ¨ stman and Heldin, 2001). atherosclerosis and various fibrotic conditions (O In the case of cancer, PDGF receptor activation can drive tumor growth directly by autocrine PDGF stimulation or activating mutations in PDGF receptors (Fig. 1). However, PDGF produced by cancer cells, which themselves do not respond to PDGF, can also act in a paracrine manner on nontumor cells, such as cells in tumor blood vessels and stromal fibroblasts, which may also be important for tumor growth and homeostasis (Fig. 1) (Pietras et al., 2003a).

PDGF

1

3 2

Fig. 1 Effects of PDGF in tumors. PDGF can, in an autocrine or paracrine manner, stimulate the growth and survival of certain types of tumor cells (1). PDGF can also stimulate stromal fibroblasts (2), and cells of blood vessels, particularly pericytes (3). For references, see the text.

PDGF Receptors as Targets in Tumor Treatment

249

The purpose of the present communication is to review the development of PDGF antagonists, and their use in preclinical animal models and patients for the treatment of tumors characterized by autocrine or paracrine PDGF stimulation.

A. PDGF Isoforms and PDGF Receptors The PDGF family consists of five isoforms that are homodimers of A‐, B‐, C‐, and D‐polypeptide chains, that is PDGF‐AA, ‐BB, ‐CC, and ‐DD, and a heterodimer PDGF‐AB (Heldin et al., 2002). The A‐ and B‐chains are synthesized as inactive precursors, but are cleaved during secretion from the producer cell, and are thus present extracellularly in active forms. In contrast, the C‐ and D‐chains are secreted as inactive forms containing N‐terminal CUB domains, which have to be removed before these isoforms can bind to receptors (Li and Eriksson, 2003). It has been shown that PDGF‐CC and ‐DD are activated by tissue plasminogen activator and urokinase plasminogen activator, respectively (Fredriksson et al., 2004; Ustach and Kim, 2005). The PDGF isoforms exert their cellular effects by binding to structurally similar ‐ and ‐tyrosine kinase PDGF receptors. Each receptor contains five extracellular Ig‐like domains to which ligands bind, and an intracellular tyrosine kinase domain which has a characteristic inserted sequence of about 100‐amino acid residues without similarity to kinase domains. The receptors are activated by ligand‐induced receptor dimerization. Since the A‐, B‐, and C‐chains of PDGF bind to the ‐receptors, whereas the B‐ and D‐chains bind to the ‐receptor, different types of receptor dimers are formed depending on which PDGF isoform that binds, and on which receptor isoforms the target cell expresses. Ligand‐induced receptor dimerization allows autophosphorylation in trans between the receptor subunits in the dimer. Phosphorylation of a tyrosine residue in the activation loop of the kinase and of specific tyrosine residues in other regions of the cytoplasmic parts of the receptors leads to an increase in the catalytic activity of the kinase. Moreover, the phosphorylated tyrosines provide docking sites for downstream signaling molecules containing SH2 domains.

B. Signaling via PDGF Receptors The PDGF ‐ and ‐receptor homo‐ and heterodimers induce similar, but not identical, cellular effects. Whereas all dimeric receptor complexes mediate potent mitogenic effects, only  homodimers and  heterodimers mediate chemotaxis of smooth muscle cells and fibroblasts; activation of 

250

Arne O¨stman and Carl‐Henrik Heldin

homodimeric receptors, in fact, inhibits chemotaxis. Moreover, whereas all receptor dimers mediate rearrangements of the actin filament system of cells, only ‐ and ‐receptor dimers stimulate the formation of circular actin structures of the dorsal surface of the cell (Eriksson et al., 1992). The cellular effects of PDGF isoforms are mediated by activation of several signal transduction pathways at the respective receptor complexes. Many of these signaling pathways are initiated by the docking of SH2‐ domain‐containing molecules at specific autophosphorylated residues in the receptors. In total, at least 11 and 12 autophosphorylation sites have been identified in the PDGF ‐ and ‐receptors, respectively (Heldin et al., 1998). They bind, in a specific manner, about 10 different types of SH2‐ domain‐containing molecules, including tyrosine kinases of the Src family, phosphatidylinositol‐30 ‐kinase (PI3K), phospholipase C‐1 (PLC‐1), the Grb2/Sos1 complex that activates Ras and the Erk MAP kinase pathway, GTPase‐activating protein for Ras (RasGAP), the tyrosine phosphatase SHP‐2, and transcription factors of the STAT family. In addition, the activated receptors bind several SH2‐domain‐containing adaptor molecules, including Nck, Shc, and Crk, that do not have any enzymatic activity and whose function is to form bridges between the receptors and other signaling molecules. Accumulating observations have shown that PI3K and PLC‐ are particularly important for PDGF‐induced actin reorganization and chemotaxis, whereas the Erk MAP kinase pathway and Src are particularly important for stimulation of cell proliferation. However, there is an extensive cross‐ talk between the different signaling pathways, and the ultimate effect of activation of different signaling pathways may differ between different cell types. An interesting aspect of signaling via PDGF receptors, and other receptors, is that modulatory signals are often induced in parallel to stimulatory signals. Thus, activation of Ras occurs by docking to the PDGF receptors of the Grb2/Sos1 complex. However, the PDGF ‐receptor, but not the ‐receptor, also binds RasGAP which counteracts the Sos1‐induced Ras activation. Since Grb2/Sos1 and RasGAP bind to different phosphotyrosines, the efficiency in Ras activation is determined by the stoichiometry of phosphorylation of these residues. Interestingly, the phosphorylation of Tyr771 in the PDGF ‐receptor, which binds RasGAP, is higher in a ‐receptor homodimer than in an ‐receptor heterodimer, leading to different efficiencies in Ras activation by the different dimeric receptor complexes (Ekman et al., 2002). Moreover, activation of PI3K by PDGF receptors is modulated by the docking to the receptor, via the PDZ‐domain‐containing protein Naþ/Hþ exchanger isoform 3 regulatory factor (NHERF), of the phosphoinositide phosphatase PTEN which dephosphorylates the PI3K product (Takahashi et al., 2006). In addition, the tyrosine phosphatase SHP‐2 binds to ‐ and

PDGF Receptors as Targets in Tumor Treatment

251

‐receptors and counteracts the phosphorylation induced by the receptors, and thus modulates the strength of the signaling from the receptors. Importantly, different tyrosine phosphatases can selectively dephosphorylate different autophosphorylation sites in the PDGF receptors, and thereby modulate signaling (Klinghoffer and Kazlauskas, 1995; Kovalenko et al., 2000; Persson et al., 2004). In fact, efficient PDGF signaling is dependent on transient and reversible inhibition of tyrosine phosphatases, which is obtained by PDGF‐induced production of H2O2 which oxidizes and thereby inactivates phosphatases (Bae et al., 2000; Sundaresan et al., 1995). Interestingly, peroxiredoxin type II, a cellular peroxidase that eliminates H2O2, was shown to associate with the PDGF receptor and to suppress tyrosine phosphatase inactivation, thereby suppressing PDGF receptor activation (Choi et al., 2005). Signaling via PDGF receptors has also been shown to be modulated by certain membrane proteins. Thus, binding of urokinase to its receptor (UPAR) induces an association with and activation of the PDGF ‐receptor in a PDGF‐independent manner (Kiyan et al., 2005). Moreover, the PDGF ‐receptor forms a complex with the hyaluronan receptor CD44; binding of hyaluronan to CD44 suppresses PDGF ‐receptor activation most likely by recruiting a tyrosine phosphatase to the receptor (Li et al., 2006).

II. PHYSIOLOGICAL ROLES OF PDGF During the embryonal development, PDGF isoforms are often expressed by epithelial cells in various organs, and PDGF receptors on neighboring mesenchymal cells, suggesting paracrine roles for the PDGF isoforms in the development of different types of mesenchymal cell types in different organs. Detailed insights into the physiological roles of PDGF isoforms have come from the knockout of genes for PDGF isoforms and receptors in mice (Betsholtz, 2004). Inactivation of the gene for PDGF‐B (Leveen et al., 1994) or the ‐ receptor (Soriano, 1994) gives similar phenotypes. On one hand, mesangial cells of kidneys do not develop, resulting in poor filtration in the glomeruli. On the other hand, there is a defect in the development of smooth muscle cells of the blood vessel walls, resulting in bleedings at the time of birth, which is the cause of death of these animals. The fact that the phenotypes of B‐chain and ‐receptor knockout mice are so similar suggests that during embryonal development, PDGF‐BB is a more important ligand for this receptor than PDGF‐DD. Mice in which the PDGF A‐chain gene has been knocked out die at about 3 weeks of age because of lung emphysema (Bostro¨m et al., 1996). The defect involves poor development of lung alveoli because of lack of spreading of

252

Arne O¨stman and Carl‐Henrik Heldin

alveolar smooth muscle cell progenitors. There is also an abnormal mucosal lining in the gastrointestinal tract with fewer villus clusters in A‐chain knockout mice (Karlsson et al., 2000). Knockout of the PDGF C‐chain gene causes perinatal death due to feeding and respiratory difficulties (Ding et al., 2004). When both the A‐ and C‐chain genes were inactivated a severe phenotype was observed, including cleft palate, subepidermal blistering, deficiency of renal cortex mesenchyme, spina bifida, and skeletal and vascular defects (Ding et al., 2004). Interestingly, these defects phenocopies the loss of the PDGF ‐receptor (Soriano, 1997), consistent with the notion that the ‐receptor in vivo is stimulated by both PDGF‐AA and PDGF‐CC. The different roles of the ‐ and ‐receptors during embryogenesis can largely be explained by their different expression patterns. However, differences in their signaling capacities also contributes, as shown in an in vivo experiment in which the intracellular parts of the receptors were exchanged (Klinghoffer et al., 2001). Interestingly, the intracellular ‐receptor part failed to mediate the ‐receptor’s effect on vascular development. In the adult, PDGF has been shown to promote wound healing (Robson et al., 1992). Topical application of PDGF‐BB in the wounded area increases the amount of granulation tissue through stimulation of mitogenicity and chemotaxis of fibroblasts and smooth muscle cells, and chemotaxis of neutrophiles and macrophages, as well as through stimulation of the production of various matrix molecules. The importance of PDGF for the connective tissue is also reflected by the fact that it regulates the IFP of tissues (Rodt et al., 1996). The mechanism probably involves the ability of PDGF to enhance the formation of contacts between connective tissue components and stromal fibroblasts, and to stimulate contractility of these cells. Signaling via PI3K was found to be particularly important for this effect (Heuchel et al., 1999).

III. ROLES OF PDGF RECEPTORS IN TUMORS A. PDGF Stimulation of Malignant Cells 1. MALIGNANCIES ASSOCIATED WITH MUTATIONAL ACTIVATION OF PDGF SIGNALING In general, mutational activation of PDGF receptor signaling is rare in solid tumors and hematologic malignancies. However, some rare diseases are associated with mutations in PDGF ligands and receptors (Table I). Dermatofibrosarcoma protuberans (DFSP) and the juvenile giant cell fibroblastoma (GCF) are rare skin tumors of intermediate malignancy

Table I

Tumors with Known Genetic Alterations of PDGF or PDGF Receptor Genes

Tumor Dermatofibrosarcoma protuberans (DFSP)

Genetic alteration

Estimated frequency of genetic alteration in the disease (%)

Activating mechanism

Imatinib sensitivity

Clinical effect of imatinib treatment

Translocation of the PDGFB gene to the COL1A1 gene Mutations in the PDGFRA gene

100

Autocrine PDGF‐BB stimulation

Yes

Yes

10

Fusion of the PDGFRB gene with the TEL gene