Adoptive Immunotherapy Ludewig Hoffmann (Methods Molec Medicine 109 Humana 2005)

M E T H O D S I N M O L E C U L A R M E D I C I N E TM
Adoptive
Immunotherapy
Methods and Protocols
Edited by
Burkhard Ludewig
Matthias W. Hoffmann
Adoptive Immunotherapy
METHODS IN MOLECULAR MEDICINE TM
113.
113 Multiple Myeloma: Methods and Protocols, 96.
96 Hepatitis B and D Protocols: Volume 2,
edited by Ross D. Brown and P. Joy Ho, Immunology, Model Systems, and Clinical
2005 Studies, edited by Robert K. Hamatake and
112.
112 Molecular Cardiology: Methods and
Johnson Y. N. Lau, 2004
Protocols, edited by Zhongjie Sun, 2005 95.
95 Hepatitis B and D Protocols: Volume 1,
Detection, Genotypes, and
111.
111 Chemosensitivity: Volume 2, In Vivo
Characterization, edited by Robert K.
Models, Imaging, and Molecular Regulators, Hamatake and Johnson Y. N. Lau, 2004
edited by Rosalyn D. Blumethal, 2005
94.
94 Molecular Diagnosis of Infectious
110.
110 Chemosensitivity: Volume 1, In Vitro Diseases, Second Edition, edited by Jochen
Assays, edited by Rosalyn D. Blumethal, Decker and Udo Reischl, 2004
2005 93.
93 Anticoagulants, Antiplatelets, and
109.
109 Adoptive Immunotherapy, Methods and Thrombolytics, edited by Shaker A.
Protocols, edited by Burkhard Ludewig and Mousa, 2004
Matthias W. Hoffmann, 2005 92.
92 Molecular Diagnosis of Genetic Diseases,
108.
108 Hypertension, Methods and Protocols, Second Edition, edited by Rob Elles and
edited by Jrme P. Fennell and Andrew H. Roger Mountford, 2004
Baker, 2005 91.
91 Pediatric Hematology: Methods and
107.
107 Human Cell Culture Protocols, Second Protocols, edited by Nicholas J. Goulden
Edition, edited by Joanna Picot, 2005 and Colin G. Steward, 2003
106.
106 Antisense Therapeutics, Second Edition, 90.
90 Suicide Gene Therapy: Methods and
edited by M. Ian Phillips, 2005 Reviews, edited by Caroline J. Springer, 2004
105.
105 Developmental Hematopoiesis: Methods 89.
89 The BloodBrain Barrier: Biology and
and Protocols, edited by Margaret H. Research Protocols, edited by Sukriti Nag,
Baron, 2005 2003
104.
104 Stroke Genomics: Methods and Reviews, 88.
88 Cancer Cell Culture: Methods and
edited by Simon J. Read and David Virley, Protocols, edited by Simon P. Langdon, 2003
2005 87.
87 Vaccine Protocols, Second Edition, edited
103.
103 Pancreatic Cancer: Methods and Protocols, by Andrew Robinson, Michael J. Hudson,
edited by Gloria H. Su, 2005 and Martin P. Cranage, 2003
102.
102 Autoimmunity: Methods and Protocols, 86.
86 Renal Disease: Techniques and Protocols,
edited by Andras Perl, 2004 edited by Michael S. Goligorsky, 2003
101.
101 Cartilage and Osteoarthritis: Volume 2, 85.
85 Novel Anticancer Drug Protocols, edited
Structure and In Vivo Analysis, edited by by John K. Buolamwini and Alex A. Adjei,
Frederic De Ceuninck, Massimo Sabatini, 2003
and Philippe Pastoureau, 2004 84.
84 Opioid Research: Methods and Protocols,
100.
100 Cartilage and Osteoarthritis: Volume 1, edited by Zhizhong Z. Pan, 2003
Cellular and Molecular Tools, edited by 83.
83 Diabetes Mellitus: Methods and Protocols,
Massimo Sabatini, Philippe Pastoureau, edited by Sabire zcan, 2003
and Frederic De Ceuninck, 2004 82.
82 Hemoglobin Disorders: Molecular
99.
99 Pain Research: Methods and Protocols, Methods and Protocols, edited by Ronald
edited by David Z. Luo, 2004 L. Nagel, 2003
98.
98 Tumor Necrosis Factor: Methods and 81.
81 Prostate Cancer Methods and Protocols,
Protocols, edited by Angelo Corti and edited by Pamela J. Russell, Paul Jackson,
Pietro Ghezzi, 2004 and Elizabeth A. Kingsley, 2003
97.
97 Molecular Diagnosis of Cancer: Methods 80.
80 Bone Research Protocols, edited by Miep
and Protocols, Second Edition, edited by H. Helfrich and Stuart H. Ralston, 2003
Joseph E. Roulston and John M. S. Bartlett, 79.
79 Drugs of Abuse: Neurological Reviews and
2004 Protocols, edited by John Q. Wang, 2003
M E T H O D S I N M O L E C U L A R M E D I C I N E
Adoptive
Immunotherapy
Methods and Protocols
Edited by
Burkhard Ludewig
Research Department, Kantonsspital St. Gallen, Switzerland
and
Matthias W. Hoffmann
Department of Abdominal and Transplant Surgery,
Hannover Medical School, Hannover, Germany
2005 Humana Press Inc.
999 Riverview Drive, Suite 208
Totowa, New Jersey 07512
www.humanapress.com
All rights reserved.
No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any
means, electronic, mechanical, photocopying, microfilming, recording, or otherwise without written permis-
sion from the Publisher. Methods in Molecular Medicine is a trademark of The Humana Press Inc.
All papers, comments, opinions, conclusions, or recommendations are those of the author(s), and do not
necessarily reflect the views of the publisher.
This publication is printed on acid-free paper.

ANSI Z39.48-1984 (American Standards Institute) Permanence of Paper for Printed Library Materials.
Production Editor: Tracy Catanese

Cover design by Patricia F. Cleary
Cover illustration: Figure 1 from Chapter 3, Delivery of Tumor Antigens to Dendritic Cells Using Biode-
gradable Microspheres, by Ying Waeckerle-Men, Bruno Gander, and Marcus Groettrup. Schematic
representation of the regulation of dendritic cell persistence within secondary lymphoid organs by effector
T-cells. Immune responses are positively regulated by dendritic cells mediating antigen influx into secondary
lymphoid organs from peripheral sites such as tumor tissues.
Photocopy Authorization Policy:

Authorization to photocopy items for internal or personal use, or the internal or personal use of specific
clients, is granted by Humana Press Inc., provided that the base fee of US $25.00 per copy is paid directly
to the Copyright Clearance Center at 222 Rosewood Drive, Danvers, MA 01923. For those organizations that
have been granted a photocopy license from the CCC, a separate system of payment has been arranged and
is acceptable to Humana Press Inc. The fee code for users of the Transactional Reporting Service is: [0-58829-
406-4/05 $25.00].
Printed in the United States of America. 10 9 8 7 6 5 4 3 2 1
e-ISBN: 1-59259-862-5
ISSN: 1543-1894
Library of Congress Cataloging-in-Publication Data
Adoptive immunotherapy : methods and protocols / edited by Burkhard Ludewig and Matthias W. Hoffmann.
p. ; cm. -- (Methods in molecular medicine ; 109)
Includes bibliographical references and index.
ISBN 1-58829-406-4 (alk. paper)
1. Cancer--Immunotherapy. 2. Cellular therapy. 3. Immunocompetent cells--Therapeutic use.
[DNLM: 1. Immunotherapy, Adoptive--methods. 2. Antibodies--therapeutic use. 3. Cytotoxicity,
Immunologic. 4. Dendritic Cells. 5. Stem Cell Transplantation. 6. T-Lymphocytes.
QW 940 A239 2005] I. Ludewig, Burkhard. II. Hoffmann, Matthias W. III. Series.
RC271.I45A335 2005
616.99'4061--dc22
2004010540
Preface
Over the last decade, the advances in cellular and molecular immunology have
been tremendous. Our continuously improving understanding of the immune
system and the appreciation of the mechanisms by which tumors and viral or
bacterial infections are controlled have led to promising new treatment strate-
gies. Adoptive transfer of tailored antigen-specific immune cells and/or opti-
mally designed immunological effector molecules is an elegant and promising
approach to the establishment or restoration of protective immune responses. At
this point, it appears timely to publish the present volume on Adoptive Immuno-
therapy for the Methods in Molecular Medicine series.
Experts from various fields contributed to this comprehensive collection of
state-of-the-art methods for adoptive immunotherapy. Recent technical advances
and specific methods from the areas of dendritic cell therapy, generation of anti-
gen-specific cytotoxic T cells, and antitumor treatment with specific antibodies
have been compiled. Particular emphasis is placed on preclinical and clinical
applications. These chapters are complemented by articles describing the
progress of selected approaches in clinical trials. Leading experts in the field
provide theoretical overviews and point out future directions for the improve-
ment of adoptive immunotherapy. Additional chapters on the molecular defini-
tion of target antigens, mathematical modeling approaches to immunotherapy,
and the utilization of regulatory T cells complete the volume. We were fortunate
to receive support from leading scientists from all over the world who contrib-
uted their most recent experimental protocols and shared their practical advice.
Special emphasis was placed on the Notes section to include helpful trouble-
shooting advice for daily work at the lab bench, which is a unique feature of the
Methods in Molecular Medicine series. It is our hope that we have achieved
our main objective to provide helpful information for scientists just entering the
field of adoptive immunotherapy. Furthermore, this volume attempts to encour-
age more experienced researchers to try novel experimental approaches.
We are indebted to all our authors for generously making their time avail-
able in contributing to this volume and for the excellent quality of their work.
Marcel Kremers help in the final editing process is greatly appreciated.
Finally, we want to thank the Series Editor, John Walker, and the staff of
Humana Press for their support and continuous encouragement.
Burkhard Ludewig
Matthias Hoffmann
v
vi Preface
Contents
Preface .............................................................................................................. v
Contributors .....................................................................................................xi
1 Exploiting Dendritic Cells for Active Immunotherapy
of Cancer and Chronic Infection
David ONeill and Nina Bhardwaj ....................................................... 1
2 A Mathematical Approach for Optimizing Dendritic
Cell-Based Immunotherapy
Gennady Bocharov, Neville J. Ford, and Burkhard Ludewig ............. 19
3 Delivery of Tumor Antigens to Dendritic Cells
Using Biodegradable Microspheres
Ying Waeckerle-Men, Bruno Gander, and Marcus Groettrup ........... 35
4 RNA Transfection of Dendritic Cells
Frank Grnebach, Martin R. Mller, and Peter Brossart ................... 47
5 Isolation and Generation of Clinical-Grade Dendritic Cells
Using the CliniMACS System
John D. M. Campbell, Christoph Piechaczek, Gregor Winkels,
Edith Schwamborn, Daniela Micheli, Sonja Hennemann,
and Jrgen Schmitz ......................................................................... 55
6 Generation of Clinical Grade Monocyte-Derived Dendritic Cells
Using the CliniMACS System
Thomas Putz, Hubert Gander, Reinhold Ramoner,
Claudia Zelle-Rieser, Andrea Rahm, Walter Nussbaumer,
Georg Bartsch, Lorenz Hltl, and Martin Thurnher ...................... 71
7 Adenoviral Transduction of Dendritic Cells
Rienk Offringa, Kitty Kwappenberg, Martijn Rabelink,
Delphine Rea, and Rob Hoeben ..................................................... 83
8 Generation of Autologous Peptide- and Protein-Pulsed
Dendritic Cells for Patient-Specific Immunotherapy
David ONeill and Nina Bhardwaj ..................................................... 97
9 Phenotypical and Functional Characterization
of Clinical-Grade Dendritic Cells
I. Jolanda M. de Vries, Gosse J. Adema, Cornelis J. A. Punt,
and Carl G. Figdor ........................................................................ 113
10 Dendritic Cells in Clinical Trials for Multiple Myeloma
Volker L. Reichardt and Peter Brossart ............................................ 127
vii
viii Contents
11 Identification of Tumor-Associated Autoantigens With SEREX

zlem Treci, Dirk Usener, Sandra Schneider,
and Ugur Sahin ............................................................................. 137
12 Application of Proteomics and Protein Analysis for Biomarker
and Target Finding for Immunotherapy
Petra Weingarten, Petra Lutter, Andreas Wattenberg,
Martin Blueggel, Sonja Bailey, Joachim Klose,
Helmut E. Meyer, and Christoph Huels ....................................... 155
13 Isolation and Expansion of Tumor-Reactive Cytotoxic
T-Cell Clones for Adoptive Immunotherapy
Helga Bernhard, Burkhard Schmidt, Dirk H. Busch,
and Christian Peschel ................................................................... 175
14 Tracking Adoptively Transferred Antigen-Specific T-Cells
With Peptide/MHC Multimers
Norbert Meidenbauer and Andreas Mackensen ............................... 185
15 Gene Transfer of MHC-Restricted Receptors
Helmut W. H. G. Kessels, Monika C. Wolkers,
and Ton N. M. Schumacher .......................................................... 201
16 Immunotherapy With CTL Restricted by Nonself MHC
Liquan Gao, Anne-Marie Downs, and Hans J. Stauss ....................... 215
17 Designing TCR for Cancer Immunotherapy
Ralf-Holger Voss, Jrgen Kuball, Matthias Theobald ....................... 229
18 Isolation and Expansion of Tumor-Specific CD4+ T-Cells
by Means of Cytokine Secretion
Christian Becker ................................................................................ 257
19 Methods for the Ex Vivo Characterization of Human CD8+ T
Subsets Based on Gene Expression and Replicative
History Analysis
Nathalie Rufer, Patrick Reichenbach, and Pedro Romero ............... 265
20 Regulatory T Cells in Antitumor Therapy: Isolation
and Functional Testing of CD4+CD25+ Regulatory T-Cells
Helmut Jonuleit and Edgar Schmitt .................................................. 285
21 Monoclonal Antibody-Based Strategies in Autoimmunity
and Transplantation
Lucienne Chatenoud ......................................................................... 297
22 Producing Bispecific and Bifunctional Antibodies
Dipankar Das and Mavanur R. Suresh .............................................. 329
23 AntibodyCytokine Fusion Proteins for the Therapy of Cancer
Gustavo Helguera and Manuel L. Penichet ...................................... 347
Contents ix
24 Synthesis and Biological Evaluation of a Paclitaxel

Immunoconjugate
Ahmad Safavy and Kevin P. Raisch ................................................... 375
25 Cytotoxic Tumor Targeting With scFv Antibody-Modified
Liposomes
Cornelia Marty and Reto A. Schwendener ....................................... 389
26 Intravenous Immunoglobin Treatment for Fibrosis,
Atherosclerosis, and Malignant Conditions
Ilan Krause and Yehuda Shoenfeld ................................................... 403
27 Study of T-Cell Costimulatory Blockade In Vivo
at a Single Cell Level
Minh Diem Vu and Xian Chang Li .................................................... 409
28 Stem Cell Transplantation: Graft-Mediated Antileukemia Effects
William J. Hogan and Hans Joachim Deeg ....................................... 421
29 Influence of Radiation Protocols on Graft-vs-Host Disease
Incidence After Bone-Marrow Transplantation
in Experimental Models
Sebastian Schwarte and Matthias W. Hoffmann .............................. 445
30 Induction of Chimerism and Tolerance Using Freshly Purified
or Cultured Hematopoietic Stem Cells
in Nonmyeloablated Mice
Nikos Emmanouilidis and Christian P. Larsen .................................. 459
31 Induction of Mixed vs Full Chimerism to Potentiate GVL
Effects After Bone-Marrow Transplantation
Markus Y. Mapara and Megan Sykes ................................................ 469
32 Application of Donor Lymphocytes Expressing a Suicide Gene
for Early GVL Induction and Later Control of GVH Reactions
After Bone-Marrow Transplantation
Attilio Bondanza, Fabio Ciceri, and Chiara Bonini .......................... 475
Index ............................................................................................................ 487
Contributors
GOSSE J. ADEMA Department of Tumor Immunology, Nijmegen Center for
Molecular Life Sciences, University Medical Center Nijmegen,
Nijmegen, the Netherlands
SONJA BAILEY Protagen AG, Dortmund, Germany
NINA BHARDWAJ NYU Cancer Institute Tumor Vaccine Center, New York
University School of Medicine, New York, NY
GEORG BARTSCH Department of Urology, University of Innsbruck,
Innsbruck, Austria
CHRISTIAN BECKER Department of Dermatology, Universittsklinik Mainz,
Mainz, Germany
HELGA BERNHARD Department of Hematology/Oncology, Klinikum rechts
der Isar, Technical University of Munich, Munich, Germany
MARTIN BLUEGGEL Protagen AG, Dortmund, Germany
GENNADY BOCHAROV Institute of Numerical Mathematics, Russian
Academy of Sciences, Moscow, Russia, and Department of Mathematics,
University College Chester, Chester, UK
ATTILIO BONDANZA Experimental Hematology Laboratory, Cancer
Immunotherapy and Gene Therapy Program, San Raffaele Scientific
Institute, Milan, Italy
CHIARA BONINI Experimental Hematology Laboratory, Cancer
PETER BROSSART Department of Hematology, Oncology, Immunology, and
Rheumatology, University of Tbingen, Tbingen, Germany
DIRK H. BUSCH Department of Microbiology/Immunology, Klinikum rechts
JOHN D. M. CAMPBELL Miltenyi Biotec GmbH, Bisley, Surrey, UK
LUCIENNE CHATENOUD Immunologie biologie-IRNEM-INSERM U580,
Hpital Necker Enfants Malades, Paris, France
FABIO CICERI Experimental Hematology Laboratory, Cancer
DIPANKAR DAS Faculty of Pharmacy and Pharmaceutical Sciences,
University of Alberta, Edmonton, Alberta, Canada
xi
xii Contributors
I. JOLANDA M. DE VRIES Department of Tumor Immunology, Nijmegen

Center for Molecular Life Sciences, University Medical Center Nijmegen,
Nijmegen, The Netherlands
HANS JOACHIM DEEG Clinical Research Division, Fred Hutchinson Cancer
Research Center, Seattle, WA
ANNE-MARIE DOWNS Tumour Immunology Section, Department of
Immunology, Division of Medicine, Imperial College London, London,
UK
NIKOS EMMANOUILIDIS Department of Abdominal and Transplant Surgery,
Hannover Medical School, Hannover, Germany
CARL G. FIGDOR Department of Tumor Immunology, Nijmegen Center for
Molecular Life Sciences, University Medical Center Nijmegen, Nijmegen,
The Netherlands
NEVILLE J. FORD Department of Mathematics, University College Chester,
Chester, UK
BRUNO GANDER Institute of Pharmaceutical Sciences, Swiss Federal
Institute of Technology (ETH) Zrich, Zrich, Switzerland
HUBERT GANDER Department of Urology, University of Innsbruck,
Innsbruck, Austria
LIQUAN GAO Tumour Immunology Section, Department of Immunology,
Division of Medicine, Imperial College London, London, UK
MARCUS GROETTRUP Research Department, Cantonal Hospital St. Gallen,
St. Gallen, Switzerland, and Division of Immunology, Department of
Biology, University of Constance, Konstanz, Germany
FRANK GRNEBACH Department of Hematology, Oncology, Immunology,
and Rheumatology, University of Tbingen, Tbingen, Germany
GUSTAVO HELGUERA Department of Microbiology, Immunology, and
Molecular Genetics, and the Molecular Biology Institute, University of
California, Los Angeles, CA
SONJA HENNEMANN Miltenyi Biotec GmbH, Bergisch Gladbach, Germany
ROB HOEBEN Department of Molecular Cell Biology, Leiden University
Medical Center, Leiden, The Netherlands
MATTHIAS W. HOFFMANN Department of Abdominal and Transplant
Surgery, Hannover Medical School, Hannover, Germany
WILLIAM J. HOGAN Clinical Research Division, Fred Hutchinson Cancer
Research Center, Seattle, WA, and Division of Hematology, Mayo Clinic,
Rochester, MN
LORENZ HLTL Department of Urology, University of Innsbruck, Innsbruck,
Austria
Contributors xiii
CHRISTOPH HUELS Protagen AG, Dortmund, Germany

HELMUT JONULEIT Department of Dermatology, Universittsklinik Mainz,
Mainz, Germany
HELMUT W. H. G. KESSELS Department of Immunology, The Netherlands
Cancer Institute, Amsterdam, The Netherlands
JOACHIM KLOSE Institute for Human Genetics, Charit, Berlin, Germany
ILAN KRAUSE Center for Autoimmune Diseases, Sheba Medical Center, Tel-
Hashomer, and Sackler Faculty of Medicine, Tel-Aviv University,
Tel-Aviv, Israel
JRGEN KUBALL Department of Hematology and Oncology, Johannes
Gutenberg-University Mainz, Mainz, Germany
KITTY KWAPPENBERG Department of Immunohematology and Blood
Transfusion, Leiden University Medical Center, Leiden, The Netherlands
XIAN CHANG LI Division of Immunology, Department of Medicine, Harvard
Medical School, Boston, MA
CHRISTIAN P. LARSEN Emory Transplant Center and Department of Surgery,
Emory University School of Medicine, Atlanta, GA
BURKHARD LUDEWIG Research Department, Kantonsspital St. Gallen,
Switzerland
PETRA LUTTER Protagen AG, Dortmund, Germany
ANDREAS MACKENSEN Department of Hematology/Oncology, University of
Regensburg, Regensburg, Germany
MARKUS Y. MAPARA Division of Hematology/Oncology, University of
Pittsburgh Cancer Institute, Pittsburgh, PA
CORNELIA MARTY Molecular Cell Biology, Paul Scherrer Institute, Villigen,
Switzerland
NORBERT MEIDENBAUER Department of Hematology/Oncology, University of
Regensburg, Regensburg, Germany
KEES MELIEF Leiden University Medical Center and University of Leiden,
Leiden, The Netherlands
HELMUT E. MEYER Medical Proteom-Center, Ruhr-University Bochum,
Bochum, Germany
DANIELA MICHELI Miltenyi Biotec GmbH, Bergisch Gladbach, Germany
MARTIN R. MLLER Department of Hematology, Oncology, Immunology,
WALTER NUSSBAUMER Central Institute of Blood Transfusion, University of
Innsbruck, Innsbruck, Austria
DAVID ONEILL NYU Cancer Institute Tumor Vaccine Center, New York
University School of Medicine, New York, NY
xiv Contributors
RIENK OFFRINGA Department of Immunohematology and Blood

Transfusion, Leiden University Medical Center, Leiden, The Netherlands
MANUEL L. PENICHET Department of Microbiology, Immunology, and
Molecular Genetics, and the Molecular Biology Institute, University of
California, Los Angeles, CA
CHRISTIAN PESCHEL Department of Hematology/Oncology, Klinikum rechts
des Isar, Technical University of Munich, Munich, Germany
CHRISTOPH PIECHACZEK Miltenyi Biotec GmbH, Bergisch Gladbach,
Germany
CORNELIS J. A. PUNT Departments of Tumor Immunology and Medical
Oncology, University Medical Center Nijmegen, Nijmegen, The
Netherlands
THOMAS PUTZ Department of Urology, University of Innsbruck, Innsbruck,
Austria
MARTIJN RABELINK Department of Molecular Cell Biology, Leiden
University Medical Center, Leiden, The Netherlands
ANDREA RAHM Department of Urology, University of Innsbruck, Innsbruck,
Austria
KEVIN P. RAISCH Division of Radiation Biology, Department of Radiation
Oncology, and Comprehensive Cancer Center, University of Alabama at
Birmingham, Birmingham, AL
REINHOLD RAMONER Department of Urology, University of Innsbruck,
Innsbruck, Austria
DELPHINE REA Department of Immunohematology and Blood Transfusion,
Leiden University Medical Center, Leiden, The Netherlands
PATRICK REICHENBACH Swiss Institute for Experimental Cancer Research
(ISREC), Epalinges, Switzerland
VOLKER L. REICHARDT Department of Hematology, Oncology, Immunology,
PEDRO ROMERO Division of Clinical Onco-Immunology, Ludwig Institute
for Cancer Research, Lausanne Branch, University Hospital, Lausanne,
Switzerland
NATHALIE RUFER NCCR Molecular Oncology, Swiss Institute for
Experimental Cancer Research (ISREC), Epalinges, Switzerland
AHMAD SAFAVY Division of Radiation Biology, Department of Radiation
Oncology, and Comprehensive Cancer Center, University of Alabama at
Birmingham, Birmingham, AL
UGUR SAHIN Department of Internal Medicine, Johannes Gutenberg-
University, Mainz, Germany
Contributors xv
BURKHARD SCHMIDT Department of Hematology/Oncology, Klinikum rechts

EDGAR SCHMITT Institute of Immunology, University of Mainz, Germany
JRGEN SCHMITZ Miltenyi Biotec GmbH, Bergisch Gladbach, Germany
SANDRA SCHNEIDER Research and Development, Ganymed
Pharmaceuticals, Mainz, Germany
TON N. M. SCHUMACHER Department of Immunology, The Netherlands
Cancer Institute, Amsterdam, The Netherlands
EDITH SCHWAMBORN Miltenyi Biotec GmbH, Bergisch Gladbach, Germany
SEBASTIAN SCHWARTE Department of Radiooncology, Hannover Medical
School, Hannover, Germany
RETO A. SCHWENDENER Paul Scherrer Institute, Molecular Cell Biology,
Villigen, Switzerland
YEHUDA SHOENFELD Sackler Faculty of Medicine, Tel-Aviv University,
Israel
HANS J. STAUSS Tumour Immunology Section, Department of Immunology,
Division of Medicine, Imperial College London, London, UK
MAVANUR R. SURESH Faculty of Pharmacy and Pharmaceutical Sciences,
University of Alberta, Edmonton, Alberta, Canada
MEGAN SYKES Bone Marrow Transplantation Section, Transplantation
Biology Research Center, Massachusetts General Hospital, Harvard
MATTHIAS THEOBALD Department of Hematology and Oncology, Johannes
Gutenberg-University, Mainz, Germany
MARTIN THURNHER Department of Urology, University of Innsbruck,
Innsbruck, Austria
ZLEM TRECI Department of Internal Medicine, Johannes Gutenberg-
University, Mainz, Germany
DIRK USENER Research and Development, Ganymed Pharmaceuticals,
Mainz, Germany
RALF-HOLGER VOSS Department of Hematology and Oncology, Johannes
Gutenberg-University, Mainz, Germany
MINH DIEM VU Division of Immunology, Department of Medicine, Harvard
YING WAECKERLE-MEN Research Department, Cantonal Hospital St.
Gallen, St. Gallen, Switzerland
ANDREAS WATTENBERG Protagen AG, Dortmund, Germany
PETRA WEINGARTEN Protagen AG, Dortmund, Germany
GREGOR WINKELS Miltenyi Biotec GmbH, Bergisch Gladbach, Germany
Contributors xvi
MONIKA C. WOLKERS Department of Immunology, The Netherlands Cancer

Institute, Amsterdam, The Netherlands
CLAUDIA ZELLE-RIESER Department of Urology, University of Innsbruck,
Innsbruck, Austria
Dendritic Cells and Immunotherapy 1
1
Exploiting Dendritic Cells for Active
Immunotherapy of Cancer and Chronic Infection
David ONeill and Nina Bhardwaj
Summary
Dendritic cells (DC) are important antigen-presenting cells (APC) that can prime naive
T-cells and control lymphocyte-mediated adaptive immune responses with respect to magni-
tude, memory, and self-tolerance. Understanding the biology of these cells is central to the
development of new generation immunotherapies for cancer and chronic infection. This chap-
ter presents a brief overview of DC biology and the preparation and use of DC-based vaccines.
Key Words: Dendritic cells; antigen presentation; cancer vaccine; apoptotic cells; cyto-
toxic T-lymphocytes.
1. DC Biology
Protective immunity results from the concerted action of innate and adaptive immune
systems (1). The innate immune system includes phagocytic cells, natural killer (NK)
cells, and complement, and has evolved to respond rapidly to pathogens to protect the
host early in infection. The adaptive immune system, which consists of B- and T-lym-
phocytes, is required for the eventual clearance of many infections and for the genera-
tion of immunologic memory. Both innate and adaptive immunity apparently function
to protect against the development of malignant tumors (24). APC form an important
link between innate and adaptive immunity. APC process intracellular and extracellular
proteins into antigenic peptides, which are then presented on cell-surface major histo-
compatibility complex (MHC) molecules to cells of the adaptive immune system (5).
Because of their ability to express co-stimulatory molecules and cytokines, APC can
stimulate the expansion of lymphocytes that recognize the displayed peptides, initiating
an adaptive immune response. Although monocytes, macrophages, B-cells, and DC can
all function as APC, DC are thought to be the principle initiators of immune responses
(6). In culture with lymphocytes, relatively few DC and very little antigen are needed to
stimulate T-cell responses, and primary responses to antigens may be achieved using
DC in long-term culture (7,8). In both in vitro and in vivo assays, DC are by far the most
potent APCs (9).
From: Methods in Molecular Medicine, vol. 109: Adoptive Immunotherapy: Methods and Protocols
Edited by: B. Ludewig and M. W. Hoffmann Humana Press Inc., Totowa, NJ
1
2 ONeill and Bhardwaj
1.1. DC Subtypes
DC are bone marrow-derived cells that are present in trace amounts in the blood
(<0.1% of blood mononuclear cells) and virtually every tissue. Cell-surface pheno-
typing has shown that in mice and apparently in humans there are as many as five
distinct subpopulations of DC (10,11). It is not completely clear, however, whether all
of these represent distinct cell lineages or are a reflection of functional plasticity. In
humans, the three best-characterized DC subsets may be derived by culturing precur-
sor cells obtained from the blood (10). Cells closely resembling epidermal Langerhans
cells may be obtained from CD34+ hematopoietic progenitor cells (HPC) cultured with
granulocyte/macrophage colony-stimulating factor (GM-CSF), tumor necrosis factor
(TNF)-, and transforming growth factor (TGF)-. Cells resembling so-called dermal
or interstitial DC (also known as DC1) may be obtained by culturing monocytes in
GM-CSF and interleukin (IL)-4 followed by stimulation with proinflammatory
cytokines such as TNF- or with microbial products such as lipopolysaccharide (LPS).
Circulating CD11c BDCA2+ so-called plasmacytoid DC precursors in the blood may
be differentiated into a third type of DC (plasmacytoid DC, or DC2) following expo-
sure to viruses or bacterial (unmethylated CpG motif) DNA. Plasmacytoid DC are
unique in that they travel to the lymph nodes directly from the blood (instead of through
the lymphatics), and that upon stimulation they produce very high levels of type I
interferon (interferon [IFN]- and -) (10). Although it has been thought that most DC
are of myeloid origin, with the exception of plasmacytoid DC, the precise origin of the
different DC subtypes is not completely clear. In mice, which have been studied in
more detail than humans, there is evidence that all DC types can be derived from both
common myeloid and common lymphoid progenitors (12,13), as well as from a third
progenitor cell type that does not have myeloid or lymphoid potential (14).
1.2. Antigen Uptake and Maturation
DC exist in a resting state in virtually every tissue and are recruited to sites of
inflammation by chemokines such as MIP1- (CCL3), MIP1- (CCL4), and RANTES
(CCL5). In tissues, DC and DC precursors capture antigens from a wide range of
sources including bacteria, viruses, dead or dying cells, and extracellular proteins,
peptides, and immune complexes. DC have a host of different antigen uptake recep-
tors for this purpose, including toll-like receptors (TLR) (15,16), Fc receptors (17,18),
and C-type lectins (19). Some of these receptors induce simultaneous uptake and stimu-
latory signals, whereas others are inhibitory receptors that induce suppressive signals
upon antigen uptake.
DC continuously sample their environment for antigens, but they must be stimu-
lated or matured before they can become strong stimulators of the immune system
(9). In fact, there is accumulating evidence that in the absence of maturation stimulus,
DC function to maintain immunologic tolerance to captured antigens (2022).
DC maturation is induced by two classes of stimuli, or danger signals, termed
exogenous and endogenous (15,23,24). Exogenous stimuli are associated with micro-
bial infections and are mediated by signaling through DC pattern-recognition recep-
tors such as TLR (9,15,25). There are at least 10 different TLR, each recognizing
different sets of pathogen-associated molecules. For example, TLR3 recognizes

double-stranded RNA, TLR4 recognizes LPS, and TLR9 recognizes bacterial CpG
motif DNA. Myeloid (as opposed to plasmacytoid) DC express TLRs 2 through 8,
whereas plasmacytoid DC express TLRs 7 and 9, so these different DC subtypes
respond differentially to different TLR ligands. DC may secrete a number of important
inflammatory cytokines following TLR stimulation, including high levels of IL-12,
IFN-, and TNF-, but the specific cytokine profile induced depends upon the subtype
of DC stimulated as well as the nature of the stimulus. For example, plasmacytoid DC
secrete abundant IFN- when stimulated, whereas myeloid dermal DC do not.
Endogenous maturation stimuli originate from inflammatory molecules produced
by cells of the host immune system or by damaged tissues (9,24), and stimulate signal-
ing through specific receptors on the DC. Examples of endogenous stimuli include
TNF- and related molecules expressed on activated lymphocytes such as CD40L and
TRANCE, proinflammatory cytokines such as IL-1, heat shock proteins, growth fac-
tors such as TSLP, adhesion molecules, costimulatory molecules, ligation of Fc recep-
tors, and contact with activated T-cells, T-cells, and NKT-cells.
DC maturation is characterized by decreased phagocytic capacity, enhanced pro-
cessing and presentation of antigens, induced ability to migrate to T-cell areas of lymph
nodes, and increased ability to stimulate T-cell proliferation and cytokine production
(26). In culture, it is at this stage that DC acquire their characteristic dendritic appear-
ance, with numerous cytoplasmic processes, or veils. Maturation is accompanied by
phenotypic changes that include increased cell-surface expression of MHC and co-
stimulatory molecules, including members of the TNF receptor (CD40), TNF (OX40L,
CD27L), and B7 (CD80 and CD86, B7-H3) families. There is downregulation of
chemokine receptors such as CCR2 and CCR5 that direct DC to sites of inflammation
(via MCP1, MIP1-, MIP1-, and RANTES) and upregulation of CCR7, which tar-
gets the DC to lymphatic vessels and lymph nodes via interaction with CCL19 (MIP-
3-, ELC), and CCL21 (SLC).
Antigen processing is highly regulated and controlled by maturation. For example,
following receipt of a maturation signal, DC reduce the pH of endosomes (this facili-
tates processing of exogenously acquired antigens through activation of endosomal
proteases), remove invariant chains from the antigen binding pockets of MHC class II
molecules (so that processed peptides can access the empty pockets), and exocytose
peptide-bound MHC class II molecules to the cell surface together with co-stimula-
tory molecules. Maturation also upregulates the activity of certain proteasome mem-
bers to create immunoproteasomes that may enhance the processing of antigens that
access the MHC class I pathway.
The process of maturation is also accompanied by upregulation of adhesion mol-
ecules such as CD54 (ICAM1), cytokines (e.g., TNF-, IL-12, IL-18), and chemokines
(RANTES, MIP1, IP-10). The latter enable the recruitment of T-cells, monocytes, and
other DC into the local environment. In their mature state, DC express other markers
such as CD83 (a molecule involved in thymic T-cell selection and DCDC interac-
tions) and DC-LAMP, a lysosomal protein that can distinguish mature from immature
DC. In the T-cell areas of lymphoid organs, mature DC live for only 12 d (27), but
their life span can be prolonged if they encounter T-cells that are activated and express-
ing membrane-bound activators such as CD40L and TRANCE (9).
1.3. Antigen Presentation

Through their T-cell antigen receptor (TCR), T-lymphocytes specifically recognize
peptide antigens bound to the highly polymorphic MHC molecules on the APC sur-
face. MHC class I molecules present peptide antigens to CD8+ T-cells, whereas MHC
class II molecules present peptide antigens to CD4+ T-cells. In addition, other, less
polymorphic antigen-presenting molecules structurally similar to MHC class I are also
found on DC. This would include CD1 molecules, which function to present microbial
lipids to antigen-specific T-cells (28,29). As with TLR and endocytic receptors, dif-
ferent DC subtypes have different sets of CD1 molecules. For example, CD1a is found
on epidermal Langerhans cells, whereas CD1b and CD1c are found on dermal den-
dritic cells. CD1d presents specific glycolipids (galactosyl ceramides) to NKT-cells,
cytokine-secreting cells that are important mediators of T-cell immunity (30,31). CD1
molecules have not been found on plasmacytoid DC.
Infection of host cells with viruses results in the processing of viral peptides onto
MHC class I via the so-called classical or endogenous pathway (26,32). DC and most
other cell types can process peptides this way from endogenously synthesized cyto-
plasmic proteins. In this pathway, viral proteins are ubiquitinated and degraded by the
proteasome, and the resulting peptides are transported via transporters for antigen pre-
sentation (TAP) molecules into the endoplasmic reticulum (ER), where they are loaded
onto MHC class I. The MHC class I/peptide complexes exit the ER via the trans-Golgi
network and are transported to the cell surface, where the bound peptide is presented
to the TCR on CD8+ T-cells.
In contrast, processing of antigens onto MHC class II occurs only in APC (26,32).
MHC class II/peptide complexes are formed through endocytosis of extracellular
sources of protein. Endosomes containing the ingested protein mature and fuse with
lysosomes, where acid proteases degrade the proteins into peptide fragments that are
then loaded onto MHC class II molecules. The MHC class II/peptide complexes are
transported to the cell surface within specialized vesicles, where the bound peptide is
presented to the TCR on CD4+ T-cells.
DC are also uniquely capable of so-called exogenous presentation onto MHC class
I, also known as cross-presentation, when processing extracellular antigens (26,32
35). Through this pathway biosynthesis does not need to take place within the DC for
presentation onto MHC class I to occur, permitting DC to elicit CD8+ T-cell responses
to exogenous as well as endogenous antigens. Because of cross-presentation, DC can
display exogenous antigens on both MHC class I and class II, because these antigens
are also processed by the endosomal/lysosomal pathway simultaneously. Cross-pre-
sentation by DC has been shown to be necessary for inducing T-cell immunity to
viruses or other intracellular pathogens that do not infect DC directly (35). In addition,
dead cells and immune complexes can specifically target the cross-presentation path-
way through dedicated receptors on the DC surface (e.g., the Fc receptor for immune
complexes and LOX1 for necrotic cell debris) (17,18,36,37).
1.4. Lymphocyte Activation

DC can migrate very rapidly to lymph nodes, where they can dynamically interact
with many T-cells at a time (38,39). In the lymph node, triggering of a T-cell response
is dependent upon the intensity and length of the DCT-cell interaction (40). Activa-
tion requires two types of signal at the immunological synapse between a DC and a
T-cell, the first between the MHC/peptide complex on the DC and the T-cell antigen
receptor, and the second through assorted co-stimulatory molecules and cytokines.
Through these signals, mature DC can induce T-cells to expand clonally and to differ-
entiate into memory and effector cells. There is a plasticity of this response that is
dependent on many factors, including the antigen dose, the nature of the DC matura-
tion stimulus (in the form of pathogen-derived products and the DC microenviron-
ment), and the state of maturation of the DC. All of these influence the secreted
cytokine profile of the DC and the polarization of responding T-helper (Th) cells in
the form of a Th1 or Th2 response (27,41,42). DC maturation is critical for this induc-
tion of immunity. Immature DCs that circulate through lymph nodes in the steady state
can induce T regulatory responses and are important at maintaining immune tolerance
toward self antigens (21,22,43).
The generation of effector CD8+ T-cells (cytotoxic T-lymphocytes, or CTL) is
particularly important in establishing immunity to tumors and intracellular infections,
but the type of T-helper response is important as well, especially for maintaining
immunologic memory (44). Th1 cells, which produce IFN- and TNF-, support CTL
responses, whereas Th2 cells (which produce IL-4, IL-5, and IL-13) support humoral
immunity and down-regulate Th1 responses. Th1 polarization is potently induced by
DC secretion of IL-12, although IL-12 is not an absolute requirement for this, per-
haps because of other DC-generated cytokines such as IL-23 or IL-27 (27,45). In
addition to activating T-cells, it is now known that DCs can also directly interact with
and activate B-cells (46,47), NK cells (48), and NKT-cells (49). All these cells may
play a critical role in the induction of antitumor immunity.
2. Ex Vivo Manipulation of DCs for Active Immunotherapy

Vaccines against microbial pathogens are traditionally prepared by isolating an
attenuated or killed version of the pathogen and mixing it with an adjuvant that serves
to boost the immune response. The success of this type of vaccine often depends on
its ability to stimulate the production of neutralizing antibodies (50). For a number of
chronic intracellular infections such as HIV, hepatitis C, tuberculosis, or malaria,
however, this approach has not proved sufficient to generate protective immunity. To
control these disorders, the generation of Th1 and CTL responses appears to be very
important (50).
The induction and maintenance of CTL-mediated immunity also appears to be of
great importance for immunotherapies for patients with cancer. (2,3,51). By studying
cellular immune responses in cancer patients, a variety of tumor-associated antigens
have been identified that are recognized by T-cells. These antigens can be used to
vaccinate individuals against their own tumors (5254). Attempts to vaccinate can-
cer patients with killed tumor cells, cell lysates, or tumor antigen proteins or peptides
have produced immunologic and clinical responses, some of them complete and long-
lasting (5560). However, it remains to be demonstrated in large, prospective, ran-
domized trials that these antigen/adjuvant preparations can provide a clinically
significant benefit in the form of improved survival (58,61,62).
Cancer vaccine adjuvants such as QS-21, GM-CSF, or Incomplete Freunds Adju-
vant (IFA, or Montanide) have all been shown to boost CTL responses to tumor-asso-
ciated antigens, but these responses are often weak and often require in vitro
restimulation of T-cells to be detected (57,63). Because of their potent
immunostimulatory capacity and their ability to prime nave T-cells, the manipulation
of DC as a natural vaccine adjuvant has generated great interest (9,50,6367). To
date, the most common approach has been to isolate large numbers of DC by culturing
bone marrow-derived progenitors ex vivo in the presence of cytokines, loading the DC
with antigens and re-injecting them back into the subject. In mice, bone marrow-
derived DC loaded ex vivo with tumor antigens by a variety of methods have been
shown to induce potent antigen-specific CTL responses. These vaccines can protect
mice from challenge with tumors bearing the antigenic peptide in association with
cell-surface MHC, and can cure animals bearing established tumors (6871).
In human subjects, vaccination of healthy volunteers with peptide-pulsed DC is
well tolerated and clearly induces antigen-specific T-cell immunity in the form of
CTL and Th1 responses (72,73). In addition, CTL responses and tumor regressions
have been noted in small clinical trials in cancer patients vaccinated with tumor-asso-
ciated antigens loaded by a variety of methods onto ex vivo generated DC. Clinical
and immune responses have been reported in patients with metastatic melanoma (74
77), metastatic renal cell carcinoma (78), B-cell lymphoma (79,80), prostate cancer
(8186), breast and ovarian cancer (87), and colon and lung cancer (88). Because
early results have been encouraging, larger trials are now underway to determine
whether DC vaccines can improve survival in cancer patients (89).
Three general methods have been described for the preparation of DC from human
subjects for use in clinical trials. The methods differ in the starting population of blood
cells used and in the different mixtures of cells obtained in the finished product. The
methods involve, respectively: (1) differentiating DC from nonproliferating monocyte
precursors; (2) differentiating DC from proliferating CD34+ hematopoietic progenitor
cells; or (3) directly isolating DC from peripheral blood. The optimal route of admin-
istration for the cells has not yet been determined. They are typically injected intrader-
mally, subcutaneously, or intravenously. Other routes such as direct injection into
lymph nodes or tumors have also been described.
2.1. Monocyte-Derived DC
By far the most popular method is to prepare DCs from blood monocytes. Mono-
cyte-derived DCs (MoDC) can be generated from peripheral blood mononuclear cells
(PBMC) obtained from whole blood, although to obtain yields sufficient for the pro-
duction of multiple vaccines, PBMC are usually obtained by leukapheresis (90). To
prepare MoDC, CD14+ monocytes are first selected from PBMC, using either a simple
plastic adherence step (monocytes adhere to plastic, whereas lymphocytes do not), or
by large-scale cell sorting using immunomagnetic beads. The monocytes are induced
to differentiate into immature CD14 CD83 DC by culturing for several days in the
presence of IL-4 and GM-CSF. The DC are then typically stimulated to mature by
culturing for an additional period of time (12 d) in the presence of proinflamma-
tory cytokines or soluble CD40L. Mature MoDC are large, CD14 CD83+ cells that
express high levels of MHC class I and II molecules and high levels of co-stimulatory
molecules such as CD40, CD80, and CD86. Peptide antigens may be loaded or
pulsed onto the DC either before or after maturation. Cell lysates or purified or
recombinant proteins are typically loaded just before the maturation stimulus. Viruses
or RNA may be added either before or after maturation.
2.2. DC Derived From CD34+ Hematopoietic Progenitor Cells
This method begins with the collection of CD34+ proliferating progenitors from the
peripheral blood (91,92). This requires mobilizing CD34+ progenitor cells from the bone
marrow by treating patients with granulocyte colony-stimulating factor (G-CSF) prior to
leukapheresis (63,76). The cells are expanded in culture for a week or more in the pres-
ence of GM-CSF and TNF-. The final product is a more complex mixture of cells than
MoDC. It includes a population similar to MoDC and another population phenotypically
identical to epidermal Langerhans cells. Differentiation may be skewed toward Langer-
hans cells by adding TGF- to the culture. A fairly large percentage of the final product
consists of myeloid cells at varying stages of differentiation, including some CD14+
cells. These do not function as antigen-presenting cells but do not appear to interfere
with the potency of the vaccine, although this has not been well studied. CD34+ progeni-
tor cell-derived DC may be matured and loaded with antigens similarly to Mo-DC.
2.3. DC Enriched From Peripheral Blood
In the first clinical trial to use DC-based immunotherapy for cancer, DC were puri-
fied directly from PBMC by a series of density gradient centrifugation steps (80). In
this procedure, the PBMC were first depleted of monocytes by centrifugation through
discontinuous Percoll gradients. The monocyte-depleted PBMC were then cultured
for 24 h in the presence of antigen, and then the DC were separated from lymphocytes
by sequential centrifugation through 15% and then 14% metrizamide gradients. The
low-density fraction containing the DC was then cultured overnight again in the pres-
ence of antigen, washed, and injected back into the patient (80). The final product
contained 5090% DC, and in subsequent studies it was found that although the DC
obtained at the time of leukapheresis had an immature phenotype, by the end of the
2-d procedure they express maturation markers such as CD80, CD86, CD83, and CCR7
(88). DC preparations based on this method are currently being used in phase III trials
for the immunotherapy of prostate cancer (89).
The yield of DC obtained by this procedure can be significantly enhanced by stimu-
lating the patient with FMS-like tyrosine kinase-3 ligand (Flt3 ligand, or FL) prior to
leukapheresis (88). FL mobilizes and expands peripheral blood DC, and increases the
final DC yield approx 20-fold. Taking advantage of the increased numbers of DC
obtained in the leukapheresis product following administration of FL, a modification
of the DC isolation procedure has been described that omits the initial Percoll gradient
and the 14% metrizamide gradient (only the 15% metrizamide gradient is used) (88).
Unfortunately, pharmaceutical-grade FL is no longer commercially available. Still,

the development of commercial closed systems that take advantage of cell-separation
technology based on immunomagnetic beads, and the short preparation time involved,
may make the enrichment of DC directly from blood an attractive option in the future.
2.4. Choice of Cell Type
All three types of DC preparation have been shown to stimulate proliferation and
cytokine production of antigen-specific T-cells, and have been associated with clini-
cal responses in trials involving human subjects with cancer. However, no direct com-
parisons of these cells have been performed in human subjects. In vitro data from one
group has indicated an advantage of CD34+ cell-derived DC over MoDC in the pre-
sentation of peptides (93), but MoDC have many advantages for clinical use. They are
considerably simpler to prepare and are a well characterized, homogenous population
of cells. In addition, patients do not need to be prestimulated with cytokines prior to
leukapheresis to prepare MoDC.
2.5. Ex Vivo Maturation of DC
Whatever the method used to prepare the DC, it is clearly important that the DCs be
matured prior to clinical use (43). Clinical studies with DC-based vaccines indicate
that DCs that have been matured ex vivo more effectively stimulate T-cell responses
(94), and there is accumulating evidence that the use of antigen-loaded immature DC
in vaccines actually leads to immune tolerance, perhaps due to anergy or T-regulatory
cell mechanisms (20,21,95).
In early work on DC-based vaccines, DC were matured using supernatants from
cultured monocytes (monocyte-conditioned medium, or MCM) as a source of pro-
inflammatory cytokines (90,96,97). Currently, the most popular method is to use a
cocktail of three cytokines (IL-1-, IL-6, and TNF-) plus prostaglandin E2 (PGE2),
originally described by Jonuleit et al., which mimics the effect of MCM on DC matu-
ration (98). One concern with this method is that DC matured in the presence of PGE2
do not secrete IL-12 (99). However, the addition of PGE2 is important for inducing
DC migratory ability (100), and DC matured with this cocktail still strongly stimulate
CTL responses both in vitro and in vivo. This may be due to the action of other IL-12-
related cytokines, such as IL-23 and IL-27 (45).
2.6. Loading DCs With Antigens Ex Vivo: Peptide Antigens
There are many ways to load DC with antigens. One of the more commonly used
methods is to co-culture them directly with peptides (72,7477). This has been made
possible because of progress in identifying immunodominant peptide epitopes for
tumor-associated and microbial antigens that are recognized by T-cells (52,54). Use
of peptides requires knowledge of a patients HLA type and the existence of relevant
MHC-restricted peptide epitopes. HLA-A2-restricted peptides are commonly used,
often in combinations, although other MHC class I-restricted epitopes and even MHC
class II-restricted T-helper epitopes may be used as well. Altered or enhanced pep-
tide antigens have also been used to boost immunity to less immunogenic self anti-
gens (50,88). Peptides from microbial antigens such as influenza matrix protein are
frequently included to test for recall immune responses. Peptides may be loaded onto
DC either before or after DC maturation, since short peptides can apparently bind
directly to MHC molecules on the DC surface without having to be processed within
the cell (9). In our hands, pulsing peptides onto mature DC has resulted in somewhat
better specific T-cell stimulation in vitro.
The optimal peptide dose with which to load DC is not entirely clear. Intuitively
it might be thought that increasing the concentration of peptide would result in better
antigen loading and more potent stimulation of T-cells. The opposite, however,
appears to be true. Alexander-Miller et al. showed that APC loaded with very low
concentrations of peptide (as little as 0.1 nM) stimulate high-avidity T-cell clones
that much more effectively recognize and lyse antigen-expressing target cells (50,
101). For clinical use it will be very important to load DC with a peptide dose that
stimulates expansion of high-avidity T-cell clones. Most protocols currently pulse
DC with peptide concentrations of 1 to 10 M, although the optimal dose will most
likely need to be determined through dose escalation studies in human subjects.
2.7. Proteins and Cell Lysates
One disadvantage of using peptide-pulsed DC is that the dominant epitopes of the
antigen of interest must be known for the HLA type of the patient. This often restricts
peptide vaccination studies to individuals with common HLA types. Alternatively,
DC can be loaded prior to maturation with purified or recombinant proteins (79), or
even with whole tumor cell lysates (74,102,103). This method allows host HLA mol-
ecules to select epitopes from an antigens entire sequence, and in the case of loading
with tumor lysates permits vaccination with the complete antigenic content of the
tumor.
The immunogenicity of protein-loaded DC can be enhanced by using proteins
coupled to cytokine or carrier-protein sequences (79,104), or using proteins of xeno-
geneic origin (83). Keyhole limpet hemocyanin (KLH), a powerful immunogen
derived from a marine mollusk, is one such protein that is commonly used to
nonspecifically boost the immunogenicity of a vaccine. It is also added as a control to
test for the ability of a vaccine to prime T-helper responses to a neoantigen (6,32).
One disadvantage of loading DC with purified protein antigens is that the MHC
class I pathway is not specifically targeted. In practice, however, pulsing DC with
crude protein mixtures such as tumor cell lysates seems to produce vaccines that
generate clinical responses (74,102,103). This may be because of the ability of DC
to cross-present exogenous antigens onto MHC class I.
2.8. DNA, RNA, and Viruses
Loading DC with antigens by DNA transfection has been successfully performed
(105,106); however, it is often associated with a significant amount of cell death. DC
tolerate RNA transfection well, however, and this has been used to successfully load DC
with RNA encoding specific antigens or even whole tumor RNA (68,81,107109). Typi-
cally, DC are transfected with RNA by co-culture with naked RNA or by RNA electro-
poration prior to maturation. DC can also be loaded with antigens by infecting with
nonreplicating viral vectors such as recombinant adenovirus or pox viruses (110114).
As with loading DC with proteins or RNA, this method allows vaccines to be gener-
ated for patients of any HLA type, since the encoded proteins are cleaved and pro-
cessed onto MHC molecules within the host cell.
2.9. Apoptotic Cells, Immune Complexes, and Other Methods
Apoptotic tumor cells can be used as an antigen source to take advantage of cross-
presentation of extracellular antigens onto MHC class I (34,115,116). Cross-presenta-
tion of specific protein antigens may also be targeted by loading DC with IgG immune
complexes of those antigens, which are taken up via Fc receptors on the DC (17,18,36).
A number of other methods have also been employed to manipulate DCs ex vivo to
create cellular or cell-derived vaccines. Cell fusions between DC and tumor cells cre-
ate heterokaryons that can stimulate anti-tumor T-cell responses in vitro, in laboratory
animals, and in humans with cancer (78,117,118). DC can also be used as a source of
exosomes, antigen-presenting vesicles which can be loaded with antigens and used to
stimulate T-cell immunity (119121).
3. Treatment Strategies for the Use of Dendritic Cell Vaccines
It is clearly established that antigen-loaded DC generated by a number of different
methods can induce antigen-specific immunity in the form of CTL and Th1 responses
in human subjects. Numerous technical variables need testing before this approach is
optimized, however, and large controlled studies are needed to prove efficacy in the
treatment of cancer or serious chronic infections.
Most of the reports on the therapeutic use of DC vaccines published to date have
been on patients with advanced cancer. Although occasional dramatic clinical responses
have been noted, it is likely that DC immunizations may be most effective in the adju-
vant setting when the patient is in remission or the tumor burden is low but risk of
disease progression and death is high. This needs to be investigated in controlled clini-
cal trials. In addition, the possibility of using DC vaccines in combination with novel
therapies such as anti-angiogenesis agents (122), antibodies that target tumor antigens,
or immunostimulatory agents such as anti-CTLA-4 (123125) has not been addressed.
Such synergistic approaches could significantly enhance the vaccines anti-tumor effect
and should be tested in clinical trials.
One significant drawback of DC vaccines is that the ex vivo production of individu-
ally tailored cellular therapies is both laborious and expensive. For this reason the use
of novel in situ approaches that take advantage of the biological properties of DC in
vivo has generated great interest. Approaches that can mobilize DC to an accessible site
where they can be matured and pulsed with antigens in vivo are being developed that
may hopefully lead to potent therapies that do not require expensive facilities or labor-
intensive cell processing (9). Such approaches include the systemic mobilization of DC
using Flt3 ligand, the local injection of chemokines such as MIP-3, the use of DNA
vaccines containing bacterial CpG motifs (which can target DC and B cells via TLR9),
or the use of topical compounds such as Imiquimod (a TLR7 ligand) in conjunction
with peptide vaccines (16,50,126,127). As more is discovered about the biological prop-
erties of DC, we should be better able to manipulate these important cells to take advan-
tage of their potent immunoregulatory function for the treatment of human disease.
References
1.
1 Palucka, K. and Banchereau, J. (1999) Linking innate and adaptive immunity. Nat. Med.
5, 868870.
2.
2 Dunn, G. P., Bruce, A. T., Ikeda, H., Old, L. J., and Schreiber, R. D. (2002) Cancer
immunoediting: from immunosurveillance to tumor escape. Nat. Immunol. 3, 991998.
3.
3 Smyth, M. J., Godfrey, D. I., and Trapani, J.A. (2001) A fresh look at tumor
immunosurveillance and immunotherapy. Nat. Immunol. 2, 293299.
4.
4 Pardoll, D. (2003) Does the immune system see tumors as foreign or self? Annu. Rev.
Immunol. 21, 807839.
5.
5 Sant, A. and Yewdell, J. (2003) Antigen processing and recognition. Curr. Opin. Immunol.
15, 6668.
6.
6 Banchereau, J., Briere, F., Caux, C., et al. (2000) Immunobiology of dendritic cells. Annu.
Rev. Immunol. 18, 767811.
7.
7 Bhardwaj, N., Friedman, S. M., Cole, B. C., and Nisanian, A. J. (1992) Dendritic cells are
potent antigen-presenting cells for microbial superantigens. J. Exp. Med. 175, 267273.
8.
8 Fonteneau, J. F., Larsson, M., Somersan, S., et al. (2001) Generation of high quantities of
viral and tumor-specific human CD4+ and CD8+ T-cell clones using peptide pulsed mature
dendritic cells. J. Immunol. Methods. 258, 111126.
9.
9 Steinman, R. M. and Pope, M. (2002) Exploiting dendritic cells to improve vaccine effi-
cacy. J. Clin. Invest. 109, 15191526.
10.
10 Shortman, K. and Liu, Y. J. (2002) Mouse and human dendritic cell subtypes. Nat. Rev.
Immunol. 2, 151161.
11. MacDonald, K.P., Munster, D.J., Clark, G.J., Dzionek, A., Schmitz, J., and Hart, D.N.
(2002) Characterization of human blood dendritic cell subsets. Blood 15, 15.
12.
12 Manz, M. G., Traver, D., Miyamoto, T., Weissman, I. L., and Akashi, K. (2001) Dendritic
cell potentials of early lymphoid and myeloid progenitors. Blood 97, 33333341.
13.
13 Wu, L., DAmico, A., Hochrein, H., OKeeffe, M., Shortman, K., and Lucas, K. (2001)
Development of thymic and splenic dendritic cell populations from different hemopoietic
precursors. Blood 98, 33763382.
14.
14 del Hoyo, G. M., Martin, P., Vargas, H. H., Ruiz, S., Arias, C. F., and Ardavin, C.
(2002) Characterization of a common precursor population for dendritic cells. Nature
415, 10431047.
15.
15 Medzhitov, R. (2001) Toll-like receptors and innate immunity. Nat. Rev. Immunol. 1,
135145.
16.
16 Takeda, K., Kaisho, T., and Akira, A. (2003) Toll-like receptors. Annu. Rev. Immunol. 21,
335376.
17.
17 Rafiq, K., Bergtold, A., and Clynes, R. (2002) Immune complex-mediated antigen presen-
tation induces tumor immunity. J. Clin. Invest. 110, 7179.
18.
18 Schuurhuis, D. H., Ioan-Facsinay, A., Nagelkerken, B., et al. (2002) Antigen-antibody
immune complexes empower dendritic cells to efficiently prime specific CD8+ CTL
responses in vivo. J. Immunol. 168, 22402246.
19 Figdor, C. G., van Kooyk, Y., and Adema, G. J. (2002) C-type lectin receptors on den-
19.
dritic cells and Langerhans cells. Nat. Rev. Immunol. 2, 7784.
20.
20 Steinman, R. M., and Nussenzweig, M. C. (2002) Avoiding horror autotoxicus: the impor-
tance of dendritic cells in peripheral T cell tolerance. Proc. Natl. Acad. Sci. USA 99,
351358.
21.
21 Steinman, R. M., Hawiger, D., and Nussenzweig, M. C. (2003) Tolerogenic dendritic cells.
Annu. Rev. Immunol. 21, 685711.
22.
22 Hawiger, D., Inaba, K., Dorsett, Y., et al. (2001) Dendritic cells induce peripheral T cell
unresponsiveness under steady state conditions in vivo. J. Exp. Med. 194, 769779.
23.
23 Aderem, A. and Ulevitch, R. J. (2000) Toll-like receptors in the induction of the innate
immune response. Nature 406, 782787.
24.
24 Gallucci, S. and Matzinger, P. (2001) Danger signals: SOS to the immune system. Curr.
Opin. Immunol. 13, 114119.
25.
25 Pulendran, B., Palucka, K., and Banchereau, J. (2001) Sensing pathogens and tuning
immune responses. Science 293, 253256.
26.
26 Mellman, I. and Steinman, R. M. (2001) Dendritic cells: specialized and regulated antigen
processing machines. Cell 106, 255258.
27.
27 Lanzavecchia, A. and Sallusto, F. (2001) The instructive role of dendritic cells on T cell
responses: lineages, plasticity and kinetics. Curr. Opin. Immunol. 13, 291289.
28.
28 Moody, D. B. and Porcelli, S. A. (2002) Intracellular pathways of CD1 antigen presenta-
tion. Nature Rev. Immunol. 3, 1122.
29.
29 Porcelli, S. A. and Modlin, R. L. (1999) The CD1 system: antigen presenting molecules
for T cell recognition of lipids and glycolipids. Annu. Rev. Immunol. 17, 297329.
30.
30 Fujii, S., Shimizu, K., Kronenberg, M., and Steinman, R. M. (2002) Prolonged IFN-
gamma-producing NKT response induced with alpha-galactosylceramide-loaded DCs.
Nat. Immunol. 3, 867874.
31.
31 Kronenberg, M. and Gapin, L. (2002) The unconventional lifestyle of NKT cells. Nature
32 Bhardwaj, N. (2001) Processing and presentation of antigens by dendritic cells: implica-
32.
tions for vaccines. Trends Mol. Med. 7, 388394.
33.
33 Heath, W. R. and Carbone, F. R. (1999) Cytotoxic T lymphocyte activation by cross-
priming. Curr. Opin. Immunol. 11, 314318.
34.
34 Albert, M. L., Sauter, B., and Bhardwaj, N. (1998) Dendritic cells acquire antigen from
apoptotic cells and induce class I- restricted CTLs. Nature 392, 8689.
35.
35 Heath, W. R. and Carbone, F. R. (2001) Cross-presentation in viral immunity and self-
tolerance. Nat. Rev. Immunol. 1, 126134.
36.
36 Kalergis, A. M. and Ravetch, J. V. (2002) Inducing tumor immunity through the selec-
tive engagement of activating Fcgamma receptors on dendritic cells. J. Exp. Med. 195,
16531659.
37.
37 Savill, J., Dransfield, I., Gregory, C., and Haslett, C. (2002) A blast from the past: clear-
ance of apoptotic cells regulates immune responses. Nat. Rev. Immunol. 2, 965975.
38.
38 Stoll, S., Delon, J., Brotz, T. M., and Germain, R. N. (2002) Dynamic imaging of T cell
dendritic cell interactions in lymph nodes. Science 296, 18731876.
39.
39 Norbury, C. C., Malide, D., Gibbs, J. S., Bennink, J. R., and Yewdell, J. W. (2002) Visu-
alizing priming of virus-specific CD8+ T cells by infected dendritic cells in vivo. Nat.
Immunol. 3, 265271.
40.
40 Langenkamp, A., Casorati, G., Garavaglia, C., Dellabona, P., Lanzavecchia, A., and Sallusto,
F. (2002) T cell priming by dendritic cells: thresholds for proliferation, differentiation and
death and intraclonal functional diversification. Eur. J. Immunol. 32, 20462054.
41.
41 Palucka, K. and Banchereau, J. (2002) How dendritic cells and microbes interact to elicit
or subvert protective immune responses. Curr. Opin. Immunol. 14, 420431.
42.
42 Boonstra, A., Asselin-Paturel, C., Gilliet, M., et al. (2003) Flexibility of mouse classical
and plasmacytoid-derived dendritic cells in directing T helper type 1 and 2 cell develop-
ment: dependency on antigen dose and differential toll-like receptor ligation. J. Exp. Med.
197, 101109.
43.
43 Lutz, M. and Schuler, G. (2002) Immature, semi-mature and fully mature dendritic cells:
which signals induce tolerance or immunity? Trends Immunol. 23, 445.
44.
44 Kaech, S. M. and Ahmed, R. (2003) CD8 T cells remember with a little help. Science 300,
263265.
45.
45 Trinchieri, G. (2003) Interleukin-12 and the regulation of innate resistance and adaptive
immunity. Nat. Rev. Immunol. 3, 133146.
46.
46 Balazs, M., Martin, F., Zhou, T., and Kearney, J. (2002) Blood dendritic cells interact with
splenic marginal zone B cells to initiate T-independent immune responses. Immunity 17,
341352.
47.
47 Jego, G., Palucka, A. K., Blanck, J. P., Chalouni, C., Pascual, V., and Banchereau, J.
(2003) Plasmacytoid dendritic cells induce plasma cell differentiation through type I
interferon and interleukin 6. Immunity 19, 225234.
48.
48 Ferlazzo, G., Tsang, M. L., Moretta, L., Melioli, G., Steinman, R. M., and Munz, C. (2002)
Human dendritic cells activate resting natural killer (NK) cells and are recognized via the
NKp30 receptor by activated nk cells. J. Exp. Med. 195, 343351.
49.
49 Fujii, S., Shimizu, K., Kronenberg, M., and Steinman, R. M. (2002) Prolonged IFN-
gamma-producing NKT response induced with alpha-galactosylceramide-loaded DCs.
Nat. Immunol. 3, 867874.
50.
50 Berzofsky, J. A., Ahlers, J. D., and Belyakov, I. M. (2001) Strategies for designing and
optimizing new generation vaccines. Nat. Rev. Immunol. 1, 209219.
51 Shankaran, V., Ikeda, H., Bruce, A. T., et al. (2001) IFNgamma and lymphocytes pre-
51.
vent primary tumour development and shape tumour immunogenicity. Nature 410,
11071111.
52.
52 Boon, T. and Old, L.J. (1997) Cancer tumor antigens. Curr. Opin. Immunol. 9, 681683.
53.
53 Houghton, A. N., Gold, J. S., and Blachere, N. E. (2001) Immunity against cancer: lessons
learned from melanoma. Curr. Opin. Immunol. 13, 134140.
54.
54 Stevanovic, S. (2002) Identification of tumour-associated T-cell epitopes for vaccine
development. Nat. Rev. Cancer 2, 514520.
55.
55 Bendandi, M., Gocke, C. D., Kobrin, C. B., et al. (1999) Complete molecular remissions
induced by patient-specific vaccination plus granulocyte-monocyte colony-stimulating
factor against lymphoma. Nat. Med. 5, 11711177.
56 Livingston, P. (2001) The unfulfilled promise of melanoma vaccines. Clin. Cancer Res. 7,
56.
18371838.
57.
57 Nestle, F. O. (2002) Vaccines and melanoma. Clin. Exp. Dermatol. 27, 597601.
58.
58 Machiels, J. P., van Baren, N., and Marchand, M. (2002) Peptide-based cancer vaccines.
Semin. Oncol. 29, 494502.
59.
59 Marchand, M., van Baren, N., Weynants, P., et al. (1999) Tumor regressions observed in
patients with metastatic melanoma treated with an antigenic peptide encoded by gene
MAGE-3 and presented by HLA A1. Int J. Cancer 80, 219230.
60.
60 Rosenberg, S. A., Yang, J. C., Schwartzentruber, D. J., et al. (1998) Immunologic and
therapeutic evaluation of a synthetic peptide vaccine for the treatment of patients with
metastatic melanoma. Nat. Med. 4, 321327.
61.
61 Sabel, M. S., and Sondak, V. K. (2002) Tumor vaccines: a role in preventing recurrence in
melanoma? Am. J. Clin. Dermatol. 3, 609616.
62.
62 McClay, E. F. (2002) Adjuvant therapy for patients with high-risk malignant melanoma.
Semin. Oncol. 29, 389399.
63.
63 Banchereau, J., Schuler-Thurner, B., Palucka, A. K., and Schuler, G. (2001) Dendritic
cells as vectors for therapy. Cell 106, 271274.
64.
64 Brossart, P., Wirths, S., Brugger, W., and Kanz, L. (2001) Dendritic cells in cancer vac-
cines. Exp. Hematol. 29, 12471255.
65.
65 Fong, L. and Engleman, E. G. (2000) Dendritic cells in cancer immunotherapy. Annu. Rev.
Immunol. 18, 245273.
66.
66 Jefford, M., Maraskovsky, E., Cebon, J., and Davis, I. D. (2001) The use of dendritic cells
in cancer therapy. Lancet Oncol. 2, 343353.
67.
67 Nestle, F. O., Banchereau, J., and Hart, D. (2001) Dendritic cells: on the move from bench
to bedside. Nat. Med. 7, 761765.
68.
68 Ashley, D. M., Faiola, B., Nair, S., Hale, L. P., Bigner, D. D., and Gilboa, E. (1997) Bone
marrow-generated dendritic cells pulsed with tumor extracts or tumor RNA induce antitu-
mor immunity against central nervous system tumors. J. Exp. Med. 186, 11771182.
69.
69 Celluzzi, C. M. and Falo, L. D., Jr. (1998) Physical interaction between dendritic cells and
tumor cells results in an immunogen that induces protective and therapeutic tumor rejec-
tion. J. Immunol. 160, 30813085.
70.
70 Gilboa, E., Nair, S. K., and Lyerly, H. K. (1998) Immunotherapy of cancer with dendritic-
cell-based vaccines. Cancer Immunol. Immunother. 46, 8287.
71.
71 Porgador, A., Snyder, D., and Gilboa, E. (1996) Induction of antitumor immunity using
bone marrow-generated dendritic cells. J. Immunol. 156, 29182926.
72.
72 Dhodapkar, M. V., Steinman, R. M., Sapp, M., et al. (1999) Rapid generation of broad T-
cell immunity in humans after a single injection of mature dendritic cells. J. Clin. Invest.
104, 173180.
73.
73 Dhodapkar, M. V., Krasovsky, J., Steinman, R. M., and Bhardwaj, N. (2000) Mature den-
dritic cells boost functionally superior CD8(+) T-cell in humans without foreign helper
epitopes. J. Clin. Invest. 105, R9R14.
74.
74 Nestle, F. O., Alijagic, S., Gilliet, M., et al. (1998) Vaccination of melanoma patients with
peptide- or tumor lysate-pulsed dendritic cells. Nat. Med. 4, 328332.
75.
75 Thurner, B., Haendle, I., Roder, C., et al. (1999) Vaccination with mage-3A1 peptide-
pulsed mature, monocyte-derived dendritic cells expands specific cytotoxic T cells and
induces regression of some metastases in advanced stage iv melanoma. J. Exp. Med. 190,
16691678.
76 Banchereau, J., Palucka, A. K., Dhodapkar, M., et al. (2001) Immune and clinical responses
76.
in patients with metastatic melanoma to CD34(+) progenitor-derived dendritic cell vaccine.
Cancer Res. 61, 64516458.
77.
77 Schuler-Thurner, B., Schultz, E. S., Berger, T. G., et al. (2002) Rapid induction of tumor-
specific type 1 T helper cells in metastatic melanoma patients by vaccination with mature,
cryopreserved, peptide-loaded monocyte-derived dendritic cells. J. Exp. Med. 195, 1279
1288.
78.
78 Kugler, A., Stuhler, G., Walden, P., et al. (2000) Regression of human metastatic renal
cell carcinoma after vaccination with tumor celldendritic cell hybrids. Nat. Med. 6,
332336.
79 Timmerman, J. M., Czerwinski, D. K., Davis, T. A., et al. (2002) Idiotype-pulsed den-
79.
dritic cell vaccination for B-cell lymphoma: clinical and immune responses in 35 patients.
Blood 99, 15171526.
80.
80 Hsu, F. J., Benike, C., Fagnoni, F., et al. (1996) Vaccination of patients with B-cell lym-
phoma using autologous antigen-pulsed dendritic cells. Nat. Med. 2, 5258.
81.
81 Heiser, A., Coleman, D., Dannull, J., et al. (2002) Autologous dendritic cells transfected
with prostate-specific antigen RNA stimulate CTL responses against metastatic prostate
tumors. J. Clin. Invest. 109, 409417.
82.
82 Fong, L., Brockstedt, D., Benike, C., Wu, L., and Engleman, E. G. (2001) Dendritic cells
injected via different routes induce immunity in cancer patients. J. Immunol. 166, 4254
4259.
83.
83 Fong, L., Brockstedt, D., Benike, C., et al. (2001) Dendritic cell-based xenoantigen vacci-
nation for prostate cancer immunotherapy. J. Immunol. 167, 71507156.
84.
84 Small, E. J., Fratesi, P., Reese, D. M., et al. (2000) Immunotherapy of hormone-refractory
prostate cancer with antigen-loaded dendritic cells. J. Clin. Oncol. 18, 38943903.
85.
85 Burch, P. A., Breen, J. K., Buckner, J. C., et al. (2000) Priming tissue-specific cellular
immunity in a phase I trial of autologous dendritic cells for prostate cancer. Clin. Cancer
Res. 6, 21752182.
86.
86 Tjoa, B. A., Simmons, S. J., Bowes, V. A., et al. (1998) Evaluation of phase I/II clinical
trials in prostate cancer with dendritic cells and PSMA peptides. Prostate 36, 3944.
87.
87 Brossart, P., Wirths, S., Stuhler, G., Reichardt, V. L., Kanz, L., and Brugger, W. (2000)
Induction of cytotoxic T-lymphocyte responses in vivo after vaccinations with peptide-
pulsed dendritic cells. Blood 96, 31023108.
88.
88 Fong, L., Hou, Y., Rivas, A., et al. (2001) Altered peptide ligand vaccination with Flt3
ligand expanded dendritic cells for tumor immunotherapy. Proc. Natl. Acad. Sci. USA 98,
88098814.
89.
89 Valone, F. H., Small, E., MacKenzie, M., et al. (2001) Dendritic cell-based treatment of
cancer: closing in on a cellular therapy. Cancer J. 7(Suppl. 2), S53S61.
90 Thurner, B., Roder, C., Dieckmann, D., et al. (1999) Generation of large numbers of fully
90.
mature and stable dendritic cells from leukapheresis products for clinical application. J.
Immunol. Meth. 223, 115.
91.
91 Caux, C., Massacrier, C., Dezutter-Dambuyant, C., et al. (1995) Human dendritic Langer-
hans cells generated in vitro from CD34+ progenitors can prime naive CD4+ T cells and
process soluble antigen. J. Immunol. 155, 54275435.
92.
92 Caux, C., Dezutter-Dambuyant, C., Schmitt, D., and Banchereau, J. (1992) GM-CSF and
TNF-alpha cooperate in the generation of dendritic Langerhans cells. Nature 360, 258261.
93.
93 Mortarini, R., Anichini, A., Di Nicola, M., et al. (1997) Autologous dendritic cells derived
from CD34+ progenitors and from monocytes are not functionally equivalent antigen-pre-
senting cells in the induction of melan-A/Mart-1(2735)-specific CTLs from peripheral
blood lymphocytes of melanoma patients with low frequency of CTL precursors. Cancer
Res. 57, 55345541.
94.
94 Jonuleit, H., Giesecke-Tuettenberg, A., Tuting, T., et al. (2001) A comparison of two types
of dendritic cell as adjuvants for the induction of melanoma-specific T-cell responses in
humans following intranodal injection. Int. J. Cancer 93, 243251.
95.
95 Dhodapkar, M. V., Steinman, R. M., Krasovsky, J., Munz, C., and Bhardwaj, N. (2001)
Antigen-specific inhibition of effector T cell function in humans after injection of imma-
ture dendritic cells. J. Exp. Med. 193, 233238.
96.
96 Bender, A., Sapp, M., Schuler, G., Steinman, R. M., and Bhardwaj, N. (1996) Improved
methods for the generation of dendritic cells from nonproliferating progenitors in human
blood. J. Immunol. Meth. 196, 121135.
97.
97 Romani, N., Reider, D., Heuer, M., et al. (1996) Generation of mature dendritic cells from
human blood. An improved method with special regard to clinical applicability. J.
98.
98 Jonuleit, H., Kuhn, U., Muller, G., et al. (1997) Pro-inflammatory cytokines and prostag-
landins induce maturation of potent immunostimulatory dendritic cells under fetal calf
serum-free conditions. Eur. J. Immunol. 27, 31353142.
99.
99 Lee, A.W., Truong, T., Bickham, K., et al. (2002) A clinical grade cocktail of cytokines
and PGE(2) results in uniform maturation of human monocyte-derived dendritic cells:
implications for immunotherapy. Vaccine 20(Suppl. 4), A8A22.
100.
100 Scandella, E., Men, Y., Gillessen, S., Forster, R., and Groettrup, M. (2002) Prostaglandin
E2 is a key factor for CCR7 surface expression and migration of monocyte-derived den-
dritic cells. Blood 100, 13541361.
101.
101 Alexander-Miller, M. A., Leggatt, G. R., and Berzofsky, J. A. (1996) Selective expansion
of high- or low-avidity cytotoxic T lymphocytes and efficacy for adoptive immunotherapy.
Proc. Natl. Acad. Sci. USA 93, 41024107.
102.
102 Geiger, J., Hutchinson, R., Hohenkirk, L., McKenna, E., Chang, A., and Mule, J. (2000)
Treatment of solid tumours in children with tumour-lysate-pulsed dendritic cells. Lancet
356, 11631165.
103.
103 Chang, A. E., Redman, B. G., Whitfield, J. R., et al. (2002) A phase I trial of tumor lysate-
pulsed dendritic cells in the treatment of advanced cancer. Clin. Cancer Res. 8, 10211032.
104.
104 Kim, D. T., Mitchell, D. J., Brockstedt, D. G., et al. (1997) Introduction of soluble proteins
into the MHC class I pathway by conjugation to an HIV tat peptide. J. Immunol. 159,
16661668.
105.
105 Tuting, T., Wilson, C. C., Martin, D. M., et al. (1998) Autologous human monocyte-
derived dendritic cells genetically modified to express melanoma antigens elicit pri-
mary cytotoxic T cell responses in vitro: enhancement by cotransfection of genes
encoding the Th1 biasing cytokines IL-12 and IFN-alpha. J. Immunol. 160, 11391147.
106.
106 Yang, S., Vervaert, C. E., Burch, J., Jr., Grichnik, J., Seigler, H. F., and Darrow, T. L.
(1999) Murine dendritic cells transfected with human gp100 elicit both antigen-specific
CD8(+) and CD4(+) T-cell responses and are more effective than DNA vaccines at gener-
ating anti-tumor immunity. Int. J. Cancer 83, 532540.
107.
107 Boczkowski, D., Nair, S. K., Snyder, D., and Gilboa, E. (1996) Dendritic cells pulsed with
RNA are potent antigen-presenting cells in vitro and in vivo. J. Exp. Med. 184, 465472.
108.
108 Nair, S. K., Heiser, A., Boczkowski, D., et al. (2000) Induction of cytotoxic T cell responses
and tumor immunity against unrelated tumors using telomerase reverse transcriptase RNA
transfected dendritic cells. Nat. Med. 6, 10111017.
109 Sullenger, B. A. and Gilboa, E. (2002) Emerging clinical applications of RNA. Nature
109.
418, 252258.
110.
110 Jenne, L., Schuler, G., and Steinkasserer, A. (2001) Viral vectors for dendritic cell-based
immunotherapy. Trends Immunol. 22, 102107.
111.
111 Dietz, A. B., and Vuk-Pavlovic, S. (1998) High efficiency adenovirus-mediated gene trans-
fer to human dendritic cells. Blood 91, 392398.
112.
112 Chaux, P., Luiten, R., Demotte, N., et al. (1999) Identification of five MAGE-A1 epitopes
recognized by cytolytic T lymphocytes obtained by in vitro stimulation with dendritic
cells transduced with MAGE-A1. J. Immunol. 163, 29282936.
113.
113 Larsson, M., Fonteneau, J.F., Somersan, S., et al. (2001) Efficiency of cross presenta-
tion of vaccinia virus-derived antigens by human dendritic cells. Eur. J. Immunol. 31,
34323442.
114.
114 Engelmayer, J., Larsson, M., Lee, A., et al. (2001) Mature dendritic cells infected with
canarypox virus elicit strong anti-human immunodeficiency virus CD8+ and CD4+ T-cell
responses from chronically infected individuals. J. Virol. 75, 21422153.
115.
115 Kotera, Y., Shimizu, K., and Mule, J. J. (2001) Comparative analysis of necrotic and apop-
totic tumor cells as a source of antigen(s) in dendritic cell-based immunization. Cancer
Res. 61, 81058109.
116.
116 Berard, F., Blanco, P., Davoust, J., et al. (2000) Cross-priming of naive CD8 T cells against
melanoma antigens using dendritic cells loaded with killed allogeneic melanoma cells. J.
Exp. Med. 192, 15351544.
117.
117 Gong, J., Avigan, D., Chen, D., et al. (2000) Activation of antitumor cytotoxic T lympho-
cytes by fusions of human dendritic cells and breast carcinoma cells. Proc. Natl. Acad.
Sci. USA 97, 27152718.
118.
118 Gong, J., Koido, S., Chen, D., et al. (2002) Immunization against murine multiple
myeloma with fusions of dendritic and plasmacytoma cells is potentiated by interleukin
12. Blood 99, 25122517.
119.
119 Zitvogel, L., Regnault, A., Lozier, A., et al. (1998) Eradication of established murine
tumors using a novel cell-free vaccine: dendritic cell-derived exosomes. Nat. Med. 4,
594600.
120.
120 Wolfers, J., Lozier, A., Raposo, G., et al. (2001) Tumor-derived exosomes are a source of
shared tumor rejection antigens for CTL cross-priming. Nat. Med. 7, 297303.
121.
121 Schartz, N. E., Chaput, N., Andre, F., and Zitvogel, L. (2002) From the antigen-presenting
cell to the antigen-presenting vesicle: the exosomes. Curr. Opin. Molec. Ther. 4, 372381.
122.
122 Kerbel, R., and Folkman, J. (2002) Clinical translation of angiogenesis inhibitors. Nat.
Rev. Cancer 2, 727739.
123.
123 Egen, J. G., Kuhns, M. S., and Allison, J. P. (2002) CTLA-4: new insights into its biologi-
cal function and use in tumor immunotherapy. Nat. Immunol. 3, 611618.
124.
124 Hodi, F. S., Mihm, M. C., Soiffer, R. J., et al. (2003) Biologic activity of cytotoxic T
lymphocyte-associated antigen 4 antibody blockade in previously vaccinated metastatic
melanoma and ovarian carcinoma patients. Proc. Natl. Acad. Sci. USA 100, 47124717.
125.
125 Phan, G. Q., Yang, J. C., Sherry, R. M., et al. (2003) Cancer regression and autoimmunity
induced by cytotoxic T lymphocyte-associated antigen 4 blockade in patients with meta-
static melanoma. Proc. Natl. Acad. Sci. USA 100, 83728377.
126. Homey, B., Muller, A., and Zlotnik, A. (2002) Chemokines: agents for the immunotherapy
of cancer? Nat. Rev. Immunol. 2, 175184.
127. Schwarz, K., Storni, T., Manolova, V., et al. (2003) Role of toll-like receptors in
costimulating cytotoxic T cell responses. Eur. J. Immunol. 33, 14651470.
Mathematical Approach to DC Immunotherapy 19
2
A Mathematical Approach for Optimizing
Dendritic Cell-Based Immunotherapy
Gennady Bocharov, Neville J. Ford, and Burkhard Ludewig
Summary
Adoptive dendritic cell (DC)-based immunotherapy represents a promising approach to
overcome peripheral tolerance against autologous tumor antigens and to maintain protective
antitumor immunity. The translation of successful preclinical studies, however, appears to be
hampered by new complexities associated with the clinical situation. Mathematical modeling
provides the means for qualitative and quantitative analysis, predictions for complex dynamic
systems in immunology, and for the design and improvement of therapeutic approaches. We
present here a workable computational methodology for developing meaningful data- and hypo-
thesis-driven mathematical models for DC-based immunotherapy with a particular focus on
numerical parameter estimation and sensitivity analysis.
Key Words: Mathematical model; immune response; computer simulation; data fitting;
parameter estimation; sensitivity analysis; predatorprey dynamics; dendritic cell; cytotoxic
T-lymphocyte; numerical software.
1. Introduction
Immunotherapeutic approaches based on adoptive transfer of dendritic cells (DC)
expressing relevant antigens may be used for active mobilization of cellular immune
responses (cytotoxic T-lymphocytes [CTL], T-helper cells, and natural killer [NK]
cells) against tumors (1,2). The efficacy of this active immunization depends on the
complex biology of the DC life cycle and their interaction with T-cells. The kinetics of
this interaction and its sensitivity to relevant parameters are still incompletely under-
stood. These parameters include antigen loading, DC maturation stage, frequency and
route of DC injection, frequency and activation status of T-cells, and the homing rate
of DC to and their persistence within lymphoid tissues. Mathematics provides the
means for an integrative description of simplified models of various immunological
phenomena, including those arising in the context of adoptive immunotherapy. Its
primary role is to assist in parameter estimation (e.g., the cellular interaction kinetic
rates, life spans, delays, activation thresholds) and analysis (the numbers game),
and to make testable predictions in advance of experiments and clinical applications.
19
20 Bocharov, Ford, and Ludewig
Table 1
Decision Making Before and During the Modeling Process a
Modeling objectives
Determine the focus of the investigation.
General understanding of the rules underlying the observed behavior (e.g., causality
inferences and/or critical parameters of system dynamics)
Reliable calculations of system dynamics
Specific predictions about new observationsexperimental design
Characterize and analyze the data
Modeling approaches
Decide on the approach to develop the model.
Building block (models are formulated by applying the governing physical laws/con-
straints and constitute relations to the subsystems)
Black box (models are formulated on the basis of the input-output characteristics of
the system, no consideration of its internal functioning)
Model types
Decide on the types of equations that provide the appropriate compromise between simplicity
and tractability on the one hand, and accuracy on the other.
Continuous in time
Discrete in time
Deterministic
Stochastic
Types of building block models
Microscopic (spatially distributed)
Macroscopic (lumped)
a Summarized from refs. 35
Mathematical modeling involves a number of distinct steps. In order to establish a

mathematical model, many simplifying assumptions have to be specified and deci-
sions must be made, either explicitly or implicitly (35) (Table 1). The translation
process, starting from a particular (immunological) phenomenon and ending in a math-
ematical formalism (i.e., a set of equations), is the hardest part of applying mathemat-
ics, because it involves the conversion of imprecise assumptions into formulas. We
describe here one possible approach to modeling the interaction of DC with CTL. In
the presented model, the DCCTL interaction is described by adapting different theo-
retical frameworks, such as predatorprey models from population biology (6) and
Monod-type kinetics with saturation which are applied in biochemistry (7). We are
considering these deterministic models at the macroscopic level of the whole immune
system, and neglect details such as spatial structures of lymphoid organs or cytokine
networks. The overall aim of this approach is to produce a scientifically meaningful
mathematical model that is both descriptive and predictive.
In the Methods section we explain the use of the computational techniques for in
silico simulations, using the models, parameter estimation, information-theoretical
assessment of the mathematical model given the data, and sensitivity analysis. These
represent the major aspects of the general problem of valid inference (8). Although
our main interest has been the mathematical modeling of the population dynamics of
DCCTL interaction in mice, the computational methodology has widespread appli-
cability. To describe the kinetics of the population densities of immune cells, we uti-
lize either ordinary or delay differential (in the case of memory effects) equations,
which describe the rate of change of cell population densities as a function of time t.
2. Materials
2.1. Experimental Data
Reliable input from experimental or clinical research in terms of precise and com-
prehensive datasets is a core part of an interdisciplinary modeling approach. The data
set presented in Table 2 was generated using established protocols (9,10). Briefly,
major histocompatibility complex (MHC) class I tetramers complexed with the
immunodominant CTL epitope (gp33) derived from the glycoprotein of the lympho-
cytic choriomeningitis virus (LCMV-GP) (11) were used to follow activation of gp33-
specific CTL after immunization with DC. DC derived from transgenic mice
ubiquitously expressing the first 60 aa of LCMV-GP including gp33 (H8-DC) (12)
were injected intravenously (i.v.) into nave C57BL/6 recipient mice. At the indicated
time points following immunization, the densities of the following cell populations as
a function of time t were determined:
Activated CD8+62L- T-cells staining with the gp33-tetramer (tet+) in spleen that have
downregulated the CD62L molecule, Ea(t)
Quiescent CD8+CD62L+tet+ cells in spleen Em(t).
Furthermore, to assess the availability of adoptively transferred DC for productive
interaction with T-cells within secondary lymphoid organs, 51Cr-labeled H8-DC were
injected i.v. into nave recipient mice, and the accumulated radioactivity was deter-
mined in spleen at different time points using established protocols (13). The data set
for homing of adoptively transferred DC from blood to spleen has been published
elsewhere (14).
2.2. Simulation Software
The practical way to study mathematical models is based on using numerical tech-
niques to approximate the solutions. To understand the DCCTL interaction dynam-
ics, model equations were evaluated under different parameter settingsa direct
approach. Differential equations of the model were programmed using the universal
computer language FORTRAN. Efficient and reliable software packages such as
MATLAB (Website: http://www.mathworks.com) or Berkeley Madonna (Website:
http://www.berkeleymadonna.com) represent more convenient alternatives to the
classical programming approach, because they automate a number of steps (e.g.,
programming of equations, selection of the numerical method) of the process of solv-
ing initial value problems for ordinary or delay differential equations.
22
Table 2
Example of the Original Data Set (Mean Standard Deviation) for the Kinetics
of tet+ CD8 T-Cells Expansion Induced after Intravenous (i.v.) Injection of 2 105 H8-DCa
Derived data used for fitting
of the model variable
Ea E a+ E m
Days after Splenocytes tet+ (activated gp33- (total gp33-
22
i.v. injection (cells) CD8+ T-cells CD8 T-cells % CD62L specific CTL) specific CTL)
d4 (7.0 0.66) 107 (1.0 0.02 ) 107 (1.4 0.19) 105 43 6.7 % 1.4 105
d7 (4.0 1.5) 107 (7.0 2.7) 106 (3.6 1.8) 105 82 5.6 % 3.6 105
d12 (6.1 0.67) 107 (6.1 0.71) 106 (1.7 0.55) 105 71 8.2 % 1.7 105
Bocharov, Ford, and Ludewig

d18 (5.5 0.73) 107 (5.1 1.2) 106 (6.6 1.3) 104 49 5.2 % 6.6 104
a The data set used for fitting the model predictions for activated cytotoxic T-lymphocytes (CTL) (E ) and the total CTL (E + E ) variables of the
a a m
mathematical model.
Fig. 1. Conceptual model for the predatorprey type induction/regulation of CD8+ T-cell
responses by dendritic cells (DC). Antigen-expressing DC migrate from blood to spleen, where
they induce clonal expansion of nave antigen-specific cytotoxic T-lymphocytes (CTL),
whereas activated CTL eliminate DC. Arrows indicate the modeled processes.
3. Methods
3.1. Generation of a Conceptual Scheme (Model) of the System
The first task in a data- and hypothesis-driven modeling approach is to choose/
specify the functional form of the model to represent the data of the system under analy-
sis (3). Although many variables and parameters might seem necessary to describe the
effect of DC on CTL, one must a priori restrict the model to the most important interac-
tions. We quote from Burnham and Anderson (8): Development of the a priori set of
candidate models often should include a global model: a model that has many param-
eters, includes all potentially relevant effects, and reflects causal mechanisms thought
likely, based on the science of situation. This type of a global model for a systemic
dynamics of DCCTL interaction has recently been developed by Ludewig et al. (14).
Our conceptual model is based on the spatio-temporal view of the organization and
predatorprey type regulation of CD8+ T-cell responses by DC in vivo. Antigen-
expressing DC migrate irreversibly from blood to the spleen, where they induce clonal
expansion of nave antigen-specific CTL. Activated CTL eliminate the DC and recircu-
late among spleen, blood, and peripheral tissues. The key element of the model is the
predatorprey-type interaction of DC with antigen-specific CTL (Fig. 1). We present
here our modeling methodology for the splenic compartment of the immune system.
3.2. Specifying Assumptions and Selecting Quantities
To formulate equations for DCCTL interaction following i.v. injection, we make
the following simplifying biological assumptions. Such a list is also helpful for the
evaluation of the modeling results from the viewpoint of the underlying biology.
1. DC do not re-circulate from lymphoid organs into the blood after intravenous injection
(15).
2. Adoptively transferred DC are in mature state (10).
3. DC-mediated induction of antigen-specific CTL is due to their interaction in the spleen
(14).
4. DC do not divide in secondary lymphoid organs (16).
5. DC decay due to a short life span (17) and their killing by activated CTL (9,18).
6. The population of antigen-specific CTL in spleen is split into quiescent (nave or central
memory-like) and activated CTL (effector or effector memory-like) (19).
7. CTL re-circulate among spleen, blood, and peripheral organs (e.g., liver).
The required quantities for the specific model depend on the conceptual scheme
and the assumptions. The quantities include here time as an independent variable,
time-dependent variables (population densities of cells), and the parameters that char-
acterize the kinetics of the specified processes. The biological meaning and units of
the parameters considered in the presented model of DCCTL interaction are listed in
Table 3.
3.3. Derivation of Model Equations

The cell population dynamics can be represented by the following prototype mass
balance equation (see Note 1):
Change rate of number of cells of j-th type at time t =
transfer between compartments + cell division cell death transition between states (1)
We model the localized population dynamics of DCCTL interaction, i.e., in the
spleen, and ignore the DC re-circulation between spleen and blood.
The rate of change in the density of DC in the spleen is modeled as:
d Q
D ( t ) = BS Blood DBlood ( t ) D D ( t ) bDE Ea ( t ) D ( t ) (2)
dt QSpleen
The first term represents the trafficking of DC from blood to spleen (QBlood and
QSpleen being the volumes of the blood and spleen compartments, respectively), and
the other two take into account the natural death of the cells and their elimination by
activated CTL. We substitute the formulae identified in (14) for the kinetics of DC in
the blood DBlood(t).
The dynamics of activated CTL is modeled by the following equation:
d D ( t d ) Ea ( t d )
Ea ( t ) = Ea E naive Ea ( t ) + b p ram Ea ( t ) + ba D ( t ) E m ( t ) (3)
dt D + D (t d )
The first term considers the homeostasis of nave CTL in the spleen, the second
term represents the DC-induced division of CTL proceeding at the rate that saturates
at a high number of DC. The time lag between the cognate interaction of CTL with DC
represents the duration of preprogramming of CTL for division and differentiation
(20). The last two terms take into account the silencing of activated CTL into quies-
cent memory cells (third term) and the activation of the memory cells by DC.
Mathematical Approach to DC Immunotherapy
Table 3
Parameters of the DCCTL Model in Spleen
Estimate (units)
Notation Biological Definition [99% Confidence Interval]
BS Transfer rate of H8-DCs from blood to spleen 2.832 d1
t1/2 = ln2/D Half-life of gp-33-expressing DC 3 d (ad hoc fixed value)
bDE Per capita elimination rate of H8-DCs by activated CTL 0.487 105 mL/cell/d
[0.13 106, 0.6 105]
Enaive The number of nave gp-33-specific CTL contributing to primary clonal expansion 370 cells
(ad hoc fixed value)
25
d Duration of antigen preprogrammed CTL divisions 1 d (ad hoc fixed value)

E
t1/a2 = ln 2 / Ea Half-life of activated CTL 5.78 d [0.12, + )
E
t1/m2 = ln 2 / E m Half-life of resting memory CTL 69 d (ad hoc fixed value)
bp Maximal expansion factor of activated CTL per d 12 d1 [10, 85]
D DC density in the spleen for half-maximal proliferation rate of CTL 2.12 103 cell/mL
[7.5 102, 1.2 104]
ram Fraction of activated CTL reverting into resting memory cells per day 0.01 d1 [0.4 103, 1.2]
ba Activation rate of quiescent CTL by DCs 103 mL/cell/d
25
The equation for the dynamics of quiescent memory CTL is:

d
E m ( t ) = ram Ea ( t ) E m + ba D ( t ) E m ( t ) (4)
dt
which considers the transition of the activated CTL into the quiescent memory state,
the death of memory CTL at some slow rate, and the activation of memory CTL de-
pending on the availability of DC.
3.4. Computer Simulation of the System Dynamics
A typical computer simulation includes the following steps:
1. Program the set of differential equations of the model, either following protocols speci-
fied in specialized simulation packages (see Methods) or using universal computer lan-
guages such as FORTRAN or C++.
2. Set initial values for the time-dependent population densities of DC and CTL.
3. Set the values for the model parameters (see Note 2 for an example of how to derive the
initial guess for CTL proliferation parameters bp, D).
4. Set run-time parametersthe start and finish times as well as the report times for the
solution and the error-per-step tolerance in the solver.
There are several difficulties in obtaining efficiently a numerical approximation to
the solution of the model equations (see Note 3). It follows that the computer simula-
tions require specification of values for the model parameters. The parameter values
can be obtained from experimental data sets either directly, or using special numerical
procedures for the inverse modeling approach.
3.5. Estimation of Parameters and Confidence Intervals Via Data-Fitting
Estimation of adjustable parameters of a mathematical model with specified archi-
tecture is another central task of data-driven modeling (3). The purpose of data-fitting
is to calculate values of the model parameters that optimize some objective function
which is a measure of the fit (agreement) between the simulated values (predictions of
the model) and the data (see Note 4). Figure 2 shows the experimental data and the
corresponding best-fit solution of the mathematical model.
3.5.1. Least-Squares Fitting Functions
Let the set of observation times {t j } j =1 and observations { y j } j =1 for M (1 i
N i N
M) of the model variables be specified. An example of the experimental data set repre-
senting the kinetics of tet+ CD8 T-cells expansion after i.v. injection of 2 105 H8-DC
used in parameter estimation is given in Table 2. Simulation of the system with the
model suggests prediction (a regression function) for the observed variable yi(tj,p),
which depends on the adjustable parameters p, i.e., the model parameter vector with L
components. The common ordinary least-squares (LSQ) error measure of the match
between the model and data is given by the objective function determined by the square
of the absolute deviation between the model and data:
N M 2
( p) = y ij y i ( t j , p )
(5)
j =1 i =1
Fig. 2. Data vs model description for the population dynamics of dendritic cells (DC) and
cytotoxic T-lymphocytes (CTL) in spleen induced by intravenous injection of 2 105 gp33-
presenting H8-DC. The symbols represent averages of 3 mice standard deviation. The smooth
lines predict the population dynamics of the total tet+, activated tet+, and quiescent memory tet+
CTLs and H8-DCs for the best-fit estimates of the model parameters. DC elimination follows a
biphasic kinetics; the first, slower phase reflects their life span and the accelerated decay phase
results from the killing effect by activated CTLs.
The weighted LSQ objective function
N M 2
(p) = w ji y ji y i (t j , p ) (6)
j =1 i =1
can be used when different weights have to be given to particular observations. A

( )
2
natural choice for the weights could be the inverse variances w ij = ji of the obser-
vations. The LSQ approach proves to be efficient when the variation in scale of the
data set over the observation time interval is within one order of magnitude. This is
exactly the case with the data for the initial biodistribution kinetics of DC after intra-
venous injection (see ref. 14 and Note 5).
3.5.2. Minimization of the Least-Squares Function

The numerical technique for finding the best-fit parameter estimates for a given
data set, mathematical model, and objective function requires (1) providing an initial
guess for the model parameters; (2) solving the model equations to compute (p);
and (3) adjusting the parameter values by some minimization routine available, for
example, in the MATLAB library ( fmins). Because the parameters of the model are
constrained to be nonnegative, we used a log10-transformation for the parameters, so
that the resulting sequence of reduced order data-fitting problems were treated as an
unconstrained minimization (see Note 6).
There are software packages such as Berkeley Madonna that can automatically find
the values of a number of parameters in a model that minimize the deviation between
the models output and a data set. In principle, these facilities greatly enhance the
capability of treating the parameter estimation problem. However, data-fitting with
non-linear differential equations-based models is a type of expertise that can be learned
only in actual practice (see refs. 2123 for extensive discussion of various aspects of
statistical techniques and computational methods).
3.5.3. Computation of Confidence Intervals
To characterize the precision or reliability of best-fit parameter estimates (p*) a
number of approaches exist, of which we consider (1) the variance-covariance matrix
for estimated parameters (23); and (2) the profile-likelihood-based method (24) (see
Note 7). Alternatively, a bootstrap approach may be used (21,25). The confidence
interval analysis is a computation-intensive procedure. The simplest approach is to
approximate the 95% confidence intervals using estimates of the standard errors
p*l 1.96 l , where l = vll , 1 l L stand for the diagonal elements of the vari-
ance-covariance matrix. In turn, the matrix elements are determined via the residual
sum of LSQ function, the number of degrees of freedom, and the Hessian matrix
(22,23), according to the formula
1
2 2 2 2
( p* ) ( p* ) ( p* ) ( p* )
p1
2 p1p2 p1p3 p1p L
2
v11 c12 c13 c1L 2 2 2
c ( p* ) ( p* ) ( p* ) ( p* )
v22 c23 c2 L p 2 p1 p22 p2 p3 p2 p L
21 ( p* )
c31 c32 v33 c3 L = 2 2 2
2 2
N obs L ( p* ) ( p* ) ( p* ) ( p* )

p3p1 p3p2 p32 p3p L (7)
c L1 cL 2 cL 3
v LL

2 2 2 2
( p* ) ( p* ) ( p* ) ( p* )
p L p1 p L p2 p L p3 p L2
where Nobs and L stand for the total number of scalar observations and the number of
model parameters, respectively. The Hessian matrix can be approximated using the
complete information matrix or via numerical differentiation provided in MATLAB.
The estimates of the parameters for DCCTL interaction in spleen and their confi-
dence intervals are shown in Table 3 (see Note 7).
3.5.4. Evaluation of the Model Parsimony and Accuracy
In general, the biological model architectures are not comprehensively justified on
the basis of proven mechanisms, but reflect a parsimonious characterization of the
system under study (26). The mathematical model presented here is not the only plau-
sible model for studying the dynamics of DCCTL interaction in vivo. Other formu-
lations may be preferable given a different context (e.g., data) or a different set of
goals. It is important, therefore, not only to rank plausible models with respect to
their consistency with the data (as measured by the best-fit objective function), but
also to consider their distance to an unknown true model underlying the data set.
The information-theoretic framework provides a basis for such assessment of infor-

mation complexity of the model (the model parsimony) and the selection of the best
ones with respect to the information in the data (8). Using the maximized-likelihood
function, one can compute the value of the corrected Akaikes information criterion
(AIC) (appropriate for data sets smaller than 40 units):
L+2
cAIC = 2 ln ( p* , ) + 2 ( L + 1) 1 + (8)
N obs ( L + 2 )
with (p*,) being the maximized likelihood function, or the revised relative indicator
L+2
cAIC = N obs ln ( p* ) + 2 ( L + 1) 1 + (9)
N obs ( L + 2 )
in which extraneous terms are discarded and the relationship between the maximum-
likelihood estimation and the LSQ estimation (either ordinary- or Log-LSQ) is used.
The best model is considered to be that which yields the lowest value of the indicators.
The indicators can be regarded as taking parsimony into account.
3.6. Sensitivity Analysis
The major advantage of a mathematical model is that the model parameters can be
easily altered and the effect on system dynamics can be examined via sensitivity analy-
sis (see Note 8). We present an example in which the initial numbers of injected DC
and nave CTL were changed to predict the effect on the peak expansion of CTL. In
this process, the model equations were run using various combinations of the numeri-
cal values of the initial numbers of DC and CTL; other parameter values are shown in
Table 3. The computer experiments summarized in Fig. 3 suggest that the peak CTL
expansion is a conserved feature of the DCCTL interaction dynamics once the initial
numbers of DC and CTL have reached a saturation level. Furthermore, this simulation
indicates that increasing the initial number of CTL above a certain threshold does not
lead to further expansion because of the accelerated elimination of the antigen-pre-
senting DC.
4. Notes
1. Various functional forms can be suggested for process terms in the above equation, as
illustrated for the DC-induced proliferation rate of T-cells in Table 4. To model dose-
response curves, there is a broad set of phenomenological equations available that allow
one to get the desired shape, including one-site saturation, sigmoidal with variable slope,
four-parameter logistic equation, and so on. It has to be determined which of these forms
is most consistent with the system under consideration, and this can be judged either by
using a priori biological arguments, or by confronting against experimental data as pre-
sented, for example in (27), or using the information-theoretic model selection criteria
(see Subheading 3.5.4.).
2. Experimental data on CTL priming with DC (10) indicate that 1001000 DC have to
reach the spleen to achieve protective levels of CTL activation. These numbers directly
suggest the range for the initial estimates of the threshold parameter of half-maximal
CTL activation, i.e., 100 D 1000 (cells). The doubling time of CTL during the
Fig. 3. Sensitivity analysis of CTL expansion to parameters variation. The maximal expan-
sion of gp33-specific CTL as a function of the initial number of intravenously injected H8-DC
and the initial number of the specific CTL.
Table 4
Some Functional Forms for DC-Induced CTL Growth
and the Associated Qualitative Implications
Process
DC-induced per capita T cell division rate
Form1 Dose/Effect Form2 Dose/Effect Form3 Dose/Effect
DC DC
~DC ~ ~
+ ( DC )
2
+ DC
expansion phase ranges from 6 to 24 h. This suggests a plausible range for the CTL
amplification per day parameter 1 bp 24 = 16 (d1). Data-fitting procedures are
required to further refine these initial estimates (see Subheading 3.5.).
3. Voit (7) presents a useful practical guide to doing simulations with mathematical models
formulated using systems of ordinary differential equations. They offer a basic, easy to
install and start software for Power Law Analysis and Simulation (PLAS). Some differ-
ential equations model processes with widely separated decay times (called stiff) require
special stiff solvers (based upon either implicit Runge-Kutta or Gears methods) to pro-
duce simulations. These are available, for example, in the software package Berkeley
Madonna. In the case of models formulated with delay (rather than ordinary) differential
equations, software development is still an active area of research. Some interesting guide-
lines and recipes on how to conduct in silico experiments on the model system are dis-
cussed in (7) for biochemical models.
4. We have available observations as input data, which typically arise from a multiple series
of observations, in which case the individual data have to be summarized in some way
(e.g., as means and standard errors). Since the basis of the data-fitting usually includes
assumptions about distributions (e.g., normal or Log-normal) of the errors in observations,
it is important to ensure that those assumptions are realistic. A general statistical frame-
work is the Bayesian approach, which under some natural assumptions reduces to Maxi-
mum A Posteriori Estimation and, further down, to Maximum Likelihood (ML) Estimation
(3,23). The choice depends on the knowledge of the statistical features of the data, expec-
tations in advance about the model parameter values, and the selected set of models. If one
uses a single model and assumes a uniform prior on the parameter values, then the ML
approach would be the natural choice (3,23). The LSQ parameter estimation is equivalent
to the ML estimation under the set of assumptions (often made implicitly) that (1) the
observational errors are normally distributed, (2) equivalent positive and negative devia-
tions from expected values differ by equal amounts; and (3) the errors between samples are
independent and identically distributed. Other powers of the deviation between the model
and the data can be used depending on the error distributionfor example, the first power
would correspond to an exponential distribution of the errors (3).
5. Immunological characteristics may vary over several orders of magnitude. In such cases
logarithmic scaling should be applied in order to make the data-fitting problem equally
sensitive to differences between the model and data in the lower end of the observed
values. Statistically, this implies an assumption of geometric normality of observational
errors, in which equivalent deviations differ by equal proportions. Minimizing the Log-
LSQ function is equivalent to maximizing the likelihood function under the assumption
that the observational errors are independent and Log-normally distributed. The param-
eter estimation problem corresponds to a choice of Log-LSQ (relative deviation) objective
function
{ln ( yij ) ln yi (t j , p)}

N M 2
LogLS ( p ) = (10)
j =1 i =1
The DC-induced population dynamics of tet+ CD8+ T-cells is exactly such a case. To
estimate the model parameters characterizing the DCCTL interaction, we minimized the
Log-LSQ function in an iterative way over sequentially expanding time intervals domi-
nated by different sets of parameters.
6. In terms of efficiency, it may be advantageous to try a number (or a combination) of
minimization methods. Whereas the initial estimates of model parameters can first be
improved by a computationally simple, slowly converging method (e.g., derivative-free
Simplex method), the resulting better guess can further be improved using a computa-
tionally extensive but rapidly converging procedure (e.g., quasi-Newton method). Good
starting values for parameter estimates can sometimes be obtained by a sequential pro-
cess of refining the estimates via fitting the subsets of the data, which are obtained by
subdividing the observation interval. As the size of the subinterval increases, the best-fit
parameter values can be improved in a step-by-step manner. The limited accuracy of the
numerical solution of the mathematical model and, therefore, the correct number of digits
in the value of the LSQ function must be accounted for in the minimization process,
especially if it uses finite-difference approximations to the derivatives of the objective
functions.
7. For complex mathematical models the variance-covariance matrix approach to assess the
precision of parameter estimates is more difficult to implement because of the complexity
(this is the case for the mathematical model of systemic DCCTL interaction). We used a
different approach: approximate (e.g., 99%) confidence regions for every best-fit param-
eter estimate (*) of the whole parameter vector p* p*( L 1) , * were obtained using
the profile-likelihood-based method (24), by searching for those values of that satisfied
the following inequality
( [ ] ) e N
2
LogLS p*L 1 ,
1,1 0.01
obs
LogLS ( p*L 1 , * )
[ ] (11)
The value of 12, 0.99 , which stands for the 0.99th quantile of the 2-distribution on one
degree of freedom is 6.635 (see standard statistical textbooks), Nobs stands for the number
of observations. Therefore, numerically one needs to find the two extreme values of
(the endpoints of the confidence range) that enable
6.635
(
LogLS p*
[ L 1]
)
, = e N obs
( [ ]
LogLS p*L 1 , * ) (12)
8. Sensitivity analysis in modeling of biochemical and bioengineering systems has been dis-
cussed elsewhere (7,28). To summarize, the sensitivity analysis has multiple functions: it
can indicate the controllability of the system dynamics, the parameter redundancy in the
model, the effect of errors in the data on parameter estimates, and the predictability of the
system, as well as assisting in ranking the importance of the modeled processes. In addi-
tion to the simplest form of sensitivity analysisi.e., how a permanent perturbation in
parameter(s) around the best-fit value affects the solution (transient or steady-state)one
may study the effect of time-varying perturbation(s), or the sensitivity of more complex
system features, like the clonal burst size, the transition time to a particular state, and so
on. The practical details of sensitivity analysis with PLAS software created by Ferreira
are discussed in detail by Voit (7). The simplest sensitivity computations require that the
parameter of interest be independent of the remaining parameters in the model; otherwise
one has to take into account all functional dependences upon other parameters.
Acknowledgments
We thank Hans Hengartner and Rolf Zinkernagel for continued support. This work
was supported by the Swiss National Science Foundation, the Kanton of St. Gallen,
the Wellcome Trust, the Alexander von Humboldt Foundation, the Leverhulme Trust,
and the Russian Foundation for Basic Research.
References
1.
1 Schuler, G., Schuler-Thurner, B., and Steinman, R. M. (2003) The use of dendritic cells in
cancer immunotherapy. Curr. Opin. Immunol. 15, 138147.
2.
2 Steinman, R. M. and Pope, M. (2002) Exploiting dendritic cells to improve vaccine effi-
cacy. J. Clin. Invest. 109, 15191526.
3. Gershenfeld, N. (2002) The Nature of Mathematical Modelling. Cambridge University
Press, Cambridge, UK.
4. Kutz, M. (2002) Standard Handbook of Biomedical Engineering and Design. McGraw-
Hill Professional Publishing, New York, NY.
5. Blanchard, P., Devaney, R. L., and Hall, G. R. (1998) Differential Equations. Brooks/Cole
Publishing Company, Pacific Grove, CA.
6. Volterra, V. (1926) Variations and fluctuations in the numbers of co-existing animal spe-
cies. In: The Golden Age of Theoretical Biology 19231940. Scudo, F. M., and Ziegler, J.
R., eds. Springer, Berlin, Heidelberg, Germany, pp. 65236.
7. Voit, E. O. (2000) Computational Analysis in Biochemical Systems. A Practical Guide for
Biochemists and Molecular Biologists. Cambridge University Press, Cambridge, UK.
8. Burnham, K. P. and Anderson, D. R. (2002) Model Selection and Multimodel Inference
a Practical Information-Theoretic Approach, 2nd Ed. Springer, New York, NY.
9.
9 Ludewig, B., Bonilla, W. V., Dumrese, T., Odermatt, B., Zinkernagel, R. M., and
Hengartner, H. (2001) Perforin-independent regulation of dendritic cell homeostasis by
CD8(+) T cells in vivo: implications for adaptive immunotherapy. Eur. J. Immunol. 31,
17721779.
10.
10 Ludewig, B., Ehl, S., Karrer, U., Odermatt, B., Hengartner, H., and Zinkernagel, R. M. (1998)
Dendritic cells efficiently induce protective antiviral immunity. J. Virol. 72, 38123818.
11 Gallimore, A., Glithero, A., Godkin, A., et al. (1998) Induction and exhaustion of lympho-
11.
cytic choriomeningitis virus-specific cytotoxic T lymphocytes visualized using soluble
tetrameric major histocompatibility complex class Ipeptide complexes. J. Exp. Med. 187,
13831393.
12.
12 Ehl, S., Hombach, J., Aichele, P., et al. (1998) Viral and bacterial infections interfere with
peripheral tolerance induction and activate CD8+ T cells to cause immunopathology. J.
Exp. Med. 187, 763774.
13.
13 Eggert, A. A., Schreurs, M. W., Boerman, O. C., et al. (1999) Biodistribution and vaccine
efficiency of murine dendritic cells are dependent on the route of administration. Cancer
Res. 59, 33403345.
14.
14 Ludewig, B., Krebs, P., Junt, T., et al. (2004) Determining control parameters for dendritic
cellcytotoxic T lymphocyte interaction. Eur. J. Immunol. 34, 24072418.
15.
15 Fossum, S. (1988) Lymph-borne dendritic leukocytes do not recirculate, but enter the
lymph node paracortex to become interdigitating cells. Scand. J. Immunol. 27, 97105.
16.
16 Ardavin, C., Martinez, del Hoyo, G., Martin, P., et al. (2001) Origin and differentiation of
dendritic cells. Trends Immunol. 22, 691700.
17.
17 Ruedl, C., Koebel, P., Bachmann, M., Hess, M., and Karjalainen, K. (2000) Anatomical
origin of dendritic cells determines their life span in peripheral lymph nodes. J. Immunol.
165, 49104916.
18.
18 Hermans, I. F., Ritchie, D. S., Yang, J., Roberts, J. M., and Ronchese, F. (2000) CD8+ T
cell-dependent elimination of dendritic cells in vivo limits the induction of antitumor
immunity. J. Immunol. 164, 30953101.
19.
19 Wherry, E. J., Teichgraber, V., Becker, T. C., et al. (2003) Lineage relationship and pro-
tective immunity of memory CD8 T cell subsets. Nat. Immunol. 4, 225234.
20.
20 van Stipdonk, M. J., Lemmens, E. E., and Schoenberger, S. P. (2001) Naive CTLs require
a single brief period of antigenic stimulation for clonal expansion and differentiation. Nat.
Immunol. 2, 423429.
21. Armitage, P., Berry, G., and Matthews, J. N. S. (2002) Statistical Methods in Medical
Research. Blackwell Science, Oxford, UK.
22. Baker, C. T. H., Bocharov, G., Paul, C. A. H., and Rihan, F. A. (2004) Computational
modeling with functional differential equations: identification, selection and sensitivity,
Appl. Num. Math., in press.
23. Bard, Y. 1974. Nonlinear Parameter Estimation. Academic, New York, NY.
24. Venzon, D. J. and Mooogavkor, S. H. (1988) A method for computing profile-likelihood-

based confidence intervals. Appl. Statistician 37, 8794.
25.
25 De Boer, R. J., Oprea, M., Antia, R., Murali-Krishna, K., Ahmed, R., and Perelson, A. S.
(2001) Recruitment times, proliferation, and apoptosis rates during the CD8(+) T-cell
response to lymphocytic choriomeningitis virus. J. Virol. 75, 10,66310,669.
26.
26 Wood, S. N. and Thomas, M. B. (1999) Super-sensitivity to structure in biological models.
Proc. R. Soc. Lond. B Biol. Sci. 266, 565570.
27. Borghans, J. A., Taams, L. S., Wauben, M. H., and De Boer, R. J. (1999) Competition for
antigenic sites during T cell proliferation: a mathematical interpretation of in vitro data.
Proc. Natl. Acad. Sci. USA 96, 10,78210,787.
28. Rabitz, H. (1981) Chemical sensitivity analysis theory with applications to molecular
dynamics and kinetics. Comput. Chem. 5, 167180.
Microspheres for Targeting Antigens to DC 35
3
Delivery of Tumor Antigens to Dendritic
Cells Using Biodegradable Microspheres
Ying Waeckerle-Men, Bruno Gander, and Marcus Groettrup
Summary
Poly(D,L-lactide-co-glycolide) (PLGA) polymers have been used for the production of bio-
degradable medical sutures and for controlled drug release for decades. Useful characteristics
such as in vivo biodegradability, an adjustable release profile, and the very high encapsulation
capacity have stimulated immunologists to explore PLGA microspheres (MS) as antigen deliv-
ery systems for vaccination for more than 15 yr. In previous studies aiming at the development
of single-dose vaccines, direct immunization with PLGA MS containing various antigens
induced strong and sustained immune responses. We have observed that human immature
monocyte-derived dendritic cells (MoDC) prepared for clinical application are able to internal-
ize high numbers of MS without negative effects on their pivotal properties. Furthermore,
PLGA-MS-incorporated antigens are effectively processed for presentation on major histo-
compatibility complex (MHC) class I and MHC class II molecules by dendritic cells (DCs) in
vitro and induced strong cytotoxic T-lymphocyte (CTL) responses in vivo. Taken together,
PLGA MS is a promising delivery vehicle for the improvement of current DC-based tumor
vaccine protocols.
Key Words: Antigen processing; dendritic cells; immunotherapy; microspheres;
poly(lactide-co-glycolide); vaccine delivery.
1. Introduction
DC-based immunotherapy for various cancers is increasingly being tested in clini-
cal trials. In some cases, vaccination with tumor antigen-pulsed autologous DC corre-
lated well with tumor regression, although the average response rates (1020 %) were
quite low (1,2). The high rate of clinical failures is most likely due to the inability to
induce strong and sustained cellular immune responses. Although it has emerged that
mature rather than immature DC should be applied, most parameters of DC-based
tumor vaccine development still need to be optimized in order to improve the clinical
efficacy. Among these parameters are the vaccination schedule, the route of injection,
the cell number used for inoculation, tumor antigen selection, and, especially, the mode
of antigen loading into DC. In many current DC-based cancer vaccine trials, short
35
36 Waekerle-Men, Gander, and Groettrup
Fig. 1. Microscopic image of poly(D,L-lactide-co-glycolide) microspheres (PLGA MS) and

chemical formula of PLGA. (A) Scanning electron micrograph of PLGA MS, and (B) formula
of PLGA, composed of lactic acid (left bracket) and glycolic acid (right bracket).
tumor antigen-derived peptide ligands of certain MHC alleles, e.g., HLA-A*0201, are
used for antigen loading. This strategy excludes many cancer patients from the therapy
because they lack the specific MHC allele. Another problem resides in the short half-
life of peptide/MHC class I molecule complexes on the cell surface of externally loaded
DC (3). The use of tumor-derived lysate or mRNA as a source of antigens may over-
come the problem of MHC restriction and represent a useful alternative when tumor-
specific antigens are difficult to define. However, sufficient amounts of autologous
tumor material are not easily obtained, and the in vitro cultivation of tumors for long-
term vaccine supply often fails. In addition, cross-presentation of soluble proteins by
antigen-presenting cells (APC) is extremely inefficient.
1.1. PLGA Microspheres for Antigen Delivery
An interesting alternative to the exogenous supply of soluble peptides is to admin-
ister antigens in particulate form. One prominent example is the incorporation of anti-
gens into biodegradable PLGA MS (Fig. 1). PLGA polymers have been widely used
for the manufacturing of surgical sutures and controlled drug-delivery systems for
human use. They are aliphatic polyesters composed of varying proportions (50:50 up
to 100:0) of lactate and glycolate (Fig. 1B). In contact with aqueous media, PLGA is
partly hydrated and slowly hydrolyzed to lactic and glycolic acids, which are further
metabolized via the citric acid cycle. Medical-grade PLGA polymers differ in molecu-
lar mass and co-polymer composition, two properties that determine mainly the extent
of swelling, degradation kinetics, and the release of entrapped drugs or antigens. We
mostly use a relatively low molecular weight (1417 kDa) PLGA 50:50 or PLGA
75:25, either with free carboxylic end groups (uncapped) or esterified (capped) end
groups (see Note 1). These fairly hydrophilic PLGA polymers were found suitable for
the release of antigens from MS over a period of 3060 d, a period of time after which
the polymer is largely hydrolyzed (4).
Microencapsulation of antigens into PLGA can be achieved by several methods.

Most of them involve the dissolution of PLGA in an organic solvent, formation of
PLGA solution droplets, and extraction of the organic solvent into an aqueous phase
or warm air. Most commonly employed methods encompass solvent extraction from
a water-in-oil-in-water (w/o/w) dispersion, polymer coacervation, spray-drying, or
spray freeze-drying (5). In this chapter, we describe the spray-drying technique at
laboratory scale, which yields small MS with diameters of less than 10 m, a size
range ideally suited for the uptake by macrophages and DC (6,7). Further, we also
describe methods to determine the content of microencapsulated tetanus toxoid (TT)
in PLGA MS and a test to assess the in vitro release of antigens from PLGA MS.
1.2. Evaluation of PLGA-MS Uptake by MoDC

The presence of PLGA MS inside phagocytic cells can be observed through con-
ventional light microscopy (6) or by fluorescence microscopy when the MS contain a
fluorescent dye such as coumarin 6 (8). In addition, the proportion of cells that have
ingested coumarin 6-labeled MS (MS-C6) can be determined by flow cytometry (FL1-
H channel for green fluorescence). Here, we outline the determination of PLGA MS
uptake by human MoDC. Since PLGA MS tend to adhere to the cell surface, a method
to remove externally adhering MS is also described. Detailed protocols for MoDC
preparation can be found in Chapters 6 and 8 of this book.
1.3. Quality Control of Monocyte-Derived DC After PLGA-MS Uptake

As the hydrolysis of PLGA MS produces lactic and glycolic acids, one might
expect that the resulting acidification during PLGA degradation alters the cellular
functions of MoDC such as their maturation capability, recovery and survival rate,
the capability to stimulate nave T-cells, and the pattern of proinflammatory cytokine
secretion. Therefore, the quality control of PLGA-MS-loaded MoDC is necessary,
because a sensitive and long-lasting presentation of CTL and T helper epitopes can
occur only when the essential functional properties of MoDC are not adversely
affected after MS loading. In addition, for use as cellular tumor vaccines, MS-loaded
DC should still be able to migrate from the site of injection to the secondary lymphoid
organs, where they may encounter and stimulate antigen-specific T-cells. We have
recently found that immature DC loaded with PLGA MS can be fully matured with
different DC maturation stimuli, and that the functional and phenotypical hallmarks
of MoDC maturation were not changed through PLGA MS uptake. We also observed
that the migration capacity of matured, PLGA-MS-loaded MoDC did not decline.
Interestingly, the survival of these cells was even consistently improved by approx
50% after treatment with Fas mAb, when the cells were matured with lipopolysaccha-
ride (LPS). In addition, MS-carrying MoDC maintained the capacity to stimulate nave
allogeneic T helper cells and secrete amounts of the cytokines interleukin (IL)-12, IL-
10, and tumor necrosis factor (TNF)- comparable to those of untreated control cells
(8). The phenotypic and functional characterization of clinical-grade DC is covered in
Chapter 9. Therefore, only special parameters related to PLGA MS loading are illus-
trated in this chapter.
1.4. Presentation of Antigens Contained in PLGA MS by Dendritic Cells

Previous studies in our laboratories demonstrated that immunization with various
antigens, including natural proteins and synthetic antigenic peptides, gave rise to
high antibody titers as well as strong and sustained T helper and CTL responses in
vivo (912). Protein antigens were presented both on MHC class I and class II mol-
ecules when administered to macrophages or DC in association with PLGA MS (6).
Based on these findings and the evidence that DC are essential for T-cell priming,
we have evaluated PLGA MS as antigen-delivery and -targeting devices, specifi-
cally regarding their potential to prolong antigen presentation by DC and, hence, to
improve DC-based tumor vaccines.
Although presentation of MHC class II-restricted antigens by DC is, in general,
rather efficient with soluble exogenous proteins such as TT, we could still observe
that a 10-fold lower concentration of antigen was sufficient to obtain the same level
of T-cell proliferation when TT was offered to DC by PLGA MS loading. In addi-
tion, when we determined the kinetics of antigen presentation with MoDC by using
a low concentration of TT (e.g., 0.01 g/mL), MoDC pulsed with soluble TT ceased
to stimulate TT-specific T-cells after 6 d, whereas MS-TT-charged DC were still
able to activate T-cells 10 d after MS-TT loading (Y. Waeckerle-Men, manuscript in
preparation). Thus, antigen delivery via PLGA MS can lead to a prolongation of
MHC class II-restricted antigen presentation.
Cross-presentation of soluble proteins by DC can occur, but is extremely ineffi-
cient, while strong cross-presentation of exogenous protein antigens in particulate
forms such as iron oxide beads (13), latex beads (14), and polysaccharides (15) or in
PLGA MS (6) has been observed in phagocytic APC. We have shown that MoDC
loaded with microencapsulated FluM5866 peptide prolonged the antigen presentation
time for up to 9 d, as compared with the soluble peptide (for up to 3 d) (16, Fig. 2).
This prolongation of antigen presentation might be due to the slow hydrolysis of PLGA
MS within DC, which provide a continuous supply of peptide ligands for newly syn-
thesized MHC class I molecules (Fig. 3). In this chapter, we describe two examples of
studying the presentation of MS-encapsulated antigens to specific CD4+ or CD8+ T-
cells by MoDC, using MS charged with the TT protein and FluM5866 peptide (desig-
nated MS-TT and MS-Flu).
2. Instruments and Materials
2.1. Microencapsulation by Spray-Drying
1. Spray-dryer, e.g., Mini Spray-Dryer 191 (Bchi, Flawil, Switzerland).
2. Ultrasonicator, e.g., Vibra-cell (Sonics & Materials, Danbury, CT).
3. Resomer RG502H (Boehringer-Ingelheim, Ingelheim, Germany, see Note 1).
4. Ethyl formate (analytical grade, Fluka, Buchs, Swizerland).
5. Fluorescent dye, e.g., Coumarin 6 (Fluka).
6. Tetanus toxoid (Pasteur Mrieux, Paris, France).
7. Membrane filters (0.2 m) of regenerated cellulose (RC 58, Schleicher and Schuell,
Dassel, Germany).
2.2. Uptake of Microspheres by Dendritic Cells
1. Immature monocyte-derived dendritic cells and DC culture medium (for details see Chap-
ters 6 and 8).
Fig. 2. Prolonged presentation of a cytotoxic T-lymphocyte (CTL) epitope by monocyte-

derived dendritic cells (MoDC) after antigen delivery via PLGA MS. Immature MoDC gener-
ated from HLA-A*0201 positive donors were loaded with 1 g/mL of Flu peptide in poly-
(D,L-lactide-co-glycolide) microspheres (PLGA MS) (MS-Flu peptide), in soluble form (sol. Flu
peptide), with empty MS (MS-E) or left untreated. The cells were then matured with lipopolysac-
charide (5 g/mL) for 48 h. After washing and being irradiated (20 Gy), MoDC were seeded into
96-well flat bottom plates (2 104/well). Subsequently, Flu peptide-specific cytotoxic T-lym-
phocytes (CTL) (1 105 per well) were added to MoDC either immediately (d 0), or on d 1, 4, or
7 after peptide/MS loading. After 20 h of co-incubation of MoDC and CTL, supernatants were
collected for determining interferon- production by ELISA. The data represent the mean of
duplicate cultures standard error of the mean.
2. 24-well Cell-culture plates (Cellstar, Greiner Bio-One, Frickenhausen, Germany).

3. Sterilized 15-mL centrifugation tubes.
4. Dimethylsulfoxide (DMSO) (Sigma D-2650, Switzerland).
5. Phosphate-buffered saline (PBS), pH 7.4.
6. FACS buffer (2% FCS in PBS containing 0.05% sodium azide).
7. Optical microscope.
8. Fluorescent microscope, e.g., Leica DMR fluorescence microscope with digital camera.
9. Flow cytometer, e.g., FACScan flow cytometer (Becton Dickinson).
2.3. Quality Control of PLGA MS-Loaded MoDC
1. FITC-labeled anti-human CD80, CD83, CD86, HLA-DR, CD95, and the respective
isotype controls FITC-labeled mouse IgG1, IgG2a, IgG2b (BD Pharmingen, Basel, Swit-
zerland).
2. Rat antihuman-CCR7 mAb (Note 3) and FITC-conjugated goat-antirat IgG (Jackson
Immunoresearch, La Roche, Switzerland) was used as secondary antibody for anti-CCR7
staining.
3. Human chemokines CCL19 and CCL21 (R&D Systems, Wiesbaden-Nordenstadt, Ger-
many).
4. Annexin V-FITC (BD Pharmingen).
Fig. 3. Schematic illustration of antigen processing of poly(D, L -lactide-co-glycolide)

microspheres (PLGA MS)-encapsulated proteins. Proteins encapsulated in PLGA MS are phago-
cytosed by dendritic cells (DC). PLGA MS can escape the endosome and are hydrolyzed in the
cytoplasm, thereby slowly releasing the protein for degradation by the proteasome. Resulting
peptides are transported via TAP transporter into the endoplasmic reticulum, where they bind
major histocompatibility complex (MHC) class I heavy chain/2-microglobulin complexes and
migrate via the Golgi apparatus to the cell surface for recognition by cytotoxic T-lymphocytes
(CTL). Alternatively, PLGA MS are hydrolzyed in endo-lysosomes; incorporated proteins are
processed by cathepsins for presentation on MHC class II molecules on the cell surface, where
they are recognized by T helper cells.
5. Propidium iodide (BD Pharmingen).

6. Transwell plates with polycarbonate filters of 5 m pore size (e.g., Corning Costar,
Cambridge, MA).
7. CD4+/CD45RO- Multi Sort Kit (Miltenyi, Bergisch Gladbach, Germany).
8. ELISA kits for TNF-, IL-10, and IL-12p70 (BD Pharmingen).
2.4. Presentation of Microencapsulated Antigens to T-Cells
1. Immature monocyte-derived dendritic cells and the relevant DC medium (e.g., AIM V
medium with GM-CSF and IL-4) (see Chapters 6 and 8).
2. Lipopolysaccharide (LPS, from Salmonella abortus equi, Fluka).
3. Soluble TT (1 mg/mL).
4. MS-TT (e.g., 10 g TT per mg of MS).
5. Soluble HPLC-purified FluM5866 (amino acid sequence: GILGFVFTL).
6. TT-specific CD4+ cell lines generated by repeated in vitro re-stimulation with TT in the
presence of human IL-2 by using a standard protocol.
7. T-cell culture medium: Iscoves Modified Dulbeccos Medium (IMDM) with
GlutaMAX-1 supplemented with 10% human AB serum, 10 mM HEPES and 1% peni-
cillin/streptomycin (Gibco-BRL Life Technologies, Paisley, UK).
8. ELISA kit for human interferon (IFN)- (BD Pharmingen).
3. Methods
3.1. Microsphere Preparation by Laboratory Spray-Drying
Spray-drying is a continuous process suitable for small- and large-scale PLGA MS
production. Here, we specially outline the preparation of three types of MS: MS with-
out content (empty or placebo MS, MS-E); MS containing coumarin 6 (MS-C6), a
green fluorescent dye used for evaluation of MS uptake by DC; and MS containing
tetanus toxoid (MS-TT), which have been used for antigen presentation studies. For
spray-drying, 1 g of Resomer RG502H is first dissolved in 20 mL ethyl formate to
obtain a 5% (m/v) polymer solution. This polymer solution is then spray-dried at a
flow rate of 2 mL/min, and at inlet and outlet temperatures of 48C and 40C, respec-
tively, to yield MS-E. Further technical details on the process are described elsewhere
(4). MS-C6 is prepared by directly dissolving 5 g/mL coumarin 6 in a solution of 5%
PLGA in ethyl formate. Spray-drying is performed as described above for MS-E. For
preparing MS-TT, 1 mL of TT in phosphate-buffered saline (PBS) (10 mg/mL) is
added to 20 mL of 5% PLGA in ethyl formate; the aqueous TT-PBS is dispersed in
the organic PLGA solution under ultrasonication for 2 30 s on ice. Then, the water
in oil emission dispersion is spray-dried under the conditions outlined above. The
spray-dried MS are collected, washed with water to remove non-encapsulated pro-
tein, and dried under vacuum for 24 h. Numerous methods have been proposed for the
quantification of the actual antigen content of PLGA MS (17). The choice of method
depends on the properties of particular antigen and the desired or required assay (see
Note 2).
The in vitro release of antigen from PLGA MS can be monitored by incubating 30
40 mg of MS in 4 mL PBS containing 0.02% sodium azide at 37C under mild rota-
tional movement. At regular intervals, 1 mL of release test medium is withdrawn for
analysis and replaced by fresh buffer. The sampled test medium is assayed photo-
metrically, fluorometrically, or by ELISA for protein concentration. During release
testing, the pH of the release medium should be monitored and adjusted if necessary.
All determinations should be performed in triplicate. In addition, the amount of
unreleased antigen remaining inside the PLGA MS can also be monitored at different
time points by antigen extraction and assay (see Note 2).
3.2. Evaluation of PLGA-MS Uptake by MoDC
Fluorescent PLGA MS (e.g., MS-C6) are dispersed in DC culture medium and sub-
sequently added to MoDC at a concentration of 0.2 mg of particles per 1 106 cells/
well in 24-well cell-culture plates. Incubation is performed at 37C for 418 h, or at
4C as control. After incubation, MoDC are collected and washed three times with
PBS. The MS-C6 uptake can be examined directly by fluorescence microscopy. Phase-
contrast and fluorescence micrographs are acquired for a semi-quantitative estimation
of MS uptake.
The uptake of fluorescent MS can also be evaluated with flow cytometry. After
co-incubation of MS-C6 and MoDC, as described previously, 1 105 MoDC are col-
lected, washed, and resuspended in 500 L of FACS buffer. The MS uptake efficiency
is analyzed by flow cytometry. MoDC that have not been co-incubated with PLGA MS
are used as negative control. The whole DC population is first gated in the forward
scatter/side scatter dot plot. Then, the number of FL1-H-positive cells observed within
the forward scatter/FL1-H dot plot is scored as DC loaded with PLGA MS.
Flow-cytometric analysis is a very convenient method to quantify the uptake of fluo-
rescent MS. However, this technique per se cannot distinguish between externally
adhering and intracellularly localized PLGA MS. To remove PLGA MS adhering to
MoDC surface after MoDC/MS co-incubation, pure DMSO is added to the cells and
left for exactly 3 s, after which the mixture is immediately diluted with growth medium.
This treatment with DMSO dissolves quantitatively externally adhering PLGA MS
without affecting the viability or integrity of the MoDC. After overnight incubation,
DC are collected for uptake analysis by FACS as described.
3.3. Quality Control of PLGA-MS-Loaded MoDC
3.3.1. Phenotype and Maturation of DC Loaded With PLGA MS
Immature MoDC loaded with PLGA MS can be fully matured with different DC
maturation stimuli such as LPS, CD40 ligand, poly I:C, or a cocktail of the proinflam-
matory cytokines IL-1, IL-6, TNF- (8). The phenotype of MoDC after MS loading
is characterized, in our studies, by the expression of the maturation markers CD80,
CD83, CD86, and HLA-DR. Another important DC maturation marker is the
chemokine receptor CCR7, which is up-regulated after DC stimulation in order to
enable and promote DC migration towards the lymph-node-derived chemokines
CCL19 and CCL21.
3.3.2. Survival of DC Loaded With PLGA MS

One important issue in DC vaccination is the duration of DC survival after MS-
loading. After vaccination with DC, the cells should remain alive for a sufficiently
long period to fully activate T-cells. DC viability can be determined by staining with
annexin V-FITC and propidium iodide (PI), as live cells resist annexin V and PI stain-
ing. We have observed that PLGA-MS loading does not trigger the apoptosis of imma-
ture or matured MoDC. Interestingly, the survival of matured, PLGA-MS-loaded
MoDC were even consistently improved by approx 50% after treatment with Fas mAb,
when the cells were matured with LPS (8).
3.3.3. Migration Capacity of DC Loaded With PLGA MS

The migration capacity of DC can be tested with an in vitro chemotaxis Transwell
migration assay described previously (18). Briefly, 600 L of DC medium alone or
DC medium containing either 250 ng/mL CCL19 or 250 ng/mL CCL21 are added to
the bottom chamber of 24-well Transwell plates with polycarbonate filters of 5 m
pore size. In each well, 1 105 MoDC are added to the upper chamber in a total
volume of 100 L of DC medium. The plates are then incubated for 3 h at 37C. After
careful removal of the upper chamber, 500 L of cells (migrated cells) in the bottom
chamber are collected to count for 60 s by flow cytometry. The percentage of migrated
cells can be calculated as: % Migrated DC = (number of migrated DC/total DC num-
ber) 100%.
3.3.4. T-Cell Stimulatory Functions of MoDC Loaded With PLGA MS

T-cell stimulation by MoDC can be measured by the mixed lymphocyte response
(MLR). Briefly, nave CD4+CD45RA+ T-cells are purified by negative selection using
the CD4+/CD45RO- Multi Sort Kit. Then, allogeneic MoDC are -irradiated (30 Gy)
before or after PLGA MS ingestion and cultured with CD4+CD45RA+ T-cells in 96-
well flat-bottom plates at different T-cell/DC ratios in duplicate. On d 4, 3H-thymidine
(1 Ci/well) is added and its incorporation determined after 1618 h. Culture medium
from immature MoDC and from MoDC stimulated under indicated experimental con-
ditions in the presence or absence of PLGA MS is collected for evaluating the cytokine
production pattern by TNF-, IL-10, and IL-12p70 ELISA.
3.4. MHC Class II-Restricted Antigen

Presentation by MoDC Charged With PLGA MS
One million immature MoDC are loaded with 2 g/mL of TT in solution or entrapped
in PLGA MS, and the loaded MoDC are cultured at 37C in 24-well cell-culture plates.
After overnight co-culture, the uptake of PLGA MS is checked by microscopic inspec-
tion. Then, the cells are collected and extensively washed. The DC loaded with PLGA
MS are counted and resuspended at 2 105/mL in IMDM medium. Next, 2 104 DC
(100 L/well) are seeded into 96-well flat-bottom plates in the presence of 2 g/mL LPS.
TT-specific CD4+ T-cells are added to the DC culture immediately, or 1, 3, 5, or
7 d later, for co-incubation in duplicate. After 3 d of co-culture, 3H-thymidine (1 Ci/
well) is added, and the incorporation determined after 1618 h.
3.5. MHC Class I-Restricted Antigen

Presentation by MoDC Charged With PLGA MS
One million immature MoDC are loaded with 2 g/mL of Flu M1 peptide in solution
or entrapped in PLGA MS, and MoDC are cultured at 37C in 24-well cell-culture
plates. After overnight co-culture, the uptake of PLGA MS is checked by visual inspec-
tion under a microscope. The cells are collected and extensively washed to remove free
antigen. DC loaded with PLGA MS are counted and resuspended at 2 105/mL in
IMDM medium. Aliquots of 100 L (2 104 DC) per well are seeded in 96-well round-
bottom plates in the presence of 2 g/mL LPS. Flu-specific CTL (105/well) are added
to DC culture immediately, or 1, 3, 5, or 7 d later, for co-incubation in duplicate. After
24 h co-culture, the supernatants are collected and determined for IFN- production by
ELISA.
4. Future Perspectives
In our earlier studies on immunization in animals (912), we observed that micro-
encapsulated antigens elicited strong and concomitant antibody, T helper, and CTL
responses, which emphasized the potential of PLGA MS as adjuvant. It is now evident
that immature DC can engulf a large number of PLGA MS, and that this phagocytic
process does not adversely affect DC maturation in terms of DC phenotype and capac-
ity to stimulate T-lymphocytes. Considering that PLGA particles can escape from
endo-lysosomes into the cytosol (19), we may expect that the direct targeting of
MS-formulated antigens into DC not only prolongs antigen presentation, but may also
mediate strong cross-priming of CTL in vivo. We have indeed observed that immuni-
zation of mice with bone-marrow-derived dendritic cells (BMDC) that had been loaded
with microencapsulated -galactosidase (-gal) protein induced strong CTL responses,
whereas injection of BMDC externally charged with soluble -gal failed to do so
(Y. Waeckerle-Men, manuscript in preparation). Our data, combined with a recent
report that vaccination with BMDC loaded with a fragment from HER-2 in polysac-
charide microparticles has achieved complete tumor eradication (15), indicate that the
delivery of tumor antigens via MS is clearly more potent than exogenous supply of
proteins or peptides to DC. In vivo studies have to be performed now in mice with
BMDC loaded with PLGA MS in order to evaluate the potential of this approach in a
therapeutic setup of tumor immunotherapy; the efficacy of this system will have to be
compared to that of other approaches. Given that PLGA is approved for clinical use, a
successful outcome will set the stage for testing PLGA MS in a phase I clinical trial of
tumor immunotherapy. Some efforts are still needed to produce sterile PLGA MS.
Terminal sterilization may be performed by gamma irradiation of the antigen-contain-
ing PLGA MS. However, we will have to ascertain that this treatment does not hamper
the properties of the antigen-containing PLGA MS. If gamma irradiation fails, the
aseptic production of PLGA MS in a sterile environment will be a feasible but costly
alternative.
Concerns have been raised that vaccination with autologous MoDC in general will
turn out to be too labor intensive and costly to be established as a standard therapy. It
has been argued that antigen loading of DC would be more practical in vivo. Along
this avenue, PLGA MS may turn out to be a valuable tool, as it should be possible
to co-microencapsulate chemo-attractants of DC along with the antigen. The slow
release of chemo-attractants may lead to the recruitment of immature DC, which could
ingest intradermally deposited PLGA MS. Since PLGA MS loading per se does not
lead to DC maturation, the co-encapsulation of a maturation stimulus like poyI:C or
CpG oligos should be necessary. A continuous flow of matured and antigen-loaded
DC to the draining lymph nodes could ensue, which would be an ideal stimulus for the
T-cell response. The development of PLGA MS as an in vivo antigen delivery device
for DC would be even more attractive for treating malignant diseases.
5. Notes
1. PLGA 50:50 or PLGA 75:25 polyesters are available from, e.g., Boehringer Ingelheim,
Germany, under the trade name of Resomer RG502(H) or Resomer RG752.
2. For water-soluble antigens like TT, total antigen content in MS can be determined by
dissolving 1030 mg of PLGA MS in dichloromethane, collecting the undissolved TT on
a 0.2 m cellulose membrane filter, eluting the TT from the filter with PBS, and quantify-
ing the total ultraviolet-spectrophotometrically, fluorimetrically, or by a colorimetric pro-
tein assay. To estimate the amount of antibody-reactive TT, we recommend the use of
antigen-specific ELISA. The encapsulation efficiency is calculated as the determined
amount of encapsulated antigen relative to nominal antigen content. All determinations
should be performed in triplicate.
3. Rat antihuman-CCR7 mAb clone 3D12 commercially available from BD Biosciences.
Acknowledgments
We thank Dr. Burkhard Ludewig for very useful scientific discussion and advice,
and Dr. Edith Uetz-von Allmen for experimental support. This work was funded by
Cancer League St. Gallen-Appenzell, Swiss Cancer League, Foundation Propter Hom-
ines Vaduz Liechtenstein, Cancer Research Institute, and Deutsche Krebshilfe.
References
1.
2.
2 Thurner, B., Haendle, I., Roder, C., et al. (1999) Vaccination with mage-3A1 peptide-
induces regression of some metastases in advanced stage IV melanoma. J. Exp. Med. 190,
16691678.
3.
3 Ludewig, B., McCoy, K., Pericin, M., et al. (2001) Rapid peptide turnover and inefficient
presentation of exogenous antigen critically limit the activation of self-reactive CTL by
dendritic cells. J. Immunol. 166, 36783687.
4.
4 Thomasin, C., Corradin, G., Men, Y., Merkle, H. P., and Gander, B. (1996) Tetanus toxoid
and synthetic malaria antigen containing poly(lactide)/poly(lactide-co-glycolide) micro-
spheres, importance of polymer degradation and antigen release for immune response.
J. Control. Release 41, 131145.
5.
5 Johansen, P., Men, Y., Merkle, H. P., and Gander, B. (2000) Revisiting PLA/PLGA
microspheres: an analysis of their potential in parenteral vaccination. Eur. J. Pharm. Bio-
pharm. 50, 129146.
6.
6 Men, Y., Audran, R., Thomasin, C., et al. (1999) MHC class I- and class II-restricted
processing and presentation of microencapsulated antigens. Vaccine 17, 10471056.
7.
7 Walter, E., Dreher, D., Kok, M., et al. (2001) Hydrophilic poly(DL-lactide-co-glycolide)
microspheres for the delivery of DNA to human-derived macrophages and dendritic cells.
J. Control. Release 76, 149168.
8.
8 Waeckerle-Men, Y., Scandella, E., Uetz-von Allmen, E., et al. (2004) Phenotype and func-
tional analysis of human monocyte-derived dendritic cells loaded with biodegradable
poly(lactide-co-glycolide) microspheres for immunotherapy. J. Immunol. Meth. 287(12),
109124.
9.
9 Men, Y., Thomasin, C., Merkle, H., Gander, B., and Corradin, G. (1995) A single adminis-
tration of tetanus toxoid in biodegradable microspheres elicits T cell and antibody responses
similar or superior to those obtained with aluminum hydroxide. Vaccine 13, 683689.
10.
10 Men, Y., Gander, B., Merkle, H. P., and Corradin, G. (1996) Induction of sustained and
elevated immune responses to weakly immunogenic synthetic malarial peptides by encap-
sulation in biodegradable polymer microspheres. Vaccine 14, 14421450.
11.
11 Men, Y., Tamber, H., Audran, R., Gander, B., and Corradin, G. (1997) Induction of a
cytotoxic T lymphocyte response by immunization with a malaria specific CTL peptide
entrapped in biodegradable polymer microspheres. Vaccine 15, 14051412.
12.
12 Peter K., Men, Y., Pantaleo, G., Gander, B, and Corradin, G. (2001) Induction of a cyto-
toxic T-cell response to HIV-1 proteins with short synthetic peptides and human compat-
ible adjuvants. Vaccine 19, 41214129.
13.
13 Kovacsovics-Bankowski, M., Clark, K., Benacerraf, B., and Rock, K. L. (1993) Efficient
major histocompatibility complex class I presentation of exogenous antigen upon phago-
cytosis by macrophages. Proc. Natl. Acad. Sci. USA 90, 49424946.
14.
14 Shen, Z., Reznikoff, G., Dranoff, G., and Rock, K. L. (1997) Cloned dendritic cells can
present exogenous antigens on both MHC class I and class II molecules. J. Immunol. 158,
27232730.
15.
15 Ikuta, Y., Katayama, N., Wang, L. J., et al. (2002) Presentation of a major histocompat-
ibility complex class 1-binding peptide by monocyte-derived dendritic cells incorporating
hydrophobized polysaccharide-truncated HER2 protein complex: implications for a poly-
valent immuno-cell therapy. Blood 99, 37173724.
16.
16 Audran, R., Peter, K., Dannull, J., et al. (2003) Micro-encapsulation of peptides prolongs
their presentation to cytotoxic T cells by antigen presenting cells in vitro. Vaccine 21,
12501255.
17.
17 Sah, H. 1997. A new strategy to determine the actual protein content of poly(lactide-co-
glycolide) microspheres. J. Pharm. Sci. 86, 13151318.
18. Scandella, E., Men, Y., Gillessen, S., Frster, R., and Groettrup, M. (2002) Prostaglandin
19. Panyam, J., Zhou, W. Z., Prabha, S., Sahoo, S. K., and Labhasetwar, V. (2002) Rapid
endo-lysosomal escape of poly(D,L-lactide-co-glycolide) nanoparticles: implications for
drug and gene delivery. FASEB J. 16, 12171226.
RNA Transfection of Dendritic Cells 47
4
RNA Transfection of Dendritic Cells
Frank Grnebach, Martin R. Mller, and Peter Brossart
Summary
Dendritic cells (DC) are the most powerful antigen-presenting cells that induce and main-
tain primary immune responses in vitro and in vivo. The development of protocols for the
ex vivo generation of DC provided a rationale to design and develop DC-based vaccination
studies for the treatment of infectious and malignant diseases. The efficacy of antigen loading
and delivery into DC is pivotal for the optimal induction of T-cell-mediated immune responses.
Recently it was shown that DC transfected with RNA coding for a tumor-associated anti-
gen (TAA) or whole-tumor RNA are able to induce potent antigen- and tumor-specific T-cell
responses directed against multiple epitopes. The latter technique does not require the defini-
tion of the TAA or HLA haplotype of the patients and has the potential of broad clinical appli-
cation. Such a polyvalent vaccine might be able to reduce the probability of clonal tumor escape
and to elicit CTL responses directed against naturally processed and presented immuno-
dominant tumor antigens. Additional targeting of HLA class II restricted epitopes may further
amplify and prolong the induced T-cell responses.
Key Words: Dendritic cells; RNA transfection; electroporation; CTL induction; immuno-
therapy.
1. Introduction
DC are the most powerful antigen-presenting cells, with the unique ability to ini-
tiate and maintain primary immune responses in vitro and in vivo (1,2). The develop-
ment of protocols for the ex vivo generation of DC provided a rationale to design and
develop DC-based vaccination strategies for the treatment of infectious and malignant
diseases (36). The efficacy of antigen delivery into DC is pivotal for the optimal
induction of T-cell-mediated immune responses.
Various viral and nonviral transfection methods are available for introducing DNA
into mammalian cells, depending on the purpose of the experiment. Retroviruses are
particularly useful in this respect, since their life cycle involves the stable integration
of viral DNA into the genome of infected cells, making them an important vehicle for
a broad range of applications including gene therapy. Nonviral transfection methods
include direct microinjection of DNA into the cell nucleus, incorporation of DNA into
47
48 Grnebach, Mller, and Brossart
lipid vesicles (liposomes) that fuse with the plasma membrane, and exposure of cells
to a brief electric pulse that transiently opens pores in the plasma membrane (electro-
poration). However, plasmid DNA transfer into DC is not very efficient and the use of
viral vectors for transduction of DC implies a more complex and laborious manipula-
tion associated with safety issues. Moreover, DC transfected by viral vectors have
been reported to have impaired capacity in stimulating T-lymphocytes (79).
In contrast, it was recently shown that DC transfected with RNA coding for a TAA
or whole-tumor RNA are able to induce potent antigen- and tumor-specific T-cell
responses directed against multiple epitopes (1012). We analyzed the efficiency of
RNA transfection into DC using different sources of RNA as well as applying differ-
ent delivery strategies, including electroporation, lipofection, and CD71 receptor-
based endocytosis, and found that electroporation was most effective (13).
Electroporation is a technique used to permeabilize cell membranes by means of an
electric discharge applied directly to living cells. This permeabilization is probably
due to the creation of pores in the cell membrane. Whenever a cell is exposed to the
action of an electrical field the components of its membrane polarize until the estab-
lishment of a transmembranous difference of potential. If the potential difference pro-
duced in this way reaches a certain threshold (variable according to the cell type), the
membrane will be disorganized or degraded at various spots, and the cell will become
permeable to different exogenous molecules (14). Nevertheless, the actual mechanism
is still not fully understood.
To evaluate the sensitivity of these different approaches, we utilized in vitro RNA
transcripts from a vector coding for the enhanced green fluorescence protein (EGFP)
and RNA isolated from a cell line that was transfected with an EGFP plasmid. EGFP is
a GFP variant that contains chromophore mutations that make the protein 35 times
brighter than the wild-type and that is codon optimized for higher expression in mam-
malian cells. It is readily detectable using techniques of fluorescence microscopy, flow
cytometry, or macroscopic imaging (15,16).
2. Materials
1. Ficoll Paque Plus, density 1.077 g/mL (Amersham Biosciences, Freiburg, Germany).
2. BD Falcon 6-well Multiwell plate. Tissue-culture treated polystyrene, flat bottom,
with low-evaporation lid (BD Biosciences, Heidelberg, Germany).
3. PBS: phospate-buffered saline (Invitrogen, Karlsruhe, Germany).
4. PBS/FCS: PBS containing 0.5% heat-inactivated fetal calf serum (FCS).
5. RPMI Medium 1640.
6. RP10 medium: RPMI 1640 supplemented with 10% heat-inactivated FCS, 2 mM L-
glutamine, 50 M 2-mercaptoethanol, and antibiotics streptomycin (100 g/mL) and peni-
cillin (100 U/mL).
7. X-VIVO 20 medium (Cambrex Bio Science, Apen, Germany) .
8. Granulocyte/macrophage colony-stimulating factor (GM-CSF; Leukomax; Novartis,
Nrnberg, Germany).
9. Interleukin (IL)-4 (Genzyme, Cambridge, MA).
10. Tumor necrosis factor (TNF)- (Genzyme).
11. Propidium iodide (PI) staining solution: 50 g PI/mL in PBS (pH 7.4) (BD Biosciences).
Fig. 1. Process flow of in vitro generation of dendritic cells (DC) from adherent peripheral
blood mononuclear cells (PBMC) and RNA electroporation. IL-4, interleukin 4; GM-CSF,
granulocyte macrophage-colony stimulating factor; TNF-, tumor necrosis factor-; CTL,
cytotoxic T-lymphocytes; EGFP, enhanced green fluorescence protein.
12. FACSCalibur flow cytometer (BD Biosciences).

13. RNeasy anion-exchange spin columns (Qiagen, Hilden, Germany).
14. 10X MOPS-EDTA-sodium acetate buffer (MESA buffer): 3-[N-morpholino]propanesul-
fonic acid (MOPS), 0.4 M, pH 7.0, 0.1 M sodium acetate, 10 mM EDTA.
15. Agarose-LE, RNase-free (Ambion, Huntingdon, UK).
16. 1% Ethidium bromide stock solution (10 mg/mL).
17. RNA sample loading buffer: 62.5% deionized formamide, 1.14 M formaldehyde, 1.25 X
MOPS-EDTA-sodium acetate buffer, 200 g/mL bromphenol blue, 200 g/mL xylene
cyanole, 50 g/mL ethidium bromide.
18. Plasmid pSP64-EGFP-2 (V. F. I. Van Tendeloo).
19. Restriction enzyme NdeI .
20. QIAEX II Gel Extraction Kit (Qiagen).
21. 10 mM Tris-Cl, pH 8.5.
22. SP6 Cap Scribe Kit (Roche Applied Science, Mannheim, Germany).
23. RNA Millenium Markers (Ambion).
24. Easyject Plus electroporator (EquiBio; Peqlab, Erlangen, Germany).
25. 4-mm electroporation cuvets.
3. Methods
The methods described below outline (1) cell isolation and generation of DC from
adherent peripheral blood mononuclear cells (PBMC), (2) the RNA isolation from
animal cells and formaldehyde agarose gel electrophoresis, (3) the generation of EGFP
mRNA by in vitro transcription, (4) the electroporation of DC, and (5) the detection of
EGFP fluorescence by flow cytometry (Fig. 1).
3.1. Cell Isolation and Generation of Human DC

From Adherent Peripheral Blood Mononuclear Cells (PBMC)
1. Dilute fresh human heparinized blood by addition of an equal volume of PBS. Alterna-
tively, leukocyte-rich buffy coat not older than 8 h is diluted 1:4 with PBS.
2. Pipet 12 mL Ficoll-Paque Plus in a 50-mL sterile conical centrifuge tube.
3. Carefully layer the diluted blood sample on Ficoll-Paque Plus.
4. Centrifuge at 650g for 20 min at 1820C in a swing-out rotor without brake.
5. After centrifugation the PBMC (lymphocytes and monocytes) form a distinct band at the
sample/medium interface. Carefully transfer the interphase cells to a new 50-mL conical
tube. The cells are best removed from the interface using a sterile plastic Pasteur pipet
without removing the upper layer.
6. Wash the cells twice or thrice: add PBS/FCS (see Note 1) to 50 mL, mix and centrifuge
120g for 10 min at 1820C (see Note 2).
7. Resuspend the cell pellet in RP10 medium (see Note 3).
8. Seed the cells at a concentration of 1 107 cells/3 mL/well into 6-well plates.
9. Incubate for 2 h in a humidified 37C/5% CO2 incubator for adherence of monocytes.
10. Gently shake and remove medium containing non-adherent lymphocytes.
11. Wash the adherent monocytes three times with 3 mL PBS (see Note 4).
12. Add RP10 medium supplemented with human recombinant GM-CSF (100 ng/mL) and
IL-4 (20 ng/mL) for the generation of immature DC (see Note 3). Every other d add
cytokines.
13. Culture cells in a humidified 37C/5% CO2 incubator for 6 d. After 6 d the loosely adherent
or nonadherent cells should display immature DC morphology and phenotype (CD14 +/,
CD1a++, CD40+, CD83, CD86+, HLA-DR++).
14. Harvest the immature monocyte-derived DC and electroporate with RNA.
15. For maturation, culture DC with TNF- (10 ng/mL) for an additional 24 h (after d 6).
16. Analyze the phenotype of DC by flow cytometry after 7 d of culture: CD14, CD1a+,
CD80+, CD83+, CD86++, HLA-DR+++.
3.2. Total RNA Isolation From Human Cells

and Formaldehyde Agarose Gel Electrophoresis
Total RNA is isolated from cell lysates using RNeasy Midi or Maxi anion-
exchange spin columns according to the protocol for isolation of total RNA from ani-
mal cells provided by the manufacturer (see Note 5). To avoid exogenous sources of
RNase contamination, always apply RNase-free techniques when working with RNA
(see Note 6).
Quantity and purity of RNA is determined by measuring the absorbance at 260 nm
in a spectrophotometer. The ratio of the readings at 260 nm and 280 nm provides an
estimate of the purity of RNA with respect to contaminants that absorb in the ultravio-
let (UV), such as protein. Pure RNA has an A260/A280 ratio of 1.92.1. Furthermore,
RNA samples are routinely checked by formaldehyde/agarose gel electrophoresis for
size and integrity and stored at 80C in small aliquots.
Preparation of 1.2% formaldehyde/agarose gel of size 10 14 0.7 cm:
1. Add 1.2 g RNase-free agarose, 10 mL 10X MESA buffer, and RNase-free water to 100 mL.
2. Heat the mixture to melt agarose and cool to 65C.
3. Add 1.8 mL of 37% (12.3 M) formaldehyde and 1 L of a 10 mg/mL ethidium bromide
stock solution.
Fig. 2. (A) 1.2% formaldehyde agarose gel of total human RNA showing the 18S and 28S
ribosomal RNA (rRNA). (B) 1.2% formaldehyde agarose gel of enhanced green fluorescence
protein in vitro transcript (EGFP-IVT). M, RNA Millenium Markers (Ambion).
4. Pour the mixture into the gel tray.

5. Equilibrate the gel prior to running in 1X MESA buffer for at least 30 min.
6. Add RNA sample loading buffer to the RNA sample in ratio 1:2 to 1:5. The RNA marker
is treated equally.
7. Heat the RNA samples just before loading onto the gel to 65C for 10 min and chill on
ice.
8. Run the gel at < 5 V/cm in 1X MESA buffer. When resolved, the 18S and 28S ribosomal
RNA (rRNA) bands should appear as sharp bands on the stained gel. 28S rRNA bands
should be present with an intensity approximately twice that of the 18S rRNA band, indi-
cating that no degradation of RNA has occurred (Fig. 2A).
3.3. Generation of EGFP mRNA by In Vitro Transcription

For generation of EGFP mRNA, the plasmid pSP64-EGFP-2 (see Note 7), which
allows in vitro transcription under the control of the phage SP6 promoter, is used
(Fig. 3) (16). The plasmid is linearized with the restriction enzyme NdeI (see Notes 8
and 9) and then purified with the QIAEX II Gel Extraction Kit using the protocol for
desalting and concentrating DNA solutions (see Note 10). In vitro transcription is
done with the SP6 Cap Scribe kit according to the protocol provided by the manufac-
turer, except that 2.0 g linearized plasmid DNA is used per transcription reaction
and the synthesis time is extended to 3 h. Purification of in vitro transcripts is per-
formed with RNeasy Mini anion-exchange spin columns according to the RNA
Fig. 3. Schematic drawing of plasmid pSP64-EGFP-2 for in vitro transcription of EGFP.

EGFP, enhanced green fluorescence protein; ori, origin of replication; AmpR, -lactamase cod-
ing region.
cleanup protocol provided by the manufacturer. RNA in vitro transcripts (IVT) are
routinely checked by formaldehyde/agarose gel electrophoresis for correct size and
integrity, and stored at 80 in small aliquots (Fig. 2B).
3.4. Electroporation of DC
Electroporation is done on d 6 of DC generation (Fig. 1).
1. Prepare 6-well plates with 3 mL RP10 medium supplemented with cytokines and pre-
warm in the incubator (see Note 3).
2. Wash immature DC twice with serum-free X-VIVO 20 medium and resuspend to a final
concentration of 2 107 cells/mL.
3. Mix 200 L of the cell suspension with 10 g RNA (2.5 g RNA/106 DC, see Note 11)
and transfer into a 4-mm electroporation cuvet (see Note 12). EGFP in vitro transcribed
RNA is recommended as a positive control.
4. Adjust the parameters of the electroporator apparatus as follows: voltage of 300 V, capaci-
tance of 150 F, and resistance of 1540 .
5. Insert the cuvet into the electroporation chamber and trigger the pulse.
6. Transfer the cells immediately into the 6-well plate with prewarmed RP10 medium and
return it to the incubator (see Note 13).
7. Treat the control cells equally but use PBS instead of RNA.
3.5. EGFP Analysis by Flow Cytometry

EGFP-RNA-transfected DC are checked for EGFP expression 2430 h post
electroporation by flow cytometric analysis. Briefly, 2 105 cells are washed once
with 5 mL PBS and are then resuspended in 200 L of PBS/FCS buffer. Gating is
performed on cells exhibiting a large forward scatter (FSC) and side scatter (SSC)
profile in order to allow exclusion of contaminating autologous lymphocytes. Gated
DC are then evaluated for EGFP expression (excitation maximum = 488 nm; emission
maximum = 507 nm). As a control, nontransfected DC (PBS control) are used.
The vitality of the cells after electroporation can be assessed by propidium iodide
(PI) staining (see Note 14). 10 L of PI staining solution are added to the cell sample
in ratio 1:10. PI is detected in the orange range of the spectrum using a 562588 nm
band-pass filter.
4. Notes
1. For clinical use, FCS can be omitted.
2. The washings remove thrombocytes, which are not spun down at this centrifugal force.
3. For clinical use the RP10 medium can be replaced with serum-free X-VIVO 20 medium.
4. In our experience the purity of the adherent monocytes after the wash steps varies consid-
erably from donor to donor.
5. RNeasy kits provide a fast and simple method for preparation of total RNA from animal
cells and tissues, but other methods producing high-quality RNA can be applied.
6. See manuals of the RNA isolation kits.
7. An in vitro transcription vector can be constructed by subcloning the coding sequence of
any gene into the vector pSP64 Poly(A) (Promega) by standard cloning procedures (17).
For linearization of the vector, a singular restriction site should be present approx 50200
bp downstream of the termination codon of the respective gene. If this is not the case, an
adequate restriction site must be created by site-directed mutagenesis.
8. Plasmid DNA must be linearized with a restriction enzyme downstream of the insert to be
transcribed, since linearized templates give transcripts of defined length and usually
higher yields of transcripts. Circular plasmid templates will generate extremely long, het-
erogeneous RNA transcripts.
9. Remove an aliquot for an analytical gel in order to confirm that cleavage is complete.
10. Plasmid DNA from some isolation procedures may be contaminated with residual
RNase A. The plasmid template used for transcription should be sufficiently purified
and not contain high amounts of bacterial RNA or RNases. The QIAEX II Gel Extrac-
tion Kit avoids phenol/chloroform extraction and ethanol precipitation. The DNA is
eluted with an appropriate volume of 10 mM Tris-Cl, pH 8.5, to achieve a concentration
of 0.51.0 g/L.11. The preparation can be downscaled, but keep the ratio between
RNA and cells of 2.5 g /106.
12. The RNA must be added to the cell suspension shortly before the pulse to prevent deg-
radation due to RNases.
13. During the electroporation pulse, electrolysis of the medium and electrodes is a major
event. This electrolysis strongly and unfavorably affects the characteristics of the medium,
notably the pH. Therefore it is absolutely advisable to transfer the electroporated cells to
fresh medium as soon as possible after the pulse (within seconds); failure to do so will
considerably increase the mortality rate.
14. PI intercalates into double-stranded nucleic acids. It is excluded by viable cells but can
penetrate cell membranes of dying or dead cells.
Acknowledgments
The authors thank Dr. Viggo F.I. Van Tendeloo (University of Antwerp, Antwerp,
Belgium) for providing the vector pSP64-EGFP-2. This work was supported in part by
a grant from the Deutsche Forschungsgemeinschaft (SFB 510, Project B2) and the
fortne-program of the University of Tbingen.
References
1.
1 Banchereau, J. and Steinman, R. M. (1998) Dendritic cells and the control of immunity.
Nature 392, 245252.
2.
2 Cella, M., Sallusto, F., and Lanzavecchia, A. (1997) Origin, maturation and antigen pre-
senting function of dendritic cells. Curr. Opin. Immunol. 9, 1016.
3.
3 Sallusto, F. and Lanzavecchia, A. (1994) Efficient presentation of soluble antigen by cul-
tured human dendritic cells is maintained by granulocyte/macrophage colony stimulating
factor plus interleukin 4 and down regulated by tumour necrosis factor. J. Exp. Med. 179,
11091118.
4.
4 Zhou, L. and Tedder, T. F. (1996) CD14 blood monocytes can differentiate into function-
ally mature CD83+ dendritic cells. Proc. Natl. Acad. Sci. USA 93, 25882592.
5.
5 Brossart, P., Grnebach, F., Stuhler, G., et al. (1998) Generation of functional human
dendritic cells from adherent peripheral blood monocytes by CD40 ligation in the absence
of granulocyte-macrophage colony-stimulating factor. Blood 92, 42384247.
6.
6 Brossart, P., Wirths, S., Brugger, W., and Kanz, L. (2001) Dendritic cells in cancer vac-
cines. Exp. Hematol. 29, 12471255.
7.
7 Brossart P., Goldrath, A. W., Butz, E. A., Martin, S., and Bevan, M. J. (1997) Virus-
mediated delivery of antigenic epitopes into dendritic cells as a means to induce CTL. J.
Immunol. 158, 32703276.
8.
8 Reeves, M. E., Royal, R. E., Lam, J. S., Rosenberg, S. A., and Hwu, P. (1996) Retroviral
transduction of human dendritic cells with a tumor-associated antigen gene. Cancer Res.
56, 56725677.
9 Van Tendeloo, V. F., Snoeck, H. W., Lardon, F., et al. (1998) Nonviral transfection of
9.
distinct types of human dendritic cells: high-efficiency gene transfer by electroporation
into hematopoietic progenitorbut not monocyte-derived dendritic cells. Gene Ther. 5,
700707.
10.
10 Boczkowski, D., Nair, S. K., Snyder, D., and Gilboa, E. (1996) Dendritic cells pulsed with
RNA are potent antigen-presenting cells in vitro and in vivo. J. Exp. Med. 184, 465472.
11.
11 Nair, S. K., Boczkowski, D., Morse, M., Cumming, R. I., Lyerly, H. K., and Gilboa, E.
(1998) Induction of primary carcinoembryonic antigen (CEA)-specific cytotoxic T lym-
phocytes in vitro using human dendritic cells transfected with RNA. Nat. Biotechnol. 16,
364369.
12. Mller, M. R., Grnebach, F., Nencioni, A., and Brossart, P. (2003) Transfection of den-
dritic cells with RNA induces CD4- and CD8-mediated T cell immunity against breast
carcinomas and reveals the immunodominance of presented T cell epitopes. J. Immunol.
170, 58925896.
13. Grnebach, F., Mller, M. R., Nencioni, A., and Brossart, P. (2003) Delivery of tumor-
derived RNA for the induction of cytotoxic T-lymphocytes. Gene Ther. 10, 367374.
14 Andreason, G. L. and Evans, G. A. (1989) Optimization of electroporation for transfection
14.
of mammalian cell lines. Anal. Biochem. 180, 269275.
15.
15 Zhang, G., Gurtu, V., and Kain, S. R. (1996) An enhanced green fluorescent protein
allows sensitive detection of gene transfer in mammalian cells. Biochem. Biophys. Res.
Commun. 227, 707711.
16. Van Tendeloo, V. F., Ponsaerts, P., Lardon, F., et al. (2001) Highly efficient gene delivery
by mRNA electroporation in human hematopoietic cells: superiority to lipofection and
passive pulsing of mRNA and to electroporation of plasmid cDNA for tumor antigen load-
ing of dendritic cells. Blood 98, 4956.
17. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory
Manual, 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
DC Isolation and Generation by CliniMACS 55
5
Isolation and Generation of Clinical-Grade
Dendritic Cells Using the CliniMACS System
John D. M. Campbell, Christoph Piechaczek,

Gregor Winkels, Edith Schwamborn, Daniela Micheli,
Sonja Hennemann, and Jrgen Schmitz
Summary
Dendritic cells (DC) can either be generated from progenitors such as stem cells or CD14+
monocytes, or isolated directly from the blood. Blood-derived DC are present as at least two
distinct populationsmyeloid and plasmacytoid DC. Here we describe methods for the clini-
cal-grade isolation of blood DC and DC precursors using the CliniMACS. We describe the
isolation of ultra-pure monocytes in order to generate large numbers of monocyte-derived DC,
and also new methods for the direct isolation of blood DC. Isolation of blood DC in large
numbers means that natural DC with different properties can be investigated for their clinical
function for the first time.
Key Words: Dendritic cells; myeloid DC; plasmacytoid DC; magnetic separation;
CliniMACS.
1. Introduction
DC are the most efficient antigen-presenting cells (APC) of the immune system.
DC are present at low levels in most tissues, acting as sentinels that patrol the bound-
aries against foreign enemies, such as infectious agents, toxins, and allergens (1). The
potent antigen (Ag)-presenting ability of the DC is currently of great interest in the
clinical manipulation of immune responses. Vaccination with DC, particularly in can-
cer and viral-induced disease, may lead to the establishment of Ag-specific immune
responses in the patient, which ultimately may eliminate the pathogen or tumor.
In humans, DC have been identified that develop from at least two different pro-
genitor cell typesDC of myeloid origin (mDC, or DC1), and DC of plasmacytoid
origin (pDC, or DC2) (2,3). There is also heterogeneity within each DC group (2,3).
The mDC and pDC subsets react to differing maturation signals, and produce different
cytokines in response to stimulus. DC are present in relatively small numbers in the
bloodstream (less than 1% of mononuclear cells) (4), and until recently, no suitable
55
56 Campbell et al.
specific markers were available to isolate these cells. DC for therapy were therefore
generated in large numbers from precursor cells, such as CD34+ cells or CD14+ mono-
cytes (5,6). Using CD14 selection on the CliniMACS device, it is possible to isolate
very large numbers of precursors for the generation of monocyte-derived DC (MoDC)
for therapy. MoDC are analogous to mDC, and have already been successfully used in
early clinical trials.
We have also now generated novel reagents to isolate blood DC for therapy, using
the antibodies BDCA-1 (CD1c) and BDCA-4 to isolate mDC and pDC respectively.
This means that for the first time, the potent Ag-presenting abilities of natural DC can
be exploited for clinical use, and the specific functions of mDC and pDC (710) can
be investigated in clinical trials.
In this chapter we detail the three methods for isolation of DC for therapydirect
isolation of blood mDC and pDC, and the purification of CD14+ cells for generation
of MoDC. The application of CD14-derived DC for vaccination of tumor patients is
described elsewhere in Chapter 6 of this volume. Here we will describe:
1. The isolation of ultra-pure CD14+ cells for MoDC generation.
2. The optimal culture conditions for the generation of large numbers of MoDC.
3. The isolation of mDC and pDC.
Each method is described separately, as the specific materials required differ for
each isolation.
2. Materials
2.1. Clinical-Scale Separation of CD14+ Monocytes
1. Leukapheresis product from unmobilized donor containing 1020 109 leukocytes (white
blood cells [WBC]).
2. CliniMACS CD14 Reagent (Miltenyi Biotec, cat. no. 281-01).
3. CliniMACS plus Instrument (Miltenyi Biotec, cat. no.155-01); software version 2.3.1.
4. CliniMACS Tubing Set (Miltenyi Biotec, cat. no.161-01).
5. PreSystem-Filter (Miltenyi Biotec, cat. no.181-01).
6. Luer/Spike Interconnector, (Miltenyi Biotec, cat. no.187-01).
7. CliniMACS PBS/EDTA Buffer (Miltenyi Biotec, cat. no. 700-25).
8. Human serum albumin (HSA) (e.g., Aventis Behring).
9. Transfer Packs 600 mL (Miltenyi Biotec, cat. no. 190-01).
10. Transfer Set Coupler/Coupler (Miltenyi Biotec, cat. no.186-01).
11. Centrifuge, suitable for bag processing.
12. Balance for weighing bags.
13. Plasma extractor.
14. Orbital shaker.
15. Sampling Site Coupler (Miltenyi Biotec, cat. no.189-01).
16. Tubing slide clamps/scissor clamps.
17. Sample tubes, size 2 mL.
18. CD14-FITC (e.g., Miltenyi Biotec GmbH, cat. no. 130-080-701).
19. CD15-PE (e.g., BD-Pharmingen).
20. Propidium iodide solution (1 mg/mL).
21. Ammonium chloride red blood cell lysing solution (final concentration: 155mM NH4Cl; 10
mM KHCO3; 0.1 mM EDTA, pH 7.3). Make fresh before use from a 10X stock solution.
2.2. Generation of Immature Dendritic Cells From Monocytes

1. Cell culture bags (e.g., CellGenix) or flasks (e.g., NUNC).
2. 50- or 250-mL conical centrifuge tubes (e.g., Falcon).
3. Syringes.
4. Medium, e.g., X-VIVO 15 (BioWhittaker).
5. Human AB serum (e.g., Sigma).
6. Granulocyte/macrophage colony-stimulating factor (GM-CSF; e.g., Leucomax, Sandoz).
7. Interleukin (IL)-4 (e.g., R&D Systems).
8. DC generation mediumX-VIVO 15 medium supplemented with 1% AB serum; 1000
IU/mL GMCSF; 1000 IU/mL IL-4.
9. Trypan blue solution for viability testing.
10. Biological safety cabinet.
11. CO2 incubator with 5 % CO2.
12. Variable volume pipettors.
13. Microscope.
14. Anti-CD14-FITC (e.g., Miltenyi Biotec GmbH).
15. Anti-CD80-FITC (e.g., BD-Pharmingen).
16. Anti-CD83-PE (e.g., BD-Pharmingen).
17. Anti -CD86-PE (e.g., BD-Pharmingen).
18. Anti -HLA-DR-FITC (e.g., BD-Pharmingen).
19. Isotype controls.
20. Propidium iodide (1 mg/mL).
2.3 Blood DC
2.3.1 Plasmacytoid Dentritic Cells (pDC)
1. Leukapheresis product from unmobilized donor containing 1040 109 WBC.
2. CliniMACS Anti-BDCA-4 MicroBeads (Miltenyi Biotec GmbH, cat. no. 291-01).
3. CliniMACS plus instrument (Miltenyi Biotec GmbH , cat. no. 155-01 or 155-02), software
version 2.3.1+.
4. CliniMACS Tubing Set (Miltenyi Biotec GmbH, e.g., cat. no. 165-01 or 161-01).
5. CliniMACS PBS/EDTA buffer (P/E; 3 1L; Miltenyi Biotec, ref. no. 700-25).
6. Pre-System Filter (Miltenyi Biotec GmbH, cat. no. 181-01).
7. Luer/Spike Interconnector (Miltenyi Biotec GmbH, cat. no. 187-01).
8. Transfer packs, 150 mL and 600 mL.
9. Centrifuge, suitable for bag processing, bags to balance centrifuge.
11. Sterile tubing welder.
12. Orbital shaker.
13. Sampling site coupler.
15. Flow cytometer (properly equipped and compensated to allow simultaneous analysis of
FITC, PE, and PI. If equipped with a 633-nm laser, additionally APC).
16. Automated hematology analyzer or Neubauer hemocytometer.
17. Human serum albumin (HSA) or bovine serum albumin (BSA) as supplement to P/E
(final concentration 0.5%, P/E/H or P/E/B) (see Note 7).
18. Ammonium chloride lysing solution (155 mM NH4Cl, 10 mM KHCO3, 0.1 mM EDTA).
19. FcR-Blocking Reagent (Miltenyi Biotec, ref. no. 130-059-901). Titer 1:5.
58 Campbell et al.
20. CD45.FITC (Miltenyi Biotec, ref. no. 130-080-202). Titer 1:11.

21. Anti-BDCA-2.PE (Miltenyi Biotec, ref. no. 130-090-511). Titer 1:11.
22. CD123.APC (Miltenyi Biotec ref. no. 130-090-901). Titer 1:11.
23. Propidium iodide, stock: 1mg/mL (Sigma, ref. no. P4170). Titer 1:1000.
24. 12 75 mm (4 mL) tubes (Falcon-2052 or equivalent).
2.3.2 Myeloid Dentritic Cells (mDC)

1. Leukapheresis product from unmobilized donor containing 1040 109 WBC.
2. CliniMACS CD1c (Anti-BDCA-1)-Biotin (Miltenyi Biotec GmbH, CliniMACS Anti-
Biotin MicroBeads, ref. no. 192-01).
3. CliniMACS CD19 (MicroBeads, ref. no. 193-01).
4. CliniMACS plus instrument (Miltenyi Biotec GmbH , cat. no. 155-01 or 155-02), software
version 2.3.1+.
5. 2X CliniMACS Tubing Set (Miltenyi Biotec, e.g., ref. no. 165-01 or 161-01 or 162-01).
6. 2X CliniMACS PBS/EDTA buffer (3 1L; Miltenyi Biotec, e.g., ref. no. 700-25 or
705-25).
7. 2X Pre-System Filter (Miltenyi Biotec, ref. no. 181-01).
8. 2X Luer/Spike Interconnector (Miltenyi Biotec, ref. no. 187-01).
9. Transfer packs, 150 mL and 600 mL.
10. Centrifuge suitable for bag processing, bags to balance centrifuge.
12. Sterile-tubing welder.
13. Orbital shaker.
14. Sampling site coupler.
16. Flow cytometer (properly equipped and compensated to allow simultaneous analysis of
FITC, PE, and PI. If equipped with a 633-nm laser, additionally APC).
17. Automated hematology analyzer or Neubauer hemocytometer.
18. Human serum albumin (HSA) or bovine serum albumin (BSA) as supplement to P/E
(final concentration 0.5%, P/E/H or P/E/B) (see Note 7).
19. Ammonium chloride lysing solution (155 mM NH4Cl, 10 mM KHCO3, 0.1 mM EDTA).
20. FcR-Blocking Reagent (Miltenyi Biotec, ref. no. 130-059-901). Titer 1:5.
21. CD1c.PE (Clone AD5-8E7; Miltenyi Biotec GmbH 130-090-508). Titer 1:11.
22. CD20.APC (Clone LT20; Miltenyi Biotec GmbH). Titer 1:11.
23. CD45.FITC (Miltenyi Biotec, ref. no. 130-080-202). Titer 1:11.
24. Propidium iodide, stock: 1 mg/mL (Sigma, ref. no. P4170). Titer 1:1000.
25. 12 75 mm FCM tubes (Falcon-2052 or equivalent).
3. Methods
3.1. Isolation of CD14+ Cells for the Generation of Immature MoDC
In this section we describe the separation of CD14+ monocytes; the cultivation pro-
cedure for generation of immature DC (iDC) from monocytes; flow cytometry analy-
sis of the CD14 separation and iDC.
3.1.1. Isolation of CD14+ Cells Using the CliniMACS
Isolation of CD14+ monocytes is performed by immunomagnetic labeling of target
cells followed by an automated separation process using the CliniMACSplus.
3.1.1. Immunomagnetic Labeling

Determine the total number of leukocytes in your sample. This protocol is designed
for leukapheresis products containing 1020 109 WBC (see Notes 13). Before start-
ing, it is useful to determine the number of monocytes in the sample (either by auto-
mated hematology analyzer or flow cytometry), as this number of target cells must be
entered into the CliniMACS cell-separation software before separation. Supplement
CliniMACS PBS/EDTA buffer with HSA (final concentration 0.5%) before starting.
The HSA is required to inactivate and to coat the surfaces of tubing and bags.
1. Determine the weight of an empty 600-mL transfer pack to be used as a cell-preparation
bag. Transfer the leukapheresis product into the cell-preparation bag and record the weight.
2. Fill the bag with HSA-supplemented CliniMACS PBS/EDTA buffer to a total volume of
600 mL and centrifuge the cell suspension for 15 min at 300g (room temperature, 19
25C), no brake.
3. Remove supernatant carefully using the plasma expressor to a final sample volume of
50 mL, taking care not to resuspend the cell pellet.
4. Resuspend cell pellet and add the entire content of one vial CliniMACS CD14 Reagent
(7.5 mL), and mix carefully (see Note 4).
5. Incubate the cell-preparation bag for 15 min at room temperature (1925C) on an orbital
shaker at 25 rpm maximum.
6. Add HSA-supplemented CliniMACS PBS/EDTA buffer to a final volume of 600 mL for
cell washing and centrifuge for 15 min at 300g (room temperature), no brake.
7. Remove as much supernatant as possible from the cell pellet, resuspend the cells, and add
CliniMACS PBS/EDTA/HSA buffer to a final volume of 100 mL.
8. Retain a 0.5-mL sample from the cell suspension and place into a sample tube for analy-
sis. It is recommended to determine at least total cell number, viability, and frequency/
number of CD14+ cells.
3.1.2. Automated Separation
1. Determine the weight of an empty 600-mL transfer pack with a Luer/Spike Interconnector
inserted and attach this bag to the cell collection port of the CliniMACS Tubing Set.
Ensure that all slide clamps are open. Switch on the instrument and choose selection
program ENRICHMENT 1.1.
2. Confirm your choice and enter the parameters requested on the screen. Check the calcu-
lated number of labeled total and target cells and confirm by pressing ENT. Check
buffer and bag requirements and confirm displayed data after preparation of tubing set
and buffer bags as mentioned above.
3. Follow the tubing-set installation instructions on the screen and start the automated sepa-
ration program.
4. Determine the weight of the cell collection bag after the separation and take a 0.5-mL
sample for cell-count determination and flow-cytometry analysis. The weight can be used
to determine the volume of the positive fraction.
5. Take a 0.5-mL sample from the negative fraction collection bag for cell-count determina-
tion and flow-cytometry analysis.
3.1.3. Analysis of CD14 Separation (Fig. 1)

Analysis of all four fractions obtainedoriginal, positive, negative and buffer waste
bagsallows the most accurate determination of yield (see Notes 5 and 6).
60 Campbell et al.
Fig. 1. Determination of CD14+ cell purity and yield. Representative flow cytometry analy-
sis of a leukapheresis sample before and after CliniMACS CD14 separation. Samples are first
gated on Forward vs Side Scatter (FSC vs SSC) to exclude any debris (R1). Results from R1 are
then plotted CD15-PE vs PI to exclude dead cells (R2). A combined gate is then made (e.g., G3
= R1*R2 in Cell Quest software, BD biosciences). G3 therefore contains all viable cells.
CD14 is plotted against CD15 to evaluate the numbers of live CD14+CD15 cells (viable mono-
cytes) in the samples (R3). In this example, the start population has 13.2% CD14+CD15 cells,
which is enriched to >99% after CliniMACS sorting.
To calculate yield:
total WBC * % viaable monocytes ( purity )
Total, viable monocytes =
100%
Example:
Total cell number positive fraction: 1.9 109
Viability of positive fraction = 99.9% (G3)
Purity of positive fraction = 99.3% (R3)
Percentage viable monocytes = 99.9% * 99.3% = 99.2%
1.9 10 9 * 99.2%
Total, viable monocytes = = 1.885 10 9 viable CD14+ monocytes
100%
1. Determine the cell count from the samples and ensure that the total amount of WBC is no
greater than 1 106/100 L. If necessary, dilute the samples with PBS/EDTA/0.5% HSA.
2. Determine the viability of each fraction either using trypan blue exclusion or propidium
iodide exclusion during the flow-cytometry analysis (recommended).
3. Label four tubes as follows: (1) original fraction; (2) positive fraction; (3) negative frac-
tion; and (4) wash fraction. Place 100 L of cell suspension in each tube.
4. Add 20 L of CD14-PE and 10 L of CD15-FITC to each tube.
5. Incubate the tubes for 10 min in the dark at 28C.
6. Add 1 mL of ammonium chloride red blood cell lysing solution and incubate for 10 min at
room temperature.
7. Spin down the cells for 5 min at 300g (10g) at room temperature.
8. Wash cells with 1 mL of PBS/EDTA/0.5% HSA. Spin down the cells for 5 min at 300g
(10g) at room temperature. Remove supernatant completely.
9. Resuspend cells in an adequate volume of buffer for flow-cytometric analysis according
to routine laboratory practice. To assess cell viability and to exclude the effects of non-
specific binding of antibodies by dead cells, add propidium iodide (PI) in a concentration
of 1g/mL to each tube immediately prior to acquisition.
10. Acquire at least 10,000 events per samplesee Fig. 1 for detailed gating strategy.
3.2. Immature Monocyte-Derived Dentritic Cells (MoDCs)
3.2.1. Generation of Immature DC From CD14+ Cells
1. Calculate the required volume of medium to achieve a final concentration of 2 106
CD14+ cells per mL.
2. Transfer the CD14+ cells of the positive fraction to centrifuge tubes or a 150-mL collec-
tion bag and centrifuge for 10 min at 300g without brake.
3. Freshly prepare DC generation medium.
4. Resuspend cells in DC generation medium, prewarmed to 37C, at a density of 2 106/
mL and transfer them into culture bags or flasks. Incubate cells in a CO2 incubator with
5% CO2 and 37C.
After 72 h:
5. Prepare fresh DC generation medium.
Add a volume of DC generation medium, prewarmed to 37C, to each flask equal to
one third of the original culture volume. Replace cells in incubator.
After 120 h (5 d):
6. Repeat step 5.
After 7 d:
7. Gently mix cultures and pool their contents into centrifuge tubes or centrifugable bags. If
cells remain in culture vessel, add 10 mL PBS/EDTA buffer, prewarmed to 37C, into
each flask or bag and incubate for 5 min at 37C. Add all supernatants into centrifuge
tubes.
8. Centrifuge for 10 min at 300g without brake at room temperature; remove supernatant.
Pool cells into a single tube and suspend with PBS/EDTA solution. Remove a small
sample for cell count, viability testing, and flow-cytometry analysis.
3.2.2. Flow-Cytometry Analysis of Immature Dendritic Cells (Fig. 2)
A minimum suggested panel of Ab to evaluate the generation of DC from CD14+
cells is anti-CD14, anti-HLA-DR, anti-CD86, anti-CD80, and anti-CD83. Immature
62 Campbell et al.
Fig. 2. Generation of iDC from CD14+ cells. Representative flow cytometry analysis of
freshly isolated CD14+ monocytes, and the same cells after 7-d culture with granulocyte-mac-
rophage colony-stimulating factor and interleukin-4. The cells rapidly lose expression of CD14
(in this case from a mean fluorescence of 2850 for the monocytes reduced to only 5 for the
dendritic cells [DC]). The DC also do not express CD80 or CD83, and only low levels of CD86
and HLA-DR, indicating that these are homogeneous, immature DC.
DC should have a CD14 low/negative, CD83, CD80, CD86+, HLADR+ phenotype

see Fig. 2.
Cells should be stained for flow cytometry as detailed under Subheading 3.1.3.,
without the RBC lysis step. Acquire at least 10,000 events; see Fig. 2 for detailed gat-
ing strategy.
3.3. Blood DC
This section details the isolation of mDC and pDC. Each method is detailed sepa-
rately, but the notes are combined in a single section at the end.
3.3.1. Plasmacytoid Dendritic Cells (pDC)

pDC were previously identified as CD11c/ CD123+ DC. We have developed a
panel of new antibodies against blood DC. pDC are specifically recognized by the
antibodies BDCA-2 and BDCA-4. These reagents are used in tandem to carry out, and
determine the success of, pDC selectionBDCA-4 is used to isolate the cells, and
BDCA-2 to validate the purity of the isolated pDC.
This method describes the separation of BDCA-4+ cells from a leukapheresis prod-
uct containing up to 40 109 cells. The expected yield from this starting population is
up to 0.2 109 BDCA-4+ pDC. The range of pDC in peripheral blood is 0.170.22%
of leukocytes (median 0.21% 1.1 104 cells/mL blood).
3.3.1.1. IMMUNOMAGNETIC LABELING
Determine the total number of leukocytes. The following protocol is designed for
leukapheresis products containing 1040 109 WBC. Use only CliniMACS PBS/
EDTA buffer (referred to as PBS/EDTA buffer) supplemented with HSA or BSA
(final concentration 0.5%), which is required to inactivate and to coat the surfaces of
tubing and bags.
1. Determine the weight of an empty 600-mL transfer pack to be used as a cell preparation
bag. Transfer the leukapheresis product into the cell preparation bag and record the
weight.
2. Dilute the leukapheresis product with HSA- or BSA-supplemented PBS/EDTA buffer to
maximal filling volume and centrifuge the cell suspension for 15 min at 200g (room tem-
perature, 1925C), no brake.
3. Discard supernatant carefully and repeat washing step.
4. Resuspend the cell pellet and adjust to a final sample volume of 95 mL.
5. Add the entire contents of one vial CliniMACS Anti-BDCA-4 MicroBeads (7.5 mL) and
mix carefully.
6. Incubate the cell preparation bag for 30 min at room temperature (1925C) on an orbital
shaker at 25 rpm maximum (see Note 8).
7. Add HSA- or BSA-supplemented CliniMACS PBS/EDTA buffer to a final volume of
600 mL for cell washing and centrifuge for 15 min at 300g (1925C), no brake. Remove
as much supernatant as possible from the cell pellet and carefully resuspend cells. Repeat
washing procedure (see Note 9).
8. Remove supernatant and then adjust volume to 100 mL.
9. Retain a 0.5-mL sample from this original cell suspension (see Note 10) and place into a
sample tube for analysis (see Subheading 3.3.1.3.). Store at room temperature until ana-
lyzed.
3.3.1.2. AUTOMATED SEPARATION

1. Determine the weight of an empty 150-mL transfer pack with a Luer/Spike Interconnector
inserted and attach this bag (cell collection bag) to the cell collection port. Ensure that all
slide clamps are open.
2. Switch on CliniMACSplus instrument and select separation program ENRICHMENT 3.1.
3. Confirm your choice by pressing ENT and enter the parameters requested on the screen.
4. Follow the tubing set installation instructions on the screen and start the automated separa-
tion program.
64 Campbell et al.
5. Determine the weight of the filled cell collection bag after separation and take a well-
mixed sample for flow cytometric analysis and determination of the total cell number.
6. The enriched pDC can be washed further in the cell collection bag and then transferred
into cell-culture bags or flasks for culturing. They also can be frozen in HSA containing
10% DMSO.
3.3.1.3. FLOW CYTOMETRIC ANALYSIS OF ISOLATED CELLS (FIG. 3)

The yield of pDC is determined from cell counts and phenotype analysis of the
original and positive fractions. Determine the cell count from the samples and
ensure that the total amount of WBC is no greater than 1 106/100 L. If necessary,
dilute the samples with PBS/EDTA/0.5% HSA. For accurate determination of pDC
numbers in peripheral blood, samples are stained with anti-CD45 to positively iden-
tify leukocytes, and BDCA-2 to identify pDC. Additionally, anti-CD123 staining as
a second pDC marker may be used if sufficient channels are available on the flow
cytometer.
1. Centrifuge 5 min at 300g, remove supernatant, and resuspend pellet in 20 L P/E/B.
2. Staining volume is 110 L. Add 22 L FcR-blocking reagent, 58 L P/E/B, 10 L
CD45.FITC, 10 L anti-BDCA-2.PE, and 10 L CD123.APC or P/E/B to each sample.
Incubate for 10 min at 4C.
3. Add 900 L of ammonium chloride lysing solution to all fractions and incubate for
another 10 min in the dark at room temperature.
4. Centrifuge 5 min at 300g, remove supernatant, and add 1 mL P/E/B. Mix and centrifuge
again 5 min at 300g.
5. Remove supernatant and resuspend cells in an adequate volume of buffer (e.g., 0.5 mL)
for flow-cytometric analysis. To assess cell viability and to exclude the events of non-
specific binding of antibodies by dead cells, add PI in a concentration of 1g/mL to each
tube immediately prior to acquisition.
6. For optimal analysis of the original fraction, a minimum of 100 BDCA-2+ events should
be counted (i.e., at a frequency of 0.2% BDCA-2+ among unseparated cells, 50,000 viable
cells have to be acquired). For the highly pure positive fraction, it is sufficient to acquire
10,000 events. For detailed gating strategy see Fig. 3.
7. For a sample calculation of pDC purity and yield see Fig. 3.
3.3.2. Myeloid DC
Isolation of CD1c+ myeloid dendritic cells is performed by two magnetic separation
steps. In the first step, CD1c-expressing cells are indirectly labeled with CliniMACS
CD1c(anti-BDCA-1)-biotin. In the same step, CD19-expressing B-cells are directly
labeled with CliniMACS CD19 MicroBeads. There is a small population of CD1c-
expressing B-cells, and this subpopulation of B-cells and all other CD19+ cells are
depleted using the CliniMACS device in the first step. In the second step, the CD19-
depleted fraction is magnetically labeled by CliniMACS Anti-Biotin MicroBeads to
enrich CD1c+ dendritic cells.
This method describes the separation of CD1c+ cells from a leukapheresis product
containing up to 40 109 cells. The expected yield from this starting population is up
to 0.2 109 CD1c+ mDC. The range of mDC in peripheral blood is 0.170.24% of
leukocytes (median 0.21% 1.1 104 cells/mL blood).
Fig. 3. Determination of PDC purity and yield. Representative flow cytometry analysis of a
leukapheresis sample before and after CliniMACS Anti-BDCA-4 separation. Samples are first
gated on CD45.FITC vs Side Scatter to exclude debris and focus on white blood cells (WBC).
Results from R1 are plotted Anti-BDCA-2.PE vs PI to exclude dead cells (R2). A combined
gate is then made (e.g., G3 = R1*R2 in Cell Quest software, BD biosciences). G3 therefore
contains all viable WBC. Anti-BDCA-2.PE is plotted against CD123.APC to evaluate the num-
bers of live BDCA-2+CD123+ cells (viable PDC) in the samples (R3). In this example, the
start population has 0.31% PDC cells, which is enriched to 92.5% after CliniMACS sorting.
To calculate yield:
total WBC * % viable PDC ( purity )
Total, viable PDC =
100%
Example:
Percentage viable PDC = 97.6% * 92.5% = 90.28%
2.84 10 7 * 90.28%
Total, viable PDC = = 2.56 10 7 viable PDC
100%
66 Campbell et al.
3.3.2.1. IMMUNOMAGNETIC LABELING1

Determine the total number of leukocytes. The following protocol is designed for
leukapheresis products containing 1040 109 WBC. Use only PBS/EDTA buffer
supplemented with HSA or BSA (final concentration 0.5%), which is required to inac-
tivate and to coat the surfaces of tubing and bags. During the depletion phase, it is
necessary to know the percentage of CD19+ cells in the samplethis must be deter-
mined by flow-cytometry analysis beforehand.
1. Determine the weight of an empty 600-mL transfer pack to be used as a cell preparation
bag. Transfer the leukapheresis product into the cell preparation bag and record the
weight.
2. Dilute the leukapheresis product (1:3) with CliniMACS PBS/EDTA buffer (supplemented
with 0.5% HSA) and centrifuge the cells at 200g for 15 min without brake.
3. Remove supernatant, taking care not to resuspend the pellet, and adjust sample to a final
volume of 90 mL.
4. Resuspend cell pellet, add entire contents of one vial of CliniMACS CD1c (anti-BDCA-1)-
biotin (5 mL) and one vial CliniMACS CD19 MicroBeads (7.5 mL) and mix carefully.
6. Add HSA- or BSA-supplemented CliniMACS PBS/EDTA buffer to a final volume of
600 mL for cell washing and centrifuge for 15 min at 300g (1925C), no brake.
7. After centrifugation, remove supernatant, taking care not to resuspend the cell pellet, and
adjust sample to a volume of 100 mL.
8. Remove a sample of 0.5 mL with a 10-mL syringe, taking care to get a well mixed aliquot
(see Note 10), and place into a sample tube for analysis (see Subheading 3.3.2.5.). Store
at room temperature (1925C) until analyzed.
3.3.2.2. AUTOMATED SEPARATION1

1. Connect a 600-mL transfer pack with a Luer/Spike Interconnector to the cell-collection
port of the CliniMACS tubing set. Ensure that all Luer/Spike slide clamps are open.
2. Switch on CliniMACSplus instrument and select the separation program called DEPLE-
TION 2.1.
3. Confirm your choice by pressing ENT and enter the Ref-number of the CliniMACS LS
tubing set (ref. no. 162-01). Follow instructions given on the instrument screen for instal-
lation of the tubing set (see Note 11).
4. Enter WBC concentration of the sample after CD19 labeling, ready to be loaded to the
instrument. Then enter frequency of labeled (CD19+) cells and sample loading volume.
5. Start the separation process.
3.3.2.3 IMMUNOMAGNETIC LABELING2

1. After finishing the depletion process, centrifuge the cell preparation bag (CD19-depleted
fraction) at 300g for 15 min without brake.
2. Remove supernatant, taking care not to resuspend the pellet, and adjust sample to a final
volume of 95 mL.
3. Add the entire contents of one vial CliniMACS Anti-Biotin MicroBeads (7.5 mL) and
mix carefully.
5. Add HSA-supplemented PBS/EDTA buffer to a final volume of 600 mL for cell washing,
and centrifuge for 15 min at 300g, 1925C, no brake.
6. After centrifugation, adjust the volume of the leukapheresis product for loading on the
CliniMACSplus instrument to 100 mL.
7. Resuspend cells, and remove a 0.5-mL sample using a 10-mL syringe, taking care to get
a well mixed aliquot for analysis. Determine at least the cell concentration, viability, and
BDCA-1 cell frequency/cell number.
3.3.2.4. AUTOMATED SEPARATION2

1. Connect a 150-mL or 600-mL transfer pack with a Luer/Spike Interconnector to the cell
collection port of the CliniMACS tubing set. Ensure that all Luer/Spike slide clamps are
open.
2. Switch on CliniMACSplus instrument and select separation program ENRICHMENT 3.1.
3. Confirm your choice by pressing ENT and enter the parameters requested on the screen.
4. Follow the tubing set installation instructions on the screen and start the automated sepa-
ration program.
5. When the separation has finished, determine the weight of the filled cell collection bag,
and take a well mixed sample for flow-cytometric analysis and determination of total cell
number.
6. The enriched mDC can be washed further in the cell collection bag and then transferred
into cell-culture bags or flasks for culturing. They also can be frozen in HSA containing
10% DMSO.
3.3.2.5. FLOW-CYTOMETRIC ANALYSIS OF ISOLATED CELLS (FIG. 4)

Analysis of the following fractionsoriginal, CD19-depleted, and BDCA-1+
allows the most accurate determination of yield. Determine the cell count from the
samples and ensure that the total amount of WBC is no greater than 1 106/100 L. If
necessary, dilute the samples with PBS/EDTA/0.5% HSA. For accurate determination
of mDC numbers, cells are stained for analysis with a combination of CD1c (PE),
CD20 (APC), and CD45 (FITC) monoclonal antibody reagents. Samples are stained
with anti-CD45 to positively identify leukocytes; CD1c (BDCA-1) is a marker spe-
cific for mDC and a subpopulation of B-cells in blood. CD20 is used to stain B-cells
to control the depletion efficiency, since the antigen is specifically expressed upon
B-cells and labeling with a CD19 fluorochrome is not optimal after CD19 Micro-
Beads incubation.
1. Place each sample in a suitable tube for flow cytometry. Centrifuge 5 min at 300g,
remove supernatant, and resuspend pellet in 20 L P/E/B.
2. Staining volume is 110 L. Add 58 L P/E/B, 22 L FcR-blocking reagent, 10 L
CD45.FITC, 10 L CD1c.PE, and 10 L CD20.APC. Incubate for 10 min at 4C.
3. Add 900 L of ammonium chloride lysing solution to the fractions and incubate for
another 10 min in the dark at room temperature.
4. Centrifuge 5 min at 300g, remove supernatant, and add 1 mL P/E/B. Mix and centrifuge
again 5 min at 300g.
5. Remove the supernatant, resuspend the cell pellet, add 1000 L P/E/B and 1 L PI imme-
diately before FCM analysis.
6. For optimal analysis of the original fraction, a minimum of 100 BDCA-1+ events should
be counted (i.e., at a frequency of 0.2% BDCA-1+ among unseparated cells, 50,000 viable
68 Campbell et al.
Fig. 4. Determination of MDC purity and yield. Representative flow cytometry analysis of a
leukapheresis sample before separation, after CliniMACS CD19 depletion and following
CliniMACS CD1c (Anti-BDCA-1) separation. Samples are first gated on CD45.FITC vs Side
Scatter to exclude debris and focus on white blood cells (WBC). Results from R1 are plotted
CD1c.PE vs PI to exclude dead cells (R2). A combined gate is then made (e.g., G3 = R1*R2 in
Cell Quest software, BD biosciences). G3 therefore contains all viable WBC. CD1c.PE is
plotted against CD20.APC to evaluate the numbers of live CD1c+CD20 cells (viable MDC) in
the samples (R3). In this example, the start population has 0.61% MDC cells, which is pre-
enriched to 1.34% after CD19 depletion and finally enriched using CD1c (Anti-BDCA-1)-
Biotin and Anti-Biotin MicroBeads up to 93.0%.
To calculate yield:
total WBC * % viable MDC ( purity )

Total, viable MDC =
100%
Example:
Percentage viable MDC = 97.5% * 93.0% = 90.68%
7.41 10 7 * 90.68%
Total, viable MDC = = 6.72 10 7 viable MDC
100%
cells have to be acquired). For the highly pure positive fraction, it is sufficient to acquire
10,000 events. For detailed gating strategy, see Fig. 4.
7. For a sample calculation of mDC purity and yield, see Fig. 4.
4. Notes
1. Leukapheresis procedures can lead to the activation of monocytes, which ultimately
decreases the yield of DC obtained from the isolated cells. Reducing the inlet flow rate
of leukapheresis (ideally to 50 mL/min) substantially reduces monocyte activation.
2. Storage of leukapheresis samples decreases the quality of the CD14 separation and the
yield of DC. Preparation and separation of the leukapheresis product should be per-
formed immediately after leukapheresis. If the leukapheresis product must be stored, it
should be kept at controlled room temperature (1925C) and should not be stored for
more than 24 h before separation. Storage of isolated monocytes is also not recom-
mended.
3. One vial of CliniMACS CD14 Reagent is capable of labeling up to 4 109 CD14+ mono-
cytes out of a total leukocyte number of up to 20 109 cells.
4. The CliniMACS CD14 Reagent is for single use only. Storage at +2C to +8C is recom-
mended. Do not freeze. Do not use after expiration date.
5. If samples are also kept from the negative and waste fractions, in the unlikely event of
lower yield, it is easier to trace the CD14+ cells.
6. The target cells for the separation are the CD14high monocytes. For accurate analysis, any
CD14dim cells have to be excluded by using CD15 as an exclusion marker for granulo-
cytes.
7. During preclinical work-up, BSA may be substituted for HSA on the grounds of cost.
8. As an alternative, cell bags can be gently shaken every 5 min by hand during magnetic
labeling.
9. To get highest purities of PDC, it is essential to get rid of platelets, which otherwise block
the anti-BDCA-4 reagent. Never omit one of the initial two washing steps to abbreviate
the protocol.
10. To get a representative aliquot from a cell fraction out of a bag, use at least a 10-mL
syringe and refill it repeatedly before taking the sample.
11. Make sure that the tubing set is correctly inserted. If one valve is blocked, the whole
separation can fail. All bags (cell collection bag, negative fraction bag, buffer waste bag)
must be unsealed when procedure is started.
References
1.
1 Banchereau, J. (1997) Dendritic cells: therapeutic potentials. Transfus. Sci. 18, 313326.
2.
2 Dzionek, A., Fuchs, A., Schmidt, P., et al. (2000) BDCA-2, BDCA-3, and BDCA-4: three
markers for distinct subsets of dendritic cells in human peripheral blood. J. Immunol. 165,
60376046.
3.
3 Dzionek, A., Inagaki, Y., Okawa, K., et al. (2002) Plasmacytoid dendritic cells: from spe-
cific surface markers to specific cellular functions. Hum. Immunol. 63, 11331148.
4.
4 Fay, J. (1998) Dendritic cells in the treatment of cancer. Baylor Univ. Med. Cent. Proc.
11(4), 217219.
5.
5 Pickl, W. F., Majdic, O., Kohl, P., et al. (1996) Molecular and functional characteristics of
dendritic cells generated from highly purified CD14+ peripheral blood monocytes. J.
Immunol. 157, 38503859.
70 Campbell et al.
6.
6 Thurner, B., Rder, C., Dieckmann, D., et al. (1999) Generation of large numbers of fully
7.
7 Cella, M., Facchetti, F., Lanzavecchia, A., and Colonna, M. (2000) Plasmacytoid den-
dritic cells activated by influenza virus and CD40L drive a potent TH1 polarization. Nat.
Immunol. 1, 305310.
8.
8 Cella, M., Jarrossay, D., Facchetti, F., et al. (1999) Plasmacytoid monocytes migrate
to inflamed lymph nodes and produce large amounts of type I interferon. Nat. Med. 5,
919923.
9. Fong, L., Hou, Y., Rivas, A., et al. (2001) Altered peptide ligand vaccination with Flt3
ligand expanded dendritic cells for tumor immunotherapy. PNAS 98, 88098814.
10. Ito, T., Inaba, M., Inaba, K., et al. (1999) A CD1a+/CD11c+ subset of human blood den-
dritic cells is a direct precursor of Langerhans cells. J. Immunol. 163, 14091419.
Clinical-Grade MoDC by CliniMACS 71
6
Generation of Clinical-Grade Monocyte-Derived
Dendritic Cells Using the CliniMACS System
Thomas Putz, Hubert Gander, Reinhold Ramoner,

Claudia Zelle-Rieser, Andrea Rahm, Walter Nussbaumer,
Georg Bartsch, Lorenz Hltl, and Martin Thurnher
Summary
We describe the generation of monocyte-derived dendritic cells (MoDC) from leukapheresis
products using the CliniMACS system from Miltenyi Biotec. In a clinical setting, the method
turned out to be feasible for the generation of clinical-grade MoDC from patients with meta-
static renal-cell carcinoma. MoDC generated with this system exhibited a fully mature pheno-
type as well as high migratory and T-cell stimulatory capacity.
Key Words: Monocyte-derived dendritic cells; immunomagnetic isolation; leukapheresis;
purity; CD14; CD83; CCR7; migration; proliferation.
1. Introduction
Dendritic cells (DC) are antigen-presenting cells (APC) that can induce vigorous
T-lymphocyte responses (1). DC pick up and process antigens to present antigenic
peptides in a major histocompatibility complex (MHC)-restricted fashion to the T-cell
receptor of T-cells. DC also provide co-stimulatory signals required for T-cell activa-
tion (1) and bias T helper cell differentiation (2). In addition, DC can enhance B-lym-
phocyte and natural killer (NK)-cell responses (1,3). The in vivo immunogenicity of
DC has meanwhile been tested and confirmed in numerous clinical trials (4).
One way to generate DC in vitro is to differentiate them from CD14+ monocytes in
a two-step culture system (5,6) (Fig. 1). In a first step, granulocyte/macrophage colony-
stimulating factor (GM-CSF) and interleukin (IL)-4 promote the development of
immature DC with high antigen-capturing capacity. In the second step, proinflamma-
tory factors induce the terminal maturation of CD83+ DC (7,8), which stably express
MHC-peptide complexes (9), abundant T-cell adhesion, and co-stimulatory molecules
(CD80 and CD86) (10) , as well as chemokine receptors such as CCR7, which direct
DC migration to lymph nodes (11).
71
72 Putz et al.
Fig. 1. Monocyte-derived dendritic cell generation in the two-step granulocyte/macrophage

colony-stimulating factor (GM-CSF)/interleukin (IL)-4 culture system.
Various methods such as adherence to plastic or centrifugal elutriation have been

used to isolate monocytic precursors for subsequent DC differentiation (12,13). Yet
another approach to isolate monocytes is based on the immunomagnetic isolation of
CD14+ cells. In the present work we have used the immunomagnetic isolation system
from Miltenyi Biotec to generate monocyte-derived DC (MoDC). We analyzed MoDC
with regard to phenotype as well as migratory and T-cell stimulatory capacity. In a
clinical setting, we also used the CliniMACS system to generate MoDC from
leukapheresis products.
2. Materials
2.1. Leukapheresis
1. Cobe Spectra Apheresis System (Gambro BCT, Lakewood, CO).
2. Anticoagulant Citrate Dextrose Solution (ACD-A) (Baxter, Austria).
3. Cell-culture flasks, 162-cm2 (Costar, Germany).
4. Luer/Spike Interconnector (Miltenyi Biotec, Bergisch Gladbach, Germany).
5. CliniMACS PBS/EDTA buffer (Miltenyi Biotec).
6. Heat-inactivated human AB plasma: human AB plasma (local Institute of Blood Transfu-
sion) from individuals tested negative for various viruses 6 mo after donation. In addi-
tion, after heat-inactivation (30 min at 56C) plasma was subjected to sterility testing at
the local Institute of Hygiene (bacteria, fungi).
7. Neubauer chamber.
8. Plaxan (Gatt, Austria).
2.2. Enrichment of Monocytes
1. CliniMACS CD14 Reagent (Miltenyi Biotec).
2. CliniMACS Tubing Set 600 (Miltenyi Biotec).
3. CliniMACS instrument (Miltenyi Biotec).
4. Luer/Spike Interconnector (Miltenyi Biotec).
5. Cell-culture flasks, 162-cm2 (Costar).
6. AIM-V (Invitrogen, Carlsbad, CA).

7. Heat-inactivated human AB plasma.
8. HEPES (Bio Whittaker, Europe).
9. 2-Mercaptoethanol (Merck, Germany).
10. Recombinant human GM-CSF (Leucomax, Novartis).
11. Recombinant human IL-4 (Strathmann Biotech, Hannover, Germany).
12. Dimethylsulphoxide (DMSO) (Sigma, Germany).
13. Six-well plates (Costar).
14. Renal-cell carcinoma lysate (12).
15. Keyhole limpet hemocyanin (KLH), endotoxin-free/pyrogen-free (Calbiochem-Nova-
biochem, San Diego, CA).
16. Tumor necrosis factor (TNF)-* (R&D Systems, Minneapolis, MN).
17. IL-1* (R&D Systems).
18. IL-6* (R&D Systems).
*These cytokines can now be purchased in GMP quality.
19. Prostaglandin E2 (Prostin E2) (Pharmacia & Upjohn, Vienna, Austria).
20. Lactated Ringers solution (Mayerhofer Pharmazeutika, Linz, Austria).
21. Luer/Spike Interconnector (Miltenyi Biotec).
2.3. Flow-Cytometric Analysis

1. PE conjugated antihuman CD83 (cat. no. 556855, BD, Mountain View, CA).
2. PE conjugated mouse IgG1- monoclonal immunglobulin isotype control (BD, cat. no.
555749).
3. Rat IgG2a anti-human CCR7 (Dr. Reinhold Frster, Hannover, Germany, or BD).
4. FITC conjugated mouse anti-rat IgG2a monoclonal antibody (BD, cat. no. 553896).
5. Rat IgG2a- monoclonal immunoglobulin isotype control (BD, cat. no. 555841).
6. FACSCalibur and CellQuest software (BD).
2.4. Migration Assay

1. Twenty-four-well Transwell chambers (Corning Costar, Cambridge, MA) with polycar-
bonate filters of 6.5 mm in diameter and 5 m pore size.
2. AIM-V (Invitrogen).
4. Macrophage inflammatory protein (MIP)-3 (CCL19) (R&D Systems).
5. Lipopolysaccharide (LPS) from Salmonella abortus equi (Sigma, Austria).
6. FACSCalibur and CellQuest software (BD).
2.5. Mixed Leukocyte Reaction

1. Peripheral blood mononuclear cells (PBMC) from healthy donors for T-cell isolation.
2. CD4 MicroBeads (Miltenyi Biotec).
3. Irradiated MoDC (3000 rad).
4. Ninety-six-well flat-bottomed tissue-culture plates (BD Falcon).
5. AIM-V (Invitrogen).
7. [3H]thymidine (1 Ci/well = 37 kBq/well, ICN Biomedicals).
8. Skatron cell harvester, glass fiber filters and scintillation vials (Skatron, Lier, Norway).
9. OptiPhase Hi-Safe2 Scintillator (cat. no. 0197435241 Wallac, Fisher).
10. Liquid-scintillation counter (Wallac 1410, Pharmacia).
74 Putz et al.
Fig. 2. Purity of monocytes isolated with the CliniMACS system. A representative sample
with 94% CD14+ cells is shown.
3. Methods
3.1. Leukapheresis
At the local Institute for Blood Transfusion, renal-cell carcinoma (RCC) patients
underwent a standard leukapheresis procedure performed with the Cobe Spectra cell
separator. Using the MNC program at a continuous whole-blood inlet flow rate rang-
ing from 50 to 70 mL/min (see Note 1), 35 L of whole blood were processed. ACD-
A at a 1:12 ratio was used for anticoagulation.
Transfer the apheresis product (50100 mL) to a cell-culture flask using a Luer/
Spike Interconnector and adjust to 200 mL using CliniMACS PBS/EDTA buffer
supplemented with 2% heat-inactivated human AB plasma. Determine the total cell
number with a Neubauer chamber in the presence of Plaxan (to lyse red blood cells).
Table 1 lists the numbers of cells obtained from 16 RCC patients.
3.2. Generation of MoDC From Leukapheresis
Products Using the CliniMACS System
3.2.1. Enrichment of Monocytes
1. CD14+ monocytes were separated from apheresed cells by positive selection using the
CD14 reagent, the CliniMACS Tubing Set 600 (for up to 20 109 cells), and the Clini-
MACS instrument. All steps were carried out according to the manufacturers instructions.
(For a detailed protocol see Chapter 5 in this volume.)
2. Remove an aliquot of your final isolated positive fraction using the Luer/Spike
Interconnector for flow-cytometric determination of monocyte purity (Fig. 2). Yields of
CD14+ cells from RCC patients are listed in Table 1.
Clinical-Grade MoDC by CliniMACS
Table 1
Monocyte Enrichment and MoDC Generation From PBMC of RCC Patients Using the CliniMACS System
PBMC CD14+ cells MoDC CD14+ MoDC MoDC yield lymphocyte contamination
Patients (108) (108) (108) (% of PBMC) (% of PBMC) (% of CD14+) of MoDC (%) a
Patient 1 15.00 2.25 1.80 15 12 80 1.7
Patient 2 19.00 3.50 1.60 18 8 46 1.1
Patient 3 20.00 5.80 1.00 29 5 17 2.7
Patient 4 79.00 6.00 2.65 8 3 44 4.6
Patient 5 50.00 8.20 1.40 16 3 17 6.7
Patient 6 100.00 7.48 2.97 7 3 40 3.2
Patient 7 23.00 7.00 2.30 30 10 33 1.7
Patient 8 60.00 7.30 2.00 12 3 27 1.8
75
Patient 9 39.00 4.00 0.60 10 2 15 2.8

Patient 10 90.00 8.00 6.50 9 7 81 1.9
Patient 11 50.00 5.00 1.70 10 3 34 6.5
Patient 12 90.00 20.00 8.00 22 9 40 1.8
Patient 13 90.00 13.00 3.36 14 4 26 5.1
Patient 14 48.00 6.50 2.00 14 4 31 1.5
Patient 15 118.00 6.70 0.88 6 1 13 1.6
Patient 16 74.00 7.00 5.90 9 8 84 ND
mean 60.31 7.36 2.79 14 5 39 3
SD 31.30 4.00 2.09 7 3 23 2
a Lymphocyte contamination was determined by selective forward/sideward scatter gating (see also Fig. 3B).
ND, Not determined; MoDC, monocyte-derived dendritic cells; PBMC, peripheral blood mononuclear cells; RCC, renal-cell carcinoma; SD, standard
deviation.
75
76 Putz et al.
3. CD14+ cells (50 106 in 50 mL) were cultured in 162-cm2 cell-culture flasks in AIM-V
containing 1% heat-inactivated human AB plasma, 10 mM HEPES, and 50 M 2-mercap-
toethanol as well as a combination of recombinant human GM-CSF (1000 U/mL) and
recombinant human IL-4 (1000 U/mL).
4. After 2 d of culture, 50 mL of fresh medium containing supplements were added.
5. On d 5, MoDC were harvested and frozen in liquid nitrogen using a standard protocol
(50% AIM-V, 40% human AB plasma, 10% DMSO).
6. Two days before vaccination, d-5 MoDC were thawed, counted, and replated in 6-well
plates at 1.8 106 cells per well in 6 mL of fresh medium containing GM-CSF (1000
U/mL) and IL-4 (250 U/mL) as well as 1% heat-inactivated human AB plasma (see
Notes 24). All cells were pulsed with autologous RCC lysate (10 g/mL) and half of
the cells also with KLH (10 g/mL). Maturation of thawed d-5 MoDC was induced
simultaneously with the RCC lysate/KLH pulses by stimulation with a cocktail con-
sisting of 1000 U/mL of recombinant human TNF-, IL-1 (5 ng/mL), IL-6 (10 ng/
mL), and 1 M prostaglandin E2 (14,15).
7. After 48 h, MoDC were harvested, washed, counted, and resuspended in lactated Ringers
solution containing 1% heat-inactivated human AB serum.
8. Remove an aliquot of your final MoDC suspension using the Luer/Spike Interconnector
for flow-cytometric determination of CD83 expression.
Only samples with CD83 expression >85% were released (Fig. 3A). Purity of MoDC
ranged from 93 to 99% (mean: 97%). Sterility testing was performed at the local Insti-
tute of Hygiene; all tests were negative (100% sterility).
3.3. Flow-Cytometric Analysis
To determine surface expression of CD83 or CCR7, cell suspensions were labeled
for 30 min on ice with mouse anti-CD83-PE or rat anti-CCR7 followed by anti-rat
IgG2a-FITC. The samples were analyzed using the FACSCalibur and CellQuest soft-
ware from BD.
3.4. Migration Assay
Chemotactic capacity of MoDC was measured by assessing migration through a
polycarbonate filter (6.5 mm in diameter and 5 m pore size) in 24-well Transwell
chambers (16).
1. 600 L AIM-V supplemented with 1% heat-inactivated human AB plasma with or with-
out 300 ng/mL MIP-3 (CCL19) was added to the lower chamber. Immature MoDC or
MoDC matured with either LPS or cocktail for 48 h were added to the upper chamber
(1 105 cells in a total volume of 100 L) and incubated for 3 h at 37C.
Fig. 3. (opposite page) Purity and homogeneity of monocyte-derived dendritic cells (MoDC)
generated with the CliniMACS system. Monocytes were propagated for 5 d in the granulocyte/
macrophage colony-stimulating factor (GM-CSF)/interleukin (IL)-4 cell-culture system, pulsed
with autologous tumor lysate, and stimulated for 2 d with the maturation cocktail. A representa-
tive example is shown. (A) Ninety percent of all MoDC generated expressed the maturation
marker CD83. (B) The MoDC portion represented 97% of all cells (gate 1); contaminating lym-
phocytes were approx 3% (gate 2). (C) MoDC displayed a homogenous morphology with typi-
cal veils.
78 Putz et al.
Fig. 4. Monocyte-derived dendritic cells (MoDC) generated with the CliniMACS system
express CCR7 and migrate in response to macrophage inflammatory protein (MIP)-3. Mono-
cytes were maintained for 5 d in the granulocyte/macrophage colony-stimulating factor (GM-
CSF)/interleukin (IL)-4 cell-culture system and matured for 2 d with either the maturation
cocktail or lipopolysaccharide (LPS), or left unstimulated (immature). (A) MoDC treated with
the maturation cocktail exhibited high levels of CCR7. (B) LPS treatment of MoDC resulted in
low CCR7 expression. (C) Migration of MoDC was determined in a Transwell chamber assay.
Accumulation of migrating MoDC in the lower chamber containing MIP-3/CCL19 was deter-
mined by counting cells for 1 min in a flow cytometer.
Fig. 5. Monocyte-derived dendritic cells (MoDC) generated with the CliniMACS system
exhibit high CD4+ T-cell stimulatory capacity. Monocytes were propagated for 5 d in the
granulocyte/macrophage colony-stimulating factor (GM-CSF)/interleukin (IL)-4 cell-culture
system and stimulated for 2 d with the maturation cocktail or lipopolysaccharide. Subse-
quently, MoDC were irradiated and co-cultured with CD4+ T-cells. Proliferation of CD4+
T-cells was determined by 3H-thymidine incorporation. Background proliferation of CD4+
T-cells was negligible (data not shown).
2. A 500-L aliquot of the cells that migrated to the bottom chamber was counted by flow
cytometry as the number of events acquired during 1 min. Each experiment was per-
formed in duplicates.
The mean number of spontaneously migrated MoDC (in the absence of MIP-3)
was subtracted from the total number of MoDC that migrated in response to MIP-3
(see Note 4).
Fig. 4A,B shows CCR7 expression of MoDC and Fig. 4C shows MoDC migration
in response to MIP-3.
3.5. Mixed Leukocyte Reaction
1. CD4+ T-cells were positively selected using CD4 MicroBeads according to the manu-
facturers instructions.
2. CD4+ T-cells (2 105/well) were stimulated with various numbers of allogeneic, irradi-
ated (3000 rad) MoDC. Cells were cultured in triplicates in 96-well flat-bottomed tissue-
culture plates in a final volume of 200 L/well in medium containing 1% heat-inactivated
human AB plasma.
3. T-cell proliferation was measured as [3H]-thymidine incorporation (1 Ci/well = 37 kBq/
well). Cells were pulsed during the last 16 h of a 5-d culture period, harvested on glass
fiber filters using a Skatron cell harvester, and analyzed in a liquid-scintillation counter.
Results are expressed as the mean counts per minute (cpm) of six independent experi-
ments.
Fig. 5 confirms that MoDC generated with this system have high T-cell stimulatory
capacity.
80 Putz et al.
4. Notes
1. Although rare, early apoptosis may cause considerable cell loss during the first step of
MoDC culture (immature MoDC, Fig. 1). The reason for this phenomenon is unclear, but
flow rates during leukapheresis may be critical and should therefore not exceed 80 mL/
min (50 mL/min is recommended by Miltenyi Biotec).
2. Immature MoDC, 18 106, are thawed to prepare one vaccine. This usually results in 11
14 106 mature, antigen-loaded MoDC, corresponding to a recovery of 6177%. The
stress induced by the pro-inflammatory cocktail, as well as the addition of tumor lysate,
which signals massive necrosis, is likely to be responsible for apoptotic cell loss.
3. Thawing of MoDC is also critical. Cryo-vials should be transferred directly from liquid
nitrogen to 37C. As soon as the cell suspension is thawed, it is diluted with plasma-
supplemented buffer. However, dilution has to be performed dropwise with gentle shak-
ing of the cryo-tube, to allow the DMSO to diffuse from the MoDC. Otherwise, the DMSO
in the MoDC would cause a rapid influx of water and osmotic lysis of the MoDC.
4. The maturation step is carried out at a cell density of 3 105 MoDC/mL (i.e., 1.8 106
cells per 6 mL). We have previously shown that efficient MoDC maturation requires cell
densities below 5 105 cells/mL (8). Both LPS, a component of the outer cell wall of
Gram-negative bacteria, as well as a cocktail consisting of TNF-, IL-1, IL-6, and
prostaglanding E2 induce MoDC maturation. However, it has to be noted that the use of
different maturation stimuli can cause phenotypic and functional differences. As shown
in Fig. 4A, the cocktail is more potent at inducing the expression of CCR7, the chemokine
receptor that directs migration of DC to MIP-3-producing lymph nodes (17). Conse-
quently, MoDC matured with the cocktail migrated much more efficiently towards MIP-
3 in a Transwell chamber migration assay (Fig. 4C). Furthermore, the T-cell stimulatory
capacity of MoDC treated with the maturation cocktail was higher compared to MoDC
treated with LPS (Fig. 5).
5. MoDC prepared with the CliniMACS system have been administered to 16 patients
with metastatic RCC (Hltl et al., manuscript in preparation). No adverse effects were
observed. Mean follow up time was 9 mo (range: 216 mo). Taken together, our data
indicate that the CliniMACS system is feasible for the generation of functional, clini-
cal-grade MoDC. Preliminary clinical data also suggest that MoDC prepared with the
CliniMACS system are well tolerated.
Acknowledgments
This work was supported by a grant of the Competence Center Medicine Tyrol
(KMT) to Martin Thurnher.
References
1.
Nature 392, 245252.
2.
2 Rissoan, M. C., Soumelis, V., Kadowaki, N., et al. (1999) Reciprocal control of T helper
cell and dendritic cell differentiation. Science 283, 11831186.
3.
3 Fernandez, N. C., Lozier, A., Flament, C., et al. (1999) Dendritic cells directly trigger NK
cell functions: cross-talk relevant in innate anti-tumor immune responses in vivo. Nat.
Med. 5, 405411.
4.
4 Schuler, G., Schuler-Thurner, B., and Steinman, R. M. (2003) The use of dendritic cells in
cancer immunotherapy. Curr. Opin. Immunol. 15, 138147.
5.
5 Sallusto, F. and Lanzavecchia, A. (1994) Efficient presentation of soluble antigen by cul-
tured human dendritic cells is maintained by granulocyte/macrophage colony-stimulating
factor plus interleukin 4 and downregulated by tumor necrosis factor alpha. J. Exp. Med.
179, 11091118.
6.
7.
7 Lechmann, M., Zinser, E., Golka, A., and Steinkasserer, A. (2002) Role of CD83 in the
immunomodulation of dendritic cells. Intl. Arch. Allergy Immunol. 129, 113118.
8.
8 Thurnher, M., Papesh, C., Ramoner, R., et al. (1997) In vitro generation of CD83+ human
blood dendritic cells for active tumor immunotherapy. Exp. Hematol. 25, 232237.
9.
9 Cella, M., Engering, A., Pinet, V., Pieters, J., and Lanzavecchia, A. (1997) Inflammatory
stimuli induce accumulation of MHC class II complexes on dendritic cells. Nature 388,
782787.
10.
10 Carreno, B. M. and Collins, M. (2002) The B7 family of ligands and its receptors: new
pathways for costimulation and inhibition of immune responses. Ann. Rev. Immunol. 20,
2953.
11.
11 Moser, B. and Loetscher, P. (2001) Lymphocyte traffic control by chemokines. Nat.
Immunol. 2, 123128.
12.
12 Holtl, L., Zelle_Rieser, C., Gander, H., et al. (2002) Immunotherapy of metastatic renal
cell carcinoma with tumor lysate-pulsed autologous dendritic cells. Clin. Cancer Res. 8,
33693376.
13.
13 Rieser, C., Papesh, C., Herold, M., et al. (1998) Differential deactivation of human den-
dritic cells by endotoxin desensitization: role of tumor necrosis factor-alpha and prostag-
landin E2. Blood 91, 31123117.
14.
15.
15 Rieser, C., Bock, G., Klocker, H., Bartsch, G., and Thurnher, M. (1997) Prostaglandin E2
and tumor necrosis factor alpha cooperate to activate human dendritic cells: synergistic
activation of interleukin 12 production. J. Exp. Med. 186, 16031608.
16. Scandella, E., Men, Y., Gillessen, S., Forster, R., and Groettrup, M. (2002) Prostaglandin
17. Robbiani, D. F., Finch, R. A., Jager, D., Muller, W. A., Sartorelli, A. C., and Randolph, G.
J. (2000) The leukotriene C(4) transporter MRP1 regulates CCL19 (MIP-3beta, ELC)-
dependent mobilization of dendritic cells to lymph nodes. Cell 103, 757768.
Adenoviral Transduction of Dendritic Cells 83
7
Adenoviral Transduction of Dendritic Cells
Rienk Offringa, Kitty Kwappenberg, Martijn Rabelink,

Delphine Rea, and Rob Hoeben
Summary
Recombinant adenoviral (rAd) vectors are highly suitable for efficient genetic modification
of dendritic cells (DC). In certain cases, the high immunogenicity of rAd may be a disadvan-
tage. This chapter describes the essential aspects of optimal rAd-mediated gene transfer into
DC, discusses the consequences of the immunogenicity of rAd vectors, and provides sugges-
tions to minimize these complications.
Key Words: Dendritic cells; recombinant adenovirus; genetic transduction; monocytes.
1. Introduction
DC are generally considered the crucial antigen-presenting cells (APC) for the
induction and regulation of T-cell-dependent immune responses (1). This notion
has prompted extensive research aiming at genetic modification of human DCfor
instance, to program these APC for presentation of large amounts of peptide antigens
of interest into their surface HLA molecules, or to secrete immuno-regulatory
cytokines. rAd vectors are widely used for genetic modification of a variety of cell
types, including DC (see Note 1). Adenoviral vectors can harbor one, or even mul-
tiple transgenes, as well as strong promoter/enhancer sequences to direct high
expression of the transgene(s) of interest. They are considered attractive and rela-
tively safe viral vectors for clinical applications, because rAd batches free of replica-
tion-competent adenovirus (RCA) can reliably be obtained, high titers of up to 1
1011 infectious particles/mL can be reached with relative ease, and rAd batches can
be stored for prolonged periods without significant loss of infectivity (2,3). A promi-
nent advantage of rAd as vehicles for gene transfer into DC is that, unlike for
retroviral vectors, this process does not depend on mitotic activity of the targeted
cells. This is crucial, because (until proven otherwise) human DC, even in their
immature stage, constitute terminally differentiated cells that have lost the capacity
for cell division. Furthermore, unlike the case with poxvirus vectors, high transduc-
83
84 Offringa et al.
tion efficiencies ranging between 50 and 90% can readily be achieved without com-
promising the vitality and the key immunological functions of DC.
In spite of these advantages, certain drawbacks of rAd-mediated gene transfer into
DC should not be ignored. Even though high transduction rates can be obtained, this
generally requires a high multiplicity of infection (MOI), in the range of 2001000
infectious particles per DC (4). Notably, rAd-infection will therefore not only intro-
duce the rAd genome harboring the transgene into the DC, but also a considerable
amount of adenoviral antigens. Peptides derived from these highly immunogenic pro-
teins, for which the majority of the human population exhibits potent memory T-cell
immunity, are processed into class I and II MHC molecules at the DC surface. As a
result, rAd-infected DC readily trigger Ad5-specific responses by human T-cells both
in vitro and in vivo, which may very well dominate over immunity against transgene-
encoded antigens of interest. The HLA-restricted presentation of adenoviral antigens
by rAd-modified DC is not restricted to the capsid antigens derived from the rAd
particles that have infected the DC, but can also include various other adenoviral anti-
gens encoded by the viral backbone. Even though replication-deficient rAd generally
lack the E1 region required for efficient expression of the adenoviral gene program,
the other viral genes that were retained in the vector can nevertheless be expressed in
cells that are infected at high MOI (5,6), as is the case for rAd-modified DC. Impor-
tantly, the contribution of rAd vector components to the antigenic load of rAd-modi-
fied DC can be limited by employing rAd vectors that use other cell-surface molecules
as receptor and that, as a result, infect DC with higher efficiency, permitting the use of
a lower MOI. Alternatively, rAd vectors can be used in which additional genes have
been deleted from the viral backbone (see Note 2).
In conclusion, rAd are highly efficient and versatile vectors for genetic modifica-
tion of DC, but like most other viral vectors they are also highly immunogenic. Conse-
quently, it is recommended to weigh the pros and cons of rAd vectors against those of
alternative vector systems before embarking on genetic modification of DC.
2. Materials
2.1. Preparation of rAd Stocks
1. Infection medium: DMEM (Gibco-BRL, Gaithersburg, MD) supplemented with 0.3 g/L
glucose, 0.6 g/L glutamine, 0.1 g/L biotine, 10 mM MgCl2 (for PER.C6 helper cells only),
and 2% heat-inactivated horse serum (HS).
2. Postinfection medium: DMEM (Gibco-BRL) supplemented with 0.3 g/L glucose, 0.6 g/L
glutamine, 0.1 g/L biotine, and 8% heat-inactivated fetal bovine serum (FBS).
3. PBS/HS: phosphate-buffered saline supplemented with 2% heat-inactivated horse serum
(HS).
4. 2X F15 medium: For preparation of 5 L, dissolve into 4 L of milli-Q water: 1 jar of MEM
powder (MEM, minimal essential medium for 10 L, with Earles salts, with L-glutamine,
without NaHCO3; Gibco-BRL cat. no. 61100-087). Subsequently add and dissolve 22 g
NaHCO3 and 6 g glutamine. Finally add 200 mL of a 0.1 g/L biotine solution, 120 mL of
a 15% glucose solution, 200 mL MEM amino acids solution (50X) (Gibco-BRL), 100 mL
MEM nonessential amino acids solution (100X) (Gibco-BRL), 100 mL Basal Medium
Eagle (BME) vitamins solution (100X) (Gibco-BRL). Adjust volume to 5 L, set pH to
7.15 with 3N NaOH, sterilize through a 0.22 m filter, and aliquot.
5. One M HEPES pH 7.4, sterilized through a 0.22 m filter.

6. Agar solution, 1.7%: 8.5 g agar noble (Difco) in 475 mL milli-Q water; sterilized through
autoclave.
7. Agar 1.7%, pH 7.4: cool freshly autoclaved agar to <60C, add 25.0 mL 1 M HEPES (pH
7.4), prewarmed at 37C, per 475 mL agar.
8. Cesium chloride (CsCl) stock solutions: prepare 500 mL of a 1.45 g/cm3 (3.6 M) CsCl
solution by dissolving 305 g CsCl (MW 168.37) in 500 mL T10E1 (10 mM Tris, 1 mM
EDTA , pH 8.1). Use part of this solution to prepare the other CsCl solutions: 300 mL +
106.7 mL T10E1 = 1.33 g/cm3 CsCl; 100 mL + 120.2 mL T10E1 = 1.20 g/cm3 CsCl.
9. TD dialysis buffer (pH 7.8): prepare a 10X TD stock solution by dissolving into 1000 mL
milli-Q water 80 g NaCl (1.37 M), 3.8 g KCl (51 mM), 1.0 g Na2HPO42H2O (5.6 mM)
and 30 g Tris-HCl (248 mM). Also prepare a 200X Ca/Mg stock solution by dissolving
into 200 mL milli-Q water 5.3 g CaCl2 (180 mM) and 4.0 g MgCl26H2O (98 mM). Now
prepare 1000 mL of 1X TD dialysis buffer by adding to 900 mL milli-Q water 100 mL
10X TD and 5mL Ca/Mg stock solution.
10. Sucrose dialysis buffer (pH 7.8): add to 4000 mL of milli-Q water 40.9 g NaCl (140 mM),
4.4 g Na2HPO4.2H2O ( 5,0 mM), 1.0 g KH2PO4 (1.5 mM), 250 g sucrose (730 mM, 5 %)
and adjust the final volume to 5000 mL.
2.2. Isolation, rAd-Infection, and Phenotypical
Analysis of Immature Monocyte-Derived DC
1. MACS buffer: PBS supplemented with 2 mM EDTA and 0.5% BSA.
2. Antibodies for analysis of the DC surface phenotype by flow cytometry: phycoerythrin
(PE)-anti-CD1a (Caltag Laboratories, Burlingame, CA), PE-anti-CD80 (BB1), PE-anti-
CD86 (FUN-1), FITC-anti-CD40 (5C3) (all from Pharmingen, San Diego, CA), PE-anti-
CD14 (L243), PE-anti-HLA-DR (mfP9), and PE- and fluorescein isothio-cyanate
(FITC)-conjugated isotype controls (all from Becton Dickinson), PE-anti-CD83
(Immunotech), and FITC-anti-HLA class I (Serotec Inc., Raleigh, NC). Of course, this
selection should merely be regarded as a suggestion, and in many cases several alterna-
tive antibodies that also work very well will be available.
3. The origin and/or composition of other materials used is described under Subheading 3.
3. Methods
The methods described below outline (1) the production of high-titer batches of
rAd; (2) the isolation of immature DC from peripheral blood mononuclear cells
(PBMC); (3) the infection of DC with rAd; and (4) the phenotypical and functional
analysis of uninfected and rAd-infected DC.
3.1. Production of High-Titer rAd
Efficient gene transfer into human DC with rAd derived from human adenovirus
type 5 can require an MOI of up to 1000 (4). It is therefore essential to use rAd stocks
of high titer (>1 1010 plaque-gorming units (pfu) per mL) and purity that have
been isolated by two-step cesium chloride (CsCl) density centrifugation. Although the
introduction into the DC of highly immunogenic vector-derived antigens is inherent to
the procedure of rAd-mediated gene-transfer, the balance between the transgenic and
vector components can be optimized by making use of rAd batches that have a mini-
mal content of defective rAd particles. To check the quality of CsCl-purified rAd
batches, the concentration of viral particles (VP/mL) can be assessed next to the con-
86 Offringa et al.
centration of pfu/mL. High-quality rAd batches typically display a VP:pfu ratio of

smaller than 30. Please note that a VP:pfu ratio of 30 does not mean that 29 out of 30
viral particles are defective. Instead, this value indicates that during standard infection
conditions, using highly permissive target cells, 1 out of 30 viral particles will be able
to elicit a productive infection (7). It should be noted that the ratio obtainable may
depend on the rAd construct used, and can be negatively affected if the transgene
encodes a gene product that interferes with efficient rAd production. Furthermore,
care should be taken to refrain from repeated freeze-thawing of the purified rAd stocks,
as this may lead to irreversible vector inactivation and aggregation.
Virus stocks are routinely tested for the presence of RCA by polymerase chain
reaction (PCR). The absence of RCA is especially important if the rAd used harbor
transgenes encoding potentially hazardous proteins, such as (proto-) oncogenes or
growth factors, and/or if the rAd-mediated gene transduction is part of a procedure
that is aimed at clinical applications.
All handling of rAd and rAd-infected cells must be performed in functional
biosafety cabinets in an appropriate biological containment laboratory.
3.1.1. Large-Scale rAd Multiplication
PER.C6 cells are the host of choice for generating batches of rAd that are free of
RCA (8). As alternatives cell lines 911 (9) or 293 (10) helper cells can be used, pro-
vided that suitable RCA checks are performed on all stocks. PER.C6 cells are rou-
tinely split in a ratio of 1:5 twice a week. When setting up cultures for infection, a ratio
of 1:4 is recommended. As an example, use six 15-cm dishes (133 cm2) of PER.C6
cells to seed 20 T-180 flasks (180 cm2). This quantity of cells will usually produce 1
20 1010 pfu of virus, depending on the vector and insert.
Approximately 3 d after seeding, at the time when each flask contains 3 107 cells,
PER.C6 is infected with an MOI of 3 to 5.
1. Add 2 109 pfu virus to 120 mL infection medium (see Subheading 2.1.), and mix well.
2. Replace the medium in the flasks with 6 mL of inoculum and gently rock the flasks to
spread the medium over the cells.
3. Incubate for 2 h at 37C in a well leveled 5% CO2 incubator. To enhance viral uptake it is
recommended to gently rock the flasks several times during these 2 h.
4. Add 12 mL/flask of postinfection medium (see Subheading 2.1.) and put flasks back into
incubator.
Usually 4048 h postinfection a cytopathic effect (CPE) is achieved. All cells are
rounded up, the medium has changed from red to orange/yellow, and about half of the
cells can be detached by gentle agitation of the flasks. It is important to check the cells
regularly, because some rAd require a longer incubation time to induce CPE. Harvest-
ing of virus-infected cells is performed as follows:
1. Gently rock flasks until all cells are detached. Avoid heavy shaking and formation of
foam and aerosols.
2. Transfer medium with cells to 50-mL conical tubes. Centrifuge at 250g for 5 min.
3. Remove supernatant. This medium usually contains approx 15 108 pfu/mL of rAd. It
can be stored for future use, for instance as viral stock for starting up new, large-scale
amplifications.
4. Use 40 mL of this supernatant to collect all cell pellets into one 50-mL tube, and centri-
fuge this tube again.
5. Remove supernatant, resuspend cells in 3 mL PBS supplemented with 2% HS, and trans-
fer to smaller conical tubes.
6. Release the virus from the cells by processing the cell suspension through three rapid
freeze/thaw cycles, centrifuge at 1100g for 10 min, and collect supernatant.
This rAd preparation can be stored at 20C and can be used for CsCl purification,
infecting a new batch of cells, or for RCA screening by PCR. Occasionally, a large
fraction of the rAd produced is not associated with the cell pellet, but rather resides in
the supernatant. This virus can be concentrated from the supernatant with ammonium
sulphate precipitation (11).
3.1.2. CsCl Purification of rAd
The rAd preparation is first purified over a discontinuous CsCl gradient. For this
purpose use open (no cap) 10-mL Ultra Clear tubes (Beckmann, cat. no. 344059).
1. First pour 2.0 mL of the high-density CsCl (1.45 g/cm3) solution in each tube. Carefully
overlay with 4.5 mL of the second CsCl (1.20 g/cm3) solution on top of the first one. The
final virus purity depends on the quality of the gradient.
2. Finally, load your viral stock on top of the gradient. This volume should be less than 4 mL,
and less than 1 109 cells, otherwise the gradient will be overloaded.
3. Centrifuge at 30,000 rmp (110,000g) in a Beckman Sw41 rotor for 2 h at 16C (brake
switched off).
4. Carefully place the rotor into a biosafety cabinet, take out the tubes, and fix these in a
clamp attached to a stand.
5. After centrifugation, the gradient should reveal a pellet of high-density cell components
at the bottom, a turbid layer of low-density cell components at the top, and two bands in
the middle. The upper band is yellow-white and consists of defective viral particles. The
lower one, bluish-white, contains the infectious virus.
6. Puncture the tubes just below the lowest band containing the virions using a 2.5-mL sy-
ringe with a 19G needle. Gently aspirate the virus. Avoid collecting other bands and
impurities, and keep the volume as small as possible (usually less than 2 mL).
The virus is further purified on a continuous CsCl gradient:
7. Transfer the virus isolated from first gradient to a 14-mL Quick Seal tube (Beckman, cat.
no. 342413).
8. Fill up the tube with additional 1.33 g/cm3 CsCl solution.
9. Spin sealed tubes for approx 18 h at 48,000 rpm (160,000g) in a Beckman 70.1Ti rotor
(brake switched off).
10. After centrifugation, only one clear band should be visible. Isolate this band as described
in step 6, but before extracting the virus, puncture a hole in the top of the tube to allow the
inflow of air into the sealed tubes.
For dialysis of the virus isolate, a dialysis cassette (Pierce; Slide-A-Lyzer Cassette;
10,000 MWCO) is used, which can hold 0.53.0 mL of solution. For smaller volumes
(approx 0.1 up to 0.5 mL), dialysis tubes (Pierce; Slide-A-Lyzer Mini Dialysis Unit;
10,000 MWCO) are used.
11. Fill the dialysis cassette using a 5-mL syringe with a 23G needle. Dialyze against 500 mL
TD dialysis buffer at 4C. Refresh the buffer at the end of the day, and continue dialysis
overnight.
88 Offringa et al.
Table 1
Dilution Scheme for Virus Titration
Volume of infection Volume of virus (L)
Dilution 10Ex medium in tube (mL) [from tube 10X]
2 1.98 20 [stock]
3 1.80 200 [102]
4 1.80 200 [103]
5 1.80 200 [104]
6 1.80 200 [105]
7 1.80 200 [106]
8 1.80 200 [107]
9 1.80 200 [108]
10 1.80 200 [109]
11 1.80 200 [1010]
12. The next morning, replace buffer with a 5% sucrose buffer. Dialyze for 23 h (not longer).
13. Take the virus from the cassette with a 5-mL syringe with a 23G needle.
14. Aliquot in small portions of 100 L or less. In addition, make two or three portions of
15 L, which can be used to determine virus titer in plaque assays.
15. Store aliquots at 80C as soon as possible.
3.1.3. Analysis of rAd Titer
Analyses of rAd titers are best performed with 911 cells, as these cells, in contrast
to Per.C6 and 293, grow well under agar overlays (9). From a subconfluent culture of
911 cells, seed six wells of a six-well plate for each plaque assay and allow the cells
to adhere overnight. Next day, make sure that the cultures are approx 8090%
confluent. Infect cells with six different dilutions of the viral stock, starting with one
of the 15-L aliquots. When testing CsCl-purified preparations, it is recommended to
test dilutions in the range of 106 to 1011. For nonconcentrated rAd preparations, test
dilutions in the range of 104 to 1010 (see Table 1 for dilution scheme).
Infect host cells with viral dilutions as follows:
1. Remove all medium from the wells and replace by 1.0 mL of the selected viral dilutions;
six serial dilutions are tested on one six-well plate. Spread medium over surfaces of the
wells.
2. Incubate for 2 h at 37C in 5% CO2 incubator. To enhance viral uptake it is recommended
to gently rock the plate several times during these 2 h.
3. Subsequently, the inoculum of each well is replaced by 2 mL of a 1:1 agar/F15 overlay.
The solutions for this overlay need to be prepared in advance as follows: Prepare a 1.7%
agar, pH 7.4 (see Subheading 2.1.). Before use, while incubating the cells with viral
dilutions, melt the agar in a microwave oven. Keep the molten agar at 42C in a water
bath. Also prepare 2X F15 complete medium (2xF15c). This should be prepared fresh (on
d of use) (see Table 2). Keep at 37C in a water bath. To prepare the overlay, mix melted
agar and prepared 2xF15c in 1:1 ratio (2.0 mL needed for each well).
Table 2
Scheme for Preparation of 2X F15 Complete
(2XF15c) Medium and 1:1 Agar/F15 Overlay a
Volumes per number
of plaque assays 1 2 3 4 5 6 7 8
2xF15 (mL; see Subheading 2.) 7 15 20 25 30 40 45 50
Glutamine 30g/L (L) 140 300 400 500 600 800 900 1000
MgCl2 1 M (L) 175 375 500 625 750 1000 1125 1250
Horse serum (HS) (L) 280 600 800 1000 1200 1600 1800 2000
Pen./strep. (L) 50 50 100 100 150 150 200 200
Fungizone 250 g/mL (L) 50 50 100 100 150 150 200 200
2xF15c (mL) 7.6 16.3 21.8 27.2 32.7 43.6 49.0 54.5
Agar pH 7.4 (mL) 7.6 16.3 21.8 27.2 32.7 43.6 49.0 54.5
Agar/2xF15c (mL) 15.2 32.6 43.6 65.5 65.4 87.2 98.0 109.0
a The upper part of this table lists the volumes of the various stock solutions (in L, unless indi-
cated otherwise) that need to be added to make up the amount of 2xF15c sufficient for 18 plaque
assays (each plaque assay consists of 6 wells on a 6-well plate). After mixing, the 2xF15c should be
warmed to 37C in a waterbath. The lower part of the table lists the volume 2F15c obtained. Just
before overlaying the wells, add 1 volume of agar pH 7.4, prewarmed at 42C, mix and apply to the
wells (2 mL/well).
Plaques will start to appear from d 5. At d 712, depending on the virus used, the
number of plaques will reach its maximum. Calculate the titer as follows: number of
plaques/dilution = plaque forming units per mL (pfu/mL). For CsCl-purified rAd
batches, expect titers to range between 1 108 and 1011; for nonconcentrated batches,
expect titers between 1 105 and 109.
3.1.4. Determination of the Concentration of VP/mL
The viral particle concentration of CsCl-purified adenovirus (after dialysis, in 5%
sucrose buffer) can be determined by an OD260 nm measurement. Before measurement,
the virus needs to be inactivated with NaOH/SDS.
Measurement of virus sample (for cuvets with a capacity of 100 L):
1. Add 20 L of a virus sample to an Eppendorf tube filled with 60 L 5% sucrose buffer
and 20 L of an 0.1 M NaOH/1% SDS solution.
2. Transfer to a quartz cuvet.
3. Measure optical density (OD ) at a wavelength of 260 nm.
4. Use a solution of 80 L sucrose + 20 L of NaOH/SDS as a blank.
5. If the OD reading exceeds the linear range, repeat measurement with an appropriate dilu-
tion of the virus stock sucrose buffer.
Calculation:
1.0 OD260 nm = dilution factor 1.1 1012 particles/mL
90 Offringa et al.
Table 3
PCR Reactions
For the polymerase chain reactions (PCR), add per tube:
Volume (L) Solution Stock concentration
4.5 PCR buffer, Perkin Elmer 10X
3.5 MgCl2 25 mM
2.0 dNTP 10 mM total (2.5 mM each)
2.0 F1 forward primer 10 pmol/L
2.0 R2 reverse primer 10 pmol/L
1.0 tRNA 5 mg/mL
0.3 Taq-polymerase, Perkin Elmer 5 U/L
33.7 Milli Q
3.1.5. RCA Screening

Screening of rAd stocks for RCA is routinely performed by PCR on nonconcen-
trated preparations isolated from freeze-thawed rAd-infected cells that result from the
procedure described under Subheading 3.1.1. PCR reactions on purified or medium
samples are less sensitive. Before performing PCR reactions, the virus samples need
to be inactivated. This will be done by a proteinase K (prot. K) treatment, followed by
heat inactivation.
1. Heat inactivation: in 1.5-mL Eppendorf tube, mix 45.0 L virus sample with 5.0 L 10X
PCR buffer (Perkin Elmer) and 0.5 L prot. K (40 mg/mL) (ICN).
2. Incubate at 44C overnight.
3. Tighten caps of tubes very well by plastic multicap locks or tape, and incubate the tubes
in boiling water for 5 min.
4. Centrifuge for 5 min at maximum speed in Eppendorf centrifuge.
5. The supernatant can be directly used in the PCR assay or stored at 20C until further use.
6. The PCR reactions are performed with primers that are specific for the Ad5E1 region,
which should be absent for rAd, but which is present in RCA as well as in the host cells
used for virus multiplication:
Forward primer: 5'-GGG.TGG.AGT.TTG.TGA.CGT.G (19-mer)
Reverse primer: 5'-TCG.TGA.AGG.GTA.GGT.GGT.TC (20-mer)
These primers correspond to nt 5270 and 674693 of the adenovirus type 5 sequence
(GenBank gi:33694637), respectively. PCR reactions are set up according to the scheme
in Table 3. Add 1.0 L of virus sample into one of these tubes. Add 1.0 L of water to
one of the tubes as negative control. DNA isolated from a wild-type Ad5 stock can be
used as a positive control. A series of 1, 10, 100, and 1000 pfu/L (or similar range) of
control DNA is to be used. Perform 30-cycle PCR reactions as follows:
Step 1: block temp 94.0C, time 30 s.
Step 2: block temp 63.0C, time 45 s.
Step 3: block temp 72.0C, time 1 min (8 min for final cycle).
7. Analyze PCR products by electrophoresis on a standard 2% agarose gel. Before loading

onto the gel, remove tRNA from samples by adding 1 L RNAse (10 mg/mL stock) and
incubating at 37C for 10 min. If a sample is contaminated with RCA, a band of 640 bp will
appear. The sensitivity of the assay can be checked on the basis of the wild-type Ad5 con-
trol samples. The 10-pfu sample should result in a visible band, while the samples contain-
ing 100 pfu and more should give rise to heavy bands. Note: this PCR reaction is not suitable
for 293-derived rAd stocks, because 293 cells contain the entire amplified region.
3.2. Isolation of Immature, Monocyte-Derived DC

PBMC from healthy blood donors constitutes the most readily accessible source of
DC precursors: the monocytes. For the proper interpretation of most immunologically
oriented experiments, it is advisable to make use of PBMC from fully HLA-typed
blood donors.
3.2.1. Isolation of Human PBMC From Buffy Coat
We routinely isolate PBMC from a 5060 mL buffy coat, derived from 500 mL of
peripheral blood, as follows:
1. Add 15 mL Ficoll-Hypaque solution (1.077 g/L) to each of six Leucosep tubes (Greiner,
cat. no. 227290) and spin tubes for 2 min at 650g to force the ficoll under the filter.
2. Apply 1015 mL (maximally; do not overload gradient) onto the filter and fill up tubes
with 2530 mL of PBS. Mix buffy coat and PBS by pipetting the solution up and down.
3. Centrifuge at room temperature for 10 min at 650g, brake set to low.
4. Collect the cells from the ficoll/PBS interphase in a 50-mL tube, fill up this tube with
PBS, mix and spin down cells for 5 min at 350g.
5. Wash the cells with PBS two more times.
6. For healthy donor buffy coats, this isolation typically results in the isolation of 510 108
PBMC that are free of erythrocytes. We routinely freeze these PBMC in aliquots of 0.5
1 108 cells.
3.2.2. Isolation of CD14+ Monocytes

Before proceeding, it is important to consider how many DC will be needed. On
average, the preparation of 10 106 immature DC requires the isolation of 20 106
monocytes, which requires approx 200 106 PBMC. Please note that recovery rates
may vary and are certainly expected to be lower if one intends to isolate DC from
diseased subjects such as cancer patients.
Although monocytes can be isolated from both fresh and frozen PBMC samples,
best results are usually obtained when starting from freshly isolated cells, in particular
because thawed cells tend to form clumps that will negatively affect the efficiency of
the isolation procedure. Therefore, we generally use part of the PBMC to immediately
proceed with the monocyte isolation, while the remainder of the PBMC is frozen in
aliquots to be used as autologous responder cells in subsequent immunological assays.
In order to obtain a highly homogenous DC population, we highly recommend iso-
lating monocytes through MACS with anti-CD14-coated beads, as follows:
1. Spin down the required amount of PBMC in a 50-mL conical tube (5 min at 350g). It is
recommended to start with at least 20 106 PBMC, because with lower amounts the
isolation procedure tends to become less efficient.
92 Offringa et al.
2. To the cell pellet add MACS buffer (see Subheading 3.2.2.) and CD14 beads (Milteny
Biotech, cat. no. 130-050-201). Per 1 107 PBMC add respectively 90 L and 10 L of
these compounds.
3. Resuspend cells and incubate at 4C (not on ice) for 30 min.
4. Fill up the tube with MACS buffer (04C), spin down cells in a cooled centrifuge, and
resuspend cell pellet in MACS buffer: 1 mL of buffer per 2 108 cells.
5. Place an LS MACS column (Milteny Biotech, cat. no. 130-042-401) into the magnetic
device, and equilibrate it with 3 mL of MACS buffer. Elute column and apply cell sus-
pension to column.
6. Wash column with three loads of 3 mL MACS buffer.
7. Remove column from magnetic device and elute the beads with the CD14-positive cells
with 3 mL of MACS buffer (see Subheading 3.2.2. for expected recovery rates). At
this stage, the monocytes can either be differentiated into DC, or aliquotted and frozen
for future use. We have not found that the presence of the magnetic beads has any
negative influence on monocyte growth, function, survival, or their capacity to differ-
entiate into DC.
3.2.3. Preparation of Monocyte-Derived DC

In order to allow the CD14+ monocytes to differentiate into DC, set up cultures in a
six-well plate (2 mL culture per well) of 1 106 cells/mL in RPMI medium supple-
mented with 10% FBS, 500 U/mL interleukin (IL)-4 (Sanvertech), and 800 U/mL
granulocyte/macrophage colony-stimulating factor (GM-CSF; Leucomax, Novartis).
After 2 d, add 2 mL of fresh medium to each well containing 1000 U/mL IL-4 and
1600 U/mL GM-CSF. After another 2 d of culture, take 2 mL of medium out off each
well and replace it by 2 mL of fresh medium containing 1000 U/mL IL-4 and 1600
U/mL GM-CSF. Repeat this procedure after another 2 d.
Fully differentiated, immature DC are obtained after 67 d of culture. The quality
of this cell preparation can be analyzed through FACS analysis for a limited set of
critical surface markersin particular, CD1a, CD14, CD83, CD86, and HLA class II
(see Subheading 2.2. for suitable antibodies). The desired DC population should ex-
clusively consist of cells that are positive for CD1a and negative for CD14 and CD83,
while expressing low levels of CD86 and moderate levels of CD80, CD40, HLA-class
II, and HLA-class I (4). The presence of a subpopulation of CD14+ cells usually indi-
cates that the differentiation process was incomplete, while increased levels of CD86
and CD83 usually indicate that a fraction of the DC has (partially) matured.
3.2.4. rAd Infection of DC

For most applications, it is preferable to infect DC in their immature stage (see
Note 3). After 67 d of differentiation, the immature DC cultures can be infected with
rAd as follows:
1. Harvest cells, transfer to conical 1015 mL tube and spin down for 5 min at 650g.
2. Resuspend cell pellet in 10 mL of serum-free RPMI, count vital cells, and spin cells again.
3. Resuspend cells in serum-free RPMI at a concentration of 2.5 106 cells/mL.
4. Plate cell suspension in 24-well plate at 200 L/well and add the desired amount of rAd
suspension to each well. In our hands, optimal results (maximal gene transduction effi-
ciency in combination with minimal cytotoxicity) for rAd5- and rAd2-based vectors is
usually obtained at an MOI of 5001000. To reach this MOI, add 2.5 108 pfu rAd per
well. Infection is performed in a minimal volume to optimize infection efficiency. In
view of this, we recommend using rAd stocks with high titer. Efficient rAd-mediated
gene transduction can also be obtained at lower MOI if one makes use of certain modified
rAd vectors (see Note 2).
5. Incubate for 6 h at 37C in tissue-culture incubator.
6. After 6 h, add to each well 1 mL of RPMI medium supplemented with 10% FBS, 500 U/
mL IL-4, and 800 U/mL GM-CSF. If full maturation of rAd-transduced DC is desired,
certain DC-maturation agents can be added to the cultures at this stage (see Note 4).
We generally perform phenotypical and functional analysis of our rAd-infected DC
on the following d, after overnight culturing. Assessment of the results of rAd infec-
tion involves two primary analyses:
7. The efficiency of rAd-mediated gene transduction is preferably analyzed through FACS
analysis, and may require extracellular or intracellular staining of the DC, depending on the
transgen-encoded antigen. In order to validate the rAd-mediated gene-transfer procedure,
rAd vectors encoding either EGFP or -galactosidase are ideal, because gene-transfer effi-
ciency can be readily assessed at the level of individual cells without the need for antibody-
mediated detection. At MOI of 5001000, gene transduction efficiencies are expected to
range between 50 and 80%. However, due to the character of the transgene-encoded protein
concerned, certain rAd constructs may result in lower transduction efficiencies, a lower
recovery of vital rAd-infected DC, and/or a decreased lifespan of these DC.
8. Analysis of a limited set of surface markers, in particular CD83, CD86, and HLA-class II.
Our previous work has shown that rAd infection of DC induces partial maturation of
these cells, as compared to the full maturation induced by LPS or CD40 ligand, in that the
expression levels of CD86 and HLA-class II are strongly increased, whereas only very
modest levels of CD83 expression are reached (4) (see Notes 3 and 4 for additional infor-
mation on DC maturation and polarization).
We use our rAd-infected DC for functional analyses and immunological experi-
ments anywhere between 18 and 48 h after rAd infection. Although for most applica-
tions we expect this timing not to be critical, certain applications may require the use
of rAd-infected DC at an earlier time point, 618 h after infection (see Note 3).
4. Notes
1. An overview of the cumulative published work on rAd-transducted DC lies beyond the
scope of this chapter, but information on this subject can readily be obtained through a
MedLine/PubMed search with the Boolean term adenovirus AND dendritic. This search,
which at present results in more than 350 hits, can be narrowed down, for instance by
providing in addition the name of the transgene of interest.
2. The initial attachment of rAd to the majority of human cell types is mediated by its
fiber capsid protein, which binds to the high-affinity coxsacky-adenovirus receptor
(CAR). The absence of CAR on human DC is in accordance with the fact that rAd
infection of these cells requires high MOI (4). In view of this notion, several labs have
constructed rAd vectors with modified fiber proteins that target other cell-surface pro-
teins that are expressed on DC, and thereby allow efficient rAd infection of DC at 100-
fold lower MOI (1217) and references therein). Alternatively, bispecific antibodies
directed against fiber and DC-surface markers, in particular CD40, have be success-
fully used to reduce the effective dose of rAd (see ref. 18 and references therein).
94 Offringa et al.
A major advantage of the use of lower MOI is that a lower dose of adenoviral antigens
is applied to the DC and, therefore, that the balance between the immunogenicity of
transgene-encoded and rAd-derived antigens is shifted towards the transgene-encoded
antigens of interest (see also Subheading 3.2.4.). Furthermore, these modifications can
be used to improve rAd-mediated gene targeting of DC in vivo.
3. It is generally preferable to infect DC at the immature stage in view of the following
considerations. Although both immature and mature DC can be infected to high efficien-
cies with rAd, efficiencies reached with immature DC are usually higher (4). If rAd infec-
tion is intended to introduce presentation of transgene-encoded antigenic peptide epitopes
into surface HLA molecules, access to especially the HLA class II processing route is
best before DC maturation (1). If DC infected with rAd-encoding antigens of interest are
to be injected in vivo with the aim of inducing (or modulating) T-cell immune responses
against these antigens, most injection routes (except intranodal injection) will require the
DC to home to lymphoid organs where encounter with antigen-specific T-cells can take
place. DC migration is guided by the programmed expression of chemokine and adhesion
receptors, which evolves during DC maturation (19). It is conceivable that fully matured
DC have already lost some of the chemotactic homing qualities that they displayed at
an earlier stage after receiving their maturation triggers (20). Moreover, if the rAd-
infected DC are to be used for induction of T-helper (Th)1-type CD4+ T-cell responses
and/or Th1-dependent cellular immunity (i.e., by CD8+ cytotoxic T-lymphocytes), maxi-
mal secretion of the Th1-polarizing cytokine IL-12 is desired. DC that have been trig-
gered for maturation secrete large quantities of this cytokine during the first 2436 h, but
lose this capacity afterwards (21).
4. As originally described by us (4) and confirmed in studies by many others, rAd infection of
DC induces partial maturation of DC in the absence of polarization. The latter essentially
means that DC triggering by rAd infection, unlike through ligation of its CD40 receptor,
does not cause the DC to secrete IL-12. When DC are to be used for inducing potent T-cell
immunity against rAd-transgene-encoded antigens, the latter is usually desired. Therefore, it
is usually recommendable to induce full maturation of rAd-infected DC by stimulating their
CD40 receptors or their Toll-like receptors (22). The combination of these triggers with rAd
infection will result in an exacerbation of the matured DC phenotype and secretion of very
high levels of IL-12 (4,12). When inducing DC maturation, either by rAd infection and/or
other stimuli, one should always keep in mind that certain important features of DC matura-
tion such as expression of homing receptors and cytokine genes are only transiently induced,
and are lost at later time points after initiation of the maturation process (see Note 3).
References
1.
2.
2 Russell, W. C. (2000) Update on adenovirus and its vectors. J. Gen. Virol. 81, 2573604.
3. St. George, J. A. (2003) Gene therapy progress and prospects: adenoviral vectors. Gene
Ther. 10, 11351141.
4.
4 Rea, D., Schagen, F. H., Hoeben, R. C., et al. (1999) Adenoviruses activate human den-
dritic cells without polarization toward a T-helper type 1-inducing subset. J. Virol. 73,
10,24510,253.
5.
5 Nevins, J. R., Imperiale, M. J., Kao, H. T., Strickland, S., and Feldman, L. T. (1984)
Detection of an adenovirus E1A-like activity in mammalian cells. Curr. Top. Microbiol.
Immunol. 113, 1519.
6.
6 Amalfitano, A., Hauser, M. A., Hu, H., Serra, D., Begy, C. R., and Chamberlain, J. S.
(1998) Production and characterization of improved adenovirus vectors with the E1, E2b,
and E3 genes deleted. J. Virol. 72, 926933.
7.
7 Nyberg-Hoffman, C., Shabram, P., Li, W., Giroux, D., and Aguilar-Cordova, E. (1997)
Sensitivity and reproducibility in adenoviral infectious titer determination. Nat. Med. 3,
808811.
8.
8 Fallaux, F. J., Bout, A., van der Velde, I., et al. (1998) New helper cells and matched early
region 1deleted adenovirus vectors prevent generation of replication-competent adeno-
viruses. Hum. Gene Ther. 9, 19091917.
9.
9 Fallaux, F. J., Kranenburg, O., Cramer, S. J., et al. (1996) Characterization of 911: a new
helper cell line for the titration and propagation of early region 1deleted adenoviral vec-
tors. Hum. Gene Ther. 7, 215222.
10.
10 Graham, F. L., Smiley, J., Russell, W. C., and Nairn, R. (1977) Characteristics of a human
cell line transformed by DNA from human adenovirus type 5. J. Gen. Virol. 36, 5974.
11.
11 Schagen, F. H., Rademaker, H. J., Rabelink, M. J., et al. (2000) Ammonium sulphate
precipitation of recombinant adenovirus from culture medium: an easy method to increase
the total virus yield. Gene Ther. 7, 15701574.
12.
12 Rea, D., Havenga, M. J., van Den Assem, M., et al. (2001) Highly efficient transduction of
human monocyte-derived dendritic cells with subgroup B fiber-modified adenovirus vec-
tors enhances transgene-encoded antigen presentation to cytotoxic T cells. J. Immunol.
166, 52365244.
13.
13 Belousova, N., Korokhov, N., Krendelshchikova, V., et al. (2003) Genetically targeted
adenovirus vector directed to CD40-expressing cells. J. Virol. 77, 11,36711,377.
14.
14 Fontana, L., Nuzzo, M., Urbanelli, L., and Monaci, P. (2003) General strategy for broad-
ening adenovirus tropism. J. Virol. 77, 11,09411,104.
15.
15 Gaggar, A., Shayakhmetov, D. M., and Lieber, A. (2003) CD46 is a cellular receptor for
group B adenoviruses. Nat. Med. 9, 14081412.
16.
16 Okada, N., Masunaga, Y., Okada, Y., et al. (2003) Dendritic cells transduced with gp100
gene by RGD fiber-mutant adenovirus vectors are highly efficacious in generating anti-
B16BL6 melanoma immunity in mice. Gene Ther. 10, 18911902.
17.
17 Vogels, R., Zuijdgeest, D., van Rijnsoever, R., et al. (2003) Replication-deficient human
adenovirus type 35 vectors for gene transfer and vaccination: efficient human cell infec-
tion and bypass of preexisting adenovirus immunity. J Virol. 77, 82638271.
18.
18 de Gruijl, T. D., Luykx-de Bakker, S. A., Tillman, B. W., et al. (2002) Prolonged matura-
tion and enhanced transduction of dendritic cells migrated from human skin explants after
in situ delivery of CD40-targeted adenoviral vectors. J. Immunol. 169, 53225331.
19.
19 Sallusto, F. and Lanzavecchia, A. (2000) Understanding dendritic cell and T-lymphocyte
traffic through the analysis of chemokine receptor expression. Immunol. Rev. 177, 134140.
20.
20 Watanabe, S., Kagamu, H., Yoshizawa, H., et al. (2003) The duration of signaling through
CD40 directs biological ability of dendritic cells to induce antitumor immunity. J.
Immunol. 171, 58285836.
21. Langenkamp, A., Messi, M., Lanzavecchia, A., and Sallusto, F. (2000) Kinetics of den-
dritic cell activation: impact on priming of TH1, TH2 and nonpolarized T cells. Nat.
Immunol. 1, 311316.
22. Kaisho, T. and Akira, S. (2002) Toll-like receptors as adjuvant receptors. Biochim.
Biophys. Acta 1589, 113.
DC for Patient-Specific Immunotherapy 97
8
Generation of Autologous Peptide- and Protein-Pulsed
Dendritic Cells for Patient-Specific Immunotherapy
David ONeill and Nina Bhardwaj
Summary
This chapter presents a detailed protocol for the generation of mature, monocyte-derived
dendritic cells (DC) that are loaded or pulsed with tumor-associated peptide antigens for use
as patient-specific immunotherapy. The protocol can easily be adapted for the treatment of
patients with a variety of tumors or chronic viral infections by simply changing the peptide
antigens used. The vaccine may use major histocompatibility complex (MHC) class I- or class
II-restricted peptides to elicit stimulation of CD8+ cytotoxic T-lymphocytes (CTL) or CD4+
T-helper cells, respectively. It also provides an example of loading DC with a purified protein
antigenin this instance, keyhole limpet hemocyanin (KLH). KLH may be used to boost the
immunogenicity of the vaccine and as a control to test for the induction of CD4+ T-helper-cell
responses. Other antigenic proteins may be added or substituted. Once prepared, the DC are
frozen in aliquots and samples are tested for identity, purity, viability, and sterility. After these
release criteria are met, vaccine aliquots can be thawed, drawn up into syringes, and injected
back into the patient.
Key Words: Cancer; dendritic cells; immunotherapy; melanoma; monocytes; vaccine.
1. Introduction
This protocol is adapted from the method of Thurner et al. for the large-scale
generation of monocyte-derived DC (Mo-DC) from leukapheresis products (1).
Mo-DC are the most commonly used DC preparation for clinical use, although DC
derived from CD34+ hematopoietic progenitor cells or enriched directly from peri-
pheral blood have also been used successfully (2,3). Our modifications to the Thurner
method include omitting an initial overnight incubation of the monocytes in the
absence of interleukin (IL)-4 and granulocyte/macrophage colony-stimulating factor
(GM-CSF), and the use of a cocktail of three proinflammatory cytokines (IL-1, IL-6,
tumor necrosis factor [TNF]-) and prostaglandin E2 (PGE2) to mature the DC in the
place of monocyte-conditioned medium (MCM). The cocktail, first described by
Jonuleit et al. (4) and referred to here as MCM mimic, allows potent, reproducible
DC maturation in less than 24 h (5).
97
The protocol begins with the preparation of frozen aliquots of peripheral blood
mononuclear cells (PBMC) and heat-inactivated plasma from apheresis products
obtained from the patient (1). To prepare the DC, frozen PBMC are thawed and
plated in medium containing 1% autologous plasma into 10-cm tissue-culture dishes.
Monocytes adhere to the plastic dishes during a 1-h incubation, after which the dishes
are washed to remove lymphocytes, which do not adhere to plastic. The monocytes
(CD14+ cells) are then induced to differentiate into immature DC by culturing for
5 d in the presence of IL-4 and GM-CSF. During this time they become non-adher-
ent and differentiate into immature DC (CD14 CD83 cells). The DC are harvested
then and transferred to six-well plates, where they are cultured for an additional d.
They are then pulsed with protein antigens and stimulated to mature by culturing for
another 18 to 24 h in the presence of MCM mimic. On the final d of culture the DC,
which have matured into CD14 CD83+ cells with characteristic cytoplasmic projec-
tions, or veils, are loaded with peptide antigens for 1 h, washed, and frozen in
aliquots in a controlled-rate freezing device. After testing control aliquots for speci-
fied release criteria, the cells may be thawed and injected directly into the patient
from whom they were prepared. Following this protocol we routinely obtain cell
yields of 5 to 10% relative to the number of PBMC plated, with 80 to 90% of the
cells having morphologic and phenotypic characteristics of mature DC.
Earlier methods for the generation of Mo-DC for clinical use relied on the use of
whole human blood as starting material (6,7). The advantage of using leukapheresis
products as opposed to whole blood is that over 3 billion PBMC can routinely be
obtained from one product. Whole blood yields approx 1 million PBMC per mL of
blood drawn, so a 100-mL blood sample typically yields only 5 to 10 million mature
DC, enough for just one injection. Our protocol uses an open system of tissue-cul-
ture dishes and plates, although it is potentially adaptable to closed or semi-closed
culture systems using bags (8) or cell factories (9). There is no general consensus as
to the best culture medium to use. Various laboratories use plasma, serum, or even
serum-free medium. We currently use RPMI 1640 medium with 1% autologous heat-
inactivated plasma, and have not observed significant improvements in yield using
serum or serum-free medium.
DC-based vaccines are currently used primarily for the immunotherapy of cancer.
One very common method to generate a DC-based cancer vaccine is to load MHC-
restricted peptides carrying dominant epitopes of tumor-associated antigens directly
onto DC (2,3,10,11). HLA-A2-restricted peptides are commonly used, often in com-
binations. Other MHC class I-restricted epitopes and even MHC class II-restricted
T-helper epitopes are also available. Control peptides carrying dominant epitopes from
antigens the patient has been previously exposed to may also be included to test for
recall immune responses. Influenza virus peptides are commonly used for this purpose
(12). Peptides may be loaded onto DC before or after DC maturation, although in our
hands pulsing mature DC with peptides has resulted in somewhat better specific T-cell
stimulation in vitro.
DC are also commonly loaded with protein antigens to produce vaccines, and we
have included an example of this in the protocol. We use KLH, a purified protein that
is a powerful immunogen derived from a marine mollusk. It may be used to boost the
immunogenicity of the vaccine by inducing a T-helper response (proteins taken up
from the extracellular fluid are largely processed by the endosomal , or MHC class II,
pathway) (13,14). Because it is a xenoantigen, KLH may also be used as a control
antigen to test the vaccines ability to prime nave T-cells (2,10,12). The DC are pulsed
with KLH just prior to maturation, since immature DC are most adept at acquiring
complex antigens exogenously.
DC prepared following this protocol may also be loaded with antigens by RNA
transfection (15,16) or by co-culturing with whole-cell lysates (10,17,18), immune
complexes (1921), or apoptotic cells (2224). Regardless of the antigen source, it is
important to mature the DC prior to injection into patients, since there is increasing
evidence that injection of antigen-loaded immature DC results in ineffective immuni-
zation or even immune tolerance (25,26).
For clinical use, many groups prepare DC fresh from PBMC prior to each vaccine
injection. One problem with this approach is that freshly prepared cells cannot be
fully tested for sterility prior to injection into the patient. For improved safety, reli-
ability, and efficiency of large-scale production, several laboratories have begun
preparing DC-based vaccines in large batches that can be aliquotted and tested for
sterility prior to use (27). In our hands, cryopreserved DC work just as well as fresh
DC in in vitro assays measuring antigen-specific T-cell stimulatory capacity and
migration in response to chemokines (unpublished observations). We have outlined
our method for DC cryopreservation in this protocol.
2. Materials (see Note 1)

2.1. Special Equipment (see Note 2)
1. Controlled-Rate Freezer (Planer Kryo 10 Series III, Planer Products, Sunbury-on-Thames,
UK).
2. Portable 50-L liquid nitrogen dewar with roller base and transfer hose (Cryo-Cyl 50 LP;
Minnesota Valley Engineering, Burnsville, MN).
2.2. Supplies
1. Ten-cm tissue-culture dishes (Falcon 353003; Becton Dickinson, Franklin Lakes, NJ).
2. Six-well tissue-culture plates (Falcon 353046, Becton Dickinson).
3. Fifteen-mL conical tubes (Falcon 352097, Becton Dickinson).
4. Fifty-mL conical tubes (Falcon 352098, Becton Dickinson).
5. Conical tubes, 250 mL (430776, Corning, Corning, NY).
6. Tube support cushions, 250 mL (430236, Corning,).
7. Thirteen-mL sterile screw cap tubes (60.540, Sarstedt, Newton, NC).
8. Nunc CryoTube vials, 1.8 mL (375418, Nalge Nunc, Rochester, NY).
9. One-mL Nunc CryoTube vials (377224, Nalge Nunc).
10. Millex-LG 25-mm syringe-driven filter unit (SLLG025SS, Millipore, Bedford, MA).
2.3. Blood Products (see Note 3)

1. Leukapheresis product (200 mL peripheral blood mononuclear cells, obtained from patient).
2. Plasmapheresis product (100 to 200 mL, obtained from patient).
2.4. Cell Culture Reagents (see Note 4)

1. RPMI 1640 medium with L-glutamine (12-702F, Cambrex, Walkersville, MD).
2. X-VIVO 15 serum-free medium with gentamicin and phenol red (04-418Q, Cambrex).
3. One M HEPES-buffered saline (17-737E, Cambrex).
4. Phosphate-buffered saline (PBS) without Ca++ and Mg++ (17-516F, Cambrex).
5. Trypan blue vital stain, 0.4% (17-942E, Cambrex).
6. Albumin (human) 25%, USP (Plasbumin-25, Bayer, Elkhart, IN).
7. Ficoll-Paque Plus (Amersham Biosciences, Piscataway, NJ).
8. DMSO, USP (Cryoserv, CRY-VB00, Edwards Life Sciences, Irvine, CA).
9. Sterile 0.9% sodium chloride for injection, USP (Abbott Laboratories, North Chicago, IL).
10. Gentamicin, 10 mg/mL (American Pharmaceutical Partners, Los Angeles, CA).
2.5. Cytokines
1. Leukine (recombinant human GM-CSF), 500 g/mL, 5600 IU/g (Berlex, Seattle, WA).
2. Recombinant human IL-4, lyophilized, 50 g vial, 6700 IU/g (CellGenix, Gaithers-
burg, MD).
3. Recombinant human TNF-, lyophilized, 25 g vial (CellGenix).
4. Recombinant human IL-1, lyophilized, 50 g vial (CellGenix).
5. Recombinant human IL-6, lyophilized, 150 g vial (CellGenixD).
6. One mg/mL PGE2 (Prostin E2, Pharmacia & Upjohn, Puurs, Belgium).
2.6. Protein Antigen
1. VACMUNE (keyhole limpet hemocyanin subunits, 20 mg/mL, Biosyn, Carlsbad, CA).
2.7. MHC-Restricted Peptide Antigens
1. Pharmaceutical-grade peptides carrying MHC class I- or MHC class II-restricted immuno-
dominant epitopes for a variety of viral or tumor-associated antigens may be obtained
aliquotted in lyophilized form from Clinalfa, Laufelfingen, Switzerland. Other compa-
nies also provide this service.
2.8. Solutions
1. Freezing solution 1 (X-VIVO 15 medium with 10% human albumin and 10% DMSO):
Mix 25 mL of X-VIVO 15 medium with 20 mL of 25% human albumin and 5 mL of
DMSO. Sterile filter using a Millex-LG filter unit attached to a 60-mL syringe. Chill the
filtered solution on ice if using right away or store refrigerated (see Note 5). Expires in
1 mo.
2. RPMI/1% autologous plasma (RPMI 1640 medium with 1% autologous heat-inactivated
plasma, 10 mM HEPES-buffered saline, and 20 g/mL gentamicin): Mix 489 mL
of RPMI-1640 medium with 5 mL of thawed autologous heat-inactivated plasma, 5 mL
of 1 M HEPES-buffered saline, and 1 mL of 10 mg/mL gentamicin. Sterile filter and store
refrigerated (28C). Expires in 8 d. We use autologous plasma in place of fetal calf
serum so that the cells may be injected back into human subjects (FDA does not permit
biologic products from ruminants to be used in the preparation of therapeutic cells).
3. PBS/5% albumin. Mix 40 mL of phosphate-buffered saline with 10 mL of 25% human
albumin. Sterile filter and store refrigerated. Expires in 1 mo.
4. 400 IU/L IL-4. Dissolve the contents of a frozen 50-g vial of lyophilized IL-4 in
840 L PBS/5% albumin. Mix thoroughly but gently. Do not vortex excessively. Store
refrigerated (28C). Keep this solution on ice when using. The solution expires in 8 d.
5. 100 IU/L GM-CSF. Combine 36 L of leukine with 964 L of PBS/5% human albumin
in a sterile 1-mL Nunc tube. Mix thoroughly but gently. Do not vortex excessively. Store
refrigerated (28C). Keep this solution on ice when using. It expires in 8 d.
6. RPMI/1% autologous plasma with cytokines. For every 10 mL of medium with cytokines
needed, combine 10 mL RPMI/1% autologous plasma with 5 L of 400 IU/L IL-4 and
10 L of 100 IU/L GM-CSF. Mix thoroughly but gently. The final concentration of
cytokines in this solution is 200 IU/mL of IL-4 and 100 IU/mL of GM-CSF. Use this
solution immediately (do not prepare and store in advance).
7. RPMI/1% autologous plasma with 10X cytokines. For every 1 mL of medium with 10X
cytokines needed, combine 1 mL RPMI/1% autologous plasma with 5 L of 400 IUL
IL-4 and 10 L of 100 IU/L GM-CSF. Mix thoroughly but gently. This is a 10X work-
ing solution with a concentration of 2000 IU/mL IL-4 and 1000 IU/L GM-CSF. It is
used to refresh the cytokines in dendritic cell cultures on d 3. Use this solution immedi-
ately (do not prepare and store in advance).
8. Frozen stock solutions of TNF-, IL-1, and IL-6. Working aseptically, reconstitute lyo-
philized TNF- in 0.5 mL of PBS/5% albumin (concentration of 50 g/mL). Store frozen
(20 to 80C) in 20-L aliquots. Reconstitute lyophilized IL-1 in 1 mL of PBS/5%
albumin (concentration of 50 g/mL). Store frozen (20 to 80C) in 20-L aliquots.
Reconstitute lyophilized IL-6 in 1 mL of PBS/5% albumin (concentration of 150 g/mL).
Store frozen (20 to 80C) in 200-L aliquots.
9. 100X monocyte conditioned medium mimic (100X MCM mimic): Combine 200 L of
PGE2 (Prostin E2) with 1560 L of PBS/5% albumin in a 1.8-mL Nunc tube. Mix thor-
oughly (the alcohol in Prostin E2 will cause some of the albumin to precipitate, but this
goes back into solution after thorough mixing). Use this solution to rinse out one vial
each of frozen TNF-, IL-1, and IL-6 stock solutions, pooling back into the 1.8-mL
Nunc tube. The final volume should be 2 mL. 100X MCM mimic contains 0.5 g/mL
TNF-, 0.5 g/mL IL-1, 15 g/mL IL-6, and 100 g/mL PGE2.
10. Freezing solution 2 (90% autologous plasma, 10% DMSO): Mix 9 mL of autologous
plasma with 1 mL of DMSO. Sterile filter using a Millex-LG filter unit attached to a
10-mL syringe. Chill the filtered solution on ice if using right away or store refrigerated
(see Note 5). Expires in 8 d.
3. Methods (see Note 6)

3.1. Preparation of Frozen Stocks of Autologous Peripheral
Blood Mononuclear Cells (PBMC) and Heat-Inactivated Plasma
3.1.1. PBMC (allow 4 to 5 h)
1. Fill the portable 50-L dewar with liquid nitrogen to between 1/2 and 3/4 full. Prepare
freezing solution 1 as outlined under Subheading 2.8. and pre-heat a water bath to 56C.
2. Record the volume stated on the leukapheresis product bag. Store the plasmapheresis
product bag at 4C until ready to process.
3. After gentle mixing, sterilely transfer the entire contents of the leukapheresis bag into a
labeled 250-mL conical tube. Record the volume collected.
4. Distribute the leukapheresis product in 15-mL aliquots into labeled 50-mL conical tubes.
Then add 25 mL of RPMI 1640 medium to each tube, cap, and gently mix.
5. For each tube, remove the cap and gently layer 12 mL of Ficoll-Paque Plus beneath the
cell suspension, starting by placing the tip of your pipet at the bottom of the tube. Be
careful not to disturb the interface as it develops and not to blow bubbles into the Ficoll
layer. Re-cap the tubes.
6. Centrifuge the tubes at 900g for 30 min at room temperature (20C) with no brake.
7. For each tube, gently aspirate the cloudy PBMC layer with a new pipet. Transfer to new
labeled 250-mL tubes, pooling the aspirated PBMC (add up to 60 mL of pooled PBMC
per tube). When finished, discard the 50-mL tubes containing the residual Ficoll, granu-
locytes, red blood cells, plasma, and platelets.
8. Dilute the contents of each tube of pooled PBMC with approx 3 vol of RPMI 1640
medium. Cap and mix gently by inversion.
9. Centrifuge at 400g for 10 min at room temperature (20C) with full brake.
10. Remove the supernatant and resuspend each pellet in approx 100 mL RPMI 1640 medium.
Centrifuge again at 400g for 10 min at room temperature (20C) with full brake (see
Note 7).
11. Resuspend each pellet in a small volume of RPMI 1640 medium and pool all of the cells
into one 50-mL tube. Bring up to a final volume of 50 mL with more medium.
12. Thoroughly mix to ensure a uniform cell suspension by capping the tube tightly and
inverting the tube several times. Transfer 10 L of cell suspension to a small tube con-
taining 990 L of medium. Mix thoroughly. Mix 10 L of this diluted suspension with
10 L of 0.4% Trypan blue solution and count the number of viable cells using a hema-
cytometer. Be sure not to count residual red cells and platelets.
13. Calculate the percent viable PBMC, the number of viable PBMC/mL, and the total num-
ber of viable PBMC. Then calculate the number of 1-mL aliquots to be frozen assuming a
concentration of 200 million viable PBMC/mL (see Note 8). Prepare labels for this many
aliquots and apply the labels to sterile 1.8-mL Nunc cryovials.
14. Prepare the controlled-rate freezing device so that it is ready to perform a run, following
the manufacturers instructions. This usually involves attaching the liquid nitrogen dewar
using a hose connection and initializing the machines freezing program. It should take
2 to 3 min for the machine to stabilize at its start temperature (4C).
15. Centrifuge the PBMC at 400g for 10 min at 4C.
16. Aspirate the supernatant and loosen the pelleted cells by flicking or slapping the tube.
Resuspend the cells at a final concentration of 200 million cells/mL in cold freezing solu-
tion 1.
17. Aliquot 1 mL of cell suspension per tube into the labeled Nunc tubes.
18. Load the filled cryovials into the freezer chamber of the mechanical freezing device and
start the freezing run.
19. At the end of the run, remove the frozen samples and place in a transfer container with
crushed dry ice. Transfer the cryovials immediately into a liquid-nitrogen freezer for long-
term storage (cells for clinical use should be stored in the vapor phase of the freezer, not
submerged in liquid nitrogen).
3.1.2. Heat-Inactivated Plasma (allow at least 1 h)

1. Record the volume stated on the plasmapheresis product bag.
2. After gentle mixing, sterilely transfer the entire contents of the bag into labeled 50-mL
conical tubes.
3. Centrifuge the 50-mL tubes at 2000g for 20 min at 4C to pellet platelets.
4. Remove the supernatant plasma, transferring 40-mL volumes into labeled 50-mL conical
tubes.
5. Heat-inactivate the plasma by incubating in a water bath at 56C for 35 min.
6. Centrifuge the heat-inactivated plasma at 2000g for 20 min at 4C to remove any precipi-
tated material.
7. Transfer 10-mL volumes of the supernatant to labeled 13-mL sterile Sarstedt tubes. Place
the tubes in a 80C freezer for long-term storage.
3.1.3. Quality-Control Testing of Frozen Intermediates

1. Aliquots of PBMC and plasma should be tested for sterility and for the presence of bacte-
rial endotoxin. It is best to outsource sterility testing to a qualified microbiology labora-
tory. A commercially available kit (Limulus amoebocyte lysate assay, Cambrex) may be
used to test for the presence of endotoxin. Since high concentrations of plasma interfere
with the endotoxin assay, samples should be diluted at least 1:10 before testing, and a
sensitive kinetic chromogenic assay should be used.
3.2. Dendritic Cell Processing: Day 0 (allow 3 to 4 h)

1. Thaw one 10-mL aliquot of autologous heat-inactivated plasma in a 37C water bath
with constant agitation. Make sure the cap of the tube is fully tightened while thawing.
Periodically remove the tube from the water bath during this process and invert the
tube several times to mix the contents thoroughly. When completely thawed, put the tube
on ice.
2. Prepare two 500-mL batches of RPMI/1% autologous plasma as directed in Subhead-
ing 2.8.
3. Prepare PBS/5% albumin as outlined in Subheading 2.8.
4. Prepare a 400 IU/L (60 ng/L) working solution of interleukin-4 (IL-4) by dissolving
the contents of a frozen vial of lyophilized IL-4 in PBS/5% albumin as outlined in Sub-
heading 2.8.
5. Prepare a 100 IU/L (18 ng/L) working solution of GM-CSF by diluting a refrigerated
aliquot of Leukine in PBS/5% albumin as described in Subheading 2.8.
6. Thaw six vials (1.2 billion cells) of frozen autologous PBMC in a 37C water bath with
constant agitation. Make sure the caps of the tubes are fully tightened while thawing.
When completely thawed, sterilely transfer and pool the thawed cells into a labeled ster-
ile 50-mL tube containing 30 mL of RPMI/1% autologous plasma. Mix well.
7. Centrifuge the cells for 5 min at 500g. Pipet off the supernatant and loosen the pellet by
slapping the bottom of the tube with your fingers until the pellet is loosely coating the
lower portion of the tube. Resuspend the cells in 30 mL RPMI/1% autologous plasma,
tighten the cap of the tube, and mix thoroughly by swirling the tube and by inversion (do
not vortex cells excessively).
8. Centrifuge the cells again for 5 min at 500g. Pipet off the supernatant, loosen the pellet as
described above, and resuspend in 50 mL RPMI/1% autologous plasma. Tighten the cap
of the tube and mix well by inversion.
9. Dilute 10 L of cell suspension with 190 L RPMI/1% autologous plasma in a 1-mL
Nunc tube. Mix well. Remove 10 L of this diluted cell suspension and mix with 10 L
0.4% Trypan blue solution for cell counting. Count the cells and calculate the percent
viability, the number of viable cells/mL, and the total number of viable cells.
10. Plate 35 million viable PBMC/dish on Falcon 353003 10-cm tissue-culture dishes (see
Note 9). Adjust the volume in each dish to 10 mL by adding RPMI/1% autologous plasma.
Mix the contents of each dish thoroughly by gently swirling and rocking.
11. Incubate the cells for 1 to 2 h at 37C in the tissue-culture incubator. Dont stack the
tissue-culture dishes for this step so that they can rapidly equilibrate to 37C.
12. During the incubation, pre-warm an unopened bottle of RPMI 1640 medium in a 37C
water bath. This will be used for washing off nonadherent cells.
13. Prepare RPMI/1% autologous plasma with cytokines as directed in Subheading 2.8. You
will need 10 mL for each tissue-culture dish (see Note 10). This will be used to culture the
adherent monocytes following the wash step.
14. When the incubation is complete, remove a tissue-culture dish from the incubator and
wash twicefirst with the medium already present in the plate, then with the prewarmed
RPMI 1640 medium. When washing, pipet up and down several times, gently rocking the
dish. When pipetting up, tilt the plate toward you while aspirating up the medium. When
pipetting down, direct the stream of medium over the entire surface of the dish by moving
your pipet. Dont scrape the surface of the dish with the pipet. Pipet out all of the medium
after each wash, but work quickly so that you dont let the cells dry out between washes.
15. After the second wash, add 10 mL of RPMI/1% autologous plasma with cytokines. Check
the dish under the inverted microscope after adding the medium with cytokines. There
should be very few freely floating cells (these can be identified by tapping the plate and
looking for moving cells). You may need to wash again if many cells are still free-float-
ing. When finished, return the dish to the tissue-culture incubator on a new shelf.
16. Repeat the previous two steps (washing twice, adding medium with cytokines, and return-
ing to the incubator) with all of the tissue-culture dishes.
17. Incubate the plates for 3 d at 37C in the tissue-culture incubator.
3.3. Dendritic Cell Processing: Day 3 (allow at least 30 min)

1. Prepare RPMI/1% autologous plasma with 10X cytokines as directed in Subheading 2.8.
You will need 1 mL for each tissue-culture dish.
2. View the cells under an inverted microscope and assess on a scale of 0 (worst) to 5 (best)
for cell density, cell viability/relative lack of debris, relative number of non-adherent
cells, relative number of large cells, and for dendritic or veiled appearance (see Note 11).
3. Add 1 mL of medium with 10X cytokines to each plate. Mix each plate well by swirling
and gentle rocking.
4. Incubate the plates for two more days at 37C.
3.4. Dendritic Cell Processing: Day 5 (allow 1 to 2 h)

as described in Subheading 3.3., Note 2 (see Note 11).
2. Harvest the cultures by pipetting up and down with a 10-mL pipet to resuspend all non-
adherent cells (see Note 11). Transfer and pool the harvested non-adherent cells into a
labeled 250-mL tube. Discard the empty tissue-culture dishes.
3. Cap the tube and record the total volume of harvested cells. Mix the tube thoroughly.
Remove 10 L of cell suspension and mix with 10 L of 0.4% Trypan blue solution for
cell counting.
4. Count the cells using a hemacytometer and calculate the percent cell viability, the number
of viable cells per mL, and the total number of viable cells. Calculate the d 5 yield of
viable cells by dividing the total number of viable cells harvested by the number of cells
plated on d 0 (read Note 12 before proceeding).
5. The harvested cells are to be washed, resuspended at approx 500,000 cells/mL in me-
dium with cytokines, and then plated at 3 mL/well in six-well tissue-culture plates (see
Note 13). Calculate the volume (in mL) of medium you will need to resuspend the cells
in by dividing the total number of viable cells harvested by 500,000 (see Note 13).
Measure out this amount of RPMI/1% autologous plasma and add IL-4 and GM-CSF to
make RPMI/1% autologous plasma with cytokines as outlined in Subheading 2.8.
6. Centrifuge the cell suspension at 500g for 5 min. Discard the supernatant and loosen the
pellet(s) as described previously. Resuspend the cells in all of the medium with cytokines
prepared and mix thoroughly.
7. Transfer 3 mL of cell suspension per well into labeled six-well plates. Incubate the plates
in the 37C tissue-culture incubator overnight.
3.5. Dendritic Cell Processing: Day 6 (allow at least 1 h)

as described in Subheading 3.3., Note 2.
2. Add 1.5 L of 20 g/L KLH to each well (final KLH concentration of 10 g/mL). Mix
the cultures thoroughly by swirling the plates. Return the plates to the tissue-culture incu-
bator.
3. If not already done, prepare frozen stocks of TNF-, IL-1, and IL-6 as outlined in Sub-
heading 2.8.
4. Prepare 100X monocyte-conditioned medium mimic (100X MCM mimic) as described in
Subheading 2.8. You will need 30 L for each well. Keep 100X MCM mimic on ice
while working with it (see Note 14).
5. Add 30 L of 100X MCM mimic to each well. Mix thoroughly by swirling the plates.
Return the plates to the tissue-culture incubator and incubate overnight.
3.6. Dendritic Cell Processing: Day 7 (allow at least 5 h)

1. Fill the portable 50-L dewar with liquid nitrogen to between 1/2 and 3/4 full.
2. Prepare freezing solution 2 (90% autologous plasma, 10% DMSO) as outlined in Sub-
heading 2.8. Chill on ice.
3. Prepare labels for Nunc cryovials, with five labels designated for injection and two
labels designated QC. Apply the labels to sterile 1.8-mL Nunc cryovials.
as described in Subheading 3.3., Note 2.
5. Return the plates to the 37C incubator.
6. Aseptically transfer 1 mL sterile 0.9% NaCl, USP, to a sterile 1-mL Nunc tube labeled
saline.
7. Aseptically transfer 1 mL DMSO, USP, to a sterile 1-mL Nunc tube labeled DMSO.
8. Reconstitute lyophilized peptides at a concentration of 1 mM (for MHC class I-restricted
peptides, this would be a concentration of approx 1 g/L, since most class I-restricted
peptides have 9 to 10 amino acids and a molecular weight of approx 1000 Da). This
results in a 1000X peptide working solution (final concentrations of peptides in culture
will be 1 M). Use saline to reconstitute peptides that are soluble in water and DMSO
for peptides that are poorly soluble in water (solubility for each peptide must be deter-
mined empirically). For peptides obtained frozen in solution, thaw quickly in a 37C
water bath.
9. Keep all peptide solutions on ice except for those brought up in DMSO (DMSO will
freeze when placed on ice).
10. Remove the plates from the incubator. On the lid of each plate, use a marking pen to mark
each well with the name of the peptide it will receive. Each well is to receive only a single
peptide, 3 L per well (see Note 13).
11. Add the peptides to their designated wells, 3 L per well. Mix the cultures thoroughly
by swirling the six-well plates. The final peptide concentration should be 1 M (see
Note 15).
12. Place the plates back in the 37C incubator and incubate for at least 1 h.
Table 1
Resuspending Day 7 DC in Freezing Solution 2
Day 7 cell yield Vol. freezing Final DC conc.
(millions) solution 2 (mL) (millions/mL)
130 8.7 15
120 8.0 15
110 7.3 15
100 6.7 15
90 6.0 15
80 6.2 13
70 6.4 11
60 6.0 10
50 6.2 8
40 6.7 6
30 6.0 5
13. When the incubation is finished, use a 10-mL pipet to harvest the non-adherent cells from
each plate (by pipetting up and down as described for d 5). Pool the harvested cell suspen-
sion into a labeled sterile tube of appropriate size.
14. Centrifuge the harvested cells at 500g for 5 min. Pipet off the supernatant and loosen the
pellet. Resuspend the cells in 50 mL RPMI/1% autologous plasma. Transfer to a new
50-mL tube if desired.
15. Centrifuge again at 500g for 5 min. Pipet off the supernatant, loosen the pellet, and resus-
pend the cells in 50 mL sterile PBS.
16. Cap the tube and mix thoroughly. Remove 10 L of cell suspension and mix with 10 L
of 0.4% Trypan blue solution for cell counting.
17. Count the cells using a hemacytometer and calculate the percent cell viability, the number
of viable cells per mL, and the total number of viable cells for each batch. Calculate the
d 7 yield of viable cells by dividing the total number of viable cells harvested by the
number of cells plated on d 0.
18. Prepare the controlled-rate freezing device so that it is ready to perform a run, following
the manufacturers instructions. This usually involves attaching the liquid-nitrogen dewar
using a hose connection and initializing the machines freezing program. It should take 2
to 3 min for the machine to stabilize at its start temperature (4C).
19. Centrifuge the cells again at 500g for 5 min. Pipet off the supernatant and loosen the
pellet. Resuspend the cells in 10 mL sterile 0.9% NaCl, USP. Transfer to a new labeled
15-mL conical tube and centrifuge again at 500g for 5 min. Discard the supernatant and
resuspend the cell pellet in freezing solution 2 as outlined in Table 1.
20. Mix the cell suspensions thoroughly. Transfer 1 mL of cell suspension to each Nunc tube
designated for injection, and transfer the remaining cell suspension equally to the two
remaining tubes designated QC.
21. Load the filled cryovials into the freezer chamber of the mechanical freezing device and
start the freezing run.
22. At the end of the run, remove the frozen samples and place in a transfer container with
crushed dry ice. Transfer the cryovials immediately into a liquid-nitrogen freezer for long-
term storage (cells for clinical use should be stored in the vapor phase of the freezer, not
submerged in liquid nitrogen).
Table 2
Sample Controlled-Rate Freezing Program
Step Start temp. End temp. Rate Comment
a. 4C 6C 1C/min Chamber temp. lowered to 6C at 1C/min
b. 6C 6C Hold at 6C for 10 min.
c. 6C 50C 25C/min Chamber temp. lowered rapidly to absorb
heat of fusion.
d. 50C 14C +15C/min Chamber temp. allowed to approx sample temp.
e. 14C 60C 1C/min Sample temp. lowered to 60C at 1C/min.
f. 60C 90C 5C/min Sample temp. lowered to 90C at 5C/min.
g. Hold at 90C and signal end of run.
3.7. Quality-Control (Release) Testing

1. One QC aliquot for each batch of DC must be thawed and tested to ensure that the DC
preparation conforms to specifications for identity, purity, and viability before the cells
can be injected into human subjects (see Note 16). The second QC aliquot for each batch
is to be kept frozen in case future testing is needed.
3.8. Injection Preparation
1. Thaw an aliquot of DC designated for injection in a 37C water bath filled with sterile
water. Working in the BSC, use a blunt cannula to draw the cell suspension up equally
into two labeled 1-mL syringes. Place the syringes on ice and transport to the clinic for
injection (see Note 17).
4. Notes
1. We have not included in the materials list required items that are routinely found in a
modern cell-culture laboratory. This list would include a class II biological safety cabi-
net; a water-jacketed, 37C, 5% CO2 tissue-culture incubator; a refrigerated centrifuge
(Beckman Coulter Allegra 6KR or equivalent) with a rotor and buckets/carriers suitable
for 15-, 50-, and 250-mL conical tubes; a water bath; standard and inverted microscopes;
a vortex; a hemacytometer; a refrigerator; and 20C, 80C, and liquid-nitrogen freez-
ers. The laboratory should be supplied with assorted sterile filtration units, pipets and
pipetting devices, pipet tips, needles, syringes, and cell-culture flasks. Sources of liquid
nitrogen, dry ice, and wet ice should be available. We would also like to note that the
sources of itemized equipment, reagents, and supplies are only suggested; other manufac-
turers may offer equivalent products that give excellent results.
2. We have achieved superior results with cryopreservation of PBMC and dendritic cells
using a commercially available mechanical controlled-rate freezing device (Planer Kryo-
10). In our hands, antigen-loaded DC frozen using such a device have significantly better
immunostimulatory capacity than cells frozen using inexpensive isopropanol freezing
baths (such as the Nalgene Cryo 1C chamber). DC prepared using the controlled-rate
freezer are identical to freshly prepared DC when assessed for cell-surface phenotype,
immunostimulatory capacity, and migratory response to chemokines (unpublished obser-
vations). The freezer should be programmed by a qualified service technician authorized
by the manufacturer. A sample program is shown in Table 2. The program takes approx
75 min to run.
3. Leukapheresis and plasmapheresis products must be obtained by a qualified cell-collec-

tion facility or contractor. To obtain PBMC we typically request that the facility process
10 L of blood (approx two blood volumes) with the blood cell separator set to obtain
mononuclear cells. The cells and plasma should be processed as soon as possible follow-
ing collection, although we have obtained good results with products that have been
shipped overnight on wet ice.
4. Cytokines, peptides and protein antigens. In addition to cytokines from CellGenix, we
have obtained excellent results with cytokines from R&D Systems, Minneapolis, MN. It
is important that cytokines used to prepare clinical-grade DC have a detailed certificate of
analysis and have been tested for sterility. It is also desirable to purchase reagents that did
not come into contact with animal products during their production; otherwise, they may
need to be tested for the presence of adventitious viral agents before they can be used in
clinical trials. Clinalfa (Laufelfingen, Switzerland) is a good source for a variety of phar-
maceutical-grade peptides that carry CD8 or T-helper epitopes for viral and tumor-asso-
ciated antigens. Alternatively, if accompanied by a detailed certificate of analysis and
tested for sterility, peptides manufactured by somewhat less exacting standards may be
acceptable if they are not to be directly injected into patients. KLH is an animal product
and must be tested for the presence of adventitious viral agents before use in human
subjects. Biosyn KLH comes tested for adventitious viral agents and in the United States
is approved for use by the FDA.
5. For the preparation of freezing media we specify the use of the Millex-LG filter unit
because the membrane and filter housing are compatible with DMSO. Do not substitute
an incompatible filter when filtering any DMSO-containing solution. As an alternative,
since the DMSO stock solution is certified sterile, the solution may be prepared without
DMSO, then filtered, and then the DMSO can be added. Freezing solutions must be chilled
completely to 0 to 4C prior to use to achieve best results.
6. In the United States, clinical protocols using antigen-loaded dendritic cells for immuno-
therapy in human subjects must first be approved by the Food and Drug Administration
and your institutional review board. All cell-processing steps should be carried out in an
appropriate controlled environment (i.e., a clean room with HEPA-filtered air) following
current good manufacturing practices (cGMP). This is especially true when working with
an open culture system such as described here. Work with open vessels should be carried
out in a class II biological safety cabinet (BSC) using aseptic technique. Workers should
be fully gowned and gloved, including shoe covers, a head cover, and a face mask. Be
mindful that the finished cell product is for autologous use only: throughout the proce-
dure, all tubes and plates should be labeled with two unique identifiers (such as a patient
ID number and a product batch number) to avoid mix-ups and cross-contamination with
other products.
7. The three soft spins at 400g remove most of the platelets, which remain in the super-
natant.
8. We typically obtain yields of about 3 billion PBMC, enough for 15 frozen aliquots.
9. We plate 35 million cells per 10-cm tissue-culture plate because in our hands this is the
lowest number of cells that can reproducibly saturate the plate (these plates will only
yield a maximum of 5 to 6 million immature DC on d 5). There may be donor to donor
variability in the optimal number of cells to plate (we have plated up to 50 million PBMC/
plate), but the objective is always to use just enough cells to yield at least 5 million cells/
plate on d 5.
10. IL-4 and GM-CSF induce the differentiation of monocytes (large CD14+ cells) into
immature DC (which are large, nonadherent, CD14 CD83 cells). It is a good idea to
carry out titration experiments to determine the optimal amount of these cytokines
needed to achieve the desired effect. GM-CSF is required to keep the cells viable in
culture, whereas IL-4 prevents monocyte differentiation into macrophages. Macro-
phages are easily identified as large, adherent cells with a fried egg appearance. In
our hands, 100 IU/mL GM-CSF and 200 IU/mL IL-4 are sufficient to induce monocyte
differentiation into immature DC. Increasing the concentration of these cytokines has
no adverse effects, but tends not to improve the yield of immature DC on d 5.
11. It is important to understand the morphological changes to expect in the cell culture.
After washing out most of the lymphocytes on d 0, the remaining monocytes can be seen
adhering tightly to the plastic surface of the tissue-culture dish, many of them with a
flattened appearance. A fair number of adherent platelets will also be seen, which are
much smaller than monocytes or lymphocytes. Most of the platelets will not survive the
wk-long culture period, and platelets are not noticeably present by d 7. During the first
5 d of culture, the monocytes will detach from the plastic as they differentiate into imma-
ture DC. There will be a fair amount of cell death and debris visible, but it is reasonable to
expect a viable cell yield of 10 to 15% (relative to the number of PBMC plated) by d 5.
Immature DC are large, relatively round cells with few or no cytoplasmic projections.
They are two to three times the size of lymphocytes. Over 90% of the cells should be non-
adherent by d 5, with the vast majority having the morphological characteristics of imma-
ture DC (most of the remaining cells are lymphocytes). Once the cells are washed and
re-plated on d 5, much less cell debris should be seen. You should expect about two thirds
of the cells to survive from d 5 to d 7. By d 5, some of the DC may begin to show early
changes associated with DC maturationthe characteristic cytoplasmic projections (den-
drites) or veils. These changes are much more evident on d 7 following the overnight
maturation stimulus (culture in MCM mimic). By d 7, approx 80% of the cells in the
culture should be large, nonadherent cells with numerous cytoplasmic projections and/or
veils. Most of the remaining cells will be lymphocytes (small round cells). There will be
some cell debris and evidence of cell death, but this should not be a predominant feature.
It is reasonable to expect a 5 to 10% yield of mature DC (relative to the number of PBMC
plated) on d 7.
12. Planning for d 6. KLH is a very potent immunogen, so it may not be desirable to inject a
patient with KLH for every injection in a series of immunizations. In our current trial of
DC for immunotherapy of malignant melanoma, we prepare two parallel batches of DC,
one loaded with peptides and KLH to be used in the first injection, and one loaded with
peptides alone that is used for the following three injections. This way all of the injections
are prepared at the same time but in two parallel batches. If you wish to similarly inject
differently pulsed DC at different time points during a clinical trial, we recommend that
you also break the preparation up into parallel batches at this point.
13. It is a good idea to plan ahead for the peptide-loading step on d 7 at this point. If you plan
to pulse the DC with more than one peptide, it is best to pulse with only one peptide per
welli.e., do not pool peptides before adding to the DC culture. By not pooling the pep-
tides you avoid the possibility that some of the peptides may be preferentially recognized
on a given DC due to competition (28). To efficiently pulse DC with more than one
peptide, plate the day 5 DC into a number of wells that is a multiple of the number of
peptides you will be adding. This allows each peptide to be distributed to the same num-
ber of wells. You can adjust the concentration of DC to anywhere from 1.5 to 2 million
cells/well to ensure that you plate into the multiple of wells needed.
14. MCM mimic is a potent cocktail of proinflammatory cytokines and PGE2 that reproduc-
ibly results in high-level stimulation of dendritic cell maturation following overnight
(18 to 24 h) culture (4,5,9). Most laboratories use this cocktail in place of monocyte-
conditioned medium because it is better defined and provides more reproducible DC matu-
ration. Of its four components, we use final concentrations in tissue-culture medium of
5 ng/mL TNF-, 5 ng/mL IL-1, 150 ng/mL IL-6, and 1 g/mL PGE2.
15. The optimal concentration of peptides to use to load DC is not currently known. In our
hands, mature d-7 DC pulsed with peptide concentrations of 1 and 10 M are clearly able
to stimulate antigen-specific T-cell clones in vitro. We have obtained better immuno-
stimulatory capacity using a 1- to 2-h pulse on d 7 than with an overnight pulse beginning
on d 6 (unpublished results). A discussion on the dosing of DC with peptides may be
found in Chapter 1.
16. In the United States, the Food and Drug Administration (FDA) requires that therapeutic
cells be tested or validated for identity, purity, potency, and stability before they can be
used in human subjects. We have validated this method to ensure the identity, potency,
and stability of the cells by testing frozen/thawed DC for cell-surface phenotype (high-
level expression of HLA-DR, CD86, CCR7, CD83, negative staining for CD14), immuno-
stimulatory capacity, and viability under transport conditions. Frozen/thawed DC
routinely and specifically stimulate MHC-restricted antigen-specific T-cell clones in lym-
phocyte proliferation and interferon Elispot assays, with activity in these assays indis-
tinguishable from freshly prepared DC. The frozen/thawed DC are stable for at least 2 h
in freezing solution 2 when kept on ice. In addition, we test a QC aliquot of each batch of
DC for viability (Trypan blue staining), sterility (bacterial, fungal, and mycoplasma cul-
tures, mycoplasma DNA fluorescence), absence of bacterial endotoxin (Limulus amoeb-
ocyte lysate assay), and identity (flow cytometry). Plasma and DMSO will interfere with
the Limulus amoebocyte lysate (LAL) assay, so samples need to be diluted 10- to 20-fold
and tested using a sensitive kinetic chromogenic assay (this can be outsourced or done
with a commercially available kit from Cambrex). Our current QC (release) criteria
include >70% viable cells, negative results for all sterility and endotoxin tests, and >50%
of the cells having the characteristics of mature DC by flow cytometry (large, CD14
CD83+ cells). We routinely obtain close to 90% mature DC, with cell viability usually
>90%. Most of the remaining cells are lymphocytes.
17. The optimal route of administration of dendritic cell vaccines is still a matter of debate.
DC vaccines may be administered by subcutaneous, intradermal, or intravenous injec-
tion, or may be injected directly into lymph nodes or tumors. We currently administer
bilateral subcutaneous injections, 0.5 mL each, into the medial aspect of the upper arm (a
region that drains into the axillary lymph nodes). If the axillary lymph nodes have been
removed surgically, we inject into a similar location in the thighs. The freezing solution
contains 10% DMSO, USP, in autologous plasma, which is compatible with injection
into human subjects.
References
1 Thurner, B., Roder, C., Dieckmann, D., et al. (1999) Generation of large numbers of fully
1.
Immunol. Methods 223, 115.
2 Banchereau, J., Palucka, A. K., Dhodapkar, M., et al. (2001) Immune and clinical responses
2.
in patients with metastatic melanoma to CD34(+) progenitor-derived dendritic cell vaccine.
Cancer Res. 61, 64516458.
3 Fong, L., Hou, Y., Rivas, A., et al. (2001) Altered peptide ligand vaccination with Flt3
3.
ligand expanded dendritic cells for tumor immunotherapy. Proc. Natl. Acad. Sci. USA 98,
88098814.
4.
5.
5 Lee, A. W., Truong, T., Bickham, K., et al. (2002) A clinical grade cocktail of cytokines
and PGE(2) results in uniform maturation of human monocyte-derived dendritic cells:
implications for immunotherapy. Vaccine 20(Suppl 4), A8A22.
6 Bender, A., Sapp, M., Schuler, G., Steinman, R. M., and Bhardwaj, N. (1996) Improved
6.
methods for the generation of dendritic cells from nonproliferating progenitors in human
blood. J. Immunol. Methods 196, 121135.
7.
8.
8 Pullarkat, V., Lau, R., Lee, S. M., Bender, J. G., and Weber, J. S. (2002) Large-scale
monocyte enrichment coupled with a closed culture system for the generation of human
dendritic cells. J. Immunol. Methods 267, 173183.
9 Berger, T., Feuerstein, B., Strasser, E., et al. (2002) Large-scale generation of mature
9.
monocyte-derived dendritic cells for clinical application in cell factories. J. Immunol.
Methods 268, 131.
10.
11.
11 Schuler-Thurner, B., Schultz, E. S., Berger, T. G., et al. (2002) Rapid induction of
tumor-specific type 1 T helper cells in metastatic melanoma patients by vaccination with
mature, cryopreserved, peptide- loaded monocyte-derived dendritic cells. J. Exp. Med.
195, 12791288.
12.
12 Dhodapkar, M. V., Steinman, R. M., Sapp, M., et al. (1999) Rapid generation of broad
T-cell immunity in humans after a single injection of mature dendritic cells. J. Clin.
Invest. 104, 173180.
13.
14.
14 Bhardwaj, N. (2001) Processing and presentation of antigens by dendritic cells: implica-
tions for vaccines. Trends Mol. Med. 7, 388394.
15.
15 Mitchell, D. A. and Nair, S. K. (2000) RNA-transfected dendritic cells in cancer immuno-
therapy. J. Clin. Invest. 106, 10651069.
16.
16 Sullenger, B. A. and Gilboa, E. (2002) Emerging clinical applications of RNA. Nature
418, 252258.
17.
17 Geiger, J., Hutchinson, R., Hohenkirk, L., McKenna, E., Chang, A., and Mule, J. (2000)
Treatment of solid tumours in children with tumour-lysate-pulsed dendritic cells. Lancet
356, 11631165.
18.
18 Chang, A. E., Redman, B. G., Whitfield, J. R., et al. (2002) A phase I trial of tumor lysate-
pulsed dendritic cells in the treatment of advanced cancer. Clin. Cancer Res. 8, 10211032.
19.
19 Rafiq, K., Bergtold, A., and Clynes, R. (2002) Immune complex-mediated antigen presen-
tation induces tumor immunity. J. Clin. Invest. 110, 7179.
20.
20 Kalergis, A. M. and Ravetch, J. V. (2002) Inducing tumor immunity through the selec-
tive engagement of activating Fcgamma receptors on dendritic cells. J. Exp. Med. 195,
16531659.
21.
21 Schuurhuis, D. H., Ioan-Facsinay, A., Nagelkerken, B., et al. (2002) Antigen-antibody
immune complexes empower dendritic cells to efficiently prime specific CD8+ CTL
responses in vivo. J. Immunol. 168, 22402246.
22.
22 Larsson, M., Fonteneau, J. F., Somersan, S., et al. (2001) Efficiency of cross presenta-
tion of vaccinia virus-derived antigens by human dendritic cells. Eur. J. Immunol. 31,
34323442.
23.
23 Albert, M. L., Sauter, B., and Bhardwaj, N. (1998) Dendritic cells acquire antigen from
apoptotic cells and induce class I- restricted CTL. Nature 392, 8689.
24.
24 Kotera, Y., Shimizu, K., and Mule, J. J. (2001) Comparative analysis of necrotic and apop-
totic tumor cells as a source of antigen(s) in dendritic cell-based immunization. Cancer
Res. 61, 81058109.
25.
ture dendritic cells. J. Exp. Med. 193, 233238.
26.
26 Jonuleit, H., Giesecke-Tuettenberg, A., Tuting, T., et al. (2001) A comparison of two types
of dendritic cell as adjuvants for the induction of melanoma-specific T-cell responses in
humans following intranodal injection. Int. J. Cancer 93, 243251.
27. Feuerstein, B., Berger, T. G., Maczek, C., et al. (2000) A method for the production of
cryopreserved aliquots of antigen- preloaded, mature dendritic cells ready for clinical use.
J. Immunol. Methods 245, 1529.
28. Palmowski, M. J., Choi, E. M., Hermans, I. F., et al. (2002) Competition between CTL
narrows the immune response induced by prime- boost vaccination protocols. J. Immunol.
168, 43914398.
Characterization of Clinical-Grade DCs 113
9
Phenotypical and Functional Characterization
of Clinical-Grade Dendritic Cells
I. Jolanda M. de Vries, Gosse J. Adema,

Cornelis J. A. Punt, and Carl G. Figdor
Summary
Dendritic cells (DC) are the most potent antigen-presenting cells and form a promising new
treatment modality. Fully activated DC loaded with antigen are very useful in stimulating
immune responses, in particular those to combat cancer. Immature DC can either cause immu-
nological tolerance or induce regulatory T-cells, opening up future application in transplanta-
tion, autoimmunity, and perhaps chronic inflammation.
For these clinical applications, generation of DC under Good Manufacturing Practice condi-
tions and without the use of animal products is a main prerequisite. To date, there are many
different DC culture protocols, which subsequently lead to different types of DC. Protocols
differ not only in DC preparation techniques, but also many other variables, like maturation
status, dose and timing interval, route of administration, and antigen loading. DC vaccination,
although very promising, is far from standardized. In this chapter we discuss the usefulness of
standardized clinical and immunological criteria and the need for careful study design to further
optimize the use of dendritic cells and to unequivocally prove their efficacy.
Key Words: Quality control; Good Manufacturing Practice; maturation; migration; immu-
notherapy.
1. Introduction
A role for dendritic cells (DC) has been implicated in several diseases, including
cancer. DC loaded with tumor antigens have been shown to be potent inducers of
T-cell responses in vitro. The ability to culture large amounts of DC from monocytes in
vitro combined with the availability of class I-restricted peptides derived from tumor-
associated antigens, such as gp100, tyrosinase, MAGEs, and NY-ESO-1, gives the
opportunity to vaccinate cancer patients with peptide-loaded DC (15). Recently, DC
pulsed with tumor antigens have been successfully used in vivo for the induction of
antitumor T-cell reactivity in melanoma patients (1,2). Therefore, DC have the poten-
tial to be used as vaccine adjuvants in immunotherapy against cancer. Currently, many
113
114 de Vries et al.
clinical trials using DC-based immunotherapies are ongoing for the treatment of a wide
range of malignancies. It is not surprising that in these international trials, huge varia-
tions occur in both culturing of DC and in study design (6). DC can be directly isolated
from blood and by culturing from peripheral blood monocytes or CD34+ bone-marrow
cells in the presence of interleukin (IL)-4 and granulocyte/monocyte colony-stimulat-
ing factor (GM-CSF), tumor necrosis factor- (TNF-), stem cell factor, and FLT3
ligand, respectively (79).
So far, most DC vaccines have been used to stimulate immune responses, in par-
ticular those to combat cancer (10,11). Recent findings indicate that immature DC can
either cause immunological tolerance or induce regulatory T-cells, opening up future
application in transplantation, autoimmunity, and perhaps chronic inflammation. Fur-
thermore, generation of DC under Good Manufacturing Practice (GMP) conditions is
a main prerequisite for their use in clinical trials. In this chapter we will discuss cur-
rent issues relating to monocyte-derived DC (MoDC) culturing under GMP guidelines
as well as the function of these cells for specific goals.
2. Generation of Monocyte-Derived DC Under GMP Conditions

Today, all DC vaccines appear to be safe and nontoxic, with only occasional and
mild side effects. However, generation of MoDC according to GMP-guidelines, with-
out the use of animal products like fetal calf serum (FCS), is necessary for their use in
clinical trials. Especially the use of animal-derived products is reported by Mackensen
et al. to be of risk (12). They show the presence of antibodies to FCS and bovine serum
albumin (BSA) in patients vaccinated with DC cultured in FCS/BSA. Moreover, in
their study one patient had a type I hypersensitivity anaphylactic reaction after repeti-
tive vaccination with autologous peptide-pulsed DC. On the basis of these data, DC
cultured either in human serum or plasma or under serum-free conditions are recom-
mended for therapeutic applications in vivo.
To generate DC in the absence of FCS for clinical applications, many laboratories
set out to optimize the differentiation of monocytes to DC in the presence of clinical-
grade media either serum-free or supplemented with human serum/plasma (pooled or
autologous). This has led to the use of many different protocols and subsequently to
different types of DC. Protocols differ not only in DC preparation but also many other
variables, such as maturation status, dose and timing interval, route of administration,
and antigen loading. DC vaccination, although very promising, is far from standard-
ized, preventing the introduction of this type of therapy as standard cancer treatment.
To make this possible, there is a strong need for standardization and quality control
of ex vivo-generated DC and for the implementation of large two-pronged double-
blind clinical studies to unequivocally prove clinical efficacy. This is needed to let this
form of immunotherapy move away from bulky disease to patients with low tumor
burden. Most studies conducted so far concern late-stage cancer patients with meta-
static lesions. The current lack of standardization prevents the inclusion of early-stage
patients. In these patients, long-term side effects due to prolonged survival might be
foreseen, and therefore DC must be grown in molecularly well-defined culture media,
cytokine cocktails, and/or molecularly defined triggers for DC activation (e.g., toll-
like receptor ligands). Additionally, DC have to be cultured in closed culture systems,

which will simplify handling, thus limiting the use of expensive and labor-intensive
GMP facilities.
3. Culturing and Phenotypical Characterization

of Immature and Mature Monocyte-Derived DC
To isolate monocytes after a leukapheresis, several options are available. Counter-
flow centrifugation (centrifugal elutriation), adherence of peripheral blood mono-
nuclear cells to plastic, or isolation of CD14+ cells by magnetic sorting, using
clinical-grade anti-CD14 microbeads. We used counter-flow centrifugation (centrifu-
gal elutriation) and compared these DC with those generated from adherent peripheral
blood mononuclear cells. Both GMP-approved methods yield highly purified mono-
cytes, from which, with the use of IL-4 and GM-CSF, immature DC (7) could be
generated that express MHC I and II and low levels of costimulatory molecules. Strik-
ingly, maturation of the immature DC generated from elutriated monocytes using
either monocyte-conditioned medium (MCM) or MCM/prostaglandin (PG)E2/ TNF-
appeared to be difficult and resulted in a significantly lower yield. This lack of matu-
ration of elutriated DC might be due to the fact that the elutriated monocytes are first
activated by the leukapheresis process, a phenomenon also described by Thurner and
colleagues (2), and subsequently by the counter-flow centrifugation procedure, in
which they are subjected to shear stress.
The optimal maturation stimulus to obtain mature DC from immature DC is still
unclear. Recent clinical trials have tended to use either MCM (2) or the cytokine TNF-
, either alone (1315)or in combination with IL-1, IL-6, and PGE2 (1618).
Cytokines may be preferable to MCM, as they can be standardized and are molecu-
larly defined. The combination of TNF-, IL-1, IL-6, and PGE2 has become popular
as a result of analyses in vitro demonstrating enhanced capacity to stimulate T-cells
under FCS-free conditions (16). Despite the previously observed lack of IL-12 pro-
duction and subsequent Th1 development of PGE2-matured DC in vitro (FCS-cul-
tured) (19), after vaccination with MCM/TNF-/PGE2-matured DC we observed
polarization of T-cells towards interferon (IFN)- production in patients (20). Further-
more, exposure to PGE2 is required for optimal migration of mature DC towards the
chemokine receptor (CCR)7 ligands CC-chemokine ligand (CCL)19 and CCL21,
chemokines which attract DC to lymph nodes (21,22).
A significant issue to be taken into account when discussing DC maturation is the
possibility of exhaustion. DC that are maximally stimulated have a relatively short
window in which to produce Il-12on the order of 48 h (19,23). Maturation of DC
in vitro sets the clock ticking, and therefore it is hoped that, after migration of antigen-
expressing DC from the site of injection to the lymph node, these DC are still capable
of stimulating Th1-type tumor-specific response. The ideal maturation cocktail there-
fore may be one that induces migratory capacity yet leaves them responsive to further
stimulation by CD40 ligand through engagement of antigen-specific T-cells in vivo,
resulting in release of bioactive IL-12 in the microenvironment where it is most needed.
Maturation with IFN- appears to cause this effect (24), and more recent studies sug-
116 de Vries et al.
Table 1
General Quality Criteria for DC to be Used in Clinical Trials
Microbiological controls Negative for bacteria and fungal contamination
Viability >50% determined by tryphan blue exclusion
Purity >70% determined by light scatter or staining with non-DC
lineage markers
Morphology Immature: Adherent, stretched cells with some extensions
Mature: Loosely attached, veiled and clustered cells
Phenotype Immature: CD14neg/low, CD83neg, CD80neg/low, CD86low,
MHC class Ipos, MHC class IIpos, DC-SIGNpos,
CCR5pos
Mature: CD83pos, CD80pos, CD86pos, MHC class Ipos,
MHC class IIpos, DC-SIGNpos, CCR7pos
Induction of immune response Mixed lymphocyte reaction:
T cell proliferation at DC:PBMC ratio 1:20 in at least
one donor.
Recognition of the loaded antigen by T cells:
Cytotoxicity assay or cytokine production
gest that transient exposure to maturation stimuli may also be useful (25). Another
approach for the generation of mature DC is based on an alternate DC activation path-
way, using calcium-mobilizing agents to drive differentiation of fully mature DC from
monocytes in only 2 d (26). The validity of these approaches awaits testing in a con-
trolled clinical setting.
4. Quality Control
DC have to be evaluated for multiple DC markers and meet release criteria. There
is an urgent task for the research community to define these criteria. The DC prepara-
tions should be tested for sterility/pyrogenicity, viability (after thawing), purity, phe-
notype (surface antigens), migratory capacity, and function. Here, we would like to
introduce general quality criteria that DC should meet before use in a clinical setting
(Table 1, Fig. 1).
4.1. Sterility/Pyrogenicity, Viability, and Purity
Samples of the DC vaccine harvested at day 4 and on the day of the first DC vacci-
nation should be sent for microbiologic analysis for clearance.
For release, viability of the DC vaccine, determined for example by trypan blue
exclusion, should be more than 50%. This is a percentage that can easily be reached at
the first vaccination day, when DC are harvested, loaded with antigen, and then imme-
diately administrated to the patient. However, subsequent DC vaccines are often fro-
zen and thawed at the required vaccination day. In these samples, the viability is lower,
Fig. 1. Flow-cytometric analysis of immature dendritic cells and mature dendritic cells (cul-
tured from adherent monocytes). Gray lines represent isotype-matched controls, whereas the
dark overlays are the markers as indicated on the right.
but it is possible to reach the criterion of 50% when optimized freezing and thawing
methods are used (21,27). These methods describe viabilities up to 80% of frozen/
thawed DC (27).
The purity of the DC preparation can be evaluated based on light scatter (flow
cytometry) as well as staining for specific markers (non-DC lineage markers) for con-
taminating cells. Berger et al. have reported that DC preparations are often contami-
nated with T-cells (CD3+), NK cells (CD56+), and especially by B-cells (CD19+), but
they still reached purities of >85% DC (27). We propose that DC preparations with
purities of >70% can be released for clinical use.
118 de Vries et al.
4.2. Morphology, Phenotype, and Function

Before administration of DC, these cells should be evaluated extensively by flow
cytometry and microscopy. The morphology of the generated DC can be analyzed
under an inverted phase-contrast microscope. The typical morphology (nonadherent
veiled cells) of mature DC is easy to distinguish from the adherent and stretched mor-
phology of immature DC and from the very large round adherent macrophages (20).
To determine the stability of the DC, the wash-out test is proposed. In this test, DC
are washed and deprived of cytokines. Under these conditions, only fully mature DC
remain nonadherent, veiled, and viable (28).
Phenotypically, the quality of the mature DC vaccine could be monitored by flow
cytometry, evaluating surface expression of the maturation marker CD83, as well as
the costimulatory molecules CD80 and CD86. These markers should be low or even
absent on immature DC preparations. Both DC types should have high levels of DC-
SIGN and MHC class I and class II; on mature DC, an increase of the latter two mol-
ecules is preferred.
Migration of DC to lymph nodes is a prerequisite in the induction of an immune
response. Therefore the expression of CCR molecules, allowing directed migration, is
highly important. Immature DC do not express CCR7, but do express CCR5, whereas
mature DC are lacking CCR5 but do express (when fully mature) CCR7, allowing
them to migrate to the T-cell areas of lymph nodes and carry out their function (20
22).
Next to reaching the T-cells, DC samples should be able to stimulate T-cell prolif-
eration in a mixed lymphocyte reaction in vitro. A drawback of this assay is that it is
not only dependent on the quality of DC but also on the responder cells, in particular
the extent of the HLA-mismatch, and therefore the data generated in different indi-
viduals cannot easily be compared. However, DC should display a potent stimulatory
capacity at low DC/peripheral blood mononuclear cell ratios.
Mixed lymphocyte reactions provide information about the competence of DC to
stimulate an immune response. However, in immunotherapy we are aiming at induc-
ing immune responses towards well-defined antigens. Therefore, DC should be recog-
nized by cytotoxic T-cells with specificity for the loaded antigens. This should be
tested on freshly harvested and antigen-loaded DC as well as on frozen/thawed DC
with preloaded antigen. Since this testing requires a lot of the mostly rare cytotoxic T-
cells, we propose to load antigenic peptides after thawing of the DC to ensure optimal
peptide loading (21).
5. Functional Characterization of Immature and Mature DC
5.1. DC Function In Vivo
In vivo, immature DC originating from the bone marrow and present in
nonlymphoid tissues are potent in antigen uptake and processing (29). Upon antigenic
stimulation under inflammatory conditions, DC change their chemokine receptor ex-
pression pattern and migrate to the secondary lymphoid tissues (3032). Here, DC
become fully mature and secrete an array of chemokines to recruit B-cells, T-cells,
and other DC to efficiently induce an immune response (29,33).
5.2. Antigen Uptake and T-Cell Activation

DC can take up a wide array of antigens and present them to T-cells as peptides
bound to both MHC class I and II. They mediate diverse functions in the immune
system with high efficiency. These functional stages are related to the maturation state
of the DC. Immature cells are adept at endocytosis and express relatively low levels of
surface MHC class I and II and costimulatory molecules. Therefore, it is useful to
maintain cells in an immature state when preparing for antigen loading. Exogenous
antigens such as proteins or apoptotic cells are taken up by immature DC and pro-
cessed and subsequently retained for use as immunogenic peptides days later. After
stimulation with microbial products or inflammatory cytokines, immature DC trans-
form into mature DC. These cells have a reduced capacity for antigen uptake but are
extremely well equipped for T-cell stimulation: high MHC class I and II and
costimulatory molecules on their surface as well as T-cell adhesion molecules (34).
Maturation also leads to the extension of long dendrites, which may increase the capa-
city to capture T-cells. Furthermore, upon maturation DC remodel their chemokine
receptor repertoire and thereby gain the ability to migrate to lymph nodes and stimu-
late T-cells, the ultimate goal in DC-based immunotherapy (31).
5.3. Migration
Upon exposure to pathogens in peripheral tissues, resident immature DC become
activated through toll-like receptors (35) and take up and process antigen, while migrat-
ing to the draining lymph nodes to present their antigenic load (36). To facilitate migra-
tion, CCR5 is down-regulated, while CCR7 is upregulated, the latter being required for
trafficking and entry into the T-cell areas of the lymph node (31,3739,40).
DC injected into patients must actively migrate to lymph nodes and subsequently
into the T-cell areas of lymph nodes. Mature DC are migratory both in vitro and in vivo,
irrespective of whether they are administered into the skin or intranodally (20). By
contrast, in vitro-generated immature DC are unable to leave the skin after intradermal
injection. The low migratory capacity of immature DC relates directly to their strong
adhesive properties in vitro, which is mediated by highly expressed 1 integrins, in
particular 51 (20,41). Similar findings have been reported by Gunzer et al., who
show in a collagen matrix model that the immature state of murine DC is characterized
by low migration, whereas mature murine DC exhibit high migratory activity (42).
In vitro-generated monocyte-derived immature DC lack the chemokine receptor
CCR7, required for migration into the T-cell areas of lymph nodes (43,44). Hence,
immature DC generated in vitro do not co-localize and interact with nave T-cells in the
lymph nodes, a prerequisite for the induction of an effective immune response (20). In
contrast to immature DC, monocyte-derived mature DC express CCR7 and, as expected,
these cells migrate into the T-cell areas of the lymph nodes (20). The inability of imma-
ture DC to migrate into the T-cell areas could explain why vaccination with antigen-
loaded immature DC in melanoma patients fails to induce proliferative as well as DTH
responses. The failure of immature DC to induce an immune response correlates with
the observation that immature DC, when used as a vaccine adjuvant, give rise to T-cells
that display the typical properties of regulatory T-cellsnon-proliferative, IL-10-pro-
120 de Vries et al.
ducing T-cells that can dampen pre-existing antigen-specific effector T-cell function
(45,46).
5.4. Chemokine Production
DC are known to secrete a variety of chemokines that allow them to attract and
interact with lymphocytes to induce an immune response (40,61).We have showed
that the chemokine profiles of clinical-grade mature and immature DC are similar, as
previously observed for DC cultured in the presence of FCS (47,48), including high
expression of DC-CK1 that preferentially attracts nave T-cells in mature DC (21,49).
DC also express high levels of TARC and MDC, which have recently been reported to
attract memory T-cells in addition to NK- and T-helper-2 cells (5052).
6. Application of Immature and Mature DC

6.1. DC Immunotherapy to Induce Tolerance
An increasing number of preclinical studies focus on the capacity of immature DC
to induce antigen-specific unresponsiveness or tolerance following antigen capture.
DC in the steady state are immature and can silence immunity in an antigen-specific
manner through two recently identified mechanisms (11). In one tolerance mecha-
nism, due to the absence of DC maturation stimuli, interacting T-cells proliferate but
are soon deleted. Effector functions and, therefore, memory do not develop, resulting
in peripheral tolerance. A second tolerance mechanism involves the induction of
IL-10-producing T-cells by immature DC, which in turn can silence other effector
T-cells. Immature DC incubated with immunosuppressive agents such as dexametha-
sone (53), vitamin D (54), and more recently the Rel B inhibitor Bay (55) all result in
IL-10-producing T-cells and thereby in the induction of peripheral tolerance.
While most clinical DC-vaccination studies so far are aimed at stimulating immune
responses, the finding that immature DC are extremely suitable to silence immune
responses opens up new clinical applications in transplantation, allergy, autoimmu-
nity, and perhaps chronic inflammation.
6.2. DC Immunotherapy to Induce Immunity

Most DC vaccines have been generated from myeloid precursors or monocytes
grown in GM-CSF/IL-4, further matured with cytokine cocktails and loaded with anti-
gen ex vivo. Comparative studies with immature and mature DC clearly demonstrated
that only mature DC stimulate protective T-cell responses (20,45,46). Additionally,
only mature DC show enhanced migration to draining lymph nodes when compared
with immature DC (20). The inability of immature DC to induce an immune response
also indicates that monocyte-derived DC do not mature after injection in vivo (56).
Thus, for active immunization in cancer patients it appears to be critical that DC are
matured in vitro before injection (6).
Most studies conducted so far concern late-stage cancer patients with metastatic
lesions. Clinical results are variable, although in several studies remarkable long-last-
ing objective clinical responses were observed. Therefore, DC vaccination represents a
promising approach, which does not meet expectations yet due to the current lack of
knowledge. Evidence is accumulating that the dose and timing interval, route of admin-
istration, and antigen loading play a critical role in determining the quality and quantity
of the immune response (38,57). Especially the route of injection and the migration of
DC from the injection site might highly influence clinical outcome. We recently showed
that migration of DC is highly dependent on their maturation status (20). Furthermore,
we demonstrated that the majority of injected monocyte-derived DC matured in vitro,
reside at the injection depot, despite the fact that they express CCR7 and are highly
motile in vitro (20,5861). Gunzer et al. reasoned that the microenvironment largely
influences emigration (42). In addition, the relatively high local density of DC at the
site of injection may affect this microenvironment, thereby having a major impact on
emigration. Matrix metalloproteinases are good candidates to enhance the emigration
of DC (7). Certainly a better understanding of this lack of emigration is of importance
for future DC vaccine development.
With regard to optimization of antigen loading of DC, one should keep in mind that
monovalent antigen specificity is likely to be insufficient in patients with large tumor
burden, also in view of the tumor minus variants that arise. Vaccination against mul-
tiple tumor antigens or against tumor antigens plus tumor-associated antigens expressed
by stromal or endothelial cells in the tumor is currently being explored in preclinical
models. Vaccination with these mixtures of antigens could induce a combined immune
response that could have an additive if not synergistic effect.
7. Perspectives
To further optimize clinical efficacy, DC vaccination strategies need to be opti-
mized, standardized, and quality controlled. DC must be grown in molecularly well-
defined culture media, cytokine cocktails, and/or molecularly defined triggers of DC
activation. Most importantly, DC have to be cultured in closed culture systems, which
will simplify handling, thus limiting the use of expensive and labor-intensive GMP
facilities. In the future, implementation of DC in large two-pronged double-blind stud-
ies are necessary to unequivocally prove clinical efficacy.
References
1. Nestle, F. O., Alijagic, S., Gilliet, M., et al. (1998) Vaccination of melanoma patients with
peptide- or tumor lysate-pulsed dendritic cells. Nat. Med. 4(3), 328332.
2. Thurner, B., Haendle, I., Roder, C., et al. (1999) Vaccination with mage-3A1 peptide-
induces regression of some metastases in advanced stage IV melanoma. J. Exp. Med.
190(11), 16691678.
3. Hsu, F. J., Benike, C., Fagnoni, F., et al. (1996) Vaccination of patients with B-cell lym-
phoma using autologous antigen-pulsed dendritic cells. Nat. Med. 2(1), 5258.
4. Cochand, L., Isler, P., Songeon, F., and Nicod, L. P. (1999) Human lung dendritic cells
have an immature phenotype with efficient mannose receptors. Am. J. Respir. Cell Mol.
Biol. 21(5), 547554.
5. Vissers, J. L. M. De Vries, I. J. M., Oosterwijk, E., Figdor, C. G., and Adema, G. J. (2000)
Towards specific immunotherapy employing dendritic cells in melanoma and renal cell car-
cinoma, in Peptide-Based Cancer Vaccines. Eurekah.com, Austin, TX, pp. 200214.
122 de Vries et al.
6. McIlroy, D. and Gregoire, M. (2003) Optimizing dendritic cell-based anticancer immuno-

therapy: maturation state does have clinical impact. Cancer Immunol. Immunother. 52(10),
583591.
7. Steinman, R. M. (1991) The dendritic cell system and its role in immunogenicity. Annu.
8. Hart, D. N. (1997) Dendritic cells: unique leukocyte populations which control the pri-
mary immune response. Blood 90(9), 32453287.
9. Levin, D., Constant, S., Pasqualini, T., Flavell, R., and Bottomly, K. (1993) Role of den-
dritic cells in the priming of CD4+ T lymphocytes to peptide antigen in vivo. J. Immunol.
151(12), 67426750.
10. Banchereau, J., Schuler-Thurner, B., Palucka, A. K., and Schuler, G. (2001) Dendritic
cells as vectors for therapy. Cell 106(3), 271274.
11. Steinman, R. M. and Nussenzweig, M. C. (2002) Avoiding horror autotoxicus: the impor-
tance of dendritic cells in peripheral T cell tolerance. Proc. Natl. Acad. Sci. USA 99(1),
351358.
12. Mackensen, A., Drager, R., Schlesier, M., Mertelsmann, R., and Lindemann, A. (2000)
Presence of IgE antibodies to bovine serum albumin in a patient developing anaphylaxis
after vaccination with human peptide-pulsed dendritic cells. Cancer Immunol.
Immunother. 49(3), 152156.
13. Marten, A., Flieger, D., Renoth, S., et al. (2002) Therapeutic vaccination against meta-
static renal cell carcinoma by autologous dendritic cells: preclinical results and out-
come of a first clinical phase I/II trial. Cancer Immunol. Immunother. 51(1112),
637644.
14. Lin, C. L., Lo, W. F., Lee, T. H., et al. (2002) Immunization with Epstein-Barr Virus
(EBV) peptide-pulsed dendritic cells induces functional CD8+ T-cell immunity and may
lead to tumor regression in patients with EBV-positive nasopharyngeal carcinoma. Can-
cer Res. 62(23), 69526958.
15. Stift, A., Friedl, J., Dubsky, P., et al. (2003) Dendritic cell-based vaccination in solid
cancer. J. Clin. Oncol. 21(1), 135142.
16. Jonuleit, H., Kuhn, U., Muller, G., et al. (1997) Pro-inflammatory cytokines and prostag-
serum-free conditions. Eur. J. Immunol. 27(12), 31353142.
17. Schuler-Thurner, B., Schultz, E. S., Berger, T. G., et al. (2002) Rapid induction of tumor-
specific type 1 T helper cells in metastatic melanoma patients by vaccination with mature,
cryopreserved, peptide-loaded monocyte-derived dendritic cells. J. Exp. Med. 195(6),
12791288.
18. Holtl, L., Zelle-Rieser, C., Gander, H., et al. (2002) Immunotherapy of metastatic renal
cell carcinoma with tumor lysate-pulsed autologous dendritic cells. Clin. Cancer Res.
8(11), 33693376.
19. Kalinski, P., Hilkens, C. M., Wierenga, E. A., and Kapsenberg, M. L. (1999) T-cell prim-
ing by type-1 and type-2 polarized dendritic cells: the concept of a third signal. Immunol.
Today 20(12), 561567.
20. De Vries, I. J., Krooshoop, D. J., Scharenborg, N. M., et al. (2003) Effective migration of
antigen-pulsed dendritic cells to lymph nodes in melanoma patients is determined by their
maturation state. Cancer Res. 63(1), 1217.
21. Scandella, E., Men, Y., Gillessen, S., Forster, R., and Groettrup, M. (2002) Prostaglandin
dritic cells. Blood 100(4), 13541361.
22. Luft, T., Jefford, M., Luetjens, P., et al. (2002) Functionally distinct dendritic cell (DC)
populations induced by physiologic stimuli: prostaglandin E(2) regulates the migratory
capacity of specific DC subsets. Blood 100(4), 13621372.
23. Langenkamp, A., Messi, M., Lanzavecchia, A., and Sallusto, F. (2000) Kinetics of den-
dritic cell activation: impact on priming of TH1, TH2 and nonpolarized T cells. Nat.
Immunol. 1(4), 311316.
24. Vieira, P. L., Heystek, H. C., Wormmeester, J., Wierenga, E. A., and Kapsenberg, M. L.
(2003) Development of Th1-inducing capacity in myeloid dendritic cells requires envi-
ronmental instruction. J. Immunol. 164(9), 45074512.
25. Spisek, R., Bougras, G., Ebstein, F., et al. (2003) Transient exposure of dendritic cells to
maturation stimuli is sufficient to induce complete phenotypic maturation while preserv-
ing their capacity to respond to subsequent restimulation. Cancer Immunol. Immunother.
52(7), 445454.
26. Czerniecki, B. J., Carter, C., Rivoltini, L., et al. (1997) Calcium ionophore-treated periph-
eral blood monocytes and dendritic cells rapidly display characteristics of activated den-
dritic cells. J. Immunol. 159(8), 38233837.
27. Berger, T. G., Feuerstein, B., Strasser, E., et al. (2002) Large-scale generation of mature
monocyte-derived dendritic cells for clinical application in cell factories. J. Immunol.
Methods 268(2), 131140.
28.
human blood. An improved method with special regard to clinical applicability. J Immunol
Methods 196(2), 137151.
29.
Nature 392(6673), 245252.
30.
30 Caux, C., Dezutter-Dambuyant, C., Schmitt, D., and Banchereau, J. (1992) GM-CSF and
TNF-alpha cooperate in the generation of dendritic Langerhans cells. Nature 360(6401),
258261.
31.
31 Sallusto, F., Schaerli, P., Loetscher, P., et al. (1998) Rapid and coordinated switch in
chemokine receptor expression during dendritic cell maturation. Eur. J. Immunol. 28(9),
27602769.
32.
32 Zlotnik, A. and Yoshie, O. (2000) Chemokines: a new classification system and their
role in immunity. Immunity 12(2), 121127.
33.
33 Sallusto, F., Mackay, C. R., and Lanzavecchia, A. (2000) The role of chemokine recep-
tors in primary, effector, and memory immune responses. Annu. Rev. Immunol. 18,
593620.
34 Mellman, I. and Steinman, R. M. (2001) Dendritic cells: specialized and regulated anti-
34.
gen processing machines. Cell 106(3), 255258.
35.
35 Cumberbatch, M., Dearman, R. J., and Kimber, I. (1997) Langerhans cells require signals
from both tumour necrosis factor-alpha and interleukin-1 beta for migration. Immunology
92(3), 388395.
36.
36 Hirao, M., Onai, N., Hiroishi, K., et al. (2000) CC chemokine receptor-7 on dendritic cells
is induced after interaction with apoptotic tumor cells: critical role in migration from the
tumor site to draining lymph nodes. Cancer Res. 60(8), 22092217.
37.
37 Kukutsch, N. A., Rossner, S., Austyn, J. M., Schuler, G., and Lutz, M. B. (2000) Forma-
tion and kinetics of MHC class I-ovalbumin peptide complexes on immature and mature
murine dendritic cells. J. Invest. Dermatol. 115(3), 449453.
38.
38 Nestle, F. O., Banchereau, J., and Hart, D. (2001) Dendritic cells: On the move from bench
to bedside. Nat. Med. 7(7), 761765.
124 de Vries et al.
39.
39 Cox, A. L., Skipper, J., Chen, Y., et al. (1994) Identification of a peptide recognized by
five melanoma-specific human cytotoxic T cell lines. Science 264(5159), 716719.
40.
40 Dieu, M. C., Vanbervliet, B., Vicari, A., et al. (1998) Selective recruitment of immature
and mature dendritic cells by distinct chemokines expressed in different anatomic sites. J.
Exp. Med. 188(2), 373386.
41. DAmico, G., Bianchi, G., Bernasconi, S., et al. (1998) Adhesion, transendothelial migra-
tion, and reverse transmigration of in vitro cultured dendritic cells. Blood 92(1), 207214.
42.
42 Gunzer, M., Friedl, P., Niggemann, B., Brocker, E. B., Kampgen, E., and Zanker, K. S.
(2000) Migration of dendritic cells within 3-D collagen lattices is dependent on tissue
origin, state of maturation, and matrix structure and is maintained by proinflammatory
cytokines. J. Leukoc. Biol. 67(5), 622629.
43.
Res. 59(14), 33403345.
44.
44 Parlato, S., Santini, S. M., Lapenta, C., et al. (2001) Expression of CCR-7, MIP-3beta,
and Th-1 chemokines in type I IFN-induced monocyte-derived dendritic cells: impor-
tance for the rapid acquisition of potent migratory and functional activities. Blood
98(10), 30223029.
45.
45 Grabmaier, K., Vissers, J. L., De Weijert, M. C., et al. (2000) Molecular cloning and
immunogenicity of renal cell carcinoma-associated antigen G250. Int. J. Cancer 85(6),
865870.
46.
ture dendritic cells. J. Exp. Med. 193(2), 233238.
47.
47 Vissers, J. L., Hartgers, F. C., Lindhout, E., Teunissen, M. B., Figdor, C. G., and Adema,
G. J. (2001) Quantitative analysis of chemokine expression by dendritic cell subsets in
vitro and in vivo. J. Leukoc. Biol. 69(5), 785793.
48.
48 Lindhout, E., Vissers, J. L., Hartgers, F. C., et al. (2001) The dendritic cell-specific CC-
chemokine DC-CK1 is expressed by germinal center dendritic cells and attracts CD38-
negative mantle zone B lymphocytes. J. Immunol. 166(5), 32843289.
49.
49 Adema, G. J., Hartgers, F., Verstraten, R., et al. (1997) A dendritic-cell-derived C-C
chemokine that preferentially attracts naive T cells. Nature 387(6634), 713717.
50.
50 Sallusto, F., Lenig, D., Forster, R., Lipp, M., and Lanzavecchia, A. (1999) Two subsets of
memory T lymphocytes with distinct homing potentials and effector functions. Nature
401(6754), 708712.
51.
51 Godiska, R., Chantry, D., Raport, C. J., et al. (1997) Human macrophage-derived
chemokine (MDC), a novel chemoattractant for monocytes, monocyte-derived dendritic
cells, and natural killer cells. J. Exp. Med. 185(9), 15951604.
52.
52 Inngjerdingen, M., Damaj, B., and Maghazachi, A. A. (2000) Human NK cells express CC
chemokine receptors 4 and 8 and respond to thymus and activation-regulated chemokine,
macrophage-derived chemokine, and I-309. J. Immunol. 164(8), 40484054.
53.
53 Rea, D., van Kooten, C., van Meijgaarden, K. E., Ottenhoff, T. H., Melief, C. J., and
Offringa, R. (2000) Glucocorticoids transform CD40-triggering of dendritic cells into an
alternative activation pathway resulting in antigen-presenting cells that secrete IL-10.
Blood 95(10), 31623167.
54.
54 Adorini, L., Penna, G., Giarratana, N., and Uskokovic, M. (2003) Tolerogenic dendritic
cells induced by vitamin D receptor ligands enhance regulatory T cells inhibiting allograft
rejection and autoimmune diseases. J. Cell Biochem. 88(2), 227233.
55.
55 Martin, E., OSullivan, B., Low, P., and Thomas, R. (2003) Antigen-specific suppression
of a primed immune response by dendritic cells mediated by regulatory T cells secreting
interleukin-10. Immunity 18(1), 155167.
56.
56 De Vries, I. J., Lesterhuis, W. J., Scharenborg, N. M., et al. (2003) Maturation of dendritic
cells is a prerequisite for inducing immune responses in advanced melanoma patients.
Clin. Cancer Res. 9(14), 50915100.
57 Fong, L., Brockstedt, D., Benike, C., Wu, L., and Engleman, E. G. (2001) Dendritic
57.
cells injected via different routes induce immunity in cancer patients. J. Immunol.
166(6), 42544259.
58.
58 Barratt-Boyes, S. M., Zimmer, M. I., Harshyne, L. A., et al. (2000) Maturation and traf-
ficking of monocyte-derived dendritic cells in monkeys: implications for dendritic cell-
based vaccines. J. Immunol. 164(5), 24872495.
59.
Res. 59(14), 33403345.
60 Morse, M. A., Coleman, R. E., Akabani, G., Niehaus, N., Coleman, D., and Lyerly, H. K.
60.
(1999) Migration of human dendritic cells after injection in patients with metastatic ma-
lignancies. Cancer Res. 59(1), 5658.
61. Lindhout E., Vissers J. L. M. Figdor C. G. Adema G. J. (1999) Chemokines and lympho-
cyte migration. The Immunologist 7, 147152.
Clinical Trials for Multiple Myeloma 127
10
Dendritic Cells in Clinical Trials for Multiple Myeloma
Volker L. Reichardt and Peter Brossart
Summary
Due to the existence of the truly specific tumor antigen idiotype in multiple myeloma and
based on encouraging data from dendritic cell vaccinated B-cell non-Hodgkins lymphoma
(NHL) patients, dendritic cell-based vaccination was first initiated in myeloma patients in 1995.
This overview will summarize published and ongoing clinical trials in patients with multiple
myeloma who are treated with idiotype-based dendritic cell (Id/DC) vaccination. All groups of
investigators have found that Id/DC vaccination of multiple-myeloma patients is feasible and
that myeloma-specific immunity can be induced in heavily pretreated individuals. In future
trials, new dendritic cell-based immunization strategies will be investigated based on tech-
niques like RNA transfection of DC.
Key Words: Dendritic cells; immunotherapy; multiple myeloma; clinical trial; idiotype;
T-cells.
1. Introduction
Multiple myeloma (MM) is a malignant monoclonal B-cell disorder characterized
by plasma cell expansion in the bone marrow compartment and production of a unique
monoclonal immunoglobulin (Ig). The diagnosis is often delayed due to unspecific clini-
cal findings, resulting in more advanced disease in most of the patients at presentation.
Despite conventional chemotherapy and even high-dose chemotherapy protocols with
autologous stem-cell transplantation (1,2), which have markedly improved the response
rate, the time to progression, and the overall survival, the course of most multiple
myeloma patients remains ultimately fatal. New drugs targeting not only the malignant
B and plasma cells but also the tumor environmentsuch as thalidomide and its
derivatives, and very recently, the proteasome inhibitorshave gained much attention
and have yielded promising results in early clinical trials (3,4). A different approach of
targeted therapy is a therapeutic vaccination of MM patients with the goal of inducing
immunological control of residual disease. For such an approach, dendritic cell (DC)-
based vaccination strategies have been developed and will be discussed.
127
128 Reichardt and Brossart
2. Idiotype and Tumor-Associated Antigens in MM

The immunoglobulin (Ig) produced by each B-cell and its progeny is homogeneous
and unique. The differences responsible for the vast Ig molecule variety are based on
the variable regions of Ig heavy and light chains. Individual heavy and light chains are
composed of different antigenic determinants in the hyper-variable regions, which
form so-called idiotopes . The combination of all individual idiotopes of any given Ig
molecule is called the idiotype. Therefore, an idiotype (Id) is a specific protein prod-
uct and marker of an individual B-cell clone. The amount of secreted Id protein is
strongly correlated with the amount of myeloma plasma cells and therefore with the
tumor mass. The idiotype protein can be isolated from serum or plasma in the majority
of patients if a pretreatment serum sample with an appropriate amount of Id protein is
available. Idiotypes are clone specific and have successfully been used as targets of
immunotherapy in animal models of MM since the 1970s by Eisen et al. (5), and will
be discussed in depth in this chapter.
Tumor-associated antigens in MM comprise a large group of either over-expressed
proteins such as mucin 1 (MUC1); telomerase, with its catalytic subunit hTERT; the
gene product of PRAME or the sperm protein 17; so-called cancer testis antigens like
MAGE, BAGE, GAGE, and NY-ESO-1; or some mutated gene products, such as Ras,
as recently reviewed and discussed by Pellat-Deceunynck (6). After thorough investi-
gation in MM cell lines, several groups have also found relevant tumor-associated
antigen (TAA) expression in fresh MM cells, especially in increasing the stage of MM
in the case of cancer testis antigens (7). An important next step before the clinical
application of TAA vaccination in MM is that TAA-specific cytolytic T-cell responses
can be induced in vitro, as shown for MUC1 by our group (8) and for sperm protein 17
by Lim et al. (9). Many more MM-associated antigens are likely to be identified using
either RNA array technology, proteomics, or SEREX technology (10). To the best of
our knowledge, no clinical phase I studies have been reported on DC-based vaccina-
tion using TAA or TAA peptides in MM patients so far.
3. Id Vaccination in MM
For the unique expression of individual Id on MM B-cells and plasma cells, sev-
eral groups have focused on this truly tumor-specific antigen for vaccination pur-
poses. Based on encouraging immunological and clinical results of Id vaccination in
follicular NHL, as reported initially by Stanford et al. (11,12) and continued by
Bendandi et al. (13), the first Id vaccination in MM was reported by Mellstedt et al.
(14,15). Purified Id protein was either precipitated in aluminum phosphate (Id/alum)
and vaccinated subcutaneously (s.c.) or co-injected with granulocyte macrophage
colony-stimulating factor (GM-CSF). Id/alum and Id/GM-CSF vaccines induced
Id-specific T-cell responses as measured by interferon (IFN)- enzyme-linked
immunospot (ELISPOT) assay in early-stage MM patients (n = 5), proving the con-
cept that the MM-specific self-antigen is able to induce immune responses in the
tumor-bearing patient. Based on these promising data, Massaia et al. have performed
a phase I clinical trial (n = 12) of Id/keyhole limpet hemocyanin (Id/KLH) vaccina-
tion in MM patients in first remission after high-dose chemotherapy (16). The ratio-
nale for immunizing MM patients in a status of minimal residual diseasei.e.,

following high-dose chemotherapy and autologous stem-cell transplantationhas
been discussed before (17). All vaccines were given s.c. and co-injected with GM-
CSF or interleukin (IL)-2. Only transient local reactions at the injection sites were
reported, and cytokine-associated systemic toxicity was mild. The authors found
in all patients T-cell proliferative responses to the highly immunogenic xeno-anti-
gen KLH, but only in 2 out of 11 patients could Id-specific T-cell proliferation be
detected after in vitro re-stimulation. In 8 of 10 patients studied, an Id-specific
delayed-type hypersensitivity (DTH) skin reaction was documented after seven
injections of Id/KLH. Five of 11 patients who received all planned vaccines remained
in remission 930 mo following the first immunization; all other patients relapsed
within 936 mo.
4. DC in MM Patients
Based on strongly encouraging data from Id/DC vaccination in NHL (18,19),
which proved feasibility and immunological responses in B-cell NHL, several inves-
tigators studied the feasibility of generating DC vaccines from MM patients either
from peripheral blood leukapheresis products (2023), from peripheral blood mono-
cytes (21,24,25), or from CD34+ precursors (26). All authors found it feasible to
generate DC vaccines from heavily pretreated MM patients and demonstrated a com-
parable phenotypic profile in peripheral blood DC (PB-DC) as compared to normal
donors (21,22). While Raje et al. reported comparable in vitro T-cell stimulatory
capacity in MM patients DC, Brown and Ratta detected functional defects of PB-DC
in MM patients. Most authors agree upon the notion that monocyte-derived DC
(MoDC) of MM patients either generated by leukapheresis and plastic adherence or
by positive selection of CD14+ blood monocytes are phenotypically and functionally
normal (23,25,27). In vitro studies have found that MoDC from MM patients can
internalize and present idiotype proteins, and can be used to induce Id-specific T-cell
responses in vitro (25,28). Thorough in vitro investigations by Tarte et al. have been
reported on the generation of MoDC without the use of xenogeneic fetal calf serum to
avoid the sensitization against bovine immunogens with repetitive immunizations in
a clinical setting (29). We have extended these investigations and have developed an
optimized protocol for the generation of serum-free MoDC in MM patients (30). Raje
et al. reported on comparable yields, similar phenotype and function of CD34+-
derived DC vs MoDC (21), but Ratta et al. found lower yields of CD34+-derived DC
in MM patients (25).
5. Clinical Application of Id/DC Vaccination in MM

The Stanford group was the first to report on the initiation of DC-based Id vacci-
nation in MM patients who were responsive to high-dose therapy (HDT) followed
by autologous peripheral blood stem-cell transplantation (PBSCT) (31). This proto-
col followed very closely the very successful approach of using Id-pulsed PB-DC
for patients with low-grade B-cell NHL, as earlier published by Hsu et al. (18) and
updated recently by Timmerman (19).
The first scientific report on the clinical application of Id/DC vaccination in a

single MM patient used immature MoDC pulsed with autologous Id and KLH (24).
The Id-KLH/DC vaccines were comprised of at least 20% pure MoDC and were
injected three times intravenously (i.v.). The authors found an Id-specific T-cell pro-
liferative response in this refractory MM patient and were able to induce Id-specific
cytotoxic T-cells after seven rounds of in vitro re-stimulation. Despite large amounts
of circulating Id, the authors were able to detect Id-specific IgM antibody responses
after vaccination. No clinical response, as determined by significant changes of the
circulating Id, could be determined.
In another trial, 12 MM patients who were treated with HDT and PBSCT to induce
a status of minimal residual disease were enrolled in a clinical study of Id/DC vacci-
nation followed by Id/KLH booster immunizations (20). All Id/DC vaccines were
generated using PB-DC harvested by leukapheresis and given i.v. Id/KLH vaccines
were given s.c. and co-injected with the chemical adjuvant ISAF. The authors dem-
onstrated the feasibility of isolating PB-DC from MM patients even 37 mo after
HDT and PBSCT. The purity of the PB-DC vaccines varied between 3% and 63%,
which is a PB-DC enrichment of approx 100-fold as compared to peripheral blood.
Id-specific T-cell proliferative responses were detected repeatedly in 2 out of 12
patients. Eleven of 12 patients mounted strong KLH-specific T-cell responses, sug-
gesting immunocompetence to a neo-antigen at the time of vaccination. Cytolytic
Id-specific T-cell responses could be found in only one of three patients studied when
a single round of in vitro re-stimulation was performed. Id-specific antibody
responses were not analyzed because all but two patients had large amounts of circu-
lating Id at the time of vaccination. Interestingly, the two patients who mounted
Id-specific T-cell proliferative responses were in clinical complete remission (CR) at
the time of vaccination and remained in CR 30 and 17 mo after PBSCT. One patient
who stabilized her disease only after HDT and PBSCT and failed to mount a measur-
able Id-specific T-cell response had declining Id levels in the subsequent 2 yr without
any further cytoreductive treatment.
In the next cohort of MM patients treated with DC vaccines after HDT and
PBSCT, the Stanford group examined the effect of pulsing the PB-DC with a conju-
gate of Id and KLH, trying to increase the immunogenicity of the weak self-antigen
Id (32). The PB-DC were isolated using a new methodology based on sequential
density centri-fugation in a device developed and patented by Dendreon Corpora-
tion, Seattle, WA. While the authors achieved markedly higher cell numbers than in
leukapheresis products, the purity of infused PB-DC fell significantly. Importantly,
the rate of Id-specific immune responses remained comparable to the first cohort
(20), with 2 out of 14 patients developing T-cell proliferative responses. In line with
the previous observations, all DC infusions were well tolerated, regardless of the
much higher dose of cells infused. The authors found decreasing amounts of Id pro-
tein in the serum of several patients after Id vaccination, and described two cases of
conversion from partial remission (PR) to CR following Id/DC vaccination. Still,
the number of patients remains too small to establish a correlation between a modest
rate of Id-specific immune response and clinical course.
The same Dendreon device for the enrichment of PB-DC as mentioned above has
been used by researchers from Sacramento, CA (M. McKenzie et al.) and by a group at
the Mayo Clinic in Rochester (M. Lacy et al.), who have both reported on clinical
feasibility, immune responses, and early clinical results. As of February 2004, these
results have not been published.
A series of six MM patients vaccinated with Id-pulsed MoDC was reported by
Lim et al. (33), who chose MoDC generated after steady-state leukapheresis. All
MoDC were immature as judged by their FACS phenotype, were pulsed with Id Fab
fragments and KLH, and were given either freshly or thawed for vaccination 2 and 3.
Id protein-specific T-cell proliferative responses were detected in five of six patients
following Id/DC vaccination, and Id-specific cytotoxic T-cell precursor frequency
was increased in some patients. In line with their initial report (24), the authors also
detected a specific humoral response to Id despite circulating Id in the postvaccina-
tion serum. The authors did not evaluate humoral or cellular immune responses to
bovine proteins, which might have been induced due to the repetitive use of fetal calf
serum in generating the MoDC. All patients developed proliferative T-cell and anti-
body responses to KLH. One patient experienced a drop of his Id protein level by
25%; all other patients remained stable with a follow-up of 8 mo or progressed.
A report by Cull and collaborators summarizes the results of two MM patients who
received four Id Fab fragment and KLH pulsed immature MoDC vaccines, which were
given i.v. (34). Both patients, with advanced refractory disease, developed Id-specific
proliferative T-cell responses and KLH-specific T-cell responses. No Id-specific cyto-
toxicity could be measured in vitro, but in one patient an Id-specific antibody response
was noted. One patient progressed quickly after completion of the vaccine trial and the
other patient remained stable for some time following the vaccination.
The only report on Id/DC vaccination using DC generated from CD34+ precursors
is a study in relapsed and refractory MM patients (n = 11) who received a single s.c.
Id/DC injection followed by Id/GM-CSF s.c. booster vaccinations (n = 9) or Id/DC
vaccination again (n = 2) (26). All patients received immature DC that were loaded
with Id F(ab')2 fragments and proteinase K digested Id peptides. The same randomly
digested Id peptides were used for s.c. booster injections with GM-CSF as adjuvant.
Id-specific T-cell reactivity, as determined by IFN- ELISPOT, was detected in 4 out
of 10 patients studied. The authors also found increased anti-Id antibody titers follow-
ing vaccination in 3 of 10 patients investigated. No significant clinical responses were
seen following Id/DC vaccination in this cohort of refractory or relapsed MM patients.
In one small (n = 5) but very promising study of Id/DC vaccination of MM patients,
Yi et al. reported on Id-pulsed MoDC vaccination (35). All patients were in stable PR
4 to 33 mo following single or double HDT and PBSCT, and received three Id/DC
vaccines. To enhance the efficacy of vaccination, low-dose interleukin-2 was given
s.c. for 5 consecutive days following each s.c. DC vaccination. The phenotype of the
MoDC preparations was clearly that of mature DC. Id-specific T-cell responses were
evaluated with IFN- and IL-4 ELISPOT and T-cell proliferation assays. Two of
five patients developed Id-specific T-cell proliferative responses and four of five
patients had a positive IFN- immune response following Id/DC vaccination. Anti-Id
B-cell reactivity as determined by ELISPOT technology was found in all five patients.
One patient experienced a significant drop of his serum Id protein, possibly due to
vaccination or due to a delayed effect of HDT and PBSCT, which he received 4 mo
prior to vaccination.
We have developed a protocol to generate Mo-DC from MM patients under serum-
free tissue-culture conditions to avoid the immunization with bovine antigens by repeti-
tive vaccinations. Using this technology, a series of MM patients (n = 12) responding to
single or double HDT and autologous PBSCT have been vaccinated at our institution
with Id-pulsed MoDC followed by s.c. Id/KLH booster immunizations (30). Mature
Id-pulsed MoDC were given i.v. as in our previous protocol (20) and Id/KLH was given
with GM-CSF as adjuvant. Two of 12 patients mounted detectable Id-specific T-cell
proliferative responses, and one of these patients developed Id-specific cytotoxic T-cell
reactivity. This cytotoxic T-lymphocyte (CTL)-responsive patient experiences an ongo-
ing very good PR 40 mo following initiation of the Id/DC vaccination. All patients
developed cellular and delayed humoral immune responses to KLH.
6. Perspectives and Newer Protocols
A current clinical trial of DC vaccination in MM patients focuses on tumor lysate-
pulsed DC based on encouraging in vitro data as published by the Arkansas group
(36). This new approach holds the advantage of potentially targeting a host of indi-
vidual MM-specific targets and circumvents the necessity of isolation of Id protein. A
disadvantage is the need for a certain amount of fresh MM cells to generate individual
MM cell lysates.
We and others (R. Raymakers, Nijmengen, Netherlands) have begun to offer Id/DC
vaccination to MM patients after HDT and PBSCT who have achieved a clinical CR
but still have measurable disease on a molecular level. The goal of vaccinating this
group of patients is the induction of a molecular CR, which is most likely to be
achieved in a situation of minimal residual disease.
DC-based vaccination with idiotype peptide-loaded DC, as proposed by the studies
of Fagerberg and Hansson, is still not being clinically evaluated (37,38). Data gener-
ated by Trojan et al. provide evidence that Id peptides are rare in MM patients (39).
Very provocative data from the Boston group regarding vaccination with fusion cells
of autologous MM cells with DC will be clinically evaluated in the near future (40).
We have succeeded in the induction of MM cell line-specific cytotoxicity in vitro
using RNA transfection of DC (41). These techniques will soon be evaluated in a
phase I clinical trial in our institution.
References
1.
1 Attal, M., Harousseau, J. L., Stoppa, A. M., et al. (1996) A prospective, randomized trial
of autologous bone marrow transplantation and chemotherapy in multiple myeloma.
Intergroupe Francais du Myelome. N. Engl. J. Med. 335, 9197.
2.
2 Child, J. A., Morgan, G. J., Davies, F. E., et al. (2003) High-dose chemotherapy with
hematopoietic stem-cell rescue for multiple myeloma. N. Engl. J. Med. 348, 18751883.
3.
3 Singhal, S., Mehta, J., Desikan, R., et al. (1999) Antitumor activity of thalidomide in
refractory multiple myeloma. N. Engl. J. Med. 341, 15651571.
4.
4 Richardson, P. (2003) Clinical update: proteasome inhibitors in hematologic malignan-
cies. Cancer Treat. Rev. 29(Suppl. 1) 3339.
5.
5 Lynch, R. G., Graff, R. J., Sirisinha, S., Simms, E. S., and Eisen, H. N. (1972) Myeloma
proteins as tumor-specific transplantation antigens. Proc. Natl. Acad. Sci. USA 69,
15401544.
6 Pellat-Deceunynck, C. (2003) Tumour-associated antigens in multiple myeloma. Br. J.
6.
Haematol. 120, 39.
7.
7 van Baren, N., Brasseur, F., Godelaine, D., et al. (1999) Genes encoding tumor-specific
antigens are expressed in human myeloma cells. Blood 94, 11561164.
8.
8 Brossart, P., Schneider, A., Dill, P., et al. (2001) The epithelial tumor antigen MUC1 is
expressed in hematological malignancies and is recognized by MUC1-specific cytotoxic
T-lymphocytes. Cancer Res. 61, 68466850.
9.
9 Chiriva-Internati, M., Wang, Z., Salati, E., Bumm, K., Barlogie, B., and Lim, S. H. (2002)
Sperm protein 17 (Sp17) is a suitable target for immunotherapy of multiple myeloma.
Blood 100, 961965.
10 Tureci, O., Sahin, U., and Pfreundschuh, M. (1997) Serological analysis of human tumor
10.
antigens: molecular definition and implications. Mol. Med. Today 3, 342349.
11.
11 Kwak, L. W., Campbell, M. J., Czerwinski, D. K., Hart, S., Miller, R. A., and Levy, R.
(1992) Induction of immune responses in patients with B-cell lymphoma against the
surface-immunoglobulin idiotype expressed by their tumors. N. Engl. J. Med. 327, 1209
1215.
12.
12 Hsu, F. J., Caspar, C. B., Czerwinski, D., et al. (1997) Tumor-specific idiotype vaccines in
the treatment of patients with B-cell lymphomalong-term results of a clinical trial. Blood
89, 31293135.
13.
13 Bendandi, M., Gocke, C. D., Kobrin, C. B., et al. (1999) Complete molecular remissions
induced by patient-specific vaccination plus granulocyte-monocyte colony-stimulating
factor against lymphoma. Nat. Med. 5, 11711177.
14.
14 Bergenbrant, S., Yi, Q., Osterborg, A., et al. (1996) Modulation of anti-idiotypic immune
response by immunization with the autologous M-component protein in multiple myeloma
patients. Br. J. Haematol. 92, 840846.
15.
15 Osterborg, A., Yi, Q., Henriksson, L., et al. (1998) Idiotype immunization combined with
granulocyte-macrophage colony-stimulating factor in myeloma patients induced type I,
major histocompatibility complex-restricted, CD8- and CD4-specific T-cell responses.
Blood 91, 24592466.
16.
16 Massaia, M., Borrione, P., Battaglio, et al. (1999) Idiotype vaccination in human my-
eloma: generation of tumor-specific immune responses after high-dose chemotherapy.
Blood 94, 673683.
17.
17 Reichardt, V. L., Okada, C. Y., Stockerl-Goldstein, K. E., Bogen, B., and Levy, R. (1997)
Rationale for adjuvant idiotypic vaccination after high-dose therapy for multiple myeloma.
Biol. Blood Marrow Transplant. 3, 157163.
18.
18 Hsu, F. J., Benike, C., Fagnoni, F., et al. (1996) Vaccination of patients with B-cell lym-
phoma using autologous antigen-pulsed dendritic cells. Nat. Med. 2, 5258.
19 Timmerman, J. M., Czerwinski, D. K., Davis, T. A., et al. (2002) Idiotype-pulsed den-
19.
dritic cell vaccination for B-cell lymphoma: clinical and immune responses in 35 patients.
Blood 99, 15171526.
20 Reichardt, V. L., Okada, C. Y., Liso, A., et al. (1999) Idiotype vaccination using dendritic
20.
cells after autologous peripheral blood stem cell transplantation for multiple myelomaa
feasibility study. Blood 93, 24112419.
21.
21 Raje, N., Gong, J., Chauhan, D., et al. (1999) Bone marrow and peripheral blood dendritic
cells from patients with multiple myeloma are phenotypically and functionally normal
despite the detection of Kaposis sarcoma herpesvirus gene sequences. Blood 93, 1487
1495.
22.
22 Brown, R. D., Pope, B., Murray, A., et al. (2001) Dendritic cells from patients with
myeloma are numerically normal but functionally defective as they fail to up-regulate
CD80 (B7-1) expression after huCD40LT stimulation because of inhibition by trans-
forming growth factor-beta1 and interleukin-10. Blood 98, 29922998.
23.
23 Ratta, M., Fagnoni, F., Curti, A., et al. (2002) Dendritic cells are functionally defective in
multiple myeloma: the role of interleukin-6. Blood 100, 230237.
24.
24 Wen, Y. J., Ling, M., Bailey-Wood, R., and Lim, S. H. (1998) Idiotypic protein-pulsed
adherent peripheral blood mononuclear cell-derived dendritic cells prime immune system
in multiple myeloma. Clin. Cancer Res. 4, 957962.
25.
25 Ratta, M., Curti, A., Fogli, M., et al. (2000) Efficient presentation of tumor idiotype to
autologous T cells by CD83(+) dendritic cells derived from highly purified circulating
CD14(+) monocytes in multiple myeloma patients. Exp. Hematol. 28, 931940.
26.
26 Titzer, S., Christensen, O., Manzke, O., et al. (2000) Vaccination of multiple myeloma
patients with idiotype-pulsed dendritic cells: immunological and clinical aspects. Br. J.
Haematol. 108, 805816.
27.
27 Tarte, K., Lu, Z. Y., Fiol, G., Legouffe, E., Rossi, J. F., and Klein, B. (1997) Generation of
virtually pure and potentially proliferating dendritic cells from non-CD34 apheresis cells
from patients with multiple myeloma. Blood 90, 34823495.
28.
28 Butch, A. W., Kelly, K. A., and Munshi, N. C. (2001) Dendritic cells derived from mul-
tiple myeloma patients efficiently internalize different classes of myeloma protein. Exp.
Hematol. 29, 8592.
29.
29 Tarte, K., Fiol, G., Rossi, J. F., and Klein, B. (2000) Extensive characterization of den-
dritic cells generated in serum-free conditions: regulation of soluble antigen uptake, apop-
totic tumor cell phagocytosis, chemotaxis and T cell activation during maturation in vitro.
Leukemia 14, 21822192.
30.
30 Reichardt, V.L., Milazzo, C., Brugger, W., Einsele, H., Kanz, L, and Brossart, P. (2003)
Idiotype vaccination of multiple myeloma patients using monocyte-derived dendritic cells.
Haematologica 88,11391149.
31. Reichardt, V., Okada, C., Benike, C., et al. (1996) Idiotypic vaccination using dendritic
cells for multiple myeloma patients after autologous peripheral blood stem cell transplan-
tation. Blood 88(Suppl 1), 481a.
32.
32 Liso, A., Stockerl-Goldstein, K. E., Auffermann-Gretzinger, S., et al. (2000) . Idiotype
vaccination using dendritic cells after autologous peripheral blood progenitor cell trans-
plantation for multiple myeloma. Biol. Blood Marrow Transplant. 6, 621627.
33.
33 Lim, S. H. and Bailey-Wood, R. (1999) Idiotypic protein-pulsed dendritic cell vaccination
in multiple myeloma. Int. J. Cancer 83, 215222.
34.
34 Cull, G., Durrant, L., Stainer, C., Haynes, A., and Russell, N. (1999) Generation of anti-
idiotype immune responses following vaccination with idiotype-protein pulsed dendritic
cells in myeloma. Br. J. Haematol. 107, 648655.
35.
35 Yi, Q., Desikan, R., Barlogie, B., and Munshi, N. (2002) Optimizing dendritic cell-based
immunotherapy in multiple myeloma. Br. J. Haematol. 117, 297305.
36.
36 Wen, Y. J., Min, R., Tricot, G., Barlogie, B., and Yi, Q. (2002) Tumor lysate-specific
cytotoxic T lymphocytes in multiple myeloma: promising effector cells for immuno-
therapy. Blood 99, 32803285.
37.
37 Fagerberg, J., Yi, Q., Gigliotti, D., et al. (1999) T-cell-epitope mapping of the idiotypic
monoclonal IgG heavy and light chains in multiple myeloma. Int. J. Cancer 80, 671680.
38.
38 Hansson, L., Rabbani, H., Fagerberg, J., Osterborg, A., and Mellstedt, H. (2003) T-cell
epitopes within the complementarity-determining and framework regions of the tumor-
derived immunoglobulin heavy chain in multiple myeloma. Blood 101, 49304936.
39.
39 Trojan, A., Schultze, J. L., Witzens, M., et al. (2000) Immunoglobulin framework-derived
peptides function as cytotoxic T-cell epitopes commonly expressed in B-cell malignan-
cies. Nat. Med. 6, 667672.
40. Gong, J., Koido, S., Chen, D., et al. (2002) Immunization against murine multiple myeloma
with fusions of dendritic and plasmocytoma cells is potentiated by interleukin 12. Blood 99,
25122517.
41. Milazzo, C., Reichardt, V. L., Muller, M. R., Grunebach, F., and Brossart, P. (2003)
Induction of myeloma specific cytotoxic T cells using dendritic cells transfected with
tumor-derived RNA. Blood 101, 977982.
Identification of Autoantigens With SEREX 137
11
Identification of Tumor-Associated Autoantigens With SEREX
zlem Treci, Dirk Usener, Sandra Schneider, and Ugur Sahin
Summary
Serological analysis of tumor antigens by recombinant cDNA expression cloning (SEREX)
allows the systematic cloning of tumor antigens recognized by the spontaneous autoantibody
repertoire of cancer patients. For SEREX, cDNA expression libraries are constructed from
fresh tumor specimens, packaged into -phage vectors, and expressed recombinantly in Es-
cherichia coli. Recombinant proteins expressed during the lytic infection of bacteria are
transferred onto nitrocellulose membranes to be probed with diluted autologous patient serum
for identification of clones reactive with high-titered IgG antibodies. This chapter describes the
SEREX technology in detail.
Key Words: Tumor antigens; expression screening; -phage libraries; autoantibodies.
1. Introduction
One of the cardinal challenges in tumor immunology is the identification of anti-
genic structures in human tumors that are recognized by the hosts immune system.
By using tumor-specific cytotoxic T-lymphocyte (CTL) clones as probes for cDNA
expression cloning, a growing number of tumor antigens has been identified (1,2).
The necessity of established precharacterized CTL clones with tumor-cell restricted
reactivity is a critical obstacle of this methodology and was the main reason why the
majority of antigens defined hitherto were identified in malignant melanoma.
A large body of evidence points to a coordinated recruitment of CD4+, CD8+, and
B-cell responses to the same tumor antigen, and suggests that once immune recogni-
tion of a gene product is elicited, it is not restricted to merely one effector system.
Furthermore, it is frequently argued that the CTL repertoire of cancer patients is
deleted for many relevant CTL precursors. However, it is quite unlikely that a con-
comitant antibody response towards antigens (in particular intracellular ones) for
which respective CTLs have been be deleted, would also be erased (3). Thus, specific
antibodies may be the persisting hallmark of a substantial interaction between tumor
and immune system, and may help to trace back not only relevant antigens but also
deleted CTL specificities. Based on these rationales, we designed SEREX, a proce-
137
138 Treci et al.
dure allowing the systematic and unbiased cloning of tumor antigens recognized by
the IgG autoantibody repertoire of cancer patients (4).
To this aim, cDNA expression libraries are constructed from fresh tumor speci-
mens, expressed recombinantly and probed with autologous patient serum to identify
clones reactive with high-titered IgG-antibodies.
The SEREX approach is technically characterized by several features:
1. There is no need for established tumor cell lines and precharacterized CTL clones.
2. The use of fresh tumor specimens restricts the analysis to genes that are expressed by the
tumor cells in vivo and circumvents in vitro artifacts associated with short- and long-term
tumor cell culture
3. The use of the polyclonal (polyspecific) patients serum as a probe for immunoscreening
allows for the identification of multiple antigens with one screening course.
4. The screening is restricted to clones against which the patients immune system has raised
high-titered IgG responses indicating the presence of concomitant T-helper lymphocyte
responses in vivo.
5. As both the expressed antigenic protein and the coding cDNA are present in the same
plaque of the phage immunoscreening assay, identified antigens can be subjected imme-
diately to molecular characterization.
6. The release of periplasmatic proteins involved in protein folding during phage-induced
bacterial lysis allows at least partial folding of recombinant proteins and provides the
basis for the identification of linear as well as nonlinear epitopes. This has been confirmed
by showing that enzymes expressed in lytic plaques do exert their catalytic activity (our
unpublished results). In contrast, epitopes derived from eucaryotic posttranslational modi-
fication (e.g., glycosylation) are not detected by the page immunoscreening assay.
The proliferation of the technology to many other laboratories and the coordinated
analysis of different types of human cancers has resulted in a multitude of tumor-
associated antigens identified by the SEREX technique (57). Most of these antigens
are deposited in the SEREX Database, which harbors more than 2000 entries and pro-
vides a comprehensive molecular insight into what is called the cancer immunome
(see Websites: www2.licr.org/CancerImmunomeDB and www.eurice.de/EUCIP) (8).
The use of reverse T-cell immunology on SEREX-discovered antigens led to the
identification of epitopes recognized by CD8+ as well as CD4+ T-cells isolated from
cancer patients (9,10). This not only confirmed the original rationale but made these
antigens attractive candidates in experimental clinical trials for specific vaccination of
cancer patients (11). Moreover, SEREX has opened up new possibilities in the field of
tumor serology (12), where it has long been hoped to develop antibody-based screen-
ing tests for general use in the monitoring, diagnosis, and prognosis of cancer.
However, it soon became apparent that sera of cancer patients also contain IgG anti-
bodies that are unrelated to their disease and contribute whenever SEREX is applied in
a not yet quantified extent to the cloning of antigens. These are, for example, natural
antibodies as well as antibodies elicited by infections with pathogens. In addition, anti-
body entities may be elicited consecutive to necrosis or tissue destruction as accompa-
niments of tumor growth, with no direct relation at all to the molecular alterations
specific for the cancer cell.
Extracting the authentically cancer-related antigens has proven to be a major chal-
lenge.
This chapter not only provides a detailed protocol for the powerful SEREX tech-
nology. It also outlines how conceptually the validation process downstream of the
high-output discovery of antigens should be designed to allow for extraction of the
relevant ones.
2. Materials
2.1. Disposables
1. Disposable gloves.
2. Small and large Petri dishes (diameter 8 cm, 13.5 cm), (Greiner Bio-One, Kremsmnster,
Austria).
3. Fifty-mL Conical Falcon tube (BD Biosciences, Heidelberg, Germany).
4. Fourteen-mL Sterile Falcon round-bottom tube white caps (BD Biosciences).
5. Tubes:
a. Safe Lock 0.5 mL (Eppendorf, Hamburg, Germany).
b. Safe Lock 1.5 mL (Eppendorf).
6. Nitrocellulose Blotting Membranes: 45 m, (Sartorius AG, Gttingen, Germany).
7. 3MM blotting filters (Whatman, Maidstone, UK).
8. Square Petri dish 24.5 24.5 cm (VWR, Darmstadt, Germany) .
9. Self-made plastic spacers, size 23 2 0.1 cm to fit into a square Petri dish forming 24
sections of 23 1 cm .
10. Self-made incubator of 3-mm thick plexiglas to allow parallel processing of 24 nitrocel-
lulose membrane slips each in a different serum; floor space of 25 56 cm; separated in
2 24 cm incubation segments.
2.2. Chemicals and Reagents
1. IPTG (AppliChem, Darmstadt, Germany) in water, store 1 M stock at 20C .
2. Nitroblue-tetrazoliumchloride (NBT) (AppliChem), 75 mg/mL stock in 70% N-N dimeth-
ylformamide .
3. 5-Bromo-4-chloro-3-indol-phosphate-toluidine (BCIP) (AppliChem), 50 mg/mL in 100%
N-N dimethylformamide .
4. Goat anti-human IgG labeled with alkaline phosphatase (Dianova, Hilden Germany).
5. Affinity adsorbent 669580 (Roche Diagnostic, Mannheim, Germany).
6. Na-acid NaN3 (AppliChem): prepare a 10% (w/v) solution in water, store at 4C.
7. Thimerosal: prepare a 10% (w/v) solution in water, store at 4C.
8. MMLV reverse transcriptase (Stratagene, La Jolla, US).
9. MgSO4 (MgSO4 7H2O) (Roth, Karlsruhe, Germany): 1 M in water, sterile filtered.
10. MgCl2 (MgCl2 6H2O) (AppliChem): 1 M in water, sterile filtered.
11. Maltose (AppliChem): 10% (w/v) in water.
12. X-Gal (AppliChem): 250 mg/mL in DMF.
13. Ethanol, 96% p.a. (Roth).
14. Tetracycline (Sigma, Taufkirchen, Germany): 5 mg/mL stock in 96% ethanol, store at
20C.
15. Ampicillin (Sigma): 100 g/mL stock in water, store at 20C.
16. Kanamycin (Sigma): 50 /mL stock in water, store at 20C.
2.3. Kits
1. Messenger RNA Isolation Kit (Stratagene).
2. ZAP cDNA Synthesis Kit (Stratagene).
140 Treci et al.
3. Uni Zap XR Vector Kit (Stratagene).

4. Sizesep 400 Spun columns (Amersham-Bioscience, Uppsala, Sweden).
5. ZAP-cDNA Gigapack III Gold Cloning Kit (Stratagene).
6. Max-Plax packaging extracts (Epicentre, Madison, WI).
2.4. Buffers and Media

1. LB broth: 10 g NaCl, 10 g tryptone, 5 g yeast extract, add deionized H2O to a final vol-
ume of 1 L .
2. LB agar: 10 g NaCl, 10 g tryptone, 5 g yeast extract, 20 g agar, add deionized H2O to a
final volume of 1 L .
3. LB-tetracyline agar: 1 L autoclaved LB agar, cool to 55C and add tetracycline to a final
concentration of 12.5 g/mL .
4. LB-ampicillin agar: Prepare 1 L LB agar, autoclave, cool to 55C, and add ampicillin to a
final concentration of 150 g/mL.
5. LB-kanamycin agar: Prepare 1 L LB agar, autoclave, cool to 55C, and add kanamycin to
a final concentration of 50 g/mL.
6. LB top agar: Prepare 1 L LB broth, add 0.4% (w/v) agarose (Applichem) and 0.4% agar.
7. SM buffer: 5.8 g NaCl, 2.0 g MgSO4, 50 mL of 1 M Tris-HCl (pH 7.5), 5 mL of 2% (w/v)
gelatin, add deionized H2O to a final volume of 1 L .
8. 10X TBS: 0.5 M Tris base, 1.5 M NaCl, adjust pH 7.5 with concentrated hydrochloric
acid .
9. 10X TBST: add 0.5% Tween-20 to 10X TBS.
10. Blocking buffer: 1X TBST with 5% low-fat milk powder (Saliter, Obergnzburg, Ger-
many).
11. 20X PBS: 2.8 M NaCl, 54 mM KCl, 162 mM Na2HPO4, 30 mM KH2PO4, adjust to pH
7.47.6 with NaOH or HCl if necessary.
12. Tris-HCl, pH 3.0: 0.05 M Tris base, adjust to pH 3 with concentrated HCl.
13. 10X MOPS: 200 mM MOPS, 50 mM NaAc, 10 mM EDTA in 1 L, adjust to pH 7.0 with
NaOH.
14. 10X CDS (color development solution): 0.1 M Tris-HCl (pH 9.5), 5 mM MgCl2, 0.1 M
NaCl (pH 9.5).
3. Methods
Basic methods of molecular biology are described elsewhere (13).
3.1. Construction of a Phage Expression Library

For construction of a large, representative cDNA library, high quality mRNA as
starting material is of utmost importance. We propose the guanidium isothiocyanate-
phenol-chloroform extraction method (14), which produces large amounts of high-
quality, non-degradation total RNA. For subsequent purification of poly(A) RNA, a
number of comparable commercial kits are available. We generally use the messenger
RNA Isolation Kit (Stratagene). All precautions should be taken to avoid contamina-
tion with RNAses. Once high-quality mRNA is obtained, it is subjected to the proce-
dure described in Subheading 3.1.1. This protocol for construction of a phage library
follows basically the instructions of the manufacturer (Uni-ZAP XR cDNA library
synthesis kit, Stratagene) with minor modifications (see Note 1).
3.1.1. Synthesis of Doubled-Stranded cDNA for Directional Ligation

The cDNA-library-synthesis kit (e.g., Stratagene, cat. no. 200400) allows the gen-
eration of adaptor-ligated double-stranded cDNA for directional cloning. Fragments
suitable for directional cloning are generated using an oligo(dT) primer with an inter-
nal XhoI site. 5-methyl-dCTP is used instead of dCTP for first-strand synthesis. The
use of RNAse MMLV-reverse transcriptase (e.g., Superscript, Stratascript) improves
amount and quality of the first-strand cDNA. The second-strand synthesis is performed
with DNA polymerase I in combination with RNaseH, which generates internal prim-
ing sites. Unmodified dNTPs are used here, resulting in the generation of
hemimethylated double-stranded cDNA with blocked internal XhoI sites and a single
unmethylated XhoI site at the initial 3'-priming region. Pyrococcus furiosus poly-
merase lacking terminal nucleotide transferase activity fills up and blunts cDNA ter-
mini. A second restriction site is introduced into the cDNA fragments by ligating
adaptors with sticky unphosphorylated EcoR1 overhangs and phosphorylated blunted
ends. XhoI digestion of adaptor-ligated cDNA results in the generation of fragments
containing a 3' XhoI site (introduced during priming) and a 5' EcoR1 site (introduced
by adaptor ligation).
1. Preheat a 37C water bath. Thaw, vortex, and spin down the contents of the nonenzymatic
tubes of the Uni-ZAP XR cDNA library synthesis kit and place them on ice. MMLV-RT
is temperature sensitive and should be kept at 20C until needed.
2. The final volume of the first-strand synthesis reaction is 50 L. The volume of added
reagents and enzymes is 12.5 L, which allows the mRNA and DEPC-treated water to be
added in a volume up to 37.5 L. Add the following reagents in this order: 5 L of 10X
first-strand buffer, 3 L of first-strand methyl nucleotide mixture, 2 L of linker-primer
(1.4 g/L), X L of DEPC-treated water, 1 L of RNase Block Ribonuclease Inhibitor
(40 U/L). Mix the reaction and then add X L of poly (A) RNA (5 g). Mix gently.
3. After primer annealing for 10 min at room temperature, add 1.5 L of MMLV-RT (50
U/L) to the first-strand synthesis reaction. The final volume of the first-strand synthe-
sis reaction should now be 50 L. Mix the sample gently and spin down the contents in
a microcentrifuge.
4. Incubate the first-strand synthesis reaction at 37C for 1 h. To stop reaction, place tube
on ice.
5. For synthesis of second-strand cDNA, thaw, vortex, and spin all nonenzymatic second-
strand components and place them on ice. All reagents should be <16C when DNA poly-
merase I is added.
6. Add the following components on ice to the first-strand synthesis reaction in this order:
20 L of 10X second-strand buffer, 6 L of second-strand dNTP mixture, 114 L of sterile
distilled water, 2 L of RNase H (1.5 U/L), 11 L of DNA polymerase I (9.0 U/L).
7. After vortexing and spinning, incubate for 2.5 h at 16C. Temperatures above 16C can
cause the formation of hairpin structures and decrease efficiency of insertion of correctly
synthesized cDNA into the prepared vector.
8. For blunting the cDNA termini, place the reaction tube on ice and add the following to the
second-strand synthesis reaction: 23 L of blunting dNTP mix, 2 L of cloned Pfu DNA
polymerase (2.5 U/L).
9. After vortexing and spinning, incubate the reaction at 72C for exactly 30 min.
10. Prepare a 1:1 (v/v) emulsion of phenol (equilibrated to pH 7.08.0) and chloroform, and
add 200 L of it to the reaction. Vortex and spin the reaction briefly at maximum speed at
142 Treci et al.
room temperature and transfer the upper aqueous layer, containing the cDNA, to a new
tube. Do not remove any interface that may be present.
11. Add an equal volume of chloroform and vortex. Spin the reaction briefly at maximum
speed at room temperature and again transfer the upper layer.
12. Precipitate the cDNA by adding 20 L of 3 M sodium acetate and 400 L of 100% etha-
nol. Flick the bottom of the tube and precipitate overnight at 20C.
13. Spin in a microcentrifuge at maximum speed for 60 min at 4C. Remove and discard the
supernatant carefully without disturbing the pellet. Sometimes the pellet is not clearly
visible. Keep in mind that it may taper off along the side of the tube.
14. Wash the pellet by adding 500 L of 70% (v/v) ethanol to the side of the tube opposite
to the precipitate. Do not mix or vortex. Spin in a microcentrifuge at maximum speed
for 2 min at room temperature in the same orientation as before.
15. Aspirate the ethanol and dry the pellet.
16. Resuspend the pellet in 9 L of EcoR I adapters and incubate at 4C for at least 30 min.
Repeat mixing to ensure that the cDNA is completely in solution.
17. For adapter ligation, add the following components to the tube containing the blunted
cDNA and the EcoR I adapters: 1 L of 10X ligase buffer, 1 L of 10 mM rATP, 1 L of
T4 DNA ligase (4 U/L). Spin down and incubate overnight at 8C.
18. Heat-inactivate the ligase at 70C for 30 min. Thereafter, spin the reaction briefly and
cool it to room temperature for 5 min.
19. Phosphorylate the adapter ends by adding the following components: 1 L of 10X ligase
buffer, 2 L of 10 mM rATP, 6 L of sterile water, 1 L of T4 polynucleotide kinase
(10 U/L). Incubate the reaction for 30 min at 37C.
20. Heat-inactivate the kinase for 30 min at 70C. Spin down briefly and allow the reaction to
equilibrate to room temperature for 5 min.
21. For restriction enzyme digest with Xho I, add the following components to the reaction:
28 L of Xho I buffer supplement and 3 L of Xho I (40U/L). Incubate the reaction for
1.5 h at 37C.
22. Add 50 L of a 1:1 (v/v) emulsion of phenol (equilibrated to pH 7.08.0) and chloroform
to the reaction. Vortex and spin the reaction briefly at maximum speed at room tempera-
ture and transfer the upper aqueous layer, containing the cDNA to a new tube. Do not
remove any interface that may be present.
23. Add an equal volume of chloroform and vortex. Spin as above and again transfer the
upper layer.
3.1.2. cDNA Size Fractionation

Complete elimination of residual adaptors, which are present in great molar excess,
is crucial for efficient ligation of the cDNA fragments with the phage vector arms. We
recommend the use of Sizesep separation columns, which allow the removal of adap-
tors as well as size selection of long cDNA fragments.
1. Place the SizeSep 400 spin column in 14-mL Falcon tubes to collect the flow-through.
2. Overlay the column with 1X STE buffer.
3. Centrifuge (700g) for 12 min and discard the flow-through. Repeat this washing step.
4. To dry the column, centrifuge for 3 min at 700g.
5. Before inserting the spin column, place a 1.5-mL safe lock tube into the 14-mL Falcon
tube and load the cDNA reaction (Subheading 3.1.1., step 23) on the middle of the col-
umn without touching it.
6. Spin 3 min at 700g and collect the flow-through (1st fraction).
7. Pipet 50 L 1X STE buffer onto the column, spin (3min, 700g), and collect the flow-
through as 2nd fraction.
8. Repeat this twice to obtain the 3rd and 4th fractions.
9. Add 5 L 3 M NaAc and 55 L 96% ethanol to each fraction and precipitate the reaction
overnight at 20C.
10. After precipitation, spin the reaction at maximum speed for 60 min at 4C.
11. Wash the pellet by carefully adding 250 L of 70% (v/v) ethanol without disturbing the
pellet. Centrifuge at maximum speed for 5 min at room temperature in the same orienta-
tion as before.
12. Air-dry the pellets and resuspend them in 8 L of sterile H2O.
13. For quantification, take 1 L from each tube and add it to 4 L H2O (1:5 dilution). Trans-
fer 0.5 L of these into new tubes with 4.5 L H2O (1:50 dilution). Dot 2.5 L from both
dilutions of each fraction onto an agarose/ethidium-bromide gel. Dot in parallel increas-
ing dilutions of a DNA marker of known concentration.
By comparison on a UV transilluminator, the amount of cDNA in each reaction can be
estimated.
3.1.3. Ligation of cDNA into the Phage Vector

1. To prepare the sample ligation, add the following components in the described order:
X L resuspended cDNA (approx 100 ng), 0.5 L of 10X ligase buffer, 0.5 L of
10 mM rATP (pH 7.5), 1 L of the Uni-ZAP XR vector (1 g/L), 0.5 L of T4 DNA
ligase (4 U/L), and X L water for a final volume of 5 L.
2. Incubate the reaction tubes for 2 d at 4C.
3.1.4. Packaging Procedure

For packaging of methylated DNA, commercial kits are available from different
manufacturers (Gigapack Gold, Stratagene; MaxPlax, Invitrogen). These systems are
of comparable convenience and efficacy. We usually use Gigapack III Gold packag-
ing extracts. We suggest performing test packaging to determine the appropriate
amount of packaging extract.
1. For test packaging, which should be done after one day of ligation, remove one tube of
packaging extract from the 80C freezer and place it on dry ice.
2. Quickly thaw the packaging extract between your fingers, add 1 L of the ligated DNA to
the packaging extract immediately. Mix well by stirring with a pipet tip, avoiding air
bubbles.
3. Incubate the tube at room temperature for not longer than 2 h. Extended packaging times
may decrease efficacy considerably.
4. Add first 500 L of SM buffer followed by 20 L of chloroform, and mix the contents of
the tube gently.
5. Make serial dilutions of the packaged phage in SM buffer (102106) for plating and
titering of this test ligation, as described under Subheading 3.1.6.
6. Count the plaques and determine the titer (pfu/mL) and packaging efficiency according to
the following equation:
Number of plaques dilution factor total reaction volume
volume of dilution plated amount of DNA packaged (1)
7. The remaining ligation reaction has to be packaged at d 2 of ligation, repeating steps 16.
The number of packaging extracts needed has to be calculated according to the efficiency
determined by the test ligation described above in step 6.
144 Treci et al.
Spin the tubes briefly to sediment the debris. The supernatants containing the phage
are ready for titering and can be stored at 4C.
3.1.5. Preparing the Host Bacteria
1. Streak the host strain XLl-Blue MRF onto an LB agar plate containing the appropriate
antibiotic and incubate the plate overnight at 37C.
2. Inoculate LB broth supplemented with 10 mM MgSO4 and 0.2% (w/v) maltose, with a
single colony and grow at 37C, shaking for 46 h to an OD600 of 0.70.9.
3. Spin the cells at 500g for 10 min and discard the supernatant.
4. Gently resuspend the cells in sterile 10 mM MgSO4 to obtain an OD600 of 0.5.
3.1.6. Plating and Titering of the Library

The color assay allows determination of the percentage of recombinants within a
newly constructed library and is part of the quality control.
1. To plate the packaged ligation product, mix the following components: 1 L of the final
packaged reaction as well as 1 L of its 1:10 dilution each with 200 L of XL1-Blue
MRF cells (an OD600 of 0.5).
2. Allow the phage to attach to the cells by incubating them at 37C for 15 min. Add the
following components: 4 mL of LB top agar (melted and cooled to approx 47C), 15 L
of 0.5 M IPTG, 50 L of X-gal.
3. Plate immediately onto the LB-tetracycline agar plates in small Petri dishes (diameter 8
cm) and allow the plates to set for 10 min. Invert the plates and incubate at 37C over-
night.
4. Count the plaques. Calculate the efficiency using the following equation:
Number of plaques dilution factor total reaction volume
Total number of micrograms packaged number of microliters plated (2)
Background plaques are blue, while recombinant plaques will be white and should
be 50- to 100-fold above the background.
3.1.7. Amplification of the Library

The titer of a primary phage library may drop with time, resulting in the loss of
rare cDNA species. Therefore, one round of library amplification is suggested to
obtain a large and stable high-titer stock of the library (see Note 2). Make sure that as
long as the library has not gone through amplification, XL1-Blue MRF cells are
used as host strain (see Note 3).
1. Plate the entire library on large Petri dishes. Follow Subheading 3.1.6., steps 13. IPTG
is not required, since protein expression is not intended. Subconfluency of phage plaques
is ensured at 105 pfu per plate. Plates should not be lytic to ensure representation also of
slow-growing clones.
2. After overnight incubation, overlay the plates with 8 mL of SM buffer and incubate at
4C for 4 h with gentle rocking. Recover the bacteriophage suspension from each plate
and pool it into a sterile polypropylene container. Do not use glassware, to which the
phages would stick.
3. Rinse the plates thoroughly with additional 2 mL of SM buffer and add this to the polypro-
pylene container.
4. Add chloroform to 5% (v/v) final concentration. Mix well and incubate for 15 min at
room temperature.
5. Remove the cell debris by centrifugation for 10 min at 500g. Recover and transfer the
supernatant to a new polypropylene container. Add chloroform to a final concentration of
0.3% (v/v) and store at 4C. Titer before use.
6. The phage library could be stored for long periods at 80C with 7% DMSO.
Make sure that you assess and document all quality criteria of your library (see
Note 4).
3.2. Immunoscreening of Phage Expression Library
3.2.1. Preparation of Patient Serum
Polyclonal human sera contain antibodies that are reactive with E. coli and phage
proteins. These contaminating antibodies will increase background and affect sensi-
tivity and reliability of the assay. For elimination of such antibody specificities we
have developed an affinity purification protocol consisting of three consecutive steps
(see Note 5). For preabsorption we prefer a batch rather than a column procedure; the
batch procedure promotes not only efficient binding of serum to the affinity adsorbent
but is also faster and more convenient, in particular for processing of a larger number
of different sera.
3.2.1.1. PREABSORPTION AGAINST MECHANICALLY DISRUPTED BACTERIA

1. Inoculate LB broth supplemented with 10 mM MgSO4 and 0.2% (w/v) maltose, with a
single colony of XL1-Blue MRF and grow overnight, shaking at 37C.
2. Harvest the bacteria by centrifugation (3500g, 15 min) and resuspend pellet in 5 mL 1X
MOPS buffer.
3. Place cells on ice and disrupt by sonication five for 20 s (Branson Sonic Power Co. A
Smithkline).
4. For each individual serum fill a 50-mL Falcon tube with 2 g of affinity adsorbent, add
3 mL 1X PBS/0.01% NaN3, incubate for 10 min. Spin the tube and discard supernatant.
5. Add lysate obtained by mechanical disruption of bacteria to the affinity adsorbent and
rotate on an overhead shaker for 4 h at room temperature or overnight at 4C.
6. Spin down the matrix (100g, 1 min) and discard supernatant.
7. Wash column twice for 10 min with 30 mL 1X PBS/0.01% NaN3 on an overhead shaker.
Discard washing buffer.
8. Incubate affinity adsorbent for 2 h at 4C with 1 M glycin and wash subsequently twice
for 10 min with 1X PBS/0.01% NaN3. You have now obtained a resin saturated with
bacterial proteins.
9. Dilute human serum 1:10 in dilution buffer (0.01% NaN3, 0.01% thimerosal and 0.5%
low-fat milk powder in 1X TBS). Transfer the diluted serum (not more than 40 mL per
tube) to the 50-mL Falcon tubes with the prepared resin. Incubate for 4 h at room tem-
perature or overnight at 4C on an overhead shaker.
10. Centrifuge to spin down the resin (100g, 1 min). Harvest the preabsorbed sera, which are
stored at 4C until introduction into the second purification step as described under Sub-
heading 3.2.1.2.
11. This resin as well as the one generated under Subheading 3.2.1.2. can be regenerated and
re-used. For regeneration, wash the resin first three times for 10 min with 30 mL 50 mM
146 Treci et al.
Tris-HCl, pH 3. Subsequently, wash twice for 10 min with 30 mL TBS and 10 min with
30 mL TBS and 0.01% NaN3 and thimerosal. Store the affinity adsorbent in TBS and
0.01% NaN3 and thimerosal at 4C. Maximal shelf life of regenerated resin is 3 mo. The
binding capacity is sufficient for processing of 6 mL undiluted serum in total (equivalent
to 60 mL 1:10 prediluted serum).
3.2.1.2. PREABSORPTION AGAINST BACTERIOPHAGE

PROTEINS AND LYTICALLY INFECTED BACTERIA
1. Inoculate LB broth supplemented with 10 mM MgSO4 and 0.2% (w/v) maltose with a
single colony of XL1-Blue MRF, and grow overnight, shaking at 37C.
2. Harvest the bacteria by centrifugation (3500g, 15 min). Discard supernatant and resus-
pend the pellet in 2 mL 10 mM MgSO4. Transfer 200 L of it to 4 mL fresh LB broth
supplemented with 10 mM MgSO4 and 0.2% (w/v) maltose. The rest is kept at 4C until
use in step 4.
3. Infect freshly inoculated LB broth with 104 pfu of a wild-type ZAPII phage. For this
purpose, you can harvest blue phages obtained under Subheading 3.1.5. Shake at 37C
for 4 h.
4. Add remaining bacterial suspension from step 2 and shake 2 h at 37C. Under these
conditions you will obtain amplification of bacteriophages and lytic infection of bacteria
in culture.
5. Place cells on ice and disrupt by sonication five times for 20 s (Branson Sonic Power Co. A,
Smithkline).
6. Fill a 50-mL Falcon tube with 2 g of affinity adsorbent, add 3 mL 1X PBS/0.01% NaN3,
incubate for 10 min. Spin the tube and discard supernatant.
7. Add lysate obtained by phage lysis of bacteria to the affinity adsorbent and rotate on an
overhead shaker for 4 h at room temperature or overnight at 4C.
8. Spin down the matrix (100g, 1 min) and discard supernatant.
9. Wash column twice for 10 min with 30 mL 1X PBS/0.01% NaN3 on an overhead shaker.
Discard washing buffer.
10. Incubate affinity adsorbent for 2 h at 4C with 1 M glycin and wash subsequently twice
for 10 min with 1X PBS/0.01% NaN3. You have now obtained a resin saturated with
phage and bacterial proteins.
11. Transfer the sera processed through the first preabsorption step (Subheading 3.2.1.1.) to
the 50-mL Falcon tubes with the prepared resin. Incubate for 4 h at room temperature or
overnight at 4C on an overhead shaker.
12. Centrifuge to spin down the resin (100g, 1 min). Harvest the sera, which are stored at 4C
until introduction into the third purification step as described under Subheading 3.2.1.3.
3.2.1.3. PREABSORPTION AGAINST LYTIC FILTERS

1. Prepare E. coli XL1-Blue MRF as described under Subheading 3.1.5.
2. For each large Petri dish agar plate (diameter 13.5 cm) infect 600 L bacteria (OD600 =
0.5) with 105 pfu of wild-type ZAP II phage (blue phagesee Subheading 3.2.1.2.,
step 3). Incubate at 37C for 15 min. Add 40 L 1 M IPTG and 8 mL LB top agar (melted
and cooled to 42C).
3. Poor onto prewarmed LB-tetracycline agar plates, allow the top agar to solidify, and
incubate upside down overnight at 37C. Due to the high multiplicity of infection (MOI)
used, the plates will be lytic next morning without spared bacterial lawn.
4. Place nitrocellulose filters on top of the bacteria plaques, incubate 4 h at 37C.

5. Lift filters and replace each by a second one, repeat incubation for 6 h at 37C.
6. All filters are immersed in a large container with 1X TBST and incubated with gentle
agitation. Take care the top agar is removed entirely.
7. Wash three times for 10 min in 1X TBST.
8. Place each filter into an individual Petri dish. Incubate them in blocking buffer for 1 h at
room temperature with gentle agitation to block nonspecific protein-binding sites.
9. Wash filters three times for 10 min in a large container with 1X TBS. The filters can be
either air dried and stored at room temperature for up to 6 mo or used directly for the third
step of preabsorption.
10. Place each filter blotted side facing upward into a Petri dish. Transfer each individual
human serum that has been processed as described under Subheadings 3.2.1.1. and
3.2.1.2. to a filter and incubate for 4 h at room temperature on a horizontal shaker.
11. Remove the membrane, recover the serum completely, and incubate it with a fresh lytic
membrane.
12. Transfer serum into a 500-mL bottle or Falcon tubes. The serum is now ready to use for
immunoscreening.
3.2.2. Primary Immunoscreening of cDNA Expression Library

If the library has been generated from tissue contaminated with a considerable num-
ber of B-lymphocytes, the immunoscreening protocol should be modified either on the
level of primary or secondary immunoscreening (see Note 6).
1. Inoculate a single colony from the freshly plated XL1-Blue MRF cells in 20 mL LB
medium containing 0.2% maltose and 10 mM MgSO4 and grow overnight at 30C to an
OD600 of 0.70.9.
2. Spin the cells gently in Falcon tubes, remove the supernatant, and resuspend cells in 10 mM
MgSO4 to an OD600 of 0.5.
3. Add 5 104 pfu of the library to 0.6 mL cells. Mix well and incubate at 37C for 15 min
without shaking to allow phages to adhere to cells.
4. Add 15 L of 1 M IPTG and 910 mL of melted LB top agar, mix well, and pour
the mixture onto a prewarmed, dry LB-tetracycline agar plate (diameter 13.5 cm) (see
Note 7).
5. After the top agar is solidified, invert the plates and incubate at 37C overnight.
6. Number nitrocellulose filters with a waterproof pencil. Soak them in TBS until they are
completely wet. Discard nitrocellulose that does not wet properly. Place filters on blot-
ting paper to drain residual fluid. Take plates out of the incubator and place filters care-
fully onto the top agar, starting at an edge. Incubate plates at 37C overnight.
7. Chill plates at 4C for at least 1 h. Mark the orientation of the filters relative to the agar
plates with a needle by piercing in several places.
8. Remove filters carefully from plates, immerse them in a large container with 1X TBST
(see Note 8). Shake them vigorously on a platform shaker for 3060 min.
9. Remove any remaining top agar by lightly rubbing filters submerged in 1X TBST with
gloved fingers.
10. Transfer filters into a large container with fresh 1X TBST buffer and wash them for an
additional 1530 min under vigorous agitation.
11. Place each filter into an individual Petri dish. Immerse them in blocking buffer for 1 h at
room temperature with gentle agitation to block nonspecific protein-binding sites.
148 Treci et al.
12. Wash filters three times for 10 min in a large container with 1X TBS. Make sure that
precipitates of milk powder are rubbed off with gloved fingers.
13. Transfer each filter again into an individual Petri dish. Add preabsorbed serum prepared
as described under Subheading 3.2.1. Incubate on the shaker at room temperature over-
night.
14. Wash filters three times for 10 min in a large container with 1X TBS.
15. Transfer each filter again into an individual Petri dish. Add goat anti-human IgG labeled
with alkaline phosphatase (1:2500 dilution) in 1X TBS with 1% dry milk. Incubate for
1.5 h on the shaker at room temperature.
16. Wash filters three times for 10 min in a large container with 1X TBS.
17. Prepare staining solution by adding nitroblue tetrazolium chloride (NBT) to a final con-
centration of 0.3 mg/mL and 5-bromo-4-chloro-3-indol-phosphate-toluidine (BCIP) to a
final concentration of 0.15 mg/mL to color development solution under continuous stir-
ring. Filter this staining solution to avoid precipitates.
18. Transfer each filter into an individual Petri dish and immerse them in staining solution.
Allow reaction to proceed in the dark until positive reactions are clearly visible. Substrate
reaction occurs at the circumference of a phage plaque, so that positive clones appear as
darker halo.
19. Identify and mark positives on filters. Determine the respective plaque on the agar plate
using needle holes for orientation.
20. Pick positive plaques with two to three adjacent negative plaques using a pipet tip and
place each agar plug in a separate tube containing 1 mL of SM and 20 L of chloroform.
Vortex vigorously and incubate at 4C overnight.
3.2.3. Secondary Round of Immunoscreening
and Monoclonalization of Phages
This oligoclonal phage suspension is introduced to a secondary round of immuno-
screening. Negative plaques have been included deliberately as internal negative con-
trols for the secondary round of immunoscreening. The objective of the secondary
round is to confirm positivity and to enrich positive phages to monoclonality.
1. Divide the LB-tetracycline agar plate in a large Petri dish (diameter 13.5 cm) into four
quarters by carving out cross-shaped borders with a scalpel.
2. Dilute the primary stock of oligoclonal phages at 1:500 and 1:100 in SM buffer, add 2 L
from these dilutions and IPTG to 23 mL top agar. Pour each of these dilutions on one
quarter. Thus two different titers of two different oligoclonal phage stocks can be assayed
on one 13.5-cm Petri plate.
3. Repeat Subheading 3.2.2., steps 519 for primary immunoscreening.
4. Confirmation of positivity should be done only for clones resulting in subconfluent phage
plaques. If the entire lawn is lytic with both dilutions, repeat the test with a higher dilu-
tion.
5. Pick two monoclonal phages for each clone confirmed as positive, as described under
Subheading 3.2.2., step 20. Transfer them into 0.5 mL SM with 20 L chloroform for in
vivo excision.
3.3. Serological Characterization of Clones
Differential serology, which means testing of a clone with sera from tumor patients
as well as appropriate control groups, is an important tool to determine the antigens
with tumor-associated sero-responses (see Note 9).
3.3.1. Conventional Petite Serology

1. Divide agar-tetracycline plates in 15-cm Petri dishes into four quarters by carving out
cross-shaped borders with a scalpel.
2. Mix monoclonal phages representing individual antigens of interest 5:1 with wild-type
ZAP phage as a control. Infect 0.6 mL E. coli XL1-Blue MRF host strain (OD600 = 0.5)
in the presence of 0.8 mM IPTG with a final pfu of 2 103.
3. Add 23 mL LB top agar and pour onto one quarter of the prewarmed LB agar plates for
overnight incubation. Thus four different clones can be processed in parallel on one plate.
4. After plaque formation, recombinant proteins are blotted onto nitrocellulose filters and
processed as described under Subheading 3.2.2., steps 720. Each filter is incubated
with serum from a different donor preabsorbed as described under Subheading 3.2.1. In
total, 1520 sera from cancer patients suffering from the tumor under investigation and
1520 sera of healthy controls are used on the respective set of four clones under investi-
gation.
3.3.2. SeroGRID for Petite Serology

Compared to conventional petite serology, SeroGRID is a rapid, high-throughput
procedure that allows multiplex analysis of an entire set of antigens with a selection of
different sera (15).
1. Instead of round large Petri dishes (diameter 13.5 cm), use 24.5 24.5 2.5 cm square
Petri dishes to prepare an LB-tetracycline plate.
2. Before pouring phage-infected bacteria in the top agar layer onto the plate, insert 22 plas-
tic spacers into the agar, producing slots 1 cm 24 cm in size to obtain a separate slot for
each individual phage clone.
3. Titer each phage clone, mix approximately in a ratio of 5:1 with wild-type phage as inter-
nal negative control, adjust to a final pfu of 6 102 and infect E. coli XL1-Blue MRF.
Add 3 mL of top agar in the presence of IPTG and pour into the prepared slots.
4. One slot is reserved for wild-type phage as a negative control. As a positive control, a
phage containing PINCH (GenBank, accession no. U09284), which reacts with most
human sera, is used.
5. After overnight incubation, spacers are removed. For plaque lift of recombinantly expressed
proteins, a 24 23 cm nitrocellulose membrane is prepared and placed on the plate.
6. Wash and block as described in Subheading 3.2.2., steps 712
7. After washing and blocking, cut the membranes in 1 23 cm multiantigen strips with
the cutting edge at right angles to the direction of individual phage slots. Each strip is
placed in an individual segment of the self-made incubation chamber and incubated with
a different serum overnight.
8. Proceed as described in Subheading 3.2.2., steps 920 to obtain immunostained filters.
3.4. Molecular Characterization of Clones

3.4.1. In Vivo Excision of the Phagemid From the Phage Vector
The Uni-ZAP XR vector is designed to allow simple, efficient in vivo excision
from the vector, and recircularization to form an ampicillin-resistant pBluescript
phagemid containing the cloned insert. The cloning site of the vector is flanked by
fl bacteriophage origins of replication. Simultaneous infection of E. coli with both the
vector and the fl bacteriophage leads to nicking at these sites and synthesis of a
150 Treci et al.
single-stranded DNA molecule. This molecule includes all sequences of the pBlue-
script SK() phagemid and the insert, if one is present and is circularized by specific
proteins from the fl phage. In addition, a functional fl origin is generated as found in
fl bacteriophage or phagemids. The ExAssist helper phage used for this procedure
contains an amber mutation that prevents replication of the phage genome in a
nonsuppressing E. coli strain such as SOLR cells. This allows only the excised
phagemid to replicate in the host, avoiding co-infection by ExAssist helper phage.
1. Streak the host strain E. coli XL1-Blue MRF and E. coli XL1-Blue SOLR onto an LB
agar plate containing tetracycline/kanamycin, and incubate the plate overnight at 37C.
2. Grow separate overnight cultures of XL1-Blue MRF and SOLR cells in LB broth, supple-
mented with 0.2% (w/v) maltose and 10 mM MgSO4, at 30C to an OD600 of 0.70.9.
3. Spin the cells down and resuspend them to obtain an OD600 of 1.0 in 10 mM MgSO4.
4. Combine the following components in a Falcon 2059 polypropylene tube: 200 L of
XL1-Blue MRF cells at an OD600 of 1.0, 250 L of monoclonal secondary positive
phage stock (containing >1 105 phage particles), 1 L of the ExAssist helper phage (>1
106 pfu/L).
5. Incubate the Falcon 2059 polypropylene tube at 37C for 15 min.
6. Add 3 mL of LB broth and incubate the Falcon 2059 polypropylene tube for 2.53 h at
37C.
7. Heat the Falcon 2059 polypropylene tube at 6570C for 20 min to inactivate the ExAssist
helper phage and then spin the tube at 1000g for 15 min.
8. Decant the supernatant into a sterile Falcon 2059 polypropylene tube. This stock contains
the excised pBluescript phagemid packaged as filamentous phage particles, and may be
stored at 4C for 12 mo.
9. To plate the excised phagemids, add 10 L and 100 L of the phage supernatant from step
7 each to 200 L of freshly grown SOLR cells from step 2 (OD600 = 1.0) in 1.5-mL micro-
centrifuge tubes.
10. Incubate the tubes at 37C for 15 min.
11. Streak 1020 L of the cell mixture from microcentrifuge tubes on LB-ampicillin agar
plates (50 g/mL) and incubate the plates overnight at 37C. Colonies appearing on the
plates contain the pBluescript double-stranded phagemid with the cloned DNA insert.
Colonies can now be picked, grown, and subjected to plasmid purification. For long-term
storage, prepare a bacterial glycerol stock and store at 80C.
3.4.2. Further Molecular Characterization of In Vivo Excised Clones
In vivo excised phagemids can be conveniently subjected to state-of-the-art tech-
nologies for molecular, structural, and functional analysis (see Note 10).
4. Notes
1. Three different phage expression library systems have been reported for SEREX
screening: (1) Uni ZAP vector, which allows in vivo excision of pBluescript phagemid
using the SOLR strain (Stratagene); (2) ZAP Express vector, which allows in vivo
excision of pBK-CMV phagemid using the XLOLR strain (Stratagene); (3) TRIPLX
vector (Clontech Laboratories Inc., Palo Alto, CA).
Protocols described here refer to the ZAP II-based system, which we prefer since
it results in large and clearly visible lytic plaques with high amounts of recombinant
protein.
2. If primary phage count is >4 106, primary immunoscreening can be conducted without
library amplification. Even so, we recommend one round of amplification of a fraction of
the library containing at least 2 106 phages. We prefer amplification on agar plates
rather than amplification in soluble medium, since this procedure reduces the risk of dra-
matic changes in the representation of individual clones. Amplification of the library
allows generation of large amounts of phages for multiple screening rounds as well as for
storing. It may be advisable to conduct an initial screening round directly after the library
has been packaged but before amplification, which will allow a more representative cov-
erage of the diversity of transcript species.
3. Messenger RNA is primed in the first-strand synthesis with a Xho I site containing linker
primer, and is reverse transcribed using RNase H negative MMLV-RT and 5-methyl
dCTP. The hemimethylated cDNA is protected from digestion of the cDNA with Xho I,
used later to cleave the linker primer.
Hemimethylated DNA introduced in a bacterial strain would be efficiently digested
by the mcrA and mcrB restriction systems. It is necessary to pass the library through a
McrA McrB host (e.g., XL1-Blue MRF). After that, the library is no longer hemimeth-
ylated and can be grown on McrA+ McrB+ strains (e.g., XL1-Blue).
4. Three surrogate parameters(1) number of clones; (2) rate of recombinants; and (3)
median and maximal insert size of insertsare used to determine the quality of an expres-
sion library. The number of primary plaques is a good indicator for the representation of
individual transcripts in the library. cDNA clones in -ZAP libraries are expressed as
fusion proteins containing N-terminal -galactosidase fragments. Having in mind that
only one third of directionally cloned cDNA are in the appropriate frame and some cDNA
fragments are cloned with a leading 5'-UTR possibly introducing premature stop codons,
we recommend that a good library should contain >2 106 primary plaques. To esti-
mate the average and maximum fragment size, we perform agarose gel analysis of inserts,
either amplified by PCR using oligonucleotides binding to insert flanking regions of the
vector backbone and >105 phages as template, or by analyzing 20 randomly picked, in
vivo excised restriction enzyme-digested clones. The median size of human mRNA-
derived inserts should be >1 kb, and the presence of inserts with size >45 kb indicates
sufficient representation of large cDNAs. The rate of recombinants (white plaque num-
ber/total plaque number) is generally clearly above 95%. Lower rates point to inefficient
ligation of the cDNA fragments (e.g., mismatched amounts of vector and insert, defective
cDNA-fragment or vector ends, contamination with adaptors). Even though rates around
85% are tolerable, we recommend performing sequencing of randomly picked clones to
ensure the correctness of inserts (correct direction, presence of adaptor sequences at 5'
end, presence of poly(A) tail at the 3' end), if rates of recombinants are below 95%.
5. The quality of the preabsorbed serum is critical. Usually the serum can be used several
times (515 rounds of screening). It should be stored at 4C and thimerosal or NaN3
should be added each week to prevent bacterial contamination. Reuse often results in
reduced background due to absorption of nonspecific components rather than in dimin-
ished activity. This increases sensitivity and specificity. As a consequence, weak positive
reactivities are detected in later rounds of the immunoscreening.
Our protocol of preabsorption uses not only mechanically disrupted bacteria, which
deplete antibodies against intracellular components and cell-wall elements of bacteria.
We also preabsorb against bacteria lysed by phages. The latter reagent contains addition-
ally phage proteins as well as bacterial proteins induced secondary to infection (e.g.,
shock proteins).
152 Treci et al.
Many human sera can be used after having been subjected to one round of the
described tripartite preabsorption process (Subheadings 3.2.1.1.3.2.1.3.). However, if
this does not reduce unspecific cross-reactivity sufficiently, parts of the process can be
repeated. If mainly the bacterial lawn in between plaques is contributing to the back-
ground, Subheading 3.2.1.1. should be repeated. If the plaque circumferences are stained
unspecifically, repeat Subheading 3.2.1.2. or 3.2.1.3.
6. Libraries derived from tissue specimens contaminated with B-lymphocytes will contain
inserts expressing segments of genes encoding human immunoglobulins. Immunoscreen-
ing of such libraries derived from tissue will result in the identification of false positives
by binding of the second antibody to recombinantly expressed human IgG derived from
the B-cell transcriptome. Such inserts may represent up to 90% of all positive clones.
These clones should be eliminated early from further characterization. One option is to
use a modified primary screening as described previously (16). To this aim, the filters are
handled according to the protocol including the blocking step (Subheading 3.2.2., steps
112), but then preincubated first with enzyme-conjugated anti-human IgG and stained to
detect the immunoglobulin inserts, which are marked with a pencil. Subsequently, these
membranes go through the standard process starting with incubation with the autologous
serum, second antibody, and thereafter staining (Subheading 3.2.2., steps 1220). Only
clones appearing now de novo and not marked in context of the prior staining step are
recovered. If contamination is modest, this pre-testing by direct incubation with the anti-
IgG antibody may be done on the level of secondary immunoscreening after individual
phages have been picked and oligo-clonalized.
A substantial a priori reduction of such false positives can be achieved by depletion of
immunoglobulin sequences during library construction using subtractive cDNA library
approaches (17).
7. A reduced moisture content of plates results in higher quality plaque lifts. Agar plates
should be prepared at least 24 h prior to use. Preincubate plates at 37C 1 h before use. If
necessary, wipe dry with a paper tissue before pouring the top agar. If despite these
endeavors the top agar and plaques stick to the filter and are removed together with it, the
phage traces pertinent on the agar plate under the top agar layer, even though not visible,
are sufficient to recover phage clones of interest.
8. Nitrocellulose filters should be handled with care. Never touch with bare hands, always
use gloves. Do not allow filters to dry after antigen absorption.
Because inter-lot differences of nitrocellulose may affect the quality of the assay, we
prefer to test batches prior to use. Washes have to be thorough and generous. Top agar has
to be removed without scratching but diligently. Parts of the filter covered by gelatinized
contents of lytic plaques or sticking traces of top agar are not eligible for immunodetection
by antibodies. As described, we prefer to separate filters and process them blotted side
facing upward in individual Petri dishes for the critical steps of blocking, antibody incu-
bation, and staining, to ensure adequate exposure. These Petri dishes do not need to be
sterile and can be re-used.
9. SEREX is a high-output technology. Comprehensive mapping of even a single patient
specimen results in cloning of multiple different antigens. However, as pointed out in
the introduction, a considerable proportion of clones detected by SEREX are not prima-
rily cancer related. Evaluation of multiple SEREX-defined clones in serological assays
using panels of allogeneic sera from cancer patients as well as appropriate control groups
is an important step towards focusing on the relevant antigens worth further extended
analysis (18).
Usually, a two-step strategy is used. First, a small-scale serological study based on

the plaque lift assay is conducted to define those antigens without reactivity with con-
trol sera of healthy individuals but exhibiting reactivity with tumor sera (petite serol-
ogy). Subsequently, those antigens that appear to be cancer related are subjected to
ELISA based on purified recombinant protein for further confirmation with a large
panel of sera (grande serology) (19). Since this process is labor intensive, several
modified protocols have been established (2022), one of which is revisited in Sub-
heading 3.3.2. (15).
10. SEREX aims at the identification of tumor-related antigens of potential diagnostic or
therapeutic use. Conceptually, besides state-of-the-art procedures for characterization of
new genes, the molecular characterization should tackle in particular which of the gene
products are (1) selectively expressed and/or (2) of tumor biological relevance (18,23).
A first step is sequencing of the antigen encoding cDNAs and subsequent BLAST
search to determine clones identical to known autoantigens or to proteins with known
tumor biological relevance. Expression profiling by quantitative RT-PCR in a panel of
normal and tumor tissues determines genes with tumor-associated tissue distribution.
References
1.
1 Boon, T., Cerottini, J. C., van den Eynde, B., van der Bruggen, P., and van Pel, A. (1994)
Tumor antigens recognized by T-lymphocytes. Annu. Rev. Immunol. 12, 337366.
2.
2 van der Bruggen, P., Traversari, C., Chomez, P., et al. (1991) A gene encoding an antigen
recognized by cytolytic T lymphocytes on a human melanoma. Science 254, 16431647.
3.
3 Hartley, S. B., Booke, M. P., Fulcher, D. A., et al. (1993) Elimination of self reactive B
lymphocytes proceeds in two stages: arrested development and cell death. Cell 72, 325335.
4.
4 Sahin, U., Treci, O., Schmitt, H., et al. (1995) Human neoplasms elicit multiple specific
immune responses in the autologous host. Proc. Natl. Acad. Sci. USA 92, 11,81011,813.
5.
5 Sahin, U., Treci, O., and Pfreundschuh, M. (1997) Serological identification of human
tumor antigens. Curr. Opin. Immunol. 9, 709716.
6. Treci, O., Sahin, U., and Pfreundschuh, M. (1997) Serological analysis of human tumor
antigens: Molecular definition and implications. Mol. Med. Today 3, 342349.
7.
7 Boon, T. and Old, L. J. (1997) Cancer tumor antigens. Curr. Opin. Immunol. 9, 681683.
8.
8 Jongeneel, V. (2001) Towards a cancer immunome database. Cancer Immun. 1, 3.
9.
9 Zarour, H. M., Maillere, B., Brusic, V., et al. (2002) NY-ESO-1 119-143 is a promiscuous
major histocompatibility complex class II T-helper epitope recognized by Th1- and Th2-
type tumor-reactive CD4+ T cells. Cancer Res. 62, 213218.
10 Zeng, G., Li, Y., El-Gamil, M., et al. (2002) Generation of NY-ESO-1-specific CD4+ and
10.
CD8+ T cells by a single peptide with dual MHC class I and class II specificities: a new
strategy for vaccine design. Cancer Res. 62, 36303635.
11.
11 Jager, E., Gnjatic, S., Nagata, Y., et al. (2000) Induction of primary NY-ESO-1 immunity:
CD8+ T lymphocyte and antibody responses in peptide-vaccinated patients with NY-ESO-
1+ cancers. Proc. Natl. Acad. Sci. USA 97, 12,19812,203.
12.
12 Old, L. J. and Chen, Y. T. (1998) New paths in human cancer serology. J. Exp. Med. 187,
11631167.
13. Maniatis, T., Sambrook, J., and Fritsch, E. F., eds. (1989) Molecular Cloning: A Labo-
ratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY,
pp.7.117.12.
14.
14 Chomczynski, P. and Sacchi, N. (1987) Single-step method of RNA isolation by acid
guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156159.
154 Treci et al.
15.
15 Krause, P., Treci, ., Micke, P., Buhl, R., Huber, C., and Sahin, U. (2003) SeroGRID:
An improved method for the rapid selection of antigens with disease related immunoge-
nicity. J. Immunol. Methods 283, 261267.
16. Treci, ., Schmitt, H., Fadle, N., Pfreundschuh, M., and Sahin, U. (1997) Molecular
definition of a novel human galectin which is immunogenic in patients with Hodgkins
disease. J. Biol. Chem. 272, 64166422.
17.
17 Diatchenko, L., Lukyanov, S., Lau, Y. F., and Siebert, P. D. (1999) Suppression subtrac-
tive hybridisation: A versatile method for identifying differentially expressed genes. Meth-
ods Enzymol. 303, 349380.
18. Treci,O., Sahin, U., Zwick, C., Neumann, F., and Pfreundschuh, M. (1999) Exploitation
of the antibody repertoire of cancer patients for the identification of human tumor anti-
gens. Hybridoma 18, 2328.
19.
19 Stockert, E., Jger, E., Chen, Y. T., et al. (1998) A survey of the humoral immune response
of cancer patients to a panel of human tumor antigens. J. Exp. Med. 187, 13491354.
20.
20 Lagarkova, M. A., Koroleva, E. P., Kuprash, D. V., et al. (2003) Evaluation of humoral
response to tumor antigens using recombinant expression-based serological mini-arrays
(SMARTA). Immunol. Lett. 85, 7174.
21.
21 Scanlan, M. J., Welt, S., Gordon, C. M., et al. (2002) Cancer-related serological recogni-
tion of human colon cancer: identification of potential diagnostic and immunotherapeutic
targets. Cancer Res. 62, 40414047.
22. Treci, ., Luxemburger, U., Heinen, H., et al. (2004) CrELISA: A fast and robust enzyme
linked immuno sorbent assay bypassing the need for purification of recombinant protein. J.
Immunol. Methods, in press.
23. Treci, ., Sahin, U., Zwick, N., Neumann, F., and Pfreundschuh, M., (1999) Identifica-
tion of human tumor antigens using the antibody repertoire of cancer patients. Gann Mono-
graph Cancer Res. 48, 93101.
Proteomics for Biomarker and Target Finding 155
12
Application of Proteomics and Protein Analysis
for Biomarker and Target Finding for Immunotherapy
Petra Weingarten, Petra Lutter, Andreas Wattenberg, Martin Blueggel,

Sonja Bailey, Joachim Klose, Helmut E. Meyer, and Christoph Huels
Summary
Regulatory T-cells play a central role in the maintenance of the immunological balance and
are powerful inhibitors of T-cell activation both in vivo and in vitro. The enhancement of sup-
pressor-cell function might be a target for immunotherapeutic approaches for the treatment of
immune-mediated diseases like multiple sclerosis and Crohns disease.
The method of choice to elucidate the still unclear effector functions of regulatory T-cells is
the differential proteome analyses performed with human and murine T-cell populations. To
this end, whole-protein extracts of conventional and regulatory T-cells are separated by high-
resolution two-dimensional gel electrophoresis according to Klose (1). The proteomes are ana-
lyzed by a 2DE gel image analysis software, ProteomWeaver. The protein spots that are
found differentially expressed are picked from the gels and prepared for matrix-assisted laser
desorption/ionization (MALDI) mass spectrometrical analysis automatically. The high-resolu-
tion 2DE-PAGE and the automated spot handling and protein identification allows one to rap-
idly find new potential candidate proteins that are of functional relevance for regulatory T-cells,
to be used as targets for drug development or as biomarkers for research and diagnostic pur-
poses.
Key Words: Proteome analysis; T-lymphocytes; mass spectrometry; electrophoresis; image
analysis; protein identification; in-gel digestion; silver staining; Coomassie staining; protein
database; 2DE; spot picking; sample preparation; T-regulatory cells; peptide mass fingerprint;
MALDI; peptide fragmentation; drug target; suppressor T-cells; drug development.
1. Introduction
Proteins as targets are dominating pharmaceutical research and development
(R&D). Forty-five percent of the drugs on the market modulate receptorligand inter-
actions, and 28% are directed to enzymes. Only 2% of the drugs in therapeutic use
interfere directly or indirectly with DNA (2). There is no evidence for a dramatic
change of this fact in the future, since antisense and DNA-intercalating drug candi-
dates have yet to show their broader applicability.
155
156 Weingarten et al.
Additionally, the pharmaceutical industry is under strong pressure to reduce over-

all R&D costs. One major chance to achieve this goal is to get access to proprietary
and validated targets with a proven role in diseased states and new mode of action
(MOA). This allows development of a competitive advantage and longer and stron-
ger patent protection for new drugs. Validated targets with new MOA and corre-
sponding lead compounds are the key for future development of the pharmaceutical
industry.
Huge efforts have been put into the sequencing of the human genome, setting the
basis for the todays activities in functional genomics. Technological progress led to
the generation of vast amounts of data in a short time, and to a high number of new
potential targets that have to be validated for the pharmaceutical R&D process. The
development of all new technologies for target discovery, and in particular the techno-
logical revolution in proteomics, led to new approaches in pharmaceutical R&D and
made access to validated targets faster and more cost effective.
The classical way of performing proteomics studies via two-dimensional gel elec-
trophoresis (2DE) and mass spectrometry in combination with sophisticated bio-
informatics is still dominating the proteomics world. Reproducible and sensitive 2DE
systems are mandatory for the process. The 2DE system from one of the inventors of
the 2DE, Joachim Klose, is the most reliable and reproducible compared to the other
systems used in the community. It uses the ampholine technology for the first dimen-
sion in combination with long-distance separation, leading to superior performance
(1,3,4).
The protocols presented here are depicted from studies focused on early drug
development. Several mechanisms control the discrimination between self and non-
self, including the thymic deletion of auto-reactive T-cells and the induction of anergy
in the periphery. In addition to these passive mechanisms, evidence has accumulated
for the active suppression of auto-reactivity by a population of regulatory T-cells that
co-express CD4 and CD25. Regulatory T-cells play a central role in the maintenance
of the immunological balance and are powerful inhibitors of T-cell activation both
in vivo and in vitro (5). The enhancement of suppressor-cell function might be a target
for immunotherapeutic approaches for the treatment of immune-mediated diseases.
By contrast, the elimination of regulatory T-cells or the inhibition of their functions
might prove to be beneficial for the induction of tumor immunity or the treatment of
chronic infectious diseases.
To elucidate the still unclear effector functions of regulatory T-cells, we performed
differential proteome analyses with diverse human and murine T-cell populations (6).
The whole protein extracts of conventional and regulatory T-cells of four individual
human donors and BALB/c mouse pools were separated by high-resolution two-
dimensional gel electrophoresis. The proteomes of resting as well as anti-CD3/CD28-
stimulated CD4+ and CD4+ CD25+ T-cells were compared. The identification of can-
didate proteins that are of functional relevance for both human and murine regulatory
T-cells contributes to the understanding of the causes for and mechanisms involved in
autoimmune diseases, allergy, and cancer, and will lead to the development of new
drugs to manipulate the activity of regulatory T-cells (7).
2. Materials
2.1. Equipment
2.1.1. Sample Preparation
1. Sonication bath to perform sonication in a water bath, e.g., Sonorex super RK 52H
(Bandelin Electronic, Berlin, Germany).
2. Glass beads with a diameter of 1.02.0 mm (Roth, Karlsruhe, Germany).
3. Complete protease inhibitor cocktail (Roche Diagnostics, Penzberg, Germany).
4. Phenylmethylsulfonyl fluoride (PMSF) in ethanol.
5. Pepstatin A.
6. Urea.
7. Thiourea.
8. Dithiothreitol (DTT).
9. Ampholine Servalyt 2-4 (Serva, Heidelberg, Germany).
10. Benzonase.
11. Magnetic stirrer with 5 1 mm stirring bars.
12. Fine tweezers.
2.1.2. Isoelectric Focusing (IEF)

1. Glass tubes for IEF gels: length 40 cm or 20 cm; inner diameter: 0.7 mm for analytical
gels or 0.9 mm for preparative gels; outer diameter: 7.0 mm (Schott Glaswerke, Mainz,
Germany).
2. Nylon strings (fishing line, 0.8 mm thick; one end is made thicker by melting the end of
the string and fitting it quickly into the glass tube).
3. IEF chamber: the apparatus consists of an upper and a lower cylindrical buffer chamber
and is equipped with intermediate cylinders in various heights to allow gel tubes of vari-
ous lengths. Up to eight glass tubes can be inserted into the apparatus and fixed with
screws (for detailed view see ref. 3).
4. Power supply: EPS 3500 XL (Amersham Biosciences Europe, Freiburg, Germany).
5. Gel loader tips (Eppendorf, Hamburg, Germany).
6. IEF gel holder (for detailed view see ref. 3).
7. Plastic boxes to store the IEF gels on the holder at 18C to 80C.
2.1.3. SDS-PAGE
1. Desaphor VA 300 system with equipment (Sarstedt, Nuembrecht, Germany).
2. Plastic spacers, 0.75 mm or 1.5 mm, for analytical or preparative gels (Sarstedt).
3. Power Supply: EPS 3500 XL (Amersham Biosciences Europe).
4. HydroTech Gel Drying System (Bio-Rad Laboratories, Mnchen, Germany).
5. Cellophane film (Ptz Folien, Taunusstein, Germany, or Bio- Rad Laboratories, Mnchen,
Germany).
2.1.4. Gel Staining

1. Staining trough (30 cm 40 cm).
2. Horizontal gel shaker type 3020 (GFL, Burgwedel, Germany).
2.1.5. Image Analysis

1. Flatbed scanner PowerLook 2100 (Umax Data Systems, Taiwan).
2. ProteomWeaver Vers. 2.0 (Definiens, Mnchen, Germany).
2.1.6. Spot Excision

1. Spot Picking Robot PROTEINEER sp (Bruker Daltonik, Bremen, Germany).
2. Pierced microtiter plates (Bruker Daltonik).
2.1.7. Digestion
1. Digesting robot PROTEINEER dp (Bruker Daltonik).
2. Digestion Kit: dp 384 Reagent Kit (Bruker Daltonik).
3. Acetonitrile (p.a.), acetone (p.a.) (Merck, Darmstadt, Germany).
2.1.8. MALDI Analysis

1. Ultraflex TOF/TOF Mass Spectrometer (Bruker Daltonik).
2. MALDI 600 m AnchorChip Target (Bruker Daltonik).
2.1.9. Database Analysis

1. Proteinscape Database (Bruker Daltonik).
2. ProFound (Genomic Solutions, Ann Arbor, MI).
3. Mascot (Matrix Science, London, UK).
4. Sequest (Thermo Finnigan, San Jose, CA).
2.2. Buffers and Solutions

2.2.1. Sample Preparation
1. Protease inhibitor solution: 0.2 M KCl, 10% (w/v) glycerol, 0.1 M phosphate buffer
(NaH2PO4/Na2HPO4, 1:2) pH 7.1, 1 mM MgSO4H2O, 0.13 M CHAPS, 2.6 M pepstatin
A, 1.8 mM PMSF, 1 pill Complete/5 mL.
2.2.2. IEF
1. Separation gel solution: The composition is given in Table 1. The gel solution is filtered
(GF/C glass microfiber filter, diameter 7 cm, Whatman Scientific., Maidstone, Kent, UK),
aliquoted into 2-mL portions, and stored at 70C. To remove acrylic acid, 20 mL of the
acrylamide solution (containing 3.5% [w/v] acrylamide and 0.3% [w/v] piperazine
diacrylamide) is mixed with 1 g prewashed Amberlite MB-1A and agitated for 1 h. There-
after, the solution is filtered to remove the ion-exchange material.
2. Cap gel solution: The composition is given in Table 1. The gel solution is filtered (GF/C
glass microfiber filter, diameter 7 cm, Whatman Scientific), aliquoted into 750-L por-
tions, and stored at 70C. To remove acrylic acid, 20 mL of the acrylamide solution
(containing 3.5% [w/v] acrylamide and 0.3% [w/v] piperazine diacrylamide) is mixed
with 1 g prewashed Amberlite MB-1A and agitated for 1 h. Thereafter, the solution is
filtered to remove the ion-exchange material.
3. Ammonium persulfate solution: 0.08 g ammonium persulfate is dissolved in bidistilled
water to give a final volume of 10 mL. The solution is filtered, aliquoted into 100-L
portions, and stored at 70C.
4. Sephadex suspension: 20 g Sephadex G-75 superfine (Amersham Biosciences Europe) is
swollen in 500 mL bidistilled water for 5 h at 90C. The supernatant water is decanted
and the Sephadex resuspended into 1000 mL 25% (w/v) glycerol and gently stirred for
2 h. The glycerol solution is changed once during the 2-h period. The supernatant water is
then sucked off from the suspension by using a filter funnel. The Sephadex is aliquoted
into 272-mg portions and stored at 20C.
Table 1
Solutions for IEF
Separation gel solution pH 2.011.0 Cap gel solution pH 2.011.0
3.5% (w/v) Acrylamide 12.3% (w/v) Acrylamide
0.3% (w/v) Piperazine diacrylamide 0.13% (w/v) Piperazine diacrylamide
4% (v/v) Carrier ampholyte solution 4% (v/v) Carrier ampholyte solution
pH 2.011.0 pH 2.011.0
9M Urea 9M Urea
5% (w/v) Glycerol 5% (w/v) Glycerol
0.06% (v/v) TEMED 0.06% (v/v) TEMED
Carrier ampholyte solution pH 2.011.0 Sample protection solution

8 mL Ampholine pH 3.510.0 30% (w/v) Urea
8 mL Servalyt pH 2.011.0 5% (w/v) Glycerol
16 mL Pharmalyte pH 5.08.0 2% (v/v) Servalyt 24
24 mL Pharmalyte pH 4.06.5
8 mL Servalyt pH 6.59.0
The chemicals were dissolved in bidistilled water. Servalytes are from Serva (Heidelberg, Ger-
many) and Pharmalytes are from Amersham Biosciences Europe (Freiburg, Germany).
5. Sample protection solution: The composition is given in Table 1. The resulting solution
is filtered, aliquoted into 100-L portions, and stored at 70C.
6. Anode solution: 0.742 M phosphoric acid, 3 M urea (prepared freshly).
7. Cathode solution: 5% (v/v) ethylenediamine, 5% (v/v) glycerol, 9 M urea. The electrode
solutions are prepared freshly before each IEF run. The cathodic solution can be heated
up to 30C to better dissolve the urea (prepared freshly).
8. Equilibration solution: 125 mM Tris-base, 40% (v/v) glycerol, 3% (w/v) SDS, 65 mM
DTT, pH 6.8. The final solution is filtered, aliquoted into 10-mL portions, and stored at
70C.
2.2.3. SDS-PAGE
1. Gel solution: 1.5 L of the ready-made mixture of 30% acrylamide and 0.4% bisacryla-
mide (Serva) are added to 1.5 L buffer solution containing 1.5 M Tris/Tris-HCl, pH 8.8
(138.0 g Tris and 56.8 g Tris-HCl dissolved in 2 L H2O will give a pH of 8.8), 0.4%
SDS (w/v), and 0.12% (v/v) TEMED. The mixture is divided into two portions, filtered,
and degassed for 30 min. The two portions are combined carefully, avoiding a new
intake of air. The gel solutions are aliquoted into 70-mL portions for analytical gels and
140-mL portions for preparative gels, and stored at 20C.
2. Ammonium persulfate solution: 40 g ammonium persulfate are dissolved in bidistilled
water to give a final volume of 100 mL. The solution is filtered, aliquoted into 1-mL
3. Underlayering solution: 40 g glycerol are dissolved in distilled water to give a final vol-
ume of 100 mL and stored at 4C.
4. Overlayering solution A: 1-Butanol saturated with Electrode buffer. B: 17.25 g Tris,
7.1 g TrisHCl, and 0.5 g SDS are dissolved in distilled water to give a final volume of
500 mL. The solution is aliquoted into 20-mL portions and stored at 20C.
Table 2
Solutions for Silver Staining
Solution Composition
Fixation solution (A) 50% (v/v) Ethanol
10% (v/v) Acetic acid
Incubation solution (B) 0.5 M Sodium acetate
0.2% (w/v) Sodium thiosulfate
30% (v/v) Ethanol
0.5% (v/v) Glutardialdehyde
Silver nitrate solution (C) 0.1% (w/v) Silver nitrate
0.01% (v/v) Formaldehyde
Wash solution (D) 2.5% (w/v) Sodium carbonate
Developer solution (E) 2.5% (w/v) Sodium carbonate
0.1% (v/v) Formaldehyde
Stop solution (F) 0.05 M EDTA
Table 3
Solutions for Colloidal Coomassie Staining
Solution Composition
Fixation solution (A) 50% (v/v) Methanol
2% (v/v) Phosphoric acid
Incubation solution (B) 34% (v/v) Methanol
2% (v/v) Phosphoric acid
17% (w/v) Ammonium sulfate
5. Agarose buffer: 0.2 g SDS (final concentration 0.1%) is dissolved in bidistilled water to
give a final volume of 174 mL. 25 mL 125 mM Tris buffer (pH 6.8, titrated with phospho-
ric acid) are added to this solution. The resulting solution is filtered, aliquoted into 10-mL
6. Agarose solution: 1% low-melt agarose and a few crystals bromphenol blue (both elec-
trophoresis grade, Bio-Rad Laboratories) are added to the agarose buffer. The suspension
is dissolved by melting at 70C. The final solution is used at 40C.
7. Electrode buffer: The buffer is prepared freshly for each 2DE run. 0.025 M Tris, 0.192 M
glycine, and 0.1% (w/v) SDS are dissolved in distilled water. A 10X buffer solution can
be prepared and stored for at least one week. Dilute the 10X buffer immediately before
the 2DE run.
2.2.4. Gel Staining
2.2.4.1. SOLUTIONS FOR SILVER STAINING OF 2DE GELS
Six solutions are needed for the silver staining according to Heukeshoven (8). The
composition of the solutions is given in Table 2. All solutions are prepared freshly.
2.2.4.2. SOLUTIONS FOR COLLOIDAL COOMASSIE STAINING
Two solutions are needed for the colloidal Coomassie staining according to Neuhoff
(9). Both solutions are prepared freshly. The composition is given in Table 3.
3. Methods
3.1. Sample Preparation
Cell lysis and protein preparation are important steps with much influence on the
quality and reproducibility of 2D gels. The sample preparation is performed by a
modified protocol according to Klose and Kobalz (10). The cells can either be pre-
pared immediately after harvesting and PBS washing or from frozen samples (see
Note 1).
1. The cell pellet is mixed rapidly with the protease inhibitor solution. The amount of the
applied solution is calculated as follows: mg of the pellet (wet weight) 1.25 = L of the
protease inhibitor solution (see Notes 2 and 3).
2. The resulting sample volume is estimated by adding the wet weight of the cell pellet (mg)
and the volume of the protease inhibitor solution in L.
3. The cells are lysed by sonication in the presence of small glass beads (12 mm diameter).
The number of glass beads should be one per 30 L sample solution. Sonication is per-
formed 6 times for 10 s, each time leaving intervals of 50 s at 4C.
4. Ribonucleic acids are digested with benzonase (0.5 units/L sample volume) for 20 min
at room temperature with gentle stirring (see Note 4).
5. Then 7 mM urea, 2 mM thiourea, 70 mM DTT, and 2% carrier ampholytes Servalyt pH
2.04.0 (Serva) are added and the proteins are solubilized at room temperature.
6. The glass beads and the magnetic stirring bars are removed with fine tweezers.
7. The protein concentration is determined by amino acid analysis or other suitable methods
(e.g., Bradford [11]) (see Note 5). The cell lysate is frozen and stored at 70C.
3.2. IEF
3.2.1. Preparation of IEF Gels
1. The nylon strings are inserted into the glass tubes, which are held vertically.
2. Separation gel solution, cap gel solution, and ammonium persulfate solution are thawed.
The separation gel solution, as well as the cap gel solution, are degassed for 5 min.
3. Then 51 L of ammonium persulfate solution are mixed into the separation gel solution,
and the glass tubes are filled by dipping the end of the tubes into the gel solution and
lifting the nylon string to suck the separation gel solution to a height of 35.5 cm (for 40-
cm IEF gels) or 15.5 cm (for 20-cm IEF gels) into the tubes.
4. Thereafter, 19 L of ammonium persulfate solution are mixed into the cap gel solution,
and the cap gel solution is sucked into the glass tubes for an additional 0.8-cm height.
5. Immediately afterwards, the nylon string is lifted for a further 0.5 cm to obtain at the end
of the tube a small space free of gel solution.
6. The filled gel tubes are left in vertical position at room temperature for 20 min to allow
polymerization. Afterwards, the gels are stored in the dark in a humid atmosphere. To this
end the gels are stored in a sealed plastic bag in the presence of a wet filter paper. The gels
are used for isoelectric focusing after keeping them at room temperature for 3 d in the
dark.
3.2.2. Sample Application and IEF Run
1. The glass tubes are inserted into the IEF apparatus (cap gel on the cathodic side) and
fixed with the screws (see Note 6). The cathodic buffer is poured into the lower chamber.
2. The gel-free ends of the glass tubes are filled with cathodic buffer and placed into the
cathodic buffer. Avoid any air bubbles.
Table 4
Electric Current Gradient for Isoelectric Focusing
20 cm IEF gels 40 cm IEF gels
60 min 100 V 60 min 100 V
60 min 200 V 60 min 300 V
17.5 h 400 V 23 h 1000 V
60 min 650 V 30 min 1500 V
30 min 1000 V 20 min 2000 V
10 min 1500 V
10 min 2000 V
= 9233 Vh = 24,817 Vh
3. 330 g of a fine mixture of urea/thiourea (5.1:2, w/w), 25 L DTT solution, and 25 L of

carrier ampholyte solution pH 2.011.0 (see Table 1) are added to the Sephadex suspen-
sion.
4. 12 mm of the final Sephadex suspension are pipetted onto the upper (anodic) gel sur-
face. This is best done with fine Pasteur pipets.
5. The protein sample is pipetted onto the Sephadex and carefully overlaid with 1 mm
sample protection solution. This is best done with the gel loader tips. The sample volume
depends on the protein concentration. For gels used for the image analysis (analytical
gels), 60100 g protein in 10 to 15 L are applied to a gel. For preparative gels used for
protein identification, 150300 g protein in 20 to 35 L are applied.
6. The glass tubes are filled with anodic buffer, and the rest of the buffer is supplied to the
upper chamber covering the glass tubes.
7. Isoelectric focusing is performed using the current gradient, as indicated in Table 4.
8. After the focusing is finished, the excess solutions from the lower and upper side of the
glass tubes are removed by a Pasteur pipet.
9. The nylon string is used to gently push the IFF gels off the tubes onto a gel holder covered
with equilibration solution. The gels are incubated for 10 min in the solution. Then the
solution is removed with a pipet and filter paper. Touching the gels with the filter paper is
to be avoided.
10. For 40-cm gels, mark on the glass tube the half-length of the IEF gel before extruding the
gel. The gels are cut into two halves prior to SDS-PAGE.
11. The IEF gels are stored at 70C.
3.3. SDS-PAGE
3.3.1. Preparation of Gels for SDS-PAGE
1. All glass plates are cleaned prior to use to avoid point streaking during the silver stain.
The gel cassettes are prepared by placing the spacer covered with a thin layer of silicone
grease between two identical glass plates to seal the cassette. The plates are fixed accord-
ing to the manufacturers instructions with two clamps.
2. The cassettes are placed into the polymerization stand.
3. Ammonium persulfate solution (144 L) is added to 70 mL of the thawed gel solution and
gently mixed by shaking.
Table 5
Silver Staining Procedure
Volume Duration
Gel treatment of solution/gel of treatment Remarks
Fixation 0.5 L Solution A 2 h or overnight
Incubation 0.5 L Solution B 2h
Wash 2 L Distilled water 2 20 min
Silver reaction 0.5 L Solution C 30 min
Wash 0.5 L Distilled water few s Remove the water immediately.
Wash 0.5 L Solution D 1 min Distribute arising silver clouds quickly.
Development 0.5 L Solution E 310 min Watch the staining intensity
Stop reaction 0.5 L Solution F 20 min
4. The gel solution is poured into the cassette, air bubbles are removed, and the cassette is
filled up just 0.5 cm below the upper side of the glass plate. The gel cassettes are carefully
filled up with the overlayering solution A.
5. The gels are allowed to polymerize for at least 30 min at room temperature.
6. The overlayering solution A is replaced by overlayering solution B and the gels are stored
at 4C overnight.
3.3.2. Protein Separation
1. For the protein separation in the second dimension, the tube gels are thawed slowly and
washed once with SDS running buffer. The buffer is removed with a pipet.
2. The gels should be left wet for a better transfer to the SDS gels.
3. The washed tube gels are transferred onto the SDS gel surface without being stretched or
compressed. The inclusion of air or solution should be avoided.
4. The IEF gels are overlaid with 40C agarose solution, and the agarose is gelatinized at
room temperature.
5. The gel cassettes are inserted into the electrophoresis apparatus, and the electrophoresis
chamber is filled up with electrode buffer.
6. The temperature during SDS-PAGE run should be kept constant at 15C.
7. The analytical gels are run for 15 min at 65 mA and than at 100 mA until the bromphenol
blue reaches the end of the gel. For preparative gels, the gels are run for 15 min at 75 mA
and then at 200 mA.
8. After the run, the gels are transferred into the appropriate fixation buffer.
3.4. Gel Staining
3.4.1. Silver Staining
1. The 2DE gels are fixed in the fixation solution for at least 2 h, better overnight. The gels
are shaken gently to avoid the gels sticking together during the fixation process.
2. Afterwards, each gel is transferred into a trough. The staining protocol is listed in Table 5.
During the staining process, the gels stay in the same trough while the solutions are being
sucked away.
A silver-stained 2DE gel of human regulatory T-cells is displayed in Fig. 1.
Fig. 1. Large-gel 2DE pattern of human CD4+ CD25+ T-regulatory cells. The sample prepa-
ration and production of the 2DE gel was performed as described in the text. The sample load
was 60 g protein /10 L sample. The gel was stained with silver according to Heukeshoven.
Approximately 3500 protein spots were detected using the ProteomWeaver image analysis
software.
3.4.2. Colloidal Coomassie G-250 Staining

The colloidal Coomassie G-250 staining according to Neuhoff (9) is less sensitive
than silver staining but compatible with mass spectrometry based protein analysis
methods. Depending on the protein properties, the detection limit is approx 10 ng
protein per spot.
1. The 2DE gels are fixed in the fixation solution for at least 2 h, better overnight. The gels
are shaken gently to avoid the gels sticking together during the fixation process.
2. Afterwards, the gels are transferred into a large trough. The staining protocol is listed in
Table 6.
3. For the destaining process, the gels are placed into a new trough and the water is changed
several times.
4. For highest sensitivity in the mass spectrometrical analysis, the gels should be processed
immediately after the destaining, but they can be stored for several days at 4C in a plastic
foil.
Table 6
Staining Process With Colloidal Coomassie G-250
Volume Duration
Gel treatment of solution/gel of treatment Remarks
Fixation 0.5 L Solution A overnight
Wash 0.5 L Distilled water 3 30 min
Incubation 0.5 L Solution B 1h
Staining 0.5 L Solution C up to 6 d
(add 660 mg Coomassie
G-250 to solution B)
Destaining Distilled water several h Change the water several times during
the destaining process.
3.5. Image Analysis

Image analysis is perhaps the most crucial element of proteome analysis. There
are many systems available for image capturee.g., CCD cameras and white light or
laser scanners. For visible stains like silver, or Coomassie stain, conventional scan-
ners can be used successfully if they have a resolution of 16-bit grayscale or better. A
typical work flow of 2DE gel image analysis comprises the following steps: scanning
of the gel image (image capture), image capture (data acquisition), spot detection, gel
matching, data analysis and interpretation, and database creation or export to a data-
base.
1. The gels are dried between two cellophane films prior to the image-capture process with
a gel dryer. To digitize large 2D gels, a 440 355 mm flatbed scanner is utilized.
2. In case of gels used for the image analysis and determination of spot intensities, the gel is
placed in the scanner and scanned as a transmissive 16-bit grayscale image with 200-dpi
image resolution and a value of 1.0. Coomassie-stained gels are scanned without drying
as transmissive 48-bit-depth color image with 200-dpi resolution (see Note 7).
3. To ensure high reproducibility, the scanning parameters should always be kept constant.
4. Gels are analyzed semi-automatically with the ProteomWeaver software. Gels of the
same sample (replicates) are sorted into the same group. Different groups are compared
according to the analytical request.
5. Spot detection and pre-matching normalization are performed within the same experi-
ment automatically. To eliminate spots with low signal-to-noise ratios, pre-matching fil-
ters are set and saturated spots are eliminated manually. Then the gel images are matched
with each other. Obvious mismatches and incorrect spot detection are corrected manu-
ally.
6. To achieve more precise results, a pair-match basis normalization is carried out after spot
matching.
7. Spots with a standard deviation >50% within the same group are eliminated by the spot
filter tool for reliable analysis and interpretation of differential spots.
8. For inter-group comparison, several statistical filters are used to localize spots that
increased or decreased or are present uniquely in one group. Spot ratios >2 for increased
and <0.5 for decreased spots are considered as differential if confirmed by the soft-
ware-implemented statistical tests.
9. The calibration is performed on the basis of a set of protein spots with known isoelectric
points (pI) and molecular weights (MW).
In Fig. 2, enlargements of gel regions of the different T-cell populations are shown.
Using the ProteomWeaver software, the marked spots were analyzed as differen-
tially expressed, comparing the T-cell populations.
3.6. Spot Excision
The spots of interest as identified by image analysis are then located on a prepara-
tive gel, excised, and digested (12,13). The spot excision, digestion, and MALDI analy-
sis are fully automated in order to minimize keratin contamination (see Note 8), to
maximize reproducibility, and to facilitate consistent sample tracking and data han-
dling. The identification of proteins from a 2DE gel is a multistep process that is out-
lined in Fig. 3 and described in detail below.
The spot-picking robot (PROTEINEER sp) is employed to excise the protein
spots from the preparative 2DE gels and to deposit them into specially pierced
microtiter plates for further analysis (see Notes 9,10). The robot has a high-resolution
scanner integrated on its platform for acquisition of high-quality images. A spot-de-
tection algorithm integrated into the software localizes the exact position of the spots
of interest.
1. The preparative gel is laid into a so-called sandwich assembly (see Fig. 4), which consists
of a glass plate and a metal frame. This assembly simplifies the subsequent handling espe-
cially of unbacked gels, and prevents the gel from moving during the picking process.
2. The gel in this sandwich assembly is then fitted onto the scanner surface, and an image of
the gel is acquired on the spot-picker platform and stored in a relational database.
3. Four pierced PCR plates in a plate holder (see Note 11) are put onto the robot deck, and
accommodate the picked gel plugs. All samples are given a unique identifier that is writ-
ten to a transponder attached to the plate holder.
4. Then the spots of interest are selected either by using a simple point-and-click interface
or by employing the integrated spot-detection capability.
5. Before initiating the picking process, all processing steps for the samples, including the
digestion and the MALDI acquisition method, are defined. For the spot picker, an opti-
mized picking protocol is used that is provided by the manufacturer.
6. The picking is then initiated. Using this setup, usually >99% of the gel plugs from the
2DE gels are successfully ejected into the microtiter plate. Using such an automated
setup, keratin contamination can be minimized in this step and the exact x-y coordinates
of every picked spot can be stored in a database.
3.7. Digestion
After picking has ended, the microtiter plates with the gel plugs are transferred to the
digestion robot (PROTEINEER dp). This robot is capable of washing and digesting
384 gel cores at a time. The microtiter plates from the spot picker are put onto the appro-
priate positions of the digestion robot (see Note 12).
1. A reagent kit containing all reagents needed for the process (apart from organic solvents)
in sealed vessels is mounted in the reagent rack. This reagent kit is supplied by the vendor
Proteomics for Biomarker and Target Finding
167
Fig. 2. Selected gel regions showing different proteins expressed in murine resting (i) and anti-CD3/CD28 stimulated conventional CD4+
and CD4+ CD25+ T-regulatory cells. The differential proteins were found by image analysis with ProteomWeaver image analysis software.
The expression profiles of the four cell types of two independent mouse pools were analyzed.
167
Fig. 3. Schematic representation of the protein identification process from 2DE gels.
Fig. 4. The sandwich assembly used for the spot picking robot. The 2DE gel is put on the
glass plate and held in place by the frame on top. Two handles allow the easy removal of the
assembly from the scanner surface.
of the robot, together with optimized protocols. The organic solvents are poured into
troughs (see Note 13).
2. A clean MALDI target is fitted into the appropriate position of the robot deck.
3. As the protocol used for digestion was already defined when picking the spots, the user
only needs to press the Full-Automode button.
4. The robot then automatically identifies the samples on the deck and in a sample-tracking
database and processes them accordingly. Again, an optimized digestion protocol is sup-
plied by the manufacturer.
a. The gel plugs are washed briefly to remove the Coomassie stain and dried using aceto-
nitrile.
b. Trypsin is applied to the gel plugs and the samples are incubated for 48 h at 30C.
c. The resulting peptides are extracted and spotted onto an AnchorChip MALDI target
that was precoated with -cyano-4-hydroxy-cinnamic acid and washed.
3.8. MALDI Analysis

1. When the digestion process is finished, the MALDI target is taken from the digestion
robot and inserted into the MALDI mass spectrometer. In a first analysis, the masses of
the intact peptides generated in the digestion process are measured. As each protein gives
rise to a unique set of peptides, these sets are as unique as a fingerprint; thus the spectra
are termed peptide mass fingerprint (PMF) spectra.
2. The target spots are analyzed by summing up 200 shots per spot while the laser attenua-
tion is adjusted automatically to an optimal level by utilizing a fuzzy-logic feedback algo-
rithm.
3. The peaks in the spectrum are labeled using the SNAP algorithm provided by the vendor
of the mass spectrometer (Bruker Daltonik, Bremen, Germany) in order to identify the
monoisotopic peak of each peptide signal. The spectra are then exported to the
proteinscape database for protein identification (Fig. 5).
4. In a second round of analysis, the individual peptides from the digest are fragmented in
order to obtain PFF spectra that allow identification of a protein from only one peptide.
The acquisition of these PFF spectra is more time-consuming than the acquisition of PMF
spectra. Therefore, the acquisition of PFF spectra is triggered by an intelligent algorithm
in the database that can choose peptides suitable for fragmentation based on different
criteria, such as intensity or protein identification.
3.9. Database Analysis

1. The spectra that are imported into the database are automatically linked to the appropri-
ate spots in the 2DE gel image that were imported from the spot picker. The fully auto-
mated flow of information ensures the integrity of the data set, as it eliminates user
intervention.
2. In the database, the spectra are recalibrated using known contaminant peaks (from kera-
tin, trypsin, and so on) in order to obtain precise peptide masses. The masses of the con-
taminant peaks are then removed from the spectra so that only peaks with unknown masses
are used for database searching (see Note 14).
3. Protein identification is achieved by searching the mass spectra against a protein database
using several external search algorithms (ProFound, Mascot). The results are col-
lated into one table with a unified scoring system (see Note 15) that allows a comparison
of the different algorithms. The significance of the search result is then evaluated based
on the unified score.
4. Notes
1. Normally, cell-culture media contain serum (e.g., fetal calf serum). To remove the serum
proteins from the sample, the cells are washed well with PBS.
2. The amount of all reagents for the sample preparation is calculated on the basis of the wet
weight of the cells. After harvesting and washing the cells, they are centrifuged at max.
300g at 4C and the PBS is removed. Make sure that no excess buffer dilutes the sample.
3. Proteases are included in all cells. To inhibit their activity, suitable protease inhibitors
against all types of proteases should be added to the sample. They are best added before
or during the thawing process or immediately prior to cell lysis.
4. The sample preparation can be performed without nucleic acid digestion, but 2DE gels
will exhibit much better performance (well-defined spots, less streaking) if the nucleic
acids are digested either with benzonase or other RNases and DNases.
5. For the determination of the protein concentration, the methods of Bradford (11), Lowry
(14), or Popov (15) can be used. The most accurate method for the determination of pro-
tein concentration is amino acid analysis.
6. Within one proteome study, always run the same number of gels in parallel to yield maxi-
mum reproducibility during IEF. Different numbers of gels result in different values of
resistance and current, and therefore in different migration distances of the proteins dur-
ing the isoelectric focusing process.
7. If 40-cm IEF gels are used, the 2DE gels consist of two halves. After the gels are scanned,
they are merged together digitally with suitable software (Corel Photo-Paint or Adobe
Photoshop). The merged gel images allow easier spot matching than the analysis of indi-
vidual halves.
8. When analyzing 2D gel spots, especially of low-abundance proteins, it is of great impor-
tance to keep the contamination of the sample at an absolute minimum. When handling
the preparative 2D gel and for all subsequent sample handling steps, gloves should be
worn, as keratin (from the skin) is one of the most persistent contaminants. Also, the
reagents should be stored in a way that minimizes exposure to the environment, and regu-
larly checked for keratin contamination.
9. Disposables such as tips or plates should not be washed and re-used. In our experience,
cross-contamination cannot be ruled out. This is especially true for 2D gel samples, where
the concentration of protein in a spot can vary by several orders of magnitude. Thus,
when re-using disposables, peptides from an abundant protein can contaminate the subse-
quent analysis of a low-concentration sample.
10. The quality of plastic used for disposables can vary widely from manufacturer to manu-
facturer. For example, when considering which brand of pipet tip to use, it is advisable
to spot a few microliters of organic solvent onto a MALDI target plate together with
matrix and to check the MALDI spectrum for polymer peaks. We recommend using tips
obtained from Eppendorf, Hamburg, Germany.
11. The microtiter plates are enclosed in a special storage container, which carries a tran-
sponder in order to identify the samples and which utilizes a special lid that minimizes the
exposure of the samples to contamination in the lab. The gel pieces are processed in
specially designed microtiter plates that function as flow-through reaction vessels, where
the liquid is applied from the top and is removed through small orifices in the bottom of
the wells by applying nitrogen gas pressure.
12. The order of the four-sample carrier on the robot deck of the digester does not have to be
the same as on the spot picker. As the digest robot reads the unique identifier of each
carrier prior to starting the protocol, a mismatch of the sample is prevented.
13. When storing organic solvents, special care must be taken not to contaminate these. It is
therefore strongly advisable to use the original bottle for long-term storage and to portion
the liquids by pouring only. For portioning we recommend the use of teflon-coated plas-
tic bottles (Nunc, Wiesbanden, Germany).
14. The recalibration and peak filtering is an important feature of the proteinscape data-
base used. For a typical data set, the identification rate for the MALDI PMF analysis can
be raised by 50100%.
Fig. 5. (opposite page) Schematic representation of the protein identification from MALDI
PMF data. The spectrum (top) is imported into the proteinscape database and processed by
different algorithms (such as ProFound or Mascot). The sequence of the identified protein
is then reported for further analysis.
15. Algorithms that can identify proteins from mass spectra are available from different ven-
dors. All these algorithms give slightly different results, especially in their ranking. As
each algorithm uses a different approach to calculate the quality of the protein identifica-
tion, the results cannot be compared directly. The database used in our setup can there-
fore calculate a separate score that is independent of the algorithm used in order to restore
the comparability.
Acknowledgments
We thank HD Dr. H. Jonuleit, Department of Dermatology, University of Mainz,
Germany, and Prof. Dr. E. Schmitt, Institute for Immunology, University of Mainz,
Germany, for isolating and providing the human and murine T-cells. This work was
partially supported by the European Funds for Regional Development (Europische
Fonds fr regionale Entwicklung, EFRE), and the government of North Rhine-
Westphalia.
References
1.
1 Klose, J. (1975). Protein mapping by combined isoelectric focusing and electrophoresis of
mouse tissues. A novel approach to testing for induced point mutations in mammals.
Humangenetik. 26, 231243.
2.
2 Drews, J. (2000) Drug discovery: a historical perspective. Science 287, 19601964.
3.
3 Klose, J. (1999) Large-gel 2-D electrophoresis. Methods Mol. Biol. 112, 147172.
4. Lutter, P. (2003) Untersuchung des Einflusses vasoaktiver Substanzen auf das Proteom
humaner Endothelzellen, Dissertation. Ruhr-University Bochum, Bochum Germany.
5.
5 Jonuleit, H., Schmitt, E., Kakirman, H., Stassen, M., Knop, J., and Enk, A. H. (2002)
Infectious tolerance: human CD25+ regulatory T cells convey suppressor activity to con-
ventional CD4+ T helper cells. J. Exp. Med. 196, 255260.
6. Jonuleit, H. and Schmitt, E. (2004) Regulatory T Cells in Antitumor Therapy: Isolation
and Function of CD4+CD25+ Regulatory T Cells. In: Methods in Molecular Medicine,
vol. 109: Adoptive Immunotherapy: Methods and Protocols. (Ludewig, B. and Hoffman,
M. W., eds.) Humana, Totowa, NJ, pp. 285296.
7. Lutter, P., Jonuleit, H., Schmitt, E., et al. (2003) Proteome analysis of regulatory t cells
promising targets for innovative immunotherapeutical approaches. Mol. Cell. Proteomics
9, 990.
8.
8 Heukeshoven, J. and Dernick, R. (1988). Improved silver staining procedure for fast stain-
ing in PhastSystem Development Unit. I. Staining of sodium dodecyl sulfate gels. Electro-
phoresis 9, 2832.
9.
9 Neuhoff, V., Arold, N., Taube, D., and Ehrhardt, W. (1988). Improved staining of proteins in
polyacrylamide gels including isoelectric focusing gels with clear background at nanogram
sensitivity using Coomassie Brilliant Blue G-250 and R-250. Electrophoresis 9, 255262.
10.
10 Klose, J. and Kobalz, U. (1995) Two-dimensional electrophoresis of proteins: an updated
protocol and implications for a functional analysis of the genome. Electrophoresis 16, 1034
1059.
11.
11 Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quan-
tities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248254.
12.
12 Lutter, P., Meyer, H. E., Langer, M., et al. (2001) Investigation of charge variants of
rViscumin by two-dimensional gel electrophoresis and mass spectrometry. Electrophoresis
22, 28882897.
13.
13 Shevchenko, A., Wilm, M., Vorm, O., and Mann, M. (1996) Mass spectrometric sequenc-
ing of proteins from silver-stained polyacrylamide gels. Anal. Chem. 68, 850858.
14.
14 Lowry, O. H., Rosebrough, N. J., Farr A. L., and Randall, R. J. (1951) Protein measure-
ment with the folin phenol reagent. J. Biol. Chem. 193, 265275.
15. Popov, N., Schmitt, M., Schulzeck, S., and Matthies, H. (1975), Eine strungsfreie
Mikromethode zur Bestimmung des Proteingehaltes in Gewebehomogenaten. Acta. Biol.
Me. Germ. 34, 14411461.
CTL Clones for Adoptive Immunotherapy 175
13
Isolation and Expansion of Tumor-Reactive
Cytotoxic T-Cell Clones for Adoptive Immunotherapy
Helga Bernhard, Burkhard Schmidt,

Dirk H. Busch, and Christian Peschel
Summary
Attempts to treat patients with tumor-reactive cytotoxic T-lymphocytes (CTL) have been
limited. This is due to the difficulty of isolating and expanding functionally active T-cells,
which are present at extremely low frequencies in the peripheral blood. Recently developed
multimers of the HLApeptide complex mimic the natural ligand of the T-cell receptor and,
therefore, fluorochrome-labeled multimers allow visualization and isolation of rare T-cells with
defined specificity. Multimer-guided T-cell sorting permits the in vitro culture of antigen-spe-
cific T-cells as lines or clones. Cytolytic T-cells capable of recognizing HLApeptide com-
plexes endogenously processed by tumor cells are selected for further expansion, since lysis of
tumor cells in vitro is a prerequisite for effective tumor elimination in vivo. The expansion of
tumor-reactive CD8+ T-cells yields cell numbers sufficient for adoptive transfer. Tumor-reac-
tive T-cells retain the functional activity in terms of cytolysis after expansion, encouraging
their use in the immunotherapy of cancer patients.
Key Words: Cytotoxic T-lymphocytes; human; antigens/peptides/epitopes; MHC; HLA;
tumor immunity.
1. Introduction
Patients with advanced cancer have an impaired cellular immune system (1). The
rationale of adoptive T-cell therapy is based on the attempt to circumvent this preex-
isting tolerance by taking out the anergic, potentially tumor-reactive T-cells from the
tolerizing environment and subsequently activating these T-cells ex vivo. Following
the expansion of tumor-reactive T-cells in vitro, large numbers of T-cells can be adop-
tively transferred to the patient (for review see ref. 2). With this approach, a high fre-
quency level of ex vivo activated tumor-reactive T-cells can be achieved in vivo, which
might not be accomplished by vaccine strategies.
One limitation of this approach is that the number of antigen-specific T-cells in the
peripheral blood is exceedingly low. Recently, a novel method of identifying antigen-
175
176 Bernhard et al.
specific T-lymphocytes has been described (3). Tetrameric MHCpeptide complexes,

called tetramers or multimers, have been shown to bind stably and specifically to
appropriate MHC-peptide-specific T-cells. This technique permits both detection and
isolation of antigen-specific T-cells present at low frequencies (4). If the natural
T-cell frequency in the blood is too low to be detectable by multimers, the number of
antigen-specific peripheral blood T-lymphocytes may be enhanced by repetitive stimu-
lation with peptide-loaded autologous dendritic cells (DC) in vitro. Peptide-specific
T-cells are visualized with fluorochrome-labeled HLA-peptide multimers and sorted
with a fluorescence-activated cell sorter (FACS). Following multimer-guided sorting,
T-cells are cultured as bulk or cloned by limiting dilution, both in the presence of
irradiated allogeneic peripheral blood mononuclear cells (PBMC), B-lymphoblastoid
cell lines (LCL), anti-CD3 monoclonal antibody, and interleukin (IL)-2 (5). Follow-
ing expansion, the specificity and the function of the sorted T-cells are confirmed by
the recognition of HLA-matched, antigen-expressing tumor cells. The expansion of
the tumor-reactive T-cells yields cell numbers sufficient for adoptive immunotherapy
in tumor patients. The generation of Melan-A-specific melanoma-reactive CTL clones
will be presented in order to illustrate the methods for the stimulation, isolation, clon-
ing, and expansion of peptide-specific, tumor-reactive CTL clones (see Fig. 1).
2. Materials
2.1. Chemicals and Buffers
1. Phosphate-buffered saline (PBS; Gibco-BRL, Karlsruhe, Germany).
2. Ficoll separation solution, density 1.077 g/mL (Biochrom, Berlin, Germany).
3. Erythrocyte lysis buffer: 8.3 g NH4Cl, 1 g KHCO3, 0.0372 g EDTA in 1 L H2O.
4. FACS buffer: PBS 0.1% bovine serum albumin (BSA; Sigma, Milwaukee, WI), pH 7.3.
5. Propidium iodide (PI; Sigma, Milwaukee, WI).
6. 2-microglobulin (Sigma, Taufkirchen, Germany).
7. Chromium51 (Cr51; ICN Biomedicals, Eschwege, Germany).
8. Nonidet P40 (NP40) buffer: PBS 1% NP40 (Sigma, St. Louis, MO).
2.2. Media and Supplements

1. RPMI 1640 (Gibco-BRL).
2. X-VIVO 15 (BioWhittaker, Walkersville, MD).
3. L-glutamine (Gibco-BRL).
4. Penicillin/streptomycin (Gibco-BRL).
5. Fetal calf serum (FCS; sera plus; Pan Biotech GmbH, Aldenbach, Germany).
6. Human serum albumin (HSA) solution: 0.9% NaCl solution with 20% HSA (20% Immuno;
Baxter, Unterschleiheim, Germany).
7. Serum from healthy donors with blood group AB (AB serum; Valley Biomedical, Win-
chester, VA).
8. Monocyte medium: RPMI 1640 supplemented with 1% heat-inactivated AB serum, 2 mM
L-glutamine, 100 U/mL penicillin, 100 g/mL streptomycin.
9. DC medium: X-VIVO 15 supplemented with 1% heat-inactivated AB serum, 2 mM
10. T-cell medium: RPMI 1640 supplemented with 10% heat-inactivated AB serum, 2mM
Fig. 1. Isolation, cloning, and expansion of melanoma-reactive cytotoxic T-lymphocytes

recognizing the differentiation antigen Melan-A.
For the detection of HLA-A2-restricted Melan-A2735-specific T-cells in the peripheral
blood, the HLA-peptide multimer consisting of the Melan-A2635A27L analog (ELAGIGILTV)
(9) was used, which has a higher binding affinity to HLA-A2 and a higher immunogenicity
than the natural nonapeptide Melan-A2735 (AAGIGILTV) (10). Dot plots are shown for CD8
versus HLA-A2-peptide multimer staining (A,B,E). Dot plots are gated on PI-negative cells.
The percentages of CD8+ multimer+ T-cells gated on PI-negative cells are stated in the upper
right quadrant (A,B). A low frequency of circulating Melan-A-specific T-cells was detected in
peripheral blood (A). The frequency of multimer+ T-cells increased after three stimulations
with Melan-A2735-pulsed autologous dendritic cells (B). Following sorting with fluorochrome-
labeled HLA-peptide multimers, CD8+ multimer+ T-cells were cloned (C). Proliferating T-cell
clones (e.g., 6/96, C) were screened for Melan-A-specific lytic activity using 51Cr-labeled T2
cells as target cells loaded with the relevant peptide Melan-A2735 respective an irrelevant pep-
tide derived from the HIV reverse transcriptase (HIV476484) (D). Following expansion of
Melan-A-specific cytotoxic T-lymphocyte (CTL) clones, antigen-specificity was confirmed by
HLA-A2-Melan-A multimer staining (E). Staining of cultured CTL clones did not reach 100%
of CD8+ multimer+ cells due to activated T-cells having downregulated the CD8 molecule and
due to remaining feeder cells. Expanded CTL clones retained lytic function, as documented in
a standard 51Cr release assay (F). The CTL clones lysed T2 cells loaded with the relevant
peptide Melan-A2735 (), but not T2 cells pulsed with the irrelevant peptide HIV476484 ().
Of importance, the CTL clone 1 was able to recognize endogenously processed peptides, as
documented by lysis of an HLA-A2+, Melan-A+ melanoma cell line (). CTL clone 1 did not
lyse an HLA-A2, Melan-A+ melanoma cell line (), confirming HLA-A2 as the restriction
element.
178 Bernhard et al.
11. Lymphoblastoid B-cell line (LCL) medium: RPMI 1640 supplemented with 10% heat-
inactivated FCS, 2 mM L-glutamine, 100 U/mL penicillin, 100 g/mL streptomycin.
12. Freezing medium: 20% HSA supplemented with 10% dimethylsulfoxide (DMSO; Serva
GmbH).
2.3. Cytokines and Prostaglandin

1. Recombinant human granulocyte macrophage-colony stimulating factor (GM-CSF; Cell-
Genix, Freiburg, Germany).
2. Recombinant human IL-1 (R&D Systems, Wiesbaden, Germany).
3. Recombinant human IL-2 (Chiron, Emeryville, CA).
4. Recombinant human IL-4 (R&D Systems).
8. Recombinant human IL-15 (Sigma, Milwaukee, WI).
9. Prostaglandin E2 (PGE2; Pharmacia, Erlangen, Germany).
10. Recombinant human tumor necrosis factor (TNF)- (Roche AG, Grenzach-Wyhlen, Ger-
many).
2.4. Monoclonal Antibodies

1. Anti-CD3 monoclonal antibody (OKT3; Janssen-CILAG, Neuss, Germany).
2. FITC-labeled mouse antihuman CD8 monoclonal antibody (clone 3B5; Caltag Laborato-
ries, Burlingame, CA).
3. Methods
3.1. Generation of DC From Monocytes
DC are generated from peripheral blood monocytes according to the method pub-
lished by Jonuleit et al. (6):
1. Draw peripheral blood in a heparinized syringe.
2. Split blood in aliquots (15 mL of blood per 50-mL conical centrifuge tube).
3. Add 15 mL PBS to 15 mL blood and mix.
4. Carefully layer 15 mL ficoll under the blood/PBS mixture.
5. Centrifuge cells at 900g for 15 min at room temperature (RT) in a swing-out rotor centri-
fuge without brakes.
6. After centrifugation the PBMC form a buffy coat at the ficoll/plasma interface. Transfer
PBMC located in the interphase carefully to a new 50-mL tube.
7. Fill the PBMC-containing tube with PBS to 40 mL, mix, and centrifuge at 600g for 10
min at RT.
8. Discard the supernatant and then resuspend all of the cell pellets with an appropriate
amount of PBS in order to pool the PBMC in a 50-mL conical tube.
9. Centrifuge at 450g for 8 min at RT and discard the supernatant.
10. Resuspend the cell pellet in 10 mL erythrocyte lysis buffer and incubate for 10 min at RT.
11. Fill cell-containing tube with 30 mL of PBS. Then mix and count the PBMC.
12. Centrifuge at 350g for 7 min, discard the supernatant, and resuspend the PBMC in mono-
cyte medium.
13. Seed PBMC in six-well flat-bottom plates at 1 107 cells/well in 3 mL monocyte medium
and incubate for 2 h in a 37C, 5% CO2, humidified atmosphere.
14. Remove nonadherent cells by rinsing and adding fresh monocyte medium (3 mL/well).
15. Incubate for 1020 h at a 37C, 5% CO2, humidified atmosphere.
16. Remove monocyte medium and add DC medium supplemented with 800 U/mL GM-CSF
and 1000 U/mL IL- 4 (3 mL/well) (d 1).
17. Culture monocytes in a 37C, 5% CO2, humidified atmosphere.
18. Add 1 mL/well DC medium supplemented with 1600 U/mL GM-CSF and 1000 U/mL
IL-4 on d 2, 4, and 6.
19. On d 7, harvest and pool nonadherent immature DC (see Note 1) in a 50-mL conical tube.
Count the DC and centrifuge at 350g for 5 min at RT.
20. Resuspend the cell pellet and seed the DC in a new six-well plate at a concentration of
5 105 cells/well in 2 mL DC medium containing GM-CSF (800 U/mL) and IL-4
(500 U/mL).
21. Incubate in a 37C, 5% CO2, humidified atmosphere.
22. After 4 h, add 1 mL/well DC medium with the following supplements at final concen-
trations: IL-1 (10 ng/mL), TNF- (10 ng/mL), IL-6 (1000 U/mL) and PGE2 (1 g/mL).
23. Culture for 23 d in a 37C, 5% CO2, humidified atmosphere.
24. On d 910, harvest the mature DC. The phenotype of mature DC is routinely checked by
flow cytometry (CD14, CD80+, CD83+, CD86+, and HLA-DR+).
3.2. Stimulation of T-Cells With Peptide-Pulsed DC

Autologous dendritic cells are incubated with peptides and used as antigen-present-
ing cells for stimulating autologous T-cells (7).
3.2.1. First T-Cell Stimulation In Vitro
1. Incubate the freshly harvested mature DC with 10 g peptide in 1 mL T-cell medium in a
15-mL conical tube. Incubating with peptide is also known as pulsing.
2. Incubate for 2 h at RT and mix gently every 20 min.
3. Following peptide pulsing, wash DC three times with RPMI, count, and resuspend DC in
T-cell medium.
4. Isolate PBMC from the same donor as the DC, either from fresh blood or from thawed
frozen PBMC. Resuspend the PBMC in T-cell medium and determine the cell concentra-
tion.
5. Co-culture autologous peptide-pulsed DC and PBMC in a 96-well round-bottom plate at
the following concentration: 5 103 DC and 1 105 PBMC per well (DC:PBMC ratio of
1:20). Culture DC and PBMC in 150 L/well T-cell medium supplemented with 2 ng/mL
IL-12 in a 37C, 5% CO2, humidified atmosphere.
6. On d 1, add 25 L T-cell medium with 5 ng/mL IL-7 (final concentration) per well.
7. On d 2 and 5, add 25 L T-cell medium with 20 U/mL IL-2 (final concentration) per well.
3.2.2. Restimulations of T-Cells In Vitro

The first restimulation is done in the same plate. For further restimulations, the cell
cultures are pooled and seeded with 2 105 cells per well in fresh 96-well round-
bottom plates. Weekly restimulations are done with a DC:PBMC ratio of 1:40.
1. Pulse fresh or thawed frozen DC with peptide, wash, and resuspend in T-cell medium
(see Subheading 3.2.1., step 1).
2. For the first restimulation, remove 100 L/well and add 5 103 DC in 50 L of T-cell
medium with 1 ng/mL IL-12 (final concentration) per well.
180 Bernhard et al.
3. For further restimulations, pool T-cells, wash with T-cell medium, and plate 2 105
T-cells plus 5 103 peptide-pulsed DC in 150 L/well; no IL-12 is added.
4. On d 1, add 25 L T-cell medium with 5 ng/mL IL-7 (final concentration) per well.
5. On d 2 and 5, add 25 L T-cell medium with 20 U/mL IL-2 (final concentration) per well.
6. HLA-peptide multimer staining may be performed 710 d after each restimulation.
3.3. HLA-Peptide Multimer Staining

3.3.1. Specificity Analysis of Stimulated T-Cells
Following repetitive T-cell stimulations with peptide-loaded DC, the frequency of
peptide-specific T-cells is measured by HLA/peptide multimer (see Note 2) staining.
1. The staining procedure has to be performed on ice in the dark.
2. Use 5 105 T-cells as a negative control and 5 105 T-cells for positive staining.
3. Bring up 1 106 T-cells in 200 L FACS buffer and then pipet 100 L/well into two
wells of a 96-well round-bottom plate.
4. Centrifuge the plate at 350g at 4C, discard the supernatant, and resuspend the cell pellets
in 50 L FACS buffer.
5. Add 1.25 L of multimer stock solution (1 mg/mL) to each cell-containing well for a 1:40
dilution, and then mix gently with the pipet.
6. Incubate the cells in the 96-well round-bottom plate for a minimum of 45 to a maximum
of 120 min at 4C.
7. Add 0.5 L of anti-CD8-FITC mAb (see Note 3) to each well for a 1:100 dilution, resus-
pend, and incubate cells for 15 min at 4C.
8. Wash the stained T-cells three times with 200 L FACS buffer per well.
9. Resuspend T-cells in 500 L FACS buffer containing 2 g/mL PI.
10. Analyze by flow cytometry.
3.3.2. HLA-Peptide Multimer-Guided T-Cell Sorting

1. The staining procedure must be performed in sterile tubes.
2. Up to 2 107 stimulated cells are stained in 500 L with a 1:20 to 1:40 dilution of multi-
mer stock solution (1 mg/mL).
3. For sorting, resuspend cells at 1 107 cells/mL in FACS buffer.
4. Sort multimer+ CD8+ T-cells with a cell sorter into a sterile tube containing T-cell medium.
3.4. T-Cell Cloning

Following the sorting, peptide-specific T-cells are cloned by limiting dilution. T-cell
clones are cultured in the presence of irradiated allogeneic LCL and PBMC (so-called
feeder cells), IL-2, and anti-CD3 mAb.
1. Harvest allogeneic LCL. Wash and bring these cells up in 20 mL LCL medium. Irradiate
LCL with 10 Gy and wash these cells twice with LCL medium; 1 105 irradiated LCL/
well are required.
2. Isolate or thaw, and then pool and wash allogeneic PBMC from three different donors.
Irradiate the pooled PBMC with 3 Gy in 20 mL of T-cell medium and then wash these
cells twice with T-cell medium; 5 104 irradiated PBMC are required per well.
3. Seed the required amount of irradiated feeder cells in a 96-well round-bottom plate in
100 L of T-cell medium supplemented with 50 U/mL IL-2 and 30 ng/mL anti-CD3
mAb per well.
4. Sorted T-cells are diluted in T-cell medium supplemented with 50 U/mL IL-2 and 30 ng/
mL anti-CD3 mAb. Add diluted T-cells to feeder cells (100 L/well). Seed T-cells in 96-
well round-bottom plates at a concentration of 0.3 and 1 T-cell per well; prepare five
plates per concentration.
5. Culture the T-cells together with the irradiated feeder cells in a 37C, 5% CO2, humidi-
fied atmosphere.
6. On d 7 and 14, remove 100 L supernatant from each well and add 100 L of fresh T-cell
medium supplemented with 100 U/mL IL-2 (50 U/mL final concentration).
7. Check for clonal T-cell growth (see Note 4) by microscopy beginning on d 12.
8. Perform screening cytotoxicity assay between d 14 and 21.
3.5. Screening Cytotoxicity Assay

Lysis of chromium51-labeled target cells by T-cells is quantified by measuring the
amount of chromium51 released to the supernatant after the co-culture. Peptide-pulsed
T2 cells can be used as target cells for analyzing lytic activity of HLA-A2-restricted
peptide-specific T-cells. Maximum lysis is calculated by adding 1% NP40 to the tar-
get cells, and minimum lysis, also known as spontaneous Cr51 release, is calculated by
measuring the amount of Cr51 released from untreated target cells.
1. Up to 5 105 T2 cells (see Note 5) are suspended in 100 L of 37C prewarmed FCS.
2. Add 100 Ci Cr51 and incubate for 90 min at 37C. Mix gently every 20 min.
3. Wash Cr51-labeled T2 cells twice with LCL medium and count the cells.
4. Incubate Cr51-labeled T2 cells in 1 mL of LCL medium containing 10 g/mL of either
relevant or irrelevant peptide. 10 g/mL 2-microglobulin is added to each tube. Incubate
for 1 h at RT. Mix gently every 20 min.
5. Meanwhile prepare effector T-cells:
a. Resuspend the CTL clones in the 96-well round-bottom culture plate.
b. Transfer 50 L of the resuspended T-cell clone from the culture plate into one well of
a fresh 96-well v-bottom plate and transfer another 50 L to a second well.
c. Refill the culture plate containing CTL clones with 100 L/well T-cell medium con-
taining IL-2 (50 U/mL final concentration). Culture at 37C (5% CO2, humidified
atmosphere) until antigen-specific T-cell clones are identified and can be expanded.
6. After peptide pulsing, wash T2 cells twice with LCL medium.
7. Resuspend peptide-pulsed Cr51-labeled T2 cells in LCL medium and then add to the T-cell
clones, which each had been divided into two wells of a 96-well v-bottom plate (see
step 5c). Add T2 cells pulsed with the relevant peptide to the first and T2 cells pulsed with
the irrelevant peptide to the second well (1 103 T2 cells in 150 L/well).
8. Add 50 L 1% NP40 buffer to each maximum control well containing 1 103 T2 cells (in
total six wells). For the spontaneous release control, add 50 L LCL medium to each
well, containing 1 103 T2 cells (in total six wells).
9. Incubate for 4 h at 37, 5% CO2.
10. Following the incubation, 100 L of the supernatant is collected from each well and trans-
ferred to tubes appropriate for the gamma counter. Cr51 released to the supernatant is
measured with a gamma counter.
11. Calculate lysis:
(experimental Cr 51 release spontaneous Cr 51 release) 100
percent specific Cr 51 release =
maximum Cr 51 release spontaneous Cr 51 release
182 Bernhard et al.
3.6. Expansion of Antigen-Specific T-Cell Clones

Antigen-specific cytotoxic T-cell clones are expanded in the presence of irradiated
allogeneic feeder cells, anti-CD3 mAb, and IL-2 (5). This approach leads to a 100- to
500-fold expansion of T-cells after 2 wk of culture.
1. The required amount of LCL are harvested and washed. LCL are transferred to a 50-mL
tube with 20 mL LCL medium and irradiated with 10 Gy. 5 106 LCL are needed per
flask.
2. Thaw, pool, and then wash PBMC (see Note 6) from three different donors and irradiate
with 3 Gy in 20 mL T-cell medium. Calculate 2.5 107 PBMC for each tissue culture
flask (25 cm2).
3. After the irradiation, wash the feeder cells twice. Count the washed cells.
4. Pool and resuspend the feeder cells in 25 mL T-cell medium supplemented with 30 ng/
mL anti-CD3 mAb per flask.
5. Harvest T-cells from one clone and add to the flask.
6. Place the culture flask lid up in the incubator (37C, 5% CO2, humidified atmosphere).
7. Add 50 U/mL IL-2 (final concentration) on d 1.
8. On d 3, 5, 8, 11, and 14, remove 15 mL supernatant without swirling the cells and add 15
mL T-cell medium supplemented with IL-2 (50 U/mL final concentration).
9. Starting on d 14, the T-cells are ready for analysis.
10. For further expansions, 12 105 T-cells are cultured with 2.5 107 irradiated PBMC and
5 106 LCL per tissue-culture flask.
11. T-cells can be frozen at 12 106/mL in freezing medium. Pelleted cells are resuspended
in freezing medium and transferred into cryovials. Immediately afterwards, the cryovials
are placed in a cryo freezing container and frozen at 80C overnight.
12. Transfer the cryovials to liquid nitrogen.
3.7. Cytotoxicity Assay of Expanded T-Cell Clones

This cytotoxicity assay is performed in order to determine whether the expanded
T-cell clones retain peptide-specific lytic activity and whether they are able to lyse
tumor cell lines presenting the endogenously processed peptide antigen. The differ-
ence between the chromium release assay for screening CTL clones (see Subhead-
ing 3.5.) and the chromium release assay for testing the expanded T-cell clones are as
followed:
1. Target cells other than T2 cells (e.g., tumor cells) are labeled for 1 h with 100 Ci Cr51.
Use 1 103 target cells per well.
2. Prepare T-cells for the following effector:target ratios by sequential dilution: 90:1, 30:1,
10:1, 3:1, 1:1, and 1:3 in a volume of 100 L LCL medium.
3. Add Cr51-labeled target cells in 100 L/well LCL medium and resuspend carefully with
T-cells.
4. Add 100 L 1% NP40 buffer to maximum control wells and 100 L LCL medium to
minimum control wells.
4. Notes
1. If immature DC are not detached on d 7, the reason may be the cytokine concentration of
IL-4. Different bioassays to determine the biological activity of cytokines are used by the
companies and, therefore, international units may not be comparable.
2. An alternative to using HLA-peptide multimers for T-cell staining and isolation is the
usage of reversible HLA-peptide multimers, so-called streptamers (8). We are currently
investigating the efficacy of reversible HLA-peptide multimers, which can be dissociated
from the T-cells and, therefore, may prevent T-cells from HLA-peptide multimer-induced
cell death. Streptamers are commercially available through the company IBA, Gttingen,
Germany.
3. Some monoclonal antibodies against CD8 may interfere with the multimer binding and,
therefore, lead to false-negative multimer staining. The herein recommended mAb anti-
CD8 (clone 3B5; Caltag Laboratories, Burlingame, CA) does not interfere with multimers.
4. T-cells do not survive when feeder cells are infected with mycoplasma. Therefore, it is
recommended to regularly check for mycoplasma infection (e.g., PCR).
5. Split or feed target cells one day before performing the chromium release assay in order
to prevent high spontaneous Cr51 release. In addition, resuspend target cells in prewarmed
FCS during Cr51 labeling to avoid non-specific cell death. High spontaneous Cr51 release
may also be a sign of mycoplasma infection of target cells.
6. PBMC can be collected by leukapheresis if a high amount of PBMC is needed.
Acknowledgments
We thank Heinke Conrad, Kerstin Gebhard, Kerstin Holtz, Evelyn Schulz, Kathrin
Hofer, Carolin Kraft, and Julia Neudorfer for excellent technical assistance. We also
acknowledge Matthias Schiemann for expert assistance with multimer-guided T-cell
sorting. We thank Wendy Batten for critical reading of the manuscript. This work has
been supported by grants from the Research Council of Germany and the GSF National
Research Center for Environment and Health-Clinical Cooperation Group Vaccino
Copy.
References
1.
1 Lee, P. P., Yee, C., Savage, P. A., et al. (1999) Characterization of circulating T cells
specific for tumor-associated antigens in melanoma patients. Nat. Med. 5, 677685.
2.
2 Yee, C., Riddell, S. R., and Greenberg, P. D. (1997) Prospects for adoptive T cell therapy.
Curr. Opin. Immunol. 9, 702708.
3.
3 Altman, J. D., Moss, P. A., Goulder, P. J., et al. (1996) Phenotypic analysis of antigen-
specific T lymphocytes. Science 274, 9496.
4.
4 Busch, D. H., Philip, I. M., Vijh, S., and Pamer, E. G. (1998) Coordinate regulation of
complex T cell populations responding to bacterial infection. Immunity 8, 353362.
5 Riddell, S. R., Watanabe, K. S., Goodrich, J. M., Li, C. R., Agha, M. E., and Greenberg, P.
5.
D. (1992) Restoration of viral immunity in immunodeficient humans by the adoptive trans-
fer of T cell clones. Science 257, 238241.
6.
6 Jonuleit, H., Kuhn, U., Mller, G., et al. (1997) Pro-inflammatory cytokines and prostag-
7.
7 Fleischer, K, Schmidt, B, Kastenmller, W., et al. (2004) Melanoma-reactive class I-re-
stricted cytotoxic T cell clones are stimulated by dendritic cells loaded with synthetic
peptides, but fail to respond to dendritic cells pulsed with melanoma-derived heat shock
proteins in vitro. J. Immunol. 172, 162169.
8 Knabel, M., Franz, T. J., Schiemann, M., et al. (2002) Reversible MHC multimer staining
8.
for functional isolation of T-cell populations and effective adoptive transfer. Nat. Med. 8,
631637.
184 Bernhard et al.
9. Valmori, D., Fonteneau, J. F., Lizana, C. M., et al. (1998) Enhanced generation of specific
tumor-reactive CTL in vitro by selected Melan-A/MART-1 immunodominant peptide ana-
logues. J. Immunol. 160, 17501758.
10. Schneider, J., Brichard, V., Boon, T., Meyer zum Bschenfelde, K. H., and Wlfel, T.
(1998) Overlapping peptides of melanocyte differentiation antigen Melan-A/MART-1 rec-
ognized by autologous cytolytic T lymphocytes in association with HLA-B45.1 and HLA-
A2.1. Int. J. Cancer 75, 451458.
Tracking Antigen-Specific T-Cells 185
14
Tracking Adoptively Transferred Antigen-Specific
T-Cells With Peptide/MHC Multimers
Norbert Meidenbauer and Andreas Mackensen
Summary
Advances in immunological monitoring provide the means to track tumor antigen-specific
cytotoxic T-lymphocytes (CTL) in humans after adoptive transfer with greater specificity and
sensitivity than before. Novel tools can be used not only to detect antigen-specific CTL, but
also to evaluate the function and phenotype of individual T-cells. Peptide major histocompat-
ibility complexes (MHC) class I multimeric complexes are proving invaluable as fluorescent
reagents for tracking of antigen-specific T-cells. The multimeric complex is constructed of four
synthetic and biotinylated peptide-loaded MHC molecules, which are linked by a fluorochrome-
labeled streptavidin molecule. In contrast to indirect assays such as limiting dilution analysis
(LDA) or 51Cr release assays requiring in vitro stimulation, this method allows direct monitor-
ing of very low numbers of peptide-specific T-cells without the need for in vitro sensitization.
By combining the use of multimers with anticytokine antibodies, a more detailed picture of the
tracked T-cell can be obtained. This chapter summarizes the application of these protocols to
monitor ex vivo the frequency and functional activity of tumor-specific CTL after adoptive
transfer. Furthermore, it details potential problems interfering with correct staining and pro-
vides guidance for interpretation of results.
Key Words: Adoptive immunotherapy; CTL; tumor immunity; antigens; cell trafficking;
MHC multimers.
1. Introduction
T-cell therapy has attracted a great deal of interest in the treatment of viral infec-
tions (1), EBV-associated lymphoma (2), and relapse of hematologic diseases after
allogeneic bone-marrow transplants. The persistence of adoptively transferred T-cells
and reconstitution of virus-specific immunity was demonstrated in the absence of
noticeable side effects (1). These data provided compelling proof of principle for the
clinical use of adoptive T-cell therapy.
Immunotherapy with autologous tumor-specific T-cells for solid tumors is much
more difficult, since specific tumor-associated antigens (TAA) have been defined in
185
186 Meidenbauer and Mackensen
only a minority of tumors. In malignant melanoma, several TAA have been recently
identified and used as targets for immunotherapy (3).
A major drawback of this treatment modality was the fact that TAA-specific T-cells
had to be selected from tumor-infiltrating lymphocytes (TIL), and often only a rela-
tively low number of cells could be produced. Recently, adoptive T-cell therapy as a
therapeutic option has regained interest by improved in vitro stimulation techniques,
allowing the generation of high numbers of antigen-specific T-cells (4). However, the
knowledge about the fate of antigen-specific T-cells after transfer into patients with
solid tumors is still very limited. Yee and colleagues reported on T-cell infiltration into
both skin and tumor tissue after adoptive transfer of a Melan-A-specific CTL clone (5).
We have demonstrated recently that in vitro generated Melan-A-specific CTL survive
intact in vivo for several weeks and localize preferentially to tumor (6).
The in vitro and in vivo monitoring and quantitation of antigen-specific T-cell
functions during the interaction with the cognate MHC/peptide ligand represent an
essential tool for detailed analysis of specific immune responses. They include the
measurement of cytokine secretion using monoclonal antibodies (mAbs) (ELISA,
intracellular cytokine staining, or ELISPOT), flow-cytometric staining of surface
molecules, and the analysis of the proliferative responses by 3H-thymidine incorpo-
ration. To determine the cytolytic activity of T-cells, the 51Cr-release assay performed
over 45 h is recognized as the standard method.
This chapter summarizes a new approach to the analysis of T-cell-mediated func-
tional activity, the MHC multimer technology. Multimers allow for direct visualiza-
tion of antigen-specific T-cells by flow cytometry (7). The multimeric complex is
usually constructed of four synthetic and biotinylated peptide-loaded HLA molecules
that are linked by a fluorochrome-labeled streptavidin molecule. The use of
multimers has a major impact on immunology. It is possible to look at T-cell func-
tions as well as the presence of the receptor. Furthermore, multimer staining may
differentiate between high- and low-affinity T-cells, thus allowing purification of
the most effective T-cells, which can be expanded in vitro and adoptively trans-
ferred to the patients. In contrast to indirect assays such as limiting dilution analy-
sis (LDA), or 51Cr-release assays requiring in vitro stimulation, this method allows
direct monitoring of very low numbers of peptide-specific T-cells without the need
for in vitro sensitization. The sensitivity of HLA class I multimers is usually 1/3000
CD8+ T-cells. To define the functional activity of antigen-specific T-cells, multimers
can be used together with antibodies targeted against surface markers or intracellu-
lar or surface-bound cytokines (8,9).
HLA class II multimers have been more difficult to produce but have already been
demonstrated to detect low frequencies of antigen-specific CD4+ T-cells (10).
In a recent study, we have used HLA class I multimers to demonstrate that Melan-
A-specific CTL generated by in vitro stimulation with peptide-pulsed dendritic cells
(DC) survive physiologically intact in vivo for several weeks (6). Combination of
multimer analysis with the cytokine secretion assay demonstrated unimpaired func-
tional activity of the transferred T-cells in the peripheral blood for at least 24 h post
infusion.
In this chapter we summarize the application of HLA class I multimers to track ex

vivo the frequency and functional activity of tumor-specific CTL after adoptive transfer.
2. Materials
2.1. Generation of Antigen-Specific CTL Lines
1. Cells I (effector cells): Peripheral blood mononuclear cells (PBMC) were freshly isolated
from heparinized peripheral blood of healthy donors or patients by Ficoll density gradient
centrifugation.
2. Cells II (stimulator cells): Monocytes were enriched from PBMC either by counter-cur-
rent elutriation or adherence.
3. Medium: Lymphocytes and monocytes were cultured in RPMI 1640 supplemented with
300 mg/L L-glutamine, 50 M -mercapthoethanol, 1 mM sodium pyruvate, 40 g/mL
streptomycin, 40 U/mL penicillin, 1X MEM vitamins, and 1X nonessential amino acids
(= standard culture medium, M').
4. CD8+ T-cell isolation kit: The CD8+ T-cell isolation kit II (Miltenyi Biotec, Bergisch
Gladbach, Germany), is an indirect magnetic labeling system for the isolation of un-
touched CD8+ CTL from peripheral blood.
5. Cytokines. The following recombinant human cytokines and proteins are used for in vitro
differentiation of DC at concentrations indicated in parentheses: interleukin (IL)-1
(10 ng/mL), IL-4 (500 U/mL), IL-6 (1000 U/mL), tumor necrosis factor (TNF)- (10 ng/
mL), transforming growth factor (TGF)- (5 ng/mL) (all from CellGenix, Freiburg, Ger-
many), granulocyte/macrophage colony-stimulating factor (GM-CSF) (500 U/mL) (Essex
Pharma and Novartis Pharma, Basel, Switzerland), prostaglandin E2 (PGE2) (1 g/mL)
(Pharmacia & Upjohn, Erlangen, Germany).
6. The following HLA-A2 binding peptides were used: modified Melan-A2635 A27L
(ELAGIGILTV) (designated as Melan-A26-35L), natural Melan-A2635 (EAAGIGILTV),
and gp100280288 (YLEPGPVTA) (Bachem Biochemica GmbH, Heidelberg, Germany).
The identity of each peptide was confirmed by mass spectral analysis. The peptides were
>98% pure as assessed by high-pressure liquid chromatography analysis. The endotoxin
level was <0.1 endotoxin units/mL.
7. Human 2-microglobulin (2m) (10 g/mL), (ICN, Eschwege, Germany).
8. T-cell growth factor (TCGF): preparation of TCGF was described previously (11). TCGF
was produced by stimulating 2.5 106/mL PBMC for 2 h with 5 g/mL phytohemagglu-
tinin (Murex, Dartford, England), 5 ng/mL phorbol myristate acetate (Sigma-Aldrich,
Taufkirchen, Germany), and 5000 rad-irradiated EBV-transformed B-cells. The cells were
then washed to remove the mitogens and resuspended in RPMI 1640 supplemented with
2.5% human AB serum. After 40 h of incubation, supernatants were harvested, passed
through 0.2 m filters, and stored at 70C.
2.2. Multimer Staining
1. Cells I: PBMC were freshly isolated from heparinized peripheral blood of healthy donors
or patients by Ficoll density gradient centrifugation. Immune monitoring after adoptive
transfer is recommended before and at least at 1, 6, and 24 h or later after T-cell infusion.
2. Cells II (control cells): Antigen-specific control T-cell lines or clones (4).
3. PE or APC-labeled HLA-A*0201 multimers that had been folded around ELAGIGILTV
(Melan-A2635L), SLLMWITQC (NY-ESDA157165), YMDGTMSQV (Tyrosinase369377),
IMDQVPFSV (gp100 209217M), and SLYNTVATL (HIVgag) (e.g., synthesized by
Beckman Coulter, Fullerton, CA).
4. Multimer washing buffer: phosphate-buffered saline (PBS) + 0.1% bovine serum albu-
min (BSA).
5. Five-milliliter polystyrene round-bottom tubes (Falcon).
6. Ninety-six-well V-bottom microtiter plates (Falcon).
7. FITC- or APC-labeled anti-CD3, anti-CD4, and anti-CD8 monoclonal antibodies (mAbs).
8. PerCP-labeled anti-CD14 mAb for monocyte exclusion.
9. FITC-labeled annexin-V (e.g., PharMingen, San Diego, CA).
10. Annexin-V buffer (PharMingen).
11. Paraformaldehyde (PFA), 0.5% (Merck, Darmstadt, Germany).
2.3. Interferon-Secretion Assay

1. Cells I (responder cells): Peripheral blood mononuclear cells (PMNC) freshly isolated
from heparinized peripheral blood of healthy donors or patients by Ficoll density gradient
centrifugation. Immune monitoring should be performed before T-cell transfer and at
24 h after T-cell infusion. Antigen-specific T-cell lines or clones (4) were used as posi-
tive controls.
2. Cells I (stimulator cells): autologous monocyte-derived DC pulsed with the appropriate
peptide (30 g/mL) and human 2m (10 g/mL) in serum-free medium, were used as
stimulators.
3. The following HLA-A2-binding peptides were used: modified Melan-A2635 A27L
(ELAGIGILTV) (designated as Melan-A2635L), natural Melan-A2635 (EAAGIGILTV),
and gp100280288 (YLEPGPVTA) (Bachem Biochemica GmbH).
4. Buffer: PBS containing 0.5% BSA and 2 mM EDTA, must be used ice cold.
5. Interferon (IFN) secretion assay detection kit APC human (Miltenyi) containing:
a. Cytokine catch reagent: anticytokine mAb conjugated to cell surface (CD45)-specific
mAb (Miltenyi).
b. Cytokine detection antibody: anticytokine mAb conjugated to PE or APC (Miltenyi).
6. Medium: any standard medium may be used, e.g., RPMI-1640 or Iscove containing 10%
AB or autologous serum. Medium is required both ice cold and at 37C. Do not use fetal
calf serum (FCS) as this gives high nonspecific background responses.
7. Fifty-mL centrifuge tubes (Corning GmbH, Wiesbaden, Germany).
8. PE- or APC-labeled HLA-A*0201 multimers that had been folded around ELAGIGILTV
(Melan-A2635L), YMDGTMSQV (Tyrosinase369377), IMDQVPFSV (gp100209217M), and
SLYNTVATL (HIVgag) (Beckman Coulter).
9. Ninety-six-well round-bottom plates (Falcon) and 48-well flat-bottom plates (Corning).
10. Staphylococcal enterotoxin B (SEB) (Sigma-Aldrich).
11. Five-milliliter polystyrene round-bottom tubes (Falcon).
12. FITC- or APC-labeled anti-CD3, ant-CD4, and anti-CD8 mAbs.
13. PerCP-labeled anti-CD14 mAb for monocyte exclusion.
14. Propidium iodide (PI) or 7-AAD to exclude dead cells from the flow-cytometric analysis.
15. PFA, 0.5 %.
3. Methods
The methods described below outline (1) the in vitro generation of antigen-specific
T-cells; (2) multimer staining for detection of antigen-specific T-cells; and (3) the
combination of the multimer technology with a functional assay, as represented by the
cytokine secretion assay.
3.1. In Vitro Generation of Antigen-Specific T-Cell Lines

Antigen-specific CTL lines were generated as described previously (4,6).
1. CD8+ T-lymphocytes are enriched from PBMC by depletion of CD4+, CD11b+, CD16+,
CD19+, CD20+, and CD56+ cells with magnetic cell sorting using a midiMACS device
(Miltenyi).
2. DC are generated from monocytes enriched by counter-current elutriation or adherence
and then cultured with M' medium plus 2% autologous serum, supplemented with human
GM-CSF, IL-4, and TGF-. On d 6, fresh medium containing GM-CSF, IL-4, TNF-,
IL-6, IL-1, and PGE2 is added to the culture for an additional 48 h.
3. The DC are then harvested and pulsed for 2 h at 37C with the appropriate peptide (30 g/
mL) and human 2m (10 g/mL) in serum-free M'-medium.
4. Co-culture 25 104 effector cells/well and 5 103 peptide-pulsed autologous DC/well
in 96-well round-bottom plates in M' medium supplemented with 10% human AB serum
and 12% TCGF (12).
5. Medium exchange is performed twice a wk; effector cells are restimulated weekly with
fresh peptide-pulsed DC.
6. After four cycles of stimulation, phenotypic and functional analysis of T-cells should be
performed.
3.2. Tracking Antigen-Specific T-Cells
Using MHC Peptide Multimer Staining
Optimal temperature and incubation time should be determined for each multimer
prior to analysis. To minimize nonspecific staining, each multimer has to be titered and
used at the lowest concentration that showed a clearly distinguishable positive popula-
tion in antigen-specific CTL lines. We recommend the use of in vitro generated anti-
gen-specific CTL lines (the protocol has been described above in Subheading 3.1.) or
antigen-specific T-cell clones (some of them are commercially available, e.g.,
ProImmune, Oxford, UK).
As with mAb staining, the intensity of multimer staining is dependent on the con-
centration of ligand in solution and not on the cell concentration or receptor density on
the cells. This is due to the fact that ligand is present in enormous excess compared to
the total number of receptor molecules. We therefore recommend conserving reagents
by staining cells suspended in a minimal volume.
3.2.1. Multimer Staining
1. Typically we stain cells in 5-mL polystyrene round-bottom tubes.
2. Add PBMC or T-cells (5 106 in a volume of 150 L) to the tube.
3. Centrifuge for 10 min at 400g (1600 rpm) and remove the supernatant.
4. Add multimer to the cell pellet and mix gently.
5. Coincubate cells with the multimer on a cell shaker. For the majority of the multimers,
incubation time should be 30 min at room temperature (RT) (see Note 1).
6. Add mAbs for surface staining and incubate cells in the dark for 30 min at 4C (see Note 2).
7. Wash twice and then fix the cells with 0.5% PFA, if not immediately analyzed.
8. For combination of multimer and annexin-V staining, wash cells once with flow buffer
and once with annexin-V buffer (PharMingen) after incubation with surface mAbs.
9. Resuspend cells in 100 L annexin buffer and add 1 L annexin-V. After incubation of
20 min at RT in the dark, fix cells with 0.5% PFA.
10. We recommend the analysis of the following mAb/multimer panels for proper multimer
staining of PBMC:
a. CD3-FITC, multimer-PE, CD14-PerCP, CD8-APC;
b. CD3-FITC, multimer-PE, CD14-PerCP, CD4-APC (control for specific staining be-
cause CD4+ T-cells should not stain with MHC class I multimers);
c. CD3-FITC, PBS, CD14-PerCP, CD8-APC (exclusion of random events in the quad-
rant of interest);
d. Annexin-V-FITC, multimer-PE, CD3-PerCP, CD8-APC;
e. CD8-FITC, control multimer-PE, CD14-PerCP, CD3-APC (control of background
staining of multimers);
f. Isotype controls.
In most common flow-cytometry applications, the phenotypic markers of interest are
expressed on 5% or more of the cells to be analyzed. In contrast, as shown in Figs. 1 and
2, multimer+ T-cells are often present at frequencies below 1%, and thus multimer stain-
ing can be considered as a rare-event problem. As in all rare-event problems, a large
denominator is required. For detection of multimer frequencies in the range of 1/1000 to
1/8000 it is advisable to routinely require >106 total events.
3.2.2. Exclusion of Cells and Other Known Sources
of Unspecific Multimer Binding and Autofluorescence
In common with any other low-frequency analysis, to properly identify positively
stained cells, positive staining (e.g., with CD3 and/or CD8) is required and exclusion
of unwanted cells from analysis is vital. Multimer-positive events are evaluated within
a Boolean gate including cells within an extended lymphoid light scatter gate and a
CD3+ gate (Fig. 1). Furthermore, CD14+ monocytes, which do bind non-specifically
with flurochrome-conjugated multimers, should be excluded (see Fig. 1).
Similarly, exclusion of the dead cells (which are present in even the most accu-
rately gated lymphocyte population) using propridium iodide (PI) or 7-AAD will elimi-
nate nonspecific background staining (data not shown).
3.2.3. Competition Between Multimer and Anti-CD3 mAb Binding
To determine the specificity of the multimer staining, we quantified the extent to
which multimer binding competed with CD3 binding on CD8+ cells, using the formula
described previously (13):
% competition =
(CD3Mnl ) (CD3MnlCD3+CD8+multimer+ )
CD3+CD8+multimer
CD3MnlCD3+CD8+multimer (1)
where CD3Mnl is the log mean fluorescence intensity (MFI) of CD3 staining of the
population designated by the subscript. The usefulness of this phenomenon was con-
firmed with multimers for several antigens (6,13) (see also Fig. 3), but multimers have
been described that did not compete for CD3 binding (14).
3.2.4. Example of In Vivo Tracking
of Antigen-Specific CTL After Adoptive Transfer
Using the multimer technology, we assessed the frequency of circulating Melan-A-
specific CTL in advanced melanoma patients during adoptive T-cell therapy (6).
Fig. 1. Compound gating strategy used to identify tetramer+ CD8+ T-cells. PBMC from an
HLA-A2.1+ melanoma patient were stained with Melan-A multimers followed by anti-CD14,
anti-CD3, and anti-CD8 mAbs using four-color flow cytometry. To obtain a clearly discernible
cluster of multimer+ cells (see upper right quadrant of D), a Boolean gate including cells within
an extended lymphoid light scatter gate (A) and a generous CD3+ gate (C) and excluding CD14+
events (B) was used.
Melan-A-specific CTL were generated from HLA-A2.1+ patients by in vitro stimula-

tion of CD8+ T-cells with DC pulsed with a mutated HLA-A2-binding Melan-A
(ELAGIGILTV) peptide. PBMC obtained immediately before and 1 h, 24 h, and 14 d
after transfer (i.e., immediately before the subsequent transfer) were stained with PE-
conjugated Melan-A-multimer and APC-conjugated anti-CD8 mAb.
The multimer staining procedure and gating strategy for elimination of non-spe-
cific multimer-binding events, described in Subheadings 3.2.1. and 3.2.2., resulted in
a clear population of specific multimer+/CD8+ T-cells (see Fig. 2).
As demonstrated in Fig. 2, the pre-infusion frequency of Melan-A-multimer+ cells
within CD8+ peripheral blood lymphocytes (PBL) in the melanoma patient was <0.1%
(6). This particular melanoma patient received an i.v. infusion of a total number of 0.5
108 Melan-A-multimer+ CTL. One h after infusion, a marked increase in the fre-
quency of Melan-A-multimer+/CD8+ T-cells up to 1.4% of total CD8 cells could be
observed (Fig. 2). Between 24 h and 14 d post infusion, the frequency of Melan-A-
TM+/CD8+ T-cells in this particular patient decreased to a level in the range of the pre-
infusion value.
192
192
Meidenbauer and Mackensen

Fig. 2. Kinetics of the frequencies of Melan-A-multimer+ cells within CD8+ peripheral blood lymphocytes (PBL) in a melanoma patient after
adoptive transfer of Melan-A-specific CTL. Peripheral blood mononuclear cells taken at the indicated time points following adoptive transfer
were stained with PE-conjugated Melan-A-multimer and APC-conjugated anti-CD8 mAb. Dot plots are shown on gated CD3+, CD14- small
lymphocytes (by forward and side scatter). Numbers in the upper right quadrants represent percentages of Melan-A-multimer+/CD8+ cells
within total CD8+ PBL.
Tracking Antigen-Specific T-Cells
193
Fig. 3. IFN- secretion-based detection of Melan-A-specific CD8+ T-cells after in vitro stimulation with Melan-A peptide-pulsed DC. Melan-
A-specific T-cells were incubated with autologous dendritic cells pulsed with the specific Melan A2635L peptide or irrelevant gp100 peptide.
After 4 h, cells were labeled as described in protocol and then directly analyzed by flow cytometry. For analysis of IFN- secretion, cells were
gated for CD3+ Melan-A-multimer+ cells (left dot plot). Please note the competition of CD3 binding in the multimer+ population compared to
CD3 binding in the multimer- T-cell population (left dot plot). The right 2 dot plots show staining with APC-conjugated anti-IFN- mAb and
PerCP-conjugated anti-CD8 mAb. Numbers in the upper right quadrants represent percentages of IFN-+/CD8+ cells within CD3+/Melan-A-
multimer+ T-cells.
193
194
3.2.5. Combination of Annexin-V and Multimer Staining

Apoptosis of transferred antigen-specific T-cells has been described as an impor-
tant mechanism for the failure of adoptive T-cell therapy in HIV disease (15). Combi-
nation of annexin-V and multimer staining represents an interesting technical approach
to answer this question. The frequency of apoptotic Melan-A-multimer+ T-cells in
circulating patients PBL was analyzed before and after 1 h and 24 h after T-cell infu-
sion. An example is shown in Fig. 4, demonstrating that the frequency of Melan-A-
specific T-cells undergoing apoptosis was not significantly elevated. Please note that
T-cells can die from apoptosis after multimer labeling (see Note 3).
3.3. Combination of Cytokine Secretion Assay
With Peptide/MHC Multimer Staining
One major disadvantage of multimer analysis is the lack of information about func-
tional activity of the T-cells. This problem can be solved by combining the multimer
technology with a functional assay, as represented by the cytokine secretion assay
(e.g., provided by Miltenyi, Bergisch Gladbach, Germany) or intracellular staining for
cytokines or other relevant functional proteins.
The combination of multimer staining with cytokine secretion was first described
by Pittet and colleagues (8). We have adopted this assay to monitor the function of
adoptively transferred Melan-A-specific T-cells in patients with malignant melanoma
(6). T-cells are stimulated with the cognate peptide and secreted IFN- molecules are
retained on the cell surface by a bispecific mAb against IFN- and CD45 (leukocyte
marker) (see Fig. 5). A fluorochrome-labeled anti-IFN- mAb allows flow-cytometric
analysis.
3.3.1. Protocol of the Combined Multimer/Cytokine Secretion Assay
1. Autologous or HLA-A2-matched DC are cultured and matured as described before.
2. Harvest DC from the culture, wash, and resuspend in complete medium (e.g., RPMI +
10% AB-serum), adjusted to 3 106/mL.
3. Add relevant and control peptide (final concentration: 30 g/mL) and 2m (final concen-
tration: 10 g /mL), and incubate at 37C for 45 min.
4. Before the end of the incubation, add responder cells (5 106 in a volume of 150 L) to
a 5-mL polystyrene round-bottom tube.
Fig. 4. (opposite page) Annexin-V staining of circulating Melan-A-multimer+/CD8+ gated

T-cells after adoptive transfer. Apoptosis of transferred antigen-specific T-cells before and
after 1 h and 24 h post T-cell infusion was detected by a combined approach of multimer and
annexin-V staining. PBMC taken from melanoma patient at the indicated time points following
adoptive transfer were stained with PE-conjugated Melan-A-multimer, APC-conjugated anti-
CD8, PerCP-conjugated anti-CD3 mAbs, and FITC-conjugated annexin-V, respectively. For
analysis of annexin-V, cells were gated for CD8+ Melan-A-multimer+ cells (left dot plot). The
right dot plots show staining with FITC-conjugated annexin-V and PerCP-conjugated anti-
CD3 mAb. Numbers in the upper right quadrants represent percentages of annexin-V+/CD3+
cells within CD8+/Melan-A-multimer+ T-cells.
Fig. 5. Principles of the combined multimer and cytokine secretion analysis for detection of
functionally active antigen-specific T-cells.
5. Wash responder cells (e.g., T-cells, PBMC) in PBS + 0.1% AB serum.

6. Centrifuge for 10 min at 400g (1600 rpm) and remove the supernatant.
7. Add multimer to the cell pellet and mix gently.
8. Co-incubate cells with the multimer on a cell shaker. For the majority of the multimers,
incubation time should be 30 min at RT (see Note 4).
9. Wash pulsed DC, count, and resuspend in complete medium.
10. Wash stained responder cells and resuspend them in complete medium into 4 aliquots:
a. Aliquot 1: Add DC pulsed with relevant peptide (ratio DC:responder cells = 1:3).
b. Aliquot 2: Add DC pulsed with irrelevant peptide (ratio DC:responder cells = 1:3).
c. Aliquot 3: Add SEB to get a final concentration of 50 g/mL.
d. Aliquot 4: nonstimulated sample. For PBMC we recommend 4 106 cells/aliquot, for
purified T-cells 2 106 cells/aliquot in order to be able to acquire about 1 106
events/sample.
11. Incubate each aliquot of 500 L in a 48-well plate for 2 h at 37C.
12. Harvest the aliquots, wash with excess of cold buffer, and add cold complete medium
(90 L/1 106 cells) and cytokine catch reagent (anti-IFN--catch-mAb [10 L/1 106
cells]).
13. Mix well and incubate cells for 5 min on ice.
14. Add 10 mL warm complete medium/1 106 cells and incubate for 30 min at 37C under
slow continuous agitation/rotation, or mix every 5 min to avoid sedimentation of the cells.
15. Wash with cold buffer in excess, add cold complete medium (90 L/1 106 cells) and
cytokine-detection antibody (e.g., anti-IFN- detection mAb [10 L/1 106 cells]) and
additional staining reagent, such as anti-CD3 and/or anti-CD8 mAbs.
16. Mix well and incubate for 10 min on ice in the dark.
17. Wash cells with cold buffer in excess, centrifuge at 300g for 5 min at RT, remove super-
natant.
18. Wash cells with cold buffer in excess, centrifuge at 300g. Resuspend cells in 500 L of
cold buffer and proceed with flow-cytometric analysis or perform fixation with PFA (1%).
3.3.2. Example for a Combination of the Cytokine Secretion

Assay With Peptide/MHC Multimer Staining
for Functional Analysis of Antigen-Specific CTL
CD8-purified T-cells from a melanoma patient were stimulated once a week with
Melan-A2635L peptide-pulsed DC. Seven days after the fourth stimulation (d 28),
T-cells were incubated with autologous DC pulsed with the specific Melan A2635L
peptide or irrelevant gp100 peptide. Additional controls included unstimulated
T-cells (negative control) or T-cells unspecifically stimulated with SEB (positive
control). As shown in Fig. 3, 37.8% of Melan-A-multimer+ T-cells secreted IFN-
upon stimulation with Melan-A-pulsed autologous DC. Upon stimulation with the
control peptide, gp100, there was no increase in the IFN- secretion compared to the
unstimulated T-cells. Of note is a down-regulation of Melan-A-multimer staining
upon specific stimulation, showing a reduction in the frequency of Melan-A-
multimer+ T-cells in the stimulated samples compared to unstimulated controls. This
phenomenon has already been described and provides indirect evidence for the speci-
ficity of the generated T-cells (6,16).
4. Notes
1. Temperature influences multimer binding: Whelan et al. examined the fine specificity of
multimer staining using a well-characterized set of HIV epitope variants (17). Peptide vari-
ants that elicit little or no functional cellular response from CTL stained these cells at 4C
but not at 37C when incorporated into multimers. These results suggest that some studies
reporting multimer incubations at 4C could detect cross-reactive populations of CTL with
minimal avidity for the multimer peptide, especially in the multimer-low population.
The data are confirmed by our studies (Fig. 6A), demonstrating that incubation with
Melan-A tetramers at 37C improves the staining intensity of specific CTLs, resulting in
improved separation of tetramer-high CD8+ cells.
However, this observation does not necessarily apply to all multimer specificities. As
shown in Fig. 6B, NY-ESO-1 multimer staining at 4C resulted in a brighter expression
and higher percentage than staining at 37C. Temperature had a minimal effect on the
binding of peptides derived from MAGE10 (18), gp100 (19), and p53 (20).
In conclusion, the optimal temperature, concentration, and incubation time should be
determined for each multimer. However, conditions that yield the highest percentage of
multimer+ events or the brightest multimer binding are not necessarily those that give the
most biologically relevant information.
2. Sequence of staining with MHC peptide multimers and mAbs is critical: Denkberg et al.
have demonstrated that T-cell receptor (TCR)-binding to multimers is coreceptor depen-
dent (19). They could show with a human CTL clone specific for a tumor-associated MHC-
peptide complex that the binding of multimers to the TCR on these cells is completely
blocked by antihuman CD8 mAbs. This blockage was mediated by anti-CD8 mAbs but not
anti-CD3 mAbs. The blocking effect of the anti-CD8 Abs was attributable to directly inhib-
iting multimer binding and was not attributable to mAb-mediated TCR-CD8 internalization
and down-regulation (19). However, inhibition of multimer binding was not observed when
cells were incubated with multimer followed by anti-CD8 mAb. There is ample evidence
that binding of the MHC class I molecule to the CD8 coreceptor is not required for avid
multimer binding, and indeed, may even increase the frequency of false-positive events.
198
198
Meidenbauer and Mackensen

Fig. 6. Temperature influence multimer binding. Melan-A+ and NY-ESO-1+ CD8+ T-cell lines were generated by repetitive antigen-specific
in vitro stimulation. The frequency of both CTL lines was determined by multimer staining. Melan-A+ and NY-ESO-1+ CTL were stained
directly with APC-conjugated anti-CD8 mAb and PE-conjugated Melan-A (A) or NY-ESO-1 (B) multimers, respectively, and incubated for 30
min either at 4C or 37C. Percentages of CD8+/multimer+ T-cells are shown in the upper right quadrant of the dot plot.
MHC class I multimers developed and marketed by Beckman-Coulter Immunomics cir-

cumvent this potential problem by using a modified MHC class I molecule carrying a muta-
tion in the 3-domain, which does not bind to the CD8 coreceptor.
We therefore recommend staining with multimer first, followed by antibodies.
3. Sequence of multimer staining and T-cell stimulation is critical: Upon stimulation with
antigen, the TCR on the antigen-specific T-cells can be rapidly down-regulated. This
dramatically reduces peptide-MHC multimer staining. Therefore, the peptide-MHC
multimer labeling should be done prior to the stimulation with peptide. After staining
with peptide/MHC multimer, cells are stimulated with the specific peptide for 23 h fol-
lowed by the standard secretion assay procedure.
Depending on the type of multimer used, the peptide-MHC multimer may or may not
stimulate cytokine secretion. If the multimer does not stimulate, then the controls may be
carried out as normali.e., stain with multimer as for the stimulated samples, then incu-
bate with no or control peptide. However, if the multimer does stimulate, an extra control
should be included, where the cells are first cultured with no peptide or control peptide,
and then incubated with multimer.
4. Reversible MHC multimer staining for functional isolation of antigen-specific T-cells:
It has been shown that T-cells die from apoptosis or are functionally defective after
multimer labeling (21). An interesting technology of reversible staining appears to leave
the T-cells intact, and could improve the quality of functional studies performed after
multimer labeling (22). The authors describe multimers constructed with a mutated
streptavidin molecule that binds a short peptide sequence with a moderate affinity (Kd
106 M) and biotin with a high affinity (Kd 1013 M). The streptavidin-peptide bonds of
the multimer can be competitively disrupted by adding free biotin.
References
1.
1 Walter, E. A., Greenberg, P. D., Gilbert, M. J., et al. (1995) Reconstitution of cellular
immunity against cytomegalovirus in recipients of allogeneic bone marrow by transfer of
T-cell clones from the donor. N. Engl. J. Med. 333, 10381044.
2.
2 Papadopoulos, E. B., Ladanyi, M., Emanuel, D., et al. (1994) Infusions of donor leuko-
cytes to treat Epstein-Barr virus-associated lymphoproliferative disorders after allogeneic
bone marrow transplantation. N. Engl. J. Med. 330, 11851191.
3.
3 Rosenberg, S. A. (2001) Progress in human tumour immunology and immunotherapy.
Nature 411, 380384.
4.
4 Oelke, M., Moehrle, U., Chen, J. L., et al. (2000) Generation and purification of CD8+
melan-A-specific cytotoxic T lymphocytes for adoptive transfer in tumor immunotherapy.
Clin. Cancer Res. 6, 19972005.
5.
5 Yee, C., Thompson, J. A., Roche, P., et al. (2000) Melanocyte destruction after antigen-
specific immunotherapy of melanoma. Direct evidence of t cell-mediated vitiligo. J. Exp.
Med. 192, 16371644.
6.
6 Meidenbauer, N., Marienhagen, J., Laumer, M., et al. (2003) Survival and tumor localiza-
tion of adoptively transferred Melan-A- specific T cells in melanoma patients. J. Immunol.
170, 21612169.
7.
7 Altman, J. D., Moss, P. A. H., Goulder, P. J. R., et al. (1996) Phenotypic analysis of
antigen-specific T lymphocytes. Science 274, 9496.
8.
8 Pittet, M. J., Zippelius, A., Speiser, D. E., et al. (2001) Ex vivo IFN-gamma secretion by
circulating CD8 T lymphocytes: implications of a novel approach for T cell monitoring in
infectious and malignant diseases. J. Immunol. 166, 76347640.
9.
9 Appay, V. and Rowland-Jones, S. L. (2002) The assessment of antigen-specific CD8+ T
cells through the combination of MHC class I tetramer and intracellular staining. J.
10.
10 Kwok, W. W., Ptacek, N. A., Liu, A. W., and Buckner, J. H. (2002) Use of class II tetram-
ers for identification of CD4+ T cells. J. Immunol. Methods 268, 7181.
11.
11 Hercend, T., Meuer, S., Brennan, A., et al. (1983) Identification of a clonally restricted 90
kD heterodimer on two human cloned natural killer cell lines. Its role in cytotoxic effector
function. J. Exp. Med. 158, 15471560.
12.
12 Mackensen, A., Carcelain, G., Viel, S., et al. (1994) Direct evidence to support the immuno-
surveillance concept in a human regressive melanoma. J. Clin. Invest 93, 13971402.
13.
13 Hoffmann, T. K., Donnenberg, V. S., Friebe-Hoffmann, U., et al. (2000) Competition of
peptide-MHC class I tetrameric complexes with anti-CD3 provides evidence for specific-
ity of peptide binding to the TCR complex. Cytometry 41, 321328.
14.
14 Meidenbauer, N., Hoffmann, T. K., and Donnenberg, A. D. (2003) Direct visualization of
antigen-specific T cells using peptide-MHC-class I tetrameric complexes. Methods 31,
160171.
15.
15 Tan, R., Xu, X., Ogg, G. S., et al. (1999) Rapid death of adoptively transferred T cells in
acquired immunodeficiency syndrome. Blood 93, 15061510.
16.
16 Nielsen, M. B., Monsurro, V., Migueles, S. A., et al. (2000) Status of activation of circu-
lating vaccine-elicited CD8+ T cells. J. Immunol. 165, 22872296.
17.
17 Whelan, J. A., Dunbar, P. R., Price, D. A., et al. (1999) Specificity of CTL interactions
with peptide-MHC class I tetrameric complexes is temperature dependent. J. Immunol.
163, 43424348.
18.
18 Dutoit, V., Rubio-Godoy, V., Doucey, M. A., et al. (2002) Functional avidity of tumor
antigen-specific CTL recognition directly correlates with the stability of MHC/peptide
multimer binding to TCR. J. Immunol. 168, 11671171.
19.
19 Denkberg, G., Cohen, C. J., and Reiter, Y. (2001) Critical role for CD8 in binding of MHC
tetramers to TCR: CD8 antibodies block specific binding of human tumor-specific MHC-
peptide tetramers to TCR. J. Immunol. 167, 270276.
20.
20 Hoffmann, T. K., Loftus, D. J., Nakano, K., et al. (2002) The ability of variant peptides to
reverse the nonresponsiveness of T lymphocytes to the wild-type sequence p53(264272)
epitope. J. Immunol. 168, 13381347.
21. Xu, X. N. and Screaton, G. R. (2002) MHC/peptide tetramer-based studies of T cell func-
tion. J. Immunol. Methods 268, 2128.
22. Knabel, M., Franz, T. J., Schiemann, M., et al. (2002) Reversible MHC multimer staining
for functional isolation of T-cell populations and effective adoptive transfer. Nat. Med. 8,
631637.
Gene Transfer of MHC-Restricted Receptors 201
15
Gene Transfer of MHC-Restricted Receptors
Helmut W. H. G. Kessels, Monika C. Wolkers,

and Ton N. M. Schumacher
Summary
Adoptive therapy with allogeneic or tumor-specific T-cells has shown substantial clinical
effects for several human tumors, but the widespread application of this strategy remains a
daunting task. The antigen specificity of T-lymphocytes is solely determined by the T-cell
receptor (TCR) and chains. Consequently, genetic transfer of TCR chains may form an
alternative and potentially appealing strategy to impose a desirable tumor-antigen specificity
onto cytotoxic or helper T-cell populations. In this strategy, autologous or donor-derived T-cell
populations are equipped with a TCR of defined reactivity in short-term ex vivo cultures, and
re-infusion of the redirected cells is used to supply T-cell reactivity against defined tumor-
specific antigens.
We have previously described the genetic introduction of T-cell receptor genes into peripheral
T-cells in mouse model systems. Here we discuss the requirements for the successful genetic
modification of murine T-lymphocytes and the subsequent use of such genetically modified cells
in in vivo models.
Key Words: T-cell receptor; gene therapy; adoptive transfer; immunotherapy; retroviral
vectors.
1. Introduction
The immune system sometimes falls short in battling diseases because T-cells that
can recognize the disease-associated antigens are reduced in number or activity. This
deficiency of antigen-specific T-cell immunity can be the result of a more general
immunodeficiency, be it acquired (e.g., AIDS, chemotherapy), or congenital. Alterna-
tively, the absence of disease-specific T-cell reactivity can be due to a lack of immu-
nogenicity of the affected cell type, as is the case for most nonviral tumors. In cases
where the disease-specific T-cells are absent or reduced in number/activity, therapies
that introduce the desired T-cell reactivity through adoptive transfer provide a concep-
tually attractive approach. The adoptive transfer of T-lymphocytes that contain T-cell
receptors (TCR) that recognize tumor-specific targets has provided encouraging
results for the treatment of melanoma (1,2). Likewise, the antileukemic effect of HLA-
201
202 Kessels, Wolkers, and Schumacher
matched allogeneic bone-marrow transplantation is based on the transfer of novel

T-cell reactivities, in this case specific for recipient minor histocompatibility antigens.
The main drawback of adoptive therapy with defined T-cell populations is that it is
difficult to produce sufficient quantities of antigen-specific T-cells for subsequent
infusion. As an alternative to the adoptive transfer of T-cell immunity at the cellular
level, T-cell immunity can also be transferred at the level of the T-cell receptor (3,4).
In this strategy, autologous or donor-derived T-cell populations are equipped with a
TCR of defined specificity in short-term ex vivo cultures, and re-infusion of the redi-
rected cells is used to supply T-cell reactivity against defined antigens. TCR gene
therapy is currently developed in a number of preclinical models, and clinical trials for
TCR gene therapy will likely commence within several years. Here we discuss the
current technology for the genetic modification of T-lymphocytes with exogenous
TCR and the in vitro and in vivo analysis of the resulting TCR-modified cell popula-
tions in murine preclinical models.
2. Materials
2.1. Production of Retrovirus
1. Phoenix-Eco packaging cell line.
2. Tissue-culture treated 3003 Petri dish (Falcon, BD).
3. FuGENE 6 transfection reagent (Roche Molecular Biochemicals).
4. pMX retroviral vector containing both the TCR and TCR coding sequences, separated
by an internal ribosomal entry site (IRES) (pMX-TCR-IRES-TCR).
5. Iscoves modified Dulbeccos medium (Life Technologies BV, Scotland) supplemented
with 5% fetal calf serum (FCS; BioWhittaker, Belgium), 0.5 M -mercaptoethanol
(Merck, Germany), penicillin (100 U/mL), and streptomycin (100 g/mL) (Boehringer
Mannheim, Germany).
6. Saponin (Merck).
7. pCLEco (Imgenex).
2.2. Retroviral Infection of Mouse T-Lymphocytes

1. Mouse splenocytes.
2. Concanavalin A (Calbiochem).
3. Interleukin (IL)-7 (Santa Cruz).
4. Twenty-four-well cell-culture plates, flat bottom, tissue-culture treated (Costar).
5. Twenty-four-well cell-culture plates, flat bottom, non-tissue-culture treated (BD).
6. Recombinant human fibronectin fragment CH-296 (RetroNectin; Takara, Otsu, Japan).
7. Bovine serum albumin (BSA) (Sigma).
8. Phosphate-buffered saline (PBS).
9. RPMI 1640 medium (Life Technologies BV, Scotland) supplemented with 8% FCS (Bio-
Whittaker), penicillin (100 U/mL), and streptomycin (100 g/mL) (Boehringer Mann-
heim).
10. Erylysis buffer: 155 mM NH4Cl, 10 mM KHCO3, 0.1 mM EDTA, pH 7.4.
2.3. Adoptive Transfer and Tracing of Genetically Modified Cells

1. Recipient mice.
2. Hanks buffered salt solution (HBSS, Gibco).
3. Allophycocyanin (APC)-labeled major histocompatibility complex (MHC) tetramers.
4. FITC- or PE-conjugated anti-TCRV chain and anti-TCRV chain mAbs, PE- or APC-
conjugated anti-CD8b.2 mAb, PE- or Cy5-conjugated anti-CD4 mAb, FITC-conjugated
anti-interferon (IFN)- mAb (all mAbs from PharMingen), PE- or Cy5-conjugated F4/80
mAb (Serotec).
5. Propidium iodide (Sigma).
6. rhIL-2 (Chiron).
7. Cytofix/Cytoperm plus kit (PharMingen).
8. AutoMACS or varioMACS (Miltenyi) plus LS/LD columns, anti-PE micro beads
(Miltenyi) (optional).
9. MACS buffer: PBS, 0.5% BSA, 2mM EDTA (optional).
10. PBA: 1X PBS, 0.5% BSA, and 0.02% sodium azide.
11. Ficoll (Merck).
12. FacsCalibur (Becton Dickinson, MountainView, CA), CellQuest software.
13. Antigenic stimulus (e.g., virus, tumor cells, vaccine).
3. Methods
The methods described below outline (1) the production of retroviruses that contain
RNA that encodes the desired TCR chains; (2) the procedure to infect murine T-cells
with TCR-encoding retroviruses and the in vitro analysis of the resulting cells; and (3)
the transfer of transduced T-cells to recipient mice and the subsequent detection and
functional characterization of TCR-transduced T-cells in these mice.
3.1. Production of Retrovirus
The production of retroviruses that, upon infection, provide T-cells with a new set
of TCR genes is described and includes (1) a description of the retroviral vector and
the conditions such a vector should meet for TCR gene transfer; and (2) a description
of the procedure to generate retrovirus by transfection of producer cell lines.
3.1.1. Retroviral Plasmids
The pMX expression system is a Moloney leukemia virus (MLV)-based retroviral
expression system developed by Kitamura (5). A sequence encoding both the TCR
and the TCR chain separated by an IRES is inserted into the pMX vector between the
viral long terminal repeats (LTRs). In this setting, expression of the transgene is driven
by the MLV-LTR promoter. All current TCR transfer strategies make use of retroviral
vector systems for gene transfer because retroviral integration into the host genome
allows for the stable transmission of the transgene through cell division (see Note 1).
This ensures that the transgene will remain resident in T-cells undergoing extensive
proliferation upon activation. The pMX retroviral vector is a minimal retroviral expres-
sion vector that does not encode an antibiotic resistance marker or other markers to
detect genetically modified cells, in order to minimize the immunogenicity of the
retrovirally modified T-cells in in vivo experiments (see Note 2).
3.1.2. Transfection of Producer Cells
For the generation of retroviruses, we transiently transfect the retroviral packaging
cell line Phoenix, created by the Nolan laboratory (see Website: http://www.stanford.
edu/group/nolan/). This helper-defective packaging line is capable of producing all
the necessary trans proteins (gag, pol, and env) that are required for packaging, pro-
cessing, reverse transcription, and integration of recombinant genomes. For the pro-
duction of retroviruses, either the Phoenix-Ampho or Phoenix-Eco cell line can be
used. The retroviral particles that are produced in these systems carry different enve-
lope proteins (amphotropic vs ecotropic) and thereby determine the tropism of the
resulting recombinant retrovirus. The ecotropic packaging system is capable of deliv-
ering genes to dividing murine or rat cells, whereas the amphotropic system is capable
of delivering genes to dividing cells of most mammalian species, including human
cells. Pseudotyping with other envelope proteins is possible and can lead to enhanced
transduction efficiency, depending on the cell type that is targeted (see Note 3).
In our TCR gene-transfer experiments, where viruses are produced to infect mouse
splenocytes, the Phoenix-Eco cell line is transiently transfected with the pMX-TCR-
IRES-TCR using the FuGENE 6 transfection reagent according to the manufacturers
protocol (FuGENE is a multi-component lipid-based transfection reagent that com-
plexes with DNA and mediates transport of DNA into the cell during transfection). In
short:
1. Twenty-five microliters FuGENE reagent is added to 800 L serum-free Iscoves medium.
2. After a 5-min incubation at room temperature, 20 g purified retroviral plasmid DNA is
added. At this stage, 7.5 g of pCLEco DNA can be added for improved virus titers (6).
3. After a 15-min incubation at room temperature, the FuGENE/DNA suspension is added
to a culture dish (10 cm diameter) containing Phoenix cells grown to approx 30%
confluency in 10 mL Iscoves medium.
4. Twenty-four h after transfection, the medium of the Phoenix culture is replaced with 7 to
12 mL Iscoves medium, depending on the amount of virus needed.
5. Another 24 h later, at the peak of virus production, the retroviral supernatant is collected
and spun (425g, 10 min) to remove all producer cells.
The retroviral supernatant is directly used for the transduction procedure. The super-
natants can also be stored at 70C, although virus titers may drop as a result of the
freeze-thaw cycle.
The efficiency of transfection of producer cells is often determined by flow cyto-
metry in cases where proteins are generated that are expressed at the cell surface, or in
case a fluorescent marker gene is used. However, TCR gene products are not trans-
ported to the cell surface in virus producer cell lines due to the absence of CD3 compo-
nents, and we therefore use an alternative assay to monitor transfection efficiency
(Calogero et al., unpublished). In this protocol, TCR or TCR chain expression is
determined by intracellular antibody staining. The transfected Phoenix cells are col-
lected and permeabilized with 1 mg/mL saponin, which permeabilizes both plasma
and internal membranes. The cells are washed and stained with a fluorochrome-
labeled TCR-specific antibody, and subsequently analyzed by flow cytometry. Trans-
fection efficiencies can vary between 10 and 100% of Phoenix cells, depending on
both the fitness of the Phoenix cell culture and purity of the DNA. High-level trans-
duction of mouse T-cells is generally achieved only with retroviral supernatant of
Phoenix cells transfected with greater than 40% efficiency. It is noted that depending
on the experimental setup it can be desirable to generate stable transfectants of pro-
ducer cell lines (see Note 4).
3.2. Retroviral Infection of Mouse T-Lymphocytes

Retroviruses can be used to stably transduce mouse T-cells with a desired TCR.
The procedure to infect mouse T-cells with such retroviruses, and methods to test
transgene expression are described under Subheadings 3.2.2. and 3.2.3.
3.2.1. Generation of Spleen Cell Cultures
Moloney-based retroviruses can integrate into the genome only of dividing cells
(see Note 5), and in line with this, the transduction efficiency of resting T-cells is
negligible. To obtain actively dividing T-cells, mouse splenocytes are stimulated with
concanavalin A (ConA) and IL-7 2 d prior to transduction. ConA is a lectin that acti-
vates T-cells by cross-linking membrane glycoproteins, and IL-7 is added to improve
the survival of T-cells. For the isolation of splenocytes, a freshly removed spleen is
dispersed by pressing the spleen through a cell strainer using a plunger. Erythrocytes
are removed by resuspending the spleen cells in 3 mL erylysis buffer and incubating
for 3 min on ice. The splenocytes are washed and resuspended in RPMI 1640 medium
containing 2 g/mL concanavalin A (ConA) and 1 ng/mL IL-7, and are cultured at
37C for 48 h in tissue culture-treated 24-well plates at 3 106 cells per well. Total
splenocytes rather than purified T-cells are used for ConA/IL-7 stimulation, as T-cell
activation is dependent on the presence of non-T accessory cells. In addition, the pres-
ence of activated CD4+ cells appears to benefit the survival of CD8+ T-cells upon
ConA activation. In experimental setups where it is desirable to use selected CD8+
T-cell populations, such cell populations should be isolated subsequent to the trans-
duction procedure (see Subheading 3.3.1.). In contrast, the depletion of CD8+ T-cells
before transduction does not affect the survival or transduction efficiency of CD4+
T-cells.
3.2.2. Infection of Spleen Cell Cultures
1. Non-tissue-culture-treated 24-well plates are coated with 0.5 mL of 50 g/mL recombi-
nant human fibronectin fragment CH-296 at room temperature for 2 h.
2. The CH-296 solution is removed and replaced with 1 mL 2% bovine serum albumin in
PBS for 30 min at room temperature, and plates are washed with PBS.
3. Two d after ConA stimulation, the splenocytes are plated on CH-296-coated plates (3
106 cells/well) in 0.51 mL of retroviral supernatant. The plates are subsequently centri-
fuged for 90 min at 425g (spin infection), after which the cells are cultured at 37C.
3.2.3. In Vitro Analysis of Transgene Expression

Twenty-four to forty-eight h after transduction, transgene expression can be moni-
tored. Maximal transgene expression is observed after 48 h, but analysis at an earlier
time point may be preferred, as prolonged in vitro culture adversely affects the subse-
quent in vivo survival of modified T-cells (7).
To detect whether T-cells express the introduced TCR, the cells can be stained with
antibodies that are specific for the TCRV and TCRV subfamily used by the intro-
duced TCR, and analyzed by flow cytometry. This approach is not feasible for all
T-cell receptors because the required V- and V-specific antibodies may be lacking.
An additional limitation of this approach is that even when expression of both the TCR
and chain can be detected, this does not provide definitive proof for expression of
the correct TCR heterodimer, as it is (at least formally) possible that the observed
staining reflects assembly of the exogenous TCR chains with endogenous TCR sub-
units (see Note 6). Transduced T-cells are preferably stained with soluble MHC-tet-
ramers (8) in combination with CD8- or CD4-specific antibodies to provide direct
proof for transgenic TCR expression. MHC class I tetramer staining is performed in
PBA at room temperature for 15 min. MHC class II tetramer staining is performed in
culture medium at 37C for 3.03.5 h. Staining with (CD4- or CD8-specific) antibod-
ies is performed during the last 20 min of tetramer staining. In case of MHC class II
tetramer staining, cells are also stained with the macrophage marker F4/80 specific
mAbs to reduce background staining (9). Propidium iodide (1 g/mL) is included prior
to analysis to be able to exclude dead cells. In our hands, transduction efficiencies for
both CD4+ and CD8+ T-cells as judged by TCR expression can vary between 5 and
40% of total T-cells. This likely forms an underestimate of the true frequency of modi-
fied T-cells, as some of the resulting T-cells express the introduced TCR at a level that
is below the threshold for detection (see Note 7). Of the currently available systems
for retroviral transfer, the typing of retroviral particles with the ecotropic receptor
results in the highest transduction efficiency for mouse T-cells. Transduction of
human T-cells (with a GFP marker gene) has been reported to be most efficient with
the use of the MLV 10A1 receptor (10), although in our hands transduction frequen-
cies up to 60% have also been reached with the Phoenix-Ampho system (Jorritsma
et al., unpublished observations).
3.2.4. In Vitro Analysis of Transgene Function

To test whether antigen recognition by the introduced TCR results in successful
signaling and subsequent activation of T-cells, expression of cell-surface markers
associated with T-cell activation (e.g., CD69, CD40L) can be determined by cell-
surface staining. Alternatively, antigen-induced cytokine production (e.g., IFN-,
IL-2, tumor necrosis factor [TNF]-) or production of cytotoxic granule components
(e.g., granzyme B) can be determined by intracellular staining. Because ConA-stimu-
lated T-cells have a very limited lifespan in vitro (see Note 9), the in vitro analysis of
transduced T-cells is preferably performed within 13 d after transduction.
Protocol for intracellular cytokine staining of CD8+ T-cells:
1. Incubate transduced cells for 5 h at 37C in the presence of rhIL-2 (10 CU/mL), Brefeldin-
A (1 L/mL), and (different concentrations of) antigenic peptide.
2. After incubation, cells are surface stained with PE-conjugated anti-CD8b.2 mAb for 15
min on ice, washed, and incubated in Cytofix/Cytoperm solution for 20 min on ice.
3. After washing, the cells are stained with FITC-conjugated anticytokine mAb on ice for
20 min.
4. Cells are washed and resuspended in PBA, and analyzed by flow cytometry.
Because in most cases only a fraction of antigen-specific T-cells produce detect-
able levels of a given cytokine upon antigen encounter, the percentage of cells that is
positively stained for this cytokine will provide an underestimate of the total percent-
age of transduced T-cells.
3.3. Adoptive Transfer and Tracing of Genetically Modified Cells

The functional capacity of redirected T-cells can be analyzed in vivo in mouse
models of viral infection and tumor growth. The monitoring of redirected T-cell func-
tion in such models is described in Subheadings 3.3.2. and 3.3.3.
3.3.1. Infusion of Genetically Modified Cells

T-cell transfer into mice combined with subsequent antigenic stimulation allows
the examination of the in vivo behavior of redirected T-cells. Since a large fraction of
spleen cells do not survive the procedure of retroviral transduction, dead cells are
removed by a Ficoll step before infusion into mice, 13 d after transduction. Viable
transduced spleen cells can be directly used for infusion, but in cases where purified
CD4+ or CD8+ T-cell populations are preferred, CD4+ or CD8+ T-cells can either be
isolated or depleted from the splenocyte population prior to adoptive transfer by mag-
netic cell separation (MACS).
To this purpose:
1. Stain the splenocyte population with PE-conjugated (CD4- or CD8-specific) antibodies
following the Ficoll step for 15 min on ice in culture medium.
2. Cells are washed twice and anti-PE microbeads are added in culture medium according to
the manufacturers protocol (Miltenyi; anti-FITC and anti-APC MACS beads are also
available).
3. Cells are washed twice in culture medium, resuspended in MACS-buffer, and loaded on a
positive or negative selection column (LS or LD columns, respectively).
4. The desired cell fraction is collected and washed twice with HBSS.
5. The cells are resuspended in a small volume of HBSS, such that each mouse receives the
desired amount of transduced cells in an injection volume of 200 L.
6. The cell suspensions are transported on ice, and administered immediately by intrave-
nous injection. Prior to injection, the recipient mice are fixated and exposed to ionizing
radiation (IR) light for approx 3 min (time frame depending on the mouse strain), until
dilation of the tail veins is visible (note: long exposure to IR light will cause dehydration
of mice).
The number of redirected T-cells that needs to be infused to induce a significant
immune response depends both on the quality of the introduced TCR and the antigenic
challenge administered, but in most cases very small numbers of cells will suffice. For
the influenza A/NT/60/68-specific F5 TCR, and for the chicken ovalbumin-specific
OT-I TCR, infusion of only 103 to 105 TCR-transduced CD8+ T-cells yields a large
population of redirected cells (330% of the total CD8+ T-cell population) upon influ-
enza A infection, or upon tumor challenge with a transplantable tumor cell line that
expresses the corresponding T-cell epitope.
3.3.2. Tracing of Genetically Modified Cells

T-cell maintenance or expansion can be determined either by flow-cytometry analy-
sis of peripheral blood cells, which allows the longitudinal analysis of T-cell popula-
tions within individual mice, or by analysis of lymphoid or peripheral organs at specific
time points. To collect peripheral blood T-cells for analysis:
1. Harvest approx 50 L blood in 1 mL PBS/heparin solution.

2. Cells are spun, resuspended in 1 mL erylysis buffer, and incubated for 15 min on ice to
remove erythrocytes.
3. After spinning for 5 min at 153g, the supernatant (containing the erythrocytes) is removed,
and the cells are resuspended in PBA.
4. The staining procedure is performed as described below or in Subheading 3.2.3.
Lymphocytes may also be isolated from peripheral organs such as spleen, secondary
lymphoid organs, lung, and liver. To prepare cell suspensions from spleen or secondary
lymphoid organs, organs are dispersed by pressing through a cell strainer as described
under Subheading 3.2.1. When necessary, erythrocytes are removed from organs by
incubation in erylysis buffer for 23 min on ice. Isolation of lymphocytes from other
organs is described elsewhere (11). Redirected T-cell responses in peripheral blood or
solid organs can be followed through the use of donor cells that express an allotype
marker that differs from the recipient allotype, or by using retroviral vectors that carry a
selection marker or reporter molecule (see Note 2). However, both allotype markers and
retroviral markers can be regarded as non-self by the host immune system. Although
the magnitude of the immune response induced by different markers varies (1215),
immune responses directed against the redirected T-cells can result in a decreased effi-
ciency of TCR gene transfer, especially when long-term reactivity is required.
To circumvent possible host rejection of redirected T-cells, a setting is preferred
where the transduced cells are congenic (or autologous) and where retroviral marker
genes are absent. V/V or MHC tetramer staining may be used in such situations (see
Subheading 3.2.3.). However, these techniques do not distinguish endogenous and
exogenous T-cell responses. As an alternative strategy, TCR-transduced T-cells can be
traced by monitoring dual V expression. In contrast to the sizable population of
T-cells that expresses two V chains, the vast majority (approx 99%) of human and
mouse T-cells express only one V chain (1618). Consequently, responses of TCR-
modified T-cells can be followed by flow-cytometric analysis of T-cells that express
the V element used by the transgenic TCR in addition to one of the other V elements.
Lymphocytes are stained for 15 min at room temperature with a mixture of all available
FITC-conjugated V chains except the one that is used by the donor TCR (this mixture
is referred to as V pool). Co-staining is performed with a PE-conjugated antibody
specific for the V element used by the transgenic TCR, and APC-labeled anti-CD4 or
anti-CD8 lineage markers. Before analysis by flow cytometry, PI is added for exclusion
of dead cells from analysis. The percentage of TCR-transduced T-cells is determined
by calculating the fraction of V pool+ T-cells that is also positive for the V element
of the transgenic TCR.
3.3.3. In Vivo Function of Genetically Modified Cells
Analysis of the in vivo function of genetically modified cells focuses on two sepa-
rate issues: (1) The capacity of the modified cells to expand and home to the site of
inflammation; and (2) the capacity of the modified cells to contribute to clearance of
antigen-expressing cells.
Viral infection (e.g., influenza A infection) is a good means to analyze the basic
function of redirected T-cells because it provides a strong antigenic stimulus. For TCR
with specificities other than influenza antigens, recombinant influenza strains can be
generated that contain the desired T-cell epitope (19). Influenza A infection is carried
out intranasally. Upon anesthesia of the mice with diethylether, the viral mixture is
dropped onto the nose in a volume of 50 L HBSS. The mouse takes up the viral fluid
while inhaling. The influenza infection can be performed the same d as the T-cell
transfer, but may also be performed 13 d later.
Expansion of modified T-cells upon encounter with the antigen of interest can be
followed as described in Subheading 3.3.2. The T-cell response may peak earlier than
the endogenous T-cell response, either because the transferred T-cells have been acti-
vated prior to adoptive transfer or because of the increase in precursor frequency.
The effectiveness of transferred T-cells in clearing antigen-expressing cells may in
theory be determined by quantifying the virus titer in the infected organ (i.e., in lung
for the influenza A model). However, such assays are useful only in viral infection
models where the contribution of T-cell immunity in viral clearance is significant,
and in wild-type mice, immune control of influenza A infection is largely due to the
humoral immune response. Consequently, control of influenza A infection by redi-
rected T-cells can be assessed only in mice that lack the endogenous B-cell compart-
ment, such as RAG-1-deficient mice (20).
T-cell dependent models for tumor rejection (or autoimmunity) are better suited to
analyze the contribution of redirected T-cells in clearing the antigen-expressing cell
population. For transplantable tumor models (see Note 10), the effectiveness of T-cell
therapy can be determined either in a prophylactic setting by injecting T-cells prior to
tumor injection, or in a curative setting that involves T-cell transfer several days after
tumor inoculation, when the tumor is measurable. Prior to inoculation, tumor cells are
washed twice with HBSS and are injected subcutaneously, intravenously, or intraperi-
toneally, depending on the tumor model used. The growth of subcutaneous tumors can
easily be followed by conventional caliper measurements. In order to monitor tumor
growth in intraperitoneal or intravenous model systems, tumor cells may be modified
by the introduction of a luciferase gene. Intraperitoneal injection of the luciferase sub-
strate luciferin can then be used to monitor tumor growth in internal organs in a non-
invasive fashion with a CCD camera (Jorritsma et al., unpublished).
4. Notes
1. Dr. Alain Fischer and coworkers have successfully used retroviral-based gene transfer for
the treatment of children that suffer from severe combined immunodeficiency (SCID).
SCID-X1 disorder is caused by mutations in a single gene, coding for the cytokine recep-
tor common gamma ( c) chain. Restoration of gamma c expression in stem cells of these
patients by gene transfer led to a near complete immune reconstitution. Unfortunately, the
initial success of this gene therapy trial has been clouded by the occurrence of leukemic
disorders in two of the treated patients (21). Although still under investigation, insertional
mutagenesis by the retroviral transgenes is likely to have been one of the critical factors in
the observed cellular transformation. Whether the incidence of cellular transformation is
dependent on the type of transgene used or/and the number of divisions that the trans-
duced cells undergo will require further evaluation. Nonetheless, these serious side effects
are likely to at least temporarily limit the use of retrovirus-mediated human gene therapy
trials to life-threatening conditions where alternative treatment options are lacking. In

addition, the adverse events in the SCID-X1 trial provide a strong incentive for the contin-
ued development of gene-transfer systems in which the chances of cellular transformation
due to insertional mutagenesis are reduced. Current candidates form systems for site-spe-
cific integration and gene-transfer systems that are maintained episomally.
2. To distinguish donor T-cells from the host T-cell population, extracellular allotype mark-
ers such as Ly5, Ly1, and Thy1, or intracellular biochemical markers such as glucose-
phosphate-isomerase (Gpi) can be used. Ly5 (CD45) is often used as surface marker in
cell transfer experiments, as all bone-marrow-derived cells (except erythrocytes and eryth-
roblasts) express this molecule. The transferred donor cells (e.g., Ly5.1) can be distin-
guished from the host cells (e.g., Ly5.2) by staining tissue samples from the mice ex vivo
with Ly5.1- or Ly5.2-specific antibodies. However, minor immunological disparities
between donor and host cells can affect cell survival when allotype markers such as Ly5
are used in transplantation studies (22). Alternatively, autologous T-cells may be equipped
with an intracellular reporter molecule (e.g., GFP) or an antibiotic resistance marker (e.g.,
neomycin, hygromycin) through the inclusion of marker genes in the retroviral vector.
However, expression of such intracellular markers in gene-modified T-cells has likewise
been shown to make the transferred cells immunogenic to the host T-cell immune system
(1214).
3. A large series of alternative envelopes has been developed in particular for the transduc-
tion of human cells. A Phoenix cell line called Phoenix-gp is available that expresses only
gag-pol and that can be used for pseudotyping of retroviral virions by co-transfection of
the gene encoding one of the alternative envelope proteins (e.g., gibbon ape leukemia
virus envelope or vesicular stomatitis VSV-G protein). Producer cell lines that already
carry one of the alternative envelopes are also available from a number of commercial
and non-commercial sources. A recent study demonstrated that transduction of primary
human CD8+ and CD4+ T-cells occurs with very high efficiency with retroviral vectors
pseudotyped with the envelope of the amphotropic murine leukemia virus 10A1 (MLV-
10A1) (10,23). The superior efficiency of MLV-10A1 seen in this study could be corre-
lated to the longer half-life of this pseudotype in comparison to A-MLV, and to the usage
of both the A-MLV (Pit2) and the GaLV receptor (Pitl) for cell entry.
4. To facilitate the production of viral supernatants, and to increase viral titers, stable
transfectants of producer cell lines can be generated by including a selection marker in
the retroviral vector. Such markers may be included either between the viral LTRs or in
the vector backbone. Only in the former case will the marker gene also be present in the
retroviral genome and therefore be expressed in retrovirally transduced cells (see Note 2).
Alternatively, the producer cell line can be co-transfected with a separate vector that only
contains the selection marker. In this case, the vector that contains the marker gene should
be present at substoichiometric amounts during transfection (10- to 20-fold lower) com-
pared to the vector that contains the TCR, in order to increase the chance that the selected
producer cells also contain the TCR construct of interest. The nuclear replication and
retention functions of the Epstein-Barr virus have been used to maintain retroviral vectors
episomally within retroviral packaging cell lines (24). These hybrid EBV/retroviral vec-
tors carry a selection marker and can be used to rapidly generate high-titered retroviral
stocks. A disadvantage of these hybrid vector systems is that due to their size, recombina-
tion of vector sequences is sometimes observed during growth in E. coli.
5. The T-cell activation and proliferation required for retroviral transduction with Moloney-
based vector systems may affect the viability and possibly immune competence of the
transduced cells. HIV-derived, lentiviral vectors have been put forward as an alternative,
as these retroviral systems effectively infect resting T-cells. It is noted that also in this
case stimulation of nave T-cells with cytokines is required to obtain high transduction
efficiencies (25,26).
6. The introduction of TCR chains into mature T-cells will result in pairing of introduced
chains with endogenous TCR chains. In cases where these mixed TCR dimers will be
formed preferentially over the desired dimer of both exogenous chains, the transduced
cells will express both TCR and TCR on their surface, but may not possess reactivity
towards the antigen of interest.
7. In the Moloney-based retroviral transduction system used by us, the surface expression
level of the introduced TCR is about three- to fivefold lower than that of the endogenous
TCR. This lower TCR expression may reduce the sensitivity of the resulting T-cells,
although the magnitude of this effect is likely to be modest (27). In addition, the lower
expression level of the introduced TCR can in some cases complicate efforts to detect the
TCR-modified cells. When necessary, the expression level of introduced TCR may be
improved by using constructs that have an increased promoter activity or by enhancing
transgene expression at the post-transcriptional level (28). Surface expression of the de-
sired TCR heterodimer is likely to also be reduced as a consequence of the formation of
mixed TCR dimers between endogenous and exogenous TCR chains (29), an issue that
remains unresolved to date (see Note 8).
8. Pairing of introduced and endogenous TCR chains will lead to the formation of mixed
TCR dimers with unanticipated specificities (29), and may possibly lead to the generation
of auto-reactive T-cells. Although no incidences of auto-immunity upon TCR gene trans-
fer in a large number of mouse experiments have been observed to date (Wolkers et al.,
unpublished), the risk of autoimmunity due to inadvertent pairing is clearly TCR depen-
dent and should be re-evaluated for every single TCR used. Efforts to remodel the inter-
face of TCR - interaction in such a way that introduced TCR chains preferentially pair
with each other, are currently being undertaken and may offer a solution to this problem.
In addition, several groups have developed chimeric receptors for gene transfer into
T-cells. These receptors consist of a single-chain antibody that is coupled to signaling
modules that are normally present in the T-cell receptor (TCR) complex, or are composed
of recombinant (e.g., single chain) T-cell receptors (30). The rationale for the production
of these receptors is that they can confer antigen-recognition potential to T-cells that is
not complicated by the formation of mixed TCR dimers, and in the case of antibody-
based receptors, can be MHC independent. While chimeric receptors form an appealing
concept, the available data suggest that the current-generation chimeric receptor systems
do not function as well as conventional T-cell receptors in redirecting T-cell immunity.
9. ConA-stimulated T-cells are not all that happy. Three d after ConA/IL-7 stimulation (1 d
after transduction), approx 3050% of splenocytes has died, while 6 d after stimulation
(3 d after transduction), up to 90% of the cells are no longer viable. Both in vitro and in
vivo studies with retrovirally transduced murine T-cells are therefore preferably per-
formed shortly after the transduction procedure. This inconvenience is mouse-related;
human T-cells have been shown to be much more robust and can be cultured for pro-
longed periods in vitro after being activated and transduced.
10. The tumor model systems used to date in TCR gene-transfer studies (and described here)
rely on transplantable tumor cell lines, but it is clear that the concepts that will emerge
from these studies will require further validation in spontaneous tumor models that much
more closely mimic the human situation.
Acknowledgments
Work in the authors laboratory is supported by NWO pionier grant 00-03 and the
Dutch Cancer Society grants NKB 2001-2419 and NKB 2003-2860. We would like to
thank the members of the Schumacher group and in particular Marly van den Boom,
Annelies Jorritsma and Anna Calogero for their comments and suggestions.
References
1.
1 Dudley, M. E., Wunderlich, J. R., Robbins, P. F., et al. (2002) Cancer regression and
autoimmunity in patients after clonal repopulation with antitumor lymphocytes. Science
298, 850854.
2.
2 Yee, C., Thompson, J. A., Byrd, D., et al. (2002) Adoptive T cell therapy using antigen-
specific CD8+ T cell clones for the treatment of patients with metastatic melanoma: in vivo
persistence, migration, and antitumor effect of transferred T cells. Proc. Natl. Acad. Sci.
USA 99, 16,16816,173.
3.
3 Dembic, Z., Haas, W., Weiss, S., et al. (1986) Transfer of specificity by murine alpha and
beta T-cell receptor genes. Nature 320, 232238.
4. Kessels, H. W., Wolkers, M. C., and Schumacher, T. N. (2002) Adoptive transfer of T-cell
immunity. Trends Immunol. 23, 5, 264269.
5.
5 Kitamura, T. (1998) New experimental approaches in retrovirus-mediated expression
screening. Int. J. Hematol. 67, 351359.
6.
6 Naviaux, R. K., Costanzi, E., Haas, M., and Verma, I. M. (1996) The pCL vector sys-
tem: rapid production of helper-free, high-titer, recombinant retroviruses. J. Virol. 70,
57015705.
7.
7 Kolen, S., Dolstra, H., van de Locht, L., et al. (2002) Biodistribution and retention time of
retrovirally labeled T lymphocytes in mice is strongly influenced by the culture period
before infusion. J. Immunother. 25, 385395.
8.
8 Altman, J. D., Moss, P. A., Goulder, P. J., et al. (1996) Phenotypic analysis of antigen-
specific T lymphocytes. Science 274, 9496.
9.
9 Schepers, K., Toebes, M., Sotthewes, G., et al. (2002) Differential kinetics of antigen-
specific CD4+ and CD8+ T cell responses in the regression of retrovirus-induced sarco-
mas. J. Immunol. 169, 31913199.
10.
10 Uckert, W., Becker, C., Gladow, M., et al. (2000) Efficient gene transfer into primary
human CD8+ T lymphocytes by MuLV-10A1 retrovirus pseudotype. Hum. Gene Ther. 11,
10051014.
11.
11 Masopust, D., Vezys, V., Marzo, A. L., and Lefrancois, L. (2001) Preferential localization
of effector memory cells in nonlymphoid tissue. Science 291, 24132417.
12.
12 Riddell, S. R., Elliott, M., Lewinsohn, D. A., et al. (1996) T-cell mediated rejection of
gene-modified HIV-specific cytotoxic T lymphocytes in HIV-infected patients. Nat. Med.
2, 216223.
13.
13 Jung, D., Jaeger, E., Cayeux, S., et al. (1998) Strong immunogenic potential of a B7
retroviral expression vector: generation of HLA-B7-restricted CTL response against select-
able marker genes. Hum. Gene Ther. 9, 5362.
14.
14 Stripecke, R., Carmen Villacres, M., Skelton, D., Satake, N., Halene, S., and Kohn, D.
(1999) Immune response to green fluorescent protein: implications for gene therapy. Gene
Ther. 6, 13051312.
15.
15 Skelton, D., Satake, N., and Kohn, D. B. (2001) The enhanced green fluorescent protein
(eGFP) is minimally immunogenic in C57BL/6 mice. Gene Ther. 8, 18131814.
16.
16 Davodeau, F., Peyrat, M. A., Romagne, F., et al. (1995) Dual T cell receptor beta chain
expression on human T lymphocytes. J. Exp. Med. 181, 13911398.
17.
17 Padovan, E., Giachino, C., Cella, M., Valitutti, S., Acuto, O., and Lanzavecchia, A. (1995)
Normal T lymphocytes can express two different T cell receptor beta chains: implications
for the mechanism of allelic exclusion. J. Exp. Med. 181, 15871591.
18.
18 Balomenos, D., Balderas, R. S., Mulvany, K. P., Kaye, J., Kono, D. H., and Theofilopoulos,
A. N. (1995) Incomplete T cell receptor V beta allelic exclusion and dual V beta-express-
ing cells. J. Immunol. 155, 33083312.
19.
19 Topham, D. J., Castrucci, M.R., Wingo, F.S., Belz, G.T., and Doherty, P.C. (2001) The
role of antigen in the localization of naive, acutely activated, and memory CD8(+) T cells
to the lung during influenza pneumonia. J. Immunol. 167, 69836990.
20.
20 Moskophidis, D. and Kioussis, D. (1998) Contribution of virus-specific CD8+ cytotoxic
T cells to virus clearance or pathologic manifestations of influenza virus infection in a T
cell receptor transgenic mouse model. J. Exp. Med. 188, 223232.
21.
21 Hacein-Bey-Abina, S., von Kalle, C., Schmidt, M., et al. (2003) A serious adverse event
after successful gene therapy for X-linked severe combined immunodeficiency. N. Engl.
J. Med. 348, 255256.
22.
22 van Os, R., Sheridan, T. M., Robinson, S., Drukteinis, D., Ferrara, J. L., and Mauch, P. M.
(2001) Immunogenicity of Ly5 (CD45)-antigens hampers long-term engraftment follow-
ing minimal conditioning in a murine bone marrow transplantation model. Stem Cells 19,
8087.
23.
23 Gladow, M., Becker, C., Blankenstein, T., and Uckert, W. (2000) MLV-10A1 retrovirus
pseudotype efficiently transduces primary human CD4+ T lymphocytes. J. Gene Med. 2,
409415.
24.
24 Kinsella, T. M. and Nolan, G. P. (1996) Episomal vectors rapidly and stably produce high-
titer recombinant retrovirus. Hum. Gene Ther. 7, 14051413.
25.
25 Unutmaz, D., KewalRamani, V. N., Marmon, S., and Littman, D. R. (1999) Cytokine sig-
nals are sufficient for HIV-1 infection of resting human T lymphocytes. J. Exp. Med. 189,
17351746.
26.
26 Cavalieri, S., Cazzaniga, S., Geuna, M., et al. (2003) Human T lymphocytes transduced by
lentiviral vectors in the absence of TCR-activation maintain an intact immune compe-
tence. Blood 102(2), 497505.
27.
27 Labrecque, N., Whitfield, L. S., Obst, R., Waltzinger, C., Benoist, C., and Mathis, D.
(2001) How much TCR does a T cell need? Immunity 15, 7182.
28.
28 Zufferey, R., Donello, J. E., Trono, D., and Hope, T. J. (1999) Woodchuck hepatitis virus
posttranscriptional regulatory element enhances expression of transgenes delivered by
retroviral vectors. J. Virol. 73, 28862892.
29. Schumacher, T. N. (2002) T-cell-receptor gene therapy. Nat Rev Immunol 7, 512519.
30. Eshhar, Z. (1997) Tumor-specific T-bodies: towards clinical application. Cancer Immunol.
Immunother. 45, 131136.
Nonself MHC-Restricted CTL 215
16
Immunotherapy With CTL Restricted by Nonself MHC
Liquan Gao, Anne-Marie Downs, and Hans J. Stauss
Summary
In the past years a number of target antigens recognized by cytotoxic T-lymphocytes (CTL)
have been identified in human malignancies. In most cases, the CTL-recognized antigens did
not arise from mutations, but were instead encoded by genes that were identical in normal and
tumor cells. Gene expression in normal tissues may result in tolerance of high avidity CTL,
leaving behind low avidity CTL that cannot provide effective immunity against tumors express-
ing the relevant target antigens. In this chapter we describe a strategy to circumvent immuno-
logical tolerance that can be used to generate high avidity CTL against self-proteins, including
human tumor-associated antigens.
Key Words: Cytotoxic T-lymphocytes; tumor; immunotherapy; tolerance; allorestriction;
avidity.
1. Introduction
One of the most significant advances in the field of tumor immunology has been the
identification and characterization of tumor antigens that are recognized by CTL (1
3). Whilst the expression of some tumor antigens is restricted solely to the tumor cell,
many tumor antigens are also expressed in normal tissues (46). Expression in the
thymus is likely to cause deletion of high-avidity CTL specific for peptides presented
by self-MHC class I molecules. As a result, low-avidity CTL will be released into the
periphery. These low-avidity tumor antigen-specific CTL may be incapable of recog-
nizing naturally processed antigen on the tumor cell surface and are thus ineffective in
tumor rejection. Some high-avidity CTL may escape thymic deletion but are likely to
be rendered unresponsive by antigen expression in the periphery. In this chapter we
will describe a strategy that can be used to circumvent tolerance and to generate high-
avidity CTL specific for tumor antigens that are also expressed in normal tissues.
As T-cell tolerance is self-MHC restricted, the T-cell repertoire is rendered tolerant to
self-peptides presented by self-MHC class I molecules but not to peptides presented by
allogeneic class I molecules (7,8). It is therefore possible to generate high-avidity pep-
tide-specific CTL from allogeneic donors (912). The generation of such allo-
215
216 Gao, Downs, and Stauss
restricted CTL specific for tumor antigens provides a unique opportunity to circumvent
tolerance and to produce high-avidity tumor-specific CTL for adoptive immunotherapy,
despite the antigen being expressed on both normal and transformed cells (9,1316).
The frequency of allo-restricted CTL against a given peptide of interest is usually
low. In order to isolate and expand these small numbers of CTL, cultures need to be
manipulated in order to favor stimulation of peptide-specific CTL and avoid the
expansion of allo-reactive CTL, which are present at a much higher frequency. In this
chapter, we describe the culture conditions that support the isolation and expansion
of low-frequency peptide-specific CTL isolated from allogeneic donors. We focus on
the production of allogeneic CTL that are restricted by HLA-A2 class I molecules,
because it is the most common class I allele in the Caucasian population. Thus, the
protocol describes the isolation of peptide-specific, HLA-A2-restricted CTL from
the peripheral blood of HLA-A2-negative healthy donors. The major steps in this
process are summarized in Fig. 1 and comprise (a) bulk stimulation of HLA-A2-
peripheral blood mononuclear cells (PBMC) with TAP-deficient HLA-A2+ antigen-
presenting cells (APC) coated with HLA-A2-binding peptides; (b) limiting dilution
plating to identify peptide-specific microcultures; and (c) expansion of microcultures
and establishment of peptide-specific CTL lines.
2. Materials
2.1. Cells
1. Responder cells: PBMC (separated from buffy coat blood packs).
2. Stimulator cells:
a. T2 (a TAP [transporter associated with antigen processing]-deficient, HLA-A2+ EBV-
transformed human B-cell line) (17)
b. Drosophila cells transfected with HLA-A2, human 2-microglobulin, B7.1, and
ICAM-1 molecules.
3. HLA-A2+ cells that express the relevant antigen endogenously.
4. HLA-A2+ cells that do not express the relevant antigen endogenously.
5. HLA-A2 cells that express the relevant antigen endogenously.
2.2. Media, Chemicals, and Apparatus

1. RPMI 1640 (Life Technology, UK).
2. Complete T-cell culture medium (CM): RPMI 1640 containing 10% fetal calf serum
(FCS) (see Note 1), 100U/mL penicillin, 100 g/mL streptomycin, and 2 mM L-glutamine.
3. Schneiders insect cell-culture medium (Invitrogen, UK) containing 5% heat-inacti-
vated FCS.
4. Copper sulphate (Sigma, UK).
5. T75 tissue-culture flasks with filter caps (Helena Biosciences, UK).
6. T75 tissue-culture flasks with sealed caps (Nalgene Nunc, UK).
7. Lymphoprep (Axis Shield, Norway).
8. Dimethyl sulphoxide (DMSO) (Sigma-Aldrich, Poole, UK).
9. Cryovials (Nalgene Nunc).
10. Mouse anti-HLA-A2 monoclonal antibody HB-54 (American Type Culture Collection
(ATCC), Manassas, VA) (18).
Fig. 1. Flow-chart summarizing the process of generating human allo-restricted peptide-

specific cytotoxic T-lymphocytes.
11. Mouse anti-HLA-A2 monoclonal antibody HB-117 (ATCC) (19).

12. Sheep antimouse IgG-FITC (Sigma).
13. Phosphate-buffered saline (PBS).
14. Peptides:
a. HLA-A2-binding relevant peptide of interest.
b. HLA-A2-binding irrelevant control peptide.
15. Dithiothreitol (DTT) (Sigma).
16. Anti-CD4 monoclonal antibody QS4120 (20).
17. Twenty-four-well tissue-culture plates (Nalgene Nunc).
18. Human recombinant interleukin (IL)-2 (Chiron, Amersterdam-Zuidoost, Netherlands).
19. Human IL-7 (R&D Systems, Abingdon, UK).
20. Round-bottom 96-well tissue-culture plates (Nalgene Nunc).
21. Two-mercaptoethanol (2-ME) (Sigma-Aldrich).
22. 51Cr (sodium chromate aqueous solution) (Amersham, Bucks, UK).
23. One percent trifluoroacetic acid (TFA) (Sigma-Aldrich).

24. LP2 round-bottom tubes (Life Sciences International, UK).
25. Forty-eight-well tissue-culture plates (Nalgene Nunc).
26. Access to: irradiator, radiation facilities for 51Cr work, and gamma counter.
3. Methods
The methods described below outline (1) the in vitro stimulation of HLA-A2
PBMC with peptide-loaded HLA-A2+ APC; (2) re-stimulation of bulk cultures; (3)
limiting dilution plating; (4) the identification of peptide-specific microcultures; and
(5) the expansion and validation of peptide-specific microcultures.
3.1. In Vitro Stimulation of HLA-A2 PBMC

With Peptide-Loaded HLA-A2+ APC
CTL lines are generated by incubating PBMC of HLA-A2 donors with TAP-defi-
cient HLA-A2+ APC. The APC used are unable to present endogenous peptides and
can efficiently present exogenously loaded peptides. The commonly used APC are
Drosophila cells transfected with HLA-A2, human 2-microglobulin, B7.1, and
ICAM-1 molecules, T2 cells (TAP-deficient and HLA-A2+) and RMA-S-A2 cells
(TAP-deficient murine cell line [RMA-S] transfected with the human HLA-A2 mol-
ecule) (see Note 2). For the initial stimulation, IL-7 is added to the cultures, and for
further stimulations, recombinant IL-2 and autologous feeders are also added to the
cultures (see Note 3).
How to set up the initial in vitro stimulation culture is described in Subheadings
3.1.1.3.1.4. . This includes details about the culture of the cell lines used as APC
(stimulators) (Subheading 3.1.1.), PBMC preparation (responders) (Subheading
3.1.2.), preparation of the peptides used (Subheading 3.1.3.), and setting up the stimu-
lation cultures (3.1.4.).
3.1.1. Culture of Cell Lines To Be Used As Stimulators

All tissue-culture procedures should be performed in a class II tissue-culture cabinet.
3.1.1.1. T2 CELLS
T2 cells should be cultured in T75 tissue-culture flasks in Comple T-cell culture
medium (CM), which should be warmed to 37C prior to use. Split the cells approx
twice a wk to a cell density of 15 105/mL. Culture the cells in an upright flask in a
humidified 37C incubator with 5% CO2.
3.1.1.2. DROSOPHILA CELLS
Drosophila cells should be cultured in Schneiders insect cell-culture medium
supplemented with 5% FCS. Split the cells approx twice a wk by re-seeding the cells
to a density of 23 105/mL in a fresh T75 flask (with sealed cap). Drosophila cells
should be incubated at 24C (with atmospheric CO2), and the flasks should be placed
horizontally. Drosophila cells need to be induced with 100 mM copper sulphate 48 h
prior to peptide loading to induce expression of the transfected HLA-A2 and co-stimu-
lator molecules. After induction, wash cells three times with RPMI in order to remove
the copper sulphate before peptide loading.
Screen all cell cultures for mycoplasma regularly.
3.1.2. PBMC Preparation

Described below is the separation of lymphocytes from buffy coat blood packs, the
freezing of PBMC for use as feeders, and the staining of PBMC to identify HLA-A2
PBMC.
3.1.2.1. SEPARATION OF LYMPHOCYTES FROM BUFFY COAT BLOOD PACKS
Lymphocytes are separated from buffy coat blood packs by lymphoprep density
centrifugation.
1. Ensure blood packs are sterile by spraying with 70% ethanol.
2. Drain the blood (approx 5080 mL) into a T75 tissue-culture flask.
3. Add an equal volume of room temperature RPMI (containing no additives) and mix well
but gently.
4. Pipet 20 mL of lymphoprep (room temperature) into a 50-mL tube and gently apply 30 mL
of diluted blood onto the lymphoprep surface.
5. Centrifuge for 20 min at 800g (room temperature) with the brake turned off.
6. Collect the interface fraction (approx 10 mL) into a 50-mL tube.
7. Add 40 mL cold RPMI (containing no additives) and centrifuge for 10 min at 800g.
8. Repeat this wash step twice.
9. Count the cells and resuspend in CM at a concentration of 5 107/mL.
3.1.2.2. FREEZING CELLS

Preparation of cryopreserved stocks of PBMC should be well planned in order to
provide enough feeders for re-stimulation of bulk cultures, setting up of microcul-
tures, and establishment of CTL lines.
Freeze:
1. Ten vials of 5 106 cells per vial for use as feeders during the early stages of CTL culture.
2. Ten vials of 1 107 cells per vial for use as feeders during later stages of CTL culture.
3. The remaining PBMC at 5 107 per vial for use as feeders for CTL expansion.
PBMC are cryopreserved in freezing solution containing 20% DMSO, 50% FCS,
and 30% RPMI. The freezing solution should be chilled prior to use by storing on ice.
1. Add an equal volume of cold freezing solution to the cell suspension (therefore diluting
freezing solution 1:2).
2. Place 1 mL of the cell suspension mixture into a prelabeled cryovial and place the vials
on ice.
3. Place the cryovials in a freezing box and store at 80C overnight.
4. The next d, transfer vials to liquid nitrogen.
3.1.2.3. DETECTION OF HLA-A2 PBMC

HLA status is determined by staining the cells with antibodies that bind to HLA-
A2. To determine HLA status, two antibodies are used for staining PBMCHB-54
(anti-HLA-A2, B17) (18) and HB-117 (HLA-A2, A28) (19). The antibodies are
detected using secondary staining with antimouse IgG-FITC antibody. Only if both
antibodies stain positive, are the PBMC from a buffy coat blood pack defined as HLA-
A2+. If only one of the antibodies binds but not the other, the PBMC are defined as
HLA-A2. Only HLAA2 PBMC should be used as responders and feeders.
3.1.3. Preparation of Synthetic Peptides
Peptides are most commonly synthesized using fluorenylmethoxycarbonyl chemis-
try. Peptide purity is an important factor and should be greater than 90%. Some pep-
tides can be diluted in PBS and should be diluted to a working stock concentration of
2 mM. Hydrophobic peptides should first be dissolved in DMSO, in a volume that will
be one-tenth of the final volume, and then diluted in PBS to produce a working stock
concentration of 2 mM. Peptides containing cysteines or methionines should be dis-
solved in neat DMSO first, in a volume that will be 10% of the final volume. DTT is
then added to prevent oxidation and is added at a maximal final concentration of 5 mM
(high concentrations of DTT can be toxic to T-cells and APC). Finally, PBS (pH 7.4)
is added to produce a final peptide concentration of 2 mM. Peptides should be stored at
20C in 200-L aliquots.
3.1.4. In Vitro Stimulation (Day 0)
3.1.4.1. PEPTIDE-LOADING OF STIMULATOR CELLS
1. Induce the Drosophila transfectants with copper sulphate 2436 h prior to setting up the
in vitro stimulation (see Subheading 3.1.1.).
2. On d of setting up the stimulation, wash the required number of Drosophila cells three
times with RPMI and resuspend the cells (up to 2 106) in 200 L RPMI containing 5%
boiled FCS (see Note 4).
3. Add 10 L of 2 mM peptide to produce a final peptide concentration of 100 M and mix
well.
4. Incubate at 37C for 2 h.
5. Wash cells once with RPMI.
3.1.4.2. SETTING UP THE STIMULATION CULTURE

1. Set up two wells on a 24-well tissue-culture plate in 2 mL CM containing the following:
2 106 HLA-A2 PBMC (responders); 2 105 peptide-loaded Drosophila transfectants
(stimulators) (see Note 5); 2.5 ng/mL human IL-7; 10% mycoplasma-free QS4120 (anti-
CD4 monoclonal antibody) (20) (see Note 6).
2. Culture the cells in a humidified incubator at 37C with 5% CO2.
3.2. Re-Stimulation of the Bulk Culture
The bulk CTL culture should be re-stimulated with fresh stimulators, feeders, and
cytokines on d 5.
1. Re-suspend up to 2 106 T2 cells in 200 L RPMI containing 5% boiled FCS. Add 10 L
of 2 mM peptide to produce a final peptide concentration of 100 M and mix well.
2. Incubate at 37C for 2 h to peptide-load the T2 cells.
3. Irradiate the T2 cells with 80 Gy and wash once with RPMI.
4. Defrost autologous PBMC for use as feeders: thaw the cells quickly in a 37C water bath
and pipet gently into 10 mL of cold RPMI containing no additives (see Note 7). Centrifuge
for 5 min at 400g and resuspend the cell pellet in cold RPMI (no additives) for irradiation.
Irradiate the PBMC with 30 Gy, then spin and resuspend the cells in warm CM.
5. Harvest and count the T-cells from the bulk culture (see Note 8).
6. Place the following cell numbers and cytokines into a fresh 24-well tissue-culture plate
(see Note 9) in 2 mL CM per well:
a. 1 106 T-cells from the bulk culture.
b. 2 105 peptide-loaded T2 cells (irradiated, 80 Gy).
c. 2 106 autologous PBMC (irradiated, 30 Gy).
d. 10 U/mL human IL-2.
e. 2.5 ng/mL human IL-7.
7. Culture the cells in a humidified incubator at 37C with 5% CO2.
To confirm that the irradiation dose used is sufficient to prevent proliferation of
the irradiated cells, always set up one well with irradiated cells only. Add an aliquot
of 1 105 cells in 1 mL CM in a well of a 24-well tissue-culture plate and culture for
5 d at 37C. If proliferation of the irradiated cells is observed, the CTL culture must
be discontinued and restarted using a higher dose of irradiation.
3.3. Limiting Dilution and Re-Stimulation
This section describes the setting up of limiting dilution cultures (see Subheading
3.3.1.) and the subsequent re-stimulation of microcultures (see Subheading 3.3.2.).
3.3.1. Limiting Dilution Cultures
The aim of this procedure is to isolate peptide-specific microcultures. Fig. 2 shows
that without limiting dilution, it is difficult to detect peptide-specific CTL in bulk
cultures. Limiting-dilution plating is performed after two and three rounds of bulk
stimulation (see Fig.1.).
1. Harvest and count the T-cells from the bulk culture.
2. Using round-bottom 96-well tissue-culture plates, plate out three plates of 1 responder
CTL per well and one plate with 10 responder CTL per well, in 100 L CM. At this stage,
while each well contains very low CTL numbers, CM should contain 5 105 M 2-ME.
3. Add 100 L CM (containing 5 105 M 2-ME) per well, containing:
a. 2 104 peptide-loaded T2 cells (irradiated, 80 Gy).
b. 2 105 autologous PBMC (recovered from frozen and irradiated, 30 Gy).
Fig.2. Cytotoxicity assay of a typical bulk culture. Cytotoxic T-lymphocytes (CTL) from a
HLA-A2 donor were stimulated with HLA-A2+ APC coated with pWT126 peptide. pWT126
and pWT235 (13,21) are peptides derived from the tumor-associated antigen Wilms tumor 1
(WT1), which is a zinc finger transcription factor that is expressed at low levels in normal cells
and overexpressed in several malignancies (2224). CTL from the pWT126-stimulated bulk
culture were tested for killing ability after two rounds of stimulation and prior to limiting dilu-
tion. The graph shows the percentage of specific killing of T2 cells loaded with relevant
(pWT126) and irrelevant (pWT235) peptide (both HLA-A2-binding) and a HLA-A2+ leukemia
line that expresses WT1 endogenously (BV173). Effector to target cell (E:T) ratios ranged
from 100:1 to 3:1. The data demonstrate that prior to limiting dilution there is a high level of
HLA-A2 allo-reactivity but no peptide-specific killing of target cells.
c. 20 U/mL human IL-2 (10 U/mL final concentration).

d. 5 ng/mL human IL-7 (2.5 ng/mL final concentration).
4. Re-stimulate the remaining bulk culture and perform a second limiting dilution one wk
later. This is to increase the likelihood of isolating the desired CTL. All subsequent pro-
cedures then apply as for the first round of limiting-dilution plating.
3.3.2. Re-Stimulation of Microcultures in 96-Well Tissue-Culture Plates
Fourteen d after setting up the limiting-dilution plates, peptide-specific CTL are
still in small numbers and need to be expanded by re-stimulating the cells. Remove
100 L medium from each well and replace with the following in 100 L CM (con-
taining 5 105 M 2-ME):
1. 2 104 peptide-loaded T2 cells (irradiated, 80 Gy).
2. 2 105 autologous PBMC (recovered from frozen and irradiated, 30 Gy).
3. 20 U/mL human IL-2 (10 U/mL final concentration).
4. 5 ng/mL human IL-7 (2.5 U/mL final concentration).
Fig. 3. A typical split-well analysis of 96-well microcultures (shown are 24 wells only).
These microcultures were derived from limiting dilution of the bulk cultures described in Fig. 2.
Wells A1A6 show good killing of T2 cells loaded with the relevant peptide (pWT126) and
BV173 cells. There is little background killing of T2 cells loaded with irrelevant peptide
(pWT235). Wells B1B12 show high levels of HLA-A2 allo-reactivity, while wells C1C6 show
no killing activity against any of the three targets.
3.4. Screening of Microcultures for Specificity (Split-Well Analysis)

The limiting-dilution plates are screened to determine the ability of the microcul-
tures to kill target cells loaded with the relevant peptide. This is done using a chromium
release cytotoxicity assay which is performed 56 d after the last stimulation, as this
is when CTL killing activity peaks. A sample of CTL are removed from each of the
microcultures and tested for their ability to kill T2 loaded with the relevant peptide and
irrelevant control peptide. The process of screening limiting-dilution plates by this
method is referred to as split-well analysis and is described in this section. Fig. 3
shows a typical result of split-well analysis of 96-well microcultures.
1. Peptide load the T2 cells (targets) (see Subheading 3.2.) with 100 M relevant/irrelevant
peptide for 1 h at 37C.
2. Resuspend peptide-loaded T2 cells in 200 L CM and add 50 Ci 51Cr per 1 106 cells.
Chromium-label slightly more cells than required to allow for loss of cells during subse-
quent washing steps. Incubate at 37C for 1 h.
3. After labeling, wash the target cells three times with CM.
4. Count and resuspend target cells to 5 104/mL.
5. Thoroughly mix the CTL in each well of the limiting-dilution plates with a multichan-
nel pipet.
6. Transfer 50 L of CTL from each well into the corresponding wells of a fresh round-
bottom 96-well plate and do the same for a second 96-well plate. Two plates are set up
per limiting-dilution plateone as a control to test the killing ability of the CTL against
T2 loaded with irrelevant peptide (plate A) and the second to test the killing ability of
T2 loaded with the relevant peptide (plate B).
7. Spin the plates containing the transferred cells at 340g for 3 min and discard all the super-
natant by flicking the plate once (see Note 10).
8. Resuspend the cell pellets in 100 L CM per well.
9. Add 100 L (5 103) T2 cells loaded with irrelevant peptide to each well of the control
plate A and 100 L (5 103) T2 cells loaded with the relevant peptide to each well of
plate B.
10. Set up eight wells to measure the spontaneous chromium release by adding 100 L of
chromium-labeled T2 cells and 100 L of CM per well.
11. Set up eight wells to measure the maximum release by adding 100 L of chromium-
labeled T2 cells and 100 L of 1% TFA per well.
12. Spin the plates at 340g for 3 min, then place into a humidified 37C incubator with 5%
CO2 for 4 h.
13. Harvest 100 L of supernatant from each well and place into a round-bottom LP2 tube.
14. Measure the radioactivity using a Wallac Gamma Counter (Wallac, UK).
The specific lysis is calculated by the equation:
( experimental release spontaneous release )
100 (1)
maximum release spontaneous release
If the specific lysis of the T2 cells loaded with the relevant peptide is greater than
10% of that of T2 loaded with irrelevant peptide, the well is defined as positive and
will be expanded in the next step.
3.5. Expansion and Validation of Micro-CTL Cultures

This section describes the expansion of micro-CTL cultures from 96-well plate to
48-well plate and from 48-well plate to 24-well plate ( Subheading 3.5.1.); the valida-
tion of expanded CTL, including details of the cytotoxicity assay ( Subheading 3.5.2.);
and the expansion and maintenance of validated CTL ( Subheading 3.5.3.).
3.5.1. Expansion of Micro-CTL Cultures

The d following split-well analysis, transfer all CTL from each of the wells deter-
mined to be positive into a 48-well plate. Add 1 mL of CM containing:
3. 10 U/mL human IL-2.
Seven days after expansion to 48-well plates, transfer wells which are proliferating
to 24-well plates and re-stimulate. To do this, remove 0.5 mL of medium (trying not to
Fig. 4. Cytotoxicity assay demonstrating the specificity of high-avidity allo-restricted CTL

generated against the pWT126 peptide. pWT126-specific cytotoxic T-lymphocytes (CTL) were
isolated by limiting dilution of bulk cultures and expanded by stimulation with T2 cells coated
with pWT126 peptide. Isolated CTL lines killed the TAP-deficient T2 target cells loaded with the
relevant peptide (pWT126), C1R-A2 cells (HLA-A2+ and WT1) loaded with the relevant peptide,
and HLA-A2+ WT1+ leukemic cell lines BV173 and LAMA81. The CTL lines did not kill T2 cells
or C1R-A2 cells loaded with the irrelevant HLA-A2-binding pWT235 control peptide.
remove any cells) from each well of the 48-well plate and discard. Transfer the remain-
ing 0.5 mL containing CTL into a 24-well plate and add 1.5 mL CM containing:
3. 10 U/mL human IL-2.
3.5.2. Validation of the CTL Lines
All CTL lines must be tested for their killing ability against a panel of peptide-
coated targets and target cells expressing the antigen endogenously, using a cytotoxic-
ity assay. It is important to use both sets of target cells, as low-avidity CTL may kill
peptide-loaded targets efficiently but not targets expressing the antigen endogenously.
3.5.2.1. CYTOTOXICITY ASSAY
All proliferating CTL in 24-well cultures should be validated using the cytotoxic-
ity assay on d 5 or 6 after re-stimulation. Fig. 4 shows a typical result from a cyto-
toxicity assay in which high-avidity allo-restricted CTL generated against the
pWT126 peptide were tested against a panel of targets. Details of the cytotoxicity
assay are described in Subheading 3.4. The same procedure applies at this stage;
however, in order to obtain more information about the sensitivity of the expanded
CTL, CTL are tested for their killing ability at a range of effector-to-target-cell ratios
(E:T ratios). Start the E:T ratio at 25:1 (using 5 103 target cells) and perform double
dilutions down a 96-well plate (i.e., eight dilutions). Target cells should include:
1. T2 cells loaded with the relevant peptide.
2. T2 loaded with irrelevant peptide.
3. HLA-A2+ cells that express the relevant antigen endogenously.
4. HLA-A2+ cells that do not express the relevant antigen endogenously.
5. HLA-A2 cells that do express the relevant antigen endogenously.
CTL lines that specifically kill HLA-A2+ cells expressing the relevant antigen and
T2 loaded with the relevant peptide and not the control cells (see Fig. 4) will be further
expanded.
Phenotypic profiling is also carried out, in order to further validate the CTL lines.
This is done by staining the CTL with fluorescence-labeled antibodies and analyzing
by flow cytometry. For characterization of a CTL line, the most common antibodies to
use are anti-CD3, anti-CD4, and anti-CD8. Anti-V antibodies can be used to deter-
mine the T-cell receptor (TCR)-V usage, and tetramer staining is also used to provide
additional information regarding the specificity of the TCR. Tetramer staining should
be performed on d 810 following re-stimulation.
3.5.3. Expansion and Maintenance of Allo-Restricted Peptide-Specific CTL

Large-scale expansion of peptide-specific allo-restricted CTL lines is technically
straightforward but not always successful. CTL lines that have been repeatedly stimu-
lated in culture do not expand as efficiently as recently established lines. We recom-
mend performing one large-scale expansion of the CTL lines and to make sufficient
numbers of freezings for later use. This should be done as soon as possible after the
CTL lines have been expanded from 48-well to 24-well plate. In our experience, CTL
grow much better in 24-well plates than flasks. Re-stimulate the CTL every 712 d, by
adding the following in 2 mL CM per well:
1. 5 105 CTL.
3. 2 106 autologous or HLA-A2- donor PBMC (recovered from frozen and irradiated, 30 Gy)
(see Note 3).
4. 10 U/mL IL-2.
4. Notes
1. FCS (both heat-inactivated and non-heat-inactivated) from a range of suppliers should be
batch tested prior to use in these experiments, to see which is most suitable for supporting
the growth and proliferation of human T-cells. This can be done by stimulating a T-cell
line in CM supplemented with the FCS being tested and measuring the proliferation on
d 5 (by 3H-thymidine incorporation) or by counting the T-cell numbers at time points up
until d 10 after stimulation. Alternatively, a mixed lymphocyte reaction can be set up in
CM containing the different batches of FCS and the proliferation measured on d 5.
2. Stimulators can be switched during stimulation of the bulk culture to minimize the gen-
eration and re-stimulation of CTL with unwanted peptide specificities. The most efficient
stimulator to use for the initial stimulation is the Drosophila transfectant. Stimulators for
subsequent re-stimulations of the bulk culture can then be alternated between T2 and
RMA-S-A2 cells up until limiting dilution and subsequent expansions, when T2 cells
should be used.
3. It is better to use autologous feeders during the initial phase of CTL culture and limiting-
dilution plating. Once the CTL line has been established and validated, third-party HLA-
A2 feeders can be used.
4. Boiled FCS is used in the culture medium during peptide loading. This includes peptide
loading for stimulations and also for the cytotoxicity assay. FCS contains enzymes that
may degrade the peptides, and therefore boiling the FCS may help to reduce this enzyme
activity. To boil the FCS, it must first be diluted to 50% in RPMI (containing no addi-
tives) and then heated to 95C for 10 min. The FCS can then be further diluted in RPMI
(to 5%) for use in peptide loading.
5. Drosophila cells do not need to be irradiated, as they will die after a short period of time
in CM at 37C.
6. Addition of hybridoma culture supernatant that contains anti-human CD4 antibody has
proved to be beneficial in inhibiting proliferation of the CD4+ population and favoring
expansion of the desired CD8+ CTL population. However, the presence of some CD4+
cells is desirable. In the conditions described here, CD8+ clones have a limited lifespan in
the absence of CD4+ T-cells, and therefore long-term CTL lines usually contain CD4+
cells.
7. When PBMC are defrosted, they must be resuspended in cold RPMI containing no addi-
tives, to prevent clumping and subsequent reductions in recoverable cell numbers.
8. CTL cultures are very sensitive to changes in environmental conditions. For this reason,
do not leave cultures out of the incubator longer than necessary.
9. Do not reuse the same wells of tissue-culture plates for re-stimulation, as the cells do not
grow well in the old wells. Also, tissue-culture plasticware may vary in terms of toxicity
to cultures; therefore, use plasticware from the same supplier throughout.
10. The plates must be flicked only once in order to remove supernatant, as cells will be lost
if the plates are flicked more than this.
References
1.
1 van der Bruggen, P., Traversari, C., Chomez, P., et al. (1991) A gene encoding an anti-
gen recognizedby cytolytic T-lymphocytes on a human melanoma. Science 254, 1643
1647.
2.
2 Kawakami, Y., Eliyahu, S., Delgado, C. H., et al. (1994) Identification of human mela-
noma antigen recognized by tumor infiltrating lymphocytes associated with in vivo tumor
rejection. Proc. Natl. Acad. Sci. USA 91, 64586462.
3.
3 Finn, O. J. (1993) Tumor-rejection antigens recognized by T lymphocytes. Curr. Opin.
Immunol. 5, 701708.
4.
4 Theobald, M., Biggs, J., Dittmer, D., Levine, A. J., and Sherman, L. A. (1995) Targeting
p53 as a general tumor antigen. Proc. Natl. Acad. Sci. USA 92, 11,99311,997.
5.
5 De Visser, K. E., Schumacher, T. N., and Kruisbeek, A. M. (2003) CD8+ T cell tolerance
and cancer immunotherapy. J. Immunother. 26, 111.
6.
6 Morris, E. C., Bendle, G. M., and Stauss, H. J. (2003) Prospects for immunotherapy of
malignant disease. Clin. Exp. Immunol. 131, 17.
7.
7 Matzinger, P., Zamoyska, R., and Waldmann, H. (1984) Self tolerance is H-2 restricted.
Nature 308, 738741.
8.
8 Rammensee, H. G. and Bevan, M. J. (1984) Evidence from in vitro studies that tolerance
to self antigens is MHC-restricted. Nature 308, 741744.
9.
9 Sadovnikova, E. and Stauss, H. J. (1996) Peptide-specific cytotoxic T lymphocytes restricted
by nonself major histocompatibility complex class I molecules: Reagents for tumor immu-
notherapy. Proc. Natl. Acad. Sci. USA 93, 13,11413,118.
10.
10 Obst, R., Munz, C., Stevanovic, S., and Rammensee, H. G. (1998) Allo- and self-restricted
cytotoxic T lymphocytes against a peptide library: evidence for a functionally diverse
allorestricted T cell repertoire. Eur. J. Immunol. 28, 24322443.
11.
11 Munz, C., Obst, R., Osen, W., Stevanovic, S., and Rammensee, H. G. (1999) Alloreactivity
as a source of high avidity peptide-specific human CTL. J. Immunol. 162, 2534.
12.
12 Moris, A., Teichgraber, V., Gauthier, L., Buhring, H. J., and Rammensee, H. G. (2001)
Cutting edge: characterization of allorestricted and peptide-selective alloreactive T cells
using HLA-tetramer selection. J. Immunol. 166, 48184821.
13.
13 Gao, L., Bellantuono, I., Elsasser, A., Marley, S. B., Gordon, M. Y., Goldman, J. M., and
Stauss, H. J. (2000) Selective elimination of leukemia CD34+ by CTL specific for WTI.
Blood 95, 21982203
14 Sadovnikova, E., Jopling, L. A., Soo, K. S., and Stauss, H. J. (1998) Generation of human
14.
tumor-reactive cytotoxic T cells against peptides presented by non-self HLA class I mol-
ecules. Eur. J. Immunol. 28, 193200.
15.
15 Stauss, H. J. (1999) Immunotherapy with CTLs restricted by nonself MHC. Immunol.
Today 20, 180183.
16.
16 Mutis, T., Blokland, E., Kester, M., Schrama, E., and Goulmy, E. (2002) HLA class II
restricted T-cell reactivity to a developmentally regulated antigen shared by leukemic cells
and CD34+ early progenitor cells. Blood 100, 547552.
17.
17 DeMars, R., Chang, C. C., Shaw, S., Reitnauer, P. J., and Sondel, P. M. (1984) Homozy-
gous deletions that simultaneously eliminate expressions of class I and class II antigens of
EBV-transformed B-lymphoblastoid cells. I. Reduced proliferative responses of autolo-
gous and allogeneic T cells to mutant cells that have decreased expression of class II
antigens. Hum. Immunol. 11, 7797.
18.
18 McMichael, A. J., Parham, P., Rust, N., and Brodsky, F. (1980) A monoclonal antibody
that recognizes an antigenic determinant shared by HLA A2 and B17. Hum. Immunol. 1,
121129.
19.
19 Parham, P. and Bodmer, W. F. (1978) Monoclonal antibody to a human histocompatibility
alloantigen HLA-A2. Nature 276, 397399.
20.
20 Healey, D., Dianda, L., Moore, J. P., et al. (1990) Novel anti-CD4 monoclonal antibodies
separate human immunodeficiency virus infection and fusion of CD4+ cells from virus
binding. J. Exp. Med. 172, 12331242.
21.
21 Bellantuono, I., Gao, L., Parry, S., Marley, S., Dazzi, F., Apperley, J., Goldman, J. M.,
and Stauss, H. J. (2002) Two distinct HLA-A0201-presented epitopes of the Wilms tu-
mor antigen 1 can function as targets for leukemia-reactive CTL. Blood 100, 3835
3837.
22.
22 Oji, Y., Ogawa, H., Tamaki, H., et al. (1999) Expression of the Wilms tumor gene WT1
in solid tumors and its involvement in tumor cell growth. Jpn. J. Cancer Res. 90, 194204.
23. Menssen, H. D., Bertelmann, E., Bartelt, S., et al. (2000) Wilms tumor gene (WT1) expres-
sion in lung cancer, colon cancer and glioblastoma cell lines compared to freshly isolated
tumor specimens. J. Cancer Res. Clin. Oncol. 126, 226232.
24. Scharnhorst, V., van der Eb, A. J., and Jochemsen, A. G. (2001) WT1 proteins: functions
in growth and differentiation. Gene 273, 141161.
Designing TCR for Cancer Immunotherapy 229
17
Designing TCR for Cancer Immunotherapy
Ralf-Holger Voss*, Jrgen Kuball*, and Matthias Theobald
*Both authors contributed equally to this work.

Summary
Reprogramming T-cell populations by T-cell receptor (TCR) gene transfer is a new thera-
peutic tool for adoptive tumor immunotherapy. Gene transfer of human leukocyte antigen
(HLA)-transgenic mice-derived TCR into human T-cells allows the circumvention of tolerance
to tumor-associated (self) antigens (TAA). This chapter reports on the identification of the _
and ` chains of the heterodimeric TCR derived from a mouse T-cell clone. The related DNA
fragments are inserted into a retroviral vector for heterologous expression of the TAA-specific
TCR in human T-cells. Polymerase chain reaction (PCR)-based cloning protocols are provided
for the tailor-made customization of murine TCR. We describe the humanization and
chimerization of such TCR as well as their expression in human T-cells.
Key Words: Cancer therapy; immunotherapy; T-cell receptor; high-affinity TCR; high-
avidity T-cell; tolerance; human leukocyte antigen; major histocompatibility complex; HLA-
A*0201 restriction; tumor-associated antigen; HLA-A2 transgenic mice; reverse transcription
and polymerase chain reaction; rapid amplification of cDNA ends; single-chain TCR and single-
chain Fv; retroviral transduction; packaging cell line 293T and phoenix-ampho; adoptive trans-
fer of T-cells.
1. Introduction
1.1. T-Cell Receptors As Molecular Tools for Cancer Immunotherapy
Immunotherapy of cancer is becoming feasible by the recent break-through in both
the understanding of the principles of molecular aspects of tumor immunology (1)
and the development of biochemical techniques for the rapid identification, modifi-
cation, and in vitro as well as in vivo reimplementation of biologically relevant mol-
ecules (2). On a molecular basis, the T-cell receptor (TCR) represents the direct link
between the effector T-cell and the target cell (3). The TCR triggers a cascade of
signaling events directed towards the interior of the T-cell after successfully recog-
nizing the major histocompatibility complex (MHC)-buried peptidein the case of
cancer, the tumor-associated antigen (TAA) (4). However, TAA-specific T-cells with
TCR of sufficient affinity cannot frequently be recruited by immunization or are
absent from the human T-cell repertoire due to mechanisms of self-tolerance (5,6).
229
230 Voss, Kuball, and Theobald
Thus, for a successful T-cell-based tumor immunotherapy, it is necessary to extend

the T-cell repertoire of cancer patients by adoptively transferring activated T-cells
harnessed with TAA-specific TCR of sufficient affinity (7,8). The introduced TCR
should fulfill the following biochemical criteria:
1. TCR affinity has to exceed a minimal threshold critical for the recognition of even very
low amounts of tumor antigen on tumor cells (9,10). T-cells of sufficient avidity are
available against a small number of defined TAAs (11). Once cloned, their TCR can be
isolated and retrovirally imported into patients T-cells and subsequently expanded in
vitro, thereby preventing the cumbersome establishment of TAA-specific T-cell clones
for adoptive transfer (Fig. 1). However, for most of the TAAs, either the avidity of
specific T-cells is low or they are even absent from the repertoire. As a consequence,
many laboratories have attempted to increase the quality of TAA-specific yet low-
affinity TCR by different biotechnological approaches, such as yeast (12) or TCR dis-
play (13). In iterative cycles of affinity sampling, high-affinity TCR are generated and
filtered out to yield affinities in the nanomolar range. In addition, the generation of
allo-peptide-reactive T-cells is a source for high-quality TCR, and is discussed else-
where in this book (7,14). As an alternative, the HLA-A2(Kb)-transgenic mouse model
allows the generation of high-affinity TCR specific for human TAAs (6,15). A2Kb-
transgenic mice contain the human _1 and _2 domains of HLA-A*0201 and the mouse
_3 domain of H-2Kb, which allows the interaction of murine CD8 with the A2Kb trans-
gene product. The peripheral T-cell repertoire of these transgenic mice has not been
depleted of high-affinity TCR directed against human antigens, and therefore repre-
sents a diverse source of HLA-A2-restricted, high-avidity T-cells and their respective
TCR ([16], Fig. 1; this method will be described under Subheading 3.1.). However, a
major disadvantage arises from the xenogenic mouse origin of the TCR, which may
require their partial (7) or almost complete humanization to affect their potential immu-
nogenicity in humans. The successful humanization of mouse antibodies as a prototype
for immunoglobulin-like folded proteins (17) and their subsequent clinical benefit have
been documented in a multitude of published trials ([18] and J. Saldanhas web-site:
Humanization by Design [http://people.cryst.bbk.ac.uk/~ubcg07s/]). The structural
homology of antibodies and T-cell receptors enables the embodiment of antibody-
related characteristics in TCR constructs with respect to affinity maturation, humaniza-
tion, and chimerization. A protocol for partial humanization of a mouse TCR will be
described under Subheading 3.3.1.
Not only to overcome the affinity barrier, T-bodies have been designed that make
use of antibody-derived antigen-specific high-affinity variable domains covalently linked
to invariant domains of different origin. Such T-bodies are ideally suited to produce high-
avidity, MHC-independent T-cells (19).
2. Exogenous TAA-specific TCR introduced into a T-cell face a significant proportion of
endogenous TCR (14 104 per cell, [20]). Thereby, the exogenous TCR have to com-
pete with the endogenous TCR to become part of the functional TCR/CD3 complex, con-
sisting of the heterodimeric TCR_` and the heterohexameric CD3 complex, formed
during post-translational assembly in the endoplasmatic reticulum and subsequent trans-
port to the cell surface (21). Mouse TCR are of sufficient structural homology to be incor-
porated into the human CD3 complex (7) and to rescue surface expression in mutant
T-cells (22). Surprisingly, exogenous mouse TCR for different TAA specificities are
prone to be more favorably incorporated into the human signaling complex as compared
to an exogenous gp100-specific human TCR (unpublished results).
Designing TCR for Cancer Immunotherapy
231
231
Fig. 1. T-cell receptor (TCR) gene transfer into human T-cells for adoptive immunotherapy. TCR*: endogenous TCR.
In case of partially humanized TCR harboring human constant domains, the assembly
machinery is not able to dissect endogenous and exogenous TCR due to the invariant
nature of the constant domains. Hence, combinatorial preferences, if they exist at all, can
be solely assigned to the nature of the subfamily-specific variable domains. Although the
major determinants for interchain pairing are believed to be located within the constant
domains (23), the stability of the TCR duplex may also be influenced by their favorable
subfamily interchain pairing (2426). Nevertheless, it is not ultimately clarified to what
extent structural constraints or thymic selection processes determine the combinatorial
TCR diversity of an individual (27). This restriction does not seem to pose any problems
on the sufficient coverage of functional diversity, as only 2 106 clones, orders of mag-
nitude less than the maximum theoretical size of 1015, have been estimated in the spleen
of immunocompetent mice (28).
3. One potential problem is the undesired event of mixed chain pairing between endo-
genous and transferred TCR, which is likely to happen predominantly with the delivery
of human or humanized TCR. First, self-reactive specificities could be generated, and
second, the correct heterodimeric TCR-surface expression with specificity for the desired
TAA will be decreased and may thus impair functional responsiveness. To avoid this
event, the single chain TCR concept has been developed as an attractive alternative,
which aims to covalently attach the variable _/` domains to each other via a flexible and
thermodynamically stable linker (29,30). This molecule can be chimerized with a con-
stant domain that allows membrane anchoring and assembly with the CD3 complex (31).
To enhance signaling efficiency, these chimeric molecules have also been hooked up to
the CD3-c chain as a substrate for proximal signaling functions, such as tyrosine phos-
phorylation by Syk/Fyn kinases (32), or by fusion to the downstream molecules them-
selves (33). Apart from the ITAM (immunoreceptor tyrosine-based activation
motif)-containing cytoplasmic domain, CD3-c contributes its transmembrane and short
ectodomain (34), and is thus responsible for disulfide-bridge linkage-driven
homodimerization. Double chain TCR (dcTCR) could be similarly modified in order to
increase the probability of a TCR chain pairing with its desired antigen-specific counter-
part: for this purpose, the CD3-c chain not only endows the chimeric dcTCR_/`-c con-
struct with immediate signaling potency, but is also supposed to support preferential
pairing due to the dimerization motif located inside its transmembrane domain (35). As
an alternative, point mutations that introduce novel amino acid interactions at the inter-
face of the adjacent chains are able to manipulate the combinatorial preferences (unpub-
lished data, [36]). The study of TCR crystal structures facilitates the selection of amino
acid positions suited for manipulation of interchain pairing. This is further supported by
the observations that both the human and mouse TCR backbones can be closely super-
imposed and that the amino acid sequences, in particular those of the constant domains,
are highly conserved (37). However, the primary structure of the transferred TCR chains
has to be carefully modified in order to synergistically combine maintenance of function
with the implementation of preferential pairing. This should be kept in mind to avoid
any impairment of the signaling machinery by unreasonably omitting functionally rel-
evant motifs or even domains that are indispensable for proteinprotein interactions (38),
and secondary structural elements, such as glycosylation sites (39), believed to stabilize
the framework of the TCR/CD3 complex. Assembly with CD3 and the co-receptors CD4
or CD8 has to be taken into consideration (40), as low- to intermediate-affinity TCR, the
most prominent representatives of the TCR repertoire of an individual, require assis-
tance by these co-receptors to surmount an intrinsic threshold for T-cell activation (9,41).
2. Materials
1. 1 kb DNA ladder (New England Biolabs, Beverly, MA).
2. 100 bp DNA ladder (New England Biolabs).
3. 5'/3' RACE Kit (Roche, Basel, Switzerland).
4. Agarose DNA-grade (Biomol, Plymouth Meeting, PA).
5. Alkaline phosphatase, calf intestinal (CIP) (New England Biolabs).
6. Anti-CD3 antibody: Orthoclone OKT3 (Janssen-Cilag, Neuss, Germany).
7. Calcium Phosphate Transfection System (Invitrogen, Carlsbad, CA, former Life Tech-
nologies, cat. no. 18306-019).
8. Dextrane sulfate (Sigma, St. Louis, MO).
9. Freunds adjuvant (Difco, Detroit, MI).
10. DMEM (Cambrex, Walkersville, MD).
11. Dulbeccos PBS (Cambrex).
12. Dynabeads CD3/CD28 T-cell Expander (Dynal, Oslo, Norway).
13. Fetal calf serum (Cambrex).
14. Fugene 6 (Roche).
15. Geniticin (G418) (Sigma).
16. Gentamicin (Cambrex).
17. HEPES (Cambrex).
18. High Pure PCR Product Purification Kit (Roche).
19. Human AB serum (Bloodbank Clinical University Mainz, Germany).
20. Human CD8-PC5 (Beckman-Coulter, Miami, FL).
21. Interleukin-2 (Proleukin, Chiron, Emeryville, CA).
22. L-Glutamine (Cambrex).
23. Lipopolysaccharide (LPS) (Sigma).
24. Mastercycler Gradient (Eppendorf, Hamburg, Germany).
25. `-Mercaptoethanol (Sigma).
26. Murine V`6-antibody (Becton Dickinson, San Jose, CA).
27. Nonidet P-40, 10% (w/v) (Roche).
28. Oligonucleotides (MWG, Ebersberg, Germany).
29. Penicillin-Streptomycin (Cambrex).
30. Pfu DNA Polymerase (Stratagene, La Jolla, CA).
31. Platinum Pfx DNA Polymerase (Invitrogen).
32. pcDNA3.1() (Invitrogen).
33. Plasmocin (Invivogen, San Diego, CA).
34. pNEB193 (New England Biolabs).
35. Polybrene (Sigma).
36. Puromycin (Sigma).
37. Restriction enzymes (New England Biolabs).
38. RNase-Inhibitor (Roche).
39. Rneasy Mini Kit (Qiagen,Valencia, CA).
40. RPMI without phosphate (PAA, Pasching, Austria).
41. RPMI (Cambrex).
42. T4 DNA Ligase (New England Biolabs).
43. T4 DNA Polymerase (New England Biolabs).
44. T4 Polynucleotide Kinase (New England Biolabs).
45. Titan One Tube RT-PCR System (Roche).
46. RetroNectin (TaKaRa, Shiga, Japan).
Chemicals, unless otherwise stated, were purchased from Sigma; 10X reaction buffers of
the respective enzymes were provided by the supplier.
3. Methods
3.1. Generation of High-Avidity T-Cells
With Specificity for Human TAA in HLA-Transgenic Mice
A2Kb-transgenic mice are injected subcutaneously at the base of the tail with 100
g of an A2-presented synthetic peptide and 150 g of an I-Ab-binding synthetic T-
helper peptide representing residues 128140 of the hepatitis B virus core protein (42)
emulsified in 150 L of incomplete Freunds adjuvant. Seven days later, spleen cells
(1 106/mL) of nonimmunized A2(Kb)-transgenic mice are activated by 25 g/mL of
LPS and 7 g/mL of dextrane sulfate in an upright T75 tissue flask. After 3 d, LPS-
activated spleen cells are irradiated (3000 rad). Ten days following immunization,
spleen cells of primed mice are cultured with irradiated A2(Kb)-transgenic LPS-acti-
vated spleen-cell stimulators that are pulsed with the respective peptide at 5 g/mL
and human `2-microglobulin at 10 g/mL in complete RPMI medium (RPMI 1640
containing 10% [v/v] fetal calf serum, 25 mM HEPES, 2 mM glutamine, 5 10-5 M `-
mercaptoethanol, and 50 g/mL gentamicin, [15]). The resultant polyclonal T-cell
lines are weekly restimulated in 2 mL in a 24-well plate with irradiated (20,000 rad)
JA2 (Jurkat cells transfected with HLA-A*0201 cDNA) stimulators (0.5 106/well)
that had been pulsed with 5 g of the primary stimulating peptide, irradiated C57BL/
6 spleen feeder cells (6 106/well), and 210% (v/v) of rat concanavalin A superna-
tant (43). T-cells are cloned by limiting dilution.
3.2. Isolation and Cloning of Murine TCR

The _- and `-T-cell receptor chains (TCR_, TCR`) are 280-310 amino acids in
length. Each of the upstream 130 amino acids comprise the variable domains that
form the TCR_/` heterodimeric complex and specifically recognize the MHC-pre-
sented antigen. The constant domains are invariant and can be classified as C_, C`1,
and C`2, the latter differing by a few non-silent base sequence alterations at the very
end of the reading frame (44). The variable regions are highly degenerated and can
be classified into distinct subfamilies (IMGT, the international ImMunoGeneTics
information system http://imgt.cines.fr; Initiator and coordinator: Marie-Paule
Lefranc, Montpellier, France [4548]). The signal peptide-encoding sequences vary
as well. A time-consuming approach is the development of degenerate primers to
cover as much as possible of the whole panel of more than 20 subfamily sequences
of each TCR chain. This strategy has initially been designed for the identification of
both TCR chains by RT-PCR, but finally led to success only in case of the TCR`
chain. Only a non-functional TCR_ chain has been identified, originating from the
leaky allelic exclusion mechanism for TCR_ chains (49). This _ chain had a frame-
shift mutation within the CDR3 segment, generated by genetic rearrangements
inside the TCR_ locus during thymic maturation. The identification of this TCR
domain likely resulted from the bias of the developed degenerate primer for the 5'
sequence of the nonfunctional TCR reading frame. An RT-PCR protocol will be
provided for the isolation of the invariant C_ and C` regions of TCR_/` (Subhead-
ing 3.2.3.).
Fig. 2. Strategy for cloning full-length TCR_ and TCR` using 5'-RACE-PCR as illustrated
for both chains in a superimposed fashion. Base sequence lengths of both TCR are indicated:
the constant region of TCR` is more than 100 bases longer (indicated by a dashed box C`) than
C_. The variable regions are almost identical in size (400 bases). The cloning procedure of
each chain is composed of reverse transcription (primer SP1_/`-AscI), polyA tailing, and two
subsequent nested 5'-RACE-PCRs (primer pair oligo-dT-anchor / SP2_/` and anchor-SalI/
SP3_/`-AscI). Relevant enzyme restriction sites, whether (AscI, EcoRI, SalI) or not (NcoI)
being introduced by PCR primers, are highlighted. Their relative positions are indicated by
dashed lines that end in rhombic symbols. The truncated TCR are cloned via SalI/AscI in
pNEB193. In parallel, each full-length constant region is amplified by RT reaction (SP1_/`-
AscI) and one-step PCR (primer pair TCR_/`-con-EcoRI/SP1_/`-AscI), and cloned via EcoRI/
AscI in pNEB193. Due to an extended overlap that harbors a common NcoI site, subcloning of
an NcoI/AscI fragment finally complements (dotted boxes) the reading frame (RF) of each
truncated TCR_/`. The NcoI site to be silenced at almost the 3'-end of the C`2 RF is indicated
by square brackets.
Thus, a strategy had to be chosen to match all TCR chains of a T-cell clone. A
straight-forward technique to isolate full-length coding sequences is the rapid ampli-
fication of cDNA ends (RACE-PCR ; Fig. 2): mRNA is transcribed in vitro into a
complementary cDNA strand (primer SP1). The 3' moiety that ends up with the non-
translated region of unknown base sequence is extended with a short homopolymeric
dA-oligonucleotide by the terminal transferase reaction. This specific tail serves as
template for a specific complementary oligonucleotide (primer oligo-dT-anchor),
harboring several restriction sites at its 5' end, to enzymatically synthesize the second
strand cDNA with a thermostable DNA polymerase. In combination with a reverse
primer (SP2) binding to the 3' end of the invariant region, but nested to SP1 to
strengthen specificity, the double-stranded DNAs are amplified. The first PCR prod-
uct may be further amplified by a nested second PCR (Anchor primer/SP3), albeit
this will again truncate the sequence of the constant domain (Note 1) and will necessi-
tate the independent cloning of the constant regions (Subheading 3.2.3.). A unique
NcoI site in the overlapping region of both DNA fragments enables the rapid
subcloning of full-length TCR. Using this approach, each TCR chain has been inserted
into the expression vector pcDNA3.1() (Invitrogen) by SalI at its 5' and by BamHI at
its 3' end. The protocol and the 5' primers have basically been used as outlined in the
5'/3'-RACE Kit from Roche, whereas the 3' primers SP1SP3 have been assigned to
each TCR_/` chain individually.
3.2.1. RNA Preparation

The preparation of fully matured cytoplasmic mRNA has been performed as described
in the protocol for the isolation of cytoplasmic RNA from the cytoplasm of animal cells
of the Rneasy Mini Kit (Qiagen). To determine the functional reading frames of the TCR,
the mRNA has to be fully processed by excising the non-coding intron sequences out of
the exons. For this purpose, the detergent NP-40 has been used in order to solubilize
exclusively the cell membrane and to separate the cytoplasmic content from the nuclei
(see Note 2). 415 106 T-cells of a mouse T-cell clone have been lysed in 0.5% (w/v) of
the nonionic detergent Nonidet-P-40 (NP-40; Roche) containing lysis buffer; formation
of viscous solution by the release of chromosomal DNA is avoided. The yield of RNA
has been optimized by raising the recommended time for solubilization from 5 min to 30
min on ice. On average, 4 106 murine T-cells produce 15 g of RNA, which is sufficient
for an experimental series of reverse transcription and PCR analysis. Raising the absolute
T-cell number progressively improves RNA yield. RNA has been frozen in 5-g aliquots
and stored at 20C for short-term analysis to assess PCR product specificity at five
separate annealing temperatures in RT/RACE-PCR assays (see Note 3).
3.2.2. RACE-PCR
RACE-PCR includes the reverse transcription of mRNA into cDNA and subse-
quently the PCR amplification thereof to yield sufficient amounts of DNA for cloning
and further characterization.
The following protocol has been applied to both the generation of the TCR_ chain as
well as the TCR` chain. The RNA was prepared as described under Subheading 3.2.1.
and diluted to 15 g/10 L RNase-free water. Reverse transcription was performed in
20 L with 9 L RNA, 4 L 5X cDNA synthesis buffer (250 mM Tris-HCl pH 8.5
[20C], 40 mM MgCl2, 150 mM KCl, 5 mM DTT), 2 L dNTP (stock: 10 mM each),
1 L RNase-Inhibitor (40 U/L from human Placenta [Roche]), 1 L rev-TCR_-SP1-
AscI (12.5 M, 5'- AGG CGC GCC TTC AAC TGG ACC ACA GCC TCA GCG T -3')
or rev-TCR`-SP1-AscI (12.5 M, 5'- AGG CGC GCC TTC AGG AAT TYT TTY TYT
TGA CCA T -3' wherein Y = C/T and TCA refers to the stop codon), 1 L AMV-RT
(20 U/L), 2 L RNase-free water (see Note 4). The TCR-specific reverse primers are
identical to those that have been used to clone full length constant regions (Subheading
3.2.3.). Base composition and length of primers have to be considered for optimal tem-
perature in RT reaction. This has been worked out for a PCR device as follows: the lid
was equilibrated to 102C; a hold step at 55C enables positioning of 0.2-mL PCR
tubes of the RT-assay, followed by a 60-min reverse transcription at 55C, and 10 min
65C RNase heat inactivation of AMV-RT. The samples were equilibrated to room tem-
perature for subsequent purification to separate them from excess primers that would
otherwise interfere with amplification steps. The purification was performed as described
in the High Pure PCR Product Purification Kit (Roche).
The tailing of the first strand cDNA took place in a PCR device preceded by a short
heat denaturation step to make the 3'-noncoding region accessible to terminal transferase:
19 L of the purified cDNA was added to 2.5 L tailing 10X reaction buffer (100 mM
Tris-HCl pH 8.3 [20C], 15 mM MgCl2, 500 mM KCl), 2.5 L dATP (2 mM dATP in
Tris-HCl pH 8.3 [20C]) at 94C for 3 min and then rapidly cooled to 4C to minimize
rehybridization. The sample was prewarmed to 37C before addition of 1 L of terminal
transferase (10 U/L, hold step) and incubated for 20 min to allow for enzymatic synthe-
sis of the homopolymeric deoxy-adenosine tail. A 10-min heat inactivation step at 70C
finishes the tailing reaction. 5-L aliquots can be directly used for PCR amplification in
a temperature-dependent annealing gradient (see Note 5). For this purpose, 5-L aliquots
of the tailing product were combined with 45 L of a master mix composed of 180 L
redistilled water, 5 L of the SalI-oligo-dT-anchor primer (5'- GAC CAC GCG TAT
CGA TGT CGA CTT TTT TTT TTT TTT TTV -3' wherein V = A/C/G; stock: 37.5 M),
which is complementary to the enzymatically appended oligo-dA tail, 5 L of reverse
TCR-specific primers, rev-TCR_-SP2 (5'- AGG TTC ATA TCT GTT TCA A -3'; stock
12.5 M) or rev-TCR`-SP2 (5'- GGA TCT CAT AGA GGA TGG T -3', 12.5 M), 5 L
dNTP (10 mM each), 25 L Pfu-DNA-polymerase reaction buffer (10X stock: 200 mM
Tris-HCl [pH 8.8], 100 mM KCl, 100 mM [NH4]2SO4, 20 mM MgSO4, 1 % Triton X-100,
1 mg/mL BSA), and 5 L Pfu-DNA polymerase (2.5 U/L). The five vials evenly span
the temperature gradient and were run on the thermal cycler Mastercycler Gradient
(Eppendorf): After an initial denaturation step at 94C for 120 s, 10 PCR cycles follow,
consisting of denaturation for 30 s at 94C, slow (1C/s) annealing at 45C 10C, exten-
sion for 60 s at 72C. This is followed by 25 cycles of denaturation for 30 s at 94C, slow
annealing at 45C 10C, extension for 60 s at 72C in the eleventh cycle that will be
prolonged by 20 s for each subsequent cycle to compensate for nucleotide depletion, and
finally a 7-min extension step to blunt-end all 3' termini for efficient cloning. The exten-
sion temperature had to be raised to 72C to meet the temperature optimum criteria of
the high-fidelity Pfu-DNA polymerase (Stratagene).
In order to strengthen specificity and product yield, 1 L of the PCR product was
submitted to a second round of amplification with nested reverse primers: for TCR_, a
primer harboring an AscI cloning site, rev-TCR_-SP3-AscI (5'- AGG CGC GCC TGG
CGT TGG TCT CTT TGA A -3'); for TCR`, rev-TCR`-SP3-AscI (5'- AGG CGC GCC
TGG CCA CTT GTC CTC CTC TGA A -3') in combination with the forward SalI-
containing anchor primer (5'- GAC CAC GCG TAT CGA TGT CGA C -3') was used
as described for the first PCR. One-fifth of the PCR product was analyzed on a 1.0%
agarose gel (Fig. 3); the remainder was processed for cloning via SalI/AscI and NcoI/
AscI to produce full-length TCR as outlined under Subheading 3.2.4.
Fig. 3. RACE-PCR (1st PCR) and subsequent 3'-nested PCR (2nd PCR) for mouse TCR_ and
` chains as outlined in Fig. 2. The different annealing temperatures are indicated. The 1st PCR
produced faint PCR bands, whereas the nested PCR of the 1st PCR product amplified trace
amounts for virtually all temperatures. The PCR products were longer than the hypothetical
length of the coding regions for TCR_ and -` due to the integration of the 5'-non-coding regions.
3.2.3. Isolation of the Invariant TCR_ and -` Regions by RT-PCR

RT-PCR basically consists of the reverse transcription of isolated mRNA to first
strand cDNA and subsequent PCR amplification(s), provided that the flanking sequences
of the gene of interest are known or can at least be covered by degenerate primers. The
relevant components have been assembled in the Titan One Tube RT-PCR System
(Roche).
The optimal production of PCR fragments has been further refined by five indi-
vidual temperatures in the annealing step, as already specified for RACE-PCR. There-
fore, a master mix has been pipetted (250 L) that was aliquoted into five 0.2-mL PCR
tubes and placed evenly among the temperature gradient (20C) of a thermal cycler
(Mastercycler gradient, Eppendorf). The one-step RT-PCR comprises two master
mixes: Mastermix 1 was pipetted in the following sequence: 90 L RNAse free water
(DEPC-treated and autoclaved twice), 5 L dNTP (25mM stock), 2.5 L for-C`1/2-
or C_-specific primer (100 pmol/L stock, for sequences see below), 2.5 L rev-
C`1/2- or C_-specific primer (100 pmol/L stock), 12.5 L DTT (100 mM stock),
2.5 L RNase inhibitor (40U/L stock), and 10 L RNA (5 g). The source of RNA
was identical to the mouse T-cell clone preparation, as already mentioned (RACE-
PCR). The components are mixed with RNase-free plasticware and stored on ice. The
Mastermix 2 is composed of 10 L RNase-free water, 50 L 5X RT-PCR buffer,
60 L MgCl2 (25 mM stock), and 5 L AMV-RT/Taq/Pwo-DNA-polymerase enzyme
mix. The Mastermix 2 is homogenized as well, then mixed together with Mastermix 1,
and aliquoted to 50 L into individual PCR tubes.
The thermal cycler is programmed to continuously link reverse transcription with
polymerase reaction: initially, the device is pre-equilibrated to the preferred tempera-
ture for the RT reaction (55C, hold function). The five PCR tubes are placed among
the temperature gradient in an 8 12-well format. Reverse transcription proceeds for
45 min and, after a 2-min heat-inactivation step, is followed by the PCR reaction. The
default cycling program consists of 10 cycles of denaturation at 94C for 30 s, anneal-
ing at 45 10C for 60 s, and extension at 68C for 90 s. To compensate for nucleotide
depletion, 25 cycles of denaturation at 94C for 30 s, annealing at 45 10C for 60 s,
and extension at 68C for 90 s follows, prolonged by 4 s for each cycle. A final exten-
sion at 68C for 7 min and cooling to 4C (hold function) terminates the protocol.
RT-PCR with for-TCR`-con-EcoRI (5'- GGA ATT CCG AGG ATC TGA GAA
ATG TGA -3') and the degenerated rev-TCR`-con-AscI (5'- AGG CGC GCC TTC
AGG AAT TYT TTY TYT TGA CCA T -3' wherein Y = C/T) to cover the constant
regions C`1 and C`2 (see Note 6), elicited a strong PCR band of 0.5 kB without subse-
quent nested PCR. The constant TCR_ segment has been isolated in parallel from the
same RNA source with the primers for-TCR_-con-EcoRI (5'- GGA ATT CCC CAG
AAC CCA GAA CCT GCT GTG TA -3') and rev-TCR_-con-AscI (5'- AGG CGC GCC
TTC AAC TGG ACC ACA GCC TCA GCG T -3').
3.2.4. Cloning of TCR-Related PCR Products

Cloning of DNA fragments was performed according to common protocols and
therefore is only summarized (explicit guidelines in (50); Note 7): One-fifth of the
(nested) PCR product was analyzed on a regular 1% (w/v) agarose gel in 0.5X TBE
electrophoresis buffer. The PCR product was purified by the High Pure PCR Product
Purification Kit (Roche) to eliminate excess primer and was spectroscopically quanti-
fied by UV light absorption at 260 nm. At least 2 g were used for cloning into
pNEB193 (NEB) by subsequent SalI (or EcoRI) and AscI restriction (NEB) of frag-
ment and vector, dephosphorylation of the vector by calf intestinal phosphatase (CIP),
and directional ligation of the fragment into the cut vector with T4 DNA polymerase.
Transformation was performed with the CaCl2-competent Escherichia coli strain XL1-
Blue (Stratagene). Small-scale plasmid preparations of individual clones were pre-
pared with the QIAprep Spin Miniprep Kit (Qiagen). TCR sequences have been
confirmed in both directions by a commercial supplier.
3.2.5. Subcloning of TCR Fragments

to Yield Full-Length TCR Reading Frames
The outlined concept of RACE-PCR did not provide a complete coding sequence
for the TCR_/` chains. The TCR clones comprising the almost full-length TCR (see
Subheading 3.2.2.) and those encoding the constant domains (3.2.3.) share a common
NcoI site inside their (residual) constant regions (Fig. 2). In the case of TCR_, simple
subcloning of an NcoI/AscI-digested C_ fragment into the NcoI/AscI-cut vector har-
boring the truncated TCR_ results in a full-length TCR_ chain ready for functional
expression studies. Cloning of TCR` turned out to be a little more complex due to an
additional internal NcoI site at almost the 3' end of the C`2 region; this was dealt with
by partial enzyme restriction to produce predominantly a complete NcoI/AscI frag-
ment instead of a truncated NcoI/NcoI fragment (see Note 8).
3.2.6. Subcloning of TCR_/` Genes

in Retroviral Expression Vectors
The full-length TCR have been introduced into the retroviral vector system pBullet by
NcoI/BamHI ligation. This vector was kindly provided by Ralph Willemsen (Depart-
ment of Clinical and Tumor Immunology, Daniel den Hoed Cancer Center, Rotterdam,
The Netherlands; refs. 32,51,52). The vector is ideally suited for transient transfection of
the packaging cell line 293T and, depending on the envelope used, for reproducible trans-
duction of species-specific T-cells.
The pBullet vector system belongs to the class of splicing vectors that make use of
splice donor and acceptor sites to improve expression of the gene of interest (53). An
NcoI site located directly downstream of the splice acceptor element allows the inser-
tion of the transgene into the same sequence context as compared to the wild-type
retroviral envelope protein. This concept requires the integration of the start codon
ATG into the NcoI-site CCATGG.... Occasionally, this will modify the proximal
amino acid as an integral part of the signal peptide.
The transgenes have been amplified by PCR and their 5' and 3' ends trimmed to NcoI
and BamHI. In the case of NcoI, the internal NcoI sites of mouse TCR disable direct
digestion; the vector was therefore cut with NcoI, filled to blunt end with T4-DNA
polymerase, and subsequently BamHI digested and dephosphorylated (50). The for-
ward primer for PCR was designed at its 5' end in such a way as to amend the missing
G... to the blunt-ended vector ...CCATG to restore NcoI ...CCATGG... in the recombi-
nant vector. In addition, the 5' end was chemically (oligonucleotide synthesis) and/or
enzymatically phosphorylated by T4-DNA-polynucleotide kinase (PNK, ref. 50). This
cloning procedure is similar to directional insertion via blunt/sticky end ligation (see
Note 9).
3.3. Modification of TCR Constructs

3.3.1. Humanization of TCR
Each chain of the heterodimeric TCR_/` comprises two immunglobulin-like folded
domains, of which the variable domain harbors the three CDR loops responsible for the
recognition of the peptide/MHC complex. The constant domain is invariant and prima-
rily serves as distance holder to expose the variable domain to the peptide/MHC com-
plex of the antigen-presenting cell (APC) and to enable the association to the CD3
complex and to the coreceptors CD4 or CD8. A disulfide bridge close to the transmem-
brane region covalently links both chains. A transmembrane region with a very short
cytoplasmic tail provides membrane anchoring and accurate association with CD3 sub-
units. Charged pairs of residues inside the hydrophobic core of the TCR/CD3 complex
direct the asymmetric assembly of the subunits (54).
Although mouse and human TCR are of considerable sequence identity with respect
to their constant domains, the usefulness of murine TCR in man may be limited by their
potential immunogenicity. Both the high degree of sequence and structural homology
and the invariant character of mouse constant domains permit their entire replacement
by human C_/` regions without interfering with TCR function. Humanization beyond
that initial step of domain replacement (deimmunization) requires several rounds
of mutational refinements with regard to structural and functional aspects of TCR
biology.
The variable and constant domains of either TCR_/` chain, each composed of a
twisted anti-parallel ` sheet, are linked by an extended loop and interact with each other
via a few residues. The interactions of both the juxtaposed variable V_ and V` as well as
the constant domains C_ and C` are complex, and the tight interactions of the C_/C`
pair in particular govern interchain affinity (23). V_/C` and V`/C_ interactions are
almost absent and thus facilitate domain replacement.
On a genetic basis, an approach towards humanization took advantage of singular
restriction sites located inside the loop sequence that joins the variable and constant
domains to produce chimerized, so-called partially humanized TCR; in the case of the
mouse TCR_ chain, a conserved AlwNI-site CAGNNNCTG (wherein N = A/C/G/T),
translationally located aminoterminally to the first C_ `-strand and embedded inside
the conserved PEPA sequence, allows the domain exchange (Fig. 4A).
A full-length human constant C_ has been isolated irrespective of its subfamily
specificity by coupled RT-PCR/nested PCR from the human Jurkat cell line. The RT-
PCR protocol was identical to the isolation of huC_ and huC`, and has been slightly
modified as described under Subheading 3.2.3. by increasing the annealing tempera-
ture to 55 10C (RT-PCR: primer pair for_huC_TCR: [5'- ATA TCC AGA ACC
CTG ACC CT -3'] / rev_huC_TCR [5'- GGG AGC ACA GGC TGT CTT ACA -3']).
For nested PCR, the RT-PCR protocol has been adopted, omitting the RT reaction and
using the high-fidelity Pfu-DNA polymerase whose optimal extension temperature
lies at 72C (nested PCR: primer pair for_huC_TCR_AlwNI [5'- CAG AAC CCA
GAA CCT GCC GTG TAC CAG CTG AGA GAC TCT A -3'] / rev_huC__BamHI
[5'- CG GGAT CCT CAG CTG GAC CAC AGC CGC AGC GTC ATG A -3']). In
order to get a full-length constant region, the reverse primers for RT-PCR have been
located at the 3'-noncoding region close to the stop codons of TCR_ (as for TCR`).
By introducing two mutations into a primer suited for nested PCR, an AlwNI site has
been designed at the corresponding position in the human as compared to the mouse
residue sequence (Fig. 4A).
242
242
Fig. 4. (A) Nucleotide and amino acid sequence alignment of a mouse MDM2(8188)-specific and a human gp100(280288)-specific
Voss, Kuball, and Theobald

TCR_ at the related regions of domain replacement to exemplify the high degree of amino acid homology. Highly conserved amino acids are
indicated in bold. GxG refers to the C-terminal consensus sequence of the CDR3 loop, and IQNPxPAVY to the start of the highly conserved C_
domain. Additionally, boxed residues delineate the amino acid sequence of the resulting chimerized molecule. Differing residues are under-
lined. The AlwNI restriction site CAGNNNCTG (italic for the recognized bases, bold for the whole sequence) is located close to the start of the
C_ domain. Nucleotides that have to be changed in the PCR primer for complementarity to the mouse V_ region are underlined. (B) Nucleotide
and amino acid sequence alignment of an A2.1-restricted mouse MDM2(8188)-specific and a human gp100(280288)-specific TCR`. Abbre-
viations are as described in (A). EDL refers to the start of a loop connecting both domains, followed by the highly conserved C` domain. The
BstYI restriction site PuGATCPy is located within the EDL consensus sequence. For cloning strategy refer to Subheading 3.3.1.
Due to internal AlwNI sites in the vector, DNA fragment ligation had to be per-
formed: the cloned mouse TCR was digested with EcoRI/AlwNI and ligated to the
AlwNI/BamHI-cut nested PCR product. The heteromultimerized products were sepa-
rated from the unwanted homomultimers (generated as an artifact by head to tail
multimerization) by redigestion with EcoRI/BamHI to yield the chimerized DNA frag-
ment (see Note 10). The resulting 0.83-kB band was ligated into a cloning vector
(Subheading 3.2.4.) and subsequently cloned via NcoI/BamHI into the retroviral
pBullet vector (Subheading 3.2.6.).
The generation of the chimerized TCR` chain has been designed the same way
(Fig. 4B): a nonunique BstYI site PuGATCPy (Pu = purine-base/Py = pyrimidine-
base) has been identified in the DNA sequence of the conserved EDL stretch in the inter-
vening loop. The additional location of BstYI inside the constant domain of human TCR
requires the blunt-end ligation of the nested PCR product to circumvent truncation: the
BstYI fragment encoding the mouse variable domain has been blunt-ended by the T4-
DNA polymerase fill reaction, and subsequently cut with EcoRI. The resulting EcoRI/
[BstYI]-fragment has been blunt-ligated to the nested PCR fragment 5'- p-[BstYI]-huC`-
BamHI -3', thereby restoring the nonpalindromic BstYI recognition sequence: the T4-
DNA polymerase reaction results in a GGATC-3'-end of muV` that will complement
with the 5'-T-end of PCR-amplified huC` to a BstYI site (see Note 11). Before liga-
tion, the blunt PCR product has to be enzymatically phosphorylated by T4 DNA-PNK
(Subheading 3.2.6.). The mixed ligation product has to be recut with EcoRI/BamHI
(for the same reasons as mentioned above) and cloned into appropriate expression vec-
tors for functional studies (RT-PCR: for_huC`TCR [5'- TAG AGG ACC TGA ACA
AGG TGT T -3'] / rev_huC`TCR [5'- AAC CAT GAC GGG TTA GAA GCT CCT A -
3']; nested PCR: for_huC`TCR_[BstYI] [5'- TGA ACA AGG TGT TCC CAC CCG
A -3']/ rev_huC`TCR_BamHI [5'- CGG GAT CCT CAG AAA TCC TTT CTC TTG
ACC ATG GCC ATC A -3']). The related nucleotide sequences for human TCR_/` has
been taken from Genbank of NCBI (National Center for Biotechnology Information,
Bethesda, MD; Entrez nucleotide in http://www.ncbi.nlm.nih.gov/) and refers to the
entry M12423 for TCR_ in Jurkat and M15564 as a clonal representative of a TCR`
sequence that will be found in Jurkat cell line. There exist a few sequence differences
(due to 2 C` alleles) when comparing multiple TCR` sequences (e.g., to gp100(280
288)-specific TCR` in Fig. 4) in this region relevant for primer design, but this does not
impair the approach.
Due to the location of the restriction sites AlwNI and BstYI centered in the loops
that link the variable and constant domains, they are found in all murine TCRs irre-
spective of their subfamily membership, including C`1 and C`2 domains. The human-
ization approach is affected by these restriction sites when additionally located inside
the coding segments of the variable domains: this is likely to occur, in particular, with
respect to the highly degenerate CDR3-loop segments. As a consequence, truncated
variable regions have to be avoided by partial restriction.
More elaborate PCR techniques work with an intermediate megaprimer: in a first
PCR, this strategy produces a large fragment of the gene to be fused (i.e., the mouse
variable region) by utilizing an oligonucleotide that anneals to a short stretch of bases
Fig. 5. Ligation PCR consists of two consecutive PCRs: the first PCR produces the double-
stranded megaprimer (SP-V_-Li-V`) that downstream is complementary to the 5'-end of the
V` region (V`). An aliquot of the 1st PCR product that already includes the 2nd template is
transferred to the second PCR mix containing the nested primers. Intermediate purification is
usually not necessary, as the nested primers are of sufficient specificity.
at the sequence of interest (i.e., the ample region between the variable and the constant
region). The chosen position has to be of sufficient homology with the 5' end of the
second gene fragment (i.e., the human constant region) to allow annealing of the
megaprimer with its 3' end in a subsequent PCR. The advantages of this approach are
its independence from restriction enzymes and its ability to simply join gene segments
at a precise nucleotide and at any amino acid position (see also Subheading 3.3.2.).
3.3.2. Generation of Single-Chain TCR

Single-chain TCR (scTCR) are chimeric TCR consisting of the variable V_ and
V` domains of an antigen-specific TCR covalently linked via a 1520 amino acid
sequence, the so-called linker, to bridge the C-terminal of the preceding V_ domain
to the N-terminal of the V` domain. The linkage should provide sufficient length to
maintain the sterical orientation between the domains and to allow for some kind of
flexibility required for antigen recognition and signal transduction into the cytoplasm
(55). The linker has to be resistant against proteolysis (29), and thus must have an
amino acid composition that combines flexibility, solubility, thermodynamic stabil-
ity, and a sustainment of defined oligomeric state (30); a triplicate redundancy of the
glycine-rich motif GlyGlyGlyGlySer proved to be effective for generating functional
scTCR and scFv fragments of antibodies. The variable V` domain is followed by the
C` domain to displace the antigen-recognition domains far enough from the polar
lipid of the cell membrane. The linker has to be specifically inserted at the borders of
the V_ and V` domains, thus at the end of the last ` strand of V_ (see also Figs. 4 &
5) and the beginning of V`. The flanking base sequences are likely non-homologous
and can therefore be easily linked by a modified PCR that utilizes a chimeric primer
consisting of identical base sequences for successful priming to the domains as tem-
plates. In addition, any arbitrary sequence, in our case the 218 linker sequence ([29];
NH2-Gly Ser Thr Ser Gly Ser Gly Lys Pro Gly Ser Gly Glu Gly Ser Thr Lys Gly-
COOH), can be inserted between the two sequence-related parts. Because the up-
stream V_ region is much smaller than the residual construct, it is more convenient to
design the chimeric primer in reverse and use it in combination with a forward primer
specific for the signal peptide sequence that precedes V_. In this case, this leads to
the production of conceivably the shortest megaprimer (Fig. 5). The first PCR prod-
uct encodes the signal peptide of V_, V_, the artificial linker, and a short tail that
specifically recognizes the 5' end of V` directly downstream of the signal peptide
encoding sequence, as defined by the mouse TCR variable gene segment families
([56]; first PCR: for_pBullet [5'- GGA TAC ACG CCG CCC ACG TGA -3'] /
rev_pBullet_IRES [5'- GAG GGA GTA CTC ACC CCA ACA -3']; chimeric primer:
rev_Chim_218 [5'- TTT GGG TGT CTG AGT AAT GAT GCC ACC GCC CTT
GGT GCT GCC CTC GCC GCT GCC GGG TTT GCC GCT ACC GGA AGT AGA
GCC TGC TAT AAC CTT CAG TCT GGT ACC TGC -3']). A nested PCR using the
double-stranded megaprimer and a nested primer as forward primer mix combined
with a reverse primer specific for the 3' end of the C` region will yield a full-length
scTCR of any desired amino acid sequence and linkage (see Note 12). Second PCR:
megaprimer: first PCR product, comprising 5'- SP-V_-Li-V`'-3'; primer: for_SP-
V__[NcoI] (5'- GAA AGG CTG CTG TGC TCT CT -3') / rev_C`_BamHI (5'- CGG
GAT CCG CTC AGG AAT TTT TTT TCT TGA CCA T -3'). Even minimal changes
in the annealing temperature significantly influence the amount of PCR product. In
case of trouble, base composition, GC content, and possible secondary structures that
may form inside the chimeric primer have to be taken into consideration. The second
PCR uses nested primers encoding enzyme restriction sites that turn the PCR product
suitable for digestion and efficient cloning into expression vectors. We cloned the
PCR product directly with NcoI/BamHI into the retroviral transfer vector pBullet
(32,51,52), which depends on NcoI site-specific insertion at the 5' end of the scTCR
transgene. We have modified the ligation PCR method somewhat to conveniently
yield scTCR in two consecutive PCRs. A primer pair that hybridizes to plasmid sequences
flanking the templates amplifies them and turns them into more ideally suited PCR
templates for the subsequent, nested PCR. As a prerequisite, both templates have to
be embedded in the same band of vectors. Only 2 pmol of chimeric primer are added,
in contrast to 50 pmol for the flanking primers, in order to improve specificity of the
intermediate megaprimer product (see Note 13). For a sample at a particular annealing
temperature, 1 L (100 ng) of each plasmid encoding the templates to be chimerized
are pipetted to 0.5 L of the plasmid-specific forward and reverse primer (each 100
pmol/L), 2 L chimeric primer (1 pmol/L), 0.5 L dNTP (each 10 mM), 1 L
MgSO4 (50 mM), 10X Pfx amplification buffer (Invitrogen), 10X Pfx enhancer solu-
tion, 33 L sterile water, and 0.5 L Pfx polymerase (2.5 U/L). The cycling param-
eters were chosen to be 2 min at 94C, followed by 34 cycles of denaturation for 30 s
at 94C, annealing for 45 s at 55 10C, extension for 60 s/kB template at 68C, and
Fig. 6. Ligation PCR for MDM2-specific TCR_ and TCR` as templates to generate a scTCR
with an 18-mer linker. One-fifth of each PCR has been loaded on a regular 1% agarose gel. In
general, higher annealing temperatures (65C) produce sufficient megaprimer (0.5 kB) for sub-
sequent PCR to gain a full-length scTCR (1.3 kB). Although the first PCR does not produce
even a faint band resulting from the scTCR construct, the second PCR visibly enhances ampli-
fication of tiny amounts of the scTCR construct. This gel photography was kindly provided by
Simone Thomas (Theobald Laboratory, Department of Hematology and Oncology, Johannes
Gutenberg University, Mainz, Germany). M1 = 100-bp marker, M2 = 1-kB marker (NEB).
final extension for 7 min at 68C. For the nested PCR, template-specific primers
were applied, which included blunt NcoI and BamHI restriction sites in the forward
and reverse direction, respectively. The chimeric primer and the plasmid templates
were replaced by water, and for each corresponding temperature, 1 L of the first
PCR product was added. Again, a temperature gradient was used for the same three
independent temperatures (see Note 14). Cycling parameters were identical to the
first PCR. Generally, higher annealing temperatures (>60C) resulted in a unique
ethidium bromidestained band of almost 1.3 kB, specific for the scTCR (Fig. 6).
Ligation took place as described under Subheading 3.2.6.
3.4. Heterologous Expression of Mouse TCR in Human T-Cells
3.4.1. Selection Markers for In Vitro Enrichment of TCR-Transduced T-Cells
For in vitro analysis of defined TCR-transduced T-cells and evaluation of subtle
TCR modifications, it is essential to analyze pure TCR-transduced T-cell subpopula-
tions. Therefore, TCR cassettes are linked by an internal ribosome entry site (IRES)
to a resistance marker. The IRES element located between two genes allows the bicis-
tronic mRNA transcripts to be translated simultaneously. Because both the protein of

interest and the antibiotic resistance protein are produced from a single transcript, it
is more likely that a single cell will express either protein. Furthermore, cells will
express high levels of the desired protein when an attenuated IRES is used that mini-
mizes translation of the selection marker relative to the gene of interest and thus
raises the selective pressure to survive. Hence, high doses of antibiotics will select
for cells expressing higher levels of the gene of interest. We use the IRES-puro2 and
-neo cassettes (BD-Clontech), which contain the IRES element of the encephalo-
myocarditis virus (ECMV). For this purpose, TCR transgenes are cloned upstream
and the selection cassettes downstream of the IRES element (Fig. 7; [57,58]). For
routine application, the IRES cassettes were integrated into the retroviral vector
pBullet in order to insert and to select for any TCR chain (see Note 15). The insertion
of the selection cassettes between the LTRs reduces exposure to the antibiotics to one
week, subsequently ensuring stable transgene expression.
3.4.2. Transduction, Selection, and Expansion Procedure
The principal method relies on the transient transfection of the packaging cell line
293T and its derivative phoenix-ampho for rapid virus production. The transduction
protocol differs somewhat from that of Ralph Willemsen (32,51,52) or Garry P. Nolan
(see Note 16). The major advantage of this routine protocol is the reproducible yield
of virus titers in the range of 1 1061 107 GFU/mL (GFP forming units/mL), which
is a critical threshold for successful transduction (see Note 17). One may keep in mind
that each TCR chain is encoded on a separate plasmid (Fig. 7). Poor transfection will
drop the detection of transduced T-cells that have incorporated both transgenes at the
same time. The original calcium phosphate transfection method was more sensitive to
physicochemical changes (temperature, pH) than the Fugene transfection agent
(Roche), which is based on a proprietary blend of lipids:
1. Day 1. 293T and phoenix-ampho cell lines (ATCC) will be coseeded, each at 1.2 106, in
a T75 tissue flask (DMEM for 293T/phoenix-ampho [DMEM, 10% heat-inactivated FCS,
1X glutamine, 1X NEAA, 25 mM HEPES, 1X Pen/Strep]) (see Note 18).
2. Day 2. Medium will be removed and replaced (10 mL) by fresh DMEM 10% FCS. Four h
later, 60 L of Fugene 6 will be added to prewarmed (RT) 900 L DMEM (without FCS)
as well as 5 g of each of the plasmids: pHit60, pCOLT-GALV (kindly provided by R.
Willemsen; refs. 32,51,52), and both, the TCR_/IRESpuromycin- as well as the TCR`/
IRESneomycin-recombinant plasmid. The mix will be incubated for 20 min and then
slowly dropped into the medium of the gently swirled 293T/phoenix-ampho coculture
(see Note 19).
PBMCs will be seeded at 2 106/mL in 24-well plates (2 mL each well) and activated
for 2 d with soluble anti-CD3 antibody at 20 ng/mL in huRPMI/10% (medium composi-
tion for human T-cells: RPMI, phosphate-free, 10% pooled and heat-inactivated [30 min,
55C] human AB-serum, 1X glutamine, 25 mM HEPES, 1X Pen/Strep) (see Note 20).
3. Day 3. DMEM medium will be removed and replaced by 8 mL of huRPMI/10%.
4. Day 4. Retroviral supernatant of 293T/phoenix-ampho will be harvested and replaced
the same way as on d 3, and centrifuged for 1 min (324g) to remove nonadherent 293T
and phoenix-ampho cells. 48 106 of the pooled preactivated T-cells (depending on
availability) will be suspended in 48 mL of retroviral supernatant (1 106 T-cells/mL)
248
248
Voss, Kuball, and Theobald

Fig. 7. Vector design. Insertion of the selection cassettes into retroviral vectors to generate bicistronic mRNAs for coupled
translation of TCR_/puromcin-N-acetyl-transferase and TCR`/neomycin phosphotransferase. CMV = cytomegalovirus; ECMV-
IRES = internal ribosomal entry site of the encephalomyocarditis virus; kb = kilobases; LTR = long terminal repeats. Synthetic
intron refers to an mRNA-stabilizing element (BD/Clontech). Cloning sites are indicated. The polyA signal is located in the 3'
LTR.
adding Polybrene (4 g/mL) and IL2 (40100 U/mL) and seeded in one well of a 6-well
plate (Becton Dickinson Falcon, Cat. No. 353046), centrifuged at 30C for 60 min
(1300g), and co-incubated overnight at 37C in a humidified incubator (5% CO2),
thereby initiating the first cycle of retroviral transduction (see Note 21).
5. Day 5. T-cells will be separated from exhausted retroviral supernatant by centrifugation
(324g, 5 min, 21C) and suspended in freshly prepared supernatant. The second cycle is
identical with the first one (see Note 22).
6. Day 6. T-cells will be removed from old retroviral supernatant and supplemented with
fresh huRPMI/10% medium at an equal volume (48 mL). For sustained activation of
T-cells, 2 L CD3/28 beads/1 106 T-cells (Dynal) and 40100 U IL2/mL will be added.
7. Day 8. For selection of transduced T-cells, 600800 g/mL (f.c.) geniticin will be added.
8. Day 12. Five g/mL (f.c.) puromycin will be added.
9. Day 15. CD3/28 beads will be removed by magnetic depletion and cells will be cultured
at 0.5 106/mL in fresh huRPMI/10%, weekly supplemented with 2 L CD3/28 beads/
1 106 cells and 40100 U IL2/mL (up to twice a wk) (see Note 23).
This protocol allows rapid and efficient selection as well as expansion of TCR-
transduced T-cells up to some orders of magnitude. The cell numbers are sufficient to
isolate subpopulations (e.g., CD4+/CD8+) for further expansion and functional analy-
ses (see Fig. 8). However, the expansion of T-cells by CD3/28 beads selectively pro-
motes the expansion of CD28+ T-cells. Also, CD8+ T-cell growth is favored (400-fold
expansion within 6 wk, unpublished results). As the expansion of CD8+ T-cells corre-
lates with the amount of CD4+ T-cells in the culture, the expansion of T-cells without
any other cell-based feeder system is limited. Depending on the individual donor, cells
can be kept in culture for more than 3 mo. We recommend freezing an aliquot of
freshly selected T-cells at that early time for convenient reculturing.
4. Notes
1. Alternatively, one might try to exploit the invariant nature of the 3'-noncoding regions for
the primers SP1/2 and to position SP3 at the 3' end of the TCR reading frame. This would
enable the isolation of the full-length TCR gene, in particular TCR_, which is repre-
sented by only one constant region gene locus.
2. Mouse T-cells will be restimulated weekly with excess of irradiated C57BL/6-derived
spleen feeder cells and peptide-pulsed HLA-A2.1-transfected human Jurkat cells (JA2).
This will not impede the sufficiently homogenous purification of the RNA pool. Never-
theless, the mouse T-cells have to be carefully counted and morphologically distinguished.
3. In case of insufficient T-cell numbers (growth, availability), one might try to scale down
the input RNA since PCR is very sensitive.
4. Alternatively, control RNA and primer may be added as internal standards to control
reverse transcription and subsequent PCRs (for details, see protocol 5'/3'-RACE Kit
[Roche]).
5. As the amount of the tailing product is not a limiting factor, it is strongly recommended to
assay at least three different annealing temperatures in the range of 5565C. The optimal
annealing temperature turned out to be in the elevated range, since this tends to minimize
unspecific reactions without loss of the specific product. The same guideline applies to
the second PCRkeep the temperature gradient unchanged and deliver each sample of
the first PCR to the same temperature in the subsequent PCR. In case of trouble, one may
try to fine tune a PCR product of one particular temperature of the first PCR on a tem-
Fig. 8. Human T-cells transduced with empty vector (mock) (A), green fluorescent protein
(GFP) (B), mouse wild-type MDM2-specific TCR (Mu Wt TCR MDM2) (C), and chimeric
MDM2-specific TCR with humanized constant domains (Mu Chim TCR MDM2) (D). The left
panel shows the bulk transduction efficiency 2 d after two consecutive transduction cycles and
before geneticin/puromycin selection; the right panel shows the overall selection efficiency for
all constructs 2 wks after cessation of the selective pressure. In general, TCR that were struc-
turally modified, as exemplified for the partially humanized TCR (D) tend to be expressed at
lower amounts according to anti-v`6 or tetramer staining (not shown).
perature gradient in the second PCR. The setup of master mixes guarantees uniformity
and comparability of the individual temperatures. As a general rule, the rise or decline of
PCR bands within a temperature gradient is a reliable indicator for the optimal annealing
temperature.
6. The introduction of multiple ambivalent base positions into PCR primers has to be con-
sidered carefully, because aberrant amino acid sequences may be generated due to pos-
sible permutations of adjacent base tripletsin this case, the three Y positions will only
result in the two allowed motifs LysArgLys and LysLysLys of the C-terminal tail
(C`1 and C`2), which might serve as ER retention signals during TCR recycling (59).
7. There is a high degree of variability in choosing the appropriate ligation strategy: the
restriction sites for cloning have to be scrutinized to avoid cutting inside the TCR_/`
genes. When uncertain about the sequences of the variable gene regions, one may choose
eight-base-pair cutters like Asc (NEB) to minimize the probability of undesired diges-
tion, or select A/T cloning to prevent trouble with restriction sites. Altogether, how-
ever, it is of advantage to use restriction enzymes, because any construct can be easily
subcloned or routinely checked by test digestion for accuracy.
8. More conveniently, the reverse primer for RT-PCR (Subheading 3.2.3.) amplification of
the C`2 region could be extended to silence the NcoI site by a base mutation without
changing the amino acid sequence. Mouse C`1 has no terminal NcoI site.
9. For routine cloning, it is much more comfortable to precede ligation by silencing any
internal NcoI sites. NcoI/BamHI cloning allows higher ligation and transformation effi-
ciency.
10. For optimal fragment ligation, either DNA fragment has to be quantified in analytical
agarose gels to get an estimate of equivalent stoichiometry. The multimers can easily be
distinguished from the heterodimer by molecular weight and should vanish after the
EcoRI/BamHI cut.
11. The major drawback of this approach is the reduced ligation efficiency of blunt-ended
DNA and the frequent generation of 5' overhangs by T4 DNA polymerase, which will
reduce compatibility of the ligation educts. Strictly adhere to recommended blunt-end
cloning protocols and screen more bacterial clones than usual.
12. One limitation is the synthetic length of the chimeric primer (120mers have been probed
in our laboratory); each end of the chimeric primer that is complementary to its template
should be of at least half the sequence length as compared to the DNA linker sequence.
For example, the 18 amino acids of a linker will be encoded by 54 bases. As a conse-
quence, the complementary sequences on both sides of the chimeric primer should com-
prise a minimum of 27 bases, resulting in a 108mer.
13. Raising the amount of the chimeric primer did not ameliorate both the amount and speci-
ficity of the final chimeric PCR product, as determined by extensive primer titration.
14. It is strongly recommended to assess PCR at three different annealing temperatures at
once. A temperature gradient of 20C overall, with an average temperature of 55C,
proved to be a useful vantage point.
15. For any cell line and antibiotics used, killing curves have to be determined in order to
guarantee lethal doses. In the case of bulk IL-2 treated T-cells, 400 g/mL geneticin and
2.5 g/mL puromycin (f.c.) proved to kill the cells. During the 1 wk of treatment, the
concentration is doubled to be fail-safe. The stock concentration of the antibiotics is cor-
rected for biological activity and stored in frozen aliquots.
16. Useful hints/tutorials and protocols for reproducible transfection and transduction are
given at the homepage of the Garry P. Nolan laboratory, which is mainly responsible for
the advanced 293T/phoenix-ampho retroviral transduction system (see Website: http://

www.stanford.edu/group/nolan/). Material transfer agreement (MTA) forms for deliver-
ing the packaging cell lines can be downloaded; after being signed by the principal inves-
tigator, these have to be returned for ATCC orders. In our hands, the packaging cell lines
have initially been expanded to the maximum and Plasmocin treated (Invivogen) before
being frozen down in aliquots of 5 106 per vial. Cell morphology and adherence capa-
bility were checked routinely (protrusions that form cell/cell contacts as monolayers, no
clumps, and striking granularity). Frequent passaging (2 106 /T75 flask for 2.5 d) and
confluence were avoided. Ideally, the cells were not taken in culture until 3 d before
starting a new transduction experiment (at d 1). Extensive but careful (293T are semi-
adherent) washing with PBS before trypsinization is important to avoid the formation of
clumps and to favor uniform adherence as a monolayer when seeded. The incubator and
the sterile bench were routinely deactivated to minimize the risk of mycoplasma contami-
nation, which would have a detrimental effect on transduction.
17. The retroviral particle-containing supernatant has been titrated by transduction of the
susceptible cell line CF2TH (52) with 10-fold dilutions of a retroviral supernatant origi-
nating from a green fluorescent protein (GFP)-containing vector construct. The propor-
tion of positive cells has been quantified by FACS in the green-emitting FITC channel
upon excitation by 488 nm. One drawback of this transient technique is the significant
drop in transduction efficiency of thawed cultures of recently transfected packaging cells.
This necessitates the redundant transfection with the same vector constructs in cases
where the experiment has to be repeated. Alternatively, one may attempt to establish
stable packaging cell lines/clones. For high throughput of novel TCR constructs, how-
ever, the transient approach is sufficient and less time-consuming.
18. Prevent cells from confluence until virus supernatant harvest, as this will dramatically
reduce virus titer and transduction efficacy. The addition of HEPES supports pH stability.
19. Ratio of Fugene (L) to plasmid (g) has to be titrated in order to increase transfection
efficiency. In our hands, a ratio of 3:1 proved to be feasible for the simultaneous transfec-
tion of up to 4 plasmids (equals 20 g). Adhere to the recommendations of the commer-
cial supplier for handling of the transfection reagent.
20. RPMI depleted of phosphate not only improves transduction efficiency due to an increase
of phosphate receptor expression as target for the GALV envelope (60), but also improves
performance, survival, and antigen specificity of transduced bulk T-cell cultures in vitro.
21. In our hands, the use of RectoNectin (TaKaRa) as a mediator for cell/virus-particle
adherence improves transduction somewhat. For critical vector constructs, one may
enhance expression of the transgene product by taking advantage of fibronectin-coated
6-well plates.
22. This procedure may be repeated as long as retroviral supernatant of exponentially grow-
ing packaging cells will be harvested. Once seeded, the packaging cells will become
confluent at latest on d 5. In case of transient transfection, the schedule is therefore
limited to two daily cycles. For a more convenient Monday-to-Friday protocol, the sec-
ond cycle may be dispensable. The transfected cells can be stored at 80C and thawed
again in order to isolate new supernatant for further transductions, albeit with lower
efficiency as long as they have not been confluent before freezing.
23. Achievable cell densities depend on their actual expansion potency. The older the culture,
the more limited cell division is. Cell density should not fall below 0.5 106/mL and
should not exceed 1.5 106/mL, as this would lead to unfavorable growth and the eventual
death of the cell culture. These events favor the outgrowth of oligoclonal T-cell popula-
tions of uncertain specificities that differ from those of bulk cultures and may lead to a
functional bias.
References
1.
1 Yu, Z. and Restifo, N. P. (2002) Cancer vaccines: progress reveals new complexities.
J. Clin. Invest. 110(3), 289294.
2.
2 Kessels, H. W., Wolkers, M. C., and Schumacher, T. N. (2002) Adoptive transfer of T-cell
immunity. Trends Immunol. 23(5), 264269.
3.
3 Bankovich, A. J. and Garcia, K. C. (2003) Not just any T cell receptor will do. Immunity
18(1), 711.
4.
4 Kane, L. P., Lin, J., and Weiss, A. (2000) Signal transduction by the TCR for antigen.
Curr. Opin. Immunol. 12(3), 242249.
5.
5 Theobald, M., Biggs, J., Hernandez, J., Lustgarten, J., Labadie, C., Sherman, L. A.
(1997) Tolerance to p53 by A2.1-restricted cytotoxic T lymphocytes. J. Exp. Med.
185(5), 833841.
6.
6 Kuball, J., Schuler, M., Antunes, FE., et al. (2002) Generating p53-specific cytotoxic T
lymphocytes by recombinant adenoviral vector-based vaccination in mice, but not man.
Gene Ther. 9(13), 833843.
7.
7 Stanislawski, T., Voss, R. H., Lotz, C., et al. (2001) Circumventing tolerance to a human
MDM2-derived tumor antigen by TCR gene transfer. Nat. Immunol. 2(10), 962970.
8.
8 Kessels, H. W., Wolkers, M. C., van den Boom, M. D., van der Valk, M. A., and
Schumacher, T. N. (2001) Immunotherapy through TCR gene transfer. Nat. Immunol.
2(10), 957961.
9.
9 Holler, P. D. and Kranz, D. M. (2003) Quantitative analysis of the contribution of TCR/
pepMHC affinity and CD8 to T cell activation. Immunity 18(2), 255264.
10.
10 Lanzavecchia, A., Lezzi, G., and Viola, A. (1999) From TCR engagement to T cell activa-
tion: a kinetic view of T cell behavior. Cell 96(1), 14.
11.
11 Rosenberg, S. A. (1999) A new era for cancer immunotherapy based on the genes that
encode cancer antigens. Immunity 10(3), 281287.
12.
12 Holler, P. D., Holman, P. O., Shusta, E. V., OHerrin, S., Wittrup, K. D., and Kranz, D. M.
(2000) In vitro evolution of a T cell receptor with high affinity for peptide/MHC. Proc.
Natl. Acad. Sci. USA 97(10), 53875392.
13.
13 Kessels, H. W., van den Boom, M. D., Spits, H., Hooijberg, E., and Schumacher, T. N.
(2000) Changing T cell specificity by retroviral T cell receptor display. Proc. Natl. Acad.
Sci. USA 97(26), 14,57814,583.
14.
14 Stauss, H. J. (1999) Immunotherapy with CTLs restricted by nonself MHC. Immunol.
Today 20(4), 180183.
15.
15 Theobald, M., Biggs, J., Dittmer, D., Levine, A. J., and Sherman, L. A. (1995) Targeting
p53 as a general tumor antigen. Proc. Natl. Acad. Sci. USA 92(26), 11,99311,997.
16.
16 Sherman, L. A., Hesse, S. V., Irwin, M. J., La Face, D., and Peterson, P. (1992) Selecting
T cell receptors with high affinity for self-MHC by decreasing the contribution of CD8.
Science 258(5083), 815818.
17.
17 Katayama, C. D., Eidelman, F. J., Duncan, A., Hooshmand, F., and Hedrick, S. M. (1995)
Predicted complementarity determining regions of the T cell antigen receptor determine
antigen specificity. EMBO J. 14(5), 927938.
18.
18 Breedveld, F. C. (2000) Therapeutic monoclonal antibodies. Lancet 355(9205), 735740.
19.
19 Eshhar, Z. (1997) Tumor-specific T-bodies: towards clinical application. Cancer Immunol.
Immunother. 45(34), 131136.
20.
20 Labrecque, N., Whitfield, L. S., Obst, R., Waltzinger, C., Benoist, C., and Mathis, D.
(2001) How much TCR does a T cell need? Immunity 15(1), 7182.
21.
21 Carson, G. R., Kuestner, R. E., Ahmed, A., Pettey, C. L., and Concino, M. F. (1991) Six
chains of the human T cell antigen receptor.CD3 complex are necessary and sufficient for
processing the receptor heterodimer to the cell surface. J Biol Chem. 266(12), 78837887.
22.
22 Gouaillard, C., Huchenq-Champagne, A., Arnaud, J., Chen, C. L., and Rubin, B. (2001)
Evolution of T cell receptor (TCR) alpha beta heterodimer assembly with the CD3 com-
plex. Eur J. Immunol. 31(12), 37983805.
23 Li, Z. G., Wu, W. P., and Manolios, N. (1996) Structural mutations in the constant region
23.
of the T-cell antigen receptor (TCR)beta chain and their effect on TCR alpha and beta
chain interaction. Immunology 88(4), 524530.
24.
24 Saito, T., Sussman, J. L., Ashwell, J. D., and Germain, R. N. (1989) Marked differences in
the efficiency of expression of distinct alpha beta T cell receptor heterodimers. J. Immunol.
143(10), 33793384.
25.
25 Uematsu, Y. (1992) Preferential association of alpha and beta chains of the T cell antigen
receptor. Eur J. Immunol. 22(2), 603606.
26.
26 Burns, R. P., Jr., Natarajan, K., LoCascio, N. J., et al. (1998) Molecular analysis of skewed
Tcra-V gene use in T-cell receptor beta-chain transgenic mice. Immunogenetics 47(2),
107114.
27 Vacchio, M. S., Granger, L., Kanagawa, O., et al. (1993) T cell receptor V alpha-V beta
27.
combinatorial selection in the expressed T cell repertoire. J. Immunol. 151(3), 1322
1327.
28.
28 Casrouge, A., Beaudoing, E., Dalle, S., Pannetier, C., Kanellopoulos, J., and Kourilsky, P.
(2000) Size estimate of the alpha beta TCR repertoire of naive mouse splenocytes. J.
Immunol. 164(11), 57825787.
29.
29 Whitlow, M., Bell, B. A., Feng, S. L., et al. (1993) An improved linker for single-chain Fv
with reduced aggregation and enhanced proteolytic stability. Protein Eng. 6(8), 989995.
30.
30 Robinson, C. R. and Sauer, R. T. (1998) Optimizing the stability of single-chain pro-
teins by linker length and composition mutagenesis. Proc. Natl. Acad. Sci. USA 95(11),
59295934.
31.
31 Chung, S., Wucherpfennig, K. W., Friedman, S. M., Hafler, D. A., and Strominger, J. L.
(1994) Functional three-domain single-chain T-cell receptors. Proc. Natl. Acad. Sci. USA
91(26), 12,65412,658.
32.
32 Willemsen, R. A., Weijtens, M. E., Ronteltap, C., et al. (2000) Grafting primary human T
lymphocytes with cancer-specific chimeric single chain and two chain TCR. Gene Ther.
7(16), 13691377.
33.
33 Fitzer-Attas, C. J., Schindler, D. G., Waks, T., and Eshhar, Z. (1998) Harnessing Syk
family tyrosine kinases as signaling domains for chimeric single chain of the variable
domain receptors: optimal design for T cell activation. J. Immunol. 160(1), 145154.
34.
34 Johansson, B., Palmer, E., and Bolliger, L. (1999) The extracellular domain of the zeta-
chain is essential for TCR function. J. Immunol. 162(2), 878885.
35.
35 Bolliger, L. and Johansson, B. (1999) Identification and functional characterization of the
zeta-chain dimerization motif for TCR surface expression. J. Immunol. 163(7), 38673876.
36.
36 Atwell, S., Ridgway, J. B., Wells, J. A., and Carter, P. (1997) Stable heterodimers from
remodeling the domain interface of a homodimer using a phage display library. J. Mol.
Biol. 270(1), 2635.
37.
37 Garcia, K. C., Degano, M., Pease, L. R., et al. (1998) Structural basis of plasticity in T cell
receptor recognition of a self peptide-MHC antigen. Science 279(5354), 11661172.
38.
38 Backstrom, B. T., Milia, E., Peter, A., Jaureguiberry, B., Baldari, C. T., and Palmer, E.
(1996) A motif within the T cell receptor alpha chain constant region connecting peptide
domain controls antigen responsiveness. Immunity 5(5), 437447.
39.
39 Rudd, P. M., Elliott, T., Cresswell, P., Wilson, I. A., and Dwek, R. A. (2001) Glycosylation
and the immune system. Science 291(5512), 23702376.
40.
40 Germain, R. N. (2002) T-cell development and the CD4-CD8 lineage decision. Nat. Rev.
Immunol. 2(5), 309322.
41.
41 Naeher, D., Luescher, I. F., and Palmer, E. (2002) A role for the alpha-chain connecting
peptide motif in mediating TCR-CD8 cooperation. J. Immunol. 169(6), 29642970.
42 Sette, A., Vitiello, A., Reherman, B., et al. (1994) The relationship between class I binding
42.
affinity and immunogenicity of potential cytotoxic T cell epitopes. J. Immunol. 153(12),
55865592.
43.
43 Irwin, M. J., Heath, W. R., and Sherman, L. A. (1989) Species-restricted interactions
between CD8 and the alpha 3 domain of class I influence the magnitude of the xenoge-
neic response. J. Exp. Med. 170(4), 10911101.
44.
44 Gascoigne, N. R., Chien, Y., Becker, D. M., Kavaler, J., and Davis, M. M. (1984) Genomic
organization and sequence of T-cell receptor beta-chain constant- and joining-region genes.
Nature 310(5976), 387391.
45 Lefranc, M. P., Giudicelli, V., Ginestoux, C., et al. (1999) IMGT, the international Immuno-
45.
GeneTics database. Nucleic Acids Res. 27(1), 209212.
46.
46 Ruiz, M., Giudicelli, V., Ginestoux, C., et al. (2000) IMGT, the international Immuno-
GeneTics database. Nucleic Acids Res. 28(1), 219221.
47.
47 Lefranc, M. P. (2001) IMGT, the international ImmunoGeneTics database. Nucleic Acids
Res. 29(1), 207209.
48.
48 Lefranc, M. P. (2003) IMGT, the international ImmunoGeneTics database. Nucleic Acids
Res. 31(1), 307310.
49.
49 Niederberger, N., Holmberg, K., Alam, S. M., et al. (2003) Allelic exclusion of the TCR
alpha-chain is an active process requiring TCR-mediated signaling and c-Cbl. J. Immunol.
170(9), 45574563.
50. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning. A Laboratory
Manual, 2nd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
51.
51 Weijtens, M. E., Hart, E. H., and Bolhuis, R. L. (2000) Functional balance between T cell
chimeric receptor density and tumor associated antigen density: CTL mediated cytolysis
and lymphokine production. Gene Ther. 7(1), 3542.
52.
52 Weijtens, M. E., Willemsen, R. A., Hart, E. H., and Bolhuis, R. L. (1998) A retroviral
vector system STITCH in combination with an optimized single chain antibody chi-
meric receptor gene structure allows efficient gene transduction and expression in human
T lymphocytes. Gene Ther. 5(9), 11951203.
53. Tuan, R. S. (1997) Recombinant Gene Expression Protocols. 1st Ed. Human Press Inc.,
Totowa, NJ.
54.
54 Call, M. E., Pyrdol, J., Wiedmann, M., and Wucherpfennig, K. W. (2002) The organizing
principle in the formation of the T cell receptor-CD3 complex. Cell 111(7), 967979.
55 Rudd, P. M., Wormald, M. R., Stanfield, R. L., et al. (1999) Roles for glycosylation of cell
55.
surface receptors involved in cellular immune recognition. J. Mol. Biol. 293(2), 351366.
56.
56 Arden, B., Clark, S. P., Kabelitz, D., and Mak, T. W. (1995) Mouse T-cell receptor vari-
able gene segment families. Immunogenetics 42(6), 501530.
57.
57 Gurtu, V., Yan, G., and Zhang, G. (1996) IRES bicistronic expression vectors for efficient
creation of stable mammalian cell lines. Biochem. Biophys. Res. Commun. 229(1), 295298.
58.
58 Mizuguchi, H., Xu, Z., Ishii-Watabe, A., Uchida, E., and Hayakawa, T. (2000) IRES-
dependent second gene expression is significantly lower than cap-dependent first gene
expression in a bicistronic vector. Mol. Ther. 1(4), 376382.
59. Jackson, M. R., Nilsson, T., and Peterson, P. A. (1990) Identification of a consensus motif
for retention of transmembrane proteins in the endoplasmic reticulum. EMBO J. 9(10),
31533162.
60. Bunnell, B. A., Muul, L. M., Donahue, R. E., Blaese, R. M., and Morgan, R. A. (1995)
High-efficiency retroviral-mediated gene transfer into human and nonhuman primate pe-
ripheral blood lymphocytes. Proc. Natl. Acad. Sci. USA 92(17), 77397743.
Isolation of Tumor-Specific CD4+ T-Cells 257
18
Isolation and Expansion of Tumor-Specific
CD4+ T-Cells by Means of Cytokine Secretion
Christian Becker
Summary
A major objective of immune analysis in the setting of cancer, cancer vaccination, and
therapy is to accurately characterize and isolate functional T-cells elicited by a tumor. Secre-
tion of cytokines is an important function of activated effector and memory T-cells. The
cytokine secretion assay (CSA) directly assesses T-cells secreting cytokines after a short
restimulation in vitro. Cells can be stained for surface markers enabling characterization of
subsets. Viable cytokine-secreting T-cells can be isolated for expansion and further functional
testing or adoptive transfer. Since the method is not restricted to any antigen or major histo-
compatibility complex (MHC) haplotype, it enables us to detect CD8+ as well as CD4+ T-cells
reacting against a wide vary of tumor-associated antigens, ranging from particular known
tumor antigens to whole tumor cells and crude tumor lysates. Thus, the method provides a
valuable tool to analyze and isolate functional tumor-responsive T-cells according to one of
their effector functions.
Key Words: T-lymphocyte; immunoassay; T-cell differentiation; CD4+ T-helper cell; tumor
immunity; cytokine secretion assay.
1. Introduction
Considerable evidence suggests a central role of CD4+ T-cells in T-cell-mediated
tumor immunity (16). In nave hosts, tumor-specific CD4+ T-cells exist in very low
frequencies. During the course of an immune response against a tumor, these cells may
expand and acquire effector function. Yet, expanded effector and memory populations
remain small, and functional testing of these cells has been tedious and indirect. The
cytokine secretion assay (CSA) (7,8) overcomes these limitations by assessing a promi-
nent effector function of CD4+ T-helper cells: cytokine secretion.
Basically, only T-cells that have previously progressed towards an effector memory
phenotype respond with cytokine production to temporary ex vivo peptide/antigen
restimulation. To detect these cells, single-cell suspensions will be briefly stimulated in
vitro, and surface-coated with an antibody specific for the secreted cytokine of interest.
257
258 Becker
The cells are activated for a short period of time and labeled afterwards with a second-
ary cytokine-specific antibody for cytometric analysis and cell sorting. The secreted
cytokine binds preferentially to the affinity matrix on the secreting cells.
CSA has been used in tumor immunology to isolate human- and mouse-tumor-
reactive CD8+ T-cells from peripheral blood ( 911) and tumor-infiltrating lympho-
cytes (TILs) from tumor patients (12) or from immunized mice (12). This chapter
describes a protocol for the isolation and expansion of murine interleukin (IL)-2- and
interferon (IFN)--secreting CD4+ T-cells induced by tumor immunization in vivo.
2. Materials
2.1. Media, Buffers, and Reagents
1. Stimulation buffer: IMDM (Invitrogen, Carlsbad, CA) + 0.5% MS (mouse serum).
2. Washing buffer: PBS 1X, 4C.
3. Secretion buffer: IMDM + 10% FCS, 37C.
4. Culture medium: IMDM + 10% FCS, 1ng/mL rmIL-7 (R&D Systems, Minneapolis, MN)
+ Gentamycin.
5. Propidium iodide (PI) (Sigma-Aldrich, St. Louis, MO).
2.2. Antibodies
1. Cytokine secretion assay detection kit (Miltenyi, Bergisch Gladbach, Germany) consist-
ing of:
a. Cytokine catch reagent (anti-cytokine antibody conjugated to cell surface antibody).
b. Cytokine detection antibody conjugated to phycoerythrin (PE) (antibody against the
same cytokine as the cytokine catch antibody, but binding to a different epitope).
2. For stimulation: anti-CD28 (37.51, BD PharMingen, San Diego, CA), anti-CD3 (145-2C11,
BD Pharmingen).
3. For phenotypic characterization/sorting: FITC-conjugated anti-CD4, APC-conjugated
anti-CD45R/B220 (BD PharMingen).
2.3. Devices
1. Fluorescence-activated cell sorter (i.e., MoFlo, Dako Cytomation, Carpinteria, CA, or
other) with sample cooling system.
2. Sample mixer (i.e., MX1, Dynal, Brown Deer, WI).
3. Cell strainer, 70 m, nylon (Falcon).
4. Sample mixer (i.e., Dynal) .
5. Fifty-milliliter centrifuge tubes (i.e., Corning GmbH, Wiesbaden, Germany).
6. Twenty-four-well plates (i.e., Corning).
3. Methods
3.1. Isolation of Murine-Tumor-Specific CD4+ T-Cells
From Mice Immunized With Irradiated Tumor Cells
This protocol has been established for the isolation and expansion of hemagglutinin
(HA)-peptide-specific IFN-- or IL-2-secreting CD4+ T-cells induced by immunization
of BALB/c mice with irradiated HA-transfected tumors (see Fig. 1). It may need optimi-
zation when applied to other tumor antigens or T-cells secreting other cytokines.
1. Isolate spleens from mice 3 wks after the last immunization with tumor cells (see Note 1),
place into IMDM without serum (see Note 2) and prepare single-cell suspensions by
mincing with toothed forceps. After erythrocyte lysis, remove cell debris by passing the
cell suspension through a cell strainer. Resuspend cells in complete IMDM medium with
0.5% mouse serum at a cell density of 1 107 cells/mL.
2. Stimulate at least 1 107 cells per well in a 24-well plate with 10 g/mL peptide for 4
6 h under standard culture conditions. Additionally, always set up controls without pep-
tide and with spleen cells from mock-immunized mice (PBS-injected or injected with a
tumor not expressing the antigen of interest, sex and age matched) with and without
peptide. Add 0.2 g/mL anti-CD28 antibody (i.e., 37.51, BD Pharmingen) to all samples
to provide co-stimulation (see Note 3). While cells are stimulated, prewarm IMDM/FCS
medium at 37C.
3. Re-collect cells from wells and wash twice with ice-cold IMDM to remove previously
produced cytokines that might result in cross-staining.
4. Labeling with catch matrix: Resuspend cells in 80 L ice-cold IMDM/FCS per 107 cells,
put on ice, and add 20 L cytokine catch antibodies per 107 cells. Leave on ice for 7 min.
5. Cytokine secretion period: Transfer samples (in labeling buffer) into tubes with pre-
warmed 37C IMDM/FCS medium (use 50 mL medium for 5 107 cells) and put tube(s)
on sample mixer at 37C. Slowly tumble/mix tube(s) for 45 min. During this incubation,
cells will secrete cytokines, and these will bind to the cytokine catch matrix on the cells.
Diluting the cells avoids cross-staining of nonproducing cells.
6. Spin down cells at 4C and wash cells three times with 5 mL ice-cold PBS. Keep cells on
ice or at 4C from now on to avoid loss of cytokine staining.
7. Secondary cytokine staining: Resuspend cells in 80 L ice-cold IMDM/FCS per 107
cells, put on ice, and add 20 L cytokine-staining antibodies per 107 cells. Addition-
ally, label cells with anti-CD4 FITC and anti-B220 Allophycocyanine (APC) antibod-
ies (BD Pharmingen). Leave on ice for 10 min.
8. Wash twice with ice-cold PBS. Resuspend cells to 1 107 cells/mL in ice-cold PBS,
add PI (propidium iodide, 5 g/mL, Sigma), and subject to fluorescence-activated cell
sorting (see Note 3).
9. Sorting (see Fig. 2): Set up an FSC/SSC dot plot and draw a gate on small lymphocytes
[R1] (recently activated memory T-cells are not significantly larger than resting T-cells).
Analyze R1-gated cells in a secondary Pulse Width/FSC dot plot to gate out doublets.
Create a new gate [R1 and R2] and analyze cells in a third FL5/FL4 dot plot. Draw a gate
on FL5 cells to gate out cells with higher auto-fluorescence [gate R5]. Create a com-
bined gate [R1 and R2 and not R5] on the gate list and analyze in a fourth FL4/FL2 dot
plot, and gate out FL4 (B220+) high cells (B220-APC-labeled B-lymphocytes) [gate R3].
Finally create a gate [R1 and R2 and R3 and not R5] and analyze in a fifth Fl1/Fl2 dot
plot. Use the final dot plot to gate on FL1- and FL2-positive cells [R4] (cytokine-secret-
ing CD4+ T-cells) for sorting. Sort at least 20,000 cells using the combined gating strat-
egy [R1 and R2 and R3 and R4 and not R5].
10. After sorting, spin down tube with sorted cells and re-suspend in 200 L complete medium
with IL-7. CD4+/IFN-+ or CD4+/IL-2+ cells can be cultivated under these conditions for up
to 7 d without any visible loss of viability.
11. Cell numbers achieved by the CSA may be insufficient for functional testing, and further
expansion of the cells may thus be inevitable: For in vivo expansion, cells are transferred
into T-cell deficient, syngeneic recipient mice (RAG/ or SCID). Under these conditions,
cells expand into memory-type cells with improved effector function. Aspirate well with
260 Becker
syringe (Omnifix-F, 1 mL, 0.33 12 mm needle) and directly inject into the tail vein of a
mouse. Analyze and isolate cells from mice after at least 50 d of expansion in vivo (popu-
lation size depends on the number of transferred cells) (see Fig. 3). For an alternative in
vitro expansion procedure, see Note 4.
4. Notes
1. Irradiation dosage, cell numbers, and time schedule for protective tumor immunization
must be predetermined for each tumor individually by immunization/challenge experi-
ments. For sorting of cytokine-secreting T-cells, use animals that have completely rejected
the tumor challenge. Basically, memory CD4+ T-cells (isolated 3 wk after immunization)
are more rugged than recently activated effector T-cells and ride out the isolation proce-
dure much better. The assay can be performed with cells from a single spleen. Other tis-
sues (such as tumor-draining lymph nodes) should be pooled from several immunized
mice to achieve sufficient cell numbers for sorting.
The cytokine secretion assay does not distinguish between cells of high and low rec-
ognition efficiency. Cytokine production (i.e., IFN- by CD4+ T-cells) is not a direct
measure of other functions (i.e., cytotoxicity), since these may be regulated independently.
2. Carefully avoid xenogeneic sera and protein in media throughout cell preparation and
stimulation, as these will activate undesired T-cell specificities. However, during the
secretion phase of the assay, FCS-containing media may be used. Generally, enrichment
efficiency of antigen-specific cytokine-secreting cells depends on the amount of unre-
lated background activity in the sample. Controls should be set up to evaluate this back-
ground, and specific cells should be isolated only when background activity is low (e.g.,
on the level of peptide-stimulated cells from nave mice). Basically, the IFN- secretion
assay shows much less background activity than IL-2 and IL-4 secretion assays. When
starting experiments, also add a positive control (i.e., stimulated with anti-CD3 and anti-
CD28 [1 g/mL each or stimulate cells with the superantigen staphylococcal enterotoxin
B (Sigma) 1 g/mL]). Generally, the assay is best for detection and isolation of cytokine-
secreting populations up to 2% of the total population.
3. Although faster, magnetic enrichment of cytokine-producing cells is not recommended.
First, magnetic separation leads to many undesired cytokine-positive cells as well as
cytokine-negative cells in the positive fraction (i.e., B-cells that react with PE). Second,
magnetic enrichment was found to be only 10- to 15-fold for mouse cells. Most important,
the positive fraction usually contains cell debris that negatively influences the viability of
the isolated population. Keep the sample cooled during sorting, as the cytokine-holding
antibody matrix will easily come off at room temperature.
Fig. 1. (opposite page) Cytokine secretion of HA-peptide-reactive CD4+ T-cells induced

by immunization of BALB/c mice with irradiated HA-expressing tumor cells. Spleen cells
from individual mice isolated 3 wk after the last immunization with irradiated, HA-expressing
tumor cells were stimulated in vitro with 10 g/mL HA-peptide (SFERFEIFPKE, I-E d-
restricted hemagglutinin epitope from influenza Mount Sinai strain A/PR/8/34) for 6 h (D) or
left unstimulated (C). Spleen cells from PBS-injected control littermates, either peptide stimu-
lated (B) or unstimulated (A), served as controls. For co-stimulation, all cultures were supple-
mented with 0.2 g/mL anti-CD28 antibody. Stimulated cells were subjected to the IFN--,
IL-2-, or IL-4-secretion assay. Subsequently, cells were counterstained with anti-CD4-FITC
and anti-B220-APC antibodies and PI, and analyzed by FACS. 30,000 cell gated on B220/PI
cells are shown.
262 Becker
Fig. 2. Sorting strategy: Illustration of the sorting dot plots and gates as described under
Methods. (A) Lymphocyte gate. (B) Doublet discrimination. (C) Autofluorescence. (D) Gating
out B cells. (E) CD4+/cytokine+ gate.
Fig. 3. Sorting and in vivo expansion of IFN-- and IL-2-secreting CD4+ HA-peptide-spe-
cific T-cells. CD4+ T-cells were induced and activated in vitro as described, and cytokine-
producing CD4+ T-cells sorted out as shown in Fig. 2 on a MoFlo FACS sorter. Ten thousand
sorted cells from (A) and 20,000 cells from (C) were injected into the tail vein of individual
BALB/c RAG/ mice 24 h after sort. One wk later, mice were immunized with irradiated HA-
expressing cells from an unrelated tumor. Four wk later, spleen cells of adoptively transferred
mice were analyzed for CD4+/T-cell receptor (TCR)+ cells.
4. Keep in mind that all expansion procedures will alter the phenotype and function (although
not the specificity) of the cells. Alternatively, sorted cells may be expanded in vitro. Wash
cells after 48 h with medium and stimulate with freshly prepared syngeneic dendritic cells.
To prepare dendritic cells for in vitro expansion, spleen cells from nave, syngeneic mice
are incubated with anti-CD11c beads followed by magnetic sorting (MACS). The isolated
DC are immature but acquire the mature phenotype after overnight culture in granulocyte/
macrophage colony-stimulating factor (GM-CSF)-conditioned (R&D Systems) medium.
Use approx 10- to 20-fold more dendritic cells than sorted T-cells. Under these conditions,
264 Becker
cells will expand independent of their specificity (13). Expansion can be further amplified
by adding IL-7 (1 ng/mL) to the cultures.
References
1.
1 Goedegebuure, P. S. and Eberlein, T. J. (1995) The role of CD4+ tumor-infiltrating lym-
phocytes in human solid tumors. Immunol. Res. 14, 119131.
2.
2 Hung, K., Hayashi, R., Lafond-Walker, A., Lowenstein, C., Pardoll, D., and Levitsky,
H. (1998) The central role of CD4+ T cells in the antitumor immune response. J. Exp.
Med. 188, 23572368.
3.
3 Marzo, A. L., Kinnear, B. F., Lake, R. A., et al. (2000) Tumor-specific CD4+ T cells have
a major post-licensing role in CTL mediated anti-tumor immunity. J. Immunol. 165,
60476055.
4.
4 Surman, D. R., Dudley, M. E., Overwijk, W. W., and Restifo, N. P. (2000) Cutting edge:
CD4+ T cell control of CD8+ T cell reactivity to a model tumor antigen. J. Immunol.
164(2), 562565.
5.
5 Ossendorp, F., Mengede, E., Camps, M., Filius, R., and Melief, C. J. (1998) Specific T
helper cell requirement for optimal induction of cytotoxic T lymphocytes against major
histocompatibility complex class II negative tumors. J. Exp. Med. 187(5), 693702.
6.
6 Nishimura, T., Nakui, M., Sato, M., et al. (2000) The critical role of Th1-dominant immu-
nity in tumor immunology. Cancer Chemother. Pharmacol. 46(Suppl), S5261.
7.
7 Manz, R., Assenmacher, M., Pflger, E., Miltenyi, S., and Radbruch, A. (1995) Analysis
and sorting of live cells according to secreted molecules relocated to a cell-surface affinity
matrix. Proc. Natl. Acad. Sci. USA 92, 19211925.
8.
8 Brosterhus, H., Brings, S., Leyendeckers, H., et al. (1999) Enrichment and detection of
live antigen-specific CD4 + and CD8 + T cells based on cytokine secretion. Eur. J.
Immunol. 29, 40534059.
9 Oelke, M., Moehrle, U., Chen, J. L., et al. (2000) Generation and purification of CD8 +
9.
Melan-A-specific cytotoxic T lymphocytes for adoptive transfer in tumor immunotherapy.
Clin. Cancer Res. 6, 19972005.
10.
10 Oelke, M., Kurokawa,T., Hentrich, I., et al. (2000) Functional characterization of CD8+
antigen-specific cytotoxic T lymphocytes after enrichment based on cytokine secretion:
comparison with the MHC-tetramer technology . Scand. J . Immunol. 52, 544549.
11.
11 Pittet, M. J., Zippelius, A., Speiser, D. E., et al. (2001) Ex vivo IFN- secretion by circu-
lating CD8 T lymphocytes: implications of a novel approach for T cell monitoring in in-
fectious malignant diseases. J. Immunol. 166, 76347640.
12. Becker, C., Pohla, H., Frankenberger, F., et al. (2001) Adoptive tumor therapy with T
lymphocytes enriched through an IFN- capture assay. Nat. Med. 7(10), 11591162.
13. Ge, Q., Palliser, D., Eisen, H. N., and Chen, J. (2002) Homeostatic T cell proliferation in
a T celldendritic cell coculture system. Proc. Natl. Acad. Sci. USA 99, 29832988.
Characterization of Human CD8+ T Subsets 265
19
Methods for the Ex Vivo Characterization
of Human CD8+ T Subsets Based
on Gene Expression and Replicative History Analysis
Nathalie Rufer, Patrick Reichenbach, and Pedro Romero
Summary
The generation of an antigen-specific T-lymphocyte response is a complex multi-step pro-
cess. Upon T-cell receptor-mediated recognition of antigen presented by activated dendritic cells,
naive T-lymphocytes enter a program of proliferation and differentiation, during the course of
which they acquire effector functions and may ultimately become memory T-cells. A major goal
of modern immunology is to precisely identify and characterize effector and memory T-cell
subpopulations that may be most efficient in disease protection. Sensitive methods are required
to address these questions in exceedingly low numbers of antigen-specific lymphocytes recov-
ered from clinical samples, and not manipulated in vitro. We have developed new techniques to
dissect immune responses against viral or tumor antigens. These allow the isolation of various
subsets of antigen-specific T-cells (with major histocompatibility complex [MHC]-peptide
multimers and five-color FACS sorting) and the monitoring of gene expression in individual
cells (by five-cell reverse transcription-polymerase chain reaction [RT-PCR]). We can also fol-
low their proliferative life history by flow-fluorescence in situ hybridization (FISH) analysis of
average telomere length. Recently, using these tools, we have identified subpopulations of CD8+
T-lymphocytes with distinct proliferative history and partial effector-like properties. Our data
suggest that these subsets descend from recently activated T-cells and are committed to become
differentiated effector T-lymphocytes.
Keywords: Cytolytic T-lymphocytes (CTL); memory; effector; tumor immunity; T-cell dif-
ferentiation; melanoma; immunoassay; senescence; fibroblasts; gene expression; single-cell
RT-PCR; polyA cDNA amplification; replicative history; telomere length; flow-FISH; in situ
hybridization; nuclei.
1. Introduction
In both cancers, such as melanoma, and chronic viral infections, such as HIV, anti-
gen-specific CD8+ T-cell responses often develop, but their therapeutic effects are lim-
ited. Therefore, issues such as CD8+ T-cell subset (memory or effector) composition and
a full understanding of their functional properties in well-defined clinical situations are
265
266 Rufer, Reichenbach, and Romero
particularly pertinent. In this regard, metastatic melanoma lends itself to this type of
analysis. On one hand, this type of skin cancer has proven to be one of the most immuno-
genic among human tumors. On the other hand, a relatively high number of molecularly
defined antigens that are recognized by both MHC class I and class II T-lymphocytes
have been identified in the last 13 y (1). Some of these antigens have been targeted for
specific immunotherapy. For instance, more than 23 phase III clinical studies using
peptide-based cancer vaccines have been reported since 1995 (reviewed in [2]). In addi-
tion to vaccination, tumor antigens are targeted in other approaches such as adoptive
transfer therapy (3,4) or passive transfer of specific monoclonal antibodies (5).
Cytolytic T-lymphocyte (CTL)-defined tumor antigens have received great atten-
tion and are at the forefront of the development of cancer vaccines. Some of them
include MAGE-A3 (6), NY-ESO-1 (7), gp100 (8,9), and Melan-A/MART-1 (10). A
major focus of cancer vaccine development is the quantitative and qualitative monitor-
ing of the tumor antigen-specific CTL response in vaccinated patients (2). This type of
analysis has been greatly facilitated by the development in the 1990s of techniques to
identify single T-cells producing cytokines in response to challenge with antigen
(11,12) and of fluorescent MHC/peptide multimers (13). Despite this important
progress, monitoring of single-antigen-specific CTL remains difficult because of the
incomplete understanding of human CD8+ T-cell biology and also because most of
tumor antigen-specific CTL occur in accessible compartments, such as peripheral
blood, at frequencies below the detection limit of current immunoassays. One fortu-
nate exception is the CD8+ T-cell response to an immunodominant HLA-A2 restricted
peptide derived from the melanocyte/melanoma-associated antigen Melan-A/MART-1
(14). We have observed that the majority of healthy individuals have detectable num-
bers of HLA-A2/Melan-A/MART-1 peptide multimer+ CD8 T-lymphocytes in the cir-
culating lymphocyte compartment. Remarkably, these cells are phenotypically and
functionally naive despite their high frequency (10). This was explained by the exist-
ence of a high level of thymic output of Melan-A/MART-1-specific CD8 T-cells (15).
Importantly, this large repertoire is activated in HLA-A2 patients with metastatic mela-
noma, whose tumors drive the expression of high levels of the Melan-A/MART-1 anti-
gen. In this setting, Melan-A/MART-1-specific T-lymphocytes acquire phenotypic and
functional characteristics of antigen-experienced CD8+ T-lymphocytes and can be
found not only in peripheral blood but also among the lymphocytes infiltrating solid
tumors or in lymph node metastasis (10).
We are currently characterizing the functional diversity of Melan-A/MART-1-spe-
cific CD8+ T-lymphocytes that can be found either in blood or in tumor-infiltrated
lymph nodes (TILNs) from melanoma patients. The five-cell RT-PCR is an elegant
approach to study the function of isolated tumor-specific T-cells in patients in whom
only limited amounts of blood and of metastatic tumor material are available. This
allows us to analyze the expression level of genes involved in effector or regulatory
functions. Moreover, the precise characterization of the gene expression pattern (by
five-cell cDNA amplification RT-PCR) and of the replicative history (flow-FISH analy-
sis of average telomere length) of tumor antigen-specific T-lymphocytes provide valu-
able insights in dissecting the maturation pathways of effector and memory T-cells in
Fig. 1. Gene expression analysis of four samples processed either without (cDNA-1, cDNA-2)
or with (cDNA+-3, cDNA+-4) the amplified poly(A) reverse transcription-polymerase chain reac-
tion protocol. The starting cDNA (isolated from 102 or 10 cells) and polyA-amplified cDNA
(isolated from 10 cells) was prepared using 10-fold serial dilutions as indicated (Log10 of the
reciprocal of the dilution value). Expression of GAPDH, CD3, and CCR7 mRNAs is depicted.
the Melan-A model. Understanding the position of these subsets in the process of CD8+
T-cell differentiation and their role in immune responses will help us to improve vacci-
nation strategies for melanoma patients.
We have used a modified RT-PCR protocol that relies on the detection of specific
cDNAs after global amplification of expressed mRNAs (16,17). Because the method
yields sufficient cDNA in material harvested from as few as five isolated T-lympho-
cytes, it allows the analysis of gene expression in small purified T-cell subpopula-
tions. Figure 1 shows the levels of detection of target mRNAs among four samples
processed either without or with the poly(A)-amplified cDNA RT-PCR protocol. Typi-
cally, low to undetectable levels of expression of glyceraldehyde phosphate dehydro-
genase (GAPDH), CD3, and CCR7 mRNAs are found in cDNA samples that were not
poly(A) amplified. In contrast, amplified poly(A) cDNA allows the detection of all
genes of interest even when using a 103-fold dilution (or corresponding to 102 cells)
and gives robust PCR signals (Fig. 1). A recent study showed that this method could
detect CD3 transcripts in 60 to 90% of extracts from single CD3+ cells (18). More
recently, using the same five-cell cDNA amplification protocol, we described the prop-
erties of CD8+ T-lymphocyte subsets with functional features that are intermediate
between naive and effector T-cells. Indeed, these subsets express increasing levels of
mediators of effector CD8+ T-lymphocyte functions such as granzyme B, perforin,
interferon (IFN)-, tumor necrosis factor (TNF)-, and CD94 (19).
Telomeres are specialized structures found at the eukaryotic chromosomal ends.
They are composed of repeating oligonucleotide units, and their length progressively
decreases with age and with population doublings in vitro. The evolution of telomere
length has been extensively monitored mainly in two types of human cellsfibro-
blasts and lymphocytes. Indeed, in human T-lymphocytes, telomeres shorten by approx
3060 base pairs with every cell division. It is in principle possible to calculate the
number of divisions that a particular lymphocyte has undergone, provided one knows
the initial telomere length. The latter information is rather difficult to obtain. How-
ever, it is still possible to compare telomere lengths between different lymphocyte
populations and infer the average number of cell divisions that separate the popula-
tions with shorter telomeres from those with the longer telomeres.
In a population of cells, the average telomere length per cell would be an indicator
of the number of divisions that the cell has undergone relative to the reference popula-
tion. This requires an assay that allows one to measure the content of telomeric DNA
in individual cells. We have developed a flow-cytometry method using quantitative
fluorescence in situ hybridization (flow-FISH) with labeled peptide nucleic acid (PNA)
probes (20,21) for this purpose. This probe hybridizes to denatured telomere repeats in
fixed cells. Comparison of flow-FISH measurements with the results of conventional
telomere-length measurements by Southern blot analysis (R = 0.9) showed that the
amount of label (intensity of fluorescence signal) is a measure of the average length of
telomere repeats. Since the assay requires only 105 cells, the flow-FISH technique
permits comparison of telomere lengths in individual cells belonging to different sub-
populations. Consistent differences in telomere length in CD8+ T-cell subsets were
identified. The telomere length of memory T-cells or of HLA-A*0201-influenza-spe-
cific T-lymphocytes in normal adults is significantly shorter than that of naive T-cells,
thus supporting the notion that the telomere length reflects the replicative history of
T-lymphocytes (2023).
2. Materials
2.1. Five-Cell cDNA Amplification and RT-PCR
1. 0.1 M DTT (DiThioThreitol; Fluka, Buchs, Switzerland; at 20C).
2. Lysis buffer: for 1 mL of lysis buffer, add 8 L of 10 mg/mL, 20C tRNA (Roche Diag-
nostics Gmbh, Mannheim, Germany), 30 L of 100% Triton X-100 (Fluka), 40 L of
0.1 M DTT, and 922 L of dd-H2O. Prepare aliquots of 1 mL and store at 20C.
3. Oligo-dT primer, 20-mer (stock-solution at 1 g/L at 20C; prepare a working solution
at 100 ng/L with dd-H2O and store at 20C).
4. dNTP (Invitrogen, Paisley, UK): prepare aliquots at 10 mM and store at 20C.
5. Moloney murine leukemia virus reverse transcriptase (M-MLV RT), 200 U/L (Invitrogen).
6. Recombinant RNAsin ribonuclease inhibitor (RNAsin), 40 U/L (Promega, Madison, WI).
7. Glycogen (Roche Diagnostics Gmbh): 20 mg/mL of aliquots at 20C).
8. Ethanol, 100% (room temperature [RT], protected from light) and ice-cold 70% (20C).
9. NH4-acetate, 7.5 M (C2H7NO2; Fluka).
10. dATP (Amersham Pharmacia, Piscataway, NJ): aliquots of 10 mM at 20C).
11. Terminal deoxynucleotidyl transferase enzyme (TdT, Promega), 25 U/L.
12. Oligo-dT Iscove primer, 61-mer (HPCL purified, 1g/L in aliquots at 20C), 5'-CAT
GTC GTC CAG GCC GCT CTG GGA CAA AAT ATG AAT TCT TTT TTT TTT TTT
TTT TTT TTT T-3'.
13. KCl, 1 M; Tris-HCl (pH 8.8), 1 M; MgCl2, 1 M (Fluka).
14. PCR buffer, 5X (final 1.5 mM and 2 mM MgCl2): 250 mM KCl, 50 mM Tris-HCl (pH 8.8),
0.5 mg/mL BSA (bovine serum albumine; CalBioChem, San Diego, CA; 100 mg/mL, filter
and store at 4C), and 7.5 mM or 10 mM MgCl2. Aliquot 5X PCR buffer and store at 20C.
15. Taq DNA recombinant polymerase Platinum (Invitrogen; 5 U/L).
16. Mineral oil (Eurobio, Les Ullis, France).
2.2. Flow-FISH Analysis of Average Telomere Length

1. Bovine serum albumin (BSA; CalBioChem, 100 mg/mL = 10% (w/v); filter and store at
4C).
2. Ultrapure deionized formamide (Invitrogen). To deionize 500 mL of formamide, use 25 g
of resin (AgR 501-X8 (D); BioRad Laboratories), stir for 2 h at RT, filter twice, aliquot,
and keep at 20C. As formamide is highly toxic, the chemical hood should be used for
all manipulations.
3. Telomere-specific FITC-conjugated (C3TA2)3 peptide nucleic acid (PNA) probe (Applied
Biosystems, Rotkreuz, Switzerland). Dilute the telomere-specific PNA probe at 1 mg/mL
in 1/1 N/N' dimethylformamide/H2O (HPLC grade, Fluka). Aliquot to the equivalent vol-
ume of 20 L and store at 20C. For use, dilute in 1X TE buffer pH 8.0 (10 mM Tris, 1 mM
EDTA) at the working concentration of 30 g/mL, and store up to several months at 4C.
4. Tris-HCl, 1 M, pH 7.1.
5. Formamide for washes (Sigma-Aldrich Chemie, Gmbh, Steinheim, Germany).
6. Ten percent (w/v) Tween-20 (Sigma).
7. RNase T1 (Roche Diagnostics Gmbh): dilute 1:5 in 1X PBS (105 U/mL), heat at 80C for
20 min to inactivate DNAse, aliquot, and store at 4C.
8. Propidium iodide (Sigma): dilute at 10 g/mL in 1X PBS and store at 4C.
9. Quantum 24p Fluorescein Microbeads Standards (Bangs Laboratories, Fishers, IN).
10. Cell lysis buffer, 5X: 40 mM Tris-HCl (pH 7.5), 20 mM MgCl2, 1.28 M sucrose, and 4%
Triton X-100, store at 4C.
11. Thirty percent sucrose solution: 8 mM Tris-HCl (pH 7.5), 4 mM MgCl2, 30% sucrose, and
0.05% Na-azide, store at 4C.
3. Methods
3.1. Five-Cell cDNA Amplification and RT-PCR
The basic principle of the five-cell global cDNA amplification protocol requires that
the target sequences to be amplified be flanked by known sequences to which the ampli-
fication primers can anneal and initiate polymerization (Fig. 2). One end is initially
defined through a cDNA reaction using reverse transcriptase and an oligo(dT) primer
that will prime via the poly(A) tail present at the 3' end of most mRNA molecules. The
other end is then created by the addition of an homopolymer d(A) sequence to the
3'-OH end of the first cDNA using terminal deoxynucleotidyl transferase. Global PCR
amplification of the dA/dT flanked cDNAs is carried out using a single modified
oligo(dT) primer as previously described by Brady and Iscove (16). Priming of the
cDNA during global RT-PCR is initiated via annealing of the d(T) region of the 61-mer
oligonucleotide primer (named oligo-dT Iscove hereafter) to the homopolymeric d(A)
regions present at the termini of the cDNA molecules. We have included a purification
step before adding the poly d(A) tails, in order to get rid of free dNTPs that may inter-
fere during the tailing reaction.
Fig. 2. Basic principles and outlines of the five-cell global cDNA amplification protocol.
Adapted from Brady and Iscove (16).
Taken together, the methods described below outline (1) the five-cell sorting of
cytolytic CD8+ T-lymphocyte subsets of interest by flow cytometry; (2) the simulta-
neous lysis of five-cell aliquots and the transcription of total mRNA into cDNA; (3)
the cDNA precipitation; (4) the 3'-oligo(dA) tailing of precipitated cDNA; (5) the
global cDNA amplification by RT-PCR; and (6) the characterization of gene expres-
sion patterns in T-cell subsets by RT-PCR using specific primers.
3.1.1. Cell Preparation and Flow Cytometry
CD8+ T-cells are purified in two rounds of positive sorting by magnetic beads and
a MiniMACS device (Miltenyi Biotech, Bergish Gladbach, Germany). The resulting
cells are >98% CD3+CD8+. Cells are stained with appropriate monoclonal antibodies
in 20 mM PBS containing 0.2% BSA and 50 M EDTA for 20 min at 4C and washed.
To avoid contamination of small populations by more abundant subsets, 10 103 cells
of each subset of interest are first isolated by flow-cytometry-based cell sorting (FACS-
Vantage SE, using CellQuest software, Becton Dickinson, San Diego, CA). A repre-
sentative example shows the purity of naive(CD45RA+CCR7+CD27+), pre-effector
(CD45RA+CCR7-CD27+), and effector (CD45RA+CCR7CD27) CD8+ T-lymphocyte
subsets following flow-cytometry sorting (Fig. 3). Each subset (corresponding to
>95% of the sort window) is then resorted directly into 96-well V-bottom plates (five
cells per well) containing 15 L of a lysis/cDNA mix solution.
Characterization of Human CD8+ T Subsets
271
Fig. 3. Isolation of CD8+ T-lymphocyte subsets by flow cytometry for the five-cell cDNA amplification protocol. First, CD8+ T-lymphocytes
were purified by magnetic beads and a MiniMACS device. Second, 104 T-cells of naive (N; RA+CCR7+27+), of pre-effector (pE; RA+CCR7-
27+), and of effector (E; RA+CCR7-27-) subsets were sorted by flow cytometry. Immediate reanalysis of the various sorted subpopulations
revealed more than 95% purity. Third, five-cell aliquots of these highly purified subsets were resorted directly into 96-well V-bottom plates
containing the lysis/cDNA mix solution and further processed. Sorting regions used are indicated as R1 or R2.
271
3.1.2. Cell Lysis and cDNA Synthesis

1. The lysis/cDNA mix solution is freshly prepared while cells are being sorted. This solution
is composed of 6.3 L of lysis buffer, 3 L of 5X reverse transcriptase (RT) buffer (comes
along with the MMLV-RT enzyme), 1.5 L of DTT (0.1 M; final 10 mM), 0.75 L of
dNTPs (10 mM; final 0.5 mM), 0.25 L of oligo-dT (100 ng/L; final 1.67 ng/mL), 0.2 L
of RNAsin (40 U/L; final 8 U), 0.4 L of MMLV-RT (200 U/L; final 80 U) and of 2.6
L of dd-H2O. The total volume is 15 L per 96-well and per five-cell sorted aliquot.
2. After flow-cytometry sorting of a set of five-cell aliquots, the 96-well V-bottom plates
are covered carefully with a plastic adhesive cover (Catalys AG, Wallisellen, Switzer-
land). To allow the transcription of total mRNA into cDNA, the 96-well plates are incu-
bated 60 min at 37C and plates are spun down once (300g, 3 min).
3. The contents of each well are then transferred into 0.5-mL PCR tubes, and the M-MLV
reverse transcriptase is heat-inactivated at 90C for 3 min.
4. Tubes are placed on ice for a few min, centrifuged, and stored at 20C, or at 80C for
longer periods of time (see Note 1).
3.1.3. cDNA Precipitation

1. For further processing, the cDNA is precipitated overnight at 80C after addition of
7.5 L of 7.5 M NH4-acetate, 45 L of 100% ethanol using 2 L of 10 mg/mL glycogen
as a carrier.
2. Aliquots of 20 mg/mL of glycogen are kept at 20C and the dilution 1:2 with dd-H2O is
prepared freshly just before use.
3. Following overnight precipitation, tubes are centrifuged at 4C for 20 min (13,000g) and
the supernatant is carefully removed with a pipet. Neat white-to-transparent pellets should
be seen at the bottom of the tubes.
4. Pellets are washed in 150 L of ice-cold 70% ethanol, vortexed, and centrifuged at 4C
for 10 min (13,000g). The ethanol is removed and pellets are dried 12 h at RT.
3.1.4. Homopolymeric 3'-Oligo(dA) Tailing

To evaluate mRNA expression in a small number of cells (110 cells), the follow-
ing method was adapted from previously published protocols (16,17):
1. First, to promote 3'-oligo(dA) tailing of cDNA, the dried pellets are suspended in 5 L of
tailing solution containing 0.25 L of dATP (10 mM; final 0.5 mM), 0.05 L of terminal
deoxynucleotidyl transferase (TdT; 25 U/L; final 1.25 U), 1 L of 5X tailing buffer
(comes along with the TdT enzyme) and 3.7 L of dd-H2O.
2. Second, processed samples are incubated at 37C for 30 min (in a water bath) and the
tailing reaction is then heat inactivated at 90C for 35 min.
3. Tubes are then placed onto ice, briefly centrifuged (13,000g), and samples are ready for
global cDNA amplification RT-PCR.
3.1.5. Global cDNA Amplification by RT-PCR

1. The PCR mix is prepared on ice and contains 10 L of 5X PCR buffer (final 2 mM
MgCl2), 1 L of oligo-dT Iscove (1 g/L; final 20 ng/L), 1 L of dNTP (10 mM; final
0.2 mM), 2.5 L of Triton X-100 (10%; final 0.5%), 1 L of Taq Platinum (5 U/L; final
5 U) and 29.5 L of dd-H2O.
2. Into each tube containing 5 L of dA-tailed cDNA, 45 L of PCR mix is added, as well as
2 drops of mineral oil.
Table 1
Primer Sequences for Human CD8+ T-Lymphocytes
Name of gene Primer sequence-forward 5' 3' Primer sequence-reverse 5' 3'
GAPDH GGACCTGACCTGCCGTCTAG CCACCACCCTGTTGCTGTAG
CD3 CGTTCAGTTCCCTCCTTTTCTT GATTAGGGGGTTGGTAGGGAGTG
CD8alpha AACCACAGGAACCGAAGACGT GGACTTGCTCCCTCAAAAGGA
CD27 ACGTGACAGAGTGCCTTTTCG TTTGCCCGTCTTGTAGCATG
CCR7 CCAGGCCTTATCTCCAAGACC GCATGTCATCCCCACTCTG
Granzyme B GCAGGAAGATCGAAAGTGCGA GCATGCCATTGTTTCGTCCAT
Perforin TTCACTGCCACGGATGCCTAT GCGGAATTTTAGGTGGCCA
IFN- GCCAACCTAAGCAAGATCCCA GGAAGCACCAGGCATGAAATC
TNF- CTGCCTTGGCTCAGACATGTT CAGTTGGTCACCAAATCAGCA
CD94 GTGGGAGAATGGCTCTGCAC TGAGCTGTTGCTTACAGATATAACGA
3. Tubes are placed into a PCR machine (BioLabo Scientific Instruments, Chatel-St-Denis,
Switzerland) at 94C for 3 min to activate the Taq Platinum, and followed by 5 cycles of
PCR (50 s at 94C; 2 min at 37C; 9 min at 72C), 35 cycles (50 s at 94C; 90 s at 60C;
8 min at 72C), and 1 cycle (8 min at 72C). This cDNA is now called globally amplified
cDNA or cDNA+. It can be stored at 20C for up to 6 mo or at 80C for longer periods
of time (see Note 2).
3.1.6. Second Step PCR Reaction
The final step is to test the quality of the globally amplified cDNA. To avoid any
PCR contamination, the PCR mixes are prepared in a clean but different laboratory
area the one(s) used during steps 3.1.2. to 3.1.5.
One L of amplified cDNA is then subjected to a second round of PCR amplifi-
cation (3840 cycles, 30 s at 94C; 45 s at 58C; 1 min at 72C, followed by 1 cycle,
8 min at 72C) in 20 L volumes of 4 L of 5X PCR buffer (final 1.5 mM MgCl2),
0.4 L of dNTP (10 mM; final 0.2 mM), 0.4 L of specific primers designed to
amplify mRNA sequences of interest (100 ng/L; final 2 ng/L), 0.1 L of Taq
Platinum (5 U/L; final 0.5 U) and 13.7 L of dd-H2O. Typically, we are using
either H2O or the Daudi B-cell line for the negative PCR control, while 103 periph-
eral blood mononuclear cells (PBMC) from a healthy individual are used as positive
PCR control. We have tested the following primers for the gene expression analysis
of human CD8+ T-lymphocytes (Table 1). For all primers, with the exception of
the primer-pair for IFN- amplification (1.25 mM MgCl2), 1.5 mM MgCl2, and the
annealing temperature of 58C are used. The expected PCR products are visualized
after electrophoresis on a 1.2% agarose gel as shown in Fig. 4.
In a recent study, we have characterized and described the properties of a pre-
effector CD8+ T-lymphocyte subset, RA+CCR727+, in human peripheral blood (19).
By using five-cell RT-PCR, we found that all samples of CD8+ cells analyzed in the
present study gave a CD3-specific PCR product. All naive T-cell samples but none of
the aliquots of effector T-cells yielded a detectable CCR7-specific product (Fig. 5).
Fig. 4. Examples of specific reverse transcription-polymerase chain reaction (RT-PCR) reac-

tions on samples processed following the five-cell global cDNA amplification protocol. The PCR
products are visualized after electrophoresis on a 1.2% agarose gel, and the expected size for each
set of primers is indicated. M: 100- to 1500-bp ladder. GranzB; granzyme B. Perf; perforin.
Fig. 5. Gene expression analysis was performed on naive (N; RA+CCR7+27+), pre-effector
(pE; RA+CCR727+) and effector (E; RA+CCR727) T-cell subsets using the global cDNA
amplification protocol. Data from 10 independent five-cell aliquots as well as negative () and
positive (+) controls are depicted. The set of primers used in this experiment is described in
detail in Subheading 3.1.6. Reprinted with permission from ref. 19.
As expected, naive CD8+ T-lymphocytes, which are not cytolytic and do not pro-
duce cytokines, did not contain detectable granzyme B, perforin, FasL, IFN-, or
NK-receptor CD94 mRNA, and only rarely gave a TNF- signal. In contrast, these
mRNA transcripts were found in most (granzyme B, CD94) or in a significant fraction
of effector T-lymphocyte aliquots. Transcript analysis of RA+CCR7-27+ pre-effector
T-cells revealed the presence of all effector-function-associated mRNAs in a signifi-
cant fraction of five-cell aliquots from this population. Among RA+CCR727+ cell
aliquots there is no obvious correlation between the presence of PCR products corre-
sponding to different effector-function-associated genes. For instance, one sample
gave PCR products for granzyme B, IFN-, and CD94, while another was positive for
FasL, TNF-, and CD94 (Fig. 5). At this stage, we cannot decide whether this obser-
vation reflects a stochastic element in the PCR-based amplification from very small
transcript numbers, cell biological heterogeneity among RA+CCR727+ T-cells, or
heterogeneity of gene expression among individual cells from a homogeneous popula-
tion (24). We conclude that, unlike naive cells, a proportion of RA+CCR727+ T-cells
express genes associated with cytolytic T-cell effector functions, and we refer to this
subset as pre-effector T-lymphocytes.
3.2. Flow-FISH Analysis of Average Telomere Length
The estimation of telomere length in individual cells by flow FISH has been ini-
tially described in 1998 by two separate groups (20,25). In a recent review, Lauzon
et al. provide a complete description of both protocols with particular emphasis placed
on their similarities and their differences (26). The flow-FISH technique described
below is based on the original publication by Rufer and colleagues (20,21) and out-
lines (1) the quantitative telomere fluorescence in situ hybridization; (2) the acquisi-
tion and the analysis of telomere fluorescence by flow cytometry; and (3) the
adaptation of flow FISH to carry out telomere length measurement in nuclei obtained
from fibroblasts and various cell lines (unpublished data). The flow-FISH method has
been recently standardized to obtain highly accurate and reproducible results (27), and
some critical aspects of this technique are discussed.
3.2.1. Quantitative Fluorescence In Situ Hybridization
1. FACS-sorted human CD8+ T-lymphocytes of the desired subpopulation (2 105 cells)
are placed in 1.5-mL V-bottom tubes (Eppendorf, Hamburg, Germany) and washed in 1X
PBS (w/o Ca++ and Mg++) containing 0.1% (w/v) BSA.
2. Typically, 1 105 cells are resuspended in 100 L of hybridization mixture containing 70%
of ultrapure deionized formamide, 20 mM Tris (pH 7.1), 1% (w/v) BSA with either no
probe (unstained control) or with 0.3 g/mL telomere-specific FITC-conjugated (C3TA2)3
PNA probe for stained samples.
3. Samples are subjected to heat denaturation of DNA for 10 min at 84C in a water-bath
followed by hybridization for 2 h at RT in the dark (see Note 3).
4. Cells are then spun down and the supernatant is aspirated (leave 100 L of liquid at the
bottom of the tubes, as formamide is quite viscous and compact cell pellets cannot be
clearly observed).
5. To remove excess and non-specifically bound telomere-PNA probe, samples are washed
twice with 1 mL of wash buffer containing 70% formamide, 10 mM Tris (pH 7.1), 0.1%
BSA, and 0.1% Tween-20 at RT. The last washing step is performed with a wash solution
containing 1X PBS, 0.1% BSA, and 0.1% Tween-20. After each wash step, cells are spun
down at 16C for 7 min at 3000 rpm (1800g) for the formamide washes and at 2000 rpm
(800g) for the PBS wash.
6. Following the last wash, cells are resuspended in 300 L of 1X PBS containing 0.1%
BSA, 1000 U/mL RNAse T1, and 0.02 g/mL propidium iodide and transferred into
FACS tubes (see Note 4).
7. Cells are incubated for at least 30 min at RT prior to analysis.
3.2.2. Flow-Cytometric Analysis
The analysis of telomere fluorescence by gating on single cells (2n; FCS/FL3; R1),
excluding doublets, and gating on the lymphocyte population (FCS/SSC; R2) is done
using Cell Quest (Becton Dickinson). A representative example is shown in Fig. 6. To
correct for the auto-fluorescence of the cells in FL1, cells incubated in the hybridization
stage in the absence of the FITC-labeled telomere-PNA probe are used. The telomere
fluorescence of lymphocytes is calculated by subtracting the fluorescence of unstained
controls from the telomere fluorescence measured in cells hybridized with the FITC-
labeled telomere-PNA probe (Fig. 6A; ). To correct for daily shifts in the linearity of
the flow cytometer, fluctuations in the laser intensity and optical alignment and to allow
expression of results in standard fluorescence units, a mixture of four populations of
FITC-labeled fluorescent beads is used. This mix of four populations of beads, each
having a fluorescence corresponding to different known amounts of molecular equiva-
lents of soluble fluorochrome (MESF), is suspended in 1X PBS0.1% BSA and run in
each experiment (Fig. 6B). The mean fluorescence for each of the four different fluores-
cence peaks is used to establish a calibration curve and to convert telomere fluorescence
values into MESF values, allowing comparison among experiments. To estimate the
telomere length (in kbp) from telomere fluorescence in MESF values, we use the follow-
ing calibration curve: kbp = MESF (103) 0.5 (Fig. 6B) (see Note 5). Recently, we
measured the telomere length in CD8+ T-lymphocyte subsets representing discrete stages
of differentiation (23). Telomere shortening appears to occur progressively, correlating
with the stage of T-cell differentiation (Fig. 6C). Thus, highly differentiated effector-
type T-cells display the profile corresponding to the shortest telomeres, with lengths
(5.2 kb) equivalent to those observed in antigen-experienced CD8+ T-cells from elderly
individuals (21). In contrast, naive and memory T-cell subsets still exhibited relatively
long telomeres, with lengths of 9.1 kb and 6.8 kb respectively.
3.2.3. Flow-FISH on Nuclei of Fibroblasts
One of the major limitations of flow-FISH technology is its unique applicability to
cells that do not contain large cytoplasm, such as lymphocytes or monocytes. To enable
accurate measurement of telomere length in fibroblasts and various immortalized cell
lines (i.e., HeLa, CaCo2, HTC75), we have recently adapted the flow-FISH technique to
work with the nuclei of such cells (unpublished data). Cells grown on a monolayer are
harvested by using trypsin treatment, and all procedures to lyse cells and isolate their
nuclei are performed as follows:
1. Wash cells to be harvested once with 1X PBS.
2. Add 0.25% trypsin (Invitrogen) to the cell monolayer.
Fig. 6. (A) Flow-FISH analysis of the naive T-cell subset from an adult blood donor. Cells
following in situ hybridization were gated on FL3 (PI) and forward scatter (FSC) and then on
side scatter (SSC) vs FSC (R1 R2) to obtain the fluorescence histogram (FL1; shaded histo-
gram) derived from FITC-(C3TA2)3-labeled PNA probe. As background control, the equivalent
cell sample hybridized without the telomeric PNA probe is depicted (white histogram). Note
that the fluorescent signal in the FL1 channel is acquired on the linear scale. The specific telo-
mere fluorescence of cells is calculated by subtracting the mean background fluorescence from
the mean fluorescence obtained with the telomere probe (). (B) To allow the expression of
results in standard fluorescence units or molecules of equivalent soluble fluorochromes (MESF
103), a mixture of four populations of FITC-labeled beads is acquired and a standardized
linear regression line is generated (i.e., y = 0.025x). We have previously compared the telomere
fluorescence of 25 different lymphocyte samples measured by flow-FISH to the telomere length
(in kb) obtained from Southern blot analysis (20). The slope of the obtained calibration curve is
used to estimate the telomere length in kb from telomere fluorescence in MESF ( 103) values
(y = 0.5x, R = 0.9). (C) Example of telomere fluorescence analysis on sorted naive, memory,
and effector CD8+ T-cell subsets. Note that there is a progressive telomere shortening as cells
differentiate into effector-type of cells.
3. Wait until cells detach from the culture well.

4. Add 10% FBS-containing medium (i.e., DMEM) to stop enzymatic reaction, and transfer
the suspension to a centrifuge tube in ice.
5. Wash the culture plate with 10 mL of cold PBS and add to the centrifuge tube.
6. Recover cells by centrifuging at 300g for 510 min at 4C and discard supernatant.
7. Wash cells once again with cold PBS, discard supernatant, and resuspend to a final con-
centration of 1 1065 106 in one volume of cold PBS (1 mL).
8. Add one volume of ice-cold cell lysis buffer (1 mL) and three volumes of ice-cold dis-
tilled water (3 mL).
9. Mix by inverting the tubes several times and incubate for 10 min on ice.
10. Prepare new series of tubes containing 2.5 volumes of an ice-cold 30% sucrose solution
(2.5 mL).
11. Carefully add on the top of this sucrose cushion, the five volumes of the cell lysis buffer
containing the lysed cells (5 mL).
12. Centrifuge at 2000g for 10 min at 4C and remove the supernatant (which contains the
cytoplasm of lysed cells).
13. Carefully resuspend the pellet containing the cell nuclei in cold PBS and count them. The
recovery of nuclei from the starting intact cell population is 50 to 80%.
Nuclei from lysed cells such as fibroblasts and in vitro established cell lines are
now ready to be used for telomere length measurements by flow FISH as described in
Subheading 3.2.1. Since we have performed all experiments with 4 105 nuclei, the
volume of the hybridization mixture is doubled (200 L instead of 100 L). More-
over, isolated nuclei are more fragile than intact cells. Therefore, isolated nuclei
should be kept on ice, and all steps using pipets should be done with care. Do not
vortex nuclei containing samples, even after they have been fixed with 70%
formamide. Keep in mind that clumps and aggregates are more likely to form on
fixed nuclei than on whole fixed cells. Telomere length analysis is then performed as
described in Subheading 3.2.2.
To verify whether the same calibration slope (y = 0.5x) could be used for the analy-
sis of telomere fluorescence values obtained from nuclei of human cells, 22 different
nuclei preparations from fibroblasts and various cell lines were isolated and charac-
terized separately by Southern blot (Fig. 7A) and flow-FISH (Fig. 7B). Southern
blotting is the gold standard for measurement of telomere length (28,29), but requires
large numbers of cells and is relatively laborious. Briefly, the DNA is extracted from
cells and digested by frequent cutting restriction enzymes such as RsaI and HinfI.
These digested DNA fragments are run on an agarose gel and hybridized to a P32-
labeled (C3TA2)3 telomeric probe. The blot is exposed to photographic film, and the
telomere length is estimated from the resulting autoradiograph by densitometric
analysis. Results from flow-FISH measurements (Fig. 7B) correlated with results
from conventional telomere length analysis by Southern blot (Fig. 7C; R = 0.86).
Furthermore, the slope for the calibration curve obtained with nuclei of fibroblasts
and various cell lines was very similar to the previously described calibration curve
using intact small-size cells (kbp = MESF 0.5) (20). Finally, we have characterized
telomere length dynamics of in vitro cultured fibroblasts in relation to cell divisions.
Progressive telomere shortening was observed by both Southern blot and flow-FISH
analysis as fibroblasts were ageing in culture (Fig. 7).
4. Notes
1. Five-cell cDNA amplification and RT-PCR. We have developed a single-step procedure
for cell lysis and cDNA synthesis, since our standard goal is to start with limited cell
Characterization of Human CD8+ T Subsets
279
Fig. 7. Isolated nuclei from various cell lines (i.e., HeLa, HTC75, CaCo2, and HT29) as well as from fibroblasts at various time-points of in
vitro culture (PD23, PD36, and PD43) and from hTERT-expressing fibroblasts were subjected to (A) Southern blot analysis of telomere repeat
arrays and (B) flow-FISH analysis of average telomere length. (C) The telomere fluorescence (in MESF 103) was proportional to the terminal
restriction fragment values (in kbp) with a high correlation coefficient (R = 0.86). Note that the slope of the calibration curve described for nuclei
of cells is similar to that of the previously described calibration curve obtained on intact cells (20). PD, population doublings; MESF, molecules
of equivalent soluble fluorochromes.
279
numbers (110 cells). This direct experimental step relies on Triton X-100 to lyse the
outer cellular membrane, as well as on the presence of an RNAse inhibitor to protect the
RNAs from degradation. Using such a strategy has the following consequences. First,
cDNA strands are synthesized from the 3' poly(A) tails of mRNAs. Second, the length of
synthesized first-strand cDNAs is estimated to vary between 100 and 2500 bp. Based on
these facts, we designed our primers of interest (see Table 1) in such a way that they are
located within the first 1000 bp following the 3'-polyA end of the mRNA sequence,
and whenever possible between two exons. However, since most mRNAs have 3'-UTR
sequences that largely exceed 1 kb of length, it is almost impossible to design primers
located between two exons (with the exception of GAPDH, granzyme B, and CD8 prim-
ers). We have recently reported that specific PCR products nicely correlated with the
levels of protein expression, and have concluded that the presence of genomic DNA was
hardly detected using our selected primers (19). As shown in Fig. 4, only naive T-cell
samples yielded a detectable CCR7-specific product correlating with CCR7 cell-surface
expression. Faint contaminant PCR bandsi.e., TNF- and IFN-were observed only
in some rare cases among samples of naive T-cell origin. Thus, it is quite unlikely that
these bands represent contaminant genomic DNA. Altogether, the fact that no genomic
DNA is detected in our five-cell assay may be explained by the low number of genomic
copies present within the 5-cell samples, and by their dilution following the global cDNA
amplification procedure (10 copies per 50 L).
2. The concentration of MgCl2 in the global cDNA amplification RT-PCR appears to be a
critical factor, and it was optimized at 2 mM (data not shown). More generally, the proce-
dure is delicate and any departure from the outlined method, including changes in the
commercial sources of the reagents, should be carefully evaluated using reference posi-
tive and negative controls. The original oligo(dT) Iscove primer is a 61-mer and contains
an EcoRI restriction site prior to the dT stretch (16). Beside these features, the unique
sequence was chosen arbitrarily by Brady and colleagues (16). We tested a truncated
43-mer of the same primer sequence (5'-CTG GGA CAA AAT ATG AAT TCT TTT TTT
TTT TTT TTT TTT TTT T-3') and could observe neither major qualitative nor quantita-
tive differences in the performance of the two primers. Successful global cDNA amplifi-
cation generates samples containing a heterogeneous collection of 3'-cDNA fragments
representing all transcribed poly(A) mRNAs. The loading of 5 L of the amplified cDNA
sample on a 1.2% agarose gel migrates as a relatively strong ethidium bromide-stained
smear ranging from 100 to 2500 bp in length.
3. Flow-FISH analysis of average telomere length. There are two key aspects in the flow-
FISH technique. First, we use directly labeled PNA probes instead of DNA or RNA probes.
PNA probes are synthetic mimics of single-strand DNA with bases attached to an uncharged
peptide-like backbone. The PNA probe hybridizes to complementary DNA in low ionic
strength solutions that do not favor re-annealing of the target strands. Thus, PNA-DNA
interactions are more stable than DNA-DNA or DNA-RNA interactions under hybridiza-
tion conditions (reviewed in ref. 30). Second, the heating and formamide treatment allow
the PNA probes, which are present during the denaturation step, to get into the cells without
their prior permeabilization. In contrast to us, Hultdin and colleagues (25) employed the
standard in situ hybridization method of fixation followed by permeabilization, with the
Caltag Fix and Perm kit. Our strategy avoids problems associated with excessive fixa-
tion, such as the potential accessibility of the PNA probe to penetrate the cell.
For the flow-FISH technique, DNA needs to be completely denatured to single strands
within the cell, and the optimal denaturation temperature and time should not result in
disintegration of the cells or of the nuclei. The denaturation step takes place with the
cells suspended in 70% formamide-based hybridization solution at 84C for 10 min. For
T-lymphocytes, there is a progressive increase in telomere fluorescence signals as the
denaturation temperature is increased, reaching a plateau at 86C (data not shown). At
higher temperature (90C) or longer time periods (>15 min) (27), the overall telomere
fluorescence signals decreased due to the loss of cells or of DNA. Finally, the optimal
denaturation temperature and time also depends on the volume of hybridization solution.
The technique described here is based on the use of low numbers of isolated T-lympho-
cyte subsets (a maximum of 105 cells) and thus on low volumes (100200 L) of hybrid-
ization mixture. If higher cell numbers are used (i.e., 5 105 cells), the final volume of
300 L as well as a denaturation temperature ranging between 85 and 87C for 15 min
has been recommended (27). Finally, the hybridization time represents another parameter
that can be modified. Since the hybridization of the telomere-PNA probes requires more
than 60 min to reach the plateau of telomere fluorescence (data not shown) (27), we
propose performing the hybridization with 0.3 g/mL of PNA probe at RT and in the dark
for 90120 min.
4. Although formamide is relatively toxic and requires handling in a fume hood, the three
washing steps are critical in order to dilute the excess of the PNA probe. Various washing
protocols, including the use of 1X PBS instead of formamide, have been tested by
Baerlocher and coworkers (27), unfortunately without providing telomere fluorescence
signals equivalent to those observed with the classical protocol. Propidium iodide (PI)
staining is widely used for the distinction of single cells from doublets as well as for the
discrimination of cells in G0/G1 vs in S phase. Because telomere replication takes place
in early S phase, it is important to gate on the 2n population of cells as determined by PI
staining (see Fig. 6A). Moreover, the PI emission spectrum overlaps to some extent with
that of FITC in FL1, and using lower concentrations of the dye reduces interference with
the telomere-PNA FITC measurements (data not shown). Thus, we use for 105, 2 105,
and 4 105 cells, respectively, 0.02, 0.04, and 0.06 g/mL of PI. Others have proposed
resuspending cells in LDS 751 at a concentration of 0.01 g/mL (27).
5. Since the flow-FISH method involves many steps where relatively minor errors can trans-
late into large shifts in fluorescent readout, we have focused our technique on monitoring
intra- and inter-experimental errors by using calibration beads as well as frozen cell
samples such as lymphoma cells or PBMC. First, FITC-labeled beads are run in each
experiment, and the resulting calibration curve is used to correct for daily fluctuations in
the laser intensity, alignment, and linearity. Second, in order to analyze the day-to-day
variation in flow-FISH results, aliquots of the same frozen lymphoma cells were ana-
lyzed in each experiment over a 6-mo period (data not shown) (21). In 38 experiments,
the mean ( s.d.) fluorescent value was 12.2 1.8 MESF ( 103), with a coefficient of
variation (CV) of 15 %. Variation between multiple samples of the same cell type in the
same experiment was <5%, and variation between experiments using the same frozen
control cells ranged between 10 and 20%. Recovery of hybridized cells varied in different
experiments from 20 to 80%, with a higher range corresponding to freshly isolated cells
with a high viability (>90%) and the lower range corresponding to previously frozen
cells. In contrast, Hultdin et al. (25) used an internal control such as 1301 (T-cell lympho-
blastic leukemia cell line) cells, known to possess extremely long telomeres (>25 kb).
1301 cells are mixed with each experimental cell sample, and the telomere fluorescence
value is calculated as a ratio between the telomere signal of the experimental sample and
of the 1301 T-cells.
Acknowledgments
The authors thank Patricia Corthsy-Henrioud, Sverine Reynard, and Martine van
Overloop for excellent technical support. This work was supported by the National
Center of Competence in Research (NCCR) Molecular Oncology program of the Swiss
National Science Foundation.
References
1.
1 Van Der Bruggen, P., Zhang, Y., Chaux, P., et al. (2002) Tumor-specific shared antigenic
peptides recognized by human T-cells. Immunol. Rev. 188, 5164.
2. Speiser, D., Pittet, M. J., Rimoldi, D., et al. (2003) Evaluation of melanoma vaccines with
molecularly defined antigens by ex vivo monitoring of tumor-specific T cells. Seminars in
Cancer Biology 13, 461472.
3.
3 Dudley, M. E. and Rosenberg, S. A. (2003) Adoptive-cell-transfer therapy for the treat-
ment of patients with cancer. Nat. Rev. Cancer 3, 666675.
4.
4 Ho, W. Y., Blattman, J. N., Dossett, M. L., Yee, C., and Greenberg, P. D. (2003) Adoptive
immunotherapy: engineering T cell responses as biologic weapons for tumor mass destruc-
tion. Cancer Cell 3, 431437.
5. Waldmann, T. A., Levy, R., and Coller, B. S. (2000) Emerging therapies: spectrum of
applications of monoclonal antibody therapy. Hematology (Am. Soc. Hematol. Educ. Pro-
gram) 2000, 394408.
6.
6 Marchand, M., Punt, C. J., Aamdal, S., et al. (2003) Immunisation of metastatic cancer
patients with MAGE-3 protein combined with adjuvant SBAS-2: a clinical report. Eur. J.
Cancer 39, 7077.
7.
7 Scanlan, M. J., Gure, A. O., Jungbluth, A. A., Old, L. J., and Chen, Y. T. (2002) Cancer/
testis antigens: an expanding family of targets for cancer immunotherapy. Immunol. Rev.
188, 2232.
8 Rosenberg, S. A., Yang, J. C., Schwartzentruber, D. J., et al. (2003) Recombinant fowlpox
8.
viruses encoding the anchor-modified gp100 melanoma antigen can generate antitumor
immune responses in patients with metastatic melanoma. Clin. Cancer Res. 9, 29732980.
9.
9 Smith, J. W., 2nd, Walker, E. B., Fox, B. A., et al. (2003) Adjuvant immunization of HLA-
A2-positive melanoma patients with a modified gp100 peptide induces peptide-specific
CD8+ T-cell responses. J. Clin. Oncol. 21, 15621573.
10.
10 Romero, P., Valmori, D., Pittet, M. J., et al. (2002) Antigenicity and immunogenicity of
Melan-A/MART-1 derived peptides as targets for tumor reactive CTL in human mela-
noma. Immunol. Rev. 188, 8196.
11.
11 Ghanekar, S. A. and Maecker, H. T. (2003) Cytokine flow cytometry: multiparametric
approach to immune function analysis. Cytotherapy 5, 16.
12.
12 Pittet, M. J., Zippelius, A., Speiser, D. E., et al. (2001) Ex vivo IFN-gamma secretion by
circulating CD8 T lymphocytes: implications of a novel approach for T cell monitoring in
infectious and malignant diseases. J. Immunol. 166, 76347640.
13.
13 Altman, J. D., Moss, P. A. H., Goulder, P. J. R., et al. (1996) Phenotypic analysis of
antigen-specific T lymphocytes. Science 274, 9496.
14.
14 Kawakami, Y., Eliyahu, S., Sakaguchi, K., et al. (1994) Identification of the immuno-
dominant peptides of the MART-1 human melanoma antigen recognized by the majority of
HLA-A2-restricted tumor infiltrating lymphocytes. J. Exp. Med. 180, 347352.
15.
15 Pittet, M. J., Zippelius, A., Valmori, D., Speiser, D. E., Cerottini, J. C., and Romero, P.
(2002) Melan-A/MART-1-specific CD8 T cells: from thymus to tumor. Trends Immunol.
23, 325328.
16.
16 Brady, G. and Iscove, N. N. (1993) Construction of cDNA libraries from single cells.
Methods Enzymol. 225, 611623.
17.
17 Sauvageau, G., Lansdorp, P. M., Eaves, C. J., et al. (1994) Differential expression of
homeobox genes in functionally distinct CD34+ subpopulations of human bone marrow
cells. Proc. Natl. Acad. Sci. USA 91, 12,22312,227.
18.
18 Bigouret, V., Hoffmann, T., Arlettaz, L., et al. (2003) Monoclonal T-cell expansions in
asymptomatic individuals and in patients with large granular leukemia consist of cyto-
toxic effector T cells expressing the activating CD94:NKG2C/E and NKD2D killer cell
receptors. Blood 101, 31983204.
19.
19 Rufer, N., Zippelius, A., Batard, P., et al. (2003) Ex vivo characterization of human CD8+
T subsets with distinct replicative history and partial effector functions. Blood 102, 1779
1787.
20.
20 Rufer, N., Dragowska, W., Thornbury, G., Roosnek, E., and Lansdorp, P. M. (1998) Te-
lomere length dynamics in human lymphocyte subpopulations measured by flow
cytometry. Nat. Biotechnol. 16, 743747.
21.
21 Rufer, N., Brummendorf, T. H., Kolvraa, S., et al. (1999) Telomere fluorescence measure-
ments in granulocytes and T lymphocyte subsets point to a high turnover of hematopoietic
stem cells and memory T cells in early childhood. J. Exp. Med. 190, 157167.
22.
22 Zippelius, A., Pittet, M. J., Batard, P., et al. (2002) Thymic selection generates a large T
cell pool recognizing a self-peptide in humans. J. Exp. Med. 195, 485494.
23.
23 Papagno, L., Spina, C. A., Marchant, A., et al. (2004) Immune activation and CD8+ T-cell
differentiation towards senescence in HIV-1 infection. PLoS 2(2), e20.
24.
24 Levsky, J. M., Shenoy, S. M., Pezo, R. C., and Singer, R. H. (2002) Single-cell gene
expression profiling. Science 297, 836840.
25.
25 Hultdin, M., Gronlund, E., Norrback, K., Eriksson-Lindstrom, E., Just, T., and Roos, G.
(1998) Telomere analysis by fluorescence in situ hybridization and flow cytometry.
Nucleic Acids Res. 26, 36513656.
26.
26 Lauzon, W., Sanchez Dardon, J., Cameron, D. W., and Badley, A. D. (2000) Flow
cytometric measurement of telomere length. Cytometry 42, 159164.
27.
27 Baerlocher, G. M., Mak, J., Tien, T., and Lansdorp, P. M. (2002) Telomere length mea-
surement by fluorescence in situ hybridization and flow cytometry: tips and pitfalls.
Cytometry 47, 8999.
28.
28 Harley, C. B., Futcher, A. B., and Greider, C. W. (1990) Telomeres shorten during ageing
of human fibroblasts. Nature 345, 458460.
29. Allsopp, R. C., Vaziri, H., Patterson, C., et al. (1992) Telomere length predicts replicative
capacity of human fibroblasts. Proc. Natl. Acad. Sci. USA 89, 10,11410,118.
30. Lansdorp, P. M., Poon, S., Chavez, E., et al. (1997) Telomeres in the haemopoietic sys-
tem. Ciba Found. Symp. 211, 209218; discussion 1922.
Regulatory T-Cells in Antitumor Therapy 285
20
Regulatory T-Cells in Antitumor Therapy
Isolation and Functional Testing of CD4+CD25+ Regulatory T-Cells
Helmut Jonuleit and Edgar Schmitt
Summary
Naturally occurring CD4+CD25+ T regulatory cells originate from the thymus and play a
central role regarding the maintenance of peripheral tolerance by suppression of autoreactive
T-cell populations. However, T regulatory cells can have beneficial as well as harmful effects.
On the one hand, they prevent a variety of autoimmune and inflammatory diseases; but on the
other hand, they concomitantly inhibit antitumor immune reactions by suppressing tumor-spe-
cific T-cell responses. Therefore, these ambivalent properties of T regulatory cells require
detailed investigation especially with respect to a potential therapeutic exploitation of these cells.
A prerequisite for such analyses is the isolation of pure T regulatory cells and the establishment
of functional tests for the analysis of their suppressive properties, since no specific markers for
CD4 +CD25 + T regulatory cells are known so far. In this chapter, techniques applying
immunomagnetic beads have been used to establish an efficient method for isolation of human
as well as murine regulatory T-cells. A combination of positive and negative selection steps
using immunomagnetic beads of different sizes yields preparations of functional, active
CD4+CD25+ regulatory T-cells with high purity. Additionally, co-culture assays for functional
characterization of isolated CD4+CD25+ regulatory T-cells are described that are able to detect
their suppressive properties for conventional CD4+ T-cells with high sensitivity.
Key Words: CD4+CD25+ T-cells; regulatory T-cells; tolerance; T-cell suppression; regu-
lation.
1. Introduction
The immune system distinguishes between self and nonself structures, but also
between harmful and innocuous foreign antigens, to prevent unessential and self-
destructive immune responses. To maintain tolerance against self as well as protec-
tive immunity against pathogens, sophisticated and highly balanced immune
mechanisms have evolved to generate antigen-specific effector cell responses as well
as antigen-specific tolerance induction.
The idea of specific suppressor T-cell populations that counteract harmful
autoaggressive immune responses in the periphery was first described in the 1970s by
285
286 Jonuleit and Schmitt
Gershon et al. (1). However, at that time neither the cells nor the hypothetical soluble
suppressor factors responsible for the observed effects could be identified. In 1995,
Sakaguchi et al. described a subpopulation of CD4+ T helper cells, characterized by a
constitutive expression of the interleukin (IL)-2 receptor -chain (CD25), that is essen-
tial to control autoaggressive immune responses in mice (2). After subsequent in vitro
studies by several groups, this population is now referred to as CD4+CD25+ T regulatory
cells. This distinct T-cell population was originally described in mice. Meanwhile, com-
parable T regulatory cell populations, with identical phenotype and functional activities,
have been defined more recently in rats and humans (38). They represent 510% of
total peripheral CD4+ T-cells. Freshly isolated, CD4+CD25+ T regulatory cells do not
proliferate after allogeneic or polyclonal stimulation, but suppress the activation and
cytokine release of conventional CD4+ and CD8+ T-cells in an antigen-nonspecific and
cell contact-dependent manner (3,810). Currently, no specific markers for regulatory
T-cells are known, but under noninflammatory conditions they comprise the vast major-
ity of all CD25+CD4+ T-cells. T regulatory cells can be isolated from nave as well as
from tolerized donors, and this chapter is focused on methods that can be utilized for
isolation and functional characterization of human and murine CD4+CD25+ T regula-
tory cells.
2. Materials
2.1. Isolation and Testing of Human Regulatory T-Cells
1. Isolation buffer: PBS 1X (w/o Ca2+, Mg2+), 0.5% human serum albumin (HSA, i.e.,
Human-Albumin Kabi 20%, Octapharma, Langenfeld, Germany), 3 mM EDTA, 4C.
2. Washing buffer: PBS 1X (w/o Ca2+, Mg2+), 1 mM EDTA, 4C.
3. Depletion buffer: MEM + 5% FCS, 4C.
4. Culture medium: X-VIVO 15 (BioWhittaker, Walkersville, MD).
5. For isolation: anti-CD25-FITC (M-A251, BD PharMingen, San Diego, CA).
6. For stimulation: anti-CD3 (OKT-3, ATCC), anti-CD28 (CD28.2, BD PharMingen).
7. For phenotypical characterization: FITC- or phycoerythrin (PE)-conjugated anti-CD3,
anti-CD4, anti-CD25, anti-CD45RO, anti-CTLA-4 (Coulter/Immunotech, Marseille,
France).
8. CD4-MACS-Multisort-Kit (Miltenyi, Bergisch-Gladbach, Germany).
9. CD25 microbeads (Miltenyi).
10. Anti-FITC Multisort Kit (Miltenyi).
11. Anti-phycoerythrin (PE) microbeads (Miltenyi, optional for isolation of T regulatory cell
subsets).
12. Dynabeads M450: CD8, CD14, CD19, CD25 (Dynal, Hamburg, Germany).
13. MACS Separator (Miltenyi).
14. MACS Separation Columns LS (Miltenyi).
15. Magnetic Particle Concentrator for Dynabeads (Dynal).
16. Sample mixer (i.e., MX1, Dynal).
2.2. Isolation and Testing of Murine Regulatory T-Cells
1. Erythrocyte lysis: Geys lysis buffer: 8.29g NH4Cl, 1g KHCO3, 0.0372g EDTA, pH 7.29;
add to 1 L of H2O and filtrate with 0.2 m.
2. GM buffer: PBS 1X (w/o Ca2+, Mg2+), 0.5% BSA, 0.01% NaN3, 5mM EDTA.
3. PBS 1X (w/o Ca2+, Mg2+), 10% FCS, 4C.
4. PBS 1X (w/o Ca2+, Mg2+), 0.5% BSA, 4C.

5. Culture medium: IMDM + 5% FCS + 1 mM NaPyr + 2 mM Gln.
6. Isolation: anti-CD25 (7D4), biotinylated (BD PharMingen).
7. Depletion: anti-CD25 (PC61, BD PharMingen), biotinylated with biotinamidocaproate
N-hydroxysuccinimide (Sigma, St. Louis, MO).
8. Isolation: anti-CD4 (H129.19), biotinylated (BD PharMingen).
9. Stimulation: anti-CD3 (145-2C11), anti-CD28 (37.51), both purified using protein G
sepharose (Pharmacia, New York, NY).
10. Phenotypical characterization: FITC- or phycoerythrin (PE)-conjugated anti-CD4 and
anti-CD25 (BD PharMingen).
11. Streptavidin (SA) beads (Miltenyi).
12. Anti-phycoerythrin (PE) beads (Miltenyi).
13. Dynabeads M450:
a. CD8 (Lyt2) beads, 4 108 beads/mL (Dynal) .
b. Pan-B (B220) beads, 4 108 beads/mL (Dynal) .
c. Anti-MAC 1 beads (CD11b): generated by binding 230 g anti-MAC-1 mAb (M7/
70.15.11.5) to 4 108/mL beads (Dynal, M-450 Epoxy).
14. Streptavidin, phycoerythrin labeled (SA-PE, Dianova, Hamburg, Germany).
15. MACS Separation Columns, LS (Miltenyi).
16. Cell strainer, 70 m, nylon (Falcon).
17. MACS Separator (Miltenyi).
18. Magnetic Particle Concentrator for Dynabeads (Dynal).
19. Sample mixer (MX1, Dynal).
3. Methods
The methods described below outline the isolation and functional testing of human
CD4+CD25+ T regulatory cells (5,11) and T regulatory cell subsets as well as the isola-
tion and a functional assay for murine regulatory T-cells.
3.1. Isolation and Functional Analysis of Human Regulatory T-Cells
The isolation procedure starts with peripheral blood mononuclear cells (PBMC)
isolated by standard gradient centrifugation from normal buffy coats of healthy volun-
teers. However, it is also possible to use PBMC isolated from whole peripheral blood
or leukapheresis products (see Note 1). Importantly, positively selected CD4+CD25+
T-cell subsets have to be used in functional assays immediately after isolation, or at
least have to be preactivated for subsequent functional analysis.
3.1.1. Multisort Positive Selection of CD4+CD25+ T-Cells
The basic scheme for isolation of human CD4+CD25+ regulatory T-cells is shown
in Fig. 1.
3.1.2. Isolation of CD4+ T-Cells (First Step)
1. Before isolation, PBMC are washed two times with 50 mL washing buffer.
2. Afterwards, CD4+ T-cells can be directly isolated using the CD4-MACS Multisort Kit
according to the instructions of the manufacturer. However, to prevent contamination
(i.e., activated monocytes, which also express small amounts of CD4), do not use more
than 4 L beads per 107 PBMC for isolation of CD4+ T-cells (see Note 2). For selection
procedure, use the isolation buffer.
288
288
Jonuleit and Schmitt

Fig. 1. Overview of the isolation techniques for human (A) and murine (B) CD4+CD25+ regulatory T-cells. Human preparations start with
ficoll histopaque-isolated peripheral blood mononuclear cells (PBMC). This method can be used for isolation of regulatory T-cells from periph-
eral blood, buffy coats, or leukapheresis products. The example for isolation of murine CD4+CD25+ T-cells starts with a spleen cell suspension.
However, the method is also useful for single-cell suspensions isolated from other murine organs (i.e., lymph nodes).
3.1.3. Isolation of CD4+CD25+ T-Cells (Second Step)

1. After detaching from CD4-Multisort Microbeads, isolate the CD25+ fraction using CD25
micobeads (2 L beads per 107 CD4+ cells) (see Notes 1 and 2).
2. This method results in a purity of >97% CD4+CD25+ T-cells. However, the CD25
microbeads can not be detached. Therefore, it is not possible to use microbeads for the
subsequent isolation of T regulatory cell subsets (see Note 3).
3.1.4. Alternative Isolation of CD4+CD25+ T-Cells

Alternatively, the CD25+ T-cell subset can be isolated by the combination of anti-
CD25-FITC monoclonal antibodies (mAb) and the anti-FITC Multisort Kit (see Note 5).
1. After detaching of CD4-Multisort Microbeads, incubate the T-cells for 15 min with anti-
CD25-FITC (1020 g per 108 CD4+ T-cells).
2. Wash two times with washing buffer.
3. Select positively with anti-FITC Multisort Microbeads (3 L/107 CD4+ T-cells) accord-
ing to the manufacturers instructions.
4. After detaching, this preparation can be used for isolation of distinct CD4+CD25+ T-cell
subsets (i.e., CD45RO+ vs CD45RO cells using anti-CD45RO-PE in combination with
anti-PE microbeads).
3.1.5. Combination of Positive and Negative Selection

for Isolation of CD4+CD25+ T-Cells
Alternatively, CD4+CD25+ T-cells can be isolated from PBMC by a combination
of positive selection with microbeads (Miltenyi) and depletion of contaminating cells
with Dynabeads (Dynal, see Note 4).
3.1.5.1. POSITIVE SELECTION OF CD25+ PBMC
1. Washed PBMC are incubated with CD25 microbeads (2 L/107 PBMC) for 20 min at 4C
in isolation buffer (1 108/mL).
2. Afterwards, the cells are washed two times in PBS.
3. The CD25+ fraction is isolated using a MACS separator (Miltenyi) according to the instruc-
tions of the manufacturer.
3.1.5.2. DEPLETION OF CD4-NEGATIVE CELLS WITH DYNABEADS

The positively selected CD25+ fraction contains 6580% CD4+ T-cells and 2035%
contaminating CD19+ B-cells, CD8+ T-cells, and a few CD14+ monocytes. The con-
taminating cells can be depleted with Dynabeads. The following amounts of beads are
used: CD19 Dynabeads: 2 beads/cell, CD8 Dynabeads: 3 beads/cell, CD14 Dynabeads:
1 bead/cell.
1. Wash the collected Dynabeads two times in 15-mL tubes with depletion buffer using the
magnetic particle concentrator (Dynal).
2. Add the CD25+ PBMC fraction (5 107/mL) in depletion buffer and incubate the suspen-
sion for 20 min at 4C on a shaker (sample mixer, Dynal).
3. Deplete contaminating cells according to the manufacturers instructions by the use of
the magnetic particle concentrator.
4. Repeat this Dynabeads depletion once for higher purity of CD4+CD25+ T-cells (>98%
after two rounds of depletion).
3.1.6. Positive Isolation of Conventional CD4+ T-Cells (CD25 Depleted)

Co-expression of CD4 and CD25 is not sufficient for characterization of human regu-
latory T-cells. CD4+CD25+ regulatory T-cells are defined by their functional activity to
suppress the activation of conventional T-cells. Therefore, to prevent experimental arti-
facts, analysis of functional activities of freshly isolated CD4+CD25+ T-cell prepara-
tions is mandatory. Several protocols for isolation of conventional CD4+ T-cells are
useful and well established. In the following, we briefly describe one easy method for
isolation of CD25-depleted CD4+ T helper cells.
1. For direct isolation of CD4+ T-cells, CD4 microbeads (Miltenyi, 24 L/107 PBMC) can
be used according to the manufacturers instructions.
2. Contaminating CD4+CD25+ T-cells can be depleted in a second step by using CD25
Dynabeads (0.5 beads/cell). This depletion procedure is carried out as described under
Subheading 3.1.5.2. and results in highly purified CD4+CD25 T helper cells (>98%).
3.1.7. Functional Analysis of Human Freshly

Isolated CD4+CD25+ T-Cell Preparations
CD25 is a helpful surface marker for isolation of regulatory T-cells. However, CD25
is not specific for regulatory T-cells. This marker is also expressed on conventional
CD4+ and CD8+ T-cells soon after activation. Therefore, control of the suppressive
activities of human CD4+CD25+ T-cell preparations in suitable functional assays is
extremely important for human regulatory T-cell preparations. In the following, two
assays will be described (12).
3.1.7.1. ALLOGENEIC MIXED LEUKOCYTE REACTION (MLR)

For an optimal stimulation of freshly isolated T-cell populations in an antigen-spe-
cific (alloantigen) manner, mature dendritic cells (DC) from allogeneic donors gener-
ated by standard methods can be used (13).
1. Freshly isolated CD4+ as well as CD4+CD25+ T-cell subsets (1 105/well) are cultured in
the presence of allogeneic DC in 96-well plates (200 L/well, DC:T-cell ratio 1:10 or
1:20) in X-VIVO 15 (RPMI + 5% FCS can also be used).
2. After 4 d of incubation, T-cell proliferation is measured; for an additional 16 h, cells are
pulsed with 3H-TdR (37 kBq/well), and incorporated radioactivity is measured by using a
liquid scintillation counter.
3. Co-cultures of CD4+ and CD4+CD25+ T-cells activated with allogeneic DC have to be
performed to analyze the suppressive activity of CD4+CD25+ T-cells on the proliferation
of conventional CD4+ T-cells. Therefore, a constant number of CD4+ T-cells is co-cul-
tured with different numbers of CD4+CD25+ T-cells (ratio 1:1 to 1:4), and both T-cell
populations are activated with a constant ratio of allogeneic DC.
3.1.7.2. POLYCLONAL STIMULATION WITH ANTI-CD3 AND ANTI-CD28 MAB

As an alternative to allogeneic stimulation:
1. Activate polyclonally constant numbers of conventional CD4+ T-cells (1 105/well) with
anti-CD3 (1 g/mL, OKT-3) and CD28 mAb (2 g/mL, CD28.2) in the presence of vary-
ing numbers of CD4+CD25+ T-cells (ratio 1:1 to 1:4).
2. T-cell proliferation is measured after 3 d of culture and an additional 16 h pulse with 3H-
TdR (37 kBq/well).
Typical results of this standard assays for suppressive activity of freshly isolated
human CD4+CD25+ regulatory T-cells are shown in Figure 2.
3.2. Isolation and Functional Analysis of Murine Regulatory T-Cells

Spleens from different strains of mice can be used as a source for primary CD4+ and
CD4+CD25+ T-cells. Single-cell suspensions of spleen cells are subjected to erythrocyte
lysis (lysis buffer) and subsequently the debris is removed by passing through a cell
strainer. Cell number is determined, and after centrifugation at 600g for 10 min, cells are
suspended in cold (4C) GM buffer in a final concentration of 1 108 cells/mL.
3.2.1. Isolation of Murine CD25+CD4+ T Regulatory Cells

3.2.1.1. POSITIVE SELECTION OF CD25+ CELLS BY MACS
1. Spleen cells are stained with biotinylated anti-CD25 mAb (7D4) in a concentration of
0.5 g/mL and incubated for 20 min at 4C.
2. Cells are washed twice with GM buffer and suspended in cold GM buffer (1 108 cells/mL).
3. Afterwards, streptavidine-phycoerithrine (PE), in a dilution of 1/400, is added, and cells
are incubated for an additional 20 min at 4C.
4. Cells are washed twice with GM buffer and diluted in cold GM buffer (5 108 cells/mL).
Anti-PE beads (50 L) are added to 1 mL cell suspension (5 108 cells/mL) and cells are
incubated for 30 min on a shaker at room temperature. The suspension is filled up to 5 mL
with GM buffer and rinsed through a cell strainer to remove cell clusters.
5. Afterwards, the cell suspension is loaded slowly on top of a MACS Separation Column
(LS) that is fixed in a MACS separator, and CD25+ cells are positively enriched accord-
ing to the manufacturers instructions.
Repeat this positive selection step once. Finally the cells are washed and suspended in
1 mL of PBS + 0.5% BSA. (see Note 6).
3.2.1.2. DEPLETION OF CD4 CELLS BY DYNABEADS

The resulting cell population consists of 7080% CD25+CD4+ T-cells and 2030%
contaminating cells, mainly B-cells, CD8+ T-cells, and macrophages. These contami-
nating cell types are depleted using Dynabeads directed against B220, CD8, and MAC-
1 (see Notes 7 and 8).
1. Add the cell suspension in a fourfold excess regarding the estimated numbers of the con-
taminating cell types, and incubate on a shaker for 20 min at room temperature (RT).
2. Depletion is performed using a magnetic particle concentrator according to the manu-
facturers instructions.
3. Repeat this depletion step once.
4. Finally, the cells are washed and suspended in culture medium. Typically, this procedure
yields a highly pure (>97%) population of murine CD25+CD4+ T regulatory cells (Fig. 1B).
3.2.2. Preactivation of Murine CD25+CD4+ T Regulatory Cells

Freshly isolated CD25+CD4+ T regulatory cells cannot inhibit the activation of dif-
ferentiated T helper cells, indicating that they have a limited suppressive capacity.
292
On the other hand, preactivated CD25+CD4+ T regulatory cells exert a potent suppres-
sive activity and also inhibit differentiated T helper cells (Fig. 2B). Thus, preactivation
is indispensable for the generation of highly suppressive CD25+CD4+ T regulatory cells.
Preactivation of CD25+CD4+ T regulatory cells is performed using a combination
of plate-bound anti-CD3 mAb and anti-CD28 mAb.
1. Twenty-four-well culture plates are coated with anti-CD3 mAb (145-2C11, 5 g/mL,
diluted in PBS) at 37C.
2. After 30 min, the wells are washed (PBS), secondarily coated with anti-CD28 mAb (37.51,
10 g/mL, diluted in PBS), and incubated for additional 30 min at 37C.
3. After an intensive wash (PBS, three times) of coated plates, the CD25+CD4+ T regulatory
cells are stimulated on these plates in culture medium, harvested after 48 h, and used as
preactivated CD25+CD4+ T regulatory cells.
3.2.3. Isolation of Murine Conventional CD4+ T-Cells (CD25 Depleted)
So far there is no unique marker defined for murine CD25+CD4+ T regulatory cells,
and CD25 is also highly expressed by activated conventional T-cells. Thus, murine
CD25+CD4+ T regulatory cells have to be characterized using a bioassay. In the fol-
lowing, we briefly describe one easy method for isolation of CD25-depleted CD4+
conventional T-cells for the functional analysis of regulatory T-cells.
3.2.3.1. DEPLETION OF CD25+ T-CELLS
1. Spleen cells are stained with biotinylated anti-CD25 mAb (PC61; 0.5 g/mL) and incu-
bated for 20 min at 4C.
2. Cells are washed twice with cold GM buffer and suspended in cold GM buffer (1
10 8/mL).
3. Add streptavidine beads (Miltenyi), dilute 1/20, and incubate for 20 min at 4C. Fill up
the suspension to 5 mL with GM buffer and rinse the cells through a cell strainer to
remove cell clusters.
Load this suspension slowly on top of a MACS Separation Column (LS) that is fixed in
a MACS separator, and deplete the CD25+ cells according to the manufacturers instruc-
tions.
Fig. 2. (opposite page) Examples of phenotypic and functional characterization of human

(A) and murine (B) CD4+CD25+ T regulatory cells. Phenotypically, freshly isolated human T
regulatory cells express CD4 and CD25 homogeneously. The majority of T-cells also express
CD45RO and intracellulary CTLA-4. The marker CD45RO can be used for further purification
of human T regulatory cells as described in the methods section. CTLA-4 will be upregulated
and expressed on the surface after activation of CD4+CD25+ T regulatory cells in vitro. On the
right side, two examples for functional analysis of freshly isolated human CD4+CD25+ T-cells
are shown. After allogeneic as well as after polyclonal stimulation, proliferation of CD4+CD25+
T regulatory cells themselves is marginal. Nevertheless, in co-culture they suppress the activa-
tion of conventional T-cells dose dependently.
Murine Th2 cells are generated by stimulating CD4+ T-cells in the presence of interleukin
(IL)-4 and anti-interferon (IFN)- mAb for 6 d. Restimulation of these cells leads to a consider-
able production of IL-4 (detected by an ELISA) that cannot be suppressed by freshly isolated
CD25+ T regulatory cells but by preactivated CD25+ T regulatory cells (preCD25+).
3.2.3.2. POSITIVE SELECTION OF CD4+ T-C-ELLS

1. The CD25-depleted cell suspension is stained with biotinylated anti-CD4 mAb (H129.19;
0.5 g/mL) and incubated for 15 min at 4C.
2. Cells are washed twice with GM buffer and suspended in cold GM buffer (1 108/mL).
Afterwards, streptavidine beads (Miltenyi, dilution 1/40) are added and incubated for 15
min at 4C.
3. Fill up the suspension to 5 mL with GM buffer and rinse the cells through a cell strainer to
remove cell clusters.
4. Load this suspension slowly on top of a MACS Separation Column (LS) that is fixed in a
MACS separator, and positively enrich the CD4+ T-cells according to the manufacturers
instructions.
5. Repeat this positive selection step.
6. Finally, the cells are washed and suspended in culture medium.
3.2.3.3. FUNCTIONAL ANALYSIS OF MURINE CD25+CD4+ T REGULATORY CELLS

1. CD4+ T-cells (1 106/mL), depleted of CD25+ T-cells (anti-CD25 mAb: PC61) and iso-
lated by two rounds of positive selection using a MACS column, are polyclonally acti-
vated by anti-CD3 mAb (13 g/mL) in the presence of mitomycin C-treated A20
B-tumor cells (1 105/mL) as accessory cells (see Note 9). Simultaneously, the CD25-
depleted CD4+ T-cells are activated in the presence of CD25+CD4+ T regulatory cells
(ratio 1:1); only the purified CD25+CD4+ T regulatory cells are activated.
2. After 4 d, 3H-TdR (18.5 kBq/well) is added for 18 h and thymidine uptake of the T-cells
is monitored by liquid scintillation counting.
As shown in Fig. 2B, CD4 + T-cells undergo a strong proliferation, whereas
CD25 +CD4 + T regulatory cells alone and the combination of CD4 + T-cells and
CD25+CD4+ T regulatory cells exhibit no or only a minimal thymidine incorporation,
indicating the strong suppressive capacity of CD25+CD4+ T regulatory cells.
4. Notes
1. CD4 and CD25 are nonspecific surface markers for regulatory T-cells, and the ratio of
CD25+ cells in peripheral blood depends on the immune status of the individual donor.
Therefore, percentages of CD4+CD25+ T-cells vary among individual donors. To prevent
artificial preparation of conventional (preactivated) CD25+ T-cells, the ratio of CD25+
PBMC should be analyzed before isolation, and only PBMC preparations that contain
less than 5% CD25+ cells should be used for isolation of CD4+CD25+ regulatory T-cells.
2. The described technique is optimized for isolation of human CD4+CD25+ regulatory T-
cells with high purity. Use of higher concentrations of antibodies or beads for isolation
might enhance the recovery but decreases the purity. If large numbers of CD4+CD25+
regulatory T-cells are required, use the combination of positive and negative selection as
described under Subheading 3.1.5. This technique results in higher recoveries compared
to positive selection using Multisort beads. Note that the enzymatic release of microbeads
is necessary if the separated cells will be used for additional steps of magnetic cell sort-
ing. However, this enzymatic release decreases the recovery by 25%.
3. CD25 microbeads are very efficient for isolation of CD25+ T-cells with high purity. How-
ever, it is difficult to stain positively selected CD25+ T-cells with anti-CD25 mAb (most
of the CD25 epitopes are engaged by microbeads). Therefore, for phenotypical control of
the purity of CD25+ separated cells, anti-CD25 mAb (FITC- or PE-conjugated) can be
mixed with CD25 microbeads (ratio of microbeads to anti-CD25 mAb = 3:1) before sepa-
ration. By this technique, the cells will be stained directly during the separation pro-
cedure.
4. The described technique is optimized for high purity of CD4+CD25+ T-cells and not for
the highest recovery of positive cells. The mean recovery of CD4+CD25+ T-cells is
50% of total marker-positive cells, depending on the individual donor and the method
of isolation. As described earlier, the use of Multisort beads reduces the recovery sig-
nificantly but results in preparations of high purity. The combination of positive selec-
tion with CD25 microbeads and depletion of contaminations by Dynabeads is a
compromise between recovery and purity of CD4+CD25+ T-cells, and this technique
reduces the costs compared to the Multisort positive selection.
5. Preparations of human CD4+CD25+ T-cells contain regulatory T-cells as well as conven-
tional (preactivated) T helper cells, and the ratio of contamination varies from donor to
donor. For higher purity of CD4+CD25+ regulatory T-cells, the CD45RA+ contaminating
cells (1025%) can be depleted by anti-CD45RA mAb in combination with anti-mouse-
IgG Dynabeads (Dynal) directly after positive selection step using CD25 microbeads.
Alternatively, the CD45RO+ fraction can be positively selected using CD45RO micro-
beads (Miltenyi). However, for positive separation, use Multisort-selected cells only. This
step enhances the purity of isolated regulatory T-cells. Preparations isolated by anti-
CD25-FITC mAb plus anti-FITC Multisort beads also offers the possibility of separating
distinct subsets of human CD4+CD25+ regulatory T-cells by positive selection with
microbeads (i.e., 7-integrin+ vs 7-integrin- regulatory T-cells using 7-PE mAb plus
anti-PE-microbeads).
6. Spleen cells should be incubated in as small as possible a volume of buffer (e.g., 2 109
spleen cells in 2 mL). A second step of CD25 positive selection leads to a considerable
enrichment of CD25+ cells, whereas a third step does not, but results in a substantial loss
of cells.
7. The CD25+CD4+ T-cells are usually contaminated with 2030% other cell types, mainly
B-cells, CD8+ T-cells, and macrophages. These cells can be depleted using Dynabeads,
since the Magnetic Particle Concentrator from Dynal does not affect the microbeads
from Miltenyi, so that microbead-loaded CD25+CD4+ T-cells are not retarded in the
concentrator.
8. Based on the assumption that the contaminating cell population comprises about 30% of
total CD25+ cells and consists of 50% B-cells, 25% CD8+ T-cells, and 25% macrophages,
the cell numbers of the different T-cell types are calculated and the specific Dynabeads
(anti-CD8, -B220, -MAC-1) are used at a ratio of 4 Dynabeads:1 cell.
9. For the functional analysis of the CD25+CD4+ T-cells, it is important to use only limited
concentrations of anti-CD3 mAb, roughly in the range of 13 g/mL. In the presence of
higher concentrations of anti-CD3 mAb, the suppressive capacity of the CD25+CD4+ T-
cells will not be sufficient to inhibit the proliferation of the co-cultured conventional
CD4+ T-cells. The same holds true in case anti-CD28 mAb are used as an additional
costimulator.
Acknowledgments
The authors are grateful to Drs. M. Stassen, S. Fondel, and Ch. Mller for critical
reading of this manuscript and helpful discussions. This work was supported by the
Deutsche Forschungsgemeinschaft grant A6SFB548 (to E. Schmitt) and grant
A8SFB548 (to H. Jonuleit).
References
1.
1 Gershon, R. K. (1975) A disquisition on suppressor T cells. Transplant. Rev. 26, 170185.
2.
2 Sakaguchi, S., Sakaguchi, N., Asano, M., Itoh, M., and Toda, M. (1995) Immunologic
self-tolerance maintained by activated T cells expressing IL-2 receptor alpha-chains
(CD25). Breakdown of a single mechanism of self-tolerance causes various autoimmune
diseases. J. Immunol. 155, 11511164.
3.
3 Jonuleit, H., Schmitt, E., Stassen, M., Tuettenberg, A., Knop, J., and Enk, A. H. (2001)
Identification and functional characterization of human CD4(+)CD25(+) T cells with regu-
latory properties isolated from peripheral blood. J. Exp. Med. 193, 12851294.
4.
4 Dieckmann, D., Plottner, H., Berchtold, S., Berger, T., and Schuler, G. (2001) Ex vivo
isolation and characterization of CD4(+)CD25(+) T cells with regulatory properties from
human blood. J. Exp. Med. 193, 13031310.
5.
5 Ng, W. F., Duggan, P. J., Ponchel, F., et al. (2001) Human CD4+ CD25+ cells: a naturally
occurring population of regulatory T cells. Blood 98, 27362744.
6.
6 Seddon, B. and Mason, D. (2000) The third function of the thymus. Immunol. Today 21,
9599.
7.
7 Seddon, B. and Mason, D. (1999) Peripheral autoantigen induces regulatory T cells that
prevent autoimmunity. J. Exp. Med. 189, 877882.
8.
8 Thornton, A. M. and Shevach, E. M. (1998) CD4+CD25+ immunoregulatory T cells sup-
press polyclonal T cell activation in vitro by inhibiting interleukin 2 production. J. Exp.
Med. 188, 287296.
9.
9 Suri-Payer, E., Amar, A. Z., Thornton, A. M., and Shevach, E. M. (1998) CD4+CD25+ T
cells inhibit both the induction and effector function of autoreactive T cells and represent
a unique lineage of immunoregulatory cells. J. Immunol. 160, 12121218.
10.
10 Piccirillo, C. A. and Shevach, E. M. (2001) Cutting edge: control of CD8+ T cell activa-
tion by CD4+CD25+ immunoregulatory cells. J. Immunol. 167, 11371140.
11.
11 Jonuleit, H., Schmitt, E., Kakirman, H., Stassen, M., Knop, J., and Enk, A. H. (2002)
Infectious tolerance: human CD25(+) regulatory T cells convey suppressor activity to con-
ventional CD4(+) T helper cells. J. Exp. Med. 196, 255260.
12. Jonuleit, H., Schmitt, E., Schuler, G., Knop, J., and Enk, A. H. (2000) Induction of
interleukin 10-producing, nonproliferating CD4(+) T cells with regulatory properties by
repetitive stimulation with allogeneic immature human dendritic cells. J. Exp. Med. 192,
12131222.
13. Jonuleit, H., Kuhn, U., Mller, G., et al. (1997) Pro-inflammatory cytokines and prostag-
Monoclonal Antibody-Based Strategies 297
21
Monoclonal Antibody-Based Strategies
in Autoimmunity and Transplantation
Lucienne Chatenoud
Summary
Monoclonal antibodies are homogeneous sets of immunoglobulins with well-defined speci-
ficity and biochemical characteristics. They were introduced into clinical practice in the early
1980s, and since then their use has rapidly expanded.
Most of the side effects observed with first-generation murine (mouse or rat) antibodies
have been successfully overcome with the advent of humanized (chimeric or CDR-grafted) and
more recently fully human antibodies. Our aim is to review the major steps in the development
of therapeutic monoclonal antibodies in the fields of transplantation and autoimmunity, and
discuss the salient features of antibodies presently used in the clinicnotably anti-T-cell and
anticytokine antibodies. The discussion will also focus on the unique capacity of some mono-
clonal antibodies to induce immune tolerance.
Key Words: Monoclonal antibody; immune tolerance; immunosuppression; autoimmune
diseases; transplantation.
1. Introduction
Major human diseases are recognized to have an autoimmune origin, placing auto-
immunity as the third major cause for morbidity and mortality in developed countries
after cancer and atherosclerosis. Moreover, the frequency of autoimmune diseases is
steadily increasing. Organ and bone-marrow transplantation represent the only thera-
peutic possibility for a variety of life-threatening conditions. It is now clear that the
fundamental immune mechanisms underlying both autoimmune diseases and allograft
rejection are similar. In both, the immune response exploits functionally distinct cellular
subsets (macrophages, dendritic cells, B-cells, effector and regulatory T-cells) that inter-
act through an intricate network of membrane receptors and soluble mediators (cyto-
kines and chemokines). For several years, in both situations, therapeutic approaches
were essentially based on immunosuppression, which is non-specificin other words,
is unrelated to the antigens involved. The mainstays of such non-specific immuno-
supression are small molecules (corticosteroids, azathioprine, methotrexate, cyclophos-
297
298 Chatenoud
phamide, cyclosporin, FK506, mycophenolate mofetil, rapamycin), whose activities are

targeted to intracellular processes. Their major drawback is their relative ineffectiveness
in the long term, with the likely risk of recurrence of the pathogenic immune process
once the drug is withdrawn, necessitating indefinite drug administration with the atten-
dant problems of recurrent infection and drug toxicity.
One radical solution would be to discover ways in which the immune system can be
selectively immunosuppressed just to those antigens unique to the target tissue, avoid-
ing the hazards of long-term nonspecific immunosuppression. Operationally, this
means the establishment of immunological tolerance in a mature immune system
that is, a state of lasting antigen-specific unresponsiveness in the absence of general-
ized immunosuppression. More than 20 yr of experience in using immunosuppressive
monoclonal antibodies both in the experimental and clinical fields have clearly dem-
onstrated their unique advantages over conventional immunosuppressants, to combine
both an extremely efficacious immunosuppressive potency and, under particular cir-
cumstances, a tolerance-promoting ability. Thus, antibodies or genetically engineered
molecules specific for distinct functionally relevant T-cell receptors (i.e., CD4, CD3,
CD80, CD86, CD40 ligand [CD40L], adhesion receptors, or adhesins) were shown to
induce immune tolerance to foreign-tissue alloantigens and to autoantigens.
Our aim will be to review the present state of the art on the use of monoclonal
antibodies in autoimmunity and transplantation, presenting both the specificities that
are part of the routine clinical armamentarium as well as those that are presently under
development. In doing so, we shall also address the mode of action of these biological
agents, which constitute very effective tools aimed at controlling deleterious immune
responses to allografts and preventing or even treating established autoimmune dis-
eases through their capacity to restore self-tolerance.
2. Uniqueness of Monoclonal Antibodies As Drugs
Several characteristics relying on both molecular and pharmacological properties
of monoclonal antibodies clearly distinguish them from conventional chemicals. Some
of these features are significant advantages, while others constitute important prob-
lems that have frequently hampered more rapid development and transfer to the clinic.
The factors that will be discussed are listed as pros and cons in Table 1.
2.1. Pharmacological and Biological Versatility
As opposed to conventional chemicals, which mostly act through the removal and/
or functional inhibition of their targets, monoclonal antibodies present a wide spec-
trum of pharmacological and biological activities that impact the capacity to redirect
immune responses. Thus, depending on their fine specificity, monoclonal antibodies
can: (1) neutralize soluble mediators such as cytokines and chemokines, which play
an essential role in orchestration of immune responses and immune cell recruitment;
(2) physically remove target cells (i.e., the use of either nonmanipulated monoclonal
antibodies or toxin or radioisotope-labeled antibodies) thus raising the concern of the
turnover of the depleted cell population; (3) inhibit or block the functional capacity
of target cells without physically eliminating them; and (4) function as pharmaco-
logical agonists that signal through specialized immune cell receptors and their asso-
Table 1
Uniqueness of Monoclonal Antibodies As Drugs
Pro Con
Pharmacological versatility Cumbersome and expensive production
Fine specificity for any target Need for extensive quality controls
Rapidity of clinical development (biochemical, virology...)
Long half-life for humanized and human Sensitization
monoclonal antibodies
ciated intracellular pathways to trigger immunoregulatory T-cells that are able to con-
trol the pathogenic activity of autoreactive or alloreactive lymphocyte effectors.
2.1.1. Cytokine Neutralization
In order to neutralize the biological effect of cytokines, one must effectively inter-
fere with their binding to specific receptors. This can be achieved by different means,
including the use of cytokine-specific antibodies, soluble cytokine receptors (13),
antagonists of cytokine receptors (as in the case if interleukin [IL]-1) (4), and geneti-
cally engineered immunoadhesins, including two soluble receptor fragments linked to
an immunoglobulin constant frame (5,6).
Pathological conditions that may benefit from therapeutic strategies aimed at cyto-
kine neutralization are, first, diseases associated with overproduction of pro-
inflammatory cytokines, i.e., tumor necrosis factor (TNF), IL-1, and IL-6, and, secondly,
pathological situations involving an immunologically mediated target tissue damage
in which cytokines driving T helper 1 (Th1) (i.e., interferon [IFN]-, IL-12) or Th2
responses (i.e., IL-4, IL-10) exert a central immunoregulatory role.
Concerning proinflammatory cytokines, the use of agents that neutralize TNF was,
as we shall discuss in detail below, a major breakthrough in the clinical management
of some autoimmune diseases, such as rheumatoid arthritis and psoriasis, as well as in
immune-mediated disease such as inflammatory bowel disease (7).
Concerning the blockade of immunoregulatory cytokines, the case of IFN- is of
particular interest, given the compelling data obtained from animal models and the
present ongoing phase II trial in inflammatory bowel disease (810). In explaining the
therapeutic effect of anti-IFN- antibodies, one should consider not only their capacity
to dampen Th2 responses, but also their ability to block the IFN--mediated
upmodulation of major histocompatibility complex (MHC) class I and class II antigen
expression. In fact, this capacity of IFN- to promote aberrant class II MHC anti-
gens on cells that do not normally express them (astrocytes, keratinocytes, thyroid,
and pancreatic islet cells) represents a major triggering factor in autoimmune diseases
(11,12). In human allograft recipients, acute rejection was often associated with the
de novo expression of class II HLA antigens on cells within the allograft, such as
kidney tubular cells or myocytes (13,14). Such de novo MHC class II antigen expres-
sion increases the antigenicity of the graft and the suceptibility of such cells to specific
cytotoxic T-cell effectors.
300 Chatenoud
2.1.2. Cellular Depletion

Following the in vivo injection of polyclonal anti-T-cell antibodies, cell lysis mainly
occurred through complement-dependent lysis in the circulation or cell opsonization
and removal in the liver and the spleen by reticuloendothelial cells (15,16).
On the other hand, with cell-directed monoclonal antibodies, the mechanisms
involved in antibody-mediated cell destruction in vivo are more varied and complex.
Actually, not only the antibody isotype (i.e., the capacity of the Fc antibody portion to
interact with Fc receptors to mediate antibody-dependent cell cytotoxicity [ADCC]
and to activate complement) but also its fine specificity can influence the lytic capac-
ity. Thus, it is not sufficient to select the adequate immunoglobulin constant domain to
achieve the desired in vivo depleting or nondepleting effect (17,18). First-generation
mouse and rat monoclonal antibodies to cell antigens mostly expressed, when injected
into humans, a rather poor lytic capacity. This was due to the now well-characterized
membrane-bound factors that inhibit activation of human complement in a species-
restricted manner. The situation of course evolved with the advent of humanized, chi-
meric, or complementarity determining region (CDR)-grafted monoclonal antibodies,
and more recently of fully human monoclonals.
However, it is interesting to discuss at this point that, independently from their
isotype, some monoclonal antibody specificities, for reasons still ill defined, are highly
depleting. This is the case for antibodies to CD52 (i.e., CAMPATH-1 antigen) a small
glycosylphosphatidylinositol (GPI)-linked protein abundantly expressed on B-cells,
T-cells, and monocytes that were extensively used in bone-marrow transplantation
and hematologic malignancies (17). Using DNA technology, sets of recombinant
monoclonal antibodies with identical specificities but different constant regions were
probed for the sequences that determine antibody function, and ranked in human
complement-mediated lysis and ADCC with regard to efficiency (18). These studies
showed first that both the fine specificity (independently from the isotype) and the
density of distribution of the target antigen-influenced lytic capacity of cell-binding
monoclonal antibodies. Secondly, among given sets of monoclonal antibodies sharing
the same specificity (i.e., variable region), a hierarchy in terms of lytic capacity could
be established that in the case of human immunoglobulins is as follows: IgG1>IgG2
>IgG3>>IgG4.
For obvious reasons, a cell antigen that easily undergoes antigenic modulation,
namely the disappearance of a receptor from the cell membrane upon binding of a
specific ligand (in this case the monoclonal antibody), will be a poor target for lysis.
This is the case for T-cell antigen receptors (CD3-TCR) (see Subheading 2.1.3.). The
lytic potential may, in this case, be significantly improved using univalent antibodies
that avoid antigenic modulation (19,20).
Among mechanisms involved in cell lysis and linked to monoclonal antibody fine
specificity, one must mention redirected T-cell lysis occurring upon the bridging of
cytotoxic T-cells to the target (21) and, in particular, for antibodies to CD3 and CD4
faced with activated T-cells, the induction of apoptosis, or programmed cell death (22).
Finally, one may add here immunotoxins that are conjugates between a toxin
(diphteria toxin or ricin) and a monoclonal antibody or a cytokine (23,24). The toxin is
specifically directed to the target cell through the antibody or the cytokine, and kills
the target following internalization.
The immunological effects of cell depletion will closely depend on the cellular
selectivity of the monoclonal antibody (i.e., generally affecting leukocytes [antibodies
to CD52], T-cells [antibodies to CD3], T-cell subsets [antibodies to CD4, CD8], or
only activated T-cells [antibodies CD25]) and on the renewal rate of the target cellular
population. Pools of short-lived lymphocytes will rapidly reappear after short-term
depletion. Conversely, recovery of recirculating long-lived T-cells will take several wk
in the mouse and several mo in man. In this context, the data using depleting mono-
clonal antibodies to CD4 is both intriguing and deserves a word of clinical caution.
Thus, the chimeric (humanized) antibody to CD4 MT-512 (human IgG1), used in both
transplanted and autoimmune patients, induced, especially when administered at high
doses (700 mg cumulative dosage), extremely pronounced and prolonged (up to 60%
from baseline for over 1830 mo) peripheral CD4+ cell depletion (25).
2.1.3. Functional Inhibition

Among the mechanisms accounting for antibody-mediated inhibition of cell func-
tion are (a) antibody-mediated cell coating or blindfolding, interfering with cell-to-
cell interactions; (b) antigenic modulation of a functionally relevant target cell
receptor; and (c) induction of cell unresponsiveness (anergy) through modulation of
activation signals delivered.
The relevance of antibody-mediated cell coating was initially described with
polyclonal anti-lymphocyte antibodies, since the induced lymphocytopenia did not
appear as the only factor mediating their therapeutic effect. In fact, T-cells that reap-
peared after treatment were functionally impaired both in vitro and in vivo. Mono-
clonal antibodies to human, mouse, and rat CD4 have been described expressing both
immunosuppressive and tolerogenic capacities that are nondepleting and coat their
cell target (2630). It is likely that CD4 coating, given the contribution of CD4 to
TCR/CD3-mediated signaling, not only interferes with lymphocyte binding to MHC
class II on antigen-presenting cells, but also triggers therapeutically relevant negative
signals (3136). Thus, independent ligation of CD3 and CD4 in vitro, as it may occur
in vivo following injection of monoclonal antibodies to CD4, significantly decreases
proliferative responses and lymphokine production induced by stimulation with
soluble antigens, alloantigens, and polyclonal stimulants.
Some cell-directed antibodies also induce antigenic modulationthe redistribu-
tion of their ligands or receptors on the cell surface. The antigenantibody complex
will eventually form a cap at a pole of the cell and disappear upon internalization or
shedding (37). Such loss of expression of a functionally relevant membrane receptor
will render the target cell functionally inactive. This is well documented for antibodies
to CD3 (3739). Depletion is only partial in both mice and humans at usual doses. In
terms of phenotype, cells undergoing antigenic modulation are CD3-TCR-CD4+ or
CD3-TCR-CD8+. The phenomenon is reversible within 812 h once the antibody is
cleared from the surrounding medium. Cells undergoing CD3 antibody-mediated anti-
genic modulation are unresponsive to in vitro antigen-specific or mitogen stimulation
302 Chatenoud
(37,39), a pattern that correlates well with the significant clinical immunosuppressive
effect of CD3 antibodies.
As previously mentioned, sustained T-cell activation upon engagement of CD3-
TCR needs the simultaneous delivery of costimulatory signals through specialized
receptors, such as CD28, CTLA-4, CD40L, ICOS interacting with their ligands B7.1,
B7.2, CD40, ICOSL at the surface of antigen-presenting cells. Several in vitro models
using both murine and human clones showed that occupancy of the TCR alone, in the
absence of adequate costimulation, induced a state of unresponsiveness termed anergy
(40,41). The role of anergic cells in physiological or induced immune tolerance is still
a matter of debate. However, one cannot exclude that monoclonal antibodies targeting
some of the receptors listed above exert their in vivo therapeutic effect at least in part
through T-cell anergy.
2.1.4. Triggering of Immunoregulatory T Cells
Although less well defined in cellular and molecular terms, another crucial in vivo
effect of some anti-T-cell monoclonal antibodies is the triggering of immuno-
regulatory pathways leading in the long term to the establishment of an operational
immune tolerance relying on T-cell immune deviation and/or dominant or infectious
tolerance, depending on the models.
The unique tolerogenic capacity of antibodies to CD4 has been known for a long
time, since antibodies to CD4 injected into mice did not trigger an antiglobulin response,
as was the case with the majority of rodent monoclonal antibodies against other T-cell
antigens (42,43). Moreover, specific tolerance could be induced into normal adult
euthymic animals if soluble antigens were administered under the cover of depleting or
non-depleting antibodies to CD4 (42,43). Such immune tolerance was maintained
through the delivery of the antigen alone, at regular intervals, in the absence of any
further monoclonal antibody treatment. These findings using antibodies to CD4 were
rapidly extended to tissular antigens such as alloantigens (especially using antibodies to
CD4 and CD8 in association) and autoantigens (26,28,44,45). In parallel, compelling
evidence has been accumulated to show that targeting of other T-cell receptors, and
in particular of CD3 and costimulatory receptors, was also very effective in inducing
tolerance in both transplantation and autoimmune settings (29,44,4655).
In terms of mechanisms, the immune tolerance induced in several of these models is
active or dominant, i.e., mediated by subsets of CD4+ immunoregulatory or sup-
pressor T-cells that control pathogenic effectors. It can be transferred to nave recipients
upon transfusion of such CD4+ immunoregulatory subsets, and can be maintained only
in the presence of the tolerogenic antigen (29,44,4656). The precise phenotype of these
regulatory T-cells varies depending on the model; some of them express the chain of
the IL-2 receptor (CD25), as naturally occurring suppressor T-cells shown to maintain
physiological tolerance do, whereas others appear to emerge from peripheral
CD4+CD25 T-cells (29,5763). The same variability applies to their cytokine depen-
dency: only in some models a role for immunoregulatory cytokines (i.e., IL-4 or TGF-)
has been demonstrated (55,64). Several groups are presently involved in the character-
ization of genes or molecules selectively expressed by such regulatory T-cells that may
lead to a more precise understanding of their mode of action (27,65).
2.2. Monoclonal Antibodies Can Be Produced Against Almost Any Target

New protein targets are continuously emerging from genomics and proteomics. One
of the main aims of these approaches is to discover new drug classes targeting these
new molecules. This has proven difficult, however, for conventional small molecules,
in spite of the accessibility to large libraries, because of the need in each case for good
and reliable bioassays.
The monoclonal antibody approach has the major advantage of initially avoiding
such bioassays. One may directly produce monoclonals against a new protein once it
has been adequately expressed, and test the pharmacological (functional) properties of
the antibodies in conventional in vitro and in vivo assaysnamely, various immune
functions in the case of antibodies targeting molecules related to the immune system. In
fact, it has become a common practice of many pharmaceutical companies to produce
monoclonals against newly discovered molecules, notably using the very versatile
phage assay techniques. In this way, functionally relevant antibodies can be produced
against molecules whose physiological and physiopathological roles are unknown.
As a meaningful example, one may quote the case of TNF. Despite all the experi-
mental suggestions arguing for a central role for TNF in the pathophysiology of rheu-
matoid arthritis, only anti-TNF monoclonal antibodies and fusion proteins provided
the tools to effectively and specifically neutralize its biological activity. Until then, all
conventional anti-inflammatory agents had failed to provide such a radical effect.
2.3. Factors That Impact on Clinical Development
2.3.1. Less Toxicity
Due to their very strict target specificity (with the exception of rare and unexpected
undesirable cross-reactions), therapeutic monoclonal antibodies usually do not express
side effects unrelated to their pharmacological recognition of the initial target mol-
ecule. This is a major advantage over small chemicals that very often display pharma-
cological activities (some of which lead to significant side effects) unrelated to the
target responsible for the relevant pharmacological effect. Thus, in the case of immu-
nosuppressive agents, many drugs inhibit the proliferation of many cell types other
than lymphocytes, notably other white blood cells, and many of them show direct
toxicity for non-lymphoid organs, as in the case of cyclosporin nephrotoxicity.
Finally, the possibility of assessing the toxicity of monoclonal antibodies is often
limited by their species specificity.
2.3.2. High-Scale Production
This still represents a real problem for the clinical development of monoclonal
antibodies. Only a few centers worldwide are able to produce high amounts of good
manufacturing practice (GMP)-grade antibodies for clinical trials. Due to present
regulations, this in fact implies exclusive in vitro production using adapted bioreactors
to grow validated cell lines harboring validated expression vectors that allow the
production of the antibodies from their coding genes. The various steps needed to go
from the initial antibody to the adequate cell line producing it in vitro in high amounts
are all cumbersome and costly.
304 Chatenoud
Once these steps are taken, the purification procedures are simple per se, though
they are submitted to extensive quality-control tests at each of their steps, evaluating
not only the physicochemical characteristics of the antibody but also its safety, nota-
bly to exclude the presence of any viral contamination of the final product.
2.4. The Problem of Sensitization

Monoclonal antibodies that were initially used therapeutically were of xenogeneic
origin, mainly mouse or rat anti-human antibodies. Even in rodents, antibodies were
also most often xenogeneic hamster and rat anti-mouse antibodies. Administration
of such xenogeneic antibodies regularly led to the appearance of an antimonoclonal
humoral response, often endowed with a potent neutralizing activity. Studies initially
performed with human antibodies and subsequently with rodent ones have allowed a
better definition of the pattern of the antiglobulin response. Various methods have
been proposed to circumvent the problem.
Antimonoclonal globulins express two major specificities that are antiisotypic and
antiidiotypic (42,66,67). Antiidiotypic antibodies represent the majority if not the total-
ity of neutralizing antibodies. Their detection is based on ELISAs and immunofluores-
cence assays testing for their capacity to inhibit the binding of the monoclonal antibody
to the target cell (66). Studies performed on immunized patients and monkey affinity
purified antiidiotypic antibodies showed that the response is oligoclonal (68). Anti-
isotypic antibodies are essentially nonneutralizing (69). Thus, the neutralizing potential
of the humoral response to monoclonal antibodies relies more on its specificity than
on the overall amount of antibodies produced, since only a few specific clones were
recruited. Serum sickness was not observed because the amounts of immune complexes
formed were insufficient to elicit a generalized reaction. The antimonoclonal humoral
response may also include IgE antibodies (associated with the potential risk of anaphy-
laxis), but in practice this has appeared to be an extremely rare observation (70,71).
Thus, from the clinical point of view, sensitization essentially posed the problem of the
abrogation of the therapeutic efficacy of the monoclonal antibody.
In practice, the adjunction of adequate dosages of chemical immunosuppressants to
the monoclonal antibody significantly decreased the frequency and intensity of sensi-
tization (72).
Although this approach has been quite useful, in order to get more complete protec-
tion from sensitization, a common trend has been to generate humanized antibodies, or
more recently human monoclonal antibodies.
2.4.1. Humanized Antibodies

Humanized chimeric or CDR-grafted monoclonal antibodies were produced in the
attempt to reduce the amount of xenogeneic determinants (73). Chimeric antibodies
present intact variable regions (Fab) from the parental rodent antibody, coupled to
human immunoglobulin constant regions. These chimeric immunoglobulins were ini-
tially derived from chemical procedures and subsequently by molecular engineering. In
the case of recombinant CDR-grafted antibodies, the rodent CDRs (i.e., the hypervari-
able regions specifically interacting with the antigen) are inserted into human heavy-
and light-chain immunoglobulin frameworks. These two approaches allow the selec-
tion of the desired human Fc fragment that will impact the final antibody effector
capacities (complement fixation, opsonization, and ADCC) and, in the case of CD3
antibodies (46,74,75), mitogenic capacity (discussed later in more detail).
The risk of a deleterious antiidiotypic response still occurring with these human-
ized antibodies is a valid concern. In fact the clinical data available so far show that
chimeric and reshaped humanized antibodies are immunogenic in some patients, but
only when the monoclonal is administered alone (in the absence of associated immu-
nosuppressants) and after more than two to three repeated antibody courses (76).
However, it is the general experience that the association to conventional immuno-
suppressants, even in low doses, significantly decreases the incidence of sensitiza-
tion (7,77,78). From a fundamental point of view, these data are the first indication of
the crucial role of the monoclonal antibody Fc fragments in the induction of the
antiidiotypic response.
In the particular case of CD3 antibodies, as previously discussed, humanization has
offered the double advantage of circumventing the problems linked to sensitization
and the mitogenicity that is responsible for the cytokine-related flu-like syndrome,
which is a major side effect (46). Thus in the first few hours after anti-CD3 administra-
tion, a massive but transient systemic release of several cytokines is observed (TNF,
IFN, IL-2, IL-3, IL-4, IL-6, IL-10, granulocyte/macrophage colony-stimulating factor
[GM-CSF]), which induces high fever, chills, and headache together with repeated
episodes of vomiting and diarrhea; patients become prostrated through massive fluid
and electrolyte loss (46,79,80). A minor proportion of patients may develop more
severe and potentially life-threatening conditions such as severe respiratory distress
(related to pulmonary edema in patients with significant fluid overload), neurotoxicity,
and hypotension, but only after the first injection (46,79,80). As we already mentioned,
this mitogenic capacity is due to the ability of the Fc portion of CD3 antibodies to
interact with monocyte Fc receptors (46). Thus, by mutating or aglycosylating
(46,74,75) the Fc portion of engineered humanized nonmitogenic CD3, antibodies have
been obtained. Results from pilot trials are very encouraging, since they confirmed that
they were well tolerated as well as having the capacity to reverse ongoing kidney allo-
graft rejection.
2.4.2. Human Antibodies

Three approaches have recently been made available to produce human therapeutic
monoclonal antibodies.
The first one consists of the use of mice whose endogenous immunoglobulin genes
have been knocked out and that have been made transgenic for human constant and
variable immunoglobulin encoding genes. Almost all the B-cells in these animals
express human immunoglobulin chains and produce human antibodies. The number of
human immunoglobulin encoding sequences introduced proved sufficient to allow the
generation of a significant diversity. This method has the advantage of simplicity (once
the transgenic mice are available) and of the production of high-affinity antibodies
using an in vivo antigen-driven selection upon immunization.
306 Chatenoud
An alternative approach is phage display using cDNA libraries expressed on fila-

mentous phages, allowing a rapid, fully in vitro selection of antibodies with high speci-
ficity (81).
The third approach consists in establishing human-mouse chimeras. Irradiated mice
are reconstituted with bone-marrow cells from SCID mice. Human lymphocytes from
presensitized donors are then injected into the reconstituted mice, where they are
boosted with defined antigens and used for fusion with myeloma cells (82).
3. A Wide Array of Specificities Already Available

Among the vast panel of monoclonal antibodies interfering with immune cells
cooperation pathways, the ones of particular interest for therapeutic purposes are listed
in Table 2. These mainly include: monoclonal antibodies directed at receptors and
coreceptors involved in antigen recognition, such as CD3, CD4, and MHC molecules;
costimulatory molecules; receptors exclusively expressed on activated T-cells (such
as CD25); cell adhesion receptors; and cytokines.
The case of antibodies to CD52 (a small GPI-anchored protein expressed at the
surface of human B cells and T-cells as well as monocyte/macrophages whose func-
tion is still ill-defined), is also interesting, since they have been extremely potent at
promoting long-term acceptance of organ allografts as well as prolonged remission of
established, and otherwise intractable, autoimmune diseases such as multiple sclero-
sis, vasculitis, and rheumatoid arthritis (8386).
4. Special Ability to Induce Immune Tolerance

The experiments of R. Billingham, L. Brent, and P. Medawar were the first to
establish that immunological tolerance was an acquired state (87). The authors
showed that mouse neonates injected with allogeneic cells accepted skin grafts shar-
ing the same haplotype as the cells injected at birth indefinitely, as adults. Impor-
tantly, this tolerant state was specific, since the animals could reject third-party
allogeneic skin (87). At variance with what was initially thought, central deletion of
alloreactive T-cells is not the only mechanism explaining neonatal tolerance. Periph-
eral tolerance mechanisms also operate to sustain the unresponsiveness. Thus, lym-
phoid cells from treated animals can transfer tolerance to adult syngeneic recipients
(88). A Th2-type immune deviation has also been evidenced in tolerant animals,
since upon specific activation, alloreactive T-cells produce low IFN- but high IL-4
levels (89,90). Moreover, neonatal tolerance is abrogated when IL-4-neutralizing
antibodies or IL-12 are administered in combination with the allogeneic cells at birth
(89,90).
Among the strategies attempted to recreate in adult hosts an immune environment
permissive to tolerance induction, those using polyclonal or monoclonal antibodies
seem to hold great promise for foreseeable future clinical applications. The
groundbreaking experiments performed by Monaco and Woods in the 1960s showed
that a combination of anti-lymphocyte (ALS) treatment and posttransplantation donor
bone-marrow infusion could induce specific unresponsiveness to skin allografts
(91,92).
Table 2
Targets of Therapeutic Monoclonal Antibodies and Fusion Proteins a
Efficacy in Efficacy in
experimental Use in clinical experimental Use in clinical
Target transplantation transplantation autoimmunity autoimmunity
CD2 +/ Organ Tx ND ND
CD3 +++ Organ Tx +++ IDD
CD4 +++ Organ Tx +++ RA, PS, MS, SLE,
Vasculitis
CD4+CD8 +++ ND +++ ND
CD20 ND PTLD ND RA, MS, SLE
CD25 ++ Organ Tx ++ Uveitis
B7.1, B7.2 ++ ND ++ RA, PS, SLE
(use of CTLA4-Ig)
CD40L ++ Organ Tx ++ SLE
CD52 ND Bone marrow, ND RA, MS
Organ Tx
LFA-1 ++ Organ Tx +++ ND
ICAM-1
VLA-4 ND ND ++? MS
MHC Class I ND ND + ND
MHC Class II + ND + ND
TNF +++ experimental ND +++ RA, IBD, PS
GVHD
IFN- ND ND +++ IBD ongoing
IL-1 ND ND ++ RA
a Experimental and clinical applications.
ND, not done; RA, rheumatoid arthritis; MS, multiple sclerosis; IBD, inflammatory bowel disease;
PS, psoriasis; Tx, transplantation; IDD, insulin dependent diabetes; SLE, systemic lupus erythemato-
sus; GVHD, graft-vs-host disease; PTLD, posttransplant lymphoproliferative disease.
Since then, various groups have reproduced and extended these findings, showing
that:
1. Tolerance to vascularized and nonvascularized allografts was not only induced by ALS
but also by different monoclonal antibodies or fusion proteins interfering with T-cell
surface receptors or their ligandsi.e., CD3, CD4, and CD8, costimulation receptors
such as CD28, CTLA-4, and CD40 ligand (29,44,4656).
2. Several of these biological immunosuppressants (i.e., ALS, CD3 immunotoxin, CD4
antibodies, CD40 ligand antibodies alone or associated to CTLA-4 Ig) reproduced the
same result in larger animals such as monkeys, which are valuable preclinical models
(50,51,54,9396).
3. Donor bone marrow was not indispensable in both the rodent and monkey models
(50,51,96,97).
4. Some anti-T-cell monoclonal antibodies also represent invaluable tools to restore self-
tolerance in established autoimmunity (46).
308 Chatenoud
At present, the precise cellular and molecular mechanisms underlying this immune
tolerance are not fully elucidated. What is known, however, is that in nearly all the
models we mentioned, tolerance is not deletionali.e., not based on the physical elimi-
nation of the alloreactive or autoreactive effector T-cells. Instead, the tolerant hosts
frequently show potent T-cell-mediated regulatory mechanisms that effectively con-
trol, downmodulate, or suppress the pathogenic activity of alloreactive or autoreactive
effectors. Tolerance, once induced, is frequently dominant or infectious (the term
initially coined by Gershon [98]) and can be acquired by nontolerant T-cells in the
absence of further immunosuppression (26,99). In some transplantation models, it was
also shown that the tolerance can also spread to additional antigenic determinants,
provided that all determinants are presented together on the same antigen-presenting
cell, a phenomenon termed linked suppression (100). The underlying principle is that
tolerance to one antigen in one tissue allograft will facilitate tolerance to other alloan-
tigens on other transplants, provided that the determinants are presented by the same
antigen-presenting cell; in other words, that there is, at least partly, alloantigen sharing
between the different grafts (100).
Despite the fact that most studies concerning tolerance induction to organ allografts
have focused on antibodies to CD4, other specificities also deserve interestfor
instance, antibodies to CD3, the combination of antibodies to LFA-1 and ICAM-1,
and antibodies to CD25 (especially in rat transplant models). Combined administra-
tion of LFA-1 and ICAM-1 to mice resulted in permanent engraftment of fully mis-
matched vascularized heart allografts, but also specific tolerance to donor-type skin
allografts (101). In a mouse heart and skin allograft model, the simultaneous but not
independent blockade of the CD28 and CD40 costimulatory pathways promoted long-
term, although not indefinite, allograft survival and, importantly, also inhibited the
development of chronic vascular rejection in the grafted hearts (102). Non-human pri-
mates treated with antibodies to CD40 ligand also significantly prolonged survival of
fully mismatched renal allografts (54). Unfortunately, attempts to transfer this strat-
egy to the clinic were unsuccessful because of severe thromboembolic complications
due to the expression of CD40L on platelets.
Based on the Th1/Th2 paradigm (103), several authors attempted to explain most
of these phenomena by an exclusive immune deviation from a Th1 allo or auto-
destructive response to a Th2 nondestructive or even protective response. The data
from some models of CD4 antibody-induced tolerance to cardiac and renal allografts
were in support of this assumption, but this was far from a general conclusion (99,104).
As a whole, results obtained using different transplantation and autoimmunity models
argue for a significant heterogeneity of immunoregulatory T-cells operational in sus-
taining the long-term effects. As discussed above, their precise phenotype and their
cytokine dependency also varies depending on the model (i.e., CD4 + CD25 + ,
CD4+CD25 T-cells (29,55,5764).
5. Efficacy in Advanced Disease

Unlike transplantation, in which tolerance-inducing drugs may be given at the time
of the organ implantaion, any such treatment for clinical autoimmune disease can hap-
pen only at the time of diagnosis, when the immune system is already primed and once
the target tissue has already undergone significant damage. Even at such a late stage,
some monoclonal antibodies (in particular antibodies to CD3) have been shown to be
very promising tools to restore self-tolerance in established autoimmunity. Initial data
were obtained in non-obese diabetic (NOD) mice, which are a good model for human
auto-immune insulin-dependent diabetes. In NOD mice, disease spontaneously devel-
ops by 4 mo of age and is preceded, starting at 3 wks of age, by a progressive infiltra-
tion of the islets of Langerhans with mononuclear cells (i.e., insulitis).
Diabetogenic T-cells that transfer acute diabetes into immunoincompetent synge-
neic recipients are present in high frequency in the spleen of diabetic NOD mice, and
include both CD4+ (essentially IFN--producing Th1 cells) and CD8+ lymphocytes
(105110). In parallel to these effector cells, co-transfer experiments have identified
subsets of immunoregulatory T-cells (in the spleen and the thymus of prediabetic ani-
mals) that express CD25 and/or L-selectin (CD62L) mediating active tolerance,
namely, exerting an active control or a down-regulatory effect on diabetogenic lym-
phocytes or their precursors that fully prevents the transfer of disease by diabetogenic
cells (61,62,111113).
A low-dose CD3 treatment (5 to 20 g for 5 consecutive d) administered to overtly
diabetic NOD mice (i.e., presence of glycosuria and glycemia = 4g/L) induced per-
manent disease remission in 6080% of mice (46,47,114). The remission was long last-
ing, and the effect was not related to generalized long-standing immunosuppression,
because 8 to 10 wk after treatment all mice responded normally to exogenous antigens.
The long-term effect was specific to -cell-associated antigens, since mice show-
ing remission after CD3 antibody treatment did not destroy syngeneic islet grafts as
untreated diabetic NOD females normally did, even though they were not immuno-
compromised as shown by their capacity to rapidly reject allogeneic skin grafts (46,47).
Nonmitogenic F[ab']2 fragments of 145 2C11 were as effective as the whole mitogenic
antibody in promoting permanent remission of overt diabetes (46,47,114). Interestingly,
in contrast with the major therapeutic effect observed in diabetic mice, at an advanced
stage in the disease, CD3 antibody treatment was totally ineffective for disease preven-
tion when applied to young prediabetic NOD mice (46,114). Concerning the mecha-
nisms mediating the therapeutic effect, one may distinguish two distinct consecutive
phases. The first is temporary, elapses during the time of antibody treatment, and is
associated with complete clearance of the insulitis. The second is long-lasting, correlat-
ing with the presence of a non-destructive pancreatic infiltrate (i.e., peripheral insulitis)
and of CD4+CD25+/CD62L+ immunoregulatory T-cells, whose numbers and functional
capacity are increased as compared to those observed in diabetic untreated NOD mice
(46,47,61,62,64,114). More recent data have shown the central role of transforming
growth factor (TGF-) in this state of CD3 antibody-induced T-cell-mediated active
tolerance. In particular, neutralizing antibodies to TGF- fully prevent disease remis-
sion when administered concomitantly with CD3 antibodies (46,64).
These experimental data paved the way for clinical trials that are presently in
progress using non-mitogenic engineered antibodies to human CD3 (74,115) to pre-
serve a functional insulin-secreting cell mass in patients presenting recent-onset
310 Chatenoud
insulin-dependent diabetes. The recent data from the first Phase I pilot clinical trial are
very promising, and show that a significant preservation of cell functional capacity
can be observed at a 1-yr follow-up after treatment of only a few days applied within
6 wk of the first diagnosis (116). A Phase II randomized placebo-controlled trial is
presently in progress in Europe that includes a total of 80 patients. The results of this
study will be available before the end of this 2004.
6. Monoclonal Antibodies and Fusion Proteins in Phase II/III Clinical Trials

6.1. CD3 Antibodies
The clinical use of CD3 antibodies essentially developed in solid-organ transplan-
tation. Between 1981 and 1985, a first pilot trial including a few patients, immedi-
ately followed by a large, randomized, multicenter trial, clearly established the
efficacy of OKT3 in reversing early acute renal allograft rejection episodes (117
119), a use for which the antibody was rapidly licensed both in the United States and
Europe. The dose regimen was based on the data from the pilot trial, in which escalat-
ing doses were applied (117119). Thus, 5 mg/d were administered for 10 to 15 con-
secutive d. The general practice was also to decrease the dosages of other concurrent
medications during OKT3 treatment.
The use of OKT3 rapidly expanded from the treatment of early acute rejection to
that of rejection episodes unresponsive to conventional high-dose corticosteroids and
polyclonal anti-lymphocyte globulins, also termed rescue treatment. Various studies
with large numbers of patients have demonstrated the efficacy of the antibody in this
setting as well (120122). In addition, the interesting results obtained in renal trans-
plantation also encouraged the use of the antibody for liver and heart transplant rejec-
tion. In both situations, OKT3 was essentially developed as a rescue treatment
(120,123125).
OKT3 has also been used in several centers as an induction therapy, namely, as a
first line treatment to prevent or delay the advent of acute rejection episodes and to
improve long-term allograft survival. Moreover, particularly in renal transplantation,
the use of OKT3 in the early post-transplant period was of interest to avoid complica-
tions linked to cyclosporin nephrotoxicity. Various controlled, single-center studies
demonstrated the efficacy of OKT3 in this setting when compared to control regi-
mens, some of which included polyclonal anti-T-cell antibodies (126128). In this
context, OKT3 was administered for 1014 consecutive days (starting on the day of
transplant) in association with corticosteroids, azathioprine, and cyclosporin. In most
cases, cyclosporin was applied starting a few days before the end of the antibody
therapy. This is a clinically relevant issue, since the data from the European registry
suggest that long-term allograft survival is significantly reduced when cyclosporin
treatment is introduced concomitantly and not consecutively to OKT3 (129).
OKT3 induction therapy was also applied with success in liver transplantation
(130133).
Despite these results, OKT3 is not adopted any more as a routine prophylactic treat-
ment, because of the risk of sensitization, which could preclude the use of the antibody
for a steroid-resistant rejection, and secondly, because of the occurrence of the flu-like
syndrome that, as compared to ATG or conventional triple therapy with cyclosporin,

could complicate the management of the patients (46,79,80,132,134,135).
The future will be the use of humanized nonmitogenic CD3 antibodies that are
already available. They have been used in transplantation in pilot clinical trials for the
treatment of established renal allograft rejection episodes (136,137). Results confirmed
that these antibodies were well tolerated and that, despite the low numbers of patients
included, the therapeutic effectiveness (reversal of rejection) was comparable to that
observed with OKT3 (136,137). These antibodies are also presently used in the autoim-
munity setting, to treat patients with recent-onset insulin-dependent diabetes (46,116).
6.2. CD25 Antibodies
As compared to OKT3, clinical experience with other murine monoclonal antibod-
ies was limited and never reached the point of large-scale distribution, though encour-
aging results were reported. This was the case for antibodies to CD25 or LFA-1
(138142).
Here again, the panorama rapidly evolved with the advent of humanized antibod-
ies. Some years ago, two humanized antibodies to CD25 (one chimeric [basiliximab/
Simulect] and one CDR-grafted [daclizumab/ Zenapax]), which are very well toler-
ated and are effective as part of induction regimens to prevent organ allograft rejec-
tion, were approved for use as induction therapies in organ transplantation
(77,78,143,144). Treatment with CD25 antibodies is started at the time of transplant
for a limited number of injections (i.e., covering a 2- to 3-wk therapeutic window) on
a background double therapy (cyclosporin and azathioprine) or triple therapy
(cyclosporin, azathioprine, and corticosteroids). With the two antibodies, a significant
reduction was observed in the incidence of biopsy-proven acute rejection as compared
to placebo-treated patients. Results from Phase III trials showed that graft survival at
3 yr was not different in antibody- vs placebo-treated patients; pooled patient survival
was >9193% in both groups, and the incidence of posttransplant lymphoproliferative
disease (PTLD) was comparable. These conclusions were recently confirmed in a
meta-analysis of eight randomized controlled trials (total of 1858 patients) (145). Stud-
ies with follow-ups longer than 3 yr may be needed to see statistical differences in
graft and patient survival.
Recent data also indicated that humanized CD25 antibodies could also be benefi-
cial for the treatment of established rejection (145). These data are of particular inter-
est since they argue for a greater efficacy of humanized CD25 antibodies as compared
to the parent murine molecules. Thus, mouse CD25 antibodies were without effect for
treatment of established rejection. This may also explain that humanized CD25 anti-
bodies appear potentially interesting for the treatment of some severe autoimmune
diseases, such as uveitis (146,147).
6.3. CD4 Antibodies

As already discussed, a significant amount of data have accumulated showing the
remarkable capacity of monoclonal antibodies to CD4 to suppress immune responses
to soluble and tissular antigens, i.e., alloantigens and autoantigens (2630,148,149).
312 Chatenoud
In humans, CD4 is expressed not only at the surface of a subset of lymphocytes but
also on monocytes, macrophages, Langerhans cells, eosinophils, endothelial cells of
hepatic sinusoids, sperm, and brain cells. The first pilot trials used mouse antibodies to
human CD4 applied to patients presenting long-standing rheumatoid arthritis, psoria-
sis, inflammatory bowel disease, and uveitis (150157). The treatment was well toler-
ated. Only rare cases of mild side effects after the first injection (linked to minor
cytokine release) have been reported (150157). The antibodies essentially induced
partial and transient disappearance of circulating CD4+ cells. Coating of CD4+ cells
was also observed, along with dose-dependent saturation of CD4 binding sites (150
157). When present, antigenic modulation affected only a minor proportion of CD4
receptors. The results were encouraging in terms of therapeutic effectiveness, but in
most cases the effects were brief (150156).
The drawback in all these trials was the sensitizing effect of the murine
monoclonals, which justified trials of humanized CD4 antibodies. Several of the pub-
lished studies focused on the use of the chimeric cM-T412 antibody (human IgG1). In
rheumatoid arthritis, results from open studies seemed encouraging, especially when
cumulative doses ranging from 350 to 700 mg were used (158161). There were no
major signs of acute toxicity reported; some authors describe self-limited adverse
events such as fever associated with myalgia, malaise, and asymptomatic hypoten-
sion, which appear to correlate with transient elevations of serum IL-6 (158,160,161).
Mild sensitization was reported in a majority of the patients (158,160,161). However,
the major problem was that the therapeutic effect could not be confirmed in the con-
text of a large randomized, double-blind, placebo-controlled study including patients
with early rheumatoid arthritis (159). Also of concern was the significant and long-
lasting CD4+ cell depletion (up to 60% from baseline for over 1830 mo) (25) noted
with cM-T412, contrary to what had initially been reported with the parental mouse
M-T151 anti-CD4 antibody (154). It has been proposed that apoptosis could mediate,
at least in part, this massive depletion (162).
The cM-T412 antibody has also been used in patients with multiple sclerosis, but
here again without any evident clinical effect (163).
The non-depleting humanized antibody to CD4 OKT4cdr4A (the CDR-grafted ver-
sion of OKT4A) was used in patients presenting with severe psoriasis; one open pilot
study and a placebo-controlled one were conducted, both of which provided promis-
ing data (157,164).
Results from open pilot studies suggested that CD4 antibody therapy in combina-
tion (i.e., with antibodies to CD52) could be effective to treat severe forms of vasculi-
tis (84,85). One may hope that controlled studies will be actively promoted in cases
with such indications to definitely prove the validity of the approach.
6.4. Antibodies to TNF
The introduction into clinical practice of humanized antibodies to TNF was a major
step in the development of novel therapeutic strategies for treatment of rheumatoid
arthritis. This was the consequence of the pioneer experimental and clinical work con-
ducted by the groups of M. Feldmann and T. Maini. The first crucial in vitro finding
was that neutralizing antibodies to TNF significantly decreased the production of most
of the pro-inflammatory cytokines (i.e., IL-1, IL-6, IL-8, GM-CSF) normally found in
in vitro cultures of cells infiltrating synovial membranes from rheumatoid arthritis
patients (7,165). The in vivo relevance of this hypothesis was validated in two experi-
mental settings. Mice expressing a human TNF transgene develop chronic arthritis that
may be fully prevented by treatment with antibodies to TNF (7,165). In addition, in a
model of collagen type II-induced arthritis, in which the animals develop an erosive
form of chronic arthritis, CD4 antibodies can prevent disease but are ineffective once
the inflammatory and destructive process is established (7,165). At variance with this,
neutralizing antibodies to murine TNF applied once the disease has started are able to
decrease the severity of objective and histopathological symptoms (e.g., swollen joints,
bone erosions) (7,165). Another unexpected but potentially relevant observation was
that in established arthritis, combining a sub-optimal dose of antibody to TNF, which
had no significant effect per se, with CD4 antibody therapy greatly improved limb
involvement and helped heal paw-swelling and joint erosion. Furthermore, optimal anti-
TNF combined with the CD4 antibody significantly improved the results obtained with
anti-TNF alone (166). Thus, in this model, anti-TNF and anti-CD4 are synergistic in
ameliorating established disease. One interpretation of these results is that in order to
sensitize the system to the effect of T-cell-directed immunointervention, it is manda-
tory to neutralize the inflammation that is the hallmark of self-perpetuating ongoing
autoimmunity. In terms of strategy, this could be relevant for most autoimmune dis-
eases, not only for established arthritis.
The results of the first randomized placebo-controlled double-blind study showing
the effectiveness of the chimeric neutralizing antibody to TNF cA2 (human IgG1), now
termed infliximab (and subsequently marketed as Remicade), for the treatment of long-
standing rheumatoid arthritis were reported in 1994 (167). They fully confirmed the data
from the initial phase I/II open trial showing no acute toxicity and significant therapeutic
benefits, as assessed by the clinical and laboratory parameters, lasting for several weeks
after the end of treatment. Some of the patients having disease relapse underwent two or
more cycles of retreatment; disease flare-ups were sensitive to retreatment, but the mean
duration of the induced remissions progressively diminished, probably due to the
appearance of antiallotypic sensitization (76). Presently trials are being conducted that
associate these molecules to methotrexate in an attempt to obtain longer-lasting clinical
remissions (7,168,169). Part of the beneficial effect of combining with methotrexate is
probably due to a reduction of the immunogenicity of the anti-TNF antibody, as con-
firmed by the reduced response to the chimeric antibody in patients receiving the com-
bined therapy (7,168,169). These results were then confirmed in a subsequent Phase III
study. Importantly, data from this trial showed that anti-TNF therapy was effective in
protecting against joint destruction; arrest of damage was observed in more than 50% of
patients by 6 mo after beginning treatment and lasted for the 2 yr of the study. There was
also evidence of repair in a significant proportion of patients (170,171). Based on these
results, infliximab was approved for use, in combination with methotrexate, both in the
United States and Europe. In the attempt to obtain a better clinical benefit, especially in
terms of the duration of the remissions, trials have been conducted enrolling patients
with recent onset disease (at about 6 mo from first symptoms and 2 mo from formal
diagnosis). Recently reported results appear very promising.
314 Chatenoud
With the disclosure of the very promising results using the cA2/ infliximab anti-
body, other biological agents targeting TNF were developed. Another chimeric anti-
TNF monoclonal antibody named CDP571 was also shown to be clinically effective
(172). Two fusion proteins linking the TNF receptor molecules p55 or p75 to a human
IgG constant frame (lenercept or etanercept/Enbrel) also showed effectiveness in
clinical rheumatoid arthritis (173175). However, only etanercept had an active clini-
cal development and has been approved both in the United States and Europe. Other
interesting candidates are on their way, among which the fully human D2E7 antibody
(adalimumab) showed good efficacy in both Phase II and III trials (176,177).
Antibodies to TNF were also used with success for treatment of severe Crohns
disease and are presently also approved for this indication (178,179). Although the
pathophysiology of this disease remains unclear, there is evidence for the significant
role played by inflammatory cytokines (180). The therapeutic benefit obtained upon
TNF antibody treatment correlated with a decrease in IFN- production by mono-
nuclear cells infiltrating the lamina propria (181).
More recently, there were also interesting data reported from trials using TNF
blocking agents in juvenile rheumatoid arthritis (182), ankylosing spondylitis (183),
psoriatic arthritis (184), and psoriasis (185).
One side effect reported, especially in patients undergoing repeated treatments with
infliximab, was the increased incidence of tuberculosis (186). This is one reason why
combination therapy with drugs aimed at neutralizing TNF and IL-1 has recently been
disapproved by the Food and Drug Administration.
6.5. Antibodies to CD52

CD52 is a low molecular weight (12 amino acids) GPI-anchored protein expressed
at the surface of human B-cells, T-cells, as well as monocytes/macrophages. The first
rat antibody to CD52, Campath-1H, was characterized in 1983 and was initially used
to deplete mature T-cells from bone-marrow transplants. A fully reshaped humanized
version, Campath-1H (human IgG1), was derived by genetic engineering (73) and is
available for clinical use as alemtuzumab. It has been extensively used in vivo to treat
CD52+ hematologic malignancies and in vitro to purge bone-marrow transplants to
prevent graft-vs-host disease. Campath-1H is highly depleting in vivo; CD4+ cells do
not return to normal for several consecutive months. Upon the first injection, Campath-
1H triggers an acute self-limited cytokine release that is the cause of a transient flu-
like syndrome.
Campath-1H has been applied in patients with rheumatoid arthritis. At 3 and 6 mo
of treatment, an improvement in the Paulus score was observed in about half of the
patients (187). Another indication in which Campath-1H seems to be very effective is
severe systemic vasculitis (8486). The antibody was initially applied to rare forms of
vasculitis, the pathogenesis of which is thought to involve mainly T-cell-mediated
mechanisms. Subsequently, patients with Wegeners granulomatosis have also been
treated with success. Particularly impressive in this clinical context were the long-
term remissions that were obtained when combining antibodies to CD52 and to CD4
(8486).
Very promising results have been obtained in multiple sclerosis (188190). The
initial trials included patients with relapsing/remitting disease evolving over many
years and unresponsive to conventional treatments. Data from 8-yr follow-up are avail-
able. In a vast majority of the patients, a stabilization of clinical symptoms has been
observed over 612 mo. Over the long-term follow-up, a marked decrease in the
appearance of new lesions in the central nervous system as assessed by nuclear mag-
netic resonance (NMR) scanning was evidenced that seemed to correlate with the
observed long-lasting and significant peripheral depletion of CD4+ T-lymphocytes
(188190). An unexpected side effect of the treatment was the appearance of Graves
disease in a significant proportion of the patients successfully treated for their mul-
tiple sclerosis (190). Multicenter trials are presently ongoing, including patients with
more recent onset disease.
As a whole, Campath-1H seems quite unique in its capacity to promote long-lasting
remission of life-threatening autoimmune diseases that are unresponsive to conven-
tional treatments. Results from ongoing trials will help to draw more definitive con-
clusions.
Of particular interest are also the data reported by the group of R. Calne with the
CD52 antibody Campath-1H in transplantation (83). Campath 1H was administered to
31 renal allograft recipients at a dose of 20 mg (i.v.) on d 0 and 1 after transplant. The
patients were maintained on low-dose monotherapy with cyclosporin started 72 h after
transplant. At 1528 mo of follow-up, 29 patients presented with functioning grafts.
Six rejection episodes were scored that were responsive to conventional steroid
therapy. More recently, two pilot studies attempted a similar strategy using Campath-
1H and rapamycin monotherapy (191,192). The data are interesting, but larger studies
are needed to draw definitive conclusions.
6.6. Antibodies to VLA-4

Integrins are adhesion molecules of fundamental importance to the recruitment of
leucocytes in inflammation. The 41 integrin VLA-4 is a leucocyte ligand for endo-
thelial vascular cell adhesion molecule-1 (VCAM-1), fibronectin, and osteopontin.
The interaction between VLA-4, expressed at the surface of activated lymphocytes
and monocytes, with its ligand, VCAM-1, is essential for cell migration into inflamed
parenchyma. Based on promising data obtained in experimental models of VLA-4
blockade, a specific humanized monoclonal antibody, natalizumab, was tested in ran-
domized placebo-controlled trials. A first indication was multiple sclerosis (193). A
total of 213 patients presenting relapsingremitting or relapsing secondary progres-
sive multiple sclerosis received natalizumab or placebo treatment every 28 d for 6 mo.
Results showed a marked reduction in the number of new brain lesions, as assessed by
gadolinium-enhanced magnetic resonance imaging, in natalizumab- vs placebo-treated
patients over the 6-mo observation period (193). The same antibody was applied in
Crohns disease (194). Here again, a placebo-controlled randomized trial was con-
ducted including a total of 248 patients presenting with moderate to severe disease.
Two infusions of the antibody were administered 4 wk apart. Patients who received
natalizumab had higher remission rates than the placebo group, and a significant
316 Chatenoud
improvement in Crohns Disease Activity Index. The quality of life improved in

natalizumab-treated patients, and no major adverse events were reported (194).
6.7. Biological Ligand of CD2: the LFA-3Ig Fusion Protein
The human LFA-3/human IgG1 fusion protein (alefacept) binds to CD2 at the sur-
face of T-lymphocytes. Controlled multicenter studies were conducted using this agent
in patients with chronic plaque psoriasis. Depending on the trial, one or two 12-wk
courses of alefacept were administered once a wk (7.5 to 15 mg/injection). In all cases,
the drug was well tolerated and at least a 75% reduction of the disease score (PASI
index) was observed at 2 wk from the last dose that in a significant proportion of the
patients lasted for 7 mo (194,195). Based on these positive results, alefacept has been
approved by regulatory authorities for this indication. The immunological monitoring
of the patients evidenced a selective reduction of memory-effector CD45R0+ cells in
the circulation, while lymphocytes expressing a nave phenotype were spared (195).
These reductions in memory T-cell counts correlated with measurements of disease
activity, which fits the assumption that pathogenic T-cells are included among memory
T-lymphocytes (194).
6.8. Antibodies to CD20
Rituximab (Rituxan) is a human-mouse chimeric monoclonal antibody specific for
CD20, a B-cell CD20 antigen. It causes rapid and specific B-cell depletion. Rituximab
was approved in the United States in 1997 and in Europe in 1998 (MabThera) to treat
severe refractory CD20-positive B-cell non-Hodgkins lymphoma. The use of this anti-
body has now been extended as first-line therapy and maintenance therapy to lymphoma
and stem-cell transplantation procedures.
More recent studies are also evaluating the use of rituximab in autoimmune disor-
ders, such as rheumatoid arthritis, immune thrombocytopenic purpura, autoimmune
hemolytic anemia, systemic lupus erythematosus, and multiple sclerosis (196200).
In rheumatoid arthritis, the treatment appears safe and associated with major improve-
ment. Based on these results, a phase II controlled trial is presently in progress.
From a more fundamental point of view, these results highlight the pathogenic role
of B-cells in rheumatoid arthritis.
7. Future Perspectives
The progress achieved on therapeutic monoclonal antibodies over the last decade is
remarkable. They are already well established in several clinical settings, fulfilling
needs that had not been met with conventional chemical immunosuppressants. One
may assume in view of the numerous antibodies and fusion proteins presently under
development, especially humanized and fully human molecules, that their importance
will rapidly expand both in transplantation and in autoimmunity.
Today, monoclonal antibodies represent the most promising approach for tolerance
induction in animal models (both induction of tolerance to alloantigens in transplanta-
tion and restoration of self-tolerance in autoimmune diseases). One may reasonably
hope that these spectacular results will be rapidly transferred to the clinic. Moreover,
understanding the molecular basis of these unique tolerogenic properties would clear
the way for the development of chemicals or recombinant receptor agonists and/or
antagonists that would mimic the desired therapeutic effect.
References
1.
1 Fanslow, W. C., Sims, J. E., Sassenfeld, H., et al. (1990) Regulation of alloreactivity in
vivo by a soluble form of the interleukin-1 receptor. Science 248, 739742.
2.
2 Jacobs, C. A., Lynch, D. H., Roux, E. R., et al. (1991) Characterization and pharmacoki-
netic parameters of recombinant soluble interleukin-4 receptor. Blood 77, 23962403.
3.
3 Ozmen, L., Gribaudo, G., Fountoulakis, M., Gentz, R., Landolfo, S., and Garotta, G. (1993)
Mouse soluble IFN gamma receptor as IFN gamma inhibitor. Distribution, antigenicity,
and activity after injection in mice. J. Immunol. 150, 26982705.
4.
4 Dinarello, C. A. (1991) Interleukin-1 and interleukin-1 antagonism. Blood 77, 1627
1652.
5.
5 Peppel, K., Crawford, D., and Beutler, B. (1991) A tumor necrosis factor (TNF) receptor-
IgG heavy chain chimeric protein as a bivalent antagonist of TNF activity. J. Exp. Med.
174, 14831489.
6.
6 Lesslauer, W., Tabuchi, H., Gentz, R., et al. (1991) Recombinant soluble tumor necrosis
factor receptor proteins protect mice from lipopolysaccharide-induced lethality. Eur. J.
Immunol. 21, 28832886.
7 Feldmann, M. (2002) Development of anti-TNF therapy for rheumatoid arthritis. Nat. Rev.
7.
Immunol. 2, 364371.
8.
8 Skoglund, C., Scheynius, A., Holmdahl, R., and Van Der Meide, P. H. (1988) Enhance-
ment of DTH reaction and inhibition of the expression of class II transplantation antigens
by in vivo treatment with antibodies against gamma-interferon. Clin. Exp. Immunol. 71,
428432.
9.
9 Nicoletti, F., Meroni, P. L., Landolfo, S., et al. (1990) Prevention of diabetes in BB/Wor
rats treated with monoclonal antibodies to interferon-gamma. Lancet 336, 319.
10.
10 Bach, J. F. (1993) Strategies in immunotherapy of insulin-dependent diabetes mellitus.
Ann. N. Y. Acad. Sci. 696, 364376.
11.
11 Pujol-borrell, R., Todd, I., Doshi, M., et al. (1987) HLA class II induction in human
islet cells by interferon-gamma plus tumour necrosis factor or lymphotoxin. Nature
326, 304306.
12.
12 Sarvetnick, N., Liggitt, D., Pitts, S. L., Hansen, S. E., and Stewart, T. A. (1988) Insulin-
dependent diabetes mellitus induced in transgenic mice by ectopic expression of class II
MHC and interferon-gamma. Cell 52, 773782.
13.
13 Hall, B. M., Bishop, G. A., Duggin, G. G., Horvath, J. S., Philips, J., and Tiller, D. J.
(1984) Increased expression of HLA-DR antigens on renal tubular cells in renal trans-
plants: relevance to the rejection response. Lancet 2, 247251.
14.
14 Fuggle, S. V., Mcwhinnie, D. L., and Morris, P. J. (1987) Precise specificity of induced
tubular HLA-class II antigens in renal allografts. Transplantation 44, 214220.
15 Greaves, M. F., Tursi, A., Playfair, J. H., Torrigiani, G., Zamir, R., and Roitt, I. M. (1969)
15.
Immunosuppressive potency and in vitro activity of antilymphocyte globulin. Lancet 1,
6866.
16.
16 Martin, W. J. and Miller, J. F. (1968) Cell to cell interaction in the immune response. IV.
Site of action of antilymphocyte globulin. J. Exp. Med. 128, 855885.
17.
17 Bindon, C. I., Hale, G., and Waldmann, H. (1988) Importance of antigen specificity for
complement-mediated lysis by monoclonal antibodies. Eur. J. Immunol. 18, 15071514.
318 Chatenoud
18.
18 Isaacs, J. D., Clark, M. R., Greenwood, J., and Waldmann, H. (1992) Therapy with mono-
clonal antibodies. An in vivo model for the assessment of therapeutic potential. J. Immunol.
148, 30623071.
19. Abbs, I. C., Clark, M., Waldmann, H., Chatenoud, L., Koffman, C. G., and Sacks, S. H.
(1994) Sparing of first dose effect of monovalent anti-CD3 antibody used in allograft
rejection is associated with diminished release of pro-inflammatory cytokines. Therapeu-
tic Immunology 1, 325331.
20.
20 Routledge, E. G., Lloyd, I., Gorman, S. D., Clark, M., and Waldmann, H. (1991) A hu-
manized monovalent CD3 antibody which can activate homologous complement. Eur. J.
Immunol. 21, 27172725.
21.
21 Wong, J. T. and Colvin, R. B. (1991) Selective reduction and proliferation of the CD4+
and CD8+ T cell subsets with bispecific monoclonal antibodies: evidence for inter-T cell-
mediated cytolysis. Clin. Immunol. Immunopathol. 58, 236250.
22.
22 Wesselborg, S., Janssen, O., and Kabelitz, D. (1993) Induction of activation-driven
death (apoptosis) in activated but not resting peripheral blood T cells. J. Immunol. 150,
43384345.
23 Kronke, M., Depper, J. M., Leonard, W. J., Vitetta, E. S., Waldmann, T. A., and Greene,
23.
W. C. (1985) Adult T cell leukemia: a potential target for ricin A chain immunotoxins.
Blood 65, 14161421.
24.
24 Walz, G., Zanker, B., Brand, K., et al. (1989) Sequential effects of interleukin 2-diphthe-
ria toxin fusion protein on T-cell activation. Proc. Natl. Acad. Sci. USA 86, 94859488.
25.
25 Moreland, L. W., Pratt, P. W., Bucy, R. P., Jackson, B. S., Feldman, J. W., and Koopman,
W. J. (1994) Treatment of refractory rheumatoid arthritis with a chimeric anti-CD4 mono-
clonal antibody. Long-term followup of CD4+ T cell counts. Arthritis Rheum. 37, 834838.
26.
26 Qin, S., Cobbold, S. P., Pope, H., Elliott, J., Kioussis, D., Davies, J., and Waldmann, H.
(1993) Infectious transplantation tolerance. Science 259, 974977.
27 Cobbold, S. P., Nolan, K. F., Graca, L., et al. (2003) Regulatory T cells and dendritic cells in
27.
transplantation tolerance: molecular markers and mechanisms. Immunol. Rev. 196, 109124.
28.
28 Waldmann, H. and Cobbold, S. (2001) Regulating the immune response to transplants. a
role for CD4+ regulatory cells? Immunity 14, 399406.
29.
29 Bushell, A., Karim, M., Kingsley, C. I., and Wood, K. J. (2003) Pretransplant blood trans-
fusion without additional immunotherapy generates CD25+CD4+ regulatory T cells: a
potential explanation for the blood-transfusion effect. Transplantation 76, 449455.
30.
30 Bushell, A., Morris, P. J., and Wood, K. J. (1995) Transplantation tolerance induced by
antigen pretreatment and depleting anti-CD4 antibody depends on CD4+ T cell regulation
during the induction phase of the response. Eur. J. Immunol. 25, 26432649.
31 Bank, I. and Chess, L. (1985) Perturbation of the T4 molecule transmits a negative signal
31.
to T cells. J. Exp. Med. 162, 12941303.
32 Emmrich, F., Kanz, L., and Eichmann, K. (1987) Cross-linking of the T cell receptor
32.
complex with the subset-specific differentiation antigen stimulates interleukin 2 recep-
tor expression in human CD4 and CD8 T cells. Eur. J. Immunol. 17, 529534.
33 Jabado, N., Le Deist, F., Fisher, A., and Hivroz, C. (1994) Interaction of HIV gp120 and
33.
anti-CD4 antibodies with the CD4 molecule on human CD4+ T cells inhibits the binding
activity of NF-AT, NF-kappa B and AP-1, three nuclear factors regulating interleukin-2
gene enhancer activity. Eur. J. Immunol. 24, 26462652.
34 Jabado, N., Pallier, A., Jauliac, S., Fischer, A., and Hivroz, C. (1997) gp160 of HIV or
34.
anti-CD4 monoclonal antibody ligation of CD4 induces inhibition of JNK and ERK-2
activities in human peripheral CD4+ T lymphocytes. Eur. J. Immunol. 27, 397404.
35.
35 Jabado, N., Pallier, A., Le Deist, F., Bernard, F., Fischer, A., and Hivroz, C. (1997) CD4
ligands inhibit the formation of multifunctional transduction complexes involved in T cell
activation. J. Immunol. 158, 94103.
36.
36 Jauliac, S., Mazerolles, F., Jabado, N., et al. (1998) Ligands of CD4 inhibit the association
of phospholipase Cgamma1 with phosphoinositide 3 kinase in T cells: regulation of this
association by the phosphoinositide 3 kinase activity. Eur. J. Immunol. 28, 31833191.
37 Chatenoud, L. and Bach, J. F. (1984) Antigenic modulation: a major mechanism of anti-
37.
body action. Immunol. Today 5, 2025.
38.
38 Hirsch, R., Eckhaus, M., Auchincloss, J. R., Sachs, D. H., and Bluestone, J. A. (1988)
Effects of in vivo administration of anti-T3 monoclonal antibody on T cell function in
mice. I. Immunosuppression of transplantation responses. J. Immunol. 140, 37663772.
39. Chatenoud, L. and Bach, J. F. (1982) The use of monoclonal antibodies for the follow-up
of transplanted patients. Heart Transplant. 2, 4751.
40.
40 Schwartz, R. H. (1990) A cell culture model for T lymphocyte clonal anergy. Science 248,
13491356.
41 Schwartz, R. H. (1996) Models of T cell anergy: is there a common molecular mecha-
41.
nism? [see comment] J. Exp. Med. 184, 18.
42.
42 Benjamin, R. J., Cobbold, S. P., Clark, M. R., and Waldmann, H. (1986) Tolerance to rat
monoclonal antibodies. Implications for serotherapy. J. Exp. Med. 163, 15391552.
43.
43 Gutstein, N. L., Seaman, W. E., Scott, J. H., and Wofsy, D. (1986) Induction of immune
tolerance by administration of monoclonal antibody to L3T4. J. Immunol. 137, 11271132.
44.
44 Cobbold, S. P., Adams, E., Marshall, S. E., Davies, J. D., and Waldmann, H. (1996) Mecha-
nisms of peripheral tolerance and suppression induced by monoclonal antibodies to CD4
and CD8. Immunol. Rev. 149, 533.
45.
45 Cobbold, S. P., Adams, E., Graca, L., and Waldmann, H. (2003) Serial analysis of gene
expression provides new insights into regulatory T cells. Semin. Immunol. 15, 209214.
46 Chatenoud, L. (2003) CD3-specific antibody-induced active tolerance: from bench to bed-
46.
side. Nat. Rev. Immunol. 3, 123132.
47.
47 Chatenoud, L., Thervet, E., Primo, J., and Bach, J. F. (1994) Anti-CD3 antibody induces
long-term remission of overt autoimmunity in nonobese diabetic mice. Proc. Natl. Acad.
Sci. USA 91, 123127.
48.
48 Nicolls, M. R., Aversa, G. G., Pearce, N. W., et al. (1993) Induction of long-term specific
tolerance to allografts in rats by therapy with an anti-CD3-like monoclonal antibody.
Transplantation 55, 459468.
49.
49 Pearson, T. C., Madsen, J. C., Larsen, C. P., Morris, P. J., and Wood, K. J. (1992) Induc-
tion of transplantation tolerance in adults using donor antigen and anti-CD4 monoclonal
antibody. Transplantation 54, 475483.
50.
50 Thomas, F. T., Ricordi, C., Contreras, J. L., et al. (1999) Reversal of naturally occuring
diabetes in primates by unmodified islet xenografts without chronic immunosuppression.
51.
51 Thomas, J. M., Neville, D. M., Contreras, J. L., et al. (1997) Preclinical studies of allograft
tolerance in rhesus monkeys: a novel anti-CD3-immunotoxin given peritransplant with
donor bone marrow induces operational tolerance to kidney allografts. Transplantation
64, 124135.
52.
52 Niimi, M., Pearson, T. C., Larsen, C. P., et al. (1998) The role of the CD40 pathway in
alloantigen-induced hyporesponsiveness in vivo. J. Immunol. 161, 53315337.
53.
53 Larsen, C. P., Elwood, E. T., Alexander, D. Z., et al. (1996) Long-term acceptance of skin
and cardiac allografts after blocking CD40 and CD28 pathways. Nature 381, 434438.
320 Chatenoud
54.
54 Kirk, A. D., Burkly, L. C., Batty, D. S., et al. (1999) reatment with humanized monoclonal
antibody against CD154 prevents acute renal allograft rejection in nonhuman primates
[see comments]. Nat. Med. 5, 686693.
55.
55 Bushell, A., Niimi, M., Morris, P. J., and Wood, K. J. (1999) Evidence for immune regu-
lation in the induction of transplantation tolerance: a conditional but limited role for IL-4.
J. Immunol. 162, 13591366.
56 Plain, K. M., Chen, J., Merten, S., He, X. Y., and Hall, B. M. (1999) Induction of specific
56.
tolerance to allografts in rats by therapy with non-mitogenic, non-depleting anti-CD3
monoclonal antibody: association with TH2 cytokines not anergy. Transplantation 67,
605613.
57.
57 Wood, K. J., Luo, S., and Akl, A. (2004) Regulatory T cells: potential in organ transplan-
tation. Transplantation 77, S68.
58.
58 Karim, M., Kingsley, C. I., Bushell, A. R., Sawitzki, B. S., and Wood, K. J. (2004) Alloan-
tigen-induced CD25+CD4+ regulatory T cells can develop in vivo from CD25-CD4+ pre-
cursors in a thymus-independent process. J. Immunol. 172, 923928.
59 Graca, L., Le Moine, A., Cobbold, S. P., and Waldmann, H. (2003) Antibody-induced
59.
transplantation tolerance: the role of dominant regulation. Immunol. Res. 28, 181191.
60.
60 Graca, L., Thompson, S., Lin, C. Y., Adams, E., Cobbold, S. P., and Waldmann, H. (2002)
Both CD4(+)CD25(+) and CD4(+)CD25(-) regulatory cells mediate dominant transplan-
tation tolerance. J. Immunol. 168, 55585565.
61.
61 Chatenoud, L., Salomon, B., and Bluestone, J. A. (2001) Suppressor T cellstheyre back
and critical for regulation of autoimmunity! Immunol. Rev. 182, 149163.
62.
62 Bach, J. F. and Chatenoud, L. (2001) Tolerance to islet autoantigens and type I diabetes.
Annu. Rev. Immunol. 19, 131161.
63.
63 Hall, B. M., Pearce, N. W., Gurley, K. E., and Dorsch, S. E. (1990) Specific unresponsive-
ness in rats with prolonged cardiac allograft survival after treatment with cyclosporine.
III. Further characterization of the CD4+ suppressor cell and its mechanisms of action. J.
Exp. Med. 171, 141157.
64.
64 Belghith, M., Bluestone, J. A., Barriot, S., Megret, J., Bach, J. F., and Chatenoud, L. (2003)
TGF- dependent mechanisms mediate restoration of self-tolerance induced by antibodies
to CD3 in overt autoimmune diabetes. Nature Medicine 9, 12021208.
65.
65 Zelenika, D., Adams, E., Humm, S., et al. (2002) Regulatory T cells overexpress a subset
of Th2 gene transcripts. J. Immunol. 168, 10691079.
66.
66 Chatenoud, L., Baudrihaye, M. F., Chkoff, N., Kreis, H., Goldstein, G., and Bach, J. F.
(1986) Restriction of the human in vivo immune response against the mouse monoclonal
antibody OKT3. J. Immunol. 137, 830838.
67.
67 Chatenoud, L. (1986) The immune response against therapeutic monoclonal antibodies.
Immunol. Today 7, 367368.
68. Chatenoud, L., Jonker, M., Villemain, F., Goldstein, G., and Bach, J. F. (1986) The human
immune response to the OKT3 monoclonal antibody is oligoclonal. Science 232, 14061408.
69.
69 Baudrihaye, M. F., Chatenoud, L., Kreis, H., Goldstein, G., and Bach, J. F. (1984) Unusu-
ally restricted anti-isotype human immune response to OKT3 monoclonal antibody. Eur.
J. Immunol. 14, 686691.
70.
70 Abramowicz, D., Crusiaux, A., and Goldman, M. (1992) Anaphylactic shock after
retreatment with OKT3 monoclonal antibody. N. Engl. J. Med. 327, 736.
71.
71 Abramowicz, D., Crusiaux, A., Niaudet, P., Kreis, H., Chatenoud, L., and Goldman, M.
(1996) The IgE humoral response in OKT3-treated patients - Incidence and fine specific-
ity. Transplantation 61, 577581.
72.
72 Hricik, D. E., Mayes, J. T., and Schulak, J. A. (1990) Inhibition of anti-OKT3 antibody
generation by cyclosporine-results of a prospective randomized trial. Transplantation
50, 237240.
73.
73 Riechmann, L., Clark, M., Waldmann, H., and Winter, G. (1988) Reshaping human anti-
bodies for therapy. Nature 332, 323327.
74 Bolt, S., Routledge, E., Lloyd, I., et al. (1993) The generation of a humanized, non-mito-
74.
genic CD3 monoclonal antibody which retains in vitro immunosuppressive properties.
Eur. J. Immunol. 23, 403411.
75.
75 Alegre, M.L., Peterson, L.J., Xu, D., et al. (1994) A non-activating humanized anti-CD3
monoclonal antibody retains immunosuppressive properties in vivo. Transplantation 57,
15371543.
76.
76 Elliott, M. J., Maini, R. N., Feldmann, M., et al. (1994) Repeated therapy with monoclonal
antibody to tumour necrosis factor alpha (cA2) in patients with rheumatoid arthritis. Lan-
cet 344, 11251127.
77.
77 Nashan, B., Moore, R., Amlot, P., Schmidt, A. G., Abeywickrama, K., and Soulillou, J. P.
(1997) Randomised trial of basiliximab versus placebo for control of acute cellular rejec-
tion in renal allograft recipients. CHIB 201 International Study Group [published erratum
appears in Lancet 1997 350(9089), 1484]. Lancet 350, 11931198.
78.
78 Vincenti, F., Kirkman, R., Light, S., et al. (1998) Interleukin-2-receptor blockade with
daclizumab to prevent acute rejection in renal transplantation. Daclizumab Triple Therapy
Study Group [see comments]. N. Engl. J. Med. 338, 161165.
79.
79 Chatenoud, L., Ferran, C., Legendre, C., et al. (1990) In vivo cell activation following
OKT3 administration. Systemic cytokine release and modulation by corticosteroids. Trans-
plantation 49, 697702.
80.
80 Abramowicz, D., Schandene, L., Goldman, M., et al. (1989) Release of tumor necrosis
factor, interleukin-2, and gamma-interferon in serum after injection of OKT3 monoclonal
antibody in kidney transplant recipients. Transplantation 47, 606608.
81.
81 Marks, J. D., Hoogenboom, H. R., Bonnert, T. P., Mccafferty, J., Griffiths, A. D., and
Winter, G. (1991) By-passing immunization. Human antibodies from V-gene libraries dis-
played on phage. J. Mol. Biol. 222, 581597.
82.
82 Lubin, I., Segall, H., Marcus, H., et al. (1994) Engraftment of human peripheral blood
lymphocytes in normal strains of mice. Blood 83, 23682381.
83.
83 Calne, R., Moffatt, S. D., Friend, P. J., et al. (1999) Campath IH allows low-dose cyclo-
sporine monotherapy in 31 cadaveric renal allograft recipients. Transplantation 68,
16131616.
84.
84 Mathieson, P. W., Cobbold, S.P., Hale, G., et al. (1990) Monoclonal-antibody therapy in
systemic vasculitis. N. Engl. J. Med. 323, 250254.
85.
85 Lockwood, C. M., Thiru, S., Isaacs, J. D., Hale, G., and Waldmann, H. (1993) Long-term
remission of intractable systemic vasculitis with monoclonal antibody therapy. Lancet 341,
16201622.
86.
86 Lockwood, C. M., Thiru, S., Stewart, S., et al. (1996) reatment of refractory Wegeners
granulomatosis with humanized monoclonal antibodies [see comments]. QJM 89, 903912.
87 Billingham, R. E., Brent, L., and Medawar, P. B. (1953) Actively acquired tolerance to
87.
foreign cells. Nature 172, 603606.
88.
88 Roser, B. J. (1989) Cellular mechanisms in neonatal and adult tolerance. Immunol. Rev.
107, 179202.
89.
89 Donckier, V., Flamand, V., Desalle, F., et al. (1998) IL-12 prevents neonatal induction of
transplantation tolerance in mice. Eur. J. Immunol. 28, 14261430.
322 Chatenoud
90.
90 Donckier, V., Wissing, M., Bruyns, C., et al. (1995) Critical role of interleukin 4 in the
induction of neonatal transplantation tolerance. Transplantation 59, 15711576.
91. Monaco, A. P., Wood, M. L., and Russell, P. S. (1966) Studies on heterologous antilym-
phocyte serum in mice. III. Immunological tolerance and chimerism produced across the
H2-locus with adult thymectomy and antilymphocyte serum. Ann. N. Y. Acad. Sci. 129,
190209.
92 Wood, M. L., Monaco, A. P., Gozzo, J. J., and Liegeois, A. (1971) Use of homozygous
92.
allogeneic bone marrow for induction of tolerance with antilymphocyte serum: dose and
timing. Transplant. Proc. 3, 676679.
93.
93 Thomas, F. T., Carver, F. M., Foil, M. B., et al. (1983) Long-term incompatible kidney
survival in outbred higher primates without chronic immunosuppression. Ann. Surg. 198,
370378.
94.
94 Thomas, J. M., Carver, F. M., Foil, M. B., Hall, W. R., Adams, C., and Fahrenbruch, G. B.
(1983) Renal allograft tolerance induced with ATG and donor bone marrow in outbred
rhesus monkeys. Transplantation 36, 104106.
95 Thomas, J., Alqaisi, M., Cunningham, P., et al. (1992) The development of a posttransplant
95.
TLI treatment strategy that promotes organ allograft acceptance without chronic immuno-
suppression. Transplantation 53, 247258.
96.
96 Hamawy, M. M. and Knechtle, S. J. (1998) Strategies for tolerance induction in nonhuman
primates. Curr. Opin. Immunol. 10, 513517.
97.
97 Qin, S. X., Wise, M., Cobbold, S. P., et al. (1990) Induction of tolerance in peripheral T
cells with monoclonal antibodies. Eur. J. Immunol. 20, 27372745.
98.
98 Gershon, R. K. and Kondo, K. (1971) Infectious immunological tolerance. Immunology
21, 903914.
99.
99 Cobbold, S. and Waldmann, H. (1998) Infectious tolerance. Curr. Opin. Immunol. 10,
518524.
100 Davies, J. D., Leong, L. Y., Mellor, A., Cobbold, S. P., and Waldmann, H. (1996) T cell
100.
suppression in transplantation tolerance through linked recognition. J. Immunol. 156,
36023607.
101.
101 Isobe, M., Yagita, H., Okumura, K., and Ihara, A. (1992) Specific acceptance of cardiac
allograft after treatment with antibodies to ICAM-1 and LFA-1. Science 255, 11251127.
102.
102 Larsen, C. P. and Pearson, T. C. (1997) The CD40 pathway in allograft rejection, accep-
tance, and tolerance. Curr. Opin. Immunol. 9, 641647.
103.
103 Mosmann, T. R. and Coffman, R. L. (1989) TH1 and TH2 cells: different patterns of lym-
phokine secretion lead to different functional properties. Annu. Rev. Immunol. 7, 145173.
104.
104 Nickerson, P., Steurer, W., Steiger, J., Zheng, X., Steele, A. W., and Strom, T. B. (1994)
Cytokines and the Th1/Th2 paradigm in transplantation. Curr. Opin. Immunol. 6, 757764.
105.
105 Bendelac, A., Carnaud, C., Boitard, C., and Bach, J. F. (1987) Syngeneic transfer of
autoimmune diabetes from diabetic NOD mice to healthy neonates. Requirement for both
L3T4+ and Lyt-2+ T cells. J. Exp. Med. 166, 823832.
106.
106 Wicker, L. S., Miller, B. J., and Mullen, Y. (1986) Transfer of autoimmune diabetes mel-
litus with splenocytes from nonobese diabetic (NOD) mice. Diabetes 35, 855860.
107 Yagi, H., Matsumoto, M., Kunimoto, K., Kawaguchi, J., Makino, S., and Harada, M.
107.
(1992) Analysis of the roles of CD4+ and CD8+ T cells in autoimmune diabetes of NOD
mice using transfer to NOD athymic nude mice. Eur. J. Immunol. 22, 23872393.
108.
108 Haskins, K., Portas, M., Bergman, B., Lafferty, K., and Bradley, B. (1989) Pancreatic
islet-specific T-cell clones from nonobese diabetic mice. Proc. Natl. Acad. Sci. USA 86,
80008004.
109.
109 Katz, J. D., Benoist, C., and Mathis, D. (1995) T helper cell subsets in insulin-dependent
diabetes. Science 268, 11851188.
110.
110 Healey, D., Ozegbe, P., Arden, S., Chandler, P., Hutton, J., and Cooke, A. (1995) In vivo
activity and in vitro specificity of CD4+ Th1 and Th2 cells derived from the spleens of
diabetic NOD mice. J. Clin. Invest. 95, 29792985.
111 Herbelin, A., Gombert, J. M., Lepault, F., Bach, J. F., and Chatenoud, L. (1998) Mature
111.
mainstream TCR alpha beta(+)CD4(+) thymocytes expressing L-selectin mediate active
tolerance in the nonobese diabetic mouse. J. Immunol. 161, 26202628.
112.
112 Lepault, F. and Gagnerault, M. C. (2000) Characterization of peripheral regulatory CD4(+)
T cells that prevent diabetes onset in nonobese diabetic mice. J. Immunol. 164, 240247.
113.
113 Lepault, F., Gagnerault, M. C., Faveeuw, C., Bazin, H., and Boitard, C. (1995) Lack of L-
selectin expression by cells transferring diabetes in NOD mice: insights into the mecha-
nisms involved in diabetes prevention by Mel-14 antibody treatment. Eur. J. Immunol. 25,
15021507.
114.
114 Chatenoud, L., Primo, J., and Bach, J. F. (1997) CD3 antibody-induced dominant self
tolerance in overtly diabetic NOD mice. J. Immunol. 158, 29472954.
115 Alegre, M. L., Collins, A. M., Pulito, V. L., et al. (1992) Effect of a single amino acid
115.
mutation on the activating and immunosuppressive properties of a humanized OKT3
monoclonal antibody. J. Immunol. 148, 34613468.
116.
116 Herold, K. C., Hagopian, W., Auger, J. A., et al. (2002) Anti-CD3 monoclonal antibody in
new-onset type 1 diabetes mellitus. N. Engl. J. Med. 346, 16921698.
117.
117 Cosimi, A. B., Burton, R. C., Colvin, R. B., et al. (1981) Treatment of acute renal allograft
rejection with OKT3 monoclonal antibody. Transplantation 32, 535539.
118.
118 Cosimi, A. B., Colvin, R. B., Burton, R. C., et al. (1981) Use of monoclonal antibodies to
T-cell subsets for immunologic monitoring and treatment in recipients of renal allografts.
N. Engl. J. Med. 305, 308314.
119. Ortho Multicenter Transplant Study Group (1985) A randomized clinical trial of OKT3
monoclonal antibody for acute rejection of cadaveric renal transplants. N. Engl. J. Med.
313, 337342.
120.
120 Goldstein, G., Kremer, A. B., Barnes, L., and Hirsch, R. L. (1987) OKT3 monoclonal anti-
body reversal of renal and hepatic rejection in pediatric patients. J. Pediatr. 111, 10461050.
121.
121 Hricik, D. E., Zarconi, J., and Schulak, J. A. (1989) Influence of low-dose cyclosporine on
the outcome of treatment with OKT3 for acute renal allograft rejection. Transplantation
47, 272277.
122.
122 Norman, D. J., Barry, J. M., Bennett, W. M., et al. (1988) The use of OKT3 in cadaveric
renal transplantation for rejection that is unresponsive to conventional anti-rejection
therapy. Am. J. Kidney Dis. 11, 9093.
123.
123 Colonna Jo, D., Goldstein, L. I., Brems, J. J., et al. (1987) A prospective study on the use
of monoclonal anti-T3-cell antibody (OKT3) to treat steroid-resistant liver transplant re-
jection. Arch. Surg. 122, 11201123.
124.
124 Woodle, E. S., Thistlethwaite Jr, J. R., Emond, J. C., et al. (1991) OKT3 therapy for he-
patic allograft rejection. Differential response in adults and children. Transplantation 51,
12071212.
125 Gilbert, E. M., Dewitt, C. W., Eiswirth, C. C., et al. (1987) Treatment of refractory cardiac
125.
allograft rejection with OKT3 monoclonal antibody. Am. J. Med. 82, 202206.
126 Frey, D. J., Matas, A. J., Gillingham, K. J., et al. (1992) Sequential therapya prospective
126.
randomized trial of MALG versus OKT3 for prophylactic immunosuppression in cadaver
renal allograft recipients. Transplantation 54, 5056.
324 Chatenoud
127.
127 Debure, A., Chkoff, N., Chatenoud, L., et al. (1988) One-month prophylactic use of OKT3
in cadaver kidney transplant recipients. Transplantation 45, 546553.
128.
128 Abramowicz, D., Goldman, M., De Pauw, L., Vanherweghem, J. L., Kinnaert, P., and
Vereerstraeten, P. (1992) The long-term effects of prophylactic OKT3 monoclonal anti-
body in cadaver kidney transplantationa single-center, prospective, randomized study.
129 Opelz, G. (1995) Efficacy of rejection prophylaxis with OKT3 in renal transplantation.
129.
Collaborative Transplant Study. Transplantation 60, 12201224.
130.
130 Farges, O., Ericzon, B. G., Bresson-hadni, S., et al. (1994) A randomized trial of OKT3-
based versus cyclosporine-based immunoprophylaxis after liver transplantation. Long-
term results of a European and Australian multicenter study. Transplantation 58, 891898.
131.
131 Millis, J. M., Mcdiarmid, S. V., Hiatt, J. R., et al. (1989) Randomized prospective trial of
OKT3 for early prophylaxis of rejection after liver transplantation. Transplantation 47,
8288.
132. Eason, J. D. and Cosimi, A. B. (1999) Biologic immunosuppressive agents. In: Transplan-
tation. Ginns, L., Cosimi, A., and Morris, P., eds. Blackwell Science, Malden, MA, pp.
196224.
133.
133 Robbins, R. C., Oyer, P. E., Stinson, E. B., and Starnes, V. A. (1992) The use of mono-
clonal antibodies after heart transplantation. Transplant. Sci. 2, 2227.
134.
134 Chatenoud, L., Ferran, C., Reuter, A., et al. (1989) Systemic reaction to the anti-T-cell
monoclonal antibody OKT3 in relation to serum levels of tumor necrosis factor and inter-
feron-gamma. N. Engl. J. Med. 320, 14201421.
135.
135 Cosimi, A. B. (1987) Clinical development of Orthoclone OKT3. Transplant. Proc. 19,
716.
136.
136 Friend, P. J., Hale, G., Chatenoud, L., et al. (1999) Phase I study of an engineered
aglycosylated humanized CD3 antibody in renal transplant rejection. Transplantation 68,
16321637.
137.
137 Woodle, E. S., Xu, D., Zivin, R. A., et al. (1999) Phase I trial of a humanized, Fc receptor
nonbinding OKT3 antibody, huOKT3gamma1(Ala-Ala) in the treatment of acute renal
allograft rejection. Transplantation 68, 608616.
138.
138 Le Mauff, B., Hourmant, M., Rougier, J. P., et al. (1991) Effect of anti-LFA1 (CD11a)
monoclonal antibodies in acute rejection in human kidney transplantation. Transplanta-
tion 52, 291296.
139.
139 Maraninchi, D., Mawas, C., Stoppa, A. M., et al. (1989) Anti LFA1 monoclonal antibody
for the prevention of graft rejection after T cell-depleted HLA-matched bone marrow trans-
plantation for leukemia in adults. Bone Marrow Transplant. 4, 147150.
140.
140 Soulillou, J. P., Cantarovich, D., Le Mauff, B., et al. (1990) Randomized controlled trial
of a monoclonal antibody against the interleukin-2 receptor (33B3.1) as compared with
rabbit antithymocyte globulin for prophylaxis against rejection of renal allografts. N. Engl.
J. Med. 322, 11751182.
141 Kriaa, F., Hiesse, C., Alard, P., et al. (1993) Prophylactic use of the anti-IL-2 receptor
141.
monoclonal antibody LO-Tact-1 in cadaveric renal transplantation: results of a random-
ized study. Transplant. Proc. 25, 817819.
142.
142 Kirkman, R. L., Shapiro, M. E., Carpenter, C. B., et al. (1991) A randomized prospective
trial of anti-Tac monoclonal antibody in human renal transplantation. Transplantation 51,
107113.
143.
143 Waldmann, T. A. and OShea, J. (1998). The use of antibodies against the IL-2 receptor in
transplantation. Curr. Opin. Immunol. 10, 507512.
144.
144 Bumgardner, G. L., Hardie, I., Johnson, R. W., et al. (2001) Results of 3-year phase III
clinical trials with daclizumab prophylaxis for prevention of acute rejection after renal
transplantation. Transplantation 72, 839845.
145.
145 Adu, D., Cockwell, P., Ives, N. J., Shaw, J., and Wheatley, K. (2003) Interleukin-2 recep-
tor monoclonal antibodies in renal transplantation: meta-analysis of randomised trials.
BMJ 326, 789.
146 Nussenblatt, R. B. (2002) Bench to bedside: new approaches to the immunotherapy of
146.
uveitic disease. Int. Rev. Immunol. 21, 273289.
147.
147 Guex-Crosier, Y., Raber, J., Chan, C. C., et al. (1997) Humanized antibodies against
the alpha-chain of the IL-2 receptor and against the beta-chain shared by the IL-2 and
IL-15 receptors in a monkey uveitis model of autoimmune diseases. J. Immunol. 158,
452458.
148.
148 Shizuru, J. A., Taylor-Edwards, C., Banks, B. A., Gregory, A. K., and Fathman, C. G.
(1988) Immunotherapy of the nonobese diabetic mouse: treatment with an antibody to T-
helper lymphocytes. Science 240, 659662.
149 Wofsy, D. and Seaman, W. E. (1987) Reversal of advanced murine lupus in NZB/NZW
149.
F1 mice by treatment with monoclonal antibody to L3T4. J. Immunol. 138, 32473253.
150.
150 Horneff, G., Burmester, G. R., Emmrich, F., and Kalden, J. R. (1991) Treatment of rheu-
matoid arthritis with an anti-CD4 monoclonal antibody. Arthritis Rheum. 34, 129140.
151.
151 Wendling, D., Wijdenes, J., Racadot, E., and Morel-Fourrier, B. (1991) Therapeutic use of
monoclonal anti-CD4 antibody in rheumatoid arthritis. J. Rheumatol. 18, 325327.
152.
152 Morel, P., Revillard, J. P., Nicolas, J. F., Wijdenes, J., Rizova, H., and Thivolet, J. (1992)
Anti-CD4 monoclonal antibody therapy in severe psoriasis. J. Autoimmun. 5, 465477.
153.
153 Emmrich, J., Seyfarth, M., Fleig, W. E., and Emmrich, F. (1991) Treatment of inflamma-
tory bowel disease with anti-CD4 monoclonal antibody. Lancet 338, 570571.
154.
154 Reiter, C., Kakavand, B., Rieber, E. P., Schattenkirchner, M., Riethmuller, G., and Kruger,
K. (1991) Treatment of rheumatoid arthritis with monoclonal CD4 antibody M-T151.
Clinical results and immunopharmacologic effects in an open study, including repeated
administration. Arthritis Rheum. 34, 525536.
155.
155 Goldberg, D., Chatenoud, L., Morel, P., et al. (1990) Preliminary trial of an anti-CD4
monoclonal antibody (MoAb) in rheumatoid arthritis (RA). Arthritis Rheum. 33, S153.
156. Goldberg, D., Morel, P., Chatenoud, L., et al. (1991) Immunological effects of high dose
administration of anti-CD4 antibody in rheumatoid arthritis patients. J. Autoimmun. 4,
617630.
157.
157 Bachelez, H., Flageul, B., Dubertret, L., et al. (1998) Treatment of recalcitrant plaque
psoriasis with a humanized non-depleting antibody to CD4. J. Autoimmun. 11, 5362.
158.
158 Moreland, L. W., Bucy, R. P., Tilden, A., et al. (1993) Use of a chimeric monoclonal
anti-CD4 antibody in patients with refractory rheumatoid arthritis. Arthritis Rheum. 36,
307318.
159.
159 Van Der Lubbe, P. A., Dijkmans, B. A., Markusse, H. M., Nassander, U., and Breedveld,
F. C. (1995) A randomized, double-blind, placebo-controlled study of CD4 monoclonal
antibody therapy in early rheumatoid arthritis. Arthritis Rheum. 38, 10971106.
160.
160 Van Der Lubbe, P. A., Reiter, C., Breedveld, F. C., et al. (1993) Chimeric CD4 mono-
clonal antibody cM-T412 as a therapeutic approach to rheumatoid arthritis. Arthritis
Rheum. 36, 13751379.
161.
161 Van Der Lubbe, P. A., Reiter, C., Miltenburg, A. M., et al. (1994) Treatment of rheumatoid
arthritis with a chimeric CD4 monoclonal antibody (cM-T412): immunopharmacological
aspects and mechanisms of action. Scand. J. Immunol. 39, 286294.
326 Chatenoud
162.
162 Choy, E. H., Adjaye, J., Forrest, L., Kingsley, G. H., and Panayi, G. S. (1993) Chimaeric
anti-CD4 monoclonal antibody cross-linked by monocyte Fc gamma receptor mediates
apoptosis of human CD4 lymphocytes. Eur. J. Immunol. 23, 26762681.
163.
163 van Oosten, B. W., Lai, M., Hodgkinson, S., et al. (1997) Treatment of multiple sclerosis
with the monoclonal anti-CD4 antibody cM-T412: results of a randomized, double-blind,
placebo-controlled, MR-monitored phase II trial. Neurology 49, 351357.
164 Gottlieb, A. B., Lebwohl, M., Shirin, S., et al. (2000) Anti-CD4 monoclonal antibody
164.
treatment of moderate to severe psoriasis vulgaris: results of a pilot, multicenter, multiple-
dose, placebo-controlled study. J. Am. Acad. Dermatol. 43, 595604.
165.
165 Brennan, F. M. and Feldmann, M. (1992) Cytokines in autoimmunity. Curr. Opin.
Immunol. 4, 754759.
166.
166 Williams, R. O., Mason, L. J., Feldmann, M., and Maini, R. N. (1994) Synergy between
anti-CD4 and anti-tumor necrosis factor in the amelioration of established collagen-in-
duced arthritis. Proc. Natl. Acad. Sci. USA 91, 27622766.
167.
167 Elliott, M. J., Maini, R. N., Feldmann, M., et al. (1994) Randomised double-blind com-
parison of chimeric monoclonal antibody to tumour necrosis factor alpha (cA2) versus
placebo in rheumatoid arthritis. Lancet 344, 11051110.
168.
168 Maini, R. N., Breedveld, F. C., Kalden, J. R., et al. (1998) herapeutic efficacy of multiple
intravenous infusions of anti-tumor necrosis factor alpha monoclonal antibody combined
with low-dose weekly methotrexate in rheumatoid arthritis [see comments]. Arthritis
Rheum. 41, 15521563.
169.
169 Feldmann, M. and Maini, T. (2001) Anti-TNF alpha therapy of rheumatoid arthritis: what
have we learned? Annu. Rev. Immunol. 19, 163196.
170.
170 Maini, R., St Clair, E. W., Breedveld, F., et al. (1999) Infliximab (chimeric anti-tumour
necrosis factor alpha monoclonal antibody) versus placebo in rheumatoid arthritis patients
receiving concomitant methotrexate: a randomised phase III trial. ATTRACT Study
Group. Lancet 354, 19321939.
171.
171 Lipsky, P. E., van der Heijde, D. M., St Clair, E. W., et al. (2000) Infliximab and methotr-
exate in the treatment of rheumatoid arthritis. Anti-Tumor Necrosis Factor Trial in
Rheumatoid Arthritis with Concomitant Therapy Study Group. N. Engl. J. Med. 343,
15941602.
172.
172 Rankin, E. C., Choy, E. H., Kassimos, D., et al. (1995) The therapeutic effects of an engi-
neered human anti-tumour necrosis factor alpha antibody (CDP571) in rheumatoid arthri-
tis. Br. J. Rheumatol. 34, 334342.
173.
173 Furst, D. E., Weisman, M., Paulus, H. E., et al. (2003) Intravenous human recombinant
tumor necrosis factor receptor p55-Fc IgG1 fusion protein, Ro 45-2081 (lenercept):
results of a dose-finding study in rheumatoid arthritis. J. Rheumatol. 30, 21232126.
174.
174 Moreland, L. W., Baumgartner, S. W., Schiff, M. H., et al. (1997) reatment of rheumatoid
arthritis with a recombinant human tumor necrosis factor receptor (p75)-Fc fusion protein
[see comments]. N. Engl. J. Med. 337, 141147.
175.
175 Moreland, L. W., Margolies, G., Heck, Jr., L. W., et al. (1996) Recombinant soluble tumor
necrosis factor receptor (p80) fusion protein: toxicity and dose finding trial in refractory
rheumatoid arthritis. J. Rheumatol. 23, 18491855.
176 Kempeni, J. (1999) Preliminary results of early clinical trials with the fully human anti-
176.
TNFalpha monoclonal antibody D2E7. Ann. Rheum. Dis. 58(Suppl 1), I7072.
177 den Broeder, A., van de Putte, L., Rau, R., et al. (2002) A single dose, placebo controlled
177.
study of the fully human anti-tumor necrosis factor-alpha antibody adalimumab (D2E7) in
patients with rheumatoid arthritis. J. Rheumatol. 29, 22882298.
178.
178 Van Dullemen, H. M., Van Deventer, S. J., Hommes, D. W., et al. (1995) Treatment of
Crohns disease with anti-tumor necrosis factor chimeric monoclonal antibody (cA2).
Gastroenterology 109, 129135.
179.
179 Ghosh, S., Goldin, E., Gordon, F. H., et al. (2003) Natalizumab for active Crohns disease.
N. Engl. J. Med. 348, 2432.
180.
180 Van Deventer, S. J. (1997) Tumour necrosis factor and Crohns disease Gut 40, 443448.
181 Plevy, S. E., Landers, C. J., Prehn, J., et al. (1997) A role for TNF-alpha and mucosal T
181.
helper-1 cytokines in the pathogenesis of Crohns disease. J. Immunol. 159, 62766282.
182.
182 Lovell, D. J., Giannini, E. H., Reiff, A., et al. (2000) Etanercept in children with pol-
yarticular juvenile rheumatoid arthritis. Pediatric Rheumatology Collaborative Study
Group. N. Engl. J. Med. 342, 763769.
183.
183 Brandt, J., Haibel, H., Cornely, D., et al. (2000) Successful treatment of active
ankylosing spondylitis with the anti-tumor necrosis factor alpha monoclonal antibody
infliximab. Arthritis Rheum. 43, 13461352.
184.
184 Mease, P. J., Goffe, B. S., Metz, J., VanderStoep, A., Finck, B., and Burge, D. J. (2000)
Etanercept in the treatment of psoriatic arthritis and psoriasis: a randomised trial. Lancet
356, 385390.
185 Chaudhari, U., Romano, P., Mulcahy, L. D., Dooley, L. T., Baker, D. G., and Gottlieb, A.
185.
B. (2001) Efficacy and safety of infliximab monotherapy for plaque-type psoriasis: a
randomised trial. Lancet 357, 18421847.
186.
186 Gomez-Reino, J. J., Carmona, L., Valverde, V. R., Mola, E. M., and Montero, M. D. (2003)
Treatment of rheumatoid arthritis with tumor necrosis factor inhibitors may predispose to
significant increase in tuberculosis risk: a multicenter active-surveillance report. Arthritis
Rheum. 48, 21222127.
187.
187 Isaacs, J. D., Watts, R. A., Hazleman, B. L., et al. (1992) Humanised monoclonal antibody
therapy for rheumatoid arthritis. Lancet 340, 748752.
188.
188 Moreau, T., Thorpe, J., Miller, D., et al. (1994) Preliminary evidence from magnetic reso-
nance imaging for reduction in disease activity after lymphocyte depletion in multiple scle-
rosis [published erratum appears in Lancet 1994 344(8920), 486]. Lancet 344, 298301.
189.
189 Coles, A. J., Wing, M. G., Molyneux, P., et al. (1999) Monoclonal antibody treatment
exposes three mechanisms underlying the clinical course of multiple sclerosis. Ann.
Neurol. 46, 296304.
190.
190 Coles, A. J., Wing, M. G., Smith, S., et al. (1999) Pulsed monoclonal antibody treatment
and autoimmune thyroid disease in multiple sclerosis. Lancet 354, 16911695.
191.
191 Knechtle, S. J., Pirsch, J. D., Fechner, J. H., et al. (2003) Campath-1H induction plus
rapamycin monotherapy for renal transplantation: results of a pilot study. Am. J. Trans-
plant. 3, 722730.
192.
192 Kirk, A. D., Hale, D. A., Mannon, R. B., et al. (2003) Results from a human renal allograft
tolerance trial evaluating the humanized CD52-specific monoclonal antibody alemtuzumab
(CAMPATH-1H). Transplantation 76, 120129.
193.
193 Miller, D. H., Khan, O. A., Sheremata, W. A., et al. (2003) A controlled trial of natalizumab
for relapsing multiple sclerosis. N. Engl. J. Med. 348, 1523.
194.
194 Lebwohl, M., Christophers, E., Langley, R., Ortonne, J. P., Roberts, J., and Griffiths, C. E.
(2003) An international, randomized, double-blind, placebo-controlled phase 3 trial of intra-
muscular alefacept in patients with chronic plaque psoriasis. Arch. Dermatol. 139, 719727.
195.
195 Krueger, G. G. (2003) Clinical response to alefacept: results of a phase 3 study of intrave-
nous administration of alefacept in patients with chronic plaque psoriasis. J. Eur. Acad.
Dermatol. Venereol. 17 Suppl 2, 1724.
328 Chatenoud
196.
196 Rastetter, W., Molina, A., and White, C. A. (2004) Rituximab: expanding role in therapy
for lymphomas and autoimmune diseases. Annu. Rev. Med. 55, 477503.
197.
197 Shaw, T., Quan, J., and Totoritis, M. C. (2003) B cell therapy for rheumatoid arthritis: the
rituximab (anti-CD20) experience. Ann. Rheum. Dis. 62(Suppl 2), ii5559.
198.
198 Silverman, G. J. and Weisman, S. (2003) Rituximab therapy and autoimmune disorders:
prospects for anti-B cell therapy. Arthritis Rheum. 48, 14841492.
199. Leandro, M. J., Edwards, J. C., and Cambridge, G. (2002) Clinical outcome in 22 patients
with rheumatoid arthritis treated with B lymphocyte depletion. Ann. Rheum. Dis. 61,
883888.
200. De Vita, S., Zaja, F., Sacco, S., De Candia, A., Fanin, R., and Ferraccioli, G. (2002)
Efficacy of selective B cell blockade in the treatment of rheumatoid arthritis: evidence
for a pathogenetic role of B cells. Arthritis Rheum. 46, 20292033.
Bispecific and Bifunctional Antibodies 329
22
Producing Bispecific and Bifunctional Antibodies
Dipankar Das and Mavanur R. Suresh
Summary
Bispecific antibodies are artificially engineered monoclonal antibodies (MAbs) that consist
of two distinct binding sites and are capable of binding two different antigens noncovalently.
They can be produced by chemical cross-linkage, genetic engineering, or somatic hybridiza-
tion. This chapter describes a rapid method using somatic fusion to generate hybrid hybridomas
(quadromas). Two fluorescence-labeled hybridoma cell lines were fused with polyethylene gly-
col (PEG) to generate the quadroma. Generation of a quadroma secreting bsMAb against biotin
and HRPO is described, along with a benzhydroxamic acid-agarose affinity chromatography
procedure to purify the bsMAbHRPO complex. This bsMAb can be used for ultrasensitive
ELISA detection of biotinylated antigens. Essentially a similar method can be used for fusing
any two hybridomas for therapeutic applications.
Bifunctional antibodies are colinear molecules with one or more paratopes linked with diag-
nostic or therapeutic molecules. There are some limitations of therapeutic monoclonal antibod-
ies in the clinic that can be overcome by engineering smaller and more effective antibody
fragments. Here we describe a stepwise procedure for developing a bifunctional ScFv (bfScFv).
We constructed a bfScFv from a hybridoma cell line using PCR strategies. The VL and VH
gene segments are linked with a 45-bp linker and fused with a biotin mimic sequence at the 3'
end. This engineered bifunctional antibody fragment gene could be expressed and the protein
purified on a large scale in Escherichia coli as inclusion bodies. Such bifunctional antibody
molecules could have useful applications in the area of immunodiagnostics and immunotherapy.
Similar strategies can be used to incorporate a second single-chain antibody or any nonantibody
entity such as a cytokine for therapeutic applications.
Key Words: Bispecific antibodies; FITC; TRITC; polyethylene glycol (PEG) fusion;
ELISA; bifunctional antibodies; ScFv; PCR; inclusion bodies; refolding.
1. Introduction
In the 1950s, cancer immunotherapy and radioimmunotherapy emerged as a new
strategy based on polyclonal antibody preparations (1). The heterogeneous nature of
polyclonals was a major limitation. With the development of the hybridoma tech-
nology for producing monoclonal antibodies (MAbs) by Kohler and Milstein (2) in
329
330 Das and Suresh
1975, it became possible to make large quantities of pure, identical antibodies against
a particular antigen. This landmark discovery has revolutionized the use of MAb in
the area of immunodetection and immunotherapy of cancers, infectious, and other
diseases. In the last 30 yr, with the rapid developments in recombinant DNA tech-
nology, facile preparation of recombinant antibodies has been achieved (3,4). This
technology allowed the design flexibility to alter size, configuration, valency, effec-
tor functions, and species-specific domains in the construction of MAbs. These
include whole chimeric, humanized antibodies and their small antibody reductants,
such as Fab, F(ab')2, and single-chain variable (ScFv) and domain antibodies. A
single-chain antibody molecule consists of the two variable domains (VH and VL)
of an antibody molecule joined together by a designed artificial polypeptide linker
such that the antigen-binding site is retained in a single co-linear protein (5,6).
Bispecific antibodies (bsMAb) are artificially engineered immunoglobulins incor-
porating two distinct binding specificities, and were first generated by chemical
methods (7). A bispecific antibody incorporates two distinct paratopes, which are
capable of binding two different epitopes or antigens noncovalently. In contrast, a
bifunctional antibody (bfMAb) refers to the colinear molecules that have one or
more paratopes linked to a diagnostic or therapeutic non-antibody entity (8). With
the specificity of antigen binding coupled with unique functionalities, applications
range from immunohistochemistry, ultrasensitive immunodetection, radioimmuno-
therapy, radioimmunodiagnosis, and immunotherapy to targeted drug delivery.
Bispecific monoclonal antibodies have been extensively studied in both diagnostic
(9,10) and therapeutic areas (1119). These types of nanoprobes can be made by
chemical fusion, somatic fusion, or genetic engineering methods (20).
In this chapter, we will describe in a stepwise fashion how to develop (a)
bispecific MAb for diagnostic applications using quadromas (hybrid hybridoma);
and (b) bifunctional ScFv (bfScFv) for therapeutic applications using recombinant
antibody technology.
2. Materials
2.1. Quadroma Development
1. Mouse IgG1 antibiotin hybridoma cell line (P54).
2. Rat IgG2a antihorseradish peroxidase (HRPO) hybridoma cell line (YP4).
3. Fluorescein isothiocyanate (FITC; Sigma, St. Louis, MO).
4. Tetramethyl rhodamine isothiocyanate (TRITC; Sigma).
5. DMEM medium (Gibco-BRL, Roockville, MD).
6. Fetal bovine serum (FBS; Gibco-BRL).
7. Penicillin, streptomycin, and glutamine (PSG; Gibco-BRL).
8. Ninety-six-well sterile tissue-culture plate (Nunc, Naperville, IL).
9. Ninety-six-well ELISA plate (Nunc).
10. Goat antimouse horseradish peroxidase (GAM-HRPO; Sigma).
11. 10% BSA (dialyzed in 1X PBS, pH 7.3).
12. 3,3',5,5'-Tetra-methyl benzidine (TMB) peroxidase substrate (Kirkegaard & Perry Labo-
ratories, Gaithersburg, MD).
13. Polyethylene glycol (PEG; Hybrimax, Sigma).
14. Benzhydroxamic acid agarose (BHA; Upfront Chromatography, Copenhagen, Den-

mark).
15. Dimethyl sulfoxide (DMSO, Sigma).
16. 0.22 M syringe filter (Millipore, Bedford, MA).
2.2. Bifunctional Single-Chain Variable-Fragment Antibody Development

1. Anti-WEE (western equine encephalitis) producing hybridoma (11D2) cell lines (De-
fence Research and Development Canada, Medicine Hat, AB, Canada).
2. DMEM cell-culture medium (Gibco-BRL).
3. Fetal bovine serum (FBS, Gibco-BRL).
4. Penicillin, streptomycin, and glutamine (PSG, Gibco-BRL).
5. TRizol reagent (Gibco-BRL).
6. Oligonucleotide primers.
7. PCR machine (Robocycler 40, Stratagene, La Jolla, CA).
8. 100 mM dNTPs (Pharmacia, New York, NY).
9. RNasin (Promega, Madison, WI).
10. Diethyl pyrocarbonate (DEPC, Sigma).
11. Agarose (Gibco-BRL).
12. TAE, 50X: 242 g Tris base, 57.1 mL glacial acetic acid, 100 mL 0.5 M EDTA (pH 8.0)
per 1 L water.
13. Ethidium bromide (EtBr, Sigma).
14. DNA mol wt markers, T4 DNA ligase, Superscript II, restriction enzymes (Gibco-BRL).
15. E. coli strain TOP 10 F (Invitrogen, Carlsbad, CA).
16. E. coli strain BL21 (DE3) (Novagen, Madison, WI).
17. Expression vector pET22b+ (Novagen).
18. Isopropyl -D-thiogalacto pyranoside (IPTG, Gibco-BRL).
19. Lysozyme (Sigma).
20. Prestained protein mol wt markers, Biorad protein assay reagent (Bio-Rad, Hercules,
CA).
21. Vent DNA polymerase (New England Biolabs [NEB], Beverly, MA).
22. L-arginine (Sigma).
23. Urea (Sigma).
24. Sarcosyl (Sigma).
25. Carbenicillin (Sigma).
26. SOC medium: 2% tryptone, 0.5% yeast extract, 10 mM NaCl, 2.5 mM KCl, 10 mM MgCl2,
10 mM MgSO4, 20 mM glucose.
27. QIAprep plasmid DNA isolation kit (Qiagen, Valencia, CA).
28. QIAquick gel extraction kit (Qiagen).
29. Lysis buffer: 50 mM Tris-HCl (pH 8.0), 200 mM NaCl, 1 mM EDTA.
30. Refolding buffer: 50 mM Tris-HCl (pH 8.0), 0.4 M L-arginine.
31. Dialysis membrane (MWCO: 12,00014,000 Da).
32. Petri dish.
33. Luria Bertani (LB) medium: 1% tryptone, 0.5% yeast extract, 0.5% NaCl, pH 7.3.
34. TMB peroxidase substrate (Kirkegaard & Perry Laboratories Inc).
35. 10% BSA; dialyzed in 1X PBS, pH 7.3.
36. Streptavidin horseradish peroxidase (St-HRPO) (Sigma).
37. PBST: PBS with 0.1% Tween-20, pH 7.3.
38. ELISA plate (Nunc).
332 Das and Suresh
3. Methods
3.1. Quadroma Development
Somatic fusion for the development of bispecific MAb involves (1) fusion of two
established hybridomas, generating a quadroma (9,21); and (2) fusion of one esta-
blished hybridoma (instead of a nonsecreting myeloma) with lymphocytes derived
from a mouse immunized with a second antigen generating a trioma (22). Here we will
describe the stepwise protocol for the development of a quadroma from two well-
established, fluorescence-labeled hybridomas (23,24). The method described below
includes the different steps of (1) cell labeling; (2) PEG fusion; (3) FACS; (4) screen-
ing for quadromas; (5) affinity purification of bsMAb; and (6) characterization of
bsMAb.
3.1.1. Cell Labeling
1. Prepare the fluorescent dye (FITC and TRITC) solution as a 1 mg/mL stock in DMSO
and sterile filter with a 0.22-m syringe filter.
2. Wash separately 2 107 hybridoma cells of P54 and YP4 in 10 mL serum-free DMEM
(SFDMEM) medium three times at 800 rpm (114g) for 5 min at room temperature. Resus-
pend the P54 cells in 10 mL SFDMEM medium, pH 7.4, and the YP4 cells in 10 mL
SFDMEM medium, pH 6.8. Calculate the viability of P54 and YP4 hybridoma cells by
the trypan blue exclusion method prior to fusion. Ideally, viability should be approx 90
95% (see Note 1).
3. Label the P54 cells with TRITC solution (final concentration of 2 g/mL in SFDMEM,
pH 7.4) and the YP4 cells with FITC solution (final concentration of 1 g/mL in SFDMEM,
pH 6.8) (25).
4. Incubate cells with the respective fluorescent dyes for 30 min at 37C with gentle shaking
(see Note 2).
5. Wash with SFDMEM medium to remove any free fluorescent dye.
6. Resuspend in 10 mL of SFDMEM medium.
3.1.2. PEG Fusion
1. Mix 5 mL of TRITC-labeled P54 cells (107 cells) with 5 mL of FITC-labeled YP4 cells
(107 cells) and centrifuge at 800 rpm (114g) for 5 min at room temperature.
2. Decant the supernatant completely and resuspend the cells in remaining medium (50
100 L) (see Note 3).
3. Fuse approx 107 cells from the TRITC-labeled P54 hybridoma with the same number of
cells from the FITC-labeled YP4 hybridoma by using 500 L of 50% (w/v) polyethylene
glycol (PEG) solution for 2 min in a CO2 incubator (Fig. 1) (see Note 4).
4. Centrifuge the cells at 800 rpm (114g) for 5 min at room temperature to remove PEG and
resuspend in 15 mL of DMEM medium with 1015% FBS.
5. Wash the cells three times in 15 mL DMEM with 1015% FBS and finally suspend in
10 mL of DMEM medium with 1015% FBS (see Note 5).
6. Transfer the cells to a 25-cm2 tissue-culture flask and incubate for 1 h at 37C with 5%
CO2 in a standard tissue-culture incubator.
3.1.3. FACS
1. Perform flow cytometric cell sorting using Epics Elite Cell Sorter (Coulter Corporation,
Hialeah, FL) or a similar instrument.
Fig. 1. Outline of PEG fusion.
2. Sort cells with dual fluorescence and clone one to two cells per well in a 96-well tissue-
culture plate (see Note 6).
3. Culture the sorted cells at 37C with 5% CO2 for 13 wk and monitor for the appearance
of clones.
4. Screen the resulting putative quadromas for bispecific monoclonal antibody secretion
after 1020 d as described below in Subheading 3.1.4.
3.1.4. Screening for Quadromas Secreting Bispecific Antibodies
1. Coat NUNC microtiter plates with biotin-BSA (1 g/well, in 100 L of PBS using a stock
of 10 g/mL) overnight at 4C (10,26).
2. Wash the plate three times with 200 L of PBST (PBS with 0.1% Tween-20).
3. Block free binding sites with 200 L of 1% BSA at 37C for 1 h.
4. Dilute (1:10, 1:100, and 1:1000) various cell-culture supernatants with 1% BSA in PBS.
Add 100 L to the antigen-coated well along with 50 L of 10 g/mL HRPO in 1% BSA
per well, and incubate for 1 h at 37C (see Note 7).
5. Wash the plate three times with PBST and add 100 L of TMB peroxidase substrate to
each well.
6. Select positive quadromas secreting bsMAb after 1545 min of color development by
measuring OD at 650 nm using an ELISA plate reader.
334 Das and Suresh
7. This bridge ELISA measures only the bsMAb activity, by binding to the solid-phase
antigen with one paratope (antibiotin) and in solution to the other.
8. Select strongly positive clones and subclone by limiting dilution three successive times to
get the best-producing quadroma (see Note 8).
3.1.5 Affinity Co-Chromatography of bsMAb

1. Harvest the quadroma supernatant containing the bsMAb at 1500 rpm (403g) for 7 min at
room temperature to remove the cells.
2. Add ammonium sulphate to 50% saturation to the supernatant and precipitate the bsMAb.
Resuspend the pellet in a minimum volume of PBS and dialyze against PBS, pH 7.0 (27).
3. Combine approx 150 mg of crude bsMAb with 10 mg of HRPO in PBS, pH 7.0, and
incubate for 20 min at 37C.
4. Load HRPO-bsMAb complex to a 15-mL benzhydroxamic acidagarose (BHA) (1.25
30 cm column) matrix and adjust flow rate to 20 mL/ h (28) (see Note 9).
5. Wash the column with six to eight column volumes of PBS and elute bound protein as a
preformed HRPObsMAb complex in 2-mL fractions with 0.1 M borate buffer, pH 9.0.
6. Check eluted fractions for bsMAb activity on an antigen-coated plate (see Note 10).
7. Pool the most active fractions and dialyze against PBS, pH 7.3.
3.1.6 Characterization of bsMAb Activity

1. Perform a two-step bridge ELISA with 96-well microtiter plates. This is based on the
activity of the bsMAb capable of bridging the two antigens.
2. Coat wells with biotin-BSA (1 g/well, in 100 L of PBS using a stock of 10 g/mL)
overnight at 4C.
3. Wash the plate three times with 200 L of PBST. Block free binding sites with 200 L of
1% BSA and incubate for 30 min to 1 h at 37C.
4. Wash the plate with PBST, add 100 L different dilutions of bsMAbHRPO complex in
PBS, and incubate for 1 h at 37C.
5. Wash thrice with PBST, incubate the plate with 100 L/well of TMB peroxidase sub-
strate.
6. Take the OD at 650 nm within 1530 min in a microplate reader to determine the titer of
the bsMAb activity.
7. Test purity of the affinity-purified material by SDS-PAGE and estimate protein concen-
tration by the Bradford method.
3.2. Bifunctional Single-Chain Variable-Fragment Antibody Development

3.2.1. Bifunctional ScFv Cloning by Recombinant DNA Technology
The initial cloning of specific variable immunoglobulin regions (VH and VL) is done
by PCR strategies. The Kabat database is a valuable source for design of primers and
sequence predictions (29). Designed PCR primers have been used to anneal many VH
and VL genes in segments encoding their FR1 and FR4 framework region sequences
(30). At the 5' end of VH and VL, one may use primers for FR1, whereas at the 3' end,
PCR primers for the constant region are appropriate. Following the elucidation of the
nucleotide sequence (by manual or automated sequencers), one can engineer appropriate
restriction sites in a second PCR reaction. The amplification of VH and VL with desired
restriction site(s) is followed by linking or assembly of VH and VL with a peptide linker
using an overlapping PCR procedure. Here we outline the design of 11D2-biotin mimic
bifunctional ScFv fragment, consisting of VL domain linked by a 15-amino-acid linker

to the VH domain for eventual expression. The protocol for the isolation and cloning of
bifunctional ScFv is a modification of a previously described method (31).
3.2.2. Cloning of ScFv
All DNA manipulations are performed according to standard cloning procedures (32).
3.2.2.1. TOTAL RNA ISOLATION AND CDNA SYNTHESIS
1. Centrifuge 11D2 hybridoma cells (56 106cells/mL) at 1500 rpm (403g) for 7 min at
20C and remove the supernatant.
2. Lyse hybridoma cells by adding 1 mL of TRizol reagent and pass the cell lysate several
times through a 1-mL pipet.
3. Incubate the lysed sample for 5 min at 1530C to allow complete dissociation of nucle-
oprotein complex.
4. Add 0.2 mL of chloroform, mix well by shaking vigorously, and incubate at 1530C for
23 min.
5. Centrifuge the sample at 12,000g for 15 min at 28C.
6. Remove the RNA-containing aqueous phase, avoiding interface material.
7. Add 0.5 mL isopropyl alcohol to the aqueous phase and incubate at 1530C for 10 min.
8. Pellet the RNA by centrifugation at 12,000g for 10 min at 28C.
9. Wash the RNA pellet with 75% ethanol and centrifuge at 7500g for 5 min at 28C (see
Note 11).
10. Briefly dry the RNA pellet (510 min) and dissolve in RNase free water by incubating at
55 to 60C for 10 min.
11. Read A260/A280 of 1:100 dilution of the RNA (see Note 12).
12. Aliquot and store RNA at 70C (see Note 13).
13. Synthesize the first strand of cDNA by mixing 2 g total RNA, 100 pmoles specific primers
(PL for light chain and PH for heavy chain), 200 units SuperscriptII (RNase H reverse
transcriptase) in a total volume of 20 L containing 50 mM Tris (pH 8.3), 75 mM KCl,
3 mM MgCl2, 10 mM DTT, 0.5 mM dNTPs, and 1 unit RNasin (Fig. 2A) (see Note 14).
14. Incubate the reaction at 42C for 1 h.
15. Generate VL or VH by mixing 1 L of cDNA reaction mixture, 0.5 M each primer
(P1 and P2 for VL, P3 and P4 for VH), 200 M dNTPs, and 1 unit vent DNA poly-
merase in 20 mM Tris (pH 8.8), 10 mM (NH4)2SO 4, 2 mM MgSO4, and 0.1% Triton X-
100 (Fig. 2B) (see Note 15).
16. Set the PCR cycle at 94C/55C/72C, 45 s/45 s/45 s for 25 cycles. At the beginning of
the first cycle, incubate for 3 min at 94C, and at the end of the last cycle, incubate for 7
min at 72C.
17. Analyze PCR products by 1% agarose/TAE gel electrophoresis prestained with EtBr
(0.25 g/mL) (see Note 16).
18. Purify amplified PCR products in 1% agarose/TAE gel by using QIAquick gel extrac-
tion kit.
3.2.2.2. ASSEMBLY OF VH AND VL GENES WITH LINKER

1. Perform assembly of ScFv by mixing equimolar amounts of VL, linker oligonucleotide
(P5), VH, 200 M dNTPs, 1 unit vent DNA polymerase, and 5 L 10X PCR buffer in a
50 L volume (Fig. 2C) (see Note 17).
2. Set the PCR at 94C/55C/72C, 45 s/45 s/45 s for 10 cycles.
336 Das and Suresh
Fig. 2. (A) First strand cDNA synthesis. (B) Amplification of VL and VH.
3.2.2.3. REENGINEERING OF SCFV TO SCFV-BIOTIN MIMIC TAG

1. Transfer a small aliquot of assembled ScFv (1 L) to a new PCR mixture containing 0.5
M of P1 and P6 primers, 200 M dNTPs, 1 unit vent DNA polymerase and 5 L 10X
PCR buffer in a 50-L volume (Fig. 2D) and continue amplification at 94C/60C/72C,
45 s/45 s/45 s for 25 more cycles. At the beginning of the first cycle, incubate for 3 min at
94C, and at the end of the last cycle, incubate for 7 min at 72C (33) (see Note 18).
Fig. 2. (C) Assembly of VL and VH. (D) Incorporation of Biotin Mimic tag.
2. Analyze the amplified product by 1% agarose/TAE gel and purify bifunctional ScFv band
from the gel by QIAquick gel extraction kit.
3. Digest the purified bifunctional ScFv fragment with BamHI and SalI.
4. Digest E. coli expression vector pET22b+ with BamHI and SalI, and purify from gel as
described above (see Note 19).
338 Das and Suresh
3.2.2.4. LIGATION AND TRANSFORMATION

1. Analyze both purified insert (bifunctional ScFv) and vector by 1% agarose/TAE gel.
2. Perform ligation of bifunctional ScFv with vector by mixing 2 L 10X ligase buffer, 50
ng vector DNA, 50 ng insert DNA, 1 unit T4 DNA ligase in a 20-L volume.
3. Incubate the ligation mixture at room temperature for 1 h or at 14C overnight.
4. Increase the ligation mixture volume up to 50 L with sterile water and inactivate T4
DNA ligase at 65C for 10 min.
5. Add 500 L n-butanol to precipitate the DNA and isolate by centrifugation at 12,000g for
10 min at room temperature. Redissolve precipitated DNA in 5 L sterile water.
6. Use 2 L of the redissolved ligation mixture to transform 40 L freshly prepared Top10F'
E. coli competent cells by electroporation (E. coli Pulser, Bio-Rad) (see Note 20).
7. Add 1 mL SOC medium and incubate at 37C in incubator shaker for 1 h.
8. Plate 100 L of the culture in LB plate containing 100 g/mL carbenicillin and incubate
in a 37C incubator for 16 h (overnight).
3.2.2.5. SCREENING OF RECOMBINANT CLONES

AND TRANSFORMATION IN E. COLI EXPRESSION HOST
1. Pick single colonies and grow overnight at 37C in 23 mL LB medium containing 100
g/mL carbenicillin.
2. Isolate DNA from individual recombinant clones by QIAprep plasmid DNA isolation kit.
Digest plasmid DNA by BamHI + SalI enzymes and analyze by 1% agarose/TAE gel elec-
trophoresis.
3. Select correctly oriented clones by restriction digestion fragment mapping for expression
studies.
4. Transform plasmid containing the 11D2-biotin mimic ScFv in E. coli BL21 (DE3) com-
petent cells by heat shock method (see Note 21).
5. Inoculate an overnight fresh colony of E. coli BL21 (DE3) transformants into 5 mL LB
medium containing 100 g/mL carbenicillin and shake at 250 rpm in a 37C incubator
until OD at 600 nm reaches 0.40.5 (see Note 22).
6. Induce bacterial culture by adding 1 mM IPTG and allow to grow at 30C for 3 h.
7. Harvest the bacterial culture by centrifugation at 4000g for 10 min at 4C and isolate total
cell protein by treating 1 mL bacterial pellet with sample buffer (50 mM Tris (pH 6.8),
100 mM DTT, 2% SDS, 0.1% bromophenol blue, 10% glycerol) at 95C for 5 min.
8. Analyze total cell protein by SDS-PAGE (Fig. 3A) using 10% polyacrylamide geland
proceed according to Laemmli (34) with a Mini Protean II apparatus (Bio-Rad).
9. Stain protein gels with Coomassie Brilliant Blue R-250 (0.25 g of Coomassie Brilliant Blue
R250 in 45 mL of methanol, 10 mL of acetic acid, and 45 mL of water) and destain with
10% acetic acid and 30% methanol. Visually look for the expected protein band at approx
32 kDa upon induction with IPTG and its corresponding absence in uninduced culture.
3.2.2.6. WESTERN BLOT ANALYSIS OF EXPRESSED 11D2-BIOTIN MIMIC SCFV

1. Electrophoretically separate total cell protein on SDS-PAGE using 10% polyacrylamide
gel and transfer onto an ECL Hybond nitrocellulose membrane (35) using Trans blot
apparatus (Bio-Rad) following manufacturers instructions.
2. Block the membrane with 5% skim milk in PBST for 2 h.
3. Wash the membrane four times with PBST and react with streptavidin-HRPO (1:1,000 in
1% dialyzed BSA) for 1 h.
Fig. 3. Total cell protein of induced and uninduced bacterial culture separtated on SDS-
PAGE and stained with Coomassie Brillant Blue (A). Western Blot probed with streptavidin-
HRPO (B). Lane 1: St. mol wt markers; lane 2: induced with IPTG; lane 3: uninduced.
4. Finally, wash the membrane four times with PBST and perform enhanced chemilumines-
cent (ECL)-based detection (Fig. 3B) according to manufacturers instructions (Amersham
Pharmacia, Piscataway, NJ).
3.2.2.7. SEQUENCE ANALYSIS

1. Conduct DNA sequencing analysis of the functional clone.
2. Compare the deduced DNA sequence with Kabat Database (see Website: http://immuno.
bmc.nwu.edu).
3.2.3. Protein Extraction
The steps described here in Subheadings 3.2.3.1. and 3.2.3.3. outline the stepwise
procedure for isolation of soluble protein and insoluble protein from 250 mL induced
E. coli culture.
3.2.3.1. SOLUBLE PROTEIN
1. Suspend the bacterial pellet in 5% of culture volume of periplasmic extraction buffer (50
mM Tris [pH 8.0], 20% sucrose, 1 mM EDTA [pH 8.0]) and incubate on ice for 45 min
with occasional shaking.
2. Centrifuge the suspension at 11,000g for 15 min at 4C and collect supernatant as peri-
plasmic soluble protein.
3. Resuspend the pellet in 5 mL of lysis buffer, 0.1 mg/mL lysozyme, and incubate for 30
min on ice with occasional shaking.
4. Homogenize the suspension by sonication (6 10 s) and centrifuge at 11,000g for 15 min
at 4C. Collect the supernatant as the cytoplasmic soluble protein and the pellets as the
insoluble fraction.
5. Use the insoluble fraction to isolate inclusion bodies of the expressed protein.
6. Analyze the cytoplasmic and periplasmic soluble proteins by SDS-PAGE.
7. Perform Western blot as described earlier in Subheading 3.2.2.6.
340 Das and Suresh
3.2.3.2. ISOLATION OF INCLUSION BODIES

1. Resuspend the insoluble fraction in 35 mL lysis buffer, incubate at room temperature for
15 min, and repeat sonication (6 10 s) (see Note 23).
2. Adjust the sample to 2% sodium deoxycholate and mix well by inversion.
3. Incubate the sample at room temperature for 30 min and centrifuge at 11,000g for 30 min
at 4C.
4. Remove the supernatant and further wash pellet twice at 11,000g for 20 min at 4C with
3 mL lysis buffer.
5. Estimate the purity of the inclusion bodies by SDS-PAGE using 10% polyacrylamide gel
under reducing conditions.
3.2.4. Protein Refolding

Here we describe different steps of how to perform proper refolding that yields
significant amounts of soluble functional protein (see Note 24).
3.2.4.1. REFOLDING OF THE 11D2-BIOTIN MIMIC SCFV
1. Solubilize inclusion bodies by suspending in 100 mL 2% sarcosyl, 50 mM Tris (pH 10.0),
and stirring at room temperature for 2 h.
2. Perform air oxidation by continued stirring of the sample in the presence of 50 M CuSO4
for 20 h at room temperature.
3. Add solid urea to the sample to make a final concentration of 6 M. Remove sarcosyl from
the sample by absorption twice with 10% Dowex 1X8. Prepare Dowex 1X8 200400
mesh (Bio-Rad) by sequential washings in 1 M NaOH, water, 4 M acetic acid, water,
50 mM Tris (pH 10.0), and equilibrate in 6 M urea, 50 mM Tris (pH 8.0) (36).
4. Upon removal of the sarcosyl, determine the protein concentration by the Bradford assay
(37) (see Note 25).
5. Adjust solubilized denatured protein to 100 g/mL in refolding buffer and dialyze over a
period of 3 d with three changes.
6. Finally, dialyze the sample against 50 mM Tris-HCl (pH 8.0) for 48 h at 4C.
7. Upon completion of dialysis, remove any aggregates by centrifugation at 8000g for 10
min at 4C and determine protein concentration by Bradford method.
3.2.5. Characterization
The refolded and purified ScFv can be tested for its activity toward the WEE anti-
gen by ELISA.
3.2.5.1. ELISA
1. Direct solid phase ELISA is performed by coating WEE antigen on a Nunc Maxisorb
ELISA plate.
2. Coat wells at 4C overnight with inactivated WEE antigen (38) (400 ng/well) or 1% dia-
lyzed BSA as negative control in PBS.
3. Wash the plate three times with 200 L of PBST and block with 200 L of 1% BSA for
2 h at room temperature.
4. Subsequently wash the plate as described above and react with different concentrations of
refolded 11D2-biotin mimic ScFv or irrelevant antibody as negative control (100 L/
well) for 2 h at room temperature.
5. Wash the plate and incubate 1 h with St-HRPO (1:1000 in 1% dialyzed BSA, 100 L/
well) or GAM-HRPO (1:10,000 in 1% dialyzed BSA, 100 L/well) for negative control
(see Note 26).
6. Following a final wash with PBST, develop the color by adding 100 L of TMB peroxi-
dase substrate to each well and read OD at 650 nm within 2030 min using an ELISA
plate reader.
3.2.6. Factors Influencing In Vitro Refolding

The mechanism of refolding of antibodies is similar to that of other proteins con-
taining disulfide bonds (3941). The disulphide bond formation is promoted while
minimizing aggregate formation. Urea and guanidine-HCl are the most commonly
used denaturants to solubilize the inclusion bodies. However, it has been reported
that treatment with mild detergent can also be used to solubilize inclusion body pro-
tein (36,42). In addition, reducing agents such as -mercaptoethanol or dithiothreitol
(DTT) are used to reduce inter- or intra-chain disulphide bonds that might have
formed during lysis of the bacteria and during solubilization of the inclusion bodies.
In some cases, promotion of the correct disulfide bond formation is enhanced by the
presence of metal ions or inclusion of redox coupling reagents, made of a mixture of
reduced and oxidized thiol groups (43). It has been recognized that protein folding,
formation of disulphide bonds, and proper association of different domains are very
much dependent on folding conditions (44,45). Every protein species is unique in
nature, and a refolding protocol for each needs to be optimized. There is no universal
renaturation protocol that can be used for all recombinant proteins (46).
4. Notes
1. Prepare 50 mL of SFDMEM medium with two different pH values of 6.8 and 7.4. Filter
sterilize the medium and resuspend respective cell lines in two different media. pH of the
medium is important for proper labeling by the fluorescent dye.
2. This cell-labeling protocol was similar to a previously described method (25), with some
modifications.
3. Remove medium by aspiration and mix gently to resuspend the cells.
4. This is a modified fusion protocol using PEG. Instead of electrofusion, we optimized the
condition for generation of quadroma using PEG fusion, which is less time consuming
and more efficient.
5. Do not spin the fused cells more than 800 rpm (114g), as cells are loosely attached to each
other. Subsequent washing steps are also performed at 800 rpm (114g).
6. Keep proper controls (unlabeled, singly labeled cells) to select the dual fluorescence cells.
7. Testing neat and diluted supernatant is essential to mitigate inhibition by monospecific
antibodies. It is very common to see higher ELISA activity in the diluted vs neat superna-
tants.
8. Quadromas are fused products of two hybridomas. The high ploidy levels of these cells
require recloning more times to isolate a good cell line that is stable.
9. HRPO with a heme cavity in its substrate binding site has affinity for BHA. If a specific
affinity ligand is not readily available, an immunopurification strategy using sepharose
linked with specific heavy- or light-chain reactive antibodies (Sigma) can be used to pu-
rify only the bispecific molecules.
342 Das and Suresh
10. Purification of bsMAb-HRPO complex by BHA agarose will also co-purify monospe-
cific antibodies directed towards HRPO. However, the co-purified contaminants and free
excess HRPO should be of little concern, as they will not bind to the target antigen and
will be eliminated in any subsequent wash steps.
11. All mRNA isolation reagents were prepared with 0.1% DEPC-treated autoclaved water.
12. A260 of 1 is equivalent to 40 g/mL of RNA. Good quality RNA yields A260/A280 of 1.82.0.
13. RNA can be stored at 70C in DEPC-treated water.
14. Instead of total RNA, mRNA can be used for first strand cDNA synthesis. Poly A+-
containing transcripts can be purified from total RNA by oligo (dT) beads, which are
commercially available.
15. It is not necessary to precipitate cDNA. Samples from the RT reaction mixture can be
used directly as template for PCR. Degenerate primers for VH and VL can be generated
according to (4751).
16. EtBr is carcinogenic. Wear gloves and take precautions during agarose gel electrophoresis.
17. Measure A260 of the PCR-purified DNA. A260 of 1 is equivalent to 50 g/mL of DNA.
Linker oligo is designed in such a way that there are 24-base overlaps both at the 3' end of
VL and the 5' end of VH.
18. Biotin mimic oligo is synthesized according to (33) and cloned at the 3' end of ScFv by PCR.
19. Digest isolated PCR products and vector DNA with restriction endonucleases suitable
for cloning.
20. Electroporation was done in a 0.1-cm cuvette at 180 kV.
21. Preparation of chemical-competent cells of E. coli BL21 (DE3) and transformation were
done according to (32).
22. It has been observed that the pET22b+ expression vector in E. coli BL21 (DE3) cells can
be unstable. Therefore, instead of expressing the ScFv from a frozen glycerol stock of
transformed E. coli BL21 (DE3) cells, transformation of the E. coli BL21 (DE3) cells may
have to be performed by the heat-shock method just prior to scaling up for expression.
23. Alternative method to isolate the inclusion bodies from E. coli is the use of a French Press.
24. In our protocol, urea was used to denature the inclusion bodies, but 46 M guanidine
hydrochloride can also be used. We have discussed different issues in Subheading 3.3.5.,
which should be considered during refolding of inclusion bodies.
25. For protein estimation, use the Bradford dye-binding assay, because it is very fast, sensi-
tive, and easy to use.
26. Ensure that stock streptavidin-HRPO is diluted in extensively dialyzed BSA to remove
traces of free biotin, which dampens the signal.
Acknowledgments
M. R. S. thanks the CIHR-Industry (Biomira Inc.) Chair award for salary support and
CBDN and NSERC-DND for grant support. YP4 was a gift from C. Milstein, MRC
Laboratory of Molecular Biology, Cambridge, UK. Thanks are also due to Dr. Fred
Jacobs for critical review of the manuscript. M.R.S. would like to dedicate this article to
Prof. A. A. Novjaim, who has been an inspiration and mentor for more than 20 years.
References
1. White., C. A., Weaver, R. L., and Grillo-Lopez, A. J. (2001) Antibody-targeted immuno-
therapy for treament of malignancy. Annu. Rev. Med. 52, 125145.
2. Kohler, G. and Milstein, C. (1975) Continuous cultures of fused secreting antibody of
redefined specificity. Nature 256, 495497.
3. Winter, G. and Milstein, C. (1991) Man-made antibodies. Nature 349, 293299.

4. Pluckthun, A. and Pack, P. (1997) New protein engineering approaches to multivalent and
bispecific antibody fragments. Immunotechnology 3, 83105.
5. Bird, R. E., Hardman, K. D., Jacobson, J. W., et al. (1988) Single-chain antigen-binding
proteins. Science 242, 423426.
6. Huston, J.S., Levinson, D., Mudgett-Hunter, M., et al. (1988) Protein engineering of
antibody binding sites: recovery of specific activity in an anti-digoxin single-chain Fv
analogue produced in Escherichia coli. Proc. Natl. Acad. Sci. USA 85, 58795883.
7. Nisonoff, A. and Rivers, M. M. (1961) Recombination of a mixture of univalent antibody
fragments of different specificity. Arch. Biochem. Biophys. 93, 460462.
8. Kriangkum, J., Xu, B., Nagata, L. P., Fulton, R. E., and Suresh, M. R. (2001) Bispecific
and bifunctional single chain recombinant antibodies. Biomol. Eng. 18(2), 3140.
9. Suresh, M. R., Cuello, C., and Milstein, C. (1986) Advantages of bispecific hybridomas
in one-step immunocytochemistry and immunoassays. Proc. Natl. Acad. Sci. USA 83,
79897993.
10. Cao, Y., Christian, S., and Suresh, M. R. (1998) Development of bsMAb anti-biotin x
anti-HRPO as a universal immunoprobe for detecting biotinylated macromolecules. J.
11. Van Ojik, H. H. and Valerius, T. (2001) Preclinical and clinical data with bispecific anti-
bodies recruiting myeloid effector cells for tumor therapy. Crit. Rev.Oncol. Hematol. 38,
4761.
12. Withoff, S., Helfrich, W., de Leij, L. F. M. H., and Molema, G. (2001) Bi-specific anti-
body therapy for the treatment of cancer. Curr. Opin. Mol. Ther. 3, 5362.
13. Talac, R. and Nelson, H. (2000) Current perspectives of bispecific antibody-based immu-
notherapy. J. Biol. Regul. Homeost. Agents 14, 175181 .
14. van Spriel, A. B., van Ojik, H. H., and van De Winkel, J. G. (2000) Immunotherapeutic
perspective for bispecific antibodies. Immunol. Today 21, 391397.
15. Kroesen, B. J., Helfrich, W., Molema, G., and de Leij, L. (1998) Bispecific antibodies for
treatment for of cancer in experimental animal models and man. Adv. Drug. Deliv. Rev.
31, 105129.
16. Wang, H., Liu, Y., Wei, L., and Guo, Y. (2000) Bi-specific antibodies in cancer therapy.
Adv. Exp. Med. Biol. 465, 369380.
17. Koelemij, R., Kuppen, P. J., van de Velde, C. J., Fleuren, G. J., Hagenaars, M., and
Eggermont, A. M. (1999) Bispecific antibodies in cancer therapy, from the laboratory to
the clinic. J. Immunotherapy 22, 514524.
18. Segal, D. M., Weiner, G. J., and Weiner, L. M. (1999) Bispecific antibodies in cancer
therapy. Curr. Opin. Immunol. 11, 558562.
19. Cao, Y. and Lam, L. (2003) Bispecific antibody conjugates in therapeutics. Adv. Drug.
Deliv. Reviews55, 171197.
20. Cao, Y. and Suresh, M. R. (1998) Bispecific antibodies as novel bioconjugates. Biocon-
jugate Chemistry 9 (8), 635644.
21. Milstein, C. and Cuello, A. C. (1983) Hybrid hybridomas and their use in immunohis-
tochemistry. Nature 305, 537540.
22. Nolan, O. and Kennedy, O. R. (1990) Bifunctional antibodies: concept, production and
applications. Biochem. Biophys. Acta 1040, 111.
23. Karawajew, L., Behrsing, O., Kaiser, G., and Micheel, B. (1988) Production and ELISA
application of bispecific monoclonal antibodies against fluorescein isothiocyanate (FITC)
and horseradish peroxidase (HRP). J. Immunol Methods 111, 9599.
344 Das and Suresh
24. Karawajew, L., Micheel, B., Behrsing, O., and Gaestel, M. (1987) Bispecific antibody-
producing hybrid bybridomas selected by a fluorescence activated cell sorter. J. Immunol
Methods 96, 265270.
25. Kreutz, F. T., Xu, D. Z., and Suresh, M. R. (1998) A new method to generate quadromas
by electrofusion and FACS sorting. Hybridoma 17, 267273.
26. Gupta, S. and Suresh, M. R. (2002) Affinity chromatography and co-chromatography of
bispecific monoclonal antibody immunoconjugates. J. Biochem. Biophys. Methods 51,
203216 .
27. Husereau, D. R. and Suresh, M. R. (2001) A general affinity method to purify peroxidase-
tagged antibodies. J. Immunol. Methods 249 (12), 3341.
28. de Ropp, J. S., Mandal, P. K., and La Mar, G. N. (1999) Solution 1H NMR investigation of
the heme cavity and substrate binding site in cyanide-inhibited horseradish peroxidase.
Biochemistry 38, 10771086.
29. Kabat, E. A., Wu, T. T., Perry, H. M., Gottesman, K. S., and Foeller, C. (1991) Sequences
of Proteins of Immunological Interest, 5th Ed., U.S. Department Health and Human Ser-
vices, Public Health Service, National Institute of Health, Publication No. 81-3242.
30. Orlandi, R., Gussow, D. H., Jones, P. T., and Winter, G. (1989) Cloning immunoglobulin
variable domains for expression by the polymerase chain reaction. Proc. Natl. Acad. Sci.
USA 86, 38333837.
31. Xu, B., Kriangkum, J., Nagata, L. P., Fulton, R. E., and Suresh, M. R. (1999) A single
chain Fv specific against western equine encephalitis virus. Hybridoma 18(4), 315323.
32. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory
Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
33. Luo, D., Geng, M., Schultes, B., et al. (1998) Expression of a fusion protein of scFv-biotin
mimetic peptide for immunoassay. J. Biotechnol. 65, 225228.
34. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of
bacteriophage T4. Nature 277, 680685.
35. Towbin, H., Staehelin, T., and Gordon, J. (1979) Electrophoretic transfer of proteins from
polyacrylamide gels to nitrocellulose sheets. Proc. Natl. Acad. Sci. USA 76, 43504354.
36. Kurucz, I., Titus, J. A., Jost C. R., and Segal. D. M. (1995) Correct disulfide pairing and
efficient refolding of detergent-solublilized single chain Fv proteins from bacterial inclu-
sion bodies. Molecular Immunology 32 (17/18), 14431452.
37. Bradford, M. M. (1976) A rapid and sensitive method for the quantitation of microgram
quantities of protein utilizing the principle of dye binding. Anal. Biochem. 72, 248254.
38. Long, M. C., Jager, S., Mah, D. C. W., et al. (2000) Construction and characterization of
a novel recombinant single-chain variable fragment antibody against western equine
encephalitis virus. Hybridoma 19(1), 113.
39. Rudolph, R. (1990) Renaturation of recombinant disulfide -bonded protein from inclu-
sion bodies. In: Modern Methods in Protein and Nucleic Acid Research. Tschesche, H.,
ed., Walter de Druyter, Berlin, Germany, pp. 149171.
40. Buchner, J. and Rudolph, R. (1991) Renaturation, purification and characterization of re-
combinant Fab-fragments produced in Escherichia coli. Biotechnology 9(2), 157162.
41. Buchne, J., Pastan, I., and Brinkmann, U. (1992) A method for increasing the yield of
properly folded recombinant fusion proteins: single-chain immunotoxins from renatur-
ation of bacterial inclusion bodies. Anal. Biochem. 205(2), 263270.
42. Lacks, S. A. and Springhorn, S. S. (1980) Renaturation of enzymes after polyacrylamide
gel electrophoresis in the presence of sodium dodecyl sulphate. J. Biol. Chem. 255,
74677473.
43. Saxena, V. P. and Wetlaufer, D. B. (1970) Formation of three-dimensional structure in

proteins. I. Rapid nonenzyme reactivation of reduced lysozyme. Biochemistry 9(25),
50155023.
44. Suttnar, J., Dyr, J. E., Hamsikova, E., Novak, J., and Vonka, V. (1994) Procedure for
refolding and purification of recombinant proteins from Eschrichia coli inclusion bodies
using a strong anion exchanger. J. Chromatogr. B 656, 123126.
45. Wei, C., Tang, B., Zhang, Y., and Yang, K. (1999) Oxidative refolding of recombinant
prochymosin. Biochem. J. 340, 345351.
46. Verma, R., Boleti, E., and George, A. J. T. (1998) Antibody engineering: Comparison of
bacterial, yeast, insect and mammalian expression system. J. Immunol. Methods 216,
165181.
47. Clark, A.M. (2002) Standard protocols for the construction of Fab libraries. In: Methods in
Molecular Biology, vol. 178: Antibody Phage Display: Methods and Protocols. OBrien,
P. M., and Aitken, R., eds. Humana, Totowa, NJ, pp. 3958.
48. Dziegiel , M., Nielsen, L, K., Adersen, P. S., Blancher, A., Dickmeiss, E., and Engber, J.
(1995) Phage display used for gene cloning of human recombinant antibody against the
erythrocyte surface antigen, rhesus D. J. Immunol. Methods 182(1), 719.
49. Marks, J. D., Tristem, M., Karpas, A., and Winter, G. (1991) Oligonucleotide primers for
polymerase chain reaction amplification of human immunoglobulin variable genes and
design of family-specific oligonucleotide probes. Eur. J. Immunol. 21(4), 985991.
50. Kettleborough, C. A., Saldanha, J., Ansell, K. H., and Bendig, M. M. (1993) Optimization
of primers for cloning libraries of mouse immunoglobulin genes using the polymerase
chain reaction. Eur. J. Immunol. 23(1), 206211.
51. Burton, D. R. and Barbas, C. F. (1994) Human antibodies from combinatorial libraries.
Adv. Immunol. 57, 191280.
AntibodyCytokine Fusion Proteins 347
23
AntibodyCytokine Fusion Proteins for the Therapy of Cancer
Gustavo Helguera and Manuel L. Penichet
Summary
In recent years the development of tumor-specific recombinant antibodies fused to immuno-
stimulatory cytokines such as interleukin-2 (IL-2), interleukin-12 (IL-12), and granulocyte/mac-
rophage colony-stimulating factor (GM-CSF) has provided a promising novel approach to cancer
immunotherapy. The combined properties of specific targeting of antibodies and the
immune stimulation of cytokines results in high cytokine concentration in the tumor microenvi-
ronment, and as a consequence, in an improved tumoricidal activity of the antibody and/or in a
secondary effective immune response against the tumor. In the present chapter we describe strat-
egies for the construction, expression, and in vitro characterization of antibodycytokine fusion
proteins, with particular emphasis on antibody/IL-2 fusion proteins.
Key Words: Antibody engineering; antibody therapy; antibody; cancer; constant region;
cytokine; expression vector fusion protein; IgG3; HER2/neu; immunocytokine; immunoglobu-
lin; immunoligand; immunotherapy; interleukin-2; tumor targeting; variable region.
1. Introduction
Progress in genetic engineering and expression systems has led to the development
of tumor-specific antibodies fused to immunostimulators, such as the cytokines inter-
leukin (IL)-2, IL-12, and granulocyte/macrophage colony-stimulating factor (GM-CSF)
(1,2). The goal of using these tumor-specific antibodies is to concentrate the cytokines
in the tumor microenvironment and enhance the tumoricidal activity of the antibody
and/or the host immune response against the tumor while limiting the severe toxicity
associated with high doses of cytokines. These antibodycytokine fusion proteins, also
called immunocytokines or immunoligands (1,3), have shown significant antitumor
activity in mice bearing tumors leading to clinical trials (411).
Antibodies or immunoglubulins (Igs) are hetero-tetramers comprised of two identical
light chain (L chain) and heavy chain (H chain) polypeptides held together by disulfide
bonds (H2L2) (Fig. 1). The antigen-binding variable region (V-region) is formed by the
amino-terminal domains of both the H (VH) and L chain (VL). The other domains of the
H and L chains show much less variability and constitute the constant region (C-region).
347
348 Helguera and Penichet
Fig. 1. Schema of an immunoglobulin G (IgG) molecule and the genes that encode its H and
L chains. The genes encoding for IgG are represented on the top of the figure. Boxes represent
exons encoding for discrete domains. The central line represent introns that are present in the
primary transcript but not in the mature mRNA. Both H and L chains contain hydrophobic
leader sequences (indicated by the black exon) necessary for their processing and secretion.
The H chain is divided into four discrete functional domains (VH, CH1, CH2, and CH3), while
the L chain consists of two domains (VL and CL). The V region domains form the Fv region,
which is responsible for antigen binding. The C-region domains are responsible of the effector
functions of the antibody. The CH2 domains present glycosylation sites (dark circles) that con-
tribute to the functional properties of the antibody. The hinge region provides flexibility to the
molecule. The fragments product of the papain enzyme cleavage are also indicated: two Fab
(responsible for antigen binding) and one Fc (responsible of the effector functions).
Variations in the H chain C region characterize different isotypes and define five classes
of human antibodies: IgA ( chain), IgD ( chain), IgG ( chain), IgE ( chain), and IgM
( chain). In humans, IgG is the most abundant immunoglobulin present in serum, with
four subclasses: IgG1, IgG2, and IgG4, exhibiting a mol wt of approx 150 kDa, and
IgG3, exhibiting a mol wt of approx 165 kDa (12). Differences in the CL region result in
two L-chain isotypes, and . The domains after the hinge of the C region form the Fc
region, which is responsible for the antibody-related effector functions such as antibody-
dependent cellular cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC)
(4,5,13,14). The Fc region also binds the FcRn (Brambell receptor) that results in an
extended half-life in serum (15,16). The domain structure of antibodies facilitates pro-
tein engineering, allowing the separation of their different functional regions and the
genetic fusion of nonantibody molecules in different positions of the antibody molecule
(Fig. 2). The nonantibody partner can be fused at the carboxy-terminus (C-terminus)
Fig. 2. Construction of antibody fusion proteins with the nonantibody partner or ligand
fused to different regions of the H chain. Panels A, B, and C show C-terminus fusion proteins
with the nonantibody partner (represented as a black circle) fused after the CH3 domain (A),
after the hinge (B), or after the CH1 domain (C). Panels D, E, and F denote antibody fusion
proteins with the ligand fused to the N-terminus of the complete H chain (D), the CH1 domain
(E), or immediately before the hinge (F). Panels G and H show scFv fragments with the ligand
fused to the C-terminus (G) or N-terminus (H).
(Fig. 2A,B,C) or at the amino-terminus (N-terminus) of the H chain (Fig. 2D,E,F) or the
L chain (not shown). In some cases, the nonantibody partner can be fused via a flexible
linker, depending on the specific requirements of conformation, processing, or accessi-
bility to the antigen or the cytokine receptor (1721). When the cytokine is fused to
an intact antibody (Fig. 2A,D), the antibodycytokine fusion protein can deliver both
the biological activity associated with the genetically fused cytokine, and the antibody
effector functions. Fusion of the ligand immediately after the hinge (H-ligand) (Fig. 2B)
or to the CH1 domain (CH1-ligand) (Fig. 2C) results in a loss of antibody effector func-
tion. This may be useful when the Fc effector functions are unnecessary or harmful.
In addition, for many applications such as targeting well-established solid tumors, the
small size of H-ligand and CH1-ligand may be an advantage over the larger CH3-ligand
(3). The smallest antibody fragments, known as scFv fragments, are composed of the
variable domains of H and L chains fused by a synthetic flexible linker peptide. Cyto-
kines can also be fused to both the C- and N-terminus of scFv (Fig. 2G,H) (20,21).
Alternatively, cytokines can replace the VH domain or the VH-CH1 domains (Fig. 2E,F)
of the antibody. Such molecules lose the ability to bind antigen but retain the antibody
Fc effector properties and possess improved pharmacokinetics (22,23).
Multiple expression systems have been used for antibody production in bacteria,
yeast, plants, baculovirus, insect cells, and mammalian cells (2426). Each system has
advantages and shortcomings, depending on the characteristics and expected use of
the protein. Bacteria and yeast have been successfully used to express scFv antibody
fragments, and complete functional antibodies have been expressed in plant and insect
cells (24,25,27). However, mammalian cells have been the most effective in express-
ing full-length, functional recombinant antibodies and antibody fusion proteins, since
these cells possess the machinery necessary for the correct assembly, post-transla-
tional modification, and secretion of these proteins (26).
In recent years, the field of antibodycytokine fusion proteins has expanded to
include a variety of cytokines fused to antibodies of different structures (1,2), making
it virtually impossible to cover protocols about the construction, expression, and char-
acterization of every antibody-cytokine fusion protein. Among this family of recombi-
nant proteins, antibody/IL-2 fusion proteins have been the best characterized and most
broadly used in successful antitumor experiments. In this chapter, we describe the
methods for the construction, expression, and in vitro characterization of antibody
fusion proteins, using as an example a well-studied human IgG3/IL-2 fusion protein
developed in our laboratory (28) specific for the human tumor-associated antigen
HER2/neu. However, it is important to stress that similar procedures can be applied to
many combinations of antibodies and cytokines.
2. Materials
2.1. Construction of AntibodyCytokine Fusion Proteins
2.1.1. Cloning of the V Region Sequence
into the Antibody Expression Vectors
1. Mammalian expression vectors for human L- and H-chain C regions (pAG4622 and
pAH4802) (34).
2. Primers 5' and 3' to amplify the DNA encoding for H- and L-chain V regions.
3. PCR equipment (PTC 100 Thermal Controller; M. J. Research, Watertown, MA).
4. Taq-DNA polymerase and reaction buffers (New England Biolabs, Beverly, MA).
5. TA cloning vector (Invitrogen, Carlsbad, CA).
6. Restriction endonucleases: EcoRV, NheI, SalI, and reaction buffers (New England Biolabs).
7. T4 DNA ligase and reaction buffers (New England Biolabs).
8. Additional equipment and reagents for agarose gel electrophoresis, transforming E. coli,
minipreps, restriction endonuclease digestion, and purification of DNA restriction frag-
ments.
2.1.2. Construction of the AntibodyCytokine Fusion Protein

1. Jurkat cells (American Type Culture Collection [ATCC], Rockville, MD).
2. Primers 5' and 3' to amplify the cDNA encoding for mature form of human IL-2.
3. Avian myeloblastosis virus (AMV) reverse transcriptase and reaction buffers (New
England Biolabs).
4. Mammalian expression vectors for human L chain and H chain with desired V regions.
5. TA cloning vector (Invitrogen).
6. Intermediate cloning vector (pAT3462) (29).
7. Restriction endonucleases: EcoRV, NheI, SalI, BamHI, SspI, NsiI, EcoRI, PmlI, and re-
action buffers (New England Biolabs).
8. T4 DNA ligase and reaction buffers (New England Biolabs).
2.2. Transfection into Expression Systems and Initial Characterization

2.2.1. Transfection into Mammalian Expression System
1. Mammalian expression vectors encoding for human L and H chains fused to the cytokine.
2. PvuI restriction endonuclease and reaction buffers (New England Biolabs).
3. Recipients of transfection: murine myeloma cell lines such as Sp2/0-Ag14, P3X63
Ag8.653, or NS0/1 (ATCC).
4. Phosphate-buffered saline (PBS): 154 mM NaCl, 1.9 mM NaH2PO4, 8.1 mM Na2HPO4,
pH 7.2.
5. Bio-Rad Gene Pulser Electroporator and 0.4-cm gap electroporation cuvet (Bio-Rad
Laboratories, Hercules, CA).
6. Iscoves modified Dulbeccos medium (IMDM) (Irvine Scientific, Irvine, CA), supple-
mented with 2 mM L-glutamine, 10 U/mL penicillin, and 10 g/mL streptomycin (Sigma
Chemical, St. Louis, MO).
7. Growth medium: 10% calf serum (Atlanta Biologicals, Norcross, GA), 0.1 mg/mL
gentamycin, 100 U/mL nystatin suspension (Sigma Chemical) in IMDM.
8. Selection medium: 10% calf serum, selection marker (histidinol or HXM) in IMDM.
9. One M histidinol in 0.1 M HEPES, pH 7.4 (Sigma Chemical).
10. HXM stock 30X: 3.75 mg/mL hypoxanthine, 225 g/mL xanthine, and 90 g/mL
mycophenolic acid (Sigma Chemical).
11. Ninety-six-well tissue-culture flat-bottom plates (Falcon; Becton-Dickinson, Lincoln
Park, NJ).
2.2.2. Enzyme-Linked Immunosorbent Assay (ELISA) Screening

1. Ninety-six-well flat-bottom microtiter plates (Immunolon II; Costar, Corning, NY).
2. Carbonate coating buffer: 0.015 M Na2CO3, 0.035 M NaHCO3, pH 9.3.
3. Blocking buffer: 3% w/v BSA, 0.02% w/v sodium azide, in PBS.
4. Goat antihuman IgG (Zymed Laboratories, San Francisco, CA).
5. Goat antihuman antibody (free & bound) (Sigma Chemical).
6. Alkaline phosphatase (AP) goat antihuman L chain (Sigma Chemical).
7. Diethanolamine buffer: 9.6% v/v diethanolamine, 0.24 mM MgCl2 (Sigma Chemical),
pH 9.8.
8. AP substrate: 1 pellet p-nitrophenyl phosphate disodium (Sigma Chemical) dissolved in
5 mL diethanolamine buffer.
9. ELISA plate reader (MR700 Plate Reader; Dynatech Labs, Chantilly, VA).
10. Twenty-four-well tissue-culture plate (Falcon; Becton-Dickinson).
2.2.3. 35S Biosynthetic Antibody Labeling and Immunoprecipitation

1. Petri dishes 60 15-mm (Falcon; Becton-Dickinson).
2. Twenty-four-well tissue-culture plate (Falcon; Becton-Dickinson).
3. Deficient Dulbeccos Modified Eagle Medium (DMEM) in L-glutamine, methionine, and
cysteine (Cellgro; Mediatech, Herndon, VA), with 1:100 GlutaMAX (Gibco-BRL, Life
Technologies, Rockville, MD).
4. Labeling medium: 0.1% (v/v) calf serum and EasyTag Labeling Mix composed of
[35S]-methionine + [35S]-cysteine (NEN Life Science Products, Boston, MA) in defi-
cient DMEM.
5. NDET: 1% v/v NP 40, 0.4% (w/v) deoxycholate, 66 mM EDTA, 10 mM Tris-HCl,
pH 7.4.
6. 0.3% (w/v) Sodium dodecyl sulfate (SDS) in NDET.
7. Polyclonal antihuman IgG antibody, e.g., rabbit hyperimmune antihuman IgG (30).
8. Sucrose pad: 30% (w/v) sucrose, 0.3% SDS in NDET.
9. Staph A: 10% (w/v) suspension of formaldehyde-fixed Staphylococcus aureus with mem-
brane-bound protein A (IgGSorb, The Enzyme Center, Malden, MA), with 0.3% SDS in
NDET.
10. Distilled H2O.
11. SDS-PAGE sample buffer: 2% (w/v) SDS, 10% (w/v) glycerol, 0.008% (w/v) bromophe-
nol blue, 25 mM Tris-HCl, pH 6.7.
12. -mercaptoethanol.
13. Rainbow [14C]-methylated protein mol wt marker (Pharmacia, UK).
14. J774.2 supernatant: sterile filtered supernatant of J774.2 cells (ATCC) grown in log phase
with 10% v/v calf serum in IMDM.
15. Subcloning medium: 20% v/v calf serum and 10% v/v J774.2 supernatant in IMDM.
16. Ninety-six-well tissue-culture flat-bottom plates (Falcon; Becton-Dickinson).
2.2.4. Growing Cells in Roller Bottles

1. Petri dishes 100 20-mm (Falcon; Becton-Dickinson).
2. Fetalclone serum (HyClone, Logan, UT).
3. Tissue-culture roller bottle, 2 L (Falcon; Becton-Dickinson).
4. Cell production roller apparatus (Bellco Biotechnology, Vinland, NJ).
5. GlutaMAX-1 (Gibco, Invitrogen, France).
6. Whatman filter paper 1 (Whatman International, Maidstone, UK).
2.2.5. Purification, Quantification, and Storage

1. Protein A, recombinant; immobilized on 6% fast-flow beaded agarose (Sigma Chemical).
2. LKB IBF 11 mini-columns (Sepracor, Marlborough, MA).
3. Sodium azide in PBS, 0.02% (w/v).
4. Elution buffer A: 0.1 M citric acid-NaOH, pH 4.5. Elution buffer B: 0.1 M glycine-HCl,
pH 2.5. Elution buffer C: 0.1 M glycine-HCl, pH 2.0.
5. Neutralization buffer: 2 M Tris-HCl, pH 8.0.
6. Centricon centrifugal filter concentrator (Millipore, Billerica, MA).
7. Dialysis buffer: 150 mM NaCl, 50 mM Tris-HCl, pH 7.8.
8. Dialysis tubing (Sigma Chemical).
9. Bicinchoninic acid-based protein assay (BCA protein assay; Pierce Chemical, Rock-
ford, IL).
2.3. Further In Vitro Characterization of the Antibody Fusion Protein

2.3.1. Antigen Binding Studies
2.3.1.1. ELISA TO TEST ANTIBODYCYTOKINE FUSION PROTEIN ANTIGEN BINDING
1. Purified antigen.
2. Same materials described in Subheading 2.2.2.
2.3.1.2. FLOW CYTOMETRY

1. Tumor cell line expressing antigen on its surface.
2. Flow cytometry buffer: 2% (v/v) calf serum, 0.02% (w/v) sodium azide in PBS.
3. Biotinylated goat antihuman IgG (Pharmingen, San Diego, CA).
4. Phycoerythrin (PE)-labeled streptavidin (Pharmingen).
5. Flow cytometer equipped with a blue laser excitation of 15 mW at 488 nm (FACScan;
Becton-Dickinson, Mountain View, CA).
2.3.2. Study of the Cytokine Biological Activity

1. Ninety-six-well tissue-culture flat-bottom plates (Falcon, Becton-Dickinson).
2. CTLL-2 cell line (ATCC).
3. Recombinant human interleukin-2 (rhIL-2) (PeproTech, Rocky Hill, NJ).
4. [Methyl-3H]-thymidine (ICN, Costa Mesa, CA).
5. Complete RPMI 10: RPMI 1640 (Gibco-BRL, Life Technologies), 10% (v/v) calf serum,
2 mM L-glutamine, 50 M -mercaptoethanol, 100 U/mL penicillin, and 100 g/mL strep-
tomycin sulfate.
6. Glass fiber filter 90 120-mm printed Filtermat (Wallac Oy, Turku, Finland).
7. Ninety-six-well-plate cell harvester (Tomtec, Hamden, CT).
8. EcoLume scintilliation medium (Amersham, Arlington Heights, IL).
9. Liquid scintillation counter (Microbeta 1450; Wallac Oy).
3. Methods
In this section we describe: (a) the construction of plasmids for expression of a hu-
man antibody fused to a cytokine; (b) transfection of the expression vector into myeloma
expression systems, screening, scale-up culture, antibody fusion protein purification,
and initial characterization; and (c) in vitro characterization of the antibody fusion pro-
tein antigen-binding properties and cytokine biological activity. Finally, we briefly men-
tion some examples of in vivo studies using antibody-cytokine fusion proteins. As an
example we present the anti-HER2/neu human IgG3/IL-2 antibodycytokine fusion pro-
tein developed in our laboratory (31).
3.1. Construction of AntibodyCytokine Fusion Proteins

3.1.1. Cloning of the V-Region Sequence
into the Antibody-Expression Vectors
The production of functional antibodies requires the simultaneous expression of
genes coding for both L and H chains. Several expression vectors encoding for L and
H chains on separate plasmids are available for the construction of antibodycytokine
fusion proteins (29,3234). Alternatively, expression vectors containing the H and L
chain in one plasmid have been described (5,13,35,36). Although the single plasmid
has the advantage that co-transfection or sequential transfection is not required, large
constructs may be difficult to manipulate (26).
The first step to construct an antibodycytokine fusion protein is to prepare the
plasmids encoding the desired immunoglobulin isotype. Figure 3 shows the expres-
sion vectors developed in Dr. Sherie Morrisons laboratory that contain the human
CL region (Fig. 3A), and the human 1 (Fig. 3B) and human 3 (Fig. 3C) CH regions,
respectively. The expression vectors are designed as cassettes and facilitate the clon-
ing of PCR-amplified VL and VH regions in the corresponding CL and CH regions
(32,34).
V-region sequences targeting a tumor antigen can be obtained from hybridomas by
RT-PCR using as template the mRNA and a set of primers that anneal to the DNA
flanking regions encoding for the V regions (32,34,37), from a phage display library
(38) or from other sources. In the current example (anti-HER2/neu IgG3/IL-2), the
DNA encoding the VL and VH domains of the humanized antibody hum4D5-8 (39),
also known as Herceptin (Trastuzumab), specific for the tumor-associated antigen
HER2/neu was obtained from its developers at Genentech (San Francisco, CA) and
cloned by PCR into the L chain and 3 H chain PCR expression vectors.
The L-chain expression vector (Fig. 3A) contains the murine VH promoter for
expression in murine myeloma cells, the pBR322 origin of replication (see Note 1),
the -lactamase gene for selection in Escherichia coli, and the marker gpt (40), which
encodes xanthine-guanine phosphoribosyltransferase for selection in eukaryotic cells
with HXM. The L-chain vector has the genomic sequences corresponding to the
human CL region and includes, flanking the region encoding for the VL domain, the
EcoRV-SalI cloning sites allowing the direct replacement by a different VL region
(34). The insertion of the VL region into the L-chain expression vector requires stan-
dard methods of genetic engineering (41). To begin, amplify by PCR the DNA encod-
ing for the desired V-region sequence (in our case, anti-HER2/neu VL chain), to
generate an EcoRV blunt site at the 5' end and a SalI at the 3' end (32). Table 1 shows
the structure of the forward and reverse primers for PCR cloning of the VL region.
Ligate the PCR product to the TA cloning vector and sequence to identify potential
errors. Then, digest the PCR clone and the L-chain expression vector using the
EcoRV-SalI restriction endonucleases. Isolate the digested VL region fragment and
the L-chain expression vector, purify, and ligate. Finally, verify by restriction analy-
sis that the novel vector has the right structure.
The H-chain expression vectors have the same promoter, origin of replication, and
prokaryotic selection marker as the L-chain vector. However, instead of gpt, the
H-chain expression vectors have the eukaryotic selection marker his for resistance to
histidinol (42). Figures 3B and C show the vectors with the genomic sequence of the
CH region of IgG1 and IgG3, respectively. The vector has EcoRV and NheI restriction
sites separated by a short fragment that allows the cloning of different VH regions
(32). Antibody-cytokine fusion proteins have been successfully made with IgG1
(4,5,13); however, in the current example, IgG3 was used because it has an extended
hinge region. This long hinge region confers more flexibility and spacing to the fusion
proteins, facilitating simultaneous binding to antigen and cytokine receptor (3).
Fig. 3. Expression vectors for cloning VL and VH regions by PCR. (A) Scheme of the PCR
L-chain expression vector. Thick black lines with arrows indicate the exons of human CL
region and a VL sequence flanked by EcoRV-SalI cloning sites. The point of the arrow indi-
cates the direction of transcription. P refers to the promoter sequence, En the enhancer, 3'utr
the 3' untranslated region, amp the bacterial selective marker, and gpt the eukaryotic selec-
tive marker. (B,C) Maps of the PCR H-chain expression vectors 1 and 3, respectively.
References are similar to panel A except that the thick black lines with arrows indicate the
exons encoding for human CH region, his refers to the eukaryotic selective marker, and VH
region is absent; only the cloning sites EcoRV-NheI are present.
Table 1
Structure of the Primers for Cloning the VL Region
PCR primer Sequence structure
5' VL a 5'-GGGGATATCCACCATGNNN...3'
3' VL b 5'-AGCGTCGACTTAGCNNN...3'
a The 5' V sense primer contains the EcoRV site (under-
L
lined), the Kozak sequence (bold), the initiation codon (ATG),
and part of the leader of the human or mouse L chain (NNN...).
b The 3' V antisense primer include the SalI site (under-
L
lined) and a short fragment of a splicing donor site (bold). Re-
member that the vectors used have the antibody genomic
sequence. After the splicing site, include a fragment of the se-
quence complementary to the one that encode the J segment of
the desired VL region (NNN...).
Table 2
Structure of the Primers for Cloning the VH Region
5' VH a 5'-GGGGATATCCACCATGNNN...3'
3' VH b 5'-GGGGCTAGCNNN...3'
a The 5' V sense primer has the EcoRV site (underlined),
H
the Kozak sequence (bold), the initiation codon (ATG), and
part of the leader of the human or mouse H chain (NNN...).
Note that in the case of using a blunt site different than EcoRV,
the restriction site will be lost once it is ligated to the expres-
sion vector.
b The 3' V antisense primer includes the NheI site (under-
H
lined). After that sequence, include the sequence complemen-
tary to the one that encodes the J segment of the desired VH
region (NNN...).
The procedure for cloning a VH region (in our case anti-HER2/neu VH region) into
the PCR H chain 3 expression vector is similar to the one described for the VL
region. However, for the H-chain expression vector the cloning sites are EcoRV-NheI.
The 3' primer requires an NheI site because the first two amino acids of the CH1
domain of IgG3 and IgG1 are encoded in that site. Table 2 shows the structure of the
forward and reverse primers for PCR cloning of VH region.
3.1.2. Construction of the AntibodyCytokine Fusion Protein
Using our expression vectors, the fusion of the antibody and the cytokine sequences
requires the use of intermediate vectors. The nature of the intermediate steps required
would depend on the antibody region to which the nonantibody partner will be fused.
In the current example, the cytokine human IL-2 is fused at the 3' end of the anti-
HER2/neu IgG3-CH3. The cytokine cDNA can be obtained commercially (ATCC) or
cloned from a known source. For example, PMA-stimulated Jurkat cells can be used
Table 3
Forward and Reverse Primers for Cloning of Mature Human IL-2
5' (IL-2) a 5'- ACAACACGTGCACCTACTTCAAGTTC-3'
3' (IL-2) b 5'- GCACTGAATTCTCAAGTCAGTGTTGA-3'
a The 5' (IL-2) sense primer includes the PmlI blunt site (underlined) and
the first nucleotides of mature human IL-2 sequence.

b The 3' (IL-2) antisense primer includes the EcoRI site (underlined), the
sequence complementary to the stop codon (TCA, in bold) and the last nucle-
otides of human IL-2 sequence.
to generate cDNA by reverse transcription using an oligo dT primer (14). To clone the
cDNA encoding human IL-2, amplify this cDNA by PCR using primers flanking the
IL-2 coding sequence with appropriate restriction sites for subcloning (see Table 3).
Clone the cytokine PCR product into TA cloning vector and sequence to identify
potential errors. Using the restriction sites introduced during the PCR reaction (PmlI
and EcoRI), subclone the IL-2 cDNA into the intermediate vector created for the
fusion of the cytokine to the human IgG3-CH3-3' end (29). The digestion of the inter-
mediate vector with SspI results in a blunt end one nucleotide after the last CH3 codon,
while the IL-2 cDNA fragment digested with PmlI results in a blunt end two base
pairs before the first codon of mature IL-2. The ligation of the CH3 blunt end SspI and
the IL-2 PmlI result in a fusion protein linking the last amino acid of CH3 to the first
amino acid of mature IL-2 with a codon encoding a cysteine between them (Fig. 4A)
(see Note 2). The product of this ligation is an intermediate vector with the IgG3-CH3
domain of fused to the IL-2 and the 3' un-coding region (pAT3462/IL-2). The strategy
for the construction of IL-2 fused to anti-HER2/neu IgG3 (vector pAH4802-VH) is
illustrated in Fig. 4B. In this case, the presence of multiple restriction sites makes a
single ligation difficult, and a three-way ligation is required. Digestion of the vector
pAH4802-VH with NheI/NsiI results in the fragment CH 3 (fragment 1). The diges-
tion of pAH4802-VH with NheI, and partial digestion (43) of the linearized vector
with BamHI result in fragment 2, containing the expression vector backbone with VH
(longest fragment). Digestion of the intermediate vector pAT3462/IL-2 with BamHI/
NsiI produces fragment 3, which contains the DNA encoding for the last 20 amino
acids of IgG3-CH3 fused to IL-2 and the 3' un-translated region. The ligation of the
three fragments yields the antihuman HER2/neu IgG3/IL-2 H-chain expression vec-
tor (see Fig. 4B).
Our expression vectors are developed as cassettes, making it straightforward to
produce the same fusion proteins with variable regions of different affinities and/or
specificities (7,14,31) or produce novel fusion proteins with the same variable regions
but with a different nonimmunoglobulin partner (19, 446). Another alternative is the
creation of families of antibody fusion proteins with the nonantibody partner fused at
different positions of the CH region (29) or at the N-terminus of the VH region via a
flexible linker (17,19,44).
Fig. 4. Construction of anti-HER2/neu IgG3/IL-2 expression vector. Panel A shows the

nucleotide and amino acid sequence of the 3 CH3/IL-2 fusion region. The ligation of the blunt
sites SspI and PmlI results in a single amino acid (cysteine). Panel B show the strategy for the
construction of the expression vector for anti-HER2/neu IgG3/IL-2. Three-way ligation of the
fragments containing the DNA sequence encoding for the IgG3 constant regions from pAH4802
VH (fragment 1), the VH anti-HER2/neu and the expression vector backbone from pAH4802
VH (fragment 2), and the last 60 base pairs of 3 CH3 fused to IL-2 and the 3' untranslated
region from vector pAT3462/IL-2 (fragment 3). A dotted line outside the plasmid indicates the
fragment used in the three-way ligation.
3.2. Transfection into Expression Systems and Initial Characterization

The use of expression vectors with transcriptional control through immunoglobulin
promoter requires the use of cell lines from the B-cell lineage, such as murine myeloma
cells. This system has been used successfully in the production of several antibody
cytokine fusion proteins (26). Alternatively, the human cytomegalovirus (CMV) pro-
moter has been used for antibody production in both myeloma cell lines and mammalian
cells from other lineages, including Chinese hamster ovary (CHO) cells (5,4749).
3.2.1. Transfection into Mammalian Expression System

Co-transfection of murine myeloma cell lines (Sp2/0-Ag14, P3X63Ag8.653, or
NS0/1) with H- and L-chain expression vectors can result in successful producers
(44); however, we prefer to initially develop a L-chain producer as a recipient for
transfection of different H-chain constructs. This procedure saves time and favors the
development of cell lines expressing high levels of antibody fusion proteins (19,28,
45). To begin, prepare the L-chain producer by transfecting the L-chain expression
vector by electroporation (see step 5) and select stable transfectants by resistance to
HXM. Screen by ELISA (see Subheading 3.2.2.), select the best L-chain producer,
and use this stable cell line as a recipient for transfection with H-chain expression
vectors.
1. Following standard genetic engineering techniques (41), linearize 10 g of the expression
vectors with PvuI. Confirm digestion with ethidium bromide agarose gel (see Note 3).
Extract the DNA with phenol/chloroform (see Note 4). Precipitate with ethanol and air dry
for 30 min. Resuspend in 50 L cold sterile filtered PBS and keep at 4C until transfection.
2. Count 2 106 cells/plate in log-phase growth of the myeloma cell lines (in case of H-chain
transfection, use L-chain producers). For each transfection it is convenient to prepare five
plates plus one negative control plate that is not transfected (see Note 5).
3. Pellet cells by centrifugation at 220g (g corresponds to gravity and is equivalent to relative
centrifugal force or rcf) for 5 min at 4C, pour off supernatant, and resuspend in 0.5 mL
cold PBS.
4. Transfer the cells into the electroporation cuvet and add the linearized DNA.
5. Electroporate with a pulse of 960 Fd capacitance and 0.2 V pulse. Incubate on ice for
10 min.
6. Transfer to 10 mL 5% calf serum in IMDM at 4C. Pellet the cells by centrifugation
(220g, 5 min, 4C) and pour off the supernatant.
7. Resuspend the transfected and control cells in 12.5 mL/plate growth medium. Dispense
125 L/well in 96-well tissue-culture plates. Incubate at 37C and 5% CO2.
8. After 2 d, add 125 L of selection medium to the transfected and control cells. For L-chain
producers, use selection medium with HXM instead of histidinol. For P3X63Ag8.653
cells, the selection medium for H-chain transfectant is 5 mM histidinol or for L-chain
transfectant is HXM: 125 g/mL hypoxanthine, 7.5 g/mL xanthine, and 3 g/mL
mycophenolic acid (see Note 6). Note that the final concentration of selection medium in
the well is half.
9. Three d later, remove half of medium from the wells by aspiration and add 125 L more
of selection medium. The final concentration in the well of the drug used for selection
after this final application is 75%.
10. After 10 to 14 d, the transfected clones should have visible colonies. The control plate
should have no colonies; otherwise, you will have a high background of nonproducer
clones in your transfectants. Do not let the cells become too dense (indicated when the
medium turns yellow), since this may stress the cells and result in either selection for
low-level producers or loss of the clone. Screen for the secreted antibody fusion proteins
via ELISA.
3.2.2. ELISA Screening
1. Coat 96-well Immunolon II plate with 50 L/well of 5 g/mL goat antihuman IgG (for
H-chain producers) or 5 g/mL goat antihuman antibody (for L-chain producers) in
freshly made carbonate buffer using a multi-channel pipet (32). The following steps are
the same for screening both L-chain and H-chain producers.
2. Wash the plate wells four times with PBS and add 100 L of blocking buffer to each well
(see Note 7). For immediate use, leave at room temperature 30 min. For later use, leave at
4C wrapped carefully to avoid evaporation.
3. Wash four times with PBS, apply 50 L/well of supernatant from the samples to be
screened, and incubate either 2 h at room temperature (RT) in rotator or overnight at 4C.
Include a positive (any supernatant from a good producer) and negative control (superna-
tant from untransfected cells).
4. Wash four times with PBS and incubate for 1 h at 37C with AP-labeled goat antihuman
according to manufacturers instructions in 1% BSA in PBS.
5. Wash four times with PBS. Add freshly made AP substrate. Incubate at room temperature
30 min to 1 h and read the plates at 410 nm.
6. Compare the magnitude of the signal with the size of the clone, and identify a small clone
with high signal strength. With this criterion, pick the five to eight best producers.
7. Carefully resuspend the cells and transfer to a 24-well plate with 0.5 mL 10% calf
serum in IMDM to expand the clone. Transfer 2 to 6 d later to 60 15-mm Petri dishes
(see Note 8).
3.2.3. 35S Biosynthetic Antibody Labeling and Immunoprecipitation
This procedure is designed to determine the proper assembly, secretion, and mol wt
of the antibodycytokine fusion protein and to compare the productivity of different
clones in order to select the optimal producer.
1. Count 106 cells from the clones that seem to be the best producers by ELISA. Pellet by
centrifugation (5 min, 220g, 4C) and remove the supernatant.
2. Wash the cells twice, resuspending in 2 mL of deficient DMEM, and pellet by centrifuga-
tion (5 min, 220g, 4C).
3. Remove supernatant and add 1 mL of [35S]-labeling medium containing 16 Ci/mL to
each cell line (see Note 9).
4. Under sterile conditions, move the cells into 24-well plate and incubate overnight at 37C
in 5% CO2. The supernatant should contain the secreted [35S]-labeled antibody.
5. Resuspend the cells and transfer to 1.5-mL Eppendorf tube. Pellet by centrifugation (3
min, 1500g, RT).
6. Transfer the supernatant (secreted fraction) to a new 1.5-mL tube and add 0.5 mL of
NDET to the pellet containing the cells (cytoplasmic fraction). Resuspend by pipetting up
and down to lyse the cells. Pellet the cell lysate by centrifugation (5 min, 21,000g, 4C)
and transfer the supernatant to a new tube. Discard the pellet (see Note 10).
7. Add 2.5 L of hyperimmune rabbit antihuman IgG to both sets of tubes (the secreted
fraction and the cytoplasmic fraction). Incubate on ice for 1 h (see Note 11).
8. Add 100 L of Staph A (mix the suspension well before using it). Incubate 5 min at RT.
Pellet by centrifugation (1 min, 21,000g, RT) and remove supernatant.
9. Resuspend the immunoprecipitates with 0.5 mL 0.3% SDS in NDET. Layer samples on
top of a 1.0-mL sucrose pad in a 1.5-mL Eppendorf tube. Pellet by centrifugation (1 min,
21,000g, RT). Discard supernatant.
10. Wash the precipitate, resuspending it in 300 L 0.3% SDS in NDET and pellet by cen-
trifugation (1 min, 21,000g, RT). Discard supernatant. Wash again (optional).
11. Resuspend the immunocomplex in 0.5 mL distilled H2O to remove detergents, and pellet
by centrifugation (1 min, 21,000g, RT). Drain well.
12. Add 50 L of SDS-PAGE sample buffer, vortex, and incubate 2 min in boiling water to
separate radiolabeled antibody from Staph A. Pellet by centrifugation and transfer the
supernatant with the soluble radiolabeled antibody into a new tube.
13. For SDS-PAGE analysis, run half of the radiolabeled sample under nonreducing condi-
tions on 5% phosphate gel and the other half under reducing conditions (30 min incuba-
tion at 37C in the presence of 0.15 M -mercaptoethanol) on 12% Tris-glycine gels (50)
with radiolabeled mol wt marker. Dry the gels and expose for autoradiography to check
the correct mol wt and assembly of the transfected proteins (32,51). Always include as a
negative control the untransfected cell line and (if available) a good producer as a posi-
tive control (see Note 12). As the productivity of the clone is proportional to the intensity
of the signal, this will be a criterion for the selection of the best producer.
14. Once one cell line is selected, resuspend the cells and transfer it to 24-well plate to scale up.
15. It is necessary to subclone the cells by limiting dilution to assure the isolation of a pure
clone. Take 100 L of the log-phase cell culture and make serial dilutions in three tubes
with 12.5 mL of subcloning medium. Plate the three different dilutions into 96-well tis-
sue-culture plates, 125 L/well. Incubate at 37C and 5% CO2. The next day, add 125 L/
well of subcloning medium, and incubate 7 to 10 d until colonies become visible. Screen
the plates that have colonies by ELISA (see Subheading 3.2.2.) in less than 33% of the
wells to minimize the chance of selecting a well with multiple colonies. Pick the best
producer based on the criterion of small-size colony and high signal.
16. Expand the selected producer into roller bottles. Take a sample to prepare frozen
stocks (52).
3.2.4. Growing Cells in Roller Bottles

1. Take about 5 107 cells growing in 50 mL of medium (two 100 20-mm Petri dishes).
Pellet by centrifugation for 5 min, 220g, at 4C.
2. Resuspend in 100 mL of 10 % (v/v) Fetalclone serum in IMDM and seed into a 2-L roller
bottle. It is not necessary to add the drug used for selection with subcloned cell lines. If a
room with gas interchange is not available, briefly bubble the medium in the bottle with
CO2 and tighten the cap immediately. Put on a roller apparatus at 37C (see Note 13).
3. Check the cells daily. When the medium turns orange-yellow due to the density of the
cells, double the volume with IMDM alone. The cells often adhere to the sides of the flask
and then grow in suspension.
4. Add IMDM up to 1 L to a final concentration of 1% Fetalclone serum. The full process
can take 10 d to 4 wk, depending on the growth rate of the transfectant. As the cells
continue to grow in roller bottles, the medium may be depleted of glutamine, resulting in
lower antibody fusion protein yield. It is convenient then to add 35 mL of GlutaMAX-1
once the final volume of 1 L is reached.
5. If you wish to reuse the cells, harvest the supernatant just as the cells begin to look sick.
Pellet by centrifugation as described in step 6. Add fresh medium with 1% Fetalclone
serum and the cells. Harvest the supernatant as before. Alternatively, let the cells die; this
will increase your yield of antibody.
6. Pellet the cells by centrifugation in 250-mL centrifuge bottles at 3500g for 10 min at 4C.
7. Filter through Whatman 3MM paper or the equivalent before running through the chro-
matography column. To stabilize the supernatant, it is convenient to add 0.02% sodium
azide, 2 mM EDTA, 10 mM PO4 buffer and 0.45 M NaCl. Store at 4C.
8. If after several preparations the cell line looses protein expression, it is convenient to
subclone (see Subheading 3.2.3., step 15) to isolate again a good producer. To avoid this
problem, it is convenient to have a stock of good producers. Depending on the quality of

the producers and the amount of protein required, the number of roller bottles required
may differ. It is convenient to start with one or two roller bottles, determine the yield, and
scale up if needed.
3.2.5. Purification, Quantification, and Storage

Different types of columns can be used for purification of antibodycytokine fu-
sion proteins, including protein A, protein G, or columns with the antigen (53). Al-
though human IgG3 does not bind protein A through its Fc fragment, the HER2/neu
IgG3/IL-2 fusion protein that we have developed is able to bind protein A through its
V region, since we use the V region of the humanized antibody hum4D5-8, also known
as Herceptin (Trastuzumab; Genentech, San Francisco, CA) (28,54).
1. Flow the roller bottle culture supernatants (up to 4 L) through a mini-column containing
1 mL protein A on 6% fast-flow beaded agarose by gravity. Repeat the procedure two to
four times.
2. Prior to the elution, set up the 1.5-mL fraction tubes (Eppendorf tubes) with 150 L of
2 M Tris-HCl (pH 8.0) to neutralize the eluates, and keep them on ice.
3. To collect the antibody fusion protein from the column, first wash the column with 10 mL
of PBS with 0.02% sodium azide.
4. Elute the column with 2 mL buffer A, 4 mL buffer B, and 2 mL buffer C. In all cases
(buffers A, B, and C) add up to 1 mL of buffer, collect the corresponding fraction, mix the
tube gently to avoid bubbles and to neutralize the low pH, and repeat the procedure until
the second mL of the buffer C passes through the column. Store the fraction tubes at 4C.
5. After the elution of buffer C, wash the column again with 10 mL PBS with 0.02% sodium
azide as in step 2. To store the column, fill with PBS with 0.02% sodium azide, seal with
parafilm, and keep at 4C.
6. Take 10 L of each fraction, add sample buffer, and run an SDS-PAGE 5% phosphate gel
under nonreducing conditions. Stain with Coomasie blue to detect the fractions with the
band corresponding to the fusion protein. The fusion protein should be collected during
the elution with buffer B (pH 2.5) (Fig. 5).
7. Repeat the procedure from steps 16 until no more antibodies are collected. To assess the
rate of antibody fusion protein depletion from the roller-bottle supernatants, take samples
before proceeding to the elution and compare by quantitative ELISA (see Note 14).
8. Dialyze the eluted fractions using dialysis bags of appropriate molecular mass cut-off in
a 2-L volume of dialysis buffer or PBS without sodium azide (51). Change three times
every 12 h. In case the concentration of eluted fractions is too low, concentrate using
Centricon centrifugal filter concentrator with a molecular mass cut-off appropriate to the
size of the antibody fusion protein prepared. Later proceed to dialysis.
9. Check the integrity of the protein before the quantification and storage by SDS-PAGE.
Run the sample under nonreducing conditions on 5% phosphate gel, and under reducing
conditions on 12% Tris-glycine gel (see Subheading 3.2.3., step 13). Stain with
Coomasie Blue (51). Figure 6 shows a Coomasie blue-stained gel comparing the anti-
body alone with the antibodycytokine fusion protein under nonreducing (Fig. 6A) and
reducing conditions (Fig. 6B). The pattern should be similar to the autoradiography in
Subheading 3.2.3., step 13.
10. Determine protein concentrations by bicinchoninic acid (BCA)-based protein assay using
BSA as standard. As an alternative, use quantitative ELISA if available.
Fig. 5. Profile of anti-HER2/neu IgG3/IL-2 elution fractions from protein A column ana-
lyzed by SDS-PAGE under nonreducing conditions and stained by Coomasie blue. In the first
fractions (1 to 3) at high pH, traces of bovine IgG can be found that correspond to a faint band
at 150 kDa. With the elution at pH 2.5 (fractions 4 to 7), the 200-kDa band appears, which
exhibits a mol wt corresponding to our antibody-cytokine fusion protein. At pH 2.0 (fractions 8
and 9), no protein is collected.
Fig. 6. SDS-PAGE analysis of anti-HER2/neu IgG3/IL-2. Secreted antibodies anti-HER2/

neu IgG3 (lane 1) and anti-HER2/neu IgG3/IL-2 (lane 2) were purified from culture superna-
tants and analyzed by SDS-PAGE stained with Coomasie blue under nonreducing (A) and
reducing (B) conditions. The estimated mol wt corresponds to 170 kDa for the antibody alone
(A, lane 1) vs 200 kDa for the fully assembled antibody-cytokine fusion protein (A, lane 2)
(H2L2). The 60-kDa size of the H chain alone (B, lane 1) shows an expected increase to 75 kDa,
corresponding to the cytokine fused with the H chain (B, lane 2) (H), while no changes in mol
wt are detected in the 25-kDa L-chain band (L).
11. The antibody fusion proteins can be stored in dialysis buffer or PBS at 4C if they will be
used in the short term (i.e., less than 6 mo). For longer storage, the fusion proteins can be
aliquoted, snap frozen, and stored at 80C (52). In both cases (short and long term), a
concentration of 1 mg/mL is recommended (see Note 15).
3.3. Further In Vitro Characterization of the Antibody Fusion Protein

Antibodycytokine fusion proteins consist in two moieties: the antibody moiety
that targets the tumor antigen, and the cytokine moiety that possess immunostimulato-
ry activity. Their coexistence in the same molecule may affect the functional activity
of each other. In fusion proteins that include complete antibodies, the Fab region must
be tested for antigen binding while, depending on the isotype, the antibody Fc region
can be tested using standard assays to study effector functions such as ADCC and
CDC (4,5,13,14,55). The fused cytokine also requires specific functional assays that
are necessary to test its effective activity. Below we describe the procedures to study
binding of the antibody to the antigen and the activity of the cytokine (IL-2).
3.3.1. Antigen-Binding Studies

3.3.1.1. ELISA TO TEST ANTIBODYCYTOKINE FUSION PROTEIN ANTIGEN BINDING
The binding properties of the antibody may be affected by its fusion to the cytokine.
ELISA is a convenient method to study the capacity of the antibody fusion protein to
bind the antigen and estimate its comparative avidity in respect to the parental anti-
body (see Note 16). The current procedure corresponds to the binding of antihuman
HER2/neu IgG3/IL-2 fusion protein to its antigen, the extracellular domain of HER2/
neu (ECD-HER2), but can be adjusted to study antibody fusion proteins with different
specificities.
1. Coat the 96-well Immunolon II plates with 50 L of the antigen ECD-HER2, 1 g/mL, in
carbonate buffer, overnight at 4C.
2. Block the plates according to Subheading 3.2.2., step 2.
3. Wash four times with PBS and apply serial dilutions of the antibodycytokine fusion
protein, the parental antibody as positive control, and a nonspecific isotype antibody as
negative control (in our case, we use anti-DNS IgG3 antibody) in 50 L PBS containing
1% BSA. Incubate overnight at 4C.
4. Continue the protocol according to Subheading 3.2.2., steps 4 and 5.
3.3.1.2. FLOW CYTOMETRY

The binding properties of the antibodycytokine fusion protein to the antigen coated
on a solid surface may not reflect effective binding to the antigen expressed on the
surface of tumor cells. Flow cytometry can be used to determine effective binding of
the antibodycytokine fusion protein to cells expressing the target antigen on its sur-
face (see Note 17). The same method can be adapted to study other antibodycytokine
fusion proteins.
1. Count 106 cells expressing the antigen on the surface (in this case, CT26-HER2/neu).
Pellet cells by centrifugation at 220g, 5 min, at 4C, pour off supernatant, resuspend in
1.0 mL of cold flow-cytometer buffer, and transfer to 1.5-mL tube. Keep the cells on ice
through the procedure.
2. Pellet the cells by centrifugation (3 min at 1500g and RT) and remove the supernatant.
Incubate with 1 g of the antibodycytokine fusion protein, anti-HER2/neu IgG3/IL-2,
in 0.1 mL of flow-cytometry buffer for 2 h at 4C. Shake gently every 15 min. Use anti-
HER2/neu IgG3 (44) and isotype control (in our case, we use anti-DNS IgG3 antibody)
as positive and negative isotype controls, respectively. In addition, if available, the

parental cell line not expressing the antigen should also be used as control to test bind-
ing specificity.
3. Wash the cells adding 1 mL of flow-cytometry buffer, mix, centrifuge for 3 min at 1500g
at RT, and remove the supernatant. Repeat the washing step one more time and remove
the supernatant.
4. Incubate for 30 min at 4C with 0.5 g of biotinylated goat antihuman IgG in a volume of
0.1 mL of flow-cytometry buffer. Shake gently every 15 min.
5. Wash the cells with 1 mL of flow-cytometry buffer, centrifuge for 3 min at 1500g and
RT, and remove the supernatant. Incubate with PE-labeled streptavidin according to the
manufacturers instructions in a volume of 0.1 mL of flow-cytometry buffer for 30 min at
4C in the dark. Shake gently every 15 min.
6. Wash the cells with 1 mL of flow-cytometry buffer, centrifuge for 3 min at 1500g and
RT, and remove the supernatant. Resuspend in 0.2 mL flow-cytometry buffer to read
immediately in flow cytometer. To store the cells for a period of days, resuspend the cells
in 0.2 mL 2% paraformaldehyde, cover the tubes with aluminum foil, and keep at 4C
(see Note 18).
7. Perform analysis by flow cytometry with a FACScan equipped with a blue laser excita-
tion of 15 mW at 488 nm (Fig. 7).
3.3.2. Study of the Cytokine Biological Activity

The cytokine bioactivity can be tested in vitro by the magnitude of the biological
effect that the cytokine has on sensitive cell lines. Biological activities induced by
cytokines include induction of cell proliferation, degranulation, and cytotoxicity,
among others. A convenient type of test is the induction of cell proliferation depen-
dent on the cytokines. Ideally, it requires the use of a cell line dependent on the
cytokine being studied, as is the case of the murine macrophage cell line FDC-P1 for
murine GM-CSF, or the murine cell lines CTLL-2 and HT-2 cells for human and
murine IL-2 (4,5,13,14,45). Here we describe a protocol to determine functional
activity of IL-2 genetically fused to an antibody, using CTLL-2 proliferation assay.
1. Load into tissue-culture flat-bottom 96-well plates quadruplicate serial dilutions from
10 ng/mL to 0.01 ng/mL of rhuIL-2 (positive control), anti-HER2/neu IgG3/IL-2, and
anti-HER2/neu IgG3 (negative control) in 50 L complete RPMI 10 medium.
2. Seed with 50 L of complete RPMI 10 medium containing 5 103 CTLL-2 cells (see
Note 19). Incubate for 18 h at 37C 5% CO2 (see Note 20).
3. Pulse with 0.5 Ci/well [3H]-thymidine in 25 L complete RPMI 10 medium for 6 h (see
Note 21).
4. Harvest with cell harvester through glass microfiber filter to collect the radiolabeled DNA
(see Note 22).
5. Air-dry the filters and seal in envelope with EcoLume scintillation medium. Quantitate in
a liquid scintillation counter.
6. Derive a linear regression from IL-2 control within the range where the growth is linearly
dependent on IL-2 concentration. Determine sample activity by plotting points that fall
within the linear range on the linear regression.
In general, similar proliferation profile in rhuIL-2 and the IL-2 antibody-fusion protein
has been observed; however, some cases reveal fusion proteins with lower activity than
the free cytokine (5,7,13,14).
Fig. 7. Flow cytometry demonstrating binding of anti-HER2/neu IgG3/IL-2 to the HER2/

neu expressed on the surface of CT26-HER2/neu. CT26-HER2/neu cells were incubated with
anti-DNS human IgG3 (negative control) (A), anti-HER2/neu human IgG3 (positive control)
(B), or anti-HER2/neu IgG3/IL-2 (C), followed by biotinylated goat antihuman IgG and PE-
labeled streptavidin. FL2-H indicates PE fluorescence.
3.4. In Vivo Assays for AntibodyCytokine Fusion Proteins

Although in vivo studies are not the scope of this chapter, they are necessary to
determine the effectiveness of the fusion proteins as cancer therapeutic tools. In gen-
eral, these studies include half-life, biodistribution, tumor targeting, antitumor activ-
ity, and toxicity. A basic in vivo study required to understand the biological activity
of the antibody-cytokine fusion protein is the pharmacokinetics. The half-life of anti-
bodies in serum is in the range of d; the half-life of cytokines ranges from min to h
(56); for antibody-cytokine fusion proteins, the half-life is intermediate between that
of the antibody and of the cytokine (5,57,58). The biodistribution patterns indicate
both the areas in which the antibodycytokine fusion protein concentrates in the body
and whether it can effectively target tumors (5,7,59). The antitumor activity is highly
dependent on the model used, the targeting antigen, and the fused cytokine. Synge-
neic tumor models are important to understand the immunological response triggered
by treatment with the antibodycytokine fusion proteins (1,3). These tumor models
can be local subcutaneous (s.c.) tumors (4,7) or metastases induced in lung or liver
(10). In general, the cell lines used to induce tumors express the antigen either natu-
rally or artificially (transduction or transfection) and are able to grow in mice while
maintaining the expression of the antigen (7,8,10,46,60). Experiments can be designed
to use the antibodycytokine fusion protein directly to target the tumor (4,7,10,61,62)
or as adjuvants of a tumor antigen vaccination (6). In all cases, it is important to
confirm that the treatment is more effective than the mixture of free antibody and
cytokine or a nonspecific antibodycytokine fusion protein (1,3,10,58,62). It is nec-
essary to study the mechanism responsible for antitumor activity, since the presence
of the cytokine in the antibody may cause toxicity, and to conduct toxicology studies,
which are critical if the antibody-cytokine fusion protein is intended to be used in
clinical trials.
4. Notes
1. Because pBR322 is a low-copy-number plasmid, it is recommended to amplify the num-
ber of copies of the plasmid by adding chloramphenicol to the bacteria culture (63). We
use the E. coli strain HB101 to amplify our intermediate and expression vectors.
2. Fully functional antibodycytokine fusion proteins, such as IgG1/IL-2 and IgG1/GM-
CSF have been constructed with a peptide linker between the H-chain C-terminus and the
cytokine (4,5,13). IL-2 or GM-CSF has also been fused to the C-terminus of the CH3
domain in the absence of a peptide linker, and the cytokine remains active (14,45). How-
ever, it may be critical to include a flexible linker between the cytokine and the antibody
when the nonantibody partner is fused to the N-terminus of the H chain, to facilitate
antigen binding (17,19,44,64).
3. CAUTION: Ethidium bromide is a strong mutagen and a possible carcinogen or terato-
gen. Wear a lab coat, chemical splash goggles, and nitrile gloves.
4. CAUTION: Phenol/chloroform is poisonous. Take precautions to avoid contact or inges-
tion. Chloroform is a suspected carcinogen. Perform all work in a chemical fume hood to
avoid exposure by inhalation; wear protective clothes and gloves.
5. It is convenient to transfect different cell lines such as Sp2/0-Ag14, P3X63Ag8.653, or
NS0/1 in parallel. Different antibody fusion proteins may be produced more successfully
in one cell line than the other.
6. Previous to the transfection, it is convenient to make a titration curve in order to establish
the minimum concentration of drug necessary to kill untransfected cells. Once the cell
lines have been transfected, do not use more than one drug simultaneously, because the
toxicity may be too high for the cell line to tolerate. In case of co-transfection, we have
effectively selected clones using only the H-chain selection marker (histidinol).
7. CAUTION: Contains sodium azide, which is poisonous. It is a blocker of the cellular
respiratory chain. Ingesting very small amounts can cause death. Avoid exposure by in-
halation and skin contact. Wear protective clothes and gloves.
8. Nonadherent cell lines derived from Sp2/0-Ag14, P3X63Ag8.653, or NS0/1 myeloma
cell lines can grow in Petri dishes; however, this is not the case for adherent cell lines
such as CHO cells, which require the use of tissue-culture dishes to grow properly.
9. CAUTION: This procedure requires the use of 35S and should be performed in Nuclear
Regulatory Commission (NRC)-licensed sites by personnel trained in the use of this iso-
tope.
10. We recommend analyzing both the cytoplasm and the secretions, especially the first time
the transfection is made. The analysis of the cytoplasmic fraction allows the identifica-
tion of antibody molecules during the synthesis process. The analysis of the supernatant
fraction reveals whether the protein is being properly secreted and assembled. Most of the
protein should be in the supernatant.
11. It is important to stress that the protein-A binding site for IgG is in the Fc region and that
it is present in all wild-type human IgG except IgG3 (65). For immunoprecipitation of
IgG3 fusion proteins with Staph A, a prior incubation with a polyclonal rabbit antihu-
man IgG allows binding to protein A. In the case of L-chain producers, the L chain lacks
a protein-A binding site; hence, co-immunoprecipitation with a rabbit antihuman Fab
capable of binding the human L chain is required. To prepare these types of polyclonal
antibodies, see ref. 30.
12. In our laboratory, for research use in mice, we define a good producer as the cell line
able to secrete at least 1 g/106 cells/24 h. However, for large-scale production of anti-
bodycytokine fusion proteins for clinical use, higher levels of antibody fusion protein
production may be required.
13. Fetalclone serum is partially depleted of bovine IgG, which is convenient to avoid con-
tamination with bovine IgG during the antibody fusion protein purification process. As
an alternative, cells can be adapted to grow in serum-free media.
14. Quantitative ELISA can be used to determine the amount of antibody secreted by the
transfectant. To perform a quantitative ELISA, follow the directions in Subheading
3.2.2., steps 15. In the same plate, prepare a calibration curve with serial dilution using
an isotype quantified by BCA as standard (see Subheading 3.2.5., step 10).
15. Since freezing and thawing the antibody fusion protein may be harmful to both the anti-
body and the ligand, we recommend determining the antibody binding properties as well
as the ligand specific activity of each new fusion protein before and after freezing. In
addition, aggregates can also occur, resulting in a decrease of activity. To assess the pres-
ence of aggregates, analyze a sample using fast-performance liquid chromatography
(FPLC) (45).
16. Binding assays such as BIAcore or IAsys allow exact determination of affinity. With
these assays, the kinetics of the antibodycytokine fusion protein binding to the antigen
can be studied in real time (17). This technology can also be used to study the binding of
the fused cytokine to its receptor (66).
17. An alternative to flow cytometry to study the binding of antibody fusion protein to the
surface of cells is cell ELISA (67).
18. CAUTION: Paraformaldehyde is used in these procedures and is poisonous and a sus-
pected carcinogen. It may cause heritable genetic damage and irreversible damage to
sight, and is corrosive. Perform all work with in a chemical fume hood to avoid exposure
by inhalation. Wear protective clothes, goggles, and gloves.
19. The process of thawing can be very traumatic for the CTLL-2 cells. The d after thawing,
it may seem like all the cells have died. However, after 7 to 10 d with 20% calf serum in
IMDM supplemented with 100 U/mL IL-2, the cells can be recovered. Once the density
increases, decrease the calf serum to 5% and IL-2 to 5 to 10 U/mL.
20. For measurement of proliferation, alternatives to the use of radioisotope include the
colorimetric methods MTT and MTS/PMS (68,69). These methods detect cell meta-
bolic activity proportional to the cell number. The MTS/PMS assay Cell Titer 96 aque-
ous nonradioactive colorimetric assay (Promega, Madison, WI) is more sensitive than
the [3H]-thymidine incorporation in determining IL-2-induced CTLL-2 proliferation
(69). The higher sensitivity results from the fact that low concentrations of IL-2 are able
to increase the metabolism of CTLL-2 cells without de novo DNA synthesis. If the
MTS/PMS assay is used, the range of standard rhuIL-2 concentration for serial dilution
has to be from 1 ng/mL to 0.001 ng/mL, one order of magnitude lower than for [3H]-
thymidine incorporation. Drawbacks of this method that may affect the final interpreta-
tion of the data include proportionally higher background levels, and increased MTS/
PMS signal from dying cells (69).
21. CAUTION: This procedure requires the use of 3H and should be performed in NRC-
licensed sites by personnel trained in the use of this isotope.
22. For your convenience, you may freeze the plate at 20C to stop the proliferation and
harvest up to 2 wk later. To thaw the plates, leave 45 min at RT or 10 min at 37C.
Acknowledgments
We would like to thank Dr. Sherie L. Morrison and Mr. K. Ryan Trinh for their
professional advice. Our work using antibody fusion proteins was supported in part by
grants CA86915 from NIH/NCI, the 2002 AACR-California Department of Health
Services Career Development Award in Gender-related Cancer Research, the 2003
UCLA Jonsson Comprehensive Cancer Center Interdisciplinary Grant, and the 2004
Brian D. Novis International Myeloma Foundation (IMF) Senior Grant Award.
References
1.
1 Lode, H. N., Xiang, R., Becker, J. C., Gillies, S. D., and Reisfeld, R. A. (1998)
Immunocytokines: a promising approach to cancer immunotherapy. Pharmacol. Ther. 80,
277292.
2.
2 Helguera, G., Morrison, S. L., and Penichet, M. L. (2002) Antibody-cytokine fusion pro-
teins: harnessing the combined power of cytokines and antibodies for cancer therapy. Clin.
Immunol. 105, 233246.
3. Penichet, M. L., Shin, S. U., and Morrison, S. L. (1999) Fab fusion proteins: Immuno-
ligands. In: Antibody Fusion Proteins. Chamow, S. M. and Ashkenazi, A., eds. John Wiley
& Son, New York, NY, pp. 1552.
4.
4 Liu, S. J., Sher, Y. P., Ting, C. C., Liao, K. W., Yu, C. P., and Tao, M. H. (1998) Treat-
ment of B-cell lymphoma with chimeric IgG and single-chain Fv antibody-interleukin-2
fusion proteins. Blood 92, 21032112.
5.
5 Hornick, J. L., Khawli, L. A., Hu, P., Lynch, M., Anderson, P. M., and Epstein, A. L.
(1997) Chimeric CLL-1 antibody fusion proteins containing granulocyte-macrophage
colony-stimulating factor or interleukin-2 with specificity for B-cell malignancies
exhibit enhanced effector functions while retaining tumor targeting properties. Blood 89,
44374447.
6.
6 Dela Cruz, J. S., Lau, S. Y., Ramirez, E. M., et al. (2003) Protein vaccination with the
HER2/neu extracellular domain plus anti-HER2/neu antibody-cytokine fusion proteins
induces a protective anti-HER2/neu immune response in mice. Vaccine 21, 13171326.
7.
7 Penichet, M. L., Harvill, E. T., and Morrison, S. L. (1998) An IgG3-IL-2 fusion protein
recognizing a murine B cell lymphoma exhibits effective tumor imaging and antitumor
activity. J. Interferon Cytokine Res. 18, 597607.
8.
8 Becker, J. C., Pancook, J. D., Gillies, S. D., Furukawa, K., and Reisfeld, R. A. (1996) T
cell-mediated eradication of murine metastatic melanoma induced by targeted interleukin
2 therapy. J. Exp. Med. 183, 23612366.
9.
9 Becker, J. C., Varki, N., Gillies, S. D., Furukawa, K., and Reisfeld, R. A. (1996) Long-
lived and transferable tumor immunity in mice after targeted interleukin-2 therapy. J. Clin.
Invest. 98, 28012804.
10 Xiang, R., Lode, H. N., Dolman, C. S., et al. (1997) Elimination of established murine
10.
colon carcinoma metastases by antibody-interleukin 2 fusion protein therapy. Cancer Res.
57, 49484955.
11. Hank, J. A. A. M., Gan, J., Sternberg, A., et al. (2003) Clinical Administration of anti-
body-cytokine hu14.18-IL-2 induces IL-2-mediated immune activation. In: Proceedings
of the American Association for Cancer Research, 94th Annual Meeting. 2nd Ed., Vol. 44,
Washington, DC, p. 1156.
12. Janeway, C. A., Travers, P., Walport, M., and Schlomchick, M. (2001) Immunobiology:
the Immune System in Health and Disease. 5th Ed. Garland, New York, NY.
13 Gillies, S. D., Reilly, E. B., Lo, K. M., and Reisfeld, R. A. (1992) Antibody-targeted
13.
interleukin 2 stimulates T-cell killing of autologous tumor cells. Proc. Natl. Acad. Sci.
USA 89, 14281432.
14.
14 Harvill, E. T. and Morrison, S. L. (1995) An IgG3-IL2 fusion protein activates comple-
ment, binds Fc gamma RI, generates LAK activity and shows enhanced binding to the
high affinity IL2-R. Immunotechnology 1, 95105.
15.
15 Brambell, F. W., Hemmings, W. A., and Morris, I. G. (1964) A theoretical model of
gamma-globulin catabolism. Nature 203, 13521354.
16.
16 Ward, E. S., Zhou, J., Ghetie, V., and Ober, R. J. (2003) Evidence to support the cellular
mechanism involved in serum IgG homeostasis in humans. Int. Immunol. 15, 187195.
17 Challita-Eid, P. M., Abboud, C. N., Morrison, S. L., et al. (1998) A RANTES-antibody fusion
17.
protein retains antigen specificity and chemokine function. J. Immunol. 161, 37293736.
18.
18 McGrath, J. P., Cao, X., Schutz, A., et al. (1997) Bifunctional fusion between nerve growth
factor and a transferrin receptor antibody. J. Neurosci. Res. 47, 123133.
19.
19 Peng, L. S., Penichet, M. L., and Morrison, S. L. (1999) A single-chain IL-12 IgG3 anti-
body fusion protein retains antibody specificity and IL-12 bioactivity and demonstrates
antitumor activity. J. Immunol. 163, 250258.
20.
20 Lustgarten, J., Marks, J., and Sherman, L. A. (1999) Redirecting effector T cells through
their IL-2 receptors. J. Immunol. 162, 359365.
21.
21 Halin, C., Gafner, V., Villani, M. E., et al. (2003) Synergistic therapeutic effects of a
tumor targeting antibody fragment, fused to interleukin 12 and to tumor necrosis factor
alpha. Cancer Res. 63, 32023210.
22.
22 Moreland, L. W., Baumgartner, S. W., Schiff, M. H., et al. (1997) Treatment of rheuma-
toid arthritis with a recombinant human tumor necrosis factor receptor (p75)-Fc fusion
protein. N. Engl. J. Med. 337, 141147.
23.
23 Moreland, L. W., Schiff, M. H., Baumgartner, S. W., et al. (1999) Etanercept therapy in
rheumatoid arthritis. A randomized, controlled trial. Ann. Intern. Med. 130, 478486.
24.
24 Verma, R., Boleti, E., and George, A. J. (1998) Antibody engineering: comparison of bacte-
rial, yeast, insect and mammalian expression systems. J. Immunol. Methods 216, 165181.
25.
25 Potter, K. N., Li, Y., and Capra, J. D. (1993) Antibody production in the baculovirus
expression system. Int. Rev. Immunol. 10, 103112.
26.
26 Yoo, E. M., Chintalacharuvu, K. R., Penichet, M. L., and Morrison, S. L. (2002) Myeloma
expression systems. J. Immunol. Methods 261, 120.
27.
27 Wright, A., Shin, S. U., and Morrison, S. L. (1992) Genetically engineered antibodies:
progress and prospects. Crit. Rev. Immunol. 12, 125168.
28.
28 Penichet, M. L., Dela Cruz, J. S., Shin, S. U., and Morrison, S. L. (2001) A recombinant
IgG3-(IL-2) fusion protein for the treatment of human HER2/neu expressing tumors. Hum.
Antibodies 10, 4349.
29 Shin, S. U., Friden, P., Moran, M., and Morrison, S. L. (1994) Functional properties of
29.
antibody insulin-like growth factor fusion proteins. J. Biol. Chem. 269, 49794985.
30. Harlow, E. and Lane, D. (1989) Anti-immunoglobulin antibodies. In: Antibodies: A
Laboratory Manual. 1st Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY,
pp. 623631.
31.
31 Penichet, M. L., Dela Cruz, J. S., Challita-Eid, P. M., Rosenblatt, J. D., and Morrison, S.
L. (2001) A murine B cell lymphoma expressing human HER2 / neu undergoes spontane-
ous tumor regression and elicits antitumor immunity. Cancer Immunol. Immunother. 49,
649662.
32. Morrison, S. L. (1994) Cloning, expression, and modification of antibody V regions. In: Cur-
rent Protocols in Immunology, Vol. 1. Coligan, J. E., Kruisbeek, A. M., Margulies, D. H.,
Shevach, E. M., and Strober, W., eds. John Wiley & Sons, New York, NY, pp. 2.12.12.12.16.
33.
33 Shin, S. U. and Morrison, S. L. (1990) Expression and characterization of an antibody
binding specificity joined to insulin-like growth factor 1: potential applications for cellu-
lar targeting. Proc. Natl. Acad. Sci. USA 87, 53225326.
34.
34 Coloma, M. J., Hastings, A., Wims, L. A., and Morrison, S. L. (1992) Novel vectors for
the expression of antibody molecules using variable regions generated by polymerase chain
reaction. J. Immunol. Methods. 152, 89104.
35.
35 Norderhaug, L., Olafsen, T., Michaelsen, T. E., and Sandlie, I. (1997) Versatile vectors for
transient and stable expression of recombinant antibody molecules in mammalian cells. J.
36 Preston, M. J., Gerceker, A. A., Reff, M. E., and Pier, G. B. (1998) Production and
36.
characterization of a set of mouse-human chimeric immunoglobulin G (IgG) subclass
and IgA monoclonal antibodies with identical variable regions specific for Pseudomo-
nas aeruginosa serogroup O6 lipopolysaccharide. Infect. Immun. 66, 41374412.
37.
37 Larrick, J. W., Danielsson, L., Brenner, C. A., Abrahamson, M., Fry, K. E., and
Borrebaeck, C. A. (1989) Rapid cloning of rearranged immunoglobulin genes from human
hybridoma cells using mixed primers and the polymerase chain reaction. Biochem.
Biophys. Res. Commun. 160, 12501256.
38. Powers, D. B. and Marks, J. (1999) Monovalent phage display of Fab and scFv fusions. In:
Antibody Fusion Proteins. Chamow, S. M. and Ashkenazi, A., eds. John Wiley & Son,
New York, NY, pp. 151188.
39.
39 Carter, P., Presta, L., Gorman, C. M., et al. (1992) Humanization of an anti-p185HER2
antibody for human cancer therapy. Proc. Natl. Acad. Sci. USA 89, 42854289.
40 Mulligan, R. C. and Berg, P. (1981) Selection for animal cells that express the Escherichia
40.
coli gene coding for xanthine-guanine phosphoribosyltransferase. Proc. Natl. Acad. Sci.
USA 78, 20722076.
41 Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Subcloning. In: Molecular Cloning:
41.
A Laboratory Manual. 2nd. Ed. Cold Spring Harbor Laboratory Press, Cold Spring Har-
bor, NY, pp. F.1F.11.
42 Southern, P. J. and Berg, P. (1982) Transformation of mammalian cells to antibiotic resis-
42.
tance with a bacterial gene under control of the SV40 early region promoter. J. Mol. Appl.
Genet. 1, 327341.
43. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Analysis and cloning of eukaryotic
genomic DNA. In: Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Har-
bor Laboratory Press, Cold Spring Harbor, NY, pp. 9.19.62.
44.
44 Challita-Eid, P. M., Penichet, M. L., Shin, S. U., et al. (1998) A B7.1-antibody fusion
protein retains antibody specificity and ability to activate via the T cell costimulatory
pathway. J. Immunol. 160, 34193426.
45 Dela Cruz, J. S., Trinh, K. R., Morrison, S. L., and Penichet, M. L. (2000) Recombinant
45.
anti-human HER2/neu IgG3-(GM-CSF) fusion protein retains antigen specificity and
cytokine function and demonstrates antitumor activity. J. Immunol. 165, 51125121.
46.
46 Penichet, M. L., Challita, P. M., Shin, S. U., Sampogna, S. L., Rosenblatt, J. D., and
Morrison, S. L. (1999) In vivo properties of three human HER2/neu-expressing murine
cell lines in immunocompetent mice. Lab. Anim. Sci. 49, 179188.
47.
47 Deans, R. J., Denis, K. A., Taylor, A., and Wall, R. (1984) Expression of an immunoglobulin
heavy chain gene transfected into lymphocytes. Proc. Natl. Acad. Sci. USA 81, 12921296.
48.
48 Foecking, M. K. and Hofstetter, H. (1986) Powerful and versatile enhancer-promoter unit
for mammalian expression vectors. Gene 45, 101105.
49 McLean, G. R., Nakouzi, A., Casadevall, A., and Green, N. S. (2000) Human and murine
49.
immunoglobulin expression vector cassettes. Mol. Immunol. 37, 837845.
50. Harlow, E. and Lane, D. (1989) Electrophoresis. In: Antibodies: A Laboratory Manual.
1st Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 635658.
51. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Commonly Used Techniques in
Molecular Cloning. In: Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring
Harbor Laboratory, Cold Spring Harbor, NY, pp. E.1E.39.
52. Harlow, E. and Lane, D. (1989) Storing and purifying antibodies. In: Antibodies: A Labo-
ratory Manual. 1st Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp.
285318.
53. Harlow, E. and Lane, D. (1989) Immunoaffinity purification. In: Antibodies: A Laboratory
Manual. 1st Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp. 511551.
54 Meininger, D. P., Rance, M., Starovasnik, M. A., Fairbrother, W. J., and Skelton, N. J.
54.
(2000) Characterization of the binding interface between the E-domain of Staphylococcal
protein A and an antibody Fv-fragment. Biochemistry 39, 2636.
55. Nelson, D. L., Kurman, C. C., and Serbousek, D. E. (1994) 51Cr release assay of anti-
body-dependent cell-mediated cytotoxicity (ADCC); In Current Protocols in Immunol-
ogy, Vol. 2. Coligan, J. E.,Kruisbeek, A. M., Margulies, D. H., Shevach, E. M., and Strober,
W., eds. John Wiley & Sons, New York, NY, pp. 7.27.17.27.8.
56.
56 Donohue, J. H. and Rosenberg, S. A. (1983) The fate of interleukin-2 after in vivo admin-
istration. J. Immunol. 130, 22032208.
57.
57 Kendra, K., Gan, J., Ricci, M., et al. (1999) Pharmacokinetics and stability of the ch14.18-
interleukin-2 fusion protein in mice. Cancer Immunol. Immunother. 48, 219229.
58 Harvill, E. T., Fleming, J. M., and Morrison, S. L. (1996) In vivo properties of an IgG3-IL-
58.
2 fusion protein. A general strategy for immune potentiation. J. Immunol. 157, 31653170.
59.
59 Becker, J. C., Pancook, J. D., Gillies, S. D., Mendelsohn, J., and Reisfeld, R. A. (1996)
Eradication of human hepatic and pulmonary melanoma metastases in SCID mice by anti-
body-interleukin 2 fusion proteins. Proc. Natl. Acad. Sci. USA 93, 27022707.
60.
60 Lode, H. N., Xiang, R., Dreier, T., Varki, N. M., Gillies, S. D., and Reisfeld, R. A. (1998)
Natural killer cell-mediated eradication of neuroblastoma metastases to bone marrow by
targeted interleukin-2 therapy. Blood 91, 17061715.
61.
61 Becker, J. C., Varki, N., Gillies, S. D., Furukawa, K., and Reisfeld, R. A. (1996) An anti-
body-interleukin 2 fusion protein overcomes tumor heterogeneity by induction of a cellu-
lar immune response. Proc. Natl. Acad. Sci. USA 93, 78267831.
62.
62 Sabzevari, H., Gillies, S. D., Mueller, B. M., Pancook, J. D., and Reisfeld, R. A. (1994) A
recombinant antibody-interleukin 2 fusion protein suppresses growth of hepatic human
neuroblastoma metastases in severe combined immunodeficiency mice. Proc. Natl. Acad.
Sci. USA 91, 96269630.
63. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Plasmid vectors. In: Molecular Clon-
ing: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor,
NY, pp. 1.11.110.
64.
64 Chen, T. T., Tao, M. H., and Levy, R. (1994) Idiotype-cytokine fusion proteins as cancer
vaccines. Relative efficacy of IL-2, IL-4, and granulocyte-macrophage colony-stimulat-
ing factor. J. Immunol. 153, 47754787.
65. Harlow, E. and Lane, D. (1989) Bacterial cell wall proteins that bind antibodies. In: Anti-
bodies: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY,
pp. 615623.
66.
66 Harvill, E. T. and Morrison, S. L. (1996) An IgG3-IL-2 fusion protein has higher affinity
than hrIL-2 for the IL-2R alpha subunit: real time measurement of ligand binding. Mol.
Immunol. 33, 10071014.
67.
67 Liu, Z., Gurlo, T., and von Grafenstein, H. (2000) Cell-ELISA using beta-galactosidase
conjugated antibodies. J. Immunol. Methods 234, P153167.
68. McKenzie, A. N. J. and Zurawski, G. (1994) Measurement of Interleukin-13. In: Current
Protocols in Immunology, Vol. 1. Coligan, J. E., Kruisbeek, A. M., Margulies, D. H.,
Shevach, E. M., and Strober, W., eds. John Wiley & Sons, New York, NY, pp. 6.18.1
6.18.5.
69. Gieni, R. S., Li, Y., and HayGlass, K. T. (1995) Comparison of [3H]-thymidine incorpo-
ration with MTT- and MTS-based bioassays for human and murine IL-2 and IL-4 analy-
sis. Tetrazolium assays provide markedly enhanced sensitivity. J. Immunol. Methods
187, 8593.
Synthesis and Evaluation of a PTX Immunoconjugate 375
24
Synthesis and Biological Evaluation
of a Paclitaxel Immunoconjugate
Ahmad Safavy and Kevin P. Raisch
Summary
Targeted cancer therapy is a promising strategy for the treatment of this disease. In this
approach, a cytotoxic agent (CA), such as a drug or a radionuclide, is attached, usually
covalently, to a targeting vehicle (TV), which in turn is capable of recognizing specific
receptor motifs on the surface of the tumor cells. Once administered systemically, the construct
would localize on the tumor through the TV moiety and would release the CA cargo, resulting
in the destruction of the malignant tissue. Small-molecule peptides as well as monoclonal anti-
bodies have been used as TVs. The synthesis, antigen binding, and cytotoxicity of a covalent
conjugate of the anticancer drug paclitaxel (taxol) to the anti-epidermal growth factor receptor
(EGFR) monoclonal antibody C225 (IMC-C225; Erbitux, ImClone Systems, Somerville, NJ)
are described in this chapter to illustrate the methods used for the construction and in vitro
evaluation of these conjugates.
Key Words: Paclitaxel, taxol; C225; monoclonal antibody; immunoconjugates; targeted
therapy; drug delivery; tumor targeting.
1. Introduction
Tumor-specific treatment of cancer has been greatly facilitated by the development
of monoclonal antibodies (MAbs) and through the use of intact, fragmented, or chemi-
cally derived MAbs. In this strategy, cell-surface antigens, overexpressed by the cells
of a malignant tissue, are targeted by the MAb molecules, which in turn tag the cell
for elimination by the organisms immune system (intact MAb or fragments), or
deliver an anticancer agent specifically to the site of the disease (immunoconjugates).
Because of the tissue-specific nature of the delivery process, a high tumor-to-normal-
tissue ratio would be expected in the target uptake of the CA, which may result in
higher antitumor efficacy and lower systemic toxicity and side effects.
Paclitaxel (PTX) is the first of the important taxane antitumor drugs; it has shown
activity against several types of human solid tumors (1,2), and has been in clinical use
since the early 1990s. The activity of this wide-spectrum drug against breast, ovarian,
and prostate cancers is impressive: response rates (RRs) of 40% to 60% have been
375
376 Safavy and Raisch
reported for PTX in cancer patients, with significant RRs in anthracyclin-resistant

disease (3). In cisplatin-treated ovarian cancer patients, PTX has shown good RRs in a
phase II study (3). In phase II and III clinical trials with non-small-cell lung cancer
patients, PTX demonstrated overall response rates of 10 to 36% (4).
Although the therapeutic advantages of taxanes warranted further preclinical and
clinical studies, some inherent shortcomings of these drugs could not be overlooked.
Prohibitively low water solubility (approx 0.25 g/mL) necessitated the addition of
allergenic excipients to the clinical formulations. On the other hand, high systemic
toxicity limited the level of the administered dose of these drugs. These, and the ulti-
mate interest of the medicinal chemist in designing more potent anticancer drugs
with concomitant preservation of the patient quality of life, have led to the design and
development of soluble tumor-targeting taxane derivatives (5).
The monoclonal antibody C225 is an anti-epidermal growth factor (EGF) receptor
monoclonal antibody that is currently in clinical trials (69). The EGF is a transmem-
brane protein of the tyrosine kinase growth factor family, and its receptor (EGFR)
(1012) is overexpressed in a number of human malignancies, including breast, ova-
rian, lung, and head and neck cancers (13,14). This overexpression of EGFR in differ-
ent human malignancies renders C225 an attractive model vehicle to study targeted
delivery of anticancer agents.
A PTX-peptide conjugate was reported to evaluate the tumor-targeted strategy for
this drug using a bombesin/gastrin-releasing peptide receptor-recognizing peptide
(15). An enhancement in the in vitro cytotoxic activity of PTX against H1299 human
non-small-cell lung cancer cells was observed in controlled experiments, which was
ascribed to the targeting mechanism of drug delivery by this conjugate. To further
expand the development of water-soluble tumor-targeting taxanes, we conjugated
PTX to C225 to build a model conjugate (Fig. 1) for evaluation of the potential of this
drug-delivery system (16). In this chapter we describe the details of the synthesis,
purification, identification, and in vitro evaluation of the PTXC225 conjugate as a
model protocol for preparation of paclitaxel-antibody covalent adducts.
2. Materials
Caution: The compounds and reagents listed below should be considered hazard-
ous to living organisms. In addition to the specific indications given below, care must
be taken in handling all of them. This may include, but may not be limited to, using
protective gear and working in a contained and ventilated space.
1. Paclitaxel (Toxic! Avoid inhalation and contact with skin, eyes, and respiratory system).
2. Monoclonal antibody C225 (may be immunogenic and/or allergenic on ingestion or
entering the circulation).
3. Phosphate buffer, 25 mM, pH 8.1.
4. Centricon-50 concentrator (Millipore, Bedford, MA).
5. Lowry reagents consisting of:
a. 2% Na2CO3 in 0.1 N NaOH.
b. 0.5% CuSO4 5H2O in 1% sodium potassium tartrate, adjusted to pH 7.0.
c. Phenol Follin-Ciocalteau reagent (Sigma Chemical, St. Louis, MO) diluted 1:3 by
distilled water.
6. Plate reader at 595 nm.
Synthesis and Evaluation of a PTX Immunoconjugate
377
Fig. 1. Schematic representation of a paclitaxel-antibody conjugate. The linkage is through the 2'-hydroxy group of the drug. Reprinted with
permission from ref. 16.
377
7. Anhydrous succinic anhydride.

8. Anhydrous pyridine (Aldrich Chemical, Milwaukee, WI).
9. Silica gel-60, 230400 mesh (Merck, Darmstadt, Germany).
10. Chloroform (HPLC grade).
11. Methanol (HPLC grade).
12. Low-pressure glass chromatography column.
13. Melting-point apparatus.
14. Anhydrous N,N-dimethyl formamide (DMF).
15. N-hydroxy succinimide (NHS).
16. 2-Ethoxy-1-ethoxycarbonyl-1,2-dihydroquinoline (EEDQ).
17. Dialysis tube or slide, 50,000 molecular-weight cut-off (MWCO).
18. Dulbeccos phosphate-buffered saline (DPBS).
19. A-431 human epidermoid cancer cells (CRL-1555, American Type Culture Collection
[ATCC], Manassas, VA) (see Note 1).
20. Dulbeccos modified Eagles medium (DMEM) (Invitrogen, Carlsbad, CA).
21. Phosphate-buffered saline (PBS): 2.7 mM KCl, 1.47 mM KH2PO4, 137 mM NaCl, 8 mM
Na2PO4, pH 7.4.
22. Fetal bovine serum (FBS) (HyClone, Logan, UT).
23. L-glutamine (200 mM, Invitrogen).
24. Trypsin/EDTA solution: 0.03% w/v trypsin, 0.01% w/v EDTA, 0.1% w/v glucose, 220 M
KH2PO4, 168 M Na2HPO4, 5.4 mM KCl, 137 mM NaCl, 0.006% w/v phenol red, pH 7.05.
25. Hemacytometer (Fisher Scientific, Pittsburgh, PA).
26. Coulter particle counter (Beckman Coulter, Fullerton, CA).
27. Tris-buffered saline with Tween-20 (TBS-T): 20 mM Tris-HCl (pH 7.5), 137 mM NaCl,
0.05% Tween-20.
28. EGF (20g/mL stock) (Sigma Chemical) (dissolve EGF in 10 mM acetic acid with 0.1%
BSA).
29. Lysis buffer: 0.025 M Tris-HCl (pH 7.5), 0.25 M NaCl, 0.005 M EDTA, 1% (v/v) NP-
40, 0.001 M phenylmethylsulfonylfluoride (PMSF), 2 g/mL aprotinin, 2 g/mL
leupeptin, 0.002 M Na orthovanadate, 0.05 M Na fluoride, and 0.03 M Na pyrophosphate
(see Note 2).
30. Eppendorf tubes, 1.5-mL, sterilized by autoclaving.
31. BCA Protein Assay (Pierce, Rockford, IL).
32. SDS-PAGE apparatus (Bio-Rad, Hercules, CA).
33. Trans-Blot system (Bio-Rad).
34. Immobilon-P membrane (Millipore, Bedford, MA).
35. Mouse monoclonal anti-EGFR, clone F4 (Sigma Chemical).
36. Mouse monoclonal antiphosphotyrosine, PY99 (Santa Cruz Biotechnology, Santa Cruz,
CA).
37. Sheep polyclonal antimouse IgG-HRP, NAV931 (Amersham Biosciences, Piscataway, NJ).
38. ECL Western Blotting Detection Kit (Amersham Biosciences).
39. BioMax X-ray film, 8 10 (Eastman Kodak, Rochester, NY).
40. BioMax film cassette (Eastman Kodak).
41. Photography darkroom.
42. X-ray film developing system.
43. Twenty-four-well cell-culture plates, 3524 (Corning, Corning, NY).
44. Isoton II diluent (Beckman Coulter).
45. Prism software (GraphPad Software, San Diego, CA).
46. Annexin V-FITC Apoptosis Detection Kit (BioVision Inc., Mountain View, CA).
47. Six-well cell-culture plates, 3516 (Corning Inc.).
48. Becton Dickinson FACSCalibur System (Becton-Dickerson, San Jose, CA).
49. CellQuest v. 3.1 software (Becton-Dickinson).
3. Methods
The methods shown in this chapter describe (1) preparation and purification of the
side-chain-activated PTX (2'-O- paclitaxel succinate, PTXSX); (2) synthesis and puri-
fication of the PTXC225 conjugate; (3) identification of the PTXC225 conjugate by
matrix-assisted LASER desorption/ionization time-of-flight mass spectrometry
MALDI-TOF MS; (4) cell growth inhibition screening on the conjugate; and (5)
apoptosis induction assay.
3.1. Preparation and Purification of 2'-O-Paclitaxel Succinate (PTXSX)
Since the antitumor activity of PTX is therapeutically significant only if the drug
is available in its native (i.e., chemically unbound) form, potential delivery strategies
may be those utilizing a prodrug mechanism. In this two-tier approach, the drug is
first delivered to the tumor as a covalent adduct to a TV followed by the timely cleav-
age of PTX inside or in the close vicinity of the tumor to exert its cytotoxic effect.
Due to the special structure of PTX and the type and accessibility of chemical
functionalities of the molecule, it was anticipated that esterification of one of the
hydroxy groups would be a logical plan. In an attempt to generate water-soluble
taxanes, Deutsch et al. had reported the synthesis of 2'-O-succinate of PTX in 1985
(17). Also, drug-release kinetics results from early work in our laboratory identified
this PTX derivative as a suitable candidate for the design of targeting peptide PTX
prodrugs (15). Thus, commercially available PTX and succinic anhydride were first
dried in a high-vacuum desiccator over phosphorus pentoxide for 12 h and at room
temperature. The dried PTX (1 eq) was then dissolved in anhydrous pyridine, and suc-
cinic anhydride (6 eq) was added in solid form and in several portions. The resulting
solution was incubated under argon in a sealed flask and at room temperature for 6 h.
Distillation of the solvent in vacuum afforded the solid product PTXSX that was
purified by column chromatography. In the procedure of Deutsch et al., a water/
acetone crystallization process had been used to purify this product. To achieve bet-
ter yields of recovery, we used silica gel (SiO2) column chromatography for this
purification. In this procedure, 20 g of SiO2 per g of the crude PTXSX was used in a
standard glass chromatography column with chloroform as solvent for column pack-
ing and product loading. A methanol gradient of 015% in chloroform was used to
elute the product. This procedure afforded near-quantitative yields of recovery and
high product purities. The identity of the PTXSX was established by comparing the
melting point and nuclear magnetic resonance spectroscopy characteristics of this
product to those reported in the literature (17) (see Note 1).
3.2. Synthesis of PTXC225
Introduction of the succinyl group to PTX generates a single carboxyl group as a
specific site for the antibody conjugation. A logical technique for conjugation of
Table 1
Experimental a Molar Drug-to-Antibody Ratios (D:A)
for PTXC225 Conjugates Prepared Under Different Conditions b
Reaction PTXSX:C225 c D:A (PBS d) D:A (carb e)
1 1.1 0 0
2 5.1 1.2 0.6 0
3 10.1 2.2 0.2 1.1 0.3
4 20.1 4.5 1.1 2.7 1.7
a Measured from MR = MWconjugate - MWC225/ MWPTXSX using
MALDI-measured MWs.
b Reprinted with permission from ref. 16.
c Reactant molar ratios
d Phosphate-buffered saline, 25 mM, pH 8.1
e Carbonate buffer, 25 mM, pH 9.1
this carboxyl group to the antibody, without interfering with other functions in the
molecule, would be through activated ester (AE) formation (18). In this strategy, the
endogenous amine groups of the antibody are selectively targeted for binding.
Although carboxylic AEs are known to react with hydroxy groups to form carboxy-
lic esters, this transformation occurs only in the presence of a nucleophilic catalyst
such as N,N-dimethyl 4-aminopyridine (19).
In the presence of both amine and hydroxyl groups, and in the absence of a catalyst,
an amide bond is preferentially formed within a typical reaction time. This property of
the AE chemistry excluded the possibility of intramolecular self-condensation of PTX.
To design a reproducible protocol for this coupling, the yields of product recov-
ery and the number of substituted drugs per C225 (the PTX:MAb ratio) were studied
under different conditions such as reaction times, stoichiometry, solvents, and tem-
peratures (Table 1).
Coupling of PTX-SX to the antibody was highly reagent selective. We examined
several coupling reagents routinely used in peptide synthesis for amide bond forma-
tion (20), and the progress and the extent of the reaction were followed by MALDI-
TOF MS (see Subheading 4.3.). Work in our (15) and other laboratories (21) has
shown that MALDI-TOF MS is a reliable and direct tool for the evaluation of conju-
gation reactions involving small molecules and antibodies or other large proteins (see
Subheading 4.3.). Of the several different reagents used, only the EEDQ/NHS com-
bination proved effective for this coupling. No significant increase in C225 MW was
noticed with other reagents when monitored by mass spectrometry. Thus, the PTXSX
(1 eq) was activated by the addition of NHS (1.1 eq) and EEDQ (1.2 eq) in dry DMF
(125 L/mol PTXSX) at 0C for 1 h. The DMF solution of the PTXSX activated
ester (containing 10 eq of PTXNHS) was then added at 0C to the solution of C225
(1 eq) in 25 mM (pH 8.1) PBS (300 L/mg C225) with gentle stirring. Stirring was
continued at this temperature for 1 h, at which time the formation of the conjugate
was confirmed by MALDI-TOF MS. The final product PTXC225 was purified by
dialysis in a cellulose membrane with a MW cut-off of 50 kDa and in DPBS (4 2 L).
The antibody content of the purified solution was measured by the standard method
of Lowry (22) (see Note 2).
3.3 Identification of the PTXC225 Conjugate

A direct and unambiguous identification of the PTX-MAb conjugate is critical. An
error in the measurement of the PTX:MAb ratio would result in errors in the calcula-
tion of the in vitro and in vivo doses, leading in turn to further complications such as
false-negative therapeutic indices or false-positive toxicities. Access to a direct method
of measurement of this ratio, therefore, will have obvious advantages (21).
The use of MALDI-TOF MS provides direct MW analysis, which yields a clear
measure of the degree of substitution in real time (23) and has been used in the conjuga-
tion of small molecules to antibodies. In the development of a protocol for the synthesis
of unprotected hydroxamic acids (HA)-MAb conjugates, we reported the application of
MALDI-TOF MS to the optimization of the reaction conditions and direct measure-
ment of the HA:MAb ratio for the bifunctional chelating agent trisuccin (24).
The PTXC225 described here, as well as other PTX-MAb conjugates not shown,
were identified, and their drug-to-antibody (D:MAb) molar ratios were determined by
the use of MALDI-TOF MS technique. Once the MWs are measured, the D:MAb is
readily calculated through Eq. 1.
MWPTXC225 MWC225
D : MAb = (1)
MWPTXSX
where MWPTXC225 is the molecular weight of the purified final product conjugate, and
MWC225 and MWPTXSX are those of the unconjugated C225 and PTXSX, respectively.
A typical increase in the MW of the MAb as a result of PTX conjugation is shown
in Fig. 2. In this conjugate, a measured molecular weight increase of 2925 Da refers to
a PTX:C225 ratio of 3 as calculated by Eq. 1.
The PTX-C225 coupling was studied under different conditions before selecting
the conditions described above. The results are shown in Table 1.
3.4. Tumor Cell Line and Maintenance

The A-431 cell line is maintained in Dulbeccos modified Eagles medium (DMEM)
containing 10% fetal bovine serum (FBS), supplemented with 2 mM L-glutamine and
incubated at 37C in 5% CO2 (see Note 3).
Cells are subcultured (split) every 57 d, at a ratio of 1:6.
1. The culture medium is removed, and the cell monolayer is washed once with 510 mL
PBS (see Note 4).
2. The cells are removed from the surface of the tissue-culture flask by adding trypsin/EDTA
solution (use 1 mL for a T-25 flask or 3 mL for a T-75 flask).
3. The cells are incubated for 515 min at 37C.
4. After the cells are released from the surface, the suspension is diluted using culture
medium three times the trypsin/EDTA volume.
5. One-sixth of the cells are transferred to a new flask, and the remaining cells are counted
for use in the following experiments (see Note 5).
382
Fig. 3. Immunoblot assay to demonstrate the inhibition of EGFR kinase activity. (Top) A
24-h exposure of A-431 cells to C225 or PTXC225 had no effect on the receptor levels.
(Bottom) EGF-activated EGF tyrosine kinase activity was reduced to the same extent by
both C225 and PTXC225. The immunoblots were screened with either anti-EGFR (top) or
anti-phosphotyrosine (bottom) antibodies. Antibody-free cells were used as controls. Sample
cells were treated with equimolar concentrations of C225 and PTXC225. Reprinted with per-
mission from ref. 16, copyright 2003 American Chemical Society.
3.5. Measurement of EGFR Expression and Receptor

3.5.1. Phosphorylation
In order to determine whether conjugation of PTX to C225 does not interfere with
the ability of the antibody to bind to the corresponding antigen (i.e., EGFR), the
conjugates ability to block EGF-induced EGFR tyrosine kinase activity is examined
by an immunoblot assay as shown in Fig. 3 (16).
1. A-431 cells are seeded in 6-well plates at a density of 100,000 cells/well in 3 mL culture
medium per well and are allowed to grow for 48 h.
2. The cells are treated at a concentration of 1 g/mL with C225 or PTXC225 for 24 h and
are exposed to 10 nM (60 ng/mL) EGF for 5 min at 37C (see Note 6).
3. The cell-culture plates are immediately placed on ice and the cell monolayers are washed
three times with ice-cold PBS.
4. The cellular proteins are solubilized in the lysis buffer for 1015 min on ice (see Note 7).
5. Enough lysis buffer is used to cover the surface of the cells, i.e., 400 L of buffer per well
in a 6-well tissue-culture plate.
Fig. 2. (opposite page) MALDI-TOF mass spectra of C225 (B) and PTXC225 conjugate
(C). The MW increase confirms the addition of PTX. BSA (MW 66 kDa) is used as a MW
calibrator. Peaks A and D are the doubly charged molecular ions showing the half-MWs of B
and C, respectively. Reprinted with permission from ref. 16, copyright 2003 American
Chemical Society.
6. Four hundred L of the cellular lysate are collected into a sterilized 1.5-mL Eppendorf
tube.
7. The cellular lysate is clarified by centrifugation at 12,000g, for 15 min at 4C, and is
transferred into a 1.5-mL Eppendorf tube for storage at 20C.
8. The protein concentration is quantitated in the clarified cellular lystates using a BCA
protein assay kit. The lysis buffer is used as a blank and for diluting the protein standards
(see Note 8).
9. Equal amounts of protein (5 g) are loaded per lane and separated by 10% SDS-PAGE.
Separated proteins are then transferred to Immobilon-P membrane using a Trans-Blot
system (as described by the manufacturer).
10. The immunoblot is blocked in 10% milk-tris-buffered saline with Tween-20 (TBS-T), for
1 h at room temperature, or overnight at 4C.
11. The blocking step and all the subsequent incubations and washes are done using a rocker
or shaker with slow agitation:
12. The blocking buffer is removed, and primary antibody, anti-EGFR, or anti-phosphoty-
rosine is diluted 1:2000 in 3% milk-TBS-T. The blot is incubated at room temperature for
2 h or overnight at 4C.
13. The primary antibody is removed and the blot is washed at room temperature with TBS-T
(three times for 510 min each).
14. The secondary antibody, anti-mouse IgG-HRP antibody, is diluted 1:2000 in TBS-T and
incubated with the blot for 1 h at room temperature.
15. The blot is washed three times at room temperature (510 min each wash) with TBS-T.
16. The blot is developed using the ECL Western Blot detection system according to the
manufacturers protocol.
17. The blot is wrapped in a single layer of plastic wrap and the package is placed in the film
cassette. The film is then applied to a blot in the darkroom and is developed after being
exposed to the blot for 1 min and 3 min. Based on the results, longer or shorter exposure
times must be determined to obtain the optimal exposure.
18. To reprobe the immunoblot with a different primary antibody, the immunoblot is incubated
in 7 M guanidine hydrochloride for 30 min at room temperature followed by washing four
times with sterile H2O. The immunoblot is blocked and probed as described above.
3.6. In Vitro Cell Proliferation Assay

A cell-proliferation assay is used to study the cytotoxic effect of PTXC225 to A-431
cells. Equal molar concentrations of the conjugate, C225 alone, PTX alone, or the com-
bination of the two unconjugated molecules, are compared (see Note 9) (16).
3.6.1. Determination of IC50 Dose

Prior to analyzing PTXC225 in cell proliferation and apoptosis assays, the IC50
dose must be determined. The IC50 dose is the concentration of the agent resulting in
50% cell growth inhibition as compared to the untreated control cells. To be statisti-
cally significant, the IC50 dose should be the average of at least three independent
experiments:
medium and allowed to grow for 48 h.
2. On d 0, the cells are treated with PTXC225 in varying concentrations (110 nM) for 24 h.
Filter-sterilized PTXC225 is diluted in culture medium to a concentration of 1 M.
One to ten L of this solution is then dispensed into each well containing 1 mL of the
culture medium to obtain a drug concentration of 110 nM.
3. After adding the drug, the plate is rocked for 12 min to completely mix the drug into the
culture medium. The plate is then incubated at 37C for 24 h.
4. The drug is removed on d 1 by aspiration, and each well is washed once with 1 mL PBS.
The PBS is aspirated off, and 1 mL drug-free medium is added to each well. The plate is
incubated at 37C for an additional 3 d.
5. The cells are counted on d 4 by aspirating the culture medium from each well, washing
once with 1 mL PBS/well, and adding 0.5 mL of trypsin-EDTA/well. After the cells are
released from the surface, the 0.5 mL of trypsinized cells is diluted into 9.5 mL Isoton II
and is counted using a Coulter particle counter: The untreated cells are normalized to
100%, to which base the percent-survival of each treated group is compared. Analyses are
done using Prism software.
6. The IC50 dose is determined by plotting the normalized cell counts versus the drug dose
and computing the dose that inhibits 50% of cell proliferation using a linear quadratic
equation.
3.6.2. Determination of Cytotoxicity
1. A-431 cells are seeded in a 24-well plate at a density of 10,000 cells/well in 1 mL culture
medium and allowed to grow for 48 h at which time (d 0) the cells are treated with
C225 (3.2 nM), PTX (6.4 nM), C225 + PTX (0.32 nM + 0.64 nM), or PTXC225 conjugate
(3.2 nM) for 24 h. All drugs are diluted in culture medium to a concentration of 1 M and
filter sterilized. The wells are spiked with the corresponding agents and are incubated
with rocking at 37C.
2. The drug is removed on d 1 by aspiration, and each well is washed once with 1 mL PBS
followed by the addition of 1 mL drug-free medium, and incubation at 37C for an addi-
tional 3 d.
3. The cells are counted on d 4 and normalized to the percentage of untreated cells as
described above in Subheading 3.6.1., step 5.
3.7. Measurement of Apoptotic Cell Death

Detection of early stages of apoptosis was determined using an annexin V-FITC
apoptosis detection kit as described previously by Safavy et al. (16):
medium per well and allowed to grow for 48 h.
2. On day 0, A-431 cells are treated with each agent for 24 h as described in Subheading
3.6.2.: PTX (6.4 nM), C225 (3.2 nM), PTX + C225 (6.4 nM + 3.2 nM), or PTXC225
conjugate (3.2 nM).
3. The drug is removed after 24 h by aspiration, and each well is washed once with 3 mL
PBS per well followed by the addition of 3 mL drug-free medium to each well. The plate
is incubated at 37C for an additional 3 d.
4. On d 4, the culture medium is collected into a 12 75 mm tube, and a 1-mL PBS wash is
added to each respective tube containing the culture medium.
5. The cell monolayer is removed by adding 0.5 mL typsin-EDTA and incubating the cells
at 37C. The trypsinized cells are added to the respective tube containing culture medium
and PBS wash.
6. The cells are separated from the medium by centrifugation at 400g for 5 min.
7. The cells are washed once with 3 mL ice-cold PBS.
8. After centrifugation, the PBS is removed and the cells are incubated in binding buffer
containing annexin V-FITC and PI as described by the manufacturers protocol.
9. The data is collected using a Becton-Dickinson FACSCalibur system and analyzed using
CellQuest v.3.1 software.
4. Notes
1. Due to the possibility of damaging the drug activity, heating of the PTX during the drying
process should be avoided. We found that removing moisture over P2O5 at room tempera-
ture and in high vacuum is sufficient.
The purity and dryness of the pyridine used is also very important. This highly hygro-
scopic liquid may be purified by storage over sodium hydroxide followed by distillation
from barium oxide. Storage over 4 molecular sieve under dry argon or nitrogen is
recommended.
2. It was found during the course of this study that conjugation of up to three PTX molecules
per molecule of the antibody completely preserves the solubility of the MAb in aqueous
media at concentrations normally used for in vitro and in vivo experiments. For this work,
we used conjugate solutions within a C225 concentration range of 0.252.2 mg/mL. Fur-
thermore, we have prepared completely soluble PTXC225 solutions for therapeutic experi-
ments in animal models containing up to 10 mg/mL of the conjugate. Conjugate solubility,
however, seemed to be limited by the PTX:C225 ratio. Conjugates with a PTX:C225 of
>3 caused either immediate or delayed loss of the protein in the form of a heavy precipi-
tate. At the same time, a PTX:C225 of 2 seemed to be sufficient for effective cell killing as
indicated by the in vitro cytotoxicity experiments described in Subheading 3.2.
3. A-431 cells were used in this protocol, because they are a very well studied EGFR-
overexpressing cell line (24). However, while many other cell lines can be used, the initial
cell-plating numbers may have to be adjusted according the growth characteristics of each
particular cell line.
4. All cell-culture manipulations are carried out inside a biological hood. Culture medium
and buffers containing glucose are 0.22-m filter sterilized, and all other buffers are ster-
ilized by autoclaving. Antibodies, antibody-drug conjugates, and chemotherapy drugs are
0.22-m filter sterilized prior to use.
5. For cell counting, a particle counter by Beckman Coulter is very helpful due to the volume
of samples to be counted. To set up the Beckman Coulter counter, first count the cells
using a hemacytometer, following the manufacturers protocol. The cell suspension may
have to be diluted 1:10 or 1:100 in PBS in order to have an accurate count. After the
number of cells per milliliter have been determined, five identical dilution vials are set up
for the Beckman Coulter counter. Each vial can be used for three or four cell counts. Cells
are counted at a low-threshold setting. Between cell counts, the lower threshold is adjusted
up or down and the cells are counted again. Adjusting the lower threshold and cell count-
ing is continued until the counter has accurately counted the same number of cells/mL as
determined by the hemacytometer. This lower threshold setting should be recorded if it is
to be used for counting this cell line. The lower threshold setting must be determined for
each cell line used in these experiments.
6. EGF is added quickly to each well in a biological hood and mixed into the culture me-
dium by gently making a figure-eight motion with the tissue-culture plates. Then the plate
is incubated at 37C for no more than 5 min. Reports have indicated that between 5 and
10 min after exposure to a high concentration of EGF, the EGF receptor begins to be
internalized and degraded.
7. All the ingredients for the lysis buffer are to be mixed together except for the inhibitors
leupeptin, Na-orthovanadate, aprotinin, and PMSF. The buffer should be autoclaved and
stored at 4C. On the d the lysis buffer is to be used, add the inhibitors to an aliquot of the
buffer and PMSF, and store on ice.
8. Optimally, protein concentrations should be determined on the cellular lysates prior to
storage at 20C. However, this may not be practical if lysates are collected at different
time points. Therefore, the cellular lysates can be quickly thawed at 37C, and then placed
on ice until used for the protein assay. If a precipitate is visible after thawing, the incuba-
tion of the lysates should be continued at 37C followed by vortexing until the precipitate
becomes soluble.
9. Due to solubility of PTX in aqueous solutions, the stock solution of the drug is prepared
in high-purity DMSO. The amount of free PTX should be calculated from the measured
PTX:C225 in the conjugate. A PTX:C225 of 2 translates into 6.4 nM of free PTX at a
conjugate concentration of 3.2 nM with respect to C225.
References
1.
1 Rowinsky, E. K. and Donehower, R. C. (1995) Paclitaxel (taxol). N. Engl. J. Med. 332,
10041014.
2. Donehower, R. C., Rowinsky, E. K., Grochow, L. B., Longnecker, S. M., and Ettinger, D.
S. (1987) Phase I trial of taxol in patients with advanced cancer. Cancer Treat. Rep. 71,
11711177.
3. Khayat, D., Antoine, E. C., and Coeffic, D. (2000) Taxol in the management of cancers of
the breast and the ovary. Cancer Invest. 18, 242260.
4. Goldspiel, B. R. (1997) Clinical overview of the taxanes. Pharmacotherapy 17, 110S
125S.
5. Safavy, A. (2001) Taxane derivatives for targeted therapy of cancer. US patent no.
6.191.290.
6. Sridhar, S. S., Seymour, L., and Shepherd, F. A. (2003) Inhibitors of epidermal-growth-
factor receptors: a review of clinical research with a focus on non-small-cell lung cancer.
Lancet Oncol. 4, 397406.
7. Kies, M. S. and Harari, P. M. (2002) Cetuximab (Imclone/Merck/Bristol-Myers Squibb).
Curr. Opin. Investig. Drugs 3, 10921100.
8. Goldstein, N. I., Prewett, M., Zuklys, K., Rockwell, P., and Mendelsohn, J. (1995) Bio-
logical efficacy of a chimeric antibody to the epidermal growth factor receptor in a human
tumor xenograft model. Clin. Cancer Res. 1, 13111318.
9. Naramura, M., Gillies, S. D., Mendelsohn, J., Reisfeld, R. A., and Mueller, B. M. (1993)
Therapeutic potential of chimeric and murine anti-(epidermal growth factor receptor) an-
tibodies in a metastasis model for human melanoma. Cancer Immunol. Immunother. 37,
343349.
10.
10 Thompson, D. M. and Gill, G. N. (1985) The EGF receptor: structure, regulation and
potential role in malignancy. Cancer Surv. 4, 767788.
11.
11 Schlessinger, J. (1988) The epidermal growth factor receptor as a multi-functional allos-
teric protein. Biochem. 27, 31193123.
12.
12 Carpenter, G. and Cohen, S. (1979) Epidermal growth factor. Ann. Rev. Biochem. 48,
193216.
13.
13 Pathak, M. A., Matrisian, L. M., Magun, B. E., and Salmon, S. E. (1982) Effect of epider-
mal growth factor on clonogenic growth of primary human tumor cells. Int. J. Cancer 30,
745750.
14.
14 Singletary, S. E., Baker, F. L., Spitzer, G., et al. (1987) Biological effect of epidermal
growth factor on the in vitro growth of human tumors. Cancer Res. 47, 403406.
15.
15 Safavy, A., Raisch, K. P., Khazaeli, M. B., Buchsbaum, D. J., and Bonner, J. A. (1999)
Paclitaxel derivatives for targeted therapy of cancer: toward the development of smart
taxanes. J. Med. Chem. 42, 49194924.
16.
16 Safavy, A., Bonner, J. A., Waksal, H., et al. (2003) Synthesis and biological evaluation of
paclitaxel-C225 conjugate as a model for targeted drug delivery. Bioconjug. Chem. 14,
302310.
17.
17 Deutsch, H., Glinski, J., Hernandez, M., et al. (1989) Synthesis of congeners and prodrugs:
water soluble prodrugs of taxol with potent antitumor activity. J. Med. Chem. 32, 788792.
18.
18 Wilbur, D. S. (1992) Radiohalogenation of proteins: An overview of radionuclides, label-
ing methods, and reagents for conjugate labeling. Bioconjug. Chem. 3, 433470.
19.
19 Neises, B. and Steglich, W. (1978) Simple method for the esterification of carboxylic
acids. Angew. Chem. Int. Enl. 17, 522523.
20. Lloyd-Williams, P., Albericio, F., and Giralt, E. (1997) Chemical Approaches to the Syn-
thesis of Peptides. CRC, Boca Raton, FL.
21. Siegel, M. M., Hollander, I. J., Hamann, P. R., et al. (1991) Matrix-assisted uv-laser des-
orption/ionization mass spectrometric analysis of monoclonal antibodies for the determi-
nation of carbohydrate, conjugated chelator, and conjugated drug content. Anal. Chem.
63, 24702481.
22. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) Protein measure-
ment with the folin phenol reagent. J. Biol. Chem. 193, 265275.
23. Fenselau, C. (1997) MALDI MS and strategies for protein analysis. Anal. Chem. 69,
661A665A.
24. Saleh, M. N., Raisch, K. P., Stackhouse, M. A., et al. (1999) Combined modality therapy
of A-431 human epidermoid cancer using anti-EGFr antibody C225 and radiation. Cancer
Biother. Radiopharm. 14, 451463.
Cytotoxic Tumor Targeting 389
25
Cytotoxic Tumor Targeting
With scFv Antibody-Modified Liposomes
Cornelia Marty and Reto A. Schwendener
Summary
Specific targeting of liposome-formulated cytotoxic drugs or antigens to receptors expressed
selectively on target cells represents an effective strategy for increasing the pharmacological
efficacy of the delivered molecules. We have developed a feasible technique to selectively
attach antibodies and fragments thereof, but also small-mol-wt ligands such as peptides, carbo-
hydrates, or any molecules that recognize and bind target antigens or receptors to the surface of
small unilamellar liposomes. Our concept is based on the site-specific functionalization of the
ligands to be attached to the liposomes by thiol groups. These thiol groups can easily be intro-
duced to antibodies or peptides by addition of cysteines, preferably at sites that do not interfere
with the receptor binding domains. Optimally, the site-specific modification is introduced at
the C-terminal end of the ligand, separated by an inert spacer sequence located between the
thiols and the specific part of the ligand. The thiol-reactive molecules on the liposome surface
are maleimides that are linked to phospholipids composing the liposome bilayer membrane.
We illustrate the coupling method of a functionalized single-chain antibody fragment with bind-
ing specificity to ED-B fibronectin, an isoform of fibronectin exclusively expressed in tumor
tissues, to long circulating small unilamellar poly(ethylene glycol) liposomes.
Key Words: Ligand-mediated liposome targeting; site-directed ligand modification; ligand
functionalization; liposomes; immunoliposomes; thiol-maleimide coupling; cysteine modifi-
cation; single-chain antibody fragment; peptides; targeted tumor therapy; vaccines.
1. Introduction
Antibody or small-molecule ligand-mediated targeting of liposome-encapsulated
anticancer or antiangiogenic drugs to receptors expressed selectively on tumor and/or
tumor endothelial cells represents an effective strategy for increasing the therapeutic
efficacy of the liposomal drugs. Small unilamellar ligand-targeted liposomes are
intended to increase the specificity of interaction with target cells and to increase
the amount of drug delivered to these cells (1,2). We have developed a feasible tech-
nique to attach antibodies and fragments thereof, but also small-mol-wt ligands such
as peptides, carbohydrates, or any molecules that selectively recognize and bind to
389
390 Marty and Schwendener
target antigens or receptors, to the surface of small unilamellar liposomes (SUV) (3).
The random introduction of reactive groups to a proteine.g., via amino group modi-
ficationleads to uncontrollable attachment and random orientation of the attached
molecules on the liposome surfaces, causing decrease of binding efficiency. By intro-
ducing reactive groups at a distinct site on the ligand molecule, the orientation on the
liposome surface can be defined. Our concept is based on the site-specific func-
tionalization of the ligands to be attached to the liposomes by thiol groups. These
thiol groups can easily be introduced onto antibodies or peptides by addition of cys-
teines, preferably at sites that do not interfere with the receptor binding domains.
Optimally, the site-specific modification is introduced at the C-terminal end of the
ligand, separated by an inert spacer sequence located between the thiols and the spe-
cific part of the ligand. The thiol-reactive molecules on the liposome surface are
maleimides that are linked to phospholipids composing the liposome bilayer mem-
brane. We illustrate the coupling method of a functionalized single-chain antibody
fragment (scFv) with binding specificity to ED-B fibronectin, an isoform of fibronec-
tin exclusively expressed in tumor tissues, to long circulating small unilamellar
poly(ethylene glycol)-liposomes (PEG-liposomes).
2. Materials
2.1. Liposome Preparation and Modification
1. Soy phosphatidylcholine (SPC) (L. Meyer GmbH, Hamburg, Germany).
2. Cholesterol Fluka (Buchs, Switzerland) (see Note 1).
3. D,L--tocopherol (Merck, Darmstadt, Germany).
4. 2-Dipalmitoyl-sn-glycero-3-phosphatidylethanolamine (DPPE) (Sygena, Liestal, Swit-
zerland).
5. Methoxy-poly(ethylene glycol)-phosphatidylethanolamine (PE-PEG-OMet) (Sygena).
6. Amino-poly(ethylene glycol)-phosphatidylethanolamine (PE-PEG-NH2) (Shearwater
Polymers, Enschede, The Netherlands).
7. 3,3'-dioctadecyloxacarbocyanine perchlorate (DiO) (Molecular Probes, Leiden, The Neth-
erlands).
8. Phosphate buffer (PB): 13 mM KH2PO4, 54 mM NaHPO4, pH 7.4.
9. Sulfosuccinimidyl N-(4-carboxycyclohexylmethyl)maleimide (sulfo-SMCC) (Pierce,
Rockford, IL).
10. Round-bottom flasks (25250 mL).
11. Sodium cholate.
12. Rotatory evaporator, e.g., Rotavap (Buechi AG, Flawil, Switzerland).
13. Mini-Lipoprep instrument (Harvard Apparatus Limited, Edenbridge, Kent, UK).
14. Lipex extruder (Lipex Biomembranes, Vancouver, Canada).
15. Nuclepore membranes of defined pore sizes: 0.4, 0.2, 0.1 m (Whatman, Maidstone,
Kent, UK).
16. Dialysis tubing, 10,000 mol wt Cut off.
17. Fluorescence spectrometer.
2.2. Antibody Coupling

1. HBSE buffer: 10 mM HEPES, 150 mM NaCl, 9.1 mM EDTA, pH 7.5.
2. Tributylphosphine (Fluka, Buchs, Switzerland).
3. Metrizamide (Sigma , St. Louis, MO).

4. Ultracentrifuge tubes (Beckmann).
5. Ultracentrifuge.
2.3. In Vitro Binding of scFv-Liposomes to Tumor Cells

1. Cell-culture medium (DMEM; Gibco-BRL, Basel, Switzerland) supplemented with 10%
fetal bovine serum (FBS), 1% L-glutamine, 1% nonessential amino acids, 1 mM sodium
pyruvate, 100 U/mL penicillin, and 0.1 mg/mL streptomycin (all from Gibco-BRL).
2. Phosphate buffered saline (PBS): 137 mM NaCl, 2.68 mM KCl, 8.09 mM Na2HPO4, 1.76
mM KH2PO4, pH 7.4.
3. Glass cover slips (20-mm diameter).
4. Collagen I from rat tails (Sigma) or extracted from rat tails (see Note 2).
5. Glycerin 10%.
6. Fluorescence microscope.
3. Methods
The methods described below outline (1) the preparation of liposomes by detergent
dialysis and filter extrusion; (2) the liposome modification with maleimide molecules;
(3) the attachment of functionalized scFv antibody fragments to liposomes via thiol-
to-maleimide coupling; (4) the separation of liposomes from unbound scFv fragments;
and (5) the in vitro binding of scFv-modified liposomes to cancer cells.
3.1. Liposome Preparation
In the past 30 yr a large number of methods of liposome preparation have been
developed and refined. For comprehensive information, we refer to the corresponding
literature. We use the two methods described below because of their ease and versatil-
ity, as well as the high quality of the liposomes they produce.
3.1.1. General Liposome Compositions
Liposomes can be composed of a large selection of phospholipids and additional
lipophilic compounds, such as cholesterol, poly(ethylene glycol) lipids (PEG), and
antioxidants. Depending on the intended application of the liposomes, different lipid
compositions have to be selected. The state-of-the-art liposomes used for intrave-
nous applicationse.g., liposomes carrying cytotoxic antitumor drugsare those
composed of lipids containing hydrophilic carbohydrates or polymers, usually a
poly(ethylene glycol)-modified phopholipid. Such PEG- or stealth liposomes evade
fast absorption in the mononuclear phagocyte system (1). Liposome formulations car-
rying antigens intended as vaccines are administered by subcutaneous or intradermal
injection and do not require PEG modification, since the targets are phagocytosing
cellsmacrophages and dendritic cells localized at the site of injection (4). The lipid
compositions used for the preparation of conventional scFv-liposomes and pegylated
scFv-liposomes are given in Table 1.
3.1.2. Model Calculations of Numbers of Liposomes and Reactive Groups
Based on experimentally determined mean hydrodynamic diameters of the lipo-
somes (see Note 3) and from assumptions on vesicle geometry parameters made by
392
Table 1
Lipid Composition in Milligrams per Milliliter of the Liposome Types A, B, C, D
Type Liposomes SPC a Cholesterol DPPE DPPE-PEG-Omet DPPE-PEG-NH2
A Conventional control liposomes 40 mg 4 mg

100 mol% 20 mol%
52 mol 10 mol
B Conventional liposomes for modification 40 mg 4 mg 2.52 mg
100 mol% 20 mol% 7 mol%
52 mol 10 mol 3.5 mol
392
C Control PEG liposomes 40 mg 4 mg 9.8 mg

100 mol% 20 mol% 7 mol%
D PEG liposomes for modification 40 mg 4 mg 9.8 mg
100 mol% 20 mol% 7 mol%
Marty and Schwendener

a The compositions of the lipid mixtures are always calculated by taking the main lipid component (SPC, or other lipids) as 100% in moles. The
additional components are given as mole percents referred to the main lipid. To lipid mixtures containing unsaturated fatty acids antioxidants such
as D,L--tocopherol are added at 0.10.2 mol% (see Note 3).
Table 2
Calculated Numbers of Liposomes As a Function
of Experimentally Measured Mean Diameters
Mean Phospholipid molecules Liposomes per 1 mg Trapped volume per 1 mg
diameter (nm) per liposome phospholipid (E13) phospholipid (microliters)
25 4,250 18.400 0.52

50 19,930 3.930 1.60
80 54,000 1.450 2.90
100 86,070 0.911 3.80
120 125,500 0.625 4.70
200 357,500 0.219 8.20
Huang and Mason (5), it is possible to approximate liposome numbers per volume
and, e.g., the numbers of amino groups available on one liposome of a given composi-
tion and mean diameter. In Table 2, examples of calculations based on mean liposome
sizes of 25 to 200 nm are listed. For example, a liposome with a mean diameter of
50 nm contains about 20,000 SPC molecules, based on the volume of one SPC mol-
ecule of 1.253 nm3. Thus, the same given amount of lipid in mg/mL yields more than
four times more liposomes with a diameter of 50 nm (1.6 1015/mL) than liposomes
having a diameter of 100 nm (3.6 1014/mL; see Table 2).
A graphic plot of the liposome diameters (nm) vs the numbers of liposomes formed
with 1 mg phospholipid/mL yields an exponential curve. After further transformations
the following equation is obtained:
Number of liposomes/mL = a d b c
where a = 1.594 1017 (per nm/mg), b = 2.118, c = phospholipid concentration of
the liposome solution (mg/mL), and d = mean diameter of the prepared liposomes
(nm).
Under the assumption that all other additional membrane-forming molecules
occupy the same volume of 1.253 nm3 in the bilayer, and that all molecules are evenly
distributed in both lipid monolayers, liposomes containing, for example, 2 mol% amino
groups with a mean diameter of 50 nm carry approx 400 amino groups distributed over
the bilayer membrane. Hence, 200 amino groups are located on the outer monolayer
surface, available for chemical modification. Consequentlyas shown in Table 3 for
liposomes of 40 mg SPC/mL (see Table 1) and diameters of 50 and 100 nm, and
considering losses during productionthe vesicles are composed of different num-
bers of lipid molecules and reactive groups. A liposome containing 7 mol% amino
groups and having a diameter of 50 nm carries on average 640 amino groups on its
surface, whereas on a larger liposome of 100 nm, approx 2800 amino groups are avail-
able for modification.
3.1.3. Liposome Preparation by Dialysis
Small unilamellar vesicles (SUV) of 50200 nm mean size can be prepared using
the detergent dialysis method (6). This method is based on the controlled removal of
394
Table 3
Calculation of Reactive Groups on Liposomes Prepared As Listed in Table 1
Diameter NH2-groups Initial concentration Calculated numbers Yield after modification Yield after modification
(nm) (mol%) (NH2-groups/mL) (NH2-groups/lip)a (maleimide groups/mL) b (maleimide groups/lip) c
50 15 2 3.1 1017 200 1.6 1017 120
394
7 1.0 1018 640 5.1 1017 400

100 15 2 3.1 1017 850 1.6 1017 510
7 1.0 1018 2800 5.1 1017 1670
a lip,
liposome.
b Yield calculation: [initial lipid concentration] 0.95 0.6 0.9 = 0.513 (based on yields of liposome preparation [95%], yield of sulfo-SMCC
modification calculated from lipid-NH2-group labeling with BODIPY [60%] and loss or dilution of lipid during dialysis [90%]).
c Corresponding to a sulfo-SMCC modification efficiency of 60 % at a 5X molar excess of sulfo-SMCC (see Subheading 3.1.6.).
Marty and Schwendener

detergent from mixed lipid/detergent micelles. The compositions of the lipid mixtures
are calculated by taking the main bilayer-forming lipid component (SPC, or other lip-
ids) as 100% in moles. The additional components are calculated as mole percents
referred to the main lipid:
1. For liposome preparation, a given lipid mixture (SPC, 40 mg/mL; cholesterol, 4 mg/mL,
20 mol%; DPPE, 2.5 mg, 7 mol%; D,L--tocopherol, 0.2 mg, 1 mol%; see Table 1) is
dissolved in methanol/methylene chloride (1:1 v/v) in a round-bottom flask (see Notes 4
and 5).
2. Sodium cholate (see Note 6) at a ratio of total lipids, including all lipophilic molecules, to
detergent of 0.6 moles is added.
3. This mixture is evaporated at 40C to dryness (3060 min) using a rotatory evaporator.
4. The dry lipid/detergent film is dispersed in PB and left 3060 min at room temperature
for equilibration.
5. Dialyze the formed mixed lipid/detergent micelles against 35 L of PB (volume ratio = 1
to 1000) for 1215 h at room temperature, e.g., using a Mini-Lipoprep instrument (see
Notes 7 and 8).
3.1.4. Liposome Preparation by High-Pressure Filter Extrusion

SUVs can also be prepared by sequential filter extrusion of multilamellar liposome
dispersions in PB through Nucleopore membranes (Sterico, Dietikon, Switzerland) of
0.4, 0.2, 0.1, and 0.05 m pore diameter with a Lipex extruder (Lipex Biomembranes
Inc., Vancouver, Canada) (7). Using this method, lipid films without detergents are
prepared as described in Subheading 3.1.3. and dispersed in PB, followed by sequen-
tial extrusion through Nucleopore membranes of decreasing pore sizes (see Notes 9
and 10).
3.1.5. Fluorescence Labeling of Liposomes and scFv Antibodies
The lipophilic dye DiO is used for the labeling of liposomes. DiO dissolved in
methanol/methylene chloride (1:1, v/v) is added at 0.20.4 mg/mL (0.20.4 mol%) to
the organic lipid mixture. To determine the initial lipid concentration, aliquots of the
lipid dispersion are taken before dialysis or extrusion. These aliquots and those taken
after liposome preparation are measured in a fluorescence spectrometer for the deter-
mination of the lipid concentrations after liposome preparation. Functionalized anti-
bodies or peptides can be trace labeled with fluorescent dyes such as Alexa Fluor or
BODIPY succinimidyl ester (Molecular Probes), allowing unspecific amino-group
labeling with dyes of different fluorescence properties.
3.1.6. Introduction of Maleimide Groups onto Liposome Surfaces
The liposomes are modified with the bifunctional coupling reagent sulfo-SMCC to
introduce maleimide groups onto their surface.
1. The liposome types B or D that contain amino groups introduced with corresponding
lipids (see Table 1) in 0.5 mL PB are incubated under gentle stirring with 0.6 mg crystal-
line sulfo-SMCC corresponding to a fivefold molar excess relative to the reactive amino
groups, for 3060 min at 30C.
2. Nonreacted sulfo-SMCC is removed by dialysis using a tube with 10,000 mol-wt cutoff
(see Notes 1114).
Fig. 1. (A) Introduction of maleimide groups to the surface of amino-group-containing con-

ventional or PEG-liposomes. (B) Covalent coupling of thiol-functionalized molecules R-SH
(scFv antibody fragments, peptides, and so on) to maleimide liposomes.
3.1.7. Coupling of scFv Antibody Fragments

to Maleimide-Modified Liposomes
For the method of cloning, protein expression in Pichia pastoris, and purification
of functionalized scFv antibody fragments, we refer to our publication (8). Briefly, an
original anti-ED-B-fibronectin scFv antibody was engineered by functionalization at
its C-terminal end by the introduction of a hydrophilic spacer (GGSSGGSSGS) and
the terminal cysteine -Cys-Gly-Cys-Ser-Cys sequence (see Note 15). The scheme of
the scFv-liposome coupling is shown in Fig. 1.
Fig. 2. Separation of scFv-liposomes from unbound scFv antibody molecules. Western blot
analysis of aliquots taken from top to bottom of a metrizamide gradient: lanes 18, probes from
an incubation of scFv-liposomes (positive reaction); lanes 916, probes from an incubation of
scFv with unmodified control liposomes (negative control). Lanes 13 and 911 correspond to
fractions containing liposomes.
Due to the presence of the introduced cysteines, the scFv antibodies form dimers in
solution. Thus, coupling to the liposomes has to be effected in the presence of reduc-
ing agents (see Note 16). The purified scFv antibody dimer (0.5 mg/mL) in HBSE
buffer is reduced in the presence of a 2 mM final concentration of tributylphosphine
(TBP) for 4 h at 2025C (room temperature) under an argon or nitrogen atmosphere.
Maleimide-modified and fluorescence-labeled liposomes (types B or D, conventional
or stealth) in 100 L HBSE are incubated with 100 g reduced scFv antibody in
200 L HBSE for 1020 h at 2025C (room temperature) in the presence of 2 mM
TBP under argon or nitrogen atmosphere (see Note 17). Nonreacted maleimide groups
can be blocked by addition of an excess of cysteine, followed by further incubation
(e.g., 5- to 10-fold molar excess cysteine relative to the maleimide concentration, 2 h
incubation). The modified liposomes are separated from nonreacted scFv on a
metrizamide gradient as described in Subheading 3.2.
3.2. Separation of scFv-Liposomes
From Nonreacted scFv on a Metrizamide Gradient
Nonreacted scFv fragments are separated from the scFv-liposomes by flotation on
a discontinuous metrizamide gradient.
1. The liposomescFv reaction solution (200 L) is mixed with 100 L metrizamide (60%
in HBSE [w/v]) in ultracentrifuge tubes (5 mL, nitrocellulose) and overlaid with 2 mL
metrizamide (10% in HBSE [w/v]) followed by HBSE as a top layer.
2. The density gradient is centrifuged for 7 h at 85,000g and 4C using a swing-out rotor in
an ultracentrifuge.
3. Fractions of 400 L are carefully removed from the bottom to the top of the tubes by
cautious insertion of a capillary tube or Pasteur pipet and analyzed for protein content on
a non-reducing 14% SDS-PAGE gel followed by Western blot analysis (see Note 18).
As shown in Fig. 2, fractions containing scFv-coupled liposomes (lanes 13) give a
33 kDa band on a Western blot, corresponding to the mol wt of the scFv antibody mono-
mer plus the attached PEG-lipid. Nonreacted scFv monomers and dimers remain at the
bottom of the metrizamide gradient (lanes 58 and 1416). Incubation of unmodified
Fig. 3. Binding of -ED-B scFv-PEG-immunoliposomes to human colon carcinoma cells

Caco-2 (E-DB fibronectin positive) and Co-115 (E-DB fibronectin negative) cultured on col-
lagen-I-coated cover slips. Phase contrast image (A) and immunofluorescence (B) showing
binding of DiO-labeled -ED-B scFv-PEG-immunoliposomes to ED-B-positive Caco-2 cells.
On the ED-B-negative Co-115 cells shown in phase contrast (C) and by immunofluorescence
(D), binding was not detectable. Unmodified DiO-labeled control liposomes did not bind to
both cell types (not shown).
control liposomes with the scFv antibody (lanes 916) results in no protein detection in
the liposome fractions (lanes 911).
4. To remove the metrizamide, the pooled liposome fractions are dialyzed against PB using
a 10,000-mol-wt cut-off dialysis tube (volume ratio = 1 to 1000).
3.3. Binding of scFv-Liposomes to Tumor Cells In Vitro
An example of how the binding properties of the prepared scFv-liposomes can be
evaluated in vitro is given below (see Fig. 3).
1. Human colon carcinoma cells Caco-2 (ED-B fibronectin positive) and Co-115 (ED-B
fibronectin negative) are cultured in complete DMEM medium. Glass cover slips (1520
mm diameter) are placed into 12-well tissue-culture plates and coated with 100 L rat-tail
collagen I (10 mg/mL) and incubated for 30 min at 37C.
2. Caco-2 and Co-115 cells (3 105 cells/well) are plated on the cover slips and cultured for
72 h in a humidified 5% CO2 atmosphere at 37C.
3. Then the washed cells (PBS) are incubated with 100 L DiO-labeled -ED-B scFv-lipo-
somes corresponding to 910 1013 liposomes or 12.531.5 g scFv, respectively, in
PBS for 30 min at 4C.
4. After another washing step, the cover slips are removed, treated with 10% glycerine,
placed on a microscope slide, and analyzed on a fluorescence microscope (Leica DLMB).
As negative controls, the cells are incubated with unmodified fluorescence-labeled lipo-
somes (liposome types A, C; see Table 1 and Note 19).
4. Notes
1. Cholesterol (e.g., from Fluka, purum quality, >95%) should be recrystallized from metha-
nol. Cholesterol of lower quality should be avoided, since liposome membrane stability
can be reduced.
2. Rat-tail collagen can be extracted from fresh rat tails according to the method developed
by Ellsdale (9).
3. Mean hydrodynamic diameters of vesicles (liposomes, nanospheres, nanobeads) can be
determined with dynamic laser light-scattering instruments, e.g., the NICOMP 380 par-
ticle sizer (Particle Sizing Systems, Santa Barbara, CA).
4. For steric reasons, the maximal amount of total PEG-lipids that can be incorporated into
a lipid bilayer is 7 mol%. Within this amount, the contents of PE-PEG-Omet and
PE-PEG-NH2 can be variede.g., 5 mol% PE-PEG-Omet and 2 mol% PE-PEG-NH2,
depending on the required degree of amino group modification.
5. Other lipid compositions with synthetic lipids or hydrogenated SPC (HSPC) are often
used, especially for liposome formulations intended for parenteral applications (long cir-
culating or stealth liposomes). Several analytical methods to follow loss of lipids dur-
ing the preparation and modification steps are available. Radioactively labeled lipids
(3H-DPPC, 14C-DPPC) or cholesterol (3H-cholesterol), or 3H-cholesteryl hexadecyl ether
(NEN Life Science Products, Boston, MA), or lipophilic fluorescent dyes (e.g., lipophilic
BODIPY derivatives, Molecular Probes) are added at appropriate amounts to the initial
lipid mixtures.
6. The preparation of liposomes from mixed detergent/lipid micelles can also be done with
other detergents, such as n-alkyl-glucosides (n = 6 9), octyl-thioglucoside, or N-octanoyl-
N-methylglucamin (MEGA-8, Fluka). Interestingly, the choice of detergent influences the
size of the resulting liposomes. Thus, liposomes prepared from n-octyl-glucoside/lipid
micelles have an average size of 180 nm, whereas those made with n-hexyl-glucoside are
60 nm in diameter (10).
7. When synthetic lipids are used, the dialysis has to be performed above the corresponding
transition temperature Tc of the lipid. Hence, when for example dipalmitoylphosphatidyl
choline (DPPC) is used as main liposome-forming lipid, a temperature above its Tc of
41C has to be chosen. Additional membrane-forming components (cholesterol and so
on) depress the Tc by several degrees.
8. The use of common dialysis tubes for the preparation of liposomes is not recommended
because formation of concentration gradients in the tubes leads to uncontrolled detergent
removal, resulting in nonhomogeneous liposome preparations. Controlled detergent
removale.g., using a Mini-Lipoprep instrument or, for the production of large volumes
of liposomes, a capillary dialyzer method (11)is very important to obtain homogeneous
and reproducible small unilamellar liposomes of defined mean diameters.
9. The major differences between detergent dialysis and filter extrusion are the following:
Dialysis is a very gentle method, recommended when labile and unstable molecules are
incorporated into liposomes. Furthermore, we made the observation that lipophilic dyes
(e.g., Texas-red-DPPE) are strongly absorbed by Nucleopore membranes using the extru-
sion method, whereas in detergent dialysis, such dyes are quantitatively incorporated into
the liposomes.
10. The detachment of the lipid mixtures from the glass walls of the round-bottom flasks can
be accelerated by addition of small glass beads (23 mm diameter) and vigorous shaking.
The encapsulation efficiency of hydrophilic molecules into the trapped volume of the
liposomes is significantly increased by using the freeze-thaw method, as described by
Mayer (12). Using synthetic lipids, temperatures above the corresponding Tc also have to
be applied to the extrusion process.
11. The concentration of maleimide groups linked to the liposome surface can be determined
by using either a fluorescent thio-reagent or, e.g., radiolabeled cysteine or methionine.
Alternatively, the maleimide groups are saturated with an excess of cysteine followed by
determination of unbound cysteine with the Ellmans reagent (13).
12. A large selection of bifunctional coupling reagents for amino-group modification is avail-
able from Pierce, Rockford, IL.
13. As an alternative to the coupling of bifunctional maleimide reagents, such as sulfo-
SMCC, to amino groups on the liposomes, the maleimide-PEG-lipid molecule distearoyl-
PEG-maleimide (DSPC-PEG-Mal) can be purchased from Avanti Lipid Products, Ala-
baster, AL.
14. Another alternative to the specific coupling method described here is the so-called post-
insertion method (1). Briefly, thiol-modified ligands are coupled to DSPC-PEG-Mal
micelles in buffered aqueous solutions. After removing nonreacted ligand molecules by
dialysis, the modified micelles are incubated with preformed liposomes, whereby the
DSPC-PEG-maleimide-ligand molecules are efficiently transferred to the outer lipid layer
of the liposomes.
15. The number of cysteines introduced into the ligand molecule can vary from one to three
or more. Site-specific thiol modification with a single C-terminal cysteinee.g., as used
for the attachment of peptides to liposomesis also effective (9). Ideally, the coupling
reaction of a thiolated ligand with maleimide-modified liposomes should be performed at
a molar excess of ligand. However, the limited availability of large quantities of recombi-
nant proteins or the high costs of synthetic peptides do not permit working with ideal
reaction conditions.
16. The reduction of dimers can be achieved by using reducing agents containing thiol groups
themselves, such as -mercaptoethanol or dithiothreitol. However, the thiol groups of the
reducing agent compete directly with those of the scFv antibodies for attachment to
maleimide groups on the liposomes. Therefore they have to be removed after reduction
and before liposome coupling. This problem can be circumvented by using the trialkyl-
phosphine reagents tributylphosphine (TBP) or tris(2-carboxy-ethyl)phosphine (TCEP)
for reduction (14). Thus, we use TBP at a 2 mM final concentration for reduction of the
scFv-dimers, and we recommend maintaining the reducing conditions during the cou-
pling reaction to the liposomes.
17. It is recommended to use screw-cap glass vials with small magnetic stirring bars for the
coupling reactions. Incubation time can vary from 2 to 24 h, depending on the properties
(size, solubility) of the molecule to be attached to the liposomes.
18. Instead of separating the modified liposomes from nonreacted ligand molecules by
metrizamide centrifugation, unbound scFv can be removed by dialysis using high-mol-wt
cut-off dialysis tubes. Lately, metrizamide has been difficult to obtain. Thus, it can be
replaced by Ficoll PM70 (Amersham Biosciences, Uppsala, Sweden) at the same concen-
trations as metrizamide. We recommend the Spectra/Por DispoDialyzers, which offer a
wide selection of mol wt cut-offs ranging from 100 to 300,000 Daltons (Spectrum Labo-
ratories).
19. Depending on the properties of the cells expressing target receptors for ligand-modified
liposomes, flow cytometry can be used for the detection of liposome-cell binding.
Acknowledgments
The authors thank C. Meylan and P. Weber for their valuable contributions.
References
1.
1 Sapra, P. and Allen, T. M. (2003) Ligand-targeted liposomal anticancer drugs. Progr.
Lipid Res. 42, 439462.
2 Allen, T. M. (2002) Ligand-targeted therpeutics in anticancer therapy. Nat. Rev. Cancer 2,

2.
750762.
3 Marty, C., Ballmer-Hofer, K., Neri, D., Klemenz, R., Schott, H., and Schwendener, R. A.
3.
(2002) Cytotoxic targeting of F9 teratocarcinoma tumours with anti-ED-B fibronectin scFv
antibody modified liposomes. Br. J. Cancer 87, 106112.
4.
4 Ludewig, B., Barchiesi, F., Pericin, M., Zinkernagel, R. M., Hengartner, H., and Schwen-
dener, R. A. (2000) In vivo antigen loading and activation of dendritic cells via a liposo-
mal peptide vaccine mediated protective antiviral and antitumor immunity. Vaccine 19,
2332.
5.
5 Huang, C. and Mason, J. T. (1978) Geometric packing constraints in egg phosphatidylcho-
line vesicles. Proc. Natl. Acad. Sci. USA 75, 308310.
6.
6 Rubas, W., Supersaxo, A. Weder, H. G., et al. (1986) Treatment of murine L1210 leuke-
mia and melanoma B16 with lipophilic cytosine arabinoside prodrugs incorporated into
unilamellar liposomes. Int J. Cancer 37, 149154.
7 Mayer, L. D., Hope, M. J., and Cullis, P. R. (1986) Vesicles of variable sizes produced by
7.
a rapid extrusion procedure. Biochim. Biophys. Acta 858, 161168.
8 Marty, C., Scheidegger, P., Ballmer-Hofer, K., Klemenz, R., and Schwendener, R. A.
8.
(2001) Production of functionalized single-chain Fv antibody fragments binding to the
ED-B domain of the B-isoform of fibronectin in Pichia pastoris. Protein Expr. Purif. 21,
156164.
9 Ellsdale, T. and Bard, J. (1972) Collagen substrata for studies on cell behaviour. J. Cell.
9.
Biol. 54, 626637.
10 Schwendener, R. A., Asanger, M., and Weder, H. G. (1981) The preparation of large
10.
bilayer liposomes: controlled removal of n-alkyl-glucoside detergents from lipid/deter-
gent micelles. Biochem. Biophys. Res. Commun. 100, 10551062.
11. Schwendener, R. A. (1986) The preparation of large volumes of homogeneous, sterile
liposomes containing various lipophilic cytostatic drugs by the use of a capillary dialyzer.
Cancer Drug Delivery 3, 123129.
12 Mayer, L. D., Hope, M. J., Cullis, P. R., and Janoff, A. S. (1985) Solute distributions and
12.
trapping efficiencies observed in freeze-thawed multilamellar vesicles. Biochim. Biophys.
Acta 817, 193196.
13.
13 Eyer, P., Worek, F., Kiderlen, D., et al. (2003) Molar absorption coefficients for the
reduced Ellman reagent: reassessment. Anal. Biochem. 312, 224227.
14. Console, S., Marty, C., Garcia-Echeverria, C., Schwendener, R. A., and Ballmer-Hofer,
K. (2003) Antennapedia and HIV TAT protein transduction domains promote endocy-
tosis of high Mr cargo upon binding to cell surface glycosaminoglycans. J. Biol. Chem.
278(37), 35,10935,114.
15. Burmeister Getz, E., Xiao, M., Chakrabarty, T., Cooke, R., and Selvin, P. R. (1999) A
comparison between the sulfhydryl reductants tris(2-carboxyethyl)phosphine and
dithiothreitol for use in protein biochemistry. Anal. Biochem. 273, 7380.
Intravenous Immunoglobulin Treatment 403
26
Intravenous Immunoglobulin Treatment
for Fibrosis, Atherosclerosis, and Malignant Conditions
Ilan Krause and Yehuda Shoenfeld
Summary
We describe our experience with intravenous immunoglobulin (IVIg) treatment in fibrotic
conditions and our results and experience with the effect of IVIg therapy to prevent metastases
in malignancy. We have delineated the mechanisms by which IVIg can affect atherosclerosis
(i.e., effect on MMP-9, antiidiotypes to anti-OxLDL), which led to reduced atherosclerosis in
animal models. The effect of IVIg on skin fibrosis was assessed in a murine model of sclero-
derma-like disease. Collagen expression was decreased in the skin of mice treated with mouse
IVIg, associated with decreased type I collagen gene expression, and accompanied by inhibi-
tion of transforming growth factor (TGF) and interleukin (IL)-4 secretion by splenocytes. We
also described a favorable response to IVIg treatment in patients with either systemic sclerosis
or myelofibrosis. The administration of IVIg to mice inoculated with melanoma or sarcoma
cells induced a statistically significant inhibition of metastatic lung foci and prolongation of
survival time. IVIg was found to stimulate the production of IL-12, an anti-tumor and anti-
angiogenic cytokine. Positive staining of the cytoplasm, cell membrane, and nuclear membrane
of several types of malignant tumors by IVIg was immunohistochemically demonstrated.
Key Words: intravenous immunoglobulins; fibrosis; atherosclerosis; malignancy.
1. Introduction
Recently, we were able to treat a large number of patients (>200) with various
autoimmune conditions by IVIg. Herein, we review our experience with IVIg treat-
ment in fibrotic conditions and describe our results and experience with the effect of
IVIg therapy to prevent metastases in malignancy.
2. Therapy of Atherosclerosis With IVIg

The last decade has seen a revolution in our attitude toward the etiology and patho-
genesis of atherosclerosis (AS). In addition to the conventional risk factors (e.g., smok-
ing, diabetes, obesity, and so on), infections as well as inflammatory and immune
involvements were found to be involved in the causation of AS. Accelerated athero-
403
404 Krause and Shoenfeld
sclerosis was also reported in many classical autoimmune rheumatic diseases, i.e.,
systemic lupus erythematodes (SLE), rheumatoid arthritis (RA), vasculitis, and others
(15).
The autoimmune aspects of AS have been delineated and seem to involve both the
humoral and the cellular arms of the immune system, entailing a passive transfer;
experiments with activated lymphocytes in nave mice led to accelerated atherosclero-
sis. In contrast to many of the classical autoimmune diseases, several autoantigens/
autoantibodies seem to be implicated, i.e., HSP65, OxLDL, 2GPI (4,6), Lp(a), PT,
antiendothelial cell antibodies (AECA), and more to come (6,7).
As an immune/autoimmune mediated disease, AS is subject to immunomodulation
with anti-CD40, IVIG, cytokines, chemokines, and other immunomodulators (8). We
have delineated the mechanisms by which IVIg can affect ASeffect on MMP-9 (9),
antiidiotypes to anti-OxLDL (10). It seems that all the above mentioned immuno-
modulations led to reduced atherosclerosis in animal models. It is conceivable that in
the near future, such novel therapeutic approaches (i.e., anti tumor necrosis factor
[TNF]-) will be incorporated into the routine therapy of subjects with atherosclerosis
(i.e., restenosis prevention after angioplasty), and specifically when associated with
rheumatic autoimmune diseases (11).
3. IVIg in Fibrosis
Deposition of collagen, laminin, fibrinogen, and other molecules is fundamental to
the process of inflammation and healing. However, the accumulation of excessive
amounts of these extracellular proteins may lead to malfunction of vital organs, result-
ing in myelofibrosis, cirrhosis, pulmonary fibrosis, extraperitoneal fibrosis, or skin
thickening. Systemic sclerosis (SSc) is an autoimmune connective-tissue disease
characterized by microvascular damage, extracellular matrix deposition, and fibrosis,
involving mainly the skin, lungs, and gastrointestinal tract. No effective therapy has
so far been described for the diffuse fibrotic changes. Blank et al. (12) assessed the
effect of IVIg on skin fibrosis in tight-skin (Tsk/+) mice. The Tsk/+ mouse represents
a murine model of scleroderma-like disease with heritable fibrosis resembling the skin
fibrosis seen in human SSc patients. The excessive fibrosis in these mice is the result
of increased synthesis and accumulation of collagen in the skin. Tsk/+ mice received
IVIg beginning at the age of 4 wk, administered twice weekly for 4 wks. Control mice
were infused with 2% maltose. Collagen expression was decreased in the skin of Tsk/
+ mice treated with IVIg compared with that in control mice. The decreased collagen
expression after exposure to IVIg was associated with decreased type I collagen gene
expression. The reduction in skin fibrosis upon IVIg treatment was accompanied by
inhibition of TGF and IL-4 secretion by splenocytes. Levy et al. (13) were the first
to report a response of SSc patients to treatment with IVIg. Three patients with pro-
gressive and rapidly deteriorating disease (mainly affecting the skin) were planned to
receive six monthly courses of high-dose IVIg (2 g/kg). Two of the three patients
received six IVIg courses as planned, and no adverse effects or disease progression
occurred during the therapy. The third patient received three courses, after which he
developed renal failure and later died of sepsis. All three patients had a large decrease
in their skin score after the treatment compared to that before the treatment. It seems,
therefore, that IVIg may have a role in the treatment of SSc patients with rapidly
deteriorating skin disease, and possibly other target organs fibrosis. Recently we have
expanded our experiments to an additional 14 patients (14).
Myelofibrosis has been reported as a rare cause of pancytopenia in patients with
autoimmune diseases. Aharon et al. (15) described a 54-yr-old female patient who was
admitted with severe anemia subsequently found to be due to marrow fibrosis. During
the course of her hospitalization, the diagnosis of SLE was established. The patient
was treated with high-dose steroids, but improvement of her clinical symptoms as well
as normalization of her peripheral blood count were achieved only after high-dose
therapy with IVIg was instituted. Along with the improvement in the peripheral blood
parameters, normalization of the bone-marrow architecture was recorded on a repeated
bone-marrow biopsy. We suggest that IVIg therapy may be considered in extreme
cases of bone-marrow suppression in SLE.
4. Antimetastatic Effects of IVIg (Patent Nos. 5.562.902 and 5.965.130)

There is a bi-directional relationship between autoimmunity and cancer. Malignant
conditions are frequently associated with autoimmune phenomena. Examples include:
an increased incidence of EatonLambert myasthenia-like syndrome in patients with
small-cell carcinoma of the lung; thymoma in patients with myasthenia gravis; differ-
ent types of epithelial or lymphoproliferative malignancies in patients with autoim-
mune hemolytic anemia, thrombocytopenia, or neutropenia; and melanoma associated
with vitiligo. Conversely, there is an increased risk of cancer in autoimmune condi-
tions, as exemplified by the emergence of ovarian carcinoma in patients with dermato-
myositis, lymphoproliferative diseases in patients with rheumatoid arthritis, SLE and
Sjogrens syndrome, lung cancer in scleroderma patients, and thyroid papillary carci-
noma in patients with autoimmune thyroid diseases. The cancer may appear at the
time of diagnosis of the autoimmune disease or several years later (16). Since the two
diseases are similarly treated, we studied the efficacy of IVIg as a treatment for malig-
nant conditions: The administration of IVIg to mice inoculated intravenously with
melanoma or sarcoma cells induced a statistically significant inhibition of metastatic
lung foci and prolongation of survival time. Similar results were seen with SCID mice
inoculated with SK-28 human melanoma cells (17). In a different model, a lower num-
ber of melanoma recurrences and prolongation of survival time were demonstrated in
the IVIg-treated groups. In vitro studies revealed that IVIg was found to stimulate the
production of IL-12, an antitumor and antiangiogenic cytokine. Moreover, it enhanced
natural killer (NK) cell activity, thus explaining its beneficial effect in SCID mice
(which lack B- and T-cells but possess NK cells) (18). The results indicate that IVIg
acts as an anti-tumor agent, and may be considered as a potential therapy for the pre-
vention of tumor spread in humans. In another experiment, we studied the effect of
purified IVIg on MMP-9 secretion and mRNA expression by in vitro differentiated
human monocytic cells. Degradation of the extracellular matrix (ECM) is essential for
progression and metastasis of cancer cells. The ECM-degrading enzymes, matrix
metalloproteinases (MMPs), are produced mainly by intratumor monocytes/macroph-
ages. MMPs, particularly MMP-9, are reported to be of crucial significance for both
growth and tumor invasiveness. Inhibition of the expression of MMP-9 may prevent
tumor development. We found that IVIg dose-dependently and significantly reduced
the amount of secreted MMP-9 and its mRNA expression. F(ab)2, but not Fc frag-
ments, led to suppressed MMP-9 activity (9). However, competitive experiments dem-
onstrated that Fc, but not F(ab)2 fragments, reversed the IVIg-induced inhibitory
effects. Our results suggest that the whole IgG molecule may be needed for pertinent
IVIg-induced MMP-9 down-regulation. Our study points to an additional new mecha-
nism whereby IVIg may play a beneficial role in the prevention of tumor spread in
humans. We also tried to determine whether F(ab)2 prepared from IVIg binds to cellu-
lar structures of different tumor tissues. Direct immunohistochemistry using a
streptavidin peroxidase staining method was performed on biopsy samples of 18 dif-
ferent tumor tissues. Positive staining of the cytoplasm, cell membrane, and nuclear
membrane of several types of malignant tumors by F(ab)2 from IVIg was immuno-
histochemically demonstrated. Nuclear staining of tumor cells by IVIg was rare. IVIg
bound to different tumors of epithelial origin, especially colon carcinoma, breast car-
cinoma, and squamous cell carcinoma of the lung. Malignant tumors of mesenchymal
origin such as leiomyosarcoma have also demonstrated positive staining by IVIg (19).
Hence, IVIg contains antibodies to the cytoplasm, nuclear membrane, and cell mem-
brane of different malignant tumors, especially of epithelial origin. This binding might
provide a basis for the assumption that IVIg treatment of cancer patients may induce
antibody-dependent cell-mediated cytotoxicity response against tumors, and implies
that it can be potentially beneficial as adjuvant treatment of malignant diseases.
Merimsky et al. (20) observed a patient with a malignant peripheral nerve sheath tu-
mor (MPNST) who was treated with IVIg for multiple sclerosis. Her MPNST course
was remarkably longer and more indolent than expected; she achieved a disease-free
interval (DFI) of 30 mo. Seven other patients who were not treated by IVIg had a
relatively aggressive course (median DFI 3 mo). These results led us to examine the
effect of IVIg on the growth of sarcoma in vitro and in vivo in an experimental model
of MCA-bearing mice. When added to MCA-105 sarcoma cell cultures, IVIg pro-
duced a dose-dependent inhibitory effect on [H3]-thymidine incorporation (18). The
results demonstrate that the anti-proliferative activity results from an apoptotic effect
of IVIg on the tumor cells. In a second set of experiments, we evaluated the capability
of IVIg, when administered orally or subcutaneously, to inhibit the growth of MCA-
105 sarcoma lung metastases. A decrease in the mean lung weight was observed in the
mice that were treated by subcutaneous or oral administration, the latter being more
effective. The results point to a potential role for IVIg in the treatment of MPNST and
other soft-tissue sarcomas. Recently, we reported about a patient with superficial
spreading melanoma with liver and lung metastases (21). The patient refused chemo-
therapy and was treated with monthly high-dose IVIg (2g/kg). Six mo after the initia-
tion of IVIg therapy, the liver metastases regressed significantly, while the lung
metastases remained the same. After about 9 mo from the initiation of IVIg therapy,
new subcutaneous and bony metastases appeared, which continued to grow in the next
6 mo, although there was no significant change in the lung and hepatic lesions. The
patients died in a septic state 14 mo after the start of IVIg. This is the first report
pointing to the probable efficiency of high-dose IVIg in cases of unresponsive wide-
spread metastases of melanoma, and points to the implantation of the safe IVIg in
some patients with tumor metastases.
References
1.
1 Shoenfeld, Y. (2000) Atherosclerosis and the immune system: is atherosclerosis an autoim-
mune disease? Sem. Clin. Immunol. 1, 56.
2.
2 Shoenfeld, Y., Sherer, Y., and Haratz, D. (2001) Atherosclerosis as an infectious, inflam-
matory and autoimmune disease. Trends in Immunol. 22, 293295.
3. Shoenfeld, Y., Haratz, D., and Jacob, G. (2000) Heat shock protein 60/50, beta2-glyco-
protein I and Oxidized LDL as players in murine atherosclerosis. J. Autoimmunity 15,
199202.
4.
4 Sherer, Y. and Shoenfeld, Y.. (2003) Antiphospholipid antibodies: are they pro-athero-
genic or an epiphenomenon of atherosclerosis? Immunobiol. 207, 1316.
5.
5 George, J., Afek, A., Gilburd, B., Shoenfeld, Y., and Harats, D. (2001) Cellular and humoral
immune responses to heat shock protein 65 are both involved in promoting fatty-streak for-
mation in LDL-receptor deficient mice. J. Am. Coll. Cardiol. 38, 900905.
6 George, J., Harats, D., Gilburd, B., et al. (2000) Adoptive transfer of 2-glycoprotein
6.
I-reactive lymphocytes enhances early atherosclerosis in LDL receptor-deficient mice.
Circulation 102, 18221827.
7.
7 Harats, D., Yacov, N., Gilburd, B., Shoenfeld, Y., and George, J. (2002) Oral tolerance
with heat shock protein 65 attenuates Mycobacterium tuberculosis-induced and high-fat-
diet-driven atherosclerotic lesions. J. Am. Col. Cardiol. 7, 13331338.
8.
8 Sherer, Y. and Shoenfeld, Y. (2002) Immunomodulation for treatment and prevention of
atherosclerosis. Autoimmune Rev. 1, 2127.
9.
9 Shapiro, S., Shoenfeld, Y., Gilburd, B., Sobel, E., Lahat, N. (2002) Intravenous gamma
globulin inhibits the production of matrix metalloproteinase-9 in macrophages. Cancer
95, 20322037 .
10.
10 Wu, R., Shoenfeld, Y., Sherer, Y., et al.(2003) Anti-idiotypes to oxidized LDL antibodies
in intravenous immunoglobulin preparationspossible immunomodulation of atheroscle-
rosis. Autoimmunity 36, 9197.
11.
11 Doria, A., Shoenfeld, Y., Wu, R., et al. (2003) Risk factors for subclinical atherosclerosis
in patients with systemic lupus erythematosus. Ann. Rheum. Dis. 62, 10711077.
12.
12 Blank, M., Levy, Y., Amital, H., et al. (2002) Concise communication: the role of intrave-
nous] immunoglobulin therapy in mediating skin fibrosis in tight skin mice. Arth. Rheumat.
46, 16891690.
13.
13 Levy, Y., Sherer, Y., Langevitz, P., et al. (2000) Skin score decrease in systemic sclerosis
patients treated with intravenous immunoglobulina preliminary report. Clin. Rheumatol.
19, 207211.
14.
14 Amital, H., Rewald, E., Levy, Y., et al. (2003) Fibrosis regression induced by intravenous
gammaglobulin treatment. Ann. Rheum. Dis. 62, 175177.
15.
15 Aharon, A., Levy, Y., Bar Dayan, Y., et al. (1997) Successful treatment of early secondary
myelofibrosis in SLE with IVIG. Lupus 6, 408411.
16.
16 Shoenfeld, Y. and Gershwin, E. M., eds. (2000) Cancer and Autoimmunity. Elsevier,
Amsterdam, The Netherlands.
17. Shoenfeld, Y. and Fishman, P. (1999) Gamma-globulin inhibits tumor spread in mice. Int.
Immunol. 11, 12471251.
18.
18 Fishman, P., Bar-Yehuda, S., and Shoenfeld, Y.(2002) IVIG to prevent tumor metastases
(review). Int. J. Oncol. 21, 875880.
19.
19 Bar-Dayan, T., Barshack, I., Blank, M., et al. (1999) Antibodies to the cytoplasm, cell
membrane and nuclear membrane of malignant neoplasms in pooled normal human
polyspecific immunoglobulin G. Int. J. Oncol. 15, 10911096.
20. Merimsky, O., Meller, I., Inber, M., Bar-Yehuda, S., Shoenfeld, Y., and Fishman, P. (2002)
A possible role for IVIG in the treatment of soft tissue sarcoma: a clinical case and an
experimental model. Int. J. Oncol. 20, 839843.
21. Shoenfeld, Y., Levy, Y., and Fishman, P. (2001) Shrinkage of melanoma metastases fol-
lowing high dose intravenous immunoglobulin treatment. IMAJ 3, 698699.
T-Cell Costimulatory Blockade 409
27
Study of T-Cell Costimulatory Blockade
In Vivo at a Single-Cell Level
Minh Diem Vu and Xian Chang Li
Summary
The critical role of costimulatory signals in T-cell activation and the complexity of T-cell
costimulatory pathways involved make a detailed understanding of this system a challenging
task. By taking advantage of the unique chemical properties of CFSE, we and others have
developed an in vivo model that allows quantitative analysis of T-cell activation at a single-
cell level. This model involves labeling of donor T-cells with the tracking dye CFSE and
adoptively transferring into lethally irradiated allogeneic hosts. T-cells proliferating in the
host mice can be explicitly analyzed upon recovery. By using mice deficient for certain
costimulatory molecules as a source of donor cells or by treating the host mice with reagents
that block certain costimulatory pathways, this CFSE model is extremely useful in studying
the role of T-cell costimulatory signals in activation, survival, and effector differentiation of
alloreactive T-cells in vivo.
Key Words: Costimulation; T-cell activation; alloimmunity; transplantation; in vivo assay;
CFSE.
1. Introduction
T-cells are essential in allograft rejection. Thus, blocking T-cell activation remains a
key strategy in preventing transplant rejection. In essence, T-cell activation requires at
least two signals. Signal 1 is delivered by the T-cell receptor (TCR) upon recognition of
the alloantigens, and signal 2 is provided by a set of cell-surface structures collectively
called costimulatory molecules. Both signals drive robust T-cell activation and devel-
opment of T-cell effector function. In fact, stimulation of T-cells via the T-cell receptor
in the absence of costimulatory signals often induces an anergic state or apoptotic cell
death (18). Thus, T-cell costimulatory molecules are of central importance in T-cell
activation. Consequently, blocking T-cell costimulatory signals has attracted tremen-
dous attention in tolerance induction.
Among the T-cell costimulatory molecules identified, CD28 and CD154 (CD154 is
also called CD40 ligand, or CD40L) are the classical and also most extensively studied
409
410 Vu and Li
ones (912). However, cell-surface molecules with costimulatory properties are not
confined to CD28 and CD154, and multiple alternative costimulatory molecules have
been discovered (13). For example, upon TCR stimulation, signals delivered by ICOS,
OX40 (CD134), 4-1BB (CD137), CD30, CD27, and CD70 can costimulate T-cell acti-
vation, cytokine production, and effector cell function (1419). Thus, understanding
precisely the individual and collective roles of such diverse costimulatory molecules in
regulating activation of alloreactive T-cells becomes critically important in transplan-
tation research. We and others have developed an in vivo model that allows quantita-
tive analysis of T-cell activation at a single-cell level by taking advantage of a tracking
dye called carboxyfluorescein succinimidyl ester (CFSE) (2025). In this chapter,
CFSE-labeled T-cells proliferating in the allogeneic hosts are used to illustrate the
model and the means by which T-cell activation in vivo and the impact of costimulatory
blockade on T-cell activation are analyzed.
2. Materials
1. CFSE (Molecular Probes, Portland, OR).
2. Hanks balanced salt solution (HBSS) (BioWhittaker, Walkersville, MD).
3. Phosphate-buffered saline (PBS) (BioWhittaker).
4. Fetal bovine serum (FBS) (BioWhittaker).
5. Bovine serum albumin (BSA) (Sigma, St. Louis, MO).
6. Red blood cell lysing buffer (Sigma).
7. Surgical instruments for small animals.
8. Petri dishes (Corning, Bedford, MA).
9. Lymphocyte preparation kit (Sigma).
10. Nylon cell strainer, 70 m (BD Biosciences, Bedford, MA).
11. Animal irradiator (Gammacell Exactor, Kanata, Ontario, Canada).
12. FACS machine equipped with CellQuest software (BD Biosciences, Mountain View, CA).
13. Paraformaldehyde (Sigma).
14. Staining mAbs: CyChrome anti-mouse CD4 (clone GK1.5), CyChrome anti-mouse CD8
(clone 53-6.7), PE-anti-mouse IL-2 (clone JES6-5H4), isotype control antibodies, and
PE-annexin V (PharMingen, San Diego, CA).
15. Cytofix/Cytoperm kit (PharMingen).
16. Annexin V binding buffer (PharMingen).
17. C57BL/6 (H-2b) and DBA/2 (H-2d) mice (Jackson Laboratory, Bar Harbor, ME).
18. Murine CTLA4-Ig fusion protein.
19. Anti-CD154 mAb (MR1, hamster IgG, American Type Culture Collection [ATCC],
Manassas, VA).
20. PMA and ionomycin (Sigma).
21. Penicillin and streptomycin (Sigma).
3. Methods
Technically, the in vivo CFSE assay consists of the following five steps: (1) label-
ing of donor lymphocytes with CFSE; (2) adoptively transferring CFSE-labeled cells
into an allogeneic host; (3) recovery of CFSE-labeled donor cells from the host mice
after a given period of time; (4) cell staining in vitro after recovery; (5) flow cytometry
and data acquisition/analysis (Fig. 1) (see Note 1).
Fig. 1. A cartoon illustration of the in vivo CFSE model. Key steps include (1) labeling of
donor cells with CFSE; (2) adoptive transferring of CFSE-labeled cells into irradiated hosts;
(3) cell recovery; (4) in vitro staining; and (5) data acquisition and analysis.
3.1. Labeling of Donor Lymphocytes With CFSE

3.1.1. Chemical Properties of CFSE
The precursor of fluorescent CFSE dye is carboxyfluorescein diacetate succinimidyl
ester (CFDASE), which is nonfluorescent, nonpolar, but highly lipophilic (26,27).
When CFDASE is added to cell suspensions, CFDASE passively diffuses through cell
membranes, and upon esterase hydrolysis in the cytoplasm of live cells, CFDASE is
converted to fluorescent CFSE. CFSE covalently binds to cytoplasmic proteins to form
a stable complex (22,23). When CFSE-labeled cells divide, the CFSE contents segre-
gate equally between the two daughter cells. Therefore, the fluorescent intensity of
labeled cells halves after each consecutive cell division, which allows tracking the
division history of activated T-cells.
3.1.2. Donor Lymphocyte Preparation
Lymphocytes were prepared from the spleen and the peripheral lymph nodes of
donor C57BL/6 (H-2b) mice. A detailed description of this method has been reported
in ref. 28.
412 Vu and Li
1. Sedated mice are sacrificed by cervical dislocation, and all the lymph nodes in the cervi-
cal, auxiliary, popliteal, and inguinal areas and the spleen are harvested and placed into a
sterile Petri dish containing cold HBSS.
2. Single-cell suspension is prepared by gently disrupting the lymphoid tissues and pressing
them through a 200-m metal mesh.
3. Cells are then transferred from the Petri dish into 15-mL centrifuge tubes and centrifuged
in Sorvall RTH250 rotor at 1200 rpm (300g) for 5 min at 4C.
4. After centrifugation, the supernatant was discarded, leaving behind only the cell pellet.
Red blood cells in the cell preparation are removed using red blood cell lysing buffer
(29). For this purpose, gently disrupt the cell pellet, and then add 10 mL of lysing buffer
(room temperature).
5. Five min later, the tubes are centrifuged at 1200 rpm (300g) for 5 min to pellet the lym-
phocytes.
6. Cells are washed once in cold HBSS and resuspended in HBSS for cell counting and
CFSE labeling.
3.1.3. CFSE Labeling

CFSE is prepared in DMSO at 5 mM, stored at 20C, and used as stock solutions.
Labeling of lymphocytes with CFSE consists of the following steps (see Note 2):
1. Resuspend cells in serum-free HBSS at 2 107 cells/mL in a 50-mL centrifuge tube.
2. Prepare an equal volume of CFSE in serum-free HBSS at 10 M using another 50-mL
tube.
3. Gently mix the cell suspension and the CFSE solution, so the final cell concentration for
labeling is 1 107/mL and the final CFSE concentration is 5 M.
4. Incubate the cells at room temperature for 6 min with gentle mixing.
5. After 6 min, stop the labeling reaction by adding FBS (10% of total volume).
6. Pass the cell suspension through a 70-m nylon cell strainer to remove cell debris and cell
clumps.
7. Wash cells twice in HBSS, resuspend cells in HBSS for counting and injection.
3.2. Adoptive Transfer of CFSE-Labeled Cells into Allogeneic Hosts

3.2.1. Irradiation of Host Mice
DBA/2 (H-2d) mice are given 1000 rads lethal irradiation with a GammaCell Exac-
tor (Kanata, Ontario, Canada) 4 to 5 h before cell transfer.
3.2.2. Transfer of CFSE-Labeled Cells into the Irradiated Hosts

1. Resuspend CFSE-labeled cells in HBSS.
2. Restrain the host mice in an animal restrainer.
3. Inject 60 to 80 106 cells in a total volume of 0.5 mL into each host via the tail vein (see
Note 3).
3.2.3. Treatment of Host Mice With Costimulatory Blockade Reagents

The host mice can be treated with CTLA4-Ig to block the B7/CD28 costimulatory
pathway, anti-CD154 mAb (MR1) to block the CD40/CD154 costimulatory pathway,
or a combination of both. Mice treated with isotype control antibody are included as a
control. The reagents are given at 0.5 mg i.p. daily for 3 d starting at the time of cell
transfer.
3.3. Recovery of CFSE-Labeled Donor Cells From the Host Mice

Three d after adoptive cell transfer, the host mice are killed, spleen and peripheral
lymph nodes are harvested separately, and single-cell suspensions are prepared as
described in Subheading 3.2. (see Notes 4 and 5). For cells from the host spleen, red
blood cells are lysed using the red blood cell lysing buffer as described. Cells are then
washed once in cold PBS containing 0.5% BSA (PBS/BSA), counted, and resuspended
in PBS/BSA for staining.
3.4. Cell Staining After Recovery From the Hosts
3.4.1. Cell-Surface Staining
1. Resuspend cells in PBS/BSA at 1 107/mL and aliquot 100 L into individual 1.5-mL
Eppendorf tubes (1 106 cells/100 L/tube).
2. Add 1 g CyChrome-anti-mouse CD4 or 1 g CyChrome-antimouse CD8 mAb into
each tube. For 3-color staining, PE-conjugated mAbs against other cell-surface markers,
for example, CD25, CD44, CD69, and so on can be added at the same time. Cells stained
with CyChrome-conjugated or PE-conjugated isotype-matched control antibodies are
included as controls.
3. Incubate the cells on ice for 20 min.
4. Wash cells twice in PBS/BSA after the incubation and fix the cells in 1% paraformal-
dehyde (prepared in PBS) for FACS analysis.
3.4.2. Staining for Intracellular Cytokines

1. Resuspend cells in RPMI-1640 medium supplemented with 10% FBS and 1% penicillin/
streptomycin at 5 106 cells/mL.
2. Briefly stimulate the cells in vitro with PMA (50 ng/mL) and ionomycin (500 ng/mL) at
37C for 4 h.
3. In the last 2 h of in vitro stimulation, add GolgiStop (PharMingen) at 1 g/mL into the
culture.
4. Harvest the cells and proceed to surface staining with CyChrome-anti-CD4 and
CyChrome-anti-CD8 as described.
5. After cell-surface staining, cells are incubated in Cytofix/Cytoperm solution at 4C for
10 min.
6. Wash the cells twice in Perm/Wash solution, aliquot cells into 1.5-mL Eppendorf tubes
(1 106 cells/100 L/tube), add PE-anti-IL-2, PE-anti-interferon (IFN)-, or PE-anti-
tumor necrosis factor (TNF)-, and incubate on ice for 30 min. PE-conjugated isotype-
control antibodies are included as controls.
7. Wash the cells twice in Perm/Wash solution and resuspend the cells in Perm/Wash solu-
tion for FACS analysis.
3.4.3. Staining for Apoptotic Cell Death

1. Resuspend cells in Ca2+-rich annexin V binding buffer at 1 106 cells/100 L/tube in
1.5-mL Eppendorf tubes.
2. Add 5 L of PE-conjugated annexin V into each tube and incubate at room temperature
for 30 min.
3. Wash the cells once in annexin V binding buffer and resuspend the cells in the binding
buffer for analysis.
414 Vu and Li
Fig. 2. Typical FACS parameters and setup for in vivo CFSE assay. FSC and SSC are used
to identify and to gate lymphocyte population. CD4+ or CD8+ T-cells are gated based on subset
staining, and their division history is revealed by the dilution of CFSE.
3.5. Flow Cytometry and Data Acquisition/Analysis

3.5.1. Data Acquisition
1. All samples are acquired using FACS equipped with CellQuest software (see Note 6).
2. The FACS should be set up based on the following parameters in order to identify the
total cellular events, different T-cell subsets, and their CFSE profile: (1) FSC vs SSC dot
plot; (2) FL3 (CyChrome staining) vs SSC dot plot; (3) FL1 (CFSE) vs FL2 (PE staining)
dot plot; (4) FL1 (CFSE) histogram (Fig. 2).
3. For samples with three-color staining, pay special attention to parameter compensation,
particularly the FL1 (i.e., CFSE), as the CFSE has a wide range of fluorescent intensity.
4. Collect a minimum of 500,000 events from each sample for better graphic representation.
3.5.2. Data Analysis

The following analyses can be readily accomplished using the CFSE model: (1)
division frequency of T-cells and T-cell subsets can be precisely calculated; (2) acti-
vated T-cells at distinct division cycles can be phenotyped; (3) effector programs of
activated T-cells can be systemically compared; (4) apoptotic cell death of activated
T-cells at distinct division cycles can be precisely analyzed.
1. Calculation of division frequency: distinct division cycles of T-cells, CD4+ or CD8+ T-
cell subsets are identified by their CFSE profiles and individually gated. The absolute
number of daughter cells in each cell division is given by the CellQuest software, the
number of precursors that have divided and given rise to the absolute number of daughter
cells is extrapolated using the formula: y/2n; in which y = absolute number of cells in each
cell cycle and n = number of cell divisions. For example, 16 daughter cells in the third cell
division are the progeny of two precursors, each of which have divided three times (16/23
= 2). The division frequency in the responding T-cell population is then calculated by
dividing the total number of precursors by the total number of CFSE positive cells col-
lected (24).
2. Analysis of effector functions: total T-cells or CD4+ or CD8+ subsets are gated, their
division history (i.e., CFSE profile, FL1) is plotted against intracellular cytokine staining
(FL2), and a quadrant mark can be set up based on control staining using PE-isotype
control antibody. Thus, the relative expression of effector cytokines in T-cells at distinct
division cycles can be clearly appreciated. By gating onto individual division cycles, the
absolute % of cytokine positive cells can be calculated and plotted (Fig. 3).
3. Phenotyping of dividing T-cells: within a given T-cell subset, the phenotypic changes
following cell-cycle transitions can be analyzed. Similar to intracellular cytokine analy-
sis, the division history of T-cells or CD4+ or CD8+ T-cells is identified by their CFSE
profile (FL1) and plotted against surface staining with PE-conjugated specific antibod-
ies (FL2). Thus, the phenotypic changes upon cell-cycle progression can be analyzed
(Fig. 4). The same method can be used to identify apoptotic cell death based on staining
with PE-annexin V (Fig. 5).
4. Notes
1. Depending on the questions asked and the T-cell subset examined, this CFSE model
allows the integration of several variables. First, the CFSE-labeled donor cells can be
prepared from genetic knockout micefor example, CD28 knockout, CD154 knockout,
or CD28/CD154 double knockout mice, and so on. Second, the host mice can be treated
with a variety of therapeutic reagents, e.g., monoclonal antibodies or recombinant fusion
proteins against different costimulatory pathways at the time of injecting CFSE-labeled
allogeneic cells. Thus, T-cell activation in vivo in response to such treatments can be
appreciated and compared. Third, the host mice can be prepared by lethal irradiation, or
one can simply use genetically immunodeficient mice, such as scid mice or Rag-defi-
cient mice. Also, F1 mice that share partial major histocompatibility complex (MHC)
haplotype with donor mice can be used as the hosts. Apparently, either type of hosts has
certain advantages and disadvantages, and therefore, proper experimental controls should
be included in all experiments.
2. CFSE, like any other chemical, can be toxic to cells, and its toxicity is dose dependent
and cell-type dependent. We noticed that the final CFSE concentration of 5 M consis-
tently produces excellent labeling for T-cells without any detectable side effect. How-
416
416
Fig. 3. Analysis of cytokine expression in the context of cell division. Donor cells were recovered from the host mice and stained for
intracellular cytokine expression using PE-conjugated cytokine-specific mAb. CD4+ T-cells were identified by CyChrome-anti-CD4 staining,
and their division history was revealed by the dilution of the CFSE dye. Cytokine expression in distinct division cycles was plotted and analyzed.
Vu and Li
Fig. 4. Phenotypic characterization of dividing T-cells. Donor cells were recovered from the
host mice and stained with PE-anti-CD25, PE-anti-CD44, and PE-anti-CD62L. CD4+ T-cells
were identified by CyChrome-anti-CD4 staining and their division history was revealed by the
dilution of the CFSE dye. Expression of such activation markers in distinct division cycles was
plotted and analyzed.
ever, higher CFSE concentrations often result in greater cell loss during the labeling pro-
cess. It is likely that higher CFSE concentrations may also interfere with T-cell function.
Thus, care should be taken when labeling different types of cells with CFSE.
3. Transferring T-cells into lymphopenic mice can also induce homeostatic expansion of
transferred cells. However, homeostatic expansion requires (1) empty hosts; (2) small
number of T-cells transferred; and (3) a longer period of time (>7 d). Thus, the number of
donor cells injected into the irradiated hosts is important to prevent homeostatic expan-
sion from occurring. We recommend transferring 60 to 80 106 CFSE-labeled donor
cells and performing the analysis 3 d after adoptive cell transfer. Cell proliferation under
such conditions is driven primarily by the host alloantigens (8).
418 Vu and Li
Fig. 5. Cell division and activation induced cell death. Donor cells were recovered from the
host mice and stained with PE-annexin V. CD4+ T-cells were identified by CyChrome-anti-
CD4 staining and their division history was revealed by the dilution of the CFSE dye. Apopto-
tic cell death in distinct division cycles was plotted and analyzed.
4. In the model described, a 3-d period of in vivo stimulation represents the best window of
opportunity to have a panoramic view of cell-division history in vivo, and up to eight
cell divisions can be clearly identified at this time point. Longer than 3 d of in vivo
stimulation often results in polarized cell-division profiles, i.e., a profile with two major
peaks (divided cells and undivided cells), leaving little in between.
5. In vivo T-cell response is often compartmentalized. For example, T-cells activated in
the host spleen may exhibit certain distinct characteristics other than those in the host
lymph nodes (30,31). Such differences may be due to distinct homing properties of
T-cell subsets to different lymphoid compartments, or due to differences in the quality
of stimulation in different lymphoid tissues. Thus, the host spleen and peripheral lymph
nodes should be examined and compared separately.
6. Multi-color flow cytometry can be complex, especially with a CFSE fluorescent inten-
sity across a wide spectrum. Thus, special attention should be paid to the following
points: (1) proper adjustment of compensation parameters; (2) acquisition of adequate
cell numbers, especially for analyzing rare cell subsets (e.g., 0.51.0 106 events); (3)
when identification of donor-derived and host-derived cells appears confusing, it is
helpful to incorporate additional markers such as donor MHC haplotype staining in
order to be certain.
Acknowledgments
Grant support for Xian C. Li is provided by the Juvenile Diabetes Research Foun-
dation.
References
1.
1 Jenkins, M. K., Pardoll, D. M., Mizuguchi, J., Quill, H., and Schwartz, R. H. (1987) T-cell
unresponsiveness in vivo and in vitro: fine specificity of induction and molecular charac-
terization of the unresponsive state. Immunol. Rev. 95, 113135.
2.
2 Schwartz, R. H. (1990) A cell culture model for T lymphocyte clonal anergy. Science
248(4961), 13491356.
3.
3 Matzinger, P. (1999) Graft tolerance: a duel of two signals. Nat. Med. 5, 616617.
4.
4 Gimmi, C. D., Freeman, G. J., Gribben, G. J., Gray, G., and Nadler, L. M. (1993) Human
T-cell clonal anergy is induced by antigen presentation in the absence of B7 costimulation.
Proc Natl Acad Sci USA 90(14), 65866590.
5.
5 Greenwald, R. J., Boussiotis, V. A., Lorsbach, R. B., Abbas, A. K., and Sharpe, A. H.
(2001) CTLA-4 regulates induction of anergy in vivo. Immunity 14(2), 145155.
6.
6 Wells, A. D., Walsh, M. C., Bluestone, J. A., and Turka, L. A. (2001) Signaling through
CD28 and CTLA-4 controls two distinct forms of T cell anergy. J. Clin. Invest. 108(6),
895904.
7.
7 Frauwirth, K. A., Alegre, M. L., and Thompson, C. B. (2000) Induction of T cell anergy in
the absence of CTLA-4/B7 interaction. J. Immunol. 164(6), 29872993.
8.
8 Li, Y., Li, X. C., Zheng, X. X., Wells, A. D., Turka, L. A., and Strom, T. B. (1999) Block-
ing both signal 1 and signal 2 of T-cell activation prevents apoptosis of alloreactive T cells
and induction of peripheral allograft tolerance. Nat. Med. 5(11), 12981302.
9.
9 Azuma, M., Cayabyab, M., Buck, D., Phillips, J. H., and Lanier, L. L. (1992) CD28 inter-
action with B7 costimulates primary allogeneic proliferative responses and cytotoxicity
mediated by small, resting T lymphocytes. J. Exp. Med. 175(2), 353360.
10.
10 Larsen, C. P., Elwood, E. T., Alexander, D. Z., et al. (1996) Long-term acceptance of
skin and cardiac allografts after blocking CD40 and CD28 pathways. Nature 381(6581),
434438.
11.
11 Kirk, A. D., Harlan, D. M., Armstrong, N. N., et al. (1997) CTLA4-Ig and anti-CD40
ligand prevent renal allograft rejection in primates. Proc. Natl. Acad. Sci. USA 94(16),
87898794.
12.
12 Cayabyab, M., Phillips, J. H., and Lanier, L. L. (1994) CD40 preferentially costimulates
activation of CD4+ T lymphocytes. J. Immunol. 152(4), 15231531.
13.
13 Croft, M. (2003) Costimulation of T cells by OX40, 4-1BB, and CD27. Cytokine Growth
Factor Rev. 14(34), 265273.
14.
14 Hutloff, A., Dittrich, A. M., Beier, K. C., et al. (1999) ICOS is an inducible T-cell co-
stimulator structurally and functionally related to CD28. Nature 397(6716), 263266.
15.
15 Mallett, S., Fossum, S., and Barclay, A. N. (1990) Characterization of the MRC OX40
antigen of activated CD4 positive T lymphocytes-a molecule related to nerve growth fac-
tor receptor. EMBO J. 9(4), 10631068.
16.
16 de Jong, R., Loenen, W. A., Brouwer, M., et al. (1991) Regulation of expression of CD27,
a T cell-specific member of a novel family of membrane receptors. J. Immunol. 146(8),
24882494.
17.
17 Kim, Y. J., Pollok, K. E., Zhou, Z., et al. (1993) Novel T cell antigen 4-1BB associates
with the protein tyrosine kinase p56lck1. J. Immunol. 151(3), 12551262.
420 Vu and Li
18.
18 Del Prete, G., De Carli, M., DElios, M. M., et al. (1995) CD30-mediated signaling
promotes the development of human T helper type 2-like T cells. J. Exp. Med. 182(6),
16551661.
19.
19 Kobata, T., Agematsu, K., Kameoka, J., Schlossman, S. F., and Morimoto, C. (1994) CD27
is a signal-transducing molecule involved in CD45RA+ naive T cell costimulation. J.
Immunol. 153(12), 54225432.
20.
20 Demirci, G., Gao, W., Zheng, X. X., Malek, T. R., Strom, T. B., and Li, X. C. (2002) On
CD28/CD40 ligand costimulation, common gamma-chain signals, and the alloimmune
response. J. Immunol. 168(9), 43824390.
21.
21 Lyons A.B. (2000) Analysing cell division in vivo and in vitro using flow cytometric
measurement of CFSE dye dilution. J. Immunol. Methods 243(12), 147154.
22.
22 Lyons, A. B. (1999) Divided we stand: tracking cell proliferation with carboxyfluorescein
diacetate succinimidyl ester. Immunol. Cell Biol. 77(6), 509515.
23.
23 Lyons, A. B. and Parish, C. R. (1994) Determination of lymphocyte division by flow
cytometry. J. Immunol. Methods 171(1), 131137.
24.
24 Wells, A. D., Gudmundsdottir, H., and Turka, L. A. (1997) Following the fate of individual
T cells throughout activation and clonal expansion. J. Clin. Invest. 100(12), 31733183.
25.
25 Fulcher, D. and Wong, S. (1999) Carboxyfluorescein succinimidyl ester-based prolifera-
tive assays for assessment of T cell function in the diagnostic laboratory. Immunol. Cell
Biol. 77(6), 559564.
26.
26 Weston, S. A. and Parish, C. R. (1990) New fluorescent dyes for lymphocyte migration
studies. J. Immunol. Methods 133(1), 8797.
27.
27 Parish, C. R. (1999) Fluorescent dyes for lymphocyte migration and proliferation studies.
Immunol. Cell Biol. 77(6), 499508.
28. Kruisbeek, A. M. (2000) In vitro assays for mouse lymphocyte function, In: Current
Protocols in Immunology. Coligan, J. E., Kruisbeek, A. M., Margulies, D. H., Shevack, E.
M., and Strober, W. (eds). National Institutes of Health, Wiley, New York, NY,
pp. 3.1.13.1.5.
29.
29 van Oss, C. J., Bronson, P. M., Dinolfo, E. A., and Chadha, K. C. (1981) Two methods for
the removal of erythrocytes from buffy coats for the production of human leukocyte inter-
feron. Immunol. Commun. 10(6), 549555.
30. Drake, D. R. 3rd. and Braciale, T. J. (2003) Not all effector CD8+ T cells are alike.
Microbes Infect. 5(3), 199204.
31. Cron, R. Q., Gajewski, T. F., Sharrow, S. O., Fitch, F. W., Matis, L. A., and Bluestone, J.
A. (1989) Phenotypic and functional analysis of murine CD3+,CD4,CD8 TCR-gamma
delta-expressing peripheral T cells. J. Immunol. 142(11), 37543762.
Stem Cell Transplantation 421
28
Stem Cell Transplantation
Graft-Mediated Antileukemia Effects
William J. Hogan and Hans Joachim Deeg
Summary
Graft-mediated antileukemia (GVL) activity is a major factor contributing to the success of
allogeneic hematopoietic stem transplantation (aHCT). Recent advances have permitted the
establishment of GVL activity without the need for a myeloablative conditioning regimen,
thereby permitting even older and sicker patients to avail of potentially curative therapy. Use of
adoptive immunotherapy by combining reduced intensity conditioning and donor leukocyte
infusion (DLI) has resulted in strategies that can be exploited to maximize GVL effects while
minimizing toxicity. These advances, combined with new molecularly targeted agents, creates
new possibilities to develop less toxic, curative therapy for a greater number of patients. This
review summarizes pertinent information regarding the evidence in favor of GVL effects, the
impact of disease type and mechanisms of GVL.
Key Words: GVHD; GVL effect; hematopoietic cell transplantation; minor antigens.
1. Introduction
Initial efforts to cure cancer were directed exclusively at chemotherapy or radio-
therapy dose escalation. As toxicity to normal host hematopoietic cells was dose-lim-
iting, allogeneic transplantation was seen as a way to permit further intensification of
therapy. However, after cytotoxic therapy, malignant cells die with first-order kinet-
ics, making cure an unachievable goal of dose escalation in most situations (1). In
addition, more intensive regimens are associated with more pronounced non-hemato-
logic toxicity, reducing the net benefit of additional anti-leukemic effects (2). Recent
advances in the field of transplantation biology include efforts to reduce transplant-
related mortality (TRM) by tailoring the conditioning regimen to the patient with tar-
geted dosing of cytotoxic drugs (3) or by replacing myeloablative conditioning with
much less intense regimens that rely predominantly on immunological effects of trans-
planted donor cells to eradicate malignant cells (46). Advantages of reduced inten-
421
422 Hogan and Deeg
sity or minimally ablative conditioning regimens include reduction in organ toxicity

attributable to the conditioning regimen, such as mucositis, bacterial infection, and
venoocclusive disease of the liver, in addition to decreased transfusion requirements.
There is also some evidence that the incidence of acute graft-vs-host disease (GVHD)
might be decreased (7). One possible explanation for this finding is that the less
intense conditioning causes less of a cytokine storm, thereby reducing the contribution
of one of the major putative factors in the pathogenesis of this disorder. These
approaches have allowed older or less robust patients to be considered for allogeneic
transplantation with curative potential, a critical issue, as the median ages of diag-
noses for many hematological malignancies are in the sixth to seventh decades.
2. Evidence of a Graft-vs-Leukemia (GVL) Effect in Human Patients

During allogeneic transplantation, donor hematopoietic progenitor cells and immu-
noreactive leukocytes are transferred to the recipient and establish donor-derived
hematopoiesis. As previously observed in rodent models, it was appreciated that donor
cells also had unwanted effects, mediating a stereotypical syndrome, initially dubbed
secondary disease, now known as graft-vs-host disease. Subsequent observations sug-
gested, however, that this undesirable complication could be associated with a benefi-
cial graft-vs-leukemia effect (810). These observations included (1) the inverse
correlation between GVHD and leukemia recurrence from retrospective single-institu-
tion and registry studies (11); (2) the occurrence of leukemic remissions during epi-
sodes of acute and chronic GVHD (12); (3) the ability to purposefully induce such
remissions by the withdrawal of immunosuppression (13,14); (4) the higher risk of
relapse in recipients of either syngeneic or autologous grafts compared to human leuko-
cyte antigens (HLA)-matched (but minor antigen disparate) donors (15,16); (5) the
increased risk of relapse associated with T cell depleted (TCD) grafts (1719); and,
most convincingly, (6) the ability of donor lymphocyte infusions to induce sustained
remissions after relapse following transplantation (20).
2.1. Remissions Associated With GVHD

or Withdrawal of Immunosuppression
Barnes et al. (21) predicted the possibility of a GVL effect as early as 1956, and
subsequent observations, including analyses of retrospective data from transplant reg-
istries, supported this hypothesis (814). Sullivan et al. (11) analyzed data from 1202
patients with acute myeloid leukemia (AML), acute lymphocytic leukemia (ALL), or
chronic myeloid leukemia (CML) receiving allogeneic marrow from HLA-identical
siblings. Among patients with AML and ALL transplanted in relapse and surviving in
remission to day +150 (thereby being at risk for chronic GVHD), the probability of
subsequent relapse was 74% in those who did not develop GVHD compared to 34%
in patients with chronic GVHD (p < 0.001). Actuarial survival was 25% and 62%,
respectively (p < 0.009). Among patients with CML in accelerated phase or blast cri-
sis, the probability of relapse after day +150 was 65% in patients without GVHD and
36% in those with acute and chronic GVHD (p < 0.017). Horowitz et al. reported data
on 2254 patients receiving HLA-identical sibling bone-marrow transplants for AML
in first remission, ALL in first remission, and CML in first chronic phase (17). For
T-cell-replete grafts, lower relapse rates were observed in patients who developed
acute (relative risk 0.68, p = 0.03), chronic (relative risk 0.43, p = 0.01), and both acute
and chronic (relative risk 0.33, p = 0.0001) GVHD compared to those without GVHD,
suggesting that GVHD was associated with durable anti-leukemic effects.
2.2. Outcomes With Syngeneic vs HLA-Identical Sibling Donors

Gale et al. reported International Bone Marrow Transplant Registry (IBMTR) data
comparing outcomes in 103 identical-twin transplants to 1030 concurrent HLA-iden-
tical sibling transplants matched for prognostic factors (16). Three-year probabilities
of relapse were greater after identical-twin transplants for patients with ALL (36% vs
26%), AML (52% vs 16%), and CML (40% vs 7%). Increased relapse risks in AML
and CML persisted after adjusting for GVHD (relative risk 3.1 and 5.5, respectively).
This finding was consistent with prior reports of a greater relapse rate in patients with
syngeneic donors (relative risk 2.58, p = 0.008) compared to those with HLA-identical
siblings without GVHD (17), suggesting that GVL effects may occur without clini-
cally apparent GVHD. The advantage presented by a decreased risk of relapse in HLA-
identical siblings was offset to some degree by more transplant-related mortality.
2.3. Outcomes After T-Cell Depleted Grafts

TCD was introduced in an effort to reduce the incidence and severity of GVHD.
While successful in achieving this aim, this procedure resulted in an increased risk of
disease relapse and graft failure, particularly in patients with myeloid leukemias.
Apperley et al. (18) reported on patients with CML in chronic phase, receiving either
unmanipulated (n = 106) or TCD (n = 102) donor bone marrow. The probability of
remaining in remission after receiving a T-cell-replete graft (97% at 6 yr) was superior
to a TCD graft (67% at 3 yr). There was an increased risk of early (8 vs 0 patients) and
late graft failure (10 vs 0 patients) in the TCD graft recipients. These findings are
consistent with IBMTR data on patients with CML, which demonstrated that recipi-
ents of HLA-identical, T-cell-replete grafts who did not develop GVHD were seven
times less likely to relapse than those who received TCD grafts without developing
GVHD. Even those recipients of TCD grafts who developed GVHD had a higher risk
of relapse than those who received a T-cell-replete graft without GVHD (17). These
observations provided additional evidence to support the hypothesis that the immuno-
biology of GVL effects and GVHD are distinct at some level, providing hope that it
may ultimately be possible to promote the development of a GVL effect while pre-
venting GVHD.
2.4. Adoptive Immunotherapy With Donor Lymphocyte Infusions

Observations suggesting the presence of GVL effects were confirmed with the
successful use of donor-derived lymphocytes to achieve remissions in patients with
relapsed hematological malignancies (20). Initial studies combined interferon (IFN)-
with donor lymphocyte infusion (DLI) (20); however, it was subsequently shown that
the combination was not superior to DLI alone (22). More recent data suggest that the
424 Hogan and Deeg
composition of DLI is important in determining outcome, with a suggestion that the

risk of severe GVHD can be reduced without compromising efficacy by limiting the
T-cell (CD3+) content of the infusion (23). DLI results in complete remissions (CRs)
in a high percentage of patients with relapsed chronic-phase CML; however, CRs are
observed less frequently in other situations, including advanced CML and acute leu-
kemia.
Treatment-related mortality has been reported to be 7 to 9% for DLI from related
donors, and up to 44% for alternative donor transplants (2426). GVHD and tran-
sient or irreversible marrow hypoplasia are the major complications of DLI (27).
The reported incidence of acute (2170%) and chronic GVHD (2660%) appears to
depend on several factors, including interval between hematopoietic cell transplan-
tation (HCT) and DLI, disease type, cell dose and infusion schedule, and the devel-
opment of post-DLI marrow hypoplasia (22,24,25,2730). While the occurrence of
GVHD has been highly correlated with response in some studies (27), recent reports
suggest that both the T-cell dose and the infusion schedule may be important in
determining outcome, as a schedule of dose escalation may minimize the risk of
severe GVHD while preserving GVL effects. Raiola (24) reported that after HLA-
matched related DLI, the incidence of grades IIIV acute GVHD was 1%, 6%, and
23% for CD3+ cell doses per infusion of 1 106, 1 107, and 1 108/kg respec-
tively. Grade IV acute GVHD was only seen after 1 108 CD3+ cells per kg for
related donors but occurred with infusions of 1 107 CD3+ cells per kg from alter-
native donors. Dazzi (31) compared a single large dose of donor lymphocytes (bulk
dose regimen, BDR) with an escalating dose regimen (EDR). Although the prob-
ability of achieving cytogenetic remission did not differ significantly between
the two groups, the incidence of GVHD was much lower using EDR (10% vs 44%,
p = 0.01). After correcting for cell dose, the incidence and severity of acute and
chronic GVHD were both significantly lower for recipients treated by EDR. Factors
predicting response after DLI included the type and phase of disease, chemotherapy
responsiveness, and the development of GVHD or marrow hypoplasia (24,30). Bone-
marrow aplasia occurred in approx 1118% of patients and was more likely in
patients with predominant host hematopoiesis and hematologic relapse (22,27).
3. Disease Specificity of Graft-vs-Malignancy Effects

3.1. CML
With the introduction of DLI as a strategy to eradicate residual or relapsed disease
after HCT, it was noted that certain malignancies, such as CML, were particularly
sensitive to this therapeutic approach, while responses were less common in other
diseases. Overall, complete responses are observed in 6080% of CML patients who
receive DLI without preceding chemotherapy (24,26). Response rates vary according
to the amount of residual disease and response to prior chemotherapy, and may be as
high as 100% in patients with isolated molecular relapse, 89% for cytogenetic relapse,
7579% with chronic-phase relapse, compared to 1236% of patients with acceler-
ated-phase or blast-crisis CML (22,24,26). Other pre-DLI characteristics predictive of
complete response in CML patients are post-transplant chronic GVHD, time interval
between BMT and DLI less than 2 yr, and development of GVHD after DLI. Overall,
the probability of remaining in CR after a response has been suggested to be as high as
85%; however, this may also be dependent on disease status at time of DLI (2426).
Dazzi reported that the probability of survival for patients who achieve molecular
remissions was significantly better than for those who failed to do so (95% vs 53% at
3 yr after DLI, p = 0.0001) (32). Extrapolation of these findings to the initial transplant
approach has led to the increasing use of reduced-intensity or nonmyeloablative con-
ditioning followed by HCT and exploitation of the allogeneic effect in the treatment of
CML. While this modality has not been compared to conventional myeloablative HCT
in a randomized fashion, there are encouraging signs that it may be a safe and effective
alternative. Or et al. (33) reported data on 24 patients (median age 35 yr) with CML in
first chronic phase conditioned with fludarabine, busulfan, and antithymocyte globu-
lin (ATG). With median follow-up of 42 mo, 21 of 24 patients (88%) remained alive
and disease free, suggesting that myeloablative conditioning is not necessary to cure
diseases that are particularly susceptible to GVL effects.
3.2. AML/MDS
DLI alone is somewhat less effective for patients with relapsed AML, with CR
rates ranging from 15 to 38% (22,25,26). One study of 23 patients with AML undergo-
ing unrelated donor DLI reported a remission rate of 42%, with responders having a
1-y disease-free survival probability of 23% (29). DLI administered during isolated
molecular relapse or combined with chemotherapy have been suggested as ways to
improve results (34,35). Small studies have reported CRs in 2545% of patients with
myelodysplastic syndrome, and responses in polycythemia vera (PV) have also been
reported (22,25,26).
3.3. ALL
Although the observation that acute and chronic GVHD are associated with
reduced relapse rates for patients with ALL suggested that this disease might be ame-
nable to manipulation of the immune system, the results with DLI have been disap-
pointing (3638). CR rates after DLI alone for patients with hematological relapse
have been reported to be 025% (22,25,26). Collins et al. (39) evaluated patients from
27 transplant centers with persistent or recurrent ALL who received DLI with or with-
out preceding chemotherapy. Two of 15 patients who received no pre-DLI chemo-
therapy achieved CR, lasting 1112 and >764 d, respectively. Of 25 patients who
received DLI in the nadir after chemotherapy, 5 entered remission (median 83 d; range
42193 d). Seven patients who did not respond initially received a second DLI;
however, none of these patients attained durable remissions. Eighteen of 37 evaluable
patients (49%) developed acute GVHD, and 5 of 20 (25%) evaluable patients devel-
oped chronic GVHD. Actuarial overall survival was 13% at 3 yr. In those patients who
do respond to DLI, the duration of response tends to be less than for myeloid malig-
nancies (22). Despite these disappointing results, there is some evidence to suggest
that sustained responses may be possible if DLI is administered at the time of molecu-
lar relapse (24).
426 Hogan and Deeg
3.4. Myeloma
Anecdotal reports initially suggested a graft-vs-myeloma (GVM) effect (4044).
Lokhorst subsequently reported on 27 patients, 25 of whom had a prior partially TCD
graft (30). Of the 13 patients who received pre-induction chemotherapy, 8 responded,
but there were no complete responses. Overall, 14 of 27 patients (52%) responded to
DLI, with 6 complete remissions (22%). Five responses occurred after dose escala-
tion, and responses were more likely with chemotherapy-sensitive disease and a T-cell
dose >1 108/kg. Median overall survival for the entire group was 18 mo. Other
reports have also suggested that responses occur in approx 2550% of patients (26).
3.5. Lymphoid Malignancies

Several studies have provided evidence suggesting a role for allogeneic cells in the
treatment of lymphomas and chronic lymphocytic leukemia (CLL) (4,4549). Khouri
et al. (50) reported on outcomes of 20 patients with relapsed follicular or small-cell
lymphocytic lymphoma undergoing reduced intensity conditioning followed by HCT.
The conditioning regimen consisted of fludarabine and cyclophosphamide with or with-
out rituximab, and GVHD prophylaxis consisted of tacrolimus and methotrexate. At
the time of HCT, 12 patients were in CR, all other patients achieved CR after HCT. At
follow-up (median 21 mo), no patient had relapsed, and the actuarial probability of
disease-free survival at 2 yr was 84%. Branson et al. (51) also described the use of DLI
after progression or relapse following a reduced-intensity regimen based on fludarabine,
melphalan, and alemtuzumab (CAMPATH) for patients with refractory or relapsed lym-
phoid malignancies. DLI was associated with a response in 6 of the 12 patients (50%)
with progressive or relapsed disease. One patient with mantle-cell lymphoma achieved
CR with DLI alone, while two patients with Hodgkin disease (HD) achieved CR with a
combination of chemotherapy and DLI. One patient with HD and one with multiple
myeloma (MM) remained progression free at 9 and 12 mo after DLI, while another
patient with MM achieved partial response (PR) but died from GVHD and CMV infec-
tion 7 mo after DLI. In another study by Marks et al., 8 of 13 patients with persistent or
relapsed follicular lymphoma had complete responses to DLI, suggesting that low-grade
lymphoid malignancies are more susceptible to GVL effects (27).
Reports suggesting graft-vs-tumor effects in HD led to a matched case control
study by the European Group for Blood and Marrow Transplantation (EBMT) com-
paring outcomes after allogeneic HCT (n = 45) to outcomes after autologous HCT
(n = 45) for patients with relapsed HD (52). Patients were matched for age, gender,
disease status, conditioning regimen, and time from diagnosis to HCT. TRM was
very high in the allogeneic group, with 4-yr actuarial probabilities of survival, pro-
gression-free survival (PFS), and TRM of 25%, 15%, and 48% compared to 37%,
24%, and 27% after autologous HCT. However, there was an association between
grades IIIV acute GVHD with lower relapse rates after allogeneic HCT, consistent
with a graft-vs-Hodgkins-disease effect. Overall, it did not appear that there was a
major advantage to myeloablative allogeneic HCT; however, subsequent studies have
suggested that there is a continuing risk of relapse or secondary AML/MDS more
than a decade after autologous HCT; both complications are less likely after alloge-
neic HCT (53). The advent of nonmyeloablative regimens has raised the possibility
that less toxic approaches might be more successful, and it has even been suggested
that patients with heavily pretreated HD may be so immunosuppressed that they might
benefit from DLI without preceding HCT; however, this remains to be proven (54,55).
3.6. Solid Tumors

While the majority of reports describing graft-vs-malignancy (GVM) effects relate
to hematopoietic malignancies, this phenomenon has also been observed with certain
solid tumors, especially renal cell cancer and occasionally breast or ovarian cancer
(5664). Although there have been reports of graft-vs-malignancy effects in sarcomas
(64) and tumors originating from prostate (61), colon (62), and lung (63), responses
have generally not been consistent or dramatic enough to suggest clinical benefit.
Childs et al. (57) reported on 19 patients with metastatic renal-cell carcinoma receiv-
ing peripheral blood stem cell grafts from HLA-identical or single-antigen-mismatched
siblings after reduced-intensity conditioning. Efforts to induce graft-vs-tumor effects
included early withdrawal of postgrafting immunosuppression and use of DLI if nec-
essary. There were complete responses in three patients and PRs in seven patients,
resulting in a 53% overall response rate. Two patients died of non-relapse mortality
and eight died of progressive disease. Rini et al. (58) also reported on a reduced-inten-
sity conditioning regimen using cyclophosphamide and fludarabine for renal cell car-
cinoma. Of 12 patients who had follow-up to 180 d, 9 had sustained donor engraftment
and 4 died of TRM. Four patients had partial remissions. Bregni et al. (59) reported on
seven patients with renal-cell carcinoma who were treated with a reduced-intensity
regimen consisting of thiotepa, cyclophosphamide, and Flu. Four partial remissions
were observed; at the time of reporting, only one patient was alive with ongoing PR at
208 d, one was alive with progression, and two had died of transplant-related prob-
lems. Nevertheless, the presence of occasional dramatic responses in some patients
with bulky metastatic renal-cell carcinoma provides hope that an adoptive immuno-
therapy approach might be further refined to improve outcomes in a disease with no
other effective therapy.
Murine studies have suggested that breast carcinoma might be susceptible to graft-
vs-tumor effects (65,66). Bregni et al. (59) treated six patients with breast cancer with
a reduced-intensity conditioning regimen followed by allogeneic HCT, and reported
PRs in two patients, one of whom remained alive and in PR until the time of reporting
at d 417, while the other patient died in PR from GVHD at 490 d. Bay et al. (60)
reported on five patients with ovarian cancer receiving allogeneic transplantation from
HLA-matched siblings. One patient received a myeloablative conditioning regimen,
while the other four received non-myeloablative regimens. One recipient of a
nonmyeloablative regimen died soon after transplantation from progressive disease.
The remaining four patients developed GVHD associated with tumor regression. One
patient died of GVHD at 127 d; two of the remaining patients received DLI, resulting
in GVHD in both patients and GVM effect in one patient.
A number of factors have resulted in considering patients with malignant mela-
noma for adoptive immunotherapy approaches. These include (1) the observation that
428 Hogan and Deeg
spontaneous regressions have occasionally been reported, suggesting that in certain

circumstances this tumor may be responsive to immunotherapy; (2) the identification
of melanoma-specific tumor antigens; (3) the fact that eradication of normal melano-
cytes would not be expected to result in significant organ toxicity; and (4) the absence
of another effective therapy. Childs et al. (67) reported the results of nonmyeloablative
HCT on 15 patients with refractory metastatic melanoma. Four patients experienced a
partial remission; however, this appeared to be temporally associated with the condi-
tioning regimen. No clear GVM effect was seen after withdrawal of immunosuppres-
sion or institution of DLI. At the time of reporting, 1 patient remained alive (median
survival 86 d), 12 had died from progressive melanoma, and 2 died from TRM.
3.7. Posttransplantation Lymphoproliferative Disorders
Infectious and malignant complications are commonly encountered after HCT due
to delayed immune reconstitution. There is encouraging preliminary evidence to sug-
gest that adoptive immunotherapy might be of value in the prevention and treatment of
these complications. Despite advances in monitoring and therapy, cytomegalovirus
(CMV) infection remains a major source of morbidity and even mortality after HCT.
The recognition of the importance of T-cells in controlling CMV led to the adoptive
transfer of T-cell clones to restore antiviral immunity (6870). This technique is also
being evaluated for the prevention and treatment of diseases mediated by adeno-
virus and EpsteinBarr virus (EBV) (7174). EBV is instrumental in the development
of posttransplantation lymphoproliferative disease (PTLD). This is an unusual neo-
plasm that usually arises in B-lymphocytes of donor origin after HCT and those of
recipient origin after solid-organ transplantation. These neoplastic cells proliferate in
the absence of an effective T-cell response, and reconstitution of EBV-specific T-cell
immunity by adoptive immunotherapy often results in durable remissions (7579).
4. Mechanisms of GVL
Perhaps the greatest challenge in hematopoietic cell transplantation today is under-
standing the mechanisms that underlie GVL effects and GVHD. While these two pro-
cesses generally occur in concert, there is tantalizing evidence to suggest that there
might be sufficient differences between them to exploit in an immunotherapy strategy
(Fig. 1). Although the underlying immunobiology of GVL effects is not yet fully
understood, there is evidence to suggest that a number of mechanisms may be opera-
tive. For an individual, the extent of GVL may depend on several factors, including the
quantity or quality of polymorphic differences between the donor and host, the type of
malignancy, source of stem cells, and concomitant immunosuppression or infection.
4.1. Effector Cells
It is generally accepted that T-cells play a major role in generating GVL responses;
however, other cells almost certainly provide important contributions. Certain malig-
nant cells, such as CML blasts, may be able to act as antigen-presenting cells to donor
T-cells, thereby facilitating an effective immune response. It is also increasingly recog-
nized that natural killer cells may play a role, particularly in the HLA-C mismatched
setting.
Fig. 1. Tissue-specific expression of minor histocompatibility antigens may permit selec-

tive graft-vs-leukemia activity.
4.1.1. T-Cells
Several lines of evidence suggest a major role for T-cells in the generation of GVL.
Donor T-cells have been shown to be essential for GVL in murine models of leukemia
(80). Further studies in severe combined immunodeficiency mice with established
human leukemia growth suggested that the role of T-cells in eradicating tumor cells is
also important in humans (81). The identification of CD4+ and CD8+ T-cell clones
directed specifically against leukemic cells, and retrospective studies demonstrating
increased relapse rates after TCD hematopoietic cell transplantation, provide addi-
tional support in favor of a major role for T-cells in mediating GVL effects in humans
(17,8285).
There is debate regarding the relative importance of specific T-cell subsets in the
immunobiology of GVL effects and GVHD. Evidence in support of a predominant
role for CD4+ cells include murine studies where either CD8+-depleted marrow or
CD4+-supplemented marrow was transplanted. Despite a low incidence of GVHD,
leukemia-free survival appeared to be preserved in some strain combinations (86).
Clinical studies of CD8+-depleted hematopoietic grafts or DLI also appear to have
anti-leukemic activity with less GVHD, reinforcing the contribution of the CD4+
population in generating GVL reactions (8789). Claret et al. analyzed the T-cell
receptor (TCR) repertoire in four patients with relapsed CML who achieved a com-
plete remission (3/4 without GVHD) after infusion of CD4+ lymphocytes from HLA-
identical sibling donors (90). Before DLI, all four patients were found to have a
restricted TCR V repertoire in peripheral T-cells, while after DLI, new V clones
were identified coincident with cytogenetic responses, suggesting the possibility of
selective T-cell eradication of leukemic cells. Although CD4+ T-cells from trans-
plant recipients have been shown to lyse leukemia cells, alloreactivity towards nor-
430 Hogan and Deeg
mal host cells is well described, suggesting a role for CD4+ cells in mediating both
GVL effects and GVHD (91,92). In other models, CD8+ cells appear to be the pri-
mary mediators of GVL, with CD4+ cells facilitating clonal expansion of CD8+ cells
(93,94). It is likely that at some level, GVH and GVL activities can be separated
despite the existence of some effector cells that exhibit both activities, and it has
been suggested that selective TCD methods might be feasible (82).
4.1.2. Antigen-Presenting Cells
To optimize GVL potency, the malignant cell should act as both a stimulator and a
target for donor effector mechanisms (95). Dendritic cells represent the most effective
antigen-presenting cells, with the ability to process and express antigenic peptides and
initiate cell-mediated and humoral immune responses. In certain malignancies, such
as CML, the leukemic clone is derived from a pluripotent hematopoietic cell with the
potential to generate leukemic dendritic cells. These and other antigen-presenting cells
harbor the chimeric BCR/abl fusion protein and may present leukemia-specific anti-
gens to donor T-cells via class I or II major histocompatability complex (MHC) anti-
gens (96). This may be part of the explanation for better responses to DLI in CML, and
augmentation of this ability may provide a means to achieve even greater success
(97,98).
4.1.3. Natural Killer Cells
Many tumors do not present tumor-specific antigens and have down-regulated
class I MHC, thereby avoiding recognition by T-cells (99). This means that T-cell-
mediated GVL reactions are rendered either ineffective or less specific, relying on
recognition of minor antigen targets expressed on both malignant and normal host
cells. Natural killer (NK) cells do not depend on specific antigen recognition for tar-
get-cell lysis, rendering them very interesting from a tumor immunology perspective.
While there has long been optimism that NK cells might provide a valuable contribu-
tion to successful tumor immunotherapy, this goal has been slower to realize (100).
To date, most strategies utilizing NK cells in the treatment of malignancies have
employed ex vivo generation of lymphokine-activated killer (LAK) cells or activated
NK cells, or in vivo administration of cytokines to activate NK cells against autolo-
gous cancer cells. In general these studies have been somewhat disappointing, lim-
ited by lack of efficacy or excessive toxicity, or both (101106). However, the recent
identification and characterization of cell-surface receptors including killer cell
immunoglobulinlike receptors (KIRs), C-type lectins, and natural cytotoxicity recep-
tors (NCRs) have provided encouragement that the properties of NK cells can finally
be exploited (107). There is a wealth of new information becoming available, eluci-
dating the activating and inhibitory pathways mediated through an array of NK
receptors, promising a more sophisticated approach than previous non-specific inter-
vention using cytokines such as interleukin (IL)-2. Whereas evidence to suggest that
NK cells may contribute to antileukemic effects in the HLA-matched setting is lim-
ited, a potent GVL effect of NK cells is more convincing in the haplo-identical set-
ting (100,108113). In this allogeneic situation, KIR mismatches facilitate leukemic
cell lysis without GVHD (114,115).
4.2. Antigenic Targets

Antigenic targets for GVL effects comprise two main categories: (1) antigenic dif-
ferences that occur as a result of polymorphisms between donor and host (minor
histocompatability antigens); and (2) tumor-specific antigens. Effective eradication of
malignant cells may depend on an immune response directed against either or both of
these targets.
4.2.1. Minor Histocompatability Antigens

Minor histocompatability antigens (mHA) are alloantigens capable of eliciting an
allogeneic T-cell response between HLA-matched individuals and occur due to the
presence of polymorphic peptides (116). Minor antigen disparities are presumably the
basis for the difference in relapse rates between recipients of hematopoietic grafts
from syngeneic and HLA-matched sibling donors. Multiple minor antigens, derived
from both autosomal and sex chromosomes, have already been described (117123).
Polymorphisms in proteins may result in either a previously unencountered peptide or,
alternatively, differential antigen processing, either of which may lead to presentation
of a neo-antigen (124,125). Such antigens may be restricted to certain tissues, such as
hematopoietic cells, or be expressed by a multitude of tissue types, fueling enthusiasm
that an allogeneic effect could be directed against malignant hematopoeitic cells in a
relatively specific fashion (126,127). One can imagine that either GVL or GVHD could
be seen alone or in combination, depending on the distribution of mHA for a given
donorhost combination. The eradication of malignant cells in the absence of clinical
GVHD, noted in some patients after HCT or DLI, has provided additional encourage-
ment that this approach may ultimately be successful. These findings have fostered a
major effort to characterize immunodominant mHAs to be used as targets for immu-
notherapy.
4.2.2. Tumor Antigens

As malignant cells are derived from autologous cells, it is not entirely surprising that
many tumor antigens are actually self-antigens, which are overexpressed in malig-
nant cells. In a recent review, Pardoll describes four different categories of tumor anti-
gens (128). These include:
1. Unique antigens created as a result of a mutation. These include bcr-abl, PML/RAR,
ETV6-AML, DEK-CAN, and activating mutations such as p21ras or inactivating muta-
tions in cyclin-dependent kinase 4 (CDK4) (129140).
2. Viral antigens such as human papilloma virus 16 E6 and E7, which are expressed on most
cervical cancers. Other oncogenic viruses, such as hepatitis B and C and EBV, may also
provide useful tumor antigens.
3. Tissue-specific differentiation antigens. Proteinase 3 is a serine protease with several
known polymorphisms found in high concentrations in the primary granules of acute and
chronic myeloid blasts. Cytotoxic T lymphocytes (CTL) recognizing an octameric HLA-
A2-restricted peptide (PR1) derived from proteinase 3 are capable of specific killing of
myeloid leukemic cells. Such CTL occur spontaneously after allogeneic HCT and may be
at least partially responsible for the curative potential of allogeneic transplantation in this
disease (141143). Clave et al. (144) demonstrated that peptides derived from a specific
432 Hogan and Deeg
polymorphic site bind to the HLA-A2 molecule. In HLA-A2-expressing patients with

myeloid leukemia undergoing HCT from HLA-identical donors, no relapses occurred in a
group of 4 patients who had at least one allele absent in their donor, whereas 7 of the 15
remaining patients relapsed, raising the possibility of an appropriate target for T-cell
immunotherapy. Other examples include the occurrence of T-cell responses directed
against normal melanocytes, resulting in vitiligo, an observation that has been noted after
immunotherapy for melanoma, or rarely after HCT. Putative antigens include tyrosinase,
tyrosinase-related protein-1 (TRP-1), TRP-2, gp 100, and MART1/melan-A (128,145,146)
4. Tumor-selective antigens. This group includes antigens that are expressed at very low
levels in normal tissues but are up-regulated in tumors. WT1 encodes a transcription fac-
tor expressed in embryogenesis and is rarely encountered in normal adult tissue. It is
overexpressed in many types of leukemia and can elicit CTL responses that lyse leukemia
cells in some circumstances (147,148).
4.3. Effector Mechanisms

Even after appropriate recognition and clonal expansion by the donor immune sys-
tem, GVL effects will occur only if malignant cells are susceptible to the effector
mechanisms employed. Such mechanisms include direct lysis via the secretion of
perforin and granzymes from cytotoxic cells, or induction of apoptosis via fas ligand
(FasL) (95,149151). However, reduced susceptibility of malignant cells to these
pathways have been reported (152,153). In a murine model, Schmaltz et al. examined
the relative contributions the perforin and FasL pathways in mediating GVHD and
GVL effects (154). In this model, donor T-cells mediated GVHD activity primarily
through the FasL effector pathway, and GVL effects through the perforin pathway,
suggesting a possible point of intervention to augment GVL effects while minimizing
GVHD. Some T-cell clones demonstrate strong cytotoxic activity that is not mediated
by either the FasL or perforin pathway (155). Recently additional effector mecha-
nisms have been identified, including tumor necrosis factor (TNF)-related apoptosis-
inducing ligand (TRAIL), a member of the TNF superfamily. It participates in NK
cell-mediated antitumor activity, and is also expressed on T-cells. In murine studies,
TRAIL-expressing T-cells demonstrated greater anti-tumor activity but not more
GVHD than TRAIL-deficient T-cells, suggesting another possible means to accentu-
ate GVL effects without GVHD (156,157).
Effector cells also depend on cytokines to mediate some of their anti-tumor ef-
fects. Cytokine gene polymorphisms can result in genetically fixed differences in
cytokine responses to stimuli such as lipopolysaccharide (LPS). Investigation of
cytokine gene polymorphisms has not yet been explored in detail with respect to how
they might influence GVL activity; however, it has been proposed that certain poly-
morphisms are associated with the severity of acute (TNF, IL-10, IFN-) and chronic
GVHD (IL-6) after allogeneic HCT (158161). Cytokine gene polymorphisms are
also suggested to influence rejection after solid-organ transplantation (162165), rais-
ing the possibility that they may also be important in mediating GVL activity.
5. Conclusions
The GVM effects of allogeneic HCT make this therapeutic modality the most
effective treatment for a range of malignancies. Recent progress has permitted older
and less robust patients to take advantage of this curative therapy by reducing trans-
plant-related complications and mortality associated with high-intensity condition-
ing regimens. As adoptive immunotherapy using nonmyeloablative conditioning and
DLI become established as a standard therapeutic intervention, further advances in
our understanding of transplantation immunobiology have suggested potential strat-
egies that might be exploited to maximize GVL effects while minimizing toxicity.
These include the use of targeted dosing of cytotoxic agents, the combination of
autologous HCT with delayed allografting using minimally myelosuppressive con-
ditioning, and the development of more effective nonmyeloablative conditioning
and postgrafting immunosuppression regimens. Appropriate dosing and interven-
tion have enhanced the safety and effectiveness of DLI. Strategies to reproducibly
separate the toxicity of GVHD from the beneficial GVL effect are under intense
investigation; however, this important goal has yet to be fully realized in the clinical
setting.
References
1. Porter, D. L. and Antin, J. H. (1999) The graft-versus-leukemia effects of allogeneic cell
therapy. [Review]. Annu. Rev. Med. 50, 369386.
2. Clift, R. A., Buckner, C. D., Appelbaum, F. R., et al. (1990) Allogeneic marrow transplan-
tation in patients with acute myeloid leukemia in first remission: a randomized trial of two
irradiation regimens. Blood 76, 18671871.
3. Deeg, H. J., Storer, B., Slattery, J. T., et al. (2002) Conditioning with targeted busulfan
and cyclophosphamide for hemopoietic stem cell transplantation from related and unre-
lated donors in patients with myelodysplastic syndrome. Blood 100, 12011207.
4. McSweeney, P. A., Niederwieser, D., Shizuru, J. A., et al. (2001) Hematopoietic cell trans-
plantation in older patients with hematologic malignancies: replacing high-dose cytotoxic
therapy with graft-versus-tumor effects. Blood 97, 33903400.
5. Slavin, S., Nagler, A., Naparstek, E., et al. (1998) Nonmyeloablative stem cell transplan-
tation and cell therapy as an alternative to conventional bone marrow transplantation with
lethal cytoreduction for the treatment of malignant and nonmalignant hematologic dis-
eases. Blood 91, 756763.
6. Giralt, S., Estey, E., Albitar, M., et al. (1997) Engraftment of allogeneic hematopoietic
progenitor cells with purine analog-containing chemotherapy: harnessing graft-versus-leu-
kemia without myeloablative therapy. Blood 89, 45314536.
7. Mielcarek, M., Martin, P. J., Leisenring, W., et al. (2003) Graft-versus-host disease
after nonmyeloablative versus conventional hematopoietic stem cell transplantation. Avail-
able online at First Edition Paper link, see Website: http://www.bloodjournal.org/. Blood
102(2), 756762.
8. Weiden, P. L., Storb, R., Tsoi, M. S., Graham, T. C., Lerner, K. G., and Thomas, E. D.
(1976) Infusion of donor lymphocytes into stable canine radiation chimeras: implications
for mechanism of transplantation tolerance. J. Immunol. 116, 12121219.
9. Weiden, P. L., Sullivan, K. M., Flournoy, N., Storb, R., and Thomas, E. D. (1981) Antileu-
kemic effect of chronic graft-versus-host disease: contribution to improved survival after
allogeneic marrow transplantation. New Engl. J. Med. 304, 15291533.
10. Weiden, P. L., Flournoy, N., Thomas, E. D., et al. (1979) Antileukemic effect of graft-
versus-host disease in human recipients of allogeneic-marrow grafts. New Engl. J. Med.
300, 10681073.
434 Hogan and Deeg
11. Sullivan, K. M., Weiden, P. L., Storb, R., 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.
12. Odom, L. F., August, C. S., Githens, J. H., et al. (1978) Remission of relapsed leukaemia
during a graft-versus-host reaction. A graft-versus-leukaemia reaction in man? Lancet
2, 537540.
13. Collins, R. H., Jr., Rogers, Z. R., Bennett, M., Kumar, V., Nikein, A., and Fay, J. W.
(1992) Hematologic relapse of chronic myelogenous leukemia following allogeneic bone
marrow transplantation: apparent graft-versus-leukemia effect following abrupt discon-
tinuation of immunosuppression. Bone Marrow Transplant. 10, 391395.
14. Higano, C. S., Brixey, M., Bryant, E. M., et al. (1990) Durable complete remission of
acute nonlymphocytic leukemia associated with discontinuation of immunosuppression
following relapse after allogeneic bone marrow transplantation. A case report of a prob-
able graft-versus-leukemia effect. Transplantation 50, 175177.
15. Fefer, A., Sullivan, K. M., Weiden, P., et al. (1987) Graft versus leukemia effect in man:
the relapse rate of acute leukemia is lower after allogeneic than after syngeneic marrow
transplantation. Prog. Clin. Biol. Res. 244, 401408.
16. Gale, R. P., Horowitz, M. M., Ash, R. C., et al. (1994) Identical-twin bone marrow trans-
plants for leukemia. Ann. Intern. Med. 120, 646652.
17. Horowitz, M. M., Gale, R. P., Sondel, P. M., et al. (1990) Graft-versus-leukemia reactions
after bone marrow transplantation. Blood 75, 555562.
18. Apperley, J. F., Mauro, F. R., Goldman, J. M., et al. (1988) Bone marrow transplantation
for chronic myeloid leukaemia in first chronic phase: importance of a graft-versus-leu-
kaemia effect.[erratum; Br. J. Haematol. 70(2), 261]. Br. J. Haematol. 69, 239245.
19. Marmont, A. M., Horowitz, M. M., Gale, R. P., et al. (1991) T-cell depletion of HLA-
identical transplants in leukemia. Blood 78, 21202130.
20. Kolb, H. J., Mittermuller, J., Clemm, C., et al. (1990) Donor leukocyte transfusions for
treatment of recurrent chronic myelogenous leukemia in marrow transplant patients. Blood
76, 24622465.
21. Barnes, D. W. H., 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. 2, 626627.
22. Kolb, H. J., Schattenberg, A., Goldman, J. M., et al. (1995) Graft-versus-leukemia effect
of donor lymphocyte transfusions in marrow grafted patients. European Group for Blood
and Marrow Transplantation Working Party Chronic Leukemia. Blood 86, 20412050.
23. Guglielmi, C., Arcese, W., Dazzi, F., et al. (2002) Donor lymphocyte infusion for relapsed
chronic myelogenous leukemia: prognostic relevance of the initial cell dose. Blood 100,
397405.
24. Raiola, A. M., Van Lint, M. T., Valbonesi, M., et al. (2003) Factors predicting response
and graft-versus-host disease after donor lymphocyte infusions: a study on 593 infusions.
Bone Marrow Transplant. 31, 687693.
25. Shiobara, S., Nakao, S., Ueda, M., et al. (2001) Donor leukocyte infusion for Japanese
patients with relapsed leukemia after allogeneic bone marrow transplantation: indications
and dose escalation. Ther. Apheresis 5, 4045.
26. Collins, R. H., Jr., Shpilberg, O., Drobyski, W. R., et al. (1997) Donor leukocyte infusions
in 140 patients with relapsed malignancy after allogeneic bone marrow transplantation. J.
Clin. Oncol. 15, 433444.
27. Marks, D. I., Lush, R., Cavenagh, J., et al. (2002) The toxicity and efficacy of donor
lymphocyte infusions given after reduced-intensity conditioning allogeneic stem cell trans-
plantation. Blood 100, 31083114.
28. de Lima, M., Bonamino, M., Vasconcelos, Z., et al. (2001) Prophylactic donor lympho-
cyte infusions after moderately ablative chemotherapy and stem cell transplantation for
hematological malignancies: high remission rate among poor prognosis patients at the
expense of graft-versus-host disease. Bone Marrow Transplant. 27, 7378.
29. Porter, D. L., Collins, R. H., Jr., Hardy, C., et al. (2000) Treatment of relapsed leukemia
after unrelated donor marrow transplantation with unrelated donor leukocyte infusions.
Blood 95, 12141221.
30. Lokhorst, H. M., Schattenberg, A., Cornelissen, J. J., et al. (2000) Donor lymphocyte infu-
sions for relapsed multiple myeloma after allogeneic stem-cell transplantation: predictive
factors for response and long-term outcome. J. Clin. Oncol. 18, 30313037.
31. Dazzi, F., Szydlo, R. M., Craddock, C., et al. (2000) Comparison of single-dose and esca-
lating-dose regimens of donor lymphocyte infusion for relapse after allografting for
chronic myeloid leukemia. Blood 95, 6771.
32. Dazzi, F., Szydlo, R. M., Cross, N. C., et al. (2000) Durability of responses following
donor lymphocyte infusions for patients who relapse after allogeneic stem cell transplan-
tation for chronic myeloid leukemia. Blood 96, 27122716.
33. Or, R., Shapira, M. Y., Resnick, I., et al. (2003) Nonmyeloablative allogeneic stem cell
transplantation for the treatment of chronic myeloid leukemia in first chronic phase. Blood
101, 441445.
34. Au, W. Y., Lie, A. K., Lee, C. K., Liang, R., and Kwong, Y. L.. (1999) Donor lymphocyte
infusion induced molecular remission in relapse of acute myeloid leukaemia after alloge-
neic bone marrow transplantation. Bone Marrow Transplant. 23, 12011203.
35. Lee, J. H., Lee, K. H., Kim, S., et al. (2001) Combination chemotherapy of intermediate-
dose cytarabine, idarubicin, plus etoposide and subsequent mobilized donor leukocyte in-
fusion for relapsed acute leukemia after allogeneic bone marrow transplantation. Leukemia
Res. 25, 305312.
36. Zikos, P., Van Lint, M. T., Lamparelli, T., et al. (1998) Allogeneic hemopoietic stem cell
transplantation for patients with high risk acute lymphoblastic leukemia: favorable impact
of chronic graft-versus-host disease on survival and relapse. Haematologica 83, 896903.
37. Appelbaum, F. R. (1997) Graft versus leukemia (GVL) in the therapy of acute lympho-
blastic leukemia (ALL). [Review] Leukemia 11(Suppl 4), S1517.
38. Passweg, J. R., Tiberghien, P., Cahn, J. Y., et al. (1998) Graft-versus-leukemia effects in T
lineage and B lineage acute lymphoblastic leukemia. Bone Marrow Transplant. 21, 153158.
39. Collins, R. H., Jr., Goldstein, S., Giralt, S., et al. (2000) Donor leukocyte infusions in
acute lymphocytic leukemia. Bone Marrow Transplant. 26, 511516.
40. Zomas, A., Stefanoudaki, K., Fisfis, M., Papadaki, T., and Mehta, J.. (1998) Graft-versus-
myeloma after donor leukocyte infusion: maintenance of marrow remission but extramed-
ullary relapse with plasmacytomas. Bone Marrow Transplant. 21, 11631165.
41. Tricot, G, Vesole, DH, Jagannath, S, Hilton, J, Munshi, N, and Barlogie, B. (1996) Graft-
versus-myeloma effect: proof of principle. Blood 87, 11961198.
42. Alyea, E., Weller, E., Schlossman, R., et al. (2001) T-cell-depleted allogeneic bone mar-
row transplantation followed by donor lymphocyte infusion in patients with multiple
myeloma: induction of graft-versus-myeloma effect. Blood 98, 934939.
43. Verdonck, L. F., Lokhorst, H. M., Dekker, A. W., Nieuwenhuis, H. K., and Petersen, E. J.
(1996) Graft-versus-myeloma effect in two cases. Lancet 347, 800801.
436 Hogan and Deeg
44. Bertz, H., Burger, J. A., Kunzmann, R., Mertelsmann, R., and Finke, J. (1997) Adoptive
immunotherapy for relapsed multiple myeloma after allogeneic bone marrow transplanta-
tion (BMT): evidence for a graft-versus-myeloma effect. Leukemia 11, 281283.
45. Mehta, J., Powles, R., Singhal, S., Iveson, T., Treleaven, J., and Catovsky, D. (1996) Clini-
cal and hematologic response of chronic lymphocytic and prolymphocytic leukemia per-
sisting after allogeneic bone marrow transplantation with the onset of acute
graft-versus-host disease: possible role of graft-versus-leukemia. Bone Marrow Trans-
plant. 17, 371375.
46. Rondon, G., Giralt, S., Huh, Y., et al. (1996) Graft-versus-leukemia effect after allogeneic
bone marrow transplantation for chronic lymphocytic leukemia. Bone Marrow Transplant.
18, 669672.
47. Dreger, P. and Montserrat, E. (2002) Autologous and allogeneic stem cell transplantation
for chronic lymphocytic leukemia. [Review]. Leukemia 16, 985992.
48. Sohn, S. K., Baek, J. H., Kim, D. H., et al. (2000) Successful allogeneic stem-cell trans-
plantation with prophylactic stepwise G-CSF primed-DLIs for relapse after autologous
transplantation in mantle cell lymphoma: a case report and literature review on the evi-
dence of GVL effects in MCL. [Review]. Am. J. Hematol. 65, 7580.
49. Khouri, I. F., Keating, M., Korbling, M., et al. (1998) Transplant-lite: induction of graft-
versus-malignancy using fludarabine-based nonablative chemotherapy and allogeneic
blood progenitor-cell transplantation as treatment for lymphoid malignancies. J. Clin.
Oncol. 16, 28172824.
50. Khouri, I. F., Saliba, R. M., Giralt, S. A., et al. (2001) Nonablative allogeneic hematopoietic
transplantation as adoptive immunotherapy for indolent lymphoma: low incidence of toxic-
ity, acute graft-versus-host disease, and treatment-related mortality. Blood 98, 35953599.
51. Branson, K., Chopra, R., Kottaridis, P. D., et al. (2002) Role of nonmyeloablative alloge-
neic stem-cell transplantation after failure of autologous transplantation in patients with
lymphoproliferative malignancies. J. Clin. Oncol. 20, 40224031.
52. Milpied, N., Fielding, A. K., Pearce, R. M., Ernst, P., and Goldstone, A. H. (1996) Alloge-
neic bone marrow transplant is not better than autologous transplant for patients with
relapsed Hodgkins disease. European Group for Blood and Bone Marrow Transplantation.
J. Clin. Oncol. 14, 12911296.
53. Akpek, G., Ambinder, R. F., Piantadosi, S., et al. (2001) Long-term results of blood and
marrow transplantation for Hodgkins lymphoma. J. Clin. Oncol. 19, 43144321.
54. Carella, A. M., Cavaliere, M., Lerma, E., et al. (2000) Autografting followed by
nonmyeloablative immunosuppressive chemotherapy and allogeneic peripheral-blood
hematopoietic stem-cell transplantation as treatment of resistant Hodgkins disease and
non-Hodgkins lymphoma. J. Clin. Oncol. 18, 39183924.
55. Porter, D. L., Stadtmauer, E. A., and Lazarus, H. M. (2003) GVHD: graft-versus-host
disease or graft-versus-Hodgkins disease? an old acronym with new meaning. Bone Mar-
row Transplant. 31, 739746.
56. Eibl, B., Schwaighofer, H., Nachbaur, D., et al. (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, 15011508.
57. Childs, R., Chernoff, A., Contentin, N., et al. (2000) Regression of metastatic renal-cell
carcinoma after nonmyeloablative allogeneic peripheral-blood stem-cell transplantation.
New Engl. J. Med. 343, 750758.
58. Rini, B. I., Zimmerman, T., Stadler, W. M., Gajewski, T. F., and Vogelzang, N. J. (2002)
Allogeneic stem-cell transplantation of renal cell cancer after nonmyeloablative chemo-
therapy: feasibility, engraftment, and clinical results.[comment]. J. Clin. Oncol. 20,

20172024.
59. Bregni, M., Dodero, A., Peccatori, J., et al. (2002) Nonmyeloablative conditioning fol-
lowed by hematopoietic cell allografting and donor lymphocyte infusions for patients with
metastatic renal and breast cancer. Blood 99, 42344236.
60. Bay, J. O., Fleury, J., Choufi, B., et al. (2002) Allogeneic hematopoietic stem cell trans-
plantation in ovarian carcinoma: results of five patients. Bone Marrow Transplant. 30,
95102.
61. Peccatori, J. Ciceri, F., Bernardi, M., et al. (2002) Evidence of allogeneic graft-versus-
tumor effect in prostate and ovarian cancer. Haematologica 87(Suppl 1), 1214.
62. Zetterquist, H., Hentschke, P., Thorne, A., et al. (2001) A graft-versus-colonic cancer
effect of allogeneic stem cell transplantation. Bone Marrow Transplant. 28, 11611166.
63. Moscardo, F., Martinez, J. A., Sanz, G. F., et al. (2000) Graft-versus-tumour effect in non-
small-cell lung cancer after allogeneic peripheral blood stem cell transplantation. Br. J.
Haematol. 111, 708710.
64. Pedrazzoli, P., Da Prada, G. A., Giorgiani, G., et al. (2002) Allogeneic blood stem cell
transplantation after a reduced-intensity, preparative regimen: a pilot study in patients
with refractory malignancies. Cancer 94, 24092415.
65. 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, 5963.
66. Morecki, S., Yacovlev, E., Diab, A., and Slavin, S. (1998) Allogeneic cell therapy for a
murine mammary carcinoma. Cancer Research. 58, 38913895.
67. Childs, R. W., Bradstock, K. F., Gottlieb, D., Kefford, R., and Barrett, J. (2000) Non-
myeloablative allogeneic stem cell transplantation as immunotherapy for metastatic mela-
noma: results of a pilot study. Blood 96(Part 2), 353b. .
68. Foster, A. E., Gottlieb, D. J., Marangolo, M., et al. (2003) Rapid, large-scale generation of
highly pure cytomegalovirus-specific cytotoxic T cells for adoptive immunotherapy. J.
Hematother. Stem Cell Res. 12, 93105.
69. Walter, E. A., Greenberg, P. D., Gilbert, M. J., et al. (1995) Reconstitution of cellular
immunity against cytomegalovirus in recipients of allogeneic bone marrow by transfer of
T-cell clones from the donor. New Engl. J. Med. 333, 10381044.
70. Riddell, S. R., Watanabe, K. S., Goodrich, J. M., Li, C. R., Agha, M. E., and Greenberg, P.
71. Gottschalk, S., Edwards, O. L., Sili, U., et al. (2003) Generating CTLs against the sub-
dominant Epstein-Barr virus LMP1 antigen for the adoptive immunotherapy of EBV-
associated malignancies. Blood 101, 19051912.
72. Hamel, Y., Blake, N., Gabrielsson, S., et al. (2002) Adenovirally transduced dendritic
cells induce bispecific cytotoxic T lymphocyte responses against adenovirus and cytome-
galovirus pp65 or against adenovirus and Epstein-Barr virus EBNA3C protein: a novel
approach for immunotherapy. Hum. Gene Ther. 13, 855866.
73. Regn, S., Raffegerst, S., Chen, X., Schendel, D., Kolb, H. J., and Roskrow, M.. (2001)
Ex vivo generation of cytotoxic T lymphocytes specific for one or two distinct viruses
for the prophylaxis of patients receiving an allogeneic bone marrow transplant. Bone
Marrow Transplant. 27, 5364.
74. Hromas, R., Cornetta, K., Srour, E., Blanke, C., and Broun, E. R. (1994) Donor leukocyte
infusion as therapy of life-threatening adenoviral infections after T-cell-depleted bone
marrow transplantation. Blood 84, 16891690.
438 Hogan and Deeg
75. Rooney, C. M., Smith, C. A., Ng, C. Y., et al. (1998) Infusion of cytotoxic T cells for the
prevention and treatment of Epstein-Barr virus-induced lymphoma in allogeneic trans-
plant recipients. Blood 92, 15491555.
76. Emanuel, D. J., Lucas, K. G., Mallory, G. B., Jr., et al. (1997) Treatment of posttransplant
lymphoproliferative disease in the central nervous system of a lung transplant recipient
using allogeneic leukocytes. Transplantation 63, 16911694.
77. Comoli, P., Labirio, M., Basso, S., et al. (2002) Infusion of autologous Epstein-Barr virus
(EBV)-specific cytotoxic T cells for prevention of EBV-related lymphoproliferative dis-
order in solid organ transplant recipients with evidence of active virus replication. Blood
99, 25922598.
78.
78 Haque, T., Taylor, C., Wilkie, G. M., et al. (2001) Complete regression of posttransplant
lymphoproliferative disease using partially HLA-matched Epstein Barr virus-specific
cytotoxic T cells. Transplantation 72, 13991402.
79.
79 Restrepo, A., Albrecht, F., Raez, L. E., et al. (1999) Post-liver transplantation
lymphoproliferative disorders with and without infusions of donor bone marrow cells.
[Review] Crit. Rev. Oncogen. 10, 239245.
80. Truitt, R. L., Johnson, B. D., McCabe, C., and Weiler, M. B. (1997) Graft versus leuke-
mia. In: Graft- vs-Host Disease. Ferrara, J. L. M, Deeg, H. J., and Burakoff, S. J., eds.
Marcel Dekker, New York, NY, p. 385.
81.
81 Malkovska, V., Cigel, F., and Storer, B. E. (1994) Human T cells in hu-PBL-SCID mice
proliferate in response to Daudi lymphoma and confer anti-tumour immunity. Clin. Exp.
Immunol. 96, 158165.
82.
82 van Lochem, E., de Gast, B., and Goulmy, E. (1992) In vitro separation of host specific
graft-versus-host and graft-versus-leukemia cytotoxic T cell activities. Bone Marrow Trans-
plant. 10, 181183.
83 Sosman, J. A., Oettel, K. R., Smith, S. D., Hank, J. A., Fisch, P., and Sondel, P. M. (1990)
83.
Specific recognition of human leukemic cells by allogeneic T cells: II. Evidence for HLA-
D restricted determinants on leukemic cells that are crossreactive with determinants present
on unrelated nonleukemic cells. Blood 75, 20052016.
84.
84 Kondo, Y., Shiobara, S., and Nakao, S. (2001) Identification of T-cell clones showing
expansion associated with graft-vs-leukemia effect on chronic myelogenous leukemia in
vivo and in vitro. Exp. Hematol. 29, 471476.
85.
85 Jiang, Y. Z., Kanfer, E. J., Macdonald, D., Cullis, J. O., Goldman, J. M., and Barrett, A. J.
(1991) Graft-versus-leukaemia following allogeneic bone marrow transplantation: emer-
gence of cytotoxic T lymphocytes reacting to host leukaemia cells. Bone Marrow Trans-
plant. 8, 253258.
86.
86 Korngold, R., and Sprent, J. (1987) T cell subsets and graft-versus-host disease. Trans-
plantation 44, 335339.
87.
87 Giralt, S., Hester, J., Huh, Y., et al. (1995) CD8-depleted donor lymphocyte infusion as
treatment for relapsed chronic myelogenous leukemia after allogeneic bone marrow trans-
88.
88 Nimer, S. D., Giorgi, J., Gajewski, J. L., et al. (1994) Selective depletion of CD8+ cells for
prevention of graft-versus-host disease after bone marrow transplantation. A randomized
controlled trial. Transplantation 57, 8287.
89.
89 Alyea, E. P., Soiffer, R. J., Canning, C., et al. (1998) Toxicity and efficacy of defined
doses of CD4(+) donor lymphocytes for treatment of relapse after allogeneic bone marrow
transplant. Blood 91, 36713680.
90.
90 Claret, E. J., Alyea, E. P., Orsini, E., et al. (1997) Characterization of T cell repertoire in
patients with graft-versus-leukemia after donor lymphocyte infusion. J. Clin. Invest. 100,
855866.
91.
91 Faber, L. M., van Luxemburg-Heijs, S. A., Veenhof, W. F., Willemze, R., and
Falkenburg, J. H. (1995) Generation of CD4+ cytotoxic T-lymphocyte clones from a
patient with severe graft-versus-host disease after allogeneic bone marrow transplanta-
tion: implications for graft-versus-leukemia reactivity. Blood 86, 28212828.
92.
92 Jiang, Y. Z., and Barrett, A. J. (1995) Cellular and cytokine-mediated effects of CD4-
positive lymphocyte lines generated in vitro against chronic myelogenous leukemia. Exp.
Hematol. 23, 11671172.
93.
93 Truitt, R. L., and Johnson, B. D. (1995) Principles of graft-vs.-leukemia reactivity. [Review]
Biol. Blood Marrow Transplant. 1, 6168.
94.
94 Palathumpat, V., Dejbakhsh-Jones, S., and Strober, S. (1995) The role of purified CD8+ T
cells in graft-versus-leukemia activity and engraftment after allogeneic bone marrow trans-
plantation. Transplantation 60, 355361.
95.
95 Barrett, A. J. (1997) Mechanisms of the graft-versus-leukemia reaction. [Review] Stem
Cells 15, 248258.
96.
96 Eisendle, K., Lang, A., Eibl, B., et al. (2003) Phenotypic and functional deficiencies of
leukaemic dendritic cells from patients with chronic myeloid leukaemia. Br. J. Haematol.
120, 6373.
97.
97 Eibl, B., Ebner, S., Duba, C., et al. (1997) Dendritic cells generated from blood precursors of
chronic myelogenous leukemia patients carry the Philadelphia translocation and can induce a
CML-specific primary cytotoxic T-cell response. Genes Chromosomes Cancer 20, 215223.
98.
98 Dietz, A. B., Litzow, M. R., Bulur, P. A., and Vuk-Pavlovic, S.. (2001) Transgenic inter-
leukin 2 secreted by CML dendritic cells stimulates autologous T(H)1 T cells. Cytotherapy
3, 97105.
99.
99 Trinchieri, G. (1989) Biology of natural killer cells. [Review] Adv. Immunol. 47, 187376.
100.
100 Hercend, T., Takvorian, T., Nowill, A., et al. (1986) Characterization of natural killer cells
with antileukemia activity following allogeneic bone marrow transplantation. Blood 67,
722728.
101.
101 Kruit, W. H., Goey, S. H., Lamers, C. H., et al. (1997) High-dose regimen of interleukin-
2 and interferon-alpha in combination with lymphokine-activated killer cells in patients
with metastatic renal cell cancer. J. Immunother. 20, 312320.
102.
102 Law, T. M., Motzer, R. J., Mazumdar, M., et al. (1995) Phase III randomized trial of
interleukin-2 with or without lymphokine-activated killer cells in the treatment of patients
with advanced renal cell carcinoma. Cancer 76, 824832.
103.
103 Okuno, K., Takagi, H., Nakamura, T., Nakamura, Y., Iwasa, Z., and Yasutomi, M. (1986)
Treatment for unresectable hepatoma via selective hepatic arterial infusion of lymphok-
ine-activated killer cells generated from autologous spleen cells. Cancer 58, 10011006.
104 Kimura, H. and Yamaguchi, Y. (1995) Adjuvant immunotherapy with interleukin 2 and
104.
lymphokine-activated killer cells after noncurative resection of primary lung cancer. Lung
Cancer 13, 3144.
105.
105 Lim, S. H., Newland, A. C., Kelsey, S., et al. (1992) Continuous intravenous infusion of
high-dose recombinant interleukin-2 for acute myeloid leukaemiaa phase II study. Can-
cer Immunol. Immunother. 34, 337342.
106.
106 Meropol, N. J., Barresi, G. M., Fehniger, T. A., Hitt, J., Franklin, M., and Caligiuri, M. A.
(1998) Evaluation of natural killer cell expansion and activation in vivo with daily subcu-
taneous low-dose interleukin-2 plus periodic intermediate-dose pulsing. Cancer Immunol.
Immunother. 46, 318326.
107 Farag, S. S., Fehniger, T. A., Ruggeri, L., Velardi, A., and Caligiuri, M. A. (2002) Natu-
107.
ral killer cell receptors: new biology and insights into the graft-versus-leukemia effect.
[Review] Blood 100, 19351947.
440 Hogan and Deeg
108.
108 Ruggeri, L., Capanni, M., Casucci, M., et al. (1999) Role of natural killer cell alloreactivity
in HLA-mismatched hematopoietic stem cell transplantation. Blood 94, 333339.
109.
109 Zeis, M., Uharek, L., Glass, B., et al. (1997) Allogeneic MHC-mismatched activated natu-
ral killer cells administered after bone marrow transplantation provide a strong graft-ver-
sus-leukaemia effect in mice. Br. J. Haematol. 96, 757761.
110 Glass, B., Uharek, L., Zeis, M., Loeffler, H., Mueller-Ruchholtz, W., and Gassmann, W.
110.
(1996) Graft-versus-leukaemia activity can be predicted by natural cytotoxicity against
leukaemia cells. Br. J. Haematol. 93, 412420.
111.
111 Ruggeri, L., Capanni, M., Martelli, M. F., and Velardi, A. (2001) Cellular therapy:
exploiting NK cell alloreactivity in transplantation. [Review] Curr. Opin. Hematol. 8,
355359.
112.
112 Jiang, Y. Z., Barrett, A. J., Goldman, J. M., and Mavroudis, D. A. (1997) Association of
natural killer cell immune recovery with a graft-versus-leukemia effect independent of
graft-versus-host disease following allogeneic bone marrow transplantation. Ann. Hematol.
74, 16.
113 Jiang, Y. Z., Cullis, J. O., Kanfer, E. J., Goldman, J. M., and Barrett, A. J. (1993) T cell
113.
and NK cell mediated graft-versus-leukaemia reactivity following donor buffy coat trans-
fusion to treat relapse after marrow transplantation for chronic myeloid leukaemia. Bone
Marrow Transplant. 11, 133138.
114.
114 Aversa, F., Tabilio, A., Velardi, A., et al. (1998) Treatment of high-risk acute leukemia
with T-cell-depleted stem cells from related donors with one fully mismatched HLA hap-
lotype. New Engl. J. Med. 339, 11861193.
115.
115 Ruggeri, L., Capanni, M., Urbani, E., et al. (2002) Effectiveness of donor natural killer
cell alloreactivity in mismatched hematopoietic transplants. Science 295, 20972100.
116.
116 Falkenburg, J. H., Marijt, W. A., Heemskerk, M. H., and Willemze, R. (2002) Minor his-
tocompatibility antigens as targets of graft-versus-leukemia reactions. Curr. Opin.
Hematol. 9, 497502.
117.
117 Riddell, S. R., Murata, M., Bryant, S., and Warren, E. H. (2002) Minor histocompatibility
antigenstargets of graft versus leukemia responses. Intl. J. Hematol. 2, 155161.
118.
118 Goulmy, E. (1997) Human minor histocompatibility antigens: new concepts for marrow
transplantation and adoptive immunotherapy. [Review] Immun. Revs. 157, 125140.
119.
119 Wang, W., Meadows, L. R., den Haan, J. M., et al. (1995) Human H-Y: a male-specific
histocompatibility antigen derived from the SMCY protein. Science 269, 15881590.
120.
120 Warren, E. H., Gavin, M. A., Simpson, E., et al. (2000) The human UTY gene encodes a
novel HLA-B8-restricted H-Y antigen. J. Immunol. 164, 28072814.
121.
121 Vogt, M. H., van den Muijsenberg, J. W., Goulmy, E., et al. (2002) The DBY gene codes for
an HLA-DQ5-restricted human male-specific minor histocompatibility antigen involved in
graft-versus-host disease. Blood 99, 30273032.
122.
122 Vogt, M. H., Goulmy, E., Kloosterboer, F. M., et al. (2000) UTY gene codes for an HLA-
B60-restricted human male-specific minor histocompatibility antigen involved in stem
cell graft rejection: characterization of the critical polymorphic amino acid residues for
T-cell recognition. Blood 96, 31263132.
123.
123 Vogt, M. H., de Paus, R. A., Voogt, P. J., Willemze, R., and Falkenburg, J. H. (2000)
DFFRY codes for a new human male-specific minor transplantation antigen involved in
bone marrow graft rejection. Blood 95, 11001105.
124.
124 Brickner, A. G., Warren, E. H., Caldwell, J. A., et al. (2001) The immunogenicity of a new
human minor histocompatibility antigen results from differential antigen processing. J.
Exp. Med. 193, 195206.
125.
125 Pierce, R. A., Field, E. D., Mutis, T., et al. (2001) The HA-2 minor histocompatibility
antigen is derived from a diallelic gene encoding a novel human class I myosin protein. J.
Immunol. 167, 32233230.
126.
126 de Bueger, M., Bakker, A., Van Rood, J. J., Van der Woude, F., and Goulmy, E. (1992)
Tissue distribution of human minor histocompatibility antigens. Ubiquitous versus re-
stricted tissue distribution indicates heterogeneity among human cytotoxic T lymphocyte-
defined non-MHC antigens. J. Immunol. 149, 17881794.
127.
127 Warren, E. H., Gavin, M., Greenberg, P. D., and Riddell, S. R. (1998) Minor histocompat-
ibility antigens as targets for T-cell therapy after bone marrow transplantation. [Review]
Curr. Opin. Hematol. 5, 429433.
128.
128 Pardoll, D. M. (2002) Spinning molecular immunology into successful immunotherapy.
[Review] Nat. Rev. Immunol. 2, 227238.
129.
129 Clark, S. S., McLaughlin, J., Crist, W. M., Champlin, R., and Witte, O. N. (1987) Unique
forms of the abl tyrosine kinase distinguish Ph1-positive CML from Ph1-positive ALL.
Science 235, 8588.
130 Kurzrock, R., Blick, M. B., Talpaz, M., et al. (1986) Rearrangement in the breakpoint
130.
cluster region and the clinical course in Philadelphia-negative chronic myelogenous leu-
kemia. Ann. Intern. Med. 105, 673679.
131.
131 Ben-Neriah, Y., Daley, G. Q., Mes-Masson, A. M., Witte, O. N., and Baltimore, D. (1986)
The chronic myelogenous leukemia-specific P210 protein is the product of the bcr/abl
hybrid gene. Science 233, 212214.
132.
132 Heisterkamp, N., Stam, K., Groffen, J., de Klein, A., and Grosveld, G. (1985) Struc-
tural organization of the bcr gene and its role in the Ph translocation. Nature 315, 758
761.
133.
133 Fossum, B., Gedde-Dahl, T., 3rd, Breivik, J., et al. (1994) p21-ras-peptide-specific T-cell
responses in a patient with colorectal cancer. CD4+ and CD8+ T cells recognize a peptide
corresponding to a common mutation (13GlyAsp). Int. J. Cancer 56, 4045.
134.
134 Yotnda, P., Garcia, F., Peuchmaur, M., et al. (1998) Cytotoxic T cell response against the
chimeric ETV6-AML1 protein in childhood acute lymphoblastic leukemia. J. Clin. Invest.
102, 455462.
135.
135 Ohminami, H., Yasukawa, M., Kaneko, S., et al. (1999) Fas-independent and nonapoptotic
cytotoxicity mediated by a human CD4(+) T-cell clone directed against an acute myelog-
enous leukemia-associated DEK-CAN fusion peptide. Blood 93, 925935.
136.
136 Boer, J., Mahmoud, H., Raimondi, S., Grosveld, G., and Krance, R. (1997) Loss of the
DEK-CAN fusion transcript in a child with t(6;9) acute myeloid leukemia following
chemotherapy and allogeneic bone marrow transplantation. Leukemia 11, 299300.
137 Fornerod, M., Boer, J., van Baal, S., Morreau, H., and Grosveld, G. (1996) Interaction of
137.
cellular proteins with the leukemia specific fusion proteins DEK-CAN and SET-CAN and
their normal counterpart, the nucleoporin CAN. Oncogene 13, 18011808.
138. von Lindern, M., Fornerod, M., Soekarman, N., et al. (1992) Translocation t(6;9) in acute
non-lymphocytic leukaemia results in the formation of a DEK-CAN fusion gene. [Re-
view] [74 refs]. Baillieres Clin. Haematol. 5, 857879.
139 de The, H., Lavau, C., Marchio, A., Chomienne, C., Degos, L., and Dejean, A. (1991) The
139.
PML-RAR alpha fusion mRNA generated by the t(15;17) translocation in acute
promyelocytic leukemia encodes a functionally altered RAR. Cell 66, 675684.
140 Kakizuka, A., Miller, W. H., Jr., Umesono, K., et al. (1991) Chromosomal translocation
140.
t(15;17) in human acute promyelocytic leukemia fuses RAR alpha with a novel putative
transcription factor, PML. Cell 66, 663674.
442 Hogan and Deeg
141.
141 Molldrem, J. J., Lee, P. P., Wang, C., et al. (2000) Evidence that specific T lymphocytes
may participate in the elimination of chronic myelogenous leukemia. Nat. Med. 6, 1018
1023.
142.
142 Molldrem, J. J., Clave, E., Jiang, Y. Z., et al. (1997) Cytotoxic T lymphocytes specific for
a nonpolymorphic proteinase 3 peptide preferentially inhibit chronic myeloid leukemia
colony-forming units. Blood 90, 25292534.
143 Molldrem, J., Dermime, S., Parker, K., et al. (1996) Targeted T-cell therapy for human
143.
leukemia: cytotoxic T lymphocytes specific for a peptide derived from proteinase 3 pref-
erentially lyse human myeloid leukemia cells. Blood 88, 24502457.
144.
144 Clave, E., Molldrem, J., Hensel, N., Raptis, A., and Barrett, A. J. (1999) Donor-recipi-
ent polymorphism of the proteinase 3 gene: a potential target for T-cell alloresponses to
myeloid leukemia. J. Immunother. 22, 16.
145.
145 Boon, T. and Old, L. J. (1997) Cancer tumor antigens. Curr. Opin. Immunol. 9, 681
683.
146.
146 Robbins, P. F. and Kawakami, Y. (1996) Human tumor antigens recognized by T cells.
[Review] Curr. Opin. Immunol. 8, 628636.
147 Ohminami, H., Yasukawa, M., and Fujita, S. (2000) HLA class I-restricted lysis of leuke-
147.
mia cells by a CD8(+) cytotoxic T-lymphocyte clone specific for WT1 peptide. Blood 95,
286293.
148.
148 Oka, Y., Elisseeva, O. A., Tsuboi, A., et al. (2000) Human cytotoxic T-lymphocyte responses
specific for peptides of the wild-type Wilms tumor gene (WT1 ) product. Immunogenetics
51, 99107.
149.
149 Reddy, P., Teshima, T., Hildebrandt, G., et al. (2002) Interleukin 18 preserves a perforin-
dependent graft-versus-leukemia effect after allogeneic bone marrow transplantation.
Blood 100, 34293431.
150.
150 Brahmi, Z., Hommel-Berrey, G., Smith, F., and Thomson, B. (2001) NK cells recover
early and mediate cytotoxicity via perforin/granzyme and Fas/FasL pathways in umbilical
cord blood recipients. Hum. Immunol. 62, 782790.
151.
151 Susskind, B., Shornick, M. D., Iannotti, M. R., et al. (1996) Cytolytic effector mechanisms
of human CD4+ cytotoxic T lymphocytes. Hum. Immunol. 45, 6475.
152.
152 Kunitomi, A., Hori, T., Maeda, M., and Uchiyama, T. (2002) OX40 signaling renders adult
T-cell leukemia cells resistant to Fas-induced apoptosis. Intl. J. Hematol. 76, 260266.
153.
153 Liu, J. H., Wei, S., Lamy, T., et al. (2002) Blockade of Fas-dependent apoptosis by soluble
Fas in LGL leukemia. Blood 100, 14491453.
154.
154 Schmaltz, C., Alpdogan, O., Horndasch, K. J., et al. (2001) Differential use of Fas ligand
and perforin cytotoxic pathways by donor T cells in graft-versus-host disease and graft-
versus-leukemia effect. Blood 97, 28862895.
155.
155 Tanaka, Y., Takahashi, T., Nieda, M., et al. (2002) Generation of Fas-independent CD4+
cytotoxic T-cell clone specific for p190 minor bcr-abl fusion peptide. Leukemia Res. 26,
317321.
156.
156 Schmaltz, C., Alpdogan, O., Muriglan, S. J., et al. (2003) Donor T cell-derived TNF is
required for graft-versus-host disease and graft-versus-tumor activity after bone marrow
transplantation. Blood 101, 24402445.
157 Schmaltz, C., Alpdogan, O., Kappel, B. J., et al. (2002) T cells require TRAIL for optimal
157.
graft-versus-tumor activity. Nat. Med. 8, 14331437.
158.
158 Middleton, P. G., Taylor, P. R., Jackson, G., Proctor, S. J., and Dickinson, A. M. (1998)
Cytokine gene polymorphisms associating with severe acute graft-versus-host disease in
HLA-identical sibling transplants. Blood 92, 39433948.
159.
159 Nordlander, A., Uzunel, M., Mattsson, J., and Remberger, M. (2002) The TNFd4 allele is
correlated to moderate-to-severe acute graft-versus-host disease after allogeneic stem cell
transplantation. Br. J. Haematol. 119, 11331136.
160.
160 Dickinson, A. M., Cavet, J., Cullup, H., Wang, X. N., Sviland, L., and Middleton, P. G.
(2001) GvHD risk assessment in hematopoietic stem cell transplantation: role of
cytokine gene polymorphisms and an in vitro human skin explant model. Hum. Immunol.
62, 12661276.
161.
161 Cavet, J., Dickinson, A. M., Norden, J., Taylor, P. R., Jackson, G. H., and Middleton, P.
G. (2001) Interferon-gamma and interleukin-6 gene polymorphisms associate with graft-
versus-host disease in HLA-matched sibling bone marrow transplantation. Blood 98,
15941600.
162.
162 Warle, M. C., Farhan, A., Metselaar, H. J., et al. (2002) Cytokine gene polymorphisms and
acute human liver graft rejection. Liver Transplant. 8, 603611.
163.
163 Mazariegos, G. V., Reyes, J., Webber, S. A., et al. (2002) Cytokine gene polymorphisms
in children successfully withdrawn from immunosuppression after liver transplantation.
164 Awad, M. R., Webber, S., Boyle, G., et al. (2001) The effect of cytokine gene polymor-
164.
phisms on pediatric heart allograft outcome. J. Heart Lung Transplant. 20, 625630.
165. Poole, K. L., Gibbs, P. J., Evans, P. R., Sadek, S. A., and Howell, W. M. (2001) Influence
of patient and donor cytokine genotypes on renal allograft rejection: evidence from a single
centre study. Transplant Immunol. 8, 259265.
Influence of Radiation Protocols on GVHD 445
29
Influence of Radiation Protocols on Graft-vs-Host
Disease Incidence After Bone-Marrow Transplantation
in Experimental Models
Sebastian Schwarte and Matthias W. Hoffmann
Summary
Bone-marrow transplantation is an approved curative treatment for many hemato- and
oncologic diseases. Nevertheless, the severe acute clinical course of graft-vs-host disease
(GVHD) after allogeneic bone-marrow transplantation is frequently fatal, and is to date not
curable. Acute GVHD must, therefore, be prevented from the start of the bone-marrow trans-
plantation by immunosuppressive medication, causing sometimes serious side effects. There-
fore, new preventive strategies are tested, starting with animal experiments. Often mice are
chosen for this kind of trial, and the clinical protocol of bone-marrow transplantation is trans-
ferred into the experimental settings.
The first step to induce an acute GVHD is whole-body irradiation of the recipients. Several
methods are available for this purpose: the most common is a 60cobalt source (-irradiation);
less common are a 137cesium source (-irradiation) and a linear (particle) accelerator (photons).
Differences between these radiation techniques can occur and can unexpectedly interfere with
the results of the experiments. In this chapter, the materials and methods for bone-marrow
transplantation in mice, with particular emphasis on the different radiation techniques, are
explained; furthermore, the advantages and disadvantages in regard to the underlying physical
principles will be discussed.
Key Words: Radiation; whole body irradiation; bone-marrow transplantation; cobalt;
cesium; linear accelerator; particle accelerator; incidence; GVHD; GVHR; mice.
1. Introduction
Bone-marrow transplantation is an approved curative treatment for many hemato-
oncologic diseases, such as immunodeficiency (SCID, Wiskott-Aldrich syndrome),
acute leukemia, aplastic anemia, chronic leukemia, hemoglobinopathies (thalassemia
major), myelodysplasia, enzymatic defects (Gauchers disease), lymphoma, myeloma,
and after whole-body irradiation of carcinomas with myelosuppressive side effects.
Allogeneic bone-marrow transplantation is the treatment of choice for patients with
lymphomas or leukemias. Despite significant improvements since the beginning
445
446 Schwarte and Hoffmann
of the modern era of allogeneic bone-marrow transplantation, the side effects of this
therapy can be tremendous and life-threatening. In particular, the disease caused by
the immunological reaction of immunocompetent donor cells against the immuno-
compromised host (graft-vs-host disease [GVHD]) is still a major problem. An acute
and a chronic form of GVHD can be differentiated. The acute form of GVHD can be
very dangerous and life-threatening and manifests mainly at the skin, the liver, and
the gastrointestinal tract. The chronic form resembles an autoimmune disease, and
severe complications are mainly caused by the immunosuppressive therapy.
The treatment of a patient with a severe acute clinical course of GVHD is fre-
quently not successful. Acute GVHD must, therefore, be prevented by potent immu-
nosuppressive drugs, with their inherent range of risks and side effects. Therefore, a
main goal in bone-marrow transplantation is to specifically suppress the GVH reac-
tion without impairing the physiological immune response to pathogens and residual
leukemic cells. For this reason, it is of major importance to develop more specific
therapeutic strategies in bone-marrow transplantation.
Mice are frequently used as experimental models to test new therapeutic approaches
in the treatment of GVHD after allogeneic bone-marrow transplantation. They have
the advantage that several genetically defined inbred, transgenic, and knock-out mouse
strains are available. These can be used to examine GVHD pathophysiology and
GVHD prevention in several informative combinations. In the following, a mouse
model is presented that allows the study of GVHD under defined conditions.
The first step to induce acute GVHD is whole-body irradiation of the host. In the
clinical setting, this serves three purposes: first, lethal irradiation destroys malignant
cells underlying the leukemic disease. Second, irradiation destroys host lymphocytes
and in turn prevents or ameliorates the rejection of the allogeneic bone-marrow
inoculum. Third, irradiation is thought to create space in the bone marrow by destroying
host bone-marrow cells, thereby improving the engraftment of the injected bone-mar-
row cells. Several methods are available to irradiate animals in an experimental setting:
the most commonly used radiation source is 60cobalt (-irradiation); less frequently used
are 137cesium (-irradiation) and the linear (particle) accelerator (photon irradiation).
After irradiation, bone-marrow cells are purified from the donor. In the clinical
setting, multiple bone-marrow aspirates are harvested from the donor, whereas in ani-
mal models bone-marrow cells are obtained from the donor tibia and femur.
Finally, bone-marrow cells are injected intravenously. The outcome is followed by
regularly checking the weight and inspection for signs of GVHD.
2. Materials
1. Phosphate buffered saline (PBS): 28.0 mM NaH2PO4, 7.2 mM Na2HPO4, and 0.14 M
NaCl; 9.55 g/L bidistilled water; without calcium and magnesium; may be stored indefi-
nitely without contamination at 4C.
2. Anesthesia with ketamine (ketamine hydrochloride) and Rompun (xylazine hydrochlo-
ride): mixture of 0.5 mL ketamine hydrochloride and 0.5 mL xylazine hydrochloride,
dissolved in 9 mL sterile PBS; light-sensitive; stable at room temperature for up to 1 mo.
Ketamine (10%): 0.08 mL/100 g i.p. (80 mg/kg); Rompun (2%): 0.01 mL/100 g i.p.
(2 mg/kg). Two-hundred fifty microliters per mouse.
3. Irradiation equipment: a 60cobalt source, a linear accelerator, or a 137cesium source.

4. Mouse cages made of plexiglas for irradiation.
5. Medium 199 (TC 199): Hanks salt solution with L-glutamine without NaHCO3: 10.61
g/L bidist. water; stable at 4C for 2 mo if kept sterile.
6. One liter hemolysis buffer: pH 7.3, 8.34 g NH4Cl, 0.037 g EDTA, and 1.00 g NaHCO3;
stable at 4C for up to 2 mo if kept sterile.
7. One hundred milliliters trypan blue: 8 mL of a 2% stock solution, dissolved in a 92-mL
NaCl solution; store at 4C indefinitely.
8. Surgical instruments (forceps, scissors).
9. Fifteen-milliliter plastic tubes.
10. Polystyrene cell-culture dishes (Petri dishes).
11. Sieve with 0.6-mm mesh size.
12. Sieve with 0.2-m mesh size.
13. Fifty-milliliter polypropylene tube.
14. Counting chamber (Neubauers hematocytometer).
15. BL/LB, 0.40 12 mm, 27 gage 0.5 in. disposable injection needles.
16. BL/LB, 0.90 40 mm, 20 gage 1.5 in. disposable injection needles.
17. Five-milliliter, 10-mL, and 20-mL disposable syringes.
3. Methods
Mice should be kept under clean, preferably specific pathogen free conditions.
Following lethal irradiation and bone-marrow reconstitution, the mice are temporarily
immunodeficient until the bone marrow has produced sufficient immunocompetent
cells to protect the mice from infections.
The anesthetized recipients are lethally irradiated 24 h before the injection of donor
cells. Donor animals are sacrificed about 6 h before the planned injection to allow for
the collection of bone marrow and lymph-node cells (see Note 1), and to have suffi-
cient time for cell processing.
Donors and recipients should be age and sex matched in order to avoid undesired
immune responses to minor (H-Y) antigens that might affect the GVH reaction under
study.
3.1. Irradiation of the Recipients
To prevent rejection of the allogeneic bone-marrow cells, the recipients are lethally
irradiated by a single dose of whole-body irradiation. The recipient animals are con-
tained in special cages during the irradiation.
3.1.1. Anesthesia
To achieve a homogenous radiation effect, the animals are kept still by general
anesthesia. Therefore, the animals are anesthetized with 250 L of the ketamine-
Rompun mixture, applied intraperitoneally with a 27-gage needle. Anesthesia lasts for
about 20 min and is sufficient for the duration of the irradiation procedure.
3.1.2. Cage Design
The size of the radiation cage must be large enough to accommodate the mice in
one plane, side by side, to achieve a homogenous irradiation effect. The cage should
contain air holes to provide sufficient ventilation. Usually, a cage contains a maxi-
mum of eight mice. A suitable radiation cage can measure 16 cm in length, 10 cm in

width, and 2.5 cm in height. The lid of the cage should close directly above the ani-
mals with as little space between them and the lid as possible (see Note 2). A mouse
cage for irradiation can be made of safety glass, plexiglas, or cardboard carton, and
should be secured to prevent escape of the animals.
3.1.3. Irradiation Equipment
See Note 3. Irradiation with a 60cobalt source is carried out at a distance of 75 cm
between the device and the animals for a radiation dose of 9 Gy. A 0.5-cm thick plexiglas
plate should be placed on top of the cage to achieve optimal tissue distribution.
3.2. Collection of Donor Cells
For the collection of bone-marrow and lymph-node cells the animals are sacrificed
by cervical dislocation, following anesthesia with carbon dioxide. The level of anesthe-
sia is checked by controlling for spontaneous movements, posture of the head, reflexes,
spontaneous breathing, and heartbeat of the mouse.
3.2.1. Preparation of Lymph-Node Cells
1. Donor animals are fixed on their back and disinfected with 70% EtOH.
2. Then a transverse skin incision is made, starting 1 cm below the xiphoid process.
3. The skin is manually pulled upwards and downwards, exposing the abdomen and thorax
up to the neck and front legs.
4. The submandibular, axillar, and brachial lymph nodes are removed with a curved ana-
tomic forceps, taking care not to include surrounding fat.
5. The lower part of the skin is drawn caudally towards the hind legs, and the inguinal
lymph nodes are removed.
6. Then a midline incision is made through the abdominal wall to expose the intestines. Lymph
nodes are collected from the intestinal mesentery, taking particular care to exclude fat.
7. Lymph nodes are collected in sterile, ice-cold TC 199 medium in 15-mL tubes that are
usually kept on ice in a styrofoam container.
8. Under sterile conditions, the lymph nodes are transferred to a sterile sieve (mesh size: 0.6
mm) and ground with a spade. Subsequently, the sieve is rinsed with 10 mL TC 199. The
cell suspension is collected in a cell-culture dish (Petri dish) and is flushed through another
sieve (mesh size: 0.2 m) to exclude cell debris.
9. The resulting single-cell suspension is collected in a 50-mL plastic tube and is centri-
fuged (10 min, 300g, 4C). Cells are taken up in 10 mL sterile PBS, and viable cells are
counted after trypan blue staining in a microhemocytometer.
10. The optimal number of cells for induction of GVHD depends on the chosen model (see
Note 4).
3.2.2. Preparation of Bone-Marrow Cells

After collecting the lymph-node cells, the bone marrow is harvested.
1. Bone-marrow cells are harvested from the tibia and femur of donor mice: after stripping
the skin from both hind legs, the femoral bones are disarticulated at the hip joint. Care is
taken not to damage or splinter the bones, as this makes later flushing inefficient. Subse-
quently, muscle tissue is carefully dissected away from the bones. The bones are collected
in 20 mL sterile TC 199 at 4C in 50-mL plastic tubes.
Fig. 1. Harvesting of bone-marrow cells.
2. Under sterile conditions, three cell-culture dishes are prepared, each containing 10 mL
sterile TC 199. Scissors and forceps should be disinfected with 70% EtOH regularly dur-
ing this process. First, the bones are collected in the first dish. Femur and tibia are disar-
ticulated at the knee joint (see Fig. 1) and the separated bones are collected in the second
dish. There, the distal ends of both bones are cut off with scissors to access the bone-
marrow cavity. The bone marrow is harvested by flushing the bone cavities with a 27-
gage needle on a 1-mL syringe containing TC199 into the third dish.
3. Bone-marrow cells are then repeatedly resuspended with a 20-gage needle on a 20-mL
syringe (TC 199) until the cell suspension is homogenized. Then cells are transferred to a
50-mL plastic tube and centrifuged (10 min, 4C, 300g).
4. After centrifugation, the bone-marrow cell suspension is incubated with sterile hemolysis
buffer at room temperature for 5 min. Then it is filtered through a sieve (mesh size: 0.2
m) to remove bone fragments and other particulate material. The cell suspension is col-
lected in a 50-mL plastic tube and is washed twice in sterile PBS. Thereafter, the cells are
taken up with 10 mL sterile PBS to count them with trypan blue (vitality control).
5. Similar to the clinical situation, in which 2 108 bone-marrow cells per kg body weight
are administered to the patient, mice should receive on average 5 106 cells (see Note 5).
3.3. Injection of Donor Cells

To inject the cells intravenously, the four tail veins of the mouse can be expanded
and visualized by patting them with a swab and warm water. Water should not be too
hot or boil. It is recommended to inject the tail vein with no more than 200 L of
volume. Only the lateral tail veins are suitable for injections.
For this procedure the mouse must be restrained in a holding device. At best, the
mouse is placed in a body-adjusted plastic tube or cylinder to restrain the mouse inside
the tube, and to allow the tail to hang out. Care should be taken not to exert unneces-
sary force against the animals.
The injection is carried out with a 27-gage injection needle on a 1-mL disposable
syringe.
3.4. Diagnosing GVHD

After bone-marrow reconstitution, the recipients are checked every other d for at
least 100 d (see Note 6) for weight changes, and are checked daily for survival. In
addition, the skin is inspected and the defense reaction is tested. If skin lesions such as
alopecia, desquamations, or ulcerous-erosive changes appear, the animals should be
sacrificed. Similarly, animals should be sacrificed upon severe weight loss or apa-
thetic behavior. The liver, the gastrointestinal tract, and the skin should be examined
by histology. It is recommended to take photos for documentation of skin lesions and
a hunched back. See Note 7 for the clinical course of chronic GVHD.
4. Notes
1. In contrast to human beings, in whom acute GVHD is caused by allogeneic bone-marrow
cells alone, mature T-cells must be added to the bone-marrow inoculum in most experi-
mental mouse models in order to induce acute GVHD. It is reported that with total body
irradiation of 8.5 Gy (60cobalt source) and 3 107 bone-marrow cells in a major histo-
compatibility complex (MHC)-I disparate mouse model, about 40% of recipients sur-
vived the first 3 mo after bone-marrow transplantation (1). Many studies confirm that the
murine bone marrow contains only a low frequency of mature T-cells (13% of all bone-
marrow cells) (25). In marked contrast, the proportion of immunocompetent T-cells in
humans amounts to 8% 1.6% of the bone-marrow pool (6), and is therefore consider-
ably higher than in the mouse. Another factor necessitating the addition of mature
T-lymphocytes from secondary lymphoid organs (spleen, lymph nodes) is the high pro-
portion of immunoregulatory cells in murine bone marrow. Thirty to 50% of bone-mar-
row -T-cell antigen receptor (TCR) T-cells are NK1.1+ cells (7); about 30% of CD3+
cells are CD4-CD8- double negative T-cells (8). So far, experiments in autoimmune dis-
ease and GVHD models have proven the immunoregulatory function of both of these cell
types (914). The mechanism of immunoregulation has been attributed to the secretion of
interleukin (IL)-4 by these cells (5). Therefore, NK1.1CD4-CD8- T-cells have been termed
natural suppressor cells (15). Both the low frequency of mature immunocompetent
T-cells and the high percentage of immunoregulatory T-cells could explain the unsuc-
cessful induction of GVHD in mice reconstituted with bone-marrow cells only. This is
why, even with higher cell numbers, mature T-cells cannot induce a GVHD reaction as
expectedthey are still suppressed by the concomitant increase in the number of trans-
ferred suppressor cells. In summary, in contrast to the human system, bone-marrow
cells alone are unable to reliably induce a GVH reaction in the mouse. It is necessary to
add mature T-cells from secondary lymphoid organs, such as lymph nodes or spleen, to
increase the ratio of immunocompetent vs immunoregulatory T-cells in the bone-marrow
inoculum (see Note 4 for further details).
2. Should an alternative design of an irradiation cage be used, it is important to remember
that the effect of irradiation is more diminished, the greater the distance between the
animals and the lid of the cage. The dose needed to achieve the same irradiation effect
increases with the second power of the distance between the irradiation source and the
target. Ideally, the lid should close directly on top of the back of the animals in order to
achieve perfect irradiation distribution. To achieve optimal radiation results, it is neces-
sary to optimize the radiation conditions. First, it is important to construct an optimal
cage. If the lid does not close directly above the back of the recipients, a higher radiation
dose might be necessary. In our own experience, the mortality of recipients of allogeneic
Fig. 2. Lethal graft-vs-host disease after irradiation by a particle accelerator (5 MeV) in

cages of different height (2.5 cm, 5 cm). Kb mice were reconstituted with 5 106 Des-T-cell
antigen receptor (TCR) (anti-Kb TCR) bone-marrow cells. Three experimental groups were
analyzed: (1) 12 Gy irradiation, 5 cm height, 1 106 added lymph-node cells (LNC); (2) 11 Gy,
2.5 cm, 1 106 added LNC; (3) 11 Gy, 2.5 cm, 5 106 added LNC.
bone marrow was higher in a group that was irradiated with a smaller dose (11 Gy) of a
particle accelerator in a smaller box (2.5 cm high; n = 5; survival rate: 0% at 55 d after
bone-marrow transplantation) than with a higher dose (12 Gy) of the same particle accel-
erator in a higher box (5 cm high; n = 3; survival rate: 33% after 89 d post bone marrow
transplantation) (unpublished, Fig. 2). This can be explained by the build-up effect of
radiation beams. With too much space between lid and the irradiated objects, the build-up
dose will be lost and has to be rebuilt in the object itself. In this scenario, the dose
maximum is not optimally focused, which can lead to misleading results. In summary, it
is recommended that the irradiation cage should have a maximum height of 2.5 cm.
As mentioned above, another modification to focus the build-up dose and the depth
dose, and to optimize the radiation effect, is to place a plate on the cage. Usually, a
plexiglas plate is put on top of the cage to distribute the radiation homogenously and to
focus the maximum of the radiation optimally on the body of the recipient. The optimum
thickness of such a plate can vary from 1 to 4 cm, depending on the radiation device and
the employed energy of the radiation source. Alternatively, the use of moulages has been
recommended in the literature (16,17) for oncologic radiation patients. Moulages are wax
molds (lexan) with tissue-like characteristics adapted to the body forms of the patients. In
the experimental situation, the moulages would have to be placed on top of the cage.
3. Different radiation devices are available for GVHD induction in experimental models.
The three most commonly used are 60cobalt, 137cesium, and particle (linear) accelerators.
The biological effects, for instance with respect to GVHD induction or side effects, differ
considerably depending on the radiation source used. Characteristics of these radiation
sources are summarized in Table 1.
452
Table 1
Characteristics of Different Radiation Sources
60Cobalt Linear accelerator 137Cesium
Naturally found radioactive source Electrons are accelerated by electromagnetic high frequency waves Naturally found radioactive source
Radioactive decay Photon radiation Radioactive decay
- and - radiationa Used energies (monoenergetic): between 6 MV and 25 MV - and - radiationa
Energy: 1.25 MVb Maximum doses with: Energy: 662 keVb (monoenergetic)
Maximum dose at 0.5 cm depth 23 MV at 3.0 cm depth; 6 MV at 1.5 cm depth Maximum dose at <0.5 cm depth
Formation of penumbrac No formation of penumbrac; extensive continuous spectrum Formation of penumbrac
Half-life: about 5.25 yr No half-life Half-life: about 30 ys
a -radiationis absorbed by the technical device; only the -radiation will have an effect.
b The 60cobalt source emits two equally powerful -lines of 1.17 and 1.33 MeV. For practical reasons they are calculated as one monochromatic line of
1.25 MeV. In contrast, the 137cesium source emits only one monoenergetic line.
c Within the radiation field, the dose is roughly constant, but decreases towards the margins of the radiation field. In the area within these margins, the
dose reduction is geometrically very steep; outside of these margins and the radiation field it converts into a flat runner. The dose value in the main axis of
452
the radiation field is defined as 100%. The distance between two distinct dose values in the area of the radiation field margins (dose reductions) is called
penumbra. The term penumbra therefore deals with the flat dose reduction at the margins of the radiation field. Often the distance between the 80% and
the 20% value will be given (again, 100% is according to the main field axis). The formation of the penumbra is influenced by the size of the radiation
source and its focus, and by the construction of the collimator system. Particle accelerators have a penumbra region of only a few millimeters, which are
therefore smaller than those of -irradiation devices (17).
Table 2
Schwarte and Hoffmann

Results of Different Irradiation Modalities With the Same Radiation Dose (9 Gy) and Injected Lymph-Node Cell Number a
Number of Dose Height of Cell number of injected Death rate Survival
Radiation device treated mice (Gy) cage (cm) lymph-node cells (Days after BMTx) rate (%)
Cobalt 6 9 5 5 105 7, 11 (2x), 14 (2x), 23, 27 0
Particle accelerator 12 9 5 5 105 7, 8, 12, 14, 42, 46, 73 42
Cesium 5 9 b 5 105 20, 82, 113 40
a For
legend, see Table 1.
b For
technical reasons, the mice were placed inside a special metal cylinder (18 cm high, base area 58 cm2). The metal cylinder with the mice inside
rotated around its vertical axis close to the cesium source for equivalent dose distribution.
Experiments were performed to compare the different radiation sources after alloge-
neic bone-marrow transplantation (Schwarte and Hoffmann, unpublished). Using 9 Gy
irradiation with 60cobalt and 5 105 lymph-node cells plus 5 106 bone-marrow cells, all
recipients died from acute GVHD within 30 d. With the same treatment but with the other
radiation devices, recipient mice could live longer than 120 d (Table 2).
It can be observed that with these other two radiation techniques, the mice developed
skin lesions, such as alopecia or ulcerative-erosive skin desquamations, especially after
radiation with the 137cesium source. Acute GVHD could not be induced by these two radi-
ation modalities. In contrast, it was confirmed that the dose of 9 Gy with the latter two
radiation sources was lethal in the absence of bone-marrow reconstitution. All irradiated
mice died within 14 d, irrespective of the presence or absence of an acute GVHD reaction.
The high lethality in the radiation control groups (without treatment) showed that the
radiation was sufficient to induce a lethal tissue injury, particularly of the bone marrow,
but was not able to set the recipient into a GVHD-inducable state.
The different reactions to the radiation techniques and their effects on GVHD results
can be explained by the underlying physical principles and some immunological concepts.
The main reason for these findings and different GVHD incidence, dependent on the
different radiation devices, is the principle of the build-up dose and the depth dose maxi-
mum.
One particularly important quality of these radiation devices is the dose build-up effect
with increasing tissue depth. After hitting the body surface, the radiation dose changes
with increasing depth along the central axis of the beam, which is caused by various physi-
cal effects (18). The graphic description of these dose distributions in the radiated object is
termed isodose curves.
The explanation for the phenomenon of the build-up dose is as follows: at first, as the
radiation decreases, the dose is reduced. The relative depth dose of the 60cobalt source at
the surface amounts to about 90%, of the particle accelerator with 5 MeV to 70%, and
with 16 MeV to 40% (17).
At the surface of the object, only electrons from the radiation source, the surrounding
area (mainly air), and backscattered electrons from the object have an effect; in the irradi-
ated object itself, significantly more electrons are removed and transmitted, proportional
to the total amount of available photons (so-called photonfluence).
As a consequence of the limited range of liberated electrons, the total sum of elec-
trons increases with increasing depth until they have lost their motion energy. At the end
of their course and, therefore, at the end of their defined depth range, a maximum of
liberated electrons can be observed. At the same time, the effect of each liberated, sec-
ondary electron adds up to the primary photon beam, which is absorbed by the object and
is therefore further diminished. On the whole, an increase of dose is achieved up to a
defined depth in the tissue, though the primary beam is decreased.
A depth dose maximum of 100% is achieved with the 60cobalt source at 0.5 cm, with
the particle accelerator at 1.5 cm (6 MeV) to 3.0 cm (23 MeV), and with the 137cesium
source at less than 0.5 cm. Beyond this depth, the dose decreases, mainly due to the effect
of photoabsorption (17,19).
With the 60cobalt equipment, an acute GVHD can be induced in murine recipients.
The depth dose maximum is at 0.5 cm. Beyond this point, the dose decreases steeply.
Nonetheless, this is sufficient to induce an acute GVHD. The only technical disadvantage
is that due to its short half-life (approx 5 yr) a 60cobalt source needs to be checked and
replaced at regular intervals.
The 137cesium source has a lower energy (662 KeV) than 60cobalt (1.25 MeV) and the
linear accelerators. Therefore, the dose depth maximum is <0.5 cm, which is quite close
the surface, even if the dose is built up appropriately under optimized conditions. Below
this point, the dose decreases even more steeply than with a 60cobalt source. This is why
the irradiation is most effective at the surface, but significantly decreases in effectiveness
with increasing tissue depth. Due to these considerations, 137cesium is no longer used as a
telegamma radiation device in the clinical routine, but only for brachytherapy of tumors
(20). Furthermore, the surface effects can explain why in experimental models so many
recipient mice develop skin lesions compared to other radiation devices (60cobalt, linear
accelerator): the maximum radiation effect at the surface leads to a heavy injury of the
skin, resulting in a high incidence of skin lesions, but without the effects in the depth. In
turn, the recipient organism is not sufficiently immunosuppressed to allow for the induc-
tion of acute GVHD; only a mild, chronic GVHD can be induced. Other studies prove the
interdependences between radiation source and the induction of GVHD (21).
It has previously been observed that organisms vary in their biological response to the
different radiation techniques. With 137cesium, a higher secretion of IL-1 and IL-6 could
be detected, whereas with a conventional X-ray device, a higher secretion of interferon
(IFN)- and tumor necrosis factor (TNF)- was identified (21).
In contrast, linear accelerators use more energy than the 60cobalt source. The maxi-
mum depth dose is at 1.5 cm (6 MeV) or 3.0 cm (23 MeV). Considering the small size of
an animal such as a mouse for irradiation, problems with the use of higher energies (e.g.,
23 MeV) may occur. The dose lethally damages the organism. However, irrespective of
the death of the irradiated mouse, the suboptimal distribution and depth dose, the same
effect as with the 137cesium source will be obtained. If lower energies are used, irradia-
tion will have to be optimized, i.e., the cage design and the plate for the build-up dosing
must be adjusted (2.0 cm for a 6 MeV linear accelerator and 4.5 cm for a 25 MeV linear
accelerator).
It was determined that the mouse strains respond differently to irradiation (22).
BALB/c, C3H/OuJ, C3H/HeJ, and C57BL/6 mice were exposed to 60cobalt and 137cesium
sources. The lethality rate decreased in the following sequence: BALB/c > C3H/OuJ =
C3H/HeJ > C57BL/6. The C57BL/6 mouse strain was, therefore, the most radiation resis-
tant. The reason for this difference is still unclear. It is assumed that probably DNA-
repair systems in some strains do not work as efficiently as in other strains, or that the
pool of radiosensitive stem cells differs between the mouse strains.
4. Only the addition of mature T-cells from secondary lymphatic organs, such as lymph
nodes or spleen, to the bone-marrow inoculum can induce acute GVHD in the mouse (see
Note 2 for further details).With that in mind, the number of added cells determines the
form and clinical course of the GVH-reaction (23). In a murine GVHD-model with CBA-
mice as donors and BALB/c mice as recipients, which had been irradiated 5 to 6 h before
bone-marrow transplantation with 750 cGy (-irradiation), the recipients were given 5
106 CBA bone-marrow cells with varying numbers of spleen cells. Less than 1 105
spleen cells did not lead to any GVH reaction at all, or at least only to a mild course of
GVHD; 2.5 to 4 105 spleen cells led to a moderate to severe GVHD; 1 106 or more
spleen cells were associated with a hyperacute form of GVHD in which most of the
BALB/c recipients died within 8 d after bone-marrow transplantation (4). The various
mouse strains differ in their extent of immune responses. That is why a defined number of
cells to induce an acute GVHD cannot be given in general. Nevertheless, in all mouse
strains, mature T-cells must be added to induce lethal acute GVHD.
5. In the literature, varying numbers of transferred cells can be found. Therefore, many
alternatives to this mentioned cell count are possible. The amount of 5 106 bone-mar-
row cells is related to the clinical setting, in which the patient is given 2 108 bone-
marrow cells per kg body weight, calculated with an average mouse weight of 25 g.
6. Alternatives for monitoring for a GVHD course: every d for mortality and once a wk for
weight changes. It is dependent on the question of interest and of data survey for statisti-
cal analysis. The definition of the acute form of GVHD is that it appears during the first
100 d after bone-marrow transplantation; the chronic form of GVHD by definition starts
after 100 d. That is why bone-marrow transplant recipients should be controlled for at
least 100 d if the acute form is the objective for experiments.
7. Mice developing a chronic GVHD typically show the following disease course: In the
first wk after cell transfer, the mice get visibly and evidently ill: they lose weight and
show bristled fur. Then the status of a few mice can worsen, resulting in their death
within the next 2 wk. In most cases, however, the recipients recover for a certain period
of time. These mice exhibit only marginal body weight loss, the skin lesions improve at
the end of the second wk after bone-marrow transplantation, and the mice show regular,
almost normal physical activity. After 4 wk, the mice develop a severe and progressive
body weight loss with diarrhea and a hunched back. Most of them die within the next 1 to
2 wk; the body weight loss can then be severe, sometimes up to more than 40% (24). The
described course of chronic GVHD in mice, also in the study cited above (24), is fre-
quently observed. In that study, the chronic GVHD was caused by injection of purified
CD8+ T-cells, without CD4+ T-cells. That study is seen as evidence that CD8+ T-cells
more likely induce a chronic GVHD, rather than an acute form. Even in standard text-
books this study is cited for this interrelation (25). It is striking, though, that in publica-
tions with observations of a chronic GVHD by CD8+ T-cells, the recipient mice were
lethally irradiated by a 137cesium source (2329). Therefore, the general statement that
CD8+ T-cells lead to chronic GVHD cannot be maintained. The reaction of the recipient
is also influenced by the radiation source during the conditioning for the bone-marrow
transplantation. Since the immunological reactions following irradiation and bone-mar-
row transplantation do not necessarily lead to an acute GVH reaction, but rather to a mild
and chronic GVHD, the donor cells somehow react in another way than expected. In
another study, three possible concepts were proposed as explanations for this observa-
tion (30): the activation-induced cell death, tolerance by internalization of the T-cell
receptor, and anergy of the donor T-cells. The immune response of the recipient and the
induction of the GVHD is dependent on various factors, e.g., the combination of the
mouse strains and the different radiation modalities. In summary, based on the physical-
technical characteristics, it is easiest and most reliable to work with a 60cobalt source for
the induction of an acute GVHD, at least in the experimental setting with mice. In con-
trast, linear accelerators and 137cesium sources are rather unreliable for this purpose.
Acknowledgment
The authors would like to thank Joachim Lotz for critical review of the manuscript.
References
1.
1 Waer, M., Ang, K. K., Schueren, E. van der, and Vandeputte, M. (1984) Increased inci-
dence of murine graft-versus-host disease after allogeneic bone marrow transplantation by
previous infusion of syngeneic bone marrow cells. Transplantation 38, 396400.
2.
2 Eberl, G., Lees. R, Smiley, S. T., Taniguchi, M., Grusby, M. J., and MacDonald, H. R.
(1999) Tissue-specific segregation of CD1d-dependent and CD1d-independent NK T cells.
J. Immunol. 162, 64106419.
3 Zeng, D.,Gazit, G., Dejbakhsh-Jones, S., et al. (1999) Heterogeneity of NK1.1- T cells in
3.
the bone marrow: divergence from the thymus. J. Immunol. 163, 53385345.
4 Rosario, M. L. U. del, Zucali, J. R., and Kao, K. J. (1999) Prevention of graft-versus-
4.
host disease by induction of immune tolerance with ultraviolet b-irradiated leukocytes
in H-2 disparate bone marrow donor. Blood 93, 35583564.
5 Zeng, D., Lewis, D., Dejbakhsh-Jones, S., et al. (1999) Bone marrow NK1.1 and
5.
NK1.1+ T cells reciprocally regulate acute graft versus host disease. J. Exp. Med. 189,
10731081.
6.
6 Fauci, A. S. (1975) Human bone marrow lymphocytes: I. Distribution of lymphocyte sub-
populations in the bone marrow of normal individuals. J. Clin. Invest. 56, 98110.
7.
7 Zeng, D., Hoffmann, P., Lan, F., Huie, P., Higgins, J., and Strober, S. (2002) Unique
patterns of surface receptors, cytokine secretion, and immune functions distinguish T cells
in the bone marrow from those in the periphery: impact on allogeneic bone marrow trans-
8.
8 Sykes, M. (1990) Unusual T cell populations in adult murine bone marrow. Prevalence of
CD3+CD4CD8 and TCR+NK1.1+ cells. J. Immunol. 145, 32093215.
9.
9 Gombert, J. M., Herbelin, A., Tancrede-Bohin, E., Dy, M., Carnaud, C., and Bach, J. F.
(1996) Early quantitative and functional deficiency of NK1+-like thymocytes in the NOD
mouse. Eur. J. Immunol. 26, 29892998.
10.
10 Baxter, A. G., Kinder, S. J., Hammond, K. J., Scollay, R., and Godfrey, D. I. (1997) Asso-
ciation between TCR+CD4CD8 T-cell deficiency and IDDM in NOD/Lt mice. Dia-
betes 46, 572582.
11 Hammond, K. J. L., Poulton, L. D., Palmisano, L. J., Silveira, P. A., Godfrey, D. I., and
11.
Baxter, A. G. (1998) /-T cell receptor (TCR)+CD4-CD8 (NKT) thymocytes prevent
insulin-dependent diabetes mellitus in nonobese diabetic (NOD)/Lt mice by the influence
of interleukin (IL) 4 and/or IL-10. J. Exp. Med. 187, 10471056.
12.
12 Palathumpat, V., Dejbakhsh-Jones, S., Holm, B., Wang, H., Liang, O., and Strober, S.
(1992) Studies of CD4-CD8- bone marrow T cells with suppressor activity. J. Immunol.
148, 373380.
13.
13 Sykes, M., Hoyles, K. A., Romick, M. L., and Sachs, D. H. (1990) In vitro and in vivo
analysis of bone marrow-derived CD3+, CD4, CD8, NK1.1+ cell lines. Cell. Immunol.
129, 478493.
14 Strober, S., Cheng, L., Zeng, D., et al. (1996) Double negative (CD4CD8+) T
14.
cells which promote tolerance induction and regulate autoimmunity. Immunol. Rev.
149, 217230.
15 Strober, S. (2000) Natural killer 1.1+ T cells and natural suppressor T cells in the bone
15.
marrow. J. Allergy. Clin. Immunol. 106, S113114.
16. Quast, U. and Hoederath, A. (1996) Ganzkrperbestrahlung. In: Strahlentherapie:
Radiologische Onkologie. Scherer, E. and Sack, H. eds., Springer, New York, NY, pp.
207218.
17. Richter, J. and Schwab, F. (1998) Charakterisierung und Eigenschaften von
Dosisverteilungen, in Strahlenphysik fr die Radioonkologie. Richter, J. and Flentje, M.,
eds., Thieme Stuttgart, New York, NY, pp. 4359.
18. Verhey, L. J. (1998) Principles of radiation physics. In: Textbook of Radiation Oncology.
Leibel, S. A. and Phillips, T. L. eds. WB Saunders, Philadelphia, PA, pp. 91114.
19. Laubenberger, T. and Laubenberger, J. (eds.) (1994) Technik der medizinischen

Radiologie: Diagnostik, Strahlentherapie, Strahlenschutz fr rzte, Medizinstudenten und
MTRA; mit Anleitung zur Strahlenschutzbelehrung in der Rntgendiagnostik. Deutscher
rzte-Verlag, Kln, Germany.
20. Schmidt, R. (1998) Gerte zur Erzeugung ionisierender Strahlung. In: Strahlenphysik fr
die Radioonkologie. Richter, J. and Flentje, M. eds. Thieme Stuttgart, New York, NY, pp.
2738.
21.
21 Sykes, M., Pearson, D. A., Taylor, P. A., Szot, G. L., Goldman, S. J., and Blazar, B. R.
(1999) Dose and timing of interleukin (IL)-12 and timing and type of total-body irradia-
tion: effects on graft-vs.-host disease inhibition and toxicity of exogenous IL-12 in murine
bone marrow transplant recipients. Biol. Blood Marrow Transplant. 5, 277284.
22.
22 Henning, U. G. G., Wang, Q., Gee, N. H., and von Borstel, R. C. (1996) Protection and
repair of -radiation-induced lesions in mice with DNA or deoxyribonucleoside treatments.
Mutat. Res. 350, 247254.
23.
23 Baker, M. B., Altman, N. H., Podack, E. R., and Levy, R. B. (1996) The role of cell-
mediated cytotoxicity in acute GVHD after MHC-matched allogeneic bone marrow trans-
plantation in mice. J. Exp. Med. 184, 26452656.
24.
24 Sprent, J., Schaefer, M., Gao, E. K., and Korngold, R. (1988) Role of T cell subsets in
lethal graft-versus-host disease (GVHD) directed to class I versus class II H-2 differences.
I. L3T4+ cells can either augment or retard GVHD elicited by Lyt-2+ cells in class I-
different hosts. J. Exp. Med. 167, 556569.
25.
25 Sprent, J., Schaefer, M., and Korngold, R. (1990) Role of T cell subsets in lethal graft-
versus-host disease (GVHD) directed to class I versus class II H-2 differences. J. Immunol.
144, 29462954.
26.
26 Graubert, T. A., Russell, J. H., and Ley, T. (1996) The role of granzyme B in murine
models of acute graft-versus-host disease and graft rejection. Blood 87, 12321237.
27.
27 Graubert, T. A., DiPersio, J. F., Russell, J. H., and Ley, T. J. (1997) Perforin/granzyme-
dependent and independent mechanisms are both important for the development of graft-
versus-host disease after murine bone marrow transplantation. J. Clin. Invest. 100,
904911.
28 Korngold, R. and Sprent, J. (1985) Surface markers of T cells causing lethal graft-vs-host
28.
disease to class I vs class H-2 differences. J. Immunol. 135, 30043010.
29. Sprent, J., Schaefer, M., Lo, D., and Korngold, R. (1986) Properties of purified T cell
subsets. II. In vivo responses to class I vs. class II H-2 differences. J. Exp. Med. 163,
9981011.
30. Dey, B., Yang, Y. G., Preffer, F., Shimizu, A., Swenson, K., Dombkowski, D., and Sykes,
M. (1999) The fate of donor T-cell receptor transgenic T cells with known host antigen
specificity in a graft-versus-host disease model. Transplantation 68, 141149.
Induction of Chimerism and Tolerance 459
30
Induction of Chimerism and Tolerance
Using Freshly Purified or Cultured
Hematopoietic Stem Cells in Nonmyeloablated Mice
Nikos Emmanouilidis and Christian P. Larsen
Summary
The development of protocols to induce a state of durable mixed allogeneic hematopoietic
chimerism to confer robust donor-specific transplant tolerance has been a major focus of the
transplant community for the past decade. High levels of mixed allogeneic hematopoietic chi-
merism across a full major histocompatibility complex (MHC) barrier can be achieved by total
myeloablation and transfusion of host and donor bone-marrow cells (1) or, as shown more
recently, can be achieved with nonmyeloablative preconditioning regimes in combination with
donor bone-marrow transfusions (24) or transfusion of purified hematopoietic stem cells
(HSC) (57). Here we illustrate useful experimental techniques for the study of hematopoietic
chimerism in rodents by describing a system in which chimerism can be induced using
nonmyeloablative conditioning, short-term costimulation blockade, and transplantation of
purified or cultured hematopoietic stem cells.
Key Words: Hematopoietic stem cell; Sca-1; chimerism; tolerance; magnetic bead selec-
tion; MACS; busulfan; anti-CD40-ligand; CTLA4-Ig.
1. Introduction
Three factors are crucial for the establishment and long-term maintenance of
hematopoietic chimerism across allogeneic barriers in mice. First, deletion or anergy
of donor-specific allo-reactive T-cells, which is necessary to avoid rejection of the
bone-marrow graft. In this model, this is achieved by blocking the CD28 and CD40
co-stimulatory pathways using CTLA4-Ig and anti-CD40L along with the administra-
tion of donor antigen. Second, niches for donor stem cells need to be available within
the recipients bone-marrow microenvironment. This can be achieved by admini-
stration of a single nonmyeloablative dose of busulfan, which selectively kills HSC,
thus freeing stem-cell niches and making them available for the engraftment of trans-
planted donor HSC. Third, the transfused cells must contain multipotent HSC capable
of homing, engraftment, and self-renewal.
459
460 Emmanouilidis and Larsen
It is widely recognized that HSC in mice can be identified using flow cytometry by
their expression of stem-cell antigen 1 (SCA-1) (8) and lack of expression of various
lineage markers (e.g., CD3, CD19, and so on). HSC in mice are a minor fraction within
the bone-marrow compartment and represent only about 0.050.1% of total bone-mar-
row cells, which necessitates the development of specialized techniques in order to
study hematopoietic stem cell behavior in detail. In this chapter we will describe how
to purify Sca-1posLinneg HSC using a magnetic cell-sorting procedure and how to use
these HSC for the induction of durable and robust hematopoietic chimerism in a fully
allogeneic transplant model in non-myeloablated mice. The preconditioning of recipi-
ent mice in this model takes 5 d and involves the induction of anergy in recipient
donor-reactive T-cells, the administration of a myelotoxic drug, and transfusion of
donor antigen. Anergy or deletion of donor specific allo-reactive T-cells in this model
is achieved by blocking the CD28 and CD40 co-stimulatory pathways using CTLA4-
Ig and anti-CD40L along with the administration of donor antigen (e.g., unseparated
bone-marrow cells, splenocytes, and so on). These agents prevent the early destruc-
tion of donor HSC and allow sufficient time for newly evolving donor antigen-pre-
senting cells (APC) to migrate to the thymus, where they participate in the process of
shaping the T-cell repertoire through central deletion of new donor-specific T-cells.
Nonmyeloablative preconditioning regimens, whether involving myelotoxic drugs or
-irradiation, have been proven to be sufficient and reliable for the induction of robust
long-term hematopoietic chimerism (27,9). Thus we describe a regime that makes
use of a single dose of busulfan, an agent with stem-cell selective toxicity, which, if
administered 1 d before transfusion of the engrafting HSC dose, frees stem-cell niches
within the recipient bone-marrow compartment and makes them available for donor
HSC engraftment.
Recent results have shown the ability to expand HSC ex vivo (1014) and the
potency of these cells for self-renewal and long-term bone-marrow engraftment (5,6).
Thus, we also provide a brief example of how short-term in vitro cultured HSC can be
studied within this protocol.
2. Materials
1. Adult male 68-wk-old C57BL/6 (H-2b) (1) and adult male 68-wk-old BALB/c (H-2d)
mice (JAX, Bar Harbor, ME).
2. Murine anti-CD40L (MR1) (BioExpress, West Lebanon, NH).
3. Human CTLA4-Ig (Bristol-Myers Squibb, New York, NY).
4. Biotinylated Mouse Lineage Panel (PharMingen, San Diego CA).
5. Anti-Biotin Microbeads (Miltenyi Biotec, Bergisch-Gladbach, Germany).
6. Anti-FITC Multisort Kit (Miltenyi Biotec).
7. Auto-MACS cell sorter (Miltenyi Biotec).
8. FITC-conjugated anti-Sca1 (Ly-6A/E), (PharMingen).
9. APC-conjugated anti-biotin mAB (PharMingen).
10. Fc-block (Fc_ III/II receptor) (PharMingen).
11. Whole blood cell lysis kit (R&D Systems, Minneapolis, MN).
12. R10-medium: RPMI-1640 (Mediatech, Herndon, VA) with 10% fetal calf serum (FCS)
(Mediatech).
Fig. 1. Unseparated C57/BL6 bone-marrow cells, 2 107 are administered on d 0, followed

by injection of costimulation-blocking molecules anti-CD40L and CTLA4-Ig () on d 0, 2, 4,
and 6. Busulfan () is administered on d 5, one day before donor hematopoietic stem cells are
injected on d 6.
13. MACS buffer: Dulbeccos 1X phosphate-buffered saline (1X PBS) (Mediatech) with
2 mM ethylene-diamine-tetraacetic acid (EDTA) (Sigma-Aldrich, St. Louis, MO) and
0.5% bovine serum albumin (BSA) (Sigma-Aldrich).
14. RINSE buffer: Dulbeccos 1X phosphate-buffered saline (1X PBS) (Mediatech) with 2
mM ethylene-diamine-tetraacetic acid (EDTA) (Sigma-Aldrich).
15. Histopaque 1083g/mL (Sigma-Aldrich).
16. Trypan blue solution, liquid 0.4% (Sigma-Aldrich).
17. Twenty-four-well cell-culture flat-bottom plate (Fisher Scientific, Pittsburgh, PA).
18. Serum-free expansion medium (SFEM) (Stem Cell Technology, Vancouver, Canada).
19. Mouse stem-cell factor (SCF), mouse Flt-3/Flk-2 ligand (FLT3L), recombinant mouse
interleukin (IL)-3, recombinant mouse IL-6 (all from Sigma-Aldrich): resolve lyophilized
growth factors to 100 ng/10 L SCF, 100 ng/10 L FLT3L, 20 ng/10 L IL-3, and 20 ng/
10 L IL-6 in R10-medium, aliquot to volumes of 100 L per tube, and store at 20C.
20. Scissors and microsurgical forceps (Fine Science Tools, North Vancouver, BC, Canada).
3. Methods
3.1. Preconditioning of Recipient Mice
To avoid rejection of transferred hematopoietic stem cells in allogeneic donor/
recipient settings, alloreactive recipient T-cells must be either depleted or suppressed.
A 6-d recipient preconditioning protocol is outlined in Fig. 1, utilizing costimulatory
pathway-blocking molecules anti-CD40 ligand and CTLA4-Ig as immunosupressive
drugs and busulfan as a non-myeloablative agent to create space within the recipient
bone-marrow microenvironment.
Day 1:
1. Euthanize C57BL/6 donor mice (n = 4) by cervical dislocation. Harvest tibiae, femura,
and humeri with sterile instruments.
2. Flush out the bone marrow from bone shafts onto a Petri dish with cold sterile R10
medium using a syringe and a 26-gage needle. Aspirate bone-marrow cylinders repeat-
edly to disaggregate cells.
3. Pass cells through a 40-m cell strainer into a 50-mL conical tube and add R10 medium
to a total volume of 50 mL.
4. Spin down cells at 300g and 4C, resuspend the pellet in 25 mL of RBC-lysis buffer, and
incubate for 15 min on ice.
5. Rescue cells by adding 25 mL cold R10 medium and spin down at 300g and 4C.
6. Resuspend the pellet by adding 20 mL of sterile saline. Take out 10 L of the cell suspen-
sion into a 1.5-mL Eppendorf tube and add 986 L of 1X PBS. Add 4 L of the fluores-
cent dye and count the nucleated cells on a fluorescent microscope using a standard
hematocytometer (alternatively, stain cells by adding 4 L of trypan blue and count cells
on a phase-contrast microscope).
7. Take out desired cell number (2.0 107 cells per recipient) and add sterile saline to bring
the cell suspension to a concentration of 2.0 107 cells/300 L.
8. For each recipient mouse, prepare one 1-mL syringe with a 30-gage needle and 300 L of
the cell suspension (2.0 107 cells) and store syringes on ice.
9. Anesthetize BALB/c recipient mice and administer 500 g anti-CD40L and 500 g
CTLA4-Ig per recipient intraperitoneally. Each recipient is then injected with 2.0 107
C57BL/6 bone-marrow cells via penile-vein injection.
Days 2 and 4:
Each BALB/c recipient receives 500 g anti-CD40L and 500 g CTLA4-Ig intrap-
eritoneally on d 2 and 4.
Day 5:
Busulfan is administered at 600 g/mouse intraperitoneally.
Day 6:
See Subheading 3.2.
3.2. Purification of Sca1posLinneg Hematopoietic Stem Cells
The purification procedure takes about 8 h (2) from harvest of bones to final elution
of Sca1posLinneg hematopoietic stem cells from the Auto-MACS.
3.2.1. Harvest of Cells
1. Euthanize C57BL/6 donor mice (n = 10) (3) by cervical dislocation. Harvest tibiae,
femura, and humeri with sterile instruments.
2. Flush out the bone marrow from bone shafts onto a Petri dish placed on ice with cold
sterile R10 medium using a syringe and a 26-gage needle.
3. Aspirate bone-marrow cylinders repeatedly to disaggregate cells.
4. Pass the cells through a 40-m cell strainer into a 50-mL conical tube, add R10 medium
to a total volume of 50 mL, and spin down cells at 300g and 4C.
5. Resuspend the pellet in 25 mL of RBC-lysis buffer and incubate for 15 min on ice.
6. Recover cells by adding 25 mL cold R10 medium and spin down at 300g and 4C.
7. Resuspend the pellet by adding 14 mL of rinse buffer.
8. Take two 15-mL conical tubes and fill each with Histopaque 1083 at room temperature.
Divide the cell suspension into two equal volumes and carefully load cells as a second
layer in each tube on top the Histopaque 1083 layer.
9. Spin down both tubes at 20C and 400g for 20 min with slow acceleration and centrifuge
brakes off.
10. Take out the middle layer with a Pasteur pipet and resuspend cells in 14 mL cold R10
medium.
11. Count nucleated cells: mix well, take out 10 L of the cell suspension into a 1.5-mL
Eppendorf tube, and add 986 L of MACS buffer. Add 4 L of a fluorescent dye and
count the nucleated cells on a fluorescent microscope using a standard hematocytometer
(alternatively, stain cells by adding 4 L of trypan blue and count cells on a phase-con-
trast microscope).
12. Spin down remaining cells at 300g for 15 min at 4C and resuspend the pellet in MACS
buffer at 150 L/108 cells.
3.2.2. Labeling With Magnetic Beads
1. To reduce nonspecific binding of antibodies and beads, pre-incubate cells with Fc-block
(Fc III/II receptor) by adding 5 L per 108 cells (per 150 L cell suspension) and incu-
bate at 612C for 10 min.
2. Add 4 L of anti-Sca1-FITC monoclonal antibody, mix well, and incubate in the dark at
612C for 20 min.
3. Add MACS buffer to a total volume of 15 mL and spin down at 300g for 15 min.
4. Discard supernatant and repeat the washing step a second time at 15 mL total volume,
spin down at 300g at 4C, and resuspend the pellet by adding 900 L MACS buffer per
108 cells.
5. Pass the cells through a 40-m nylon-mesh filter into a 15-mL conical tube.
6. Add anti-FITC microbeads at 100 L per 108 cells (per 900 L cell suspension), mix
well, and incubate in the dark at 612C for 20 min.
7. Divide cell suspension into equal volumes in two 15-mL conical tubes, add ice-cold
MACS buffer to each tube to a total volume of 15 mL, and spin down at 300g at 4C.
8. Repeat the washing step a second time, unite the pellets, and resuspend the cells in MACS
buffer at a total volume of 2500 mL.
9. Filter cells through a 40 m nylon-mesh filter before moving to the next step.
3.2.3. Magnetic Cell Sorting by Auto-MACS
3.2.3.1. FIRST STEP: SCA-1POS SELECTION
1. Place the tube with the labeled cells under the uptake port and place 15-mL conical tubes
under each eluent cock to save labeled (Sca1pos) and nonlabeled (Sca1neg) cells.
2. Select PosselDS as the sorting mode and hit start (4).
3. Right before the cell uptake is completed, add 500 L of MACS buffer to make sure that
all cells are loaded onto the columns.
4. After elution of cells, save cells on ice and rinse the Auto-MACS following the manu-
facturers protocol.
5. Place the negative fraction from the first magnetic separation under the uptake port and
hit PosselDS a second time to sort out any remaining Sca1pos cells. Collect labeled cells
into the same 15-mL tube as the first time.
6. Count nucleated cells from the labeled (Sca1pos) fraction: mix well, take out 10 L of the
cell suspension, and add 89 L of MACS buffer. Add 1 L of the fluorescent dye and
count the nucleated cells on a fluorescent microscope using a standard hematocytometer
(alternatively, stain cells by adding 1 L of trypan blue and count cells on a phase-con-
trast microscope).
Fig. 2. Percentage of Sca1posLinneg cells in C57/BL6 mice from unseparated bone marrow
(A), after magnetic bead selection (B), and after 6-d Sca1posLinneg in-vitro culture (C).
3.2.3.2. SECOND STEP: LINEAGE DEPLETION

1. Take Sca1pos cells, add 100 L of release agent per 4 mL of cell suspension, and incubate
for 30 min at 4C.
2. Spin down cells at 300g for 10 min and resuspend the pellet in 50 L MACS buffer per
107 cells.
3. Add 30 L of stop agent per 107 cells and incubate for 5 min.
4. Add the biotinylated lineage panel at 20 L of each antibody per 107 cells and incubate
for 20 min on ice.
5. Wash cells twice by adding MACS buffer to total volumes of 10 mL and spin down cells
at 300g for 10 min.
6. Resuspend the pellet in 80 L of MACS buffer per 107 cells, add 20 L of anti-biotin
microbeads per 107 cells, and incubate at 4C for 20 min.
7. Wash cells once by adding MACS buffer to a total volume of 14 mL, spin down at 300g
for 10 min, and finally resuspend the pellet in 1500 L MACS buffer.
8. Load the sample under the uptake port of the AutoMACS and run the sample on Deplete
mode.
9. Right before the cell uptake is completed, add 500 L of MACS buffer so that all cells
will be loaded on the columns.
10. Save the cells in 15-mL conical tubes on ice.
11. Count the nonlabeled fraction (Sca1posLinneg cells): mix well, take out 1.0 L of the cell
suspension, and add 9 L of MACS buffer. Add 0.5 L of a fluorescent dye and count the
nucleated cells on a fluorescent microscope using a standard hematocytometer (alterna-
tively, stain cells by adding 0.5 L of trypan blue and count live cells on a phase-contrast
microscope).
12. Use Sca1posLinneg cells for injection (see Subheading 3.3.) or cell culture (see Subhead-
ing 3.4.). Save a fraction of sorted cells to quantify purity and yield of HSC magnetic
bead selection by flow cytometry (see Fig. 2 and Subheading 3.5.1.).
3.3 Injection of Hematopoietic Stem Cells
1. Take desired cell number of Sca1posLinneg cells and prepare for injection: either add ster-
ile saline in order to dilute to desired cell concentration or transfer cells into 2-mL
Eppendorf tubes and spin down tubes on a microfuge for 7 min at 300g to increase cell
concentration (5).
2. Discard supernatant and add sterile saline to desired cell concentration for injection vol-
umes of 250300 L per mouse.
3. Save remaining cells on ice to determine purity of the purification procedure by flow
cytometry.
4. Anesthetize recipient mice and use 1-mL syringes with 30-gage needles for penile-vein
injections. Beginners should make use of a microscope or magnifying goggles with 35
magnification.
5. Simultaneously each BALB/c is injected with 500 g anti-CD40L and 500 g CTLA4-Ig
intraperitoneally.
3.4. HSC Short-Term Culture in Serum-Free Expansion Medium

In order to use short-term cultured HSC as engrafting cells within the described
chimerism-induction protocol, Sca1posLinneg HSC must be purified and plated in cul-
ture on the first d of the recipient preconditioning.
1. Purify HSC as described in Subheading 3.2.
2. Prepare a 24-flat-bottom-well plate by adding 460 L of fresh and sterile SFEM
medium and place in an incubator at 37C with 5% CO2 and 97% humidity.
3. Take desired number of purified Sca1posLinneg cells (20,000 cells per well), bring to a
minimum spinning volume of 2000 L by adding fresh SFEM medium, and spin down at
300g for 5 min.
4. Resuspend the pellet in SFEM medium at 20,000 cells/500 L.
5. Add 20,000 cells (500 L volume) to each well.
6. Add 100 ng SCF, 100 ng FLT3L, 20 ng IL3, and 20 ng IL6 to each well.
7. Place plate in an incubator at 37C with 5% CO2 and 97% humidity.
8. Exchange half of the medium volume after 2, 4, and 5 d by carefully aspirating 540 L of
the medium from the sides of each well and add 500 L of warmed fresh SFEM medium.
Further, renew growth factors on each medium change by adding 100 ng/10 L SCF,
100 ng/10 L FLT3L, 20 ng/10 L IL-3, and 20 ng/10 L IL-6. Examine the cell-culture
once a day on an inverted phase contrast microscope to verify the viability of the plated
cell (Fig. 3).
9. Harvest cells after 6 d of culture by washing the cloudy layer off the bottom of each
well.
10. Transfer cells into a 15-mL conical tube, spin down at 300g for 10 min, and resuspend the
pellet in sterile saline to the desired cell concentration.
11. Prepare 1-mL syringes with 30-gage needles with injection volumes of 250300 L cell
suspension.
12. Use remaining cells to characterize the cells and to quantify percentage of Sca1pos Linneg
cells by flow cytometry (proceed to Subheading 3.5.2.).
3.5. Flow Cytometry

3.5.1. Freshly Isolated HSC
1. Spin down cells at 300g for 5 min on a microfuge.
2. Discard supernatant using a Pasteur pipet.
3. Add 100 L of MACS buffer.
4. Add 1.0 L of APC-labeled anti-biotin mAB and incubate in the dark for 710 min at 4C.
Fig. 3. Sca1posLinneg hematopoietic stem cells after 4 d in serum-free expansion medium

supplemented with SCF, FLT3L, interleukin-3, and interleukin-6.
5. Wash once by adding 900 L MACS buffer and spin down for 5 min at 300g on a
microfuge.
6. Discard supernatant, resuspend the pellet in about 150 L of rinse buffer, and proceed to
flow cytometry.
3.5.2. Cultured HSC

1. Take 1.05.0 106 cells per tube and staining.
2. Spin down cells at 300g for 5 min on a microfuge.
3. Discard supernatant using a Pasteur pipet.
4. Resuspend pellet in 80 L of MACS buffer per 107 cells.
5. Add the biotinylated lineage panel at 20 L of each antibody per 107 cells and add 0.5 L
of anti-Sca1-FITC.
6. Incubate for 20 min on ice.
7. Add 500 L of wash buffer (e.g., MACS buffer) and spin down at 300g for 5 min on a
microfuge.
8. Resuspend the pellet in 100 L MACS buffer, add 1.0 L of APC-labeled anti-biotin
mAB, and incubate in the dark for 710 min at 4C.
9. Wash once by adding 900 L MACS buffer and spin down for 5 min at 300g on a micro-
fuge.
10. Discard supernatant, resuspend the pellet in about 150 L of rinse buffer, and proceed to
flow cytometry.
3.5.3. Screening BALB/c Recipient Mice

for Hematopoietic Chimerism After HSC Transfusion
1. Harvest 50 L of peripheral blood from each recipient.
2. Transfer blood onto a 96-V-bottom-well plate.
3. Add 150 L of RBC-lysis buffer per well and incubate on ice for 15 min.
Fig. 4: Expression levels of C57/BL6 H2-Kb (MHC-I) on different lineages of peripheral

blood cells in fully allogeneic preconditioned BALB/c H2-Kd-positive mice after transfusion
of 20,000 C57/BL6 Sca1posLinneg hematopoietic stem cells.
4. Spin down the plate at 300g and 4C for 5 min.

5. Discard supernatant and add 50 L of MACS buffer per well.
6. Add 0.5 L of anti-H2-Kb FITC and 0.5 L of either anti-CD3e-PE, anti-Ly6G-PE, anti-
CD11b-PE, or anti-CD45RB220-PE and incubate in the dark at 4C for 10 min.
7. Wash by adding 150 L MACS buffer and spin down at 300g for 5 min.
8. Discard supernatants, resuspend pellets in 100 L rinse buffer, and proceed to flow
cytometry. (Chimerism is measured as percentage of H2-Kb [MHC-I]-antigen-positive
cells in peripheral blood, Fig. 4.)
4. Notes
1. Expression levels of Sca-1 (Ly-6A/E) on HSC vary between different mouse strains from
high to very low and even nonexpression. Mice with a Ly-6.1 haplotype (such as BALB/c,
C3H/He, or CBA) have few Ly-6A/E-positive hematopoietic stem cells and are not suitable
for HSC purification by Sca-1 bead selection see Subheading 2.1.).
2. The viability of sorted cells greatly depends on avoiding unnecessary delays. Therefore,
work fast and have ice-cold buffers ready to use before starting the experiment.
3. The total cell number of unseparated bone-marrow cells at baseline should be at least 5.0
108 cells. Using tibiae, femura, and humeri will give a total of 0.71.0 108 bone-
marrow cells and about 0.71.0 105 HSC per mouse.
4. Magnetic columns should not be older than 1 wk. Using two new columns per experiment
will greatly increase yield and purity of the final cell sort.
5. Mark the orientation of Eppendorf tubes as inserted in microfuges to avoid accidental
loss of the pellet when discarding supernatants. Pellets of cell numbers as low as 50,000
cells are barely visible. Use a Pasteur pipet with a rubber suction ball to carefully discard
supernatants from Eppendorf tubes.
References
1.
1 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(5947), 168170.
2.
2 Wekerle, T., Sayegh, M. H., Ito, H., et al. (1999) Anti-cd154 or ctla4ig obviates the need
for thymic irradiation in a non-myeloablative conditioning regimen for the induction of
mixed hematopoietic chimerism and tolerance. Transplantation 68(9), 13481355.
3.
3 Wekerle, T., Kurtz, J., Ito, H., et al. (2000) Allogeneic bone marrow transplantation with
co-stimulatory blockade induces macrochimerism and tolerance without cytoreductive
host treatment. Nat. Med. 6(4), 464469.
4.
4 Durham, M. M., Bingaman, A. W., Adams, A. B., et al. (2000) Cutting edge: Administration
of anti-cd40 ligand and donor bone marrow leads to hemopoietic chimerism and donor-
specific tolerance without cytoreductive conditioning. J. Immunol. 165(1), 14.
5.
5 Bachar-Lustig, E., Li, H. W., Marcus, H., and Reisner, Y. (1998) Tolerance induction by
megadose stem cell transplants: Synergism between sca-1+ lin cells and nonalloreactive
t cells. Transplant. Proc. 30(8), 40074008.
6.
6 Uchida, N., Tsukamoto, A., He, D., et al. (1998) High doses of purified stem cells cause
early hematopoietic recovery in syngeneic and allogeneic hosts. J. Clin. Invest. 101(5),
961966.
7.
7 Emmanouilidis, N., Hussain, A., Adams, A. B., et al. (2002) Costimulation Blockade,
administration of Busulfan and Infusion of purified Sca1posLinneg bone marrow cells leads
to hematopoietic chimerism in a full MHC disparate mouse model. Am. J. Trans. Suppl.
3(2), 211.
8.
8 Spangrude, G. J., Heimfeld, S., and Weissman, I. L. (1988) Purification and characteriza-
tion of mouse hematopoietic stem cells. Science 241(4861), 5862.
9.
9 Adams, A. B., Durham, M. M., Kean, L., et al. (2001) Costimulation blockade, busulfan,
and bone marrow promote titratable macrochimerism, induce transplantation tolerance,
and correct genetic hemoglobinopathies with minimal myelosuppression. J. Immunol.
167(2), 11031111.
10.
10 Fraser, C. C., Szilvassy, S. J., Eaves, C. J., and Humphries, R. K. (1992) Proliferation of
totipotent hematopoietic stem cells in vitro with retention of long-term competitive in
vivo reconstituting ability. Proc. Natl. Acad. Sci. USA 89(5), 19681972.
11.
11 Petzer, A. L., Hogge, D. E., Landsdorp, P. M., Reid, D. S., and Eaves, C. J. (1996) Self-
renewal of primitive human hematopoietic cells (long-term-culture-initiating cells) in vitro
and their expansion in defined medium. Proc. Natl. Acad. Sci. USA 93(4), 14701474.
12.
12 Neben, S., Donaldson, D., Sieff, C., et al. (1994) Synergistic effects of interleukin-11 with
other growth factors on the expansion of murine hematopoietic progenitors and mainte-
nance of stem cells in liquid culture. Exp. Hematol. 22(4), 353359.
13. Koller, M. R., Oxender, M., Brott, D. A., and Palsson, B. O. (1996) Flt-3 ligand is more
potent than c-kit ligand for the synergistic stimulation of ex vivo hematopoietic cell ex-
pansion. J. Hematother. 5(5), 449459.
14. Bryder, D. and Jacobsen, S. E. (2000) Interleukin-3 supports expansion of long-term
multilineage repopulating activity after multiple stem cell divisions in vitro. Blood 96(5),
17481755.
Induction of Mixed vs Full Chimerism 469
31
Induction of Mixed vs Full Chimerism to Potentiate GVL
Effects After Bone-Marrow Transplantation
Markus Y. Mapara and Megan Sykes
Summary
Graft-vs-host (GVH) alloresponses mediated by delayed donor lymphocyte infusions (DLI)
can occur in the absence of GVHD (graft-vs-host disease). These GVH responses are confined
to the lymphohematopoietic system and mediate graft-vs-leukemia (GVL) reactions without
causing GVHD. Although interaction of donor T-cells with host-derived antigen-presenting
cells (APC) is central to the development of GVHD, we were able to show that this T-cell/APC
interaction is also critical for the induction of GVL, which can occur in the absence of GVHD
if donor T-cell administration is delayed to a time point after bone-marrow transplant (BMT)
when conditioning-induced inflammation has subsided. Induction of mixed hematopoietic chi-
merism using the method described below allows analysis of these interactions between host
APC and donor T-cells.
Key Words: Mixed chimerism; donor lymphocyte infusions; GVL; GVH; transplantation.
1. Introduction
Rodent studies have shown that administration of donor lymphocyte infusions (DLI)
to mixed hematopoietic chimeras can mediate a graft-vs-host (GVH) reaction that is
restricted to the lymphohematopoietic system, as indicated by the conversion from
mixed chimerism to full chimerism without the concomitant development of GVH
disease (GVHD) (1,2). This suggested an approach to using the GVH alloresponse to
achieve graft-vs-leukemia (GVL) without GVHD. In humans, GVL effects have been
achieved by the administration of DLI to patients with relapsed chronic myeloid leu-
kemia (CML) after allogeneic bone marrow transplant (BMT) (3,4). In a clinical trial
at our institution, mixed chimeras induced with nonmyeloablative conditioning have
enjoyed striking remissions of advanced, refractory lymphoid malignancies, both with
and without the administration of DLI (5,6). Based on these surprisingly powerful
GVL effects, we hypothesized that nontolerant donor lymphocytes might result in
stronger GVL effects when administered to mixed hematopoietic chimeras than to full
donor chimeras (see Fig. 1). DLI administration to mixed chimeras produced dramati-
469
470 Mapara and Sykes
Fig. 1. Generation of mixed and full donor hematopoietic chimeras.
cally improved leukemia-free survival compared to administration of the same DLI to

full allogeneic chimeras. DLI also converted mixed chimeras to full chimeras without
causing GVHD. The magnitude of the GVL effect was dependent on the level of major
histocompatibility complex (MHC) class I expression on recipient hematopoietic cells
in mixed chimeras. Thus, induction of mixed chimerism followed by delayed DLI
provides an approach to inhibiting GVHD that optimizes GVL effects (7).
2. Materials
2.1. Animals
1. C57BL/6 mice.
2. B10.A mice.
3. B6.2m/ mice.
2.2. Operating Tools
1. Microdissecting scissors.
2. Small forceps.
3. Small operating scissors.
2.3. Apparatus and Chemicals
1. Fifty-milliliter conical tubes.
2. Sterile spleen mesh.
3. Syringes (3 cc, 20 to 35 cc).
4. Nineteen-gage, 23-gage, 25-gage needles.
5. Petri dishes, 100 20 mm.

6. CO2 chamber for euthanasia.
7. CML medium: 500 mL RPMI 1640, 50 mL fetal calf serum (FCS; complement-inacti-
vated 30 min at 55C), 5 mL 100X nonessential amino acids, 0.146 g L-glutamine (final
concentration 2 mM), 0.05 g sodium pyruvate (final concentration 1mM), 5 mL 1 mM
HEPES buffer (final concentration 0.01 mM), 1 mL 12.5 mM 2-mercaptoethanol (final
concentration 0.025 mM), 50,000 U penicillin, 50 mg streptomycin.
8. Bone marrow medium (BMM): 500 mL Media 199, 5 mL HEPES buffer, 5 mL DNAse
(1 mg/mL in H2O), 40 L gentamycin (at 50 mg/mL).
9. FACS buffer: 100 mL 10X Hanks balanced salt solution, 900 mL D.I. H2O, 210 L of 1
N sodium hydroxide (NaOH), 1 g sodium azide (NaN3), 1 g bovine serum albumin (BSA)
fraction V.
10. Antibodies for target cell depletion:
a. In vitro T-cell depletion: a-CD4 (GK1.5), -CD-8 (2.43).
b. In vivo natural killer (NK) cell depletion: NK1.1.
3. Methods
3.1. Animals
B10.A (H-2a: Kk, Ak, Ek, Dd, Ld) female donor mice purchased from Frederick
Cancer Research Facility, NCI, were used at 8 to 12 wk of age. Female C57BL/6 (B6:
H-2b, I-E) recipient mice were purchased from Frederick Cancer Research Facility,
NCI, and used at 10 wk of age. In some experiments, B6.2m/ animals and wild-
type control B6 mice were purchased from Jackson Laboratories (Bar Harbor, ME).
All mice were housed in autoclaved micro-isolator environments, and all manipula-
tions were performed in a laminar flow hood (see Note 1).
3.2. Bone-Marrow Harvest, Transplantation,

and Generation of Mixed and Full Donor Chimeras
1. Animals were sacrificed by CO2 inhalation and dissected under sterile conditions.
2. The femora and tibiae were collected and flushed with BMM.
3. After cell counting, cells were subjected to in vitro T-cell depletion by antibody-comple-
ment lysis (see Subheading 3.3.).
4. B6 (H2b) mice received 10.25 Gy total body irradiation from a 37Cs irradiator in a single
dose (dose rate 1cGy/min, see Note 2) and were reconstituted with either a mixture of
T-cell-depleted (TCD) B6 (5 106 bone-marrow cells) plus B10.A (H2a) (15 106) bone-
marrow cells or with TCD B10.A bone marrow cells (15 106) alone by lateral tail-vein
injection 4 h after irradiation. The complete experimental setup is depicted in Fig. 1.
3.3. In Vitro T-Cell Depletion

In vitro depletion of CD4 and CD8 T-cells from the bone-marrow inoculum was
performed using mAbs and rabbit low toxic complement lysis (see Note 3) as previ-
ously described (7):
1. BMC were adjusted to a working cell concentration of 5 107 cells/mL with BMM.
2. After washing, cells were labeled with a predetermined (by separate titrations of deple-
tion activity) amount of anti-CD4 (Gk1.5) and anti-CD8 (2.43) antibodies from ascites
for 30 min at room temperature.
3. Thereafter, cells were washed twice and incubated with a predetermined amount of low-
toxic rabbit complement for 45 min at 37C.
4. Thereafter cells were washed and T-cell depletion efficiency was determined by flow
cytometry.
5. Complete T-cell depeletion is critical for establishing mixed chimeras, as residual host or
donor T-cells from the bone-marrow graft will lead to spontaneous development of either
full host or full donor chimerism, respectively (8).
3.4. In Vivo NK Cell Depletion
To demonstrate the role of direct allorecognition, it was necessary to establish
mixed chimeras in which host hematopoietic cells were devoid of MHC class expres-
sion. This was performed by reconstituting wild-type B6 mice with a mixture of
T-cell-depleted B10.A marrow and T-cell-depleted B6.2m/ marrow. It is known
that recipient NK cells reject MHC class I deficient bone-marrow cells, a phenomenon
also known as hybrid resistance, which describes the observation that F1 mice reject
the parental bone-marrow grafts due to the missing self MHC I molecules. In order to
achieve engraftment of MHC I deficient B6.2m/ marrow in animals receiving B10.A
plus B6.2m/ bone-marrow cells, recipient mice received 150 g of mAb PK136
(NK1.1) on d 6 and 1 prior to BMT.
3.5. Phenotyping of Chimeras
Chimerism in white blood cells (WBC) and bone marrow was assessed by two-color
or three-color flow cytometry (FCM) using a FACScan cytometer (Becton Dickinson,
Mountain View, CA).
1. Peripheral blood was collected into heparinized Eppendorf tubes and subjected to deion-
ized water lysis.
2. For double or triple color staining, 106 cells were incubated in the presence of directly
fluorescein-isothiocyanate (FITC)-, phycoerythrin (PE)-, or biotin (Bio)-conjugated
monoclonal antibodies (mAbs) for 30 min at 4C.
3. Development of bio-labeled mAbs was performed by subsequent incubation with phyco-
erythrin-conjugated avidin (PEA) for 10 min.
4. To reduce nonspecific binding of mAbs, 10 L of 2.4G2 (anti-Fc-RII-receptor, CDw32)
hybridoma supernatant (9) were added to all tubes. The following antibodies were used
for chimerism analyses in various cell lineages: anti-CD4-FITC, anti-CD8-FITC, anti-
B220-FITC (all purchased from PharMingen, San Diego CA), anti-Mac-1-FITC (CalTag,
San Francisco, CA), and 34-2-12-Bio (anti-H2-Dd prepared in our laboratory). Nonreac-
tive control mAb HOPC-FITC or HOPC-Bio (mouse IgG2a prepared in our laboratory)
and rat IgG2b-PE (Pharmingen) were used as negative controls.
5. The percentage of donor cells within each leukocyte population was determined using the
following formula: Net percent donor cells of a particular lineage 100%/(net percent
donor cells of a particular lineage + net percent host cells of a particular lineage), where
net refers to the percentage obtained after subtraction of staining with control antibody.
For the H-2 class I allele-specific mAb, the mouse strain (donor or host) not bearing the
allele recognized by the mAb was used as the negative control and the strain expressing
the allele was used as the positive control to determine the cutoffs for reactivity with the
H2-specific mAb. Exclusion of dead cells was performed by propidium iodide (PI) stain-
ing and live gating on PI negative cells. Ten thousand events were collected and ana-
lyzed. The different peripheral blood leukocyte populations were distinguished by their
forward scatter (FSC) and side scatter (SSC) properties: FSC low and SSC low (lympho-
cytes), SSC high (granulocytes), and FSC high and SSC low (monocytes).
3.6. Preparation of DLI

For preparation of DLI, B10.A spleens were harvested and gently teased in ACK
lysing buffer (Biowhittaker, Walkersville, MD). Single-cell suspensions were filtered
through nylon mesh. Donor lymphocytes were administered in the form of 3 107
B10.A spleen cells on d 56 post-BMT. Animals were randomized between cages to
avoid cage-related bias.
3.7. EL4 Cell Culture
The T-cell leukemia cell line EL4 (500 cells) was administered intravenously on
d 63. EL4 cells were originally obtained from ATCC and were cultured in CML
medium. A new vial of a working bank of EL4 cells was thawed for each experiment
and maintained in culture for a maximum of 2 wk.
4. Notes
1. As in all BMT experiments, it is extremely important to have an animal facility that com-
plies with specific pathogen-free (SPF) or SPF-like standards to prevent the development
of opportunistic infections after allogeneic bone-marrow transplantation.
2. In order to prevent conditioning-induced mortality, it is important to use a low dose rate
(100 rad/min) for total body irradiation.
3. Achieving complete T-cell depletion of the bone-marrow graft is crucial for obtaining
stable mixed chimerism and avoiding GVH reactions that might induce a proinflammatory
state that could persist at the time of DLI and lead to GVHD. The method described in
Subheading 3.3. is based on antibody-complement-mediated lysis of T-cells. Appropri-
ate antibody titers of Gk1.5 and 2.43 antibodies have to be predetermined by titration, to
allow optimal coating of target cells and to avoid anticomplementarity due to excess an-
tibody. It is also important to select rabbit complement batches that have a very low
background toxicity (i.e., low lysis without antibody preincubation). Alternatively, thor-
ough T-cell depletion also can be achieved using magnetic cell separation with MACS
beads from Miltenyi.
Acknowledgments
Dr. Markus Y. Mapara was supported by Deutsche Forschungsgemeinschaft (DFG-
Ma 1664/2-1) and the Deutsche Krebshilfe (70-2981-MaI). The work was supported
by NCI grant RO1 CA 79989.
References
1.
1 Sykes, M., Sheard, M. A., and Sachs, D. H. (1988) Graft-versus-host-related immunosup-
pression is induced in mixed chimeras by alloresponses against either host or donor
lymphohematopoietic cells. J. Exp. Med. 168, 23912396.
2.
2 Pelot, M. R., Pearson, D. A., Swenson, K., et al. (1999) Lymphohematopoietic graft-vs-
host reactions can be induced without graft-vs.-host disease in murine mixed chimeras
established with a cyclophosphamide-based nonmyeloablative conditioning regimen. Biol.
Blood Marrow Transplant. 5, 133143.
3.
3 Kolb, H. J., Mittrmuller, J., Clemm, C., et al. (1990) Donor leukocyte transfusions for
76, 24622465.
4.
4 Kolb, H.-J., Schattenberg, A., Goldman, J. M., et al. (1995) Graft-versus-leukemia effect
5.
5 Spitzer, T. R., McAfee, S., Sackstein, R., et al. (2000) Intentional induction of mixed
chimerism and achievement of antitumor responses after nonmyeloablative conditioning
therapy and HLA-matched donor bone marrow transplantation for refractory hematologic
malignancies. Biol. Blood Marrow Transplant. 6, 309320.
6.
6 Sykes, M., Preffer, F., McAffee, S., et al. (1999) Mixed lymphohaemopoietic chimerism
and graft-versus-lymphoma effects after non-myeloablative therapy and HLA-mismatched
bone-marrow transplantation. Lancet 353, 17551759.
7.
7 Mapara, M. Y., Kim, Y.-M., Wang, S. P., Bronson, R., Sachs, D. H., and Sykes, M. (2002)
Donor lymphocyte infusions mediate superior graft-versus-leukemia effects in mixed com-
pared to fully allogeneic chimeras: a critical role for host antigen-presenting cells. Blood
100, 19031909.
8. Ildstad, S. T., Wren, S. M., Bluestone, J. A., Barbieri, S. A., Stephany, D., and Sachs, D.
H. (1986) Effect of selective T cell depletion of host and/or donor bone marrow on lym-
phopoietic repopulation, tolerance, and graft-vs-host disease in mixed allogeneic chime-
ras (B10 + B10.D2B10). J. Immunol. 136, 2833.
9. Unkeless, J. C. (1979) Characterization of a monoclonal antibody directed against mouse
macrophage and lymphocyte Fc receptors. J. Exp. Med. 150, 580596.
GVL Induction and Control of GVH Reactions 475
32
Application of Donor Lymphocytes Expressing
a Suicide Gene for Early GVL Induction and Later Control
of GVH Reactions After Bone-Marrow Transplantation
Attilio Bondanza, Fabio Ciceri, and Chiara Bonini
Summary
Allogeneic hematopoietic stem cell transplantation (allo-SCT) is the treatment of choice for
many malignant diseases. It is recognized that the curative potential of allo-SCT relates closely
to the immune advantage conferred by allogeneic T-lymphocytes. In allo-SCT donor T-lym-
phocytes favor engraftment, provide early immune reconstitution, and fight the underlying
malignancythe so-called graft-vs-leukemia (GVL) effect. These benefits are counterbalanced
by the occurrence of a life-threatening disease: graft-vs-host disease (GVHD). A suicide gene
encodes a protein able to convert a nontoxic prodrug into a toxic product. Therefore, cells
expressing the suicide gene become selectively sensitive to the prodrug. The transfer of a sui-
cide gene into donor lymphocytes could allow, upon administration of the prodrug, the in vivo
selective elimination of transduced lymphocytes, resulting in the switch-off of GVHD, thus
allowing full exploitation of the curative potential of donor T-lymphocytes in the context of
allo-SCT. In this chapter the rationale, the materials, and the methods of the suicide-gene strat-
egy with human T-lymphocytes are described.
Key Words: Bone marrow transplantation; graft-vs-leukemia, graft-vs-host disease, stem
cells.
1. Introduction
1.1. Adoptive Transfer of Donor T-Lymphocytes in Allo-SCT
Both preclinical and clinical data suggest that the curative potential of allo-SCT
closely relates to the immune advantage of allogeneic lymphocytes. Although neither
the nature of effector cells nor the antigenic targets have been fully elucidated, the
allogeneic advantage in abating relapse probability is well documented (1). Further-
more, the administration of donor lymphocytes has become a new tool for treating
relapse of hematologic malignancies occurring after allo-SCT (2,3) and holds promise
also for the cure of solid tumors (4).
475
476 Bondanza, Ciceri, and Bonini
Before donor lymphocyte infusion (DLI) became a standard treatment, patients

relapsing after allo-SCT had a very poor prognosis. Since 1989, many patients affected
by hematologic malignancies relapsing after allo-SCT achieved complete remissions
after DLI without requiring any further cytoreductive therapy (2,510), thus demon-
strating the efficacy of the allogeneic immune advantage in allo-SCT. Unfortunately,
clinical success was often associated with the occurrence of GVHD, especially when
high numbers of T-cells were infused (3,11). New DLI regimensi.e., infusion of esca-
lating dose of lymphocyteslowered the incidence of GVHD, but this complication
remains the major obstacle for success of DLI (13).
A possible alternative strategy relies on the generation and infusion of leukemia-
specific T-lymphocytes. Up to now, adoptive immunotherapy with antigen-specific
T-cells has been used in viral disease such as cytomegalovirus (CMV) reactivation syn-
dromes (14) and EpsteinBarr virus lymphoproliferative disease (EBV-LPD) (15), and
has been investigated for the cure of metastatic solid tumors such as melanoma (16).
The overall results of adoptive immunotherapy with virus-specific T-cells are encour-
aging, with good efficacy and virtually no occurrence of GVHD (17). Thus, results of
such an approach targeted on leukemia produced high expectations, till now still to be
met (1820). Indeed, despite the fact that GVHD and GVL are potentially mediated by
donor T-lymphocytes with different specificities, we cannot yet differentiate GVL and
GVHD effector cells.
1.2. The Suicide-Gene Strategy
In allo-SCT, the transfer of a heterogeneous population of donor T-cells with a wide
range of antigen specificities offers some clear advantages over strategies involving
infusion of antigen-specific cells. First, such a population can provide more complete
immune reconstitution in immune-compromised allo-SCT patients. Furthermore, since
most targets of GVL are still unknown, the use of the entire T-cell repertoire seems to
be the best option to fully exploit GVL. To reduce the risks of uncontrolled GVHD
associated with DLI, genetic modification of donor lymphocytes to confer an inducible
suicide phenotype has been proposed and tested at clinical level. A suicide gene
encodes for a protein able to convert a nontoxic prodrug into a toxic product. Therefore,
cells expressing the suicide gene become selectively sensitive to the prodrug and can be
selectively eliminated if unwanted effects such as GVHD occur (Fig. 1).
Different suicide genes have been investigated and described in recent years. Up to
now, the thymidine kinase of herpes simplex virus (HSVtk) seems to be the most
effective and was the first tested in clinical trials (21,22).
1.3. Suicide-Gene Therapy in Allo-SCT: The Technology

1.3.1. The Vectors
Genetic modification of primary human T-lymphocytes is currently tested for the
treatment of inherited and acquired disease. This is commonly accomplished through
replication-defective retroviral vectors (RV). RV based on the Moloney murine leuke-
mia virus were the first utilized in clinical trials and remain the most effective approach
for introduction of genes into human T-lymphocytes. RV are produced by packaging
Fig. 1. Schematic representation of the suicide gene strategy. Donor T-lymphocytes are har-
vested by leukapheresis, transduced with a retroviral vector carrying a suicide gene (i.e., thymi-
dine kinase of herpes simplex virus [HSVtk]) and a cell-surface marker (i.e., LNGFR, see
text), and immunoselected to homogenous expression. If disease relapse occurs after a T-cell-
depleted bone-marrow transplant, transduced donor lymphocytes are infused into the patient in
order to obtain a graft-vs-leukemia effect. If graft-vs-host disease occurs, a specific prodrug
(i.e., ganciclovir) selectively toxic to the transduced cells is administered to the patient, aiming
at switch-off of the disease.
cell lines that are laboratory factories specialized in producing high titers of replica-
tion-defective RV (23). Practically, gene transfer by RV is usually accomplished by
high-speed centrifugation of the target cells within the packaging cell line supernatant
(spinoculation). It is also possible to preconcentrate the virus particles onto plates coated
with the CH-296 fragment of fibronectin (Retronectin) for higher transduction effi-
ciency (24).
Because RV transduce only proliferating cells, target T-cells are induced to active
cell proliferation through polyclonal stimulation (25,26) before transduction with a
suicide vector. This allows the transduction of a polyclonal T-cell population, poten-
tially representative of all T-cell precursors circulating in healthy donors (27), includ-
ing precursors able to mediate antitumor and antiviral responses (21). However, the
variability in response to activation stimuli, activation-induced cell death (AICD),
exposure to rhIL-2, and proliferation-induced T-cell exhaustion could alter the half-
life, the immune repertoire, and the immune competence of transduced cells (28). RV
transduction often results in an inversion of the physiologic CD4/CD8 ratio, enrich-
ment in memory cells associated with loss of naive T-cells (29), a skew in the T-cell
antigen receptor (TCR) repertoire to oligoclonality as assessed by the analysis of the
V families (30), and a reduced in vivo alloreactivity (31,32). Indeed, the latter modi-
Fig. 2. Schematic map of integrated SFCMM3 proviral genome. Gray boxes indicate long
terminal repeat sequences. White boxes indicate the transgenes. HSVtk codes for the herpes
simplex virus thymidine kinase and LNGFR codes for a truncated form of the low-affinity
receptor for nerve-growth factor. The latter transgene is preceded by the control of the simian
virus 40 (Sv40). Arrows indicate transcription promoters.
fications are attributed more appropriately to ex vivo culture than to the gene transfer
procedure. As a matter of fact, transduction protocols that include optimal T-cell acti-
vation and shorter culture times better preserve the immune competence of transduced
T-cells (29). To overcome alterations of immune competence secondary to in vitro
culture, new cell manipulation protocols are being investigated. Co-stimulation
through CD28 along with TCR triggering appear to reduce the skewing in the V
families of the TCR (30,33) and to partially preserve the anti-EBV immune response
(34). Alternative vectors with reduced cell manipulation requirements, such as
lentiviral vectors, may improve immunocompetence of transduced cells (35).
1.3.2. The Selection Marker
Because even the most efficient RV cannot transduce the totality of proliferating
T-lymphocytes, it is necessary, along with the gene of interest, to insert into the cell a
gene that enables for the selection of genetically modified cells (GMC) to virtually
100% purity (selection marker). The RV vector we are currently using is depicted in
Fig. 2.
There are two types of selection marker genes currently utilized in clinical trials.
The first are intracellular factors, such as the gene coding for neomycin phospho-
transferase resistance (NeoR) and its analog G418 (36). Selection of GMC requires
prolonged culture with a drug that can be toxic, albeit to a lower extent, also to GMC
(31,34). Moreover, resistance marker genes cannot be detected by easy means such as
flow cytometry (37). Newly designed resistance marker genes are expected to solve
many of the limitations of NeoR (38).
An alternative approach relies on the use of a surface marker not expressed by
target cells (26). A gene encoding for the truncated form of the low-affinity cell-sur-
face receptor of human nerve growth factor (LNGFR) was the first surface marker
successfully utilized in clinical suicide-gene therapy trials (39). New chimeric surface
markers such as the fusion gene between HSVtk and the surface molecule CD34
(40,41), or the CD20 molecule, have now been developed. Advantages of these new
markers rely on the double role of marker and suicide genes.
1.3.3. The Suicide Gene
A suicide gene encodes for a protein able to convert a nontoxic prodrug into a toxic
product. Different suicide genes have been investigated and described in recent years.
Fig. 3. Mechanism of toxic activity of the prodrug gancyclovir (GCV) on herpes simplex
thymidine kinase (HSVtk)-expressing cells. Upon entry into the cells, HSVtk metabolizes GCV
to a monophosphate intermediate derivative (pGCV). Cellular kinases further phosphorylate it
to a triphosphate compound (pppGCV), which is the toxic form. pppGCV is then incorporated
into DNA, resulting in efficient inhibition of DNA elongation and causing strand breaks. This
ends up with selective death of HSVtk-expressing cells (27).
Table 1
Overview of the Characteristics of Different Suicide Genes in Allo-SCT
Suicide Prodrug approved
gene Prodrug for clinical use Cell cycle dependence Immunogenicity
HSVtk GCV Yes Yes Yes
Fas AP1903 No No No
CD20 Rituximab Yes No No
HSVtk, Thymidine kinase of herpes simplex virus; GCV, ganciclovir.
Up to now, HSVtk seems to be the most effective (Fig. 3). The HSVtk has been suc-
cessfully transferred into different cell lines to confer ganciclovir (GCV) sensitivity,
and its efficacy has been shown both in vitro and in vivo (36,42). Only recently, a
molecular mechanism for GCV resistance in human HSVtk-expressing T-lymphocytes
was described (43,44).
HSVtk was the first suicide gene to be used in clinical trials of allo-SCT (37,45).
Despite its efficacy and safety, the HSVtk/GCV approach has some limitations (28).
Firstly, HSVtk is a gene of viral origin, which can induce a cytotoxic CD8+ T-cell
response leading to the elimination of GMC (21). Moreover, HSVtk-mediated killing
of transduced lymphocytes is cell-cycle dependent. GMC elimination can be a prob-
lem with slowly proliferating cells. Initially considered as a disadvantage, the latter
phenomenon has been recently proposed as a platform for the separation of GVHD
from the GVL effect by tuning on the kinetics of GCV administration (46,47). Finally,
the use of GCV for the treatment of CMV results in undesired clearance of HSVtk
donor lymphocytes.
To circumvent these limitations (see overview on suicide genes depicted in Table 1),
novel suicide switches were designed. The first proposed was based on the human Fas
receptor to trigger cell death in primary human T-lymphocytes (50,51). Another prom-
ising suicide gene is represented by the CD20 molecule, physiologically expressed on

the surface of B- and not T-lymphocytes (50). Since the humanized chimeric mono-
clonal antibody directed against human CD20 (Rituximab) is able to kill CD20-express-
ing cells (51), CD20 expression on T-lymphocytes can serve both as a suicide gene and
as a selection marker (52).
1.4. Suicide-Gene Therapy in Allo-SCT: Clinical Results

Suicide-gene therapy with GMC after allo-SCT proved to be safe and effective in
phase I clinical trials (22,23,5355.)
The first study performed at HSR in Milan showed that the infusion of HSVtk-
transduced donor lymphocytes to eight patients, who underwent allogeneic transplan-
tation and had subsequent disease relapse or EBV-induced lymphoproliferative
disorders, was effective in five out of eight, with three patients obtaining complete
remission. Two patients developed acute GVHD and were treated with GCV with
clearance of circulating GMC and complete regression of GVHD. In one patient, who
developed chronic GVHD, GCV administration allowed for a reduction of GMC and a
partial clinical response.
Tiberghien et al. at the University Hospital (CHU), Besancon, treated 12 patients
with escalating doses of cells transduced with HSVtk and NeoR, as prophylaxis against
disease relapse after HLA-identical T-cell-depleted marrow transplantation (23). The
presence of circulating GMC was shown in 10 out of 11 patients. The treatment with
GCV resulted in a significant reduction of GMC in all patients and in complete remis-
sion of the GVHD signs in two out of three patients, with acute GVHD of a grade = II
and in one of one patient with chronic GVHD.
In a pilot study conducted by Link et al., in which GMC with HSVtk and NeoR
were administered as DLI after allo-SCT, the percentage of circulating transduced
lymphocytes ranged from 0 to 3.8 %. Three patients responded to the infusion in terms
of antitumoral activity. One patient developed chronic GVHD that did not respond to
conventional therapy but did respond to GCV (53,54).
Champlin et al. conducted a phase I/II study with GMC with HSV-tk for relapse
after allo-SCT. Twenty-three patients were treated with one to four doses per patient
(from 0.7 to 190 106 cells/kg). In almost all the patients, it was possible to document
the presence of circulating GMC, and there was no adverse reaction to the treatment.
Two patients, affected by chronic myelogenous leukemia, had complete remission
with a dose of 50 or 85 106 cells/kg, and four had a stabilization of the disease. Only
one patient developed grade I GVHD, for which no therapy was needed (55).
Different levels of clinical response obtained in different pilot studies are in part
explained by the complexity of the ex vivo manipulation procedures. Published stud-
ies utilized different culture conditions and different degrees of in vitro expansion of
GMC. This has been shown to affect in vivo survival, expansion, and effector poten-
tial of GMC. Despite these difficulties, in all clinical studies, it was possible to kill
GMC in vivo upon administration of GCV and to control GVHD.
The full exploitation of the strategy requires validation of the technology in extended
multicentric trials. This challenging process is now ongoing.
2. Materials
2.1. Preparation of Human Lymphocytes
Retrovirally Transduced to Express a Suicide Gene
2.1.1. Activation and Transduction of Human Lymphocytes
With a Retroviral Vector Carrying the Suicide Gene and the Marker Gene
1. Lymphoprep (Nycomed, Oslo, Norway). Store at 4C.
2. Medium for T-cell culture: RPMI 1640 (BioWhittaker Europe, Verviers, Belgium). For
500 mL final: Add 5 mL of 200 mM glutamine (GibcoGrand Island, NY), 0.5 mL of
100.000 UI/mL penicillin (Pharmacia, Milan, Italy), 0.5 mL of 100.000 UI/mL
streptomicin (Bristol-Myers Sqiubb, Sermoneta, LT), 25 mL autologous serum, 1 mL
of hu-r-IL-2 (Chiron Milan, Italy), 50.000 U/mL. Store at 4C.
3. Phytoemoagglutinine (PHA) (Boehringer Mannheim-Roche GmbH, Mannheim, Ger-
many). Stock solution 1 mg/mL. Store at 20C. Use at the final concentration of 2 g/mL.
4. Anti-CD3 mAb (OKT3) (Orthoclone, Milan, Italy). Stock solution 1 mg/mL. Prepare
working solution at 100 g/mL. Store at 4C. Use at the final concentration of 30 ng/mL.
5. Anti-CD28 mAb (PharMingen, San Diego, CA). Stock solution 1 mg/mL. Store at 4C.
Use at the final concentration of 1 g/mL.
6. Polybrene (Sigma-Aldrich Chemie GmbH, Steinhgeim, Germany). Stock solution at 0.8
mg/mL. Store at 4C.
2.1.2. Cytofluorimetric Analysis and Immune Selection of Transduced Cells

1. mAb 20.4 specific for LNGFR (HB8737) (American Type Culture Collection [ATCC],
Rockville, MD). Store at 4C in the dark.
2. FITC-conjugated goat-antimouse Ab (DAKO, Glostrup, Denmark). Store at 4C in the
dark.
3. Goat antimouse-IgG1 coated magnetic beads (Dynabeads M-450, Dynal AS, Oslo, Nor-
way). Store at 4C in the dark.
2.1.3. Analysis of Transduced and Selected Cells

1. Ganciclovir (Syntex, Palo Alto, CA). Prepare aliquots at 100 M and store at 20C.
2. Radiolabeled thymidine (3H-Thy; Dupont, Boston, MA). Stored at 4C.
3. Methods
3.1. Preparation of Human Lymphocytes
Retrovirally Transduced to Express a Suicide Gene
The preparation of genetically modified cells is based on different steps:
1. Activation and transduction of human lymphocytes with a retroviral vector carrying the
suicide gene and the marker gene.
2. Immune selection of transduced cells. This step exploits the expression of the cell surface
marker gene by genetically modified cells.
3. Analysis of transduced cells in terms of purity and suicide phenotype.
3.1.1. T-Cell Transduction With SFCMM-3 Retroviral Vector

Primary human lymphocytes are transduced with a replication competent retrovirus
(RCR)-free supernatant coding for the suicide gene and a marker. As an example, we
can utilize the SFCMM-3 vector (21), encoding for the suicide gene HSV-tk and for a
truncated form of the low-affinity receptor for nerve growth factor (LNGFR) as a
cell-surface marker (27).
The protocol of T-cell manipulation is the following:
1. Day 0: Primary human lymphocytes harvested from healthy donors are isolated by
leukapheresis and Ficoll-Hypaque gradient separation (Lymphoprep Nycomed, Pharma
AS, Oslo, Norway) and activated in vitro. Different activation signals can be utilized.
2. Phytoemoagglutinine (PHA), 2 g/mL.
3. Anti-CD3 mAb (OKT3), 30 ng/mL
4. Anti-CD3 mAb 30 ng/mL plus anti-CD28 mAb, 1 g/mL.
5. Activated T-cells are cultured at 1 106 cells/mL, in medium for T-cell culture.
6. Day 2. First round of transduction is performed by one cycle of 2 h centrifugation of
T-cells in viral supernatant in the presence of polybrene (8 g/mL) at a multiplicity of
infection (MOI) = 3/1. Cells are plated in medium for T-cell culture at 37C for 18 h.
7. Day 3. Second round of transduction is performed by one cycle of 2 h centrifugation of T-
cells in viral supernatant in the presence of polybrene (8 g/mL) at an MOI = 3/1. Cells
are plated in medium for T-cell culture at 37C.
8. Day 5. Transduction efficiency is determined by FACS analysis. To this purpose, 0.2
106 transduced cells are stained with 100 L of the mouse antihuman LNGFR mAb 20.4
for 30 min at 4C and washed twice in PBS. Cells are then stained with FITC-conjugated
goat-antimouse Ab for 30 min at 4C. Transduced cells stained only with FITC-conju-
gated goat-antimouse Ab are used as negative controls. After two additional washes with
PBS, samples are analyzed on FACSCalibur (Becton Dickinson) using CellQuest soft-
ware (BD Biosciences).
3.1.2. Immuneselection of Transduced Cells
This step exploits the expression of the cell-surface marker gene by genetically
modified cells. During this step, cells are immunoselected to homogeneous transgene
expression.
1. Day 6. Transduced cells are stained with the anti-LNGFR MAb for 30 min at 4C, washed
twice with PBS, and stained with goat-antimouse-IgG1 coated magnetic beads for an
additional 30 min at 4C. After two additional washes with PBS, transduced cells are
selected on magnetic beads according to manufacturers instructions. Cells are plated in
medium for T-cell culture.
2. Day 7. Beads are eliminated from selected cells by exposure to the magnet.
3.1.3. Analysis of Transduced Cells
in Terms of Purity and Suicide Phenotype
During this step, the efficiency of the overall procedure is analyzed: this requires
the analysis of purity after T-cell selection and the analysis of ganciclovir sensitivity
of transduced/selected cells.
1. Day 8. Selection efficiency is analyzed by FACS as described in Subheading 3.1.1.,
step 8 (method utilized at d 5). A purity of at least 90% is usually obtained and allows
us to perform further experiments in the absence of an additional step of cell immune
selection.
2. Within the following wk, GCV sensitivity of transduced and selected cells is analyzed.
Since GCV-mediated killing requires cell proliferation, cells are restimulated in vitro
following the same conditions described on d 0 in the presence of increasing GCV con-
centrations (0.01 to 10 M), in 96-well plates at 0.1 106 cells/well in triplicates. After
4 d, radiolabeled thymidine (1 Ci/well; specific activity 87 Ci/mmol) is added and DNA
is harvested after an additional 16 h. Cpm are counted in a beta scintillation counter
(Wallac 1205 BetaPlate). The effect of GCV on transduced lymphocytes is expressed as
percentage of growth inhibition or as relative proliferation, referring to proliferation in
the absence of ganciclovir. A GCV concentration of 0.1 M induces more than 90%
growth inhibition in T-cells transduced with SFCMM-3.
References
1.
1 Horowitz, M. M., Gale, R. P., Sondel, P. M., et al. (1990) Graft-versus-leukemia reactions
after bone marrow transplantation. Blood 75, 555562.
2.
2 Kolb, H. J., Mittermuller, J., Clemm, C., et al. (1990) Donor leukocyte transfusions for
76, 24622465.
3.
3 Kolb, H. J., Schattenberg, A., Goldman, J. M., et al. (1995) Graft-versus-leukemia effect
4.
4 Childs, R., Chernoff, A., Contentin, N., et al. (2000) Regression of metastatic renal-cell
carcinoma after nonmyeloablative allogeneic peripheral-blood stem-cell transplantation.
N. Engl. J. Med. 343, 750758.
5.
5 Drobyski, W. R., Keever, C. A., Roth, M. S., et al. (1993) Salvage immunotherapy using
donor leukocyte infusions as treatment for relapsed chronic myelogenous leukemia after
allogeneic bone marrow transplantation: efficacy and toxicity of a defined T-cell dose.
Blood 82, 23102318.
6.
6 Szer, J., Grigg, A. P., Phillips, G. L., and Sheridan, W. P. (1993) Donor leucocyte infu-
sions after chemotherapy for patients relapsing with acute leukaemia following allogeneic
BMT. Bone Marrow Transplant. 11, 109111.
7.
7 Porter, D. L., Roth, M. S., McGarigle, C., Ferrara, J. L., and Antin, J. H. (1994) Induction
of graft-versus-host disease as immunotherapy for relapsed chronic myeloid leukemia. N.
Engl. J. Med. 330, 100106.
8.
8 van Rhee, F., Lin, F., Cullis, J. O., et al. (1994) Relapse of chronic myeloid leukemia
after allogeneic bone marrow transplant: the case for giving donor leukocyte transfu-
sions before the onset of hematologic relapse. Blood 83, 33773383.
9.
9 Slavin, S., Naparstek, E., Nagler, A., et al. (1996) Allogeneic cell therapy with donor
peripheral blood cells and recombinant human interleukin-2 to treat leukemia relapse after
allogeneic bone marrow transplantation. Blood 87, 21952204.
10.
10 Tricot, G., Vesole, D. H., Jagannath, S., Hilton, J., Munshi, N., and Barlogie, B. (1996)
Graft-versus-myeloma effect: proof of principle. Blood 87, 11961198.
11.
11 Collins, R. H., Jr., Shpilberg, O., Drobyski, W. R., et al. (1997) Donor leukocyte infusions
in 140 patients with relapsed malignancy after allogeneic bone marrow transplantation. J.
Clin. Oncol. 15, 433444.
12.
12 Mackinnon, S., Papadopoulos, E. B., Carabasi, M. H., et al. (1995) Adoptive immuno-
therapy 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, 12611268.
13.
13 Guglielmi, C., Arcese, W., Dazzi, F., et al. (2002) Donor lymphocyte infusion for relapsed
chronic myelogenous leukemia: prognostic relevance of the initial cell dose. Blood 100,
397405.
14.
14 Riddell, S. R., Watanabe, K. S., Goodrich, J. M., Li, C. R., Agha, M. E., and Greenberg, P.
15.
15 Rooney, C. M., Smith, C. A., Ng, C. Y., et al. (1995) Use of gene-modified virus-specific T
lymphocytes to control Epstein-Barr-virus-related lymphoproliferation. Lancet 345, 913.
16 Yee, C., Thompson, J. A., Byrd, D., et al. (2002) Adoptive T cell therapy using antigen-
16.
specific CD8+ T cell clones for the treatment of patients with metastatic melanoma: in
vivo persistence, migration, and antitumor effect of transferred T cells. Proc. Natl. Acad.
Sci. USA 99, 16,16816,173.
17.
17 Papadopoulos, E. B., Ladanyi, M., Emanuel, D., et al. (1994) Infusions of donor leuko-
cytes to treat Epstein-Barr virus-associated lymphoproliferative disorders after allogeneic
bone marrow transplantation. N. Engl. J. Med. 330, 11851191.
18.
18 Molldrem, J. J., Lee, P. P., Wang, C., et al. (2000) Evidence that specific T lymphocytes may
participate in the elimination of chronic myelogenous leukemia. Nat. Med. 6, 10181023.
19.
19 Bellantuono, I., Gao, L., Parry, S., et al. (2002) Two distinct HLA-A0201-presented
epitopes of the Wilms tumor antigen 1 can function as targets for leukemia-reactive CTL.
Blood 100, 38353837.
20.
20 Marijt, W. A., Heemskerk, M. H., Kloosterboer, F. M., et al. (2003) Hematopoiesis-
restricted minor histocompatibility antigens HA-1- or HA-2-specific T cells can induce
complete remissions of relapsed leukemia. Proc. Natl. Acad. Sci, USA 100, 27422747.
21.
21 Bonini, C., Ferrari, G., Verzeletti, S., et al. (1997) HSV-TK gene transfer into donor lym-
phocytes for control of allogeneic graft-versus-leukemia. Science 276, 17191724.
22.
22 Tiberghien, P., Ferrand, C., Lioure, B., et al. (2001) Administration of herpes simplex-
thymidine kinase-expressing donor T cells with a T-cell-depleted allogeneic marrow graft.
Blood 97, 6372.
23.
23 Miller, A. D. and Buttimore, C. (1986) Redesign of retrovirus packaging cell lines to avoid
recombination leading to helper virus production. Mol. Cell Biol. 6, 28952902.
24.
24 Lamers, C. H., Willemsen, R. A., Luider, B. A., Debets, R., and Bolhuis, R. L. (2002)
Protocol for gene transduction and expansion of human T lymphocytes for clinical
immunogene therapy of cancer. Cancer Gene Ther. 9, 613623.
25.
25 Culver, K. W., Anderson, W. F., and Blaese, R. M. (1991) Lymphocyte gene therapy.
Hum. Gene Ther. 2, 107109.
26.
26 Mavilio, F., Ferrari, G., Rossini, S., et al. (1994) Peripheral blood lymphocytes as target
cells of retroviral vector-mediated gene transfer. Blood 83, 19881997.
27.
27 Verzeletti, S., Bonini, C., Marktel, S., et al. (1998) Herpes simplex virus thymidine kinase
gene transfer for controlled graft-versus-host disease and graft-versus-leukemia: clinical
follow-up and improved new vectors. Hum. Gene Ther. 9, 22432251.
28.
28 Bonini, C. and Bordignon, C. (1997) Potential and limitations of HSV-TK-transduced
donor peripheral blood lymphocytes after allo-BMT. Hematol. Cell Ther. 39, 273274.
29 Marktel, S., Magnani, Z., Ciceri, F., et al. (2003) Immunologic potential of donor lympho-
29.
cytes expressing a suicide gene for early immune reconstitution after hematopoietic T-
cell-depleted stem cell transplantation. Blood 101, 12901298.
30 Ferrand, C., Robinet, E., Contassot, E., et al. (2000) Retrovirus-mediated gene transfer in
30.
primary T lymphocytes: influence of the transduction/selection process and of ex vivo
expansion on the T cell receptor beta chain hypervariable region repertoire. Hum. Gene
Ther. 11, 11511164.
31.
31 Contassot, E., Murphy, W., Angonin, R., et al. (1998) In vivo alloreactive potential of ex
vivo-expanded primary T lymphocytes. Transplantation 65, 13651370.
32.
32 Drobyski, W. R., Majewski, D., Ozker, K., and Hanson, G. (1998) Ex vivo anti-CD3 anti-
body-activated donor T cells have a reduced ability to cause lethal murine graft-versus-
host disease but retain their ability to facilitate alloengraftment. J. Immunol. 161,
26102619.
33.
33 Berger, C., Blau, C. A., Clackson, T., Riddell, S. R., and Heimfeld, S. (2003) CD28
costimulation and immunoaffinity-based selection efficiently generate primary gene-modi-
fied T cells for adoptive immunotherapy. Blood 101, 476484.
34.
34 Sauce, D., Tonnelier, N., Duperrier, A., et al. (2002) Influence of ex vivo expansion and
retrovirus-mediated gene transfer on primary T lymphocyte phenotype and functions. J.
Hematother. Stem Cell Res. 11, 929940.
35.
35 Cavalieri, S., Cazzaniga, S., Geuna, M., et al. (2003) Human T lymphocytes transduced by
lentiviral vectors in the absence of TCR activation maintain an intact immune compe-
tence. Blood 102, 497505.
36.
36 Tiberghien, P., Reynolds, C. W., Keller, J., 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, 13331341.
37 Tiberghien, P., Cahn, J. Y., Brion, A., et al. (1997) Use of donor T-lymphocytes express-
37.
ing herpes-simplex thymidine kinase in allogeneic bone marrow transplantation: a phase
I-II study. Hum. Gene Ther. 8, 615624.
38.
38 Aints, A., Belusa, R., Andersson, R. M., Guven, H., and Dilber, M. S. (2002) Enhanced
ouabain resistance gene as a eukaryotic selection marker. Hum. Gene Ther. 13, 969977.
39.
39 Bonini, C., Grez, M., Traversari, C., et al. (2003) Safety of retroviral gene marking with a
truncated NGF receptor. Nat. Med. 9, 367369.
40.
40 Fehse, B., Kustikova, O. S., Li, Z., et al. (2002) A novel sort-suicide fusion gene vector
for T cell manipulation. Gene Ther. 9, 16331638.
41.
41 Rettig, M. P., Ritchey, J. K., Meyerrose, T. E., Haug, J. S., and DiPersio, J. F. (2003)
Transduction and selection of human T cells with novel CD34/thymidine kinase chimeric
suicide genes for the treatment of graft-versus-host disease. Mol. Ther. 8, 2941.
42.
42 Moolten, F. L., Wells, J. M., Heyman, R. A., and Evans, R. M. (1990) Lymphoma regres-
sion induced by ganciclovir in mice bearing a herpes thymidine kinase transgene. Hum.
Gene Ther. 1, 125134.
43.
43 Garin, M. I., Garrett, E., Tiberghien, P., et al. (2001) Molecular mechanism for ganciclovir
resistance in human T lymphocytes transduced with retroviral vectors carrying the herpes
simplex virus thymidine kinase gene. Blood 97, 122129.
44.
44 Chalmers, D., Ferrand, C., Apperley, J. F., et al. (2001) Elimination of the truncated mes-
sage from the herpes simplex virus thymidine kinase suicide gene. Mol. Ther. 4, 146148.
45.
45 Bordignon, C., Bonini, C., Verzeletti, S., et al. (1995) Transfer of the HSV-tk gene into
donor peripheral blood lymphocytes for in vivo modulation of donor anti-tumor immunity
after allogeneic bone marrow transplantation. Hum. Gene Ther. 6, 813819.
46.
46 Cohen, J. L., Saron, M. F., Boyer, O., et al. (2000) Preservation of graft-versus-infection
effects after suicide gene therapy for prevention of graft-versus-host disease. Hum. Gene
Ther. 11, 24732481.
47.
47 Litvinova, E., Maury, S., Boyer, O., et al. (2002) Graft-versus-leukemia effect after sui-
cide-gene-mediated control of graft-versus-host disease. Blood 100, 20202025.
48.
48 Ashkenazi, A., and Dixit, V. M. (1998) Death receptors: signaling and modulation. Sci-
ence 281, 13051308.
49.
49 Thomis, D. C., Marktel, S., Bonini, C., et al. (2001) A Fas-based suicide switch in human
T cells for the treatment of graft-versus-host disease. Blood 97, 12491257.
50.
50 Introna, M., Barbui, A. M., Bambacioni, F., et al. (2000) Genetic modification of human T
cells with CD20: a strategy to purify and lyse transduced cells with anti-CD20 antibodies.
Hum. Gene Ther. 11, 611620.
51. Cerny, T., Borisch, B., Introna, M., Johnson, P., and Rose, A. L. (2002) Mechanism of
action of rituximab. Anticancer Drugs 13 Suppl 2, S310.
52.
52 Di Gaetano, N., Cittera, E., Nota, R., et al. (2003) Complement activation determines the
therapeutic activity of rituximab in vivo. J. Immunol. 171, 15811587.
53.
53 Link, C. J., Burt, R. K., Traynor, A. E., et al. (1998) Adoptive immunotherapy for leuke-
mia: donor lymphocytes transduced with the herpes simplex thymidine kinase gene for
remission induction. Hum. Gene Ther. 9, 115.
54. Link, C. J., Drobyski, W. R., Traynor, A. E., et al. (1999) Adoptive immunotherapy for
leukemia: donor lymphocytes transduced with the Herpes Simplex thymidine kinase
(HStk) gene. Blood 94, 1631a.
55. Champlin, R., Besinger, W., Henslee-Downey, J., K. et al. (1999) Phase I/II study of thy-
midine kinase (TK)-transduced donor lymphocyte infusions (DLI) in patients with hema-
tologic malignanicies. Blood 94, 1448a.
Index 487
Index
A flow cytometry characterization,

Adenoviral transduction, dendritic cells, 364366, 368
efficiency, 83, 84 production,
infection of cells, 9294 cell culture in roller bottles, 361,
limitations, 84 362
materials, 84, 85 chromatography, 362, 368
monocyte-derived dendritic cell materials, 352
preparation, 91, 92 quantification, 362
recombinant virus production, storage, 363, 368
cesium chloride gradient sulfur-35 labeling and
purification, 87, 88 immunoprecipitation, 360, 361,
multiplication, 86, 87 368
overview, 85, 86 transfection into expression systems,
replication-competent adenovirus 351, 358, 359, 367
screening, 90, 91 Antibody-modified liposomes,
titer analysis, 88, 89 functionalization, 389, 390
virus particle concentration overview, 389, 390
determination, 89 synthesis,
vectors, 83, 84, 93 coupling of ScFv antibody
Allorestriction, see Cytotoxic T- fragments, 396, 397, 400
lymphocyte liposome preparation,
Antibodycytokine fusion proteins, calculations of numbers of
bioactivity assays, 365367, 369 liposomes and reactive
construction, groups, 392, 393, 399
cloning of V-region sequence into dialysis, 393, 395, 399
expression vectors, 353356 extrusion, 395, 399
interleukin-2 fusion protein fluorescence labeling, 395
construction, 356, 357, 367 lipid compositions, 391, 392
materials, 350, 351 maleimide group introduction,
engineering principles, 347350 395, 399, 400
enzyme-linked immunosorbent materials, 390, 391, 398, 399
assay, 351, 353, 359, 360, 364, purification from nonreacted
367, 368 antibody, 397, 398
487
488 Index
tumor cell binding studies in vitro, humanized antibodies, 304,

398, 400 305
Antibody therapy, overview, 304
bifunctional antibodies, see target versatility, 303
Antibodycytokine fusion tolerance induction, 306308
proteins; Bifunctional toxicity, 303
antibodies; Paclitaxel rationale in autoimmunity and
immunoconjugate transplantation, 297, 298
bispecific antibodies, see Bispecific Atherosclerosis, intravenous
antibodies immunoglobulin therapy, 403,
cytokine fusion proteins, see 404
Antibodycytokine fusion
proteins B
immunoglobulin structure, 347, 348 B-cell, dendritic cell activation, 5, 71
intravenous immunoglobulin, see Bifunctional antibodies, see also
Intravenous immunoglobulin Antibodycytokine fusion
monoclonal antibodies, proteins; Paclitaxel
advantages and limitations, 298, immunoconjugate,
299 applications, 330
applications, definition, 330
cellular depletion, 300, 301 generation,
cytokine neutralization, 299 enzyme-linked immunosorbent
functional inhibition, 301, 302 assay, 340342
immunoregulatory T-cell materials, 331
triggering, 302 polymerase chain reaction, 334,
clinical targets, 306, 307 335
clinical trials, protein extraction, 339, 340, 342
CD3 antibodies, 310, 311 refolding of inclusion bodies,
CD4 antibodies, 311, 312 340342
CD20 antibodies, 316 ScFv cloning,
CD25 antibodies, 311 biotin mimic tag engineering,
CD52 antibodies, 314, 315 336, 337, 342
LFA-3/IgG1 fusion protein gene assembly with linker, 335
antibodies, 316 ligation and transformation,
tumor necrosis factor 338, 342
antibodies, 312314 RNA isolation and cDNA
VLA-4 antibodies, 315, 316 synthesis, 335, 342
efficacy in advanced disease, screening and transformation,
308310 338, 342
large-scale production, 303, 304 sequencing, 339
prospects, 316, 317 Western blot, 338, 339
sensitization minimization, Bispecific antibodies,
human antibodies, 305, 306 applications, 330
Index 489
definition, 330 materials, 410

generation, overview, 409, 410
affinity chromatography, 334, recovery of labeled donor cells, 413,
341, 342 418
enzyme-linked immunosorbent CD3, monoclonal antibody clinical
assay, 334 trials, 310, 311
materials, 330, 331 CD4, monoclonal antibody clinical
quadroma development, trials, 311, 312
cell labeling, 332, 341 CD4+ T-cell, see T-helper cell
fluorescence-activated cell CD4+/CD25+ T-cell, see Regulatory T-
sorting, 332, 333, 341 cell
polyethylene glycol fusion, CD8+ T-cell, see Cytotoxic T-
332, 341 lymphocyte
screening, 333, 334, 341 CD14+ monocytes, see also Monocyte-
Bone marrow transplantation, see derived dendritic cell,
Chimerism; Graft-vs-host immunomagnetic separation, see
disease; Graft-vs-leukemia CliniMACS system, dendritic
effect; Hematopoietic stem cell cell isolation
transplantation monocyte-derived dendritic cell
culture, 57, 61, 62, 71, 115, 116
C CD20, monoclonal antibody clinical
Cancer vaccines, see Dendritic cell; trials, 316
Monocyte-derived dendritic CD25, monoclonal antibody clinical
cell trials, 311
Carboxyfluorescein succinimidyl ester CD52, monoclonal antibody clinical
(CFSE), T-cell costimulatory trials, 314, 315
blockade analysis, CFSE, see Carboxyfluorescein
adoptive transfer of labeled cells into succinimidyl ester
allogeneic hosts, Chimerism,
blockade reagent treatment, 412 graft-vs-leukemia effect potentiation
cell transfer, 412, 417 after bone marrow
irradiation, 412 transplantation,
cell staining after recovery, animal care, 471, 473
apoptosis staining, 413 bone marrow harvesting and
cytokine staining, 413 transplantation, 471, 473
surface staining, 413 donor lymphocyte infusion
donor lymphocyte labeling, preparation, 473
cell division effects, 411 leukemia cell line culture, 473
labeling conditions, 412, 415, 417 materials, 470, 471
lymphocyte preparation, 411, 412 natural killer cell depletion, 472
flow cytometry, overview, 469, 470
analysis, 415 phenotyping of chimeras, 472,
data acquisition, 414, 418 473
490 Index
T-cell depletion, 471473 materials, 58, 69

hematopoietic stem cell plasmacytoid dendritic cells,
transplantation and chimerism automated separation, 63, 64
induction in nonmyeloablated flow cytometry analysis, 64, 65
mice, immunomagnetic labeling, 63, 69
flow cytometry identification of materials, 57, 58, 69
stem cells, 460, 465467 CTL, see Cytotoxic T lymphocyte
immunomagnetic purification of Cytokine secretion assay, see T-helper
stem cells, cell
cell sorting, 463, 464 Cytotoxic T lymphocyte (CTL),
harvesting, 462, 563 allo-restricted high-avidity tumor-
magnetic bead labeling, 463 specific cell generation,
injection of stem cells, 464, 465 cytotoxicity assay, 225, 226
materials, 460, 461 expansion, 224227
overview, 459, 460 limiting dilution, 221, 222
preconditioning of recipient mice, maintenance, 226, 227
461, 462 materials, 216, 218, 226
short-term culture of stem cells, overview, 215217
465 peptides,
CliniMACS system, dendritic cell loading of stimulator cells,
isolation, 220, 227
CD14+ monocyte isolation, preparation, 220
automated separation, 59 screening, 223, 224, 227
flow cytometry analysis, 5961, stimulation with HLA-A2+
69 antigen-presenting cells,
immunomagnetic labeling, 59, 69 peripheral blood mononuclear
materials, 56 cell preparation, 219, 220
monocyte-derived dendritic cell stimulation and re-stimulation
culture from precursors, 57, 61, conditions, 220222, 227
62, 115, 116 stimulator cell culture, 218,
monocyte-derived dendritic cell 219, 226, 227
generation, dendritic cell activation, 5
CD14+ monocyte enrichment, mathematical modeling of regulation
7476, 80 by dendritic cells in
flow cytometry, 76 immunotherapy,
leukapheresis, 74, 80 assumptions, 20, 23, 24
materials, 72, 73 computer simulation, 26, 30
migration assay, 76, 79, 80 confidence interval computation,
mixed leukocyte reaction, 79 28, 31, 32
myeloid dendritic cells, data sets, 21, 22
automated separation, 66, 67, 69 equation derivation, 24, 26, 29
flow cytometry analysis, 6769 least-squares fitting, 2628
immunomagnetic labeling, 66, 67, model selection, 23
69 overview, 1921
Index 491
parameters, 24, 25 cloning, 180, 181, 183

parsimony evaluation, 28, 29 cytotoxicity assays,
predator-prey type induction/ expanded clones, 182
regulation of cytotoxic T screening assay, 181, 183
lymphocytes by dendritic cells, dendritic cell generation from
23 monocytes, 178, 179, 182
sensitivity analysis, 29 expansion, 182, 183
software, 21 human leukocyte antigen peptide
memory/effector subset multimer staining, 180, 183
characterization in melanoma, materials, 176, 178
five-cell cDNA amplification, overview, 175, 176
269, 270, 272, 273, 280 stimulations with peptide-pulsed
flow cytometry, 270 dendritic cells, 179, 180
materials, 268, 269
overview, 266268 D
rationale, 265, 266 DC, see Dendritic cell
reverse transcriptase-polymerase Dendritic cell (DC), see also Monocyte-
chain reaction, 272, 273, 280 derived dendritic cell,
telomere length assay, antigen loading,
fibroblast nuclei, 276278 apoptotic tumor cells, 10
flow cytometry, 276, 281 immune complexes, 12
fluorescence in situ microspheres, see Poly(lactide-
hybridization, 275, 276, co-glycolide) microspheres
280, 281 peptides, 8, 9
proteomics analysis, proteins, 9
database analysis, 169, 171, 172 RNA transfection, see RNA
denaturing gel electrophoresis, transfection, dendritic cells,
162, 163 transduction, see Adenoviral
image analysis, 165, 166, 168, 171 transduction, dendritic cells
isoelectric focusing, 161, 162, 171 transfection and transduction, 9,
mass spectrometry, 169 10, 47, 48
materials, 157160 antigen presentation, 1, 4
sample preparation, 161, 169171 antigen uptake and maturation, 24
staining of gels, graft-vs-leukemia effect mediation,
Coomassie stain, 164, 165 43
silver stain, 163 immunotherapy applications, see
tracking with major also Multiple myeloma,
histocompatibility complex cancer vaccines, 5, 6, 10, 35, 55
multimers, see Major immunity induction, 120, 121
histocompatibility complex tolerance induction, 120
multimer lymphocyte activation, 5
tumor antigen response, 137 mathematical modeling of T-cell
tumor-reactive clone isolation and activation for immunotherapy
expansion for immunotherapy, optimization,
492 Index
assumptions, 20, 23, 24 bispecific antibody characterization,

computer simulation, 26, 30 334
confidence interval computation,
28, 31, 32 F
data sets, 21, 22 Fibrosis, intravenous immunoglobulin
equation derivation, 24, 26, 29 therapy, 404, 405
least-squares fitting, 2628 FISH, see Fluorescence in situ
model selection, 23 hybridization
overview, 1921 Flow cytomery,
parameters, 24, 25 antibodycytokine fusion protein
parsimony evaluation, 28, 29 characterization, 364366, 368
predator-prey type induction/ CD14+ monocytes, 5961, 69
regulation of cytotoxic T dendritic cell transfection assay, 52,
lymphocytes by dendritic cells, 53
23 hematopoietic stem cell
sensitivity analysis, 29 identification, 460, 465467
software, 21 major histocompatibility complex
maturation, ex vivo, 8 multimers, see Major
progenitor types, 55 histocompatibility complex
sources for ex vivo manipulation, multimer
hematopoietic progenitor cells, 7, memory/effector T-cell subset
97 characterization in melanoma,
magnetic separation, see 270
CliniMACS system, dendritic monocyte-derived dendritic cell
cell isolation uptake of microspheres, 41, 42
monocytes, 6, 7 monocyte-derived dendritic cells, 76
peripheral blood cell enrichment, myeloid dendritic cells, 6769
7, 8, 50, 53, 97, 98 plasmacytoid dendritic cells, 64, 65
preparation selection, 8 T-cell costimulatory blockade
subtypes, 2 analysis with
carboxyfluorescein
E succinimidyl ester,
Effector T-cell, see Cytotoxic T- analysis, 415
lymphocyte, data acquisition, 414, 418
Electroporation, RNA transfection in telomere length assay with
dendritic cells, 52, 53 fluorescence in situ
ELISA, see Enzyme-linked hybridization, 276, 281
immunosorbent assay T-helper cell isolation from mice
Enzyme-linked immunosorbent assay immunized with irradiated
(ELISA), tumor cells, 259263
antibodycytokine fusion protein, 351, Fluorescence in situ hybridization
353, 359, 360, 364, 367, 368 (FISH), telomere length assay
bifunctional antibody with flow cytometry, 275, 276,
characterization, 340342 280, 281
Index 493
G antigen-presenting cells, 43
Graft-vs-host disease (GVHD), natural killer cells, 430
leukemia remission association, 422 T-cells, 429, 430
radiation protocol effects in animal effector mechanisms, 432
models, tumor antigens as self antigens,
donor cell collection, 431, 432
bone marrow cells, 448, 449, potentiation by chimerism induction
455 after bone marrow
lymph node cells, 448, 454 transplantation,
evaluation, 450, 455 animal care, 471, 473
injection of donor cells, 449 bone marrow harvesting and
irradiation of mice, transplantation, 471, 473
anesthesia, 447 donor lymphocyte infusion
cage design, 447, 448, 450, preparation, 473
451 leukemia cell line culture, 473
radiation devices, 448, materials, 470, 471
451454 natural killer cell depletion, 472
materials, 446, 447 overview, 469, 470
overview, 445, 446 phenotyping of chimeras, 472,
solid tumor effects, 427, 428 473
Graft-vs-leukemia effect (GVL), T-cell depletion, 471473
clinical evidence, prospects, 432, 433
adoptive immunotherapy with rationale for therapeutic use, 421,
donor lymphocyte infusions, 422
423, 424 suicide gene therapy for early
remissions with graft-vs-host induction after bone marrow
disease, 422, 423 transplantation,
syngeneic vs HLA-identical clinical outcomes, 480
sibling donor outcomes, 423 herpes simplex thymidine kinase
T-cell depleted graft outcomes, system, 478480
423 immunoselection of transduced
disease specificity, cells, 482
acute lymphocytic leukemia, 425 materials, 481
acute myeloid leukemia, 425 overview, 475, 476
chronic lymphocytic leukemia, retroviral vector transduction of
426 T-cells, 481, 482
chronic myelogenous leukemia, transduced cell analysis, 482,
424, 425 483
Hodgkins disease, 426, 427 vectors, 476478
myeloma, 426 GVHD, see Graft-vs-host disease
solid tumors, 427, 428 GVL, see Graft-vs-leukemia effect
mechanisms,
H
antigenic targets, 431
effector cells, HD, see Hodgkins disease
494 Index
Hematopoietic stem cell Immunomagnetic separation, see

transplantation, see also Graft- CliniMACS system; Magnetic-
vs-host disease; Graft-vs- activated cell sorting
leukemia effect, Interleukin-2antibody fusion protein,
chimerism and tolerance induction in see Antibodycytokine fusion
nonmyeloablated mice, proteins
flow cytometry identification of Intravenous immunoglobulin (IVIG),
stem cells, 460, 465467 antimetastatic effects, 405407
immunomagnetic purification of atherosclerosis studies, 403, 404
stem cells, fibrosis studies, 404, 405
cell sorting, 463, 464 IVIG, see Intravenous immunoglobulin
harvesting, 462, 563
magnetic bead labeling, 463 L
injection of stem cells, 464, 465 Leukemia, see Graft-vs-leukemia effect
materials, 460, 461 LFA-3/IgG1 fusion protein, monoclonal
overview, 459, 460 antibody clinical trials, 316
preconditioning of recipient mice, Ligand-mediated liposome targeting,
461, 462 see Antibody-modified
short-term culture of stem cells, liposomes
465 Liposome, see Antibody-modified
graft-vs-leukemia effect potentiation liposomes
by chimerism induction after
bone marrow transplantation, M
animal care, 471, 473 MACS, see Magnetic-activated cell
bone marrow harvesting and sorting
transplantation, 471, 473 Magnetic-activated cell sorting
donor lymphocyte infusion (MACS), see also CliniMACS
preparation, 473 system,
leukemia cell line culture, 473 hematopoietic stem cells,
materials, 470, 471 cell sorting, 463, 464
natural killer cell depletion, 472 harvesting, 462, 563
overview, 469, 470 magnetic bead labeling, 463
phenotyping of chimeras, 472, 473 regulatory T-cells,
T-cell depletion, 471473 human cells, 289, 294
Hodgkins disease (HD), graft-vs- mouse cells, 291, 293, 295
lymphoma effect, 426, 427 Major histocompatibility complex
(MHC) multimer,
I class I vs class II multimers, 186
Immunocytokines, see Antibody rationale and applications, 186
cytokine fusion proteins tracking tumor-specific cytotoxic T-
Immunoglobulins, see Antibody lymphocytes after adoptive
therapy; Intravenous transfer,
immunoglobulin annexin-V staining, 195
Index 495
antigen-specific cell generation, MHC multimer, see Major

189 histocompatibility complex
competition between multimer multimer
and anti-CD3 monoclonal Microspheres, see Poly(lactide-co-
antibody binding, 190 glycolide) microspheres
cytokine secretion assay, 195197 Mixed leukocyte reaction,
flow cytometry gating, 190 monocyte-derived dendritic cells, 79,
materials, 187, 188 118
Melan-A-specific cell tracking, regulatory T-cell assay, 290
190, 191 MM, see Multiple myeloma
staining of multimers, 189, 190, MoDC, see Monocyte-derived dendritic
197, 199 cell
Mass spectrometry (MS), Monoclonal antibody, see Antibody
paclitaxel immunoconjugate, 381 therapy; Bispecific antibodies
proteomics analysis of T-cells, 169 Monocyte-derived dendritic cell
Mathematical modeling, cytotoxic T (MoDC),
lymphocyte regulation by adenoviral transduction, see
dendritic cells in Adenoviral transduction,
immunotherapy, see Cytotoxic dendritic cells
T lymphocyte autologous antigen-pulsed cell
Melanoma, generation for patient-specific
Melan-A-specific T-cell tracking immunotherapy,
using major histocompatibility dendritic cell processing,
complex multimers, 190, 191 103106, 108110
memory/effector T-cell subset heat-inactivated plasma
characterization in melanoma, preparation, 102, 103
five-cell cDNA amplification, injection preparation, 107
269, 270, 272, 273, 280 materials, 99101, 107, 108
flow cytometry, 270 overview, 9799
materials, 268, 269 peripheral blood mononuclear cell
overview, 266268 preparation, 101, 102, 108
rationale, 265, 266 quality control, 107, 110
reverse transcriptase-polymerase culture from CD14+ precursors, 57,
chain reaction, 272, 273, 280 61, 62, 71, 115, 116
telomere length assay, immunomagnetic separation, see
fibroblast nuclei, 276278 CliniMACS system, dendritic
flow cytomery, 276, 281 cell isolation
fluorescence in situ migration assay, 76, 79, 80
hybridization, 275, 276, mixed leukocyte reaction, 79
280, 281 quality control, see Quality control,
tumor-associated antigens, 186, 266 clinical-grade dendritic cells
Memory T-cell, see Cytotoxic T- MS, see Mass spectrometry
lymphocyte Multiple myeloma (MM),
496 Index
chemotherapy, 217 PLGA MS, see Poly(lactide-co-

dendritic cell therapy, glycolide) microspheres
idiotype pulsing, 128132 Poly(lactide-co-glycolide) microspheres
prospects, 132 (PLGA MS),
tumor-associated antigens, 128 antigen microencapsulation by
graft-vs-leukemia effect, 426 spray-drying, 37, 38, 41, 44
idiotype vaccination, 128, 129 monocyte-derived dendritic cell
Myeloid dendritic cells, uptake,
immunomagnetic separation, antigen presentation assays,
see CliniMACS system, class I antigens, 43
dendritic cell isolation class II antigens, 43
materials, 40
N overview, 38
Natural killer (NK) cell, evaluation,
depletion in mice, 472 flow cytometry assay, 41, 42
graft-vs-leukemia effect mediation, overview, 37
430 prospects, 44
NK cell, see Natural killer cell quality control,
materials, 39, 40, 44
P migration capacity of cells, 42
Paclitaxel immunoconjugate, overview, 37
characterization, phenotype and maturation of
apoptosis assay, 385, 386 cells, 42
cytotoxicity assay, 385 survival of cells, 42
epidermal growth factor receptor T-cell stimulatory functions,
kinase activity assessment by 43
Western blot, 383, 384, 386, polymer characteristics, 36
387 Polymerase chain reaction (PCR),
tumor cell line maintenance, 381, antibodycytokine fusion protein
386 construction, 356, 357, 367
design, 375377 bifunctional antibody development,
rationale for generation, 375, 376 334, 335
synthesis, memory/effector T-cell subset
conjugation reaction, 379381, characterization in melanoma,
386 272, 273, 280
mass spectrometry, 381 murine T-cell receptor isolation and
materials, 376, 378, 379, 386 cloning,
2'-O-paclitaxel succinate RACE polymerase chain reaction,
preparation, 379, 386 236, 237
PCR, see Polymerase chain reaction reverse transcriptase-polymerase
Plasmacytoid dendritic cells, chain reaction, 238, 239
immunomagnetic separation, Postttransplantation
see CliniMACS system, lymphoproliferative disease
dendritic cell isolation (PTLD), adoptive
Index 497
immunotherapy, 428 T-helper cell isolation, 287,

Proteomics, 294
approaches, 156 materials, 286, 287
T-cell studies, see Cytotoxic T- mouse cells,
lymphocyte; Regulatory T-cell depletion, 291, 293, 295
PTLD, see Postttransplantation functional analysis, 294
lymphoproliferative disease immunomagnetic sorting, 291,
293, 295
Q preactivation, 291, 293
Quadroma, see Bispecific antibodies proteomics analysis,
Quality control, clinical-grade dendritic database analysis, 169, 171, 172
cells, denaturing gel electrophoresis,
autologous cells for patient-specific 162, 163
immunotherapy, 107, 110 image analysis, 165, 166, 168, 171
clinical trial criteria, 116 isoelectric focusing, 161, 162, 171
functional characterization, mass spectrometry, 169
antigen uptake and T-cell materials, 157160
activation, 119 sample preparation, 161, 169171
chemokine production, 120 staining of gels,
migration, 119, 120 Coomassie stain, 164, 165
monocyte-derived dendritic cell silver stain, 163
generation under good Retrovirus, see Suicide gene therapy; T-
manufacturing practice cell receptor
conditions, 114, 115 RNA transfection, dendritic cells,
morphology, phenotype, and dendritic cell generation from
function, 118 peripheral blood mononuclear
sterility/pyrogenicity, viability and cells, 50, 53
purity, 116, 117 efficiency, 47, 48
electroporation, 52, 53
R green fluorescent protein as reporter,
Regulatory T-cell, flow cytometry, 52, 53
autoimmunity regulation, 156, 285 messenger RNA generation by in
isolation and functional vitro transcription, 5153
characterization, overview, 48
human cells, materials, 48, 49
immunomagnetic sorting, 289, RNA isolation, 50, 51, 53
294
mixed leukocyte reaction, 290 S
overview, 287, 288 SEREX, see Serological analysis of
polyclonal stimulation with tumor antigens by recombinant
anti-CD3 and anti-CD28, cDNA expression cloning
290, 291 Serological analysis of tumor antigens
positive and negative selection, by recombinant cDNA
289, 290, 294 expression cloning,
498 Index
database, 138 immunoselection of transduced cells,

immunoscreening of library, 482
patient serum preparation, materials, 481
overview, 145, 151 overview, 475, 476
preabsorption against retroviral vector transduction of T-
bacteriophage proteins and cells, 481, 482
lytically infected bacteria, transduced cell analysis, 482, 483
146 vectors, 476478
preabsorption against lytic
filters, 146, 147 T
preabsorption against T-cell, see also Cytotoxic T
mechanically disrupted lymphocyte; Regulatory T-cell;
bacteria, 145, 146 T-helper cell,
primary immunoscreening, 147, costimulatory blockade analysis with
148, 152 carboxyfluorescein
secondary immunoscreening, 148 succinimidyl ester,
phage expression library adoptive transfer of labeled cells
construction, into allogeneic hosts,
amplification, 144, 145, 151 blockade reagent treatment,
cDNA, 412
double-stranded synthesis for cell transfer, 412, 417
directional ligation, 141, irradiation, 412
142 cell staining after recovery,
ligation into vector, 143 apoptosis staining, 413
size fractionation, 142, 143 cytokine staining, 413
host bacteria preparation, 144 surface staining, 413
kit, 140, 150 donor lymphocyte labeling,
packaging, 143, 144 cell division effects, 411
plating and titering, 144 labeling conditions, 412, 415,
materials, 139, 140 417
molecular characterization of clones, lymphocyte preparation, 411,
149, 150, 153 412
petite serology characterization, flow cytometry,
conventional serology, 149 analysis, 415
overview, 148, 152, 153 data acquisition, 414, 418
SeroGRID, 149 materials, 410
principles, 137, 138 overview, 409, 410
Suicide gene therapy, graft-vs-leukemia recovery of labeled donor cells,
effect induction after bone 413, 418
marrow transplantation, dendritic cell activation, 5, 71
clinical outcomes, 480 graft-vs-leukemia effect mediation,
herpes simplex thymidine kinase 429, 430
system, 478480 T-cell receptor (TCR),
Index 499
antigen specificity regulation, 201 transgene expression analysis,

retroviral transduction of humanized 205, 206, 211
mouse receptors, overview, 201, 202
heterologous expression in human retrovirus production,
T-cells, plasmids, 203, 209, 210
selection markers, 246, 247, 251 producer cell transfection, 203,
transduction, selection, and 204, 210
expansion, 247, 249, tracing of cell, 207, 208
251253 TCR, see T-cell receptor
high-avidity tumor-specific T-cell T-helper cell, see also Regulatory T-
generation in transgenic mice, cell,
234 cytokine secretion assay for isolation
humanization of constructs, 240, and expansion of tumor-
241, 243, 244, 251 specific cells,
materials, 233 flow cytometry, 259263
murine receptor isolation and materials, 258
cloning, overview, 257, 258
cloning of amplification T-cell isolation from mice
products, 239, 251 immunized with irradiated
overview, 234236 tumor cells, 258, 259, 262, 263
RACE polymerase chain tumor immunity role, 257
reaction, 236, 237 TNF, see Tumor necrosis factor
reverse transcriptase- Tolerance,
polymerase chain reaction, circumvention, see Cytotoxic T-
238, 239 lymphocyte; T-cell receptor
RNA preparation, 236, 249 hematopoietic stem cell
subcloning into retroviral transplantation and tolerance
expression vectors, 240 induction in nonmyeloablated
rationale, 229, 230 mice,
receptor biochemical criteria, 230, flow cytometry identification of
232 stem cells, 460, 465467
single-chain receptor generation, immunomagnetic purification of
244246, 251 stem cells,
retroviral transduction of mouse T- cell sorting, 463, 464
cells, harvesting, 462, 563
adoptive transfer of cells, 207 magnetic bead labeling, 463
in vivo studies, 208, 209, 211 injection of stem cells, 464, 465
materials, 202, 203 materials, 460, 461
mouse T-cell studies, overview, 459, 460
cytokine staining, 206, 211 preconditioning of recipient mice,
infection, 205 461, 462
spleen cell cultures, 205, 210, short-term culture of stem cells,
211 465
500 Index
induction by monoclonal antibodies, Two-dimensional electrophoresis, see

306308 Cytotoxic T-lymphocyte;
Tumor-associated antigens, Regulatory T-cell
cell-mediated immune response, 137
identification, see Serological V
analysis of tumor antigens by
VLA-4, monoclonal antibody clinical
recombinant cDNA expression
trials, 315, 316
cloning
melanoma, 186, 266
multiple myeloma, 128 W
self antigens in graft-vs-leukemia Western blot,
effect, 431, 432 bifunctional antibodies, 338, 339
Tumor necrosis factor (TNF), epidermal growth factor receptor
monoclonal antibody clinical kinase activity assessment,
trials, 312314 383, 384, 386, 387

Adoptive Immunotherapy Ludewig Hoffmann (Methods Molec Medicine 109 Humana 2005)

Uploaded by

Copyright:

Available Formats

Adoptive Immunotherapy Ludewig Hoffmann (Methods Molec Medicine 109 Humana 2005)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Adoptive Immunotherapy Ludewig Hoffmann (Methods Molec Medicine 109 Humana 2005)

Uploaded by

Copyright:

Available Formats

M E T H O D S I N M O L E C U L A R M E D I C I N E TM

All rights reserved.

This publication is printed on acid-free paper.

Production Editor: Tracy Catanese

Photocopy Authorization Policy:

Printed in the United States of America. 10 9 8 7 6 5 4 3 2 1

Library of Congress Cataloging-in-Publication Data

11 Identification of Tumor-Associated Autoantigens With SEREX

24 Synthesis and Biological Evaluation of a Paclitaxel

I. JOLANDA M. DE VRIES Department of Tumor Immunology, Nijmegen

CHRISTOPH HUELS Protagen AG, Dortmund, Germany

RIENK OFFRINGA Department of Immunohematology and Blood

BURKHARD SCHMIDT Department of Hematology/Oncology, Klinikum rechts

MONIKA C. WOLKERS Department of Immunology, The Netherlands Cancer

David ONeill and Nina Bhardwaj

different sets of pathogen-associated molecules. For example, TLR3 recognizes

1.3. Antigen Presentation

1.4. Lymphocyte Activation

2. Ex Vivo Manipulation of DCs for Active Immunotherapy

Unfortunately, pharmaceutical-grade FL is no longer commercially available. Still,

Gennady Bocharov, Neville J. Ford, and Burkhard Ludewig

Mathematical modeling involves a number of distinct steps. In order to establish a

Bocharov, Ford, and Ludewig

3.3. Derivation of Model Equations

d Duration of antigen preprogrammed CTL divisions 1 d (ad hoc fixed value)

The equation for the dynamics of quiescent memory CTL is:

The weighted LSQ objective function

can be used when different weights have to be given to particular observations. A

3.5.2. Minimization of the Least-Squares Function

The information-theoretic framework provides a basis for such assessment of infor-

Form1 Dose/Effect Form2 Dose/Effect Form3 Dose/Effect

{ln ( yij ) ln yi (t j , p)}

24. Venzon, D. J. and Mooogavkor, S. H. (1988) A method for computing profile-likelihood-

Ying Waeckerle-Men, Bruno Gander, and Marcus Groettrup

Fig. 1. Microscopic image of poly(D,L-lactide-co-glycolide) microspheres (PLGA MS) and

Microencapsulation of antigens into PLGA can be achieved by several methods.

1.2. Evaluation of PLGA-MS Uptake by MoDC

1.3. Quality Control of Monocyte-Derived DC After PLGA-MS Uptake

1.4. Presentation of Antigens Contained in PLGA MS by Dendritic Cells

Fig. 2. Prolonged presentation of a cytotoxic T-lymphocyte (CTL) epitope by monocyte-

2. 24-well Cell-culture plates (Cellstar, Greiner Bio-One, Frickenhausen, Germany).

Fig. 3. Schematic illustration of antigen processing of poly(D, L -lactide-co-glycolide)

5. Propidium iodide (BD Pharmingen).

3.3.2. Survival of DC Loaded With PLGA MS

3.3.3. Migration Capacity of DC Loaded With PLGA MS

3.3.4. T-Cell Stimulatory Functions of MoDC Loaded With PLGA MS

3.4. MHC Class II-Restricted Antigen

3.5. MHC Class I-Restricted Antigen

Frank Grnebach, Martin R. Mller, and Peter Brossart

12. FACSCalibur flow cytometer (BD Biosciences).

3.1. Cell Isolation and Generation of Human DC

3.2. Total RNA Isolation From Human Cells

4. Pour the mixture into the gel tray.

3.3. Generation of EGFP mRNA by In Vitro Transcription

Fig. 3. Schematic drawing of plasmid pSP64-EGFP-2 for in vitro transcription of EGFP.

3.5. EGFP Analysis by Flow Cytometry

John D. M. Campbell, Christoph Piechaczek,

2.2. Generation of Immature Dendritic Cells From Monocytes

20. CD45.FITC (Miltenyi Biotec, ref. no. 130-080-202). Titer 1:11.

2.3.2 Myeloid Dentritic Cells (mDC)