Liposomes, Part F (Methods in Enzymology, Volume 464)

VOLUME FOUR HUNDRED AND SIXTY-FOUR

METHODS

IN

ENZYMOLOGY Liposomes, Part F

METHODS IN ENZYMOLOGY Editors-in-Chief

JOHN N. ABELSON AND MELVIN I. SIMON Division of Biology California Institute of Technology Pasadena, California, USA Founding Editors

SIDNEY P. COLOWICK AND NATHAN O. KAPLAN

VOLUME FOUR HUNDRED AND SIXTY-FOUR

METHODS

IN

ENZYMOLOGY Liposomes, Part F EDITED BY

¨ ZGU ¨ NES¸ NEJAT DU Department of Microbiology University of the Pacific Arthur A. Dugoni School of Dentistry San Francisco, California, USA

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

Academic Press is an imprint of Elsevier 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 32 Jamestown Road, London NW1 7BY, UK First edition 2009 Copyright # 2009, Elsevier Inc. All Rights Reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email: permissions@ elsevier.com. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made For information on all Academic Press publications visit our website at elsevierdirect.com ISBN: 978-0-12-374969-7 ISSN: 0076-6879 Printed and bound in United States of America 09 10 11 12 10 9 8 7 6 5 4 3 2 1

CONTENTS

Contributors Preface Volumes in Series

Section I. Bioactive Liposomes 1. Tubular Liposomes with Variable Permeability for Reconstitution of FtsZ Rings

xi xvii xix

1

3

Masaki Osawa and Harold P. Erickson 1. Introduction 2. Reagents 3. Bacterial Expression of Membrane Targeting FtsZ 4. Purification of FtsZ-mts and FtsZ-YFP-mts 5. Renatured Preparation of FtsZ-YFP-mts 6. Tubular Multilamellar Liposome Preparation 7. Permeability of the Multilamellar Liposomes 8. Z-ring Formation in Liposomes 9. A Crude Flow Chamber to Exchange Buffer Outside Liposomes 10. Factors Affecting Z-ring Formation in Liposomes 11. Utility of the Liposomes Beyond FtsZ References

2. Detection and Analysis of Protein Synthesis and RNA Replication in Giant Liposomes

4 4 5 5 7 7 9 11 11 14 16 16

19

Takeshi Sunami, Hiroshi Kita, Kazufumi Hosoda, Tomoaki Matsuura, Hiroaki Suzuki, and Tetsuya Yomo 1. Introduction 2. Methods 3. Analysis of the FACS Data 4. Conclusions Acknowledgments References

20 21 27 28 29 29

v

vi

Contents

3. Construction of Cell-Sized Liposomes Encapsulating Actin and Actin-Cross-Linking Proteins

31

Kingo Takiguchi, Ayako Yamada, Makiko Negishi, Makoto Honda, Yohko Tanaka-Takiguchi, and Kenichi Yoshikawa 1. Introduction 2. Experimental Section 3. Morphogenesis of Giant Liposomes Encapsulating Actin and Its Cross-linking Proteins 4. Concluding Remarks Acknowledgments References

4. Reconstitution of Membrane Budding with Unilamellar Vesicles

32 37 42 49 50 50

55

Anna V. Shnyrova and Joshua Zimmerberg 1. Introduction 2. M Protein Purification 3. Evaluation of the Membrane Activity of M Protein Through its Interaction with Intermediate-Sized Unilamellar Liposomes 4. Reconstitution of M-Protein-Driven Membrane Budding on GUVs 5. Concluding Remarks References

Section II. Liposomes and Nanotechnology 5. Detection of Antimycolic Acid Antibodies by Liposomal Biosensors

56 57 58 66 73 73

77

79

Y. Lemmer, S. T. Thanyani, P. J. Vrey, C. H. S. Driver, L. Venter, S. van Wyngaardt, A. M. C. ten Bokum, K. I. Ozoemena, L. A. Pilcher, D. G. Fernig, A. C. Stoltz, H. S. Swai, and J. A. Verschoor 1. Introduction 2. Experimental 3. Conclusion Acknowledgments References

6. Solid Lipid Nanoparticle Formulations: Pharmacokinetic and Biopharmaceutical Aspects in Drug Delivery

80 81 102 102 102

105

Eliana B. Souto and Slavomira Doktorovova´ 1. Introduction 2. Production of SLN

106 107

Contents

3. Pharmacokinetics and Pharmacodynamics 4. Modified Release Profile 5. Biopharmaceutical Aspects of Administration Routes 6. Clinical Pharmacology 7. Concluding Remarks References

vii

107 112 114 117 121 122

7. Preparation of Complexes of Liposomes with Gold Nanoparticles 131 Chie Kojima, Yusuke Hirano, and Kenji Kono 1. 2. 3. 4. 5. 6. 7.

Introduction Preparation of Complexes of EYPC Liposomes with Au NPs Time-Dependent SPR of the Complexes TEM Analysis of the Complexes DLS Analysis of the Complexes Calcein Release from the Complexes Estimation of Numbers of the Au NP and the Liposome in the Complexes 8. Optimization of Lipid Components of the Complexes 9. Concluding Remarks Acknowledgment References

8. Bio-Nanocapsule–Liposome Conjugates for In Vivo Pinpoint Drug and Gene Delivery

132 134 134 136 137 137 139 140 142 143 144

147

Takeshi Kasuya, Joohee Jung, Rie Kinoshita, Yasumasa Goh, Takashi Matsuzaki, Masumi Iijima, Nobuo Yoshimoto, Katsuyuki Tanizawa, and Shun’ichi Kuroda 1. Introduction 2. First-Generation Bio-Nanocapsules 3. Second-Generation BNCs 4. Retargeting of BNC–LP Conjugates 5. Overexpression of BNCs in S. cerevisiae 6. Conjugation of BNCs with LPs 7. Preparation of Antibody-Displaying BNC–LP Conjugates 8. Preparation of Biotin-Displaying BNC–LP Conjugates 9. Concluding Remarks Acknowledgments References

148 149 150 152 153 155 161 163 163 164 164

viii

Contents

9. Nanoliposomal Dry Powder Formulations

167

Gaurang Patel, Mahavir Chougule, Mandip Singh, and Ambikanandan Misra 1. Introduction 2. Preparation of Nanoliposomal DPFs 3. Physicochemical Characterization of NLDPFs 4. Concluding Remarks References

10. Lanthanide-Loaded Paramagnetic Liposomes as Switchable Magnetically Oriented Nanovesicles

168 169 174 187 188

193

Silvio Aime, Daniela Delli Castelli, and Enzo Terreno 1. 2. 3. 4.

Introduction Paramagnetic Ln(III)-Based Shift Reagents Preparation of Osmotically Shrunken Liposomes NMR Characterization of Magnetically Oriented Nonspherical Liposomes 5. Sample Experiments 6. Concluding Remarks Acknowledgments References

11. Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs

194 195 197 198 200 208 208 208

211

T. K. Ritchie, Y. V. Grinkova, T. H. Bayburt, I. G. Denisov, J. K. Zolnerciks, W. M. Atkins, and S. G. Sligar 1. Introduction 2. Overview of Nanodisc Technology 3. Reconstitution Considerations 4. Optimizing the Reconstitution for P-glycoprotein Acknowledgments References

12. DNA-Controlled Assembly of Liposomes in Diagnostics

212 212 218 223 228 228

233

Ulla Jakobsen and Stefan Vogel 1. Introduction 2. Probe Design 3. General Description of Materials and Techniques 4. Concluding Remarks Acknowledgment References

234 235 244 247 248 248

Contents

13. Soft Hybrid Nanostructures Composed of Phospholipid Liposomes Decorated with Oligonucleotides

ix

249

Martina Banchelli, Francesca Baldelli Bombelli, Debora Berti, and Piero Baglioni 1. Introduction 2. Materials 3. Liposome Preparation and Determination of Lipid Content 4. Incorporation of Oligonucleotides 5. Characterization of the Soft Hybrid Nanostructure 6. Applications of Oligo-Decorated Liposomes 7. Challenges and Perspectives Acknowledgments References

250 251 252 253 256 262 275 276 276

14. Synthesis, Characterization, and Optical Response of Gold Nanoshells Used to Trigger Release from Liposomes

279

Guohui Wu, Alexander Mikhailovsky, Htet A. Khant, and Joseph A. Zasadzinski 1. Introduction 2. Synthesis of HGNs 3. Optimization of HGN Dimensions for Maximum Absorption in the NIR 4. HGN Response to Femtosecond NIR Laser Pulses 5. Coupling HGN to Liposomes 6. Liposome Disruption and CF Release Due to Pulsed Laser Irradiation 7. Mechanism of Triggered Liposome Release 8. Effect of Proximity of HGNs to Liposomes 9. Conclusions Acknowledgments References

15. Complex Nanotube-Liposome Networks

280 283 286 290 294 299 300 303 304 304 304

309

Aldo Jesorka and Owe Orwar 1. Introduction 2. Network Fabrication Protocols 3. Complexity and Topology 4. Internal and Membrane Functionalization 5. Transport Phenomena and Controlled Mixing Procedures 6. Enzymatic Reactions in NVN 7. Concluding Remarks Acknowledgments References

309 310 314 315 318 320 323 323 324

x

Contents

16. Bionanotubules Formed from Liposomes

327

Josemar A. Castillo and Mark A. Hayes 1. Introduction 2. Bionanotubule Formation by Applying Electric Fields to Surface-Attached Liposomes 3. Bionanotubular Formation from Liposomes in Solution Using Electric Fields 4. Other Methods of Bionanotubular Formation from Liposomes 5. Concluding Remarks References

17. Engineering Cationic Liposome: siRNA Complexes for In Vitro and In Vivo Delivery

328 329 334 337 339 339

343

Jennifer E. Podesta and Kostas Kostarelos 1. Introduction 2. Cationic Liposome Systems for siRNA Delivery 3. Experimental Methods 4. Troubleshooting 5. Concluding Remarks References Author Index Subject Index

344 345 347 352 353 354 355 363

CONTRIBUTORS

Silvio Aime Department of Chemistry IFM and Molecular Imaging Center, University of Torino, Torino, Italy W. M. Atkins Department of Medicinal Chemistry, University of Washington, Seattle, Washington, USA Piero Baglioni Department of Chemistry and CSGI, University of Florence, Sesto Fiorentino, Florence, Italy Martina Banchelli Department of Chemistry and CSGI, University of Florence, Sesto Fiorentino, Florence, Italy T. H. Bayburt Department of Biochemistry, University of Illinois at Urbana-Champaign, Urbana, Illinois, USA Debora Berti Department of Chemistry and CSGI, University of Florence, Sesto Fiorentino, Florence, Italy Francesca Baldelli Bombelli Department of Chemistry and CSGI, University of Florence, Sesto Fiorentino, Florence, Italy Daniela Delli Castelli Department of Chemistry IFM and Molecular Imaging Center, University of Torino, Torino, Italy Josemar A. Castillo Department of Chemistry and Biochemistry, Arizona State University, Tempe, Arizona, USA Mahavir Chougule Pharmaceutics Department, College of Pharmacy and Pharmaceutical Science, Florida A&M University, Tallahassee, Florida, USA

xi

xii

Contributors

I. G. Denisov Department of Biochemistry, University of Illinois at Urbana-Champaign, Urbana, Illinois, USA Slavomira Doktorovova´ Department of Pharmaceutical Technology, Faculty of Health Sciences, Fernando Pessoa University, Porto, Portugal C. H. S. Driver Department of Chemistry, University of Pretoria, Pretoria, South Africa Harold P. Erickson Department of Cell Biology, Duke University Medical Center, Durham, North Carolina, USA D. G. Fernig School of Biological Sciences, University of Liverpool, United Kingdom Yasumasa Goh Beacle Inc., ORIC, Haga, Okayama, Japan Y. V. Grinkova Department of Biochemistry, University of Illinois at Urbana-Champaign, Urbana, Illinois, USA Mark A. Hayes Department of Chemistry and Biochemistry, Arizona State University, Tempe, Arizona, USA Yusuke Hirano Graduate School of Engineering, Osaka Prefecture University, Osaka, Japan Makoto Honda Division of Biological Science, Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Japan Current address: Stem Cell and Drug Discovery Institute, Shimogyo-ku, Kyoto, Japan Kazufumi Hosoda Department of Bioinformatics Engineering, Graduate School of Information Science and Technology, Osaka University, Suita, Osaka, Japan Masumi Iijima Institute of Scientific and Industrial Research, Osaka University, Ibaraki, Osaka, Japan, and Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa, Nagoya, Japan Ulla Jakobsen Nucleic Acid Center, University of Southern Denmark, Odense, Denmark

Contributors

xiii

Aldo Jesorka Department of Chemical and Biological Engineering, Chalmers University of Technology, Go¨teborg, Sweden Joohee Jung Institute of Scientific and Industrial Research, Osaka University, Ibaraki, Osaka, Japan Present address: Institute for Innovative Cancer Research, ASAN Medical Center, Pungnap-2, Songpa, Seoul, Korea Takeshi Kasuya Institute of Scientific and Industrial Research, Osaka University, Ibaraki, Osaka, Japan Htet A. Khant Department of Chemical Engineering, University of California, Santa Barbara, California, USA Rie Kinoshita Beacle Inc., ORIC, Haga, Okayama, Japan Hiroshi Kita Exploratory Research for Advanced Technology (ERATO), Japan Science and Technology Agency ( JST), Chiyoda-ku, Tokyo, Japan Chie Kojima Nanoscience and Nanotechnology Research Center, Research Institutes for the Twenty First Century, Osaka Prefecture University, Osaka, Japan Kenji Kono Graduate School of Engineering, Osaka Prefecture University, Osaka, Japan Kostas Kostarelos Nanomedicine Laboratory, Centre for Drug Delivery Research, The School of Pharmacy, University of London, London, United Kingdom Shun’ichi Kuroda Institute of Scientific and Industrial Research, Osaka University, Ibaraki, Osaka, Japan, and Beacle Inc., ORIC, Haga, Okayama, Japan; Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa, Nagoya, Japan Y. Lemmer Department of Biochemistry, University of Pretoria, Pretoria, South Africa, and Materials Science and Manufacturing, CSIR, Pretoria, South Africa Tomoaki Matsuura Department of Bioinformatics Engineering, Graduate School of Information Science and Technology, Osaka University, Suita, Osaka, Japan

xiv

Contributors

Takashi Matsuzaki Institute of Scientific and Industrial Research, Osaka University, Ibaraki, Osaka, Japan Alexander Mikhailovsky Department of Chemistry, University of California, Santa Barbara, California, USA Ambikanandan Misra TIFAC-CORE in NDDS, Pharmacy Department, Faculty of Technology and Engineering, The Maharaja Sayajirao University of Baroda, Kalabhavan, Vadodara, Gujarat, India Makiko Negishi Department of Physics, Graduate School of Science, Kyoto University, Sakyo-ku, Kyoto, Japan Owe Orwar Department of Chemical and Biological Engineering, Chalmers University of Technology, Go¨teborg, Sweden Masaki Osawa Department of Cell Biology, Duke University Medical Center, Durham, North Carolina, USA K. I. Ozoemena Department of Chemistry, University of Pretoria, Pretoria, South Africa Gaurang Patel TIFAC-CORE in NDDS, Pharmacy Department, Faculty of Technology and Engineering, The Maharaja Sayajirao University of Baroda, Kalabhavan, Vadodara, Gujarat, India L. A. Pilcher Department of Chemistry, University of Pretoria, Pretoria, South Africa Jennifer E. Podesta Nanomedicine Laboratory, Centre for Drug Delivery Research, The School of Pharmacy, University of London, London, United Kingdom T. K. Ritchie Department of Medicinal Chemistry, University of Washington, Seattle, Washington, USA Anna V. Shnyrova Laboratory of Cellular and Molecular Biology, Program in Physical Biology, Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, Maryland, USA Mandip Singh Pharmaceutics Department, College of Pharmacy and Pharmaceutical Science, Florida A&M University, Tallahassee, Florida, USA

Contributors

xv

S. G. Sligar Department of Biochemistry, University of Illinois at Urbana-Champaign, Urbana, Illinois, USA Eliana B. Souto Department of Pharmaceutical Technology, Faculty of Health Sciences, Fernando Pessoa University, Porto, Portugal, and Institute of Biotechnology and Bioengineering, Centre of Genetics and Biotechnology, University of Tra´s-os-Montes and Alto Douro (CGB-UTAD/IBB), Vila Real, Portugal A. C. Stoltz Department of Biochemistry, and Department of Infectious Diseases, University of Pretoria, Pretoria, South Africa Takeshi Sunami Exploratory Research for Advanced Technology (ERATO), Japan Science and Technology Agency ( JST), Chiyoda-ku, Tokyo, Japan Hiroaki Suzuki Department of Bioinformatics Engineering, Graduate School of Information Science and Technology, Osaka University, Suita, Osaka, Japan H. S. Swai Materials Science and Manufacturing, CSIR, Pretoria, South Africa Kingo Takiguchi Division of Biological Science, Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Japan Yohko Tanaka-Takiguchi Division of Biological Science, Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Japan Katsuyuki Tanizawa Institute of Scientific and Industrial Research, Osaka University, Ibaraki, Osaka, Japan A. M. C. ten Bokum Department of Biochemistry, University of Pretoria, Pretoria, South Africa Enzo Terreno Department of Chemistry IFM and Molecular Imaging Center, University of Torino, Torino, Italy S. T. Thanyani Department of Biochemistry, University of Pretoria, Pretoria, South Africa S. van Wyngaardt Department of Biochemistry, University of Pretoria, Pretoria, South Africa

xvi

Contributors

L. Venter Department of Biochemistry, University of Pretoria, Pretoria, South Africa J. A. Verschoor Department of Biochemistry, University of Pretoria, Pretoria, South Africa Stefan Vogel Nucleic Acid Center, University of Southern Denmark, Odense, Denmark P. J. Vrey Department of Biochemistry, University of Pretoria, Pretoria, South Africa Guohui Wu Department of Chemical Engineering, University of California, Santa Barbara, California, USA Ayako Yamada Department of Physics, Graduate School of Science, Kyoto University, Sakyo-ku, Kyoto, Japan Current address: Department of Chemistry, Ecole Normale Superieure, Paris, France Tetsuya Yomo Department of Bioinformatics Engineering, Graduate School of Information Science and Technology, and Graduate School of Frontier Biosciences, Osaka University, Suita, Osaka, Japan, and Exploratory Research for Advanced Technology (ERATO), Japan Science and Technology Agency ( JST), Chiyoda-ku, Tokyo, Japan Kenichi Yoshikawa Department of Physics, Graduate School of Science, Kyoto University, Sakyo-ku, Kyoto, Japan Nobuo Yoshimoto Institute of Scientific and Industrial Research, Osaka University, Ibaraki, Osaka, Japan, and Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa, Nagoya, Japan Joseph A. Zasadzinski Department of Chemical Engineering, University of California, Santa Barbara, California, USA Joshua Zimmerberg Laboratory of Cellular and Molecular Biology, Program in Physical Biology, Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, Maryland, USA J. K. Zolnerciks Department of Medicinal Chemistry, University of Washington, Seattle, Washington, USA

PREFACE

Previous Methods in Enzymology volumes on ‘‘Liposomes’’ have described methods of liposome preparation and the physicochemical characterization of liposomes (Volume 367), and the use of liposomes in biochemistry, molecular cell biology (Volume 372), immunology, diagnostics, gene delivery, and gene therapy (Volume 373). Methods involved in the production and application of antibody- or ligand-targeted liposomes, environmentsensitive liposomes, and liposomal oligonucleotides were provided in Volume 387, as were methods for studying the in vivo fate of liposomes. Finally, Volume 391 presented methods in liposomal anticancer, antibacterial, antifungal, and antiviral agents, miscellaneous liposomal therapies and electron microscopy of liposomes. The latter volume also included a short introductory chapter on ‘‘The Origin of Liposomes: Alec Bangham at Babraham.’’ This new volume includes sections focusing on bioactive liposomes and the interface of liposomes and nanotechnology. I hope that these chapters will be helpful to graduate students, postdoctoral fellows, research associates, and established scientists initiating projects on liposomes, or shifting the focus of their research. Although the chapters are not written in ‘‘protocol’’ format, they describe the experimental methods in sufficient detail that can be adopted readily for the reader’s project. In addition, the chapters provide the perspective of the authors on the field, as well as examples of results obtained with the described methods. I would like to thank all the colleagues who graciously contributed to this volume with relatively short notice, Associate Editor Tara Hoey, Editorial Services Manager Delsy Retchagar, and Production Manager Radhakrishnan Lakshmanan of Elsevier for their help in preparing this volume, and Shirley Light, formerly of Academic Press, for her initiation of the ‘‘Liposomes’’ volumes about a decade ago. I would also like to express my gratitude to my supportive and loving family. I dedicate this volume to my wife, Diana, and my curious, creative, playful, and loving children, Avery and Maxine. NEJAT DU¨ZGU¨NES¸

xvii

This page intentionally left blank

METHODS IN ENZYMOLOGY

VOLUME I. Preparation and Assay of Enzymes Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME II. Preparation and Assay of Enzymes Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME III. Preparation and Assay of Substrates Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME IV. Special Techniques for the Enzymologist Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME V. Preparation and Assay of Enzymes Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME VI. Preparation and Assay of Enzymes (Continued) Preparation and Assay of Substrates Special Techniques Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME VII. Cumulative Subject Index Edited by SIDNEY P. COLOWICK AND NATHAN O. KAPLAN VOLUME VIII. Complex Carbohydrates Edited by ELIZABETH F. NEUFELD AND VICTOR GINSBURG VOLUME IX. Carbohydrate Metabolism Edited by WILLIS A. WOOD VOLUME X. Oxidation and Phosphorylation Edited by RONALD W. ESTABROOK AND MAYNARD E. PULLMAN VOLUME XI. Enzyme Structure Edited by C. H. W. HIRS VOLUME XII. Nucleic Acids (Parts A and B) Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME XIII. Citric Acid Cycle Edited by J. M. LOWENSTEIN VOLUME XIV. Lipids Edited by J. M. LOWENSTEIN VOLUME XV. Steroids and Terpenoids Edited by RAYMOND B. CLAYTON

xix

xx

Methods in Enzymology

VOLUME XVI. Fast Reactions Edited by KENNETH KUSTIN VOLUME XVII. Metabolism of Amino Acids and Amines (Parts A and B) Edited by HERBERT TABOR AND CELIA WHITE TABOR VOLUME XVIII. Vitamins and Coenzymes (Parts A, B, and C) Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME XIX. Proteolytic Enzymes Edited by GERTRUDE E. PERLMANN AND LASZLO LORAND VOLUME XX. Nucleic Acids and Protein Synthesis (Part C) Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME XXI. Nucleic Acids (Part D) Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME XXII. Enzyme Purification and Related Techniques Edited by WILLIAM B. JAKOBY VOLUME XXIII. Photosynthesis (Part A) Edited by ANTHONY SAN PIETRO VOLUME XXIV. Photosynthesis and Nitrogen Fixation (Part B) Edited by ANTHONY SAN PIETRO VOLUME XXV. Enzyme Structure (Part B) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XXVI. Enzyme Structure (Part C) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XXVII. Enzyme Structure (Part D) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XXVIII. Complex Carbohydrates (Part B) Edited by VICTOR GINSBURG VOLUME XXIX. Nucleic Acids and Protein Synthesis (Part E) Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME XXX. Nucleic Acids and Protein Synthesis (Part F) Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME XXXI. Biomembranes (Part A) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME XXXII. Biomembranes (Part B) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME XXXIII. Cumulative Subject Index Volumes I-XXX Edited by MARTHA G. DENNIS AND EDWARD A. DENNIS VOLUME XXXIV. Affinity Techniques (Enzyme Purification: Part B) Edited by WILLIAM B. JAKOBY AND MEIR WILCHEK

Methods in Enzymology

VOLUME XXXV. Lipids (Part B) Edited by JOHN M. LOWENSTEIN VOLUME XXXVI. Hormone Action (Part A: Steroid Hormones) Edited by BERT W. O’MALLEY AND JOEL G. HARDMAN VOLUME XXXVII. Hormone Action (Part B: Peptide Hormones) Edited by BERT W. O’MALLEY AND JOEL G. HARDMAN VOLUME XXXVIII. Hormone Action (Part C: Cyclic Nucleotides) Edited by JOEL G. HARDMAN AND BERT W. O’MALLEY VOLUME XXXIX. Hormone Action (Part D: Isolated Cells, Tissues, and Organ Systems) Edited by JOEL G. HARDMAN AND BERT W. O’MALLEY VOLUME XL. Hormone Action (Part E: Nuclear Structure and Function) Edited by BERT W. O’MALLEY AND JOEL G. HARDMAN VOLUME XLI. Carbohydrate Metabolism (Part B) Edited by W. A. WOOD VOLUME XLII. Carbohydrate Metabolism (Part C) Edited by W. A. WOOD VOLUME XLIII. Antibiotics Edited by JOHN H. HASH VOLUME XLIV. Immobilized Enzymes Edited by KLAUS MOSBACH VOLUME XLV. Proteolytic Enzymes (Part B) Edited by LASZLO LORAND VOLUME XLVI. Affinity Labeling Edited by WILLIAM B. JAKOBY AND MEIR WILCHEK VOLUME XLVII. Enzyme Structure (Part E) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XLVIII. Enzyme Structure (Part F) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME XLIX. Enzyme Structure (Part G) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME L. Complex Carbohydrates (Part C) Edited by VICTOR GINSBURG VOLUME LI. Purine and Pyrimidine Nucleotide Metabolism Edited by PATRICIA A. HOFFEE AND MARY ELLEN JONES VOLUME LII. Biomembranes (Part C: Biological Oxidations) Edited by SIDNEY FLEISCHER AND LESTER PACKER

xxi

xxii

Methods in Enzymology

VOLUME LIII. Biomembranes (Part D: Biological Oxidations) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LIV. Biomembranes (Part E: Biological Oxidations) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LV. Biomembranes (Part F: Bioenergetics) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LVI. Biomembranes (Part G: Bioenergetics) Edited by SIDNEY FLEISCHER AND LESTER PACKER VOLUME LVII. Bioluminescence and Chemiluminescence Edited by MARLENE A. DELUCA VOLUME LVIII. Cell Culture Edited by WILLIAM B. JAKOBY AND IRA PASTAN VOLUME LIX. Nucleic Acids and Protein Synthesis (Part G) Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME LX. Nucleic Acids and Protein Synthesis (Part H) Edited by KIVIE MOLDAVE AND LAWRENCE GROSSMAN VOLUME 61. Enzyme Structure (Part H) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 62. Vitamins and Coenzymes (Part D) Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME 63. Enzyme Kinetics and Mechanism (Part A: Initial Rate and Inhibitor Methods) Edited by DANIEL L. PURICH VOLUME 64. Enzyme Kinetics and Mechanism (Part B: Isotopic Probes and Complex Enzyme Systems) Edited by DANIEL L. PURICH VOLUME 65. Nucleic Acids (Part I) Edited by LAWRENCE GROSSMAN AND KIVIE MOLDAVE VOLUME 66. Vitamins and Coenzymes (Part E) Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME 67. Vitamins and Coenzymes (Part F) Edited by DONALD B. MCCORMICK AND LEMUEL D. WRIGHT VOLUME 68. Recombinant DNA Edited by RAY WU VOLUME 69. Photosynthesis and Nitrogen Fixation (Part C) Edited by ANTHONY SAN PIETRO VOLUME 70. Immunochemical Techniques (Part A) Edited by HELEN VAN VUNAKIS AND JOHN J. LANGONE

Methods in Enzymology

xxiii

VOLUME 71. Lipids (Part C) Edited by JOHN M. LOWENSTEIN VOLUME 72. Lipids (Part D) Edited by JOHN M. LOWENSTEIN VOLUME 73. Immunochemical Techniques (Part B) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 74. Immunochemical Techniques (Part C) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 75. Cumulative Subject Index Volumes XXXI, XXXII, XXXIV–LX Edited by EDWARD A. DENNIS AND MARTHA G. DENNIS VOLUME 76. Hemoglobins Edited by ERALDO ANTONINI, LUIGI ROSSI-BERNARDI, AND EMILIA CHIANCONE VOLUME 77. Detoxication and Drug Metabolism Edited by WILLIAM B. JAKOBY VOLUME 78. Interferons (Part A) Edited by SIDNEY PESTKA VOLUME 79. Interferons (Part B) Edited by SIDNEY PESTKA VOLUME 80. Proteolytic Enzymes (Part C) Edited by LASZLO LORAND VOLUME 81. Biomembranes (Part H: Visual Pigments and Purple Membranes, I) Edited by LESTER PACKER VOLUME 82. Structural and Contractile Proteins (Part A: Extracellular Matrix) Edited by LEON W. CUNNINGHAM AND DIXIE W. FREDERIKSEN VOLUME 83. Complex Carbohydrates (Part D) Edited by VICTOR GINSBURG VOLUME 84. Immunochemical Techniques (Part D: Selected Immunoassays) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 85. Structural and Contractile Proteins (Part B: The Contractile Apparatus and the Cytoskeleton) Edited by DIXIE W. FREDERIKSEN AND LEON W. CUNNINGHAM VOLUME 86. Prostaglandins and Arachidonate Metabolites Edited by WILLIAM E. M. LANDS AND WILLIAM L. SMITH VOLUME 87. Enzyme Kinetics and Mechanism (Part C: Intermediates, Stereo-chemistry, and Rate Studies) Edited by DANIEL L. PURICH VOLUME 88. Biomembranes (Part I: Visual Pigments and Purple Membranes, II) Edited by LESTER PACKER

xxiv

Methods in Enzymology

VOLUME 89. Carbohydrate Metabolism (Part D) Edited by WILLIS A. WOOD VOLUME 90. Carbohydrate Metabolism (Part E) Edited by WILLIS A. WOOD VOLUME 91. Enzyme Structure (Part I) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 92. Immunochemical Techniques (Part E: Monoclonal Antibodies and General Immunoassay Methods) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 93. Immunochemical Techniques (Part F: Conventional Antibodies, Fc Receptors, and Cytotoxicity) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 94. Polyamines Edited by HERBERT TABOR AND CELIA WHITE TABOR VOLUME 95. Cumulative Subject Index Volumes 61–74, 76–80 Edited by EDWARD A. DENNIS AND MARTHA G. DENNIS VOLUME 96. Biomembranes [Part J: Membrane Biogenesis: Assembly and Targeting (General Methods; Eukaryotes)] Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 97. Biomembranes [Part K: Membrane Biogenesis: Assembly and Targeting (Prokaryotes, Mitochondria, and Chloroplasts)] Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 98. Biomembranes (Part L: Membrane Biogenesis: Processing and Recycling) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 99. Hormone Action (Part F: Protein Kinases) Edited by JACKIE D. CORBIN AND JOEL G. HARDMAN VOLUME 100. Recombinant DNA (Part B) Edited by RAY WU, LAWRENCE GROSSMAN, AND KIVIE MOLDAVE VOLUME 101. Recombinant DNA (Part C) Edited by RAY WU, LAWRENCE GROSSMAN, AND KIVIE MOLDAVE VOLUME 102. Hormone Action (Part G: Calmodulin and Calcium-Binding Proteins) Edited by ANTHONY R. MEANS AND BERT W. O’MALLEY VOLUME 103. Hormone Action (Part H: Neuroendocrine Peptides) Edited by P. MICHAEL CONN VOLUME 104. Enzyme Purification and Related Techniques (Part C) Edited by WILLIAM B. JAKOBY

Methods in Enzymology

xxv

VOLUME 105. Oxygen Radicals in Biological Systems Edited by LESTER PACKER VOLUME 106. Posttranslational Modifications (Part A) Edited by FINN WOLD AND KIVIE MOLDAVE VOLUME 107. Posttranslational Modifications (Part B) Edited by FINN WOLD AND KIVIE MOLDAVE VOLUME 108. Immunochemical Techniques (Part G: Separation and Characterization of Lymphoid Cells) Edited by GIOVANNI DI SABATO, JOHN J. LANGONE, AND HELEN VAN VUNAKIS VOLUME 109. Hormone Action (Part I: Peptide Hormones) Edited by LUTZ BIRNBAUMER AND BERT W. O’MALLEY VOLUME 110. Steroids and Isoprenoids (Part A) Edited by JOHN H. LAW AND HANS C. RILLING VOLUME 111. Steroids and Isoprenoids (Part B) Edited by JOHN H. LAW AND HANS C. RILLING VOLUME 112. Drug and Enzyme Targeting (Part A) Edited by KENNETH J. WIDDER AND RALPH GREEN VOLUME 113. Glutamate, Glutamine, Glutathione, and Related Compounds Edited by ALTON MEISTER VOLUME 114. Diffraction Methods for Biological Macromolecules (Part A) Edited by HAROLD W. WYCKOFF, C. H. W. HIRS, AND SERGE N. TIMASHEFF VOLUME 115. Diffraction Methods for Biological Macromolecules (Part B) Edited by HAROLD W. WYCKOFF, C. H. W. HIRS, AND SERGE N. TIMASHEFF VOLUME 116. Immunochemical Techniques (Part H: Effectors and Mediators of Lymphoid Cell Functions) Edited by GIOVANNI DI SABATO, JOHN J. LANGONE, AND HELEN VAN VUNAKIS VOLUME 117. Enzyme Structure (Part J) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 118. Plant Molecular Biology Edited by ARTHUR WEISSBACH AND HERBERT WEISSBACH VOLUME 119. Interferons (Part C) Edited by SIDNEY PESTKA VOLUME 120. Cumulative Subject Index Volumes 81–94, 96–101 VOLUME 121. Immunochemical Techniques (Part I: Hybridoma Technology and Monoclonal Antibodies) Edited by JOHN J. LANGONE AND HELEN VAN VUNAKIS VOLUME 122. Vitamins and Coenzymes (Part G) Edited by FRANK CHYTIL AND DONALD B. MCCORMICK

xxvi

Methods in Enzymology

VOLUME 123. Vitamins and Coenzymes (Part H) Edited by FRANK CHYTIL AND DONALD B. MCCORMICK VOLUME 124. Hormone Action (Part J: Neuroendocrine Peptides) Edited by P. MICHAEL CONN VOLUME 125. Biomembranes (Part M: Transport in Bacteria, Mitochondria, and Chloroplasts: General Approaches and Transport Systems) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 126. Biomembranes (Part N: Transport in Bacteria, Mitochondria, and Chloroplasts: Protonmotive Force) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 127. Biomembranes (Part O: Protons and Water: Structure and Translocation) Edited by LESTER PACKER VOLUME 128. Plasma Lipoproteins (Part A: Preparation, Structure, and Molecular Biology) Edited by JERE P. SEGREST AND JOHN J. ALBERS VOLUME 129. Plasma Lipoproteins (Part B: Characterization, Cell Biology, and Metabolism) Edited by JOHN J. ALBERS AND JERE P. SEGREST VOLUME 130. Enzyme Structure (Part K) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 131. Enzyme Structure (Part L) Edited by C. H. W. HIRS AND SERGE N. TIMASHEFF VOLUME 132. Immunochemical Techniques (Part J: Phagocytosis and Cell-Mediated Cytotoxicity) Edited by GIOVANNI DI SABATO AND JOHANNES EVERSE VOLUME 133. Bioluminescence and Chemiluminescence (Part B) Edited by MARLENE DELUCA AND WILLIAM D. MCELROY VOLUME 134. Structural and Contractile Proteins (Part C: The Contractile Apparatus and the Cytoskeleton) Edited by RICHARD B. VALLEE VOLUME 135. Immobilized Enzymes and Cells (Part B) Edited by KLAUS MOSBACH VOLUME 136. Immobilized Enzymes and Cells (Part C) Edited by KLAUS MOSBACH VOLUME 137. Immobilized Enzymes and Cells (Part D) Edited by KLAUS MOSBACH VOLUME 138. Complex Carbohydrates (Part E) Edited by VICTOR GINSBURG

Methods in Enzymology

xxvii

VOLUME 139. Cellular Regulators (Part A: Calcium- and Calmodulin-Binding Proteins) Edited by ANTHONY R. MEANS AND P. MICHAEL CONN VOLUME 140. Cumulative Subject Index Volumes 102–119, 121–134 VOLUME 141. Cellular Regulators (Part B: Calcium and Lipids) Edited by P. MICHAEL CONN AND ANTHONY R. MEANS VOLUME 142. Metabolism of Aromatic Amino Acids and Amines Edited by SEYMOUR KAUFMAN VOLUME 143. Sulfur and Sulfur Amino Acids Edited by WILLIAM B. JAKOBY AND OWEN GRIFFITH VOLUME 144. Structural and Contractile Proteins (Part D: Extracellular Matrix) Edited by LEON W. CUNNINGHAM VOLUME 145. Structural and Contractile Proteins (Part E: Extracellular Matrix) Edited by LEON W. CUNNINGHAM VOLUME 146. Peptide Growth Factors (Part A) Edited by DAVID BARNES AND DAVID A. SIRBASKU VOLUME 147. Peptide Growth Factors (Part B) Edited by DAVID BARNES AND DAVID A. SIRBASKU VOLUME 148. Plant Cell Membranes Edited by LESTER PACKER AND ROLAND DOUCE VOLUME 149. Drug and Enzyme Targeting (Part B) Edited by RALPH GREEN AND KENNETH J. WIDDER VOLUME 150. Immunochemical Techniques (Part K: In Vitro Models of B and T Cell Functions and Lymphoid Cell Receptors) Edited by GIOVANNI DI SABATO VOLUME 151. Molecular Genetics of Mammalian Cells Edited by MICHAEL M. GOTTESMAN VOLUME 152. Guide to Molecular Cloning Techniques Edited by SHELBY L. BERGER AND ALAN R. KIMMEL VOLUME 153. Recombinant DNA (Part D) Edited by RAY WU AND LAWRENCE GROSSMAN VOLUME 154. Recombinant DNA (Part E) Edited by RAY WU AND LAWRENCE GROSSMAN VOLUME 155. Recombinant DNA (Part F) Edited by RAY WU VOLUME 156. Biomembranes (Part P: ATP-Driven Pumps and Related Transport: The Na, K-Pump) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER

xxviii

Methods in Enzymology

VOLUME 157. Biomembranes (Part Q: ATP-Driven Pumps and Related Transport: Calcium, Proton, and Potassium Pumps) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 158. Metalloproteins (Part A) Edited by JAMES F. RIORDAN AND BERT L. VALLEE VOLUME 159. Initiation and Termination of Cyclic Nucleotide Action Edited by JACKIE D. CORBIN AND ROGER A. JOHNSON VOLUME 160. Biomass (Part A: Cellulose and Hemicellulose) Edited by WILLIS A. WOOD AND SCOTT T. KELLOGG VOLUME 161. Biomass (Part B: Lignin, Pectin, and Chitin) Edited by WILLIS A. WOOD AND SCOTT T. KELLOGG VOLUME 162. Immunochemical Techniques (Part L: Chemotaxis and Inflammation) Edited by GIOVANNI DI SABATO VOLUME 163. Immunochemical Techniques (Part M: Chemotaxis and Inflammation) Edited by GIOVANNI DI SABATO VOLUME 164. Ribosomes Edited by HARRY F. NOLLER, JR., AND KIVIE MOLDAVE VOLUME 165. Microbial Toxins: Tools for Enzymology Edited by SIDNEY HARSHMAN VOLUME 166. Branched-Chain Amino Acids Edited by ROBERT HARRIS AND JOHN R. SOKATCH VOLUME 167. Cyanobacteria Edited by LESTER PACKER AND ALEXANDER N. GLAZER VOLUME 168. Hormone Action (Part K: Neuroendocrine Peptides) Edited by P. MICHAEL CONN VOLUME 169. Platelets: Receptors, Adhesion, Secretion (Part A) Edited by JACEK HAWIGER VOLUME 170. Nucleosomes Edited by PAUL M. WASSARMAN AND ROGER D. KORNBERG VOLUME 171. Biomembranes (Part R: Transport Theory: Cells and Model Membranes) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 172. Biomembranes (Part S: Transport: Membrane Isolation and Characterization) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER

Methods in Enzymology

xxix

VOLUME 173. Biomembranes [Part T: Cellular and Subcellular Transport: Eukaryotic (Nonepithelial) Cells] Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 174. Biomembranes [Part U: Cellular and Subcellular Transport: Eukaryotic (Nonepithelial) Cells] Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 175. Cumulative Subject Index Volumes 135–139, 141–167 VOLUME 176. Nuclear Magnetic Resonance (Part A: Spectral Techniques and Dynamics) Edited by NORMAN J. OPPENHEIMER AND THOMAS L. JAMES VOLUME 177. Nuclear Magnetic Resonance (Part B: Structure and Mechanism) Edited by NORMAN J. OPPENHEIMER AND THOMAS L. JAMES VOLUME 178. Antibodies, Antigens, and Molecular Mimicry Edited by JOHN J. LANGONE VOLUME 179. Complex Carbohydrates (Part F) Edited by VICTOR GINSBURG VOLUME 180. RNA Processing (Part A: General Methods) Edited by JAMES E. DAHLBERG AND JOHN N. ABELSON VOLUME 181. RNA Processing (Part B: Specific Methods) Edited by JAMES E. DAHLBERG AND JOHN N. ABELSON VOLUME 182. Guide to Protein Purification Edited by MURRAY P. DEUTSCHER VOLUME 183. Molecular Evolution: Computer Analysis of Protein and Nucleic Acid Sequences Edited by RUSSELL F. DOOLITTLE VOLUME 184. Avidin-Biotin Technology Edited by MEIR WILCHEK AND EDWARD A. BAYER VOLUME 185. Gene Expression Technology Edited by DAVID V. GOEDDEL VOLUME 186. Oxygen Radicals in Biological Systems (Part B: Oxygen Radicals and Antioxidants) Edited by LESTER PACKER AND ALEXANDER N. GLAZER VOLUME 187. Arachidonate Related Lipid Mediators Edited by ROBERT C. MURPHY AND FRANK A. FITZPATRICK VOLUME 188. Hydrocarbons and Methylotrophy Edited by MARY E. LIDSTROM VOLUME 189. Retinoids (Part A: Molecular and Metabolic Aspects) Edited by LESTER PACKER

xxx

Methods in Enzymology

VOLUME 190. Retinoids (Part B: Cell Differentiation and Clinical Applications) Edited by LESTER PACKER VOLUME 191. Biomembranes (Part V: Cellular and Subcellular Transport: Epithelial Cells) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 192. Biomembranes (Part W: Cellular and Subcellular Transport: Epithelial Cells) Edited by SIDNEY FLEISCHER AND BECCA FLEISCHER VOLUME 193. Mass Spectrometry Edited by JAMES A. MCCLOSKEY VOLUME 194. Guide to Yeast Genetics and Molecular Biology Edited by CHRISTINE GUTHRIE AND GERALD R. FINK VOLUME 195. Adenylyl Cyclase, G Proteins, and Guanylyl Cyclase Edited by ROGER A. JOHNSON AND JACKIE D. CORBIN VOLUME 196. Molecular Motors and the Cytoskeleton Edited by RICHARD B. VALLEE VOLUME 197. Phospholipases Edited by EDWARD A. DENNIS VOLUME 198. Peptide Growth Factors (Part C) Edited by DAVID BARNES, J. P. MATHER, AND GORDON H. SATO VOLUME 199. Cumulative Subject Index Volumes 168–174, 176–194 VOLUME 200. Protein Phosphorylation (Part A: Protein Kinases: Assays, Purification, Antibodies, Functional Analysis, Cloning, and Expression) Edited by TONY HUNTER AND BARTHOLOMEW M. SEFTON VOLUME 201. Protein Phosphorylation (Part B: Analysis of Protein Phosphorylation, Protein Kinase Inhibitors, and Protein Phosphatases) Edited by TONY HUNTER AND BARTHOLOMEW M. SEFTON VOLUME 202. Molecular Design and Modeling: Concepts and Applications (Part A: Proteins, Peptides, and Enzymes) Edited by JOHN J. LANGONE VOLUME 203. Molecular Design and Modeling: Concepts and Applications (Part B: Antibodies and Antigens, Nucleic Acids, Polysaccharides, and Drugs) Edited by JOHN J. LANGONE VOLUME 204. Bacterial Genetic Systems Edited by JEFFREY H. MILLER VOLUME 205. Metallobiochemistry (Part B: Metallothionein and Related Molecules) Edited by JAMES F. RIORDAN AND BERT L. VALLEE

Methods in Enzymology

xxxi

VOLUME 206. Cytochrome P450 Edited by MICHAEL R. WATERMAN AND ERIC F. JOHNSON VOLUME 207. Ion Channels Edited by BERNARDO RUDY AND LINDA E. IVERSON VOLUME 208. Protein–DNA Interactions Edited by ROBERT T. SAUER VOLUME 209. Phospholipid Biosynthesis Edited by EDWARD A. DENNIS AND DENNIS E. VANCE VOLUME 210. Numerical Computer Methods Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 211. DNA Structures (Part A: Synthesis and Physical Analysis of DNA) Edited by DAVID M. J. LILLEY AND JAMES E. DAHLBERG VOLUME 212. DNA Structures (Part B: Chemical and Electrophoretic Analysis of DNA) Edited by DAVID M. J. LILLEY AND JAMES E. DAHLBERG VOLUME 213. Carotenoids (Part A: Chemistry, Separation, Quantitation, and Antioxidation) Edited by LESTER PACKER VOLUME 214. Carotenoids (Part B: Metabolism, Genetics, and Biosynthesis) Edited by LESTER PACKER VOLUME 215. Platelets: Receptors, Adhesion, Secretion (Part B) Edited by JACEK J. HAWIGER VOLUME 216. Recombinant DNA (Part G) Edited by RAY WU VOLUME 217. Recombinant DNA (Part H) Edited by RAY WU VOLUME 218. Recombinant DNA (Part I) Edited by RAY WU VOLUME 219. Reconstitution of Intracellular Transport Edited by JAMES E. ROTHMAN VOLUME 220. Membrane Fusion Techniques (Part A) Edited by NEJAT DU¨ZGU¨NES, VOLUME 221. Membrane Fusion Techniques (Part B) Edited by NEJAT DU¨ZGU¨NES, VOLUME 222. Proteolytic Enzymes in Coagulation, Fibrinolysis, and Complement Activation (Part A: Mammalian Blood Coagulation Factors and Inhibitors) Edited by LASZLO LORAND AND KENNETH G. MANN

xxxii

Methods in Enzymology

VOLUME 223. Proteolytic Enzymes in Coagulation, Fibrinolysis, and Complement Activation (Part B: Complement Activation, Fibrinolysis, and Nonmammalian Blood Coagulation Factors) Edited by LASZLO LORAND AND KENNETH G. MANN VOLUME 224. Molecular Evolution: Producing the Biochemical Data Edited by ELIZABETH ANNE ZIMMER, THOMAS J. WHITE, REBECCA L. CANN, AND ALLAN C. WILSON VOLUME 225. Guide to Techniques in Mouse Development Edited by PAUL M. WASSARMAN AND MELVIN L. DEPAMPHILIS VOLUME 226. Metallobiochemistry (Part C: Spectroscopic and Physical Methods for Probing Metal Ion Environments in Metalloenzymes and Metalloproteins) Edited by JAMES F. RIORDAN AND BERT L. VALLEE VOLUME 227. Metallobiochemistry (Part D: Physical and Spectroscopic Methods for Probing Metal Ion Environments in Metalloproteins) Edited by JAMES F. RIORDAN AND BERT L. VALLEE VOLUME 228. Aqueous Two-Phase Systems Edited by HARRY WALTER AND GO¨TE JOHANSSON VOLUME 229. Cumulative Subject Index Volumes 195–198, 200–227 VOLUME 230. Guide to Techniques in Glycobiology Edited by WILLIAM J. LENNARZ AND GERALD W. HART VOLUME 231. Hemoglobins (Part B: Biochemical and Analytical Methods) Edited by JOHANNES EVERSE, KIM D. VANDEGRIFF, AND ROBERT M. WINSLOW VOLUME 232. Hemoglobins (Part C: Biophysical Methods) Edited by JOHANNES EVERSE, KIM D. VANDEGRIFF, AND ROBERT M. WINSLOW VOLUME 233. Oxygen Radicals in Biological Systems (Part C) Edited by LESTER PACKER VOLUME 234. Oxygen Radicals in Biological Systems (Part D) Edited by LESTER PACKER VOLUME 235. Bacterial Pathogenesis (Part A: Identification and Regulation of Virulence Factors) Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL VOLUME 236. Bacterial Pathogenesis (Part B: Integration of Pathogenic Bacteria with Host Cells) Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL VOLUME 237. Heterotrimeric G Proteins Edited by RAVI IYENGAR VOLUME 238. Heterotrimeric G-Protein Effectors Edited by RAVI IYENGAR

Methods in Enzymology

xxxiii

VOLUME 239. Nuclear Magnetic Resonance (Part C) Edited by THOMAS L. JAMES AND NORMAN J. OPPENHEIMER VOLUME 240. Numerical Computer Methods (Part B) Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND VOLUME 241. Retroviral Proteases Edited by LAWRENCE C. KUO AND JULES A. SHAFER VOLUME 242. Neoglycoconjugates (Part A) Edited by Y. C. LEE AND REIKO T. LEE VOLUME 243. Inorganic Microbial Sulfur Metabolism Edited by HARRY D. PECK, JR., AND JEAN LEGALL VOLUME 244. Proteolytic Enzymes: Serine and Cysteine Peptidases Edited by ALAN J. BARRETT VOLUME 245. Extracellular Matrix Components Edited by E. RUOSLAHTI AND E. ENGVALL VOLUME 246. Biochemical Spectroscopy Edited by KENNETH SAUER VOLUME 247. Neoglycoconjugates (Part B: Biomedical Applications) Edited by Y. C. LEE AND REIKO T. LEE VOLUME 248. Proteolytic Enzymes: Aspartic and Metallo Peptidases Edited by ALAN J. BARRETT VOLUME 249. Enzyme Kinetics and Mechanism (Part D: Developments in Enzyme Dynamics) Edited by DANIEL L. PURICH VOLUME 250. Lipid Modifications of Proteins Edited by PATRICK J. CASEY AND JANICE E. BUSS VOLUME 251. Biothiols (Part A: Monothiols and Dithiols, Protein Thiols, and Thiyl Radicals) Edited by LESTER PACKER VOLUME 252. Biothiols (Part B: Glutathione and Thioredoxin; Thiols in Signal Transduction and Gene Regulation) Edited by LESTER PACKER VOLUME 253. Adhesion of Microbial Pathogens Edited by RON J. DOYLE AND ITZHAK OFEK VOLUME 254. Oncogene Techniques Edited by PETER K. VOGT AND INDER M. VERMA VOLUME 255. Small GTPases and Their Regulators (Part A: Ras Family) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL

xxxiv

Methods in Enzymology

VOLUME 256. Small GTPases and Their Regulators (Part B: Rho Family) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 257. Small GTPases and Their Regulators (Part C: Proteins Involved in Transport) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 258. Redox-Active Amino Acids in Biology Edited by JUDITH P. KLINMAN VOLUME 259. Energetics of Biological Macromolecules Edited by MICHAEL L. JOHNSON AND GARY K. ACKERS VOLUME 260. Mitochondrial Biogenesis and Genetics (Part A) Edited by GIUSEPPE M. ATTARDI AND ANNE CHOMYN VOLUME 261. Nuclear Magnetic Resonance and Nucleic Acids Edited by THOMAS L. JAMES VOLUME 262. DNA Replication Edited by JUDITH L. CAMPBELL VOLUME 263. Plasma Lipoproteins (Part C: Quantitation) Edited by WILLIAM A. BRADLEY, SANDRA H. GIANTURCO, AND JERE P. SEGREST VOLUME 264. Mitochondrial Biogenesis and Genetics (Part B) Edited by GIUSEPPE M. ATTARDI AND ANNE CHOMYN VOLUME 265. Cumulative Subject Index Volumes 228, 230–262 VOLUME 266. Computer Methods for Macromolecular Sequence Analysis Edited by RUSSELL F. DOOLITTLE VOLUME 267. Combinatorial Chemistry Edited by JOHN N. ABELSON VOLUME 268. Nitric Oxide (Part A: Sources and Detection of NO; NO Synthase) Edited by LESTER PACKER VOLUME 269. Nitric Oxide (Part B: Physiological and Pathological Processes) Edited by LESTER PACKER VOLUME 270. High Resolution Separation and Analysis of Biological Macromolecules (Part A: Fundamentals) Edited by BARRY L. KARGER AND WILLIAM S. HANCOCK VOLUME 271. High Resolution Separation and Analysis of Biological Macromolecules (Part B: Applications) Edited by BARRY L. KARGER AND WILLIAM S. HANCOCK VOLUME 272. Cytochrome P450 (Part B) Edited by ERIC F. JOHNSON AND MICHAEL R. WATERMAN VOLUME 273. RNA Polymerase and Associated Factors (Part A) Edited by SANKAR ADHYA

Methods in Enzymology

xxxv

VOLUME 274. RNA Polymerase and Associated Factors (Part B) Edited by SANKAR ADHYA VOLUME 275. Viral Polymerases and Related Proteins Edited by LAWRENCE C. KUO, DAVID B. OLSEN, AND STEVEN S. CARROLL VOLUME 276. Macromolecular Crystallography (Part A) Edited by CHARLES W. CARTER, JR., AND ROBERT M. SWEET VOLUME 277. Macromolecular Crystallography (Part B) Edited by CHARLES W. CARTER, JR., AND ROBERT M. SWEET VOLUME 278. Fluorescence Spectroscopy Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 279. Vitamins and Coenzymes (Part I) Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER VOLUME 280. Vitamins and Coenzymes (Part J) Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER VOLUME 281. Vitamins and Coenzymes (Part K) Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER VOLUME 282. Vitamins and Coenzymes (Part L) Edited by DONALD B. MCCORMICK, JOHN W. SUTTIE, AND CONRAD WAGNER VOLUME 283. Cell Cycle Control Edited by WILLIAM G. DUNPHY VOLUME 284. Lipases (Part A: Biotechnology) Edited by BYRON RUBIN AND EDWARD A. DENNIS VOLUME 285. Cumulative Subject Index Volumes 263, 264, 266–284, 286–289 VOLUME 286. Lipases (Part B: Enzyme Characterization and Utilization) Edited by BYRON RUBIN AND EDWARD A. DENNIS VOLUME 287. Chemokines Edited by RICHARD HORUK VOLUME 288. Chemokine Receptors Edited by RICHARD HORUK VOLUME 289. Solid Phase Peptide Synthesis Edited by GREGG B. FIELDS VOLUME 290. Molecular Chaperones Edited by GEORGE H. LORIMER AND THOMAS BALDWIN VOLUME 291. Caged Compounds Edited by GERARD MARRIOTT VOLUME 292. ABC Transporters: Biochemical, Cellular, and Molecular Aspects Edited by SURESH V. AMBUDKAR AND MICHAEL M. GOTTESMAN

xxxvi

Methods in Enzymology

VOLUME 293. Ion Channels (Part B) Edited by P. MICHAEL CONN VOLUME 294. Ion Channels (Part C) Edited by P. MICHAEL CONN VOLUME 295. Energetics of Biological Macromolecules (Part B) Edited by GARY K. ACKERS AND MICHAEL L. JOHNSON VOLUME 296. Neurotransmitter Transporters Edited by SUSAN G. AMARA VOLUME 297. Photosynthesis: Molecular Biology of Energy Capture Edited by LEE MCINTOSH VOLUME 298. Molecular Motors and the Cytoskeleton (Part B) Edited by RICHARD B. VALLEE VOLUME 299. Oxidants and Antioxidants (Part A) Edited by LESTER PACKER VOLUME 300. Oxidants and Antioxidants (Part B) Edited by LESTER PACKER VOLUME 301. Nitric Oxide: Biological and Antioxidant Activities (Part C) Edited by LESTER PACKER VOLUME 302. Green Fluorescent Protein Edited by P. MICHAEL CONN VOLUME 303. cDNA Preparation and Display Edited by SHERMAN M. WEISSMAN VOLUME 304. Chromatin Edited by PAUL M. WASSARMAN AND ALAN P. WOLFFE VOLUME 305. Bioluminescence and Chemiluminescence (Part C) Edited by THOMAS O. BALDWIN AND MIRIAM M. ZIEGLER VOLUME 306. Expression of Recombinant Genes in Eukaryotic Systems Edited by JOSEPH C. GLORIOSO AND MARTIN C. SCHMIDT VOLUME 307. Confocal Microscopy Edited by P. MICHAEL CONN VOLUME 308. Enzyme Kinetics and Mechanism (Part E: Energetics of Enzyme Catalysis) Edited by DANIEL L. PURICH AND VERN L. SCHRAMM VOLUME 309. Amyloid, Prions, and Other Protein Aggregates Edited by RONALD WETZEL VOLUME 310. Biofilms Edited by RON J. DOYLE

Methods in Enzymology

xxxvii

VOLUME 311. Sphingolipid Metabolism and Cell Signaling (Part A) Edited by ALFRED H. MERRILL, JR., AND YUSUF A. HANNUN VOLUME 312. Sphingolipid Metabolism and Cell Signaling (Part B) Edited by ALFRED H. MERRILL, JR., AND YUSUF A. HANNUN VOLUME 313. Antisense Technology (Part A: General Methods, Methods of Delivery, and RNA Studies) Edited by M. IAN PHILLIPS VOLUME 314. Antisense Technology (Part B: Applications) Edited by M. IAN PHILLIPS VOLUME 315. Vertebrate Phototransduction and the Visual Cycle (Part A) Edited by KRZYSZTOF PALCZEWSKI VOLUME 316. Vertebrate Phototransduction and the Visual Cycle (Part B) Edited by KRZYSZTOF PALCZEWSKI VOLUME 317. RNA–Ligand Interactions (Part A: Structural Biology Methods) Edited by DANIEL W. CELANDER AND JOHN N. ABELSON VOLUME 318. RNA–Ligand Interactions (Part B: Molecular Biology Methods) Edited by DANIEL W. CELANDER AND JOHN N. ABELSON VOLUME 319. Singlet Oxygen, UV-A, and Ozone Edited by LESTER PACKER AND HELMUT SIES VOLUME 320. Cumulative Subject Index Volumes 290–319 VOLUME 321. Numerical Computer Methods (Part C) Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND VOLUME 322. Apoptosis Edited by JOHN C. REED VOLUME 323. Energetics of Biological Macromolecules (Part C) Edited by MICHAEL L. JOHNSON AND GARY K. ACKERS VOLUME 324. Branched-Chain Amino Acids (Part B) Edited by ROBERT A. HARRIS AND JOHN R. SOKATCH VOLUME 325. Regulators and Effectors of Small GTPases (Part D: Rho Family) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 326. Applications of Chimeric Genes and Hybrid Proteins (Part A: Gene Expression and Protein Purification) Edited by JEREMY THORNER, SCOTT D. EMR, AND JOHN N. ABELSON VOLUME 327. Applications of Chimeric Genes and Hybrid Proteins (Part B: Cell Biology and Physiology) Edited by JEREMY THORNER, SCOTT D. EMR, AND JOHN N. ABELSON

xxxviii

Methods in Enzymology

VOLUME 328. Applications of Chimeric Genes and Hybrid Proteins (Part C: Protein–Protein Interactions and Genomics) Edited by JEREMY THORNER, SCOTT D. EMR, AND JOHN N. ABELSON VOLUME 329. Regulators and Effectors of Small GTPases (Part E: GTPases Involved in Vesicular Traffic) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 330. Hyperthermophilic Enzymes (Part A) Edited by MICHAEL W. W. ADAMS AND ROBERT M. KELLY VOLUME 331. Hyperthermophilic Enzymes (Part B) Edited by MICHAEL W. W. ADAMS AND ROBERT M. KELLY VOLUME 332. Regulators and Effectors of Small GTPases (Part F: Ras Family I) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 333. Regulators and Effectors of Small GTPases (Part G: Ras Family II) Edited by W. E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 334. Hyperthermophilic Enzymes (Part C) Edited by MICHAEL W. W. ADAMS AND ROBERT M. KELLY VOLUME 335. Flavonoids and Other Polyphenols Edited by LESTER PACKER VOLUME 336. Microbial Growth in Biofilms (Part A: Developmental and Molecular Biological Aspects) Edited by RON J. DOYLE VOLUME 337. Microbial Growth in Biofilms (Part B: Special Environments and Physicochemical Aspects) Edited by RON J. DOYLE VOLUME 338. Nuclear Magnetic Resonance of Biological Macromolecules (Part A) Edited by THOMAS L. JAMES, VOLKER DO¨TSCH, AND ULI SCHMITZ VOLUME 339. Nuclear Magnetic Resonance of Biological Macromolecules (Part B) Edited by THOMAS L. JAMES, VOLKER DO¨TSCH, AND ULI SCHMITZ VOLUME 340. Drug–Nucleic Acid Interactions Edited by JONATHAN B. CHAIRES AND MICHAEL J. WARING VOLUME 341. Ribonucleases (Part A) Edited by ALLEN W. NICHOLSON VOLUME 342. Ribonucleases (Part B) Edited by ALLEN W. NICHOLSON VOLUME 343. G Protein Pathways (Part A: Receptors) Edited by RAVI IYENGAR AND JOHN D. HILDEBRANDT VOLUME 344. G Protein Pathways (Part B: G Proteins and Their Regulators) Edited by RAVI IYENGAR AND JOHN D. HILDEBRANDT

Methods in Enzymology

xxxix

VOLUME 345. G Protein Pathways (Part C: Effector Mechanisms) Edited by RAVI IYENGAR AND JOHN D. HILDEBRANDT VOLUME 346. Gene Therapy Methods Edited by M. IAN PHILLIPS VOLUME 347. Protein Sensors and Reactive Oxygen Species (Part A: Selenoproteins and Thioredoxin) Edited by HELMUT SIES AND LESTER PACKER VOLUME 348. Protein Sensors and Reactive Oxygen Species (Part B: Thiol Enzymes and Proteins) Edited by HELMUT SIES AND LESTER PACKER VOLUME 349. Superoxide Dismutase Edited by LESTER PACKER VOLUME 350. Guide to Yeast Genetics and Molecular and Cell Biology (Part B) Edited by CHRISTINE GUTHRIE AND GERALD R. FINK VOLUME 351. Guide to Yeast Genetics and Molecular and Cell Biology (Part C) Edited by CHRISTINE GUTHRIE AND GERALD R. FINK VOLUME 352. Redox Cell Biology and Genetics (Part A) Edited by CHANDAN K. SEN AND LESTER PACKER VOLUME 353. Redox Cell Biology and Genetics (Part B) Edited by CHANDAN K. SEN AND LESTER PACKER VOLUME 354. Enzyme Kinetics and Mechanisms (Part F: Detection and Characterization of Enzyme Reaction Intermediates) Edited by DANIEL L. PURICH VOLUME 355. Cumulative Subject Index Volumes 321–354 VOLUME 356. Laser Capture Microscopy and Microdissection Edited by P. MICHAEL CONN VOLUME 357. Cytochrome P450, Part C Edited by ERIC F. JOHNSON AND MICHAEL R. WATERMAN VOLUME 358. Bacterial Pathogenesis (Part C: Identification, Regulation, and Function of Virulence Factors) Edited by VIRGINIA L. CLARK AND PATRIK M. BAVOIL VOLUME 359. Nitric Oxide (Part D) Edited by ENRIQUE CADENAS AND LESTER PACKER VOLUME 360. Biophotonics (Part A) Edited by GERARD MARRIOTT AND IAN PARKER VOLUME 361. Biophotonics (Part B) Edited by GERARD MARRIOTT AND IAN PARKER

xl

Methods in Enzymology

VOLUME 362. Recognition of Carbohydrates in Biological Systems (Part A) Edited by YUAN C. LEE AND REIKO T. LEE VOLUME 363. Recognition of Carbohydrates in Biological Systems (Part B) Edited by YUAN C. LEE AND REIKO T. LEE VOLUME 364. Nuclear Receptors Edited by DAVID W. RUSSELL AND DAVID J. MANGELSDORF VOLUME 365. Differentiation of Embryonic Stem Cells Edited by PAUL M. WASSAUMAN AND GORDON M. KELLER VOLUME 366. Protein Phosphatases Edited by SUSANNE KLUMPP AND JOSEF KRIEGLSTEIN VOLUME 367. Liposomes (Part A) Edited by NEJAT DU¨ZGU¨NES, VOLUME 368. Macromolecular Crystallography (Part C) Edited by CHARLES W. CARTER, JR., AND ROBERT M. SWEET VOLUME 369. Combinational Chemistry (Part B) Edited by GUILLERMO A. MORALES AND BARRY A. BUNIN VOLUME 370. RNA Polymerases and Associated Factors (Part C) Edited by SANKAR L. ADHYA AND SUSAN GARGES VOLUME 371. RNA Polymerases and Associated Factors (Part D) Edited by SANKAR L. ADHYA AND SUSAN GARGES VOLUME 372. Liposomes (Part B) Edited by NEJAT DU¨ZGU¨NES, VOLUME 373. Liposomes (Part C) Edited by NEJAT DU¨ZGU¨NES, VOLUME 374. Macromolecular Crystallography (Part D) Edited by CHARLES W. CARTER, JR., AND ROBERT W. SWEET VOLUME 375. Chromatin and Chromatin Remodeling Enzymes (Part A) Edited by C. DAVID ALLIS AND CARL WU VOLUME 376. Chromatin and Chromatin Remodeling Enzymes (Part B) Edited by C. DAVID ALLIS AND CARL WU VOLUME 377. Chromatin and Chromatin Remodeling Enzymes (Part C) Edited by C. DAVID ALLIS AND CARL WU VOLUME 378. Quinones and Quinone Enzymes (Part A) Edited by HELMUT SIES AND LESTER PACKER VOLUME 379. Energetics of Biological Macromolecules (Part D) Edited by JO M. HOLT, MICHAEL L. JOHNSON, AND GARY K. ACKERS VOLUME 380. Energetics of Biological Macromolecules (Part E) Edited by JO M. HOLT, MICHAEL L. JOHNSON, AND GARY K. ACKERS

Methods in Enzymology

xli

VOLUME 381. Oxygen Sensing Edited by CHANDAN K. SEN AND GREGG L. SEMENZA VOLUME 382. Quinones and Quinone Enzymes (Part B) Edited by HELMUT SIES AND LESTER PACKER VOLUME 383. Numerical Computer Methods (Part D) Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 384. Numerical Computer Methods (Part E) Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 385. Imaging in Biological Research (Part A) Edited by P. MICHAEL CONN VOLUME 386. Imaging in Biological Research (Part B) Edited by P. MICHAEL CONN VOLUME 387. Liposomes (Part D) Edited by NEJAT DU¨ZGU¨NES, VOLUME 388. Protein Engineering Edited by DAN E. ROBERTSON AND JOSEPH P. NOEL VOLUME 389. Regulators of G-Protein Signaling (Part A) Edited by DAVID P. SIDEROVSKI VOLUME 390. Regulators of G-Protein Signaling (Part B) Edited by DAVID P. SIDEROVSKI VOLUME 391. Liposomes (Part E) Edited by NEJAT DU¨ZGU¨NES, VOLUME 392. RNA Interference Edited by ENGELKE ROSSI VOLUME 393. Circadian Rhythms Edited by MICHAEL W. YOUNG VOLUME 394. Nuclear Magnetic Resonance of Biological Macromolecules (Part C) Edited by THOMAS L. JAMES VOLUME 395. Producing the Biochemical Data (Part B) Edited by ELIZABETH A. ZIMMER AND ERIC H. ROALSON VOLUME 396. Nitric Oxide (Part E) Edited by LESTER PACKER AND ENRIQUE CADENAS VOLUME 397. Environmental Microbiology Edited by JARED R. LEADBETTER VOLUME 398. Ubiquitin and Protein Degradation (Part A) Edited by RAYMOND J. DESHAIES VOLUME 399. Ubiquitin and Protein Degradation (Part B) Edited by RAYMOND J. DESHAIES

xlii

Methods in Enzymology

VOLUME 400. Phase II Conjugation Enzymes and Transport Systems Edited by HELMUT SIES AND LESTER PACKER VOLUME 401. Glutathione Transferases and Gamma Glutamyl Transpeptidases Edited by HELMUT SIES AND LESTER PACKER VOLUME 402. Biological Mass Spectrometry Edited by A. L. BURLINGAME VOLUME 403. GTPases Regulating Membrane Targeting and Fusion Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 404. GTPases Regulating Membrane Dynamics Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 405. Mass Spectrometry: Modified Proteins and Glycoconjugates Edited by A. L. BURLINGAME VOLUME 406. Regulators and Effectors of Small GTPases: Rho Family Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 407. Regulators and Effectors of Small GTPases: Ras Family Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 408. DNA Repair (Part A) Edited by JUDITH L. CAMPBELL AND PAUL MODRICH VOLUME 409. DNA Repair (Part B) Edited by JUDITH L. CAMPBELL AND PAUL MODRICH VOLUME 410. DNA Microarrays (Part A: Array Platforms and Web-Bench Protocols) Edited by ALAN KIMMEL AND BRIAN OLIVER VOLUME 411. DNA Microarrays (Part B: Databases and Statistics) Edited by ALAN KIMMEL AND BRIAN OLIVER VOLUME 412. Amyloid, Prions, and Other Protein Aggregates (Part B) Edited by INDU KHETERPAL AND RONALD WETZEL VOLUME 413. Amyloid, Prions, and Other Protein Aggregates (Part C) Edited by INDU KHETERPAL AND RONALD WETZEL VOLUME 414. Measuring Biological Responses with Automated Microscopy Edited by JAMES INGLESE VOLUME 415. Glycobiology Edited by MINORU FUKUDA VOLUME 416. Glycomics Edited by MINORU FUKUDA VOLUME 417. Functional Glycomics Edited by MINORU FUKUDA

Methods in Enzymology

xliii

VOLUME 418. Embryonic Stem Cells Edited by IRINA KLIMANSKAYA AND ROBERT LANZA VOLUME 419. Adult Stem Cells Edited by IRINA KLIMANSKAYA AND ROBERT LANZA VOLUME 420. Stem Cell Tools and Other Experimental Protocols Edited by IRINA KLIMANSKAYA AND ROBERT LANZA VOLUME 421. Advanced Bacterial Genetics: Use of Transposons and Phage for Genomic Engineering Edited by KELLY T. HUGHES VOLUME 422. Two-Component Signaling Systems, Part A Edited by MELVIN I. SIMON, BRIAN R. CRANE, AND ALEXANDRINE CRANE VOLUME 423. Two-Component Signaling Systems, Part B Edited by MELVIN I. SIMON, BRIAN R. CRANE, AND ALEXANDRINE CRANE VOLUME 424. RNA Editing Edited by JONATHA M. GOTT VOLUME 425. RNA Modification Edited by JONATHA M. GOTT VOLUME 426. Integrins Edited by DAVID CHERESH VOLUME 427. MicroRNA Methods Edited by JOHN J. ROSSI VOLUME 428. Osmosensing and Osmosignaling Edited by HELMUT SIES AND DIETER HAUSSINGER VOLUME 429. Translation Initiation: Extract Systems and Molecular Genetics Edited by JON LORSCH VOLUME 430. Translation Initiation: Reconstituted Systems and Biophysical Methods Edited by JON LORSCH VOLUME 431. Translation Initiation: Cell Biology, High-Throughput and Chemical-Based Approaches Edited by JON LORSCH VOLUME 432. Lipidomics and Bioactive Lipids: Mass-Spectrometry–Based Lipid Analysis Edited by H. ALEX BROWN VOLUME 433. Lipidomics and Bioactive Lipids: Specialized Analytical Methods and Lipids in Disease Edited by H. ALEX BROWN

xliv

Methods in Enzymology

VOLUME 434. Lipidomics and Bioactive Lipids: Lipids and Cell Signaling Edited by H. ALEX BROWN VOLUME 435. Oxygen Biology and Hypoxia Edited by HELMUT SIES AND BERNHARD BRU¨NE VOLUME 436. Globins and Other Nitric Oxide-Reactive Protiens (Part A) Edited by ROBERT K. POOLE VOLUME 437. Globins and Other Nitric Oxide-Reactive Protiens (Part B) Edited by ROBERT K. POOLE VOLUME 438. Small GTPases in Disease (Part A) Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 439. Small GTPases in Disease (Part B) Edited by WILLIAM E. BALCH, CHANNING J. DER, AND ALAN HALL VOLUME 440. Nitric Oxide, Part F Oxidative and Nitrosative Stress in Redox Regulation of Cell Signaling Edited by ENRIQUE CADENAS AND LESTER PACKER VOLUME 441. Nitric Oxide, Part G Oxidative and Nitrosative Stress in Redox Regulation of Cell Signaling Edited by ENRIQUE CADENAS AND LESTER PACKER VOLUME 442. Programmed Cell Death, General Principles for Studying Cell Death (Part A) Edited by ROYA KHOSRAVI-FAR, ZAHRA ZAKERI, RICHARD A. LOCKSHIN, AND MAURO PIACENTINI VOLUME 443. Angiogenesis: In Vitro Systems Edited by DAVID A. CHERESH VOLUME 444. Angiogenesis: In Vivo Systems (Part A) Edited by DAVID A. CHERESH VOLUME 445. Angiogenesis: In Vivo Systems (Part B) Edited by DAVID A. CHERESH VOLUME 446. Programmed Cell Death, The Biology and Therapeutic Implications of Cell Death (Part B) Edited by ROYA KHOSRAVI-FAR, ZAHRA ZAKERI, RICHARD A. LOCKSHIN, AND MAURO PIACENTINI VOLUME 447. RNA Turnover in Bacteria, Archaea and Organelles Edited by LYNNE E. MAQUAT AND CECILIA M. ARRAIANO VOLUME 448. RNA Turnover in Eukaryotes: Nucleases, Pathways and Analysis of mRNA Decay Edited by LYNNE E. MAQUAT AND MEGERDITCH KILEDJIAN

Methods in Enzymology

xlv

VOLUME 449. RNA Turnover in Eukaryotes: Analysis of Specialized and Quality Control RNA Decay Pathways Edited by LYNNE E. MAQUAT AND MEGERDITCH KILEDJIAN VOLUME 450. Fluorescence Spectroscopy Edited by LUDWIG BRAND AND MICHAEL L. JOHNSON VOLUME 451. Autophagy: Lower Eukaryotes and Non-Mammalian Systems (Part A) Edited by DANIEL J. KLIONSKY VOLUME 452. Autophagy in Mammalian Systems (Part B) Edited by DANIEL J. KLIONSKY VOLUME 453. Autophagy in Disease and Clinical Applications (Part C) Edited by DANIEL J. KLIONSKY VOLUME 454. Computer Methods (Part A) Edited by MICHAEL L. JOHNSON AND LUDWIG BRAND VOLUME 455. Biothermodynamics (Part A) Edited by MICHAEL L. JOHNSON, JO M. HOLT, AND GARY K. ACKERS (RETIRED) VOLUME 456. Mitochondrial Function, Part A: Mitochondrial Electron Transport Complexes and Reactive Oxygen Species Edited by WILLIAM S. ALLISON AND IMMO E. SCHEFFLER VOLUME 457. Mitochondrial Function, Part B: Mitochondrial Protein Kinases, Protein Phosphatases and Mitochondrial Diseases Edited by WILLIAM S. ALLISON AND ANNE N. MURPHY VOLUME 458. Complex Enzymes in Microbial Natural Product Biosynthesis, Part A: Overview Articles and Peptides Edited by DAVID A. HOPWOOD VOLUME 459. Complex Enzymes in Microbial Natural Product Biosynthesis, Part B: Polyketides, Aminocoumarins and Carbohydrates Edited by DAVID A. HOPWOOD VOLUME 460. Chemokines, Part A Edited by TRACY M. HANDEL AND DAMON J. HAMEL VOLUME 461. Chemokines, Part B Edited by TRACY M. HANDEL AND DAMON J. HAMEL VOLUME 462. Non-Natural Amino Acids Edited by TOM W. MUIR AND JOHN N. ABELSON VOLUME 463. Guide to Protein Purification, 2nd Edition Edited by RICHARD R. BURGESS AND MURRAY P. DEUTSCHER VOLUME 464. Liposomes, Part F Edited by NEJAT DU¨ZGU¨NES,

This page intentionally left blank

S E C T I O N

O N E

BIOACTIVE LIPOSOMES

This page intentionally left blank

C H A P T E R

O N E

Tubular Liposomes with Variable Permeability for Reconstitution of FtsZ Rings Masaki Osawa and Harold P. Erickson Contents 1. 2. 3. 4.

Introduction Reagents Bacterial Expression of Membrane Targeting FtsZ Purification of FtsZ-mts and FtsZ-YFP-mts 4.1. For FtsZ-mts 4.2. For FtsZ-YFP-mts 5. Renatured Preparation of FtsZ-YFP-mts 6. Tubular Multilamellar Liposome Preparation 7. Permeability of the Multilamellar Liposomes 8. Z-ring Formation in Liposomes 9. A Crude Flow Chamber to Exchange Buffer Outside Liposomes 10. Factors Affecting Z-ring Formation in Liposomes 11. Utility of the Liposomes Beyond FtsZ References

4 4 5 5 6 6 7 7 9 11 11 14 16 16

Abstract We have developed a system for producing tubular multilamellar liposomes that incorporate the protein FtsZ on the inside. We start with a mixture of spherical multilamellar liposomes with FtsZ initially on the outside. Shearing forces generated by applying a coverslip most likely distort some of the spherical liposomes into a tubular shape, and causes some to leak and incorporate FtsZ inside. We describe protocols for liposome preparation, and for preparing membrane-targeted FtsZ that can assemble contractile Z rings inside the tubular liposomes. We also describe the characterization of the multilamellar liposomes in terms of the permeability or leakiness for a small fluorescent dye and larger protein molecules. These liposomes may be useful for reconstitution of other biological systems. Department of Cell Biology, Duke University Medical Center, Durham, North Carolina, USA Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64001-5

#

2009 Elsevier Inc. All rights reserved.

3

4

Masaki Osawa and Harold P. Erickson

1. Introduction FtsZ is a bacterial tubulin homologue that forms a ring structure called the ‘‘Z ring’’ at the division plane in bacteria. The Z ring is anchored to the membrane and constricts to divide the bacteria. FtsZ recruits a dozen other essential division proteins, which are mostly involved in remodeling the peptidoglycan cell wall. We recently succeeded in reconstituting Z rings inside tubular liposomes, and found that they generated a constriction force on the liposome wall (Osawa et al., 2008). The assembly of Z rings and the generation of the constriction force were achieved with FtsZ alone, and did not require any other division protein. This was an important discovery itself for understanding the mechanism of bacterial cell division. Now the liposome system we developed should provide a simple in vitro system for studying molecular details of how FtsZ works. To achieve these results we had to overcome two technical problems. The first problem was to tether FtsZ to the membrane. Pichoff and Lutkenhaus (2005) discovered that the carboxy terminus of FtsZ binds to FtsA, and FtsA has an amphipathic helix at its carboxy terminal that inserts into the membrane. We made an FtsZ that could tether itself to the membrane by fusing an amphipathic helix (membrane targeting sequence: mts) to the carboxy terminus of FtsZ. To visualize the protein, we inserted a yellow fluorescent protein (YFP) before the mts, giving FtsZ-YFP-mts. Here we provide detailed protocols for the purification of FtsZ-YFP-mts. The second problem was how to get the protein inside liposomes. We have succeeded in getting FtsZ-YFP-mts inside spherical unilamellar liposomes using the emulsion method (Noireaux and Libchaber, 2004; Pautot et al., 2003). However, we have never found Z rings assembled in such spherical unilamellar liposomes. Eventually, we discovered a procedure that produced tubular multilamellar liposomes, and incorporated FtsZ-YFP-mts inside, where it formed Z rings. Initially this was a fortunate accident, since the cylindrical geometry was not designed, and the FtsZ was initially on the outside. We have since refined the procedures for producing tubular multilamellar liposomes, and we now understand some aspects of the permeability or leakiness that lets FtsZ inside. We describe here our protocols for producing the tubular liposomes and the tests of permeability.

2. Reagents The following reagents are used in our experiments: 

Column buffer: 50 mM Tris/HCl, pH 7.9, 50 mM KCl, 1 mM EDTA, 10% (v/v) glycerol

FtsZ in Tubular Liposomes

     

5

HMKCG buffer: 50 mM HEPES/KOH, pH 7.7, 5 mM MgAc, 300 mM KAc, 50 mM KCl 10% (v/v) glycerol HMK50-350 buffer: 50 mM HEPES/KOH, pH 7.7, 5 mM MgAc, 50–350 mM KAc DOPG:1,2-dioleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)] (Avanti) Egg PC: phosphatidylcholine (Avanti) HccA: 7-Hydroxycoumarin-3-carboxylic acid (Invitrogen) Teflon disc, 37 mm diameter

3. Bacterial Expression of Membrane Targeting FtsZ The FtsZ-YFP-mts is expressed from a pET-11b expression vector, with FtsZ366-YFP-mts or FtsZ366-mts genes inserted at NdeI/BamHI sites (366 indicates that the FtsZ was truncated there, removing the FtsA-binding C-terminal peptide). The YFP we use is the variety Venus (Nagai et al., 2002), which gave superior results in FtsZ fusions in E. coli (Osawa and Erickson, 2005). The mts used here is the amphipathic helix from E. coli MinD (Szeto et al., 2003). We have not yet tested the amphipathic helix from FtsA, which has three to five additional extra amino acids that extend the amphipathic helix (Pichoff and Lutkenhaus, 2005). The expression vector is transformed into E. coli strain C41 (Miroux and Walker, 1996), which gives better yields of soluble proteins than BL21. After transforming, colonies are selected on an LB (Luria broth) agar plate containing 100 mg/ml ampicillin. A colony is picked and cultured overnight in 50 ml LB media with 100 mg/ml ampicillin at 37  C. Five milliliters of the overnight culture is diluted in 500 ml LB and cultured at 37  C until the optical density at 600 nm reaches 0.8–1.0. Protein expression is induced by addition of 0.5 mM IPTG and at the same time the temperature of the shaker is set to 20  C (our shaker takes 1–2 h to reach 20  C). The cells are cultured overnight and spun down at 3750 rpm for 45 min in a Beckman GPR rotor.

4. Purification of FtsZ-mts and FtsZ-YFP-mts Since FtsZ-mts and FtsZ-YFP-mts are expressed as soluble proteins, we purify them using the same protocol as for wild-type FtsZ. The packed cells are resuspended in a final volume of 20 ml column buffer, and 1 mM phenylmethanesulphonylfluoride (PMSF) and 0.1–0.2 mg/ml

6

Masaki Osawa and Harold P. Erickson

lysozyme are added. The mixture is then incubated on a rotator at 4  C for 15 min. They are frozen at 80  C overnight or longer. Two cycles of freeze–thaw (fresh 1 mM PMSF is added after each thawing) method are performed. The resultant mixture is sonicated on ice until the viscosity is reduced. We usually sonicate it for three cycles of 20 s, with 1 min cooling intervals. It is then centrifuged at 32,000 rpm for 20 min at 4  C (Beckman 42.1 Ti rotor). The supernatant is collected and ammonium sulfate is added to 30% saturation (3.52 g dry ammonium sulfate to the 20 ml volume). This mixture is incubated for 20 min on ice and again centrifuged at 32,000 rpm for 20 min at 4  C (Beckman 42.1 Ti rotor). The supernatant is discarded and the pellet is resuspended in 10 ml column buffer and passed through a 0.22 mm filter. The protein is purified on an anion exchange column. A 1  10 cm Source Q column (Source 15Q, GE Healthcare) is used. The column is eluted with a 100 ml gradient from 50–500 mM KCl in column buffer.

4.1. For FtsZ-mts FtsZ has very low UV absorbance, so the peak is located by running each fraction on SDS–PAGE. The peak fractions are pooled and dialyzed into HMK350. The protein concentration is determined by the BCA method (Pierce). FtsZ produces 75% as much color as BSA (Lu et al., 1998), so it is necessary to correct for this. Aliquots are frozen and stored at 80  C.

4.2. For FtsZ-YFP-mts After elution from the Source 15Q column, the peak fractions are pooled. There are typically two peaks: a large main peak and a following small peak, and both peaks have an indistinguishable activity. These peaks can be identified by yellow fluorescence and confirmed by SDS–PAGE. They are concentrated using an Amicon Ultra-15 with centrifugation at 5000g. We have noted that incomplete boiling of FtsZ-YFP-MTS with SDS sample buffer generates two bands on the gel. The upper band (68 Kd) results from completely denatured protein and the lower band (60 Kd), which still has yellow fluorescence in the gel, is due to FtsZ-YFP-mts where the YFP is not denatured. The concentration of FtsZ-YFP-mts can be determined from its absorption at 515 nm, using the extinction coefficient for YFP-Venus 92,200 M 1 cm 1.

FtsZ in Tubular Liposomes

7

Our preferred buffer for FtsZ-YFP-mts is now HMKCG because FtsZ-YFP-mts seems to be more stable, as described below; we now use HMKCG for dialysis, dilution, reaction, and storage buffer.

5. Renatured Preparation of FtsZ-YFP-mts In our previous study (Osawa et al., 2008), we used a renaturing technique to prepare FtsZ366-YFP-mts. We developed this protocol because an early preparation of soluble protein as described above had no activity. We recently found that the soluble protein has full activity, equal to the best fractions of the renatured preparation. Because of the simplicity and much higher yield, we now use the soluble preparation for all work. One curious observation with the renatured FtsZ-YFP-mts was that it lost activity when dialyzed into HMK350. We found that addition of 10% glycerol and 50 mM chloride ion to HMK350 would preserve the activity of the renatured FtsZ-YFP-mts during dialysis. Although this precaution may only be necessary for the renatured FtsZ-YFP-mts, our preferred buffer is now HMKCG, which contains this glycerol and chloride ion.

6. Tubular Multilamellar Liposome Preparation PC and DOPG are dissolved separately in methanol at 100 mg/ml and mixed at a 4:1 ratio (20 ml/5 ml) in a 1.5 ml Eppendorf tube. The mixture is dried with an air current. 250 ml of milliQ water is added to the Eppendorf tube, and the dried lipid is suspended with vigorous vortexing. Many drops (5 ml each, total 250 ml) of the suspension are placed on a 37 mm diameter Teflon disc (Fig. 1.1A) and dried using an air current (Fig. 1.1B). The Teflon disc is placed in a beaker slightly larger than its diameter, and covered with 5 ml reaction buffer. It is then incubated at 37  C overnight to hydrate lipid (Fig. 1.1C). High salt buffers such as HMKCG or HMK350 will give aggregated multilamellar liposomes, which are desired to make tubular liposomes for Z-ring formation. The beaker is gently agitated to generate liposomes (Fig. 1.1D). Too vigorous agitation generates many small liposomes which interfere with observation of Z rings. About 1 ml of the suspension is pipetted out closest to the Teflon, which contains the most concentrated liposomes. This is placed in a 1.5 ml Eppendorf tube and left on a bench for 1 h (Fig. 1.1E). The multilamellar liposomes in HMKCG float to the surface. To obtain concentrated

8

Masaki Osawa and Harold P. Erickson

A

B

C

Liposome suspension (less than 5 ml)

Teflon disc D

Reaction buffer (HMKCG) E

Multilamellar liposome

F

HMKCG

Figure 1.1 Schematic illustration of the production of multilamellar liposomes. (A) A 37 mm Teflon disc is placed in a small beaker and 250 ml total aqueous suspension of lipid is deposited as many small drops (each drop is less than 5 ml). (B) The drops are dried with an air current and (C) the Teflon disc is covered with 5 ml of HMKCG and incubated overnight at 37  C. (D) The Teflon disc is gently agitated and multilamellar liposomes floated off. (E) One microliter of the most concentrated suspension, nearest the Teflon, is transferred to an Eppendorf tube and left on the bench for 1 h. The liposomes rise to the surface. (F) HMKCG is carefully removed from the bottom, leaving a concentrated suspension of liposomes.

liposomes, buffer from the bottom of the tube is carefully pipetted out, leaving the concentrated top layer (Fig. 1.1F). Liposomes that have a thinner wall (including unilamellar liposomes) can be easily obtained by using HMK100 with the same methods. If the KAc is reduced to 50 mM or less, many unilamellar liposomes will be obtained. Interestingly, liposomes prepared in HMK100 settle to the bottom of the Eppendorf tube over several hours. This is the opposite of multilamellar liposomes in HMKCG, which float to the top (probably because the 10% glycerol increases the density of the solution). To produce tubular liposomes, 5 ml of the aggregated multilamellar liposomes in HMKCG or HMK350 is placed on a glass slide and a coverslip is applied. To make liposomes with Z rings, liposomes are mixed first with 1 mM GTP and 4 mM FtsZ-YFP-mts. In this procedure, tubular liposomes are formed, probably generated by shear force when the suspension is spreading quickly through the narrow space between glasses. The tubular liposomes are always a minority, but with practice they can be found reproducibly. The tubular liposomes are

FtsZ in Tubular Liposomes

9

more abundant near the edges of the coverslip, perhaps because that is where shear is maximized. We have noticed that the fluorescence background is higher in the center, where the drop is initially deposited. This may reflect FtsZ-YFP-mts binding to the glass. We have not tested whether the concentration of soluble FtsZ-YFP-mts is substantially reduced by binding to the glass.

7. Permeability of the Multilamellar Liposomes To check the permeability of the multilamellar tubular liposomes that were produced by the above procedures, a polar fluorescence dye HccA is used as a small molecule indicator (MW 206.15 kDa) and FtsZ fused with YFP-Venus (FtsZ-YFP) as a large molecule indicator (MW 67 kDa). Neither of these probes can cross the lipid bilayer, so they should report on smaller or larger pores in the liposomes. Also, the FtsZ-YFP is missing the mts, and was given no GTP, so it would not make protofilaments or Z rings. It should just report the existence of pores large enough to admit the globular protein. When we mix the multilamellar liposomes with 1 mM HccA and 32 mM FtsZ-YFP and apply a coverslip, we find tubular liposomes with both HccA and YFP fluorescence. More than 80% of tubular liposomes have FtsZ-YFP inside and even more liposomes have HccA inside, indicating that multilamellar liposomes are leaky at least for some time after addition of the fluorescent probes and application of the coverslip. A simple FRAP (fluorescent recovery after photobleach) assay is used to determine the leakiness of liposomes 30 min after the coverslip is applied. The entire field of view is exposed to UV from the mercury lamp for 20 s, without any filter. This results in complete bleaching of both fluorescent molecules inside and outside the liposomes. HccA fluorescence on the outside recovers over 20–60 s as unbleached probe diffuses into the area from outside. At 20 s HccA has entered about half of the liposomes (Fig. 1.2F), and by 5 min almost all the liposomes contain fluorescent HccA. Diffusion of FtsZ-YFP is slower, requiring 5 min to recover 70% of original fluorescence on the outside. After 45 min about half of the liposomes have YFP fluorescence inside. We conclude that the multilamellar liposomes are almost all permeable or leaky to the small molecule probe, and about half have pores that will pass the larger protein probe. A curious phenomenon is the appearance of a bright ring on the inner layer of the liposome when the inside of the liposome is dark and the outside is bright (Fig. 1.2D and F). When we repeat the FRAP experiment with a confocal microscope, the bright ring is seen only rarely and is also very dim. We suggest that the bright ring may be an optical artifact seen in the wide field fluorescence microscopy.

10

Masaki Osawa and Harold P. Erickson

A pre

B 1 min

C

D 45 min

5 min

FtsZ-YFP

1 2 3

E pre

H 45 min

F 20 sec 1

HccA

2

G 5 min

3

I pre

J 45 min

DIC

20 mm

Figure 1.2 FRAP assay for checking permeability of multilamellar liposomes. The top row shows FtsZ-YFP (lacking the mts and GTP), the middle row shows HccA fluorescence, and the bottom row shows DIC. The first column shows prebleach images, and the numbers on the other panels show time after bleach. In (D), vesicles 1–3 show little or no recovery of FtsZ-YFP inside, but they do show a thin bright layer of internal fluorescence. The vesicle between 1 and 2 shows substantial recovery of FtsZ-YFP. In (F), (G) and (H) vesicles 1–3 show no recovery of HccA at 20 s, but full recovery at 5 min. The exposure of fluorescence images is kept at the same level from prebleach to postbleach. Notice also the change in shape of several vesicles over 45 min. (B) 50% magnification showing that fluorescence is completely bleached.

FtsZ in Tubular Liposomes

11

8. Z-ring Formation in Liposomes We have shown previously that FtsZ-YFP-mts can form Z rings inside tubular liposomes when FtsZ-YFP-mts plus GTP was mixed with multilamellar liposomes just before applying the coverslip (Osawa et al., 2008). To examine the leakiness of these liposomes with Z rings inside, we repeated the FRAP assay using 8 mM FtsZ-YFP-mts plus 2 mM GTP instead of 32 mM FtsZ-YFP. Most liposomes that had Z rings before the bleach showed no recovery of fluorescent Z rings, even though HccA fluorescence recovered rapidly (Fig. 1.3). Fluorescent FtsZ-YFP-mts could be seen coating the outside of the liposome but apparently did not enter the liposome, since no fluorescent Z rings were seen. We did, however, find some liposomes that recovered fluorescent Z rings after the bleach (Fig. 1.4). Figure 1.4C shows an intriguing example where HccA fluorescence invades through a hole at the top side of the tubular liposome (arrow in Fig. 1.4F). FtsZ-YFP-mts followed and fluorescent Z rings assembled as the nonbleached FtsZ-YFP-mts exchanged with the bleached (Fig. 1.4C and D), showing that occasional open liposomes can recover FtsZ from the outside and assemble Z rings. Recovery of Z rings with FtsZ-YFP-mts appeared to be less frequent than recovery of fluorescence of FtsZ-YFP. A potentially important difference is that FtsZ-YFPmts had GTP, so it should be mostly assembled into protofilaments 30 subunits long (Chen and Erickson, 2005). FtsZ-YFP was used without GTP so it should be much smaller protein monomers.

9. A Crude Flow Chamber to Exchange Buffer Outside Liposomes For various purposes one might want to change the buffer outside the liposomes, especially knowing that most of the liposomes are permeable to small molecules. The best production of tubular liposomes occurs with a small sample volume, which generates an optimal shear when the coverslip is applied. A 5 ml sample spread over a 4.8 cm2 coverslip gives a liquid layer 10 mm thick. If we try to perfuse buffer into this thin layer the liposomes are subject to high shear that results in severe elongation, loss of Z rings, and loss of liposomes. To minimize these effects, we first apply a 50 ml drop of perfusion buffer to one edge of the coverslip (Fig. 1.5B). We can then drain the liquid by applying a Kimwipe at the opposite edge, while applying a pipette tip or thin rod to prevent the coverslip from slipping. The arrangement shown in Fig. 1.5D, with a thin strip of Kimwipe applied to the coverslip and a large piece farther away, provides an optimal slow flow.

12

Masaki Osawa and Harold P. Erickson

B 1 min

C 12 min

D 45 min

E pre

F 20 sec

G 12 min

F 45 min

HccA

FtsZ-YFP

A pre

20 mm

Figure 1.3 A liposome containing multiple Z rings (arrowheads) before bleach shows recovery of HccA fluorescence both outside and inside. FtsZ-YFP-mts fluorescence is recovered on the outside surface of the liposome, but no fluorescent Z rings recover inside. The exposure of fluorescence images is kept at the same level from prebleach to postbleach. (D) We did not detect any fluorescent Z rings even when we optimized the contrast in this specific panel.

We tested the system by adding FtsZ-YFP-mts without GTP to multilamellar liposomes. Cylindrical liposomes were found with the protein inside but Z rings could not form without GTP. We then perfused HMKCG containing 1 mM GTP. As the GTP entered the liposomes, Z rings assembled (Fig. 1.5E). An additional advantage of this system is that the fluorescent protein outside the liposome is washed away during the perfusion. This removes the fluorescent protein in solution and binding to the outside surface of liposomes and enhances the contrast of the Z rings inside. Because there is no FtsZ outside to hydrolyze the GTP, the Z rings can last a long time.

13

FtsZ in Tubular Liposomes

A

B

C 1 min

D 12 min

45 min

FtsZ-YFP

pre

E

F

G 20 sec

H 12 min

45 min

HccA

pre

20 mm

Figure 1.4 A liposome with Z rings inside shows permeability to both FtsZ-YFP-mts and to HccA. (A) Arrowheads indicate Z rings prebleach. (B) YFP fluorescence is completely eliminated on the outside and inside of the liposome at 1 min after bleach. (C) About 12 min later YFP fluorescence has partially recovered on the outside, and fluorescent Z rings have formed on the inside, showing that unbleached FtsZ-YFP-mts has entered the leaky liposome. (D) At 45 min fluorescent Z rings have spread more toward the bottom. (E, F) About 20 s after bleaching, HccA has begun to enter the liposome, apparently through a hole at the top (note the zone of brighter fluorescence in the top 1/5 of the liposome). (G, H) At 12 min HccA fluorescence has fully recovered inside compared to the prebleached image in (E). Note that the FtsZ rings recover only at the top, near the hole, at 12 min. At 45 min they have spread through the top half of the liposome.

14

Masaki Osawa and Harold P. Erickson

A

E

Pipette

B

Inserted solution lifts cover glass up C

D

Thin rod

Kimwipe

5 mm

Figure 1.5 Changing buffer by flow/perfusion. (A) The normal preparation has a very thin layer of liquid between the slide and coverslip. (B) When a drop of the new buffer is placed on the side, the coverslip rises and is able to move back and forth. (C) To adsorb the liquid, the slide is tilted and a small piece of Kimwipe is placed on the edge opposite the drop. To prevent the coverslip from moving we used a thin rod or a pipette. (D) To minimize shear we slow the draining by placing a narrow piece of Kimwipe touching the coverslip and a larger piece farther away. (E) As a demonstration of buffer exchange, we prepared liposomes with FtsZ-YFP-mts but no GTP. FtsZ was trapped inside some tubular liposomes but did not form Z rings. We then flowed through a buffer containing GTP but no FtsZ. The GTP entered the liposome and initiated Z-ring assembly. This liposome was apparently leaky to GTP but not to protein, since the FtsZ did not leak out.

10. Factors Affecting Z-ring Formation in Liposomes During months of experimenting with the liposome system, we have made a number of observations that seem to be important for Z-ring formation. 

The shape of the liposome is very important, the cylindrical geometry being optimal for Z-ring assembly. We have rarely seen Z rings inside spherical liposomes. When these have internalized FtsZ-YFP-mts, the protein localizes to small patches or forms arcs that move unstably

15

FtsZ in Tubular Liposomes

(Fig. 1.6B). One possible interpretation is that when a Z ring forms inside a spherical liposome, the contractile force causes it to collapse to a small patch. Another interpretation is that FtsZ filaments cannot determine the direction to form stable Z rings. We have sometimes observed this situation in which a large round liposome had radially formed FtsZ filaments inside (Fig. 1.6B, liposome on right). With a cylindrical geometry, a contractile force will cause it to form a circle in a plane perpendicular to the axis.  The diameter of the liposome is important. Z-ring assembly is optimal in tubular liposomes less than 2 mm in diameter (bacteria are typically 1 mm in diameter). In larger diameter tubular liposomes, some Z rings are

A

B

*2

*2

*2 *2 *3 *1

*2

10 mm

*1

Figure 1.6 Behavior of FtsZ-YFP-mts in large multilamellar liposomes. (A) In a large tubular liposome (3 mm inside diameter), FtsZ-YFP-mts formed some normal Z rings indicated by two spots of equal intensity directly across from each other (labeled *1). However, many spirals were also formed, indicated by dots spaced irregularly (*2). Note the bright ring at (*3), which appears to have peeled away the inner layer of the multilamellar wall (DIC image on right). Figure 1.6A is reprinted from Osawa et al. (2008) with permission of the publisher. (B) Several incomplete Z rings, spirals or arc structures, (arrowheads) are formed in a large spherical liposome.

16

Masaki Osawa and Harold P. Erickson

formed, but many structures appear to be helices or spirals rather than closed rings (Fig. 1.6A).  A mixture of neutral (PC) and negatively charged (DOPG) lipids is important. We have obtained Z rings with 2–40% DOPG, but typically use 1:4 DOPG:PC. 100% PC or DOPG, or a 2:3 ratio of DOPG:PC failed to support Z-ring assembly. This is probably because the amphipathic helix has positively charged amino acids on the hydrophilic side, and these need some negatively charged phospholipids in the bilayer to bind to. Interestingly, tubular liposomes produced by E. coli polar lipid (Avanti) did not support Z rings. This particular batch of E. coli lipid that we used may have contained more than 40% of charged lipid.  We have made several attempts to assemble Z rings in unilamellar liposomes, without success. As mentioned above, the spherical shape of unilamellar liposomes may not support Z-ring assembly. Another factor may be the rigidity of the wall. The rigidity of the multilamellar wall may slow constriction and stabilize the Z rings.

11. Utility of the Liposomes Beyond FtsZ The liposome system that we have developed may have uses beyond the study of FtsZ. The tubular liposomes are similar in size to a bacterium, and can range up to the size of a yeast cell. Similar to our reconstitution of Z rings, they may eventually be useful in reconstituting the cytokinetic apparatus of yeast or larger animal cells. More generally, we have demonstrated that most of the liposomes are permeable or leaky to small molecules, while about half are leaky to larger proteins. They therefore constitute small femtoliter chambers in which one can enclose proteins and manipulate the small molecule environment.

REFERENCES Chen, Y., and Erickson, H. P. (2005). Rapid in vitro assembly dynamics and subunit turnover of FtsZ demonstrated by fluorescence resonance energy transfer. J. Biol. Chem. 280, 22549–22554. Lu, C., Stricker, J., and Erickson, H. P. (1998). FtsZ from Escherichia coli, Azotobacter vinelandii, and Thermotoga maritima—Quantitation, GTP hydrolysis, and assembly. Cell Motil. Cytoskel. 40, 71–86. Miroux, B., and Walker, J. E. (1996). Over-production of proteins in Escherichia coli: Mutant hosts that allow synthesis of some membrane proteins and globular proteins at high levels. J. Mol. Biol. 260, 289–298. Nagai, T., Ibata, K., Park, E. S., Kubota, M., Mikoshiba, K., and Miyawaki, A. (2002). A variant of yellow fluorescent protein with fast and efficient maturation for cellbiological applications. Nat. Biotechnol. 20, 87–90.

FtsZ in Tubular Liposomes

17

Noireaux, V., and Libchaber, A. (2004). A vesicle bioreactor as a step toward an artificial cell assembly. Proc. Natl. Acad. Sci. USA 101, 17669–17674. Osawa, M., and Erickson, H. P. (2005). Probing the domain structure of FtsZ by random truncation and insertion of GFP. Microbiology 151, 4033–4043. Osawa, M., Anderson, D. E., and Erickson, H. P. (2008). Reconstitution of contractile FtsZ rings in liposomes. Science 320, 792–794. Pautot, S., Frisken, B. J., and Weitz, D. A. (2003). Engineering asymmetric vesicles. Proc. Natl. Acad. Sci. USA 100, 10718–10721. Pichoff, S., and Lutkenhaus, J. (2005). Tethering the Z ring to the membrane through a conserved membrane targeting sequence in FtsA. Mol. Microbiol. 55, 1722–1734. Szeto, T. H., Rowland, S. L., Habrukowich, C. L., and King, G. F. (2003). The MinD membrane targeting sequence is a transplantable lipid-binding helix. J. Biol. Chem. 278, 40050–40056.

This page intentionally left blank

C H A P T E R

T W O

Detection and Analysis of Protein Synthesis and RNA Replication in Giant Liposomes Takeshi Sunami,† Hiroshi Kita,† Kazufumi Hosoda,* Tomoaki Matsuura,* Hiroaki Suzuki,* and Tetsuya Yomo*,†,‡ Contents 1. Introduction 2. Methods 2.1. Liposome preparation 2.2. Internal protein synthesis reactions followed by the reaction catalyzed by the synthesized proteins 2.3. Detecting reactions in liposomes by fluorescence-activated cell sorting 3. Analysis of the FACS Data 4. Conclusions Acknowledgments References

20 21 21 22 25 27 28 29 29

Abstract In living cells, biochemical reaction systems are enclosed in small lipidic compartments. To experimentally simulate various biochemical reactions occurring in extant cells, giant liposomes are used to reconstruct an artificial model cell. We present methods for conducting a protein synthesis reaction, followed by the reaction catalyzed by the synthesized proteins inside liposomes, and for measurement of the in liposome reaction using a fluorescence-activated cell sorter (FACS). These techniques enable us to perform detailed analysis of the biochemical reactions occurring in the microcompartments, and have the potential to reveal the role of compartmentalization in cellular systems.

* {

{

Department of Bioinformatics Engineering, Graduate School of Information Science and Technology, Osaka University, Suita, Osaka, Japan Exploratory Research for Advanced Technology (ERATO), Japan Science and Technology Agency ( JST), Chiyoda-ku, Tokyo, Japan Graduate School of Frontier Biosciences, Osaka University, Suita, Osaka, Japan

Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64002-7

#

2009 Elsevier Inc. All rights reserved.

19

20

Takeshi Sunami et al.

1. Introduction Liposomes are small vesicles composed of a lipid bilayer membrane, which are widely used in both basic and applied research in life sciences (Luisi and Walde, 2000; Torchilin and Weissig, 2003). Especially, giant liposomes (diameter > 1 mm) are of particular interest as a cell model and a microreactor containing biochemical reaction systems in a cell-sized volume (Hanczyc et al., 2003; Kaneko et al., 1998; Noireaux and Libchaber, 2004; Nomura et al., 2003; Tsumoto et al., 2001). Reconstruction of such a cell model, which we call here as an artificial cell, from defined biochemical components, with which the characteristics of the extant cells are expected to be elicited (Luisi et al., 2006; Szostak et al., 2001), is a challenging task for biochemists and chemical engineers. Researchers have been attempting to synthesize artificial cells aiming to identify various aspects of living systems. Attempts to synthesize artificial cells can contribute to a better understanding of the origin of life, as this understanding will provide a physically possible path that could have lead to primitive living cells. Furthermore, experimentally increasing the complexity of the artificial cell, starting from the simplest one, will provide an opportunity to simulate evolutionary processes to the development of more complex organisms, and eventually current organisms (Deamer, 2005; Forster and Church, 2006; Luisi et al., 2006; Szostak et al., 2001). While the experimental construction of artificial cells has been proposed many times over the last several years (Deamer, 2005; Forster and Church, 2006; Luisi et al., 2006; Szostak et al., 2001), progress toward this goal has been proceeding in discrete steps, with researchers assembling elements partially fulfilling the properties of a living system. For example, it was shown to be possible to generate artificial lipid vesicles (liposomes) of the same size as small bacteria from amphiphilic molecules (Bangham and Horne, 1964). Artificial vesicles were also shown to be capable of autocatalytic growth, and to even be able to undergo repeated cycles of growth and division (Hanczyc et al., 2003; Oberholzer et al., 1995; Takakura and Sugawara, 2004). Various types of biological reactions have been successfully performed within the environment provided by liposomes (Fischer et al., 2002; Murtas et al., 2007; Noireaux and Libchaber, 2004; Nomura et al., 2003; Tsumoto et al., 2001; Walde and Ichikawa, 2001; Yu et al., 2001). These studies represent significant steps toward the assembly of an artificial cell. What is common among most of these studies is the usage of giant liposomes (diameter > 1 mm) that are similar in size to extant living cells, which provides an environment that is biologically relevant. Researchers including us have been attempting to encapsulate various biochemical reactions, including the protein synthesis reaction, in giant

Biochemical Reactions in Giant Liposomes

21

liposomes (Fischer et al., 2002; Murtas et al., 2007; Noireaux and Libchaber, 2004; Nomura et al., 2003; Tsumoto et al., 2001; Walde and Ichikawa, 2001; Yu et al., 2001). For the protein synthesis reaction, the cell-free protein synthesis system consisting of large number of components are used as the material for encapsulation. When encapsulating this system in liposomes, it is necessary that the all components required for the reaction are being encapsulated. If the concentrations of the components are below nM, however, and the size of liposomes is below mm, it is possible that some of the components will be missing. It is often the case that components involved in the protein synthesis reaction are at sub-nM concentrations (Shimizu et al., 2001), and therefore, for the protein synthesis reaction to occur in liposomes, giant liposomes have to be used. Here, we describe the strategy for preparing giant liposomes used for encapsulating the b-glucuronidase synthesis reaction and the RNA replication by self-encoded replicase, both of which utilize the cell-free protein synthesis system. We also describe the methods to detect the reaction in individual liposomes using a fluorescence-activated cell sorter (FACS), and to analyze the data. FACS allows the evaluation of large quantities of individual liposomes, at a rate of more than 20,000 liposomes/s. By the quantitative evaluation of the internal reaction within liposomes, properties of the reactions conducted in a small compartment (Kita et al., 2008), and an interesting property of the internal structure of liposomes can be identified (Hosoda et al., 2008).

2. Methods 2.1. Liposome preparation For encapsulation of the cell-free protein synthesis system, the choice of vesicle formation method is critical, since the concentration and encapsulation efficiency of the number of individual components in such a dense suspension may vary over a wide range (Gregoriadis et al., 1999; Monnard et al., 1997; Pupo et al., 2005; Walde and Ichikawa, 2001). Among the various vesicle formation methods proposed to date, the freeze-dried empty liposomes (FDEL) method (Kikuchi et al., 1999; Kirby and Gregoriadis, 1984; Murtas et al., 2007; Torchilin and Weissig, 2003) is one of the most promising candidates. The FDEL method has long been used as a means of vesicle formation with high entrapment efficiency even for biomolecules (Kirby and Gregoriadis, 1984; Torchilin and Weissig, 2003). In this method, liposomes in suspension are lyophilized (freeze-dried) to form empty lamellae of dry lipid film. Upon rehydration, a suspension containing biochemical molecules permeates into the lamellae, which swell to form material-containing vesicles. This method provides facile, stable, and reproducible formation of lipid

22

Takeshi Sunami et al.

vesicles containing complex and dense reaction mixtures. It is also important to note that this method is applicable for almost any combination of buffer and components. For the preparation of FDEL membranes, 1-palmitoyl-2-oleoyl-snglycero-3-phosphocholine (POPC; Avanti Polar Lipids, Alabaster, AL), cholesterol (Nacalai Tesque, Kyoto, Japan), and 1,2-distearoyl-sn-glycero3-phosphoethanolamine-n-[methoxy(polyethylene glycol)-5000] (DSPEPEG5000; NOF Corporation, Tokyo, Japan) dissolved in dichloromethane/diethyl ether (1:1, v/v) are mixed at a molar ratio of 58:39:3 and dried using a rotary evaporator, followed by complete removal of solvent in the vacuum chamber. Note that the lipid composition can be altered as desired, depending on the objective of the experiments, while some lipid compositions may strongly inhibit the internal reactions. The dried lipid film is hydrated with Milli-Q water (12 mM lipid), and this suspension is subjected to vortex mixing for 20 s and sonication for 5 s. After passing through a polycarbonate filter with a pore size of 0.4 mm (Nuclepore TrackEtch Membranes; Whatman, Maidstone, Kent, UK), the suspension is dispensed into small aliquots (40 mL each), and freeze-dried overnight (Labconco Corp., Kansas, MO). After purging the tubes with Argon gas, the freeze-dried membranes are stored in a freezer. Figure 2.1A shows the scanning electron microscope image of the freeze-dried liposome membrane. This dried membrane has a highly porous structure in the micrometer scale. To encapsulate the reaction mixture in liposomes, the solution is simply injected over this dried membrane (Fig. 2.1B).

2.2. Internal protein synthesis reactions followed by the reaction catalyzed by the synthesized proteins In practice, cell-free protein synthesis systems have been used to produce functional proteins, such as green fluorescent protein (GFP) (Ishikawa et al., 2004; Sunami et al., 2006; Yu et al., 2001), T7 RNA polymerase (Ishikawa et al., 2004), Qb RNA replicase, b-galactosidase (Kita et al., 2008), and b-glucuronidase (Hosoda et al., 2008) in liposomes. Among these, we describe the synthesis of b-glucuronidase (Fig. 2.2A) (Hosoda et al., 2008) and Qb replicase (Fig. 2.2B) (Kita et al., 2008). Qb replicase has been used to construct a ‘‘self-encoding system,’’ which is described in detail below. 2.2.1. b-Glucuronidase synthesis A plasmid encoding b-glucuronidase (pET-uidA) (Hosoda et al., 2008) is mixed with the cell-free protein synthesis system (PURESYSTEM classic II (Post Genome, Tokyo, Japan); Shimizu et al., 2001), 50 mM 5-(pentafluorobenzoylamino) fluorescein di-b-D-glucuronide (PFB-FDGlcU; Invitrogen, Carlsbad, CA), and a red fluorescent protein (allophycocyanin, hereafter APC, Molecular Probes, USA) at a high concentration (500 nM final

A

B

Scale bar: 5 mm

Figure 2.1 Microscopic pictures of liposomes prepared by the FDEL method. (A) Scanning electron microscope (SEM) image of freezedried liposomes. Freeze-dried liposomes are observed in a SEM (Keyence, VE-9800, Japan) at 950
magnification with 1 kV acceleration voltage after evaporation of gold (a few hundred nanometers). (B) Microscopic images of liposomes encapsulating R-PE. Upper images: bright-field observation; lower images: fluorescence observation.

24

Takeshi Sunami et al.

A

B Rep(+)Gal(-) RNA b-sub PFB-FDGlcU

Fluorophore

b-glucuronidase

Replicase lacZ

mRNA DNA

b-galactosidase CMFDG Fluorophore

Figure 2.2 Schematic of the encapsulated reactions. (A) b-Glucuronidase synthesis. The mRNA is transcribed from the DNA template, and b-glucuronidase is translated from the mRNA using the cell-free protein synthesis system within the liposomes. Synthesized b-glucuronidase hydrolyzes the fluorogenic substrate (PFB-FDGlcU) to yield the green fluorophore PFB-fluorescein. Red fluorescent protein (APC) is added as a marker of the internal aqueous volume. (B) RNA replication by self-encoded replicase. Rep(þ)Gal() RNA encodes the Qb replicase b-subunit and an antisense sequence of the b-galactosidase gene (lacZ). Synthesized b-subunit assembles three host proteins of Escherichia coli: ribosomal protein S1, elongation factor Tu (EF-Tu), and Ts (EF-Ts) that are present in the cell-free system to generate active Qb replicase. Upon the replication of the sense strand RNA by self-encoded replicase, the sense sequence of lacZ appears, which then yield the b-galactosidase. Synthesized b-galactosidase then hydrolyzes the fluorogenic substrate (CMFDG) to yield the green fluorophore CM-fluorescein.

concentration). PFB-FDGlcU is a fluorogenic substrate of b-glucuronidase that turns into a fluorescent molecule (PFB-fluorescein) as a product of the enzyme reaction and emits green fluorescence. APC is used as a marker for the internal aqueous volume. The liposomes containing the reaction system are prepared by adding 10 mL of reaction mixture into an aliquot of freezedried liposomes (final lipid concentration 48 mM ). The liposome suspension is diluted 20-fold with the dilution buffer (reaction mixture without DNA, substrate, and APC). Protease (final concentration 1 mg/mL) is also included in the dilution buffer to suppress completely the enzyme reaction that may occur outside of the vesicles. All preparation procedures mentioned above are performed on ice. The reaction is then initiated by incubating at 37  C, and the suspension is time-sampled and subjected to FACS measurements. 2.2.2. RNA replication by self-encoded replicase In all living systems, the genome is replicated by proteins encoded within the genome itself. This universal reaction is essential for evolvability of the system. We previously constructed a simplified system termed a self-encoding system, where the genetic information is replicated by self-encoded replicase

Biochemical Reactions in Giant Liposomes

25

in liposomes (Kita et al., 2008). The self-encoding system is assembled using RNA, as shown in Fig. 2.2B. The Rep(þ)Gal() RNA encodes the sense strand of the b-subunit of Qb replicase, an RNA-dependent RNA polymerase responsible for replicating the RNA genome of coliphage Qb (Blumenthal and Carmichael, 1979), and an antisense sequence of the bgalactosidase gene (lacZ ). In addition, the RNA is designed to serve as a template for the replication by Qb replicase. When Rep(þ)Gal() RNA is added to the cell-free protein synthesis system, Qb replicase is first synthesized, which then replicates its own template to produce the antistrand RNA. The antistrand RNA also serves as a template of Qb replicase. As the RNA is replicated by its own encoding gene, we termed this as a self-encoding system. The occurrence of the replication reaction can be probed by the production of b-galactosidase, as it is encoded on the antistrand RNA. Synthesized b-galactosidase then hydrolyzes the fluorogenic substrate. In practice, Rep(þ)Gal() RNA is mixed with the cell-free protein synthesis system (PURESYSTEM, customized as described in Kita et al. (2008)), 100 mM fluorogenic substrate 5-chloromethylfluorescein di-b-Dgalactopyranoside (CMFDG; Invitrogen), and R-phycoerythrin (R-PE; Invitrogen) at high concentration (400 nM at a final concentration). CMFDG is a fluorogenic substrate of b-galactosidase that turns into a fluorescent molecule (CM-fluorescein) as a product of the enzyme reaction and emits green fluorescence. R-PE is used as a marker for internal aqueous volume. Note that both R-PE and APC can be used as an internal volume marker. Liposomes containing the reaction system are prepared and initiated essentially the same as the b-glucuronidase synthesis.

2.3. Detecting reactions in liposomes by fluorescence-activated cell sorting Fluorescence-activated cell sorting is used to measure various reactions in individual cells by detecting the emission intensities of multiple fluorescent markers upon laser irradiation. FACS can also be used to measure the internal reactions in liposomes (Kageyama et al., 2007; Taly et al., 2007). For example, green fluorescent products in liposomes can be quantified by measuring the green fluorescence intensity of individual liposomes, whereas the inner aqueous volume of each liposome can be quantified from the intensity of the red fluorescent marker protein encapsulated. As the number of green product molecules and the internal aqueous volume can be simultaneously measured, the concentration of the product molecule in individual liposomes can be estimated. The rate of measurement by FACS is more than 20,000 liposomes/s, significantly faster than that of microscopic observations. Typical results of FACS measurements are shown in Fig. 2.3. Note that FACS can be used not only to measure the internal reaction of individual liposomes but also to sort the liposomes with desired fluorescence

Internal aqueous volume (fL)

A

103

0 min

103

30 min

103

60 min

103

102

102

102

102

10

10

10

10

1 103

104

105

1 106 103

104

1 105 106 103 104 PFB-fluorescein (molecules)

105

1 106 103

90 min

104

105

106

B Internal aqueous volume (fL)

103

0 min

103

120 min

103

200 min

103

102

102

102

102

10

10

10

10

1 103

104

105

1 106 103

104

1 105 106 103 104 CM-fluorescein (molecules)

105

1 106 103

275 min

104

105

106

Figure 2.3 Time course of the reaction analyzed by FACS. Typical results of in liposome (A) b-glucuronidase synthesis and (B) RNA replication by self-encoded replicase. The product (horizontal) and internal aqueous volume (vertical) of each liposome are shown. Dots represent the data of individual liposomes. In all plots, 20,000 data points from a total of 100,000 obtained are shown for clarity. The reactions proceed only in a fraction of the liposomes. This is because not all liposomes carry the plasmid DNA or the reaction efficiency is very low (Hosoda et al., 2008; Kita et al., 2008).

Biochemical Reactions in Giant Liposomes

27

signals, and the sorted samples can be used for further analysis such as microscopic observations (e.g., Hosoda et al., 2008). For FACS measurements, the sampled liposome suspension is diluted by at least 10-fold with sheath solution (FACSFlow; Becton Dickinson, San Jose, CA) to reduce the measurement frequency to smaller than 20,000 events/s. Then, red and green fluorescence signals, reflecting the whole internal aqueous volume and the amount of product, respectively, are measured with a FACS (FACSAria; Becton Dickinson). Typically, 100,000 data are sufficient for each measurement. The time course of the internal reaction can also be measured by analyzing the liposomes at different time points. APC or R-PE, used as a measure for the internal aqueous volume, is excited with a HeNe laser (633 nm) (or semiconductor laser (488 nm) in the case of R-PE), and the emission is detected through a 660  10 nm band-pass filter (or 576  13 nm band-pass filter in the case of R-PE). The number of R-PE molecules is converted from the red fluorescence signal detected (FIR) using the linear relation between the intensity from fluorescent beads carrying a known amount of R-PE (QuantiBRITE PE Quantitation Kit; BD Biosciences Clontech, Palo Alto, CA). The number of APC molecules is converted from the red fluorescence signal detected (FIR) using the linear relation between the intensity from fluorescent beads carrying a known amount of RPE, and the linear relation between fluorescence intensity of R-PE and APC. The quantity of the whole volume (Vw) is calculated by dividing the number of APC or R-PE molecules by the concentration of encapsulated APC (500 nM) or R-PE (400 nM). The final equation used for the conversion is Vw ¼ 0.054 FIR, or Vw ¼ 0.094 FIR, with APC or R-PE, respectively. The number of product molecules is derived from the intensity of the green fluorescence signal (FIG). PFB-fluorescein and CM-fluorescein, the reaction product of b-glucuronidase and b-galactosidase, respectively, are excited with a 488-nm semiconductor laser and the emission is detected through a 530  15 nm band-pass filter. For conversion, the linear relation between the green fluorescence intensities obtained by FACS and the number of product molecules (PFB-fluorescein or CM-fluorescein) is obtained. This is obtained by determining the intensity of a known amount of fluorescent molecules, and the linear relation between the intensity of the fluorescent molecule and the product molecules. The final equation used for conversion of PFB-fluorescein and CM-fluorescein is NG ¼ 330 FIG and NG ¼ 920 FIG, respectively, where NG is the number of product molecules.

3. Analysis of the FACS Data The kinetics of the internal reaction can be obtained from the FACS data. Figure 2.4A shows the histogram (frequency distribution) of the product concentration of RNA replication by self-encoded replicase in

28

Takeshi Sunami et al.

0.1 Frequency

B

0.12 0.08 0.06

0 min 120 min 200 min 275 min

0.04 0.02 0 0.1

1 10 100 CM-fluorescein (mM )

CM-fluorescein (mM )

A

20 15 10

Internal volume (fL) 1.2-4.0: 4.0-13: 13-40: 40-130:

5 0

0 100

200 300 400 Time (min)

500

Figure 2.4 Analysis of the FACS data. (A) Histogram (frequency distribution) of the CM-fluorescein concentration in liposomes with an internal volume of 4–13 fL at different time points. This is obtained from the data shown in Fig. 2.3B. (B) Time course of the reaction in liposomes with different internal volumes.

liposomes with different internal aqueous volumes and at different time points. In the measurement data, liposomes in which reaction proceeded (reacted liposomes) can be distinguished from those remained non-reacted based on the intensity of green fluorescence. The median of the histogram of the reacted liposomes are employed as a representative product concentration in liposomes with a given aqueous volume. The median values are used to trace the representative behavior of the internal reaction. As an example, we describe the strategy to obtain the time course data from the data shown in Fig. 2.4A. First, median values of the product concentration at the end of reaction, where the population of reacted liposomes can be readily distinguished, are obtained for each liposome internal volume (1.2–4.0, 4.0–13, 13–40, 40–130 fL). The rank in the product concentration (rank 1 ¼ highest product concentration) of the liposomes exhibiting the median value is then obtained. To obtain the representative product concentration at intermediate time points, where the distributions of reacted and non-reacted liposomes overlap, the product concentrations of the liposomes of identical rank to those obtained above are subsequently obtained for each liposome internal volume at different reaction times. These give the time courses of the reaction within liposomes shown in Fig. 2.4B.

4. Conclusions The presented method for in liposome protein synthesis followed by the measurement by FACS allows us to assess the kinetics of reactions in liposomes. This method can thus be used to elucidate the reaction dynamics in a microcompartment. In all living cells, cytosolic biochemical reactions are enclosed in a lipidic compartment. The effect of the microcompartment

Biochemical Reactions in Giant Liposomes

29

properties on the internal reaction has not been fully investigated. It may be possible to elucidate this effect with the presented method, for example, by carrying out the reaction in liposomes with different lipid composition and/ or different size. Thus, by providing an environment mimicing that of living cells, and by investigating the reactions inside, it may be possible to reveal the role of compartmentalization in cellular systems.

ACKNOWLEDGMENTS This research was supported in part by ‘‘Special Coordination Funds for Promoting Science and Technology: Yuragi Project’’ and ‘‘Global COE (Centers of Excellence) Program’’ of the Ministry of Education, Culture, Sports, Science, and Technology, Japan.

REFERENCES Bangham, A. D., and Horne, R. W. (1964). Negative staining of phospholipids and their structural modification by surface-active agents as observed in the electron microscope. J. Mol. Biol. 8, 660–668. Blumenthal, T., and Carmichael, G. G. (1979). RNA replication: Function and structure of Qbeta-replicase. Annu. Rev. Biochem. 48, 525–548. Deamer, D. (2005). A giant step towards artificial life? Trends Biotechnol. 23, 336–338. Fischer, A., Franco, A., and Oberholzer, T. (2002). Giant vesicles as microreactors for enzymatic mRNA synthesis. ChemBioChem 3, 409–417. Forster, A. C., and Church, G. M. (2006). Towards synthesis of a minimal cell. Mol. Syst. Biol. 2, 45. Gregoriadis, G., McCormack, B., Obrenovic, M., Saffie, R., Zadi, B., and Perrie, Y. (1999). Vaccine entrapment in liposomes. Methods 19, 156–162. Hanczyc, M. M., Fujikawa, S. M., and Szostak, J. W. (2003). Experimental models of primitive cellular compartments: Encapsulation, growth, and division. Science 302, 618–622. Hosoda, K., Sunami, T., Kazuta, Y., Matsuura, T., Suzuki, H., and Yomo, T. (2008). Quantitative study of the structure of multilamellar giant liposomes as a container of protein synthesis reaction. Langmuir 24, 13540–13548. Ishikawa, K., Sato, K., Shima, Y., Urabe, I., and Yomo, T. (2004). Expression of a cascading genetic network within liposomes. FEBS Lett. 576, 387–390. Kageyama, Y., Toyota, T., Murata, S., and Sugawara, T. (2007). Study on structural changes in supramolecular assemblies composed of amphiphilic nicotinamide and its dihydronicotinamide derivative by flow cytometry. Soft Matter 3, 699–702. Kaneko, T., Itoh, T. J., and Hotani, H. (1998). Morphological transformation of liposomes caused by assembly of encapsulated tubulin and determination of shape by microtubuleassociated proteins (MAPs). J. Mol. Biol. 284, 1671–1681. Kikuchi, H., Suzuki, N., Ebihara, K., Morita, H., Ishii, Y., Kikuchi, A., Sugaya, S., Serikawa, T., and Tanaka, K. (1999). Gene delivery using liposome technology. J. Control. Release 62, 269–277. Kirby, C., and Gregoriadis, G. (1984). Dehydration-rehydration vesicles—A simple method for high-yield drug entrapment in liposomes. Biotechnology 2, 979–984.

30

Takeshi Sunami et al.

Kita, H., Matsuura, T., Sunami, T., Hosoda, K., Ichihashi, N., Tsukada, K., Urabe, I., and Yomo, T. (2008). Replication of genetic information with self-encoded replicase in liposomes. ChemBioChem 9, 2403–2410. Luisi, P. L., and Walde, P. (2000). Giant Vesicles. John Wiley & Sons Inc., New York. Luisi, P. L., Ferri, F., and Stano, P. (2006). Approaches to semi-synthetic minimal cells: A review. Naturwissenschaften 93, 1–13. Monnard, P. A., Oberholzer, T., and Luisi, P. (1997). Entrapment of nucleic acids in liposomes. Biochim. Biophys. Acta 1329, 39–50. Murtas, G., Kuruma, Y., Bianchini, P., Diaspro, A., and Luisi, P. L. (2007). Protein synthesis in liposomes with a minimal set of enzymes. Biochem. Biophys. Res. Commun. 363, 12–17. Noireaux, V., and Libchaber, A. (2004). A vesicle bioreactor as a step toward an artificial cell assembly. Proc. Natl. Acad. Sci. USA 101, 17669–17674. Nomura, S. M., Tsumoto, K., Hamada, T., Akiyoshi, K., Nakatani, Y., and Yoshikawa, K. (2003). Gene expression within cell-sized lipid vesicles. ChemBioChem 4, 1172–1175. Oberholzer, T., Wick, R., Luisi, P. L., and Biebricher, C. K. (1995). Enzymatic RNA replication in self-reproducing vesicles: An approach to a minimal cell. Biochem. Biophys. Res. Commun. 207, 250–257. Pupo, E., Padron, A., Santana, E., Sotolongo, J., Quintana, D., Duenas, S., Duarte, C., de la Rosa, M. C., and Hardy, E. (2005). Preparation of plasmid DNA-containing liposomes using a high-pressure homogenization-extrusion technique. J. Control. Release 104, 379–396. Shimizu, Y., Inoue, A., Tomari, Y., Suzuki, T., Yokogawa, T., Nishikawa, K., and Ueda, T. (2001). Cell-free translation reconstituted with purified components. Nat. Biotechnol. 19, 751–755. Sunami, T., Sato, K., Matsuura, T., Tsukada, K., Urabe, I., and Yomo, T. (2006). Femtoliter compartment in liposomes for in vitro selection of proteins. Anal. Biochem. 357, 128–136. Szostak, J. W., Bartel, D. P., and Luisi, P. L. (2001). Synthesizing life. Nature 409, 387–390. Takakura, K., and Sugawara, T. (2004). Membrane dynamics of a myelin-like giant multilamellar vesicle applicable to a self-reproducing system. Langmuir 20, 3832–3834. Taly, V., Kelly, B. T., and Griffiths, A. D. (2007). Droplets as microreactors for highthroughput biology. ChemBioChem 8, 263–272. Torchilin, V., and Weissig, V. (2003). Liposomes: A Practical Approach. 2nd ed. Oxford University Press, Oxford. Tsumoto, K., Nomura, S. M., Nakatani, Y., and Yoshikawa, K. (2001). Giant liposome as a biochemical reactor: Transcription of DNA and transportation by laser tweezers. Langmuir 17, 7225–7228. Walde, P., and Ichikawa, S. (2001). Enzymes inside lipid vesicles: Preparation, reactivity and applications. Biomol. Eng. 18, 143–177. Yu, W., Sato, K., Wakabayashi, M., Nakaishi, T., Ko-Mitamura, E. P., Shima, Y., Urabe, I., and Yomo, T. (2001). Synthesis of functional protein in liposome. J. Biosci. Bioeng. 92, 590–593.

C H A P T E R

T H R E E

Construction of Cell-Sized Liposomes Encapsulating Actin and Actin-Cross-linking Proteins Kingo Takiguchi,* Ayako Yamada,†,1 Makiko Negishi,† Makoto Honda,*,2 Yohko Tanaka-Takiguchi,* and Kenichi Yoshikawa† Contents 1. Introduction 1.1. Actin-cross-linking proteins coencapsulated in cell-sized liposomes 1.2. Methodologies to encapsulate actin and actin-cross-linking proteins into giant liposomes 2. Experimental Section 2.1. Preparation of actin and actin-cross-linking proteins for encapsulation 2.2. Natural swelling method for coencapsulation of actin and fascin, a-actinin, filamin, or BBMI into liposomes 2.3. Spontaneous transfer method for coencapsulation of actin and HMM (ActoHMM) into liposomes 3. Morphogenesis of Giant Liposomes Encapsulating Actin and Its Cross-linking Proteins 3.1. G-Actin and its polymerization in liposomes (natural swelling method) 3.2. Actin and fascin (natural swelling method) 3.3. Actin and a-actinin (natural swelling method) 3.4. Actin and filamin (natural swelling method) 3.5. Actin and BBMI (natural swelling method) 3.6. F-Actin (spontaneous transfer method)

* { 1 2

32 33 34 37 37 37 38 42 42 43 43 43 45 45

Division of Biological Science, Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Japan Department of Physics, Graduate School of Science, Kyoto University, Sakyo-ku, Kyoto, Japan Current address: Department of Chemistry, Ecole Normale Superieure, Paris, France Current address: Stem Cell and Drug Discovery Institute, Shimogyo-ku, Kyoto, Japan

Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64003-9

#

2009 Elsevier Inc. All rights reserved.

31

32

Kingo Takiguchi et al.

3.7. ActoHMM (spontaneous transfer method) 3.8. Actin and S-1 (spontaneous transfer method) 3.9. Mechanism determining liposome morphology 4. Concluding Remarks Acknowledgments References

45 46 46 49 50 50

Abstract To shed light on the mechanism underlying the active morphogenesis of living cells in relation to the organization of internal cytoskeletal networks, the development of new methodologies to construct artificial cell models is crucial. Here, we describe the successful construction of cell-sized liposomes entrapping cytoskeletal proteins. We discuss experimental protocols to prepare giant liposomes encapsulating desired amounts of actin and cross-linking proteins including molecular motor proteins, such as fascin, a-actinin, filamin, myosin-I isolated from brush border (BBMI), and heavy meromyosin (HMM). Subfragment 1 (S-1) is also studied in comparison to HMM, where S-1 and HMM are singleheaded and double-headed derivatives of conventional myosin (myosin-II), respectively. In the absence of cross-linking proteins, actin filaments (F-actin) are distributed homogeneously without any order within the liposomes. In contrast, when actin is encapsulated together with an actin-cross-linking protein, mesh structures emerge that are similar to those in living motile cells. Optical microscopic observations on the active morphological changes of the liposomes are reported.

1. Introduction Living cells and their organelles are compartmentalized by biomembranes in a self-organized manner, and each has a specific shape depending on its function. Cellular morphologies are thought to be determined and maintained by cytoskeletal networks (Hotani et al., 2003; Rodriguez et al., 2003). F-actin is a major component of the cytoskeleton, and is involved in a variety of cellular functions. Such functions include the extension of microspikes from cells (Albrecht-Buehler and Lancaster, 1976), the movement of filopodia in neural growth cones (Mitchison and Cramer, 1996; Mitchison and Kirschner, 1988), the extension or retraction of pseudopods during amoeboid movement (Taylor and Condeelis, 1979), and the contraction of contractile rings during cell division (Schroeder, 1973). F-actins also provide mechanical support, including stress fibers (Byers and Fujikawa, 1982). It is therefore important to study the role of the actin cytoskeleton in relation to membrane morphogenesis. Along this line, artificial model systems using giant liposomes containing actin and its associating proteins have been developed (Ba¨rmann et al., 1992; Cortese et al., 1989; Fygenson

33

Construction of Actin-Encapsulating Liposomes

et al., 1997; Ha¨ckl et al., 1998; Limozin et al., 2003, 2005; Limozin and Sackmann, 2002; Miyata and Hotani, 1992; Miyata et al., 1999; Pontani et al., 2009). In this chapter, we use the term ‘‘cell-sized liposome’’ to indicate giant liposomes with sizes of several to several tens of micrometers, which constitute closed thin membranes of phospholipids in aqueous solution. Cell-sized liposomes are expected to serve as a useful model in the real world, allowing us to make real-time observations using optical microscopy (Bangham, 1995; Cortese et al., 1989; Hotani et al., 2003; Lasic, 1995; Lipowsky, 1991; Miyata et al., 1999).

1.1. Actin-cross-linking proteins coencapsulated in cell-sized liposomes F-actins usually function in vivo in bundles or networks, and they are organized by various actin-binding, especially actin-cross-linking, proteins. Accordingly, we studied the effects of five different actin-cross-linking proteins, fascin, a-actinin, filamin, BBMI, and HMM on the morphogenesis of liposomes caused by actin assembly, by coencapsulating actin and one of those proteins together in liposomes. Actin-cross-linking proteins are characterized by the manner of cross-linking, that is, by the distance and angular flexibility between adjacent cross-linked F-actin molecules, as illustrated in Fig. 3.1. Fascin is a 55 kDa globular protein and is responsible for the tight crosslinking of F-actin to form bundles (Cant et al., 1994; Edwards and Bryan, 1995; Edwards et al., 1995; Otto et al., 1980; Yamashiro-Matsumura and Matsumura, 1985). a-Actinin is a larger protein (about 100 kDa), and Proteins

Fascin

a-actinin

Filamin

BBMI

HMM

S-1

Domain structure

Crosslinking pattern

Figure 3.1 Model of F-actin bundles or networks formed by actin-cross-linking proteins. From left to right, cross-linked F-actins mediated with fascin, a-actinin, filamin, BBMI, or HMM are shown. The upper row shows the molecular structure of each cross-linking protein. The lower row illustrates the arrangement of F-actins crosslinked by each cross-linking protein. S-1 and S-1-associating F-actins are also shown (right). F-actin is depicted as a line, while fascin, a-actinin, filamin, BBMI, HMM, and S-1 are drawn as globular monomers, dumbbell-shaped dimers, V-shaped dimers, single-headed musical note-like structures, double-headed structures, and single-headed structures, respectively.

34

Kingo Takiguchi et al.

antiparallel side-by-side associated homodimers of a-actinin form F-actin bundles with much looser packing (Imamura et al., 1988; Meyer and Aebi, 1990; Mimura and Asano, 1987). Filamin is a protein of approximately 250 kDa with a flexible elongated structure, and end-to-end associated homodimers of filamin are able to mediate flexible and high-angle linking between two adjacent F-actins (Bretscher, 1991; Hartwig and Kwiatkowski, 1991; Hock et al., 1990). The dynamics and mechanical properties of actin networks caused by these actin-cross-linking proteins have been well studied (Collins et al., 1991; Grazi et al., 1994; Hou et al., 1990; Wachsstock et al., 1993, 1994; Xu et al., 1998; Yamashiro et al., 1998). However, the relationship between the actin network organization and the morphogenesis of biological membranes has not been clarified yet. In living cells, the cytoskeleton is thought to be anchored to the membrane through membrane proteins, which probably regulate their own specific membranous morphology, position, and duration. BBMI exhibits a unique property in comparison with the other actin-cross-linking proteins mentioned above (fascin, a-actinin, filamin, or HMM). BBMI consists of a single motor domain (head) and a tail domain, and since the tail domain can bind both F-actin and a lipid membrane, BBMI can play a role as a cross-linker between F-actin and the membrane, as well as between F-actins (Coluccio, 1997). HMM has been frequently studied as a representative of double-headed myosin motors, and it can form and transform bundles and gels of F-actin (Takiguchi, 1991; Tanaka-Takiguchi et al., 2004). On the other hand, S-1 has often been studied as a representative of simple single-headed myosin motors. S-1 has only one actin-binding motor domain and is unable to cross-link F-actins, thus S-1 often has been studied in comparison to HMM.

1.2. Methodologies to encapsulate actin and actin-cross-linking proteins into giant liposomes Through the methodology of natural swelling, which is a method to obtain liposomes by the hydration of dried lipid films, we have successfully prepared liposomes containing monomeric actin (G-actin) and an actincross-linking protein (such as fascin, a-actinin, filamin, or BBMI) in the presence of 30 mM KCl (Honda et al., 1999). As the encapsulated G-actin polymerizes into F-actins, they form bundles or gels depending on the type of coencapsulated actin-cross-linking protein, causing various morphological changes of liposomes. The differences in morphology among the transformed liposomes indicate that the actin-cross-linking proteins determine the liposome shape by organizing their specific actin networks. On the other hand, in living cells, and depending on the cell type and the intracellular location, actin is expressed up to about 300 mM and undertakes its functions with the cooperation of various myosin motor proteins under physiological salt conditions (several mM Mg2þ and several tens of

Construction of Actin-Encapsulating Liposomes

35

mM Kþ) ( Janson et al., 1991; Pollard et al., 2000). However, using the natural swelling method, it is difficult or almost impossible to entrap the desired amounts of F-actin and myosin simultaneously in giant liposomes at physiological salt conditions. This is mainly due to the difficulty in preparing giant liposomes in the presence of salt, especially divalent cations such as Mg2þ, and the difficulty in controlling the amount of macromolecules to be incorporated within the liposomes, because of the passive nature of the encapsulation process. To overcome those problems, we have developed the spontaneous transfer method through an oil/water interface, by adapting water-in-oil phospholipid-coated cell-sized-droplets (W/O droplets) as precursors of giant liposomes (Yamada et al., 2006). Recently, there have been several attempts to employ W/O droplets coated by phospholipids as a model for living cells instead of liposomes (Hase and Yoshikawa, 2006), or as a precursor of liposomes (Hamada et al., 2008; Noireaux and Libchaber, 2004; Pautot et al., 2003a,b; Pontani et al., 2009; Takiguchi et al., 2008; Yamada et al., 2006). The W/O droplets are prepared easily by emulsifying an aqueous solution, together with oil, containing phospholipids. The process enables us to encapsulate biomolecules at a controlled concentration under any salt strength, into cell-sized compartments covered with a monolayer of phospholipids. By taking advantage of this, the behavior of F-actins entrapped in a closed space (Claessens et al., 2006) or conformational changes of actin molecules interacting with a phospholipid membrane in the presence of Mg2þ (Hase and Yoshikawa, 2006) have been studied. In addition, we have succeeded in developing a system for the controlled fusion of two droplets containing a substrate and an enzyme, respectively (Hase et al., 2007). Consequently, methodologies to obtain liposomes by transferring phospholipid-coated W/O droplets from an oil phase to an aqueous phase through their interface using centrifugation (Noireaux and Libchaber, 2004; Pautot et al., 2003a,b; Pontani et al., 2009) or without an external force (the spontaneous transfer method; Hamada et al., 2008; Takiguchi et al., 2008; Yamada et al., 2006) have been developed. With those methodologies, one can obtain liposomes with sizes of 10–100 mm containing desired amounts of molecules. Furthermore, by taking advantage of microfluidic techniques, new methodologies to obtain monodisperse liposomes from monodisperse droplets have been actively studied (Funakoshi et al., 2007; Shum et al., 2008; Stachowiak et al., 2008). Table 3.1 compares the various methodologies used to prepare giant liposomes. Using one of the simplest methods, that is, the spontaneous transfer method, we succeeded in constructing giant liposomes encapsulating 200 mM F-actin in the presence of 5 mM MgCl2 and 50 mM KCl. Moreover, this method enabled us to succeed in encapsulating desired amounts of HMM or S-1 with F-actin into giant liposomes (Takiguchi et al., 2008).

Table 3.1 Methods for giant liposome preparation Natural swellinga

Preparation Required time for encapsulation Yield par observation field Required amount of the sample Control of inner/ outer condition Control of the size Asymmetric membrane Physiological salt concentration Problem of oil contamination a b c d e f g h

Easy Long Low

Electroformationb

Centrifugationc

Spontaneous transferd

Jettinge

Double emulsion f

Easy Comparably long High

Easy Short

Easy Short

Small

Difficult Comparably short Comparably low Large

Difficult Comparably short Comparably low Large

High

Comparably large Difficult

Large

Comparably low Small

Difficult

Easy

Easy

Easy

Easy

Difficult

Difficult

Difficult

Easy

Easy

Difficult

Difficult

Easyg

In principle easy Easyh

Difficult

Difficult

Comparably difficult No

Difficult

Easy

Easy

Easy

Easy

No

Yes

Yes

Yes

Yes

Robust method to prepare liposomes of almost any content, but yield is sometimes low. Multilamellar liposomes tend to be obtained (Bangham et al., 1965). High yield of unilamellar liposomes. A special technique is required for salt concentrations higher than 10 mM (Angelova and Dimitrov, 1986). The principle is similar to that of the spontaneous transfer method. Liposomes are free in an aqueous phase (Noireaux and Libchaber, 2004; Pautot et al., 2003b). Easy encapsulation. Liposomes are anchored to an oil/water interface (Yamada et al., 2006). Microfluidic technique is required. Highly monodisperse liposomes can be obtained (Funakoshi et al., 2007; Stachowiak et al., 2008). Microfluidic technique and use of volatile oil are required. Highly monodisperse liposomes can be obtained (Shum et al., 2008). Pautot et al. (2003a). Hamada et al. (2008).

Construction of Actin-Encapsulating Liposomes

37

2. Experimental Section 2.1. Preparation of actin and actin-cross-linking proteins for encapsulation G-Actin is obtained by incubating dried muscle (acetone powder) prepared from rabbit skeletal muscle with cold water followed by a polymerization– depolymerization cycle (Ebashi and Ebashi, 1965; Higashi-Fujime, 1983). Fascin is isolated from porcine brain extracts using the actin gel method, following DE-52 and hydroxylapatite (Bio-Rad) column chromatographies (Honda et al., 1999; Yamashiro-Matsumura and Matsumura, 1985). a-Actinin is purified from an extract of minced rabbit skeletal muscle that had been incubated overnight at room temperature using a repeating fractionation with ammonium sulfate (Ebashi and Ebashi, 1965). Filamin is isolated from chicken gizzard by a modification (Muguruma et al., 1990) of the method of Molony et al. (1987). BBMI is purified from an extract of homogenized brush borders isolated from chicken intestines using a series of column chromatographies, Sepharose CL-4B, CM sepharose, and monoQ HR5/5 (Collins et al., 1991; Jontes and Milligan, 1997). Skeletal muscle myosin (myosin-II) was obtained from rabbit skeletal muscles (white muscles of the back and hind legs). HMM and S-1 are obtained by digestion of skeletal muscle myosin with chymotrypsin in the presence of 1 mM MgCl2/ 0.5 M KCl and of 1 mM EDTA (ethylenediamine tetraacetic acid)/120 mM KCl, respectively (Tanaka-Takiguchi et al., 2004).

2.2. Natural swelling method for coencapsulation of actin and fascin, a-actinin, filamin, or BBMI into liposomes To obtain giant liposomes encapsulating actin and its cross-linking proteins, dried phospholipid films are hydrated with an aqueous solution containing the proteins (Honda et al., 1999; Hotani and Miyamoto, 1990; Kaneko et al., 1998; Miyata and Hotani, 1992). The phospholipids used are L-aphosphatidylcholine (PC), L-a-phosphatidylethanolamine (PE) and L-aphosphatidylglycerol (PG). Each phospholipid has been isolated from egg yolk and is purchased from Sigma (St. Louis, MO) or Avanti Polar Lipid (Alabaster, AL), and is dissolved in a chloroform/methanol solution (98:2, v/v) at a concentration of 10 mM. Methanol is added to protect lipid molecules from damage resulting from the moisturization of chloroform. The phospholipid solutions are mixed in a glass test tube (total of 200 mg lipid). Typically, 20 ml of the solution is poured into a glass tube (10 mm in diameter and 30 mm in height) which has been cleaned with the chloroform/methanol solution prior to use. The organic solvent of the phospholipid solution is evaporated under a flow of nitrogen gas, and the lipids are

38

Kingo Takiguchi et al.

further dried under vacuum for at least 90 min with a vacuum pump (G-20D, ULVAC, Chigasaki, Japan). Note that the phospholipid solutions are stocked at 4  C and are protected from light, and the dried lipid films are stocked in an evacuated dessiccator at room temperature and are also protected from light. To generate liposomes from the film, 40 ml buffer A (5 mM Tris–HCl, pH 8.0, 0.2 mM ATP, 5 mM dithiothreitol, 30 mM KCl) containing various concentrations of G-actin and an actin-cross-linking protein are added to the dried lipid films on ice. The lipid films immediately started swelling to form liposomes containing the proteins, and further swelling is facilitated by occasionally agitating the test tubes by hand. After 30 min, the liposome suspensions are diluted 30-fold with buffer A to prevent the polymerization of the G-actin outside the liposomes. The polymerization of G-actin to form F-actin inside the liposomes is then initiated by raising the temperature to 25  C under dark-field microscopy (BHF, Olympus, Tokyo, Japan) (Hotani, 1984). 2.2.1. Visualization of actin filament (F-actin) To visualize F-actin localization within liposomes, rhodamine-phalloidin (R-415, Molecular Probes, Eugene, OR) in methanol is added to the lipid solution prior to the film preparation. The final concentration of rhodamine-phalloidin in the specimens is about 20 nM. The rhodamine-phalloidin specifically binds F-actin, so that actin filaments become visible by fluorescence microscopy (Miyata and Hotani, 1992). 2.2.2. Visualization of liposome membranes Giant liposomes can be visualized without any labeling by several types of optical microscopes, such as phase contrast, differential interference contrast (DIC), and dark-field. To observe liposome membranes by fluorescence microscopy, a fluorescence-labeled phospholipid is mixed with other phospholipids (about 1:200, w/w) (Honda et al., 1999). NBD- or BODIPYconjugated phospholipids are frequently used as the fluorescence-labeled phospholipid. The fluorescence-labeled liposomes are then observed with fluorescence optics usually assembled into another microscope, that is, a phase contrast, DIC, or dark-field microscope.

2.3. Spontaneous transfer method for coencapsulation of actin and HMM (ActoHMM) into liposomes A schematic representation of the experimental setup is illustrated in Fig. 3.2 (Takiguchi et al., 2008; Yamada et al., 2006).

39

Construction of Actin-Encapsulating Liposomes

Egg PC in mineral oil

A

5 mm

PDMS Buffer

Cover glass

4 mm

Objective lens B

ActoHMM

Oil

Z X

Buffer

Figure 3.2 (A) Experimental setup for construction and observation of actoHMMencapsulating giant liposomes using the spontaneous transfer method. (B) Schematic representation of the transformation from a W/O droplet in the oil phase (left) to a liposome in the aqueous phase (right). F-actin is illustrated with a line, and HMM is illustrated as a double-headed structure. The interface of the droplet is depicted as a lipid monolayer, while a multilayered interface may be generated to some extent in these experiments. The figure is adopted from Takiguchi et al. (2008).

2.3.1. Observation chamber An observation chamber is prepared consisting of a cylindrical hole (ca. 4 mm in diameter) in a poly(dimethylsiloxane) (PDMS) sheet (ca. 5 mm thick) on a glass microscope slide (0.12–0.17 mm thick, Matsunami Glass, Osaka, Japan). To obtain the PDMS sheet, the base solution and a curing agent of Silpot 184 W/C (Dow Corning Toray, Tokyo, Japan) are mixed well at a ratio of 10:1 (w/w) and are poured into a plastic Petri dish to harden. Note that bubbles generated during the mixing disappear spontaneously in about 1 h at room temperature. One can speed the hardening process by baking but the temperature must not be higher than 85  C to avoid melting the Petri dish. It takes about 24 h at room temperature or 2 h at 85  C

40

Kingo Takiguchi et al.

to harden the PDMS. A piece of the PDMS sheet is cut off and a hole with ca. 4 mm diameter is made in it with a leather-craft punch. One can make several holes in each PDMS sheet to compare different conditions. Note that if the diameter is larger than 4 mm, it is difficult to maintain a proper oil/water interface in the chamber because the water phase tends to be a spherical shape due to surface tension, since the PDMS surface is rather hydrophobic. These PDMS sheets are repeatedly used after washing with ethanol or isopropanol. The PDMS surface is cleaned with scotch tape before being set on a glass cover slide, which is cleaned by baking at 500  C for 1 h. The PDMS and the glass surface must be tightly attached. 2.3.2. Oil containing phospholipids Chloroform or a chloroform/methanol solution of PC is prepared at a concentration of 10 mM and poured into a glass test tube. Typically, 25–50 ml of the solution is poured into a glass tube (7 mm in diameter and 30 mm in height) (Maruemu, Osaka, Japan), which has been cleaned with acetone prior to use. The organic solvent of the phospholipid solution is then evaporated under a nitrogen flow and is dried under vacuum for more than 15 min to produce a dry film at the bottom of the glass tube. Typically, 500 ml of mineral oil (Nacalai Tesque, Kyoto, Japan) is then added to the tube and is sealed tightly to avoid moisture prior to ultrasonication for 60 min at 50  C and vortex mixing. For sealing, Parafilm (Pechiney Plastic Packaging, Chicago, IL) alone cannot be used because it melts under these conditions. A combination of a cap from the tube and Parafilm, or Dura Seal (Diversified Biotech, Boston, MA) and Parafilm can be used. Note that the vortex mixing must be performed immediately after ultrasonication to avoid aggregation of the phospholipid. The final concentration of PC is 0.5 or 1.0 mM, and these prepared oils must be used in 1 day. 2.3.3. Oil/water interface in the chamber Firstly, 10 ml of an aqueous phase, that is, the outer solution of liposomes, is placed at the bottom of the observation chamber. Here, we use a solution consisting of 25 mM imidazol-HCl, 5 mM MgCl2, 50 mM KCl, and 10 mM DTT at pH 7.5 (buffer B) and up to 100 mM sucrose, or the concentration of buffer B is increased up to twice instead of adding sucrose to adjust the osmotic pressure to that of the inner solution. It is recommended to touch the glass surface with the pipette tip so that one can fill the chamber from the bottom. The glass surface should be totally wet without air bubbles and the air/water interface must be rather flat and horizontal. Ten microliters of the PC-containing oil is then gently added along the PDMS wall to obtain an oil/water interface covered with

Construction of Actin-Encapsulating Liposomes

41

phospholipid between the upper oil phase and the lower aqueous phase. The observation chamber is then set on the optical microscope. An objective lens with a magnification not higher than 20 times can be employed to focus on the formed interface. 2.3.4. Preparation of W/O droplets and transfer into liposomes One hundred microliters of the PC-containing oil is placed in an Eppendorf tube and 5 ml of the inner solution, that is, the target solution to be encapsulated in liposomes, is then added to the tube. As the inner solution, buffer B containing F-actin or actoHMM is prepared. Immediately after addition of the aqueous solution, the solution is emulsified by pipetting it up and down for about 30 s. Ten microliters of the obtained W/O droplet solution is then placed on the oil phase in the chamber. One can observe that, in the oil phase, the W/O droplets gradually fall down onto the oil/water interface because of gravity. Interestingly, the droplets then spontaneously move through the interface into the aqueous solution keeping their spherical shape. In our experimental conditions, the transferred droplets, or liposomes, are anchored onto the interface (Fig. 3.2). Although it is possible to transfer the liposomes further into the bulk aqueous phase by encapsulating higher density solution than the outer phase (Hamada et al., 2008), or using centrifugation (Pontani et al., 2009), we performed the observations on liposomes anchored to the interface since we could then monitor the full process of the transfer and the subsequent transformation of each specific liposome. As for the transformation of a droplet in oil into a liposome in water, we have already discussed the full details of the process (Yamada et al., 2006). Observations are performed using a Zeiss Axiovert 100 inverted microscope equipped with an LSM 510 module for confocal microscopy (Carl Zeiss, Jena, Germany). To observe liposome membranes by fluorescence microscopy, fluorescence-labeled phospholipids are mixed with other phospholipids (1:100–1:1000, mol/mol). 2.3.5. Visualization of F-actin Actin prepared as described above is polymerized in buffer C (2 mM Tris– HCl, pH 8.0, 30 mM KCl, and 0.2 mM ATP) and then used for the experiments. To visualize F-actin entrapped within liposomes, a mixture of rhodamine-phalloidin in methanol and buffer C, or alternatively rhodamine-phalloidin redissolved in buffer C after volatilizing the methanol, is added to the actin solution before emulsification in the lipid-containing oil or mixing with HMM. The molecular ratio of rhodamine-phalloidin to the actin monomer was approximately 1:40, which depends on the amount and thickness of the actin bundles formed.

42

Kingo Takiguchi et al.

3. Morphogenesis of Giant Liposomes Encapsulating Actin and Its Cross-linking Proteins 3.1. G-Actin and its polymerization in liposomes (natural swelling method) Liposomes made of neutral and acidic phospholipids assumed spherical or tubular shapes in buffer A, and their shapes always fluctuated in solution (Fig. 3.3A) until the encapsulated actin polymerized. Liposomes containing only actin-cross-linking protein or G-actin also assumed spherical or tubular shapes in buffer A and always fluctuated in shape. The line-like images of liposomes obtained by dark-field microscopy have a uniform brightness which is almost identical to that of single-layered membranes from A

B

C

Figure 3.3 Morphological changes of actin-containing liposomes caused by actin assembly without an actin-cross-linking protein. (A) Dark-field microscopic images of liposomes made of neutral (PC) and acidic (PG) phospholipids (1:1, mol/mol). (B) Transformed liposomes obtained by the polymerization of encapsulated G-actin into F-actin. Four disk-shaped and one flat spoon-shaped liposomes are shown (viewed from the top). The lipid composition was PE and PG (1:1, mol/mol), and the concentration of encapsulated actin was 100 mM. (C) A fluorescent micrograph that shows the F-actin localization within a disk-shaped liposome obtained by actin assembly (top view). The assembled F-actin was labeled with rhodamine-phalloidin. The lipid composition was PC and PG (1:1, mol/mol), and the concentration of the encapsulated G-actin was 50 mM. All scale bars represent 5 mm. Note that here we can distinguish a disk-shaped liposome from a spherical one by shifting the focus plane of the microscope, or by continuous observation of each free-tumbling liposome in a microscope specimen. The figure is adopted from Honda et al. (1999).

Construction of Actin-Encapsulating Liposomes

43

erythrocyte ghosts whose associated proteins have been removed by protease treatment. Polymerization of the encapsulated G-actin into F-actin was achieved by raising the temperature, and the subsequent morphological changes of liposomes were monitored. When the actin polymerized, the liposomes transformed into flat disk or spoon shapes (Fig. 3.3B) (Miyata and Hotani, 1992). The F-actin that had polymerized in liposomes spontaneously aligned along the periphery of the flattened liposomes to form bundles (Fig. 3.3C). These liposomes are fairly rigid, as judged by the low fluctuation in their shapes. Their membranes were quiescent, and bending motions were restrained.

3.2. Actin and fascin (natural swelling method) When G-actin was encapsulated into liposomes together with fascin and then polymerized, the F-actin bundled tightly. Nearly 60% of liposomes transformed into lemon shapes and straight rigid projections subsequently developed from the tip(s), and eventually they became elongated straight tube(s) (Fig. 3.4A). Fluorescence labeled F-actin was always detected in the long projecting membrane regions (Fig. 3.4A, left), indicating that the fascin-mediated actin bundles are responsible for the formation of the membrane projections. The frequency of transformed liposomes and the lengths of their projections increased at higher concentrations of encapsulated actin or fascin.

3.3. Actin and a-actinin (natural swelling method) Unlike fascin, a-actinin had only a weak effect on morphological changes of actin-containing liposomes. When actin polymerized, most liposomes coencapsulating a-actinin transformed into disk shapes similar to those generated by encapsulating and polymerizing G-actin in the absence of a linking protein. The polymerized actin bundled and then aligned along the periphery. Liposomes transformed into disk shapes possessing long projections only in rare cases (<5%), and some of those projections were not straight (Fig. 3.4B). F-actin bundles always existed in those projected regions (Fig. 3.4B, left), indicating that the a-actinin-mediated actin bundles are also responsible for the formation of membrane projections and occasionally acquired enough strength and rigidity to push out the lipid bilayer membrane.

3.4. Actin and filamin (natural swelling method) When actin and filamin were encapsulated together into liposomes, they became very rigid and swollen as the encapsulated actin assembled. More than 50% of liposomes transformed into ellipsoid- or nut-like shapes

44

Kingo Takiguchi et al.

A

B

C

D

Figure 3.4 Morphological changes of liposomes caused by actin assembly with actincross-linking proteins (left: fluorescence, right: dark-field). (A) Transformed liposomes with fascin. 50 mM G-actin and 3 mM fascin were encapsulated and the lipid composition was PC and PG (1:1, mol/mol). Since the fascin-mediated actin bundles were too bright in dark-field microscope, a lower actin concentration was used in the case of fascin. There was, however, no difference in the liposome transformation between 50 and 100 mM actin. (B) Transformed liposomes with a-actinin. 100 mM G-actin and 2 mM aactinin were encapsulated and the lipid composition was PE and PG (1:1, mol/mol). (C) Transformed liposomes with filamin. 100 mM G-actin and 1 mM filamin were encapsulated and the lipid composition was PE and PG (1:1, mol/mol). (D) 10 (upper) or 25 mM (bottom) G-actin and 5 mM BBMI were encapsulated and the lipid composition was PC and PG (1:1, mol/mol). Since actin and BBMI were too difficult to coencapsulate into liposomes by the natural swelling method because they immediately formed large rigor complexes outside the liposomes, lower concentrations of actin were examined. The shape of each transformed liposome was stable and did not change. Fluorescence images show the localization of rhodamine-labeled F-actin bundles. In each case, before the polymerization of actin, the liposomes took spherical or tubular shapes and always fluctuated in shape as did the empty ones. In addition, each transformed shape started out as either spherical or tubular. All scale bars represent 5 mm. The figure is adopted from Honda et al. (1999).

Construction of Actin-Encapsulating Liposomes

45

(Fig. 3.4C) and their inner spaces were filled with three-dimensional networks of F-actin of uniform distribution (Fig. 3.4C, left). Although the F-actin gels mediated by filamin had high viscosity, they could not be observed by dark-field microscopy. Unlike liposomes containing fascin or a-actinin, liposomes with filamin never formed long membrane projections (Fig. 3.4C).

3.5. Actin and BBMI (natural swelling method) BBMI can cross-link between F-actins, as well as between F-actin and a liposomal membrane. We found that because of this property, when BBMI was coencapsulated into liposomes, it effectively transformed them into protruded shapes at a lower concentration of actin compared to the other actin-cross-linking proteins (Fig. 3.4D). In the presence of BBMI, the actin bundles formed are promptly recruited and bound to the inner surface of the liposome membrane resulting in efficient liposome transformation. Consistently, the actin bundles were observed at the membranes of the transformed liposomes (Fig. 3.4D, left).

3.6. F-Actin (spontaneous transfer method) Although we prepared liposomes encapsulating only F-actin (at 200 mM), which had already polymerized before the encapsulation, using the spontaneous transfer method, they showed no transformation (Fig. 3.5). On the other hand, Miyata et al. (1999) showed that liposomes encapsulating the same concentration of G-actin prepared by the natural swelling method grew protrusive structures as the actin polymerized. These results are consistent with the pioneering reports: the assembly of G-actins at the end of the membrane-pushing F-actin, that is, the mechanical force generated during actin polymerization and elongation, is required for the liposomal transformation (Hotani et al., 2003; Maemichi et al., 2008; Miyata et al., 1999).

3.7. ActoHMM (spontaneous transfer method) Figure 3.6 shows liposomes that have entrapped F-actin together with HMM. In contrast to the images with F-actin only (Fig. 3.6A), in the presence of HMM, F-actin assemblies, such as bundles and networks, are generated (Fig. 3.6B–D). We also confirmed the appearance of similar assemblies of F-actin in aqueous solution in control experiments. As exemplified in Fig. 3.6C and D, with increased HMM concentration, deformed liposomes tended to appear reproducibly, although a large fraction of liposomes remained spherical.

46

Kingo Takiguchi et al.

Figure 3.5 Confocal microscopic images of giant PC liposomes encapsulating F-actin obtained by the spontaneous transfer method (upper and bottom left: fluorescence, bottom right: transmission). The concentration of F-actin inside the liposomes is 200 mM. The fluorescence images show the distribution of rhodamine-phalloidinlabeled F-actin. The bottom panels show enlarged images of an individual F-actinencapsulating liposome. The small spherical objects situated on the surfaces of liposomes in the transmission images are attributed to oil droplets in the water phase (Yamada et al., 2006). Scale bars represent 50 mm (upper) or 10 mm (bottom). The figure is adopted from Takiguchi et al. (2008).

3.8. Actin and S-1 (spontaneous transfer method) When S-1 rather than HMM was coencapsulated with F-actin into giant liposomes, the liposomes showed a uniform distribution of F-actins and maintained their spherical shapes, even under conditions where excess S-1 was coencapsulated (Fig. 3.6E).

3.9. Mechanism determining liposome morphology The transformed morphology of liposomes closely correlated with the structure of actin bundles or networks (Fig. 3.1). For instance, when fascin was coencapsulated, actin bundles and liposome projections were as straight

Construction of Actin-Encapsulating Liposomes

47

A

50 mM F-actin alone

B 50 mM F-actin + 3.8 mM HMM

C 50 mM F-actin + 5.0 mM HMM

D 50 mM F-actin + 7.5 mM HMM

E 50 mM F-actin + 60 mM S-1

Figure 3.6 Confocal microscopic images of actoHMM-encapsulating giant liposomes obtained by the spontaneous transfer method (left: fluorescence, right: transmission). The concentrations of F-actin inside the liposomes are 50 mM (A–E). The fluorescence images show the distribution of F-actin. (A) Images of liposomes encapsulating F-actin alone are shown as a control. (B–D) The concentrations of HMM coencapsulated are 3.8, 5.0, and 7.5 mM, respectively (indicated at the top of each panel). In C and D, liposomes possessing nonspherical irregular shapes are shown. (E) An actoS-1encapsulating giant liposome obtained by the spontaneous transfer method. The concentration of coencapsulated S-1 is 60 mM. Transmission images show the existence of small oil droplets around the liposomes, which were squeezed out from the oil phase. Bars represent 10 mm. The figure is adopted from Takiguchi et al. (2008).

48

Kingo Takiguchi et al.

and rigid as needles, and showed no bending (Fig. 3.4A). This might arise from the higher rigidity of actin bundles mediated by fascin. On the other hand, bundles and projections were flexible and usually curved when a-actinin was coencapsulated. F-actin bundles always existed along the periphery of the central discoid portion as well as in the projecting regions (Fig. 3.4B). These results indicate that, in most cases, the elasticity of actin bundles mediated by a-actinin was not sufficiently strong to push out the lipid bilayer. Liposomes encapsulating filamin never formed projections (Fig. 3.4C), which probably results from the fact that filamin makes a three-dimensional gel rather than bundles of F-actin. Liposomes encapsulating BBMI transformed into protruded shapes even when the concentration of encapsulated actin was comparably lower, probably because the actin bundles formed with BBMI were bound to the liposome membrane efficiently by BBMI (Fig. 3.4D). These results indicate that the mechanical force generated by actin polymerization itself is strong enough to overcome the tension of liposomal membranes and can push them out if the actin filaments are tightly cross-linked since individual ones are very flexible. Fascin has a lower affinity and bundling activity to F-actin compared with a-actinin, but liposomes possessing projection(s) were formed much more frequently and their projections had higher rigidity when actin was assembled with fascin rather than with a-actinin (Fig. 3.4A and B). This might result from the fact that fascin forced the F-actin bundles to cross-link more tightly and rigidly than did a-actinin (Fig. 3.1). Moreover, liposomes encapsulating filamin never formed projections (Fig. 3.4C), even though filamin had the highest affinity to F-actin among these three proteins and had a similar stoichiometry to F-actin with a-actinin. This probably results from the fact that filamin makes a three-dimensional gel rather than bundles of F-actin (Fig. 3.1). In liposome-free conditions, it has been shown that the affinity of an actin-cross-linking protein for actin determines the mechanical properties of actin filament gels (Wachsstock et al., 1993, 1994; Xu et al., 1998). However, while the transformation frequency depends on the concentrations of actin and cross-linking proteins, the morphology does not. The results obtained from our study indicate that the manner (Fig. 3.1) rather than the quantity (such as the binding affinity or stoichiometry) of cross-linking is essential for the morphogenesis of actin-containing liposomes (Honda et al., 1999). Myosins can be classified into two groups according to their headstructure, that is, double- or single-headed myosins. HMM, which belongs to the double-headed type and has two actin-binding motor domains, is able to cross-link F-actins into actin bundles or actin gels, and moreover can transform the bundles or gels (Tanaka-Takiguchi et al., 2004). On the other hand, S-1, which belongs to the single-headed type and has only one actin-binding motor domain, is unable to cross-link F-actins. Liposomes coencapsulating HMM but not S-1 show morphological changes. From these results, one can conclude that simply bundling or redistributing

Construction of Actin-Encapsulating Liposomes

49

F-actin, that is, the inter-filament physical interactions, can be a motive force for organizing the actin networks and the liposomal morphogenesis, as well as the polymerization of G-actin and the subsequent bundling of F-actin (Fig. 3.6C and D).

4. Concluding Remarks In this chapter, we describe the experimental procedures (natural swelling and spontaneous transfer method) used to form cell-sized liposomes containing actin and actin-cross-linking proteins. Furthermore, by studying the morphogenesis of the liposomes thus obtained, we successfully demonstrated that the mechanical force generated by the assembly of actin network structures itself is strong enough to transform cell-sized giant liposomes, and that the manner, rather than the quantity, of cross-linking is essential for the morphogenesis of liposomes encapsulating actin and actin-cross-linking proteins. Especially, by the spontaneous transfer method, we successfully constructed giant liposomes encapsulating both F-actin and HMM at high concentrations in the presence of 5 mM MgCl2 and 50 mM KCl. As a result, it could be shown that the encapsulated actoHMM formed self-organized actin network-like structures, and nonspherical liposomes were obtained in a reproducible manner. Divalent cations are indispensable for life, for example, the fuel of most of the machinery of living cells is Mg-ATP, and Ca2þ is a key factor in a number of biological regulatory systems. By the natural swelling method, however, the efficiency of construction of giant liposomes significantly decreased in the presence of divalent cations. In addition, the natural swelling method cannot control the concentrations of encapsulated species. On the other hand, the spontaneous transfer method enables us to efficiently make giant liposomes encapsulating the desired amounts of F-actin up to 200 mM, even in the presence of 5 mM MgCl2. Note that 200 mM is comparable to the actin concentration expressed in living cells ( Janson et al., 1991; Pollard et al., 2000), and is the upper limit of F-actin concentration that allows handling, including pipetting, due to its very high viscosity. By evaluation of the fluorescence intensity, the encapsulated concentration of F-actin inside the liposomes is almost constant and F-actin exists in a homogeneous manner. In previous studies using the natural swelling method, the highest concentration of actin that could be encapsulated in liposomes was 200 mM, when gel-filtrated G-actin was encapsulated and then polymerized into F-actin with KCl introduced by electroporation after the formation of liposomes (Miyata et al., 1999). Although such studies are interesting, it has been impossible to efficiently encapsulate desired amounts of F-actin inside liposomes using the natural swelling method.

50

Kingo Takiguchi et al.

At this time, hundreds of actin-associating proteins and ten classes of different types of myosin motors have been identified. Usual analytical studies, however, might be unable to answer how they share roles and cooperate with each other in cells because living cells are typically a very complex system. This reconstitutive and very efficient system utilizing cell-sized giant liposomes containing actoHMM prepared by the spontaneous transfer method might represent the first critical step for developing a motile artificial cell model, which will generate spontaneous motion in a system similar to but much simpler than living cells. The important problems that remain are how to supply sufficient amounts of ATP to actoHMM existing inside the liposomes, and how to control the reactions between F-actin, HMM, and ATP. Thus, the next step may be to perform experiments by merging different liposomes, where F-actin, HMM, or ATP are encapsulated separately. Recently, a-hemolysin, a membrane-pore forming toxin from bacteria, has found application in molecular transport into vesicles (Gibrat et al., 2008; Noireaux and Libchaber, 2004; Pontani et al., 2009), and that novel methodology may prove useful to supply ATP to actoHMM-containing giant liposomes.

ACKNOWLEDGMENTS The studies described were supported by the Japan Society for the Promotion of Science ( JSPS) under a Grant-in-Aid for Creative Scientific Research (Project No. 18GS0421) and by the MEXT of Japan under a Grant-in-Aid (Scientific Research of Priority Areas, System Cell Engineering by Multi-Scale Manipulation, Project No. 20034024 and No.17076007).

REFERENCES Albrecht-Buehler, G., and Lancaster, R. M. (1976). A quantitative description of the extension and retraction of surface protrusions in spreading 3T3 mouse fibroblasts. J. Cell Biol. 71, 370–382. Angelova, M. I., and Dimitrov, D. S. (1986). Liposome electroformation. Faraday Discuss. 81, 303–311. Bangham, A. D. (1995). Surrogate cells or Trojan horses. The discovery of liposomes. BioEssays 17, 1081–1088. Bangham, A. D., Standish, M. M., and Watkins, J. C. (1965). Diffusion of univalent ions across the lamellae of swollen phospholipids. J. Mol. Biol. 13, 238–252. Ba¨rmann, M., Ka¨s, J., Kurzmeiser, H., and Sackmann, E. (1992). A new cell model-actin networks encaged by giant vesicles. In ‘‘The Structure and Conformation of Amphiphilic Membranes,’’ (R. Lipowsky, D. Richter, and K. Kremer, eds.), Vol. 66, Springer Proceedings in Physics. pp. 137–143. Springer, Heidelberg. Bretscher, A. (1991). Microfilament structure and function in the cortical cytoskeleton. Annu. Rev. Cell Biol. 7, 337–374. Byers, H. R., and Fujikawa, K. (1982). Stress fibers in cells in situ: Immunofluorescence visualization with antiactin, antimyosin, and anti-a-actinin. J. Cell Biol. 93, 804–811.

Construction of Actin-Encapsulating Liposomes

51

Cant, K., Knowles, B. A., Mooseker, M. S., and Cooley, L. (1994). Drosophila singed, a fascin homolog, is required for actin bundle formation during oogenesis and bristle extension. J. Cell Biol. 125, 369–380. Claessens, M. M. A. E., Tharmann, R., Kroy, K., and Bausch, A. R. (2006). Microstructure and viscoelasticity of confined semiflexible polymer networks. Nat. Phys. 2, 186–189. Collins, K., Sellers, J., and Matsudaira, P. (1991). Myosin I: A new insight into the mechanism and cellular significance of actin-based motility. Adv. Biophys. 27, 221–226. Coluccio, L. M. (1997). Myosin I. Am. J. Physiol. 273, C347–C359. Cortese, J. D., Schwab, B. III, Frieden, C., and Elson, E. L. (1989). Actin polymerization induces a shape change in actin-containing vesicles. Proc. Natl. Acad. Sci. USA 86, 5773–5777. Ebashi, S., and Ebashi, F. (1965). Alpha-actinin, a new structural protein from striated muscle. I. Preparation and action on actomyosin-ATP interaction. J. Biochem. 58, 7–12. Edwards, R. A., and Bryan, J. (1995). Fascins, a family of actin bundling proteins. Cell Motil. Cytoskel. 32, 1–9. Edwards, R. A., Herrera-Sosa, H., Otto, J., and Bryan, J. (1995). Cloning and expression of a murine fascin homolog from mouse brain. J. Biol. Chem. 270, 10764–10770. Funakoshi, K., Suzuki, H., and Takeuchi, S. (2007). Formation of giant lipid vesiclelike compartments from a planar lipid membrane by a pulsed jet flow. J. Am. Chem. Soc. 129, 12608–12609. Fygenson, D. K., Marko, J. F., and Libchaber, A. (1997). Mechanics of microtubule-based membrane extension. Phys. Rev. Lett. 79, 4497–4500. Gibrat, G., Pastoriza-Gallego, M., Thiebot, B., Breton, M.-F., Auvray, L., and Pelta, J. (2008). Polyelectrolyte entry and transport through an asymmetric a-hemolysin channel. J. Phys. Chem. B 112, 14687–14691. Grazi, E., Trombetta, G., Magri, E., Cuneo, P., and Schwienbacher, C. (1994). a-Actinin from chicken gizzard: At low temperature, the onset of actin-gelling activity correlates with actin bundling. Biochem. J. 298, 129–133. Ha¨ckl, W., Ba¨rmann, M., and Sackmann, E. (1998). Shape changes of self-assembled actin bilayer composite membranes. Phys. Rev. Lett. 80, 1786–1789. Hamada, T., Miura, Y., Komatsu, Y., Kishimoto, Y., Vestergaard, M., and Takagi, M. (2008). Construction of asymmetric cell-sized lipid vesicles from lipid-coated water-inoil microdroplets. J. Phys. Chem. B 112, 14678–14681. Hartwig, J. H., and Kwiatkowski, D. H. (1991). Actin-binding proteins. Curr. Opin. Cell Biol. 3, 87–97. Hase, M., and Yoshikawa, K. (2006). Structural transition of actin filament in a cell-sized water droplet with a phospholipid membrane. J. Chem. Phys. 124, Art. no. 104903. Hase, M., Yamada, A., Hamada, T., Baigl, D., and Yoshikawa, K. (2007). Manipulation of cell-sized phospholipid-coated microdroplets and their use as biochemical microreactors. Langmuir 23, 348–352. Higashi-Fujime, S. (1983). Phosphorylation of myosin light chain modulates the in vitro movement of fibrils composed of actin and myosin filaments from skeletal muscle. J. Biochem. 94, 1539–1545. Hock, R. S., Davis, G., and Speicher, D. W. (1990). Purification of human smooth muscle filamin and characterization of structural domains and functional sites. Biochemistry 29, 9441–9451. Honda, M., Takiguchi, K., Ishikawa, S., and Hotani, H. (1999). Morphogenesis of liposomes encapsulating actin depends on the type of actin-crosslinking. J. Mol. Biol. 287, 293–300. Hotani, H. (1984). Transformation pathways of liposomes. J. Mol. Biol. 178, 113–120. Hotani, H., and Miyamoto, H. (1990). Dynamic features of microtubules as visualized by dark-field microscopy. Adv. Biophys. 26, 135–156.

52

Kingo Takiguchi et al.

Hotani, H., Inaba, T., Nomura, F., Takeda, S., Takiguchi, K., Itoh, T. J., Umeda, T., and Ishijima, A. (2003). Mechanical analyses of morphological and topological transformation of liposomes. Biosystems 71, 93–100. Hou, L., Luby-Phelps, K., and Lanni, F. J. (1990). Brownian motion of inert tracer macromolecules in polymerized and spontaneously bundled mixtures of actin and filamin. J. Cell Biol. 110, 1645–1654. Imamura, M., Endo, T., Kuroda, M., Tanaka, T., and Masaki, T. (1988). Substructure and higher structure of chicken smooth muscle alpha-actinin molecule. J. Biol. Chem. 263, 7800–7805. Janson, L. W., Kolega, J., and Taylor, D. L. (1991). Modulation of contraction by gelation/ solation in a reconstituted motile model. J. Cell Biol. 114, 1005–1015. Jontes, J. D., and Milligan, R. A. (1997). Three-dimensional structure of brush border ˚ resolution by electron microscopy and image analysis. myosin-I at approximately 20 A J. Mol. Biol. 266, 331–342. Kaneko, T., Itoh, T. J., and Hotani, H. (1998). Morphological transformation of liposomes caused by assembly of encapsulated tubulin and determination of shape by microtubuleassociated proteins (MAPs). J. Mol. Biol. 284, 1671–1681. Lasic, D. D. (1995). Applications of liposomes. In ‘‘Handbook of Biological Physics, 1a, Structure and Dynamics of Membranes,’’ (R. Lipowsky and E. Sackmann, eds.), pp. 491–519. Elsevier, Amsterdam. Limozin, L., and Sackmann, E. (2002). Polymorphism of cross-linked actin networks in giant vesicles. Phys. Rev. Lett. 89, Art. no. 168103. Limozin, L., Ba¨rmann, M., and Sackmann, E. (2003). On the organization of self-assembled actin networks in giant vesicles. Eur. Phys. J. E 10, 319–330. Limozin, L., Roth, A., and Sackmann, E. (2005). Microviscoelastic moduli of biomimetic cell envelopes. Phys. Rev. Lett. 95, Art. no. 178101. Lipowsky, R. (1991). The conformation of membranes. Nature 349, 475–481. Maemichi, H., Shikinaka, K., Kakugo, A., Furukawa, H., Osada, Y., and Gong, J. P. (2008). Morphogenesis of liposomes caused by polycation-induced actin assembly formation. Langmuir 24, 11975–11981. Meyer, R. K., and Aebi, U. (1990). Bundling of actin filaments by a-actinin depends on its molecular length. J. Cell Biol. 110, 2013–2024. Mimura, N., and Asano, A. (1987). Further characterization of a conserved actin-binding 27-kDa fragment of actinogelin and a-actinins and mapping of their binding sites on the actin molecule by chemical cross-linking. J. Biol. Chem. 262, 4717–4723. Mitchison, T. J., and Cramer, L. P. (1996). Actin-based cell motility and cell locomotion. Cell 84, 371–379. Mitchison, T. J., and Kirschner, M. (1988). Cytoskeletal dynamics and nerve growth. Neuron 1, 761–772. Miyata, H., and Hotani, H. (1992). Morphological changes in liposomes caused by polymerization of encapsulated actin and spontaneous formation of actin bundles. Proc. Natl. Acad. Sci. USA 89, 11547–11551. Miyata, H., Nishiyama, S., Akashi, K., and Kinosita, K. Jr. (1999). Protrusive growth from giant liposomes driven by actin polymerization. Proc. Natl. Acad. Sci. USA 96, 2048–2053. Molony, L., McCaslin, D., Abernethy, J., Paschal, B., and Burridge, K. (1987). Properties of talin from chicken gizzard smooth muscle. J. Biol. Chem. 262, 7790–7795. Muguruma, M., Matsumura, S., and Fukazawa, T. (1990). Direct interactions between talin and actin. Biochem. Biophys. Res. Commun. 171, 1217–1223. Noireaux, V., and Libchaber, A. (2004). A vesicle bioreactor as a step toward an artificial cell assembly. Proc. Natl. Acad. Sci. USA 101, 17669–17674.

Construction of Actin-Encapsulating Liposomes

53

Otto, J. J., Kane, R. E., and Bryan, J. (1980). Redistribution of actin and fascin in sea urchin eggs after fertilization. Cell Motil. 1, 31–40. Pautot, S., Frisken, B. J., and Weitz, D. A. (2003a). Engineering asymmetric vesicles. Proc. Natl. Acad. Sci. USA 100, 10718–10721. Pautot, S., Frisken, B. J., and Weitz, D. A. (2003b). Production of unilamellar vesicles using an inverted emulsion. Langmuir 19, 2870–2879. Pollard, T. D., Blanchoin, L., and Mullins, R. D. (2000). Molecular mechanisms controlling actin filament dynamics in nonmuscle cells. Annu. Rev. Biophys. Biomol. Struct. 29, 545–576. Pontani, L.-L., van der Gucht, J., Salbreux, G., Heuvingh, J., Joanny, J.-F., and Sykes, C. (2009). Reconstitution of an actin cortex inside a liposome. Biophys. J. 96, 192–198. Rodriguez, O. C., Schaefer, A. W., Mandato, C. A., Forscher, P., Bement, W. M., and Waterman-Storer, C. M. (2003). Conserved microtubule-actin interactions in cell movement and morphogenesis. Nat. Cell Biol. 5, 599–609. Schroeder, T. E. (1973). Actin in dividing cells: Contractile ring filaments bind heavy meromyosin. Proc. Natl. Acad. Sci. USA 70, 1688–1692. Shum, H. C., Lee, D., Yoon, I., Kodger, T., and Weitz, D. A. (2008). Double emulsion templated monodisperse phospholipid vesicles. Langmuir 24, 7651–7653. Stachowiak, J. C., Richmond, D. L., Li, T. H., Liu, A. P., Parekh, S. H., and Fletcher, D. A. (2008). Unilamellar vesicle formation and encapsulation by microfluidic jetting. Proc. Natl. Acad. Sci. USA 105, 4697–4702. Takiguchi, K. (1991). Heavy meromyosin induces sliding movements between antiparallel actin filaments. J. Biochem. 109, 520–527. Takiguchi, K., Yamada, A., Negishi, M., Tanaka-Takiguchi, Y., and Yoshikawa, K. (2008). Entrapping desired amounts of actin filaments and molecular motor proteins in giant liposomes. Langmuir 24, 11323–11326. Tanaka-Takiguchi, Y., Kakei, T., Tanimura, A., Takagi, A., Honda, M., Hotani, H., and Takiguchi, K. (2004). The elongation and contraction of actin bundles are induced by double-headed myosins in a motor concentration-dependent manner. J. Mol. Biol. 341, 467–476. Taylor, D. L., and Condeelis, J. S. (1979). Cytoplasmic structure and contractility in amoeboid cells. International Review of Cytology Vol. 56, pp. 57–144. Academic Press, New York. Wachsstock, D. H., Schwartz, W. H., and Pollard, T. D. (1993). Affinity of a-actinin for actin determines the structure and mechanical properties of actin filament gels. Biophys. J. 65, 205–214. Wachsstock, D. H., Schwartz, W. H., and Pollard, T. D. (1994). Cross-linker dynamics determine the mechanical properties of actin gels. Biophys. J. 66, 801–809. Xu, J., Wirtz, D., and Pollard, T. D. (1998). Dynamic cross-linking by a-actinin determines the mechanical properties of actin filament networks. J. Biol. Chem. 273, 9570–9576. Yamada, A., Yamanaka, Y., Hamada, T., Hase, M., Yoshikawa, K., and Baigl, D. (2006). Spontaneous transfer of phospholipid-coated oil-in-oil and water-in-oil micro-droplets through an oil/water interface. Langmuir 22, 9824–9828. Yamashiro, S., Yamakita, Y., Ono, S., and Matsumura, F. (1998). Fascin, an actin-bundling protein, induces membrane protrusions and increases cell motility of epithelial cells. Mol. Biol. Cell 9, 993–1006. Yamashiro-Matsumura, S., and Matsumura, F. (1985). Purification and characterization of an F-actin-bundling 55-kilodalton protein from HeLa cells. J. Biol. Chem. 260, 5087–5097.

This page intentionally left blank

C H A P T E R

F O U R

Reconstitution of Membrane Budding with Unilamellar Vesicles Anna V. Shnyrova and Joshua Zimmerberg Contents 1. Introduction 2. M Protein Purification 3. Evaluation of the Membrane Activity of M Protein Through its Interaction with Intermediate-Sized Unilamellar Liposomes 3.1. Remarks about IUV formation and characterization 3.2. IUV preparation 3.3. Analysis of M protein binding to IUVs of different lipid compositions 3.4. M protein interaction with IUVs: Leakage of aqueous dyes 4. Reconstitution of M-Protein-Driven Membrane Budding on GUVs 4.1. GUV preparation 4.2. Interaction of M protein with GUVs 5. Concluding Remarks References

56 57 58 58 59 61 65 66 66 69 73 73

Abstract Enveloped virus particles select their lipid–protein components and egress by budding from the host cell membranes. The matrix protein of many enveloped viruses has been proposed as a crucial element for viral budding; however, molecular mechanisms behind membrane remodeling by the matrix protein are yet to be unraveled. Here, we describe a set of in vitro functional reconstitution assays that allow quantitative evaluation of both, membrane binding and creation of membrane curvature by the matrix protein isolated from Newcastle Disease Virus. Individual budding events orchestrated by the matrix protein can be resolved in real time. The assays may be applied for direct reconstitution of the on-membrane action of cellular proteins involved in membrane curvature induction upon binding in vivo. Laboratory of Cellular and Molecular Biology, Program in Physical Biology, Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, Maryland, USA Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64004-0

55

56

Anna V. Shnyrova and Joshua Zimmerberg

1. Introduction Cellular life and death depends critically on the membrane curvature creation by cellular proteins. Since the works by Palade (1953, 1975), the mechanisms of cellular trafficking and membrane deformation have been the target of extensive studies and many protein candidates for curvature induction have been proposed. However, it is not easy to extract the exact molecular mechanism behind the cellular curvature creation from in vivo observations, as cellular membranes are very crowded environments. The bottom-up approach, where isolated membrane systems are used to reconstitute cellular events with a minimum set of proteins, simplifies greatly this task. Recent reconstitution studies have proved this approach be a handy tool for a better understanding of curvature creation in cells. One illustrative example of these kinds of studies is the reconstitution of budding of enveloped viruses on giant unilamellar vesicles (GUVs) (Shnyrova et al., 2007; Solon et al., 2005). Despite the generally accepted notion that the host cell participates in viral budding, final virus particles include only small traces of cellular proteins (Briggs et al., 2003). This suggests that enveloped viruses ultimately rely on their own protein machinery for vesicle egress ion. Budding of most retroviruses (except Foamy Virus) and many paramixoviruses have been proposed to rely critically on the viral matrix protein, although other viral proteins are often enlisted as a requirement for a minimal budding machinery (reviewed by Welsch et al., 2007). The matrix (M) protein is the major component of most enveloped viruses, where it forms a tightly packed shell beneath the lipid envelope of the virus, thus resembling the caveolin coat of the endocytic vesicles, though with the opposite topology (Welsch et al., 2007). For many paramixoviruses, including Newcastle Disease Virus (NDV), expression of M protein in the absence of other viral proteins in the cell leads to formation of virus-like particles (VLPs) ( Jayakar et al., 2004; Pantua et al., 2006; Takimoto and Portner, 2004). These results indirectly suggested that NDV M protein possesses the energy and functionality necessary to create membrane shape of desired geometry. The mechanism of negative membrane curvature induction, however, still remained unresolved and the question remained open whether M alone, with just a lipid bilayer, is the minimum requirement for this process of membrane remodeling. Here, we describe an in vitro assay that allows to evaluate, to monitor in real time, and to quantify the ability of a membrane bound protein (the NDV M protein) to induce negative curvature upon its interaction with the membrane of giant liposomes. This assay offers unique insights into the molecular mechanisms driving this type of membrane remodeling in vivo.

Modeling Viral Budding with Unilamellar Vesicles

57

2. M Protein Purification An in vitro reconstitution of membrane activity of a protein requires highly purified protein in its native conformation, as misfolded proteins or sample contaminants can bind to the membrane and alter its behavior. These requirements frequently preclude and limit the use of recombinant membrane-active proteins from bacteria, as they often fold incorrectly in the absence of other components of their native environment and tend to aggregate in physiological conditions. Instead, we purify the M protein directly from its native environment: the NDV. This method yields high amounts of sufficiently purified protein. NDV LaSota strain (Charles Rivers Laboratories Inc., MA) is grown in the allantoic cavity of 11-day-old pathogen-free chick embryos (CBT Farm, MD). Upon infection, embryos are allowed to grow for 48 h at 37 ºC and 55% humidity. The allantoic fluid is then collected and stored at
4 ºC. Virus is purified the next day as described previously (Garcia-Sastre et al., 1989). Scheid and Choppin’s (1973) purification protocol is used as the reference for the M protein purification from NDV. To separate the lipid-anchored proteins from the viral nucleocapsid, the virus pellet is resuspended in 10 mL of 1 M KCl, 10 mM Tris–HCl and 5 mM CaCl2 at pH 7.4 (buffer A) with 2% the of detergent Triton X-100 and then incubated 30 min at room temperature with constant agitation. After incubation, the sample is centrifuged at 200,000g for 2 h at 4 ºC. The supernatant, containing the main components of the viral envelope is recovered and dialyzed against 2 L of buffer A, lacking KCl (five changes for 36 h). The low ionic strength of the dialysis buffer induces M protein precipitation and separation from the rest of the viral membrane proteins. The contents of the dialysis bag are centrifuged at 9000g for 30 min at 4 ºC. The pellet containing M protein is then dissolved in 1 mL of buffer B, which contains 1 M KCl, 20 mM HEPES, and 0.2 EDTA, pH 7.4. Final protein concentration is determined using the BCA protein assay kit (Pierce, IL) and is adjusted to lower than 75 mM in order to avoid protein precipitation during storage. Various viral M proteins are known for their ability to self-assemble into ordered structures at low salt conditions. The speed and efficiency of such aggregations are usually determined by the ionic strength and protein concentration (McCreedy et al., 1990). This sensitivity has been routinely used to separate matrix proteins from the rest of the viral proteins: M protein appears mostly as a monomer at high ionic strength, but aggregates slowly and irreversibly at physiological ionic strength (Sagrera et al., 1998). Accordingly, the use of high ionic strength (>1 M) solution during

58

Anna V. Shnyrova and Joshua Zimmerberg

purification is critical for the disassembly of the interior of NDV from its envelope. However, experimental observations indicate that M protein interaction with NP protein is not totally disrupted during the purification procedure based only on salt concentration changes, and that the aggregation of M protein may be in part due to the formation of this M–NP complexes. Thus, addition of calcium in millimolar concentration during M protein purification allows for a better separation of M protein from the NP complex, presumably through disruption of some electrostatic interactions, therefore leading to highly purified M protein. The protein purity is estimated from the reducing SDS–PAGE by analysis of the areas under the intensity peaks correspondent to protein bands, as was described previously (Coorssen et al., 2002).

3. Evaluation of the Membrane Activity of M Protein Through its Interaction with Intermediate-Sized Unilamellar Liposomes GUVs provide an excellent tool for studying the membrane activity of a particular protein, as they allows for real time observation of shape transformations by bright-field or fluorescence microscopy. However, experiments on GUVs may end up frustrating if the binding of the protein to a particular lipid composition is not previously analyzed by quantitative methods. In this aspect, large (200–500 nm in diameter) and intermediatesized (100–200 nm in diameter) unilamellar liposomes (LUVs and IUVs, respectively) are suitable for the initial checkup of the protein–membrane interactions.

3.1. Remarks about IUV formation and characterization There are various methods for IUV preparation described elsewhere (a great reference source is New, 1990b). The choice of a particular method mainly depends on the desired lipid composition of the vesicles and the buffer used. For the compositions and conditions described below, a combination of gentle hydration method followed by extrusion through appropriate polycarbonate membrane gives reproducible results. However, we encourage the reader to explore a variety of methods for IUV preparation in order to find an appropriate one or a combination of several that suits the particular experimental case. The success in the IUV preparation relies on achieving both the desirable size and the unilamellarity of the vesicles. The first parameter is crucial in terms of the curvature as well as the lateral membrane tension of the vesicles. For example, the membrane binding of many proteins depends on

Modeling Viral Budding with Unilamellar Vesicles

59

membrane curvature (e.g., Peter et al., 2004; Vaccaro et al., 1993). Thus, if the vesicle population exhibits high size variability, results of the binding assays may be truly confusing. Hence, the size distribution of the IUV sample is to be measured routinely, for example, by dynamic light-scattering. The lamellarity of the vesicles is extremely important when quantitative results are obtained from a binding assay. Even if the phospholipid concentration is determined for a particular IUV sample, the amount of the lipid accessible to the proteins, that is, lipids in the outer monolayer of a vesicle, may be up to orders of magnitude lower than for a unilamellar vesicle. Multilamellarity also affects the amount of encapsulated material into the vesicle and the effective membrane rigidity. This may lead to potential artifacts in experiments, for example, in measurements of the proteininduced leakage of encapsulated material by fluorescence spectroscopy. Thus, the routine of checking the unilamellarity of the IUVs should be adopted as well, for example, by TNBS assay (New, 1990a). A direct measurement of the IUVs lamellarity is offered by the electron microscopy imaging of ammonium molibdate negatively stained samples, as described previously (Bugelski et al., 1990). The use of uranyl acetate (UA) stain with this kind of samples is highly discouraged, as UA interferes with lipid membranes, especially with the charged ones (Caffrey et al., 1987; Parsegian et al., 1981).

3.2. IUV preparation For the M protein experiments, IUVs are prepared by a combination of the gentle hydration method (Reeves and Dowben, 1969) with the extrusion procedure (Mayer et al., 1986). IUV are formed from mixtures of the lipid stocks in chloroform of dioleoyl phosphatidylcholine (DOPC), palmitoyl oleoyl phosphatidylcholine (POPC), dioleoyl phosphatidylethanolamine (DOPE), dioleoyl phophatidylglycerol (DOPG) and cholesterol (chol), doped with 0.2 molar percent of lipid fluorescence marker Lissamine Rhodamine B DOPE (Rh-DOPE). All lipids are from Avanti Polar Lipids (AL). IUVs used for the binding assay with M protein are prepared from the following lipid mixtures (in molar ratios): DOPC:DOPE:Chol:Rh-PE 60:29.8:10:0.2 (PE-chol neutral mixture); POPC:Chol:Rh-DOPE 69.8:30:0.2 (chol neutral mixture); POPC:Rh-PE 99.8:0.2 (neutral mixture); POPC:DOPG:Rh-PE 84.8:15:0.2 (charged mixture). The total lipid concentration in the final IUV sample for these mixtures is 1 g/L. For the assay of the M-protein-induced IUV leakage, the PE–chol neutral mixture is employed, with total lipid concentration of 0.5 g/L in the IUV sample. Each lipid mixture (initially in chloroform) is added to a round bottom of 5 or 10 mL flask, suitable for a rotary–evaporator system (such as VV Micro Evaporator (Heidolph Brinkmann LLC, IL) or equivalent). Chloroform and methanol (9:1, v/v) are added to the flask until a final volume of

60

Anna V. Shnyrova and Joshua Zimmerberg

0.5 mL is achieved. The flask if rapidly transferred to the rotary– evaporator, where the evaporation of the organic solvents is carried out with gentle warming of the flask (30 ºC or above the phase transition temperature(s) of the lipids in the mixture) and an intermediate flask rotation speed, in order to obtain a thin and uniform lipid film. Once the visible traces of the organic solvent have disappeared, the evaporation continues for another 15–30 min. Then, the flask is transferred to a highpressure lyophilizer at room temperature for 1 h to ensure the removal of residual solvents. The desired aqueous buffer is added to the flask and the lipid film is left to hydrate overnight at 37 ºC. Special care is taken to avoid shaking the flask during the hydration process. In the case of the IUV preparation for the binding assay, the aqueous buffer is 100 mM KCl, 20 mM HEPES, and 0.2 mM EDTA, pH 7.4 (buffer C). For leakage measurements, an aqueous fluorescent dye should be entrapped inside the IUVs. Thus, buffer D (12.5 mM ANTS, 45 mM DPX, 20 mM KCl, 0.2 mM EDTA, 20 mM HEPES, pH 7.4) contains 8-aminonaphthalene-1,3,6-trisulfonate (ANTS) quenched with p-xylenebis(piridinium bromide) (DPX) (Molecular Probes, Invitrogen Corp., CA) (Du¨zgu¨nes¸, 2003a). The quenching efficiency is proportional to the concentration of ANTS/DPX. The dyes are loaded in to IUV at high concentration ensuring effective quenching of ANTS fluorescence. The leakage measurements are based on the increase of the ANTS fluorescence upon dilution. A small (<1 nm) hole in the IUV membrane is sufficient to let ANTS/DPX release to the bulk solution, thus causing the dilution. Encapsulation of dyes different in size allows estimating the size of such holes. For example, 70 kDa fluorescein isothiocyanate conjugated dextran, FITC-dextran (Sigma-Aldrich Corp., MO), contained in buffer E (75 mg/mL 70 kDa FITC-dextran dissolved in buffer C), requires a much larger (5 nm) hole in the membrane to escape. Iso-osmolarity of buffer D, buffer E, and buffer C is controlled using a Wescor 5500 (UT) vapor pressure osmometer or equivalent. The lipid cloud formed after hydration (mainly containing MLVs and LUVs) is dispersed in the buffer by gentle vortexing of the flask. Using a glass Pasteur pipette, the vesicle dispersion is loaded into an extruder (Lipex Biomembrane Inc., Canada) preassembled with a drain disc and one or two (usually 0.1 mm, if not indicated otherwise) polycarbonate filter (GE Osmonic Labstore, MN). The samples are extruded 15 times. Nonencapsulated ANTS/DPX is separated from IUVs using prepacked Sephadex PD-10 columns (GE Healthcare Biosciences Corp., NJ). Non-trapped 70 kDa FITC-dextran is removed using a column packed with Superose 6 prep grade (GE Healthcare Biosciences Corp.) at a 1 mL/min flow rate. In both cases, buffer C is used as the elution buffer. A crude estimation of the amount of lipids in the preparation is made by measuring the fluorescence signal from Rh-DOPE. For that, 96-well plate apt for

Modeling Viral Budding with Unilamellar Vesicles

61

fluorescence measurements with a plate reader instrument (Synergy 4 microplate reader (Biotek, VT) or similar) is used. Ten microliters aliquots from the IUV suspension before and after extrusion are diluted to 100 mL of buffer C. The Rhodamine (Rh) fluorescence is measured at 550 nm excitation and 590 nm emission wavelengths. The fluorescence of each sample is measured three times and the arithmetic mean of the fluorescence signals is used to estimate the amount of lipid lost during extrusion process by the following formula: lipid loss 

Fbefore
Fafter ; Fbefore

ð4:1Þ

where Fbefore stands for the mean Rh fluorescence signal before extrusion and Fafter, for mean Rh fluorescence signal after the extrusion for each sample prepared. The total lipid concentration of the IUV preparation is then estimated by applying the following formula: CIUV  Ci ð1
lipid lossÞ;

ð4:2Þ

where CIUV stands for the total lipid concentration in the IUV preparation and Ci stands for the total lipid concentration of the initial lipid mixture. We note that this method provides a rough estimation of the amount of total lipid in the final IUV preparation and cannot discriminate between lipid species in nonhomogeneous lipid mixtures. Alternatively the phosphate analysis (Bartlett, 1959; Du¨zgu¨nes¸, 2003b) or liquid chromatography combined with mass spectrometry (Byrdwell, 2005) should be employed if the amount of a particular lipid in the sample is a critical parameter. The size distribution and the lamellarity level of the IUV samples are routinely controlled before using them in M protein experiments. Only IUV preparations that show unimodal size distributions with mean diameters of 120  30 nm by dynamic light-scattering measurements with a dynamic light-scattering instrument (N4 Plus Submicron Particle Sizer (Beckman Coulter Inc., FL) or equivalent) are used in the experiments. Furthermore, the unilamellarity of these samples is also confirmed periodically by negative-staining EM as described in (Bugelski et al., 1990).

3.3. Analysis of M protein binding to IUVs of different lipid compositions Stable binding of M protein to PE–chol neutral, chol neutral, neutral and charged mixtures is assayed by ficoll gradient flotation or sedimentation methods (Fraley et al., 1980). Ficoll solution of 50% (w/v) is prepared by dissolving 25 g of Ficoll PM400 (GE Healthcare Life Sciences, NJ) in buffer C (the amount of this stock increases or decreases proportionally to the amount of samples in the binding assay). This concentration is near the

62

Anna V. Shnyrova and Joshua Zimmerberg

solubility limit of ficoll, so the solution preparation requires slight heating and extensively mixing. Solution containing smaller amount of ficoll are prepared by simple dissolving of the appropriate amount of Ficoll PM400 in buffer C. In simple two-step flotation experiments, 111 mg of M protein (2.8 nmol) is incubated with IUVs at 1:200 molar ratio of M protein to lipid for 5 min at room temperature. In the control experiments, the same amount of the protein or IUVs are diluted to the same volume as the IUV/protein mixture and similarly incubated in buffer C. After incubation, each sample is mixed with 50% (w/v) ficoll so that the ficoll is diluted to 20% (w/v). The final volume of this mixture should not exceed 1.2 mL. Then, two-step ficoll gradients are prepared in a 13  51 mm (4 mL total volume) polycarbonate centrifuge tube (Beckman Coulter Inc., CA). The 20% (w/v) ficoll solution mixed with the protein-IUV samples is carefully transferred to the bottom of the centrifuge tubes. Next, the 10% ficoll solution is layered on top (the volume of this layer is usually greater than 2.5 mL). The last gradient step is composed of 200 mL of buffer C. The gradients are then centrifuged using a swinging bucket rotor SW 55Ti (Beckmann Coulter Inc., CA) at 4 ºC and 150,000g for 30 min. IUV, seen as the Rh-labeled band, settle at the boundary between buffer C and 10% ficoll layers. Similar Rh-labeled band is seen in IUV–M protein mixture. These bands can be better observed and documented by illuminating the centrifuge tubes from the top with a flash light and acquiring a photograph (Fig. 4.1). Multistep gradients are required to resolve whether several subpopulations of IUVs are formed upon the protein adsorption. For example, a six-step gradient is prepared by layering, from the bottom to the top of the centrifuge tube, 500 mL layers of 20%, 15%, 12%, 9%, 6%, and 3% ficoll. The samples are deposited very carefully on top of the last layer. The gradients are centrifuged at 150,000g in a SW55Ti rotor (Beckmann Coulter Inc.) at 4 ºC for 90 min. The location of the protein-IUV band/s depends on the amount of protein and membrane deformation upon its binding to the IUVs. Again, the bands can be observed and documented by illuminating the centrifuge tubes from the top with a light source. These images may be very illustrative, as they give qualitative macroscopic description of the behavior of each sample as compared to the IUV control (Fig. 4.1). Independently of the method used, the gradients are fractionated into aliquots (smaller than the volume of a gradient layer, e.g., 100 mL) by using a peristaltic pump connected through tygon type tubing to 1.5 mm capillary borosilicate glass (#PG10150-4, WPI Inc., FL) carefully brought to the bottom of the centrifuge tube. A fraction collector (Model 2110 (BioRad Labs., CA) or equivalent) is attached to the output of the pump. After being collected, the 10 mL of each fraction is diluted to 100 mL with buffer C in a 96-well plate suitable for fluorescence measurements (Fig. 4.1). The Rh

550 nm em 590 nm ex

Rh fluorescence measurement of all the fractions collected IUV control IUV + protein

fmax Peak

Analysis

FLuorescence, a.u.

Fractionation

Baseline 0 10 20 30 40 50 Gradient fraction #

1.0

0.5

0.0

Normalized fluorescence

63

Modeling Viral Budding with Unilamellar Vesicles

Figure 4.1 Steps of the assay for protein binding to IUV by sedimentation (or flotation). The left panel shows the photographs of the control (only IUV) and experimental sample (IUV plus protein sample) bands after centrifugation in ficoll gradients. The samples are then fractionated and each fraction is analyzed by measuring the Rh fluorescence at 550/590 excitation/emission wavelengths using a microplate reader. The plot on the right represents the fluorescence intensity of the small aliquots from each fraction versus the fraction number. For the case shown, the IUV in the control (black squares) and in the assayed (gray circles) samples distribute predominantly in two similar populations (two fluorescence peaks). SDS–PAGE analysis of the fraction corresponding to the maximum of the smaller peak shows no significant presence of protein. Thus, the quantification of the protein binding is performed only for the fractions in the peak marked with a gray frame. The plot is normalized by subtracting the baseline (black dashed line) and dividing by the maximum value of the fluorescence peak ( fmax). The normalized scale is shown on the right side of the plot. The detailed analysis of the protein binding is further described in the text.

fluorescence, indicating the amount of lipid in each fraction, is then measured at 550 nm excitation wavelength and 590 nm emission wavelength in a plate reader instrument (Synergy 4 microplate reader (Biotek) or similar). This allows the construction of a plot of fluorescence intensity of each fraction versus the fraction number by using a suitable software (Origin 7.0 (OriginLab Corp., MA) or similar) (Fig. 4.1). The peaks in this plot show the distribution of IUV between fractions. The fractions corresponding to the peaks of the distribution of Rh fluorescence (‘‘Rh-pick fractions’’) are selected to explore for the M protein contents by SDS–PAGE. The same fractions from the control samples are also selected and analyzed by SDS–PAGE. It is recommendable to load the maximum sample volume possible into the gel (the 10-well Pierce Precise Protein Gels (Pierce) are a good option to consider, as they allow to load as much as 50 mL of a sample per line). As it is very difficult to guess a priori the percentage of protein bound to the IUV in each particular case, it is recommendable to initially use sensitive stains, such as SyproRuby Red Gel Stain (Molecular Probes, Invitrogen Corp.) or similar in the initial binding experiments. Once the binding has been characterized, less-sensitive stains, such as SimplyBlue SafeStain (Invitrogen Corp.) or equivalent may be used.

64

Anna V. Shnyrova and Joshua Zimmerberg

Equal volumes of the selected fractions are subjected to the reducing treatment and then loaded into the gel together with a line marker. The gel should also include a calibration line with a known amount of protein loaded. Stained gel images are acquired and analyzed using a gel imaging system (FluorChem FC (Alpha Innotech Corp., CA) or equivalent). M protein exhibits stronger binding to chol containing liposomes, with certain preference for the PE–chol neutral composition, likely because of less dense lipid packing in such mixtures related to high spontaneous curvature of PE (Fuller and Rand, 2001). M protein shows lower affinity for the compositions containing charged lipids. No protein signal is detected in the equivalent protein control fraction (Shnyrova et al., 2007). The areas under the intensity peaks of the protein bands are further analyzed using an appropriate software (Origin 7.0 (OriginLab Corp.) or similar). This analysis allows establishing the percentage of the protein bound to each IUV type. First, the amounts of protein in each line of the gel are found by normalizing to the area of the protein intensity peak in the calibration line: gL ¼

AL gCL ; ACL

ð4:3Þ

where gL are the grams of protein loaded in the sample line, AL is the area under the intensity peak of the protein band, ACL is the area under the intensity peak of the protein band in the calibration line, and gCL are the grams of protein loaded into the calibration line. The amount of protein calculated this way is per volume of sample loaded into the gel. Thus, the amount of protein for the fraction with the maximum Rh signal will be: gf max ¼

Vf gL ; VL

ð4:4Þ

where gfmax stands for the grams of protein in the studied fraction, Vf is the total volume of the fraction (100 mL), VL is the volume of the fraction loaded into the gel line, and gL are the grams of protein loaded in the gel line [see Eq. (3)]. However, this procedure underestimates the amount of protein bound to IUVs, as the IUV-protein band usually extends for more than one fraction collected from the centrifuge tube (plot in Fig. 4.1). To correct for the smearing of IUV over the gradient, the plot of Rh fluorescence versus the fraction collected is used. It is assumed that the distribution of M protein follows the distribution of IUV in the fractions, that is, the fractions which signals contribute to the Rh-fluorescence peak on the plot. Each point of the fluorescence represents a homogeneous fraction of IUV with bound M proteins. The maximum point of the peak corresponds to the fraction analyzed by SDS–PAGE and which protein amount is a known

65

Modeling Viral Budding with Unilamellar Vesicles

parameter (gfmax). To calculate the amount of protein in the entire band (i.e., for all the fractions that form the fluorescence peak), first, the plot showing the distribution of Rh fluorescence is baseline-corrected leading to elimination of the fractions with no detectable peaks of Rh fluorescence (plot in Fig. 4.1). Next, the fluorescence values for each fraction of the fluorescence peak are normalized by dividing for the fluorescence value of the peak maximum. By doing so, the amount of protein bound to the single population of IUVs is calculated as follows: gB ¼ gf max

n X Ff n ; F n¼1 f max

ð4:5Þ

where gB denotes the grams of protein bound to the IUVs, gfmax is calculated using Eq. (4), Ffn is the fluorescence value minus the baseline for the fraction n of the fluorescence peak and Ffmax is the fluorescence value minus the baseline for the fraction with the maximum fluorescence intensity of the peak. If peaks of the Rh fluorescence corresponding to different populations of IUV (or gradient bands) overlap, simple deconvolution analysis is needed to estimate the amount of the protein in each IUV population. Finally, the percent of the total protein bound to a single population of IUVs is estimated by: % bound protein ¼

gB  100; gi

ð4:6Þ

where gB is calculated from Eq. (4) and gi denotes the grams of protein initially in the binding assay (111 mg in the case considered). The fractions can be further analyzed for IUV morphological deformations caused by adsorbed proteins via EM imaging. Cryo-EM techniques are favorable as IUV deformations caused by the adhesion and flattering to the supporting film of the grids used for negative-staining EM are avoided altogether.

3.4. M protein interaction with IUVs: Leakage of aqueous dyes The IUV stability upon M protein adsorption is explored using IUVs loaded with aqueous fluorescent markers, either small (ANTS/DPX) or large (70 kDa FITC-dextran). Leakage of these aqueous dyes from IUVs upon the M protein additions is detected at ambient temperature and under constant stirring by spectrofluorimetric measurements using a spectrofluorometer (Aminco-Bowman SLM-2 (Spectronic Instruments Inc., NY) or equivalent). In a cell suitable for fluorescence measurements and containing a stirrer bar IUV sample is diluted with buffer C. Protein is added in an amount such as the protein/lipid molar ratio increases 0.01 units after each

66

Anna V. Shnyrova and Joshua Zimmerberg

protein addition. The increase in integral fluorescence due to the release of the dyes is detected. The normalized fluorescence intensity (FN) is recalculated from the integral IUV fluorescence intensity (F) as: FN ¼

F
Fi ; FT
F i

ð4:7Þ

where Fi corresponds to the fluorescence intensity before the protein addition and FT to the fluorescence intensity after complete disruption of IUV (infinite dilution of the fluorophores) by detergent (0.1% of Triton X-100). The excitation/emission wavelengths used are 380/520 nm for ANTS/DPX detection and 550/590 nm for Rh-DOPE detection. The experiments are carried out in triplicate. Data analysis is performed with Origin 7.0 software or similar. Starting at 0.01 protein/lipid ratio, release of both markers is observed with slower kinetics for the larger marker (Shnyrova et al., 2007). The possibility that the observed effects are caused by some artifact present in the protein solution (such as traces of detergent triton that is used during the protein purification) is ruled out by a-chymotrypsin proteolytic (a-chymotrypsin from bovine pancreas; Sigma-Aldrich Corp.). 5 mM protein solution is incubated 5-min at room temperature with a-chymotrypsin in 1:5 molar ratio. The digestion of M protein upon this treatment is confirmed by an SDS–PAGE (Shnyrova et al., 2007). After this treatment, the sample containing pre-digested M protein is added to the IUVs in the same conditions as the M protein without digestion. The efficiency of content release from IUV after protein digestion is greatly impaired (Shnyrova et al., 2007). Release efficiencies are comparable for both high- and low-molecularweight markers for the same amount of the protein added, thus most probably indicating vesicle bursting, rather than membrane pore formation. Also, inward invaginations of the IUV membrane expected upon the M protein adsorption should elevate membrane tension (Shnyrova et al., 2007), thus increasing the possibilities for membrane bursting, versus small pore formation. Thus, ‘‘massive’’ leakage with no size selectivity, indicating bursting, might be used as a rough preliminary criterion for the curvature activity of a protein.

4. Reconstitution of M-Protein-Driven Membrane Budding on GUVs 4.1. GUV preparation GUVs are prepared using a slight modification of the electroformation protocol (Angelova and Dimitrov, 1988). The design of a custom made electroformation chamber used is depicted in Fig. 4.2A. The chamber

B

10

m

m

Function generator ~ V

Pipette holder

m

m

Platinum electrodes

Pressure valve

y

35 mm plastic dish

ΔP 10

m

G

U

5 mm

V f zo orm ne at

m

Lipid application device

io

n

20

xyz micromanipulator

od Eb

Flexible tubing

Pt electrodes

PTF

PT F bi E ng

tu

Connection to function generator

A

Electroformation chamber

Microscope stage Objective

Figure 4.2 Electroformation of GUV. (A) Schematic representation of the lipid film application device and the electroformation chamber with Pt electrodes. (B) Schematic representation of the setup for observation of the GUV formation and for further manipulations of the electrode-bound GUVs.

68

Anna V. Shnyrova and Joshua Zimmerberg

consists of a polytetrafluoroethylene (PTFE) body and two 0.8 mm platinum (Pt) electrodes (Goodfellow Corp., PA) inserted into the PTFE block in such a way that on one side both electrodes are parallel to each other (this is the place for the GUVs to form), while on the other side the electrodes go in opposite directions. This is the place where the connector cables to the voltage generator are attached. Lipid mixtures used for GUV preparation are the following (in molar ratios): DOPC:DOPE:Chol:Rh-DOPE (48:43:4:5); POPC:Chol:RhDOPE (90:5:5); POPC:Rh-PE (95:5); POPC:DOPG:Rh-PE (80:15:5), similar to those used in IUV experiments. In some cases, the difference in composition between IUV and GUV is dictated by the electroformation protocol where each particular lipid mixture, upon drying, has to form a film covering platinum electrodes. When lipid mixture includes charged species, like DOPG, the mixture is dissolved in chloroform:methanol:ethyl ether (4:1:5 volume ratio), to a final concentration of 0.1 g/L of total lipid. In all other cases, the mixtures are dissolved in chloroform:methanol (9:1 volume ratio), to a final concentration of 0.1 g/L. A custom made application device is used for application of the lipid mixture to the Pt electrodes (Fig. 4.2A). This device consists of a 0.56 mm ID  1.07 mm OD PTFE microbore tubing (Cole Palmer, IL) of 2 cm length inserted into flexible tubing of 1.02 mm ID (such as Masterflex Tygon Tubing, Saint-Gobain Corp., PA) with 1 cm length that is hermetically closed on the opposite side to the PTFE tubing by a small screw at the end. This device offers the advantage of being disposable after use (alternatively, a 10 mL Hamilton syringe (Hamilton Company, NV) can be used for lipid application, but care should be taken to clean the syringe thoroughly after each use with appropriate lipid solvent combination, such as ethanol, methanol, chloroform. Sonication is also a good option for syringe cleaning). Before using the lipid solution, the device is washed with pure chloroform by repetitive cycles of squeeze/release of the flexible tubing part, while a small part of the PTFE tubing is immersed into the chloroform solvent. After washing, the PTFE end of the device is immersed into the lipid solution. By squeezing-releasing of the flexible part of the device a small amount of the solution is withdrawn. Two small drops of it are deposited on each one of the two platinum electrodes of the electroformation chamber by gently squeezing the flexible part of the application device. After the application, the electrodes are dried under vacuum for at least 1 h or overnight. Then, the electroformation chamber is assembled on a plastic disposable dish with a 0.17 mm thick cover glass bottom (Delta TPG Dish (Bioptech, PA) or similar; Fig. 4.2B) and the dish is mounted on a stage of an inverted microscope (Axiovert 200 (Carl Zeiss Inc., Germany) with a fluorescence excitation source, a filter cube combination suitable for Rh observation and a 40/0.65 and a 100/1.25 objectives (ACHROPLAN; Carl Zeiss Inc.) or equivalent microscope configuration). The electrodes are

Modeling Viral Budding with Unilamellar Vesicles

69

connected to a function generator (GFG-3015 model function generator (Instek America Corp., CA) or equivalent) preset to apply a 10 Hz and 0.2 V sinewave. Right after the initial voltage application, a buffer containing 20 mM HEPES and approximately 200 mM sucrose, pH 7.4 is gently added to the dish in such a way that it is slightly above the electrodes. This buffer should be scrupulously equilibrated osmotically with buffer C by using the procedure described above. During the following 15 min, the sinewave amplitude is gradually increased from 0.2 to 1.0 V. Electroformation times varied from 15 min to 1 h for different lipid compositions of the GUV. After GUVs become visible on the electrode by bright-field or fluorescence microscopy observation with a 40 objective, the voltage is lowered to 0.2 V and then turned off completely. If the GUVs are to be detached from the electrodes, the frequency is lowered to 0.1 Hz before lowering of the voltage. Then the GUVs are collected with 200 mL pipette and stored in a microcentrifuge tube until usage. If the GUVs are to be left on the electrodes, the plastic dish is perfused with buffer C for 15 min by using an appropriate microscope perfusion system (such as the micro-perfusion system (Bioscience Tools, CA) or equivalent). Visualization of GUVs in the attached configuration is performed directly on the inverted microscope stage immediately after perfusion of the chamber with buffer C.

4.2. Interaction of M protein with GUVs Visualization of GUVs in the attached to the electrode configuration (Fig. 4.3A) is performed directly on the inverted microscope stage immediately after perfusion of the chamber with buffer C. The biggest advantage of this configuration is the relative immobilization of the GUVs on the electrodes, which facilitates greatly the application of the protein in the proximity of the GUV, as well as allows to easily patching the GUV membrane. However, the fluorescence background from the electrode and other GUVs in the vicinity sometimes interferes with the imaging of a single GUV. Also, the proximity of the electrodes to the bottom of the dish becomes a critical factor when high-magnification objectives (such as 100 or bigger) are used. These objectives have small working distances, thus there is a requirement for the electrodes to be close to the bottom of the dish, something not easily achievable with 0.8 mm thick electrodes. To observe individual vesicles at high magnifications, the detached GUVs are routinely used. In this case, a 0.17-mm-thin glass 35 mm dish (Delta TPG Dish (Bioptech) or similar) is pretreated with 1 g/L BSA for 1 min and thoroughly washed with buffer C in order to avoid GUV spreading and bursting upon contacting the glass bottom of the dish. The dish is setup on the microscope stage and 1 mL of buffer C is added. GUV suspension of 100–200 mL collected after electroformation is added to the

70

Anna V. Shnyrova and Joshua Zimmerberg

A M protein Pipette

GUV Platinum electrode

B M protein Pipette

GUV BSA pretreated cover slip

Figure 4.3 Main configurations used in GUV experiments. (A) GUV remains bound to the electrode and the M protein is applied by patching the GUV membrane with a patch-pipette filled with the protein solution of choice. (B) A GUV detached from the electrode is situated on the glass bottom of the dish (pretreated with BSA) and the M protein is applied to the GUV through the patch-pipette brought in proximity to the GUV. Pulses of positive hydrostatic pressure cause protein efflux.

1 mL of buffer C in the dish. As GUVs are formed in the sucrose buffer, osmotically equilibrated with buffer C, after 5–10 min all the GUVs are settled in the glass coverslip of the bottom of the dish due to a difference in density between the two buffers (Fig. 4.3B). In this case, GUVs are easily visualized by high-resolution fluorescence microscopy, however, the fact that GUVs are only slightly attached to the bottom, makes the experiments with protein delivery more involved. For both configurations, prior to the protein addition, the intrinsic behavior of a target vesicle is recorded for a period of time (1–3 min). The recording is made using a digital camera (iXon (Andor, CT) or similar) attached to the camera port of the microscope. The camera is controlled through a PC with MicroManager open source microscope software (University of California, San Francisco, CA) or equivalent. The unilamellarity of the GUV may be established by fluorescence analysis of the GUV contour, as was described previously (Akashi et al., 1996). Upon establishing the stability (and unilamellarity) of the target GUV, the M protein is applied through a pipette brought to the close proximity of the GUV (Fig. 4.3B). The pipettes for M protein application are obtained by pulling a 1.5 mm capillary borosilicate glass (#PG10150-4, WPI Inc.) with a micropipette puller (Flaming Brown Micropipette puller, Model P80/PC; Sutter Instruments Co., CA). The pipette is back-filled with

Modeling Viral Budding with Unilamellar Vesicles

71

M protein solution using a 1 mL syringe equipped with MicroFilTM (WPI Inc.). Three motorized actuators (12.5 mm Travel Open-loop DC CMA Actuator; Newport Corp., CA) attached to a crossed-roller bearing linear stage, 0.5 in., with XYZ travel (461-XYZ-M, Newport Corp.) are used to control the pipette position. The pipette filled with M protein solution is inserted into a microelectrode holder with pressure port (WPI Inc.) attached to the linear stage (Fig. 4.2B). A 1-mL syringe is attached to one port in a three-way luer valve (WPI Inc.). Another port of the valve is connected through long PTFE tubing to the pressure port of the microelectrode holder. This allows controlling the pressure difference applied to the pipette (i.e., for protein application or liposome patch isolation). The luer valve should be closed (i.e., turned toward the unoccupied port) during the pipette search in the objective focus area. The delivery pipette containing a concentrated solution of the protein (4 mM) is brought carefully to the proximity of the studied GUV, the pipette motion being observed by bright-field microscopy. Once the pipette is located in the proximity of the vesicle, it is directed toward the GUV and the luer lock in the pressure loop is opened. A pulse of positive hydrostatic pressure is applied to the pipette by gently pressing the pressure syringe. Immediately, a small flow of protein can be seen exiting the pipette. Prior to the protein application, the microscope light pass is switched to the fluorescence mode and the morphological changes of the GUV membrane are monitored and documented by acquiring time lapse series. The protein concentration in the vicinity of GUV increases transiently. Shortly after protein application, invagination of the GUV membrane is detected. After each protein application the apparent size of the GUV decreases and the steady-state fluorescence in the GUV projection area increased (Shnyrova et al., 2007). Internalized membrane vesicles are clearly detected, thus justifying the fluorescence changes. Special care should be taken about the pressure applied to the delivery pipette when GUVs slightly attached to the dish bottom are used. If the pressure is too strong, the GUV can be detached from the bottom of the dish and flow away. In this case, the pressure valve of the pressure syringe should be closed, the pipette should be lift immediately and a new GUV, far enough from the initial observation point, should be found. Several experiments can be performed in the same dish. However, the pipette should be exchanged after each one of them to avoid protein dilution effects. GUVs of different compositions are used in order to establish the preferences for M-protein-induced budding. Surprisingly, and against the prediction made upon the binding preferences of M protein detected with IUVs, vesicle formation is more effective when cholesterol or charged lipids are present in the GUV composition. Thus, the budding activity of M protein is not directly related to the efficiency of its membrane binding (Shnyrova et al., 2007). Another important observation is formation of transient, membrane

72

Anna V. Shnyrova and Joshua Zimmerberg

domains upon application of M proteins to GUV. The appearance of these domains indicates that M proteins self-organize into fluid-like formations which then bud away from the membrane (Shnyrova et al., 2007). Overall, these experiments directly demonstrate that M protein implements the genetically encoded information required to create the closed spherical membrane particles through its interactions with lipid bilayer. The validity of such observations should be routinely checked by doing buffers and negative protein controls (i.e., by applying the M protein buffer containing no proteins or a reference protein, such as albumin, in similar concentration as M protein). Those are very important controls, as osmotic pressure, as well as bulk protein binding, can greatly affect the behavior of the GUVs. For example, differences in osmotic pressure between the exterior and the interior of the GUV can cause spontaneous budding and formation of vesicles in the GUV lumen without the need of a protein, as in the case described by Guedeau-Boudeville et al. (2000). Thus, iso-osmolarity of every solution used with GUVs should be particularly monitored. When experiments are performed on electrode-attached GUVs, the procedure is almost the same as with detached GUVs. However, the manipulations are eased by the fact that GUV is immobilized. Patch-clamp type experiment, hard on the slightly attached GUV, can be performed using single GUVs on the electrode (Fig. 4.3A). For these experiments, the luer pressure valve on the pressure syringe is opened before the pipette approaches a GUV closely. After touching the GUV, a very small negative pressure is applied. The GUV membrane is sucked in and a stable contact, seen as a small membrane patch inside the pipette, forms. Protein adsorption on this isolated patch inside the pipette starts instantaneously, and the morphological changes of the patch membrane, as well as of the whole liposome are observed and documented by time lapse series as described before. The membrane rearrangements quickly lead to the budding directed toward the interior of the GUV. Round vesicles of different diameters become visible near the patch and inside the GUV. The membrane outside the patched area provided a lipid reservoir to support this small vesicle formation. Thus, the GUV shrinks while more vesicles are produced. Finally, the GUV membrane detaches from the pipette and multiple vesicles are seen moving inside the GUV (Shnyrova et al., 2007). Again, buffer and negative protein controls should be performed in this configuration. The patch-clamp experiments emphasize the localized character of the M protein action. It is an important observation, as GUVs are notorious for their extreme sensitivity to the global changes in GUV membrane, such as introduction of area asymmetry (difference in area) between monolayers of GUV membrane. These experiments also indicate the power of M protein assembly: constitutive formation of membrane vesicles lead to building up the lateral membrane tension powerful enough to induce the GUV detachment from the pipette. Analyzing these observations provide valuable mechanistic insights into curvature-driven activity of M protein (Shnyrova et al. 2007).

Modeling Viral Budding with Unilamellar Vesicles

73

5. Concluding Remarks The assays described in this chapter represent a relatively easy approach for studying the activity of cellular protein involved in controlling membrane curvature. The approach is composed of two parts. The initial analysis is based upon quantification of membrane binding and leakage for membranes of different lipid compositions. Although less spectacular than the GUV-based experiments, these assays offer a significant amount of information about the mechanism of protein–lipid interaction. They can also provide ultrastructural information upon EM analysis of membrane shape transformations by bound proteins. Nevertheless, these measurements are lacking the dynamic aspects of the budding process. Microscopic observations of the GUV membrane provide a versatile tool to visualize the dynamics of membrane curvature creation. These observations reveal dynamic, transient intermediates of the budding process, such as membrane domains, which would not be resolved by structural analysis. However, these observations are limited by the optical resolution of the microscope itself. To resolve smaller structures (i.e., <200 nm), as well as to confirm fission of the vesicle neck, more sensitive techniques should be used in combination with the methods described here (such as measurements of membrane area and permeability by electrical admittance measurements on planar bilayer lipid membranes; see Shnyrova et al., 2007). Ultimately, the combination of ultrastructural, electrophysiological, and microscopy techniques form an experimental environment critical for obtaining real mechanistic insights into complex processes of membrane morphogenesis.

REFERENCES Akashi, K., Miyata, H., Itoh, H., and Kinosita, K. Jr. (1996). Preparation of giant liposomes in physiological conditions and their characterization under an optical microscope. Biophys. J. 71, 3242–3250. Angelova, M. I., and Dimitrov, D. S. (1988). A mechanism of liposome electroformation. Prog. Colloid. Polym. Sci. 76, 59–67. Bartlett, G. R. (1959). Phosphorus assay in column chromatography. J. Biol. Chem. 234, 466–468. Briggs, J. A., Wilk, T., and Fuller, S. D. (2003). Do lipid rafts mediate virus assembly and pseudotyping? J. Gen. Virol. 84, 757–768. Bugelski, P. J., Sowinski, J. M., and Kirsh, R. L. (1990). Negative stain electron microscopy. In ‘‘Liposomes: A Practical Approach,’’ (R. R. C. New, ed.), pp. 140–161. Oxford University Press, New York. Byrdwell, W. C. (2005). Methods for Lipid Analysis by Liquid Chromatography/Mass Spectrometry and Related Techniques. AOCS Press. Caffrey, M., Morris, S. J., and Feigenson, G. W. (1987). Uranyl acetate induces gel phase formation in model lipid and biological membranes. Biophys. J. 52, 501–505.

74

Anna V. Shnyrova and Joshua Zimmerberg

Coorssen, J. R., Blank, P. S., Albertorio, F., Bezrukov, L., Kolosova, I., Backlund, P. S. Jr., and Zimmerberg, J. (2002). Quantitative femto- to attomole immunodetection of regulated secretory vesicle proteins critical to exocytosis. Anal. Biochem. 307, 54–62. Du¨zgu¨nes¸, N. (2003a). Fluorescence assays for liposome fusion. Methods Enzymol. 372, 260–274. Du¨zgu¨nes¸, N. (2003b). Preparation and quantitation of small unilamellar liposomes and large unilamellar reverse-phase evaporation liposomes. Methods Enzymol. 367, 23–27. Fraley, R., Subramani, S., Berg, P., and Papahadjopoulos, D. (1980). Introduction of liposome-encapsulated SV40 DNA into cells. J. Biol. Chem. 255, 10431–10435. Fuller, N., and Rand, R. P. (2001). The influence of lysolipids on the spontaneous curvature and bending elasticity of phospholipid membranes. Biophys. J. 81, 243–254. Garcia-Sastre, A., Cabezas, J. A., and Villar, E. (1989). Proteins of Newcastle disease virus envelope: Interaction between the outer hemagglutinin-neuraminidase glycoprotein and the inner non-glycosylated matrix protein. Biochim. Biophys. Acta 999, 171–175. Guedeau-Boudeville, M. B., Bradley, J. C., Singh, A., and Jullien, L. (2000). Changes in the morphology of giant vesicles under various physico-chemical stresses. In ‘‘Giant Vesicles,’’ (P.a.W. Luigi Luigi, ed.), pp. 341–351. John Wiley & Sons, LTD, West Sussex, England. Jayakar, H. R., Jeetendra, E., and Whitt, M. A. (2004). Rhabdovirus assembly and budding. Virus Res. 106, 117–132. Mayer, L. D., Hope, M. J., and Cullis, P. R. (1986). Vesicles of variable sizes produced by a rapid extrusion procedure. Biochim. Biophys. Acta 858, 161–168. McCreedy, B. J. Jr., McKinnon, K. P., and Lyles, D. S. (1990). Solubility of vesicular stomatitis virus M protein in the cytosol of infected cells or isolated from virions. J. Virol. 64, 902–906. New, R. R.C (1990a). Characterization of liposomes. In ‘‘Liposomes: A Practical Approach,’’ (R. R. C. New, ed.), pp. 138–139. Oxford University Press, New York. New, R. R. C. (1990b). Liposomes: A Practical Approach. Oxford University Press, New York. Palade, G. E. (1953). An electron microscope study of the mitochondrial structure. J. Histochem. Cytochem. 1, 188–211. Palade, G. (1975). Intracellular aspects of the process of protein synthesis. Science 189, 867. Pantua, H. D., McGinnes, L. W., Peeples, M. E., and Morrison, T. G. (2006). Requirements for the assembly and release of Newcastle disease virus-like particles. J. Virol. 80, 11062–11073. Parsegian, V. A., Rand, R. P., and Stamatoff, J. (1981). Perturbation of membrane structure by uranyl acetate labeling. Biophys. J. 33, 475–477. Peter, B. J., Kent, H. M., Mills, I. G., Vallis, Y., Butler, P. J., Evans, P. R., and McMahon, H. T. (2004). BAR domains as sensors of membrane curvature: The amphiphysin BAR structure. Science 303, 495–499. Reeves, J. P., and Dowben, R. M. (1969). Formation and properties of thin-walled phospholipid vesicles. J. Cell Physiol. 73, 49–60. Sagrera, A., Cobaleda, C., Berger, S., Marcos, M. J., Shnyrov, V., and Villar, E. (1998). Study of the influence of salt concentration on Newcastle disease virus matrix protein aggregation. Biochem. Mol. Biol. Int. 46, 429–435. Scheid, A., and Choppin, P. W. (1973). Isolation and purification of the envelope proteins of Newcastle disease virus. J. Virol. 11, 263–271. Shnyrova, A. V., Ayllon, J., Mikhalyov, I. I., Villar, E., Zimmerberg, J., and Frolov, V. A. (2007). Vesicle formation by self-assembly of membrane-bound matrix proteins into a fluidlike budding domain. J. Cell Biol. 179, 627–633. Solon, J., Gareil, O., Bassereau, P., and Gaudin, Y. (2005). Membrane deformations induced by the matrix protein of vesicular stomatitis virus in a minimal system. J. Gen. Virol. 86, 3357–3363.

Modeling Viral Budding with Unilamellar Vesicles

75

Takimoto, T., and Portner, A. (2004). Molecular mechanism of paramyxovirus budding. Virus Res. 106, 133–145. Vaccaro, A. M., Tatti, M., Ciaffoni, F., Salvioli, R., Barca, A., and Roncaioli, P. (1993). Studies on glucosylceramidase binding to phosphatidylserine liposomes: The role of bilayer curvature. Biochim. Biophys. Acta 1149, 55–62. Welsch, S., Muller, B., and Krausslich, H. G. (2007). More than one door—Budding of enveloped viruses through cellular membranes. FEBS Lett. 581, 2089–2097.

This page intentionally left blank

S E C T I O N

T W O

LIPOSOMES AND NANOTECHNOLOGY

This page intentionally left blank

C H A P T E R

F I V E

Detection of Antimycolic Acid Antibodies by Liposomal Biosensors Y. Lemmer,*,‡ S. T. Thanyani,* P. J. Vrey,* C. H. S. Driver,† L. Venter,* S. van Wyngaardt,* A. M. C. ten Bokum,* K. I. Ozoemena,† L. A. Pilcher,† D. G. Fernig,} A. C. Stoltz,*,§ H. S. Swai,‡ and J. A. Verschoor* Contents 1. Introduction 2. Experimental 2.1. Purification of mycobacterial MA 2.2. Fluorescent labeling of MA 2.3. Determination of MA carrying capacity of liposomes 2.4. MA liposome size is influenced by cholesterol and MA content, but not by pH 2.5. MA liposome immobilization on IAsys biosensor cuvettes 2.6. Technology transfer from the IAsys waveguide to the ESPRIT SPR biosensor 2.7. The cholesteroid nature of MA demonstrated on the ESPRIT biosensor 3. Conclusion Acknowledgments References

80 81 81 82 82 84 86 89 100 102 102 102

Abstract Antibodies to mycolic acid (MA) antigens can be detected as surrogate markers of active tuberculosis (TB) with evanescent field biosensors where the lipid antigens are encapsulated in liposomes. Standard immunoassay such as ELISA, where the lipid antigen is not encapsulated, but directly adsorbed to the wellbottoms of microtiter plates, does not yield the required sensitivity and

* { { } }

Department of Biochemistry, University of Pretoria, Pretoria, South Africa Department of Chemistry, University of Pretoria, Pretoria, South Africa Materials Science and Manufacturing, CSIR, Pretoria, South Africa Department of Infectious Diseases, University of Pretoria, Pretoria, South Africa School of Biological Sciences, University of Liverpool, United Kingdom

Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64005-2

#

2009 Elsevier Inc. All rights reserved.

79

80

Y. Lemmer et al.

specificity for accurate diagnosis of TB. One reason for this is the cross-reactivity of natural anticholesterol antibodies with MAs. MAs are the major cell wall lipids of mycobacteria. Mycobacterial MA has immunomodulatory properties and elicits specific antibodies in TB patients. Liposomes were optimized for their use as carriers both for the presentation of immobilized purified mycobacterial MA on sensor surfaces, and as a soluble inhibitor of antibody binding in inhibition assays. By using an inhibition assay in the biosensor, the interference by anticholesterol antibodies is reduced. Here, we describe the MA carrying capacity of liposomes with and without cholesterol as a stabilizing agent, optimized concentration and size of liposomes for use in the biosensor assay, comparison of the methods for wave-guide and surface plasmon resonance biosensors and how the cholesteroid nature of MA can be demonstrated by the biosensor when Amphotericin B is allowed to bind to MA in liposomes.

1. Introduction As a result of the development of drug-resistant strains of Mycobacterium tuberculosis and the breakdown of the immune system of its host by HIV, tuberculosis (TB) is no longer a ‘‘controlled’’ disease and has become a major health problem in both developed and developing countries (Houghton et al., 2002). Diagnosis of TB is no longer 100% reliable due to AIDS, and was never adequate for determining extrapulmonary and child TB. HIV coinfection and drug resistance appears to shorten the lifetime of a TB patient considerably, such that it becomes a matter of life and death to be able to diagnose TB within 24 h of sampling. The decision toward treatment cannot be taken lightly, as the treatment regime has to be maintained for at least 6 months to clear all the latent TB from the body. The current drive toward new tools for TB diagnosis arose from these facts, and that the mycobacterial pathogen isolated from sputum samples is slow growing, thereby requiring several weeks to become visible during in vitro growth (Reischl, 1996; Samanich et al., 2000). Although DNA-amplifying technology has reduced the period from sampling to TB diagnosis to within days, it still uses mainly samples obtained from the lungs. Therefore, fast, affordable, and reliable diagnosis of TB has become a high priority in public health (Chan et al., 2000), and is currently actively pursued in several laboratories. Mycolic acids (MAs) appeared to be promising antigens for the design of fast immunodiagnostics for TB (Verschoor and Onyebujoh, 1999). The major cell wall lipids of mycobacteria are the MAs, which are long chain (C60–C90) a-branched, b-hydroxy fatty acids. Mycobacterial MAs are immunogenic. MA was the first nonprotein antigen shown to stimulate the proliferation of human CD4
, CD8
, and CD4þ T-lymphocytes upon CD1 presentation (Beckman et al., 1994; Goodrum et al., 2001). In addition, anti-MA antibodies could be detected in the serum of patients with active

Antibody Detection by Liposomal Biosensors

81

pulmonary TB (Pan et al., 1999) using the standard ELISA procedure, indicating that free MAs are present in the circulation during active TB disease (Beatty et al., 2000). The prevalence of anti-MA antibodies appeared to be independent of the degree to which a person was suffering from AIDS (Schleicher et al., 2002). Whereas ELISA did not detect anti-MA antibodies well enough to be considered as a basis for a diagnostic test of active TB, biosensor analysis did (Thanyani et al., 2008). It could achieve the required specificity and sensitivity to be regarded seriously as a solution to the current dilemma of standard TB diagnosis taking several weeks after sampling to produce a result. Even the highly sensitive PCR detection of mycobacterial DNA in patient sputum samples still takes a few days to deliver a diagnostic outcome. MAs are soluble in extremely nonpolar organic solvents such as chloroform, dichloromethane, and hexane. Alternatively, they can be ‘‘solubilized’’ in boiling water or aqueous buffers. For the detection of anti-MA antibodies by means of ELISA, MA was either coated from hexane solutions (Pan et al., 1999) or from boiling phosphate-buffered saline (PBS) (Schleicher et al., 2002). This presentation of MA is clearly not physiological. Presentation of MA in liposomal environments more closely resembles the way in which MA is encountered in the body of the patient. Indeed, when encapsulated into liposomes and injected into mice, MA was shown to behave as a typical pathogen-associated molecular pattern (Korf et al., 2005). It could also act as an immunomodulatory compound that suppresses experimental asthma through Treg cell intervention (Korf et al., 2006). It was only logical to derive that diagnosis of TB by detection of surrogate marker antibodies against MA should perform better when MA antigen was presented in liposomes. This was demonstrated by Thanyani et al. (2008) by making use of biosensors. Here, we describe how liposomes are optimized for their use as carriers of MA for immobilization as antigens on sensor surfaces, and as soluble inhibitors of antibody binding in inhibition assays in a TB diagnostic assay. It is dubbed the MARTI-test, as an acronym of Mycolic acid Antibody Real-Time Inhibition. Whereas the original MARTI-test is described for a wave-guide biosensor (Thanyani et al., 2008), we show here how the method is adjusted to also suit the more popularly used surface plasmon resonance (SPR) biosensors. Finally, we demonstrate the cholesteroid nature of MA with the biosensor by measuring the binding of Amphotericin B to MA in liposomes.

2. Experimental 2.1. Purification of mycobacterial MA MAs are isolated and purified from M. tuberculosis H37Rv ATTC 27294 purchased from the American Type Culture Collection (ATTC, Baltimore, MD, USA) as described by Goodrum et al. (2001). The MA is dissolved in

82

Y. Lemmer et al.

chloroform (HPLC grade; Merck; Darmstadt, Germany) and aliquotted into glass vials. The chloroform is evaporated under nitrogen gas and the dried MA is stored at 4  C until use. The purified MAs are checked for LPS contamination using the kinetic-QCL Limulus amoebocyte test kit (Sigma, St. Louis, MO). In the experiments shown in this chapter, no LPS is detected.

2.2. Fluorescent labeling of MA MA is fluorescently labeled by derivatization with 5-bromomethyl fluorescein (5-BMF; Molecular Probes, Leiden, The Netherlands), as described by Korf et al. (2005). Quality control is performed by TLC on a silica gel GHL thin layer plate. Chromatography is performed in two dimensions, with chloroform:methanol:water as the mobile phase in the first dimension, and 100% methanol (Merck) as the mobile phase in the second dimension. Iodine vapor is used to visualize the MA. The absence of free 5-BMF, not associated with the MA spot, indicates that the label is covalently bound. Fluorescently labeled 5-BMF–MA is incorporated into liposomes for assessment by the biosensor or flow cytometry.

2.3. Determination of MA carrying capacity of liposomes MAs are dissolved in chloroform and 100 mg quantities are aliquoted into amber vials and stored at 4  C. Liposomes are prepared according to the method described by Bangham (1983). This involves the deposition of a thin lipid film from an organic solvent medium on the wall of a container, followed by agitation with an aqueous medium. In short, phosphatidylcholine (PC from egg yolk, 99% pure; Sigma), cholesterol (Sigma), and dried MA are all dissolved separately in chloroform. PC (9 mg) with or without 4.5 mg cholesterol and varying amounts of MA are mixed together in a glass vial. The chloroform is evaporated under nitrogen gas in a chemical hood. PBS (1 ml) is added and the mixture is heated to 80  C until the lipid film dissociates from the wall of the vial. The hot mixture is then vortexed and sonicated for 5 min using a Branson sonifier for 50 pulses at a 20% output level. MA-carrier liposomes are extracted using a triphase partition method, modified from Dennison and Lovrein (1997). Briefly, a sample (50–200 ml) of MA-carrier liposome solution is diluted to 600 ml with H2O and extracted three times with equal volumes of tert-butanol and chloroform. The chloroform phases are collected and the chloroform is evaporated under nitrogen gas. MA remaining in the vials due to saturation of the liposomes is collected by chloroform rinses. HPLC of the extracted MA is carried out as described (Butler and Kilburn, 1990), after derivatization with p-bromophenacyl bromide. A high molecular weight internal standard

83

Antibody Detection by Liposomal Biosensors

(Ribi ImmunoChem Research Company, Hamilton, MT, USA) is added to each sample to allow quantification of the extracted MA. Liposomes are made with egg PC, with or without the addition of cholesterol. The maximum amount of MA that can be incorporated reproducibly is about 2 mg/ml for liposomes consisting of phosphatidyl choline and cholesterol, but only about 500 mg MA/ml into liposomes consisting of phosphatidyl choline only (Fig. 5.1). MA not incorporated into the liposomes can in most cases be quantitatively recovered from the wall of the vessel in which the liposomes are prepared (shaded part of the bars). At very high concentrations the recovery is not quantitative, as the MA tends to form clumps which float in the liposome suspension. The MA liposomes

A

120

% MA recovered

100 80 60 40 20 0 125 B

250

500 1000 1500 2000 MA concentration (mg/ml)

3000

4000

120

% MA recovered

100 80 60 40 20 0 125

250

500 1000 1500 2000 3000 4000 13100 MA concentration (mg/ml)

Figure 5.1 Titration of the maximum amount of MA that can be incorporated in liposomes consisting of egg phosphatidylcholine, with or without cholesterol: (A) liposomes consisting of egg phosphatidylcholine (99% pure) only; (B) liposomes consisting of egg phosphatidylcholine (99% pure) and cholesterol. The shaded part of the bars represent the percentage of MA recovered from the wall of the vessel in which the liposomes were prepared. Each bar represents the mean of three samples and the S.D.

84

Y. Lemmer et al.

can be lyophilized and reconstituted without significant loss of MA content. Heating to 80  C in the reconstitution process is an important step to ensure full recovery of MA content.

2.4. MA liposome size is influenced by cholesterol and MA content, but not by pH MA-containing liposomes are prepared with equal amounts of 5-BMF labeled and unlabeled MAs and varying amounts of cholesterol according to Table 5.1. A liposome stock suspension is obtained by adding PBS (1 ml) to the dried lipid components, dissolving it on a heat block for 60 min after vortexing and sonicating the heated (80  C) mixture with a Branson Sonifier B30 Cell Disrupter at 20% duty cycle and output control of 2 for 5 min. Both the size distribution (forward scatter, FS) and the fluorescence (FL-1) of the liposomes are measured on a flow cytometer (a Beckman Coulter Epics Ultra instrument is used to obtain the results in Fig. 5.2). For the analysis, 10 times dilutions of the stock liposomes are prepared using PBS in plastic analyzing tubes covered with foil to prevent the photobleaching of the fluorescent marker. The tubes are kept in a water bath at 37  C during the experiments. One gate is set on the analyzer to exclude background signals and debris events that have FS values below 10. A constant number of 50,000 events are counted per measurement. Flowset (3.6 mm) fluorospheres measured with each experiment is used to correlate the relative size of the liposomes to that of the beads when varying amounts of cholesterol are added to a constant amount of MA and PC. The size distribution of the liposomes changes in accordance with a change in the cholesterol concentration (Fig. 5.2A). Two different liposome states can be identified. At the two highest concentrations of cholesterol, the liposomes are 2 log units bigger than the liposomes that contain none or the lowest concentrations of cholesterol. The liposomes obtained with the addition of 11.25 ml of cholesterol stock give an intermittent size of liposomes in between that of the two states. With an increase in the cholesterol concentration and liposome size, the liposomes also show an analogous increase of 2 log units in the amount of fluorescence per liposome that is Table 5.1 Liposome compositions with various concentrations of cholesterol

Cholesterol (ml from 100 mg/ml stock) PC (ml from 100 mg/ml stock) 5-BMF–MA (mg) Unlabeled MA (mg)

0

5.63

11.25

22.5

45

90 125 125

90 125 125

90 125 125

90 125 125

90 125 125

85

Antibody Detection by Liposomal Biosensors

A

Number of liposomes

Flowset 3.6 mm 0ml 5.63 m l 11.25 m l 22.5 m l 45 m l

100 101 102 103 104 105 106 107 Size (log)

Number of fluorescent liposomes

B

0ml 5.63 m l 11.25 m l 22.5 m l 45 m l

100

101

102

103 104 105 Log fluorescence

106

107

Figure 5.2 Size (A) and fluorescence (B) of MA-containing liposomes containing various amounts of cholesterol according to Table 5.1 as determined with flow fluorometry. Flowset beads (3.6 mm) were used as a marker for the size distribution of the liposomes.

emitted (Fig. 5.2B), indicating that the change from smaller to larger liposomes is not due to swelling with water/PBS, but to an accumulation of more MA-containing material into the bigger liposomes. The results imply that cholesterol content determines two states of liposomes, where higher concentrations of cholesterol induce the disordered state of the bilayer in liposomes that increase their size 100-fold. All cholesterol-containing liposomes used in the biosensor analysis were in the disordered state at the highest concentration of cholesterol shown here.

86

Y. Lemmer et al.

In Fig. 5.2, the biggest sized liposomes measured around 30 mm and were composed of PC (9 parts), cholesterol (4.5 parts), and MA (0.25 parts). When the phosphatidyl choline is combined with either the MA or the cholesterol, significant changes in the sizes of the liposomes occur. With cholesterol alone the size sharply increases, while with MA alone the size sharply decreases (Fig. 5.3). This effect is not significantly influenced in the pH range between 4 and 10 of the PBS to which the liposomes are exposed, probably due to the unchanged surface charge of phosphatidyl choline liposomes over these pH values. The anionic phosphatidyl residues (pKa  3.5) only become neutralized below pH 4, while the cationic state of choline (pKa ¼ 13.9) is maintained over the full pH range measured (Tatulian, 1993). It is therefore expected that the stability of the liposomes described here for use in the biosensor will be quite tolerant to a broad pH window extending to both sides of the typical pH of 7.4 of PBS.

2.5. MA liposome immobilization on IAsys biosensor cuvettes A number of application notes are available for the adsorption of liposomes to hydrophobic sensor surfaces. These do not stand up as rigorous and reliable for the immobilization of MA antigen-containing liposomes aiming at antibody detection. This motivated the use of nonderivatized sensor cuvettes where the glass-like hafnium oxide surface can be made hydrophobic by treatment with the cationic detergent, cetylpyridinium chloride (CPC). CPC is well known for its use as an antiseptic agent. Its property

900 800 700

Size

600 500 400

PC/Chol PC/MA PC/MA/Chol

300 200 100 0

Figure 5.3 Relative sizes of PC/cholesterol and PC/mycolic acid liposomes and a mixture of the two liposomes. All liposome types contained 9 parts of phosphatidyl choline and either 4.5 parts of cholesterol, or 0.25 parts of MA, or both.

87

Antibody Detection by Liposomal Biosensors

of binding to glass when applied at aqueous dilutions below the critical micellar concentration (i.e., <0.1 mg/ml) is long established (Westwell and Anacker, 1959), leaving a ‘‘greasy’’ surface (Hartley and Runnicles, 1938). PBS/AE buffer consists of 8.0 g NaCl, 0.2 g KCl, 0.2 g KH2PO4, and 1.05 g Na2HPO4 per liter of double distilled, deionized water containing 1 mM EDTA and 0.025% (m/v) sodium azide and is adjusted to pH 7.4. CPC (0.02 mg/ml) and saponin (1 mg/ml) are prepared in PBS/AE. The IAsys resonant mirror biosensor system and twin-cell nonderivatized cuvettes are from Affinity Sensors (Farfield Scientific, Crewe, UK). The sensor is set for a data sampling interval of 0.4 s, temperature of 25  C, and stirring rate of 75% for all experiments. The cells are rinsed three times prior to use with 96% ethanol (Saarchem, South Africa), followed by extensive washing with PBS/AE. A 60 ml volume of PBS/AE is pipetted into each cell of the cuvette to obtain a stable baseline for 1 min. Cuvettes are washed with PBS/AE until a stable baseline could be maintained for at least 5 min. The cells are aspirated and 50 ml of a 20 mg/ml CPC in PBS/AE solution is added. After 10 min the cells are washed five times with 60 ml PBS/AE. PBS/AE (25 ml) is added until a stable baseline is achieved. Twenty-five microliters of the desired liposome solution are added and the binding response is monitored for 10 min. The MA-containing liposome concentration consisting of 9:3:1 (m/m) phosphatidyl choline:cholesterol:MA is titrated between 0.02 and 10 mg/ml of total lipid concentration. The MA-containing liposome concentration with the optimum binding capacity is found to be 500 mg total lipid/ml (Fig. 5.4). This concentration is used in all subsequent experiments.

MA-liposome binding (arc. s)

4000

3000

2000

1000

0 0

2

4 6 8 MA-liposome concentration (mg/ml)

10

Figure 5.4 Titration of the optimal concentration of MA/cholesterol-containing liposomes for immobilization on a nonderivatized IAsys biosensor cuvette.

88

Y. Lemmer et al.

After liposome binding, the cells are washed five times with 60 ml PBS/AE and immediately after that, treated five times with 60 ml 1 mg/ml saponin in PBS/AE. Due to response differences among saponin batches, it is necessary to titrate the amount of saponin required each time a new batch of saponin is used. The cells are incubated with saponin for at least 10 min and until a stable baseline is achieved before a final five times wash with 60 ml PBS/AE. Antibody interaction analysis could be continued from this point as described by Thanyani et al. (2008). The cells are washed five times with PBS/AE, the content of each cell substituted with 25 ml of PBS/AE and left for about 5–10 min to achieve a stable baseline. Inhibition studies are performed using patient’s serum that is first placed at room temperature to thaw completely. After obtaining a stable baseline, a 1/1000 dilution of serum antibodies (10 ml) in PBS/AE is added in each cell, to compare the responses of the two cells over 10 min. A preincubation of 1/500 dilutions of serum with solutions of liposomes-containing MAs and empty liposomes (PC alone) is allowed for 20 min. These are then added (10 ml) for binding inhibition studies in the two cuvette cells, one with MAs liposomes and the other with empty liposomes as a control, and allowed to bind for 10 min. Finally, dissociation of antibodies is effected with three times PBS/AE washing and measurement of the response for 5 min. Regeneration of the cuvette is effected by three washes with 96% ethanol for 1 min, followed by seven washes with 70 ml PBS/AE for 1 min. The surface is then finally treated with 50 ml potassium hydroxide (12.5 M) for 2 min followed by seven washes with 70 ml PBS/AE for 1 min. A typical profile of a positive and negative TB test using this method, but applied in the ESPRIT biosensor, is shown in Fig. 5.11. The outcome of the test in the IAsys biosensor scored an overall specificity of 48.4% (15/31) and sensitivity of 86.7% (26/30) with 61 patient sera analyzed (Thanyani et al., 2008). When adjusted for the inadequate performance of the reference culture test in HIV-infected patients, the MARTI-test scored a specificity of 76.9%. A disadvantage of the IAsys biosensor in this application is the difficulty of aligning the relative basic signal strength of the two cells in a cuvette, which must be identical during the first serum exposure before a patient diagnosis can be derived from the difference of the binding signal during the second liposome-preincubated serum exposure. We experienced an average success rate of around 30% to obtain comparable signals during the first serum exposure. For this reason and due to the fact that the IAsys cuvettes were relatively expensive, we transferred the technology from the wave-guide IAsys biosensor to the more generally used SPR biosensors.

Antibody Detection by Liposomal Biosensors

89

2.6. Technology transfer from the IAsys waveguide to the ESPRIT SPR biosensor An ESPRIT SPR biosensor (Eco Chemie B.V., Utrecht, The Netherlands) is used in this study to detect antibodies to MA in human patient sera. The principle of the SPR biosensor is based on the change in the refractive index on a thin gold film surface modified with various materials (Lee et al., 2005) to indicate the binding of ligands, in this case anti-MA antibodies. Both IAsys and ESPRIT biosensors use a cuvette system. The light is totally internally reflected from the sensing surface by means of a prism in both biosensors. SPR signals are related to the refractive index close to the sensor surface, and therefore report the amount of macromolecules bound to the sensor surface. An SPR immunosensor is composed of several important components such as a light source, detector, prism with transducer surface and flow system (Shankaran et al., 2007). The transducer surface is usually a gold film (50–100 nm, on which biomolecules, such as antibody or antigen, are immobilized) on a glass slide optically coupled to the glass prism through refractive index matching oil. The resonance conditions are influenced by the biomolecules interacting with their immobilized ligands on the gold layer. When the molecules interact, the change in the interfacial refractive index can be detected as a shift in the resonance angle. These changes are monitored over time and converted into a sensorgram, from which the kinetics and affinity constants of the interaction can be determined. There has been considerable progress in the development of new methods of immobilizing biological recognition elements onto transducer sensor surfaces (Zhang et al., 2000), a key step in the development of biosensors. The use of self-assembled mono- and multilayers (SAMs) is increasing rapidly in various fields of research, and this applies especially to the construction of biosensors (Zhang et al., 2000, 2008). The uncomplicated procedure for SAM formation and compatibility with metal substrates such as gold for electrochemical measurements enable special benefits for biosensor applications. The term self-assembly involves the spontaneous arrangement of atoms and molecules into an ordered stable form or even aggregate of functional entities (Tecilla et al., 1990). For example, the highly ordered and dense nature of the long-chain alkanethiols of SAMs mimic the cellular microenvironment of lipid bilayer structures, thereby providing novel substrates for immobilized biomolecules (Arya et al., 2006). The molecular self-assembly of long-chain alkanethiol on gold has drawn considerable attention during the past decade, since self-assembled monolayers (SAMs) have strong adhesion to a substrate, high degree of thermal and chemical stability, and mechanical strength (Kim et al., 2001). The stability of the SAMs of the alkanethiol molecules formed on the gold depends on the strength of Au–S bond and the Van der Waals force between a thiol molecule and its surrounding molecules (Han et al., 2004).

90

Y. Lemmer et al.

SAMs can be used as interface layers upon which almost all types of biological components, including proteins, enzymes, antibodies, and their receptors can be loaded (Zhang et al., 2000). Here, the preparation of octadecanethiol (ODT) in absolute ethanol to form a SAM is described. It is characterized by cyclic voltammetry (CV) and applied for the measurement of binding, or inhibition of binding of patient serum antibodies to MAs that are immobilized in liposomes onto the alkanethiol coated ESPRIT biosensor surface. The low solubility of ODT in ethanol is preferred to form the SAMs. Kim et al. (2001) demonstrated that the adsorption rate of alkanethiol onto clean gold when using a quartz crystal microbalance (QCM) biosensor depends on the thiol concentration, temperature, and solvent used. Here, a full coverage of the underivatized Au surface is observed when 10 mM of ODT is used. This is confirmed by a strongly hindered redox reaction when the surface is characterized with a CV instrument. CV experiments are carried out using an Autolab potentiostat PGSTAT 30 from Eco Chemie (Utrecht, The Netherlands) driven by the General Purpose Electrochemical Systems (GPES) data processing software version 4.9. Sodium dodecylsulphate (SDS) and absolute ethanol (analytical grade) were obtained from Merck (Gauteng, SA). ODT, ferricyanide [K3Fe (CN)6], ferrocyanide [K4Fe(CN)6], potassium chloride (KCl), and urea, all analytical grade, were obtained from Sigma-Aldrich (St. Louis, MO). Acetic acid (analytical grade), sodium hydrogen carbonate (NaHCO3), isopropanol (chemically pure), and sodium hydroxide (NaOH) were obtained from Saarchem (Gauteng, SA). ODT (10 mM) is dissolved in absolute ethanol using a water bath sonifier (Ultrasonic Cleaner, Optima Scientific CC, Model: DC150H) for 30 min. Sodium hydrogen carbonate (0.2 M), SDS (0.5%, w/v), sodium hydroxide (50 mM), 1 mM ferrocyanide/ferricyanide, 1 M KCl, and urea (6 M ) are prepared with sterile double distilled water. The coating of the gold disk cannot be monitored in real time, since the ODT is dissolved in absolute ethanol that generates too large refractive index jumps in the sensor signals when alternated with PBS/AE. The underivatized gold disk is incubated for 16 h at room temperature in 10 mM ODT. The ODT-coated gold disk is immersed in a solution of 1 mM ferrocyanide/ferricyanide (used as a redox probe) containing 1 M KCl and scanned at a rate of 25 and 50 mV/s at a potential window of
0.1 to 0.5 V (vs. Ag/AgCl, saturated KCl). The results in Fig. 5.5 show that there is a significant drop in the current response of the ODT-coated disk toward the redox probe in comparison to the uncoated gold disk, indicating the formation of a stable SAM of ODT by S–Au bonds on the surface of the gold disk. The SAM can be maintained after the exposure of the coated surface to several regeneration cycles of absolute ethanol and a mixture of 50 mM NaOH with isopropanol (2:3, v/v).

91

Antibody Detection by Liposomal Biosensors

Underivatised Au ODT coated Fourth regeneration

4.00E-04 3.00E-04 2.00E-04

I/A

1.00E-04 0.00E+00 −1.00E-04 −2.00E-04 −3.00E-04 −4.00E-04 −0.6

−0.4

−0.2

0

0.2 0.4 E/V vs AglAgCl

0.6

0.8

1

Figure 5.5 Testing of the octadecanethiol-coated ESPRIT biosensor gold surface against sequential times of regeneration with a mixture of isopropanol and 50 mM NaOH (2:3, v/v) using cyclic voltammetry. The voltammograms for the different surfaces are shaded to better identify the one from the other.

Kim et al. (2001) reported that partial ODT multilayers on the gold surface could be formed via the formation of disulfides, since thiols are oxidized to disulfides in the presence of oxygen and the solubility of disulfides in ethanol is much less than that of thiols. If a solution of ODT in ethanol is exposed to oxygen and oxidized to disulfide, the oxidized disulfide can be precipitated onto the monolayer. Here, the solution of ODT in absolute ethanol is covered with parafilm to avoid oxygen exposure. The ODT-coated disk is subsequently inserted into the biosensor on a droplet of special refractive index oil, after wiping the glass bottom surface with lens tissue. The PBS/AE is filtered through a 0.2 mm particle retention membrane and degassed with helium for 30 min before use. Degassing is required whenever a hydrophobic surface is created that is exposed to air before liposome addition to prevent the formation of a ‘‘dissolved’’ air layer onto the hydrophobic surface that prevents SPR to occur. The cuvette is flushed with 500 ml ethanol (96%) using the automatic dispenser with simultaneous draining, followed by brief (60 s) flow-washing with PBS/AE. An automated software program sequence can be created to control the addition of all the samples and liquids into the cuvette. Quality of the surfaces are monitored by determining the SPR dips after cleaning the Au ODT-coated surface with 96% ethanol and a mixture of isopropanol and 50 mM sodium hydroxide (2:3, v/v). The samples are transferred from a

92

Y. Lemmer et al.

384 multiwell plate (Bibby Sterilin Ltd, Stone, UK) to the cuvette surface by an autopipettor. First, the baseline of the ESPRIT biosensor is set with 10 ml PBS/AE, followed by addition of 50 ml MA liposomes on the disk for 20 min. The immobilized liposomes are then finally washed five times with 100 ml PBS/AE, to prepare for blocking the surface with saponin. 2.6.1. Use of degassed buffers after liposome coating of the sensor surface A recent study (Eastoe and Ellis, 2007) showed that exposure of lipids to degassed buffers resulted in a detergent effect that destabilized the lipids. The removal of hydrophobic gas by pumping created a limited ability of the degassed water or aqueous buffer to dissolve lipids. To determine whether this will also be the case here where buffers are degassed by bubbling through helium gas, we exposed a liposome-coated surface repeatedly to either degassed or nondegassed buffer. The liposomes-containing MAs are immobilized on ODT-coated gold sensor disks for 20 min. The liposomes are washed five times with degassed or nondegassed PBS/AE, and left for 5 min with mixing to achieve a baseline. This procedure is repeated three times. Figure 5.6A demonstrates how the baseline is affected during movement of degassed PBS/AE over the liposome coat; compared to when buffer is used that is not degassed (Fig. 5.6B). A stable baseline is obtained only when a nondegassed PBS/AE is used. The rest of the procedure in the MARTI-assay that follows after liposome coating is subsequently done with buffer that is not degassed, taking special care that air bubbles do not develop in the fluid lines that could affect the working of the pumps. In several attempts where we test samples with continued use of degassed buffer after liposome coating, we obtain nonreliable results, with the liposome layer often coming apart at the final step of antibody incubation. 2.6.2. Optimization of saponin concentration Different concentrations of saponin (%, m/v) prepared in PBS/AE (0.1%, 0.05%, 0.025%, 0.0125%, and 0.00625%) are tested to block the hydrophobic sites of the MA liposome layer. The stock saponin concentration is 0.1% and the subsequent dilutions are prepared from this stock solution. From the results obtained (Fig. 5.7A), there is a tendency of an increase in saponin accumulation onto MA liposomes immobilized on an ODT-coated gold surface, as the saponin concentration is increased from 0.00625% to 0.05%. At a saponin concentration of 0.05% there is an amount of net saponin accumulation after PBS/AE buffer wash (Fig. 5.7B). An unstable baseline is also obtained when 0.05% saponin is used. A saponin concentration of 0.0125% is chosen as optimal, because it gives a stable baseline and acceptable variation after PBS/AE wash (Fig. 5.7B) as compared to 0.00625% and 0.025%. The differences in optimal saponin concentration used on the IAsys (0.03%) and current ESPRIT biosensors

93

Antibody Detection by Liposomal Biosensors

A

400 350

Angle (m⬚)

300 250 200 150 100 50 Mycolic acid liposomes

0 0 B

500

1000

1500 2000 Time (s)

2500

1500 2000 Time (s)

2500

3000

3500

400 350

Angle (m⬚)

300 250 200 150 100 50 0

Mycolic acid liposomes 0

500

1000

3000

3500

Figure 5.6 Effect of degassed (A) and nondegassed (B) buffer on immobilized mycolic acids liposomes in the ESPRIT biosensor. The arrows indicate where washing cycles with PBS/AE were introduced before allowing a baseline to be reached with mixing before substitution of cell content.

(0.0125%) could be due to different batches of saponin, or that the CPC and ODT activation before immobilization of the MA produces different surface properties. 2.6.3. Optimization of first serum exposure dilution in PBS/AE After optimization of saponin concentration, the next step is to determine which concentration of serum is optimal for the MARTI-assay in the first exposure to antigen. Chung et al. (2005) indicated that serum should be diluted to minimize the nonspecific binding to the biosensor surface. Serum is a complicated protein mixture for direct application to a biosensor surface. The introduction of a first serum exposure at high dilution was previously done on the IAsys biosensor to provide a practical working dilution that did not fully saturate the antigen coat, but was still concentrated enough to give

94

Y. Lemmer et al.

A

Binding response (m⬚)

55

Angle (m⬚)

65

Saponin

45

Time (s)

35

25

15

5 PBS/AE

-5

0.00625

0.0125

0.025

0.05

0.1

Saponin concentration (%, m/v)

B

Binding response (m⬚)

55

45

Angle (m⬚)

65

Saponin Time (s)

35

25

15

5

-5

PBS/AE

0.00625

0.0125

0.025

0.05

0.1

Saponin concentration (m/v, %) Figure 5.7 Optimization of saponin concentration to avoid nonspecific binding on immobilized mycolic acids on the Au surface coated with octadecanethiol. Accumulation of saponin is performed for 5 min (A) and washed with PBS/AE (B). The error bars indicate the standard error of the mean (S.E.M.) and n ¼ 3.

a measurable signal to probe the comparability of the binding signals from the cuvette. This simultaneously blocked off the major nonspecific binding areas and hugely increased the accuracy of the MARTI-test.

95

Antibody Detection by Liposomal Biosensors

Serum samples collected for another study (Schleicher et al., 2002) were used that were obtained from 61 adult patients (aged between 18 and 65 years), who were admitted to the general medical wards of the Helen Joseph Hospital, Johannesburg, South Africa, including a number with active pulmonary TB. The TB-negative patients that were used as controls had medical conditions other than TB and were recruited from the general medical wards. The liposomes are immobilized as described earlier, the surface blocked with 0.0125% saponin, after which 50 ml of PBS/AE is added and left for 5 min to affect a stable baseline. This is followed by addition of 35 ml of either 1/500, 1/1000, 1/2000, or 1/4000 dilutions of serum in PBS/AE. For the assessment of the optimal dilution of the first serum exposure, a second exposure of serum preincubated in MAs-containing or empty liposomes is kept constant at 1/250 in all the experiments. The results show that the chosen serum dilution range of 1/4000–1/500 responded in an almost linear positive correlation between antibody-binding signal and serum concentration with a slight running out at 1/4000 that indicates that the lower limit of the serum concentration is reached. The results obtained in Fig. 5.8 gave a positive linear correlation with a coefficient (r2) of 0.9749 between the serum concentrations and their signal binding response over the range measured, which is a requirement for a successful MARTI-assay.

300

Binding response (m⬚)

250 200 150 100 50 0 1/500

1/1000

1/2000

1/4000

Serum dilution in PBS/AE

Figure 5.8 Optimization of the dilution of serum (P135) for the first exposure to antigen in the MARTI-assay, after 0.0125% saponin blocking of the mycolic acid liposome coat of the ESPRIT biosensor. The error bars indicate the standard deviation. Correlation coefficient (r2) ¼ 0.9749, n  3.

96

Y. Lemmer et al.

The 1/4000 and 1/2000 dilutions are adequate for the first serum exposure, leaving enough room for a positive binding event at the second serum exposure. 2.6.4. Second serum exposure with liposome preincubation P129 (TB-positive) serum was used to optimize the dilution of the second exposure to preincubated serum in MA-containing, or empty liposomes for inhibition studies, following on a first serum exposure to immobilized antigen at a dilution of either 1/4000 or 1/2000. The first exposure should avoid the saturation of antigen with antibody before the addition of preincubated serum. Different dilutions (1/250, 1/500, 1/1000, and 1/2000) of preincubated serum in MA-containing and empty liposomes are applied by 35 ml addition to either 1/4000 or 1/2000 of first serum exposure in PBS/AE, after 10 min of incubation. This is followed by washing away of the unbound antibody with five times 100 ml PBS/AE. The TB-positive patient P129 serum showed a significant decrease of signal when the serum was preincubated in MA-containing liposomes, compared to empty liposomes over a range of 1/250, 1/500, or 1/1000 dilution (Fig. 5.9) after a first serum exposure at 1/4000. There was no inhibition of antibody by MA preincubation observed when 1/2000 dilution of serum was used and binding response signals were also too low.

Binding inhibition response (m⬚)

140 PC

MA

120 100 80 60 40 20 0 1/250

1/500 1/1000 Serum dilution in liposomes

1/2000

Figure 5.9 MARTI antibody-binding inhibition response of preincubated TB-positive P129 serum dilutions inhibited with mycolic acids (MAs)-containing and empty (PC) liposomes after first serum exposure of 1/4000 on immobilized MAs. The error bars indicate the standard deviation. P129 showed significant MA inhibition signals at 1/250, 1/500, and 1/1000 serum dilutions, with P-values of 0.00014, 0.01411, and 0.0393, respectively, but no significant inhibition at 1/2000 serum dilution (P-value of 0.7857). A 95% (0.05) confidence limit was used, n ¼ 3.

97

Antibody Detection by Liposomal Biosensors

Binding inhibition response (m⬚)

120 PC

MA

100 80 60 40 20 0 1/250

1/500 1/1000 Serum dilution in liposomes

Figure 5.10 MARTI-binding inhibition response of various dilutions of preincubated TB-positive patient serum (P129) with mycolic acids (MA)-containing and empty (PC) liposomes after first exposure serum dilution of 1/2000 to surface immobilized MAs. The error bars indicate the standard deviation. No statistical difference (at 95% confidence limit) is obtained at 1/250 and 1/500 dilutions between MA- and PC-inhibited serum, with P-values of 0.116 and 0.356, respectively, while a significant inhibition was observed at 1/1000 (P-value of 0.0086) n ¼ 3.

This shows that the lower limit of serum concentration was reached at 1/2000 dilution to measure the inhibition of anti-MA antibody binding. The results in Fig. 5.10 indicate that inhibition values of 17%, 19%, and 41% were obtained at 1/250, 1/500, and 1/1000 dilutions of serum in liposome solution, respectively, after first serum exposure at 1/2000 dilution. At first sight, it appeared that a better value was obtained by using a first serum exposure of 1/2000 dilution, followed by a second, antigen preincubated serum dilution at 1/1000 dilution (numerical difference: 12.50 millidegrees). However, when looking at the numerical signal difference between MA- and empty liposome-inhibited serum, then the 1/4000 dilution of first serum exposure followed by second serum exposure at 1/500 still gave the best value (numerical difference: 21.53 millidegrees). In addition, the significance of the difference between antibody-binding inhibition with MA liposomes and empty liposomes was significant at 1/250, 1/500, and 1/1000 dilution of serum after first serum exposure at 1/4000 dilution, while only the 1/1000 dilution of inhibited serum produced a significant difference after a first serum exposure of 1/2000 dilution (P-value limit of 0.05). The 1/2000 dilution of first serum exposure appears, therefore, to restrict the workable range of serum dilutions at the second critical serum exposure that provides the inhibition end result. This was confirmed when another TB-positive—HIV-negative serum (P96) was tested and for which a better inhibition response was obtained at the

98

Y. Lemmer et al.

preferred serum dilutions of 1/4000 and 1/500 for first and second serum exposures, respectively, compared to the result obtained with first exposure at 1/2000. At the preferred serum dilutions of exposure, the TB-negative— HIV-negative serum P94 gave the expected zero inhibition value, with a P-value of 0.9863. 2.6.5. Regeneration of the ODT-coated gold disks After dissociation of the unbound serum antibodies to MAs, the surface is regenerated with 100 ml mixture of isopropanol and 50 mM NaOH (2:3, v/v) for 2 min and finally washed with 100 ml of 99% (absolute) ethanol. The surface is washed five times with 100 ml of PBS/AE after each regeneration step to prepare it for a next round of liposome coating on the stable ODT layer. 2.6.6. Cleaning of cuvette and needles A flow wash sequence is used to clean the needles after analyzing approximately 30 sample runs. Sequential washes with 0.5% (w/v) SDS, 6 M urea, 1% (v/v) acetic acid, 0.2 M sodium hydrogen carbonate (NaHCO2), and ddd H2O are done in order to maintain the quality of the SPR signals during repeated measurements. 2.6.7. The optimized MARTI-assay With the lesson learnt of avoiding degassed buffers after liposome coating and the conditions optimized for the blocking of the liposome layer with saponin, titrations of the optimal dilutions for first exposures to serum and second exposure to antigen-inhibited serum dilutions were done. It is concluded that best results were obtained with 1:4000 dilution of serum at first exposure and 1:500 dilution of serum at second exposure. In the second exposure, the serum is preincubated with antigen in order to effect an inhibition of binding signal, as graphically demonstrated in Fig. 5.11. The SPR dips (Fig. 5.11 inserts) between 0% and 10% reflectivity that are associated with the binding profiles proved that the sensor surfaces remained intact and fully activated during the run of the experiments. Using this optimized protocol, four serum samples were selected from the Schleicher et al. (2002) collection and assessed for the presence of anti-MA antibodies. In Table 5.2, the MARTI-assay results are presented and compared with that obtained on ELISA by Schleicher et al. (2002). From Table 5.2, P129 and P96 tested false negative on ELISA and true positive on ESPRIT biosensor, while P94 tested equivocally on ELISA and true negative on ESPRIT biosensor, as it was previously shown on the IAsys biosensor (Thanyani et al., 2008). The MARTI-assay on IAsys biosensor was successfully validated to an accuracy of 82% for the serodiagnosis of active pulmonary TB. The IAsys biosensor system applied to the

99

Antibody Detection by Liposomal Biosensors

A 400

Angle (m⬚)

300

d b

c

Reflectivity [%] 100

200

PC MA

80 60 40

100

30 -2000

a

0

0

1000

2000 Time (s)

0 2000 Angle [m⬚]

3000

4000

B 400

Angle (m⬚)

300 d

200

b

Reflectivity [%] 100

c

PC MA

80 60 40

100

30 -4000 -2000 0 2000 4000 Angle [m⬚]

a

0

0

1000

2000 Time (s)

3000

4000

Figure 5.11 Typical sensorgrams summarizing the process of measuring serum antibody (A ¼ TB-positive P129 and B ¼ TB-negative P94) binding, or inhibition of binding by mycolic acid-containing and empty liposomes, on an ESPRIT biosensor with ODT-coated gold surface and immobilized mycolic acid liposomes. Mycolic acids liposomes were immobilized on the ESPRIT biosensor surface (a), blocked with saponin (b), calibrated with a 1/4000 first exposure of serum (c), and applied to measure the binding and dissociation of 1/500 diluted sera inhibited with empty (thick line) or mycolic acid-containing (thin line) liposomes at lesser dilution (d). The arrows indicate washing with PBS/AE. Table 5.2 MARTI (ESPRIT biosensor) and ELISA analysis compared for their ability to detect antibody to MA in three selected human sera

a b

Patient no.

TB/HIV status

ELISA assaya

P129 P96 P94

TBþHIV
TBþHIV
TBþHIV


1.59 1.05 1.69

MARTI-assayb

25.09 29.53
0.23

Signal to background value of absorbance at 450 nm. Values higher than 2 are taken as positive. Percentage inhibition of antibody binding to MA liposomes. Values higher than 20% are taken as positive.

100

Y. Lemmer et al.

MARTI-test has a weakness in that the channels often do not give matching results, while the cuvettes are 10 times more expensive than the gold disks provided for the ESPRIT biosensor. The ESPRIT biosensor is provided with an adjustable laser setting to compensate for differences in the channel readings as well as an automated pipettor system that reduces variance from one sample to the next. The MARTI-assay as applied in the ESPRIT biosensor has now reached the stage where a result of sample analysis can be guaranteed within 4 h of receipt of the serum. This is the first time that such reliability has been achieved. However, more sera need to be analyzed to confirm the reproducibility of the assay among the HIV-positive population, to prove the value of the MARTI-test against the many studies reported of low sensitivity and specificity with HIV-positive samples using standard techniques of TB diagnosis (Antunes et al., 2002; Hendrickson et al., 2000; Schleicher et al., 2002).

2.7. The cholesteroid nature of MA demonstrated on the ESPRIT biosensor In our previous studies, we have provided evidence for a structural relationship and attraction between free MAs and cholesterol (Benadie et al., 2008). This was supported by demonstrating the interaction between MA and Amphotericin B—an antifungal macrolide agent known to bind to cholesterol (Baginski et al., 2002)—on the IAsys biosensor system. The same principle was confirmed with the ESPRIT biosensor as demonstrated below, in an attempt to determine what the effect of labeling of MA would be on its manifestation of cholesteroid nature. For the preparation of the different liposomes, phosphatidyl choline stock solution (90 ml, 100 mg/ml chloroform) is added to an amber glass vial containing either MA (1 mg) or an equimolar amount of 5-BMF labeled MA (1.35 mg). For the preparation of cholesterol-containing liposomes, phosphatidyl choline stock solution (60 ml, 100 mg/ml chloroform) is added to a cholesterol solution (30 ml, 100 mg/ml chloroform). The samples are mixed well until dissolved, then dried under a stream of N2 gas at 85  C. Saline (2 ml) is then added and the sample is heated on a heat block for 20 min at 85  C. The sample is then vortexed for 1 min, sonicated with a Virsonic probe sonicator until a clear solution forms to indicate vesicle formation, aliquoted at 0.2 ml per vial, lyophilized and stored at
70  C until use. Before use, lyophilized liposomes are reconstituted with buffer (2 ml). The liposomes are placed on a heat block for 30 min at 85  C, vortexed for 2 min and sonified. The final liposome concentration comes to 500 mg lipid/ml.

101

Antibody Detection by Liposomal Biosensors

The binding interaction between Amphotericin B and either cholesterol-, MA-, or 5-BMF labeled MA-containing immobilized liposomes are tested here. An ODT-coated gold disk is mounted in the ESPRIT instrument and the individual liposomes immobilized as described before. The instrument is operated and reagents used at RT. The liposomes are washed five times with nondegassed PBS/AE, and left for 5 min with mixing to obtain a baseline. Amphotericin B (1  10
4 M ) is added to the liposome layer and the direct interaction is recorded for 10 min after which the disk is washed five times with nondegassed PBS/AE, and left for 5 min. The results (Fig. 5.12) confirm the ability of the ESPRIT biosensor to demonstrate that Amphotericin B recognizes both cholesterol and MA, as was previously shown with the IAsys biosensor. In addition, it shows for the first time the intolerance of the system for fettering with the structure of MA by adding a bulky label on its carboxylic acid group.

Relative binding

120

80

40

70 60 50 40 30 20 10 0

Cholesterol

MA

Chol MA

Angle (m⬚)

Angle (m⬚)

0

5BMF-MA 70 60 50 40 30 20

Chol 5BMF-MA

10 0

5

10 15 Time (s)

20

25

0

0

5

10 15 Time (s)

20

25

Figure 5.12 Normalized AmB-binding capacity on immobilized lipid antigens cholesterol, MA, and 5BMF–MA and binding curves of MA versus cholesterol and 5BMF–MA versus cholesterol. The error bars indicate the standard deviation, n ¼ 3 for each set.

102

Y. Lemmer et al.

3. Conclusion Antibodies to MA in serum as surrogate markers of active TB can be detected by making use of liposomes as MA antigen carriers in a wave-guide evanescent field biosensor (Thanyani et al., 2008). However, SPR biosensors are more generally in use, more amenable to high-throughput screening, have lower running costs, and provide more comparable binding signals in twin-cell cuvettes. The more preferred technique of SPR biosensor technology, using standard methods to form a hydrophobic surface onto the gold disks on which the MA-containing liposomes can be immobilized, is reported here. This demonstrates the use of liposomes in biosensors to detect large biomolecules such as antibodies. By using Amphotericin B as a small ligand to bind cholesterol and MA in a wave-guide biosensor the cross-reactivity of antibodies to MAs in ELISA can be explained (Benadie et al., 2008). This can also be demonstrated using an SPR biosensor. The biosensor antibody detection approach (MARTI—TB serodiagnostic test) remains unaffected by the antibody cross-reactivity between MA and cholesterol.

ACKNOWLEDGMENTS This research was supported by a Marie-Curie Early Stage Training Fellowship (Molfun) to Y. Lemmer, the Cancer and Polio Research Fund and the North West Cancer Research Fund, Adcock Ingram Limited (South Africa), the National Research Foundation of South Africa (Technology and Human Resources for Industry Programme), the European and Developing Countries Clinical Trial Partnership Grant No. 2004.1.R.d1, and the Medical Research Council of South Africa. Dr. ten Bokum held a Claude Leon Harris Foundation postdoctoral fellowship.

REFERENCES Antunes, A., Nina, J., and David, S. (2002). Serological screening for tuberculosis in the community: An evaluation of the Mycodot procedure in an African population with high HIV-2 prevalence (Republic of Guinea-Bissau). Res. Microbiol. 153, 301–305. Arya, S. K., Solanki, P. R., Singh, R. P., Pandey, M. K., Datta, M., and Malhotra, B. D. (2006). Application of octadecanethiol self-assembled monolayer to cholesterol biosensor based on surface plasmon resonance technique. Talanta 69, 918–926. Baginski, M., Resat, H., and Borowski, E. (2002). Comparative molecular dynamics simulations of amphotericin B-cholesterol/ergosterol membrane channels. Biochim. Biophys. Acta 1567, 63–78. Bangham, A. D. (1983). Liposomes: An Historical Perspective. Marcel Dekker, New York. Beatty, W. L., Rhoades, E. R., Ullrich, H. J., Chatterjee, D., Heuser, J. E., and Russell, D. G. (2000). Trafficking and release of mycobacterial lipids from infected macrophages. Traffic 1, 235–247.

Antibody Detection by Liposomal Biosensors

103

Beckman, E. V., Porcelli, S. A., Morita, C. T., Behar, S. M., Furlong, S. T., and Brenner, M. B. (1994). Recognition of a lipid antigen by CD1-restricted ab þ T cells. Nature 372, 691–694. Benadie, Y., Deysel, M., Siko, D. G., Roberts, V. V., Van Wyngaardt, S., Thanyani, S. T., Sekanka, G., Ten Bokum, A. M., Collett, L. A., Grooten, J., Baird, M. S., and Verschoor, J. A. (2008). Cholesteroid nature of free mycolic acids from M. tuberculosis. Chem. Phys. Lipids 152, 95–103. Butler, W. R., and Kilburn, J. O. (1990). High-performance liquid chromatography patterns of mycolic acids as criteria for identification of Mycobacterium chelonae, Mycobacterium fortuitum, and Mycobacterium smegmatis. J. Clin. Microbiol. 28, 2094–2098. Chan, E. D., Reves, R., Belisie, J. T., Brennan, P. J., and Hahn, W. E. (2000). Diagnosis of tuberculosis by a visually detectable immunoassay for lipoarabinomannan. Respir. Crit. Care Med. 161, 1713–1719. Chung, J. W., Kim, D. S., Bernhardt, R., and Pyun, J. C. (2005). Application of SPR biosensor for medical diagnostics of human hepatitis B virus (hHBV). Sens. Actuator B 111, 416–422. Dennison, C., and Lovrein, R. (1997). Three-phase partitioning: Concentration and purification of proteins. Protein Exp. Purif. 11, 149–161. Eastoe, J., and Ellis, C. (2007). De-gassed water and surfactant-free emulsions: History, controversy, and possible applications. Adv. Colliod Interface Sci. 134, 89–95. Goodrum, M. A., Siko, D. G. R., Niehues, T., Eichelbauer, D., and Verschoor, J. A. (2001). Mycolic acids from Mycobacterium tuberculosis: Purification by countercurrent distribution and T cell stimulation. Microbios 106, 55–67. Han, J., Wang, X., and Kwok, D. L. (2004). Structure and stability of self-assembled monolayer for octadecanethiol adsorbed on flame annealing gold substrate and its potential application to microfluidics. ICMENS 4, 1–4. Hartley, G. S., and Runnicles, D. F. (1938). The determination of the size of paraffin-chain salt micelles from diffusion measurements. Proc. Roy. Soc. London: Math. Phys. Sci. 168, 420–440. Hendrickson, R. C., Douglass, J. F., Reynold, L. D., McNeill, P. D., Carter, D., Reed, S. G., and Houghton, R. (2000). Mass spectrometric identification of Mtb81, a novel serological marker for tuberculosis. J. Clin. Micribiol. 38, 2354–2361. Houghton, R. L., Lodes, M. J., Dillon, D. C., Reynolds, L. D., Day, C. H., McNeill, P. D., Hendrickson, R. C., Skeiky, Y. A. W., Sampaio, D. P., Badaro, R., Lyashchenko, K. P., and Reed, S. G. (2002). Use of multiepitope polyproteins in serodiagnosis of active tuberculosis. Clin. Diagn. Lab. Immunol. 9, 883–891. Kim, H. D., Noh, J., Hara, M., and Lee, H. (2001). An adsorption process study on the selfassembled monolayer formation of octadecanethiol chemisorbed on gold surface. Bull. Korean Chem. Soc. 22, 276–280. Korf, J., Stoltz, A., Verschoor, J. A., De Baetselier, P., and Grooten, J. (2005). The Mycobacterium tuberculosis cell wall component mycolic acid elicits pathogen-associated host innate immune responses. Eur. J. Immunol. 35, 890–900. Korf, J. E., Pynaert, G., Tournoy, K., Boonefaes, T., van Oosterhout, A., Ginneberge, D., Haegeman, A., Verschoor, J. A., De Baetseller, P., and Grooten, J. (2006). Macrophage reprogramming by mycolic acids promotes a tolerogenic response in experimental asthma. Am. J. Respir. Crit. Care Med. 174, 152–160. Lee, W. J., Sim, J. S., Cho, M. S., and Lee, J. (2005). Characterization of self-assembled monolayer of thiol on a gold surface and the fabrication of a biosensor chip based on surface plasmon resonance for detecting anti-GAD antibody. Biosens. Bioelectron. 20, 1422–1427. Pan, J., Fujiwara, N., Oka, S., Maekura, R., Ogura, T., and Yano, L. (1999). Anti-cord factor (trehalose 6, 60 -dimycolate) IgG antibody in tuberculosis patients recognizes mycolic acid subclasses. Microbiol. Immunol. 43, 863–869.

104

Y. Lemmer et al.

Reischl, U. (1996). Application of molecular biology-based methods to the diagnosis of infectious diseases. Front. Biosci. 1, 72–77. Samanich, K. M., Keen, M. A., Vissa, V. D., Harder, J. D., Spencer, J. S., Belisle, J. T., Zolla-Pazner, S., and Laal, S. (2000). Serodiagnostic potential of culture filtrate antigens of Mycobacterium tuberculosis. Clin. Diagn. Lab. Immunol. 7, 662–668. Schleicher, G. K., Feldman, C., Vermaak, Y., and Verschoor, J. A. (2002). Prevalence of anti-mycolic acid antibodies in patients with pulmonary tuberculosis co-infected with HIV. Clin. Chem. Lab. Med. 40, 882–887. Shankaran, D. R., Gobi, K. V., and Miura, N. (2007). Recent advancements in surface plasmon resonance immunosensors for detection of small molecules of biomedical, food and environmental interest. Sens. Actuators B 121, 158–177. Tatulian, S. (1993). Phospholipids Handbook. Marcel Dekker Inc., New York. Tecilla, P., Dixon, P. D., Slobodkin, G., Alavi, D. S., Waldeck, D. H., and Hamilton, A. H. (1990). Hydrogen-bonding self-assembly of multichromophore structures. J. Am. Chem. Soc. 112, 9408–9410. Thanyani, S. T., Roberts, V., Siko, D. G., Vrey, P., and Verschoor, J. A. (2008). A novel application of affinity biosensor technology to detect antibodies to mycolic acid in tuberculosis patients. J. Immunol. Methods 332, 61–72. Verschoor, J. A., and Onyebujoh, P. (1999). The menace of the AIDS-tuberculosis combo: Any solutions? Bioessays 21, 365–366. Westwell, A. E., and Anacker, E. W. (1959). Adsorption of cetylpyridinium chloride on glass. J. Phys. Chem. 63, 1022–1024. Zhang, S., Wright, G., and Yang, Y. (2000). Materials and techniques for electrochemical biosensor design and construction. Biosens. Bioelectron. 15, 273–282. Zhang, Y., Wang, H., Yan, B., Zhang, Y., Li, J., Shen, G., and Yu, R. (2008). A reusable piezoelectric immunosensor using antibody-adsorbed magnetic nanocomposite. J. Immunol. Methods 332, 103–111.

C H A P T E R

S I X

Solid Lipid Nanoparticle Formulations: Pharmacokinetic and Biopharmaceutical Aspects in Drug Delivery Eliana B. Souto*,† and Slavomira Doktorovova´* Contents 106 107 107 112 113

1. 2. 3. 4.

Introduction Production of SLN Pharmacokinetics and Pharmacodynamics Modified Release Profile 4.1. Short or relatively long half-life drugs 4.2. Targeting to a specific tissue by implantation for long-term release 5. Biopharmaceutical Aspects of Administration Routes 6. Clinical Pharmacology 6.1. Local effects and systemic distribution 6.2. Dose size 6.3. Concentration fluctuations and therapeutic range 6.4. Toxicological concentration 7. Concluding Remarks References

113 114 117 117 120 121 121 121 122

Abstract Solid lipid nanoparticles (SLNs) have emerged as important tools to modify the release profile for a large number of drugs including protein and peptide molecules. SLNs are produced from biocompatible and biodegradable lipid materials, making them a promising therapeutic strategy for drug targeting and delivery, and surmounting the inherent limitations of regulation acceptance. Due to their versatility in loading both lipophilic and hydrophilic molecules in the solid lipid matrix, SLNs depict the ability to prolong, extend or * {

Department of Pharmaceutical Technology, Faculty of Health Sciences, Fernando Pessoa University, Porto, Portugal Institute of Biotechnology and Bioengineering, Centre of Genetics and Biotechnology, University of Tra´s-os-Montes and Alto Douro (CGB-UTAD/IBB), Vila Real, Portugal

Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64006-4

#

2009 Elsevier Inc. All rights reserved.

105

106

Eliana B. Souto and Slavomira Doktorovova´

sustain the release profile of the loaded molecules, therefore reducing the repeated administration, and increasing the therapeutic value of a certain treatment. Additional advantages include reduction of drug toxicity and increase of drug bioavailability. To develop SLN formulations for drug targeting and delivery, a basic pharmacokinetic understanding of drug distribution is of major relevance, as well as the biopharmaceutical aspects of the administration route. This chapter provides a fundamental understanding of the pharmacokinetic properties of SLNs, which influence both biopharmaceutical and clinical profiles of the loaded molecules.

1. Introduction Lipid-based drug delivery systems (LDDS) are an interwoven field of science requiring a holistic approach of several research areas, including concepts and applications of pharmaceutical technology, pharmacology, pharmacokinetics, and biopharmaceutics. Within the LDDS, solid lipid nanoparticles (SLN) have emerged as novel modified release systems for several drugs, including protein and peptide molecules (Martins et al., 2007, 2009; Mu¨ller et al., 2006, 2008). SLN are composed of a solid matrix based on very chemically different lipid molecules (e.g., waxes, mixtures of mono-, di-, and triacylglycerols, fatty acids) generally recognized as being safe. This provides the advantage of using physiological components and/or excipients accepted for oral and topical administration, reducing the risk of acute and chronic toxicity. For the production of thermodynamically stable SLN, these particles need to be stabilized by a film of surfactant molecules surrounding the lipid matrix in aqueous dispersion. Surfactants such as polysorbates, poloxamers, and phospholipids are very common in SLN formulations for several administration routes. One can have a neat delivery system, but a major obstacle for its introduction into clinical practice is the need to undertake toxicity studies, in particular for parenteral drug delivery. The status of the raw materials used for SLN production is therefore discussed as a function of the administration route. Topical and oral administration routes for SLN are usually nonproblematic. As a result of the biodegradation studies, one can expect that the degradation velocity of SLN composed of different lipids but stabilized with the same surfactant or surfactant mixture will depend on the chemical nature of the lipid matrix. On the other hand, for SLN composed of the same lipid but stabilized by different surfactants or surfactant mixtures, the degradation velocity will be mainly determined by the chemical nature of the surfactants. As a result, the degradation velocity of the lipid matrix can

SLN Formulations: Pharmacokinetics and Biopharmaceutics

107

be adjusted (accelerated or delayed) by modifying the surfactant composition. Based on the differences in degradation velocity, SLN are expected to be a system with great flexibility to modify the release profiles for different drugs.

2. Production of SLN SLN are easily produced by a melt emulsification followed by a hot or cold high-pressure homogenization (HPH) process. Briefly, for the hot HPH the lipid phase is previously melted 5–10
C above its melting point, followed by drug dissolution or dispersion in the melted phase. Stirring this melted phase in a hot surfactant solution, a pre-emulsion is produced. The obtained pre-emulsion is homogenized under high pressure, producing a hot nanoemulsion, which is further cooled down, recrystallizing the lipid, and forming SLN. The cold HPH technique requires a previous step of melting the solid lipid so that the drug can be dissolved and/or admixed to this phase. Applying liquid nitrogen or dry ice, the lipid phase cools down rapidly, and solidifies and then, by means of mortar milling, it is ground, resulting in microparticles. These microparticles are further dispersed in a cold aqueous surfactant solution, producing a presuspension that is homogenized at or below room temperature using the HPH.

3. Pharmacokinetics and Pharmacodynamics For the majority of the modified release LDDS, a basic pharmacokinetic understanding of a given drug’s disposition in the body is of paramount importance, since LDDS are not designed only to release the drug at a controlled, delayed, or prolonged rate, but are expected to reach the target and maintain a certain drug concentration. Drug disposition is a sequentially and simultaneously occurring process, depending on the absorption, distribution, metabolism, and elimination of the drug. For modified release SLN, the occurrence of general absorbability of the drug when released from the lipid matrix must be established. The intrinsic rate of absorption will be dictated by the lipid matrix (lipid nature, surfactant composition, particle size, surface properties), which means that the rate-limiting step is the release of the drug from the SLN. From a technological point of view, modified release LDDS should be multiple-dose systems designed to depict a steady-state concentration on the target tissue. The magnitude of this concentration is dependent on the amount of drug per unit of time (dose) and on the loss of drug from the volume of distribution per unit of time (clearance).

108

Eliana B. Souto and Slavomira Doktorovova´

In SLN, the release profile is also governed by the polymorphism of the lipid matrix. Polymorphism is the ability of a compound to crystallize in more than one distinct crystalline species with different internal lattices (Souto et al., 2006). Lipid crystallization and polymorphic transformations during storage time are important parameters for the performance of SLN as delivery system (Bunjes and Unruh, 2007; Kuntsche et al., 2005). The internal structure of the lipid matrix can vary depending on the threedimensional structure of triacylglycerols (TAGs) that can adopt the following structures upon crystallization: the hexagonal (a, H) phase, with a subcell lattice spacing of 0.415–0.42 nm, the orthorhombic perpendicular (b0 , O?) phase, with a subcell strong lattice spacing of 0.42–0.43 and 0.37–0.40 nm, or as the triclinic parallel (b, T||) phase with a subcell strong lattice spacing of 0.46 nm. Other polymorphic forms may also be found with complex acylglycerols including mixed acid TAGs or partial acylglycerols; for example, multiple b0 and b, sub-a, or intermediate forms, usually mentioned as bi. However, the nomenclature and properties of monoacid TAGs can also be used for these complex acylglycerols that have similar crystal packing. Once a stabilized SLN formulation has been produced, the release is faster from a, than from b0 and b phases. Crystallization of bulk TAGs from the melt after rapid cooling usually occurs in the less stable a-form, which transforms via the b0 -form, into the most stable b-form upon heating or during storage. In colloidal dispersions such as SLN, these transformations of TAGs are faster than in the bulk material, which leads to a change in the relative fraction of the polymorphic forms. Depending on the chemical nature of the lipid and on the production conditions, different fractions of a and b0 modifications may occur. This phenomenon can lead to a reduction of the melting point, or more precisely, changes in form and shift of the melting peak. The long-term stability of these polymorphic forms is less likely to occur, leading to a gradual transformation into more stable polymorphic forms, that is, increasing the content of b0 /bi and finally b. The occurrence of polymorphic transformations is not desired because the change in lipid structure is responsible for drug expulsion during storage and changes in the release profile of the incorporated drug (Bunjes and Unruh, 2007; Johnsson et al., 2005). When a drug is administered into the systemic circulation either directly introduced by intravascular injection or reaching the blood by absorption after extravascular administration, it is distributed not only throughout the bloodstream in the organs of immediate equilibrium, but also to other organs and tissues. It may also be eliminated by metabolism and/or by excretion. Since the molecules entering first will have been distributed and perhaps eliminated before the last molecules are absorbed, the entire process is dynamic, that is, made up of simultaneously occurring individual kinetic processes.

109

SLN Formulations: Pharmacokinetics and Biopharmaceutics

Drug accumulative release wt (%)

120 Free solution

100

SLN formulation

80 60 40 20 0 0

20

40 Time (h)

60

80

Figure 6.1 Temozolomide release profiles from free solution (♦) and from SLN (e) in pH 6.8 PBS medium (modified after Huang et al., 2008a).

In modified release LDDS, the absorption of the drug after released from their matrix does not occur instantaneously, obviously delaying the distribution and elimination of the drug from the bloodstream. This can play an important role in reducing the adverse side effects of drugs with a short therapeutic window, since the drug is effectively released within the safety range (Huang et al., 2008a; Mu¨ller et al., 2006). SLN were reported to reduce the cardio toxicity and nephrotoxicity of temozolomide in comparison to a solution of free drug (Huang et al., 2008a). After i.v. administration of the drug dissolved in a liquid vehicle a high peak of drug concentration was depicted early after administration, followed by a relatively rapid decrease (Fig. 6.1). The peak concentration of temozolomide administered in SLN was lower, resulting from the fact that the drug is not immediately available for redistribution nor elimination until it is released from the nanoparticle structure. The mechanism of drug release form SLN may involve diffusion or degradation of the lipid matrix and subsequent diffusion (Manjunath and Venkateswarlu, 2006; Mu¨ller et al., 1996; Olbrich et al., 2002). Only after occurrence of one of these phenomena, the drug molecule is released into systemic circulation. As the drug is released subsequently, the maximal achieved concentration is not as high as the Cmax of drug administered in solution or emulsion. This may be particularly important in case of highly potent drugs, where initial concentration peak may reach toxic concentration and side effects may occur. Studies performed by Manjunath and coworkers using clozapine (Manjunath and Venkateswarlu, 2005) or nitrendipine (Manjunath and Venkateswarlu, 2006) as model drugs encapsulated in SLN revealed that SLN could increase the bioavailability of both drugs in comparison to free drug suspensions at all-time points. Although the concentration versus time

110 Drug plasma concentration (ng/m)

Eliana B. Souto and Slavomira Doktorovova´

400 350 300 250 200 150 100 50 0

SLN formulation Free solution 0

10

20

30 Time (h)

40

50

60

Figure 6.2 Plasma concentration versus time profiles of after administration of 10 mg/kg temozolomide as free solution () and as SLN (▲) in mice (modified after Huang et al., 2008a).

profiles were similar, the free drug suspensions appear to be less effective in delaying the distribution and elimination of the drug. Assuming first-order elimination, which is the case with most drugs, and plotting the concentration versus time profiles, the terminal slope is the elimination phase, characterized by a straight line. Figure 6.2 gives an example of such a profile obtained after i.v. administration of temozolomide-loaded SLN composed of stearic acid and lecithin (Huang et al., 2008a). The reasons behind the profile obtained as free solution may be the following: (i) the drug is rapidly absorbed and rapidly distributed between the systemic circulation (including organs of instant equilibrium) and the tissues to which the drug eventually goes; (ii) the drug is absorbed, but slowly (peripheral compartment); and (iii) the drug is slowly absorbed but rapidly distributed between the central and peripheral compartments. In SLN, the drug release usually takes longer than the distribution phase. As such, a monocompartmental model is usually obtained. In case of the peroral administration route, the uptake of particles from the gastrointestinal (GI) mucosa also needs to be considered. Due to their small size, SLN are directly taken up in the GI tract. The typical adhesiveness of SLN to the mucosal surface of the intestine walls provides longer contact of the nanoparticles with the gut, and, thus, enhanced time for drug absorption. The encapsulation of drugs into the lipid matrix also protects the drug from exposure to the environment with different pH or degradating enzymes and bacteria. Furthermore, the presence of the surfactant film used to stabilize SLN may also contribute to the higher permeability of the intestinal wall cell membranes (Manjunath and Venkateswarlu, 2005). The resulting concentration versus time profiles may therefore show higher concentrations of the drug in plasma for a longer time when compared to a solution or suspension of the same drug (Fig. 6.2). The exception may be the first few minutes after administration, when the concentration of the drug administered as a solution or suspension will rise faster (Zara et al., 2002a).

SLN Formulations: Pharmacokinetics and Biopharmaceutics

111

Improved bioavailability of poorly soluble drugs may also be achieved by formulating them into LDDS. Formulating those drugs in SLN may provide concentration versus time profiles with comparable low variations and elimination rates, but the initial burst of the drug is avoid (Mu¨ller et al., 2006). The effect of lipid matrix composition on pharmacokinetic parameters has been examined by Manjunath and Venkateswarlu (2005). Higher plasma concentrations of clozapine were achieved from SLN composed of long acyl chain length TAGs because of the extended time required for their degradation. Likewise, higher drug concentrations were achieved with positively charged SLN compared to that with SLN of lower or negative charge, even if composed of the same lipid. Elimination or terminal half-life is the time required to reduce the concentration in blood, plasma, or serum to one-half after equilibrium has been reached. This parameter can be determined from the slope of the terminal line of a semilogarithmic plot by regression analysis. This parameter is important for the selection of the type (matrix) of SLN. The shorter the terminal half-life the greater will be the amount of the drug that will be incorporated into the system. Only drugs whose terminal half-life can be correlated with the pharmacological response are candidates for modified release LDDS. The reticuloendothelial system (RES) is responsible for removal of small xenobiotics from systemic circulation. Kuppfer cells take up all xenobiotics smaller than 7 mm. The elimination of small xenobiotics from the bloodstream is facilitated by adherence of serum opsonines onto the surface of the particles, making their surface more suitable to be recognized and eliminated by phagocytic cells (Wang and Wu, 2006). Opsonization may be avoided by coating the nanoparticles with hydrophilic polymers, such as poloxamers or polyethylenglycols (PEG), producing the so-called ‘‘stealth SLN’’ (Fundaro et al., 2000; Sanjula et al., 2009; Wang and Wu, 2006; Zara et al., 2002b; Zhang et al., 2008a). Poloxamer 188 also serves as a surfactant in SLN formulations, in addition to providing a hydrophilic coating that enhances the mean residence time (MRT) of the encapsulated drug (Huang et al., 2008a). The MRT is the mean time a drug molecule resides in the body, being defined as the time corresponding to 63.2% elimination from the body. As such, modified release LDDS should have a MRT significantly longer than that obtained with conventional dosage forms. Formulation of drugs in SLN generally leads to enhanced MRT in the systemic circulation as well as in various tissues. Yang et al. (1999) reported that MRT of camptothecin in systemic circulation increased 18 times when administered as SLN, and a fourfold increase has been observed for liver, spleen, kidneys, lungs, and heart. The reason behind this enhanced MRT may be the delayed elimination of camptothecin, since it takes place only after the drug is released from the lipid

112

Eliana B. Souto and Slavomira Doktorovova´

matrix. Another factor influencing the elimination, and thus the MRT, is the stealthiness of the SLN provided by poloxamer. Presystemic loss of drug can occur regardless of whether the drug is administered systemically or locally, except when given intravascularly. It is that fraction of drug that is degraded, inactivated, or metabolized after its release from the dosage form. Tissue degradation should not be underestimated. A decrease of presystemic loss when formulating drugs in SLN may also be a reason for the enhanced bioavailability that is usually observed. Most prominent is the presystemic loss of drug upon peroral administration, namely degradation by intestinal content and enzymes, biotransformation by intestinal microbial flora, the first-pass effect, the metabolism in the gut wall, mesenteric veins, portal vein, and liver. Many drugs are given perorally or deep rectally, undergoing an extensive first-pass effect, which is the biotransformation of the drug in the gut lumen prior to absorption and in the intestinal epithelium and/or liver after permeation of the intestinal mucosa (presystemic), but before entering systemic circulation. For such compounds, the GI tract is a poor choice for administration, and other routes (transdermal, buccal, nasal) need to be used. In case of topical administration the drug may be metabolized in the skin, or in the case of ocular administration the drug may be lost in the tear fluid, which is drained into the nasal cavity.

4. Modified Release Profile One of the most important features of SLN is the modified release profile usually depicted by the loaded drugs. The localization of the drug within the SLN matrix will influence its release. Three theoretical models have been proposed to describe the SLN structure, each of them being responsible for a particular release profile (Mu¨ller et al., 2002; Souto and Mu¨ller, 2007). The drug-enriched shell model describes the situation when the drug is present at such a concentration that it is fully solubilized by the lipid. In the course of SLN production, the solidification of lipid starts first, followed by solidification of drug-enriched lipid melt around the lipid cores. The drug is therefore located mostly near the surface of the particle, thus creating a drug-enriched shell. The release of the drug from particles with this structure will be fast, and burst release of the drug shortly after the administration may occur. This is the case of glucocorticoids when formulated into SLN, where the molecules are attached to the particle surface rather than being incorporated into the particle core (Lombardi Borgia et al., 2005), thereby causing their fast release after administration (Hu et al., 2006). Extended or prolonged release of the drug occurs when the drug is located in the core of the SLN, as described by the drug-enriched core

SLN Formulations: Pharmacokinetics and Biopharmaceutics

113

model. This structure may be created when the drug is present at concentrations near or higher than its solubility limit in the lipid phase. During the production, the drug precipitates, creating the core of the particle, which is subsequently wrapped by a lipid shell as the lipid solidifies around the crystallized drug–lipid core. The drug is therefore covered by a practically drug-free lipid shell, which hinders its diffusion from the matrix. The release profiles may be prolonged for several hours, to several days (Esposito et al., 2008; Huang et al., 2008a; Kumar et al., 2007). The matrix structure is dependent on the composition and on the production parameters (Souto and Mu¨ller, 2007). The in vitro release will correlate with the concentration versus time profile in vivo. The prolonged release of the drug will delay its redistribution and elimination, thus providing greater MRT both in systemic circulation and tissues, at the same time avoiding high peak concentrations that may occur after i.v. administration of free drug in a solution or in a suspension.

4.1. Short or relatively long half-life drugs Short-half-life drugs, such as butyrate or peptides (calcitonin, insulin), were shown to be protected from immediate clearance by encapsulation into SLN. For example, butyric acid, that is normally eliminated from the bloodstream within minutes and does not reach a therapeutic level even upon large dose administration, could achieve a concentration sufficient to exhibit its anti-inflammatory and antineoplastic effects (Brioschi et al., 2008; Dianzani et al., 2006). Moreover, the enhanced protection against drug degradation and longer circulation enables reached the brain (Brioschi et al., 2007). In oral administration, cholesteryl butyrate-based SLN proved to be more efficient than the prodrug or salt of the active drug form. A relatively long half-life drug concentration could also be improved. Bromocriptine loaded in SLN could maintain the therapeutic concentration in rats for over 5 h and improve the mobility, while the effect of free drug was nearly absent within this period of time (Esposito et al., 2008).

4.2. Targeting to a specific tissue by implantation for long-term release Enhanced circulation time of SLN may favor their localization in specific tissues. Different tissue distribution was found in mice after administration of temozolomide (Huang et al., 2008a). Of particular relevance was the increased concentration in the brain and decreased concentration in heart and kidneys, where the drug may exhibit toxic effects. In general, the nanoparticles tend to accumulate in macrophages of the RES organs that remove them from the circulation. Passive targeting to RES organs is achieved solely by increasing the circulation time (Xu et al., 2009;

114

Eliana B. Souto and Slavomira Doktorovova´

Ye et al., 2008). Targeting to these organs may be even enhanced by surfacing the nanoparticles with galactosides (Lian et al., 2008). SLN may also be passively targeted to solid tumors due to the leaky vasculature of these latter. If present in the circulation during a sufficient time, SLN tend to accumulate in the tumor. This enhanced retention and permeation effects may be used for passive targeting provided that the surface characteristics of the nanoparticle enable them to escape the RES for a sufficient time (Reddy et al., 2006; Ruckmani et al., 2006b; Tausch et al., 2008).

5. Biopharmaceutical Aspects of Administration Routes Several routes have recently gained increased attention for LDDS. Based on the documentation and state-of-art of topical administration of such systems, the first lipid nanoparticle formulations have entered the market (Mu¨ller et al., 2007; Pardeike et al., 2008). These are in general designed for cosmetic purposes; however, their suitability for application on skin, documented by these products, is of high importance when considering the application of actives for skin diseases treatment. Several drugs used for the treatment of skin disorders may cause side effects either in the skin or, if they happen to enter the systemic circulation, even systemic side effects. This is of paramount importance in terms of application for long periods of time, which is the case of most skin diseases. The highly lipophilic drugs partition into the skin lipids of its horny layer and permeate toward the epidermis, and eventually to the dermis which is rich in vasculature. Although the dermis is less lipophilic than the upper parts of the skin, if the drug reaches this layer, it may enter into the bloodstream, a condition not acceptable for drugs meant for local action. Due to insolubility in water, the drugs are formulated in systems containing solubilizing agents that may subsequently increase their penetration into deeper layers of the skin. The extent of drug permeation through the skin is dependent on more factors. The penetration of the drug into the deeper layers of the skin is influenced by the skin/vehicle partition coefficient. Encapsulation of a lipophilic drug into SLN may decrease this coefficient, maintaining a higher portion of the drug in the upper parts of the skin and preventing the drug to permeate through the skin (Fang et al., 2008; Iscan et al., 2006). The simple fact that the drug is soluble in the lipid matrix of SLN explains this effect. As the horny layer and epidermis are highly lipophilic, SLN may penetrate them easily and accumulate preferentially in these upper parts of the skin. As shown for cyproterone acetate-loaded SLN, the nanoparticles accumulated mostly within the first 100 mm of the skin (Stecova et al., 2007).

SLN Formulations: Pharmacokinetics and Biopharmaceutics

115

Similar results were observed for a fluorescent dye loaded in a lipid nanoparticle formulation composed of cetyl palmitate and medium-chain TAGs (Teeranachaideekul et al., 2008a). The accumulation of drug administered in SLN in the epidermis was reported for several drugs for which systemic absorption is unwanted, for example, podophyllotoxin (Chen et al., 2006), psoralens (Fang et al., 2008), isotretinoin (Liu et al., 2007b), and several NSAIDS ( Joshi and Patravale, 2006, 2008; Puglia et al., 2008). Enhancement of skin bioavailability of a hydrophilic drug (penciclovir) was also achieved by loading it in SLN composed of glyceryl monostearate and lecithin (Lv et al., 2009). The skin penetration effects depicted by SLN are also dependent on the occlusive effect that these particles show upon topical administration, in addition to the permeation enhancing effects of the lipid materials employed for SLN production. The occlusive properties of SLN are documented by several studies (Souto and Mu¨ller, 2008). These properties can be controlled by the degree of crystallinity of the lipid matrix. The presence of small amounts of liquid lipids inside the solid lipid matrix of SLN destabilize the crystal order, leading to lower occlusion (Wissing and Mu¨ller, 2002). The surfactants employed in SLN formulations will also influence the extent to which the drug permeates through the skin, due to their own permeation enhancement effects. Lecithin-stabilized SLN showed significantly higher epidermis localization than polysorbate stabilized SLN (Chen et al., 2006). The SLN may penetrate the skin both along the stratum corneum and the hair the follicle route (Chen et al., 2006). The residence time on the skin surface of drugs delivered by means of SLN formulations was also found to be significantly longer than when administered by classical topical dosage forms (Puglia et al., 2008). With improved penetration into the upper layer of the skin, together with release of the drug prolonged over a period of time, lower doses of the drug can be administered. As the permeation through the skin is usually low, side effects of the administered drug may be avoided. Due to their small size, SLN adhere onto the mucosal surfaces, which can also be used for improvement of bioavailability, not only in skin but also in ocular and pulmonary applications. The increased residence time is of particular relevance in ophthalmic applications, where the drug solution is removed within a short time. Increased corneal permeability and precorneal retention of ibuprofen formulated in lipid nanoparticles has been reported recently (Li et al., 2008). Particle size is of utmost importance in effective pulmonary application. SLN even seem promising for application of peptides and proteins for systemic delivery through the lung, as shown for insulin-loaded SLN (Liu et al., 2008a), which increased significantly the bioavailability of insulin, in comparison to control. Uptake of SLN from alveoli to the lymph system was shown by radiolabeled SLN (Videira et al., 2002),

116

Eliana B. Souto and Slavomira Doktorovova´

providing a possibility for targeting the lymph nodes, and to reach the systemic circulation. Soybean phosphatidylcholine is used as an absorption enhancer in pulmonary delivery of proteins and peptides, and was also reported to facilitate nebulization of a nanoparticle formulation (Liu et al., 2008a). At the same time it can be employed as the surfactant stabilizing the SLN. Another promising administration route for SLN is the oral route. SLN undergo enzymatic degradation by lipases/co-lipases in the gut, the structure is disrupted and micellar structures with bile salts are assembled. These are subsequently taken up actively by intestinal epithelial cells or M-cells of Peyer’s patches (des Rieux et al., 2006). The drug incorporated into SLN will therefore be incorporated into the micelles and be absorbed together with the lipids. This effect has been suggested as the mechanism for increased absorption of drugs from the gut, as observed in the case of SLN carrying cyclosporine A (Mu¨ller et al., 2006) or salmon calcitonin (Martins et al., 2009). The absorption of SLN from the gut into systemic circulation was confirmed by Yuan et al. (2007), who monitored SLN loaded with a fluorescent dye, which was not absorbed on its own in the lymph. As much as 78% of the absorbed fluorescent SLN were detected in the thoracic duct, suggesting the lymphatic uptake as the major absorption mechanism for drugs encapsulated in SLN. Increased oral bioavailability has been reported for several drugs encapsulated in SLN. The size of the particles improves the adhesion to mucosal surfaces, providing closer and longer contact with the intestine wall. Moreover, the use of surfactants to stabilize the SLN may improve the absorption once the nanoparticles reach the gut. The great advantage is that the lipid structures are passed into the lymphatic circulation, avoiding the passage through the portal vein and the first-pass metabolism. This provides the opportunity of bioavailability enhancement upon oral administration of drugs with high first-pass effect. Lovastatin administered by means of SLN was shown to reach the systemic circulation in higher concentrations, as the first-pass loss is decreased (Suresh et al., 2007). The enhancement of oral bioavailability of drugs incorporated into SLN was proven for both lipophilic and hydrophilic drugs. Absorption rate of fenofibrate was higher and subsequent Area Under the Curve (AUC) in plasma was greater compared to micronized drug (Hanafy et al., 2007). Moreover, the interindividual variability of studied pharmacokinetic parameters decreased upon administration of this drug either in SLN or as a nanosuspension. Vinpocetine-loaded SLN were suggested as a potential formulation for this drug, as it can be protected from enzymatic degradation prior to absorption, thereby enhancing enhance its subsequent plasma concentration (Luo et al., 2006). Recently, a fivefold increase in oral bioavailability was reported for quercetin when incorporated into SLN (Li et al., 2009), and SLN proved a suitable solution for the administration of peptides and proteins via the oral route.

SLN Formulations: Pharmacokinetics and Biopharmaceutics

117

SLN are of particular interest for increasing the penetration and permeation of skin to several drugs, but may also be employed for peroral, buccal, nasal, rectal, intramuscular, and subcutaneous routes of administration. SLN may increase the amount absorbed by several orders of magnitude and they are intended to increase the absolute bioavailability of a drug for a given route of administration. Since SLN are pharmacologically inert, the mechanisms behind their role in increasing drug bioavailability may include increasing drug solubility (used as cosolvent), wettability by reducing surface tension, spreadability (increased contact surface area), miscibility with the epithelialium (mucus), and the skin (skin lipids), permeability of membranes by interacting with phospholipids (surfactant molecules), permeability of membranes by binding surface calcium and magnesium ions, as well as by changing the hydrostatic pressure to act on water-filled pores (sugars, glycols). These phenomena may be observed for both hydrophobic and hydrophilic drugs if the mechanism is based on a change a of membrane permeability.

6. Clinical Pharmacology The clinical pharmacological studies may supply pertinent information on the desired interaction between the drug molecules and receptors, whether a local effect, a general systemic concentration, or a specific target (tissue) are expected. Furthermore, they also contribute to the understanding of the desired or required drug concentration at the site of action, and its acceptable range and toxic limitation, for the development of a feasible LDDS.

6.1. Local effects and systemic distribution Classical examples of local applications are the topical, ocular, and other mucosal routes. The application of LDDS onto the skin is very well explored. Provided that the drug within the lipid nanoparticle matrix accumulates preferentially in the epidermis, the action of the drug will be limited to the local skin effects. Several examples of drugs for topical application on skin are listed in Table 6.1. Local effects are also expected for SLN for application in the eye, for example, NSAIDs applied in eye to prevent the inflammation after surgery on eye and to suppress the pain. Lipid nanoparticles containing the sodium salt of diclofenac (Attama et al., 2008) and ibuprofen (Li et al., 2008) were proposed as nonirritant and effective carriers. Cyclosporine A-loaded SLN were shown to penetrate corneal cells (Gokce et al., 2008).

118

Eliana B. Souto and Slavomira Doktorovova´

Table 6.1 Examples of drugs administered by LDDS for local distribution Drug

References

Alpha lipoic acid Ascorbyl palmitate Beta-carotene

Ruktanonchai et al. (2009) Teeranachaideekul et al. (2007a, 2008c) Hentschel et al. (2008), Triplett and Rathman (2009) Pojarova et al. (2006), Polovyanenko et al. (2008), Shahgaldian et al. (2008) Hu et al. (2006)—NLC Stecova et al. (2007) Xiang et al. (2007) Lu et al. (2008)

Calixarenes Clobetasol proprionate Cyproterone acetate Dexamethasone acetate Dexamethasone palmitate Insect repelent Ketoconazole Ketoprofen Minoxidil Naproxen Nile red Octadecylaminefluorescein isothiocynate Podophylotoxin Psoralen Retinoids

Iscan et al. (2006) Souto and Mu¨ller (2006), Souto et al. (2006) Puglia et al. (2008)—NLC Silva et al. (2009)—NLC Puglia et al. (2008)—NLC Kuchler et al. (2009), Teeranachaideekul et al. (2008b) Yuan et al. (2007)

Chen et al. (2006), Zhu et al. (2009) Fang et al. (2008) Castro et al. (2007), Jee et al. (2006), Liu et al. (2007b), Mandawgade and Patravale (2008), Pople and Singh (2006), Shah et al. (2007) Rhodamine B Wang and Wu (2006) Rhodamine 123 Chattopadhyay et al. (2008) Sunscreens Cengiz et al. (2006), Lee et al. (2007a), Scalia and Mezzena (2009), Xia et al. (2007) Tilmicisine Han et al. (2009) Triamcinolone acetonide Liu et al. (2008b) Ubidecarenone Teeranachaideekul et al. (2007b) Valdecoxib Joshi and Patravale (2006)

With respect to systemic distribution, the bloodstream must be reached usually by oral or parenteral administration. Miscellaneous drugs incorporated into LDDS reported recently are listed in Table 6.2. Systemic effects of encapsulated drugs within SLN may also be achieved by transdermal administration, upon creating a supersaturated system

SLN Formulations: Pharmacokinetics and Biopharmaceutics

Table 6.2 Examples of drugs administered by LDDS for systemic distribution Drug

References

Actarit Arthemeter Baclofen Bromocriptine Buprenorphine Calcitonin Camptothecin 10-Hydroxycamptothecin Carvedilol Chlorambucil Cholesteryl butyrate Cisplatin Ciprofloxacine Cyclosporine A Diazepam pDNA Docetaxel Doxorubicin Ferrulic acid Flavonoids Flurbiprofen Floxuridinyl diacetate Ibuprofen Insulin

Ye et al. (2008) Joshi et al. (2008) El Assawy et al. (2008) Esposito et al. (2008) Wang et al. (2009) Martins et al. (2009) Huang et al. (2008b) Liu et al. (2008c), Zhang et al. (2008b) Faisal et al. (2008), Sanjula et al. (2009) Sharma et al. (2009) Brioschi et al. (2007, 2008) Tian et al. (2008) Jain and Banerjee (2008) Mu¨ller et al. (2008) Abdelbary and Fahmy (2009) Asasutjarit et al. (2007) Xu et al. (2009) Serpe et al. (2006), Ying et al. (2008) Bondi et al. (2009b) Li et al. (2009), Wang et al. (2007) Bhaskar et al. (2009a), Han et al. (2008) Lian et al. (2008) Casadei et al. (2006), Paolicelli et al. (2008) Battaglia et al. (2007), Bi et al. (2009), Gallarate et al. (2008), Liu et al. (2008a,d), Sarmento et al. (2007) Hsu and Su (2008) Priano et al. (2007), Rezzani et al. (2009) Paliwal et al. (2008), Ruckmani et al. (2006b) Lu et al. (2006b) Bondi et al. (2009a) Bhaskar et al. (2009b), Kumar et al. (2007), Manjunath and Venkateswarlu (2006) Vivek et al. (2007) Dong et al. (2008), Lee et al. (2007b), Yuan et al. (2008) Yang et al. (2009) de Ven et al. (2009) Kuo and Chen (2009), Kuo and Su (2007) Zhang et al. (2007) Kuo and Chen (2009), Kuo and Su (2007) Zhang et al. (2007)

Magnetite Melatonin Methotrexate Mitoxantrone Nimesulide Nitrendipine Olanzapine Paclitaxel Praziquantel Saponins Saquinavir Silibinin Saquinavir Silibinin

119

120

Eliana B. Souto and Slavomira Doktorovova´

consisting of drug-loaded SLN dispersed in a drug-loaded medium, and applied under occlusion (Souto et al., 2007). The achieved bioavailability, expressed as the AUC, was reported similar to the AUC obtained after oral administration of nitrendipine (Bhaskar et al., 2009b), with the advantage of sustained and more prolonged residence of the drug in the blood. SLN were also shown to be suitable for transdermal delivery of flurbiprofen (Han et al., 2008) and melatonin (Priano et al., 2007), which showed systemic effects after encapsulation in SLN. Similar to the local effect, a desired drug concentration is often required only in a specific organ or tissue when designing a targeting approach. The LDDS may be administered intra- or extravascularly. However, the drug is either in the form of a prodrug which is activated at the target organ only, or somehow entrapped and released at the target organ. The rationale for organ targeting is to reach a therapeutic drug concentration at the receptor site with minimal systemic body burden. Examples of targeting SLN formulations include ferritin-surfaced SLN to carry 5-fluorouracil to breast cancer cells ( Jain et al., 2008), and monostearin-based SLN containing folic acid-conjugated stearic acid for paclitaxel delivery (Yuan et al., 2008). Other targeting applications include the delivery of gene therapeutic agents and diagnostics for brain tumors (Brioschi et al., 2007).

6.2. Dose size Considerations must be given to usual (noncontrolled release) dose size and dose rate. Drugs that act systemically and are therapeutically effective only in relatively large doses do not qualify a priori for modified release LDDS. A typical modified release system is essentially a multiple-dose unit. Even though the dose is not exactly the amount of a single dose multiplied by the number of dosing intervals, for usual doses equivalent to the duration of a modified release system, the total amount per unit is much higher than that of a single dose. Modified release LDDS imply a decrease of the usual dose used to achieve therapeutic concentrations. It was reported that mitoxantrone administered by means of SLN yielded the same tumor cell growth inhibition in half the dose of the free drug (Lu et al., 2006a). In another study, to assure a concentration within the therapeutic range, the lovastatin dose was lower if administered as SLN, because in this formulation it has higher bioavailability and is released over a prolonged period of time (Suresh et al., 2007).

SLN Formulations: Pharmacokinetics and Biopharmaceutics

121

6.3. Concentration fluctuations and therapeutic range Theoretically and desirably, a modified release LDDS, unless programed for a triggered mechanism, should release the drug by a zero-order process, which would result in a concentration versus time profile similar to that after i.v. constant rate infusion. However, zero-order release is rarely achieved, and if so, the resulting steady-state concentration will probably not be a straight line parallel to the abscissa, but more likely will be a soft wave due to many factors such as circadian rhythm, body activity, stress, and food intake. Nevertheless, in the design of a modified release SLN, it is necessary to select a desired (or required) target concentration or concentration range. The target concentration should usually be the same as for conventional dosage forms. The difference will be that the fluctuations of drug concentration in blood will be much smaller for the duration of the device as observed within each dosing interval of a conventional drug product. Ideally, candidates for modified release LDDS will be drugs of such wide therapeutic range that neither batch-to-batch variations in release rate nor individual physiological variations exceed the therapeutic range.

6.4. Toxicological concentration Toxicological concentration of a candidate for a modified release LDDS should be much higher than the therapeutic range. This is of utmost importance in case of accidental disruption or degradation of the LDDS, rupture or malfunction. Another consideration is the question of whether the drug is metabolized pharmacologically into highly potent structures or to toxic metabolites. Since SLN control only the release of the parent compound, but not its biotransformation, and since the rate of metabolism may vary widely among lipid materials and in the same material obtained by the same supplier, long-range predictions are difficult to make.

7. Concluding Remarks SLN play a prominent role as novel LDDS in pharmaceutical technology. The present chapter encompasses a thorough report on the use of SLN as promising LDDS to increase the bioavailability of several drugs. This chapter shortly outlined the pharmacokinetic and biopharmaceutical properties governing the modified release profile of SLN in drug delivery and summarized their applications for different routes of administration. The basics of pharmacokinetic parameters were explained in detail and examples of their use in the evaluation of SLN performance in controlling the release profile of the loaded drugs were given.

122

Eliana B. Souto and Slavomira Doktorovova´

REFERENCES Abdelbary, G., and Fahmy, R. H. (2009). Diazepam-loaded solid lipid nanoparticles: Design and characterization. AAPS PharmSciTech 10, 211–219. Asasutjarit, R., Lorenzen, S. I., Sirivichayakul, S., Ruxrungtham, K., Ruktanonchai, U., and Ritthidej, G. C. (2007). Effect of solid lipid nanoparticles formulation compositions on their size, zeta potential and potential for in vitro type pHIS-HIV-Hugag transfection. Pharm. Res. 24, 1098–1107. Attama, A. A., Reichl, S., and Mu¨ller-Goymann, C. C. (2008). Diclofenac sodium delivery to the eye: In vitro evaluation of novel solid lipid nanoparticle formulation using human cornea construct. Int. J. Pharm. 355, 307–313. Battaglia, L., Trotta, M., Gallarate, M., Carlotti, M. E., Zara, G. P., and Bargoni, A. (2007). Solid lipid nanoparticles formed by solvent-in-water emulsion-diffusion technique: Development and influence on insulin stability. J. Microencapsul. 24, 660–672. Bhaskar, K., Anbu, J., Ravichandiran, V., Venkateswarlu, V., and Rao, Y. M. (2009a). Lipid nanoparticles for transdermal delivery of flurbiprofen: Formulation, in vitro, ex vivo and in vivo studies. Lipids Health Dis. 8, 6. Bhaskar, K., Krishna Mohan, C., Lingam, M., Jagan Mohan, S., Venkateswarlu, V., Madhusudan Rao, Y., Anbu, J., and Ravichandran, V. (2009b). Development of SLN and NLC enriched hydrogels for transdermal delivery of nitrendipine: In vitro and in vivo characteristics. Drug Dev. Ind. Pharm. 35, 98–113. Bi, R., Shao, W., Wang, Q., and Zhang, N. (2009). Solid lipid nanoparticles as insulin inhalation carriers for enhanced pulmonary delivery. J. Biomed. Nanotechnol. 5, 84–92. Bondi, M. L., Azzolina, A., Craparo, E. F., Capuano, G., Lampiasi, N., Giammona, G., and Cervello, M. (2009a). Solid lipid nanoparticles containing nimesulide: Preparation, characterization and cytotoxicity studies. Curr. Nanosci. 5, 39–44. Bondi, M. L., Montana, G., Craparo, E. F., Picone, P., Capuano, G., Carlo, M. D., and Giammona, G. (2009b). Ferulic acid-loaded lipid nanostructures as drug delivery systems for Alzheimers disease: Preparation, characterization and cytotoxicity studies. Curr. Nanosci. 5, 26–32. Brioschi, A., Zenga, F., Zara, G. P., Gasco, M. R., Ducati, A., and Mauro, A. (2007). Solid lipid nanoparticles: Could they help to improve the efficacy of pharmacologic treatments for brain tumors? Neurol. Res. 29, 324–330. Brioschi, A., Zara, G. P., Calderoni, S., Gasco, M. R., and Mauro, A. (2008). Cholesteryl butyrate solid lipid nanoparticles as a butyric acid prodrug. Molecules 13, 230–254. Bunjes, H., and Unruh, T. (2007). Characterization of lipid nanoparticles by differential scanning calorimetry, X-ray and neutron scattering. Adv. Drug Deliv. Rev. 59, 379–402. Casadei, M. A., Cerreto, F., Cesa, S., Giannuzzo, M., Feeney, M., Marianecci, C., and Paolicelli, P. (2006). Solid lipid nanoparticles incorporated in dextran hydrogels: A new drug delivery system for oral formulations. Int. J. Pharm. 325, 140–146. Castro, G. A., Orefice, R. L., Vilela, J. M., Andrade, M. S., and Ferreira, L. A. (2007). Development of a new solid lipid nanoparticle formulation containing retinoic acid for topical treatment of acne. J. Microencapsul. 24, 395–407. Cengiz, E., Wissing, S. A., Mu¨ller, R. H., and Yazan, Y. (2006). Sunblocking efficiency of various TiO(2)-loaded solid lipid nanoparticle formulations(1). Int. J. Cosmet. Sci. 28, 371–378. Chattopadhyay, N., Zastre, J., Wong, H. L., Wu, X. Y., and Bendayan, R. (2008). Solid lipid nanoparticles enhance the delivery of the HIV protease inhibitor, atazanavir, by a human brain endothelial cell line. Pharm. Res. 25, 2262–2271. Chen, H. B., Chang, X. L., Du, D. R., Liu, W., Liu, J., Weng, T., Yang, Y. J., Xu, H. B., and Yang, X. L. (2006). Podophyllotoxin-loaded solid lipid nanoparticles for epidermal targeting. J. Control. Release 110, 296–306.

SLN Formulations: Pharmacokinetics and Biopharmaceutics

123

de Ven, H. V., Vermeersch, M., Shunmugaperumal, T., Vandervoort, J., Maes, L., and Ludwig, A. (2009). Solid lipid nanoparticle (SLN) formulations as a potential tool for the reduction of cytotoxicity of saponins. Pharmazie 64, 172–176. des Rieux, A., Fievez, V., Garinot, M., Schneider, Y.-J., and Pre´at, V. (2006). Nanoparticles as potential oral delivery systems of proteins and vaccines: A mechanistic approach. J. Control. Release 116, 1–27. Dianzani, C., Cavalli, R., Zara, G. P., Gallicchio, M., Lombardi, G., Gasco, M. R., Panzanelli, P., and Fantozzi, R. (2006). Cholesteryl butyrate solid lipid nanoparticles inhibit adhesion of human neutrophils to endothelial cells. Br. J. Pharmacol. 148, 648–656. Dong, X., Mattingly, C. A., Tseng, M., Cho, M., Adams, V. R., and Mumper, R. J. (2008). Development of new lipid-based paclitaxel nanoparticles using sequential simplex optimization. Eur. J. Pharm. Biopharm. 72, 9–17. El Assawy, N., Priano, L., Gasco, M., and Mauro, A. (2008). Baclofen-loaded solid lipid nanoparticles administration in rats: Pharmacokinetics and neurophysiological study. 18th Meeting of the European Neurological Society, Nice, France, p. P259. Esposito, E., Fantin, M., Marti, M., Drechsler, M., Paccamiccio, L., Mariani, P., Sivieri, E., Lain, F., Menegatti, E., Morari, M., and Cortesi, R. (2008). Solid lipid nanoparticles as delivery systems for bromocriptine. Pharm. Res. 25, 1521–1530. Faisal, S., Baboota, S., Ali, J., and Ahuja, A. (2008). Development and characterization of carvedilol-loaded solid lipid nanoparticles. J. Pharm. Pharmacol. 60, 150. Fang, J. Y., Fang, C. L., Liu, C. H., and Su, Y. H. (2008). Lipid nanoparticles as vehicles for topical psoralen delivery: Solid lipid nanoparticles (SLN) versus nanostructured lipid carriers (NLC). Eur. J. Pharm. Biopharm. 70, 633–640. Fundaro, A., Cavalli, R., Bargoni, A., Vighetto, D., Zara, G. P., and Gasco, M. R. (2000). Non-stealth and stealth solid lipid nanoparticles (SLN) carrying doxorubicin: Pharmacokinetics and tissue distribution after i.v. administration to rats. Pharmacol. Res. 42, 337–343. Gallarate, M., Trotta, M., Battaglia, L., and Chirio, D. (2008). Preparation of solid lipid nanoparticles from W/O/W emulsions: Preliminary studies on insulin encapsulation. J. Microencapsul. 1–9. Gokce, E. H., Sandri, G., Bonferoni, M. C., Rossi, S., Ferrari, F., Guneri, T., and Caramella, C. (2008). Cyclosporine A loaded SLNs: Evaluation of cellular uptake and corneal cytotoxicity. Int. J. Pharm. 364, 76–86. Han, F., Li, S. M., Yin, R., Shi, X. L., and Jai, Q. (2008). Investigation of nanostructured lipid carriers for transdermal delivery of flurbiprofen. Drug Dev. Ind. Pharm. 34, 453–458. Han, C., Qi, C. M., Zhao, B. K., Cao, J., Xie, S. Y., Wang, S. L., and Zhou, W. Z. (2009). Hydrogenated castor oil nanoparticles as carriers for the subcutaneous administration of tilmicosin: In vitro and in vivo studies. J. Vet. Pharmacol. Ther. 32, 116–123. Hanafy, A., Spahn-Langguth, H., Vergnault, G., Grenier, P., Tubic Grozdanis, M., Lenhardt, T., and Langguth, P. (2007). Pharmacokinetic evaluation of oral fenofibrate nanosuspensions and SLN in comparison to conventional suspensions of micronized drug. Adv. Drug Deliv. Rev. 59, 419–426. Hentschel, A., Gramdorf, S., Mu¨ller, R. H., and Kurz, T. (2008). Beta-carotene-loaded nanostructured lipid carriers. J. Food Sci. 73, N1–N6. Hsu, M. H., and Su, Y. C. (2008). Iron-oxide embedded solid lipid nanoparticles for magnetically controlled heating and drug delivery. Biomed. Microdevices 10, 785–793. Hu, F.-Q., Jiang, S.-P., Du, Y.-Z., Yuan, H., Ye, Y.-Q., and Zeng, S. (2006). Preparation and characteristics of monostearin nanostructured lipid carriers. Int. J. Pharm. 314, 83–89. Huang, G., Zhang, N., Bi, X., and Dou, M. (2008a). Solid lipid nanoparticles of temozolomide: Potential reduction of cardial and nephric toxicity. Int. J. Pharm. 355, 314–320.

124

Eliana B. Souto and Slavomira Doktorovova´

Huang, Z.-r., Hua, S.-c., Yang, Y.-l., and Fang, J.-y. (2008b). Development and evaluation of lipid nanoparticles for camptothecin delivery: A comparison of solid lipid nanoparticles, nanostructured lipid carriers, and lipid emulsion. Acta Pharmacol. Sin. 29, 1094–1102. Iscan, Y., Hekimoglu, S., Sargon, M. F., and Hincal, A. A. (2006). DEET-loaded solid lipid particles for skin delivery: In vitro release and skin permeation characteristics in different vehicles. J. Microencapsul. 23, 315–327. Jain, D., and Banerjee, R. (2008). Comparison of ciprofloxacin hydrochloride-loaded protein, lipid, and chitosan nanoparticles for drug delivery. J. Biomed. Mater. Res. Part B Appl. Biomater. 86B, 105–112. Jain, S. K., Chaurasiya, A., Gupta, Y., Jain, A., Dagur, P., Joshi, B., and Katoch, V. M. (2008). Development and characterization of 5-FU bearing ferritin appended solid lipid nanoparticles for tumour targeting. J. Microencapsul. 25, 289–297. Jee, J. P., Lim, S. J., Park, J. S., and Kim, C. K. (2006). Stabilization of all-trans retinol by loading lipophilic antioxidants in solid lipid nanoparticles. Eur. J. Pharm. Biopharm. 63, 134–139. Johnsson, M., Lam, Y., Barauskas, J., and Tiberg, F. (2005). Aqueous phase behavior and dispersed nanoparticles of diglycerol monooleate/glycerol dioleate mixtures. Langmuir 21, 5159–5165. Joshi, M., and Patravale, V. (2006). Formulation and evaluation of nanostructured lipid carrier (NLC)-based gel of valdecoxib. Drug Dev. Ind. Pharm. 32, 911–918. Joshi, M., and Patravale, V. (2008). Nanostructured lipid carrier (NLC) based gel of celecoxib. Int. J. Pharm. 346, 124–132. Joshi, M., Pathak, S., Sharma, S., and Patravale, V. (2008). Design and in vivo pharmacodynamic evaluation of nanostructured lipid carriers for parenteral delivery of artemether: Nanoject. Int. J. Pharm. 364, 119–126. Kuchler, S., Radowski, M. R., Blaschke, T., Dathe, M., Plendl, J., Haag, R., Schafer-Korting, M., and Kramer, K. D. (2009). Nanoparticles for skin penetration enhancement—A comparison of a dendritic core-multishell-nanotransporter and solid lipid nanoparticles. Eur. J. Pharm. Biopharm. 71, 243–250. Kumar, V. V., Chandrasekar, D., Ramakrishna, S., Kishan, V., Rao, Y. M., and Diwan, P. V. (2007). Development and evaluation of nitrendipine loaded solid lipid nanoparticles: Influence of wax and glyceride lipids on plasma pharmacokinetics. Int. J. Pharm. 335, 167–175. Kuntsche, J., Koch, M. H., Drechsler, M., and Bunjes, H. (2005). Crystallization behavior of supercooled smectic cholesteryl myristate nanoparticles containing phospholipids as stabilizers. Colloids Surf. B Biointerfaces 44, 25–35. Kuo, Y. C., and Chen, H. H. (2009). Entrapment and release of saquinavir using novel cationic solid lipid nanoparticles. Int. J. Pharm. 365, 206–213. Kuo, Y. C., and Su, F. L. (2007). Transport of stavudine, delavirdine, and saquinavir across the blood-brain barrier by polybutylcyanoacrylate, methylmethacrylate-sulfopropylmethacrylate, and solid lipid nanoparticles. Int. J. Pharm. 340, 143–152. Lee, G. S., Lee, D. H., Kang, K. C., Lee, C. I., Pyo, H. B., and Choi, T. B. (2007a). Preparation and characterization of bis-ethylhexyloxyphenolmethoxyphenyltriazine (BEMT) loaded solid lipid nano-particles (SLN). J. Ind. Eng. Chem. 13, 1180–1187. Lee, M. K., Lim, S. J., and Kim, C. K. (2007b). Preparation, characterization and in vitro cytotoxicity of paclitaxel-loaded sterically stabilized solid lipid nanoparticles. Biomaterials 28, 2137–2146. Li, X., Nie, S. F., Kong, J., Li, N., Ju, C. Y., and Pan, W. S. (2008). A controlled-release ocular delivery system for ibuprofen based on nanostructured lipid carriers. Int. J. Pharm. 363, 177–182. Li, H. L., Zhao, X. B., Ma, Y. K., Zhai, G. X., Li, L. B., and Lou, H. X. (2009). Enhancement of gastrointestinal absorption of quercetin by solid lipid nanoparticles. J. Control. Release 133, 238–244.

SLN Formulations: Pharmacokinetics and Biopharmaceutics

125

Lian, J., Zhang, S., Wang, J., Fang, K., Zhang, Y., and Hao, Y. (2008). Novel galactosylated SLN for hepatocyte-selective targeting of floxuridinyl diacetate. J. Drug Target. 16, 250–256. Liu, J., Hu, W., Chen, H. B., Ni, Q., Xu, H. B., and Yang, X. L. (2007a). Isotretinoinloaded solid lipid nanoparticles with skin targeting for topical delivery. Int. J. Pharm. 328, 191–195. Liu, J., Hu, W., Chen, H., Ni, Q., Xu, H., and Yang, X. (2007b). Isotretinoin-loaded solid lipid nanoparticles with skin targeting for topical delivery. Int. J. Pharm. 328, 191–195. Liu, H., Gong, T., Fu, H. L., Wang, C. G., Wang, X. L., Chen, Q., Zhang, Q., He, Q., and Zhang, Z. R. (2008a). Solid lipid nanoparticles for pulmonary delivery of insulin. Int. J. Pharm. 356, 333–344. Liu, J., Gong, T., Fu, H., Wang, C., Wang, X., Chen, Q., Zhang, Q., He, Q., and Zhang, Z. (2008b). Solid lipid nanoparticles for pulmonary delivery of insulin. Int. J. Pharm. 356, 333–344. Liu, K., Sun, J., Wang, Y. J., He, Y., Gao, K., and He, Z. G. (2008c). Preparation and characterization of 10-hydroxycamptothecin loaded nanostructured lipid carriers. Drug Dev. Ind. Pharm. 34, 465–471. Liu, W., Hu, M., Xue, C., Xu, H., and Yang, X. (2008d). Investigation of the carbopol gel of solid lipid nanoparticles for the transdermal iontophoretic delivery of triamcinolone acetonide acetate. Int. J. Pharm. 364, 135–141. Lombardi Borgia, S., Regehly, M., Sivaramakrishnan, R., Mehnert, W., Korting, H. C., Danker, K., Ro¨der, B., Kramer, K. D., and Scha¨fer-Korting, M. (2005). Lipid nanoparticles for skin penetration enhancement–correlation to drug localization within the particle matrix as determined by fluorescence and parelectric spectroscopy. J. Control. Release 110, 151–163. Lu, B., Xiong, S. B., Yang, H., Yin, X. D., and Chao, R. B. (2006a). Solid lipid nanoparticles of mitoxantrone for local injection against breast cancer and its lymph node metastases. Eur. J. Pharm. Sci. 28, 86–95. Lu, B., Xiong, S. B., Yang, H., Yin, X. D., and Chao, R. B. (2006b). Solid lipid nanoparticles of mitoxantrone for local injection against breast cancer and its lymph node metastases. Eur. J. Pharm. Sci. 28, 86–95. Lu, X., Howard, M. D., Mazik, M., Eldridge, J., Rinehart, J. J., Jay, M., and Leggas, M. (2008). Nanoparticles containing anti-inflammatory agents as chemotherapy adjuvants: Optimization and in vitro characterization. AAPS J. 10, 133–140. Luo, Y., Chen, D. W., Ren, L. X., Zhao, X. L., and Qin, J. (2006). Solid lipid nanoparticles for enhancing vinpocetine’s oral bioavailability. J. Control. Release. 114, 53–59. Lv, Q., Yu, A., Xi, Y., Li, H., Song, Z., Cui, J., Cao, F., and Zhai, G. (2009). Development and evaluation of penciclovir-loaded solid lipid nanoparticles for topical delivery. Int. J. Pharm. 372, 191–198. Mandawgade, S. D., and Patravale, V. B. (2008). Development of SLNs from natural lipids: Application to topical delivery of tretinoin. Int. J. Pharm. 363, 132–138. Manjunath, K., and Venkateswarlu, V. (2005). Pharmacokinetics, tissue distribution and bioavailability of clozapine solid lipid nanoparticles after intravenous and intraduodenal administration. J. Control. Release 107, 215–228. Manjunath, K., and Venkateswarlu, V. (2006). Pharmacokinetics, tissue distribution and bioavailability of nitrendipine solid lipid nanoparticles after intravenous and intraduodenal administration. J. Drug Target. 14, 632–645. Martins, S., Sarmento, B., Souto, E. B., and Ferreira, D. C. (2007). Insulin-loaded alginate microspheres for oral delivery—Effect of polysaccharide reinforcement on physicochemical properties and release profile. Carbohydr. Polym. 69, 725–731. Martins, S., Silva, A. C., Ferreira, D. C., and Souto, E. B. (2009). Improving oral absorption of samon calcitonin by trimyristin lipid nanoparticles. J. Biomed. Nanotechnol. 5, 76–83.

126

Eliana B. Souto and Slavomira Doktorovova´

Mu¨ller, R. H., Ru¨hl, D., and Runge, S. A. (1996). Biodegradation of solid lipid nanoparticles as a function of lipase incubation time. Int. J. Pharm. 144, 115–121. Mu¨ller, R. H., Radtke, M., and Wissing, S. A. (2002). Solid lipid nanoparticles (SLN) and nanostructured lipid carriers (NLC) in cosmetic and dermatological preparations. Adv. Drug Deliv. Rev. 54, S131–S155. Mu¨ller, R. H., Runge, S., Ravelli, V., Mehnert, W., Thu¨nemann, A. F., and Souto, E. B. (2006). Oral bioavailability of cyclosporine: Solid lipid nanoparticles (SLNÒ ) versus drug nanocrystals. Int. J. Pharm. 317, 82–89. Mu¨ller, R. H., Petersen, R. D., Hommoss, A., and Pardeike, J. (2007). Nanostructured lipid carriers (NLC) in cosmetic dermal products. Adv. Drug Deliv. Rev. 59, 522–530. Mu¨ller, R. H., Runge, S. A., Ravelli, V., Thunemann, A. F., Mehnert, W., and Souto, E. B. (2008). Cyclosporine-loaded solid lipid nanoparticles (SLN): Drug-lipid physicochemical interactions and characterization of drug incorporation. Eur. J. Pharm. Biopharm. 68, 535–544. Olbrich, C., Kayser, O., and Mu¨ller, R. H. (2002). Lipase degradation of Dynasan 114 and 116 solid lipid nanoparticles (SLN)—Effect of surfactants, storage time and crystallinity. Int. J. Pharm. 237, 119–128. Paliwal, R., Rai, S., Vaidya, B., Khatri, K., Goyal, A. K., Mishra, N., Mehta, A., and Vyas, S. P. (2008). Effect of lipid core material on characteristics of solid lipid nanoparticles designed for oral lymphatic delivery. Nanomedicine 5, 184–191. Paolicelli, P., Cerreto, F., Cesa, S., Feeney, M., Corrente, F., Marianecci, C., and Casadei, M. A. (2008). Influence of the formulation components on the properties of the system SLN-dextran hydrogel for the modified release of drugs. J. Microencapsul. 1–10. Pardeike, J., Hommoss, A., and Mu¨ller, R. H. (2008). Lipid nanoparticles (SLN, NLC) in cosmetic and pharmaceutical dermal products. Int. J. Pharm. 366, 170–184. Pojarova, M., Ananchenko, G. S., Udachin, K. A., Daroszewska, M., Perret, F., Coleman, A. W., and Ripmeester, J. A. (2006). Solid lipid nanoparticles of p-hexanoyl calix[4]arene as a controlling agent in the photochemistry of a sunscreen blocker. Chem. Mater. 18, 5817–5819. Polovyanenko, D. N., Bagryanskaya, E. G., Schnegg, A., Mobius, K., Coleman, A. W., Ananchenko, G. S., Udachin, K. A., and Ripmeester, J. A. (2008). Inclusion of 4-methoxy-2, 2, 6, 6-tetramethylpiperidine-N-oxyl in a calixarene nanocapsule in the solid state. Phys. Chem. Chem. Phys. 10, 5299–5307. Pople, P. V., and Singh, K. K. (2006). Development and evaluation of topical formulation containing solid lipid nanoparticles of vitamin A. AAPS PharmSciTech 7, Article 91. Priano, L., Esposti, D., Esposti, R., Castagna, G., De Medici, C., Fraschini, F., Gasco, M. R., and Mauro, A. (2007). Solid lipid nanoparticles incorporating melatonin as new model for sustained oral and transdermal delivery systems. J. Nanosci. Nanotechnol. 7, 3596–3601. Puglia, C., Blasi, P., Rizza, L., Schoubben, A., Bonina, F., Rossi, C., and Ricci, M. (2008). Lipid nanoparticles for prolonged topical delivery: An in vitro and in vivo investigation. Int. J. Pharm. 357, 295–304. Reddy, L. H., Adhikari, J. S., Dwarakanath, B. S. R., Sharma, R. K., and Murthy, R. R. (2006). Tumoricidal effects of etoposide incorporated into solid lipid nanoparticles after intraperitoneal administration in Dalton’s lymphoma bearing mice. AAPS J. 8, E254–E262. Rezzani, R., Rodella, L. F., Fraschini, F., Gasco, M. R., Demartini, G., Musicanti, C., and Reiter, R. J. (2009). Melatonin delivery in solid lipid nanoparticles: Prevention of cyclosporine A induced cardiac damage. J. Pineal Res. 46, 255–261. Ruckmani, K., Sivakumar, M., and Ganeshkumar, P. A. (2006a). Methotrexate loaded solid lipid nanoparticles (SLN) for effective treatment of carcinoma. J. Nanosci. Nanotechnol. 6, 2991–2995.

SLN Formulations: Pharmacokinetics and Biopharmaceutics

127

Ruckmani, K., Sivakumar, M., and Ganeshkumar, P. A. (2006b). Methotrexate loaded solid lipid nanoparticles (SLN) for effective treatment of carcinoma. International Conference on Nanotechnology in Advanced Drug Delivery, Nagar, India, pp. 2991–2995. Ruktanonchai, U., Bejrapha, P., Sakulkhu, U., Opanasopit, P., Bunyapraphatsara, N., Junyaprasert, V., and Puttipipatkhachorn, S. (2009). Physicochemical characteristics, cytotoxicity, and antioxidant activity of three lipid nanoparticulate formulations of alpha-lipoic acid. AAPS PharmSciTech. 10, 227–234. Sanjula, B., Shah, F. M., Javed, A., and Alka, A. (2009). Effect of poloxamer 188 on lymphatic uptake of carvedilol-loaded solid lipid nanoparticles for bioavailability enhancement. J. Drug Target. 17, 249–256. Sarmento, B., Martins, S., Ferreira, D., and Souto, E. B. (2007). Oral insulin delivery by means of solid lipid nanoparticles. Int. J. Nanomedicine 2, 743–749. Scalia, S., and Mezzena, M. (2009). Co-loading of a photostabilizer with the sunscreen agent, butyl methoxydibenzoylmethane in solid lipid microparticles. Drug Dev.Ind. Pharm. 35, 192–198. Serpe, L., Guido, M., Canaparo, R., Muntoni, E., Cavalli, R., Panzanelli, P., Della Pepal, C., Bargoni, A., Mauro, A., Gasco, M. R., Eandi, M., and Zara, G. P. (2006). Intracellular accumulation and cytotoxicity of doxorubicin with different pharmaceutical formulations in human cancer cell lines. J. Nanosci. Nanotechnol. 6, 3062–3069. Shah, K. A., Date, A. A., Joshi, M. D., and Patravale, V. B. (2007). Solid lipid nanoparticles (SLN) of tretinoin: Potential in topical delivery. Int. J. Pharm. 345, 163–171. Shahgaldian, P., Sciotti, M. A., and Pieles, U. (2008). Amino-substituted amphiphilic calixarenes: Self-assembly and interactions with DNA. Langmuir 24, 8522–8526. Sharma, P., Ganta, S., Denny, W. A., and Garg, S. (2009). Formulation and pharmacokinetics of lipid nanoparticles of a chemically sensitive nitrogen mustard derivative: Chlorambucil. Int. J. Pharm. 367, 187–194. Silva, A. C., Santos, D., Ferreira, D. C., and Souto, E. B. (2009). Minoxidil-loaded nanostructured lipid carriers (NLC): Characterization and rheological behaviour of topical formulations. Pharmazie 64, 177–182. Souto, E. B., and Mu¨ller, R. H. (2006). The use of SLN (R) and NLC (R) as topical particulate carriers for imidazole antifungal agents. Pharmazie 61, 431–437. Souto, E. B., and Mu¨ller, R. H. (2007). Lipid nanoparticles (SLN and NLC) for drug delivery. In ‘‘Nanoparticles for Pharmaceutical Applications,’’ (A. J. Domb, Y. Tabata, M. N. V. Ravi Kumar, and S. Farber, eds.), American Scientific Publishers, Los Angeles, California, USA, pp103–122. Souto, E. B., and Mu¨ller, R. H. (2008). Cosmetic features and applications of lipid nanoparticles (SLN, NLC). Int. J. Cosmet. Sci. 30, 157–165. Souto, E. B., Mehnert, W., and Mu¨ller, R. H. (2006). Polymorphic behaviour of Compritol888 ATO as bulk lipid and as SLN and NLC. J. Microencapsul. 23, 417–433. Souto, E. B., Almeida, A. J., and Mu¨ller, R. H. (2007). Lipid nanoparticles (SLNÒ , NLCÒ ) for cutaneous drug delivery: Structure, protection and skin effects. J. Biomed. Nanotechnol. 3, 317–331. Stecova, J., Mehnert, W., Blaschke, T., Kleuser, B., Sivaramakrishnan, R., Zouboulis, C. C., Seltmann, H., Korting, H. C., Kramer, K. D., and SchaferKorting, M. (2007). Cyproterone acetate loading to lipid nanoparticles for topical acne treatment: Particle characterisation and skin uptake. Pharm. Res. 24, 991–1000. Suresh, G., Manjunath, K., Venkateswarlu, V., and Satyanarayana, V. (2007). Preparation, characterization, and in vitro and in vivo evaluation of lovastatin solid lipid nanoparticles. AAPS PharmSciTech 8, 24. Tausch, C., Konstantiniuk, P., Kugler, F., Reitsamer, R., Roka, S., Postlberger, S., and Haid, A. (2008). Sentinel lymph node biopsy after preoperative chemotherapy for breast

128

Eliana B. Souto and Slavomira Doktorovova´

cancer: Findings from the Austrian Sentinel Node Study Group. Ann. Surg. Oncol. 15, 3378–3383. Teeranachaideekul, V., Mu¨ller, R. H., and Junyaprasert, V. B. (2007a). Encapsulation of ascorbyl palmitate in nanostructured lipid carriers (NLC)—Effects of formulation parameters on physicochemical stability. Int. J. Pharm. 340, 198–206. Teeranachaideekul, V., Souto, E. B., Junyaprasert, V. B., and Mu¨ller, R. H. (2007b). Cetyl palmitate-based NLC for topical delivery of Coenzyme Q(10)—Development, physicochemical characterization and in vitro release studies. Eur. J. Pharm. Biopharm. 67, 141–148. Teeranachaideekul, V., Boonme, P., Souto, E. B., Mu¨ller, R. H., and Junyaprasert, V. B. (2008a). Influence of oil content on physicochemical properties and skin distribution of Nile red-loaded NLC. J. Control. Release 128, 134–141. Teeranachaideekul, V., Souto, E. B., Mu¨ller, R. H., and Junyaprasert, V. B. (2008b). Physicochemical characterization and in vitro release studies of ascorbyl palmitate-loaded semi-solid nanostructured lipid carriers (NLC gels). J. Microencapsul. 25, 111–120. Teeranachaideekul, V., Boonme, P., Souto, E. B., Mu¨ller, R. H., and Junyaprasert, V. B. (2008c). Influence of oil content on physicochemical properties and skin distribution of Nile red-loaded NLC. J. Control. Release 128, 134–141. Tian, J., Pang, X., Yu, K., Liu, L., and Zhou, J. (2008). Preparation, characterization and in vivo distribution of solid lipid nanoparticles loaded with cisplatin. Pharmazie 63, 593–597. Triplett, M. D., and Rathman, J. F. (2009). Optimization of beta-carotene loaded solid lipid nanoparticles preparation using a high shear homogenization technique. J. Nanopart. Res. 11, 601–614. Videira, M. A., Botelho, M. F., Santos, A. C., Gouveia, L. F., Pedroso De Lima, J. J., and Almeida, A. J. (2002). Lymphatic uptake of pulmonary delivered radiolabelled solid lipid nanoparticles. J. Drug Target. 10, 607–613. Vivek, K., Reddy, H., and Murthy, R. S. (2007). Investigations of the effect of the lipid matrix on drug entrapment, in vitro release, and physical stability of olanzapine-loaded solid lipid nanoparticles. AAPS PharmSciTech 8, E83. Wang, Y., and Wu, W. (2006). In situ evading of phagocytic uptake of stealth solid lipid nanoparticles by mouse peritoneal macrophages. Drug Deliv. 13, 189–192. Wang, D., Zhao, P., Cuia, F., and Li, X. (2007). Preparation and characterization of solid lipid nanoparticles loaded with total flavones of Hippophae rhamnoides (TFH). PDA J. Pharm. Sci. Technol. 61, 110–120. Wang, J. J., Liu, K. S., Sung, K. C., Tsai, C. Y., and Fang, J. Y. (2009). Skin permeation of buprenorphine and its ester prodrugs from lipid nanoparticles: Lipid emulsion, nanostructured lipid carriers and solid lipid nanoparticles. J. Microencapsul. 1–14. Wissing, S., and Mu¨ller, R. (2002). The influence of the crystallinity of lipid nanoparticles on their occlusive properties. Int. J. Pharm. 242, 377–379. Xia, Q., Saupe, A., Mu¨ller, R. H., and Souto, E. B. (2007). Nanostructured lipid carriers as novel carrier for sunscreen formulations. Int. J. Cosmet. Sci. 29, 473–482. Xiang, Q. Y., Wang, M. T., Chen, F., Gong, T., Jian, Y. L., Zhang, Z. R., and Huang, Y. (2007). Lung-targeting delivery of dexamethasone acetate loaded solid lipid nanoparticles. Arch. Pharm. Res. 30, 519–525. Xu, Z., Chen, L., Gu, W., Gao, Y., Lin, L., Zhang, Z., Xi, Y., and Li, Y. (2009). The performance of docetaxel-loaded solid lipid nanoparticles targeted to hepatocellular carcinoma. Biomaterials 30, 226–232. Yang, S. C., Lu, L. F., Cai, Y., Zhu, J. B., Liang, B. W., and Yang, C. Z. (1999). Body distribution in mice of intravenously injected camptothecin solid lipid nanoparticles and targeting effect on brain. J. Control. Release 59, 299–307.

SLN Formulations: Pharmacokinetics and Biopharmaceutics

129

Yang, L., Geng, Y., Li, H., Zhang, Y., You, J., and Chang, Y. (2009). Enhancement the oral bioavailability of praziquantel by incorporation into solid lipid nanoparticles. Pharmazie 64, 86–89. Ye, J., Wang, Q., Zhou, X., and Zhang, N. (2008). Injectable actarit-loaded solid lipid nanoparticles as passive targeting therapeutic agents for rheumatoid arthritis. Int. J. Pharm. 352, 273–279. Ying, X. Y., Du, Y. Z., Chen, W. W., Yuan, H., and Hu, F. Q. (2008). Preparation and characterization of modified lipid nanoparticles for doxorubicin controlled release. Pharmazie 63, 878–882. Yuan, H., Chen, J., Du, Y.-Z., Hu, F.-Q., Zeng, S., and Zhao, H.-L. (2007). Studies on oral absorption of stearic acid SLN by a novel fluorometric method. Colloids Surf. B Biointerfaces 58, 157–164. Yuan, H., Miao, J., Du, Y. Z., You, J., Hu, F. Q., and Zeng, S. (2008). Cellular uptake of solid lipid nanoparticles and cytotoxicity of encapsulated paclitaxel in A549 cancer cells. Int. J. Pharm. 348, 137–145. Zara, G. P., Bargoni, A., Cavalli, R., Fundaro, A., Vighetto, D., and Gasco, M. R. (2002a). Pharmacokinetics and tissue distribution of idarubicin-loaded solid lipid nanoparticles after duodenal administration to rats. J. Pharm. Sci. 91, 1324–1333. Zara, G. P., Cavalli, R., Bargoni, A., Fundaro, A., Vighetto, D., and Gasco, M. R. (2002b). Intravenous administration to rabbits of non-stealth and stealth doxorubicin-loaded solid lipid nanoparticles at increasing concentrations of stealth agent: Pharmacokinetics and distribution of doxorubicin in brain and other tissues. J. Drug Target. 10, 327–335. Zhang, J. Q., Liu, J., Li, Y. L., and Jasti, B. R. (2007). Preparation and characterization of solid lipid nanoparticles containing silibinin. Drug Deliv. 14, 381–387. Zhang, W., Liu, J., Li, S., Chen, M., and Liu, H. (2008a). Preparation and evaluation of stealth Tashinone IIA-loaded solid lipid nanoparticles: Influence of Poloxamer 188 coating on phagocytic uptake. J. Microencapsul. 25, 203–209. Zhang, X., Pan, W., Gan, L., Zhu, C., Gan, Y., and Nie, S. (2008b). Preparation of a dispersible PEGylate nanostructured lipid carriers (NLC) loaded with 10-hydroxycamptothecin by spray-drying. Chem. Pharm. Bull. (Tokyo) 56, 1645–1650. Zhu, R. R., Qin, L. L., Wang, M., Wu, S. M., Wang, S. L., Zhang, R., Liu, Z. X., Sun, X. Y., and Yao, S. D. (2009). Preparation, characterization, and anti-tumor property of podophyllotoxin-loaded solid lipid nanoparticles. Nanotechnology 20, 55702.

This page intentionally left blank

C H A P T E R

S E V E N

Preparation of Complexes of Liposomes with Gold Nanoparticles Chie Kojima,* Yusuke Hirano,† and Kenji Kono† Contents 132 134 134 136 137 137

1. 2. 3. 4. 5. 6. 7.

Introduction Preparation of Complexes of EYPC Liposomes with Au NPs Time-Dependent SPR of the Complexes TEM Analysis of the Complexes DLS Analysis of the Complexes Calcein Release from the Complexes Estimation of Numbers of the Au NP and the Liposome in the Complexes 8. Optimization of Lipid Components of the Complexes 9. Concluding Remarks Acknowledgment References

139 140 142 143 144

Abstract Liposomes have been widely used as drug carriers. Visible liposomes have recently become more attractive as drug carriers in personalized medicine. Gold nanoparticles (Au NPs) have unique size- and shape-dependent properties based on their surface plasmon resonance. They can be visualized by computed tomography (CT) and laser optoacoustic imaging. In addition, their photothermogenic properties are useful for photothermal therapy and photoresponsive drug release from liposomes. Therefore, complexation of liposomes with Au NPs is of considerable interest. There are three types of complex: Liposomes containing Au NPs in the inner phase, liposomes with Au NPs at the lipid membrane, and liposomes modified with Au NPs on the surface. This chapter focuses on the preparation and characterization of the third type of complex that is prepared by direct mixing of a Au NP dispersion with a liposome suspension.

* {

Nanoscience and Nanotechnology Research Center, Research Institutes for the Twenty First Century, Osaka Prefecture University, Osaka, Japan Graduate School of Engineering, Osaka Prefecture University, Osaka, Japan

Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64007-6

#

2009 Elsevier Inc. All rights reserved.

131

132

Chie Kojima et al.

1. Introduction Drug delivery systems (DDS) are attractive for chemotherapy, because they reduce severe side effects and facilitate effective drug action. Drug carriers are essential for the success of DDS. There are many types of drug carriers available, such as micelles, polymers, virus particles, proteins, and liposomes. Liposomes have classically been used as drug carriers and some, such as Doxil, have already gained Food and Drug Administration (FDA) approval (Immordino et al., 2006). Liposomes are vesicle structures composed of phospholipids with a hydrophobic tail and a hydrophilic head. Due to their amphiphatic character, liposomes can encapsulate water-soluble drug molecules in the inner phase and lipid-soluble ones in the hydrophobic membrane. Improvements to the next generation of liposome drug carriers include controllable release and visibility of the liposome, obtained by addition of functional molecules. Functional molecules that have been added to liposomes include thermosensitive, pH-sensitive, and visible molecules (Al-Jamal and Kostarelos, 2007; Chilkoti et al., 2002; Immordino et al., 2006; Kono, 2001; Kono and Arshady, 2006). Gold nanoparticles (Au NPs) could also be of use as functional molecules because of their interesting shape- and size-dependent physical and chemical properties (Burda et al., 2005; Daniel and Astruc, 2004). Due to their surface plasmon resonance (SPR), Au NPs strongly absorb visible light (Link and El-Sayed, 1999). In addition, they convert this light energy to heat energy (Link and El-Sayed, 2000). Consequently, Au NPs have been considered for photothermal therapy, imaging, and photosensitive drug release (Govorov and Richardson, 2007; Jain et al., 2007; Kim et al., 2007; Pissuwan et al., 2006). Complexes of liposomes with Au NPs are attractive because they can act as both stimuli-responsive and visible drug carriers (Hong et al., 1983; Kojima et al., 2008; Li et al., 2004; Paasonen et al., 2007; Volodkin et al., 2009; Wu et al., 2008). These photochemical properties are only expressed by Au NPs ranging from approximately 2 to 100 nm in diameter (Burda et al., 2005; Daniel and Astruc, 2004). When Au NPs aggregate, they lose their unique photochemical properties. Therefore, the preparation of the complexes also has to be performed without the Au NPs aggregating. Three types of Au NP–liposome complexes exist (Fig. 7.1). The first of these contains Au NPs in the inner phase of the liposome. This type of complex can be prepared by reducing Au ions in the presence of a reductant (Wu et al., 2008), and has been used to investigate in vivo liposome distribution (Hong et al., 1983). In DDS applications, this reduction may be detrimental to drug activity. In the second type of complex, Au NPs are present in the lipid membrane (Paasonen et al., 2007; Park et al., 2006).

133

Complex of Liposomes with Gold Nanoparticles

A

B

C

Figure 7.1 A schematic image of the complexes formed between liposomes and Au NPs. (A) A liposome encapsulating Au NPs, (B) a liposome loaded with Au NPs in the membrane, and (C) a liposome modified with Au NPs at the surface (from Kojima et al., 2008).

However, as the thickness of the lipid bilayer is only about 4 nm, a limited number of Au NPs can be incorporated using this strategy. The third complex type is a liposome modified with Au NPs on its surface. This type of complex is simply prepared by mixing an Au NP dispersion with a liposome suspension (Kojima et al., 2008; Volodkin et al., 2009). In this chapter, the third preparation method is described in detail (Kojima et al., 2008), and the influence of lipid components on the liposome are described. The lipids include: egg yolk phosphatidylcholine (EYPC), distearyldimethylammonium bromide (DDAB), distearoyl-sn-glycero-3phosphoethanolamine-N-[methyl(poly(ethylene glycol))] (PEG-PE), and stearylmercaptan (C18-SH). The complexes are characterized by a variety of techniques. UV–Vis spectrometry is used to investigate SPR and its influence on the stability of Au NPs. The size and morphology of the complexes are analyzed by dynamic light scattering (DLS) and transmission electron microscopy (TEM). The release of encapsulated molecules from the liposome is examined by using the fluorescent dye calcein.

134

Chie Kojima et al.

2. Preparation of Complexes of EYPC Liposomes with Au NPs Au NPs with a diameter of 13 nm are prepared by reducing Au ions with citric acid, according to a previous report (Grabar et al., 1995). A 5-min reflux is carried out with 50 ml of 1 mM HAuCl4 (Wako Pure Chemical Industries Ltd.). Five milliliters of 38.8 mM sodium citrate (Kishida Chemical) are added with vigorous stirring. The solution changes color from yellow to dark wine red. After cooling, an Au ion dispersion with a concentration of 0.91 mM is obtained by filtration. EYPC can be obtained from a variety of sources (Avanti Polar Lipids, Sigma, etc.). For our studies, it was kindly provided by NOF Corp. (Tokyo, Japan). A chloroform solution of EYPC (10 mg/ml, 500 ml) is evaporated to remove the solvent. The obtained thin lipid membrane (5.0 mg, 6.25 mmol) is further dried under vacuum for at least 2 h, and then dispersed in 0.5 ml of phosphate-buffered saline (PBS; 20 mM Na2HPO3–NaH2PO3, 150 mM NaCl, pH 7) with sonication in a bath type sonicator for 3 min. The liposome suspension is freeze-thawed four times. The obtained liposome suspension is extruded through a polycarbonate membrane with a pore diameter of 200 nm (Kojima et al., 2008; Kono et al., 1999). Lipid concentrations are estimated by Phospholipids C-Test Wako (Wako Pure Chemical Industries Ltd.) according to the manufacturer’s instructions. Sample, blank, and standard solutions are each mixed with the color reagent, which contains phospholipase D, choline oxidase, peroxidase, 4-aminoantipyrine, 3,5-dimethoxy-N-ethyl-N-(2-hydroxy-3-sulfopropyl)aniline sodium, and ascorbic oxidase. After reaction at 37
C for 5 min, the absorbance at 600 nm is measured to estimate the EYPC concentration. The concentration-estimated liposome suspension in PBS is diluted up to 950 ml by addition of distilled water and 10 times concentrated PBS (10 PBS) in an appropriate ratio. Fifty microliters of Au NP dispersion ([Au] ¼ 0.91 mM ) is added to the liposome suspension and vortexed. The final solution is 46 mM of gold ions in 1 PBS. For use in TEM analysis, complexes with an EYPC/Au ratio of 10/1 are prepared by vortex mixing 500 ml of the Au NP dispersion ([Au] ¼ 0.91 mM) with 500 ml of the liposome suspension, diluted with distilled water and 10 PBS. The final solution is 4.6  102 mM of gold ions in 1 PBS.

3. Time-Dependent SPR of the Complexes One of characteristic properties of Au NPs is surface plasmon absorption around 523 nm. To investigate this in each of the complex dispersions, the UV–Vis absorption spectra ranging from 400 to 800 nm are measured

135

Complex of Liposomes with Gold Nanoparticles

using a Jasco Model V-560 spectrophotometer ( Jasco Inc., Japan) at 25
C (Haba et al., 2007). As a control, the spectrum of a dispersion solution without liposomes is also measured. The SPR absorption of Au NPs in the absence of liposomes almost disappeared under physiological conditions (Fig. 7.2A). This is due to aggregation resulting from shielding of electrostatic repulsion (Burda et al., 2005; Daniel and Astruc, 2004). In contrast, the SPR absorption in spectra of Au NPs in the presence of EYPC liposomes was retained; this was dependent on the EYPC/Au ratio (Fig. 7.2). At ratios of 1/10 and 1/1, the SPR signal decreased (rapidly and slowly, respectively), while at a ratio of 10/1, the signal remained after 12 h. This indicated that EYPC liposomes contributed to stable dispersion of Au NPs under isotonic conditions, preventing their aggregation.

A

B

0.2

Au NP only

0.2

EYPC/Au = 1/10 1h

Abs

Abs

0h

0.1 4h 8h 12 h

Abs

C

500 600 700 Wavelength (nm)

800 D

0.2

0.1

Liposome only 0 400

1h 4h 8h 12 h

EYPC/Au = 1/1

500 600 700 Wavelength (nm)

0.4

800

EYPC/Au = 10/1

0.3 1 h→12 h Abs

0 400

4h 8h 12 h

0.1

0.2 0.1

0 400

Liposome only

Liposome only 500 600 700 Wavelength (nm)

800

0 400

500 600 700 Wavelength (nm)

800

Figure 7.2 Time-dependent UV–Vis spectra of the complex of EYPC liposomes (200 nm in diameter) with Au NPs at different ratios in PBS. The time-dependent UV–Vis spectra of Au NPs are only shown as a control (A). The mole ratio of EYPC to Au is 1/10 (B), 1/1 (C), and 10/1 (D). [Au] ¼ 46 mM (from Kojima et al., 2008).

136

Chie Kojima et al.

4. TEM Analysis of the Complexes TEM analysis is performed as follows (Hayashi et al., 1998; Kojima et al., 2008). Collodion-coated grids are coated with a carbon thin film, using an ion sputtering device (E-1030, Hitachi High-Technologies Corp., Japan). After incubation for 1 day in a desiccator, a small drop of sample is placed on the grid for 5 min and then the excess drawn off with filter paper. The Au concentration of the complex at an EYPC/Au ratio of 0/10 and 1/10 was 46 mM. At a ratio of 10/1, the Au concentration of the complex is increased up to 4.6  102 mM, because it is difficult to identify the liposome under this diluted condition. A drop of 2% (w/v) phosphotungstic acid (pH 7) is applied to the grid, drawn off with filter paper, and the stained sample is allowed to dry. The grid is viewed under an electron microscope at 200 kV ( JEOL Ltd., JEM-2000FEX II). Figure 7.3 shows the TEM images of Au NPs in the absence and presence of EYPC liposomes. Large aggregates were observed in TEM

Figure 7.3 TEM images of the complexes of EYPC liposomes with Au NPs at the EYPC/Au ratio of 0/10 (A), 1/10 (B), and 10/1 (C) in PBS after the 6 h incubation. Arrows indicate the Au NPs. Bar ¼ 100 nm (from Kojima et al., 2008).

Complex of Liposomes with Gold Nanoparticles

137

images of both the sample comprising only Au NPs and the complex at an EYPC/Au ratio of 1/10. In contrast, no Au NPs aggregates were observed for the complex at an EYPC/Au ratio of 10/1. These findings are consistent with the SPR analysis results. In the TEM image of the complex at an EYPC/Au ratio of 10/1, many Au NPs were observed at the boundary surface within the liposomal assembly. This indicates that the Au NPs complexed with the liposomes.

5. DLS Analysis of the Complexes DLS analysis is performed in the vesicle mode of Nicomp ZLS380 (Nicomp) at room temperature using EYPC suspension (2.5 ml, approximately 0.01 mM) (Kojima et al., 2008). This is performed in both the absence and presence of the Au NP at an EYPC/Au ratio of 10/1. The size of liposome was unchanged before and after the addition of Au NPs (Fig. 7.4), suggesting that Au NPs can complex with liposomes without the aggregation. This finding is not consistent with our TEM results, in which liposomal assembly was observed. It is possible that the liposomal aggregation was an artifact of TEM sample preparation.

6. Calcein Release from the Complexes The collapse behavior of liposomes is analyzed by adding Au NPs to calcein-loaded liposomes. Although the fluorescence of the calcein encapsulated in the liposomes is essentially quenched, an intense florescence is observed after its release from the liposome. The loaded and unloaded percent of calcein is determined from the fluorescence at the initial step and the fluorescence after adding detergent to collapse the liposome, respectively. Calcein release measurements are performed according to the method previously reported (Kono et al., 1994), with some modification (Kojima et al., 2008). Fluorescence intensity is largely influenced by the photochemical properties of the Au NP, so the concentration of Au NPs was decreased to the minimum detection level of calcein. A chloroform solution of EYPC (10 mg/ml, 500 ml) is evaporated to remove the solvent. The obtained thin lipid membrane (5.0 mg, 6.25 mmol) is further dried under vacuum for at least 2 h, and then dispersed in 0.5 ml of 63 mM calcein aqueous solution (pH 7.4) with sonication for 3 min. The liposome suspensions are freeze-thawed four times. The obtained liposome suspension is extruded through a polycarbonate membrane with a pore diameter of 100 nm. Free calcein is removed by gel permeation chromatography using a

138

Chie Kojima et al.

A

1.2

Intensity (a.u.)

1 0.8 0.6 0.4 0.2 0

B

10

26

66

169 433 1110 Diameter (nm)

2848

7305

10

26

66

169 433 1110 Diameter (nm)

2848

7305

1.2

Intensity (a.u.)

1 0.8 0.6 0.4 0.2 0

Figure 7.4 Intensity-weighted distributions of the intact liposomes (A) and the complexes of liposomes with Au NPs at the EYPC/Au ratio of 10/1 (B) in PBS after the 6 h-incubation by DLS (from Kojima et al., 2008).

Sepharose 4B column and PBS (10 mM Na2HPO3–NaH2PO3, 150 mM NaCl, pH 7.4). The lipid concentrations are estimated by Phospholipids C-Test Wako (Wako Pure Chemical Industries Ltd.) according to the manufacturer’s instructions, as described earlier. An aliquot of the calceinloaded liposome dispersion is added to 3 ml of PBS containing 0.5 mM ethylenediaminetetraacetic acid (EDTA, Kishida Chemical); the final lipid concentration is 0.2 mM. The fluorescence intensity of the solution is monitored using a spectrofluorometer ( Jasco Inc., FP-6500) at excitation and monitoring wavelengths of 480 and 515 nm, respectively. Measurements are taken before and after the addition of Au NPs to the liposome

139

Complex of Liposomes with Gold Nanoparticles

suspension at an EYPC/Au ratio of 1/10. The amount of calcein remaining in the liposome is estimated from Eq. (7.1). 0

Ft  Ft  100 Fluorescence intensity ð%Þ ¼ 0 F0  F0

ð7:1Þ

where Ft and F0 refer to the fluorescence at t and 0 h after the addition of the Au 0 0 NP dispersion or PBS, and Ft and F0 refer to the fluorescence of these samples with the addition of 10% Triton X-100 solution (Kishida Chemical, final concentration 0.03%) at t and 0 h. The fluorescence intensity (%) of calcein for the intact liposome decreased after the 24-h incubation (Fig. 7.5). This suggests that the calcein absorbed on the liposome membrane might be released. The fluorescence intensity of calcein after the addition of Au NPs at an EYPC/Au ratio of 1/10 was the same as that in the intact liposome, suggesting that adding Au NPs did not promote calcein release. This implies that the liposomal membrane remained intact after interaction with Au NPs.

7. Estimation of Numbers of the Au NP and the Liposome in the Complexes

Fluorescence intensity (%)

As described earlier, complexed Au NPs at an EYPC/Au ratio of 10/1 were stably dispersed. However, complexes at ratios of 1/1 and 1/10 were not. The particle numbers for Au NPs and liposomes in the complexes can be calculated. A liposome of 200 nm in diameter was determined to contain 3.5  105 EYPC molecules, from Eq. (7.2).

100 80 60 40 20 0

0

5

10 15 20 Incubation time (h)

25

Figure 7.5 Time-dependent fluorescence intensity (%) of calcein in the liposomes with (open symbols) and without (closed symbols) of Au NPs is shown (from Kojima et al., 2008).

140

Chie Kojima et al.

4pr 2  2 ð7:2Þ A where N, r, and A refer to the number of lipids, the radius of the liposomes, and the section area of the EYPC head group (0.717 nm2), respectively (Lasic, 1993). Whereas an Au NP of 13 nm in a diameter is composed of 6.8  104 gold atoms, as determined from Eq. (7.3).    3 2 D ð7:3Þ U¼ p 3 a N¼

where U, D, and a refer to the number of Au atoms, the diameter of the Au NPs, and the edge length of a unit cell (0.40786 nm), respectively (Chithrani et al., 2006). From these calculations, the respective numbers of Au NPs and EYPC liposome in the complex at an EYPC/Au ratio of 10/1 are estimated to be 7.8  1011 and 4.0  1011, under our conditions with 45.5 nmol Au ion and 455 nmol EYPC lipid. Therefore, an EYPC liposome might interact with approximately one Au NP. Overall, these results show that the complexes at an EYPC/Au of 10/1 had a particle ratio of approximately one-to-one and were stably dispersed without disturbing the liposome structure. The enhancement of affinity between the liposome and the Au NPs is of significance for the application.

8. Optimization of Lipid Components of the Complexes As the affinity of liposome to Au NPs should be affected by the lipid component of the liposome, complexes are prepared using both cationic and PEG-modified liposomes. It is expected that cationic lipids and PEG will interact with anionic Au NPs more efficiently. These liposomes are prepared using DDAB (Tokyo Chemical Industry Co. Ltd., Tokyo, Japan) and PEG-PE (Avanti Polar Lipids Inc., Alabaster, AL), in addition to EYPC (Fig. 7.6B and C). DDAB-containing liposomes (10 mol% DDAB (0.625 mmol) and 90 mol% EYPC (5.625 mmol)) are prepared according to the same preparation method of the EYPC liposome before the sonication. The liposomes are incubated at 4
C for 1 day, followed by sonication for 10 min. The size distribution of this liposome preparation is determined to be approximately 200 nm by DLS. A PEG-PE-containing liposome with a diameter of 200 nm and composition of 5 mol% PEG-PE (0.31 mmol) and 95 mol% EYPC (5.94 mmol) is also prepared, according to the same procedure as the EYPC liposomes. The concentration of these lipids is

141

Complex of Liposomes with Gold Nanoparticles

O-

+

O

CH−O−C−R −

O

−

=

(CH3)3N−CH2CH2−O−P−O−CH2

=

EYPC

−

A

=

CH2−O−C−R O

DDAB +

N

CH3

(CH2)17CH3

PEG-PE n

C−NH−CH2CH2−O−P−O−CH2 O

O

O

=

2

−

2

−

(CH −CH −O)

=

CH3O

O− =

C

(CH2)17CH3

−−

CH3

−

Br−

−

B

CH−O−C−(CH2)16CH3 =

CH2−O−C−(CH2)16CH3 O

Figure 7.6 Structures of EYPC (A), DDAB (B), and PEG-PE (C).

estimated from the measured EYPC concentration, considering the in feed ratios. The DDAB- and PEG-bearing liposomes also inhibited the decrease of the SPR signal of the Au NPs, with a dependence on the ratio of lipid to Au (Fig. 7.7). The stability of the SPR signal for these liposomes was similar to the EYPC liposomes. This suggests that complex formation was not improved by DDAB and PEG-PE. Complex formation using liposomes with alkanethiol and different molecular weight preparations of PEG-PE may also be investigated. Liposomes are prepared using C18-SH (Tokyo Chemical Industry Co. Ltd.) and PEG2000-PE or PEG5000-PE (NOF Corp.). PEG-PE- and C18-SHcontaining liposomes with diameters of 100 nm and composition of 5 mol% PEG2000-PE or PEG5000-PE (0.31 mmol), 15 mol% of C18-SH (0.93 mmol), and 80 mol% of EYPC (5.00 mmol) is prepared, according to the same procedure as the EYPC liposome minus the reduction. Before the addition of Au NPs, the liposomes are reduced by reacting with dithiothreitol (DTT; Wako Pure Chemical Industries Ltd.) for 1 h, followed by dialysis using degassed PBS (10 mM Na2HPO3–NaH2PO3, 150 mM NaCl, pH 7.4). The PEG2000- and PEG5000-bearing liposome complexes containing C18-SH inhibited efficiently the decrease of the Au NP SPR signal in the time-dependent SPR absorption spectra, even at a lipid/Au ratio of 1/10 (Fig. 7.8). The PEG5000-PE-bearing liposome was the most stably dispersed complex, suggesting that both the alkanethiol bound to the surface of the Au NPs and PEG play a role in good dispersal of the complex.

142

A

Chie Kojima et al.

B

0.4

0.4

Lipid/Au = 10/1 1 h→12 h

0.3 Abs

Abs

0.3

Lipid/Au = 10/1

0.2 0.1

0.2 0.1

Liposome only 0.0 400

Abs

D

0.1 0.05 0 400

800

F Lipid/Au = 1/10

0.2 0.15

4h 8h 12 h

800

Lipid/Au = 1/10

0.1

1h 4h 8h 12 h

0.05

0.05 Liposome only

0.0 400

500 600 700 Wavelength (nm)

1h

Abs

Abs

Liposome only

500 600 700 Wavelength (nm)

0.2

0.1

1h 4h 8h 12 h

0.15

Liposome only

0.15

800

Lipid/Au = 1/1

0.05

E

500 600 700 Wavelength (nm)

Lipid/Au = 1/1

1h 4h 8 h 12 h

0.1

0.0 400

Liposome only

0.2

0.2 0.15

0 400

800

Abs

C

500 600 700 Wavelength (nm)

1 h→12 h

500 600 700 Wavelength (nm)

Liposome only 800

0 400

500 600 700 Wavelength (nm)

800

Figure 7.7 Time-dependent UV–Vis spectra of the complex of (A, C, E) DDAB- or (B, D, F) PEG-bearing liposomes with Au NPs at different ratios. The mole ratio of lipid to Au is (A, B) 10/1, (C, D) 1/1, and (E, F) 1/10. [Au] ¼ 46 mM.

9. Concluding Remarks We have described the preparation of various complexes of dispersions of liposomes and Au NPs. Liposomes improve the stability of Au NP dispersions under isotonic conditions. The complexes are formed without disturbing the liposome structure. In addition, complex formation is enhanced by using PEG- and thiol-modified liposomes. It is known that

143

Complex of Liposomes with Gold Nanoparticles

A

0.2 PEG-2000 0.15

Abs

1h 4h 0.1

8h 12 h

0.05

Liposome only 0 400

500

600 700 Wavelength (nm)

800

B 0.2 PEG-5000

Abs

0.15

1h 4h 8h 12 h

0.1

0.05 Liposome only 0 400

500

600 700 Wavelength (nm)

800

Figure 7.8 Time-dependent UV–Vis spectra of the complex of (A) PEG-2000- or (B) PEG-5000-bearing liposomes containing alkanethiol with Au NPs at the lipid/Au of 1/10. [Au] ¼ 46 mM.

PEG-bearing liposomes are used widely as a drug carrier due to the biocompatibility and the prolonged blood circulation of the liposomes (Greenwald et al., 2000). These types of complexes have potential diagnostic and therapeutic applications in nanomedicine.

ACKNOWLEDGMENT This work was supported by Grants-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT).

144

Chie Kojima et al.

REFERENCES Al-Jamal, W. T., and Kostarelos, K. (2007). Liposome-nanoparticle hybrids for multimodal diagnostic and therapeutic application. Nanomedicine 2, 85–98. Burda, C., Chen, X., Narayanan, R., and El-Sayed, M. A. (2005). Chemistry and properties of nanocrystals of different shapes. Chem. Rev. 105, 1025–1102. Chilkoti, A., Dreher, M. R., Meyer, D. E., and Raucher, D. (2002). Targeted drug delivery by thermally responsive polymers. Adv. Drug Deliv. Rev. 54, 613–630. Chithrani, B. D., Ghazani, A. A., and Chan, W. C. W. (2006). Determining the size and shape dependence of gold nanoparticle uptake into mammalian cells. Nano Lett. 6, 662–668. Daniel, M. C., and Astruc, D. (2004). Gold nanoparticles: Assembly, supramolecular chemistry, quantum-size-related properties, and applications toward biology, catalysis, and nanotechnology. Chem. Rev. 104, 293–346. Govorov, A. O., and Richardson, H. H. (2007). Generating heat with metal nanoparticles. Nanotoday 2, 30–38. Grabar, K. C., Freeman, R. G., Hommer, M. B., and Natan, M. J. (1995). Preparation and characterization of Au colloid monolayers. Anal. Chem. 67, 735–743. Greenwald, R. B., Conover, C. D., and Choe, Y. H. (2000). Poly(ethylene glycol) conjugated drugs and prodrugs: A comprehensive review. Crit. Rev. Ther. Drug Carrier Syst. 17, 101–161. Haba, Y., Kojima, C., Harada, A., Ura, T., Horinaka, H., and Kono, K. (2007). Preparation of poly(ethylene glycol)-modified poly(amidoamine) dendrimers encapsulating gold nanoparticles and their heat-generating ability. Langmuir 23, 5243–5246. Hayashi, H., Kono, K., and Takagishi, T. (1998). Temperature-dependent associating property of liposomes modified with a thermosensitive polymer. Bioconjug. Chem. 9, 382–389. Hong, K., Friend, D. S., Glabe, C. G., and Papahadjopoulos, D. (1983). Liposomes containing colloidal gold are a useful probe of liposome-cell interactions. Biochim. Biophys. Acta 732, 320–323. Immordino, M. L., Dosio, F., and Cattel, L. (2006). Stealth liposomes: Review of the basic science, rationale, and clinical applications, existing and potential. Int. J. Nanomed. 1, 297–315. Jain, P. K., El-Sayed, I. H., and El-Sayed, M. A. (2007). Au nanoparticles target cancer. Nanotoday 2, 18–29. Kim, D., Park, S., Lee, J. H., Jeong, Y. Y., and Jon, S. (2007). Antibiofouling polymercoated gold nanoparticles as a contrast agent for in vivo X-ray computed tomography imaging. J. Am. Chem. Soc. 129, 7661–7665. Kojima, C., Hirano, Y., Yuba, E., Harada, A., and Kono, K. (2008). Preparation and characterization of complexes of liposomes with gold nanoparticles. Colloids Surf. B 66, 246–252. Kono, K. (2001). Thermosensitive polymer-modified liposomes. Adv. Drug Deliv. Rev. 53, 307–319. Kono, K., and Arshady, R. (2006). Smart Nano and Microparticles. Kentus Books, London. Kono, K., Hayashi, H., and Takagishi, T. (1994). Temperature-sensitive liposomes: Liposomes bearing poly(N-isopropylacrylamide). J. Control. Release 30, 69–75. Kono, K., Nakai, R., Morimoto, K., and Takagishi, T. (1999). Thermosensitive polymermodified liposomes that release contents around physiological temperature. Biochim. Biophys. Acta 1416, 239–250. Lasic, D. D. (1993). Liposomes, from physics to applications. Elsevier, Amsterdam. Li, X., Li, Y., Yang, C., and Li, Y. (2004). Liposome induced self-assembly of gold nanoparticles into hollow spheres. Langmuir 20, 3734–3739.

Complex of Liposomes with Gold Nanoparticles

145

Link, S., and El-Sayed, M. A. (1999). Spectral properties and relaxation dynamics of surface plasmon electronic oscillations in gold and silver nanodots and nanorods. J. Phys. Chem. B 103, 8410–8426. Link, S., and El-Sayed, M. A. (2000). Shape and size dependence of radiative, nonradiative, and photothermal properties of gold nanocrystals. Int. Rev. Phys. Chem. 19, 409–453. Paasonen, L., Laaksonen, T., Johans, C., Yliperttula, M., Konttun, K., and Urtti, A. (2007). Gold nanoparticles enable selective light-induced contents release from liposomes. J. Control. Release 122, 86–93. Park, S.-H., Oh, S.-G., Mun, J.-Y., and Han, S.-S. (2006). Loading of gold nanoparticles inside the DPPC bilayers of liposome and their effects on membrane fluidities. Colloids Surf. B 48, 112–118. Pissuwan, D., Valenzuela, S. M., and Cortie, M. B. (2006). Gold nanoparticles as therapeutics. Trends Biotechnol. 24, 62–67. Volodkin, D. V., Skirtach, A. G., and Mohwald, H. (2009). Near-IR remote release from assemblies of liposomes and nanoparticles. Angew. Chem. Int. Ed. Engl. 48, 1807–1809. Wu, G., Mikhailovsky, A., Khant, H. A., Fu, C., Chiu, W., and Zasadzinski, J. A. (2008). Remotely triggered liposome release by near-infrared light absorption via hollow gold nanoshells. J. Am. Chem. Soc. 130, 8175–8177.

This page intentionally left blank

C H A P T E R

E I G H T

Bio-Nanocapsule–Liposome Conjugates for In Vivo Pinpoint Drug and Gene Delivery Takeshi Kasuya,* Joohee Jung,*,1 Rie Kinoshita,† Yasumasa Goh,† Takashi Matsuzaki,* Masumi Iijima,*,‡ Nobuo Yoshimoto,*,‡ Katsuyuki Tanizawa,* and Shun’ichi Kuroda*,†,‡ Contents 1. 2. 3. 4. 5.

Introduction First-Generation Bio-Nanocapsules Second-Generation BNCs Retargeting of BNC–LP Conjugates Overexpression of BNCs in S. cerevisiae 5.1. Purification of BNCs by ultracentrifugation (
10 mg protein per lot) 5.2. Purification of BNCs using column chromatography 5.3. Purification of ZZ–BNC using column chromatography 6. Conjugation of BNCs with LPs 6.1. Example 1.1: Preparation of BNC–LP conjugates containing DNA (BNC–lipoplex conjugates) 6.2. Example 1.2: In vitro transfection with BNC–lipoplex conjugates 6.3. Example 1.3: In vivo transfection with BNC–lipoplex conjugates 6.4. Example 2.1: Preparation of BNC–LP conjugates containing DOX 6.5. Example 2.2: In vitro cytotoxic effects of BNC–LP conjugates containing DOX 6.6. Example 2.3: In vivo therapeutic effects of BNC–LP conjugates containing DOX

* { { 1

148 149 150 152 153 154 154 155 155 156 156 158 158 159 161

Institute of Scientific and Industrial Research, Osaka University, Ibaraki, Osaka, Japan Beacle Inc., ORIC, Haga, Okayama, Japan Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa, Nagoya, Japan Present address: Institute for Innovative Cancer Research, ASAN Medical Center, Pungnap-2, Songpa, Seoul, Korea

Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64008-8

#

2009 Elsevier Inc. All rights reserved.

147

148

Takeshi Kasuya et al.

7. Preparation of Antibody-Displaying BNC–LP Conjugates 8. Preparation of Biotin-Displaying BNC–LP Conjugates 9. Concluding Remarks Acknowledgments References

161 163 163 164 164

Abstract A bio-nanocapsule (BNC) is an  50-nm hepatitis B virus (HBV) subviral particle comprising HBV envelope L proteins and a lipid bilayer, and is synthesized in recombinant Saccharomyces cerevisiae. When BNCs are administered intravenously in a mouse xenograft model, they can accumulate specifically in human liver-derived tissues and enter cells efficiently by the HBV-derived human liverspecific infection machinery, localized at the outer-membrane pre-S region of the L protein. BNC specificity for the human liver can be altered to other tissues by substituting the pre-S region using targeting molecules (e.g., antibodies, lectins, cytokines). BNCs can spontaneously form complexes with liposomes (LPs) by the membrane fusogenic activity of the pre-S region. LPs containing various therapeutic materials (e.g., chemicals, proteins, DNA, RNA) can therefore be covered with BNCs to form an  150-nm BNC–LP conjugate. BNC–LP conjugates injected intravenously can deliver incorporated materials to target tissues specifically and efficiently by utilizing the HBV-derived infection machinery. The stability of BNC–LP conjugates in the blood circulation is similar to that of PEGylated LPs. In this chapter, we describe the preparation and in vivo application of BNC–LP conjugates, and the potential of BNC–LP conjugates as in vivo pinpoint drug delivery systems.

1. Introduction Liposomes (LPs) are one of the most promising drug delivery system (DDS) carriers for genes and drugs, and several LP-based medicines have entered clinical use (Lorusso et al., 2007; Markman, 2006; Richardson, 2006; Rosenthal et al., 2002). Most LP-based medicines are nanoscaled. They are used for the delivery of anticancer drugs to carcinomatous lesions by taking advantage of the enhanced permeation and retention effect of solid tumors (Maeda et al., 2000). The blood vasculature in tumors is leaky, because it possesses larger sized pores than the blood vasculature in healthy tissues. A nanomedicine of diameter 100 nm can therefore readily seep out from the blood circulation and accumulate in tumors (‘‘passive targeting’’). The major obstacle for passive targeting is the immune system in the liver and spleen, that is, the reticuloendothelial system (RES) in which macrophages (including Kupffer cells) capture opsonized micro- and nanoscaled compounds (Moghimi et al., 2001). To escape the RES, recent LPs have been modified with polyethylene glycol (PEG) (i.e., PEGylated LPs),

Bio-Nanocapsule–Liposome Conjugate

149

thereby enhancing their stability in the blood circulation by avoiding LPs from binding to serum proteins (‘‘opsonization’’) (Lasic and Martin, 1995; Moghimi and Szebeni, 2003). PEGylated LPs containing anticancer drugs could dramatically reduce adverse side effects, but several reports have noted that the LP encapsulation of anticancer drugs contribute only to the reduction of side effects, not to the enhancement of therapeutic effects (Hong et al., 1999; Parr et al., 1997). This indicates that modifying LPs for the delivery of drugs to actively target foci (‘‘active targeting’’) is required. After attachment onto the cell surface, LPs are usually incorporated within target cells through an endocytotic cascade. Most of the drugs in LPs are degraded in late endosome and lysosome without exhibiting sufficient therapeutic effects. In contrast to LPs, several viral vectors (e.g., adenovirus, adeno-associated virus, lentivirus, retrovirus) are widely used in gene delivery. This is because recombinant viral vectors can efficiently pass across the plasma membrane and deliver their genomic information to the cell nucleus as well as escaping from the endocytotic cascade. These viral vectors may have unexpected severe side effects originating from the viral genome (Marshall, 2002; Savulescu, 2001), but the efficiency of gene delivery is much higher than that of lipoplexes (LP–DNA complexes) (Hama et al., 2006). Taken together, the next generation of LP-based nanomedicines must have the following features: (1) viral genome-free structure; (2) active targeting machinery; (3) escape machinery from the RES; (4) escape machinery from the endosomal degradation pathway; and (5) intracellular targeting machinery.

2. First-Generation Bio-Nanocapsules In the process of developing a hepatitis B virus (HBV) vaccine, we succeeded in overexpressing HBV subviral L particles in Saccharomyces cerevisiae (Kuroda et al., 1992). The L particle is an 50-nm hollow capsule comprising HBV envelope L proteins and a lipid bilayer (Yamada et al., 2001). In 2003, we unexpectedly found that the L particles could incorporate genes and drugs by electroporation, and deliver the incorporated materials specifically and efficiently to cell lines (in vitro) and tumors (in vivo) derived from the human liver (Yamada et al., 2003). We therefore designated the HBV L particle as a ‘‘bio-nanocapsule’’ (BNC) (Yu et al., 2005). The hepatophilic and highly efficient delivery properties of BNCs are considered to be based on the infection mechanism of HBV. BNCs and HBV display the N-terminal half of the outer-membrane pre-S region of L protein (Fig. 8.1), which is postulated to have a pivotal role in the human liver-specific attachment of HBV (Kasuya et al., 2008a,b). We succeeded in the continuous expression (>1 month) of the human blood clotting factor

150

Takeshi Kasuya et al.

Bio-nanocapsule (BNC)

Hepatitis B virus (HBV) Core protein Viral genome

DNA polymerase

Hollow (genome free) pre-S

Cellular internalization

Human hepatocytespecific binding site

S

HBV envelope L protein

Figure 8.1 HBV and BNC (schematic).

IX gene in a mouse xenograft model by a single intravenous injection of an electroporated BNC (Yamada et al., 2003). We also succeeded in obtaining a significant reduction in the size of human liver-derived tumors in a mouse xenograft model by a single intravenous injection of an electroporated BNC containing the herpes simplex virus thymidine kinase (HSV-tk) gene and sequential treatment with gancyclovir (an HSV-tk-dependent inhibitor of DNA polymerases) (Iwasaki et al., 2007). Engineered BNCs, of which the human liver-specific attachment site of the pre-S region is changed to other targeting molecules (epidermal growth factor (EGF), anti-EGF receptor (EGFR) monoclonal antibody), have been shown to accumulate in nonhuman liver tissues in vitro and in vivo (Tsutsui et al., 2007; Yamada et al., 2003). This strongly suggests that BNCs are promising in vivo pinpoint DDS carriers targeting not only human liver tissue but also other tissues.

3. Second-Generation BNCs Electroporation was used to transiently induce micropores across the lipid bilayer of BNCs. It was found that BNCs stored for long periods tended to show lower incorporation efficiency than freshly prepared BNCs. The intra- and intermolecular disulfide bonds of L proteins (14 Cys residues in three transmembrane segments of the L protein) were postulated to gradually form in a time-dependent manner, which would make BNCs

151

Bio-Nanocapsule–Liposome Conjugate

resistant to electroporation. We substituted each Cys residue with Ala or Ser by genetic modification, and identified at least eight Cys residues of the L protein unnecessary for BNC formation (Nagaoka et al., 2007). The BNC harboring the eight Cys ! Ala or Cys ! Ser mutations showed good incorporation efficiency in an electroporation method. This method was not suitable for introducing large materials into a BNC (e.g., >20-kbp plasmid for gene therapy; >10-nm fluorescence-labeled polystyrene beads for bioimaging) and adopting the electroporation method for the good manufacturing practice (GMP)-based production of BNC-based nanomedicines would be very difficult. We recently found that the N-terminal half of the L protein possesses membrane fusogenic activity (Matsuzaki et al., unpublished data). BNCs spontaneously form an 150-nm rigid complex with LPs (BNC–LP conjugate) in which multiple BNCs are embedded on the surface of LPs (Fig. 8.2) ( Jung et al., 2008). This property of BNC allows incorporation of various therapeutic materials into BNC–LP conjugates as follows. First, various materials (even 40-kbp plasmid, 100-nm fluorescent polystyrene beads) are constantly incorporated into LPs by conventional methods. Second, LPs are covered with BNCs harboring tissue specificity and high infectivity by the fusogenic activity of BNC. Third, BNC–LP conjugates can deliver various materials incorporated specifically and efficiently into human liver-derived tumors in a mouse xenograft model

Liposome

Incorporation

DNA

siRNA BNC

Conjugation Protein

Chemical compounds

BNC-liposome conjugate

Figure 8.2

Preparation of BNC–LP conjugates (bars: 50 nm).

152

Takeshi Kasuya et al.

through intravenous injection ( Jung et al., 2008). Additionally, this method can be expanded to GMP-based mass production more readily than electroporation. We therefore designated the BNC–LP conjugate as a ‘‘secondgeneration BNC’’ (Kasuya and Kuroda, 2009). With respect to the attachment of BNC–LP conjugates to human liver cells, the N-terminal half of the pre-S region has as a specific receptor for the human liver, which was demonstrated in HBV and BNCs. The intracellular drug release from BNC–LP conjugates may be mediated by the cellular internalization activity (another fusogenic activity) of the C-terminal half of the pre-S region and S region (see Fig. 8.1), as demonstrated in HBV (Glebe and Urban, 2007). In fact, an engineered BNC lacking the N-terminal half of the pre-S region can form a BNC–LP conjugate and deliver incorporated materials to target cells (Kasuya et al., 2008c). BNC–LP conjugates therefore possess the advantages of LPs and viral vectors, namely the use of versatile materials (as with LP) and the HBV-derived infection and active targeting machinery. These results indicate that BNC–LP conjugates are more promising than BNC per se as in vivo pinpoint DDS carriers.

4. Retargeting of BNC–LP Conjugates Due to the narrow tropism of HBV (because of the function of the N-terminal half of the pre-S region), systemically injected BNCs accumulate specifically in human liver-derived tissues in vivo. For expanding the indications of BNC-based nanomedicines, it is important to establish the methodology for retargeting BNC from human liver to nonliver tissues by substitution of the pre-S region by other targeting molecules (e.g., cytokines, growth factors, receptors, antibodies, glycans, lectins, aptamers). First, the pre-S (3–77) region is replaced with EGF by genetic engineering. The EGF-displaying BNC lost the specificity to human liver cells and obtained new specificity to EGFR-overexpressing A431 cells in vitro (Yamada et al., 2003). This approach needs a time-consuming step for constructing the expression system and sometimes fails to achieve the high productivity of BNCs in yeast cells. Next, a large part (50–159) of the pre-S region is replaced with the Staphylococcus aureus protein A-derived IgG Fc-binding ZZ domain. Beyond expectation, the ZZ domain-displaying BNC (ZZ–BNC) is efficiently synthesized in yeast cells. After ZZ–BNC is mixed with anti-EGFR IgG, the mixture is injected intracranially to a glioma-transplanted mouse orthograft model. The anti-EGFR IgG-displaying BNC accumulated in the EGFR-overexpressing transplanted glioma in the mouse corpus striatum (Kurata et al., 2008; Tsutsui et al., 2007), indicating that the antibody-displaying ZZ–BNC is applicable for

Bio-Nanocapsule–Liposome Conjugate

153

in vivo use. In 2008, because the ZZ domain contains many Lys residues, ZZ–BNC was modified with N-hydroxysuccinimide (NHS)-biotin to display biotin molecules onto the BNC surface. The biotinylated ZZ– BNC can be used for displaying various biotinylated targeting molecules (see above) by using an avidin (e.g., streptavidin, neutravidin) as an adaptor. For instance, the Phaseolus vulgaris agglutinin-L4 isolectin (L4-PHA)-displaying BNC (PHA–BNC) has been shown to accumulate in vivo in highly metastatic malignant tumors which overexpress b1-6-branching N-acetylglucosamine (GlcNAc), a specific ligand of L4-PHA (Kasuya et al., 2008c). PHA–BNC–LP conjugate can deliver luciferase-expressing plasmid to the b1-6GlcNAc-overexpressing cells, showing that the lack of a large part of the pre-S region does not affect the formation of BNC–LP conjugates. We recently revealed that the fusogenic activity of BNC required for LP conjugation is delineated in the short sequence at the N-terminal of pre-S region, which remains in ZZ–BNC (Matsuzaki et al., unpublished data). These data strongly suggested that the engineered BNCs (ZZ–BNC, biotinylated ZZ–BNC) and their LP conjugates can be used for lesion-specific pinpoint DDS carriers. In this chapter, we describe methods to prepare BNC–LP conjugates containing DNA, and BNC–LP conjugates containing an anticancer drug (doxorubicin (DOX)), and the effects of these BNC–LP conjugates in in vitro and in vivo systems.

5. Overexpression of BNCs in S. cerevisiae BNC is produced in S. cerevisiae AH22R (a leu2 his4 can1 cirþ pho80) strain (Kobayashi et al., 1988). The yeast cells are transformed by the spheroplast method (Hinnen et al., 1978) with the YEp plasmid pGLDLIIP39-RcT (Kuroda et al., 1992), which encodes the N-terminal chicken lysozyme signal sequence-fused HBV envelope L protein (subtype adr) under the glyceraldehyde-3-phosphate dehydrogenase (GLD; also called TDH3) gene promoter. For production of ZZ–BNC, the DNA segment encoding the pre-S region (50–159) in the plasmid pGLDLIIP39-RcT is replaced with the DNA segment encoding the IgG Fc-binding ZZ domain derived from S. aureus protein A (Tsutsui et al., 2007). The LEU2þ yeast transformants are cultured in a synthetic selection medium 8S5N-P400 (Yamada et al., 2001) at 30  C for 7 days, harvested by centrifugation, and stored at 80  C. The amount of BNC in yeast cells is estimated to be 40% of total soluble proteins (Kuroda et al., 1992).

154

Takeshi Kasuya et al.

5.1. Purification of BNCs by ultracentrifugation (
10 mg protein per lot) BNC and ZZ–BNC can be readily purified by ultracentrifugation owing to the overexpression of BNCs in yeast cells. The yeast cells (about 20 g wet weight) are disrupted with glass beads (0.5 mm in diameter, 175 ml) using a BEAD-BEATER (BioSpec Products, Bartlesville, OK, USA) in 160 ml of 0.1 M sodium phosphate buffer (pH 7.2) containing 7.5 M urea, 15 mM ethylenediaminetetraacetic acid (EDTA), 4 mM phenylmethylsulfonyl fluoride (PMSF), 0.1 mM 4-amidinophenyl-methanesulfonyl fluoride (APMSF), and 0.1% (v/v) Tween 80. The crude extract obtained by centrifugation at 34,780g at 4  C for 30 min (about 6 g of protein) is mixed with PEG 6000 solution (15%, w/v at final concentration) to precipitate BNCs. The precipitants (about 3 g of protein) are subjected to CsCl isopycnic ultracentrifugation (10–40%, w/v) using an SW28 rotor (Beckman Coulter, Inc., Fullerton, CA, USA) at 24,000 rpm at room temperature for 15 h. Fractions containing BNCs are determined by sandwich enzyme-linked immunosorbent assay (ELISA) for BNC using an IMx HBsAg assay system (Abbott Laboratories, Abbott Park, IL, USA). Positive fractions are subjected twice to sucrose density gradient ultracentrifugation (10–50%, w/v) using an SW28 rotor at 24,000 rpm at room temperature for 15 h. Fractions containing BNCs are concentrated by ultrafiltration using an Amicon Ultra 100,000 NMWL (Millipore, Billerica, MA, USA) and subjected to a Sephacryl S-500 HR (GE Healthcare, Waukesha, WI, USA) gel-filtration column equilibrated with phosphate-buffered saline (PBS) containing 1 mM EDTA. The protein concentration of BNCs is measured by a bicinchoninic acid (BCA) assay kit (Sigma-Aldrich, St Louis, MO, USA) using bovine serum albumin (BSA) as a calibration standard. Approximately 10 mg (as a protein) of BNC is obtained from recombinant yeast cells grown in 2 l of culture medium (Yamada et al., 2001).

5.2. Purification of BNCs using column chromatography The ultracentrifugation step is rate-limiting, time-consuming, and produces a low yield, so obtaining >10 mg of BNC per lot in 1 week is difficult. Based on the heat stability of BNCs (Yamada et al., 2001), a large amount of yeast crude extract can be processed immediately by heat treatment ( Jung et al., unpublished data). The crude extract of yeast cells is obtained by the method described earlier, and then dialyzed three times against PBS containing 1 mM EDTA at 4  C for 2 h to remove urea. The crude extract (about 6 g of protein) is dispensed to 40-ml plastic tubes, incubated at 70  C for 20 min, and then centrifuged at 34,780g at 4  C for 30 min to remove yeast-derived proteins. The supernatant (about 800 mg of protein) is subjected to a sulfate-cellulofine column (1.6  20 cm; Chisso Corp., Tokyo,

Bio-Nanocapsule–Liposome Conjugate

155

Japan) equilibrated with PBS containing 150 mM NaCl. After the stepwise elution of PBS containing 1 M NaCl, the fraction containing a BNC is concentrated by ultrafiltration using an Amicon Ultra 100,000 NMWL (about 120 mg of protein) and then subjected to a Sephacryl S-500 HR gel-filtration column (1.6  60 cm) equilibrated with PBS containing 1 mM EDTA. About 60 mg of the highly purified BNC is obtained from the yeast cells grown in 2 l of culture medium.

5.3. Purification of ZZ–BNC using column chromatography Porcine IgG is precipitated from porcine serum (Sigma-Aldrich) with (NH4)2SO4 (40% saturation), dissolved in PBS, and dialyzed against PBS at 4  C for 48 h. The dialyzed solution is mixed with two-times volume of 60 mM acetate buffer (pH 4.8) and then mixed with n-caprylic acid to achieve a final concentration of 6.8% (v/v). After incubation at room temperature for 30 min with gentle stirring, the supernatant containing purified IgG is obtained by brief centrifugation. Purified porcine IgG is conjugated to an NHS-activated Sepharose 4B Fast Flow (GE Healthcare) column according to manufacturer’s instructions. The IgG-conjugated Sepharose column is equilibrated with PBS. Preparation of the crude extract of yeast and heat treatment are the same as the BNC protocol. The supernatant is subjected to the porcine IgGconjugated Sepharose column (affinity chromatography) to capture ZZ– BNC specifically using the IgG–ZZ domain interaction. The column is washed extensively with 75 mM Tris–HCl (pH 7.2) containing 10 mM NaCl, and then ZZ–BNC is obtained by the stepwise elution of 10 mM Tris–HCl (pH 7.2) containing 3.5 M NaSCN, 500 mM NaCl, and 10 mM EDTA. Fractions containing ZZ–BNC are subjected to a Sephacryl S-500 HR gel-filtration column (1.6  60 cm) equilibrated with PBS containing 1 mM EDTA. About 30 mg of highly purified ZZ–BNC can be obtained from yeast cells grown in 2 l of culture medium.

6. Conjugation of BNCs with LPs The LPs used for BNC conjugation (Fig. 8.2) can be prepared by conventional methods such as solvent evaporation (Bangham and Horne, 1964), ethanol injection (Batzri and Korn, 1973), and reverse-phase evaporation (Szoka and Papahadjopoulos, 1978). The lipid composition of LPs should be optimized for the materials to be incorporated, for example, cationic lipids for the preparation of DNA- or siRNA-containing LPs (i.e., lipoplexes; Chapter 14 of volume 465, Du¨zgu¨nes¸ et al., 2002; Li and Huang, 2006), and anionic/neutral lipids for chemical compounds

156

Takeshi Kasuya et al.

containing LPs. To our knowledge, all LPs that have been employed have resulted in the formation of BNC–LP conjugates.

6.1. Example 1.1: Preparation of BNC–LP conjugates containing DNA (BNC–lipoplex conjugates) For preparation of a DNA–LP complex (lipoplex), we routinely use commercially available lyophilized cationic LPs containing O,O0 -ditetradecanoyl-N-(a-trimethylammonioacetyl)diethanolamine chloride (DC-6-14; Kikuchi et al., 1999) as the major cationic lipid, an essential component for the cellular uptake of DNA, for example, Coatsoame-EL-01-D (NOF, Tokyo, Japan) and LipoTrust series (Hokkaido System Science, Sapporo, Japan). In the case of Coatsome-EL-01-D, 1.5 mg of lyophilized LP (as lipids) is dissolved in 1 ml of 250 mg/ml luciferase (LUC) expression vector pGL3 (Promega, Madison, WI, USA) and incubated for 15 min to allow lipoplex formation. Aliquots of lipoplex (100 mg LP, 16.7 mg DNA) are mixed with freeze-dried BNC (100 mg as protein) and incubated at room temperature for 15 min. By changing the amount of LP used for 16.7 mg DNA, a series of lipoplex and BNC–lipoplex conjugates are prepared with an N/P ratio (molar ratio of nitrogen-atom content in cationic lipids to phosphorous-atom content in plasmid DNA) from 0.3 to 2.4. The z-averaged sizes and z-potentials of lipoplex and BNC–lipoplex conjugate are measured at 25  C using a Zetasizer Nano-ZS (Malvern Instruments Ltd., Worcestershire, UK). When the content of cationic lipids in lipoplex is increased, z-averaged sizes and z-potentials are apt to increase to >1 mm and >0 mV, respectively, in accordance with the increase in N/P ratio (Fig. 8.3A). BNCs are found to keep the sizes of BNC–lipoplex conjugates to <200 nm and z-potentials negative.

6.2. Example 1.2: In vitro transfection with BNC–lipoplex conjugates About 1  105 of human cells (hepatocarcinoma HepG2, colon carcinoma WiDr, cervical carcinoma HeLa) are seeded on a 24-well cell culture plate (Iwaki, Tokyo, Japan) and incubated for 24–48 h in a humidified atmosphere at 37  C in 5% (v/v) CO2. Aliquots of BNC–lipoplex conjugates containing pGL3 (N/P ratio ¼ 1.4; 100 mg LP, 16.7 mg DNA, and 100 mg BNC (as protein)) are diluted with complete cell culture medium containing 10% (v/v) fetal bovine serum and antibiotics, and applied to each well with 1 mg/ml (as protein of BNC, 0.5 mg/well) at final concentration. The medium is exchanged after incubation for 3–6 h to avoid nonspecific cellular uptake, and then incubated for 48 h to allow LUC expression.

157

Bio-Nanocapsule–Liposome Conjugate

RLU (%)

700 600 500 400 300 200 100 0 700 600 500 400 300 200 100 0

LP only

BNC-LP

0

0.5

1 1.5 N/P ratio

2

2.5

0 Zeta porential (mV)

700 600 500 400 300 200 100 0

RLU (%)

B 400 350 300 250 200 150 100 50 0

RLU (%)

Z-average (nm)

A

−10

BNC-LP

−20 −30 −40

LP only

−50 −60

0

0.5

1 1.5 N/P ratio

2

2.5

*

HepG2

HeLa

WiDr

Figure 8.3 Effect of BNC conjugation on lipoplex (DNA-containing LPs). (A) z-averaged sizes (upper panel) and z-potentials (lower panel) of BNC–lipoplex conjugates (circles with solid line) and lipoplex (squares with dashed line) at the indicated N/P ratio. (B) In vitro transfection with LUC expression plasmid by BNC–lipoplex (closed bars) and lipoplex (open bars). LUC activities in cell lysates are measured 48 h after transfection. Mean  S.D. (n ¼ 9). *p < 0.05.

After lysis with passive lysis buffer (Luciferase Assay Kit, Promega), intracellular activities of LUC can be measured by a GloMax 20/20n Luminometer (Promega) using the LUC substrate luciferin. Human liverderived HepG2 cells transfected with BNC–lipoplex conjugates containing pGL3 showed approximately five times higher LUC expression than those transfected with lipoplex containing pGL3, whereas human nonliverderived HeLa cells and WiDr cells showed no significant differences in LUC expression between BNC–lipoplex conjugates and lipoplexes (Fig. 8.3B). These data indicated that the BNC–lipoplex conjugates deliver the LUC plasmid to human liver cells specifically in vitro. This prompted us to examine if BNC–lipoplex conjugates are more suitable than conventional lipoplexes for an in vivo pinpoint gene delivery system.

158

Takeshi Kasuya et al.

6.3. Example 1.3: In vivo transfection with BNC–lipoplex conjugates Xenograft mouse models bearing tumors on their backs are generated using nude mice (5 weeks, male, BALB/c-nu/nu; CLEA Japan, Tokyo, Japan) under guidelines of the Institute of Scientific and Industrial Research, Osaka University, Japan. The cancer cell lines used are NuE and Huh-7 (human hepatocarcinoma), WiDr (human colon carcinoma), A431 (human epithelial carcinoma), PC12 (rat adrenal pheochromocytoma), and MDA-MB435 (human breast carcinoma). About 1  106 cells are mixed with 50 ml of BD Matrigel Matrix HC (BD Biosciences, Bedford, MA, USA) and subcutaneously injected into the back of mice. After 2–4 weeks, mice possessing a tumor of appropriate diameter 1 cm can be obtained. For the delivery of green fluorescent protein (GFP)-expression plasmid pEGFP-C1 (Clontech, Mountain View, CA, USA), BNC–lipoplex (N/P ratio ¼ 1.4; 100 mg LP, 16.7 mg pEGFP-C1, and 100 mg BNC (as protein)) are intravenously injected to each tumor-bearing mouse. At 1 week after injection, mice are killed, and tumors and tissues (liver, lung, spleen, brain, heart, kidney) are isolated. Tissues and tumors are embedded in plastic resin or cryomold, cut into sections of 5 mm thickness, and observed using fluorescent microscopy or confocal laser scanning microscopy. GFP expression can be observed in only human liver-derived tumors, not in other tumors and tissues ( Jung et al., 2008).

6.4. Example 2.1: Preparation of BNC–LP conjugates containing DOX Stock solutions of dipalmitoylphosphatidylcholine (DPPC, NOF), dipalmitoylphosphatidylethanolamine (DPPE, NOF), dipalmitoylphosphatidylglycerol sodium (DPPG-Na, NOF) are prepared in a chloroform/methanol (2:1, v/v) mixture at a concentration of 50 mM. A stock solution of cholesterol (Chol, NOF) is prepared in the same chloroform/methanol mixture at a concentration of 200 mM. All stock solutions are stored at 80  C and prewarmed to 60  C before use. Phospholipids and cholesterol (DPPC:DPPE:DPPG-Na:Chol ¼ 15:15:30:40 molar ratio) are dissolved in the chloroform/methanol mixture in a round-bottomed flask, and then evaporated at 60  C by utilizing a rotary evaporator to produce a thin hemispherical lipid film. The film is hydrated in buffer containing 10 mM HEPES (pH 4.0) and 120 mM (NH4)2SO4 at 60  C. The freeze–thaw cycle is repeated five times. The crude solution of LP is subjected to a Lipex extruder (Northern Lipids, Vancouver, BC, Canada) equipped with a polycarbonate filter (pore size, 200 nm) at 60  C five times, followed by seven times with a filter (pore size, 50 nm). The z-average size of LP is

Bio-Nanocapsule–Liposome Conjugate

159

measured by dynamic light scattering at 25  C using a Zetasizer Nano-ZS. The LPs are subjected to a Sephadex G-25 (GE Healthcare) gel-filtration column equilibrated with 10 mM HEPES buffer (pH 7.4) containing 100 mM NaCl and 3.4% (w/v) sucrose. Total lipid concentration is calculated from that of DPPC, which is determined by a Laboassay Phospholipid Kit (Wako, Osaka, Japan) utilizing choline oxidase and [N-ethyl-N-(2hydroxy-3-sulfopropyl)-3,5-dimethoxyaniline] (DAOS) (Takayama et al., 1977). To introduce DOX (also called adriamycin; Wako) into LPs by a remote-loading method, 200 ml of 10 mg/ml DOX–HCl solution is added to 15 mg of LP solution (as total lipids, prewarmed at 60  C), and then incubated for 20 min at 60  C with gentle stirring. After free DOX is removed by a Sephadex G-25 gel-filtration column, LPs containing DOX (about 2 mg DOX–HCl per 15 mg total lipid) can be obtained. Aliquots of LPs containing DOX (2 mg LP, 0.27 mg DOX–HCl) are gradually added to freeze-dried BNC (100 mg as protein) and incubated at room temperature for 15 min to form a BNC–LP conjugate containing DOX. The diameter of the conjugate is estimated to be 140 nm by the dynamic laser scattering method.

6.5. Example 2.2: In vitro cytotoxic effects of BNC–LP conjugates containing DOX About 5  103 cells (HepG2, Huh-7, MDA-MB-435) cultured on a 96well plate for 24–48 h are incubated with BNC–LP conjugates containing DOX at appropriate concentration (1–100 mg/ml as DOX) for 6–24 h and then incubated for 48–72 h in fresh complete medium. Cellular viability can be measured by the formation of formazan from a 3-(4,5-dimethylthiazol2-yl)-2,5-diphenyltetrazolium bromide (MTT), a tetrazole probe using a CellTiter 96 Aqueous Non-Radioactive Cell Proliferation Assay (Promega). The strong cytotoxicity of BNC–LP conjugates containing DOX, comparable to that of DOX itself, can be observed in only human liverderived HepG2 and Huh-7 cells, not in human nonliver-derived MDAMB-435 cells (Fig. 8.4A). When comparing the cytotoxicity of BNC–LP conjugates with that of LP, BNC–LP conjugates showed strong cytotoxicity only toward HepG2 cells and Huh-7 cells, whereas LP did not show severe cytotoxicity to any cells. The half maximal inhibitory concentration (IC50) of BNC–LP containing DOX for Huh-7 cells is 12 mg/ml, whereas that of LP containing DOX is 100 mg/ml. Doxil, a commercial PEGylated LP containing DOX (Markman, 2006), also showed a cytotoxicity curve similar to that of LP containing DOX. These data indicated that BNCs confer a human liver-specific active targeting function to LPs containing DOX.

A

50

Viability (%)

100 Viability (%)

Viability (%)

100

50

MDA-MB-435 0

0 10 DOX conc. (mg/ml)

1

100

10 DOX conc. (mg/ml)

100

1

10 DOX conc. (mg/ml)

100

1200

30 800

25 Body weight (g)

Tumor volume (mm3)

B

50

Huh-7

HepG2 0

100

400

20 15 10 5 0

0 0

4

8 Days

12

0

4

8

12

Days

Figure 8.4 Effect of BNC conjugation on LP containing DOX. (A) In vitro cytotoxicity curves of BNC–LP conjugates. BNC–LP conjugates containing DOX (triangles with gray lines), LP containing DOX (squares with solid lines), DOX alone (diamonds with solid lines), and doxil (crosses with dashed lines). HepG2 cells (left panel), Huh-7 cells (center panel), and MDA-MB-435 cells (right panel). Mean  SD (n ¼ 9). (B) In vivo tumor suppression by BNC–LP conjugates. BNC–LP conjugates containing DOX (triangles with gray lines), DOX alone (diamonds with solid lines), and untreated control (crosses with dashed lines). Tumor volume (left panel) and body weight (right panel). Mean  S.D. (n ¼ 9).

Bio-Nanocapsule–Liposome Conjugate

161

6.6. Example 2.3: In vivo therapeutic effects of BNC–LP conjugates containing DOX To the mouse xenograft model bearing Huh-7-derived tumors (about 1 cm in diameter), about 6 mg/kg (as DOX–HCl) of BNC–LP conjugates containing DOX is intravenously injected every 4 days under 10 ml/kg conditions. Largest (a) and smallest (b) superficial diameters of the tumor are measured every day, and tumor volume (V ) is calculated as V ¼ ab2/2. BNC–LP conjugates containing DOX efficiently suppressed tumor growth, but identical amounts of DOX injected by LP only and injected alone failed to suppress it (Fig. 8.4B). As for toxicity, mice given about 8 mg/kg (as DOX–HCl) every 4 days by a BNC–LP conjugate showed less loss of body weight than those by DOX alone. A significant change in weight was not observed in tissues (liver, spleen, kidney, heart) 12 days after the first injection (data not shown). These data indicated that BNC–LP conjugates containing DOX possess not only active targeting machinery but also less toxicity. When a mouse xenograft model bearing Huh-7-derived tumors received a single intravenous injection of BNC–LP conjugate containing DOX, LP containing DOX, doxil (PEGylated LP containing DOX), and DOX alone (8 mg/kg as DOX), blood is collected from the tail vein at 0, 30, 60, 120, and 180 min after the first injection. Plasma is collected by centrifugation at 3000 rpm at 4  C, and DOX is extracted with a five volumes of the chloroform/methanol (4:1, v/v) mixture. After evaporation of solvent with nitrogen gas, the dried DOX fraction is subjected to high-performance liquid chromatography (HPLC) equipped with an octylsilanized gel column (Nucleosil 100 5C-18, Chemco Scientific Co. Ltd., Osaka, Japan) equilibrated with a 20 mM PBS/methanol mixture (25:75, v/v) containing 40 mM sodium heptane sulfonate. Flow rate is 1 ml/min, and DOX-derived fluorescence (excitation, 475 nm; emission, 554 nm) is determined using purified DOX as the internal standard. The initial concentration of DOX (0 min) in plasma is about 100 mg/ml in all mice (Fig. 8.5). The plasma concentration of DOX after 60 min in mice injected with a BNC–LP conjugate containing DOX is much higher than those with LP containing DOX and DOX alone, which is comparable to those with doxil. BNC conjugation, like PEGylation, enhances the robustness of LP against the RES.

7. Preparation of Antibody-Displaying BNC–LP Conjugates ZZ–BNC–LP conjugates are prepared using ZZ–BNC instead of BNC by the method described earlier for in vitro and in vivo retargeting of BNC–LP conjugates according to antibody affinity. LP conjugation must

162

Takeshi Kasuya et al.

Plasma conc of DOX (mg/ml)

1000

100

10

1

0.1

0

60 120 Time after administration (min)

180

Figure 8.5 Plasma concentration of DOX. BNC–LP conjugates containing DOX (triangles with solid line), LP containing DOX (squares with solid lines), DOX alone (crosses with solid lines), and doxil (diamonds with solid lines).

be done before displaying antibodies on the surface of ZZ–BNC because antibodies displayed onto ZZ–BNC significantly reduce the fusogenic activity of ZZ–BNC (presumably by increased steric hindrance). Briefly, an aliquot of LPs (2 mg LP) containing materials of interest is gradually added to freeze-dried ZZ–BNC (100 mg as protein). We routinely add the smaller amount of antibody than ZZ–BNC (as protein) ranging from 1/5 to 1/100 (as protein) to the ZZ–BNC–LP solution. Antibodies may be spontaneously aligned on the surface of ZZ–BNC–LP conjugates by interaction of the IgG Fc domain with the ZZ domain. In the case of displaying IgG harboring weak affinity to the ZZ domain (Bjo¨rck and Kronvall, 1984), a membrane-impermeable amine-reactive cross-linking agent such as bis(sulfosuccinimidyl)suberate (BS3; Thermo Fisher Scientific, Waltham, MA, USA) was used to cross-link the binding between the IgG Fc domain and ZZ–BNC. Five millimolar BS3 dissolved in dimethylsulfoxide (DMSO) is added to the ZZ–BNC–LP displaying IgG to a concentration of 50 mM, allowed to react for 1 h, and free BS3 removed by the reaction with 0.1 mM glycine. The antibody-displaying ZZ–BNC–LP conjugate is ready for in vitro and in vivo use. To our knowledge, it is not necessary to display the antibodies as much as possible onto ZZ-BNC for maximizing the delivering ability of ZZ-BNC-LP conjugate, strongly suggesting that the amounts of antibody on ZZ-BNC should be optimized in in vitro system.

Bio-Nanocapsule–Liposome Conjugate

163

8. Preparation of Biotin-Displaying BNC–LP Conjugates For in vitro and in vivo retargeting of BNC–LP conjugates by various biorecognition molecules (e.g., cytokines, peptides, lectins, glycans), the surface of BNCs should be modified with these molecules without affecting fusogenic activity. The ZZ domain displayed on ZZ–BNC possesses many Lys residues, so the e-amino residues are modified with biotin (the smallest high-affinity tag) using sulfo-biotin-NHS-ester (Thermo Fischer Scientific) according to the manufacturer’s protocol. The number of biotins displayed on the surface of ZZ–BNC can be measured utilizing a 40 -hydroxyazobenzene-2-carboxylic acid (HABA) assay kit (Thermo Fisher Scientific). Usually, 1 mol of ZZ–BNC is modified with about 35 mol of biotin. Next, biotinylated ZZ–BNC–LP conjugate is mixed with an equimolar amount of biotinylated biorecognition molecules and avidins (e.g., avidin, streptavidin, neutravidin) to display the biorecognition molecules onto ZZ–BNC. As described in the antibody-displaying ZZ–BNC, the display of biorecognition molecules should be done after formation of the BNC–LP conjugate. The steric hindrance of biotinylated biorecognition molecules and avidins is too large for ZZ–BNC to exhibit fusogenic activity. The biorecognition molecules displaying the ZZ–BNC–LP conjugate is ready for in vitro and in vivo use. To our knowledge, it is not necessary to maximize the biorecognition molecules on ZZ–BNC to maximize the delivery capacity of the ZZ–BNC–LP conjugate, strongly suggesting that the amounts of biorecognition molecules on ZZ–BNC should be optimized in in vitro systems.

9. Concluding Remarks The studies and methods described in this chapter demonstrate that the BNC–LP conjugate, as a hybrid of a viral vector and LPs, is a promising and rational approach to achieving a pinpoint delivery system for drugs and genes in vivo. Two major problems associated with BNC–LP conjugates must be resolved before they are in clinical use. The first problem is the immunogenicity of BNC–LP. BNC was initially developed as an immunogen of a recombinant hepatitis B vaccine, so BNC per se sometimes elicits an unexpected level of immune response. Recently, we succeeded in reducing the immunogenicity by incorporating HBV escape mutants, which can proliferate in humans vaccinated with hepatitis B vaccine. The second problem is the nonspecific incorporation by the RES. Like other nanoparticle-based medicines, BNC–LP conjugates cannot fully escape from the

164

Takeshi Kasuya et al.

RES, whereas HBV can accomplish the escape. We must analyze the endogenous escape mechanism of HBV at the molecular level, and introduce it to BNC–LP conjugates.

ACKNOWLEDGMENTS The authors thank Professor M. Seno (Okayama University), Professor A. Kondo (Kobe University), and Professor M. Ueda (Keio University) for their helpful advice. We are grateful to Ms. Y. Matsushita, Ms. N. Shikaku, and Ms. Y. Matsui for their technical support. This work was supported in part by the Regional Research and Development Resources Utilization Program from the Japan Science and Technology Agency.

REFERENCES Bangham, A. D., and Horne, R. W. (1964). Negative staining of phospholipids and their structural modification by surface-active agents as observed in the electron microscope. J. Mol. Biol. 8, 660–668. Batzri, S., and Korn, E. D. (1973). Single bilayer liposomes prepared without sonication. Biochim. Biophys. Acta 298, 1015–1019. Bjo¨rck, L., and Kronvall, G. (1984). Purification and some properties of streptococcal protein G, a novel IgG-binding reagent. J. Immunol. 133, 969–974. Du¨zgu¨nes¸, N., Simo˜es, S., Pires, P., and Pedroso de Lima, M. C. (2002). Gene delivery by cationic liposome-DNA complexes. In ‘‘Polymeric Biomaterials,’’ (S. Dumitriu, ed.), pp. 943–958. Marcel Dekker, New York. Glebe, D., and Urban, S. (2007). Viral and cellular determinants involved in hepadnaviral entry. World J. Gastroenterol. 13, 22–38. Hama, S., Akita, H., Ito, R., Mizuguchi, H., Hayakawa, T., and Harashima, H. (2006). Quantitative comparison of intracellular trafficking and nuclear transcription between adenoviral and lipoplex systems. Mol. Ther. 13, 786–794. Hinnen, A., Hicks, J. B., and Fink, G. R. (1978). Transformation of yeast. Proc. Natl. Acad. Sci. USA 75, 1929–1933. Hong, R. L., Huang, C. J., Tseng, Y. L., Pang, V. F., Chen, S. T., Liu, J. J., and Chang, F. H. (1999). Direct comparison of liposomal doxorubicin with or without polyethylene glycol coating in C-26 tumor-bearing mice: Is surface coating with polyethylene glycol beneficial? Clin. Cancer Res. 5, 3645–3652. Iwasaki, Y., Ueda, M., Yamada, T., Kondo, A., Seno, M., Tanizawa, K., Kuroda, S., Sakamoto, M., and Kitajima, M. (2007). Gene therapy of liver tumors with human liver-specific nanoparticles. Cancer Gene Ther. 14, 74–81. Jung, J., Matsuzaki, T., Tatematsu, K., Okajima, T., Tanizawa, K., and Kuroda, S. (2008). Bio-nanocapsule conjugated with liposomes for in vivo pinpoint delivery of various materials. J. Control. Release 126, 255–264. Kasuya, T., and Kuroda, S. (2009). Nanoparticles for human liver-specific drug and gene delivery systems: In vitro and in vivo advances. Expert Opin. Drug Deliv. 6, 39–52. Kasuya, T., Nomura, S., Matsuzaki, T., Jung, J., Yamada, T., Tatematsu, K., Okajima, T., Tanizawa, K., and Kuroda, S. (2008a). Expression of squamous cell carcinoma antigen-1 in liver enhances the uptake of hepatitis B virus envelope-derived bio-nanocapsules in transgenic rats. FEBS J. 275, 5714–5724.

Bio-Nanocapsule–Liposome Conjugate

165

Kasuya, T., Yamada, T., Uyeda, A., Matsuzaki, T., Okajima, T., Tatematsu, K., Tanizawa, K., and Kuroda, S. (2008b). In vivo protein delivery to human liver-derived cells using hepatitis B virus envelope pre-S region. J. Biosci. Bioeng. 106, 99–102. Kasuya, T., Jung, J., Kadoya, H., Matsuzaki, T., Tatematsu, K., Okajima, T., Miyoshi, E., Tanizawa, K., and Kuroda, S. (2008c). In vivo delivery of bionanocapsules displaying Phaseolus vulgaris agglutinin-L4 isolectin to malignant tumors overexpressing N-acetylglucosaminyltransferase V. Hum. Gene Ther. 19, 887–895. Kikuchi, A., Aoki, Y., Sugaya, S., Serikawa, T., Takakuwa, K., Tanaka, K., Suzuki, N., and Kikuchi, H. (1999). Development of novel cationic liposomes for efficient gene transfer into peritoneal disseminated tumor. Hum. Gene Ther. 10, 947–955. Kobayashi, M., Asano, T., Utsunomiya, M., Itoh, Y., Fujisawa, Y., Nishimura, O., Kato, K., and Kakinuma, A. (1988). Recombinant hepatitis B virus surface antigen carrying the pre-S2 region derived from yeast: Purification and characterization. J. Biotechnol. 8, 1–22. Kurata, N., Shishido, T., Muraoka, M., Tanaka, T., Ogino, C., Fukuda, H., and Kondo, A. (2008). Specific protein delivery to target cells by antibody-displaying bionanocapsules. J. Biochem. 144, 701–707. Kuroda, S., Otaka, S., Miyazaki, T., Nakao, M., and Fujisawa, Y. (1992). Hepatitis B virus envelope L protein particles. Synthesis and assembly in Saccharomyces cerevisiae, purification and characterization. J. Biol. Chem. 267, 1953–1961. Lasic, D., and Martin, F. (eds.) (1995). In ‘‘Stealth Liposomes’’, CRC Press, Boca Raton, FL. Li, S. D., and Huang, L. (2006). Gene therapy progress and prospects: Non-viral gene therapy by systemic delivery. Gene Ther. 13, 1313–1319. Lorusso, V., Manzione, L., and Silvestris, N. (2007). Role of liposomal anthracyclines in breast cancer. Ann. Oncol. 18, 70–73. Maeda, H., Wu, J., Sawa, T., Matsumura, Y., and Hori, K. (2000). Tumor vascular permeability and the EPR effect in macromolecular therapeutics: A review. J. Control. Release 65, 271–284. Markman, M. (2006). Pegylated liposomal doxorubicin in the treatment of cancers of the breast and ovary. Expert Opin. Pharmacother. 7, 1469–1474. Marshall, E. (2002). Clinical research. Gene therapy a suspect in leukemia-like disease. Science 298, 34–35. Moghimi, S. M., and Szebeni, J. (2003). Stealth liposomes and long circulating nanoparticles: Critical issues in pharmacokinetics, opsonization and protein-binding properties. Prog. Lipid Res. 42, 463–478. Moghimi, S. M., Hunter, A. C., and Murray, J. C. (2001). Long-circulating and targetspecific nanoparticles: Theory to practice. Pharmacol. Rev. 53, 283–318. Nagaoka, T., Fukuda, T., Yoshida, S., Nishimura, H., Yu, D., Kuroda, S., Tanizawa, K., Kondo, A., Ueda, M., Yamada, H., Tada, H., and Seno, M. (2007). Characterization of bio-nanocapsule as a transfer vector targeting human hepatocyte carcinoma by disulfide linkage modification. J. Control. Release 118, 348–356. Parr, M. J., Masin, D., Cullis, P. R., and Bally, M. B. (1997). Accumulation of liposomal lipid and encapsulated doxorubicin in murine Lewis lung carcinoma: the lack of beneficial effects by coating liposomes with poly(ethylene glycol). J. Pharmacol. Exp. Ther. 280, 1319–1327. Richardson, M. (2006). AmBisome: Adds to the body of knowledge and familiarity of use. Acta Biomed. 77, 3–11. Rosenthal, E., Poizot-Martin, I., Saint-Marc, T., Spano, J. P., and Cacoub, P.DNX Study Group (2002). Phase IV study of liposomal daunorubicin (DaunoXome) in AIDS-related Kaposi sarcoma. Am. J. Clin. Oncol. 25, 57–59. Savulescu, J. (2001). Harm, ethics committees and the gene therapy death. J. Med. Ethics 27, 148–150.

166

Takeshi Kasuya et al.

Szoka, F. Jr., and Papahadjopoulos, D. (1978). Procedure for preparation of liposomes with large internal aqueous space and high capture by reverse-phase evaporation. Proc. Natl. Acad. Sci. USA 75, 4194–4198. Takayama, M., Itoh, S., Nagasaki, T., and Tanimizu, I. (1977). A new enzymatic method for determination of serum choline-containing phospholipids. Clin. Chim. Acta 79, 93–98. Tsutsui, Y., Tomizawa, K., Nagita, M., Michiue, H., Nishiki, T., Ohmori, I., Seno, M., and Matsui, H. (2007). Development of bionanocapsules targeting brain tumors. J. Control. Release 122, 159–164. Yamada, T., Iwabuki, H., Kanno, T., Tanaka, H., Kawai, T., Fukuda, H., Kondo, A., Seno, M., Tanizawa, K., and Kuroda, S. (2001). Physicochemical and immunological characterization of hepatitis B virus envelope particles exclusively consisting of the entire L (pre-S1 þ pre-S2 þ S) protein. Vaccine 19, 3154–3163. Yamada, T., Iwasaki, Y., Tada, H., Iwabuki, H., Chuah, M. K., VandenDriessche, T., Fukuda, H., Kondo, A., Ueda, M., Seno, M., Tanizawa, K., and Kuroda, S. (2003). Nanoparticles for the delivery of genes and drugs to human hepatocytes. Nat. Biotechnol. 21, 885–890. Yu, D., Amano, C., Fukuda, T., Yamada, T., Kuroda, S., Tanizawa, K., Kondo, A., Ueda, M., Yamada, H., Tada, H., and Seno, M. (2005). The specific delivery of proteins to human liver cells by engineered bio-nanocapsules. FEBS J. 272, 3651–3660.

C H A P T E R

N I N E

Nanoliposomal Dry Powder Formulations Gaurang Patel,* Mahavir Chougule,† Mandip Singh,† and Ambikanandan Misra* Contents 1. Introduction 2. Preparation of Nanoliposomal DPFs 2.1. Ternary mixture 2.2. Spray-drying 2.3. Freeze-drying 2.4. Spray freeze-drying 3. Physicochemical Characterization of NLDPFs 3.1. Assay 3.2. Flow behavior 3.3. Moisture content determination 3.4. Reconstitution time and volume 3.5. Particle size of NLDPFs 3.6. Drug retention and stability studies 3.7. Scanning electron microscopy (photomicrographs and image analysis) 3.8. Differential scanning calorimetry 3.9. In vitro lung deposition studies 3.10. In vitro drug release study 3.11. In vivo studies 4. Concluding Remarks References

168 169 170 171 172 173 174 174 174 174 175 175 175 176 176 176 179 179 187 188

Abstract Liposomal dry powder formulations (DPFs) have proven their superiority over conventional DPFs due to favorably improved pharmacokinetics and pharmacodynamics of entrapped drugs, and thus, reduced local and systemic toxicities. * {

TIFAC-CORE in NDDS, Pharmacy Department, Faculty of Technology and Engineering, The Maharaja Sayajirao University of Baroda, Kalabhavan, Vadodara, Gujarat, India Pharmaceutics Department, College of Pharmacy and Pharmaceutical Science, Florida A&M University, Tallahassee, Florida, USA

Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64009-X

#

2009 Elsevier Inc. All rights reserved.

167

168

Gaurang Patel et al.

Nanoliposomal DPFs (NLDPFs) provide stable, high aerosolization efficiency to deep lung, prolonged drug release, slow systemic dilution, and avoid macrophage uptake of encapsulated drug by carrier-based delivery of nano-range liposomes. This chapter describes methods of preparation of nanoliposomes (NLs) and NLDPFs, using various techniques, and their characterization with respect to size distribution, flow behavior, in vitro drug release profile, lung deposition, cellular uptake and cytotoxicity, and in vivo pharmacokinetics and pharmacodynamics. Some examples have been detailed for better understanding of the methods of preparation and evaluation of NLDPFs by investigators.

1. Introduction To conquer limitations of pressurized meter dose inhalers (pMDIs) and nebulizers, dry powder inhalers (DPIs) have been developed. DPIs are devices for delivering a dry powder formulation (DPF) of an active drug, including macromolecules and biotechnologicals, for local or systemic therapies (Martin and Anthony, 2005). DPFs are stable, one-phase solid blends, easy to process, very portable, patient-friendly, easy to use, do not require spacers, and have no requirement for coordination of inhalation and actuation (Ashurst et al., 2000; Geller, 2005; Hickey et al., 1994). Other benefits include the capacity to carry a high drug dose, high lung deposition (from 50% to 70%), minimal extra-pulmonary loss of drug due to low oropharyngeal deposition, low device retention, and low exhaled loss. Pulmonary disposition of drugs is associated with lung clearance by alveolar macrophages of the reticuloendothelial system (RES), irrespective of the method used for delivery. This drawback may lead to high drug dosing frequency, causing patient noncompliance (four to six times in some diseases) and increase in dose with time. Liposomal drug delivery to the respiratory tract is particularly attractive, because drugs accumulate locally in lung as the target organ. It is biocompatible (Kimelberg and Mayhew, 1978) as more than 85% of lung surfactant is composed of phospholipids. It provides controlled release of drugs for local or systemic action (Taylor and Newton, 1992), has reduced local/ systemic side effects, and thus improved therapeutic index of the drug and reduced dosing frequency, total dose, and probably cost of therapy (Courrier et al., 2002; Kimelberg and Mayhew, 1978; Konduri et al., 2005; Newman, 2004). The delivery of drugs encapsulated in liposomes in suspension form to the lung using a nebulizer or pMDIs are already under clinical investigation. In suspension form, liposomes have problems such as lipid degradation, aggregation, and fusion of liposomes resulting, in drug leakage during storage and aerosolization to lung (Hu and Rhodes, 1999). If a drug is delivered at the right site in the lung, at the required dose, at

Nanoliposomal Dry Powder Formulations

169

optimum frequency, and it stays for the required period of time, we can expect optimum therapeutic response of the drug with minimum side effects (Oku and Namba, 2005). Conventional drug DPFs fulfill the first three requisites, but fail in the fourth one. Liposomal drugs enhance the drug residence time in the lungs, prevents enzymatic degradation of the drug, and the nano-size prevents its rapid removal through the clearance mechanism, alleviating the limitations of plain drug or conventional liposomal drug. Drug encapsulated nanoliposomes (NLs) can be processed into DPF form using freeze-drying, spray-drying, and spray freeze-drying to achieve long-term stability, and overcome problems associated with the suspension form of liposomes (Courrier et al., 2002; Niven, 1995).

2. Preparation of Nanoliposomal DPFs Conventional DPFs composed of saccharide carriers containing surface adsorbed micronized drug and having particle diameter in between 3 and 5 mm (Hersey, 1975). The particle size is restricted to 3–5 mm as below 3 mm particles are engulfed by macrophages, and above 5 mm particles show extensive oropharyngeal deposition, which is often associated with widespread systemic side effects (Bisgaard, 1996; Byron and Patton, 1994; Hickey, 1992; Newman and Clarke, 1983). Desirable characteristics of ideal DPFs include enhanced powder performance, higher lung deposition, good flow properties, and efficient delivery of a wide variety of therapeutic agents. Because of including controlled drug release, biocompatibility, biodegradability, and nontoxic nature, liposomes have been extensively researched in pulmonary drug delivery (Shek and Barber, 1986). Formulation flexibility makes liposomal drug delivery more versatile than other lipid-based delivery systems. Methods such as the Bangham method, the organic solvent injection method, and the reverse-phase evaporation method (Betagiri et al., 1993a; New, 1990a) have been commonly used for the preparation of liposomes. Liposomes having a nano-size range can be obtained by passing the liposomal suspension through a high-pressure homogenizer (10,000–25,000 psi) or extrusion through polycarbonate membranes having a pore size of less than about 100 nm at 300–400 bars pressure, and 2–10 cycles above the glass transition temperature of respective lipids of liposomal membrane. Liposomes, lipid- and polymer-based nanoparticles and nanosuspensions can also be effectively formulated using the supercritical fluid technique (SCF) (Kompella and Koushik, 2001; Sunkara and Kompella, 1999). SCF technology offers advantages over the other techniques, such as narrow size distribution, uniform morphology, and assurance of complete removal of organic solvent (Kadimi et al., 2007; Misra et al., 2008). Fixed amounts of drug, phospholipid, and cholesterol are dissolved

170

Gaurang Patel et al.

in an organic solvent; the solution thus formed is loaded into a high-pressure reactor maintained below the glass transition temperature and optimized pressure (2200 psi) of supercritical fluid. The sample solution is atomized in the reaction vessel, allowing the supercritical fluid to extract the organic solvent and leave behind nano-sized liposomes (Otake et al., 2001). According to the physicochemical properties of the drug to be encapsulated into liposomes, the method of preparing liposomes is selected and the formulation and process variables are optimized to maximize the percentage of drug entrapment, and to achieve the desired vesicle size and size distribution, lamellarity, and stability. Prepared NLs are separated from unentrapped drug by centrifugation or dialysis. Unentrapped hydrophobic drug is separated at low speed (4000 rpm), and for hydrophilic drug, liposomes are separated by centrifugation at high speed (35,000 rpm). NL pellets are redispersed in aqueous medium. Required quantities of carriers (lactose, sucrose, mannitol, etc.), cryoprotectants (trehalose, mannitol, etc.), and antiadherents (glycine, L-leucine, serine, etc.) are added into drug incorporated nanoliposomal suspension and water is removed by either spray-drying, freeze-drying, or spray freeze-drying to obtain nanoliposomal DPFs (NLDPFs) (Fig. 9.1). Details of the methods for converting NL suspensions in DPF suitable for inhalation are elaborated with examples in this chapter.

2.1. Ternary mixture Freeze-dried NLDPFs are either subjected to jet milling or sieved successively through sieves # 100, 200, 350, and 500. The micronized powder is mixed with small quantity of coarse carrier such as lactose, antiadherent, and lubricant as a ternary component to get NLDPFs having improved flow behavior and enhanced lung deposition. Amphotericin B liposomes are prepared from a mixture of hydrogenated soyphosphatidylcholine; cholesterol and either saturated soyphosphatidylglycerol (7:3:0.5) for negative charge or stearylamine (1:1:0.1) using reverse-phase evaporation technique. The soyphosphatidylglycerol and stearylamine are used to prepare negatively and positively charge liposomes, respectively. Liposomes are extruded from 2 mm polycarbonate membrane to reduce size, separated from unentrapped drug and are lyophilized using sucrose as cryoprotectant. Coarse lactose (Sorbolac 400) in varying mass ratio, with or without addition of fines (500# sieved Pharmatose 325M), is used in different mixing sequence to formulate amphotericin B NLDPFs. We observed that the 10% fine with coarse lactose is found to be optimum blend at 1:6 mass ratio of liposome:lactose. The mixing sequence of the three components, that is, liposome powder, coarse carrier, and fines are also found to influence the fine particle fraction. The maximum fine particle fraction in the range of 17–28% is observed using a Rotahaler (Cipla, India)

171

Nanoliposomal Dry Powder Formulations

A

B Formulation optimization of nanoliposomes

Formulation optimization of NLDPFs

Method of preparation Thin film hydration Reverse phase evaporation Spray drying Supercritical fluid technology

Method of preparation Physical mixing of dry liposomes with carriers Freeze drying Spray drying Spray-freeze drying Supercritical fluid technology

Optimization of process parameters Temperature and time for drying of film Hydration time Speed of rotation Sonication parameters

Optimization of process parameters Total freezing time Total drying time Spray drying parameters

Optimization of formulation parameters Total drug: lipid ratio Phospholipid: cholesterol ration ratio Amount and type of charged lipids Amount and type of hydration medium

Physicochemical characterization Percentage drug entrapment Particle size and size distribution Zeta-potential Morphology Entrapped volume of aqueous phase Oxidative index

Optimization of formulation parameters Type and amount of cryoprotectant Type and amount of carrier Type and amount of antiadherent

Physicochemical characterization Flow behavior Moisture content Reconstitution time and volume Particle size of reconstituted vesicles Drug retention Stability study Morphology In vitro lung deposition study In vivo study

Figure 9.1 (A) Detailed flow chart describing formulation optimization of nanoliposomes and (B) detailed flow chart describing formulation optimization of NLDPFs.

as a delivery device at 30, 60, and 90 l/min. Although amphotericin NLDPFs are prepared from micron size liposomes and same process can be applied for NLs to process into NLDPFs (Shah and Misra, 2004b).

2.2. Spray-drying The spray-drying method to prepare NLDPFs offers many advantages, including better control over particle formation, and hence can be easily translated to large-scale production (Woolfe et al., 2001). During spraydrying, materials come into contact with hot air or gas for a fraction of a second, otherwise actual drying takes place at ambient temperature, and that

172

Gaurang Patel et al.

is why the spray-drying technique is used for the preparation of NLDPFs composed of less temperature-sensitive materials (Bosquillona et al., 2004). Respirable properties of spray-dried NLDPFs containing superoxide dismutase (Briscoe et al., 1995), ciprofloxacin (Sweeney et al., 2005), and antiasthmatic drugs (Patel and Misra, 2007; Sunkara and Kompella, 1999) have been evaluated. In our laboratory, we have developed spray-dried liposomal powder formulations of tacrolimus (Chougule et al., 2007), dapsone (Chougule et al., 2008), and amiloride hydrochloride (AH) (Chougule et al., 2006) with good inhalation properties. Physical properties such as particle size and size distribution, surface energy, surface rugosity, particle density, surface area and microviscosity greatly depend on processing and formulation variables (Chougule et al., 2006, 2008; Parmar, 2006). By controlling such parameters we can get DPI formulation having desired properties for different classes of drugs and widen the application window of pulmonary delivery. AH-incorporated multilamellar vesicles are prepared by thin film hydration. Two hundred milligrams AH, hydrogenated soyaphosphatidylcholine and cholesterol (8:2) are dissolved in CHCl3:CH3OH (1:1) and dried as a thin film under 400 mmHg vacuum and at 65
C temperature. The film is then hydrated at 65
C for 1 h with 50 ml of phosphate-buffered saline (PBS), pH 7.4. Hydration is followed by size reduction by passing through a high-pressure homogenizer (Emulsiflex-C5, Avestin Inc., Ottawa, Canada) preheated to 65
C using a thermostat for three cycles at 10,000 psi. Resultant NLs are subjected to centrifugation at 35,000 rpm, 20
C for 20 min using ultracentrifuge (Sigma Laboratory centrifuge, 3 K 30, Osterode, GmbH). Liposomal pellets are separated, resuspended, and characterized for vesicle size and percent drug entrapment. The liposome pellets equivalent to 50 mg of AH are dispersed in 200 ml PBS containing 15 mg/ml additive (lactose/sucrose/mannitol), and 10% of glycine, based on the weight of the powder at room temperature. For spray-drying, a laboratory spray-dryer (LSD-48, JISL, Mumbai, India) is used. Inlet and outlet temperature are kept at 120  5 and 65–70
C, respectively. The yield of the product is found to be 60  5%. The spray-dried product is filled in capsules (size 2) individually by weight of powder equivalent to 500 mg of AH. HDPE bottles containing capsules and silica bags are placed in a desiccator kept in a refrigerator, after purging with nitrogen and covering with PVC-coated aluminum foil until evaluation (Chougule et al., 2006).

2.3. Freeze-drying NLDPFs of highly heat-sensitive materials such as proteins, peptides, and enzymes can be prepared using freeze-drying method. NLDPFs with different aerodynamic properties can be prepared using different proportions of carriers,

Nanoliposomal Dry Powder Formulations

173

cryoprotectants, and antiadherents (Lu and Hickey, 2005; Shah and Misra, 2004a; Shahiwala and Misra, 2005). Freeze-dried LDPFs for pulmonary administration of therapeutic molecules such as budesonide, ketotifen, amphotericin B, leuprolide acetate (LA), and levonorgestral have been formulated ( Joshi and Misra, 2001a, 2003; Shah and Misra, 2004a; Shahiwala and Misra, 2004, 2005). The method is explained using leuprolide DPF as an example. The reverse-phase evaporation method is employed to prepare LA-loaded liposomes. LA, hydrogenated soyaphosphatidylcholine, and cholesterol, at 2:4:1 molar ratio, are dissolved in chlorform:methanol:acetate buffer, pH 5.2 (2:4:1) in a quick fit glass tube. After vortexing the tube for 5 min, it is attached to a rotary evaporator (Superfit Continental Pvt Ltd, Mumbai, India) to dry the contents at 55
C under vacuum (400 mmHg) until a gel formation. Again the tube is agitated on vortex mixture for 5 min to collapse the gel into fluid. Ten minutes rotary drying and 5 min vortexing is again repeated twice. The liposomal dispersion is extruded by passing through polycarbonate membranes (below 100 nm). After addition of sucrose in liposomal dispersion, it is sufficiently diluted to obtain a lipid:sugar (sucrose) ratio of 1:6. The dispersion is frozen at  40
C overnight and dried for 24 h under negative displacement pressure (model DW1 0–60E; Heto Drywinner, Birkerod, Denmark). The obtained porous cake is passed through #200 and #240 sieves. To get the final strength of 250 mg LA/10 mg of powder, equivalent amount of Sorbolac 400 is geometrically added. The DPF is filled in capsules (size 2) individually by weight of powder equivalent to 250 mg of LA. HDPE bottles containing capsules and silica bags are placed in a desiccator kept in a refrigerator after purging with nitrogen and covering with PVC-coated aluminum foil until evaluation (Shahiwala and Misra, 2005).

2.4. Spray freeze-drying Spray freeze-drying involves spraying the drug solution into a freezing medium (cryogen) followed by lyophilization of the suspension. Advantages of spray freeze-drying involves almost 100% yield, production of light and porous particles with enhanced aerosol performance, and minimum chances of aggregation of particles. Besides these advantages, disadvantages of the spray freeze-drying process include the expense and the time required (Sweeney et al., 2005). Dimyristoylphosphatidylglycerol, lactose, and ciprofloxacin in a weight percent ratio 5:17:1, respectively, form a smooth suspension upon vortexing (2  30 s to 4  30 s in 1 h), and the suspension remains stable at 4
C for several days. Spray freeze-drying is performed with a two-fluid nozzle, in which compressed nitrogen and a peristaltic pump are used to drive the suspension in fine atomized form into a flask containing liquid nitrogen. The cryogen is then allowed to evaporate, and the resulting powder is dried for 48 h in a lyophilizer. The collector is held at 52
C while the vacuum is 0.004 mbar. The powder is subsequently collected and stored in a sealed vial at 4
C (Sweeney et al., 2005).

174

Gaurang Patel et al.

3. Physicochemical Characterization of NLDPFs Prepared liposomes are characterized for percent drug entrapment, size and size distribution, zeta potential, entrapped volume, morphology, oxidative index, and differential scanning calorimetry (DSC) (Akwete and Hsu, 1993; Betagiri et al., 1993b;Chougule et al., 2006, 2007; Goel, 1988; Hyung et al., 2008; Joshi and Misra, 2001b; Kaszuba, 1999; Kent et al., 2007; Maria et al., 2004; Markus et al., 2004; Mina et al., 2002; New, 1990b; Shah and Misra, 2004a; Vyas et al., 2005), and NLDPFs are evaluated for following parameters.

3.1. Assay The amount of drug present in a fixed quantity of DPF can be obtained by dissolving the required quantity of DPF in methanol or a methanol– chloroform mixture. The resultant solution is filtered to remove the sugar part. The amount of drug in the filtrate is estimated by a suitable analytical method (Chougule et al., 2006).

3.2. Flow behavior The angle of repose is measured to determine the flow properties of a powder. The compressibility index reflects the percent voids in the powder along with the flow and dispersibility properties of formulation. Bulk and tapped density of the powder is determined for the calculation of the compressibility index and the theoretical mean mass aerodynamic diameter (MMAD) of the powder. Dispersibility index is the dispersing ability of agglomerated particles on application of pressure equivalent to inhalation air pressure. A higher dispersibility index is expected to get a higher lung deposition (Carr, 1965; Chougule et al., 2006).

3.3. Moisture content determination Stability and flow properties mainly depend on the amount of moisture present in the NLDPFs. A higher amount of moisture causes powder aggregation, leading to poor lung deposition. The moisture content of LDPF is accurately determined by the Karl Fischer titration method for abrupt idea loss on drying (LOD) moisture balance can be used (Chougule et al., 2006). Solid-state characterization and residual water content data for AH NLDPFs are tabulated in Table 9.1.

175

Nanoliposomal Dry Powder Formulations

Table 9.1 Solid-state characterization and residual water content of amiloride hydrochloride nanoliposomal dry powder formulations

Formulations

AHL AHNLDPFlactose AHNLDPFmannitol AHNLDPFsucrose

Angle of repose (
)

Carr’s compressibility index

Residual water content (%)

0.9  0.1 0.3  0.05

45.6  2.4 31.1  1.9

25.8  1.4 35.8  2.1

5.9  0.6 4.4  0.5

0.2  0.01

27.3  1.5

39.8  1.5

2.5  0.4

0.36  0.04

30.8  1.5

33.4  1.9

5.7  0.8

Tapped density (g/ml)

Reprinted with permission from Chougule et al. (2006). Copyright # American Scientific Publishers, www.aspbs.com.

3.4. Reconstitution time and volume Time and volume of phosphate buffer (pH 7.4) required to reconstitute a fixed quantity of NLDPF is measured (Chougule et al., 2007).

3.5. Particle size of NLDPFs Instruments based on laser light scattering and the photon correlation principle are mainly used to determine the particle size of NLDPFs (Chougule et al., 2006; Goel, 1988; Kaszuba, 1999). Figure 9.2 shows the particle size distribution graph of AH NLDPFs.

3.6. Drug retention and stability studies As per ICH guidelines, comparative drug retention (PDR) studies are performed on the NLDPFs for long-term stability at 5  3
C for 12 months, and accelerated testing at 25  2
C/60% RH  5% RH for 6 months (ICH Guidelines, 2003; Singh, 1999). The product in its final packing is stored at the above-mentioned storage conditions. Samples are withdrawn at definite time intervals and evaluated for assay, water content, PDR, fine particulate fraction, and emission and physical changes (Chougule et al., 2007; Singh, 1999). Data of drug retention and stability of tacrolimus NLDPFs are recorded in Table 9.2.

176

Gaurang Patel et al.

Particle size distribution

Volume (%)

12 10 8 6 4 2 0 0.01

0.1

1

10 Particle size (mm)

100

1000

3000

Figure 9.2 Particle size distribution of amiloride hydrochloride NLDPFs. (Reprinted with permission from Chougule et al. (2006). Copyright # American Scientific Publishers, www.aspbs.com.)

3.7. Scanning electron microscopy (photomicrographs and image analysis) Prepared NLDPFs are evaluated for surface morphology, size, shape, and aggregation behavior by using scanning electron microscopy (SEM) (Chougule et al., 2008). Scanning electron microphotograph and b surface texture analysis of dapsone NLDPF are depicted in Fig. 9.3. Topographical features of dapsone NLDPFs measured by SEM image analysis are tabulated in Table 9.3.

3.8. Differential scanning calorimetry Prepared NLDPFs are evaluated for thermal properties using DSC (Chougule et al., 2006). DSC curves of AH NLDPFs are contained in Fig. 9.4.

3.9. In vitro lung deposition studies Anderson cascade impactor (official in USP) and multiple-stage liquid impinger (official in BP and EP) resembles human respiratory track and thus used to evaluate target region within the lung and deposition of inhaled NLDPFs in each region of lung. From that data the emitted dose, fine particle dose, fine particle fraction, MMAD, and geometric standard deviation (GSD) can be calculated according to the procedure given in USP (USP 30 NF 25, 2007). The other parameters to be estimated include recoverable dose, emitted dose, percent emission, and percent dispersibility.

Table 9.2 Stability data of tacrolimus nanoliposomal dry powder formulations

Stability conditions

Initial

40  1M

2M

3M

6M

30  1M

2M

Description

Assay (%)

101.16  1.86 White free flowing powder 2
C and 75  5% RH 101.08  3.24 White free flowing powder 99.43  2.97 White free flowing powder 98.09  3.95 White free flowing powder 99.72  3.54 White free flowing powder
2 C and 65  5%RH 103.46  2.72 White free flowing powder 102.16  2.91 White free flowing powder

Water content

Percent drug retained

Liposomal size (VMD) (mm)

Emission (%)

Fine particle fraction (FPF)

2.8  0.3

101.38  2.47

0.140  0.02

82  3.0

71.1  2.5

2.7  0.5

95.92  3.51

0.145  0.03

80.6  2.2

70.45  2.7

3.3  0.5

91.45  3.95

0.152  0.03

78.4  3.0

67.13  2.5

3.4  0.6

87.43  3.91

0.158  0.02

77.2  2.9

62.91  1.9

3.0  0.7

85.71  3.98

0.161  0.03

75.4  2.6

60.11  3.0

2.8  0.4

100.26  3.29

0.144  0.03

81.87  3.7

70.93  2.1

2.9  0.5

101.49  3.96

0.138  0.03

81.46  3.4

71.08  2.4

(continued)

Table 9.2 (continued) Stability conditions

Description

Assay (%)

101.13  2.58 White free flowing powder 100.18  3.1 6M White free flowing powder 99.42  2.9 9M White free flowing powder 100.51  3.59 12 M White free flowing powder
25  2 C and 60  5% RH 102.19  2.53 3M White free flowing powder 101.17  3.5 6M White free flowing powder 101.83  3.92 9M White free flowing powder 102.81  4.59 12 M White free flowing powder 3M

Water content

Percent drug retained

Liposomal size (VMD) (mm)

Emission (%)

Fine particle fraction (FPF)

2.9  0.4

100.17  3.21

0.148  0.02

81.4  3.9

70.84  2.3

3.0  0.4

102.41  4.23

0.151  0.03

80.19  3.2

68.42  2.9

2.9  0.5

99.59  3.91

0.153  0.04

80.07  3.0

67.82  2.8

2.9  0.4

99.14  4.1

0.152  0.03

78.76  3.5

65.19  2.8

2.8  0.3

101.80  2.95

0.138  0.03

82.4  2.6

71.20  2.7

2.8  0.4

102.11  3.44

0.143  0.02

82.1  3.1

70.12  2.5

2.9  0.5

101.74  4.28

0.148  0.02

80.74  2.7

68.17  2.9

2.9  0.6

101.62  4.38

0.143  0.02

79.13  3.4

68.28  2.8

Reprinted with permission from Chougule et al. (2008). Copyright # Dove Medical Press.

179

Nanoliposomal Dry Powder Formulations

A

B

Figure 9.3 Scanning electron micrograph (A) and surface texture analysis (B) of dapsone NLDPF. (With kind permission from Springer Science þ Business Media: Chougule et al. (2008).

NLDPF density and geometric diameter are the key parameter which decides MMAD of NLDPF and thus lung deposition and effectiveness of the therapeutic delivery system (Chougule et al., 2007). Tacrolimus NLDPFs were evaluated for in vitro pulmonary deposition using an eight stage Andersen cascade impactor with a preseparator (Graseby-Andersen, Atlanta, GA, USA) and data obtained are tabulated in Table 9.4.

3.10. In vitro drug release study In vitro drug release study is a tool that provides an estimate of the in vivo pharmacokinetic and pharmacodynamic performance of the formulation (Fielding and Abra, 1992). Therefore an in vitro release technique is proposed, validated, and utilized for drug release studies from optimized NLDPFs. To examine the drug release kinetics and mechanism, the percent drug diffused (Shah et al., 1993), the kinetics of release, the mean steadystate flux (Chien, 1992; Robinson and Lee, 1987), and the diffusion coefficient (Chougule et al., 2006) are measured. The curve of percent drug released versus time for AH NLDPFs is graphically presented in Fig. 9.5.

3.11. In vivo studies In vivo studies are to be performed to evaluate lung absorption and the lung deposition pattern. For this purpose, the following three approaches have been utilized.

Table 9.3 Topographical features of dapsone NLDPFs measured by SEM image analysis Formulations

Roundness

Aspect ratio

Fractal dimension

Heterogeneity

Clumpiness

NLDPF-lactose NLDPF-trehalose NLDPF-sucrose

1.106  0.076 1.015  0.065 1.376  0.092

1.356  0.210 1.163  0.196 1.453  0.209

1.129  0.064 0.856  0.052 1.341  0.064

0.492  0.057 0.351  0.051 0.674  0.153

0.194  0.057 0.109  0.047 0.274  0.095

With kind permission from Springer Science þ Business Media: Chougule et al. (2008).

181

Nanoliposomal Dry Powder Formulations

40.00 a b c

DSC mW

20.00

0.00

-20.00 -40.00

200.00

100.00

300.00

Temp (C)

Figure 9.4 Differential scanning calorimetric graphs of developed amiloride hydrochloride NLDPF (a), plain drug (b), and physical mixture (c). (Reprinted with permission from Chougule et al. (2006). Copyright # American Scientific Publishers, www. aspbs.com.) Table 9.4 In vitro aerosol deposition data of tacrolimus nanoliposomal dry powder formulations

Formulations

NLDPFtrehalose NLDPFsucrose NLDPFlactose

Emitted dose (%)

Fine particle fraction (%)

Mean median aerodynamic diameter (mm)

Geometric standard deviation

82  3.0

71.1  2.5

2.2  0.1

1.7  0.2

63  5.5

53.7  3.6

2.8  0.2

2.3  0.1

74  4

62.8  3.1

2.6  0.2

2.2  0.1

Reprinted with permission from Chougule et al. (2008). Copyright # Dove Medical Press.

3.11.1. Animal models (in vivo) Evaluation of pharmacokinetic behavior and distribution in different sections of the lungs are carried out using different animal models, including rats and guinea pigs. Methods such as intratracheal instillation, micro spray, nebulization, and aerosol puff are used for dosing of NLDPFs in animals. Drug present in homogenate of various lobes of the lung and in bronchoalveolar lavage is estimated at various time intervals to get the idea of amount of drug deposition as well as pattern of distribution within the lungs. The amount of drug present in blood and homogenates of other major organs is estimated to assess systemic absorption after inhalation of NLDPFs. Validated analytical methods are developed in our laboratory to

182

% Drug release

Gaurang Patel et al.

100 90 80 70 60 50 40 30 20 10 0 0

2

SAHL

4

6 Time (h)

SLDPIM

8

SLDPIL

10

12

SLDPIS

Figure 9.5 In vitro release profile of developed amiloride hydrochloride NLDPFs and plan drug DPFs (SAHL ¼ plain drug formulation, SLDPIM ¼ NLDPF containing mannitol, SLDPIL ¼ NLDPF containing lactose, and SLDPFs ¼ NLDPF containing sucrose). (Reprinted with permission from Chougule et al. (2006). Copyright # American Scientific Publishers, www.aspbs.com.)

estimate the pharmacokinetic profiles of dapsone, rifampicin, amikacin, ketotifen, tacrolimus, AH, isoniazid, and amphotericin B in rats (Chougule et al., 2006, 2007, 2008; Joshi and Misra, 2001a,c, 2003; Shah and Misra, 2004a,c). Serum LH levels are measured to estimate the pharmacodynamic response after intratracheal administration of leuprolide and levonorgestral liposomes (Shahiwala and Misra, 2004, 2005). However, these methods require sacrifice of animals, and hence other techniques, including in vitro cell lines and ex vivo techniques, may also be used (Chougule et al., 2006). AH-loaded NLDPFs are evaluated for in vivo performance following intratracheal instillation. For each time interval, six albino rats (220–240 g) of either sex are housed in individual plastic cages at a constant temperature. Rats are fasted overnight prior to each experiment. After anesthetizing rats by intraperitoneal administration of a urethane solution (1.2 gm/kg), the trachea is exposed by blunt dissection of the sternohyoideus muscle and a small midline incision is made over the trachea between the fifth and sixth tracheal rings using a 20-gauge needle. The trachea is cannulated with PE200 tubing (5–7 cm) with the tip positioned approximately at the tracheal bifurcation. The PE50 (10– 15 cm) tubing connected to a glass Hamilton syringe (Waters, Bangalore, India) is inserted into the cannula and advanced to the bifurcation of the trachea. Prepared formulations equivalent to 100 mg of AH are administered, and animals are sacrificed at 2, 4, 8, 10, and 12 h after administration (Fig. 9.6). Bronchoalveolar lavage is performed on anesthetized and recannulated animals with 12 ml PBS, prewarmed to 37
C. To perform the lavage, the Hamilton syringe connected to the PE50 tubing is replaced with a three-way stopcock attached to two 20 ml syringes. Approximately 12 ml

Nanoliposomal Dry Powder Formulations

Figure 9.6

183

Intratracheal instillations and bronchoalveolar lavage on rat.

sterile (0.22 mm filtered) PBS is injected slowly in fractions to fill the lungs. The fluid is withdrawn by gentle aspiration; this bronchial alveolar lavage yields between 7 and 11 ml liquid, which is centrifuged at 2000 rpm for 5 min. One percent Triton X-100 is added to the supernatant and analyzed to determine unreleased AH. The lungs and portions of the trachea below the instillation site are excised and homogenized in 10 ml PBS containing 1% Triton X-100. Sulphosalicylic acid (10%) is used as a deproteinizing agent. AH released is analyzed in the supernatant after suitable dilution (Chougule et al., 2006). 3.11.2. Lung epithelial cell models (in vitro) In vitro cell line models have become more popular than in vivo models to evaluate prepared formulations for cell uptake, cell toxicity, and cell death kinetics studies (Anabousi et al., 2006; Elbert et al., 1999; Fang et al., 2004; Fuchs et al., 2003; Pang et al., 2005; Sakagami et al., 2002a,b). Intact lung studies have two limitations: (i) the contribution of different distal lung epithelial cells can not be studied separately, and (ii) the surface area for fluid absorption can only be approximated. Lung epithelial cell lines include tracheobronchial, bronchial, and alveolar cell lines and they are mainly of rat and human origin. The examples of cancer-derived cell lines include Calu-1, Calu-3, Calu-6, H441, HBE1, A427, A549, 16HBE140, and BEAS-2B. The examples of cell lines transformed from normal lung include 9HTE16o-, 16HBE14o-1HAEo-, BEAS-2BCF/T43, and AK-D. These cell lines have been tested for the change in transepithelial transport kinetics (i.e., Papp) and for predicting in vivo lung absorption (Tronde et al., 2003).

184

Gaurang Patel et al.

An in vitro cell culture model that mimics the respiratory epithelium, in particular cells that form tight monolayers and secrete components of mucus or surfactant, would be a valuable tool for NLDPFs characterization. The air-interfaced culture (AIC) or liquid-covered culture (LCC)-grown Calu3 monolayers can be used as an in vitro model to evaluate therapeutic large porous powder particles (Fiegel et al., 2003). Similar experiments can be designed for evaluation of NLDPFs. Briefly, Calu-3 cells are seeded onto Transwell clear permeable filter inserts at a density of 105 cells/cm2. After seeding, cells are grown in 1500 ml (apical) and 2600 ml (basolateral) of cell culture media at 37
C in a 5% CO2 incubator. For LCC conditions, the media bathing the cells are removed and cells are fed with 1500 ml fresh apical and 2600 ml fresh basolateral media 1 day after seeding, and then once daily. For AIC conditions, the apical medium of cells are removed 1 day after seeding and cells are fed daily with 1400 ml fresh basolateral media only after. When the cells reach sufficient confluency, the apical fluid is removed from the LCC-grown Calu-3 monolayers, and the basolateral media is reduced to 1400 ml. AIC-grown monolayers are used as a control. Filter inserts containing cell monolayers are transferred from the 6-well plate to a cut-out single well containing prewarmed media. One well is placed directly under the second-stage nozzle of an Astra-type liquid impinger (Erweka, Heusenstamm, Germany) and the impinger is sealed, as illustrated in Fig. 9.7. Dry powder particles are aerosolized onto the monolayers for 30 s at 30 l/min from #2 gelatin capsules, using a Spinhaler device. The filter inserts are then returned to the 6-well plate and placed back into an incubator at 37
C. Slightly wetted Transwell inserts without cells are placed in the impinger and treated as described earlier to obtain reproducible data. After particle impinging, 0.8 N NaOH is added to the apical and basolateral compartments of the filter to dissolve the dry powder. Following a 24 h incubation at 37
C, the complete degradation of the particles on the filter inserts are verified by light microscopy. Samples are taken and analyzed for fluorescein-sodium using a fluorescence plate reader (at excitation and emission wavelengths of 485 and 530 nm, respectively; Fiegel et al., 2003). 3.11.3. Lung tissue models (ex vivo) Ex vivo lung tissue models have been preferred when the information on the mechanisms involved in drug transport or lung disposition kinetics are not predictable from in vivo or in vitro models. The isolated perfused lung (IPL) model is most widely used for the purpose of establishing the mechanisms of drug absorption and deposition in the lungs. In this type of model, the lung are isolated from the body and placed in a simulated environment to perform the desired experiment. IPL prepared from small rodents have been used for drug deposition and organ toxicity studies (Pang et al., 2005; Sakagami et al., 2002a,b; Tronde et al., 2003). The limitations of IPL models includes. The only

185

Nanoliposomal Dry Powder Formulations

Inhaler device

Transwell containing cell monolayer

Outlet to vacuum

Figure 9.7 Astra-type liquid cascade impinger used with a Transwell under the second-stage nozzle. (With kind permission from Springer Science þ Business Media: Fiegel et al. (2003)).

drawbacks associated with these models are the short viable periods of (2–3 h), and the absence of blood and mucous circulation. The IPL preparation has been employed mostly with rats and guinea pigs rather than with mice, dogs and monkeys. The preparation consists of (a) peristaltic pumps and a tubing assembly that is used to carry the perfusate to and from the lungs and (b) a double-jacketed artificial thorax to house the IPL at a certain temperature, normally at 37
C. A flow rate of 12–15 ml/ min (rat lung) is used to pump the perfusate from the reservoir to the IPL through a central porthole of the thorax is lid. The opening at the bottom of the thorax is used to return the perfusate to the reservoir or to collection. The air-tight sealed thorax enables negative- or positive-pressure ventilation, or the maintenance of a certain pressure. Under anesthesia, the pulmonary circulation of the lung is cannulated via the pulmonary artery and lung perfusion is initiated and maintained. The pulmonary vein is also cannulated as an exit of the lung perfusion, or as an alternative, the bottom of the left auricle and both sides of the ventricles in the heart are amputated to allow the free flow of perfusate without any unnecessary back pressure in

186

Gaurang Patel et al.

A

Dosing

95%O2 + 5%CO2 Thermometer pH probe

Perfusate flow

Pump

Perfusate reservoir

Isolated lung

B Air Fluid dispenser

Temperature control pH-stat

Stirror

CO2100

Drug reservoir

Perfusion buffer

Pressurized air

Air

Artery Tracheal airflow

Air volume control unit

Nebulization catheter

Pulmonary pressure

Dose trigging

Lung mechanics

Respirator Foot switch Rat lung

Perfusion flow motor

Vein

Perfusion flow probe

Monitored Controlled

Figure 9.8 Isolated perfused rat lung (IPRL) preparations of (A) Horizontally positioned IPRL with a scheme of forced solution instillation and (B) vertically positioned IPRL with a scheme of nebulization catheter dosing. (Reprinted from Sakagami et al. (2006). Copyright 2006, with permission from Elsevier.)

the IPL. The lung and heart are excised from the body en bloc. The trachea is also cannulated and often exteriorized out of the thorax. This setup supports the IPL suspension in the thorax, and for pulmonary biopharmaceutics research, allows airway lumienal administration of test molecules. The IPL can be suspended in a horizontal or vertical fashion, as shown in Fig. 9.8. Autologous whole blood or a buffered artificial medium, such as Krebs– Ringer or Krebs–Henseleit buffer (pH 7.4) supplemented with 5 mM glucose and/or 4–5% albumin can be used as the perfusate. The perfusate medium is equilibrated with a mixture of oxygen and carbon dioxide

Nanoliposomal Dry Powder Formulations

187

throughout the experiments. Generally, the perfusion is recirculated, but can also be a single-pass system. Using sufficient surgical technique, it has been shown that the IPL viability can be maintained for a maximum of 2–3 h at 37
C, after which the lung gradually develops translucent areas, indicating edema formation (Sakagami et al., 2006). After the perfusion begins, the perfusate samples are with drawn via the tracheal port at different time intervals, enabling the characterization of the absorption profiles. The rat IPL is used for administration of different sizes of powder aerosols with the help of the exteriorized tracheal cannula port. The IPL is suspended in a physiologically natural horizontal fashion, with the help of a stainless rod through the esophagus (Fig. 9.8A). Dry powder aerosol of fluorescein is administered to the IPL via negative pressure ventilations created in the artificial thorax. The fractional fluorescein transfer to the perfusate is correlated to the depth of aerosol penetration into the lung. However, the inhalation exposures have been limited to lower aerosol concentrations, and thus necessitate an extended time of administration (Byron et al., 1986). Most existing technologies for suspending cohesive powders are continuous generators that consume large amount of powder. Examples of such technologies are the Wright dust feeder, fluidized bed generators, venturi tube injectors, and the jet mill. Powder consumption is often increased by the requirement of having a downstream impactor stage, installed to reduce the nonrespirable fraction of larger particle sizes in the exposure atmosphere. Recently, the DustGun aerosol technology has been applied to expose the isolated and perfused rat lung to dry powder aerosols of dieselsoot and the carcinogenic solute benzo(a)pyrene.

4. Concluding Remarks The methods and studies described in this chapter indicate that various drugs and peptides can be encapsulated stably into NLs, and successfully formulated into NLDPFs suitable for pulmonary administration. The preclinical studies show that NLDPFs are more effective than the free drug under both in vitro and in vivo conditions. The successful pulmonary delivery of therapeutic agents using NLDPFs suggests their potential use for treatment of pulmonary and systemic diseases. Ongoing research in the area of NLDPFs can improve clinical outcomes, and continue to expand therapeutic options in the treatment of many more diseases. Further developments focused on NL delivery in dry form to the lung are therefore warranted.

188

Gaurang Patel et al.

REFERENCES Aerosols, Nasal Sprays, Metered Dose Inhalers, and Dry Powder Inhalers. USP 30 NF 25; 2007: 601–607. Akwete, L. A., and Hsu, L. (1993). Leuprolide and other LH-RH analogues in stability and characterization of protein and peptide drugs: Case histories. In ‘‘Formation, Characterization, and Stability of Protein Drugs,’’ (Y. J. Wang and R. Pearlman, eds.), pp. 159–199. Plenum Press, New York. Anabousi, S., Bakowsky, U., Schneider, M., Huwer, H., Lehr, C. M., and Ehrhardt, C. (2006). In vitro assessment of transferrin-conjugated liposomes as drug delivery systems for inhalation therapy of lung cancer. Eur. J. Pharm. Sci. 29, 367–374. Application Note by Malvern Instruments. Size and zeta potential characterization of anionic and cationic liposomes on the Zetasizer Nano. Ashurst, I. I., Malton, A., Prime, D., and Sumby, B. (2000). Latest advances in the development of dry powder inhalers. Pharm. Sci. Technol. 3, 246–256. Betagiri, J. V., Jenkins, S. A., and Parsons, D. L. (1993a). In ‘‘Liposomes Drug Delivery System,’’ pp. 7–20. Technomic Publishing Co. Inc., Lanchaster, Basel, Switzerland. Betagiri, J. V., Jenkins, S. A., and Parsons, D. L. (1993b). In ‘‘Liposomes Drug Delivery System,’’ pp. 21–24. Technomic Publishing Co. Inc., Lanchaster, Basel, Switzerland. Bisgaard, H. (1996). Drug delivery from inhaler devices. Lung deposition, clinical effect and cost effectiveness vary. Br. Med. J. 313, 895–896. Bosquillona, C., Rouxhetb, P. G., Ahimoub, F., Simonc, D., Culotd, C., Preata, V., and Vanbever, R. (2004). Aerosolization properties, surface composition and physical state of spray-dried protein powders. J. Control. Release 99, 357–367. Briscoe, P., Caniggia, I., Graves, A., Benson, B., Huang, L., Tanswell, A. K., and Freeman, B. A. (1995). Delivery of superoxide dismutase to pulmonary epithelium via pH-sensitive liposomes. Am. J. Physiol. 268, L374–L380. Byron, P. R., and Patton, J. S. (1994). Drug delivery via the respiratory tract. J. Aerosol Med. 7(1), 49–75. Byron, P. R., Roberts, N. S. R., and Clark, A. R. (1986). An isolated perfused rat lung preparation for the study of aerosolized drug deposition and absorption. J. Pharm. Sci. 75, 168–171. Carr, R. L. (1965). Evaluating flow properties of solids. Chem. Eng. 72, 163–168. Chien, Y. W. (1992). In ‘‘Novel Drug Delivery Systems, Drugs and the Pharmaceutical Sciences,’’ 50, p. 797. Marcel Dekker, New York. Chougule, M., Padhi, B., and Misra, A. (2006). Nano-liposomal dry powder inhaler of amiloride hydrochloride. J. Nanosci. Nanotechnol. 6(9–10), 3001–3009. Chougule, M., Padhi, B., and Misra, A. (2007). Nano-liposomal dry powder inhaler of tacrolimus: Preparation, characterization, and pulmonary pharmacokinetics. Int. J. Nanomed. 2(4), 675–688. Chougule, M., Padhi, B., and Misra, A. (2008). Development of spray dried liposomal dry powder inhaler of dapsone. AAPS PharmSciTech. 9(1), 47–53. Courrier, H. M., Butz, N., and Andamme, F. (2002). Pulmonary drug delivery systems: Developments and prospects. Crit. Rev. Ther. Drug Carrier Syst. 19(4–5), 425–498. Elbert, K. J., Schafer, U., F.,.Scha¨fers, H. J., Kim, K. J., Lee, V. H., and Lehr, C. M. (1999). Monolayers of human alveolar epithelial cells in primary culture for pulmonary absorption and transport studies. Pharm. Res. 16, 601–608. Fang, X., Song, Y., Zemans, R., Hirsch, J., and Matthay, M. A. (2004). Transport across cultured rat alveolar epithelial cells: A novel in vitro system. Am. J. Physiol. (Lung Cell Mol. Physiol.) 27, L104–L110.

Nanoliposomal Dry Powder Formulations

189

Fiegel, J., Ehrhardt, C., Schaefer, U. F., Lehr, C. M., and Hanes, J. (2003). Large porous particle impingement on lung epithelial cell monolayers—Toward improved particle characterization in the lung. Pharm. Res. 20(5), 788–796. Fielding, R. M., and Abra, R. M. (1992). Factors affecting the release rate of terbutaline from liposome formulations after intratracheal instillation in the guinea pig. Pharm. Res. 9, 220–222. Fuchs, S., Hollins, A. J., Lehr, C. M., et al. (2003). Differentiation of human alveolar epithelial cells in primary culture: Morphological characterization and synthesis of caveolin-1 and surfactant protein-C. Cell Tissue Res. 311, 31–45. Geller, D. E. (2005). Comparing clinical features of the nebulizer, metered-dose inhaler, and dry powder inhaler. Respir. Care 50(10), 1313–1322. Goel, B. K. (1031). In ‘‘Medical Laboratory Technology,’’ III, p. 1031. Tata McGraw-Hill, New Delhi, India. Hersey, J. A. (1975). Ordered mixing: A new concept in powder mixing practice. Powder Technol. 11, 41–44. Hickey, A. J. (1992). In ‘‘Summary of Common Approaches to Pharmaceutical Aerosol Administration; Pharmaceutical Inhalation Aerosol Technology,’’ pp. 255–288. Marcel Dekker, New York. Hickey, A. J., Concessio, N. M., Platz, R. M., et al. (1994). Factors influencing the dispersion of dry powders as aerosols. Pharm. Technol. 8, 58–84. Hu, C., and Rhodes, D. G. (1999). Proniosomes: A novel drug carrier preparation. Int. J. Pharm. 185, 23–35. Hyung, J. L., Eun, C. C., Jongwon, S., Do-Hoon, K., Eun, J. A., and Junoh, K. (2008). Polymer-associated liposomes as a novel delivery system for cyclodextrin-bound drugs. J. Colloid Interface Sci. 320, 460–468. ICH Guidelines (2003). Stability testing of new drug substances and products. ICH Topic Q 1 A (R2) Page 7 of 20 August 2003. Joshi, M. R., and Misra, A. (2001a). Liposomal budesonide for dry powder inhaler: Preparation and stabilization. AAPS PharmSciTech. 2(4), 25–31. Joshi, M., and Misra, A. (2001b). Dry powder inhalation of liposomal ketotifen fumarate: Formulation and characterization. Int. J. Pharm. 223(1–2), 15–27. Joshi, M., and Misra, A. (2001c). Pulmonary disposition of budesonide from liposomal dry prowder inhaler. Meth. Find Exp. Clin. Pharm. 23(10), 531. Joshi, M., and Misra, A. (2003). Disposition kinetics of ketotifen from liposomal dry powder for inhalation in rat lung. Clin. Exp. Pharmacol. Physiol. 30, 153–156. Kadimi, U. S., Balasubramanian, D. R., Ganni, U. R., Balaraman, M., and Govindarajulu, V. (2007). In vitro studies on liposomal amphotericin B obtained by supercritical carbon dioxide-mediated process. Nanomedicine 3(4), 273–280. Kaszuba, M. (1999). The measurement of nanoparticles using photon correlation spectroscopy and avalanche photo diodes. J. Nano Res. 1(3), 405–409. Otake, K., Imura, T., Sakai, H., and Abe, M. (2001). Method of liposomes using supercritical carbon dioxide. Langmuir 17, 3898–3901. Kent, J., Thomas, L. A., Anders, F. V., Huiling, M., and Xuebing, X. (2007). Oxidative stability of liposomes composed of docosahexaenoic acid-containing phospholipids. J. Am. Oil Chem. Soc. 84(7), 631–637. Kimelberg, H. K., and Mayhew, E. G. (1978). Properties and biological effects of liposomes and their uses in pharmacology and toxicology. Crit. Rev. Toxicol. 6(1), 25–79. Kompella, U., and Koushik, K. (2001). Preparation of drug delivery systems using supercritical fluid technology. Crit. Rev. Ther. Drug Carrier Syst. 18(2), 173–199. Konduri, K., Nandedkar, S., Rickaby, D. A., Du¨zgu¨nes¸, N., and Gangadharam, P. R. J. (2005). The use of sterically stabilized liposomes to treat asthma. Methods Enzymol. 391, 413–427.

190

Gaurang Patel et al.

Lu, D., and Hickey, A. (2005). Liposomal dry powders as aerosols for pulmonary delivery of proteins. AAPS PharmSciTech. 6(4), E641–E648. Maria, L. I., Paola, B., Flavio, R., Silvia, A., Maurizio, C., and Luigi, C. (2004). Preparation, characterization, cytotoxicity and pharmacokinetics of liposomes containing lipophilic gemcitabine prodrugs. J. Control. Release 100, 331–346. Markus, M., Stefan, M., Christel, C., and Mu¨ller, G. (2004). Physicochemical characterisation of liposomes with encapsulated local anaesthetics. Int. J. Pharm. 274, 139–148. Martin, J. T., and Anthony, J. H. (2005). Dry powder inhaler formulation. Respir. Care 50(9), 1209–1227. Mina, A., Kanako, Y., and Kazuo, M. (2002). Oxidative stability of polyunsaturated fatty acid in phosphatidylcholine liposomes. Biosci. Biotechnol. Biochem. 66(12), 2573–2577. Misra, A., Parmar, N., Naik, S., and Patel, G. (2008). Preparation of Amphotericin B liposomes by supercritical fluid technology IN391/MUM/2008. New, R. R. C. (1990a). In ‘‘Liposomes—A Practical Approach,’’ pp. 33–103. Oxford University Press, New York. New, R. R. C. (1990b). In ‘‘Liposomes—A Practical Approach,’’ pp. 105–160. Oxford University Press, New York. Newman, S. P. (2004). Dry powder inhalers for optimal drug delivery. Expert Opin. Biol. Ther. 4(1), 23–33. Newman, S. P., and Clarke, S. W. (1983). Therapeutic aerosols I—Physical and practical considerations. Thorax 38, 881–886. Niven, R. W. (1995). Delivery of biotherapeutics by inhalation aerosol. Crit. Rev. Ther. Drug Carrier Syst. 12, 151–231. Oku, N., and Namba, Y. (2005). Glucuronate-modified, long-circulating liposomes for the delivery of anticancer agents. Methods Enzymol. 391, 145–162. Pang, Y., Sakagami, M., and Byron, P. R. (2005). The pharmacokinetics of pulmonary insulin in the isolated perfused rat lung: Implications of metabolism and regional deposition. Eur. J. Pharm. Sci. 25, 369–378. Parmar, M. (2006). Formulation of insoluble small molecule therapeutics in lipid-based carriers US20060051406A1. Patel, G., and Misra, A. (2007). Pulmonary delivery of liposomal dry powder inhaler of formoterol for effective treatment of asthma http://www.aapsj.org/abstracts/AM_2007/AAPS2007001080.PDF Robinson, J. R., and Lee, V. H. L. (1987). Controlled Drug Delivery: Fundamentals and Applications. 2nd edn. Marcel Dekker, New York. Sakagami, M., Byron, P. R., and Rypacek, F. (2002a). Biochemical evidence for transcytotic absorption of polyaspartamide from the rat lung: Effects of temperature and metabolic inhibitors. J. Pharm. Sci. 91, 1958–1968. Sakagami, M., Byron, P. R., Venitz, J., and Rypacek, F. (2002b). Solute disposition in the rat lung in vivo and in vitro: Determining regional absorption kinetics in the presence of mucociliary clearance. J. Pharm. Sci. 91, 594–694. Sakagami, M., Byron, P. R., Venitz, J., and Rypacek, F. (2006). In vivo, in vitro, and ex vivo models to assess pulmonary absorption and disposition of inhaled therapeutics for systemic delivery. Adv. Drug Deliv. Rev. 50(9–10), 1030–1060. Shah, S. P., and Misra, A. (2004a). Development of liposomal Amphotericin B dry powder inhaler formulation. Drug Deliv. 11(4), 247–253. Shah, S. P., and Misra, A. (2004b). Liposomal amikacin dry powder inhaler: Effect of fines on in vitro performance. AAPS PharmSciTech. 5(4), E65. Shah, S. P., and Misra, A. (2004c). Liposomal amphotericin B dry powder inhaler: Effect of fines on in vitro performance. Pharmazie 59(10), 812–813. Shah, V. P., Elkins, J. S., and Williams, R. L. (1993). In vitro drug release measurement of topical glucocorticoid creams. Pharmacopeial Forum 19, 5048–5059.

Nanoliposomal Dry Powder Formulations

191

Shahiwala, A., and Misra, A. (2004). Pulmonary absorption of liposomal levonorgestral. AAPS PharmSciTech. 5(1), E13. Shahiwala, A., and Misra, A. (2005). A preliminary pharmacokinetic study of liposomal leuprolide dry powder inhaler: A technical note. AAPS PharmSciTech. 6(3), E482–E486. Shek, P. N., and Barber, R. F. (1986). Liposomes: A new generation of drug and vaccine carriers. Mod. Med. 41, 314–326. Singh, S. (1999). Drug stability testing and shelf life determination according to international guidelines. Pharm. Technol. 23, 68–86. Sunkara, G., and Kompella, U. (1999). Drug delivery applications of supercritical fluid technology. Drug Deliv. Technol. 2, 33–34. Sweeney, L., Wang, Z., Loebenberg, R., Wong, J. P., Lange, C. F., and Finlay, W. (2005). Spray-freeze-dried liposomal ciprofloxacin powder for inhaled aerosol drug delivery. Int. J. Pharm. 305(1–2), 180–185. Taylor, K. M. G., and Newton, J. M. (1992). Liposomes for controlled delivery of drugs to the lung. Thorax 47, 257–259. Tronde, A., Norden, B., Jeppsson, A. B., Brunmark, P., Nilsson, E., Lennernas, H., and Bengtsson, U. H. (2003). Drug absorption from the isolated perfused rat lung—Correlations with drug physicochemical properties and epithelial permeability. J. Drug Target 11(1), 61–74. Vyas, S. P., Quraishi, S., Gupta, S., and Jaganathan, K. S. (2005). Aerosolized liposomebased delivery of amphotericin B to alveolar macrophages. Int. J. Pharm. 296, 12–25. Woolfe, A. J., Zheng, X. M., and Langford, A. (2001). Methods to produce powders for pulmonary or nasal administration WO0113885.

This page intentionally left blank

C H A P T E R

T E N

Lanthanide-Loaded Paramagnetic Liposomes as Switchable Magnetically Oriented Nanovesicles Silvio Aime, Daniela Delli Castelli, and Enzo Terreno Contents 194 195 197

1. 2. 3. 4.

Introduction Paramagnetic Ln(III)-Based Shift Reagents Preparation of Osmotically Shrunken Liposomes NMR Characterization of Magnetically Oriented Nonspherical Liposomes 5. Sample Experiments 6. Concluding Remarks Acknowledgments References

198 200 208 208 208

Abstract Magnetically oriented liposomes can be prepared by exposing unilamellar spherical systems loaded with paramagnetic lanthanide(III) complexes to hyperosmotic stress. The resulting nonspherical, lens-shaped, nanoparticles can orient within a static magnetic field, depending on the magnetic properties of their membrane bilayer. The orientation of the vesicles can be easily determined by measuring the paramagnetic contribution to the 1H chemical shift of the intraliposomal water proton resonance. As the latter shift is dominated by the bulk magnetic susceptibility contribution, its sign (negative or positive) reports about the preferred orientation adopted by the nanovesicles. The alignment within the field depends upon the magnetic susceptibility anisotropy of the liposome membrane, DwLIPO. When DwLIPO has a negative value (e.g., for nonspherical liposomes made of conventional phospholipids), the nanoparticles align with their long axis parallel to the field, whereas when DwLIPO > 0 the vesicles flip by 90
. The sign of the chemical shift of the intraliposomal water

Department of Chemistry IFM and Molecular Imaging Center, University of Torino, Torino, Italy Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64010-6

#

2009 Published by Elsevier Inc.

193

194

Silvio Aime et al.

resonance is positive in the former case and negative in the latter, respectively. The liposome orientation can be switched by incorporating in the liposome bilayer suitable amphiphilic paramagnetic lanthanide(III) complexes. The sign of DwLIPO, and consequently the magnetic alignment, will correspond to the sign of the magnetic susceptibility anisotropy of the metal complex. The magnetic susceptibility anisotropy is dependent on both the electronic configuration of the lanthanide ion and the structural characteristics of the amphiphilic complex incorporated in the liposome membrane. The magnetic orientation of such vesicles is maintained in vivo, thus opening promising perspectives for the application of nonspherical liposomes in medical imaging.

1. Introduction Liposomes are bilayered nanovesicles extensively used in the biomedical field, primarily as delivery systems. In particular, paramagnetic liposomes have recently attracted much attention as diagnostic probes for magnetic resonance imaging (MRI), thanks to their ability to carry a high number of imaging reporter to the site of interest, thus improving the sensitivity of the imaging technique for molecular imaging applications (Delli Castelli et al., 2008a). Paramagnetic liposomes are very versatile MRI probes, because they can generate contrast by exploiting different mechanisms depending on either the characteristics of the loaded paramagnetic species or the orientation of the liposome in the magnetic field. The two more important MRI contrast mechanisms rely on the shortening of the longitudinal (T1) or the transverse (T2) relaxation times of water protons. Furthermore, a new class of contrast agents has been recently introduced whose contrast mechanism deals with a saturation transfer process mediated by chemical exchange (CEST, chemical exchange saturation transfer) (Woods et al., 2006; Zhou and van Zijl, 2006). The most intriguing property of CEST agents relies on the possibility of visualizing different probes simultaneously present in the same region of the MR image (Aime et al., 2005a; Mc Mahon et al., 2008; Terreno et al., 2008a). Conventional liposomes loaded with Gd(III)- or Mn(II)-complexes can act as both T1- or T2-agents, whereas those loaded with paramagnetic Ln (III) ions different from Gd(III) lack the ability to generate T1-contrast, but maintain the possibility to act as T2-agents (Terreno et al., 2008b) and, moreover, can exhibit a very efficient CEST-contrast (lipoCEST agents) (Aime et al., 2005b). A noticeable improvement in the possibility to visualize different lipoCEST agents in the same image can be achieved upon orienting the vesicles in the magnetic field (Delli Castelli et al., 2008a,b).

Magnetically Oriented Paramagnetic Liposomes

195

Lipid-based nanosystems, such as bicelles and phospholipidic nanotubes, have peculiar magnetic properties that enable them to be oriented within a magnetic field (Karp et al., 2006; Prosser et al., 2006). The driving-force of this phenomenon is the interaction between the anisotropy of the magnetic susceptibility tensor (Dw) of the particle and the applied magnetic field (Prosser and Shyianovskaya, 2001); orientation can only occur when the particle shape is not isotropic. Systems made of phospholipids orient within the field according to their negative Dw value, but this ‘‘natural’’ alignment can be modified upon varying the Dw sign. This task has been accomplished by incorporating in the bilayer diamagnetic amphiphiles with positive magnetic anisotropy (Park et al., 2008) or paramagnetic systems, mostly lanthanide-based metal complexes, having positive Dw values (Crowell and Macdonald, 2001; Marcotte et al., 2006; Prosser et al., 1999). Although it has been reported that spherical multilamellar liposomes can orient within high magnetic fields (Brumm et al., 1992), conventional unilamellar spherical liposomes do not display a preferential orientation with respect to a magnetic field. Nonspherical unilamellar liposomes can be prepared by shrinking spherical ones through osmotic stress (Boroske et al., 1981; Menager and Cabuil, 2002; Terreno et al., 2006). In fact, when challenged with hyperosmotic stress, liposomes readily lose water, and rearranged into nonspherical systems, for example, assuming discoidal, oblate, cigar-like shapes (Hirota, 2003). As observed for other phospholipid-based membranes, the spontaneous orientation of diamagnetic, nonspherical, liposomes is to lie with the long axis parallel to the magnetic field due to the negative Dw value of the phospholipids. This orientation can be switched by 90
upon incorporating in the liposomal membrane paramagnetic species that change the Dw sign. In this chapter, we address the experimental workup that leads to the osmotically shrunken paramagnetic liposomes and describe the NMR method for determining their orientation properties.

2. Paramagnetic Ln(III)-Based Shift Reagents The paramagnetic Ln(III)-based shift reagents used in this work are illustrated in Chart 10.1. The metal complexes can be classified in two groups, depending on their hydrophilic/hydrophobic ratio. Ln–HPDO3A complexes are hydrophilic species and are the paramagnetic payload that may be hosted in the inner aqueous cavity of liposomes. Such complexes can be prepared according to the procedure reported in the literature for the Gd(III) ion (Dischino et al., 1991).

-OOC

COON

N

Ln-HPDO3A

Ln N

-OOC

-OOC N

N N

OH

COO-

COO-

N

Ln-DOTA-1

Ln -OOC

N

-OOC

N CON

Ln-DTPA-2a

CONH

Ln

N

N

CONH

COOCOON NOC

N

Ln-DTPA-1

Ln N

COOCOO-

COO-

-OOC -OOC

-OOC

N -OOC

O CONH CONH

COO-

N

O

Ln-DTPA-2b

O

Ln

N N

N N Ln N

-OOC

COO-

O

O O

N N

N N

Ln-TRIAZINO-1

Chart 10.1 Lanthanide complexes used for the preparation of magnetically oriented unilamellar liposomes.

197

Magnetically Oriented Paramagnetic Liposomes

The latter group deals with the shift reagents that have an amphiphilic character as they contain two long (typically C18) aliphatic chains covalently attached to the metal chelate. These species are poorly soluble or insoluble in water, but can be readily incorporated in the liposome bilayer (see Section 4). These compounds differ either in the structure of the coordination cage (DOTA-like, DTPA-like, and a triazino-based ligand) or in the modality through which the aliphatic chains are attached to the chelating unit (the chains are bound to a single or to different nitrogen atoms of the ligand). Ln–DOTA-1 and Ln–DTPA-2a complexes are synthesized according to the literature (Anelli et al., 2001; Prosser et al., 1999, respectively). The ligand TRIAZINO-1 is synthesized according to the procedure illustrated in Scheme 10.1, whereas the Tm(III) complex is synthesized following the procedure reported by Anelli et al. (2001). Ln–DTPA-1 complexes can be obtained from Bracco Imaging S.p.A.

3. Preparation of Osmotically Shrunken Liposomes Nonspherical large unilamellar paramagnetic liposomes suspended in isotonic buffer are prepared by using the conventional thin lipid film method. A given amount (typically 20 mg) of the lipid material (DPPC: dipalmitoylphosphatidylcholine, DSPE-PEG2000: distearoylphosphoethanolamine-polyethyleneglycol-2000, and amphiphilic Ln(III) complexes, if Cl N N

Cl N

Cl

Cl N

(C18H37)2NH

N

(C18H37)2N

Acetone H2O Na2CO3

HN NH2 NH2NH2.H2O

N

N N

(C18H37)2N

N HN NH2

Cl BrCH2COOtBu K2CO3 CH3CN COOH

COOtBu

HN N N (C18H37)2N

HN N COOH

N COOH

N HN N

CF3COOH

N

COOtBu N

(C18H37)2N N

COOtBu HN N

COOH

COOtBu

Scheme 10.1 Synthetic procedure followed for the synthesis of the ligand TRIAZINO-1.

198

Silvio Aime et al.

required) is dissolved in chloroform and the organic solution is slowly evaporated to remove the solvent until a thin film was formed on the wall of a round-bottom flask. The solvent removal is completed by keeping the film containing flask at the vacuum line for a couple of hours. The dry film is then hydrated by adding a hypotonic aqueous solution (usually 1 ml, 40 mOsm) containing the neutral Ln–HPDO3A (Ln ¼ Gd, Dy, Er, or Tm) complex. The film hydration is carried out at a temperature above the gel-to-liquid phase transition of the main phospholipid component (ca. 55
C for DPPC-based formulations) to promote the formation of a heterogeneous suspension of multilamellar vesicles entrapping the solution of the hydrophilic shift reagent. The unilamellarity is achieved by extruding the suspension several times (at least six) through polycarbonate filters with a pore diameter of 200 nm (Lipex extruder, Northern Lipids Inc., Canada) always keeping the temperature at 55
C via circulating water through the water jacket around the extruder. This workup leads to spherical liposomes, entrapping the shift reagent in their inner aqueous cavity, suspended in the hydration solution that is iso-osmolar with respect to the liposome cavity. The osmotic shrinkage of the vesicles occurs during the purification of the liposomes from the nonencapsulated shift reagent that is carried out by dialyzing the liposomes at 4
C against an isotonic (300 mOsm) HEPES/NaCl buffer (pH 7.4). Two dialysis cycles of 4 h each (volume ratio 500:1) are sufficient to remove the shift reagent. The vesicles are then subjected to a dynamic light-scattering investigation (Zetasizer NanoZS, Malvern, UK) in order to assess the mean hydrodynamic size and the polydispersity of the system.

4. NMR Characterization of Magnetically Oriented Nonspherical Liposomes The paramagnetic contribution to the chemical shift of the water protons (DINTRALIPO) that share the inner core of the liposomes with a Ln(III)-based metal complex is the sum of two contributions: DINTRALIPO ¼ DHYP þ DBMS

ð10:1Þ

The DHYP term refers to the hyperfine contribution that is mediated by the chemical coordination of the water molecules to the Ln(III) ion. Quantitatively, when the exchange rate of the metal bound water is fast, this term is directly proportional to the chemical shift of the water protons at the metal site (DLn) weighted on the molar fraction of the intraliposomal water protons that are coordinated to the metal center:

199

Magnetically Oriented Paramagnetic Liposomes

q½SR ð10:2Þ 111:2 where q indicates the number of water protons coordinated to the Ln(III) ion, [SR] is the intraliposomal concentration of the shift reagent, and 111.2 is the molar concentration of water protons in the aqueous core. In principle, DLn is the sum of a dipolar (also termed pseudocontact) and a contact term, although it has been demonstrated that the paramagnetic effect induced by Ln(III) ions on the coordinated water protons predominantly arises from the dipolar contribution. On this basis, DLn is proportional to the magnetic anisotropy of the Ln (III) complex DwLnL and to the geometric factor G that comprises the metal– proton distance r and the orientation of the metal–proton vector with respect to the magnetic axis of the metal complex (defined by the polar angles y and ’). Finally, DwLnL is proportional to the value of the Bleaney’s constant CJ (that characterizes each paramagnetic Ln(III) ion, see Table 10.1) and to the ligand field coefficient (A02 hr 2 i) of the metal complex (Babailov, 2008): DHYP ¼ DLn

DHYP  Dw LnL  G

where Dw LnL / CJ  A02 hr 2 i

and Gðr;y;’Þ ð10:3Þ

The second contribution to DINTRALIPO in Eq. (10.1) is represented by the so-called bulk magnetic susceptibility (BMS) shift term. The peculiarity of this contribution is its strong dependence on the shape and orientation within the static magnetic field of the compartment containing the paramagnetic species. This shift contribution is null when the paramagnetic complex is entrapped in a spherical region and this means that in Table 10.1 Effective magnetic moments and Bleaney’s constants for paramagnetic Ln(III) ions Ln

eff

CJ

Pr Nd Eu Gd Tb Dy Ho Er Tm Yb

3.62 3.68 3.4–3.6 7.9 9.7 10.6 10.6 9.6 7.6 4.5

 11  4.2 4.0 0  86  100  39 33 53 22

200

Silvio Aime et al.

conventional spherical liposomes encapsulating a paramagnetic shift reagent, DHYP is the only operative shift contribution for the intraliposomal water protons. An accurate quantitative description of the BMS contribution is only possible for highly symmetric compartment shapes and welldefined orientations (Chu et al., 1990). For instance, in the case of a paramagnetic SR contained in a cylinder parallel to the magnetic field, DBMS is given by the following equation (Corsi et al., 2001): 1557:23  ½SR  m2eff  s ð10:4Þ T where meff is the effective magnetic moment of the Ln(III) ion (see Table 10.1), T is the absolute temperature, and s is a ‘‘shape/orientation’’ constant that in this case is equal to 1/3. If the same cylinder is flipped by 90
and it is now perpendicular to the static field, s changes value and sign (1/6). On this basis, it is possible to predict that an SR entrapped in a nonspherical compartments oriented in the static field will display a positive BMS shift when the long axis of the compartment is parallel to the field, and a negative BMS shift when it is oriented perpendicularly. DBMS ¼

5. Sample Experiments A nice experiment for describing the effect on the osmotic shrinkage on the chemical shift of the intraliposomal water protons in liposomes encapsulating a paramagnetic shift reagent is illustrated in Fig. 10.1. The experiment was carried out on a stealth liposome (DPPC/DSPE-PEG2000 75/5 mole ratio) entrapping an hypotonic solution (40 mM) of the neutral shift reagent Tm-HPDO3A and incorporating in the bilayer the amphiphilic complex Tm-DOTA-1 (20% in moles). Upon their formation, the liposomes are suspended in an iso-osmolar buffer (40 mOsm), where the vesicles are expected to be spherical. The NMR spectrum of the suspension displays a resonance at about 0.4 ppm downfield to the bulk water. This resonance can be assigned to the intraliposomal water protons that are poorly shifted owing to the low paramagnetic payload of the vesicle. In fact, for a spherical liposome the only contribution to DINTRALIPO is the dipolar one, which is proportional to the intraliposomal concentration of the SR. The intrinsic shifting efficiency (DLn) of Tm-HPDO3A is about 20 ppm/M (Aime et al., 2006), so a 40 mM solution is expected to induce a chemical shift of ca. 0.8 ppm. The lower observed shift is likely due to the presence of the amphiphilic SR that, at least for the portion pointing inward the cavity, can affect the DINTRALIPO value. The observation that the overall observed shift is lower than that expected for the hydrophilic SR

201

Magnetically Oriented Paramagnetic Liposomes

300 mOsm (isotonic)

220 mOsm

22.3

0.6

12.1

160 mOsm

8.2 110 mOsm 4 72 mOsm

0.4

40 mOsm (iso-osmolar) 40

30

20

10

0

−10

−20

Figure 10.1 1H NMR spectra (14 T, 25
C) of liposomes made of DPPC/Tm-DOTA1/DSPE-PEG2000 (65/30/5 mol%) encapsulating 40 mM Tm-HPDO3A and suspended in a buffered medium (pH 7.4) with increasing osmolarity. The numbers indicate the DINTRALIPO values.

is an indication that the sign of the hyperfine shift contribution for the amphiphilic SR is opposite with respect to the hydrophilic one. Upon adding sodium chloride to the liposome suspension the vesicle are exposed to an osmotic stress. As a result of this stress, the resonance of the intraliposomal water protons moves away from the bulk water signal reaching a DINTRALIPO value of 22.4 ppm when the suspension is isotonic (ca. 300 mOsm). The increase of the shift is accompanied by a signal broadening and a decrease in its integral as reported in Fig. 10.2. Both observations can be accounted for in terms of the osmotic shrinkage of the nanovesicles as the water leakage (decrease in the signal integral) increases the intravesicular concentration of paramagnets that, in turn, results in an increase of the shift and of the line broadening of the intraliposomal water resonance. The data reported in Fig. 10.2 nicely fit a simple monoexponential decay that allows the estimation of the water fraction released from the liposomes undergone to the osmotic stress. On this basis, the vesicles appear to lose ca. the 75% of their entrapped water during the shrinkage and this means that the intraliposomal concentration of the Tm

202

Silvio Aime et al.

0.8

Signal integral

0.6

0.4

0.2

0.0 0

100

200 300 400 Osmolarity (mOsm)

500

600

Figure 10.2 Integral values for the intraliposomal water protons NMR signal (calculated using as reference the residual HOD signal of the deuterated solvent) as a function of the osmolarity of the liposomes suspension reported in Fig. 10.1.

(III)-based SRs in the nonspherical vesicle is almost fourfold larger than the original one. Focusing the attention on Tm-HPDO3A, its final concentration in the vesicles should be around 160 mM that corresponds to a DHYP shift of 3.2 ppm. The corresponding BMS contribution is much higher, about 16 ppm, as calculated from Eq. (4), assuming that the liposomes are oriented parallel to the static magnetic field and behave as a cylinder. Concerning the amphiphilic SR, its contribution to DHYP should have a negative sign, but the BMS contribution will be positive and, as discussed for Tm-HPDO3A, it significantly exceeds the hyperfine one. Hence, in spite of the approximations of the calculations (especially regarding the s parameter in Eq. (4)), the magnitude of the observed shift of 22.4 ppm appears reasonable. Figure 10.3 displays a cryo-TEM image of a specimen of osmotically shrunken liposomes (Terreno et al., 2009). The suspension is characterized by the presence of ellipsoidal-like vesicles whose shape in the TEM image is dependent upon their orientation with respect to the electronic beam (i.e., perpendicular to the image). The average long axis size ranged from 120 to 160 nm, in good agreement with the hydrodynamic size determined by dynamic light-scattering measurements. As already anticipated in the introduction, nonspherical liposomes align in the static magnetic field with their long axis parallel to B0 in virtue of the negative magnetic anisotropy of the self-assembled diamagnetic phospholipids (Scheme 10.2). Thus, the BMS shift contribution from such vesicles should be always positive. Figure 10.4 reports about the effect of encapsulating paramagnetic complexes in the aqueous cavity. Liposomes loaded

203

Magnetically Oriented Paramagnetic Liposomes

Electronic beam B

A

B B

A

Carbon Ice layer Lipo-CEST 200 nm

Figure 10.3 Cryo-TEM image showing a 2D projection of osmotically shrunken liposomes encapsulating 40 mM of Tm-HPDO3A and incorporating 20 mol% of Tm-DOTA-1. (A) Lens-shaped vesicle with the short axis positioned in plane with the electron beam, and (B) lens-shaped vesicle positioned with the short axis perpendicular to the electron beam. The dashed circle is displayed to guide the eye. Shrunken liposomes with all different orientations are observed.

Conventional liposome

ΔINTRALIPO = DHYP

(ΔBMS= 0)

B0 Non spherical liposome (magnetically orientable) ΔINTRALIPO = Δ HYP + ΔBMS

ΔχLIPO < 0

ΔχLIPO > 0

ΔBMS > 0

ΔBMS < 0

Paramagnetic amphiphilic metal complex (ΔχLnL < 0) Paramagnetic hydrophilic metal complex Paramagnetic amphiphilic metal complex (ΔχLnL > 0)

Scheme 10.2 Basic correlations among the physicochemical variables involved in the magnetic orientation and chemical shift of intraliposomal water protons for paramagnetic liposomes.

204

Silvio Aime et al.

12

DINTRALIPO (ppm)

10 8 6 4 2 0 -2 -4 0.0

0.3 0.4 0.1 0.2 Osmolarity of the hydration solution

Figure 10.4 DINTRALIPO values, measured at 25
C, as a function of the osmolarity of the solution used for hydrating the thin lipidic film (DPPC/DSPE-PEG 95:5 molar ratio) and containing an hydrophilic shift reagent (squares: Dy-HPDO3A, circles: Tm-HPDO3A). Filled symbols refer to spherical or slightly deformed vesicles. Open symbols refer to nonspherical liposomes. The straight lines are reported only for guiding eyes.

with Tm-HPDO3A or Dy-HPDO3A at different concentrations were purified against isotonic buffer and their DINTRALIPO values were determined and plotted as a function of the osmolarity of the hydration solution containing the metal complexes (Terreno et al., 2007). At high osmolarity values the liposomes were spherical and the DINTRALIPO values were rather small (only DHYP contribution present) and of opposite sign, as expected for complexes with the same structure (i.e., similar ligand field coefficients), but containing Ln(III) ions with CJ constants of opposite sign (Tm > 0, Dy < 0, see Table 10.1). As the osmolarity was decreased, the liposomes were subjected to the osmotic shrinkage, and the paramagnetic effect on the shift became less pronounced (mainly due to the decreased [SR] in the vesicle), but below 0.1 Osm the shrinkage was more evident and the DINTRALIPO values became positive for both the liposome types. Note also the increased absolute shift values due to the higher efficiency of the BMS contribution. To change the orientation of the nonspherical liposomes within the static magnetic field it is necessary to change the sign of the magnetic anisotropy susceptibility of the liposomal membrane (DwLIPO) from negative to positive. This task can be pursued by incorporating amphiphilic paramagnetic complexes with positive magnetic anisotropy (DwLnL > 0). Due to the much higher magnetic susceptibility of a paramagnetic species with respect to a diamagnetic system, small amounts of paramagnetic compounds are sufficient to address this task. As recalled above, the magnetic anisotropy of a paramagnetic metal complex is dependent on the metal ion, through

Magnetically Oriented Paramagnetic Liposomes

205

the CJ constant, and on the crystal field coefficient, which is mainly correlated to the structure of the coordination cage. Complexes with positive CJ values like Eu(III) or Tm(III) are excellent systems for changing the magnetic orientation of phospholipid-based assemblies as it was also reported for bicelles. Our work has shown that the incorporation of Tm– DOTA-1 complex leads to positive DINTRALIPO values, that is, the shift is along the same direction as observed for nonspherical liposomes without any paramagnetic species incorporated. Further insights into this matter has been gained by preparing a liposome incorporating the Tm–DTPA-2 complex, that is, the same complex successfully used for changing the orientation of bicelles (Prosser et al., 1999). Interestingly, the DINTRALIPO value for this system was negative (Fig. 10.5A), thus confirming the ability of this metal chelate to switch the magnetic anisotropy susceptibility of the liposomal membrane from a negative to a positive value. This result prompted us to investigate the properties of other paramagnetic amphiphiles differing in the Ln(III) ion and in the conjugation scheme of the hydrophobic tails to the ligand skeleton. The results are summarized in Fig. 10.5. By considering metal complexes with the same Ln(III) ion, for example, Tm(III) (Fig. 10.5A), the DINTRALIPO values were negative (liposomes oriented with the long axis perpendicular to B0, Scheme 10.2) for the two DTPA-2 ligands in which the two hydrocarbon chains are linked to different nitrogen atoms of the ligand, and positive (liposomes oriented with the long axis parallel to B0, Scheme 10.2) for the three ligands with different coordination cage (DOTA-1, DTPA-1, and TRIAZINO-1), but with the two hydrophobic tails bound to same nitrogen atom. This finding confirms that the orientation of nonspherical liposomes is strongly affected by the structure of the incorporated paramagnetic amphiphile. On the other hand, for a given structure, the replacement of the Ln(III) ion with a metal with a CJ value of opposite sign (e.g., Tm(III) and Dy(III)) induces the change in the orientation of the vesicles as reported in Fig. 10.5B. A peculiar behavior is expected in the presence of an amphiphilic Gd(III) complex. Gd(III) ion has a meff value (7.9) similar to Tm(III), but the symmetric distribution of its unpaired electrons makes him unable to induce a dipolar shift (CJ ¼ 0). For this reason, the incorporation of a Gd(III) complex in the membrane does not affect the DwLIPO value and the nonspherical liposomes maintain the same orientation displayed by diamagnetic vesicles and the resulting DINTRALIPO shifts are invariantly positive (Fig. 10.5C). However, the relatively high meff value allows the exploitation of the BMS shift contribution when an hydrophilic Gd(III) complex is encapsulated in an osmotically shrunken liposomes (Aime et al., 2007). In other words, the differences in the DINTRALIPO values between nonspherical vesicles oriented in the same way and loaded with Tm(III) and Gd(III)-chelates of the same ligand are primarily due to the presence of the dipolar shift contribution for the former system.

A

B

Tm-DTPA-2b

Dy-DTPA-2a Tm-DTPA-2a Tm-DTPA-2a Tm-DTPA-1 Dy-DOTA-1

Tm-TRIAZINO-1

Tm-DOTA-1

Tm-DOTA-1

20

10 0 -10 -20 DINTRALIPO (ppm)

-30

30

20

10 0 -10 -20 -30 -40 -50 DINTRALIPO (ppm)

C Gd-DTPA-1

Gd-DTPA-2b

Gd-DTPA-2a

Gd-DOTA-1

0

5

10 15 DINTRALIPO (ppm)

20

Figure 10.5 (A) (Top): DINTRALIPO values, measured at 25
C, for Tm(III)-loaded liposomes differing in the amphiphilic paramagnetic complex incorporated in the membrane (DPPC/TmL/DSPE-PEG 75:20:5 molar ratio). All systems contain Tm-HPDO3A (40 mM in the hydration solution) in the inner aqueous cavity. (B) (Middle): DINTRALIPO values, measured at 25
C, for Tm(III)- and Dy(III)-loaded liposomes for two different ligands (membrane composition DPPC/LnL/DSPE-PEG 75:20:5 molar ratio). Tm-HPDO3A or Dy-HPDO3A

207

Magnetically Oriented Paramagnetic Liposomes

A relevant issue related to the potential MRI applications of magnetically oriented liposomes deals with the maintenance of their orientation in biological systems. Figure 10.6 illustrates the results obtained from two experiments in which the lipoCEST probes were injected subcutaneously (on the left) (Terreno et al., 2008b) or intramuscularly (on the right), respectively. In both cases the CEST contrast was detected upon irradiation at the same frequency offset previously determined in aqueous solution for lipoCEST probes used in this in vivo experiment. This is a convincing demonstration that the nanovesicles maintain their magnetic orientation in the biological environment and highlights the possibility of using lipoCEST-based probes for the visualization of different biological targets in the same anatomical region.

ΔLIPO 7 ppm

0.6 0.55

−17 ppm

3.5 ppm

−17 ppm

7 ppm

ΔLIPO −17 ppm

ΔLIPO 3.5 ppm

ΔLIPO −17 ppm

0.7 0.6

0.3 0.28 0.26

0.5 0.5

0.24

0.4

0.22 0.2

0.4 0.3

0.18

0.35

0.25

0.28 0.26 0.24 0.22

0.45

0.3

0.3

0.18 0.16

0.16

0.14

0.14

0.12

0.12

0.1

0.2 0.1

0.2

Figure 10.6 In vivo MR-CEST experiments in mice. Two different lipoCEST probes are subcutaneously (left) or intramuscularly (right) injected. At the top an MRI T2w anatomical image is acquired just after the injection (the syringes indicate the injection site). At the bottom, MR-CEST parametric maps, obtained after the saturation at the frequency offset characteristics of the administered probes, are superimposed to the morphological images.

(40 mM in the hydration solution) are present in the inner aqueous cavity. (C) (Bottom): DINTRALIPO values, measured at 25
C, for Gd(III)-loaded liposomes differing in the amphiphilic paramagnetic complex incorporated in the membrane (DPPC/GdL/DSPE-PEG 75:20:5 molar ratio). Gd-HPDO3A (40 mM in the hydration solution) is encapsulated in the aqueous cavity.

208

Silvio Aime et al.

6. Concluding Remarks In this chapter, we have shown how to achieve the transformation of spherical unilamellar liposomes into nonspherical ones. Upon their loading with paramagnetic Ln(III) complexes, such systems exhibit the ability to align themselves in a static magnetic field. The parallel or perpendicular orientation is determined by the attainment of the minimum of energy of the interaction between the magnetic field and the magnetic susceptibility anisotropy of the self-assembled amphiphilic components of the bilayer membrane. The peculiar magnetic properties of paramagnetic Ln(III)based complexes can be exploited for switching from one orientation to the other and the magnetic alignment adopted by a given nanoparticles can be easily determined by measuring the 1H NMR chemical shift of the intraliposomal water protons. Currently, the most important field of application for these systems deals with the development of highly sensitive MR imaging probes in the CEST modality. Moreover, it is expected that, in analogy with other phospholipid-based magnetically oriented self-assemblies, like bicelles and nanotubes, nonspherical liposomes with their long axis perpendicular to the magnetic field could be useful systems for the determination of protein structures. Finally, they may be useful for the NMR characterization of large biomolecules associated with the membrane bilayer.

ACKNOWLEDGMENTS Economic and scientific support from local (Regione Piemonte, Nano-IGT and C 130 projects) and national government (FIRB RBNE03PX83_006, FIRB RBIP06293N, and PRIN 2005039914 projects), EC-FP6 (DiMI: LSHB-CT-2005-512146, EMIL: LSHCCT-2004-503569, and MEDITRANS: NMP4-CT-2006-026668), and EC-FP7 projects (ENCITE: 201842), EC-COST D38 action is gratefully acknowledged. Bracco Imaging S.p.A. (CRM Colleretto Giacosa (TO), Italy) is acknowledged for the longstanding and very fruitful collaboration. Discussions with S. Langereis, H. Gruell, and coworkers at Philips (Eindhoven, NL) have been very useful for the design and development of the systems reported in this contribution.

REFERENCES Aime, S., Carrera, C., Delli Castelli, D., Geninatti Crich, S., and Terreno, E. (2005a). Tunable imaging of cells labeled with MRI PARACEST agents. Angew. Chemie Int. Ed. 44, 1813–1815. Aime, S., Delli Castelli, D., and Terreno, E. (2005b). Highly sensitive MRI chemical exchange saturation transfer agents using liposomes. Angew. Chemie Int. Ed. 44, 5513–5515.

Magnetically Oriented Paramagnetic Liposomes

209

Aime, S., Geninatti Crich, S., Gianolio, E., Giovenzana, G. B., Tei, L., and Terreno, E. (2006). High sensitivity lanthanide(III) based probes for MR-medical imaging. Coord. Chem. Rev. 250, 1562–1579. Aime, S., Delli Castelli, D., Lawson, D., and Terreno, E. (2007). Gd-loaded liposomes as T1, susceptibility, and CEST agents, all in one. J. Am. Chem. Soc. 129, 2430–2431. Anelli, P. L., Lattuada, L., Lorusso, V., Schneider, M., Tournier, H., and Uggeri, F. (2001). Mixed micelles containing lipophilic gadolinium complexes as MRA contrast agents. MAGMA 12, 114–120. Babailov, S. P. (2008). Lanthanide paramagnetic probes for NMR spectroscopic studies of molecular conformational dynamics in solution: Applications to macrocyclic molecules. Prog. Nucl. Mag. Res. Sp. 52, 1–21. Boroske, E., Elwenspoek, M., and Helfrich, W. (1981). Osmotic shrinkage of giant egglecithin vesicles. Biophys. J. 34, 95–109. Brumm, T., Mops, A., Dolainsky, C., Bruckner, S., and Bayerl, T. M. (1992). Macroscopic orientation effects in broadline NMR-spectra of model membranes at high magnetic field strength. Biophys. J. 61, 1018–1024. Chu, S. C.-K., Xu, Y., Balschi, J., and Springer, C. S. Jr (1990). Bulk magnetic susceptibility shifts in NMR studies of compartmentalized samples: Use of paramagnetic reagents. Magn. Res. Med. 13, 239–262. Corsi, D. M., Platas-Iglesias, C., van Bekkum, H., and Peters, J. A. (2001). Determination of paramagnetic lanthanide(III) concentrations from bulk magnetic susceptibility shifts in NMR spectra. Magn. Res. Chem. 39, 723–726. Crowell, K. J., and Macdonald, P. M. (2001). Europium III binding and the reorientation of magnetically aligned bicelles: Insights from deuterium NMR spectroscopy. Biophys. J. 81, 255–265. Delli Castelli, D., Gianolio, E., Geninatti Crich, S., Terreno, E., and Aime, S. (2008b). Metal containing nanosized systems for MR-molecular imaging applications. Coord. Chem. Rev. 252, 2424–2443. Delli Castelli, D., Terreno, E., Carrera, C., Giovenzana, G. B., Mazzon, R., Rollet, S., Visigalli, M., and Aime, S. (2008a). Lanthanide-loaded paramagnetic liposomes as switchable magnetically oriented nanovesicles. Inorg. Chem. 47, 2928–2930. Dischino, D. D., Delaney, E. J., Emswiler, J. E., Gaughan, G. T., Prasad, J. S., Srivastava, S. K., and Tweedle, M. F. (1991). Synthesis of nonionic gadolinium chelates useful as contrast agents for magnetic resonance imaging. 1,4,7-Tris(carboxymethyl)-l0substituted- 1,4,7,10-tetraazacyclododecanes and their corresponding gadolinium chelates. Inorg. Chem. 30, 1265–1269. Hirota, S. (2003). Viscometric determination of axial ratio of ellipsoidal DNA-lipid complex. Methods Enzymol. 367, 177–199. Karp, E. S., Inbaraj, J. J., Lryukhin, M., and Lorigan, G. A. (2006). Electron paramagnetic resonance studies of an integral membrane peptide inserted into aligned phospholipid bilayer nanotube arrays. J. Am. Chem. Soc. 128, 12070–12071. Marcotte, I., Belanger, A., and Auger, M. (2006). The orientation effect of gramicidin A on bicelles and Eu3þ-doped bicelles as studied by solid-state NMR and FT-IR spectroscopy. Chem. Phys. Lipids 139, 137–149. Mc Mahon, M. T., Gilad, A. A., De Liso, M. A., Cromer Berman, S. M., Bulte, J. W. M., and van Zijl, P. C. M. (2008). New ‘‘multicolor’’ polypeptide diamagnetic chemical exchange saturation transfer (DIACEST) contrast agents for MRI. Magn. Res. Med. 60, 803–812. Menager, C., and Cabuil, V. (2002). Reversible shrinkage of giant magnetoliposomes under an osmotic stress. J. Phys. Chem. B 106, 7913–7918.

210

Silvio Aime et al.

Park, S. H., Loudet, C., Marassi, F. M., Dufourc, E. J., and Opella, S. J. (2008). Solid-state NMR spectroscopy of a membrane protein in biphenyl phospholipid bicelles with the bilayer normal parallel to the magnetic field. J. Magn. Res. 193, 133–138. Prosser, R. S., and Shyianovskaya, I. V. (2001). Lanthanide ion assisted magnetic alignment of model membranes and macromolecules. Concepts Magnetic Res. 13, 19–31. Prosser, R. S., Bryant, H., Bryant, R. G., and Vold, R. R. (1999). Lanthanide chelates as bilayer alignment tools in NMR studies of membrane-associated peptides. J. Magn. Res. 141, 256–260. Prosser, R. S., Evanics, F., Kitevski, L., and Al-Abdoul-Wahid, M. S. (2006). Current applications of bicelles in NMR studies of membrane-associated amphiphiles and proteins. Biochemistry 45, 8453–8465. Terreno, E., Cabella, C., Carrera, C., Delli Castelli, D., Mazzon, R., Rollet, S., Stancanello, J., Visigalli, M., and Ame, S. (2007). From spherical to osmotically shrunken paramagnetic liposomes: An improved generation of LIPOCEST MRI agents with highly shifted water protons. Angew. Chemie Int. Ed. 46, 966–968. Terreno, E., Delli Castelli, D., Milone, L., Rollet, S., Stancanello, J., Violante, E., and Aime, S. (2008a). First ex-vivo MRI co-localization of two LIPOCEST agents. Contrast Media Mol. Imaging 3, 38–43. Terreno, E., Delli Castelli, D., Cabella, C., Dastru`, W., Sanino, A., Stancanello, J., Tei, L., and Aime, S. (2008b). Paramagnetic liposomes as innovative contrast agents for magnetic resonance (MR) molecular imaging applications. Chem. Biodivers. 5, 1901–1912. Terreno, E., Delli Castelli, D., Violante, E., Sanders, H. M. H. F., Sommerdijk, N. A. J., and Aime, S. (2009). Osmotically shrunken LIPOCEST agents: An innovative class of magnetic resonance imaging contrast media based on chemical exchange saturation transfer. Chem. Eur. J. 15, 1440–1448. Woods, M., Woessner, D. E., and Sherry, A. D. (2006). Paramagnetic lanthanide complexes as PARACEST agents for medical imaging. Chem. Soc. Rev. 35, 500–511. Zhou, J., and van Zijl, P. C. M. (2006). Chemical exchange saturation transfer imaging and spectroscopy. Prog. Nucl. Mag. Res. Sp. 48, 109–136.

C H A P T E R

E L E V E N

Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs T. K. Ritchie,* Y. V. Grinkova,† T. H. Bayburt,† I. G. Denisov,† J. K. Zolnerciks,* W. M. Atkins,* and S. G. Sligar† Contents 1. Introduction 2. Overview of Nanodisc Technology 2.1. Structure and properties of Nanodiscs 2.2. MSP expression 2.3. MSP purification 3. Reconstitution Considerations 3.1. Preparing the reconstitution mixture 3.2. Reconstitution of bR trimer 3.3. Assembly of monomeric rhodopsin Nanodiscs 4. Optimizing the Reconstitution for P-glycoprotein 4.1. P-gp as a target for incorporation 4.2. Reconstitution of P-gp 4.3. Functional activity of P-gp in liposomes versus Nanodiscs Acknowledgments References

212 212 213 216 217 218 220 221 223 223 225 225 226 228 228

Abstract Self-assembled phospholipid bilayer Nanodiscs have become an important and versatile tool among model membrane systems to functionally reconstitute membrane proteins. Nanodiscs consist of lipid domains encased within an engineered derivative of apolipoprotein A-1 scaffold proteins, which can be tailored to yield homogeneous preparations of disks with different diameters, and with epitope tags for exploitation in various purification strategies. A critical aspect of the self-assembly of target membranes into Nanodiscs lies in the optimization of the lipid:protein ratio. Here we describe strategies for performing this optimization and provide examples for reconstituting * {

Department of Medicinal Chemistry, University of Washington, Seattle, Washington, USA Department of Biochemistry, University of Illinois at Urbana-Champaign, Urbana, Illinois, USA

Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64011-8

#

2009 Elsevier Inc. All rights reserved.

211

212

T. K. Ritchie et al.

bacteriorhodopsin as a trimer, rhodopsin, and functionally active P-glycoprotein. Together, these demonstrate the versatility of Nanodisc technology for preparing monodisperse samples of membrane proteins of wide-ranging structure.

1. Introduction As this volume highlights, model membrane systems are essential for ongoing research aimed at understanding lipid dynamics in complex biological membranes, membrane protein function, and molecular recognition between lipids and proteins or small molecules. In addition, several lipid membrane-based systems have been developed for drug delivery or other applications. Over the course of the past several decades the study of membrane proteins has been accelerated by membrane models including detergent micelles, mixed detergent/lipid micelles, bicelles, and liposomes, facilitating structural determination and functional studies. Although each of these established systems has distinct advantages, none are perfect for all applications and, in fact, each has significant limitations. Therefore, when considering methods for reconstituting membrane proteins, or designing lipid-based nanodevices, a recently established tool based on self-assembling lipid bilayer Nanodiscs is an important development (Bayburt and Sligar, 2002, 2003; Bayburt et al., 2002, 2006, 2007; Chougnet et al., 2007; Denisov et al., 2004; Marin et al., 2007; Morrissey et al., 2008; Nath et al., 2007a; Sligar, 2003). Nanodisc technology provides many advantages for controlling the physical parameters of protein–lipid particles, and they are likely to have utility as components to be incorporated into more complex nanodevices (Das et al., 2009; Goluch et al., 2008; Nath et al., 2008; Zhao et al., 2008). Here we describe the methods used for self-assembly of Nanodiscs and their application for reconstituting various membrane proteins into soluble nanoscale lipid bilayers with controlled composition and stoichiometry.

2. Overview of Nanodisc Technology Phospholipid bilayer Nanodiscs are similar in structure to nascent discoidal high-density lipoprotein particles. They consist of a circular fragment of the phospholipid bilayer encapsulated by two copies of a membrane scaffold protein (MSP) derived from apolipoprotein A-1 (Bayburt et al., 2002; Denisov et al., 2004), as illustrated in Fig. 11.1. A detailed review of the structural and biological aspects of apolipoprotein A-1 and its modification to yield MSPs has been presented (Nath et al., 2007a). Currently available MSP constructs are represented in Table 11.1. They consist of an N-terminal

Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs

213

Figure 11.1 Structure of Nanodiscs, modeled with POPC as lipid. Lipid bilayer fragment (white space filling) is encircled by two amphipathic helices of MSP (gray ribbon). The graphic was generated using the PyMOL Molecular Graphics system.

hexahistidine tag, a linker containing a protease site enabling the tag to be removed, and the main MSP sequences. Incorporation of membrane proteins into Nanodiscs with the histidine tag removed after purification of MSP enables the separation of empty disks from those containing histidine-tagged target proteins. The main MSP sequence can be varied by changing the number of amphipathic helices punctuated by prolines and glycines, to allow for disks of varying sizes. As summarized in Table 11.1, these scaffold proteins provide a collective set of tools to generate Nanodiscs ranging in outer diameter from 9.8 to 17 nm, which can accommodate a range of membrane proteins.

2.1. Structure and properties of Nanodiscs Optimization of the lipid:protein stoichiometry during the self-assembly process allows production of Nanodiscs of uniform size. The effect of scaffold protein length has been examined by determining the concentration

Table 11.1 Membrane scaffold protein constructs

Protein a

MSP1 MSP1TEV MSP1D1a MSP1D1 D73C MSP1D1(–) MSP1E1a MSP1E1D1 MSP1E2a MSP1E2D1 MSP1E3a MSP1E3D1a MSP1E3D1 D73C MSP1D1– 22 MSP1D1– 33 MSP1D1– 44 MSP2 MSP2N2 MSP2N3 MSP1FC MSP1FN

N-terminus

Disk size (nm) b

c

MW (Da)

e280 (M-1cm-1)

Features

FX TEV TEV TEV

9.7 /9.8 9.7b/10c 9.5b/9.7c 9.6b

24,608 25,947 24,662 24,650

23,950 26,930 21,430 21,430

TEV FX TEV FX TEV FX TEV TEV

9.6b/9.6c 10.4b/10.6c 10.5b 11.1b/11.9c 11.1b 12.1b/12.9c 12.1b 12.0b

22,044 27,494 27,547 30,049 30,103 32,546 32,600 32,588

18,450 32,430 29,910 32,430 29,910 32,430 29,910 29,910

TEV

9.4b

23,404

21,430

Original MSP1 (deletion 1–43 mutant of human Apo A-1) MSP1 with removable 7-his tag Deletion 1–11 mutant of MSP1TEV Cysteine in helix 2, Apo A-1 numbering, mutant of MSP1D1 MSP1D1 with removed 7-His tag Extended MSP1, helix 4 repeated Extended MSP1D1, helix 4 repeated Extended MSP1, helices 4 and 5 repeated Extended MSP1D1, helices 4 and 5 repeated Extended MSP1, helices 4, 5, and 6 repeated Extended MSP1D1, helices 4, 5, and 6 repeated Cysteine in helix 2, Apo A-1 numbering, mutant of MSP1E3D1 Deletion 1–22 mutant of MSP1TEV

TEV

9.0b

22,055

15,930

Deletion 1–33 mutant of MSP1TEV

TEV

8.6b

20,765

15,930

Deletion 1–44 mutant of MSP1TEV

FX TEV TEV TEV TEVF

9.5b 15.0b/16.5c 15.2b/17c 9.7b 9.6b

48,020 45,541 46,125 25,714 25,714

47,900 39,430 39,430 22,400 22,400

Fusion of two MSP1 with GT-linker Fusion of MSP1D1–11 and MSP1D1–22 with GT-linker Fusion of MSP1D1–11 and MSP1D1–17 with GT-linker MSP1D1 with C-terminal FLAG-tag MSP1D1 with N-terminal FLAG-tag

FX ¼ GHHHHHHIEGR; TEV ¼ GHHHHHHHDYDIPTTENLYFQG; TEVF ¼ GHHHHHHHDYDIPTTENLYFQGSDYKDDDDKG. a The plasmid is available through Addgene (http://www.addgene.org). b Stokes hydrodynamic diameter, determined by size-exclusion chromatography (Denisov et al., 2004). c Nanodisc diameter determined by SAXS (Denisov et al., 2004).

215

Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs

of radiolabeled lipid and scaffold protein in the Nanodisc-containing size exclusion peak (Denisov et al., 2004). These results, summarized in Fig. 11.2, illustrate an interesting trend. Insertion of extra helices in the central portion of the scaffold protein (MSP1E1, MSP1E2, and MSP1E3) results in Nanodiscs of increasing size, while deletions of the affinity tag and the first 22 amino acids of the N-terminus do not significantly decrease the size of the disk formed, implying that the first 22 amino acids are marginally, if at all, involved in the self-assembly process and resultant stabilization of the discoidal nanoparticle. Truncation past the first 22 amino acids leads to a gradual decrease in lipid:protein ratio accompanied by a decrease in the major monodisperse Nanodisc component and an increase in aggregated fractions. Systematic studies of the lipid:protein ratio in Nanodiscs made from different MSP constructs have shown that the number of lipids per Nanodisc, NL, and the number of amino acids in the scaffold protein, M, can be described by the following simple relationship (Eq. (11.1), modified Eq. (11.2) from Denisov et al., 2004): NL S ¼ ð0:423M  9:75Þ2

ð11:1Þ

where S represents the mean surface area per lipid used to form the ˚ 2. The quadratic relationship between the number Nanodisc, measured in A of lipid molecules per Nanodisc and the length of the scaffold protein confirms the flat two-dimensional morphology of Nanodisc particles,

200

2 2N

3 SP M

2

SP 1E M

SP 1E M

SP 1E

SP

M

M

1

100

80

60

40

20

M SP 1 SP 1T M SP E 1T V EV M ( SP −) 1Δ 1M SP 11 1Δ 1M SP 22 1Δ 1M 33 SP 1Δ 144

300

M

Number of DPPC molecules per leaflet

B

1

Number of DPPC molecules per leaflet

A

Figure 11.2 Number of DPPC molecules per Nanodisc determined experimentally using tritiated lipids. Panel A: number of lipids in Nanodiscs formed with extended MSP proteins. Panel B: number of lipids in Nanodiscs formed with truncated MSP proteins. For the description of MSP constructs, see Table 11.1.

216

T. K. Ritchie et al.

illustrated in Fig. 11.1. The size similarity of Nanodiscs formed using the same scaffold protein but different lipids clearly indicates that the length of the protein’s amphipathic helix is the sole determinant of Nanodisc diameter, while different lipid:protein stoichiometries are due to the different surface area per lipid. For example, 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) is in gel state below 314 K, with the area per lipid in the ˚ 2, while 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphochorange of 52–57 A line (POPC) is in liquid crystalline state above 268 K, with the area per lipid approximately 70 A˚2.

2.2. MSP expression MSPs are expressed using the pET expression system (Novagen) with the BL21-Gold (DE3) strain (Stratagene) as a host. The expression is very efficient, and a large amount of protein is produced in just a few hours after induction with IPTG. However, MSPs are noticeably susceptible to proteolysis, and prolonged postinduction growth results in significant decrease of the MSP yield. Different modifications of the N- and C-termini of the MSP can affect stability in vivo, and for some MSPs (e.g., fusion constructs or epitope tagged MSPs), shortening of the postinduction time and/or lowering the temperature during the growth in comparison with the standard protocol improves yield. The highest yield is achieved with a rich medium such as terrific broth (TB); however, minimal medium was used successfully for production of the isotope labeled MSP (Li et al., 2006). Relatively high oxygenation level, which is essential for good yields, can be easily maintained in a fermenter, such as Bio-Flow III. However, satisfactory yields can also be achieved in flasks by using relatively small culture volume (e.g., 500 mL in a 2-L Fernbach flask). The detailed method is outlined below: (1) A starting culture is prepared as follows: 30 mL of Luria Broth (LB) medium containing kanamycin (30 mg/L) is inoculated with a single colony from a freshly streaked plate. The suspension is incubated at 37  C with shaking at 250 rpm until the OD600 is approximately 0.4–0.6 (usually 5–6 h). At this point the culture can be used immediately or stored overnight at 4  C. (2) 2.5 L TB medium is prepared and sterilized and the fermenter parameters (37  C, 500 rpm, and air—3 L/min) are set. When the temperature reaches 37  C, 25 mg kanamycin and a few drops of antifoam are added and the fermenter is inoculated with the starting culture. (3) OD600 is checked every hour. When the OD reaches 2.5–3.0 (usually in 3–4 h), the culture is induced with 1 mM IPTG. The fermentation is stopped 3 h after induction. Typically, OD600 reaches 10–15 by the end of fermentation.

Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs

217

(4) The cells are harvested by centrifugation at 8000g for 10 min. The weight of the wet pellet collected from 2.5 L of culture grown on TB medium is usually between 50 and 60 g. The cell pellet is stored at 80  C.

2.3. MSP purification MSPs are purified using Chelating Sepharose FF (GE Healthcare), charged with Ni2þ, following the general protocol for purification of polyhistidinetagged proteins with additional washing steps using detergent-containing buffers to disrupt interaction of MSP with other proteins: (1) The metal-chelating column (3.4  6 cm) is charged by passing through 1 bed volume (50 mL) of 0.1 M NiSO4, followed by 100 mL of water. The column is equilibrated with 250 mL of 40 mM phosphate buffer, pH 7.4. (2) The cell pellet collected from 2.5 L fermentation (40–60 g) is resuspended in 200 mL of 20 mM phosphate buffer, pH 7.4. Phenylmethylsulfonyl fluoride (PMSF) is added from a stock solution in ethanol to make 1 mM. After the cells are completely resuspended, Triton X-100 is added to a final concentration of 1%. Approximately 5 mg of deoxyribonuclease I (Sigma, DN-25) is added. The cells are lysed by sonication (three 1-min rounds). The lysate is clarified by centrifugation at 30,000g for 30 min. (3) The lysate is loaded on the column. Care should be taken to make sure the flow rate does not exceed 10 mL/min (about 1 mL/min cm2). The column is washed with 250 mL of each of the following:  40 mM Tris/HCl, 0.3 M NaCl, 1% Triton X-100, pH 8.0  40 mM Tris/HCl, 0.3 M NaCl, 50 mM Na-cholate, 20 mM imidazole, pH 8.0  40 mM Tris/HCl, 0.3 M NaCl, 50 mM imidazole, pH 8.0 (4) MSP is eluted with 40 mM Tris/HCl, 0.3 M NaCl, 0.4 M imidazole. 10–14 mL fractions are collected, and protein is checked with Coomassie G-250 reagent (Pierce). The fractions containing MSP are pooled and the sample is dialyzed against buffer 1 (20 mM Tris/HCl, 0.1 M NaCl, 0.5 mM EDTA, pH 7.4) at 4  C. The protein sample is filtered using 0.22 mm syringe filter, and 0.01% NaN3 is added for storage. (5) Analyze the sample: protein purity is checked by running SDS–PAGE and performing electrospray mass spectrometry (see Table 11.1 for molecular masses). Absorbance is measured at 280 nm using 1 mm path length quartz cuvette against standard buffer, and protein concentration is calculated. If necessary, it is concentrated to 4–10 mg/mL. MSP can be stored for several days at 4  C. For long-term storage, the sample is frozen or lyophilized, and is stored at 20  C or below.

218

T. K. Ritchie et al.

(6) After purification, the column is regenerated with 50 mM EDTA, is washed with water, and is equilibrated with 20% ethanol. The column is regenerated after every purification round.

3. Reconstitution Considerations As of 2009, the list of membrane proteins reconstituted into Nanodiscs for functional studies include the cytochromes P450 (Baas et al., 2004; Bayburt and Sligar, 2002; Civjan et al., 2003; Das et al., 2007, 2009; Denisov et al., 2006, 2007; Duan et al., 2004; Grinkova et al., 2008; Kijac et al., 2007; Nath et al., 2007b) bacteriorhodopsin as a monomer and trimer (Bayburt and Sligar, 2003; Bayburt et al., 2006), G-protein coupled receptors as monomers and dimers (Bayburt et al., 2007; Leitz et al., 2006; Marin et al., 2007), other receptors (Boldog et al., 2006, 2007; Mi et al., 2008), toxins (Borch et al., 2008), blood coagulation protein tissue factor (Morrissey et al., 2008; Shaw et al., 2007), protein complexes of the translocon (Alami et al., 2007; Dalal et al., 2009), and monoamine oxidase (Cruz and Edmondson, 2007). The potential of Nanodiscs is exemplified by their utility in diverse biochemical and biophysical methodologies, including solid state NMR (Kijac et al., 2007; Li et al., 2006), single molecule fluorescence experiments (Nath et al., 2008), and solubilizing functional receptors (Bayburt et al., 2007; Boldog et al., 2007; Leitz et al., 2006; Mi et al., 2008). Importantly, these methods may be modified to accommodate other membrane proteins. As an example, we describe the methods of reconstitution of bacteriorhodopsin (bR) trimer and rhodopsin monomer. Assembly of membrane proteins into Nanodiscs follows the rules for empty Nanodiscs. Cholatesolubilized phospholipids (see Section 3.1) are mixed with MSP and detergent-solubilized membrane protein. Following detergent removal with adsorbent beads (Bio-beads SM-2, Biorad or Amberlite XAD-2; SigmaAldrich), the assembly is analyzed and purified by size-exclusion chromatography. Additional parameters to consider are the choice of detergent to initially solubilize the protein from its membrane, choice of Nanodisc size, and the lipid to MSP to membrane protein ratios. Incorporation of a membrane protein into Nanodiscs requires the protein to be initially solubilized by treatment with a detergent. For a practical guide to membrane protein solubilization, see Hjelmeland and Chrambach (1984). The crude solubilized protein can be put directly into Nanodiscs or purified beforehand. A distinct advantage of using the crude-solubilized membrane is that membrane proteins tend to be labile in detergent, and affinity purification can be done after the target is in the Nanodiscs. The use of protein purified in detergent has the advantage that the native lipid is mostly removed, thus simplifying determination of the correct MSP to phospholipid ratio.

Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs

219

When using purified protein, however, the presence of relatively high glycerol concentrations can interfere with the assembly process, so the final concentration in the reconstitution mixture should be kept below 4%. As with empty disks, the phospholipid:MSP ratio must be satisfied (see Table 11.2). Stated more precisely, the surface area of the target plus phospholipid bilayer needs to be matched to the size of Nanodisc being assembled. It should be recognized that target protein, along with any associated native lipid, will displace exogenously added phospholipid from the Nanodisc structure. ˚ 2 for DPPC, 57 A ˚ 2 for 1,2The mean surface area per lipid in Nanodiscs is 52 A 2 ˚ dimyristoyl-sn-glycero-3-phosphocholine (DMPC), and 69 A for POPC (Bayburt et al., 2002, 2006; Denisov et al., 2004). These numbers can be used as a starting point for determining the necessary amount of phospholipid, shown in Table 11.2 for empty Nanodiscs. If the structure of the target is known, an estimate of displaced lipid can be made based on cross-sectional area of the membrane domain. If the structure is not known, an estimate can be made using an area of 140 A˚2 per transmembrane helix. The Swiss-Prot database (http://www.expasy.org) annotates potential transmembrane helices for proteins in its database and ExPASy provides links to topology prediction tools for unknown proteins. One then simply subtracts the number of phospholipids displaced by the target protein, and any native lipid present, from the amount of lipid that would be used to form empty Nanodiscs of the same size and phospholipid type. Bacteriorhodopsin was found to displace 37 DMPC molecules and rhodopsin displaced 50 POPC molecules based on chemical and spectral analysis of purified Nanodiscs (Bayburt et al., 2006, 2007). The experimentally determined numbers are consistent with the cross-sectional areas of bR trimer corresponding to 40 DMPC and rhodopsin corresponding to 43 POPC estimated from the crystal structures. These results indicate that a simple subtraction of phospholipid to account for the surface area of protein is a valid approximation. Endogenous lipid must also be accounted for when reconstituting from whole solubilized membrane. A crude approximation is that the weight of lipid is equal to the weight of total protein in a membrane. We estimate the concentration of lipid using the molecular weight of POPC (MW 760). It is often convenient to use a large excess of MSP and synthetic phospholipid

Table 11.2 Reconstitution ratios for empty disks POPC

MSP1D1 65 MSP1E1D1 85 MSP1E2D1 105 MSP1E3D1 130

DPPC

DMPC

˚ 2) Bilayer area per Nanodisc (A

90 115 145 180

80 100 130 160

4400 5700 7200 8900

220

T. K. Ritchie et al.

compared to native membrane lipid so that the contribution of native lipid and membrane protein can be neglected. Once a reconstitution has been performed and analyzed by size-exclusion chromatography, the lipid:MSP ratio can be adjusted to optimize the formation of Nanodiscs. Another obvious consideration is the choice of the Nanodisc size. The bilayer area for several disk sizes is given in Table 11.2. Importantly, a critical number of phospholipids associated with the Nanodisc–protein complex may be necessary for nativelike structure. Theoretically, three bR can fit into MSP1 Nanodiscs, but the trimer only forms in the larger Nanodiscs, which suggests that sufficient phospholipid must be present to allow unperturbed oligomer formation. A final consideration is the target protein to disk ratio in the assembly mixture. Single monomeric membrane proteins will assemble into Nanodiscs as long as the ratio of Nanodisc to target is high (i.e., the number per Nanodisc follows the Poisson distribution for noninteracting target). If an oligomeric membrane protein is desired then one must consider the strength of interaction, as increasing the phospholipid component can dissociate oligomers by a surface dilution effect. For weakly interacting proteins, such as the bR homotrimer, the choice of Nanodisc to target ratio is critical (Bayburt et al., 2006). Experimentally, the ratio of bR to Nanodisc was varied to find the optimal ratio. A similar approach was used for the Tar receptor (Boldog et al., 2006). Bacteriorhodopsin trimer exhibits exciton formation that was used as a convenient assay for trimer formation. In the case of Tar, a functional assay suggested that a trimer of dimers formed at a specific reconstitution ratio. A few simple tests for assembly of a target protein with Nanodiscs can be performed to ensure efficient reconstitution. Separation of the reconstituted sample using a calibrated Superdex 200 column will allow determination of size and homogeneity of the Nanodiscs. If excess empty disks are present, column fractions can be analyzed for the presence of target by techniques such as SDS– PAGE or activity assays. Upon reinjection, the peak target fraction should elute at the same position without degradation or aggregation; size changes in the peak fraction indicate improper Nanodisc formation. The amount of phospholipid can be measured and should correspond to the expected value, as described above. For the measurement to be meaningful, however, the target-containing Nanodiscs must be isolated first from any empty Nanodiscs.

3.1. Preparing the reconstitution mixture Lipid stocks are prepared in chloroform at 25–100 mM and stored at 20  C in glass vials with Teflon-lined screw caps. The concentration of the stock solution is determined by phosphate analysis (Chen et al., 1956; Du¨zgu¨nes¸, 2003). The desired amount of chloroform lipid stock is dispensed into a disposable glass culture tube and dried using a gentle stream of nitrogen gas in a fume hood; a thin film on the lower walls of the tube can be obtained by rotating the tube while holding it at an angle. To remove

Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs

221

residual solvent, the tube is placed in a vacuum dessicator under high vacuum overnight. Buffer containing sodium cholate is added to the dried lipid film. Typically, cholate is added to twice the desired concentration of lipid, for example, if 200 mL of 100 mM lipid stock was used, 200 mL of 200 mM cholate or 400 mL of 100 mM cholate is added. The tube is vortexed, heated under hot tap water (about 60  C), and sonicated in an ultrasonic bath until the solution is completely clear, and no lipid remains on the walls of the tube. Scaffold protein is added to cholate-solubilized phospholipid to yield desired lipid:protein ratio, ensuring the final cholate concentration in the reconstitution mixture is between 12 and 40 mM, supplementing with standard buffer or cholate stock solution if necessary. The mixture is incubated at the appropriate incubation temperature, which is dependent on the lipid used, for 15 min or longer. The temperature of the self-assembly should be near the Tm of the lipid being used. Assembly with POPC is done on ice or at 4  C, DMPC at room temperature, and DPPC at 37  C. Prepared disk reconstitution mixtures can be used immediately to make Nanodiscs or incorporate membrane proteins, or lyophilized for prolonged storage. Specific examples in the following subsections demonstrate these steps with different proteins.

3.2. Reconstitution of bR trimer Purple membrane is isolated from Halobacterium salinarum JW-3 cultures and solubilized with 4% (w/v) Triton X-100 as described (Dencher and Heyn, 1978; Oesterhelt and Stoeckenius, 1974). MSP1E3 stock solutions ( 200 mM) and a DMPC/cholate mixture (200 mM/400 mM in buffer 1, prepared as described above) are added to bR (200 mM) in a microfuge tube to give MSP1E3:bR:DMPC ratio of 2:3:160. Protease inhibitors can be included in the assembly. The final concentration of DMPC should be above 7 mM, below which poor disk formation occurs (Bayburt et al., 2006). If low phospholipid concentrations are necessary, Nanodisc formation can be aided by using sodium cholate at a final concentration of 14 mM. After 1 h incubation at room temperature, detergent is removed by treatment for 3–4 h at room temperature with 500 mg wet Bio-beads SM2 per mL of solution, with gentle agitation to keep the beads suspended. Bio-beads SM-2 or Amberlite XAD-2 are prepared by suspending in methanol, washing with several volumes of methanol in a sintered glass funnel, and rinsing with large amounts of Milli-Q treated water (Millipore) to remove traces of methanol. Amberlite XAD-2 additionally requires removal of fine particles by decantation. Prepared beads are stored in water containing 0.01% (w/v) NaN3 as preservative. Incubation temperature and amount of beads are factors in the rate and completeness of detergent removal (Rigaud et al., 1998). We generally use an equal volume of beads to sample and an overnight incubation to remove detergents at

222

T. K. Ritchie et al.

4  C. Room temperature or 37  C assemblies require several hours. A table of adsorption capacities for various detergents has been compiled (Rigaud et al., 1998). If it is critical that the amount of residual detergent is known, the assembly should be tested using radiolabeled detergent. Bio-beads are removed by punching a hole in the bottom of the microfuge tube with a needle, placing the tube snugly through a hole made in the cap of a 15-mL Falcon tube (Corning), and punching a vent hole in the cap of the microfuge tube. The assembly is centrifuged briefly using the Falcon tube to collect the sample. The sample is filtered using a 0.22-mm filter and injected onto the gel filtration column run at 0.5 mL/min while monitoring A280 and A560. A typical elution profile after assembly of trimer is shown in Fig. 11.3, panel A. The reconstitution was made using optimal amount of phospholipid, yet the Nanodisc peak is still accompanied by larger aggregates that also contain bR. One possible explanation for the presence of aggregates is that multiple bR interactions promote an aggregation pathway as opposed to formation of Nanodiscs of fixed size. Fractions containing the bR Nanodiscs are pooled and the presence of trimer is assessed by measuring the visible circular dichroism spectrum which shows a positive and negative peak, due to exciton splitting (Bayburt et al., 2006).

A

B Rhodopsin nanodisc

bR nanodisc

0.5 OD

5 mOD 500 nm

Aggregates

280 nm 560 nm

0

5

10 15 20 Elution time (min)

Aggregates

25

15

20

25 30 Elution time (min)

35

Figure 11.3 Elution profile from Nanodisc reconstitutions. Panel A: elution profile of MSP1E3 bR trimer Nanodiscs after assembly. After detergent removal the sample was injected onto a Superdex 200 prep grade column at a flow rate of 0.5 mL/min. The main peak corresponds to Nanodiscs containing three bR. Panel B: elution profile of MSP1E3 rhodopsin Nanodisc assembly mixture produced from solubilized rod outer segments. The sample was injected onto a Superdex 200 HR 10/30 column run at a flow rate of 0.5 mL/min.

Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs

223

3.3. Assembly of monomeric rhodopsin Nanodiscs The assembly described herein uses whole membrane and added synthetic phospholipid to generate rhodopsin monomer Nanodiscs. Rhodopsin is handled in a darkroom under dim red light (Kodak #1 filter, 7.5 W bulb). Rod outer segments (Papermaster, 1982) are solubilized in 135 mM nonyl glucoside to give 143 mM solubilized rhodopsin. Rod outer segments contain on the order of 100 native phospholipids per rhodopsin. MSP1E3D1 (183 mM) and DMPC (0.1 M in buffer 1 containing 0.2 M cholate) are mixed with solubilized membranes at ratios of 1:168:0.05 (MSP:DMPC:rho) on ice followed by overnight removal of detergent with Bio-beads at 4  C with gentle agitation. The sample is filtered and run on a Superdex 200 HR 10/30 column run at 0.5 mL/min. The elution profile monitored at 500 nm is given in Fig. 11.3, panel B. The elution profile shows a sharp Gaussian peak, though there are small amounts of larger aggregates. The aggregates indicate that the amount of DMPC in the reconstitution could be lowered somewhat to optimize assembly of Nanodiscs.

4. Optimizing the Reconstitution for P-glycoprotein When embarking on the incorporation of a new target into Nanodiscs, one must not only consider the requirements of the Nanodisc system but also any unique requirements of the target of interest. Herein we describe the tailoring of the reconstitution to an important mammalian protein, P-glycoprotein (P-gp). P-gp is a member of the ATP-binding cassette (ABC) transporter family which has been implicated in the phenomenon of multidrug resistance in tumor cells (Higgins, 2007), as well as the absorption and disposition of many pharmaceutical compounds (Zhou, 2008), yet there is still a great deal about the mechanism and interaction with substrates that is unknown. In fact, structure–function studies of P-gp have been seriously hampered by the difficulty of obtaining large quantities of stable P-gp. Presumably, this difficulty results from the structural complexity of P-gp which comprises a 1280 amino acid protein with 12 transmembrane helices punctuated by two cytoplasmic nucleotide-binding domains (NBDs) (Higgins et al., 1997). A recent crystal structure of mouse P-gp (Abcb1a, 87% homology with human P-gp) is shown in Fig. 11.4, to illustrate the domain architecture (Aller et al., 2009). P-gp is known to be sensitive to both the lipid environment (Orlowski et al., 2006) and the detergent used during the purification process (Bucher et al., 2007). Disruption of the lipid–protein interface has been shown to

224

T. K. Ritchie et al.

TMDs

NBDs

Figure 11.4 Crystal structure of mouse P-gp (PDB: 3G5U) in the nucleotide-free state, as seen from the plane of the membrane (Aller et al., 2009). The TMDs are embedded in the membrane, while the NBDs protrude into the interior of the cell. The graphic was generated using the PyMOL Molecular Graphics system.

result in almost complete inactivation of the protein (Callaghan et al., 1997); in fact, a common practice in the purification of P-gp is to add external lipid to maintain this crucial interface (Ambudkar et al., 1998; Taylor et al., 2001). Many detergents commonly used to solubilize membrane proteins disrupt the protein–lipid interaction, and are thus detrimental for use with P-gp (Naito and Tsuruo, 1995). N-Dodecyl-b-D-maltoside (DDM) is a mild, nonionic detergent that is commonly used in the solubilization and reconstitution of P-gp (Kimura et al., 2007; McDevitt et al., 2008), and which has also previously been used in the formation of Nanodiscs (Alami et al., 2007; Boldog et al., 2006; Dalal et al., 2009). It was, therefore, chosen to use in the incorporation of P-gp into Nanodiscs. The standard lipid used during the purification and liposomal reconstitution of P-gp is an Escherichia coli total lipid extract (Kim et al., 2006; Taylor et al., 2001), which is a mixture of phosphatidylethanolamine (57.5%), phosphatidylglycerol (15.1%), cardiolipin (9.8%), and ‘‘other’’ lipids (17.6%). This mixture seems to satisfy the requirement P-gp has for the lipid content, as exemplified by high levels of

Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs

225

drug-stimulated ATPase activity in reconstituted proteoliposomes (Ambudkar et al., 1998; Taylor et al., 2001) and has also been used, with DDM, in the formation of Nanodiscs (Alami et al., 2007; Dalal et al., 2009).

4.1. P-gp as a target for incorporation There are currently four in vitro systems routinely utilized to study P-gp: whole cells overexpressing P-gp (Adachi et al., 2001; Polli et al., 2001; Takano et al., 1998; Wang et al., 2002), membrane fractions from those cells (Loo and Clarke, 2005; Loo et al., 2003; Zolnerciks et al., 2007), purified protein that has been solubilized in detergent (Liu et al., 2000; Qu et al., 2003; Rosenberg et al., 2005), and purified protein that has been reconstituted into proteoliposomes (Kim et al., 2006; Lu et al., 2001; Taylor et al., 2001). Each system has strengths and weaknesses; in the whole cell and membrane fraction systems the protein is in the most native form but there is the obvious concern about the complexity of the system. Human P-gp that has been detergent-solubilized shows no ATPase activity, whereas protein that has been reconstituted into proteoliposomes has ATPase activity (Ambudkar et al., 1998), but is not particularly stable. In fact, at room temperature P-gp-proteoliposomes have a half-life of less than 1 day (Heikal et al., 2009). Nanodiscs afford an attractive system to study P-gp because they allow for a relatively simple, controlled system in which P-gp is solubilized, yet in an active form.

4.2. Reconstitution of P-gp Baculovirus-encoding dodeca-histidine-tagged-P-gp was a generous gift from Dr. Kenneth Linton (Imperial College, London). Production of P-gp containing insect cell membranes and protein purification is performed as previously described (Taylor et al., 2001), with modifications. Briefly, insect cell membrane fractions are solubilized in solubilization buffer (20 mM Tris, 150 mM NaCl, 1.5 mM MgCl2, 20% glycerol, 0.4% lipid (80:20 E. coli total lipid:cholesterol), and 2% DDM, pH 6.8) with repeated extrusion through a 25-gauze needle. Insoluble protein is separated by centrifugation at 100,000g for 40 min. The resulting solubilized protein is incubated with ProBond Nickel-Chelating Resin (Invitrogen) for 1 h at 4  C with constant agitation, with the addition of 20 mM imidazole to reduce nonspecific binding. The resin is washed with 20 bed volumes of wash buffer (20 mM Tris, 150 mM NaCl, 1.5 mM MgCl2, 20% glycerol, 0.1% DDM, pH 8) with increasing concentrations of imidazole (80–150 mM). P-gp containing fractions are eluted with 500 mM imidazole in elution buffer (same as wash buffer, pH 6.8) and stored at 80  C until used.

226

T. K. Ritchie et al.

(1) A lipid film of 12 mmol E. coli total lipid (molar concentration determined as described above) is prepared and vacuum desiccated overnight. (2) The lipid film is resuspended in 17 mmol DDM and 1 mL buffer 1 (20 mM Tris, 100 mM NaCl, pH 7.4), then sonicated and vortexed until the solution is clear and free of lumps of lipid. (3) 500 mL of purified P-gp in elution buffer is added, as welll as, protease inhibitors (20 mM leupeptin, 1 mM benzamidine, and 1 mM pepstatin), 100 nmol MSP1E3D1, and enough buffer 1 to make a total volume of 2.5 mL, ensuring the final glycerol concentration is less than 4%. The mixture is then incubated at room temperature with constant agitation for 1 h. (4) To initiate self-assembly, 0.6 g/mL washed Bio-beads SM-2 are added and incubated at room temperature for 2 h with constant agitation. (5) Reconstituted Nanodiscs are removed from Bio-beads with a 25-gauge needle and is stored at 4  C until used. (6) Empty Nanodiscs can be made in parallel, adding 500 mL of elution buffer in place of purified P-gp.

4.3. Functional activity of P-gp in liposomes versus Nanodiscs Functional characterization of a transporter protein in Nanodiscs has unique challenges. A disadvantage of using Nanodiscs to study transporters, such as P-gp, is the inability to study true vectorial transport, per se, because there is no internal or external compartment. Fortunately, a majority of the substrates transported by P-gp stimulate ATPase activity, which can be used as a surrogate for many of the conformational and chemical processes functionally coupled to transport (Polli et al., 2001). As mentioned previously, human P-gp has no detectable ATPase activity when solubilized in DDM, but regains activity when reconstituted. The amount of lipid is stringently controlled during the reconstitution process to prevent the concurrent formation of liposomes. Thus, the activity that is determined after reconstitution can be attributed to P-gp in Nanodiscs. For an initial characterization, the activity of P-gp reconstituted in Nanodiscs was determined by measuring the basal and drug-stimulated ATPase activity in MSP1E3D1 disks and in proteoliposomes, the standard reconstitution system for P-gp. Proteoliposomes are formed as previously described, with modifications (Taylor et al., 2001). Briefly, a mixture of E. coli lipid and cholesterol (80:20, w/w) is dried to a lipid film, before rehydration in elution buffer without DDM. The solution is sonicated and vortexed to make unilamelar liposomes. DDM is added to completely solubilize the lipid, and the solution is incubated at room temperature for 1 h to equilibrate. Equal volumes of the solubilized lipid and purified P-gp are incubated with protease inhibitors for 30 min at room temperature with constant agitation. Detergent is selectively

227

Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs

removed by addition of 0.3 g/mL of Bio-beads SM-2 for 2 h at room temperature with constant agitation. Proteoliposomes are recovered with a 25-gauge needle and stored on ice until used. Basal and drug-stimulated ATPase activity was determined by phosphate release using a colorimetric assay, as previously described (Chifflet et al., 1988), at 50 mM nicardipine, with varying concentrations of ATP (Taylor et al., 2001). Empty disks or liposomes made in parallel were used as a control. Figure 11.5 shows the comparison of basal and nicardipine-stimulated activity of P-gp in MSP1E3D1 Nanodiscs and liposomes. A twofold increase in the maximum drug-stimulated ATPase activity in Nanodiscs, compared to liposomes, is seen, while the Km values are comparable. This could be due to the uniform orientation of P-gp in Nanodiscs, whereas in liposomes there are two possible orientations: right-side-out (NBDs on the interior of the liposomes, and therefore inaccessible to ATP) and inside-out (NBDs on the exterior of the liposomes, and therefore accessible to ATP). This scrambled orientation in liposomes is consistent with incorporation of the protein using completely solubilized lipid (Rigaud, 2002). An increase in basal activity is also seen in disks as compared to liposomes, where the basal activity is almost undetectable. These data not only show that P-gp is functionally active when reconstituted into Nanodiscs, but that it exhibits higher specific activity than the current standard reconstitution system as well. P-gp is a complex, integral membrane protein containing 12 transmembrane helices that was incorporated into Nanodiscs in a fairly straightforward manner, after small modifications to the standard procedure. This will facilitate a more

nmol Pi/min/mg P-gp

1600

1200

800

400

0 0.0

0.5

1.0 1.5 mM ATP

2.0

2.5

Figure 11.5 ATPase activity of P-gp in MSP1E3D1 Nanodiscs as compared to proteoliposomes. Squares represent the activity of P-gp in MSP1E3D1 Nanodiscs and circles represent activity in liposomes. Open symbols show basal activity in the absence of drug and filled symbols show activity in the presence of 50 mM nicardipine.

228

T. K. Ritchie et al.

detailed study into the mechanism of P-gp and its interaction with substrates and serves to exemplify the utility of Nanodiscs in the study of membrane proteins.

ACKNOWLEDGMENTS The work described here was supported by Grants GM 33775 and GM 31756 to S. G. S. and GM 32165 to W. M. A.

REFERENCES Adachi, Y., Suzuki, H., and Sugiyama, Y. (2001). Comparative studies on in vitro methods for evaluating in vivo function of MDR1 P-glycoprotein. Pharm. Res. 18, 1660–1668. Alami, M., Dalal, K., Lelj-Garolla, B., Sligar, S. G., and Duong, F. (2007). Nanodiscs unravel the interaction between the SecYEG channel and its cytosolic partner SecA. EMBO J. 26, 1995–2004. Aller, S. G., Yu, J., Ward, A., Weng, Y., Chittaboina, S., Zhuo, R., Harrell, P. M., Trinh, Y. T., Zhang, Q., Urbatsch, I. L., and Chang, G. (2009). Structure of Pglycoprotein reveals a molecular basis for poly-specific drug binding. Science 323, 1718–1722. Ambudkar, S. V., Lelong, I. H., Zhang, J., and Cardarelli, C. (1998). Purification and reconstitution of human P-glycoprotein. Methods Enzymol. 292, 492–504. Baas, B. J., Denisov, I. G., and Sligar, S. G. (2004). Homotropic cooperativity of monomeric cytochrome P450 3A4 in a nanoscale native bilayer environment. Arch. Biochem. Biophys. 430, 218–228. Bayburt, T. H., and Sligar, S. G. (2002). Single-molecule height measurements on microsomal cytochrome P450 in nanometer-scale phospholipid bilayer disks. Proc. Natl. Acad. Sci. USA 99, 6725–6730. Bayburt, T. H., and Sligar, S. G. (2003). Self-assembly of single integral membrane proteins into soluble nanoscale phospholipid bilayers. Protein Sci. 12, 2476–2481. Bayburt, T. H., Grinkova, Y. V., and Sligar, S. G. (2002). Self-assembly of discoidal phospholipid bilayer nanoparticles with membrane scaffold proteins. Nano Lett. 2, 853–856. Bayburt, T. H., Grinkova, Y. V., and Sligar, S. G. (2006). Assembly of single bacteriorhodopsin trimers in bilayer nanodiscs. Arch. Biochem. Biophys. 450, 215–222. Bayburt, T. H., Leitz, A. J., Xie, G., Oprian, D. D., and Sligar, S. G. (2007). Transducin activation by nanoscale lipid bilayers containing one and two rhodopsins. J. Biol. Chem. 282, 14875–14881. Boldog, T., Grimme, S., Li, M., Sligar, S. G., and Hazelbauer, G. L. (2006). Nanodiscs separate chemoreceptor oligomeric states and reveal their signaling properties. Proc. Natl. Acad. Sci. USA 103, 11509–11514. Boldog, T., Li, M., and Hazelbauer, G. L. (2007). Using Nanodiscs to create water-soluble transmembrane chemoreceptors inserted in lipid bilayers. Methods Enzymol. 423, 317–335. Borch, J., Torta, F., Sligar, S. G., and Roepstorff, P. (2008). Nanodiscs for immobilization of lipid bilayers and membrane receptors: Kinetic analysis of cholera toxin binding to a glycolipid receptor. Anal. Chem. 80, 6245–6252.

Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs

229

Bucher, K., Belli, S., Wunderli-Allenspach, H., and Kramer, S. D. (2007). P-glycoprotein in proteoliposomes with low residual detergent: The effects of cholesterol. Pharm. Res. 24, 1993–2004. Callaghan, R., Berridge, G., Ferry, D. R., and Higgins, C. F. (1997). The functional purification of P-glycoprotein is dependent on maintenance of a lipid-protein interface. Biochim. Biophys. Acta 1328, 109–124. Chen, P. S., Toribara, T. Y., and Warner, H. (1956). Microdetermination of phosphorus. Anal. Chem. 28, 1756–1759. Chifflet, S., Torriglia, A., Chiesa, R., and Tolosa, S. (1988). A method for the determination of inorganic phosphate in the presence of labile organic phosphate and high concentrations of protein: Application to lens ATPases. Anal. Biochem. 168, 1–4. Chougnet, A., Grinkova, Y., Ricard, D., Sligar, S., and Woggon, W.-D. (2007). Fluorescent probes for rapid screening of potential drug-drug interactions at the CYP3A4 level. ChemMedChem 2, 717–724. Civjan, N. R., Bayburt, T. H., Schuler, M. A., and Sligar, S. G. (2003). Direct solubilization of heterologously expressed membrane proteins by incorporation into nanoscale lipid bilayers. Biotechniques 35, 556–558, 560, 562–563. Cruz, F., and Edmondson, D. E. (2007). Kinetic properties of recombinant MAO-A on incorporation into phospholipid nanodisks. J. Neural Transm. 114, 699–702. Dalal, K., Nguyen, N., Alami, M., Tan, J., Moraes, T. F., Lee, W. C., Maurus, R., Sligar, S. G., Brayer, G. D., and Duong, F. (2009). Structure, binding, and activity of Syd, a SecY-interacting protein. J. Biol. Chem. 284, 7897–7902. Das, A., Grinkova, Y. V., and Sligar, S. G. (2007). Redox potential control by drug binding to cytochrome P 450 3A4. J. Am. Chem. Soc. 129, 13778–13779. Das, A., Zhao, J., Schatz, G. C., Sligar, S. G., and Van Dyne, R. P. (2009). Screening of type I and II drug binding to human cytochrome P450–3A4 in nanodiscs by localized surface plasmon resonance spectroscopy. Anal. Chem. 81, 3754–3759. Dencher, N. A., and Heyn, M. P. (1978). Formation and properties of bacteriorhodopsin monomers in the non-ionic detergents octyl-beta-D-glucoside and Triton X-100. FEBS Lett. 96, 322–326. Denisov, I. G., Grinkova, Y. V., Lazarides, A. A., and Sligar, S. G. (2004). Directed selfassembly of monodisperse phospholipid bilayer nanodiscs with controlled Size. J. Am. Chem. Soc. 126, 3477–3487. Denisov, I. G., Grinkova, Y. V., Baas, B. J., and Sligar, S. G. (2006). The ferrous-dioxygen intermediate in human cytochrome P450 3A4: Substrate dependence of formation of decay kinetics. J. Biol. Chem. 281, 23313–23318. Denisov, I. G., Grinkova, Y. V., McLean, M. A., and Sligar, S. G. (2007). The one-electron autoxidation of human cytochrome P450 3A4. J. Biol. Chem. 282, 26865–26873. Duan, H., Civjan, N. R., Sligar, S. G., and Schuler, M. A. (2004). Co-incorporation of heterologously expressed Arabidopsis cytochrome P450 and P450 reductase into soluble nanoscale lipid bilayers. Arch. Biochem. Biophys. 424, 141–153. Du¨zgu¨nes¸, N. (2003). Preparation and quantitation of small unilamellar liposomes and large unilamellar reverse-phase evaporation liposomes. Methods Enzymol. 367, 23–27. Goluch, E. D., Shaw, A. W., Sligar, S. G., and Liu, C. (2008). Microfluidic patterning of nanodisc lipid bilayers and multiplexed analysis of protein interaction. Lab Chip 8, 1723–1728. Grinkova, Y. V., Denisov, I. G., Waterman, M. R., Arase, M., Kagawa, N., and Sligar, S. G. (2008). The ferrous-oxy complex of human aromatase. Biochem. Biophys. Res. Commun. 372, 379–382. Heikal, A., Box, K., Rothnie, A., Storm, J., Callaghan, R., and Allen, M. (2009). The stabilisation of purified, reconstituted P-glycoprotein by freeze drying with disaccharides. Cryobiology 58, 37–44.

230

T. K. Ritchie et al.

Higgins, C. F. (2007). Multiple molecular mechanisms for multidrug resistance transporters. Nature 446, 749–757. Higgins, C. F., Callaghan, R., Linton, K. J., Rosenberg, M. F., and Ford, R. C. (1997). Structure of the multidrug resistance P-glycoprotein. Semin. Cancer Biol. 8, 135–142. Hjelmeland, L. M., and Chrambach, A. (1984). Solubilization of functional membrane proteins. Methods Enzymol. 104, 305–318. Kijac, A. Z., Li, Y., Sligar, S. G., and Rienstra, C. M. (2007). Magic-angle spinning solid-state NMR spectroscopy of nanodisc-embedded human CYP3A4. Biochemistry 46, 13696–13703. Kim, I. W., Peng, X. H., Sauna, Z. E., FitzGerald, P. C., Xia, D., Muller, M., Nandigama, K., and Ambudkar, S. V. (2006). The conserved tyrosine residues 401 and 1044 in ATP sites of human P-glycoprotein are critical for ATP binding and hydrolysis: Evidence for a conserved subdomain, the A-loop in the ATP-binding cassette. Biochemistry 45, 7605–7616. Kimura, Y., Kioka, N., Kato, H., Matsuo, M., and Ueda, K. (2007). Modulation of drug-stimulated ATPase activity of human MDR1/P-glycoprotein by cholesterol. Biochem. J. 401, 597–605. Leitz, A. J., Bayburt, T. H., Barnakov, A. N., Springer, B. A., and Sligar, S. G. (2006). Functional reconstitution of beta 2-adrenergic receptors utilizing self-assembling nanodisc technology. Biotechniques 40, 601–602, 604, 606, 608, 610, 612. Li, Y., Kijac, A. Z., Sligar, S. G., and Rienstra, C. M. (2006). Structural analysis of nanoscale self-assembled discoidal lipid bilayers by solid-state NMR spectroscopy. Biophys. J. 91, 3819–3828. Liu, R., Siemiarczuk, A., and Sharom, F. J. (2000). Intrinsic fluorescence of the P-glycoprotein multidrug transporter: Sensitivity of tryptophan residues to binding of drugs and nucleotides. Biochemistry 39, 14927–14938. Loo, T. W., and Clarke, D. M. (2005). Do drug substrates enter the common drug-binding pocket of P-glycoprotein through ‘‘gates’’? Biochem. Biophys. Res. Commun. 329, 419–422. Loo, T. W., Bartlett, M. C., and Clarke, D. M. (2003). Drug binding in human P-glycoprotein causes conformational changes in both nucleotide-binding domains. J. Biol. Chem. 278, 1575–1578. Lu, P., Liu, R., and Sharom, F. J. (2001). Drug transport by reconstituted P-glycoprotein in proteoliposomes. Effect of substrates and modulators, and dependence on bilayer phase state. Eur. J. Biochem. 268, 1687–1697. Marin, V. L., Bayburt, T. H., Sligar, S. G., and Mrksich, M. (2007). Functional assays of membrane-bound proteins with SAMDI-TOF mass spectrometry. Angew. Chem. Int. Ed. 46, 8796–8798. McDevitt, C. A., Shintre, C. A., Grossmann, J. G., Pollock, N. L., Prince, S. M., Callaghan, R., and Ford, R. C. (2008). Structural insights into P-glycoprotein (ABCB1) by small angle X-ray scattering and electron crystallography. FEBS Lett. 582, 2950–2956. Mi, L. Z., Grey, M. J., Nishida, N., Walz, T., Lu, C., and Springer, T. A. (2008). Functional and structural stability of the epidermal growth factor receptor in detergent micelles and phospholipid nanodiscs. Biochemistry 47, 10314–10323. Morrissey, J. H., Pureza, V., Davis-Harrison, R. L., Sligar, S. G., Ohkubo, Y. Z., and Tajkhorshid, E. (2008). Blood clotting reactions on nanoscale phospholipid bilayers. Thromb. Res. 122, S23–S26. Naito, M., and Tsuruo, T. (1995). Reconstitution of purified P-glycoprotein into liposomes. J. Cancer Res. Clin. Oncol. 121, 582–586.

Reconstitution of Membrane Proteins in Phospholipid Bilayer Nanodiscs

231

Nath, A., Atkins, W. M., and Sligar, S. G. (2007a). Applications of phospholipid bilayer nanodiscs in the study of membranes and membrane proteins. Biochemistry 46, 2059–2069. Nath, A., Grinkova, Y. V., Sligar, S. G., and Atkins, W. M. (2007b). Ligand binding to cytochrome P450 3A4 in phospholipid bilayer nanodiscs: The effect of model membranes. J. Biol. Chem. 282, 28309–28320. Nath, A., Koo, P. K., Rhoades, E., and Atkins, W. M. (2008). Allosteric effects on substrate dissociation from cytochrome P450 3A4 in nanodiscs observed by ensemble and singlemolecule fluorescence spectroscopy. J. Am. Chem. Soc. 130, 15746–15747. Oesterhelt, D., and Stoeckenius, W. (1974). Isolation of the cell membrane of Halobacterium halobium and its fractionation into red and purple membrane. Methods Enzymol. 31, 667–678. Orlowski, S., Martin, S., and Escargueil, A. (2006). P-glycoprotein and ‘lipid rafts’: Some ambiguous mutual relationships (floating on them, building them or meeting them by chance?). Cell. Mol. Life Sci. 63, 1038–1059. Papermaster, D. S. (1982). Preparation of retinal rod outer segments. Methods Enzymol. 81, 48–52. Polli, J. W., Wring, S. A., Humphreys, J. E., Huang, L., Morgan, J. B., Webster, L. O., and Serabjit-Singh, C. S. (2001). Rational use of in vitro P-glycoprotein assays in drug discovery. J. Pharmacol. Exp. Ther. 299, 620–628. Qu, Q., Russell, P. L., and Sharom, F. J. (2003). Stoichiometry and affinity of nucleotide binding to P-glycoprotein during the catalytic cycle. Biochemistry 42, 1170–1177. Rigaud, J. L. (2002). Membrane proteins: Functional and structural studies using reconstituted proteoliposomes and 2-D crystals. Braz. J. Med. Biol. Res. 35, 753–766. Rigaud, J. L., Levy, D., Mosser, G., and Lambert, O. (1998). Detergent removal by nonpolar polystyrene beads. Applications to membrane protein reconstitution and twodimensional crystallization. Eur. Biophys. J. 27, 305–319. Rosenberg, M. F., Callaghan, R., Modok, S., Higgins, C. F., and Ford, R. C. (2005). Three-dimensional structure of P-glycoprotein: The transmembrane regions adopt an asymmetric configuration in the nucleotide-bound state. J. Biol. Chem. 280, 2857–2862. Shaw, A. W., Pureza, V. S., Sligar, S. G., and Morrissey, J. H. (2007). The local phospholipid environment modulates the activation of blood clotting. J. Biol. Chem. 282, 6556–6563. Sligar, S. G. (2003). Finding a single-molecule solution for membrane proteins. Biochem. Biophys. Res. Commun. 312, 115–119. Takano, M., Hasegawa, R., Fukuda, T., Yumoto, R., Nagai, J., and Murakami, T. (1998). Interaction with P-glycoprotein and transport of erythromycin, midazolam and ketoconazole in Caco-2 cells. Eur. J. Pharmacol. 358, 289–294. Taylor, A. M., Storm, J., Soceneantu, L., Linton, K. J., Gabriel, M., Martin, C., Woodhouse, J., Blott, E., Higgins, C. F., and Callaghan, R. (2001). Detailed characterization of cysteine-less P-glycoprotein reveals subtle pharmacological differences in function from wild-type protein. Br. J. Pharmacol. 134, 1609–1618. Wang, E. J., Lew, K., Casciano, C. N., Clement, R. P., and Johnson, W. W. (2002). Interaction of common azole antifungals with P glycoprotein. Antimicrob. Agents Chemother. 46, 160–165. Zhao, J., Das, A., Schatz, G. C., Sligar, S. G., and Van Duyne, R. P. (2008). Resonance localized surface plasmon spectroscopy: Sensing substrate and inhibitor binding to cytochrome P450. J. Phys. Chem. C 112, 13084–13088. Zhou, S. F. (2008). Structure, function and regulation of P-glycoprotein and its clinical relevance in drug disposition. Xenobiotica 38, 802–832. Zolnerciks, J. K., Wooding, C., and Linton, K. J. (2007). Evidence for a Sav1866-like architecture for the human multidrug transporter P-glycoprotein. FASEB J. 21, 3937–3948.

This page intentionally left blank

C H A P T E R

T W E LV E

DNA-Controlled Assembly of Liposomes in Diagnostics Ulla Jakobsen and Stefan Vogel Contents 234 235 236 237 237 240 242 244 244 244 245 245 246 247 247 248

1. Introduction 2. Probe Design 2.1. Double membrane anchor single DNA-probe design 2.2. Single membrane anchor dual-probe design 2.3. Chain length dependence 2.4. Thermal denaturation experiments 2.5. Light scattering 3. General Description of Materials and Techniques 3.1. Measurement of transition temperatures (Tm) 3.2. Experimental procedure for liposome assembly 3.3. Preparation of POPC liposomes 3.4. DNA synthesis of lipid-modified DNA conjugates 3.5. HPLC purification 4. Concluding Remarks Acknowledgment References

Abstract DNA-encoding of solid nanoparticles requires surface chemistry, which is often tedious and not generally applicable. In the presented method, noncovalent attachment of DNA is used to assemble soft nanoparticles (liposomes) in solution. This process displays remarkably sharp thermal transitions from the assembled to disassembled state, thus enabling easy and fast detection of polynucleotides (e.g., DNA or RNA), including single nucleotide polymorphisms (SNPs).

Nucleic Acid Center, University of Southern Denmark, Odense, Denmark Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64012-X

#

2009 Elsevier Inc. All rights reserved.

233

234

Ulla Jakobsen and Stefan Vogel

1. Introduction DNA-controlled assembly of solid nanoparticles requires surface chemistry to stabilize colloidal nanoparticles chemically and to avoid nonspecific non-DNA-controlled aggregation. For most solid nanoparticles, surface chemistry needs to be developed specifically for each material based on the respective chemical composition of the particles. A very broad range of different chemical compositions (e.g., metals, metal oxides, metal sulfides) and possible combinations in hybrid materials exist, but only very few general procedures are available for attachment of DNA to the respective surfaces. Many of the chemical procedures used for surface stabilization and attachment of recognition sites rely on weak bonding, such as sulfur chemistry for gold surfaces, and some of the chemical procedures have to be adjusted according to the desired particle size during the particle growing process. In light of the many difficulties and tedious practical procedures, noncovalent attachment of DNA to soft nanoparticles such as liposomes becomes an attractive technology. A number of applications such as DNA-controlled tethering to surfaces toward liposome arrays and DNA-controlled fusion of liposomes have been reported (Chandra et al., 2006; Pfeiffer and Ho¨o¨k, 2004; Stengel et al., 2007; Chan et al., 2009; Yoshina-Ishii and Boxer, 2003, Yoshina-Ishii et al., 2005; Zhang et al., 1996). Our approach describes DNA-controlled assembly of liposomes in solution and on solid supported membranes ( Jakobsen et al., 2008a). Remarkably sharp thermal transitions between assembled and disassembled state and discrimination of single mismatches, deletions, and insertions in the resulting DNA-target duplex allow applications in the diagnostics of single nucleotide polymorphisms (SNPs). The basic design is based on the noncovalent attachment of DNAprobes (single-stranded DNA with terminal lipid membrane anchors, for membrane anchor structure see Fig. 12.5) to a liposome surface (Fig. 12.1). The complementary polynucleotide target can be unmodified DNA or RNA that allows the method to be used in the detection of biological polynucleotide targets ( Jakobsen et al., 2008). The conformationally flexible single-stranded DNA-probe is presumably anchored reversibly on the liposomes since DNA strands permanently anchored at both ends of the DNA strand would not allow efficient hybridization to a complementary target sequence. After hybridization of the DNA-probe to a complementary target DNA, both ends of the corresponding duplex are not able to be simultaneously anchored into the same liposome, as this would require bending of the rigid double-stranded DNA (for mechanical properties of duplex DNA, see Cloutier and Widom, 2005). Therefore, one of the membrane anchors is released into solution and subsequent interliposomal membrane anchoring occurs and is highly favored. The interliposomal anchoring of the DNA duplex leads to

DNA-Controlled Assembly of Liposomes in Diagnostics

235

Liposome(s)

= Lipid membrane anchor

= DNA target

= DNA probe

Figure 12.1 Double membrane anchor single DNA-probe design. Schematic representation of liposome aggregation upon duplex formation between a lipophilic probe DNA and an unmodified target DNA. Liposomes and DNA strands are not drawn to scale.

liposome cross-linking and liposome assembly (Fig. 12.1). The process continues until target or probe DNA strands have been consumed since liposomes are present in excess. The required duplex rigidity is not compromised by single mismatches but multiple mismatches are not tolerated. Cooperative effects such as DNA melting and the entropically favored disassembly of liposome aggregates are presumably responsible for the sharp thermal transitions observed by UV spectroscopy measurements. The narrow temperature range for the disassembly process (thermal denaturation is compared to unmodified DNA), and nonoverlapping thermal transitions for complementary and single-mismatched sequences can be utilized to turn the liposome assembly process into a powerful detection system for SNPs. The DNA-controlled assembly of liposomes is general and efficient for liposomes of sizes between 50 and 200 nm in solution and on supported membranes. In contrast to solid nanoparticles, the lipophilic DNA-probe strand is inserted noncovalently, which avoids development of tedious conjugation chemistry and enables a rapid assembly process (assembly process requires only minutes to occur at nM and mM DNAprobe concentrations). This method is, to the best of our knowledge, the first application of liposome assembly to the detection of polynucleotides with single mismatch discrimination power in solution ( Jakobsen et al., 2008).

2. Probe Design The key concept of the method is schematically illustrated in Fig. 12.1 for a double membrane anchor single DNA-probe design and in Fig. 12.2 for a single membrane anchor dual DNA-probe design. The membrane

236

Ulla Jakobsen and Stefan Vogel

Probe A Liposomes Target Probe B

= DNA probes

= DNA target

Figure 12.2 Single membrane anchor dual-probe design. Schematic representation of liposome aggregation upon duplex formation between a lipophilic probe DNA and an unmodified target DNA. Liposomes and DNA strands are not drawn to scale.

anchoring is spontaneous upon addition of liposomes to the mixture of DNA-probe and target. The membrane anchor of the DNA-probe consists of a macrocyclic building block with at least two lipid substituents (e.g., cholesteryl or saturated alkyl chains, for chemical structures see Fig. 12.5) to ensure permanent anchoring of the DNA-probe into the liposome surface. The macrocyclic membrane anchor scaffold (Vogel et al., 2003, 2006) can be substituted with lipids at two positions and subsequently incorporated into DNA using standard automated DNA-synthesis.

2.1. Double membrane anchor single DNA-probe design The method is based on lipid-modified DNA-probes and liposomes (POPC). Introduction of lipid membrane anchors in DNA-probes has been shown earlier to increase the thermal stability of the corresponding DNA duplexes ( Jakobsen et al., 2007a,b, 2008a; Rohr and Vogel, 2006). The lipid membrane anchors are incorporated at both ends of the DNAprobe strand to enable dual anchoring of the DNA-probe into a lipid bilayer (e.g., liposomes). In order to avoid self-aggregation and surfactant properties of the corresponding DNA-probes in aqueous solution, we added up to three T-nucleosides on both ends of the DNA-probe which sufficiently suppressed self-aggregation at concentrations below 2 mM. The singlestranded DNA-probe is partitioned noncovalent into the liposome bilayer and dynamically (reversible) anchored in contrast to covalently bound DNA on solid nanoparticle surfaces (e.g., Au-surfaces, Li et al., 2002; Mirkin et al., 1996). The dynamic anchoring is a requirement to enable hybridization to a

DNA-Controlled Assembly of Liposomes in Diagnostics

237

complementary target DNA since a DNA permanently anchored on both ends of the DNA strand would not be able to hybridize with a complementary target strand. Presumably, one of the two anchors is released from the membrane bilayer while hybridizing to a complementary target strand and does not anchor into the same liposome after hybridization. The corresponding double-stranded DNA duplex is mechanically very stiff (see Cloutier and Widom, 2005) and forces one of the membrane anchors to be released into the solution since anchoring of the double-stranded DNA into the same liposome would require bending of the duplex which is energetically very unfavorable for short double-stranded DNA. The released membrane anchor is subsequently anchored into another liposome and will initiate assembly of liposomes (Fig. 12.1). The exclusive interliposomal anchoring leads to cross-linking and formation of liposome aggregates with aggregate sizes depending on DNA-probe and target concentrations.

2.2. Single membrane anchor dual-probe design Another design consists of two probes (A and B) in which both DNA sequences are complementary to half of the target sequence and feature one terminal membrane anchor for each DNA-probe strand. The dualprobe design allows encoding of two batches of liposomes with separate addition of the respective DNA-probe. The liposome assembly is subsequently initiated by addition of a target with a sequence complementary to both probes A and B (Fig. 12.2). The target bridges the DNA probes A and B, and forms a mechanically stiff duplex. The process of liposome assembly, based on a design with two DNA probes, is equally efficient compared to the single DNA-probe design, but requires two separate DNA probes and introduces a DNA duplex with two double-stranded subunits of different thermal stability. The probe design becomes more complicated since a mismatch in the thermally more stable region of the duplex may not be destabilizing enough to see a difference in the overall thermal stability of the DNA-probe:target duplex. However, using appropriate sequence design, the mismatch discrimination and remarkable sharpness of the corresponding thermal transitions equals that of the double membrane anchor single DNA-probe design.

2.3. Chain length dependence Liposome assembly depends highly on the nature of the membrane anchor (chain length and chemical structure). For saturated alkyl chains as membrane anchors, lipids of a minimal chain length of 12-carbons are required to initiate assembly of liposomes; however, the process becomes significantly more efficient when 14-carbon or 16-carbon chains are used. Membrane anchors with increased chain length of up to 20-carbon chains lead to

238

Ulla Jakobsen and Stefan Vogel

0.34 5⬘-TGTGGAAGAAGTTGGT 3⬘-TTTX2C16ACACCTTCTTCAACCACX2C16TTT

Apparent absorbance at 260 nm

0.32 0.30 0.28 0.26

5⬘-TGTGGAAGAAGTTGGT 3⬘-TTTX2C12ACACCTTCTTCAACCACX2C12TTT

0.24 0.22

5⬘-TGTGGAAGAAGTTGGT

0.20

3⬘-TTTX2C10ACACCTTCTTCAACCACX2C10TTT

0.18 0.16 20

30

40

50

60

70

80

90

Temperature (⬚C)

Figure 12.3 UV-monitored thermal denaturation data for DNA-probe target duplexes in the presence of liposomes and saturated lipid membrane anchors of different chain length.

DNA-probes that are anchored so strongly into the liposome surface that hybridization does not occur for the measured DNA-probes with a duplex length up to 17 base pairs (bp). Structurally different lipids such as conformationally less flexible cholesteryl-derived membrane anchors also result in efficient membrane anchoring but display, in contrast to 16-carbon saturated lipid chains (e.g., palmityl), sequence dependency. The efficiency of DNAcontrolled assembly with different lipid membrane anchor length is shown in Fig. 12.3 (Jakobsen and Vogel, 2008b). For practical purposes, a chain length of 16-carbons has been sufficient for the permanent anchoring of DNA-probes ranging from 9-mer to 27-mer DNA probes. The lipid anchor scaffold is based on an aza crown ether and the macrocyclic core structure has two secondary amine functions available for substitution by various lipids (Fig. 12.5). Conformational restriction by a pyridino group in the macrocycle separates the corresponding lipid chains from each other (Fig. 12.5) which may enable independent anchoring of each lipid substituent into the liposome surface. Our experiments with a single 16-carbon chain lipid membrane anchor correspond to previously published data, that two lipid anchors are required to permanently anchor DNA (20-mer duplex) into lipid membranes (Pfeiffer and Ho¨o¨k, 2004). DNA-probes with a single 16-carbon chain lipid membrane anchor and an otherwise identical

239

DNA-Controlled Assembly of Liposomes in Diagnostics

5´ O

O

O N H

O

3´

Y

Figure 12.4 Single palmitoyl-chain membrane anchor building block Y. Experiments are conducted with identical sequences except for the membrane anchor (62 nM probe DNA concentration in 110 mM Naþ, 50 -TTTYTGTGGAAGAAGTTGGTGYTTT and 30 -TTTYACACCTTCTTCAACCACYTTT with the respective complementary DNA, HEPES buffer, pH 7.0).

3´ O

O−

P

H

O X2Chol :

R=

X2C10 :

R=

X2C12 :

R=

X2C16 :

R=

X2C20 :

R=

H

H

N O

O

R

N

N N

O

R

O 5´

Figure 12.5 General structure of the lipid-modified macrocyclic monomer and lipid substitution pattern.

sequence design have not been able to initiate liposome assembly in solution, which was attributed to insufficient membrane anchoring of the corresponding duplexes. However, thermal denaturation studies in the absence of liposomes have shown that hybridization occurs and the corresponding duplex is not destabilized by modification Y (Fig. 12.4). The lipid-substituted macrocycles (Fig. 12.5) can be incorporated into DNA by automated DNA-synthesis using the phosphoramidite approach at any desired position in a given sequence including the possibility for multiple incorporations.

240

Ulla Jakobsen and Stefan Vogel

The chemical synthesis of aza macrocycles is straightforward and can easily be upscaled to multigram amounts of the corresponding phosphoramidite building blocks (see Fig. 12.5) for automated DNA-synthesis ( Jakobsen et al., 2007a,b; Rohr and Vogel, 2006).

2.4. Thermal denaturation experiments The process of liposome assembly is reversible and can be monitored by ultraviolet (UV) spectroscopy and dynamic light scattering (DLS). Monitoring of the process by UV spectroscopy can be done at 260 nm or higher wavelengths (measured between 260 and 420 nm), although with decreasing signal intensity toward longer wavelengths. The assembly of liposomes is fully reversible and can be repeated by cycling the temperature around the thermal denaturation temperature (Tm) of the corresponding duplex. The thermal denaturation curves, as monitored by UV spectroscopy at 260 nm, are inversed compared to measurements in the absence of liposomes and of much higher intensity (2 orders of magnitude). This behavior is caused by the presence of liposomes which assemble at a temperature below the Tm of the corresponding DNA-probe:target duplex and disassemble at a temperature above the Tm. The process is monitored at 260 nm which represents the average absorption maximum for natural DNA bases but at 62 nM concentration, a concentration commonly used for the experiments with liposomes, the DNA absorption changes upon thermal denaturation will not be visible but only the changes in scattering intensity of the liposomes. The optical properties of the solution after addition of liposomes are no longer dominated by the UV absorption of the DNA but the apparent absorption caused by light scattering of the liposomes. At temperatures below Tm of the corresponding duplex, liposome assembly occurs and leads to high apparent absorbance through increased light scattering which is caused by increased average particle sizes of the liposome aggregates in solution. Heating of the liposome aggregates above the Tm of the corresponding duplex cause disassembly of the aggregates and results in strongly decreased apparent absorption due to decreased light scattering of individual liposomes compared to the liposome aggregates (Fig. 12.6). For diagnostic applications, the most important property of the DNAcontrolled liposome assembly processes are the remarkably sharp thermal transitions observed during denaturation of the corresponding duplexes. The narrow temperature range for the disassembly process of only 2–4  C compared to 15–20  C for thermal denaturation of duplexes in the absence of liposomes turns this process into a powerful method for the detection of SNPs. Thermal melting profiles with similar thermal transitions have only been reported for a nanoparticle system based on DNA-controlled assembly of gold colloids (Mirkin et al., 1996; Rosi and Mirkin, 2005; Taton et al., 2000).

241

DNA-Controlled Assembly of Liposomes in Diagnostics

Apparent absorbance at 260 nm

0.00

3⬘-TTTXACACCTTCTTCAACCACXTTT 5⬘-TGTGGAAGAAGTTGGTG

Mismatch position

−0.04 ΔT > 10 ⬚C

−0.08

C:G - matched sequence

C:T C:A C:C

−0.12

Single mismatches

−0.16 20

30

40

50 60 Temperature (⬚C)

70

80

Figure 12.6 UV spectroscopy Tm data of a 17-mer DNA-probe hybridized to the corresponding single mismatches (17-mer DNA-target strands) and a fully matched target sequence.

The nonoverlapping thermal denaturation curves simplify data analysis significantly which is an important advantage over fluorophore-based detection systems and allow application of this method for the analysis of weakly discriminated mismatches in DNA-targets. The current system can detect low nM concentrations of target DNA using a label-free setup. Single mismatch discrimination data in the presence and absence of liposomes resemble closely data for unmodified DNA, which allows for traditional probe sequence design (Table 12.1). Future improvements of the currently label-free method will take advantage of the structure of liposomes which allows encapsulation of labels (e.g., fluorophores) or smaller nanoparticles (e.g., quantum dots) inside the liposome to shift the detection wavelength to the spectral range of visual light or magnetic particles to separate the target detection components from a complex sample matrix (e.g., serum). Discrimination of SNPs based on DTm-values from thermal denaturation profiles as measured by UV spectroscopy is comparable to DNA in the absence of liposomes but with much sharper transitions. However, the thermal stability of the corresponding duplexes is considerably increased in the presence of liposomes as shown in Table 12.1, since the recorded Tm-values are very similar in the absence and presence of liposomes despite a much lower DNA-probe concentration of only 62 nM compared to 1 mM in the absence of liposomes. A lower DNA-probe concentration should otherwise destabilize the duplex and result in lower

242

Ulla Jakobsen and Stefan Vogel

Table 12.1 Mismatch discrimination dataa,b Tm ( C)

DTm ( C)

Tm ( C)

Duplex

No.

1 2 3 4 a b

c d

DTm ( C)

30 TTTX2C16ACACCTTCTT CAACCACX2C16TTT

With liposomesc

Without liposomesd

50 -TGTGGAAGAAGTT GGTG 50 -TGTGTAAGAAGTT GGTG 50 -TGTGAAAGAAGTT GGTG 50 -TGTGCAAGAA GTTGGTG

48

–

47

–

36

 12

37

 10

38

 10

38

 11

36

 12

37

 11

X denotes the polyaza crown ether base surrogate (see Fig. 12.5). Conditions: Tm values ( C) (DTm ¼ change in Tm value) calculated relative to the DNA:DNA reference duplex measured as the maximum of the first derivative of the melting curve (A260 vs. temperature) recorded in medium salt buffer (10 mM HEPES, 110 mM Naþ, pH 7.0). 62 nM concentrations of the two complementary strands in the presence of liposomes. 1 mM concentrations of the two complementary strands in the absence of liposomes. Exp. error: 1  C.

Tm-values but this is presumably counterbalanced by the effect of the liposomes and the increased thermal stability by introduction of the lipid membrane anchor building blocks (Rohr and Vogel, 2006).

2.5. Light scattering The readout of thermal denaturation experiments based on UV spectroscopic measurements is mainly attributed to scattered light from liposomes and liposome aggregates. The confirmation of liposome assembly upon addition of a complimentary target strand has been achieved by DLS titration studies. Addition of 0.25, 0.5, 0.75, and 1 equiv. of complementary DNA below the Tm of the corresponding DNA-probe:target duplex results in a significant increase of the average particle size from 72 nm (DNA without a complementary target) to 145 nm (1 equiv. of target, Figs. 12.7 and 12.8). The increase in average particle size corresponds linearly to the increase in target DNA concentration (Fig. 12.8) and confirms that aggregation is initiated by addition of a complementary target sequence. The DLS titration is the method of choice to follow the DNA-controlled aggregation over time and shows the fast kinetics of the process (liposome aggregation within minutes). Addition of a complementary target causes liposome aggregation within less than 15 min for a 17-mer DNA-probe (50 -TTT-X-TGTGGAAGAAGTTGGTG-X-TTT:30 -ACACCTTCTTCAACCAC, X ¼ X2C16).

243

DNA-Controlled Assembly of Liposomes in Diagnostics

0.5 eq. cDNA 0.75 eq. cDNA

0.25 eq. cDNA

100

ssDNA + liposomes

Intensity

1 eq. cDNA

50

0

50

100

150 Size (nm)

200

250

300

Figure 12.7 Dynamic light scattering measurements of liposome assembly. Singlestranded DNA (ssDNA, TTT-X-TGTGGAAGAAGTTGGTG-X-TTT, X ¼ X2C16, for the chemical structure see Fig. 12.5) is titrated with complementary DNA (cDNA, 30 -ACACCTTCTTCAACCAC, 0.25, 0.5, 0.75, 1.0 equiv.). 150 140 130

Size (nm)

120 110 100 90 80 70 0.0

0.2

0.4 0.6 Complementary DNA (equivalents)

0.8

1.0

Figure 12.8 Dynamic light scattering measurements of liposome assembly. Linear fit of average particle size versus the number of equivalents complementary target.

244

Ulla Jakobsen and Stefan Vogel

3. General Description of Materials and Techniques Materials: 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC; lyophilized powder) was purchased from Avanti Polar Lipids, and solvents (HPLC grade) were purchased commercially and used without further purification. Chemicals for DNA synthesis were purchased from Glen Research or Proligo. All other chemicals were obtained from Sigma in the best quality available. Ultrapure Milli-Q water is used in all experiments. 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid (HEPES) buffer (10 mM HEPES, 110 mM Naþ) is prepared by mixing appropriate amounts of HEPES, HEPES sodium salt, and NaCl followed by correction of pH to 7.0.

3.1. Measurement of transition temperatures (Tm) Thermal denaturation experiments are carried out on a Perkin-Elmer UV/VIS spectrometer Lambda 30 with a PTP-6 (Peltier Temperature Programmer) device by using PE TempLab 2.0 software or a Varian Cary 3E/300 equipped with a Peltier controlled 6  6 sample changer and Cary WinUV software. Melting temperatures (Tm, in  C) are determined as a first derivative of thermal denaturation curves, which are obtained by recording absorbance at 260 nm as a function of temperature at a rate of 0.5  C/min for measurements with liposomes and of 1  C/min for measurements without liposomes. The solutions are heated to 90  C, maintained for 5 min at this temperature, and then gradually cooled before performing thermal denaturation experiments. All melting temperatures are reported with an uncertainty of 1  C, as determined from multiple experiments.

3.2. Experimental procedure for liposome assembly A particular order for the mixing of components (e.g., liposomes, DNAprobes, target strand) is not required for the experiments mentioned. In a typical experiment we use POPC liposomes with an average diameter of 65  15 nm and a lipid concentration of 10 mM, prepared by extrusion through 50 nm filters. A final concentration of 0.5 mM is used for all experiments. Addition of lipophilic probe DNA in HEPES buffer solution (pH 7.0, 110 mM Naþ, 10 mM HEPES, 62 nM DNA-probe strand) followed by addition of the target sequence (62 nM DNA target) initiates assembly. Studies at higher DNA concentrations (2 mM ) results in particle aggregation as observed visually by a rapidly increasing turbidity and finally precipitation of the liposome aggregates with fully intact liposomes as

245

DNA-Controlled Assembly of Liposomes in Diagnostics

shown by an arsenazo III dye assay ( Jakobsen et al., 2008). Experiments with lower target concentrations (target concentration < 100 nM) produce stable aggregates as observed by light scattering, but without macroscopic precipitation, even after prolonged time (72 h).

3.3. Preparation of POPC liposomes POPC is suspended in aqueous HEPES buffer at 10 mM lipid concentration with application of sonication or stirring to help the formation of a uniform suspension. Liposomes are prepared by repeated extrusion (10 times) through double polycarbonate filters with a 50 or 100 nm pore size using compressed N2 (20–40 bar) and a LIPEXTM Extruder from Northern Lipids (Olson et al., 1979).

3.4. DNA synthesis of lipid-modified DNA conjugates Synthesis of lipid-modified DNA conjugates is performed in 0.2 or 1 mmol scale on an automated DNA synthesizer using the phosphoramidite approach (Caruthers, 1991). The facile chemical synthesis results in lipophilic phosphoramidite building blocks which can be used during DNA-synthesis with a slightly modified standard protocol. The solubility of the lipid-substituted macrocycles in acetonitrile is decreased with increasing chain length which requires use of solvents such as 1,2-dichloroethane instead of acetonitrile for the lipid membrane anchor phosphoramidite building blocks (Fig. 12.9).

O NC

P N O

X2C10, X2C12, X2C16, X2C20 : R = n = 1, 3, 6, 10

N O

R

O

N

N N

R

ODMTr X2Chol

H :R=

H

H

O

Figure 12.9 General structure and substitution pattern of lipophilic phosphoramidite building blocks for automated DNA-synthesis.

246

Ulla Jakobsen and Stefan Vogel

Prior to and after coupling of the phosphoramidite, a washing step with 1,2-dichloroethane is suggested to avoid precipitation of remaining phosphoramidite solution in the tubing of the DNA synthesizer. All of the phosphoramidite building blocks have a considerably higher molecular weight than standard nucleoside phosphoramidites (A,T,G,C) and sterically demanding substituents (e.g., cholesteryl, palmityl); therefore, prolonged coupling times (30 min) and more concentrated solutions (0.1 M instead of 0.05 M) are used to improve coupling yields and to shorten coupling times. Standard activators (tetrazole, dicyanoimidazol) lead to low coupling yields (<50%) and have been replaced by pyridinium hydrochloride which consistently enhances coupling yields to more than 90%. The large number of basic amine functionalities in the macrocycle requires more activator to counterbalance the buffer effect caused by the basic amino functions (5 equiv. activator). The synthesis is completed by deprotection of the oligomer (DMT-group removal), deprotection of standard DNA nucleobases and cleavage from the solid support (CPG-controlled pore glass, 500 A˚) with concentrated ammonia (30% NH3 aqueous solution) at 55  C for 12 h. The resulting solution is filtered and concentrated in vacuum.

3.5. HPLC purification The remaining solid is dissolved in 100 ml water, centrifuged (10,000 rpm) to spin down insoluble material in turbid samples and subsequently purified by HPLC on a DIONEXÒ Acclaim C18 reversed phase column. The eluent system consists of a 0.05-M triethylamine acetate buffer (A) and acetonitrile/water mixture (B, 75:25). The sample is eluted with a gradient system starting from 100% 0.05 M triethylamine acetate buffer to 70% B after 10 min to 100% B after 30 min and continued isocratic with 100% B until 60 min run time at a flow of 1 ml/min (Fig. 12.10). The fractions are freeze dried using a HETO110 freeze dryer with a  100  C freeze trap (allows direct freeze drying of samples with acetonitrile content) and the remaining solid is dissolved in 1 ml of water. DNA concentrations are determined by measuring the optical density at 260 nm. The representative HPLC chromatogram shows the major product with the desired sequence as well as the minor component with only one membrane anchor incorporation caused by incomplete coupling of the corresponding phosphoramidite building block (Fig. 12.9). In all synthesized double lipid membrane anchor DNA-probes, both products can be found in a ratio depending on the coupling yield of the second incorporation. The fractions are further characterized by MALDI-TOF mass spectrometry on a Voyager Elite (PerSeptive Biosystems) apparatus recording the spectra in positive ion mode using delayed ion extraction.

247

DNA-Controlled Assembly of Liposomes in Diagnostics

mAU

700 600

Double membrane anchor DNA-conjugate

500

5⬘-TTT-X-TGT-GGA-AGA-AGT-TGG-TG-X-TTT

400 300 Single membrane anchor DNA-conjugate 200 5⬘-TTT-X-TGT-GGA-AGA-AGT-TGG-TG

100 0 0

10

20

30 40 Retention time [min]

50

60

70

Figure 12.10 Representative HPLC chromatogram for purification of an amphiphilic DNA-probe, X ¼ X2C20 (for chemical structure of X see Fig. 12.5).

4. Concluding Remarks A general method for DNA-controlled assembly of liposomes and applications for detection of polynucleotides (DNA or RNA), based on DNA-mediated assembly of lipid bilayer membranes, is presented. The method is based on a double-membrane anchor single DNA-probe or a single-membrane anchor dual-probe design. The DNA-probes insert noncovalently into liposomes. In the presence of a complementary DNA-strand (unmodified), assembly of liposomes to larger aggregates is observed. The liposome aggregates disassemble by heating and display remarkable sharp thermal transitions which allow easy detection of SNPs. The versatile assembly strategy can be generally applied to immobilization and tethering of liposomes, as well as DNA-encoding of liposomes.

ACKNOWLEDGMENT We greatly appreciate funding from The Danish National Research Foundation.

248

Ulla Jakobsen and Stefan Vogel

REFERENCES Caruthers, M. H. (1991). Chemical synthesis of DNA and DNA analogs. Acc. Chem. Res. 24, 278–284. Chan, Y. H., van Lengerich, B., and Boxer, S. G. (2009). Effects of linker sequences on vesicle fusion mediated by lipid-anchored DNA oligonucleotides. Proc. Natl. Acad. Sci. USA 106, 979–984. Chandra, R. A., Douglas, E. S., Mathies, R. A., Bertozzi, C. R., and Francis, M. B. (2006). Programmable cell adhesion encoded by DNA hybridization. Angew. Chem. Int. Ed. Engl. 45, 896–901. Cloutier, T. E., and Widom, J. (2005). DNA twisting flexibility and the formation of sharply looped protein-DNA complexes. Proc. Natl. Acad. Sci. USA 102, 3645–3650. Jakobsen, U., Rohr, K., Madsen, R. K., and Vogel, S. (2007a). Polyaza crown ether as nonnucleosidic building blocks in DNA-conjugates. Nucleosides Nucleotides Nucleic Acids 26, 1221–1224. Jakobsen, U., Rohr, K., and Vogel, S. (2007b). Toward a catalytic site in DNA: Polyaza crown ether as non- nucleosidic building blocks in DNA conjugates. Nucleosides Nucleotides Nucleic Acids 26, 1419–1422. Jakobsen, U., Simonsen, A. C., and Vogel, S. (2008a). DNA-controlled assembly of soft nanoparticles. J. Am. Chem. Soc. 130, 10462–10463. Jakobsen, U., and Vogel, S. (2008b). Lipophilic DNA-conjugates: DNA controlled assembly of liposomes. Nucleic Acids Symp. Ser. (Oxf ). 223–224. Li, Z., Jin, R. C., Mirkin, C. A., and Letsinger, R. L. (2002). Multiple thiol-anchor capped DNA-gold nanoparticle conjugates. Nucleic Acids Res. 30, 1558–1562. Mirkin, C. A., Letsinger, R. L., Mucic, R. C., and Storhoff, J. J. (1996). A DNA-based method for rationally assembling nanoparticles into macroscopic materials. Nature 382, 607–609. Olson, F., Hunt, C. A., Szoka, F. C., Jr., Vail, W. J., and Papahadjopoulos, D. (1979). Preparation of liposomes of defined size by extrusion through polycarbonate membranes. Biochim. Biophys. Acta 557, 9–23. Pfeiffer, I., and Ho¨o¨k, F. (2004). Bivalent cholesterol-based coupling of oligonucletides to lipid membrane assemblies. J. Am. Chem. Soc. 126, 10224–10225. Rohr, K., and Vogel, S. (2006). Polyaza crown ethers as non-nucleosidic building blocks in DNA conjugates: Synthesis and remarkable stabilization of dsDNA. Chembiochem 7, 463–470. Rosi, N. L., and Mirkin, C. A. (2005). Nanostructures in biodiagnostics. Chem. Rev. 105, 1547–1562. Stengel, G., Zahn, R., and Ho¨o¨k, F. (2007). DNA-induced programmable fusion of phospholipid vesicles. J. Am. Chem. Soc. 129, 9584–9585. Taton, T. A., Mucic, R. C., Mirkin, C. A., and Letsinger, R. L. (2000). The DNAmediated formation of supramolecular mono- and multilayered nanoparticle structures. J. Am. Chem. Soc. 122, 6305–6306. Vogel, S., Rohr, K., Dahl, O., and Wengel, J. (2003). A substituted triaza crown ether as a binding site in DNA-conjugates. Chem. Commun. 1006–1007. Yoshina-Ishii, C., and Boxer, S. G. (2003). Arrays of mobile tethered vesicles on supported lipid bilayers. J. Am. Chem. Soc. 125, 3696–3697. Yoshina-Ishii, C., Miller, G. P., Kraft, M. L., Kool, E. T., and Boxer, S. G. (2005). General method for modification of liposomes for encoded assembly on supported bilayers. J. Am. Chem. Soc. 127, 1356–1357. Zhang, G. R., Farooqui, F., Kinstler, O., and Letsinger, R. L. (1996). Informational liposomes: Complexes derived from cholesteryl-conjugated oligonucleotides and liposomes. Tetrahedron Lett. 37, 6243–6246.

C H A P T E R

T H I R T E E N

Soft Hybrid Nanostructures Composed of Phospholipid Liposomes Decorated with Oligonucleotides Martina Banchelli, Francesca Baldelli Bombelli, Debora Berti, and Piero Baglioni Contents 250 251 252 253 253 254 256

1. 2. 3. 4.

Introduction Materials Liposome Preparation and Determination of Lipid Content Incorporation of Oligonucleotides 4.1. Choice of the lipid anchor 4.2. Choice of the grafting density 5. Characterization of the Soft Hybrid Nanostructure 5.1. Purification of the liposome–oligonucleotide systems (size exclusion chromatography) 5.2. Dynamic light scattering measurements 6. Applications of Oligo-Decorated Liposomes 6.1. Hybridization with complementary oligonucleotides in solution 6.2. Hybridization with self-assembled DNA nanostructures 6.3. Different preparation procedures 6.4. Kinetics aspects 7. Challenges and Perspectives Acknowledgments References

256 259 262 262 265 266 273 275 276 276

Abstract This chapter reports on the design, preparation, and characterization of liposomes decorated with synthetic lipid–oligonucleotide conjugates. Several key parameters should be considered for a successful preparation of these

Department of Chemistry and CSGI, University of Florence, Sesto Fiorentino, Florence, Italy Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64013-1

#

2009 Elsevier Inc. All rights reserved.

249

250

Martina Banchelli et al.

functional nanostructures that can be employed further as building blocks in DNA-directed assembly of nano-objects. These parameters are reviewed explicitly in this report and their contributions are discussed.

1. Introduction The insertion of synthetic lipid–oligonucleotide conjugates into fluid amphiphilic surfaces, such as lipid vesicles, is currently gaining increasing interest for DNA-directed self-assembly into nanometer-scaled arrays of functional objects. The structural and functional properties of such nanomaterials are defined through the choice of the sequence of the oligonucleotide decoration and encoded by base pairing. These hybrid soft nanomaterials merge the unique features offered by DNA structural fidelity and specificity, to the characteristics of lipid self-assembly, in terms of ease of preparation, responsiveness and hierarchical aggregation in functional arrays of nano units (Chan et al., 2008). In the design of DNA/membrane hybrids, several points yet remain to be addressed, as the guidelines for the choice of the anchoring unit, which is the lipophilic portion of the oligonucleotide conjugate. For instance, the hydrophobic portion can be directly attached to the oligonucleotide or can be separated from the functional part by a hydrophilic spacer acting as a flexible joint that guarantees conformational freedom to the oligonucleotide. In terms of its length, the choice of the oligonucleotide portion, linked to the spacer–assembler portions through its 50 or 30 end, is mainly guided by the applications purposes and is responsible for the coupling specificity and efficiency. Lipid–oligonucleotide molecules are currently being used for two different purposes: (i) to direct their assembly to soft surfaces (e.g., SLB, supported lipid bilayer or other vesicles) (Pfeiffer and Ho¨o¨k, 2004; Stengel et al., 2007) regulated by several different mechanisms as reversible clustering (Beales and Kyle Vanderlick, 2007) or vesicle fusion (Chan et al., 2008), depending on the above-mentioned structural parameters of the synthetic lipid conjugate (Chan et al., 2009) and (ii) liposomes or SLB as scaffolds of DNA nanostructures to build soft DNA/lipid hybrids (Baldelli Bombelli et al., 2009). Our group has recently studied the incorporation of a cholesteryltetraethylene glycol (TEG) functionalized oligonucleotide in phospholipid vesicles (Banchelli et al., 2008). Its hybridization with a complementary strand has also been investigated in detail. The results have been interpreted in terms of the average distance between noncovalent grafting sites onto the membrane. Both the oligonucleotide conformation and the hybridization kinetics are strongly dependent on macromolecular crowding at the liposomal surface. These investigations have been extended to the construction

Liposomes Decorated with Oligonucleotides

251

of a lipid membrane/pseudohexagonal DNA hybrid, anchored to the membrane, thanks to a cholesteryl-TEG functionalization of one of the single strands. The effects of grafting density, lipid/DNA ratio, liposome number density, and preparation procedure on the final structure and yield of the resulting hybrid nanomaterial, are demonstrated. A particularly interesting result is the fact that above a critical grafting density, connected to the lipophilic oligonucleotide excluded area, coupling kinetics are slowed down with respect to strand pairing in solution. Conversely, when the average interoligonucleotide density on the liposomal surface is low, coupling is faster than in bulk medium. In this chapter, we will describe the preparation and structural characterization of oligonucleotide-decorated liposomes with particular emphasis on the experimental procedure and on the control parameters that will ultimately lead to the preparation of well-characterized systems, intended as building blocks for further assembly into nanostructured arrays. As an applicative example, we will report on their use as scaffolding hosts for DNA nanostructures that we have investigated in our laboratory, again stressing how the experimental preparation protocol affects the final yield of the hybrid nanostructures.

2. Materials POPC (1-palmitoyl, 2-oleoyl-sn-glycero-3-phosphocholine) is purchased from Avanti Polar Lipids Inc. (Alabama). Its purity is checked by TLC, to ensure the absence of oxidation or lysis products, and the lipid is used as received if it is intact. All other chemicals (TRIS base, NaCl) are purchased from Fluka (Milan, Italy) at the highest purity commercially available. All the modified and unmodified oligonucleotides presented in this chapter are synthesized with an Applied Biosystems 394 automated DNA/RNA synthesizer at the School of Chemistry in Southampton in the group of professor Tom Brown. Chemical modifications are introduced using the appropriate phosphoramidite monomers and incorporating them during oligonucleotide assembly. Purification of oligonucleotides is carried out by reversed phase HPLC (Banchelli et al., 2008). Cholesterol-derivative oligonucleotides (18-mers) are synthesized by modifying one end of the oligonucleotide sequence (50 -end or 30 -end). An oligoethylene glycol spacer is attached between the oligonucleotide sequence and the cholesteryl group. In this way, the hydrophilic group forms a bridge between the lipophilic group and the phosphate backbone of the DNA chain and enables the oligonucleotide to interact with lipid membranes and, at the same time, to remain in a hydrophilic environment at a fixed distance from the lipid surface.

252

Martina Banchelli et al.

3. Liposome Preparation and Determination of Lipid Content Phospholipid vesicles are generally not a thermodynamically stable amphiphilic phase, and do not form spontaneously; they are metastable structures with a shelf life of several months. The physical properties of lipid vesicles very much depend on how and under which conditions they are prepared. Thus, the mean size, the lamellarity, and the physical stability of the vesicles not only depend on the chemical structure of the lipid used, but particularly on the method of vesicle preparation as well. Physical instabilities of lipid vesicle systems involve vesicle aggregation and fusion. For the study here presented, POPC has been chosen as the lipid for vesicle preparation. POPC is a major component of biological membranes and it has a lamellar gel-to-liquid crystalline phase transition temperature Tm of 2.5  C. At 25  C the POPC bilayer is fluid, hence it is characterized by high lateral and rotational lipid diffusion rather similar to a liquid (Cevc, 1993). This lipid fluid state provides in biological membranes the optimal environment to host membrane proteins. Small unilamellar POPC vesicles with a mean hydrodynamic diameter of about 60–70 nm are prepared for the incorporation of the cholesteryl– oligonucleotides. The vesicles are prepared by the FAT-VET50 method. First, the lipids are dissolved in a chloroform:methanol (5:1) mixture and the organic solvent is completely removed by rotatory evaporation and high vacuum drying in a round bottom flask. To the formed thin, dry lipid film is added a 50 mM of TRIS and 100 mM of NaCl at pH 7.5 aqueous solution. Vigorous shaking with the help of a Vortex mixer leads to the dispersion of the lipid multilayers (MLV) in the aqueous solution, which results in the formation of a heterogenous population of vesicles. Upon repetitively freezing the MLV suspension in liquid nitrogen (at 195  C) and thawing at 45  C (far above the Tm of POPC), the vesicles’ aqueous interior and the external bulk aqueous phase equilibrate and possibly a fragmentation of MLV into smaller vesicles is achieved. Freezing and thawing (FAT) cycles are repeated six times. The MLV suspension is then passed under a stream of N2 at moderate pressure repetitively (10 times) through track-etch polycarbonate filters which contain almost cylindrical pores of a defined size. The whole extrusion process leads to a mechanical transformation of the large vesicles into smaller ones. The filtration is started with filters containing larger pores (mean diameter of 100 nm), followed by a filtration through smaller pores (50 nm). The corresponding vesicle preparation is abbreviated as VET50, where ‘‘50’’ indicates the mean pore diameter used for the final extrusion. By this procedure we achieved mainly unilamellar vesicles with a mean hydrodynamic diameter of 65–70 nm and a homogenous size distribution, assessed by dynamic light scattering (DLS)

Liposomes Decorated with Oligonucleotides

253

experiments. The total membrane area and the number of lipid molecule in the outer leaflet are determined from the hydrodynamic radius considering an area per POPC of 0.72 nm2 and a membrane thickness of 3.7 nm, for the bilayer in the liquid crystalline state, respectively. Unilamellarity and low polidispersity of the liposomes are required for accurate knowledge of the grafting density (see Section 6.3.2) to interprete results from DLS. Extrusion might lead to a decrease of the lipid content caused by adsorption on the extruder walls or onto the filter. It is thus essential to determine the lipid concentration in solution after vesicles are prepared. The importance of knowing the total surface will become clear in next paragraphs: if the radius of the vesicles is precisely known, the polydispersity low (<0.07 from the second cumulant of the DLS fitting function), the total external lipid surface is determined and the oligonucleotide/lipid ratio determines the grafting density. The POPC concentration in the vesicles is determined by the Steward– Marshall method (Stewart, 1980), using an appropriate calibration curve obtained with known amounts of POPC. This colorimetric method is based on the formation of a complex between phospholipid and ammonium ferrothiocyanate that is soluble in chloroform. An aliquot of 7 ml of liposomes is mixed with 2 ml 0.1 M ammonium thiocyanate (NH4SCN) in chloroform. After shaking with vortex for 15 s, the sample is centrifuged for 5 min at 1000 rpm. The red lower layer (chloroform) is removed with a Pasteur pipette and the absorbance is read in a Lambda 900 UV–vis spectrophotometer at 485 nm. The phospholipid concentration is determined by comparison to the calibration curve.

4. Incorporation of Oligonucleotides 4.1. Choice of the lipid anchor The chemical nature and the position of the lipid anchor play a fundamental role in the attachment of the lipophilic oligonucleotides to the liposomal membrane. The anchorage determines the incorporation yield and the stability of the oligonucleotides in the membrane. The distribution of the oligonucleotides on the liposomal surface and the conformation of the attached oligonucleotides are dependent on the anchorage mode also. The main problem related to the insertion of lipophilic oligonucleotides into existing membranes is associated with their hydrophobic nature (i.e., amphiphilic properties) and their supramolecular self-organization in aqueous solution. The lipid moiety in the lipophilic oligonucleotide should be chosen to enhance partition in the liposomal bilayer rather than promote the formation of stable micelle-like aggregates in solution by self-aggregation process. Cholesterol was shown to be a good anchoring group for the insertion of lipophilic DNA into membranes: cholesterol conjugates have been shown to have a lower

254

Martina Banchelli et al.

tendency to self-aggregation with respect to alkyl chain conjugates (Gosse et al., 2004) and cholesterol-conjugated oligonucleotides can be directly incorporated into lipid bilayers through direct spontaneous insertion of the cholesteryl moiety into the bilayer. In fact, cholesterol itself embeds within the hydrophobic interior of the bilayer, forming a mobile anchor and also stabilizing the lipid bilayer. Two competing aspects have to be taken into account: the affinity for the membrane and the self-aggregation of the lipophilic oligonucleotides in solution. Of course these two points are interconnected, because, while the cholesteryl–oligonucleotide conjugate inserts into the membrane, it is subtracted from other aggregation equilibria. We have determined the aggregation threshold for cholesteryl– oligonucleotide conjugates, CAC (critical aggregation concentration), that can be conveniently monitored by light scattering or fluorescent probe techniques (Kalyanasundaram and Thomas, 1977); surface tension is not the method of choice, because the surface tension decrease is scarce and the adsorption at the interface slow. Two different anchoring strategies are pursued: a single-cholesterol anchoring group and multiple-cholesterol functions (Banchelli et al., 2009a,b) (Scheme 13.1). In this latter case, the number of cholesterol tagged is in the range 3–4, and the lipophilic units are spaced by three nucleotides along an oligonucleotide chain (12 thymidine bases). This is achieved through a novel synthetic strategy for ssDNA carrying lipophilic modified nucleotides in different positions within the oligonucleotide chain. In this way, the hydrophobic anchors can be separated by a number of nucleotides thus reducing the stability of micellar self-organization. It has been shown that membrane binding of oligonucleotides with one single-cholesterol anchor may be weak when the hydrophobic moiety is directly attached to the ssDNA (Pfeiffer and Ho¨o¨k, 2004). For this reason, we tested a new cholesterol-conjugated ssDNA where an oligoethylene glycol is introduced into the chain to act as a spacer between the cholesterol and the oligonucleotide portion. The introduction of the oligoethylene linker would improve the hybridization of the membraneanchored ssDNA with a complementary strand and the stability of the duplex at the membrane surface. Flexible spacer chains are commonly utilized to enhance the hybridization of terminally anchored oligonucleotide probes of DNA microarrays, since they create distance between the probes and the impenetrable surface, thus approaching the hybridization conditions of free chains in solution.

4.2. Choice of the grafting density Cholesteryl–oligonucleotide conjugates are added to the pre-prepared liposomal suspension (POPC 1.3 mM) from the solution at different concentrations (in the range 0.3–57 mM).

255

Liposomes Decorated with Oligonucleotides

ss-DNA 6.3 nm

TEG chol 1.4 nm 1.6 nm

6.3 nm ss-DNA

1.8 nm HEG

4.8 nm chol-ss-DNA

Scheme 13.1 Molecular diagram of single-cholesterol (ON-Chol) and multiplecholesterol oligonucleotide (ON-multichol).

In order to perform studies on the grafted liposomes it is useful to introduce the parameter hN i, defined as the average number of cholesteryl-ssDNA per vesicle. The distribution of guest molecules among colloidal hosts has been the subject of several studies that point to a Poisson distribution (Zana, 1987). Since hN i is quite high, the probability of oligonucleotide occupancy P(N ) can be well approximated by a Gaussian distribution function centered at hN i. Assuming a negligible translocation of the cholesteryl–oligonucleotide during the experimental time window, we can consider that the oligonucleotide derivative is entirely distributed in the outer vesicular leaflet. Therefore, it is meaningful to consider the stoichiometry with respect to POPC in the outer leaflet. The knowledge of liposomal size and the narrow size distribution allow a fairly accurate estimate of hN i. These hN i values, reported in the second column of Table 13.1, can be converted into an ‘‘average distance’’ (G 1/2 ) between anchoring sites onto vesicular surface. It should be stressed here that anchoring is due to intermolecular interactions of hydrophobic nature between the cholesterol of the ON-Chol and the phospholipid bilayer. The estimation of hNi is obtained by calculating the mean aggregation number (NAGG) of the POPC liposomes (35 nm radius, from DLS measurements), then the molar concentration of the liposomes, and finally the average number of cholesteryl–oligonucleotides per liposome, described in details as follows:

256

Martina Banchelli et al.

Table 13.1 Composition of the POPC liposome/cholesteryl–oligonucleotide samples Cholesteryl– oligonucleotide (mM)

POPCext leaflet/oligos

hN i

˚) G 1/2 (A

0.3 2.0 4.1 8.3 16.6 17.5 57.0

2050:1 307:1 150:1 75:1 40:1 37:1 11:1

9 60 125 250 500 525 1700

410 158 110 78 55 53 30

NAGG ¼ Total liposome surface / POPC molecular surface 2 2 ¼ (4pRext þ 4pRint ) / 0.7 nm2 Rext ¼ 35 nm, Rint ¼ 31 nm (thickness of POPC bilayer  4 nm) NAGG  39,200 [Liposome] ¼ [POPC]/NAGG hNi ¼ [cholesteryl–oligonucleotides]/[liposome] The hNi values can be then converted into an ‘‘average distance’’ (G 1/2) between anchoring sites onto vesicular surface by dividing the vesicle surface in hNi parts, and calculating their sides. The grafting density parameters reported in Table 13.1 are relative to the systems that we have investigated. The G 1/2 values must then be compared with the size of the hydrophilic portion of the lipophilic oligonucleotide. The relative values of these two structural parameters determine the conformation of the oligonucleotides and therefore their binding efficiency and kinetics.

5. Characterization of the Soft Hybrid Nanostructure 5.1. Purification of the liposome–oligonucleotide systems (size exclusion chromatography) As already mentioned, self-aggregation of lipid–oligonucleotide conjugates might compete or slow down the insertion in membranes; therefore, for a given G 1/2, a quantitative determination of oligonucleotide in solution is necessary. A quantitative determination of the ON-Chol incorporation in the lipid bilayer is assessed by gel filtration on Sephadex G-50. The technique of gel exclusion chromatography can be used to separate macromolecules by their size through columns of beads of gels that have small pores, so that smaller molecules are more retained within the pores of

Liposomes Decorated with Oligonucleotides

Figure 13.1

257

Representation of a typical size gel exclusion experiment.

the support medium, and hence elute more slowly than larger molecules (Fig. 13.1). The medium chosen for the ON-Chol/vesicles gel filtration is dextran Sephadex, which is one of the most commonly used for nucleic acids purification. In this experiment, the smaller molecule is represented by the ON-Chol monomer, whereas the larger particles consist of the oligonucleotidedecorated vesicles. The purpose of the experiment is to separate the free oligonucleotides from those anchored to the vesicles, therefore determining the percentage of oligonucleotides inserted into the vesicle bilayer. The ON-Chol-decorated vesicles are separated from unincorporated ON-Chol by a mini-column centrifugation method (Fry et al., 1978), which requires only a small amount of sample (aliquots of 200 ml by using 1 ml mini-columns) and has the advantage that the vesicles can be recovered with practically no dilution. This method is satisfactory for solutes of molecular weight less than 7000 Da; therefore, it is appropriate for separating ON-Chol (Mw ¼ 6193) from the larger POPC vesicles ( 30,000 kDa). Possible ON-Chol small oligomers may be excluded as well from the dextran pores. The Sephadex powder (10 g) is hydrated with 120 ml of TBS and stored for 24 h at 4  C before use. As an example, we report the results for minicolumn gel filtration performed on different samples simultaneously: POPC vesicles (1.3 mM ) and ON-Chol (4.1 mM ), POPC vesicles as a blank, 4.1 mM ON-Chol solution as a control. Fractions of 0.2 ml are collected and subjected to DLS and UV analysis to quantify ON-Chol concentration relatively to vesicles. UV absorption is only due to the oligonucleotides; however, scattering from the vesicles can provide a nonnegligible contribution to the extinction cross section. DLS measurements confirm that the

258

Martina Banchelli et al.

vesicles are recovered in the first fraction eluted with virtually no dilution: in fact the same value of the intensity of scattered light I(sample)/I(toluene) (where I(toluene) is the scattering of toluene measured in the same experimental conditions as the sample) is obtained from the vesicle dispersion before elution and the first eluted fraction. The light scattering contribution of the vesicles to the UV absorption spectra is evaluated by polynomial fitting of the absorption curve between 600 and 350 nm, where no absorption from the DNA bases is expected, according to the relationship (Barrow and Lentz, 1980):   I ln ¼ Clg ð13:1Þ I0 where l is the wavelength, and the exponent g (4 for Rayleigh scatterers) and C are parameters that can be adjusted in the fitting procedures. The light scattering curve has been extrapolated to 220 nm and subtracted from the absorption spectrum. UV analysis of the fractions collected in the gel separation experiments leads to the results shown in Fig. 13.2, where the absorption spectra of the vesicles/ON-Chol system are corrected for the light scattering contribution due to the vesicles. The percentage absorbance of the ON-Chol in the eluted fractions with respect to the solution before the filtration is calculated and the results are reported in Fig. 13.3. The first fraction eluted from the mixed vesicle/ONChol system contains the major amount of ON-Chol (95.8%), almost absent in the subsequent fractions. The 4.2% loss can be due to some free monomers entrapped inside the gel pores that are partially recovered by repeating the elution. The control experiment performed with free ON-Chol in solution shows that 26.5% of the oligonucleotide is inside the first eluted fraction, consistent with the occurrence of some aggregates which do not obviously form in the presence of vesicles. The presence of vesicles therefore significantly alters the relative abundance of ON-Chol monomers/oligomers. However, the relatively low cutoff of the pore size of the gel is sufficient to exclude dimers so that the first fractions of the ON-Chol gel filtration might contain a certain amount of these dimeric structures. This is not expected instead for the vesicle/ON-Chol system, since the self-aggregation process is not favored when the ON-Chol can insert into a preformed lipid membrane via hydrophobic intermolecular interactions between the cholesterol of the ON-Chol and the phospholipid bilayer. The incorporation of the cholesteryl–oligonucleotide into POPC vesicle bilayer is demonstrated to be highly efficient, and for further calculations it will be considered as complete (i.e., 100%).

259

Liposomes Decorated with Oligonucleotides

A 0.8

Abs

0.6

0.4

0.2

0.0 220

240

260 280 Wavelength (nm)

300

320

220

240

260 280 Wavelength (nm)

300

320

B 0.8

Abs

0.6

0.4

0.2

0.0

Figure 13.2 UV absorption spectra of the different fractions from the gel separation experiment: system before the separation (black), first fraction (red), second fraction (green), third fraction (blue); (A) POPC vesicles (1.3 mM) and ON-Chol (4.1 mM) and (B) free ON-Chol as the control.

5.2. Dynamic light scattering measurements DLS experiments are performed on the decorated vesicles at different lipid/ cholesteryl–oligonucleotide ratios. To test hybridization on the vesicles, the complementary sequence is also added afterwards in a 1:1 ratio with respect to the cholesteryl-ssDNA. This technique was used previously as a powerful and sensitive probe of hybridization on DNA-functionalized colloidal particles, for instance gold nanoparticles. The experimental DLS autocorrelation functions, g(t), for the decorated vesicles are analyzed by the method of cumulants (Koppel, 1972)

260

Martina Banchelli et al.

100% 90% 80% 70% 60%

Fraction 3 Fraction 2 Fraction 1

50% 40% 30% 20% 10% 0% Liposome + ON-Chol

ON-Chol

Figure 13.3 Relative values of % UV absorbance at 260 nm for the filtered fractions with respect to the absorption values before filtration.

1 1 lnð gðtÞÞ ¼ K  Gt þ m2 t2  m3 t 3 þ L 2 3

ð13:2Þ

where K is an experimental constant and mI are cumulants, adjustable constants fitting the curve. From the decay rate, G, we obtained the related z-average diffusion coefficient D ¼ G/q2 which is related to the hydrodynamic radius RH by the Stokes–Einstein equation RH ¼ kBT/(6pD). The polydispersity index (PDI) of the micellar aggregates is estimated from the m2 =m21 ratio. DLS experiments on decorated vesicles have been performed with different cholesterol-modified oligonucleotides. When the oligonucleotide portion, together with the hydrophilic spacer, protrudes outward from the membrane, the insertion should cause an increase of hydrodynamic radius of the vesicles, due to the added hydrodynamic thickness. This can be monitored using DLS to follow the time evolution of intensity autocorrelation functions. The logarithmic scale allows appreciating the slope variation that follows ON-Chol addition to the external milieu. The contribution to the scattering intensity due to possible cholesteryl– oligonucleotide aggregates must be completely negligible. In fact, if no self-aggregation of the lipophilic DNA occurs, the cholesteryl–oligonucleotide distribution onto the liposomal surface is controlled by the adsorption only, and the insertion properties of the cholesteryl function into the POPC bilayer. Therefore, it is extremely important to know the aggregation state of the lipophilic oligonucleotide in solution at various concentration because self-assembly is competitive toward the

261

Liposomes Decorated with Oligonucleotides

incorporation into the lipidic membrane and aggregation regimes should be avoided. For this reason, DLS measurements on the binary system (ON-Chol in buffer at the concentrations used with the liposomes) are required as a control experiment. The intensity autocorrelation functions are recorded at different equilibration times after dilution of the vesicular suspension with the solution containing the cholesteryl–oligonucleotides. The slope variation in the experimental curves (i.e., the relaxation rate of the concentration fluctuations, assuming a monomodal decay) can be directly correlated with a size variation, given the monodispersity of the liposomal suspension. The observed decrease of the decay rate corresponds to an increase in hydrodynamic radius with time until equilibrium is reached within 6 h. The observed radius increase corresponds to an increase of the hydrodynamic thickness around the vesicle, due to the hydrophilic portion of the guest molecules. Therefore, we have investigated the equilibrium thickness increase as a function of the added cholesteryl–oligonucleotide. The contribution of the oligonucleotide does not depend on the initial size of the vesicles, whose radius of curvature (33–35 nm) is higher than the fully extended length of the ON-Chol (9 nm). The increase of hydrodynamic thickness, termed H0, is reported as a function of ON-Chol concentration in Table 13.2. From the analysis of the dependence of the hydrodynamic layer thickness on surface coverage, we found that a conformational transition of the oligonucleotidic chain takes place (from a random coil- to a brush-state) as the surface coverage is increased and the average distance between anchoring sites on the vesicular surface is decreased (Banchelli et al., 2008).

Table 13.2 Hydrodynamic thickness H0, on the liposomes as a function of ON-Chol concentration Cholesteryl–oligonucleotide (mM)

H0 (nm)

0.3 2.0 4.1 8.3 16.6 17.5 57.0

0.45 1.25 1.55 2.15 4.20 5.25 5.50

262

Martina Banchelli et al.

6. Applications of Oligo-Decorated Liposomes 6.1. Hybridization with complementary oligonucleotides in solution As already mentioned, the grafting density on vesicles has important effects on hybridization kinetics. The coupling rates of ON-decorated liposomes with complementary strands in solution can be studied by means of UV stopped-flow experiments. The purpose of the experiment is the measure the hybridization kinetics of ON-Chol anchored to the vesicles with the free complementary strand in solution. The vesicular system is studied in parallel with the system of two oligonucleotides in solution and the effect of oligonucleotide surface density on the hybridization rate is also analyzed. UV absorbance spectroscopy is used to investigate duplex formation in a series of vesicle/ds-oligonucleotide hybrids at different POPC/oligonucleotide ratio, by varying the concentration of POPC and keeping the concentration of ds-oligonucleotide constant at 2 mM. Hybridization between ON-Chol incorporated in vesicles and complementary strand free in solution is observed by monitoring the decrease in UV absorbance at 260 nm upon rapid mixing of equal molar concentrations of the two oligonucleotide solutions in TBS saline. These solutions are mixed within approximately 103 s into a 1 cm path-length cuvette in a stopped-flow apparatus and hybridization is monitored at fixed temperature, that is, 25  C. The decrease in UV absorbance, caused by the hypochromism that occurs during duplex formation, is monitored until it has become invariant over time. Kinetic data are collected continuously and the time course of the absorbance (l ¼ 260 nm) is recorded in a range between 0.5 and 20 s from time zero, that is, when the injected solution volume reaches the instrumental beam height in the cuvette and a reasonable absorbance value can be obtained; typically, 400 pairs of data are taken in each experiment and data sets from three experiments performed under identical conditions are averaged (Fig. 13.4). Duplex formation of oligonucleotides is generally accepted as a process involving two main events (Wetmur and Davidson, 1968), initial formation of a nucleation complex followed by the sealing of the duplex. It is also generally accepted that the intermediate, initial association complex is unstable, with a marked tendency to either dissociate or seal, depending on the temperature. For short DNAs of up to several hundred base pairs, nucleation is rate-limiting at low concentrations and each duplex zips to completion almost instantly (>1000 bp s 1). The nucleation process is dependent on the concentration as well as on the complexity of the single strands at low DNA concentration nucleation is faster and for short heterogeneous oligonucleotides nucleation sites are fully extended by rapid zipping-up. For longer strands, a complex secondary structure of the

263

Liposomes Decorated with Oligonucleotides

70 ⫻ 10−3 60

A0-Ai

50 40 30 20 10 0 5

10 Time (s)

15

20

Figure 13.4 Kinetics of oligonucleotide hybridization onto POPC vesicles (35 nm hydrodynamic radius) at three different POPC/ON-Chol molar ratio: 82 (triangles), 164 (squares), and 657 (circles). Fitting curves are calculated through Eq. (5).

oligonucleotide, possibly containing intrastrand hairpin loops generated by various intramolecular base pairs, may profoundly decrease the hybridization kinetics with the complementary sequence. In fact, unimolecular intrastrand processes can be 100 times faster than the corresponding bimolecular pairing processes. Therefore, the existence of such hairpins within the oligonucleotide strand retards the rate of association, so that propagation of the duplex becomes the rate-limiting process (Gao et al., 2006). The rate constants of oligonucleotide hybridization are obtained by second-order fits to the absorbance data, for both solutions and vesicular suspensions. At time t, the concentration of ssDNA, Ct, is calculated using the following equation: Ct ¼

At  A1 C0 A0  A1

ð13:3Þ

where At is the absorbance at time t, A0 is the absorbance of the ssDNA at t ¼ 0, A1 is the absorbance of the dsDNA at equilibrium, and C0 is the initial concentration of ssDNA. C0 in our experiments is the bulk concentration, both in the case of the oligonucleotides in solution and on the vesicles. Hybridization of equal molar oligonucleotide single strands can be described by second-order reaction kinetics: 1 1  ¼ kon t Ct C0

ð13:4Þ

where kon is the association rate constant. Combining Eqs. (13.3) and (13.4), we find:

264

Martina Banchelli et al.

At ¼ A0 þ ðA1  A0 Þ

C0 kon t 1 þ C0 kon t

ð13:5Þ

To obtain quantitative kinetic information about the hybridization onto vesicles as a function of oligonucleotide grafting density, the rate constant kon is obtained by fitting the raw experimental data to Eq. (13.5). Results from stopped-flow experiments for six samples measured at different POPC/ON molar ratios and for the two oligonucleotides in solution are listed in Table 13.3. In all cases, the expression describing a second-order reaction with equal concentrations of the reactants resulted in the best fit. A clear trend in the rate of formation of the nucleation complex emerges: as [POPC]/[ON] ratio is increased and as oligonucleotide grafting density is decreased, the association rate increases. Therefore, hybridization kinetics is modulated by grafting site density. This behavior can be explained rather simply on the basis of an increased electrostatic repulsion at high surface density of oligonucleotides that may hinder the association with the complementary strand. However, another important aspect could affect the hybridization kinetics at high surface coverage as well, that is, the conformation of the single-stranded oligonucleotides at the surface. As previously shown in detail in Paragraph 3, the oligonucleotides at the vesicular surface undergo a conformational transition from mushroom to brush state when the average distance between grafting sites, G 1/2, becomes comparable to the Flory radius, RF, of the oligonucleotidic portion, which, for TEG-ss-18-mer in our experiments, varies from 40 to ˚ . At higher G 1/2 values, the oligonucleotide is in a random coil 55 A Table 13.3 Association hybridization rate constants for ON-Chol/complementary ON (1:1) POPC vesicles at different POPC/ON-Chol ratio in TBS solution (Tris 50 mM and NaCl 100 mM) Conc. POPC (mg/ml)

[POPC]/ [ON]

Number of ON per vesicle

Average distance between grafting ˚) sites (A

– 0.125 0.25 0.5 1 2 4

– 82.2 164.4 328.9 657.8 1315.6 2631.2

– 440 220 110 55 27 15

– 28 40 57 80 114 160

kon (105 M 1 s 1)

2.94  0.05 1.13  0.02 2.04  0.05 2.68  0.05 3.02  0.1 3.23  0.1 5.10  0.8

The kon values are from the stopped-flow experiments using equal concentrations, 2 mM, of the oligonucleotides.

Liposomes Decorated with Oligonucleotides

265

conformation and chains at the surface do not overlap, whereas, at G 1/2 ˚ , the oligonucleotides start to overlap and form a values lower than 55 A brush the vesicular surface. The dense packing of the oligonucleotides inside the brush may affect the hybridization rate with complementary strand more than the mushroom state, where oligonucleotide chains develop as separate coils and duplex formation might occur faster. If compared to the oligonucleotide kinetic rate in solution, we can see that beyond a [POPC]/[ON] ratio of 330, that is consistent of about 110 duplexes per vesicle, the constant rate of the oligonucleotide hybridization at the vesicular surface is lower, while above this threshold concentration the kinetic rate of hybridization on the vesicle becomes higher than the value obtained in solution. It has been shown that the associative kinetics of surface DNA hybridization on planar gold surface are suppressed by a factor of 20- to 40-fold compared to solution-phase hybridization and a 5- to 10-fold suppression in hybridization rates of 22 mers is also observed on microparticles by means of FRET measurements (Henry et al., 1999). In the present case, the oligonucleotide can hybridize with the complementary strand faster on the vesicle rather than in solution, as long as grafting density is not too high. The advantage of having nanoparticles rather than solid planar supports in the hybridization kinetics relies on the diffusion of those particles that could help the association reaction. Also, noncovalent anchorage of oligonucleotides to vesicles, rather than covalent linkage to solid supports or gold nanoparticles, allows more flexibility and mobility of the grafted molecules onto the surface. Oligonucleotides can continuously rearrange and change their distribution on the vesicle, through the spontaneous motion of the cholesterol along the lipid bilayer, probably improving the association kinetics with a complementary strand in solution as well. In addition, the conformation of the anchored oligonucleotide is, as previously discussed, an important aspect in the hybridization kinetics: the threshold [POPC]/[ON] ratio at which hybridization on the vesicles becomes faster than in solution corresponds to a calculated average grafting site distance of ˚ , which is comparable to the Flory radius of the ON-TEG portion. 50 A This might be interpreted as if the main factor controlling the kinetic properties is the conformation at the surface. This is a very promising result for the efficient fabrication of nanosystems for DNA recognition and self-assembly.

6.2. Hybridization with self-assembled DNA nanostructures The realization of addressable DNA architectures requires the design of complex motifs with sticky ends, which assemble forming extended DNA scaffolds. These subunits can be spatially organized onto functionalized lipid surfaces to possibly achieve addressable platforms with sub-nm precision. The successful realization of the DNA nanohybrid structure resides in the accomplishment of an optimized preparation protocol, where grafting

266

Martina Banchelli et al.

density, lipid/DNA ratio, liposome number density, and preparation procedure on the final structure are taken into account, as detailed in the previous paragraphs. Two main approaches can be exploited to anchor DNA nanostructures onto oligonucleotide-decorated liposome surfaces: the first method consists of the direct immobilization of the preformed DNA nanostructures (previously obtained in solution) by hybridization of complementary sticky ends set on the nano-object and on the lipid surface (single-step strategy), while the other procedure involves the stepwise addition of each strand in a predefined order to form the desired nanostructure directly on the lipid surface (stepwise strategy) (Baldelli Bombelli et al., 2009). Here, we describe these approaches for the anchoring of two different DNA nano-objects: a closed pseudohexagon and an open linear nanostructure. The choice of two different structures allows the investigator to better address all the critical parameters to be considered in the formation of such hybrid systems. Both nanostructures can be formed in solution by hybridization of six defined linear oligonucleotides in which five are identical for the two structures and the sixth is partially modified to obtain close (B0 A0 ) and open structures (B0 X), respectively (Scheme 13.2). The strand labeled as 2 in Scheme 13.2 is the sticky end complementary to the ON-Chol incorporated into the lipid membrane (see Paragraph 3).

6.3. Different preparation procedures 6.3.1. Single-step strategy The first step is the formation of the DNA nanostructures in the same buffer solution used for the liposomes preparation. The hybridization process conditions have to be optimized as a function of the length, the geometry, and the sequence of the Oligo-Decorated Nucleotide (ODN) strands used as building blocks. Generally, buffer solution with ionic strength in the range of 0.1– 0.5 M, to assure stability at room temperature (high melting temperature), and DNA concentration of 0.3–5 mM, to minimize the formation of undesired polymerization products, are used. The concentration of each ODN has to be precisely set using nearest neighboring approximation (NNA) values for the extinction coefficients and their absorbance at 260 nm. The nanostructures are formed in 50 mM Tris buffer (pH ¼ 7.5) with 100 mM NaCl (Tm  56  C) by mixing equimolar amounts of the DNA sequences to have a 3 mM final concentration of each strand in solution, which means a 3 mM concentration of the nanoconstructs. The samples are annealed by heating to 90  C and then cooled to 5  C with a constant temperature gradient over 6 h. The protocol is optimized controlling the samples by gel electrophoresis, DLS, and atomic force microscopy (AFM) experiments (Baldelli Bombelli et al., 2008; Banchelli et al., 2008; Tumpane et al., 2007).

Name

Sequence

FA-2

5’-ACGAGCCTTTGACGCTTGGA-TT-TAGTGCGTAACATAGGCTAC-TTCTGAAATTATGATAAAGA-3’

E’F’

5’-ATTTACCTGGAAGCAGCCAC-TT-TCCAAGCGTCAAAGGCTCGT-3’

ED

5’-GTGGCTGCTTCCAGGTAAAT-TT-CACTATGTAACTGGTCTCTTA-3’

D’C’

5’-TAGAGACCAGTTACATAGTG-TT-TGACCTCAGTCGCAAGGCTG-3’

CB

5’-CAGCCTTGCGACTGAGGTCA-TT-TCGGGTCAACGAATGGCTGC-3’

B’A’

5’-GCAGCCATTCGTTGACCCGA-TT-GTAGCCTATGTTACGCACTA-3’

B’X

5’-GCAGCCATTCGTTGACCCGA-TT-CCCCCCCCCTTTTTTTTTTT-3’

Pseudo-hexagonal structure D

E D⬘

C

E⬘

C⬘

F⬘ B⬘

B

F

A⬘ 5⬘

A

2 3⬘ Open structure

2

A

F⬘

E⬘

D⬘

C⬘

B⬘

F

E

D

C

B

Scheme 13.2 DNA nanostructure composing strands. Adapted from Baldelli Bombelli et al. (2008).

X

268

Martina Banchelli et al.

The binding experiment is designed on the basis of spatial considerations calculating the steric hindrance of the two nanostructures through geometrical estimates, and thus determining the maximum number of occupancy, Nmax, on the lipid surface, which corresponds to hNi mentioned earlier. The calculated Nmax are very approximate values and more likely overestimated since both orientation and conformation of the nanostructures and electrostatic repulsion contributions have not been considered in this calculation. Nonetheless, Nmax can be a first indicative parameter for the choice of the right grafting density range to use for preparing DNA/lipid hybrid complexes. As previously described, monodisperse liposome solutions of about 35 nm radius are used for which the resulting Nmax are about 90 and 1000 for the close and open DNA nanostructures, respectively. On this basis, preformed hexagons and open nanostructures are added to ON-Choldecorated vesicles in a stoichiometric fashion with respect to the anchoring sites, varying the occupancy number from 10 to 130. Therefore, the investigated surface density is safely chosen within the vesicle hosting capacity and should allow the buildup of isolated nanoconstructs. The samples, prepared either by adding the nanostructures to the decorated liposomes equilibrated overnight at room temperature, or mixing together all the components, are measured by DLS 1 and 24 h after the preparation (Fig. 13.5A). The two preparation methods provide different results in the shortest time. Indeed, after 1 h, samples obtained by simultaneously mixing give larger hydrodynamic radii and higher PDI than the corresponding ones prepared in two steps. After a few days, depending on the grafting densities, different samples reach the same values of RH and PDI (data not shown), indicating that the incorporation of the ON-Chol into the lipophilic environment is the slowest step of the process. Hence, all reported experiments are performed on samples prepared with the two steps method. Our first observation is that the attainment of an ‘‘equilibrium’’ state is time-variable, depending on the nanostructure grafting density, as shown in Fig. 13.5A. Moreover, the hydrodynamic radius of both hybrid nanostructures increases with respect to the naked liposomes and keeps growing with increasing the occupancy number (see Fig. 13.5B), while the polidispersity is almost invariant. This increase is related to the different orientation of the nanostructure as a function of the grafting density in term of steric and electrostatic repulsions between neighboring objects; these become stronger at higher packing density determining a larger hydrodynamic shell. Increasing hNi, while pseudohexagon-decorated vesicles reach a sort of saturation for N > 80, those with open nanostructures keep growing, indicating that the more flexible and less bulky open nanostructures can be more densely packed on the lipid surface. In the inset of Fig. 13.5B, we also report the hydrodynamic radius of lipid/DNA hybrids for N ¼ 30 at different

269

Liposomes Decorated with Oligonucleotides

A

44

RH (nm)

42

40

1 day

38

36

3

4

5

6 7 8 9

2

3

4

5

6 7 8 9

10

100 Time (min)

B

60

Pseudo-hexagon Open

RH (nm)

55

50 48

45

46 44

40

42 0.4

0.8

1.2

1.6

cPOPC [M]

35 0

20

40

60

80 N

100

120

140

Figure 13.5 (A) Time dependence of the hydrodynamic radius of the lipid/DNA hexagon hybrid for two grafting densities: (●) N ¼ 9, [POPC] ¼ 0.785 mM, [ODN] ¼ 0.18 mM; (□) N ¼ 30, [POPC] ¼ 0.235 mM, [ODN] ¼ 0.18 mM. The empty circle and the filled square represent the final equilibrium values reached after 1 day. (B) Hydrodynamic radii of lipid/DNA hybrids as a function of the occupancy number, [POPC] ¼ 1.33 mM. The inset shows the behavior of the hydrodynamic radius of the hybrid as a function of liposome density for N ¼ 30. (□) Pseudohexagons, (▲) open nanostructures.

liposome densities, the results show a slight increase in the absolute value, while PDI is of the same order of magnitude. This means that liposome density is not critical for the successful formation of the complexes.

270

Martina Banchelli et al.

The samples are monitored with time, and they are stable for weeks at room temperature, demonstrating the effectiveness of this method. It is important to underline that the achievement of a thermodynamically stable system requires variable equilibration times depending on the surface grafting density. 6.3.2. Stepwise strategy Ring and open nanostructures are also built on the liposome surface by a stepwise strategy, which consists of the sequential addition of equimolar amounts of each strand to ON-Chol-decorated liposomes, in stoichiometric fashion with respect to the anchoring sites, and then recruitment by the partially built hybrid. Two stepwise procedures differing by the mixing order of the oligonucleotides, illustrated in Fig. 13.7, are adopted for the construction of the more complex close nano-object to distinguish a possible kinetic control of each step due to the different shapes of the partially built nanoconstructs. Each strategy consists of eight steps where step 1 corresponds to POPC liposomes and step 2 to the insertion of ON-Chol into the lipid membrane. The spontaneous insertion of ON-Chol and the following coupling events between complementary oligonucleotides are revealed as an increase of the radius of the liposome, due to the added hydrodynamic thickness, by DLS (Fig. 13.7). The study is performed on eight different samples for each strategy representing the composing different steps (in the asymmetric strategy for close and open nanostructures the first seven steps coincide) prepared under the same conditions: the same ODN strand is added to different step samples at the same time, and the interval between two sequential additions is set at 1 h, and the final volume is adjusted with PBS buffer. ON-Chol is added previously to the liposomes, and eventually the samples are equilibrated overnight before the sequential addition of the other strands. First, samples are prepared at different grafting densities keeping constant the lipid concentration (1.33 mM) as a function of the added oligonucleotide concentration. For the open construct, the stepwise strategy gives comparable results, in terms of the hydrodynamic radius, to those obtained with the single-step procedure in all the grafting density ranges investigated. For the construction of closed nanostructures at this lipid concentration, instead, we need to distinguish between low (N < 15) and high grafting densities: pure isolated hybrid structures have only been obtained in the former condition. In fact, for low grafting densities, both symmetric and asymmetric strategies have given comparable RH values in the final step, in agreement with those obtained with the single-step, procedure, as well as a RH trend in the intermediate construction steps consistent with the expected partially built nanoconstructs (see Fig. 13.6A). However, the symmetric strategy ends up being more effective than the asymmetric one

271

Liposomes Decorated with Oligonucleotides

A = 9 (step-by-step) = 9 (single step) = 30 (step-by-step) = 30 (single step)

44

RH (nm)

42

40

38

36

34 1

2

3

4

5

6

7

8

Step B 44

42 RH (nm)

1 day

40

38 3

4

5 6 7 8 9

2

10

3

4

5

6 7 8 9

100 Time (min)

Figure 13.6 (A) Effect of the liposome number density on the formation of lipid/DNA nanohybrids for two different POPC/DNA ratios obtained with the symmetric stepwise procedure. The radii of the nanoconstructs obtained by addition of preformed hexagons are reposted with filled symbols for comparison. Readapted from Baldelli Bombelli et al. (2009). (B) Time dependence of the hydrodynamic radius of the lipid/ DNA hexagon hybrid for two grafting densities: (●) N ¼ 9, [POPC] ¼ 0.785 mM, [ODN] ¼ 0.18 mM; (□) N ¼ 30, [POPC] ¼ 0.235 mM, [ODN] ¼ 0.18 mM.

in terms of polydispersity and grafting density range. For occupancy numbers >15, although the intermediate steps are characterized by hydrodynamic radii in agreement with the expected assemblies, the addition of the

272

Martina Banchelli et al.

38

37

RH (nm)

36

35

Asymmetric FA−2 F⬘E⬘ DE D⬘C⬘ BC B⬘A⬘ *B⬘X(open)

34

Symmetric FA−2 B⬘A⬘ BC F⬘E⬘ D⬘C⬘ DE

33

32 2

4

6

8

Step

Figure 13.7 Trend of the hydrodynamic radii for the step-by-step construction of DNA pseudohexagons and open nanostructures on liposomes for N  10. Symmetric (filled circles) and asymmetric strategies (filled squares) used for close nanostructures are showed. The open nanostructures are formed by asymmetric strategy and the first seven steps coincide with those of the asymmetric strategy for the close nanostructure (empty square). The tables on the left indicate the addition order of the oligonucleotides in the two strategies (see Scheme 13.1). Adapted from Baldelli Bombelli et al. (2009).

last oligonucleotide strand induces the formation of large aggregates with a consequent significant increase in PDI. Size distribution analysis of the autocorrelation functions for these samples does not give well-defined populations, but we rather obtain a single broad population probably composed of isolated hybrid structures, together with dimers and trimers. The fact that we do not observe the same aggregation process for the addition of the last strand in the formation of the open nanostructure, suggests that the closure of the ring is the driving mechanism for the formation of aggregates. On the basis of this hypothesis, a determining factor in the aggregation process could be a too short average distance between liposomes in solution, causing the association of neighboring hybrid structures during the ring closure. This parameter can be optimized to make the symmetric strategy effective for higher occupancy numbers also. Hence, a different series of samples is prepared according to the symmetric strategy keeping constant

Liposomes Decorated with Oligonucleotides

273

DNA concentration (0.18 mM) and varying lipid concentration between 0.235 and 0.785 mM that corresponds to occupancy numbers of 30 and 9, respectively. Aggregation is not detected upon addition of the last ODN strand for these samples and the resulting hybrid nanostructures hydrodynamic sizes and PDI are comparable to those obtained for corresponding samples prepared with the single-step procedure (Fig. 13.6A). Moreover, the hybridization process is monitored with time, as reported in Fig. 13.6B. We do not observe a significant difference, as for the singlestep procedure, in the achievement of the ‘‘equilibrium’’ state between two grafting densities and after 10 min the hydrodynamic radius is invariant with time for both samples. It is important to stress that the formation of DNA nanoassemblies in solution requires a lengthy annealing process (heating at 90  C and then cooling down to 20  C with a constant temperature gradient within 6 h), while the step-by-step procedure on liposomes is performed at room temperature. The success of the stepwise strategy highlights the important role of the immobilization of ODN on the lipid surface, working as a sort of catalyzer for the hybridization reaction. We conclude that the stepwise strategy is a sucessful method for ‘‘in situ’’ construction of more complex DNA nanostructures on the liposome surface at room temperature, but, to avoid aggregation, special care is needed regarding the liposome density in solution.

6.4. Kinetics aspects DLS is shown to be an extremely powerful tool for studying the formation of soft hybrid nanostructures composed of phospholipid liposomes decorated with oligonucleotides nanostructures. DLS analysis allows the comparison of different preparation methods to infer a final optimized protocol for the preparation of monodisperse, isolated DNA/lipid nanohybrid structures. Several parameters have to be considered in the different preparation procedures and the optimal conditions may not be the same for diverse strategies. In fact, one of the main differences between single-step and stepwise procedures seems to be the kinetics of the attainment of a stable complex. Moreover, in the stepwise strategy the determining parameter is liposome crowding, that can promote aggregation during the closure step of the ring-like structure, while in the single-step protocol a longer time scale is needed to reach the equilibrium state of the system at higher grafting density. To study the kinetics of hybridization of the determining step of the formation of these hybrids, the time course of the 260 nm absorbance after mixing the solutions containing the coupling sides is followed via stoppedflow (see Section 4.1). The time-course of the absorbance is recorded between 0.04 and 17 min (longest time measurable with this equipment).

274

Martina Banchelli et al.

A 0.400

A

0.398

0.396

0.394

0 B

5

10 t (min)

15

0.400 0.398 0.396

A

0.394 0.392 0.390 0.388 0.386 0.384 0

200

400 600 t (min)

800

1000

Figure 13.8 (A) Kinetics of ON-Chol/pseudohexagon-2 hybridization onto POPC vesicles (single-step process) at two different occupancy numbers: 9 (empty circles) and 30 (filled squares). The absorbance has been normalized to 0.4 for a better comparison. (B) Kinetics of hybridization process that occurs at the step 8 of the symmetric stepwise procedure presented in Fig. 13.5 (DE addition).

In particular, for the stepwise procedure we monitor the absorbance decrease upon addition of the DE strand to the step 7 solution obtained according to symmetric strategy (Fig. 13.8A), while for the single-step procedure a preformed DNA pseudohexagon solution is added to oligoloaded vesicles (Fig. 13.8B). The final ODN concentration is kept constant in all experiments to 0.18 mM, while the lipid concentration is set to 0.235 and 0.785 mM which corresponds to N ¼ 30 and 9, respectively. The data are reported in Fig. 13.8A and B and, since the total absorption is strongly affected by vesicle scattering, the curves have been normalized to

Liposomes Decorated with Oligonucleotides

275

achieve a better comparison between different samples. Nonetheless, a quantitative estimation of the hybridization cannot be precisely determined. In the single-step procedure, hybridization kinetics curves cannot be analyzed assuming a single process with an uniform association rate constant, but are rather composed of an initial faster dynamics followed by a slower process. Presumably, the first process is driven by the diffusion of the nanostructures approaching the lipid surface, as suggested by the superimposition of the kinetics curves in Fig. 13.8A for the first minutes of the process. The higher steric and repulsive barrier of the more crowded interface comes into play later, and determines the association rate of the second process. Of course, there are several processes involved in the formation of the DNA/lipid hybrid, and a more accurate analysis is needed to be able to model it. Unfortunately, the hybridization process is not completed by 17 min, which is the longest experimental time achievable with this experimental setup. Nevertheless, 17 min seems to be a good time scale to study the kinetics of the ring closure mechanism in the symmetric stepwise preparation upon addition of the DE strand. In this case, the hybridization is mainly composed of a single association process characterized by a faster rate constant and completed by the investigated time scale. These data are in good agreement with that observed by DLS, confirming that different kinetics of binding are involved in the determining step of the two preparation procedures. This enhances the necessity to choose the right conditions depending on the chosen methodology to succeed in the preparation of DNA/lipid hybrid structures.

7. Challenges and Perspectives DNA coupling to hard nanoparticles (e.g., AuNp) to direct their hierarchical self-assembly is a relatively mature research field (Alivisatos et al., 1996; Mirkin et al., 1996; Nykypanchuk et al., 2008). Conversely, the literature on DNA coupling to soft nanoparticles (i.e., liposomes) is still at an early stage, even if interesting and innovative applications of this procedure are now being reported in the literature (Beales and Kyle Vanderlick, 2007; Jakobsen et al., 2008; Pfeiffer and Ho¨o¨k, 2004; Stengel et al., 2007). Coupling DNA to soft nanoparticles requires a noncovalent approach, that is, a lipid modification of the oligonucleotide, and its insertion into the membrane is ruled by thermodynamics, and can therefore be altered in response to experimental conditions (temperature, concentration, salinity, and so on). Hence, the preparation and characterization of DNAdecorated liposomes is probably more challenging, but should nevertheless be pursued in view of the many envisaged uses.

276

Martina Banchelli et al.

Several key factors should be taken into account: the choice of the lipidanchoring unit, the presence and nature of a spacer toward the membrane proximal end, and, of course, the length and the composition of the oligonucleotide portion, which is more directly dictated by the purpose of the application. It is difficult to rationalize these different contributions, in view of the still limited reports in the literature. The role of the lipid composition of the membrane has not investigated in detail, but we can easily predict that the affinity of the lipid-ON conjugate will be highly dependent on this parameter. It should be mentioned, however, that whatever the long-term envisaged application, the reliable knowledge of structural parameters, such as the size of the hybrid aggregates, oligonucleotide conformation, and grafting density, are vital for the use of these objects as functional building blocks for nanostructure arrays.

ACKNOWLEDGMENTS Financial support from CSGI, MIUR-PRIN, CNR-FUSINT, and the European Commission’s Sixth Framework Program (Project Reference AMNA, Contract No. 013575) are acknowledged. Dr. Alessio Innocenti is acknowledged for help in Stopped Flow Experiment. Dr. Gabriella Caminati is acknowledged for fruitful discussions.

REFERENCES Alivisatos, A. P., Johnsson, K. P., Peng, X. G., Wilson, T. E., Loweth, C. J., Bruchez, M. P., and Schultz, P. G. (1996). Organization of ‘nanocrystal molecules’ using DNA. Nature 382, 609–611. Baldelli Bombelli, F., Gambinossi, F., Lagi, M., Berti, D., Caminati, G., Brown, T., Sciortino, F., Norde´n, B., and Baglioni, P. (2008). DNA closed nanostructures: A structural and Monte Carlo simulation study. J. Phys. Chem. B 112, 15283–15294. Baldelli Bombelli, F., Betti, F., Gambinossi, F., Caminati, G., Brown, T., Baglioni, P., and Berti, D. (2009). Closed nanostructures assembled by step-by-step ss-DNA coupling assisted by phospholipid membranes. Soft Matter 5, 1639–1645. Banchelli, M., Betti, F., Berti, D., Caminati, G., Baldelli Bombelli, F., Brown, T., Wilhelmsson, L. M., Norde´n, B., and Baglioni, P. (2008). Phospholipid membranes decorated by cholesterol-based oligonucleotides as soft hybrid nanostructures P. J. Phys. Chem. B 112, 10942–10952. Banchelli, M., Gambinossi, F., Berti, D., Caminati, G., Brown, T., and Baglioni, P. (2009a). Anchoring amphiphilic DNA to phospholipid membranes. Part II: Modulation of grafting density and conformation in vesicles. J. Phys. Chem. B submitted. Banchelli, M., Gambinossi, F., Berti, D., Caminati, G., Brown, T., and Baglioni, P. (2009b). Anchoring amphiphilic DNA to phospholipid membranes. Part I: Modulation of grafting density and orientation in supported lipid bilayers. J. Phys. Chem. B manuscript in preparation. Barrow, D. A., and Lentz, B. R. (1980). Large vesicle contamination in small, unilamellar vesicles. Biochim. Biophys. Acta 597, 92–99.

Liposomes Decorated with Oligonucleotides

277

Beales, P. A., and Kyle Vanderlick, T. J. (2007). Specific binding of different vesicle populations by the hybridization of membrane-anchored DNA. J. Phys. Chem. A 111, 12372–12380. Cevc, G. (1993). Phospholipids Handbook. Marcel Dekker Inc., New York. Chan, Y.-H. M., Van Lengerich, B., and Boxer, S. G. (2008). Lipid-anchored DNA mediates vesicle fusion as observed by lipid and content mixing. Biointerphases 3, FA17–FA21. Chan, Y.-H. M., Van Lengerich, B., and Boxer, S. G. (2009). Effects of linker sequences on vesicle fusion mediated by lipid-anchored DNA oligonucleotides. Proc. Natl. Acad. Sci. 106, 979–984. Fry, D. W., White, C., and Goldman, D. J. (1978). Rapid separation of low molecular weight solutes from liposomes without dilution. Anal. Biochem. 90, 809–815. Gao, Y., Wolf, L. K., and Georgiadis, R. M. (2006). Secondary structure effects on DNA hybridization kinetics: A solution versus surface comparison. Nucleic Acids Res. 34, 3370–3377. Gosse, C., Boutorine, A., Aujard, I., Chami, M., Kononov, A., Cogne-Laage, E., Allemand, J.-F., Li, J., and Jullien, L. (2004). Micelles of lipid-oligonucleotide conjugates: Implication for membrane anchoring and base pairing. J. Phys. Chem. B 108, 6485–6497. Henry, M. R., Stevens, P. W., Sun, J., and Kelso, D. M. (1999). Real-time measurements of DNA hybridization on microparticles with fluorescence resonance energy transfer. Anal. Biochem. 276, 204–214. Jakobsen, U., Simonsen, A. C., and Vogel, S. (2008). DNA-controlled assembly of soft nanoparticles. J. Am. Chem. Soc. 130, 10462–10463. Kalyanasundaram, K., and Thomas, J. K. (1977). Environmental effects on vibronic band intensities in pyrene monomer fluorescence and their application in studies of micelar systems. J. Am. Chem. Soc. 99, 2039–2044. Koppel, D. E. (1972). Analysis of macromolecular polydispersity in intensity correlation spectroscopy: The method of cumulants. J. Chem. Phys. 57, 4814–4820. Mirkin, C. A., Letsinger, R. L., Mucic, R. C., and Storhoff, J. J. (1996). A DNA-based method for rationally assembling nanoparticles into macroscopic materials. Nature 382, 607–609. Nykypanchuk, D., Maye, M. M., Van der Lelie, D., and Gang, O. (2008). DNA-guided crystallization of colloidal nanoparticles. Nature 451, 549–552. Pfeiffer, I., and Ho¨o¨k, F. (2004). Bivalent cholesterol-based coupling of oligonucleotides to lipid membrane assemblies. J. Am. Chem. Soc. 126, 10224–10225. Stengel, G., Zahn, R., and Hook, F. (2007). DNA-induced programmable fusion of phospholipid vesicles. J. Am. Chem. Soc. 129, 9584–9585. Stewart, J. C. M. (1980). Colorimetric determination of phospholipids with ammonium ferrothiocyanate. Anal. Biochem. 104, 10–14. Tumpane, J., Sandin, P., Kumar, R., Powers, V. E. C., Lundberg, E. P., Gale, N., Baglioni, P., Lehn, J.-M., Albinsson, B., Lincoln, P., Wilhelmsson, L. M., Brown, T., et al. (2007). Addressable high-information-density DNA nanostructures. Chem. Phys. Lett. 440, 125–129. Wetmur, J. G., and Davidson, N. (1968). Kinetics of renaturation of DNA. J. Mol. Biol. 31, 349–370. Zana, R. (1987). Surfactant Solutions: New Methods of Investigation. Marcel Dekker, New York.

This page intentionally left blank

C H A P T E R

F O U R T E E N

Synthesis, Characterization, and Optical Response of Gold Nanoshells Used to Trigger Release from Liposomes Guohui Wu,* Alexander Mikhailovsky,† Htet A. Khant,* and Joseph A. Zasadzinski* Contents 280 283

1. Introduction 2. Synthesis of HGNs 3. Optimization of HGN Dimensions for Maximum Absorption in the NIR 4. HGN Response to Femtosecond NIR Laser Pulses 5. Coupling HGN to Liposomes 5.1. Pulsed laser optics 5.2. Continuous-wave laser irradiation 6. Liposome Disruption and CF Release Due to Pulsed Laser Irradiation 7. Mechanism of Triggered Liposome Release 8. Effect of Proximity of HGNs to Liposomes 9. Conclusions Acknowledgments References

286 290 294 298 299 299 300 303 304 304 304

Abstract Liposomes show great promise as intravenous drug delivery vehicles, but it is often difficult to combine stability in the circulation with rapid, targeted release at the site of interest. Targeting to specific tissues requires developing highly specific ligands with strong affinities to receptors overexpressed on diseased cells; a new cellular target requires developing new ligands and identifying new receptors. Novel photoactivated, hollow, gold nanoshell (HGN)/liposome composites provide a new approach to both controlled * {

Department of Chemical Engineering, University of California, Santa Barbara, California, USA Department of Chemistry, University of California, Santa Barbara, California, USA

Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64014-3

#

2009 Elsevier Inc. All rights reserved.

279

280

Guohui Wu et al.

release and specific targeting. HGN are extremely efficient near infrared (NIR) light absorbers, and are not susceptible to photobleaching like conventional dyes. Near-complete liposome contents release can be initiated within seconds by irradiating HGNs with an NIR pulsed laser. Targeting the drug is limited only by the dimensions of the laser beam; no specific ligands or antibodies are required, so different tissues and cells can be targeted with the same HGN/liposomes. HGNs can be encapsulated within liposomes or tethered to the outer surface of liposomes for the most efficient drug release. HGNs in liposome solutions can also trigger release, but with lower efficiency. Drug release is induced by adsorbing femto- to nanosecond NIR light pulses that cause the HGNs to rapidly increase in temperature. The resulting large temperature gradients lead to the formation of vapor microbubbles in aqueous solutions, similar to the cavitation bubbles induced by sonication. The collapse of the unstable vapor bubbles causes liposome-membrane rupture and contents release, with minimal damage to the surroundings, and little overall heating of the solution.

1. Introduction The therapeutic efficacy of many drugs can be improved by maximizing their concentration at the disease site; toxicity can be reduced simultaneously by lowering the concentration elsewhere in the body. Liposomes and other lipid-based drug carriers sequester toxic drugs within a lipid membrane to provide significant advantages over systemic therapy by altering drug biodistribution, maximizing efficacy, while minimizing damage to healthy organs and tissues (Allen and Cullis, 2004; Sengupta et al., 2005). Submicron liposomes and other lipid-based nanocarriers can remain in the bloodstream for extended periods allowing for accumulation in regions of tumor growth or inflammation due to the poorly formed and leaky vasculature, a mechanism known as the enhanced permeation and retention effect (EPR) (Allen and Cullis, 2004). However, it is difficult for a given liposome to combine the necessary physical integrity and drug retention in circulation (to maximize drug accumulation at the disease site) with rapid contents release at the disease site (to affect therapy and minimize drug resistance). For example, liposomal doxorubicin (a chemotherapy drug) reduces drug-related toxicity; however, its therapeutic activity is reduced despite its efficient delivery to tumors because of slow release from the liposome carriers (Abraham et al., 2005). Therefore, one of the current challenges for liposomes and other carriers is how to initiate the release of encapsulated drugs with both spatial and temporal controls. External signals such as ultrasound (Huang and MacDonald, 2004) and visible light (Mueller et al., 2000; Shum et al., 2001) have been used to induce contents release from liposomes, but these methods are limited to surface-accessible areas

Synthesis, Characterization, and Optical Response

281

such as the eye and skin. For delivery to deeper tissues, other strategies have emerged, ranging from liposomes sensitized to general hyperthermia (Ponce et al., 2006), and receptor-targeted (Noble et al., 2004), and pH- or enzymatically triggered liposomes (Davidsen et al., 2003; Simo˜es et al., 2006). It is difficult, however, to incorporate a destabilizing agent into the liposome membrane to promote release without compromising long-term stability and drug retention in the circulation; by contrast, liposomes optimized to be robust and resistant to leakage in the circulation are hampered by suboptimal drug release. Active targeting requires specific ligands with high affinities to receptors overexpressed on diseased cells that can lead to ‘‘bindingsite barriers’’ where the tightly bound nanocarriers prevent drug penetration into the tissue (Peer et al., 2007). In addition, targeting a different site requires identifying an appropriate receptor as well as the synthesis and characterization of new ligands. To address the joint challenges of controlled release and specific targeting, we coupled hollow gold nanoshells (HGNs) that strongly absorb near infrared (NIR) light, to remotely trigger content release from conventional liposomes and ‘‘vesosomes’’ (multicompartment lipid-based carriers, or larger liposomes encapsulating multiple smaller liposomes, Kisak et al., 2004) within seconds, using an external, pulsed laser source. The HGNs can be tethered chemically to the liposome surface, encapsulated within the liposomes, or even just be in solution with the liposomes. Absorbing pulsed laser light causes the HGNs to rapidly increase in temperature, leading to the formation of microscopic vapor bubbles (Huang et al., 2006; Tong et al., 2007) in the vicinity of the liposome bilayer, the collapse of these bubbles causes transient liposome membrane rupture and contents release (Wu et al., 2008). The effects on the liposome membrane are similar to those induced by ultrasound-induced cavitation; sonication is a commonly used method to create small, unilamellar vesicles from a lamellar dispersion and is well known to disrupt bilayer membranes. In addition to membrane disruption and fast contents release, only HGN/liposome complexes directly irradiated by the laser are ruptured, providing the necessary spatial control of contents release. Alternatively, continued irradiation of the HGNs can induce localized hyperthermia or permeabilized cell bilayers, both of which can promote drug uptake by cells, or even lead directly to cell death (Chen et al., 2007; Hirsch et al., 2003; Norman et al., 2008; Tong et al., 2007). The great advantage of NIR to activate liposome drug-release is that tissue, blood, etc. are relatively transparent to 700–1100 nm wavelength light, allowing penetration depths of several centimeters (Weissleder, 2001). Gold nanostructures designed to have a plasmon resonance at NIR wavelengths, that is, silica core/gold nanoshells (Hirsch et al., 2003; Prasad et al., 2005), gold nanorods (Huang et al., 2006; Norman et al., 2008), and HGNs or nanocages (Chen et al., 2007; Prevo et al., 2008; Sun et al., 2003) are especially effective at absorbing NIR light and converting this energy into

282

Guohui Wu et al.

heat. HGNs are similar to silica core/gold nanoshells that have been used both in vitro and in vivo to accumulate NIR light (Hirsch et al., 2003), except that HGNs have a hollow core, which allows easier synthesis, (Prevo et al., 2008) smaller overall dimensions (Prevo et al., 2008; Sun et al., 2003), and no silica to interact with tissues. Gold nanoshells and nanorods illuminated with NIR light have been used successfully to noninvasively heat and eradicate diseased cells and tissues both in vivo and in vitro (Chen et al., 2007; Hirsch et al., 2003; Huang et al., 2006; Norman et al., 2008; Tong et al., 2007). Heat-transfer analysis confirms experimental observations that absorption of nano- to femtosecond pulses of NIR light causes the temperature of the HGNs to reach the melting temperature of gold, causing the collapse of the nanoshells into solid nanospheres (Prevo et al., 2008; Wu et al., 2008). The conversion of the optical energy into heat is so fast (nanoseconds) that thermal energy dissipation to the surrounding fluids occurs after the HGN reaches its maximum temperature (Link et al., 1999a). As the high-temperature HGNs equilibrate with the surrounding fluid, large-temperature gradients induce the boiling of microscopic amounts of water within microseconds (Lapotko et al., 2006; Lin and Kelly, 1998; Wu et al., 2008). These microbubbles are unstable, and the large volume of cold water surrounding the bubbles causes them to collapse in the same way as sonication-induced cavitation bubbles. The bubble collapse induces mechanical stresses in the surrounding fluid, that can tear lipid membranes apart, thereby releasing the contents of liposomes or other lipid-based drug carriers (Wu et al., 2008). The solution returns to equilibrium less than a millisecond after the initial laser pulse (Wu et al., 2008). Depending on the laser power and the pulse repetition rate, the average solution temperature is increased by no more than a few degrees (Prevo et al., 2008). There is a minimum energy threshold for drug release and a characteristic acoustic response of the solution on irradiation (Wu et al., 2008), similar to sonication-induced cavitation (Lin and Kelly, 1998). Neither the liposomes nor their contents are degraded chemically by the irradiation and release. The potential advantages of this new photoactivated release include (Sengupta et al., 2005) (i) synergistic disease-cell targeting by combining drug-carrying particles (liposomes) and energy-absorbing particles (HGNs) (Allen and Cullis, 2004), (ii) localizing release without harmful effects on surrounding healthy tissues, with no cytotoxicity or cutaneous photosensitivity as in photodynamic therapy, since the gold nanoparticles are inert (Abraham et al., 2005), (iii) triggering up to several centimeters inside the body as most tissues are transparent to NIR light (Huang and MacDonald, 2004), and (iv) creating high-localized concentrations of drug with both spatial and temporal controls. A variety of liposome or polymeric (Discher et al., 1999) carriers could be modified by tethering or encapsulating HGNs to produce a system for rapid, targeted release on demand via NIR irradiation. In

283

Synthesis, Characterization, and Optical Response

addition, HGN-induced liposome disruption could be used to induce rapid diffusional mixing to permit the study of fast chemical kinetics in nanoenvironments mimicking cell membranes (Chiu et al., 1999).

2. Synthesis of HGNs

H )n 2 CH 2O

CH

S-(CH2CH2O)n-H

S-(

2 CH 2 O) n -H

C S-(

-H O) n

2

NIR

CH

S-(

-H O) n

2

CH

2

CH

S-(

CH

CH

CH

S-(

2

Au NP

2 CH 2 O) n -H

-H O) n

800 nm

2

CH

-H O) n

2

S-(

S-(CH 2CH2O)n-H

S-(

CH

2 CH 2 O) n -H

CH

2 CH 2 O) n -H

CH

CH

2

S-( S-(

2

CH

2

-H O) n

Au NS

CH

2

S-(CH2CH2O)n-H

H-(OCH2CH2)n-S

Ag

CH

S-(

S-(

PEG750-SH S-(CH2CH2O)n-H

+Au+3

CH H2

2

-H O) n

The templated galvanic replacement reaction of silver for gold (Chen et al., 2007; Hao et al., 2004; Liang et al., 2005; Prevo et al., 2008; Schwartzberg et al., 2006; Sun and Xia, 2004; Sun et al., 2002; Wiley et al., 2004, 2005) provides a simple and reproducible, nontoxic route to HGNs 20–50 nm in diameter for use in biomedical applications (Scheme 14.1). Silver nanoparticles with diameters of the desired HGN core size are synthesized first, then are sacrificed by adding a gold salt to the solution; the gold is reduced to metal because it has a greater standard reduction potential than the silver template, which is oxidized to a molecular solution (Sun et al., 2002). The gold plates onto the outside of the dissolving silver nanoparticle, resulting in an HGN of controlled diameter; the shell thickness is determined by the relative amount of gold salt to the silver template. The ratio of shell diameter to shell thickness governs the wavelength of the HGN absorbance (Oldenburg et al., 1998); the surface plasmon resonance (SPR) absorbance of nanoshells made in this fashion can be tuned very simply by applying Turkevich’s basic colloidal growth chemistry to the sacrificial silver nanoparticles (Link et al., 1999a; Turkevich, 1985; Turkevich et al., 1951, 1954). The emphasis here is not on making shape-specific or extremely monodisperse nanoshells, but rather on a simple and scalable route to nanoshells for practical applications, with tunable sizes and absorbance profiles that require minimal experimental footprints (reduced heating, minimal separations, etc.), and minimal exposure to toxic solvents, reagents, or intermediates that would be detrimental to biomedical applications. The major benefit of this synthesis is that it is

Scheme 14.1 Schematic illustration of the synthesis and stabilization of gold nanoshells as well as their structural change after NIR laser irradiation.

284

Guohui Wu et al.

rapid, stable, highly scalable, and in many cases is a true ‘‘one-pot’’ synthesis (Hao et al., 2004; Liang et al., 2005; Sun and Xia, 2004). Silver templates are prepared at 60
C in a well-stirred 600 ml solution of 0.2 mM silver nitrate (AgNO3; Fisher Scientific, Atlanta, GA) with 0.6 ml 1.0 M sodium borohydride (NaBH4; Fisher Scientific) in the presence of 0.5 mM sodium citrate ( J.T. Baker Chemical Co., Phillipsburg, NJ). The solution is stirred for at least 2 h to allow the NaBH4 to fully hydrolyze. The addition of sodium borohydride accelerates the chemical reactions, and the resulting nanoparticles are 15–25 nm in diameter (Prevo et al., 2008). After cooling to room temperature, larger silver nanoparticles could be grown from these stock sols, if desired. Silver particle growth is initiated by adding 0.5 ml of 2.0 M hydroxylamine hydrochloride solution (NH2OHHCl, Aldrich, Milwaukee, WI) to the silver sol, followed by stirring for 5 min (Turkevich, 1985; Turkevich et al., 1951, 1954), addition of 1.25 ml 0.1 M AgNO3 (0–1 ml), and stirring overnight. The growing silver nanoparticles turn the sol a darker yellow or orange, depending on the amount of additional AgNO3. Gold nanoshells could then be made via galvanic replacement chemistry from the template silver sols without the need to isolate the silver nanoparticles. First, a given silver sol (50 ml) is heated to 60
C and the necessary amount (3.2 ml for the nanoshells in Fig. 14.1) of 25 mM tetrachloroauric acid (HAuCl4, Aldrich) is added dropwise (depending on the initial silver template size). Silver (Agþ/Ag 0.8V, vs. SHE) has a lower redox potential than gold (AuCl4/Au 0.99V, vs. SHE) and the replacement reaction is (Sun et al., 2002): þ  3AgðsÞ þ AuCl 4ðaqÞ ! AuðsÞ þ 3AgðaqÞ þ 4ClðaqÞ

Upon the addition of the concentrated HAuCl4, the solution turns from yellow/orange to gray/yellow to blue/gray to blue/turquoise within seconds as the silver and gold are oxidized and reduced, respectively. The reactions are monitored using UV/vis/NIR spectroscopy, and stopped when the silver peak located near 400 nm vanishes (usually within a few minutes, although the reaction mixture is generally stirred for at least 1 h after gold addition). This which occurs when the gold/silver ratio in the reaction vessel approaches the stoichiometric ratio of 1:3 (to err on the side of completion, the ratio is usually 37:1). Once the reaction is complete, the samples are cooled, silver chloride is allowed to precipitate, and the supernatant containing the gold nanoshells are transferred to another vessel and stored at 4
C until further use. The size distribution and particle morphology are analyzed by transmission electron microscopy using an FEI Tecnai T20 microscope (Fig. 14.1). A typical HGN is spherical and hollow, although some have irregular shapes or incomplete shells. This particular set of concentrations gave HGNs of diameter 33  13 nm and shell

Synthesis, Characterization, and Optical Response

285

Figure 14.1 TEM images of HGNs with a variety of morphologies. All of the nanoshells have a hollow core, that is, the light gray area near the center of each nanoshell, surrounded by the gold shell, which is dark gray to black. Some nanoshells are not complete (arrow). Bar is 50 nm.

thicknesses of 2  0.9 nm, and had a maximum adsorption around 820 nm (see Fig. 14.7) using a Jasco V-530 UV/vis spectrometer ( JASCO Corp., Tokyo). The synthesized HGNs are quite stable in the dilute synthesis solutions because of electrostatic repulsion caused by the adsorption of citrate ions, resulting in an overall negative charge on the HGN. However, to improve their stability against aggregation in physiological buffers and other high-ionic strength solutions, the HGNs are further stabilized sterically by tethering poly(ethylene glycol) (PEG) of molecular weight 750 Da to the HGN via a thiol linker. Methoxypolyethylene glycol amine (750PEG-NH2, PEG molecular weight of 750 Da, Aldrich) is converted to methoxypolyethylene glycol thiol (750PEG-SH), using a twofold molar excess of 2-iminothiolane HCI (also known as Traut’s reagent; Sigma-Aldrich, St. Louis, MO) in buffer (2.28 mM Na2HPO4, pH 8.8). One-half milliliter of the prepared 0.0379 M 750PEG-SH is added to 600 ml of the HGN to achieve a 1000:1 ratio of thiol:gold. The PEG-stabilized HGN solution is centrifuged at 21,000g for 30 min and redispersed in Milli-Q water twice to remove any unattached and/or soluble chemicals. The pellet readily redisperses in

286

Guohui Wu et al.

buffer. Complete gold nanoshells and nanorods with various coatings can also be purchased directly from Nanopartz Co. (Salt Lake City, Utah). In our case, the PEG-modified HGNs are stable for at least 1 year in physiologic buffers (longer term stability is still being tested). No flocculation is observed after mixing the HGN solution with concentrated carboxyfluorescein (CF) in buffer at basic pH; severe precipitation occurs instantly for HGNs without 750PEG-SH. For in vivo applications, PEG coatings can minimize recognition by the immune system and minimize nonspecific binding to blood proteins and cells. Similar PEG-modified gold nanoparticles show little cytotoxicity in vitro (Niidome et al., 2006). After intravenous injection, PEG-modified gold nanoparticles circulate in mice with a halflife of approximately 1 h, and, for at least 72 h, there is no accumulation in major organs except for the liver (Niidome et al., 2006). The free methoxy end of the 750PEG-SH can also be modified to display various ligands or antibodies to allow the HGNs to be tethered to various lipids or proteins. Unlike gold nanorod syntheses, this approach does not require the shape-determining detergent hexadecyltrimethylammonium bromide (CTAB), which is cytotoxic. CTAB adsorbs to gold very strongly and it is difficult to replace CTAB with other ligands; removal of CTAB can result in nanoparticle aggregation (Niidome et al., 2006). An additional advantage of this galvanic replacement synthesis is the small size of the HGNs, which is especially useful when enclosing HGNs within liposomes for controlled release. In comparison, the silica core/gold shell nanoparticles pioneered by Hirsch et al. (2003) are at least 100 nm in diameter, and the dielectric core material (either silica or polystyrene) brings the additional concern of potential biological effects of silica or the products of silica degradation. An added benefit is that, compared with the one-pot synthesis of HGNs, the silica core/gold shell nanoparticle synthesis is time-consuming and laborintensive (Shi et al., 2005).

3. Optimization of HGN Dimensions for Maximum Absorption in the NIR An idealized HGN structure allows for analytical predictions of light absorption and scattering by means of classical field theories, such as the Mie scattering formalism (Kreibig and Genzel, 1985). Analytical solutions exist for the interaction of light with spherical metal nanoparticles, nanoshells, and nanorods, that is, structures possessing a high degree of symmetry. The optical properties of less symmetric particles or aggregates require numerical solutions of Maxwell’s equations. For spherical shell nanoparticles such as HGNs, analytical solutions for the far-field extinction, absorption, and scattering cross sections have been known for almost a century and can be found, for

287

Synthesis, Characterization, and Optical Response

example, in the classic book by Bohren and Huffman (1983). These cross sections are related to the attenuation of photons, N, according to N =N0 ¼ expð nsext xÞ, in which sext is the HGN extinction cross section, n is the number/volume of HGN, and x is the path length of light with initial number of photons, N0. The extinction cross section is the sum of the scattering cross section, ssca, and absorption cross section, sabs, of the incident light. To model the absorption and scattering of HGNs, analytical solutions of Mie scattering for core/shell nanoparticles can be used as described by Bohren and Huffman (1983). The dielectric function (or complex index of refraction) for gold is taken from various sources ( Johnson and Christy, 1972), but similar results are obtained for all sets of material properties. The extinction efficiency, Qext, is the ratio of the extinction cross section to the geometric cross section: Qext ¼ sext/pa2, a is the HGN radius. The absorption efficiency, Qabs, is the difference between Qext and Qsca, the scattering efficiency: Qabs ¼ Qext  Qsca ¼ sabs/pa2. Ideally, all of the radiation would be adsorbed by the HGN (Qabs >> Qsca) to provide the maximum HGN heating. The light scattered by the HGNs is lost for useful purposes. Figure 14.2 shows the absorption and scattering efficiencies as a function of wavelength for an idealized HGN with an overall diameter of 33 nm and a shell thickness of 1.7 nm. For wavelengths around 800 nm, absorption dominates scattering for these HGNs, suggesting that almost all of the laser energy is being converted to heat. The peak value of the absorption cross section for an HGN with these dimensions is 9.12  10 11 cm2. Figure 14.3 shows the calculated optimal shell thickness d as a function of the HGN diameter D. HGNs with these dimensions exhibit the maximum

12

Efficiency

10 Scattering Absorption

8 6 4 2 0 400

600 800 Wavelength (nm)

1000

Figure 14.2 Theoretical absorption and scattering efficiency for HGN with a diameter of 33 nm and shell thickness of 1.75 nm in water.

288

Guohui Wu et al.

7

Shell thickness (nm)

6

5

4

3

2

20

40

60 80 Particle diameter (nm)

100

Figure 14.3 Optimal shell thickness providing the maximum absorption efficiency at 800 nm as a function of HGN diameter.

absorption at 800 nm (optimal wavelength for a Ti:Sapphire laser). This result shows that the smaller the HGN, the thinner the gold shell must be to insure a maximum adsorption in the NIR range of the optical spectrum. For example, a 40-nm diameter HGN should have a shell thickness of 2 nm, while an 80-nm diameter HGN should have a shell thickness of  5 nm to maximize the absorption at 800 nm. Figure 14.4 shows the calculated size-dependent absorption efficiency, Qabs, and absorption cross section, sabs at 800 nm. The absorption efficiency peaks for HGNs with D  50 nm, but the absorption cross section peaks for D  85 nm. For a given mass of gold, the HGN with the highest efficiency should be chosen as the mass/HGN increases faster with HNG diameter than does the optical cross section. The peak absorption cross section is 4  10 10 cm2. For an optical energy density of 2.2 mJ/cm2 corresponding to the cavitation threshold, HGN of any size will be heated well above the melting point of bulk gold (Prevo et al., 2008). Smaller HGN will be heated to higher temperatures due to their smaller mass, but particles with D  85 nm will absorb the maximum energy/particle (Prevo et al., 2008). Thus, it is difficult to predict a priori which HGN size is optimal for a given application. While the direct implementation of Mie theory does a good job of predicting the wavelength at which the maximum absorbance occurs for the experimentally determined HGN size distribution (Fig. 14.5), it fails to

289

Synthesis, Characterization, and Optical Response

14

12 300 10 200 8 100

Absorption cross-section (cm2)

Absorption efficiency

400 ⫻ 10−12

6 20

40

60 80 HGN diameter (nm)

100

Figure 14.4 Absorption efficiency (red) and cross section (blue) at 800 nm of HGNs with optimized geometry as a function of diameter.

describe the inhomogenously broadened extinction spectrum of the HGN ensemble (Fig. 14.6). The statistical distribution of shell thickness and HGN diameter are determined from the TEM micrographs and fit by a lognormal distribution function (Fig. 14.5). The extinction cross section of the HGN ensemble has been modeled using these distribution functions. Figure 14.6 shows that the simulation predicts less broadening of the absorption spectra than what is observed experimentally. The experimental broadening may be due to the shape polydispersity, as well as possible chemical variations due to any residual silver alloy in the HGN. An additional complication is that the shell thickness in the HGNs is significantly less than the electron mean free path in gold (50 nm), and the dielectric function can be strongly affected by the scattering of electrons at the nanoparticle boundaries. To address this effect a size-dependent term is introduced into the dielectric function of gold according to Alvarez et al. (1997) and the simulation optimized numerically to achieve the best match between the experimental extinction spectra and the theoretical model. The use of the size-dependent dielectric function does not resolve completely the discrepancy between the experimental and theoretical extinction spectra. This may be explained by the fact that thin layers on the metal surface may be depleted/enriched in electron density (also the silver alloying), depending on the properties of the environment, which can change the effective dimension of the HGN (Kreibig, 1995).

290

Guohui Wu et al.

200

Occurence

150

100

50

0 0

20

40 60 Particle diameter (nm)

80

Occurence

15

10

5

0 0

1

2 3 4 Shell thickness (nm)

5

Figure 14.5 Measured diameter (left) and shell thickness (right) from TEM images for HGNs synthesized as described in text.

4. HGN Response to Femtosecond NIR Laser Pulses The HGNs prepared as discussed above have a broad absorption peak at 780–820 nm (Figs. 14.6 and 14.7), which overlaps well with the Ti: Sapphire pulsed laser (Spectraphysics Spitfire), which has an FWHM of 12 nm centered around 800 nm. Compared with conventional dyes, HGNs are much stronger than NIR absorbers and are less susceptible to photobleaching or other forms of chemical degradation. However, sufficient irradiation does lead to collapse of the HGN into solid gold

291

Synthesis, Characterization, and Optical Response

1.4

Absorbance (a.u.)

1.2 1.0 0.8

Experiment Simulation

0.6 0.4 0.2 0.0 400

500

600 700 800 Wavelength (nm)

900

1000

Figure 14.6 Measured (red) and calculated (blue) extinction spectra of the experimental distribution of HGN diameters and shell thicknesses from Fig. 14.5.

nanoparticles; this collapse leads to a loss of absorption in the NIR (Fig. 14.3). The original color of the HGN suspensions is dark blue (Fig. 14.7); after pulsed NIR laser irradiation at 800 nm at a power density of 16.1 mJ/cm2 for 8 min (120fs pulses at 1 kHz repetition rate), the HGN suspension turns a dark red color. The corresponding UV–VIS absorption spectra exhibit a shift of the absorption peak from 780 to 470 nm, indicating the conversion of the HGN to solid nanoparticles (Prasad et al., 2005; Prevo et al., 2008). TEM images of irradiated particles confirm the change in the HGN morphology; the hollow center of the nanoshell collapses and the particles anneal into the more stable solid spheres, consistent with the change in color and the shift of the adsorption maxima (Fig. 14.7). These observations confirm that HGNs reach sufficiently high temperatures after femtosecond pulses of NIR light to melt and anneal into more stable shapes. For a solid-core, spherical Au particle with size from 2 to 100 nm, the SPR is from 520 to 570 nm (Kreibig, 1977). For other gold morphologies, such as rods, cages, thin plates, and aggregates, the absorption peak due to the SPR shifts to lower energy (higher wavelength) in the NIR window. The peak at 470 nm can be explained by the formation of Au–Ag alloy nanoparticles during the laser-induced heating. Ag has an SPR at 400 nm; the plasmon absorption of Au–Ag alloy falls between the SPRs of Ag and Au nanoparticles and varies linearly with the Au mole fraction (Link et al., 1999b). X-ray photoelectron spectroscopy confirms the presence of residual silver in the HGN (Prevo et al., 2008). A second explanation for the 470-nm peak is the presence of Au clusters less than 2 nm in size

292

Guohui Wu et al.

2 NIR

Extinction (a.u.)

1.5

1

0.5

0

400

600 800 Wavelength (nm)

1000

Figure 14.7 Extinction spectra (normalized at 350 nm) of: (blue) HGN suspension; (red) HGN suspension laser-irradiated at a power density of 16.1 mJ/cm2 for 8 min (120 fs pulses at 1-kHz repetition rate). Below the peaks of extinction curves are photographs of the corresponding dispersions. The spectra are consistent with the collapse and annealing of the HGN into solid gold/silver alloy nanoparticles.

due to laser-induced particle heating and fragmentation (Link et al., 1999b). The existence of such fine Au particles has also been postulated to explain the similar absorption peak at 460 nm of Au/SiO2 after ultrasound treatments (Link et al., 1999b). Rapid nonradiative relaxation processes convert the absorbed light energy mainly to heat (Grua et al., 2003). Petrova et al. (2006) found that at temperatures higher than 250
C, nanorods anneal into spherical particles in less than 1 h; we have found that HGN also anneals into solid particles within hours at 250
C. This is significantly below the melting point of bulk gold (1064
C). Time-resolved spectroscopy studies, however, showed that nanorods (in aqueous solution) maintained their shape in spite of a rapid lattice temperature increase to 1000 K achieved by ultrafast pulsed laser excitation. Petrova et al. (2006) explained that ‘‘the difference in the temperature stability of the nanorods under continuous thermal heating compared to laser-induced heating is attributed to thermal diffusion: the rods do not stay hot for long enough after ultrafast excitation for significant structural transformation to occur.’’ On the other hand, the melting of HGN suggests that the structural changes in HGN after femtosecond laser-induced heating requires that the temperature in the HGN gold lattice must be significantly higher than 1000 K. The amount of rearrangement required to collapse an HGN into a solid sphere is also much less than that required to change a solid gold nanorod to a sphere.

Synthesis, Characterization, and Optical Response

293

The maximum temperature reached by the HGNs is also limited by the rate of energy dissipation into the surrounding liquid. However, the thermal diffusivity of water limits the conduction of heat from the HGNs, and heat dissipation from the HGN to the surrounding water is much slower (microseconds) than the electron dynamics involved in plasmon-mediated heating (nanoseconds) (Link et al., 1999a; Prasad et al., 2005; Prevo et al., 2008; Roper et al., 2007). Essentially, all of the optical energy input to the HGN goes into heating the HGN (Prevo et al., 2008). The high temperature HGN then dissipates its thermal energy into the surrounding water by conduction; large temperature gradients are generated in the vicinity of the HGN, which cause the surrounding water to boil and form rapidly growing microbubbles. These microbubbles cannot grow indefinitely as there is not sufficient energy within the HGN to raise the bulk of the solvent to the boiling temperature (Prevo et al., 2008). The bubbles become unstable and can undergo a violent collapse which produces shock waves (Pecha and Gompf, 2000) or microjets (Popinet and Zaleski, 2002). Recent time-resolved X-ray scattering experiments show that femtosecond laser excitation of gold nanoparticles leads to the compression of the solvent, which is consistent with bubble formation (Kotaidis et al., 2006). The growth and collapse of unstable vapor bubbles also produces a detectable pressure change in the bulk solution, which gives a photoacoustic signal that can be ‘‘heard’’ by a hydrophone (Model #TC4013, Reson, Goleta, CA). The bandwidth of the hydrophone is 1 Hz–170kHz. The hydrophone, which is 0.5 cm in diameter and  2 cm long, is immersed into the HGN solution 5 mm above the laser beam in a quartz cuvette with 10-mm light path. The solution volume is 2.5 ml. The sample is not stirred during the laser irradiation to minimize the acoustic noise. Data collection is synchronized with the laser cavity dumping event by using triggering signal from the laser control electronics, and the output of the hydrophone is collected by a digital oscilloscope (Tektronix TDS5032B). The acoustic transients are averaged over several hundred individual laser pulses to improve the signal/noise ratio. Figure 14.9A shows a typical acoustic signal of pressure fluctuations in a 0.142 mM HGN solution as recorded by a hydrophone between two 130 fs laser pulses (1-KHz repetition rate). No pressure fluctuations above background occurred in control solutions of phosphate buffer solution, or 4.71 mM CF solution dissolved in phosphate buffer, irradiated with the highest pulsed laser power density of 16.2 W/cm2 (Fig. 14.9B). Laser-induced heating, however, does not always produce an acoustic signal. For spherical gold particles in aqueous solution, explosive boiling happens only above a threshold temperature of 85% of the critical temperature (Tc ¼ 374
C for water) (Kotaidis et al., 2006). We have observed that a similar threshold of the laser energy density is required to generate a photoacoustic signal. The amplitude of the pressure fluctuations remains at

294

Guohui Wu et al.

Figure 14.8 TEM images showing the collapse of the HGN into solid nanoparticles after 16.1 mJ/cm2 NIR-laser irradiation.

background up to a laser power density of 2.3 mJ/cm2, above which the amplitude of the pressure fluctuations increases with increasing laser power density (Lin and Kelly, 1998) (Fig. 14.9A). Calculations show that the increased power density leads to higher HGN temperatures (Prasad et al., 2005; Prevo et al., 2008), which are then translated into larger pressure fluctuations (more water boils and larger bubbles are formed) as this energy is dissipated. The pressure fluctuations die out within a few hundred microseconds after the light pulse; the HGN and the surrounding solution equilibrate and return to ambient temperature prior to the next pulse from the laser. Although the gold nanoshells melt (Fig. 14.8), the temperature increase of the bulk solution is small; the sample reaches a steady temperature only a few degrees above ambient, which depends on the laser power density and the HGN concentration. This is because the absolute amount of energy deposited into the solution is quite small (the laser power here is 0.67 W). Therefore, any significant pulsed laser-induced heating is limited to the immediate vicinity (microns) of the HGNs. This is important for the biomedical application of HGNs as only liposomes, vesosomes (liposomes encapsulating multiple smaller liposomes), or tissues directly adjacent to the irradiated HGNs will be affected by the light pulse. However, it is well established that transient cavitation bubbles due to sonication are capable of disrupting cell and liposome membranes (Huang, 2008). The irradiated HGNs might be thought of as optically triggered, nanosonicators, which can cause the transient rupture of liposomes resulting in rapid contents release.

5. Coupling HGN to Liposomes As the extent of the heated zone is quite small, the method of coupling the HGN to liposomes or cell membranes is important. We have examined three different methods of rupturing liposomes with HGNs. First, HGNs

295

Synthesis, Characterization, and Optical Response

2

2.

0.4 0.2

0.4

) Accoustic signal (mV

m

44

m J/ cm 2 J /c 31 m 2 m 1. J/ cm 34 2 m J/ 0. cm 72 2 m J/ 0m cm 2 J/ cm

) Accoustic signal (mV

3.

7.

59

J/ cm 2 m

m

.1

48 9.

16

B J/ cm 2

A

0.2 0.0

−0.2

0.0 −0.2 −0.4

1000

Buffer,

2

J/cm

6.1 m on, 1

luti CF so

−0.4 −0.6

2

HGN

800

−0.6

2

cm

16.1 mJ/

/cm

J , 0m

(ms )

600

0

Ti

200

me

400

0

200

800 600 (ms) ime

400

1000

T

Figure 14.9 Acoustic signal amplitude of (A) 0.142 mM HGN solution as a function of pulsed laser energy is recorded by a hydrophone after a single laser pulse of various laser energy densities. The magnitude of the photoacoustic signal of pressure fluctuations associated with cavitation (Paliwal and Mitragotri, 2006) increases with increasing laser power density, consistent with larger energy absorption by the HGN and subsequent larger or more numerous cavitation bubbles. However, a threshold value of laserenergy density of 2.3 mJ/cm2 is required to induce the photoacoustic signals above background in the HGN solutions. (B) Acoustic signal amplitude in control solutions. No signal was observed in buffer or CF solution without HGN, and no signal was observed in HGN solution without laser irradiation.

can be encapsulated within dipalmitoylphosphatidylcholine (DPPC) liposomes via the interdigitated phase transition, which causes lipid membranes to form flat open sheets at low temperatures that close to form unilamellar vesicles at higher temperatures (Boyer and Zasadzinski, 2007; Kisak et al., 2002) (Scheme 14.2A). Second, the nanoshells can be tethered to the outside of preformed liposomes using a thiol/PEG–lipid linkage (Scheme 14.2B). Third, the HGN solutions can be mixed with preformed liposomes so that the HGNs are exclusively outside the liposomes (Scheme 14.2C). Scheme 14.2A: Encapsulation of HGN and CF inside liposomes. 6-Carboxyfluorescein (50 mM) (CF; Invitrogen; Eugene, OR) is dissolved in water together with 6 equiv. of concentrated NaOH, which converts the CF from its acid form to the water-soluble salt form. The CF solution is used to disperse the PEG-stabilized HGNs described previously. Liposomes are prepared as described by Boyer and Zasadzinski (2007) and Kisak et al. (2002). DPPC dried from chloroform is hydrated with Milli-Q water by vortexing at 55
C to form a lamellar dispersion, which is then cooled to room temperature. Vesicles are prepared by sonication at room temperature using a 60 Sonic Dismembrator (Fisher Scientific) for 4 min at a power of 4 W.

296

Guohui Wu et al.

A

Inside

Interdigitated lipid sheets B

Tethered NIR Liposome Lipid-PEG-SH Freely outside

C Liposome

Scheme 14.2 HGNs can be located: (A) inside the liposomes; (B) tethered to the liposomes; and (C) free outside the liposomes.

At room temperature, DPPC bilayers are in the gel or Lb0 phase; the interdigitated bilayer phase (LI) is induced by the addition of 0.106 ml of ethanol (3 M net ethanol concentration) to 0.5 ml of a 50 mg/ml DPPC vesicle suspension. The initially bluish vesicle suspension turns milky white, and its viscosity increases significantly. The vesicles fuse and burst open, forming stacks of open bilayer sheets many micrometers in size (Boyer and Zasadzinski, 2007; Kisak et al., 2002). After annealing at 4
C overnight, the interdigitated sheets are centrifuged at 3000g and dispersed in pure water three times to remove any ethanol. The pellet of interdigitated DPPC sheets is mixed with the solution of 32 mM CF and 12 mM HGN and heated at 50
C for 2 h under vortex mixing. Raising the temperature causes a phase transition from LI to the liquid crystalline La phase and the bilayer sheets become much more flexible, allowing them to close around the HGN in suspension to form interdigitation–fusion vesicles (Boyer and Zasadzinski, 2007; Kisak et al., 2002). The internal concentrations in the liposomes are 32 mM CF (110 mOsm), 12 mM HGN, and the overall lipid concentration is 22 mg/ml DPPC. Based on earlier work (Boyer and Zasadzinski, 2007), about 50–60% of the HGN is encapsulated in liposomes. Cryo-EM confirms that the HGN are encapsulated within the liposomes by this procedure (Wu et al., 2008). After encapsulation, phosphate-buffered saline (PBS; 20 mM Na2HPO4/ NaH2PO4, 34.5 mM NaCl, pH ¼ 7.4) is used to disperse the liposomes to minimize osmotic stress across the membrane. The unencapsulated CF is

297

Synthesis, Characterization, and Optical Response

removed by size-exclusion chromatography using a Sephadex G-75 column (Amersham Biosciences Corp., Piscataway, NJ) eluted with PBS buffer. The eluted suspension is centrifuged at 100g for 20 min and redispersed in PBS buffer twice to remove any unencapsulated HGN prior to irradiation. Scheme 14.2B: Tethering HGN to liposomes containing CF. A pellet of interdigitated DPPC sheets is prepared as described above. The interdigitated sheets are mixed with a 32 mM CF in water solution containing 2 mol% (relative to the amount of DPPC used) DSPE-2000PEG-NH2 powder (Avanti Polar Lipids, Alabaster, AL). The mixture is then heated at 50
C for 1 h under vortex mixing. The DSPE-2000PEG-NH2 partitions into the DPPC bilayers in the La phase as the temperature increases and the interdigitation fusion vesicles are formed (Sou et al., 2000; Zalipsky et al., 1996). Next, the amine groups at the liposome surfaces are converted to thiol by mixing with 100% excess 2-iminothiolane solution (0.29 M). The thiolated liposomes encapsulating CF are incubated with a solution of HGN and CF for 48 h to allow HGN to tether via the thiol linkages to the outer surfaces of liposomes. The final concentrations in the solution are 18 mg/ml phospholipid (98 mol% DPPC and 2 mol% DSPE-2000PEG-SH), 18 mM HGN and 32 mM CF. The liposomes with tethered HGNs are eluted through a Sephadex G-75 size-exclusion column to remove any unencapsulated CF, and centrifuged at 200g to remove untethered HGNs. Figure 14.10 shows a cryo-EM tilt series that confirms that the HGN are tethered to the liposomes (Wu et al., 2008). Thin films of the liposome/ HGN solution are spread on holey carbon TEM grids (Structure Probe, West Chester, PA) under controlled temperature and humidity conditions using a VitRobot (FEI Company, Hillsboro, OR) (Frederik and Hubert, 2005), then vitrified by rapid plunging into liquid ethane (Chiruvolu et al., 1994a,b; Frederik and Hubert, 2005; Jung et al., 2002). Cryo-EM imaging is performed on an FEI Tecnai T20 microscope, operating at 200 kV with a Gatan liquid nitrogen specimen cryo-holder. Single-axis tomographic imaging is performed with a JEOL 2010A microscope from  60
to þ60
tilt −45⬚

0⬚

+45⬚

200 nm

200 nm

200 nm

Figure 14.10 Cryo-TEM tilt series showing that the HGNs are tethered to the liposome surfaces. Arrows point out the same HGNs as they rotate with the surface of the liposome.

298

Guohui Wu et al.

angles in 2
increments with a total dose of less than 100 electrons/A˚2. Three representative images at goniometer tilt angles of 45
, 0
, and 45
are shown in Fig. 14.10. The arrows mark specific HGNs on the surface of the liposomes followed during the 90
rotation to confirm that HGNs are tethered to the surface. Scheme 14.2C: Liposomes containing CF with external HGNs. DPPC liposomes containing 32 mM CF are prepared by the same interdigitation– fusion method described earlier, except no HGNs are added to the solution prior to vesicle formation. The preformed vesicles are eluted through a Sephadex G-75 column to remove external CF, and then dispersed in different concentrations of HGN solutions as needed.

5.1. Pulsed laser optics Liposome disruption is triggered by irradiating the liposome/HGNs with the output of the femtosecond Ti:Sapphire regenerative amplifier (Spectraphysics Spitfire) running at a repetition rate of 1 kHz. The setup is illustrated in Fig. 14.11. The laser beam is collimated by a Galilean telescope to achieve a Gaussian diameter of 2.3 mm. The pulse duration is monitored by a home-built single-shot optical autocorrelator and is kept at about 120 fs. The spectral FWHM of the laser radiation is 12 nm centered around 800 nm. The laser beam is directed onto the sample by a system of mirrors; no focusing optics are used. The energy of the optical pulse is controlled by Schott neutral density glass filters. A thermopile power meter (Newport Inc., Irvine, CA) is used to measure the incident optical power. The maximum power available is 670 mW, which corresponds to

Tunable fs (ps) laser light (700–1000 nm) Beam shaping optics

Sample

Emission pickup optics and filters Two-photon absorption

Emission Spectrometer

CCD

Figure 14.11 Schematic illustration of Laser irradiation/CF release detection setup.

Synthesis, Characterization, and Optical Response

299

670 mJ/pulse and an energy density of 16.1 mJ/cm2 (or mean power density of 16.1 W/cm2). The temperature of the HGN suspensions is measured using an Omegaette HH306 digital thermometer (Omega) with a K-type thermocouple probe (Omega Engineering Inc., Stamford, CT), which is immersed into the solution 5 mm above the laser beam. The solution is stirred to ensure good mixing during irradiation. Luminescence is excited in the sample via a two-photon absorption process. The emission is collected at a 90
angle by a system of lenses and focused on the entrance slit of a monochromator (Acton Research SpectraPro 300). The laser radiation is blocked by a Schott colored glass filter (BG38). The light dispersed by the monochromator is detected by a spectroscopic CCD camera (PI Acton PIXIS-400) and transferred into a personal computer for analysis. The evolution of the photoluminescence is recorded by collecting consecutive spectra over a 600 nm bandwidth with a constant interval. To quantify the fractional release of CF, fluorescence is measured using a PTI QuantaMaster spectrofluorimeter (Photon Technology International, Lawrenceville, NJ). Any release from the liposomes is detected by an increase in fluorescence intensity (from the background) as the external concentration of CF increased. The fractional release can be quantified as fractional release ¼ (Ilaser  I0)/(Imax  I0), where Ilaser is the fluorescence intensity of the solution after laser treatment, Imax is the maximum fluorescence intensity after lysing the liposomes with reduced Triton X-100 (a nonionic surfactant which has a hydrophilic polyethylene oxide group and a hydrophobic 4-(1,1,3,3-tetramethylbutyl)-phenyl group) (Boyer and Zasadzinski, 2007), and I0 is the background fluorescence intensity before either treatment.

5.2. Continuous-wave laser irradiation To illustrate the differences between pulsed and continuous-wave irradiation, the liposome/HGN samples are also irradiated with continuous NIR light, using a Spectraphysics 3900S Ti:Sapphire CW laser. The laser wavelength is tuned to 820 nm and the output power is controlled by changing the power of the pump laser source (Spectraphysics Beamlok-2060) to be 0.7 W. The Gaussian beam diameter of the CW laser is 1.0 mm.

6. Liposome Disruption and CF Release Due to Pulsed Laser Irradiation Irradiation with femtosecond NIR light pulses with an energy density >2.2 W/cm2 triggers a near instantaneous increase in the measured fluorescence in the solution of DPPC liposomes encapsulating HGNs and CF

300

Guohui Wu et al.

Table 14.1 Comparison of triggered content release

Laser

Solution

Pulsed fs

CF CF þ Au NS Liposomes containing CF, but no Au NS Au NS suspended freely outside of liposomes containing CF Au NS and CF encapsulated inside liposomes Au NS tethered to the outer surface of liposomes containing CF Au NS and CF encapsulated inside liposomes

CW

Release (%)

02 12 1  2 28  2 71  1 93  2 1  2

(Table 14.1), but has no effect on control solutions of unencapsulated CF, a mixture of HGNs and CF, or DPPC liposomes with CF, but no HGNs. In all samples, the CF concentrations are matched to give similar concentrations averaged over the sample volume. Continuous-wave (unpulsed) laser irradiation at 800 nm with a higher average power density of 89 W/cm2 leads to no increase in the fluorescence intensity, and hence no CF release, even after 4 h of irradiation. When continuous-wave laser irradiation is used, the nanoshell is always close to being in thermal equilibrium with its surroundings, and there are insufficient temperature gradients to give rise to microbubble formation (Prasad et al., 2005; Prevo et al., 2008). From these observations, it is clear that minor changes in the solution and liposome temperature are not responsible for the rapid release of CF from the liposomes.

7. Mechanism of Triggered Liposome Release To identify the mechanism of release, the laser power density is varied as in Fig. 14.9, while comparing the total fluorescence intensity after 9 min of irradiation with 120 fs long pulses at a 1 kHz repetition rate. Figure 14.12 shows a distinct power density threshold of NIR light necessary to trigger CF release from the liposomes: no fluorescence increase is detected for a power density lower than 1.5 W/cm2, while the total release is roughly constant at about 74% for power densities greater than 4.3 W/cm2. The rate of fluorescence increase during NIR irradiation for liposomes encapsulating HGNs also increases with the laser power density (Fig. 14.13). The in situ fluorescence intensity is constant for an irradiation power density of 1.3 mJ/cm2, which is below the threshold. Above the threshold, the curves could be fit with a single exponential, with a time constant increasing with decreased

301

Synthesis, Characterization, and Optical Response

100

50

80

40

60

30

40

20

20

10

0

0 Au NS freely outside liposomes

80

40

60

30

40

20

20

10

0

Derivative

Dye release (%)

Au NS inside liposomes

0 0

2 4 6 8 10 12 14 Laser power density (mJ/cm2)

16

Figure 14.12 Effect of energy density on CR release from liposomes with HGNs encapsulated inside and suspended outside (black data points). The solid red curves are sigmoidal fits to the data; the blue curves are the derivative of the fit. The peak in the derivative gives the threshold energy for CF release, which is the same for both samples. The maximum release is quite different due to the different average distances between HGNs and the liposome bilayers.

laser power density (time constant t ¼ 5, 52, and 112 at the power density of 14.9, 13.0, and 7.1 W/cm2, respectively). At the higher power densities, release is complete within seconds. The derivative of the release versus laser-energy density curves gives an estimate of the threshold energy density for contents release (Fig. 14.12). For both HGNs inside and free outside the liposomes, the threshold value is the same, 2.2 mJ/cm2, which coincides with the threshold for the photoacoustic signal of cavitation in the solution (Fig. 14.9A). The power threshold suggests that the mechanism of triggered release is through perforation of lipid bilayers mediated by transient cavitation, that is, microbubble formation and collapse (Paliwal and Mitragotri, 2006; Pecha and Gompf, 2000; Tong et al., 2007). Several recent studies have also suggested that the effects of plasmon-resonant nanoparticles on cell membrane rupture are

302

Guohui Wu et al.

Intensity (a.u.)

3 ⫻ 104

14.9

m2

mJ/c

2 ⫻ 104

2 J/cm

m 13.0

1 ⫻ 104 7.1

2 cm

mJ/

0 2 m

J/c

m 1.3

500 400

300 ) 200 e (s m 100 Ti 0

Figure 14.13 Kinetics of in situ fluorescence shows the rate of liposomal content-release induced by encapsulated HGNs at various laser-energy densities. Time zero is the beginning of laser irradiation. Below a threshold energy density (2.2 mJ/cm2), there was no fluorescence increase. Above the threshold, there was a near instantaneous increase in the fluorescence intensity, followed by a more gradual increase. The increase of fluorescence increase can be fit by single exponential: F ¼ Fo þ A ex=t , with t ¼ 5, 52, and 112 s at laser-energy densities of 14.9, 13.0, and 7.1 mJ/cm2, respectively.

linked with cavitation dynamics and transient bubble formation (Lapotko et al., 2006; Pitsillides et al., 2003; Yao et al., 2005. While the permeability of DPPC liposomes increases near the gel to liquid crystalline transition temperature of 41
C (Ponce et al., 2006), this is not the source of the rapid increase in CF fluorescence observed here. The overall temperature increase of the solution due to the HGN absorption of NIR light is only a few degrees above ambient (Prevo et al., 2008). Even at 41
C, the permeability increase is only such that CF release would take minutes, not seconds as found here. The only way to get this rate of release is by the generation of large defects in the bilayer, as observed in liposomes after sonication (Zasadzinski, 1986). Only minor differences in liposome morphology are visible by cryo-EM after irradiation (Wu et al., 2008); the bilayers are less spherical, and hence under less tension after irradiation, consistent with a decrease in the positive osmotic pressure difference following CF release. These minor changes in the liposome shapes or sizes after irradiation suggest that cavitation induces transient defects in the bilayer, enabling drug release, after which membrane integrity is restored.

303

Synthesis, Characterization, and Optical Response

8. Effect of Proximity of HGNs to Liposomes Permeabilizing lipid membranes with microbubble cavitation should be induced by any HGNs in the solution, as long as there is an HGN within some maximum distance from a liposome bilayer. To test this hypothesis, DPPC liposomes encapsulating CF dye are mixed with increasing concentrations of HGNs, according to Scheme 14.2C. Upon pulsed laser irradiation, CF release is triggered. Figure 14.14 shows that the fractional CF release increases with HGN concentration up to an HGN concentration of 0.0315 mM (at higher HGN concentrations, the CF fluorescence is quenched by the HGNs (Dulkeith et al., 2002), data not shown). Hence, increasing the HGN concentration allows more liposomes to be within a critical distance of an irradiated HGN, which causes more liposomes to be ruptured, leading to greater CF release and an increase in the fluorescence. Minimizing and maintaining the distance between the HGN and the liposome bilayer lead us to tether HGNs to the outer surface of liposomes (Scheme 14.2B). Tethering the HGN directly to the outer surface of the liposomes increases the maximum release fraction to 96% (Table 14.1). The efficiency of phototriggered contents release is strongly affected by the proximity of HGN to the bilayer, consistent with the hypothesis that mechanical disruption by microbubbles is responsible for release (Tong et al., 2007). 60 50

Dye release (%)

40 30 20 10 0 −10

0.00

0.01 0.02 External HGN (mM)

0.03

Figure 14.14 Effect of unencapsulated HGNs on CF release from DPPC liposomes after NIR irradiation at 16.1 mJ/cm2. The solid line is a linear fit to the data.

304

Guohui Wu et al.

9. Conclusions Femtosecond pulses of NIR light absorbed by HGNs tethered to, encapsulated within, or in solution with liposomes trigger the near instantaneous release of liposome contents. The high temperatures reached by the HGN (Prasad et al., 2005; Prevo et al., 2008) induce production of unstable microbubbles, similar to the cavitation bubbles produced by ultrasound (Paliwal and Mitragotri, 2006). The mechanical and thermal effects of the microbubble collapse (Pecha and Gompf, 2000; Pitsillides et al., 2003) causes disruption of the liposome carriers, similar to the disruption caused by sonication. Neither the liposomes nor the CF appears to be altered chemically during this process, and the overall temperature rise of the bulk solution is only a few degrees. NIR light can penetrate up to 10 cm into tissue, which should allow these liposome/HGN complexes to be addressed noninvasively within a reasonable fraction of the human body. Any liposome carrier could be modified by tethering or encapsulating HGN to produce a system for rapid release on demand via NIR irradiation. This should eventually allow for better control of drug delivery to selected disease sites while minimizing systemic toxicity.

ACKNOWLEDGMENTS We thank Dr. Samir Mitragotri, Dr. Sumit Paliwal and Dr. Makoto Ogura for helpful discussions on cavitation and generously lending the hydrophone. This work was supported by the NIH Program of Excellence in Nanotechnology Grant HL080718: Nanotherapy for Vulnerable Plaques.

REFERENCES Abraham, S. A., Waterhouse, D. N., Mayer, L. D., Cullis, P. R., Madden, T. D., and Bally, M. B. (2005). The liposomal formulation of doxorubicin. Methods Enzymol. 391, 71–97. Allen, T. M., and Cullis, P. R. (2004). Drug delivery systems: Entering the mainstream. Science 303, 1818–1822. Alvarez, M. M., Khoury, J. T., Schaaff, T. G., Shafigullin, M. N., Vezmar, I., and Whetten, R. L. (1997). Optical absorption spectra of nanocrystal gold molecules. J. Phys. Chem. B 101, 3706–3712. Bohren, C. F., and Huffman, D. R. (1983). Absorption and Scattering of Light by Small Particles. Wiley, New York, 530 pp. Boyer, C., and Zasadzinski, J. A. (2007). Multiple lipid compartments slow vesicle contents release in lipases and serum. ACS Nano 1, 176–182. Chen, J., Wang, D., Xi, J., Au, L., Siekkinen, A., Warsen, A., Li, Z.-Y., Zhang, H., Xia, Y., and Li, X. (2007). Immuno gold nanocages with tailored optical properties for targeted photothermal destruction of cancer cells. Nano Lett. 7, 1318–1322.

Synthesis, Characterization, and Optical Response

305

Chiruvolu, S., Warriner, H. E., Naranjo, E., Idziak, S., Radler, J. O., Plano, R. J., Zasadzinski, J. A., and Safinya, C. R. (1994a). A phase of liposomes with entangled tubular vesicles. Science 266, 1222–1225. Chiruvolu, S., Walker, S., Leckband, D., Israelachvili, J., and Zasadzinski, J. (1994b). Higher order self-assembly of vesicles by ligand-receptor interactions. Science 264, 1753–1756. Chiu, D. T., Wilson, C. F., Ryttsen, F., Stromberg, A., Farre, C., Karlsson, A., Nordholm, S., Gaggar, A., Modi, B. P., Moscho, A., Garza-Lopez, R. A., Orwar, O., et al. (1999). Chemical transformations in individual ultrasmall biomimetic containers. Science 283, 1892–1895. Davidsen, J., Jorgensen, K., Andresen, T. L., and Mouritsen, O. G. (2003). Secreted phospholipase A(2) as a new enzymatic trigger mechanism for localised liposomal drug release and absorption in diseased tissue. Biochim. Biophy. Acta 1609, 95–101. Discher, B. M., Won, Y. Y., Ege, D. S., Lee, J.-M., Bates, F. S., and Hammer, D. (1999). Polymersomes: Tough vesicles made from diblock copolymers. Science 284, 1143–1146. Dulkeith, E., Morteani, A. C., Niedereichholz, T., Klar, T. A., Feldmann, J., Levi, S. A., van Veggel, F. C. J. M., Reinhoudt, D. N., Moller, M., and Gittins, D. I. (2002). Fluorescence quenching of dye molecules near gold nanoparticles: Radiative and nonradiative effects. Phys. Rev. Lett. 89, 203002/203001–203002/203004. Frederick, P. M., and Hubert, D. H. (2005). Cryoelectron microscopy of liposomes. Methods Enzymol. 391, 431–448. Grua, P., Morreeuw, J. P., Bercegol, H., Jonusauskas, G., and Vallee, F. (2003). Electron kinetics and emission for metal nanoparticles exposed to intense laser pulses. Phys. Rev. B Condens. Matter Mater. 68, 035424/03542.1–035424/0354.12. Hao, E., Li, S. Y., Bailey, R. C., Zou, S. L., Schatz, G. C., and Hupp, J. T. (2004). Optical properties of metal nanoshells. J. Phys. Chem. B 108, 1224–1229. Harris, N., Ford, M. J., and Cortie, M. B. (2006). Optimization of plasmonic heating by gold nanospheres and nanoshells. J. Phys. Chem. B 110, 10701–10707. Hirsch, L. R., Stafford, R. J., Bankson, J. A., Sershen, S. R., Rivera, B., Price, R. E., Hazle, J. D., Halas, N. J., and West, J. L. (2003). Nanoshell-mediated near-infrared thermal therapy of tumors under magnetic resonance guidance. Proc. Natl. Acad. Sci. USA 100, 13549–13554. Huang, S.-L. (2008). Liposomes in ultrasonic drug and gene delivery. Adv. Drug Deliv. Rev. 60, 1167–1176. Huang, S.-L., and MacDonald, R. C. (2004). Acoustically active liposomes for drug encapsulation and ultrasound-triggered release. Biochim. Biophy. Acta 1665, 134–141. Huang, X., El-Sayed, I. H., Qian, W., and El-Sayed, M. A. (2006). Cancer cell imaging and photothermal therapy in the near-infrared region by using gold nanorods. J. Am. Chem. Soc. 128, 2115–2120. Johnson, P. B., and Christy, R. W. (1972). Optical constants of the noble metals. Phys. Rev. B 6, 4370–4379. Jung, H. T., Lee, S. Y., Coldren, B., Zasadzinski, J. A., and Kaler, E. W. (2002). Gaussian curvature and the equilibrium between cylinders, discs and spheres. Proc. Natl. Acad. Sci. USA 99, 15318–15322. Kisak, E. T., Coldren, B., and Zasadzinski, J. A. (2002). Nanocompartments enclosing vesicles, colloids, and macromolecules via interdigitated lipid bilayers. Langmuir 18, 284–288. Kisak, E. T., Coldren, B., Evans, C. A., Boyer, C., and Zasadzinski, J. A. (2004). The vesosome—A multicompartment drug delivery vehicle. Curr. Med. Chem. 11, 199–219. Kotaidis, V., Dahmen, C., von Plessen, G., Springer, F., and Plech, A. (2006). Excitation of nanoscale vapor bubbles at the surface of gold nanoparticles in water. J. Chem. Phys. 124, 184702/184701–184702/184707.

306

Guohui Wu et al.

Kreibig, U. (1977). Anomalous frequency and temperature dependence of the optical absorption of small gold particles. J. Phys. (Paris) Colloq. 2, 97–103. Kreibig, U., and Genzel, L. (1985). Optical absorption of small metallic particles. Surf. Sci. 156, 678–700. Kreibig, U., and Vollmer, M. (1995). In ‘‘Optical Properties of Metal Clusters,’’ Springer Series in Materials Science, 532 pp. Lapotko, D. O., Lukianova, O. E., and Oraevsky, A. A. (2006). Selective laser nanothermolysis of human leukemia cells with microbubbles generated around clusters of gold nanoparticles. Lasers Surg. Med. 38, 631–642. Liang, H. P., Wan, L. J., Bai, C. L., and Jiang, L. (2005). Gold hollow nanospheres: Tunable surface plasmon resonance controlled by interior-cavity sizes. J. Phys. Chem. B 109, 7795–7800. Lin, C. P., and Kelly, M. W. (1998). Cavitation and acoustic emission around laser-heated microparticles. Appl. Phys. Lett. 72, 2800–2802. Link, S., Burda, C., Nikoobakht, B., and El-Sayed, M. A. (1999a). How long does it take to melt a gold nanorod? A femtosecond pump-probe absorption spectroscopic study. Chem. Phys. Lett. 315, 12–18. Link, S., Wang, Z. L., and El-Sayed, M. A. (1999b). Alloy formation of gold-silver nanoparticles and the dependence of the plasmon absorption on their composition. J. Phys. Chem. B 103, 3529–3533. Mueller, A., Bondurant, B., and O’Brien, D. F. (2000). Visible-light-stimulated destabilization of PEG-liposomes. Macromolecules 33, 4799–4804. Niidome, T., Yamagata, M., Okamoto, Y., Akiyama, Y., Takahashi, H., Kawano, T., Katayama, Y., and Niidome, Y. (2006). PEG-modified gold nanorods with a stealth character for in vivo applications. J. Control. Release 114, 343–347. Noble, C. O., Kirpotin, D. B., Hayes, M. E., Mamot, C., Hong, K., Park, J. W., Benz, C. C., Marks, J. D., and Drummond, D. C. (2004). Development of ligandtargeted liposomes for cancer therapy. Expert Opin. Ther. Targets 8, 335–353. Norman, R. S., Stone, J. W., Gole, A., Murphy, C. J., and Sabo-Attwood, T. L. (2008). Targeted photothermal lysis of the pathogenic bacteria, pseudomonas aeruginosa, with gold nanorods. Nano Lett. 8, 302–306. Oldenburg, S. J., Averitt, R. D., Westcott, S. L., and Halas, N. J. (1998). Nanoengineering of optical resonances. Chem. Phys. Lett. 288, 243–247. Paliwal, S., and Mitragotri, S. (2006). Ultrasound-induced cavitation: Applications in drug and gene delivery. Expert Opin. Drug Deliv. 3, 713–726. Pecha, R., and Gompf, B. (2000). Microimplosions: Cavitation collapse and shock wave emission on a nanosecond time scale. Phys. Rev. Lett. 84, 1328–1330. Peer, D., Karp, J. M., Hong, S., Farokhzad, O. C., Margalit, R., and Langer, R. (2007). Nanocarriers as an emerging platform for cancer therapy. Nat. Nanotechnol. 2, 751–760. Petrova, H., Perez Juste, J., Pastoriza-Santos, I., Hartland, G. V., Liz-Marzan, L. M., and Mulvaney, P. (2006). On the temperature stability of gold nanorods: Comparison between thermal and ultrafast laser-induced heating. Phys. Chem. Chem. Phys. 8, 814–821. Pitsillides, C. M., Joe, E. K., Wei, X., Anderson, R. R., and Lin, C. P. (2003). Selective cell targeting with light-absorbing microparticles and nanoparticles. Biophys. J. 84, 4023–4032. Ponce, A. M., Vujaskovic, Z., Yuan, F., Needham, D., and Dewhirst, M. W. (2006). Hyperthermia mediated liposomal drug delivery. Int. J. Hyperthermia 22, 205–213. Popinet, S., and Zaleski, S. (2002). Bubble collapse near a solid boundary: A numerical study of the influence of viscosity. J. Fluid Mech. 464, 137–163. Prasad, V., Mikhailovsky, A., and Zasadzinski, J. A. (2005). Inside-out disruption of silica/ gold core-shell nanoparticles by pulsed laser irradiation. Langmuir 21, 7528–7532.

Synthesis, Characterization, and Optical Response

307

Prevo, B. G., Esakoff, S. A., Mikhailovsky, A., and Zasadzinski, J. A. (2008). Scalable routes to gold nanoshells with tunable sizes and properties: Monitoring their response to NIR pulsed laser irradiation. Small 4, 1183–1195. Roper, D. K., Ahn, W., and Hoepfner, M. (2007). Microscale heat transfer transduced by surface plasmon resonant gold nanoparticles. J. Phys. Chem. C 111, 3636–3641. Schwartzberg, A. M., Oshiro, T. Y., Zhang, J. Z., Huser, T., and Talley, C. E. (2006). Improving nanoprobes using surface-enhanced Raman scattering from 30-nm hollow gold particles. Anal. Chem. 78, 4732–4736. Sengupta, S., Eavarone, D., Capila, I., Zhao, G., Watson, N., Kiziltepe, T., and Sasisekharan, R. (2005). Temporal targeting of tumour cells and neovasculature with a nanoscale delivery system. Nature 436, 568–572. Shi, W., Sahoo, Y., Swihart, M. T., and Prasad, P. N. (2005). Gold nanoshells on polystyrene cores for control of surface plasmon resonance. Langmuir 21, 1610–1617. Shum, P., Kim, J.-M., and Thompson, D. H. (2001). Phototriggering of liposomal drug delivery systems. Adv. Drug Deliv. Rev. 53, 273–284. Simo˜es, S., Fonseca, C., Pedroso de Lima, M., and Du¨zgu¨nes¸, N. (2006). pH responsive liposomes. State of the art from biophysics to therapy. In ‘‘Smart Nanoparticles in Nanomedicine,’’ (R. Arshady and K. Kono, eds.), Vol. 8, pp. 217–250. Kentus Books, London. Sou, K., Endo, T., Takeoka, S., and Tsuchida, E. (2000). Poly(ethylene glycol)-modification of the phospholipid vesicles by using the spontaneous incorporation of PEG into the vesicles. Bioconjug. Chem. 11, 372–379. Sun, Y. G., and Xia, Y. N. (2004). Mechanistic study on the replacement reaction between silver nanostructures and chloroauric acid in aqueous medium. J. Am. Chem. Soc. 126, 3892–3901. Sun, Y. G., Mayers, B. T., and Xia, Y. N. (2002). Template-engaged replacement reaction: A one-step approach to the large-scale synthesis of nanostructures with hollow interiors. Nano Lett. 2, 481–485. Sun, Y., Mayers, B., and Xia, Y. (2003). Metal nanostructures with hollow interiors. Adv. Mater. 15, 641–646. Tong, L., Zhao, Y., Huff, T. B., Hansen, M. N., Wei, A., and Cheng, J.-X. (2007). Gold nanorods mediate tumor cell death by compromising membrane integrity. Adv. Mater. 19, 3136–3141. Turkevich, J. (1985). A history of colloidal gold. Gold Bull. 18, 86–91. Turkevich, J., Stevenson, P. C., and Hillier, J. (1951). A study of the nucleation and growth processes in the synthesis of colloidal gold. Dis. Faraday Soc. 11, 55–65. Turkevich, J., Garton, G., and Stevenson, P. C. (1954). The color of colloidal gold. J. Colloid Sci. 9, S26–S35. Weissleder, R. (2001). A clearer vision for in vivo imaging. Nat. Biotechnol. 19, 316–317. Wiley, B., Herricks, T., Sun, Y. G., and Xia, Y. N. (2004). Polyol synthesis of silver nanoparticles: Use of chloride and oxygen to promote the formation of single crystal, truncated cubes and tetrahedrons. Nano Lett. 4, 1733–1739. Wiley, B., Sun, Y. G., Chen, J. Y., Cang, H., Li, Z. Y., Li, X. D., and Xia, Y. N. (2005). Shape controlled synthesis of silver and gold nanostructures. MRS Bull. 30, 356–361. Wu, G., Mikhailovsky, A., Khant, H. A., Fu, C., Chiu, W., and Zasadzinski, J. A. (2008). Remotely triggered liposome release by near-infrared light absorption via hollow gold nanoshells. J. Am. Chem. Soc. 130, 8175–8177. Yao, C. P., Rahmanzadeh, R., Endl, E., Zhang, Z., Gerdes, J., and Huttmann, G. (2005). Elevation of plasma membrane permeability by laser irradiation of selectively bound nanoparticles. J. Biomed. Opt. 10, 064012/064011–064012/064018. Zalipsky, S., Brandeis, E., Newman, M. S., and Woodle, M. C. (1994). Long circulating, cationic liposomes containing amino-peg-phosphatidylethanolamine. FEBS Lett. 353, 71–74. Zasadzinski, J. A. (1986). Transmission electron microscopy observations of sonicationinduced changes in liposome structure. Biophys. J. 49, 1119–1130.

This page intentionally left blank

C H A P T E R

F I F T E E N

Complex Nanotube-Liposome Networks Aldo Jesorka and Owe Orwar Contents 1. Introduction 2. Network Fabrication Protocols 3. Complexity and Topology 4. Internal and Membrane Functionalization 5. Transport Phenomena and Controlled Mixing Procedures 6. Enzymatic Reactions in NVN 7. Concluding Remarks Acknowledgments References

309 310 314 315 318 320 323 323 324

Abstract Surfactant nanotube-vesicle networks (NVN) belong to the smallest artificial devices known to date for performing controlled chemical operations with enzymes. Newly established means for transport of chemical reactants between containers, as well as advancements in initiation and control of chemical reactions in such systems have opened pathways to new devices with a resolution down to the single-molecule level. Here, we summarize the fabrication and functionalization of complex nanotube-liposome networks for such devices, and discuss related aspects of their application for studying chemical kinetics and materials transport phenomena in ultrasmall-scale biomimetic environments.

1. Introduction In biological cells, nature has achieved complex, truly nanoscale systems for computation and information processing, direct and catalyzed synthesis, sensing and other life sustaining biochemical and biophysical tasks. It is a challenging transdisciplinary endeavor to create biomimetic, Department of Chemical and Biological Engineering, Chalmers University of Technology, Go¨teborg, Sweden Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64015-5

#

2009 Published by Elsevier Inc.

309

310

Aldo Jesorka and Owe Orwar

fully or partly artificial devices that imitate some key features of biological systems, with the overall goal of providing new concepts for complex chemical operations on the nanoscale level. A key point of interest is, from a physicochemical perspective, to understand how chemical reactions proceed in a small-scale environment, and which kinetic and mechanistic principles govern these transformations. In particular, modeling and experimentally studying chemistry in ultrasmall devices containing only a few molecules is an emerging field of nanoscience and nanotechnology (Karlsson et al., 2003a; Levene et al., 2003; Xie et al., 2008), with growing importance in analytical chemistry and materials science. Methods for controlling the initiation of chemical reactions, mixing, and transport phenomena in biomimetic nanoscale compartments with designed functionalities are of considerable relevance for detailed understanding of fundamental (bio)chemical phenomena, such as enzymatic reactions, protein synthesis and transport, properties of signaling pathways, and others. An unconventional, yet fully biocompatible environment can be created by means of soft-matter fabrication methods that use phospholipid bilayer membranes. It comprises biomimetic containers as model reactors and lipotube-interconnected containers as complex network devices that have been demonstrated to be useful for investigations of chemical reactions, polymer dynamics, mass transport, and other phenomena at the micrometer and nanometer level. A set of well-defined procedures and methods has been generated from within this multidisciplinary approach, to be briefly introduced in the following five subsections.

2. Network Fabrication Protocols The concepts used for nanoscale network and device design are based on self-organized phospholipid membranes that drew inspiration from biological systems (Evans et al., 1996; Karlsson et al., 2001). A set of methods based on self-assembly of individual phospholipid molecules to membrane leaflets, self-organization, and forced shape transformations using precise micromanipulation tools can be utilized to form fully closed lipid bilayer shells and attached nanoconduits with certain structural and functional properties similar to biological micro- and nanocompartments (Karlsson et al., 2004; Lizana et al., 2009). These procedures are in fact unconventional bottom-up fabrication routes with extraordinary flexibility to yield three-dimensional soft-matter devices at a length scale that is difficult to reach with modern solid-state clean room technology. Figure 15.1 shows schematically the fabrication strategy. This fabrication concept involving forced shape transitions is interesting for creating specialized membrane-encapsulated volumes of

311

Complex Nanotube-Liposome Networks

A

B

+

C

−

Pi

Pi

VB

F = 1–10 pN

GUV

VA

MLV

ftube ~ 100 nm

fGUV ~ 10-100 mm D

Pi VC

E

F

VA VB

ΨB/A =

3rtube VB = VA 2rGUV(A)

Figure 15.1 Schematic representation of a fabrication strategy for complex nanofluidic networks composed of buffer-filled giant unilamellar phospholipid vesicles (GUV) and nanotubes. (A) Penetration of a GUV with a glass injection pipette after electroporation with a DC pulse. The injection pipette is backfilled with a solution that can be the same as or differ from the liquid content of the vesicle, for example, a buffered enzyme solution, here indicated by a different color. Attached to the GUV is an MLV, which serves as a lipid reservoir for subsequent network building. (B) By retracting the needle from the vesicle, using a micromanipulation device, a nanotube is drawn from the GUV. (C) Injecting the content of the injection pipette with positive backpressure Pi leads to formation of a second, the so-called daughter vesicle, at the end of the nanotubes. (D) The procedure can be repeated with differently backfilled injection pipettes, leading to a content-differentiated complex nanofluidic network. (E) For the creation of a new daughter vesicle, the ratio C of the two volumes involved is dependent only on the dimensions of the nanotubes connecting them and the size of the mother vesicle.

particular volume, connectivity and functionality. Several advanced micromanipulation protocols based on micropipette injection technology and electroporation have been developed, including vesicle fission, pipette writing, and vesicle inflation (Karlsson et al., 2004). In particular the vesicle inflation method, as depicted in Fig. 15.1, allows for the fabrication of fluid-state lipid nanotube-vesicle networks (NVN) of high geometrical complexity, where each node within a network can feature uniquely differentiated chemistry. The networks are composed of giant surface-immobilized phospholipid bilayer vesicles (typically 20 mm in diameter, 10–12 l, with a membrane thickness of  5 nm) interconnected by  100 nm wide lipid nanotubes. The construction and design of nanofluidic channels only a few times larger than individual biomacromolecules are thus possible. A 100-nm-radius tube has a cross-sectional area of 3.14  10 14 m2. In comparison, the size of a

312

Aldo Jesorka and Owe Orwar

common enzyme, alkaline phosphatase, is approximately 3.5 nm in diameter (Lenghaus et al., 2003), giving a cross-sectional area of 9.6  10 18 m2. The volume element of a 100-nm-long tube segment is 3.14  10 22 m3, and if such a volume is hosting a single molecule, then the concentration will be 53 nM. The total system volume of a vesicle nanotubes network is at least 6 orders of magnitude smaller than that of traditional microfluidic channels. The experimental setup for the construction of NVN is schematically depicted in Fig. 15.2. It can easily be adapted to any inverted and confocal microscope setup. The main components are a pair of water hydraulic micromanipulators (high graduation: Narishige MWH-3, coarse graduation: Narishige MC-35A). The electro-assisted injections are controlled by a microinjection pumping system (Eppendorf Femtojet) and a pulse generator (Digitimer Stimulator DS9A, 0–40 V, ms/ms/s pulse lengths, single pulse and pulse series generation). Occasionally, a manual injection pump is required, in particular when lipid material has to be removed gently from a unilamellar vesicle. This is conveniently achieved with an Eppendorf CellTram Vario piston pump. Carbon fiber microelectrodes measuring 30 mm in length and 5–10 mm in diameter (ProCFE, Axon Instruments) are employed as

Femtojet injection pump Waterhydraulic micromanipulator 1

V Waterhydraulic micromanipulator 2

Pa

DC puls generator +

Injection needle and eletroporation electrode

Carbon fiber microelectrode

GUVs Microscope cover slip Inverted microscope objective 40x

Figure 15.2 Experimental setup for the fabrication of nanotube-vesicle networks. The individual components are described in the text.

Complex Nanotube-Liposome Networks

313

counterelectrode, while a thin silver wire serves as main electrode in the backfilled capillaries. Injection tips are prepared from borosilicate capillaries (GC100TF-10, Clark Electromedical Instruments) that were carefully flame-forged in the back ends (Narishige MF 900 Microforge) to avoid that the sharp glass ends wear out the capillary holder. The capillaries are flushed with a stream of nitrogen gas to remove dust particles and tapered on a CO2-laser-puller (P-2000, Sutter Instrument) immediately before use. The outer-tip diameter of the capillaries produced by this method is typically 0.25–1 mm. Giant vesicles for network building are most conveniently prepared starting from a 5-ml droplet of 1 mg/ml L-a-phosphatidylcholine dispersed in phosphate-buffered saline (PBS) buffer (Trizma base 5 mM, K3PO4/ KH2PO4 30 mM, MgSO4 1 mM, EDTA 0.5 mM, adjusted with H3PO4 to pH 7.8). The droplet is pipetted onto a borosilicate coverslip #1, dehydrated in a vacuum desiccator under membrane pump vacuum for 15 min. The system contains 1% (v/v) glycerol as additive to avoid complete dehydration. After deposition and drying, the lipid is carefully rehydrated by covering it with 0.5 ml PBS buffer. Within a few minutes, several hundreds of cellsized unilamellar vesicles are formed. An aliquot (300 ml) is placed on the top face of a second coverslip, which is finally used in the experiment. The liposomes are immobilized on the coverslip surface by spontaneous adsorption. It is often advantageous to spincoat the negative expoxy photoresist SU-8 (Microchem) onto the cover glass prior to use, which after baking and short UV–ozone exposure forms a solid, moderately hydrophobic film on the glass surface. It prevents strong adhesion and disintegration of the surface-immobilized vesicles and extends the lifetime of the network. It has to be noted that UV exposure and cross-linking without the final oxidizing treatment renders the surface very hydrophobic, leading to lipid monolayer formation (Czolkos et al., 2007). A tapered and rear-end forged borosilicate glass micropipette is backfilled with aqueous medium, mounted onto the electroinjection system and micromanipulated onto the membrane of a surface-immobilized GUV, which itself is attached to a multilamellar lipid reservoir. With one or several anodic rectangular DC-voltage pulses of field strengths between 10 and 40 V/cm and duration of 1–4 ms applied over the micropipette, the liposome membrane can be penetrated (Fig. 15.1A). The micropipette is then pulled out and away from the liposome (about 5 mm/s), forming a lipid nanotube connection between the liposome and the pipette tip. When the nanotube has reached the desired length, aqueous medium is slowly pressure-injected (5  10 14 l/s) into the nanotubes. Drawing lipid material from the reservoir, a small vesicle is formed at the outlet of the micropipet tip. The size of the vesicle can be set by controlling injection time and volume, and the newly created structure is finally deposited on the

314

Aldo Jesorka and Owe Orwar

surface (Fig. 15.1B). Several vesicles can be fabricated successively, forming a nanotube-interconnected network (Fig. 15.1C). Already in their simplest configuration, these liposome systems can be used to study and control chemical reactions in a well-defined nanoscale environment, and their potential as constituents in advanced devices for nanofluidics and nanochemistry applications has been already partly exploited. In particular, enzymatic reactions, transport studies, and incorporation of membrane proteins will be discussed in a later section.

3. Complexity and Topology To form network structures of higher order topologies, vesicles within a network must be connected using membrane fusion induced by, for example, a focused electric field (Karlsson et al., 2002a; Stromberg et al., 2000). Circular networks as well as fully connected networks with threedimensional nanotube layers are examples of such systems (Fig. 15.3). To create a closed system, a simple network is constructed (Fig. 15.3A and B); then, a small nanotube-conjugated satellite vesicle is formed and positioned in close contact to a surface-immobilized vesicle container within the network (Fig. 15.3C). A localized electrical field is applied by placing a carbon fiber microelectrode adjacent to the fusion partners. Fusion of the vesicle containers is stimulated by application of one or several short B

A

E MLV

D

C

+

−

10 mm

Figure 15.3 Construction of complex nanotube-vesicle networks using an electroporation/membrane fusion protocol. A three-node network is constructed using the vesicle injection procedure (A, B). By means of a micropipette, a small fusion vesicle is generated and brought into contact with a surface-immobilized target vesicle. Application of a rectangular DC pulse fuses the vesicles (C, D), leading to a closed network of genus 1. (E) Differential interference contrast enhanced bright-field micrograph of an 8-GUV nanotubes-interconnected network. The highest connectivity is 6 (center vesicle), MLV denotes the multilamellar vesicle that serves as a membrane reservoir. Reprinted by permission of the American Chemical Society.

Complex Nanotube-Liposome Networks

315

rectangular DC-voltage pulses of field strengths between 40 and 80 V/cm and durations of 1–4 ms between the micropipette and the counterelectrode (Fig. 15.3D). Withdrawal of the micropipette can subsequently be performed without vesicle deformation, leakage, or disruption. Residual lipid material occasionally adhering to the pipette is removed by applying one or several cathodic DC-voltage pulses over the micropipette. The result is a closed, circular NVN (Fig. 15.3D). An extension of the procedure to very complex networks is possible, provided there is a sufficiently large membrane reservoir (MLV, multilamellar vesicle) attached (Fig. 15.3D, upper right corner). Employing the above-described micromanipulation technologies to move vesicles and produce localized electric fields and mechanical point loads, configurational transformations can be performed within and between lipid NVN. These manipulations include building and breaking connections within networks or between networks. Specifically, separate networks can be joined together using a combination of vesicle translocation and electrofusion (Karlsson et al., 2002a). Self-organization can also be utilized to create network subdomains of bifurcating lipid nanotubes with the surface-immobilized vesicles arranged at the periphery. Networks are in this instance produced from initial geometries with doubly nanotube-conjugated vesicles. Self-organization is triggered by mechanical or electromechanical action to merge two nanotubes. This is a process driven by spontaneous minimization of surface free energy of the phospholipid membrane. The dimensions of nanotubes, the network connectivity as well as the size, location, and contents of individual containers can be designed and manipulated with high precision (Lobovkina et al., 2005, 2008).

4. Internal and Membrane Functionalization Strong arguments for using lipid membranes as a material for constructing nanoscale soft-matter devices are their compatibility with biological components and their capability to host certain protein functionalities, such as ion channels, receptors, and enzymes (Criado and Keller, 1987; Eytan, 1982). Proteins can reside in the phospholipid membrane, having the fundamental bilayer structure of biological membranes. Not only water-soluble proteins and biomacromolecular assemblies, such as enzymes, but also synthetic polymers of a simpler structure can be confined to the interiors of vesicles or nanotubes. Several proteins can be extracted from cellular systems and subsequently be reconstituted into an artificial lipid host environment. Alternatively, biomacromolecules and lipids can be extracted directly from cultured

316

Aldo Jesorka and Owe Orwar

biological cells (Bauer et al., 2006). A high degree of compositional complexity can be achieved with both strategies. Davidson et al. (2003) demonstrated successfully the formation of NVN with reconstituted membrane protein from red blood cells (Fig. 15.4A and B). The network itself is prepared as described earlier. For the reconstitution, erythrocyte hosts with an eosin-5-maleimide-labeled anion exchanger AE1 are employed. The labeled material containing 1 mg/ml protein is solubilized for 1 h with 5 mM Triton X-100 at pH 7.8 in a solution consisting of 0.12 M KCl, 0.5 mM EDTA, 1 mM MgSO4, and 1% glycerol. After centrifugation (27,000g, 1 h), the supernatant is diluted two times with 10 mM Triton X-100 in the same buffer and incubated for 1 h. All steps are performed at 0  C. Finally, removal of the detergent with Biobeads SM-2 at room temperature and filtering through 200 nm membrane filters, a proteoliposome solution is obtained, which can be transformed into giant unilamellar vesicles, using the KPi-buffer dehydration/rehydration technique described earlier. The activity of the AE1 after reconstitution is verified by single-channel ion conductance measurements in excised inside-out patches from the vesicle membranes. The distribution of protein in the network membrane is established by means of fluorescence microscopy to be homogeneous. The labeled protein could diffuse via nanotube interconnections from vesicle to vesicle. A

C

B

D

Figure 15.4 Examples of internal and membrane functionalization of NVNs. (A, B) Fluorescence micrograph of a four-vesicle network with reconstituted membrane proteins (fluorescently labeled AE1 ion channels). The center vesicle is also schematically represented. The lower two vesicles in this panel are additionally filled with small unilamellar vesicles, containing the same labeled protein. (C) Fluorescence micrograph of a three vesicle network with different fluorescent dyes in each container. (D) Compartmentalization of a GUV by means of a hydrogel derived from a thermoresponsive LCST polymer hydrogel. The diameter of a GUV in each panel is  25 mm. Reprinted by permission of the American Chemical Society.

Complex Nanotube-Liposome Networks

317

Another recent example of incorporation of membrane proteins directly from cultured biological cells was demonstrated by Bauer et al. (2006). Here, a combination of dithiothreitol (DTT) and formaldehyde is employed to produce plasma membrane vesicles that are subsequently shaped into networks using the micropipette-assisted method for network fabrication. A single cell is fully sufficient to derive all material needed to build a small network. This protocol allows, similar to direct reconstitution of protein in vesicles, full control over the solution environment. Most importantly, the proteins remain in their proper orientation. However, the DTT/formaldehyde treatment may lead to loss of some protein function due to cross-linking; alternative protocols can be utilized to minimize this effect. Generally, the protein and lipid content in the networks is determined by employing different cell types, for instance by prior overexpression or the use of specialized cell types as sources for specific proteins. Aside from membrane modification and functionalization, the networks offer the unique possibility to differentiate the contents of individual vesicles. It represents a major benefit of the bottom-up production methodology, giving the ability to assemble complex chemical reaction systems (Karlsson et al., 2001). A simple example of a three-vesicle network, having each node modified with a different internalized fluorescent dye, is depicted in Fig. 15.4C. This is the foundation for the utilization of NVNs as enzymatic model reactors, or artificial cells. As already described, the exchange of injection pipettes at the time of network generation allows for fabrication of vesicles with individually defined internal composition and excellent control over the content composition and concentrations. In contrast, direct incorporation using ordinary liposome techniques, that is, enclosing the entire content upon formation of a liposome, usually suffers from low encapsulation efficiency ( Jesorka and Orwar, 2008). Several examples of this concept have been presented in the past, ranging from simple ionic solutions and colloidal particles for visualization, via small vesicles (Bolinger et al., 2004) and water-soluble polymers for compartmentalization ( Jesorka et al., 2005; Long et al., 2005), to biomacromolecules like DNA and alkaline phosphatase in order to investigate transport phenomena and enzymatic reactions on the size scale of a biological cell (Sott et al., 2006; Tokarz et al., 2005). Material is typically injected in different nodes within a network to yield an initially heterogeneous distribution of these species. Diffusional relaxation then occurs rapidly for materials small enough to traverse the nanotubes. Furthermore, well-defined internal structures can be formed in vesiclenanotube networks. An example is the controlled generation of hydrogels within the internal compartments (Markstrom et al., 2007) (Fig. 15.4D). Polymer solutions or finely dispersed suspensions can be microinjected into vesicles that subsequently undergo controlled sol–gel transitions. The thermoresponsive poly(N-isopropyl acrylamide), which undergoes a reversible

318

Aldo Jesorka and Owe Orwar

phase transition at 32  C in pure water and 27  C in 100 mM phosphate buffer solution, allows injection, mixing, and manipulation of clear solutions while enabling reversible solidification and compartmentalization as well as entrapping and release of colloidal material and larger biomacromolecules on demand. Vesicles and networks with hydrogel interiors of high density allow, for example, the study of cellular enzyme kinetics under simulated macromolecular crowding (cell-like) conditions (Helfrich et al., 2002; Johansson et al., 2000).

5. Transport Phenomena and Controlled Mixing Procedures In NVN, injected or locally generated materials can move through the network structure. Different, passive diffusional or active, tension or field-driven means of transport (Dommersnes et al., 2005; Karlsson et al., 2002b, 2005; Lizana and Konkoli, 2005; Tokarz et al., 2005) govern their mobility. Due to their minute size and high structural flexibility, NVNs allow for controlled transport of ultrasmall numbers of molecules. Size and optical properties of the networks allow generally for direct monitoring of transport processes by optical means, provided fluorescent or fluorescently labeled species are employed. Understanding and characterization of mixing and diffusion processes within such confined spaces have become the foundation of our investigations of intranetwork enzyme reaction kinetics. To initiate chemical (for example, enzymatic) reactions in the nodes of nanofluidic NVNs, control of fluid delivery is of paramount importance. The simplest and most direct method of transport and mixing of the contents of interconnected vesicles in a network is their integration by nanotube-mediated fusion, thereby delivering discrete quantities of material in a controlled fashion. A mobile, pipette-suspended vesicle is merged with and emptied into a stationary vesicle to achieve mixing, the final concentration being well defined by the two initially spherical volumes (Karlsson et al., 2003b). Another useful technique involves mixing of two individual solutions in a growing GUV conjugated by a suspended nanotube on one side and contacted by an injection pipette on the other (Fig. 15.5A). The far end of the nanotube is connected to a larger mother vesicle and a lipid membrane reservoir. The inflation of the daughter vesicle forces membrane material from the mother to traverse via the lipid nanotube to the daughter vesicle, producing a liquid flow inside the nanotube due to nonslip conditions and viscous coupling. During inflation, the different liquid contents of both pipette and nanotube will mix. At the limit of diffusion, the mixing ratio is only

319

Complex Nanotube-Liposome Networks

+p

A

rGUV

Solvent B

Solvent A

2rtube

3 mm 5 %A 15 mm 1 %A 30 mm 0.5 %A

Membrane reservoir

B

C D1

D2

MV

2.56 1.92 1.28 0.64 0 (s)

D 10 mm

Figure 15.5 Mixing and diffusional transport in NVNs. (A) Schematic drawing of mixing driven by membrane flow. Solvent A (blue arrows) from the left-hand vesicle is transported by viscous drag through the interconnecting nanotube as the right-hand vesicle grows during injection of solvent B (green arrows). The red figures represent the increasing diameter of the inflated vesicle, the blue figures are the corresponding percentages of solvent A in that vesicle. Membrane material is supplied from a membrane reservoir (red arrows). (B) Fluorescence micrograph of a three vesicle system constructed by micropipette injection according to panel (A). Mixing in a chain of consecutively microinjected vesicles leads to a dilution series that is represented by the decreasing fluorescence of the fluorescein solution originating from the mother vesicle (MV), as it is increasingly diluted with the injected buffer solution in daughter vesicles D1 and D2. (C) Schematic representation of two vesicle network interconnected by a nanotube. The red enclosure is the area of the nanotube where diffusive transport of a 30 nm single latex particle has been followed over a period of 2.56 s. (D) Time-dependent positional information of the particle in the enclosed area from panel (C). The color represents the point in time the particle was detected in the particular position. Reprinted by permission of the American Chemical Society.

dependent on the diameters of the inflated vesicle and the nanotube, but not on inflation rate (see also Fig. 15.1E). For example, in a system of several sequentially generated nanotube-conjugated vesicles, dilution series can be obtained with mixing ratios that cover almost 3 orders of magnitude between the first and last container (Fig. 15.5B). Thus, vesicles can be prepared with a well-defined volume of the two solutions at a pre-calculated mixing ratio. Three fundamental mechanisms of molecular transport within lipid nanotubes have been investigated to date. The first, Marangoni transport, is based on membrane tension gradients and the dynamic and fluid character

320

Aldo Jesorka and Owe Orwar

of the bilayer membrane (Dommersnes et al., 2005). The second mechanism is electrophoresis (Tokarz et al., 2005), a well-established way of distributing fluid and solutes in micro- and nanofluidic devices. The third and in practical terms most important mode is diffusion, a very effective means of transport over short distances (Davidson et al., 2005; Lizana and Konkoli, 2005). NVNs have several unique properties with respect to diffusion as an effective means of transport. They are sufficiently small, the chemical potential in the networks can be controlled over time by injection of different concentrations of relevant species into individual containers and, finally, the geometry of the networks can be changed over time. Reactants such as enzymes and substrates can thus be transported in a controlled manner from one end to the other and sequentially catalyze reactions in different nodes, as has been shown for diffusive-directed transport of alkaline phosphatase (Sott et al., 2006). Not only small molecules but also nanoscale objects such as submicron particles can be efficiently distributed by diffusion within NVNs. Single nanoparticles exhibited stochastic motion inside a nanotube. This movement is mainly Brownian, but contributions from random membrane tension differences are conceivable. The stochastic motion of a 30-nm-diameter particle inside a 100-nm-radius nanotube is largely one dimensional (Fig. 15.5C and D). The figure shows five different time point observations of how a particle spontaneously occupies a 6-mm segment of a lipid nanotube in a random fashion during a 2.56-s observation period (Karlsson et al., 2002b). The theoretical diffusion coefficient (Brenner and Gaydos, 1977) for a 30-nm-diameter particle at room temperature in a tube of 100 nm radius is 9.05 mm2/s, and the corresponding root-mean-square displacement of 6.81 mm during a 2.56-s time interval is in good agreement with the experiment.

6. Enzymatic Reactions in NVN In chemical reaction systems with a characteristic length scale of micro- and nanometers, which rapidly gain importance in modern technological soft-matter applications, the understanding of kinetics of chemical reactions in highly confined spaces is of increasing interest. As discussed earlier, at this small-scale, diffusion is the predominant mode of transport and mixing, thus the interplay between reactions and diffusion-dominated transport is of special relevance. Figure 15.6A and B shows as an example the progression of fluorescein diffusion in a network constructed of three vesicles. Central to the discussion about the differences between small- and large-scale reactions is the question about which transport modes dominate

321

Complex Nanotube-Liposome Networks

C

A

D MLV

V1

V2

V3

t = 0 min B V1

V2

V3 V1

V2

V1 + 2 10 mm

t = 10 min 2

3

1

4

t = 10 min

F 2

3

4

Normalized intensity

G

E

−6 −4 −2 0

1

t = 90 min

Vesicle 1 Vesicle 2 Vesicle 3 Vesicle 4

−8

0

20 40 60 80 100 120 Time (min)

Figure 15.6 Transport and enzymatic reactions in NVN. In the fluorescence micrographs, the boundary of the vesicles and the connecting nanotubes are highlighted with dashed lines. V1–V4 denotes the volumes. Fluorescence images are digitally edited to improve image quality. (A, B) Diffusion of fluorescein through an NVN at injection time (A) and after 10 min (B), followed by laser-induced fluorescence microscopy (LIF—fluorescein emission upon excitation at 488 nm). (C, D) Contrast-enhanced bright-field micrograph of the nanotube-mediated vesicle fusion technique employed to start a chemical reaction in a network. Vesicle 1, suspended by a pipette is translated toward vesicle 2, where content mixing and initiation of the reaction occurs. A multilamellar vesicle is attached to the original mother vesicle in order to provide lipid material during network construction. (E, F) LIF-images (fluorescein emission) of product formation in a four-vesicle network with linear geometry. Initially, vesicle 1 is injected with enzyme, while vesicles 2–4 are filled with FDP. Scale bar represents 10 mm. (G) Plots of normalized fluorescence intensity versus time (solid lines), representing the rate of product formation in vesicles 1–4 in the NVN shown panels (E) and (F). Dash-dotted lines show the theoretically modeled time dependencies of product formation. Reprinted by permission of the American Chemical Society.

in, and which reactions benefit from, structured spaces. There are a number of reactions that run faster in a network-like geometry. The interplay between geometry and the nature of a reaction scheme becomes important only when the individual reaction steps start to influence each other. Such reactions contain antagonistic catalytic influences in the intermediate stages of a multistep reaction scheme. This approach is, for example, an alternative way to explain certain aspects of cytoarchitecture (Konkoli, 2005).

322

Aldo Jesorka and Owe Orwar

Due to their versatility, flexibility, and favorable materials properties, vesicle-nanotube networks present a potent system to investigate chemical reactions in confined and structured space and under macromolecular crowding conditions. Reaction conditions can include interactions with biomacromolecules and cell components in their native environment, while the fluid character of the phospholipid membrane offers practical advantages, such as facile interfacing to injection equipment and active surfaces. We have investigated enzymatic reactions in unstructured and complex networks and utilized the dynamic features of the fluid membrane for reaction control. A simple technique to initiate enzymatic reactions inside NVN is nanotube-mediated merging of two interconnected vesicles (Fig. 15.6C and D). Here, vesicles are filled consecutively with the reaction partners (substrate and enzyme, respectively) by microinjection. Initially, the containers are spatially separated; diffusive mixing is insignificant due to the long distance and the narrow conduit. To initiate a reaction, the vesicles are brought in contact with each other, shortening the nanotube, and at a critical distance, the vesicles merge to form one spherical reactor with mixed contents. The concept is demonstrated on the stepwise enzymatic dephosphorylation of fluorescein diphosphate (FDP), which yields fluorescein monophosphate and finally fluorescein as products. Typical starting conditions are 10 mM FDP in a buffered solution consisting of 10 mM Trizma base and 100 mM KCl adjusted to pH 8.9; and 11 units/ml of alkaline phosphatase in a buffer of 5 mM Trizma base, 1 mM MgCl2, 0.1 mM ZnCl2, and 100 mM KCl at pH 8.9. Product formation is followed over time by fluorescence recovery after photobleaching (FRAP) microscopy. The enzyme concentration inside the vesicle after merging, here approximately 15 enzyme molecules, can be determined through comparison with bulk measurements. To gain theoretical understanding of the system, a simple enzyme substrate reaction model is developed and solved using a survival probability approach with the assumption of infinite enzyme substrate reaction rate. To further explore the potential of complex NVNs in a biomimetic context, the alkaline phosphatase/FDP reaction model was extended. Transition from a compact geometry (a single spherical container) to a structured geometry (several spherical containers connected by nanotubes) in NVNs induces the rate of the dephosphorylation reaction to display wave-like properties. By tuning the geometry of the network, the reaction dynamics can be directly influenced (Lizana et al., 2008). If, in this system, an enzyme is introduced into a terminal node of the network, it diffuses into neighboring nodes through the interconnecting nanotubes (Fig. 15.6E and F). The directionality of enzyme diffusion can be controlled exactly through establishing a concentration gradient. All the nodes except the enzymecontaining node bear FDP, which becomes fluorescent when converted to fluorescein and can thus be monitored by fluorescence microscopy.

Complex Nanotube-Liposome Networks

323

The temporal pattern of front propagation as well as the rate of reaction is dependent on geometry of the network. Figure 15.6G shows as an example the product formation dynamics in an unbranched linear NVN as wave-like propagation phenomenon (Sott et al., 2006). More extensive theoretical studies on both diffusion and reaction phenomena in similar networks have been performed, showing fascinating and nonintuitive behavior on front propagation and reaction optimization (Konkoli, 2007; Lizana and Konkoli, 2005).

7. Concluding Remarks Networks composed of phospholipid nanotubes and giant unilamellar vesicles are versatile chemical microreactors with nanofluidic interconnections. Unconventional yet comparatively facile fabrication strategies have been developed, employing phospholipids and biomacromolecules to construct enclosed membrane devices. The fabrication methodology involves self-assembly, forced shape transformations, and micromanipulation protocols such as electroinjection. The devices combine complex structure, flexibility and biocompatibility at a length scale that is difficult to reach even with the most advanced solid-state microtechniques. The networks are stable, flexible, and suitable for studies of chemical reactions in ultrasmall volumes, especially in a biologically relevant microenvironment. Reactants can be directly introduced and reliably confined in individual nodes, and strategies for microcompartmentalization and membrane modification, for example with ion channel proteins, exist. Due to the small spatial dimensions and short path lengths for molecules to travel, diffusion is the predominant, but not the only possible, material transport mode, and the impermeability of the phospholipid membrane confines all ionic and many nonpolar reactants to the network interior. Active and passive transport of small molecules, submicron particles, and biopolymers through nanotubes enables defined initiation and controlled progression of chemical reactions, while the fluid membrane boundary offers support through its structural dynamics. The uniquely flexible supramolecular architecture with its biomimetic foundation has bearing for understanding enzymatic reactions in biological systems.

ACKNOWLEDGMENTS The work was supported by the Royal Swedish Academy of Sciences, the Swedish Research Council (VR), and the Swedish Foundation for Strategic Research (SSF).

324

Aldo Jesorka and Owe Orwar

REFERENCES Bauer, B., Davidson, M., and Orwar, O. (2006). Direct reconstitution of plasma membrane lipids and proteins in nanotube-vesicle networks. Langmuir 22, 9329–9332. Bolinger, P. Y., Stamou, D., and Vogel, H. (2004). Integrated nanoreactor systems: Triggering the release and mixing of compounds inside single vesicles. J. Am. Chem. Soc. 126, 8594–8595. Brenner, H., and Gaydos, L. J. (1977). Constrained Brownian-movement of spherical particles in cylindrical pores of comparable radius—Models of diffusive and convective transport of solute molecules in membranes and porous media. J. Colloid Interface Sci. 58, 312–356. Criado, M., and Keller, B. U. (1987). A membrane fusion strategy for single-channel recordings of membranes usually non-accessible to patch-clamp pipette electrodes. FEBS Lett. 224, 172–176. Czolkos, I., Erkan, Y., Dommersnes, P., Jesorka, A., and Orwar, O. (2007). Controlled formation and mixing of two-dimensional fluids. Nano Lett. 7, 1980–1984. Davidson, M., Karlsson, M., Sinclair, J., Sott, K., and Orwar, O. (2003). Nanotube-vesicle networks with functionalized membranes and interiors. J. Am. Chem. Soc. 125, 374–378. Davidson, M., Dommersnes, P., Markstrom, M., Joanny, J. F., Karlsson, M., and Orwar, O. (2005). Fluid mixing in growing microscale vesicles conjugated by surfactant nanotubes. J. Am. Chem. Soc. 127, 1251–1257. Dommersnes, P. G., Orwar, O., Brochard-Wyart, F., and Joanny, J. F. (2005). Marangoni transport in lipid nanotubes. Europhys. Lett. 70, 271–277. Evans, E., Bowman, H., Leung, A., Needham, D., and Tirrell, D. (1996). Biomembrane templates for nanoscale conduits and networks. Science 273, 933–935. Eytan, G. D. (1982). Use of liposomes for reconstitution of biological functions. Biochim. Biophys. Acta 694, 185–202. Helfrich, M. R., Mangeney-Slavin, L. K., Long, M. S., Djoko, Y., and Keating, C. D. (2002). Aqueous phase separation in giant vesicles. J. Am. Chem. Soc. 124, 13374–13375. Jesorka, A., and Orwar, O. (2008). Liposomes: Technologies and analytical applications. Ann. Rev. Anal. Chem. 1, 801–832. Jesorka, A., Markstrom, M., Karlsson, M., and Orwar, O. (2005). Controlled hydrogel formation in the internal compartment of giant unilamellar vesicles. J. Phys. Chem. B 109, 14759–14763. Johansson, H. O., Brooks, D. E., and Haynes, C. A. (2000). International Review of Cytology—A Survey of Cell Biology 192, pp. 155–170. Academic Press Inc., San Diego, CA. Karlsson, A., Karlsson, R., Karlsson, M., Cans, A. S., Stromberg, A., Ryttsen, F., and Orwar, O. (2001). Molecular engineering—Networks of nanotubes and containers. Nature 409, 150–152. Karlsson, M., Sott, K., Davidson, M., Cans, A. S., Linderholm, P., Chiu, D., and Orwar, O. (2002a). Formation of geometrically complex lipid nanotube-vesicle networks of higherorder topologies. Proc. Natl. Acad. Sci. USA 99, 11573–11578. Karlsson, R., Karlsson, M., Karlsson, A., Cans, A. S., Bergenholtz, J., Akerman, B., Ewing, A. G., Voinova, M., and Orwar, O. (2002b). Moving-wall-driven flows in nanofluidic systems. Langmuir 18, 4186–4190. Karlsson, A., Karlsson, M., Karlsson, R., Sott, K., Lundqvist, A., Tokarz, M., and Orwar, O. (2003a). Nanofluidic networks based on surfactant membrane technology. Anal. Chem. 75, 2529–2537. Karlsson, R., Karlsson, A., and Orwar, O. (2003b). A nanofluidic switching device. J. Am. Chem. Soc. 125, 8442–8443.

Complex Nanotube-Liposome Networks

325

Karlsson, M., Davidson, M., Karlsson, R., Karlsson, A., Bergenholtz, J., Konkoli, Z., Jesorka, A., Lobovkina, T., Hurtig, J., Voinova, M., and Orwar, O. (2004). Biomimetic nanoscale reactors and networks. Annu. Rev. Phys. Chem. 55, 613–649. Karlsson, A., Sott, K., Markstrom, M., Davidson, M., Konkoli, Z., and Orwar, O. (2005). Controlled initiation of enzymatic reactions in micrometer-sized biomimetic compartments. J. Phys. Chem. B 109, 1609–1617. Konkoli, Z. (2005). Interplay between chemical reactions and transport in structured spaces. Phys. Rev. E 72, 011917. Konkoli, Z. (2007). Diffusion-controlled reactions in small and structured spaces as a tool for describing living cell biochemistry. J. Phys. Condens. Matter 19, 065149. Lenghaus, K., Dale, J. W., Henderson, J. C., Henry, D. C., Loghin, E. R., and Hickman, J. J. (2003). Enzymes as ultrasensitive probes for protein adsorption in flow systems. Langmuir 19, 5971–5974. Levene, M. J., Korlach, J., Turner, S. W., Foquet, M., Craighead, H. G., and Webb, W. W. (2003). Zero-mode waveguides for single-molecule analysis at high concentrations. Science 299, 682–686. Lizana, L., and Konkoli, Z. (2005). Diffusive transport in networks built of containers and tubes. Phys. Rev. E 72, 026305. Lizana, L., Bauer, B., and Orwart, O. (2008). Controlling the rates of biochemical reactions and signaling networks by shape and volume changes. Proc. Natl. Acad. Sci. USA 105, 4099–4104. Lizana, L., Konkoli, Z., Bauer, B., Jesorka, A., and Orwar, O. (2009). Controlling chemistry by geometry in nanoscale systems. Annu. Rev. Phys. Chem. 60, 449–468. Lobovkina, T., Dommersnes, P. G., Joanny, J. F., Bassereau, P., Karlsson, M., and Orwar, O. (2005). Pattern formation of different geometries and topologies as well as knotted structures in lipid nanotube networks. Biophys. J. 88, 208a. Lobovkina, T., Dommersnes, P. G., Tiourine, S., Joanny, J. F., and Orwar, O. (2008). Shape optimization in lipid nanotube networks. Eur. Phys. J. E 26, 295–300. Long, M. S., Jones, C. D., Helfrich, M. R., Mangeney-Slavin, L. K., and Keating, C. D. (2005). Dynamic microcompartmentation in synthetic cells. Proc. Natl. Acad. Sci. USA 102, 5920–5925. Markstrom, M., Gunnarsson, A., Orwar, O., and Jesorka, A. (2007). Dynamic microcompartmentalization of giant unilamellar vesicles by sol gel transition and temperature induced shrinking/swelling of poly(N-isopropyl acrylamide). Soft Matter 3, 587–595. Sott, K., Lobovkina, T., Lizana, L., Tokarz, M., Bauer, B., Konkoli, Z., and Orwar, O. (2006). Controlling enzymatic reactions by geometry in a biomimetic nanoscale network. Nano Lett. 6, 209–214. Stromberg, A., Ryttsen, F., Chiu, D. T., Davidson, M., Eriksson, P. S., Wilson, C. F., Orwar, O., and Zare, R. N. (2000). Manipulating the genetic identity and biochemical surface properties of individual cells with electric-field-induced fusion. Proc. Natl. Acad. Sci. USA 97, 7–11. Tokarz, M., Akerman, B., Olofsson, J., Joanny, J. F., Dommersnes, P., and Orwar, O. (2005). Single-file electrophoretic transport and counting of individual DNA molecules in surfactant nanotubes. Proc. Natl. Acad. Sci. USA 102, 9127–9132. Xie, X. S., Choi, P. J., Li, G.-W., Lee, N. K., and Lia, G. (2008). Single-molecule approach to molecular biology in living bacterial cells. Annu. Rev. Biophys. 37, 417–444.

This page intentionally left blank

C H A P T E R

S I X T E E N

Bionanotubules Formed from Liposomes Josemar A. Castillo and Mark A. Hayes Contents 328

1. Introduction 2. Bionanotubule Formation by Applying Electric Fields to Surface-Attached Liposomes 2.1. Materials and methods 3. Bionanotubule Formation from Liposomes in Solution Using Electric Fields 3.1. Materials and methods 4. Other Methods of Bionanotubule Formation from Liposomes 5. Concluding Remarks References

329 331 334 336 337 339 339

Abstract Bionanotubules are lipid-bound cylindrical structures with typical diameters in the tens of nanometers and length than can span up to hundreds of micrometers. Besides being observed in nature, bionanotubules can be prepared synthetically by various methods, some of which involve the extension of these structures from lipid vesicles. We describe the formation of lipid nanotubules from liposomes prepared with various lipid mixtures including phosphatidylcholine, phosphatidic acid, and various fluorescent phospholipids. We depict the methods used to extend bionanotubules from surface-attached vesicles, using electric fields as the driving force for bilayer extension and tubular growth. These methods include liposome preparation, surface attachment, and tubular extension by applying modest electric fields (<30 V/cm). Methods in which lipid tubules are extended from liposomes that are free in solution and subject to higher magnitude fields are also described. In addition, we summarize other protocols of bionanotubule formation from liposomes, including various modes of micromechanical manipulation of lipid vesicles. Department of Chemistry and Biochemistry, Arizona State University, Tempe, Arizona, USA Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64016-7

#

2009 Elsevier Inc. All rights reserved.

327

328

Josemar A. Castillo and Mark A. Hayes

1. Introduction Liposomes are self-assembled, nanometer to micrometer scale structures comprising one or more lipid bilayers, each of which contains an aqueous compartment. Besides being simple models for cells, their ability to carry both hydrophobic and hydrophilic components makes them ideal for therapeutic treatment, biophysical studies, food, cosmetics, constrained volume experiments, and gene transfer (Allen and Martin, 2004; de Kruijff et al., 1980; Dickson, 1993; Du¨zgu¨nes¸, 1998; Du¨zgu¨nes¸ and Nir, 1995; Ewert et al., 2005; Fenske and Cullis, 2005; Lasic, 1996; Rongen et al., 1997; Simo˜es et al., 2005; Steinberg-Yfrach et al., 1997). In addition to liposomes, three-dimensional lipid structures such as lipid nanotubules are also extremely versatile. This type of self-organizing system consists of bilayer-bound hollow cylinders with typical radial dimensions in the submicrometer range and length that can span to hundreds of microns (Castillo et al., 2009; Karlsson et al., 2004). Bionanotubules found in nature have been identified as ‘‘tunnels’’ that allow long-range communication and transport between cells (Rustom et al., 2004; Watkins and Salter, 2005). However, the natural occurrence and function of these structures are still being discovered (Brazhnik et al., 2005; Karlsson et al., 2006; Mahajan and Fang, 2005; Rustom et al., 2004; Watkins and Salter, 2005; Zimmermann et al., 2001). In addition to their use to further understand their role in biological environments, the tubules’ cylindrical geometry, low internal volumes, and bilayer-bound walls makes them ideal for many applications in nanotechnology that mimic their role in nature (Karlsson et al., 2006). Therefore, methods to fabricate synthetic nanotubules have gained popularity for both the fundamental and technological contexts. The physics of lipid bilayer-bound tubules formed via mechanical motion or biological action have been examined, and they are considered stable structures constituting local minima within available configurations of lipid bilayers (Dere´nyi et al., 2002; Karlsson et al., 2004; Rosenblatt et al., 1987). Rational control of tubular structures has been achieved by adjusting chemical composition and conditions of growth ( John et al., 2001, 2002; Jung et al., 2002; Mishra and Thomas, 2002; Singh et al., 2003; Spector et al., 1998, 2001; Svenson and Messersmith, 1999; Thomas et al., 1995, 1999; Yang et al., 2004). Many techniques have been developed to create lipid nanotubules, among these techniques there are lipid self-assembly in aqueous dispersion (Frusawa et al., 2003; Lasic, 1988; Papahadjopoulos et al., 1975), high-pressure shearing (Kulkarni et al., 2001), and the use of microfluidic devices to extend tubules from lipid films (Brazhnik et al., 2005; Mahajan and Fang, 2005; West et al., 2008a,b). Even though these techniques are useful for many applications, protocols in which tubules are

Bionanotubules Formed from Liposomes

329

extended from mother liposomes offer additional advantages. From tailoring the position of the liposomes, functionalizing the vesicle bodies, or simply studying the spontaneous formation of tubules from vesicles as a result of external forces researchers have been able to perform transport operations of small molecules, design highly complex nanoreactors, and obtain evidence of possible formation mechanisms in biological environments (Akiyoshi et al., 2003; Castillo et al., 2009; Karlsson et al., 2004, 2006; Sott et al., 2003). In general, mechanical approaches to extend lipid tubules from mother vesicle involve the application of a point load via relative translation of adhered microstructures or micropipets away from a liposome (Heinrich et al., 1999; Karlsson et al., 2004; West et al., 2008a). Using these techniques, tubules have been created from many different phospholipids such as phosphatidylcholines (PC) (Brazhnik et al., 2005; Karlsson et al., 2001; Sott et al., 2003) and sphingolipids (Kulkarni et al., 2001), as well as from liposomes embedding transmembrane proteins (Karlsson et al., 2004; Sott et al., 2003). Our laboratory has been using liposomes to study their behavior under the influence of electric fields (Hayes et al., 2007; Phayre et al., 2002; Pysher and Hayes, 2004, 2005). An interesting outcome of these studies has been the spontaneous formation of long-range complex intertwined structures including lipid nanotubules from free colloids and surface-attached vesicles (Castillo et al., 2009; Hayes et al., 2007). Other studies where electric fields are applied are generally limited to cellular systems and inducing membrane porosity and rupture ( Jones et al., 1996; Nolkrantz et al., 2001). In this chapter, we describe the methods used to investigate the formation of bionanotubules from liposomes by using electric fields, with emphasis on the use of immobilized vesicles. We also summarize other techniques used to create lipid tubules from liposomes such as various modes of mechanical manipulations.

2. Bionanotubule Formation by Applying Electric Fields to Surface-Attached Liposomes Lipid nanotubules can be extended from vesicles by applying modest (1–20 V/cm) electric field on micron-size liposomes electrostatically attached to a glass substrate (Fig. 16.1). Tubular formation, alignment, and stability vary with different liposome composition. In general, the vesicles used are 9:1 (by mass) zwitterionic-to-charged lipid ratio. However, this base ratio can be altered by changing the amount of charged lipid present,

330

A

B

Josemar A. Castillo and Mark A. Hayes

−

E.F

+

C

Figure 16.1 (A) General schematic of the formation of bionanotubules from surfaceattached liposomes using electric fields. (B) Fluorescence micrograph of 89:10:1 PC: PA:OG-DHPE liposomes before an electric field is applied. (C) Lipid nanotubules observed at an applied field of 10 V/cm. Tubules are aligned with the electric fields (bar ¼ 50 mm) (Castillo et al., 2009).

which appears to lower the magnitude of the applied potential needed to extend the tubules from the vesicles. Liposomes can also contain cholesterol to increase rigidity in the bilayer membrane, where data suggest that cholesterol amounts greater than 30% hinder tubular formation by this technique (Castillo et al., 2009). Generally, the observed behaviors of tubular growth demonstrate that the orientation and stability of the lipid nanotubules is dependent on the magnitude of the applied field. At higher voltages, tubular formation occurs in the direction of the applied field. Conversely, at lower electric fields the orientation of tubular growth is more disperse and appears random. Tubules that grow at these low-magnitude fields appear flexible and respond to local convective flow. As the magnitude of the applied field increases, the tubules appear less flexible and more stable (stability in terms of preserved geometric integrity). Images obtained at higher fields (generally higher than 8 V/cm) show lipid nanotubules mainly aligned parallel to each other with lengths of hundreds of micrometers at maximum extension (Castillo et al., 2009). Even though further investigation is necessary to understand the mechanism of vesicle formation through this technique, studies suggest that a possible mechanism is the formation of a charged lipid domain in the surface

Bionanotubules Formed from Liposomes

331

of the liposome due to lipid migration as a consequence of the applied electric field (Castillo et al., 2009; Hayes et al., 2007; Pysher and Hayes, 2005). This leading domain is thought to initiate the extension of the tubules from the mother liposome surface (Castillo et al., 2009).

2.1. Materials and methods 2.1.1. Preparation of liposomes In our laboratory, liposome preparation is accomplished by reverse-phase evaporation (Szoka and Papahadjopoulos, 1978), as described in this section. However, other techniques of liposome preparation can be employed. For tubular formation from surface-attached vesicles, giant uni- or multilamellar vesicles (diameters >10 mm) are preferred. Also, using fluorescently labeled phospholipids (charged or zwitterionic) as components of the liposomes is required to monitor the charge and the phospholipid distribution during nanotubular formation by fluorescence microscopy. 2.1.1.1. Buffer preparation Phosphate buffer is made from prepared 10 mM solutions of sodium phosphate dibasic and sodium phosphate monobasic. A specific volume of sodium phosphate monobasic is titrated with enough sodium phosphate dibasic until the desired pH of 7.4 is obtained. Sodium phosphate dibasic anhydrous (>95%) and sodium dihydrogen phosphate (99%) are obtained from Sigma-Aldrich (St. Louis, MO). 2.1.1.2. NBD-labeled liposomes Liposomes composed of PC, phosphatidic acid (PA), and N-4-nitrobenz-2-oxa-1,3-diazole phosphatidic acid (NBD-PA) (90:9:1 by mass) are also prepared by reverse-phase evaporation (Szoka and Papahadjopoulos, 1978). All phospholipids are obtained from Avanti Polar Lipids (Alabaster, AL). Lipids (10–20 mg) dissolved in chloroform (CHCl3) solutions are added to a round-bottom flask. While rotating the flask, the CHCl3 is rapidly evaporated off using a gentle stream of nitrogen gas, leaving a thin, uniform solid gel coating of lipids on the interior of the flask. Remaining CHCl3 is removed by vacuum. The dry lipids are then prehydrated with a few microliters of nanopure 18 MO water and the round-bottom flask is placed in a rotary evaporator with a water temperature of 39  C. Phosphate buffer (3.00 ml, pH ¼ 7.4) is added to the flask, which is then allowed to rotate at 39  C for 2 h. This method produces giant uni- and multilamellar vesicles. 2.1.1.3. Oregon Green-labeled liposomes Oregon GreenÒ 488, 1,2dihexadecanoyl-sn-glycerol-3-phosphoethanolamine (OG-DHPE), is obtained from Molecular Probes (Eugene, OR). Liposomes composed of various ratios of PC/PA/OG-DHPE are prepared by reverse-phase evaporation (Szoka and Papahadjopoulos, 1978). The ratios used are 89:10:1,

332

Josemar A. Castillo and Mark A. Hayes

89:30:1, and 89:60:1 PC:PA:OG-DHPE by mass, respectively. OG-DHPE is obtained as a solid and is dissolved in CHCl3 (1.00 ml) to obtain a final concentration of 1 mg/ml. In brief, 10 mg of total CHCl3-dissolved lipids are added to a round-bottom flask. While rotating the flask, the CHCl3 is rapidly evaporated off using nitrogen gas, leaving a thin, uniform solid gel coating of lipids on the interior of the flask. Remaining CHCl3 is removed by vacuum. The dry lipids are prehydrated with a few microliters of water with rotation at 39  C. Phosphate buffer (pH ¼ 7.4) is then added to the flask, which is allowed to rotate at 39  C for 2 h. 2.1.2. Surface attachment 2.1.2.1. Sample wells Rectangles of approximately 2  4 cm are defined with silicon sealant (PermatexÒ ) on a microscope glass slide (Fig. 16.2A). The height of the silicon wall is adjusted to ensure that wells are capable of holding approximately 2 ml of solution. Wells can be open to air, however, it also possible to used a second piece of glass to partially close the structure while leaving open areas to serve as access for the placement of electrodes.

A

B

− V +

Low voltage power supply

Glass slide

1.7 cm Platinum wires (1 mm diameter)

Microscope stage

Silicone well (2 × 4 cm)

Figure 16.2 (A) Schematic (top view) of the experimental apparatus used for data collection along with a fluorescence image of a tubule growing from an immobilized liposome. The image illustrates the direction of tubular growth when an electric field is applied to surface-attached vesicles. (B) View of the experimental apparatus on the microscope stage during data collection (Castillo et al., 2009).

Bionanotubules Formed from Liposomes

333

2.1.2.2. Substrate coating Wells are cleaned and treated with a cationic surfactant according to methods developed for the treatment of capillary electrophoresis surfaces (Katayama et al., 1998). After thoroughly rinsing the wells with water, they are rinsed with 10 mM NaOH for 5 min, then with water for 5 min, and gently patted dry with VWR light-duty tissue wipes (West Chester, PA). The well is then filled with 7.5% polybrene (PB) (w/v in water) for 15 min, followed by a water rinse for 5 min. Hexadimethrine bromide (PB) is obtained from Sigma-Aldrich. 2.1.2.3. Liposome attachment The liposome preparation is allowed to reside in the prepared glass slide well for 5 min. Phosphate buffer (pH ¼ 7.4) is then used to overflow the well with the purpose of displacing the nonattached liposomes from the remaining solution while preventing the drying of the attachment surface. This procedure is performed until the remaining solution is visibly clear, indicating that most of the nonattached liposomes are removed from the surface. The outside of the well and the rest of the glass slide are completely dried with tissue wipes prior to beginning the imaging with the microscope. Substrate coating procedures are performed at room temperature.

2.1.3. Apparatus for electric field application The apparatus shown in Fig. 16.2A and B is assembled in-house. It consists of a simple circuit powered by low-voltage power supply with a voltmeter attached in parallel. One centimeter sections of two platinum wires, 1-mm diameter and separated by a distance of 1.7 cm, are submerged in the sample well perpendicularly to the long axis of the microscope slide. Both electrodes should rest on the substrate to ensure stability and submersion in the solution. As shown in Fig. 16.2B, the alligator clips holding the electrodes are attached to an easily movable block to add rigidity and portability to the system as well as to minimize realignment of the electrodes prior to each experiment (Castillo et al., 2009). 2.1.4. Tubular extension After electrodes are placed in the well, tubules are extended by starting with no electric field and progressively increasing the magnitude of the voltage applied. Generally, voltage magnitude is increase by 2 V/cm every 15 min (or longer). The mother liposomes are giant uni- and/or multilamellar vesicles with typical diameters of 10–50 mm. Well-behaved and stable growth of nanotubules can be observed with modest electric fields (<20 V/cm). Before the electric field is applied, there is a relative absence of tubules and the majority of the objects observed are of spherical shape (liposomes). Tubular structures begin to extend from the immobilized liposomes approximately a few minutes after the field is initiated. Tubular orientation is influenced by the direction of the electric field applied (Castillo et al., 2009).

334

Josemar A. Castillo and Mark A. Hayes

2.1.5. Monitoring tubular growth In general, image collection is performed by fluorescence microscopy in order to monitor tubular growth in real time. An inverted microscope with dark-field and fluorescence capability (IX70, Olympus) using a 100-W high-pressure Hg lamp as a light source is used to observe the liposome and tubular networks. Light from the mercury lamp is passed through a 460–500-nm band-pass filter and through a 40 objective to the sample. Emitted light is collected through a 505-nm long-pass dichromatic mirror and a 510–560-nm band-pass filter into the camera port on the microscope. Digital image collection is performed using a QICAM CCD camera (Q Imaging Inc.) connected to a personal computer running Streampix III (Norpix). 2.1.6. Tubular quantitation A series of 10–20 fluorescence microscopy images are captured at varying times during each applied voltage. The electric field is initiated or increased several seconds before the first image at each value is recorded. Images are generally taken around the midpoint area between the two electrodes. Each capture is manually quantified for the number of attached liposomes and number of tubules formed (Castillo et al., 2009).

3. Bionanotubule Formation from Liposomes in Solution Using Electric Fields Among the diverse structures found when free lipid vesicles (in buffer solutions) are subject to electric fields there are elongated structures that extend from mother liposomes (Fig. 16.3). These nanotubular structures are obtained from liposomes in solutions across several length scales as they can be observed from scanning electron microscopy (SEM) images for smaller vesicles (200 nm diameters, Fig. 16.4) and fluorescence microscopy images of larger liposomes (1–10 mm, Fig. 16.5) (Hayes et al., 2007). For smaller liposomes (200 nm) a variety of long-range structures in the order of 10 mm are formed including slender bodies that can be considered nanotubules. These are present, however, in different modes, at various liposomes compositions including mixed PA and PC liposomes (1:9 PA:PC mol%) and pure PC vesicles. The applied fields used to obtain this type of structures are approximately 300 V/cm. For giant liposomes tubular structures are obtained with lipid compositions consisting of a 1:9 mixture of NBD-PA and PC (mol%) while undergoing electrophoresis at applied fields of approximately 50 V/cm. The nanotubules obtained through this method have shown diameters of roughly 20 nm.

335

Bionanotubules Formed from Liposomes

Acc.V Spot Magn Det WD 1.40 kV 3.0 1000x SE 11.1 Bottom

20 mm

Figure 16.3 SEM image of long-range organization of 9:1 PC:PA liposomes frozen while in the presence of a modest electric field. The average diameter of the liposomes is 200 nm and the applied field strength is approximately 300 V/cm (Hayes et al., 2007).

Det WD Acc.V Spot Magn Det WD 300 kV 3.0 25,000x SE 6.7 ASU

1 mm

1 mm

Figure 16.4 Representative SEM images of quick-frozen large unilamellar liposomes. Main image is structures frozen in the presence of an electric field. Inset is with no field applied. Membranes appear white and buffer gray in the SEM image. Note the wide diversity of structures in the main SEM image (lipid nanotubules are indicated by the white arrows) compared to the spherical nature of liposomes without the field. The suspending buffer is partially sublimed before imaging (Hayes et al., 2007).

Even though tubular and quasi-tubular structures can be obtained from liposomes in solution undergoing electrophoresis, the orientation and overall behavior of these system are much harder to control and monitor.

336

Josemar A. Castillo and Mark A. Hayes

Figure 16.5 Fluorescence image of a lipid tubule formed by applying an electric field (50 V/cm) to a 1:9 NBD-PA:PC liposome in solution. The approximate size of the structure is 5  50 mm (Hayes et al., 2007).

The methods used here are more appropriate for fundamental studies of liposome behavior under the influence of electric fields rather than for tubular fabrication purposes. Therefore, as it is more suitable for the purpose of this chapter, only a summary of these methods will be presented.

3.1. Materials and methods 3.1.1. Liposome preparation Liposomes used in the SEM studies are prepared by a standard extrusion method. Briefly, CHCl3-dissolved lipids are dried under a stream of nitrogen, the lipid film is rehydrated with buffer (2 mM tricine, 5 mM K2SO4, pH 8.8), and the liposomes sized by repeated extrusions through a 200-nm pore membrane. Liposomes used in the fluorescence imaging studies are prepared by a prehydration and incubation procedure (as described in the previous section). 3.1.2. Scanning electron microscopy Liposome electrophoresis samples are quick frozen by plunging the sample into liquid propane (this presents a significant safety hazard and proper precautions must be taken). Freezing is performed with the electric field applied. Images are acquired on a FEI XL30 EFSEM (Philips environmental field emission scanning electron microscope). The frozen samples are loaded into a cryo-prep chamber of the instrument (maintained at 140  C). The samples are first cleaved to expose the liposome suspension, and then the buffer is partially sublimed (5–20 min at  90  C). Sublimation removes the surrounding buffer, partially uncovering the liposomes and allowing them to be directly imaged. After sublimation, the samples are sputter coated with gold–palladium (typically 10 mA for 90 s at 140  C). The samples are then moved onto the instrument’s cryo-stage and imaged (Frederick et al., 1996; Hayes et al., 2007).

Bionanotubules Formed from Liposomes

337

3.1.3. Fluorescence imaging of electric field effects Fluorescence microscopy is performed on an Olympus IX70 inverted fluorescence microscope. Image collection is performed using a QICAM CCD camera (Q Imaging Inc.) connected to a personal computer running Streampix III (Norpix), and using a Sony DV color digital video camera. Liposomes undergoing electrophoresis are imaged directly through square fused silica capillaries (50 mm inner/360 mm outer dimensions).

4. Other Methods of Bionanotubule Formation from Liposomes A number of other methods have been used to form bionanotubules from liposomes. Most of these methods are based on pipette-aspiration protocols that date back to the early 1980s (Waugh, 1982). In general, they involve pulling single nanotubules from immobilized liposomes. These types of experiments where initially used to study properties of lipid membranes before the potential of these structures in the nanotechnology context are realized (Karlsson et al., 2004). In these techniques the application of point loads in the vesicle surface causes a force shape transition in the bilayer. Elongated tethers are then formed as a result of translating a pipette tip, for example, away from the mother liposome (Heinrich et al., 1999; Karlsson et al., 2004; West et al., 2008a). Outstanding work in this area has been achieved by the Orwar group. They have been able to create various geometrically sophisticated nanotubule–vesicle networks using a variety of micromanipulation techniques (Karlsson et al., 2001, 2002, 2004; Sott et al., 2003). Their approach involves not only point load application by pipette aspiration but also other protocols such as mechanical fission of vesicles (Fig. 16.6A and B). In this technique, micromanipulator-controlled carbon fibers are placed on the surface of an immobilized liposome, pressed down, and horizontally moved along the surface to create nanotubule–vesicle networks. For more detail involving these techniques, the review by Karlsson et al. (2004) and the chapter by Jesorka and Orwar (2009) are recommended. Nanotubule formation has also been achieved by micromanipulation protocols using optical tweezers, in which lasers are focused to seize onto the lipid membrane of the liposome and subsequently stretch it to obtain lipid tethers (Fig. 16.6C) (Roux et al., 2005). The observation of tubular structures formed from cell-size liposomes in the presence of gangliosides has been documented by Akiyoshi et al. (2003) (Fig. 16.6E). In this work, tubular structures and liposomes networks connected by the tubes are formed when a ganglioside is added to dioleoylphosphatidyl-choline vesicles. In addition, pulling tubules from liposomes

338

Josemar A. Castillo and Mark A. Hayes

A

F

BF

F

+P

C Tube Vesicle Pipette

Bead

5 nm

D

−

Electric field

+

E

F

Vesicle Tube

Microtubule

BODFL-HPC

Figure 16.6 Various methods of bionanotubular formation from liposomes. (A) Mechanical fission of vesicles along with fluorescence micrographs of two multilamellar lipid nanotube-vesicle networks (Karlsson et al., 2004). (B) Brief schematic showing the formation of a nanotube-connected daughter vesicle from a unilamellar mother vesicle using the vesicle inflation technique. Fluorescence micrograph of a contentsdifferentiated network is formed by loading each container with different materials (Karlsson et al., 2004). (C) Formation of a thin lipid tubule by retracting a streptavidin bead trapped by optical tweezers from a micropipette-aspirated GUV-containing biotinylated lipids (Roux et al., 2005). (D) General schematic of tubular formation from liposomes using electric fields along with a fluorescence micrograph of a tubule formed at a field strength of 8 V/cm (Castillo et al., 2009). (E) Microscopy images of tubular structures induced by adding gangliosides to dioleoylphosphatidyl-choline cell-size liposomes (Akiyoshi et al., 2003). (F) Fluorescence micrograph of tubules pulled from liposome by the action of molecular motors (Roux et al., 2005).

by the action of molecular motors has also been demonstrated by Roux et al. (2002) (Fig. 16.6F). In this study, kinesin molecules attached to giant unilamellar vesicles (GUVs) by means of small polystyrene beads give rise to network formation including tubular structures. Tubes formed from GUVs were also obtained when the vesicles were made from purified Golgi lipids and membranes. Aside from the use of electric fields described in previous sections and the micromanipulations techniques summarized here, tubules can also be extended from liposomes by pulling with hydrodynamic flows (Rossier et al., 2003). In this protocol, vesicles are linked to a substrate and unwound

Bionanotubules Formed from Liposomes

339

by flows applied either by pressure gradients or by electric fields. Tubes of various lengths (ranging from several microns to millimeters) are extruded from liposomes as a consequence of the effect that the flow has on the surface of the vesicle. It is important to note that even though electric fields are used in this technique they are applied to create flow; this which differs from the previously discussed protocol in which tubules are formed in a stationary buffer system.

5. Concluding Remarks The methods and studies included in this chapter describe the many techniques that are now available for the formation of bionanotubules from liposomes. These protocols involve both direct and indirect extension of tubules from previously prepared vesicles of various compositions. Among the various techniques discussed, micromanipulation protocols have been established for the longest. Generally, these methods are more widely used because of their advantages regarding geometry and overall variable control (e.g., vesicle surface tension). These techniques seem also beneficial when tubules need to be extended from highly rigid liposomes (e.g., cholesterolcontaining vesicles), since the application of a point load to induce shape transformation is localized within a small area on the surface of the vesicles. Indirect mechanisms of tubule formation from vesicles, such as with the use of electric fields were also described. This particular approach, when performed with immobilized vesicles, allows the use of low-magnitude electric fields (<20 V/cm) for nanotubule formation showing that lipid tubules of mixed lipid vesicles can be prepared without direct mechanical manipulation. Tubules from several microns to even millimeters in length can grow simultaneously from a mother vesicle and tubule alignment can potentially be controlled by varying the direction of the applied field. This is, however, an evolving technique for which several important issues still need to be investigated. Further studies are necessary to understand the kinetics of tubular growth and the influence that liposomes size, temperature, and adhesion chemistry have in the system.

REFERENCES Akiyoshi, K., Itaya, A., Nomura, S.-I. M., Ono, N., and Yoshikawa, K. (2003). Induction of neuron-like tubes and liposome networks by cooperative effect of gangliosides and phospholipids. FEBS Lett. 534, 33–38. Allen, T. M., and Martin, F. J. (2004). Advantages of liposomal delivery systems for anthracyclines. Semin. Oncol. 31, 5–15.

340

Josemar A. Castillo and Mark A. Hayes

Brazhnik, K. P., Vreeland, W. N., Hutchison, B., Kishore, R., Wells, J., Helmerson, K., and Locascio, L. E. (2005). Directed growth of pure phosphatidylcholine nanotubes in microfluidic channels. Langmuir 21, 10814–10817. Castillo, J. A., Narciso, D. M., and Hayes, M. A. (2009). Bionanotubule formation from surface-attached liposomes using electric fields. Langmuir 25, 391–396. de Kruijff, B., Cullis, P. R., Verkleij, A. J., Vanechteld, C. J. A., Gerritsen, W. J., Mombers, C., Noordam, P. C., and Gier, J. D. (1980). Structural and functional roles of lipids in model and biological-membranes. Eur. J. Cell Biol. 22, 217. Dere´nyi, I., Ju¨licher, F., and Prost, J. (2002). Formation and interaction of membrane tubes. Phys. Rev. Lett. 88(238101)-1-4. Dickson, D. (1993). UK scientists test liposome gene therapy technique. Nature 365, 4. Du¨zgu¨nes¸, N. (1998). Treatment of human immunodeficiency virus, Mycobacterium avium and Mycobacterium tuberculosis infections by liposome-encapsulated drugs. In ‘‘Medical Applications of Liposomes,’’ (D. D. Lasic and D. Papahadjopoulos, eds.), pp. 189–219. Elsevier, Amsterdam. Du¨zgu¨nes¸, N., and Nir, S. (1995). Liposomes as tools for elucidating the mechanisms of membrane fusion. In ‘‘Liposomes as Tools in Basic Research and Industry,’’ ( J. R. Philippot and F. Schuber, eds.), pp. 103–136. CRC Press, Boca Raton. Ewert, K. K., Ahmad, A., Evans, H. M., and Safinya, C. R. (2005). Cationic lipid-DNA complexes for non-viral gene therapy: Relating supramolecular structures to cellular pathways. Expert Opin. Biol. Ther. 5, 33–53. Fenske, D. B., and Cullis, P. R. (2005). Entrapment of small molecules and nucleic acidbased drugs in liposomes. Methods Enzymol. 391, 7–40. Frederick, P. M., Suart, M. C. A., Bomans, P. H. H., and Lasic, D. D. (1996). Cryo-electron microscopy of liposomes. In ‘‘Handbook of Non-Medical Application of Liposomes,’’ (D. D. Lasic and Y. Barenholz, eds.), 1, pp. 309322. Frusawa, H., Fukagawa, A., Ikeda, Y., Araki, Y., Ito, K., John, G., and Shimizu, T. (2003). Aligning a single-lipid nanotube with moderate stiffness. Angew. Chem. Int. Ed. 42, 72–74. Hayes, M., Pysher, M., and Kangping, C. (2007). Liposomes form nanotubules and long range networks in the presence of electric fields. J. Nanosci. Nanotechnol. 7, 2283–2286. ˇ eksˇ, B. (1999). Vesicle deformation by an axial Heinrich, V., Bozic, B., Svetina, S., and Z load: From enlongated shapes to tethered vesicles. Biophys. J. 76, 2056–2071. Jesorka, A., and Orwar, O. (2009). Complex nanotube-liposome networks. Methods Enzymol. 464, 307–323. John, G., Masuda, M., Okada, Y., Yase, K., and Shimizu, T. (2001). Nanotube formation from renewable resources via coiled nanofibers. Adv. Mater. 13, 715–718. John, G., Jung, G. H., Minamikawa, H., Yoshida, K., and Shimizu, T. (2002). Morphological control of helical solid bilayers in high-axial-ratio nanostructures through binary selfassembly. Chem. Eur. J. 8, 5494–5500. Jones, B., Lynch, P. T., Power, J. B., and Davey, M. R. (1996). Electrofusion and electrophoration equipment. In ‘‘Electrical Manipulation of Cells,’’ (P. T. Lynch and M. R. Davey, eds.), pp. 1–14. Chapman and Hall, New York. Jung, G. H., John, G., Yoshida, K., and Shimizu, T. (2002). Self-assembling structures of long-chain phenyl glucoside influenced by the introduction of double bonds. J. Am. Chem. Soc. 124, 10674–10675. Karlsson, M., Sott, K., Cans, A.-S., Karlsson, A., Karlsson, R., and Orwar, O. (2001). Micropipet-assisted formation of microscopic networks of unilamellar lipid bilayer nanotubes and containers. Langmuir 17, 6754–6758. Karlsson, M., Sott, K., Davidson, M., Cans, A.-S., Linderholm, P., Chiu, D., and Orwar, O. (2002). Formation of geometrically complex lipid nanotube-vesicle networks of higherorder topologies. Proc. Natl. Acad. Sci. USA 99, 11573–11578.

Bionanotubules Formed from Liposomes

341

Karlsson, M., Davidson, M., Karlsson, R., Karlsson, A., Bergenholtz, J., Konkoli, Z., Jesorka, A., Lobovkina, T., Hurtig, J., Voinova, M., and Orwar, O. (2004). Biomimetic nanoscale reactors and networks. Annu. Rev. Phys. Chem. 55, 613–649. Karlsson, R., Karlsson, A., Ewing, A., Dommersness, P., Joanny, J.-F., Jesorka, A., and Orwar, O. (2006). Analysis in nanoscale surfactant networks. Anal. Chem. 78, 5961–5968. Katayama, H., Ishihama, Y., and Asakawa, N. (1998). Stable capillary coating with successive multiple ionic polymer layers. Anal. Chem. 70, 2254–2260. Kulkarni, V., Wong, J., Aust, D., Wilmott, J., and Hayward, J. (2001). Lipid nanotubes as skin penetration modulators. J. Cosmet. Sci. 52, 344–345. Lasic, D. D. (1988). The mechanism of vesicle formation. Biochem. J. 258, 1–11. Lasic, D. D. (1996). On the history of liposomes. In ‘‘Handbook of Nonmedical Applications of Liposomes,’’ (D. D. Lasic and Y. Barenholz, eds.), p. 1–12. Mahajan, N., and Fang, J. (2005). Two-dimensional ordered arrays of aligned lipid tubules on substrates with microfluidic networks. Langmuir 21, 3153–3157. Mishra, B. K., and Thomas, B. N. (2002). Phospholipid/protein cones. J. Am. Chem. Soc. 124, 6866–6871. Nolkrantz, K., Farre, C., Brederlau, A., Karlsson, R., Brennan, C., Eriksson, P. S., Weber, S. G., Sandberg, M., and Orwar, O. (2001). Electroporation of single cells and tissues with an electrolyte-filled capillary. Anal. Chem. 73, 4469–4477. Papahadjopoulos, D., Vail, W. J., Jacobson, K., and Poste, G. (1975). Cochleate lipid cylinders: Formation by fusion of unilamellar lipid vesicles. Biochim. Biophys. Acta 394, 483–491. Phayre, A. N., Vanegas Farfano, H. M., and Hayes, M. A. (2002). Effects on pH gradients on liposomal charge states examined by capillary electrophoresis. Langmuir 18, 6499–6503. Pysher, M., and Hayes, M. (2004). Examination of the electrophoretic behavior of liposomes. Langmuir 20, 4369–4375. Pysher, M. D., and Hayes, M. A. (2005). Effects of deformability, uneven surface charge distribution, and multipole moments on biocolloid electrophoretic mobility. Langmuir 21, 3572–3577. Rongen, H. A. H., Bult, A., and van Bennekom, W. P. (1997). Liposomes and immunoassays. J. Immunol. Methods 204, 105–133. Rosenblatt, C., Yager, P., and Schoen, P. (1987). Orientation of lipid tubules by a magnetic field. Biophys. J. 52, 295–301. Rossier, O., Cuvelier, D., Borghi, N., Puech, P. H., Derenyi, I., Buguin, A., Nassoy, P., and Brochard-Wyat, F. (2003). Giant vesicles under flows: Extrusion and retraction of tubes. Langmuir 19, 575–584. Roux, A., Cappello, G., Cartaud, J., Prost, J., Goud, B., and Bassereau, P. (2002). A minimal system allowing tubulation with molecular motors pulling on giant liposomes. Proc. Natl. Acad. Sci. USA 99, 5394–5399. Roux, A., Cuvelier, D., Nassoy, P., Prost, J., Bassereau, P., and Goud, B. (2005). Role of curvature and phase transition in lipid sorting and fission of membrane tubules. EMBO J. 24, 1537–1545. Rustom, A., Saffrich, R., Markovic, I., Walther, P., and Gerdes, H.-H. (2004). Nanotubular highways for intercellular organelle transport. Science 303, 1007–1010. Simo˜es, S., Filipe, A., Faneca, H., Mano, M., Penacho, N., Du¨zgu¨nes¸, N., and Pedroso de Lima, M. C. (2005). Cationic liposomes for gene delivery. Expert Opin. Drug Deliv. 2, 237–254. Singh, A., Wong, E. M., and Schnur, J. M. (2003). Toward the rational control of nanoscale structures using chiral self-assembly: Diacetylenic phosphocholines. Langmuir 19, 1888–1898.

342

Josemar A. Castillo and Mark A. Hayes

Sott, K., Karlsson, M., Pihl, J., Hurtig, J., Lobovkina, T., and Orwar, O. (2003). Micropipet writing technique for production of two-dimensional lipid bilayer nanotube-vesicle networks on functionalized and patterned surfaces. Langmuir 19, 3904–3910. Spector, M. S., Selinger, J. V., Singh, A., Rodriguez, J. M., Price, R. R., and Schnur, J. M. (1998). Controlling the morphology of chiral lipid tubules. Langmuir 14, 3493–3500. Spector, M. S., Singh, A., Messersmith, P. B., and Schnur, J. M. (2001). Chiral self-assembly of nanotubules and ribbons from phospholipid mixtures. Nano Lett. 1, 375–378. Steinberg-Yfrach, G., Liddell, P. A., Hung, S.-C., Moore, A. L., Gust, D., and Moore, T. A. (1997). Conversion of light energy to proton potential in liposomes by artificial photosynthetic reaction centers. Nature 385, 239–241. Svenson, S., and Messersmith, P. B. (1999). Formation of polymerizable phospholipid nanotubules and their transformation into a network gel. Langmuir 15, 4464–4471. Szoka, F., and Papahadjopoulos, D. (1978). Procedure for preparation of liposomes with large internal aqueous space and high capture by reverse-phase evaporation. Proc. Natl. Acad. Sci. USA 74, 4194–4198. Thomas, B. N., Safinya, C. R., Plano, R. J., and Clark, N. A. (1995). Lipid tubule selfassembly: Length dependence on cooling rate through a first-order phase transition. Science 267, 1635–1638. Thomas, B. N., Lindermann, C. M., and Clark, N. A. (1999). Left- and right-handed helical tubule intermediates from a pure chiral phospholipid. Phys. Rev. Lett. E 59, 3040–3047. Watkins, S. C., and Salter, R. (2005). Functional connectivity between immune cells mediated by tunneling nanotubules. Immunity 23, 309–318. Waugh, R. (1982). Surface viscosity measurements from large bilayer vesicle tether formation. 1. Analysis. Biophys. J. 36, 19–27. West, J., Manz, A., and Dittrich, P. S. (2008a). Lipid nanotubule fabrication by microfluidic tweezing. Langmuir 24, 6754–6758. West, J., Manz, A., and Dittrich, P. S. (2008b). Massively parallel production of lipid microstructures. Lab Chip 8, 1852–1855. Yang, B., Kamiya, S., Shimizu, Y., Koshizaki, N., and Shimizu, T. (2004). Glycolipid nanotube hollow cylinders as substrates: Fabrication of one-dimensional metallic-organic nanocomposites and metal nanowires. Chem. Mater. 16, 2826–2831. Zimmermann, H., Richter, E., Reichle, C., Westphal, I., Geggier, P., Rehn, U., Rogaschewski, S., Bleiss, W., and Fuhr, G. R. (2001). Mammalian cell traces-morphology, molecular composition, artificial guidance and biotechnological relevance as a new type of ‘‘bionanotube’’ Appl. Phys. A—Mater. 73, 11–26.

C H A P T E R

S E V E N T E E N

Engineering Cationic Liposome: siRNA Complexes for In Vitro and In Vivo Delivery Jennifer E. Podesta and Kostas Kostarelos Contents 1. Introduction 2. Cationic Liposome Systems for siRNA Delivery 2.1. DOTAP/Cholesterol 2.2. DOTAP:Cholesterol:DSPE-PEG2000 3. Experimental Methods 3.1. General considerations 3.2. Materials 3.3. Cationic liposome preparation using the lipid film hydration method 3.4. PEGylated cationic liposome preparation using the lipid film hydration method 3.5. Liposome/siRNA complex formation 3.6. PEGylated liposome preparation for encapsulation of siRNA using the lipid film hydration method 3.7. In vitro gene silencing 4. Troubleshooting 5. Concluding Remarks References

344 345 346 347 347 347 348 348 349 349 349 351 352 353 354

Abstract RNA interference, the sequence-specific silencing of gene expression by introduction of short interfering RNA (siRNA) is a powerful tool that that the potential to act as a therapeutic agent and the advantage of decreasing toxic effects on normal tissue sometimes seen with conventional treatments i.e. small molecule inhibitors. Naked, unmodified siRNA is poorly taken up by cells and is subject to degradation when exposed to blood proteins during systemic administration. It has also been shown to produce non-specific immune response as well as Nanomedicine Laboratory, Centre for Drug Delivery Research, The School of Pharmacy, University of London, London, United Kingdom Methods in Enzymology, Volume 464 ISSN 0076-6879, DOI: 10.1016/S0076-6879(09)64017-9

#

2009 Elsevier Inc. All rights reserved.

343

344

Jennifer E. Podesta and Kostas Kostarelos

having the potential to generate ‘off-target’ effects. Therefore there is a requirement for a delivery system to not only protect the siRNA and facilitate its uptake, but additionally to offer the potential for targeted delivery with an aim of exploiting the high specificity afforded by RNA interference. Cationic liposomes are the most studied, non-viral delivery system used for nucleic acid delivery. As such, the use of cationic liposomes is promising for siRNA for delivery. Furthermore, polyethylene glycol (PEG) can be incorporated into the liposome formulation to create sterically stabilized or ‘stealth’ liposomes. Addition of PEG can reduce recognition by the reticuloendothelial system (RES) thereby prolonging circulation time. Here we describe a methodology for the complexation of siRNA with cationic liposomes and PEGylated liposomes using two protocols: mixing and encapsulation. Moreover, the different formulations are compared head to head to demonstrate their efficacy for gene silencing.

1. Introduction RNA interference using short interfering siRNA offers the potential for specific gene silencing, and as such, has great promise for gene function studies as well as a therapeutic agent (Castanotto and Rossi, 2009; de Fougerolles et al., 2007). RNAi is initiated by short, double-stranded RNA 19–23 nucleotides (nt) in length complementary to the gene of interest, and blocks protein expression by targeting the mRNA for nuclease degradation before translation occurs (Elbashir et al., 2001). By exploiting the RNAi pathway, synthetic siRNA can be designed to specifically silence genes for the treatment of diseases such as cancer or viral infections, for example, HIV/AIDS. Nucleic acid molecules do not readily cross the cell membrane, and therefore require a delivery vehicle to facilitate cellular uptake. During the last 20 years much has been learned from gene therapy using plasmid DNA as well as antisense oligonucleotide studies about the interaction between the delivery agents and nucleic acids (Mintzer and Simanek, 2009). Cationic liposomes have been the most studied delivery systems for nucleic acids, due to the simplicity of the electrostatic interaction between cationic liposomes and negatively charged nucleic acids to form the vector complex. Cationic liposomes have been used successfully for the intracellular delivery of siRNA. RNA interference mediated by siRNA requires delivery of siRNA into the cytoplasm of the cell, unlike plasmid DNA or plasmid- or virus-encoded short hairpin RNA that requires nuclear localization (Meister and Tuschl, 2004). Delivery to the cytoplasm removes one of the delivery hurdles by not having to cross the nuclear membrane. Previous reports utilizing naked siRNA in vivo have often relied on hydrodynamic injection (Lewis and Wolff, 2007). This technique is useful for generating liver uptake; however, this is done by inducing microperforations into the hepatocyte membrane to facilitate uptake. Moreover, the rapid injection of a large volume under high pressure is not clinically viable. Clinical siRNA therapy so far has primarily been limited to local administration of naked siRNA. This,

Engineering Cationic Liposome:siRNA Complexes for In Vitro and In Vivo Delivery

345

however, has led to a series of concerns over safety, efficacy, and specificity of the therapy (Cho et al., 2009). Activation of the innate immune system and generation of off-target effects are two serious side effects observed with siRNA. Reduction of the siRNA concentration is one way to reduce off-target effects. Shielding of siRNA from recognition by the reticuloendothelial system (RES) has the potential to prevent or limit immune activation. Furthermore, chemical modifications to the siRNA molecules have also led to more stable, less immunoreactive RNA species (Morrissey et al., 2005). For siRNA to progress as a viable therapeutic agent it requires a delivery vehicle. The vehicle should have the ability to: (a) target siRNA to the region of interest (be it a tumor or particular tissue within the body); (b) provide protection from nuclease degradation while in the blood; (c) limit excretion through the kidneys (Whitehead et al., 2009); (d) allow use of minimal siRNA doses to reduce off-target effects; and (e) shield the siRNA from the immune system. Ultimately, it should achieve enough active copies of siRNA at the target site to elicit a therapeutic effect with limited toxicity. To address all these requirements and specifications for an siRNA delivery system that can be used both in vitro and in vivo, we compared three systems based on a well-characterized cationic liposome (DOTAP:cholesterol) system and utilized two protocols for generating the liposome:siRNA complexes.

2. Cationic Liposome Systems for siRNA Delivery

Pre-formed liposomes

Mixing protocol + + + -PEG +

+

- -

+

Add siRNA +

+

+ +PEG

-

-

-

siRNA complexes to external liposome surface via electrostatic interaction

-

+

Encapsulation protocol Hydrate with siRNA solution

Prepare lipid film

siRNA interacts with internal and external side of lipid bilayer

346

Jennifer E. Podesta and Kostas Kostarelos

2.1. DOTAP/Cholesterol Liposomes are the most studied, nonviral delivery vehicle for gene therapy, and cationic liposomes in particular have been shown to have an affinity for tumor vasculature preferentially over zwitterionic and anionic liposomes (Thurston et al., 1998). Also, the growth of blood vessels within a solid tumor is rapid and disorganized, resulting in leaky vasculature that may enable liposomes to extravasate through the gaps in the endothelium into the tumor. Cationic liposomes have already been extensively studied as carriers of anticancer drugs such as paclitaxel (Endo-TAG) and doxorubicin (Campbell et al., 2009). Liposomes formed from the cationic lipid DOTAP and cholesterol have been well characterized as a delivery system. Numerous preclinical studies utilize DOTAP/Chol liposomes for delivery of plasmid DNA and antisense oligonucleotides, and more recently, siRNA (Kim et al., 2007; Ramesh et al., 2001; Zhang et al., 2008). We selected this liposome composition formulated at a 2:1 (DOTAP/Chol) molar ratio as the basis for designing and optimizing an siRNA delivery system. Cationic liposomes are able to form complexes with negatively charged siRNA via electrostatic interactions. The charge ratio (N/P: positively charged nitrogen (DOTAP) to negatively charged phosphate (siRNA)) at which the liposomes and siRNA are mixed has a measurable effect on the physicochemical properties of the resulting complex. The size differences between plasmid DNA (several kb; supercoiled configuration) and siRNA (double-stranded RNA 19–23 nt in length) suggest that they will interact differently with the liposome. Modification from existing commercial liposome transfection reagents and development of novel vector systems specifically for siRNA transfer requires characterization of these systems. For this reason, complexes generated by mixing of liposomes and siRNA are characterized over a range of charge ratios to determine the optimum N/P ratio for complex formation and subsequent transfection studies. Although useful in vitro, cationic liposomes are not a perfect delivery system in vivo. When injected systemically, they are cleared rapidly from the blood, limiting their capacity to target tissues. Incorporation of a flexible polymer such as PEG onto the liposome can significantly prolong circulation time (Allen, 1994).1

1

Size and charge measurements: The particle diameter and surface charge are measured using PCS in a nanosizer (Malvern). Liposomes are diluted 1:4 into filtered 5% dextrose to generate 500 mM (DOTAP) in 1 ml final volume. Higher dilutions may be used for size measurement; however, a higher concentration of the cationic component is required for zeta potential measurements. Diluted liposomes are pipetted into a clean (dustfree) cuvette and a series of measurements are recorded. The mean diameter and surface charge are calculated from 30 readings for both size and surface charge. The polydispersity index gives an indication of the homogeneity of the liposome population.

Engineering Cationic Liposome:siRNA Complexes for In Vitro and In Vivo Delivery

347

2.2. DOTAP:Cholesterol:DSPE-PEG2000 The addition of poly(ethylene glycol) (PEG) to create sterically stabilized liposomes (SSL), also termed ‘‘stealth’’ liposomes, has been used successfully to increase the circulation time after systemic administration. PEG has many properties that make it compatible with clinical use, such as low immunogenicity and toxicity (Allen, 1994). Moreover, it is soluble in aqueous and organic solvents, making it compatible for use in liposome formulations. Conjugation of PEG to a phospholipid, for example, 1,2-distearoyl-snglycero-3-phosphoethanolamine (DSPE) enables PEG to be anchored in the lipid bilayer. DSPE-PEG is used as a means of shielding the liposome from interaction with serum proteins and subsequent uptake by the mononuclear phagocyte system (MPS). Coating the surface of the liposome with a polymer provides a physical barrier that decreases binding of opsonins in the blood. Macrophage recognition and internalization of opsonins is responsible for the rapid clearance of conventional liposomes (Immordino et al., 2006). Additionally, the incorporation of PEG into the lipid bilayer offers a platform on which to add targeting moieties. Monoclonal antibodies, Fab’ fragments and peptides have all been successfully used for active targeting of liposomes ( Janssen et al., 2003; Stephenson et al., 2004). Here, we describe a protocol for complexation between cationic liposomes and siRNA, with particular emphasis on the differences between PEGylated and non-PEGylated liposomes, and subsequent evaluation of the siRNA transfection efficiency. There are several reports of cationic liposome:siRNA complexes for silencing experiments in the literature; however, the effect of PEGylation on the physicochemical properties as well as biological functionality of the siRNA has not been addressed systematically. We propose utilizing the encapsulation method when complexing siRNA with PEGylated liposomes for the design and engineering of vectors for systemic in vivo administration.

3. Experimental Methods 3.1. General considerations Preparation of liposomes includes a filtration step used to reduce the vesicle size. This also has the benefit of filter sterilizing the solution for use in cell culture.2 Once the liposomes have been sonicated, filtration and all 2

Cell culture: Human cervical carcinoma HeLa cells (ATCC #CCL-2TM, Manassas, VA, USA) are cultured in Eagle’s Minimum Essential Medium (MEM) supplemented with 10% (v/v) fetal bovine serum; 1 mM sodium pyruvate; 2 mM L-glutamine; 1500 mg/l sodium bicarbonate; 50 U/ml penicillin and 50 mg/ml streptomycin at 37  C and 5% CO2 in a humidified atmosphere. (All cell culture media and supplements are purchased from Invitrogen, UK.) HeLa cells are passaged 3 times weekly at 1:8; cell seeding density for silencing is determined empirically for each multiwell plate to yield 40–60% confluence at the time of siRNA transfection.

348

Jennifer E. Podesta and Kostas Kostarelos

subsequent steps are carried out in a Class 2 laminar flow cabinet under sterile conditions. Aliquots for measurement can be removed aseptically and transferred to a sterile cuvette enabling the sample to be recovered for downstream applications, if necessary. All solutions and glassware used are RNase-free; sterile, RNase- and DNase-free plasticware is used for all biological applications. Dextrose solution is prepared using RNase-free dH2O. RNase-free water is prepared by treatment with 0.1% (v/v) diethylpyrocarbonate (DEPC) for 4 h overnight, followed by autoclaving at 121  C for 20 min.

3.2. Materials 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP), cholesterol, and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000] (DSPE-PEG2000) from Avanti Polar Lipids, USA. O

Cholesterol

H N +

ClH H HO

O O

DOTAP

O

H O O O

O O P O O H O- + NH4

O N H

(OCH2CH2)45OCH3

mPEG-DSPE

3.3. Cationic liposome preparation using the lipid film hydration method 1. DOTAP and cholesterol (2:1) are dissolved in chloroform:methanol (4:1 v/v) and the organic solvent is evaporated under pressure for 30 min at 40  C using a rotoevaporator. The resulting thin lipid film is flushed with a stream of N2 to remove any trace of the organic solvent. The lipid film may be stored under nitrogen at 4  C overnight, if required. 2. The lipid film is hydrated in 5% (w/v) sterile filtered dextrose by rapid pipetting to produce large, multilamellar liposomes (MLVs). The MLVs are reduced to small, unilamellar vesicles (SUVs) by sonicating in a waterbath sonicator for 1 min followed extrusion through a 0.2 mm Anotop 10 filter (Whatman, UK). The liposome solution is then incubated at room temperature for a minimum 30 min to allow stabilization. 3. Liposome size (z-average diameter) and surface charge (z potential) are measured using laser Doppler velocimetry and photon correlation spectroscopy (PCS)/dynamic light scattering, respectively, in a Zetasizer Nano ZS (Malvern, UK), or similar instrument, at 25  C.

Engineering Cationic Liposome:siRNA Complexes for In Vitro and In Vivo Delivery

349

3.4. PEGylated cationic liposome preparation using the lipid film hydration method PEGylated liposomes are prepared using the same protocol as conventional liposomes. Briefly, DSPE-PEG2000 (5 mol%) is dissolved in the organic solvent with DOTAP and cholesterol. The PEGylated liposome is hydrated, reduced in size, and measured in the same way as the conventional, non-PEGylated liposome.

3.5. Liposome/siRNA complex formation 1. Liposomes and siRNA are diluted separately into 50% final volume. 2. The siRNA is added to the liposome by rapid pipetting to prevent localized high siRNA:liposome concentrations. This is mixed thoroughly by pipetting and brief vortexing. 3. The mixture is then incubated at room temperature for 20 min to allow complexation to occur. Complexes are formed by mixing the liposomes and siRNA at a charge ratio where there is an excess of positive charge. We prepared complexes over a range of charge ratios (1:1-6:1, N/P) to investigate the extent of association as measured by the presence of free siRNA when the L/siRNA is run on a 1% agarose/Tris-borate-EDTA (TBE) gel (Fig. 17.1). Incorporation of PEG into the liposome prevents complete association of siRNA when used at comparable charge ratios as clearly seen by the presence of free siRNA at the 3:1 charge ratio. When complexed at a 4:1 charge ratio, siRNA is completely associated with the liposomal surface in the absence of PEG, however, where PEG is present there is varying amounts of detectable siRNA. The volume of the complex will depend on the downstream application. Samples prepared for electrophoresis should have a sufficiently high siRNA concentration in a volume compatible with gel loading to ensure siRNA detection.3

3.6. PEGylated liposome preparation for encapsulation of siRNA using the lipid film hydration method 1. DOTAP, cholesterol, and DSPE-PEG2000 (2:1:0.1) are dissolved in chloroform:methanol (4:1, v/v). The organic solvent is evaporated 3

Agarose gel electrophoresis: To visualize the siRNA component of the complex, in addition to any free, noncomplexed siRNA, complexes are subjected to agarose gel electrophoresis. Agarose gels are prepared using 1% (w/v) agarose in TBE buffer. Samples are prepared such that they have the same final siRNA concentration of at least 500 ng to facilitate visualization by ethidium bromide staining. Ethidium bromide (EtBr) intercalates the minor groove of nucleic acids and fluoresces when exposed to UV light. EtBr is added to the gel prior to casting at a final concentration of 0.5 mg/ml. Samples are mixed with the appropriate volume of loading dye immediately prior to loading into the well. The gel is then run for 45 min at 70 V. Images are captured using the G:BOX system with Gene Snap software.

350

Jennifer E. Podesta and Kostas Kostarelos

1:1

2:1

3:1

4:1

siRNA

L

DOTAP:Cholesterol

(N/P)

Complexed siRNA

DOTAP: Chol:DSPE-PEG2000

Free siRNA

1:1

2:1

3:1

4:1

6:1

L

siRNA

Figure 17.1 Complexation studies of DOTAP:Cholesterol/siRNA (top) and DOTAP: Chol:DSPE-PEG2000/siRNA (bottom). Liposome/siRNA complexes are formed at the charge ratio (N/P) indicated and incubated for 30 min at room temperature before loading onto 1% agarose/TBE gel. Charge ratios are calculated for fixed siRNA concentration of 2 mg. L, liposome alone. The presence of PEG on the liposome interferes with siRNA complexation.

under pressure at 40  C for 30 min and the lipid film is flushed with N2 to remove residual solvent. 2. The lipid film is hydrated using a solution of siRNA in 5% dextrose (w/v) prepared using RNase-free dH2O. The amount of siRNA used to hydrate the film is calculated from the charge ratio. 3. Size reduction is performed as with conventional liposomes: sonication in a waterbath sonicator for 1 min, followed by extrusion through a 0.2 mm Anotop 10 filter (Whatman, UK). The PEGylated liposome/siRNA solution is then incubated at room temperature for a minimum of 30 min to allow stabilization. The complex should be maintained in a sterile environment for subsequent gene silencing experiments. 4. Liposome size (z-average diameter) and surface charge (zeta potential) are measured as before (Table 17.1).

Engineering Cationic Liposome:siRNA Complexes for In Vitro and In Vivo Delivery

351

Table 17.1 Size and surface charge of liposomes (DOTAP:Chol and DOTAP:Chol: DSPE-PEG) and liposome/siRNA complexes formed at 2:1 and 4:1 charge ratios using the mixing protocol and at 4:1 for the PEGylated liposome/siRNA complexes formed using the encapsulation protocol

Mixing protocol

N/P

Mean diameter (nm)

PdI

Zeta potential (mV)

DOTAP:Chol (2:1) DOTAP:Chol/siRNA DOTAP:Chol/siRNA Mixing protocol DOTAP:Chol:DSPEPEG2000 (2:1:0.1) DOTAP:Chol:DSPEPEG2000/siRNA DOTAP:Chol:DSPEPEG2000/siRNA Encapsulation protocol DOTAP:Chol:DSPEPEG2000 (2:1:0.1)

n/a 2:1 4:1

128.00 832.33 255.67

0.22 0.28 0.19

46.00 38.00 44.50

n/a

146.67

0.20

43.70

2:1

1946.67

0.56

9.19

4:1

203.33

0.14

17.10

4:1

247.67

0.44

25.63

Size is given as the mean diameter (nm) of at least 30 measurements, and the polydispersity index (PdI). Surface charge is the mean zeta potential (mV) for at least 45 measurements.

3.7. In vitro gene silencing 3.7.1. Day 0  

Cells are seeded in a multiwell plate in antibiotic-free media. The plate is then incubated overnight at 37  C, 5% CO2, in a humidified chamber.

3.7.2. Day 1  

Cells should be 30–50% confluent at the time of transfection. Cells are transferred to 2% serum-containing media without antibiotics and then incubated.

3.7.3. Mixing protocol Fresh complexes are prepared for cell transfection studies using aseptic techniques throughout. 

Final siRNA concentration following dilution onto the cells needs to be determined empirically for each siRNA sequence. We suggest starting in the range 20–100 nM.  Liposomes are diluted into prewarmed (25  C) serum-free medium, such as Opti-MEM or Advanced RPMI (Invitrogen, UK) that are specifically

352

Jennifer E. Podesta and Kostas Kostarelos

formulated for use without serum or with low serum concentrations, and equilibrated for 5 min.  siRNA is thawed on ice and then diluted using RNase-free water.  This is then mixed by adding the siRNA to the liposome by rapid pipetting and the mixture is incubated for 20 min at room temperature. 3.7.4. Mixing protocol and encapsulation protocol  

The complex is added dropwise to cells. Four to six hours after the addition of the complex, an equal volume of fresh media, which is supplemented with 18% serum, is added to the cells. The final serum concentration achieved is 10%. The mixture is then incubated overnight.

3.7.5. Day 2 

The medium is replaced by fresh growth medium (10% serum).

3.7.6. Assessing gene silencing RNA interference is a posttranscriptional gene silencing phenomenon, and as such, knockdown of mRNA transcripts should be assessed by real time RT-PCR. Detection of the targeted protein by Western blot or immunohistochemistry can indicate gene silencing; however, factors such as the stability of the protein and the lack of direct correlation between mRNA and protein abundance could lead to false-negative results. Silencing of a gene with a quantifiable function is much more informative than measuring protein alone. The extent and duration of silencing are variable, therefore a high degree of mRNA degradation may need to occur before a biological effect is evident, and conversely a transient reduction of only a small amount of mRNA may have a profound downstream effect (Fig. 17.2). In this example (Fig. 17.2), the siRNA used is targeted to the Polo-like kinase 1 (Plk1) gene. Plk1 is a regulator of the cell cycle and regulated mitotic progression and as such has become an interesting target for anti-cancer therapy using small molecule inhibitors, antisense and siRNA ( Judge et al., 2009). Furthermore, previous comparison with non-targeting or negative siRNA has validated the specificity of the Plk1 siRNA for inducing cell death.

4. Troubleshooting Preparation of complexes between liposomes and siRNA can result in localized precipitation, whereby the solution becomes turbid and the particles come out of solution. To prevent this, the siRNA solution is added to

Engineering Cationic Liposome:siRNA Complexes for In Vitro and In Vivo Delivery

353

120

Percent cell viability

100 80 60 40 20 0 N/P: 2:1

4:1

6:1

DOTAP:Chol /siRNA (mixed)

2:1

4:1

6:1

DOTAP:Chol:DSPE-PEG /siRNA (mixed)

2:1

4:1

6:1 Naïve

DOTAP:Chol:DSPE-PEG /siRNA (encapsulated)

Figure 17.2 Cell viability of HeLa cells transfected with siPlk1 using: DOTAP:Chol liposomes mixed with siRNA; DOTAP:Chol:DSPE-PEG2000 (5 mol%) liposomes mixed with siRNA; or DOTAP:Chol:DSPE-PEG2000 (5 mol%) liposomes hydrated with siRNA using the encapsulation protocol. Liposome/siRNA complexes are prepared at N/P 2:1, 4:1, 6:1 with a final siRNA concentration of 50 nM. Cell viability is assessed by the MTT assay 48 h posttransfection.4

the liposomes, not vice versa and rapid mixing is used. Moreover, the presence of particulate complexes can have a cytotoxic effect independent of specific siRNA activity.

5. Concluding Remarks Cationic liposomes are able to act as delivery systems for siRNA in much the same way as they have been used for plasmid DNA in gene therapy studies. As the field of RNAi expands, new synthetic siRNA molecules are being designed to improve their efficacy and stability as well as reduce off-target effects and immune stimulation. Incorporation of PEG into the lipid bilayer offers several advantages to the vector as discussed here. The problem of reduced siRNA affinity is easily overcome when using the encapsulation method for lipid film hydration without 4

MTT assay: The colorimetric MTT assay is used to measure cell viability. Yellow (3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide (MTT) is reduced by mitochondrial reductase in metabolically active cells to produce purple formazan. Samples can then be measured spectrophotometrically at wavelengths of 550–600 nm, and reduction of cell viability caused by apoptosis or necrosis can be measured by comparison to the control. Stock MTT solution is prepared at 5 mg/ml in PBS, sterile filtered and stored at 20  C. MTT is diluted 1:6 (v/v) into the cell culture medium before addition to cells in 96-well plates at a final volume of 120 ml. Cells are incubated for up to 4 h at 37  C, 5% CO2. The MTT solution is removed and 200 ml DMSO is added to each well to solubilize the formazan before absorbance at 560 nm is read on a microtiter plate reader.

354

Jennifer E. Podesta and Kostas Kostarelos

compromising biological activity. Moreover, this method provides a platform for developing actively targeted liposomes by addition of targeting sequences to PEG.

REFERENCES Allen, T. M. (1994). Long-circulating (sterically stabilized) liposomes for targeted drug delivery. Trends Pharmacol. Sci. 15, 215–220. Campbell, R. B., et al. (2009). Fighting cancer: From the bench to bedside using second generation cationic liposomal therapeutics. J. Pharm. Sci. 98, 411–429. Castanotto, D., and Rossi, J. J. (2009). The promises and pitfalls of RNA-interference-based therapeutics. Nature 457, 426–433. Cho, W. G., et al. (2009). Small interfering RNA-induced TLR3 activation inhibits blood and lymphatic vessel growth. Proc. Natl. Acad. Sci. USA 106, 7137–7142. de Fougerolles, A., et al. (2007). Interfering with disease: A progress report on siRNA-based therapeutics. Nat. Rev. Drug Discov. 6, 443–453. Elbashir, S. M., et al. (2001). Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498. Immordino, M. L., et al. (2006). Stealth liposomes: Review of the basic science, rationale, and clinical applications, existing and potential. Int. J. Nanomed. 1, 297–315. Janssen, A. P., et al. (2003). Peptide-targeted PEG-liposomes in anti-angiogenic therapy. Int. J. Pharm. 254, 55–58. Judge, A. D., Robbins, M., Tavakoli, I., Levi, J., Hu, L., Fronda, A., Ambegia, E., McClintock, K., and MacLachlan, I. (2009). Confirming the RNAi-mediated mechanism of action of siRNA-based cancer therapeutics in mice. J Clin. Invest. 119, 661–673. Kim, S. I., et al. (2007). Systemic and specific delivery of small interfering RNAs to the liver mediated by apolipoprotein A-I. Mol. Ther. 15, 1145–1152. Lewis, D. L., and Wolff, J. A. (2007). Systemic siRNA delivery via hydrodynamic intravascular injection. Adv. Drug Deliv. Rev. 59, 115–123. Meister, G., and Tuschl, T. (2004). Mechanisms of gene silencing by double-stranded RNA. Nature 431, 343–349. Mintzer, M. A., and Simanek, E. E. (2009). Nonviral vectors for gene delivery. Chem. Rev. 109, 259–302. Morrissey, D. V., et al. (2005). Potent and persistent in vivo anti-HBV activity of chemically modified siRNAs. Nat. Biotechnol. 23, 1002–1007. Ramesh, R., et al. (2001). Successful treatment of primary and disseminated human lung cancers by systemic delivery of tumor suppressor genes using an improved liposome vector. Mol. Ther. 3, 337–350. Stephenson, S. M., et al. (2004). Folate receptor-mediated targeting of liposomal drugs to cancer cells. Methods Enzymol. 387, 33–50. Thurston, G., et al. (1998). Cationic liposomes target angiogenic endothelial cells in tumors and chronic inflammation in mice. J. Clin. Invest. 101, 1401–1413. Whitehead, K. A., et al. (2009). Knocking down barriers: Advances in siRNA delivery. Nat. Rev. Drug Discov. 8, 129–138. Zhang, Y., et al. (2008). In vivo comparative study of lipid/DNA complexes with different in vitro serum stability: Effects on biodistribution and tumor accumulation. J. Pharm. Sci. 97, 237–250.

Author Index

A Abdelbary, G., 119 Abra, R. M., 179 Abraham, S. A., 280, 282 Adachi, Y., 225 Aebi, U., 34 Aime, S., 193–208 Akashi, K., 70 Akiyoshi, K., 329, 337, 338 Akwete, L. A., 174 Al-Jamal, W. T., 132 Alami, M., 218, 224, 225 Albrecht-Buehler, G., 32 Alivisatos, A. P., 275 Allen, T. M., 280, 282, 328, 346 Aller, S., 223, 224 Alvarez, M. M., 289 Ambudkar, S. V., 224, 225 Anabousi, S., 183 Anacker, E. W., 87 Anelli, P. L., 197 Angelova, M. I., 36, 66 Anthony, J. H., 168 Antunes, A., 100 Arshady, R., 132 Arya, S. K., 89 Asano, A., 34 Asasutjarit, R., 119 Ashurst, I. I., 168 Astruc, D., 132, 135 Atkins, W. M., 211–228 Attama, A. A., 117 B Baas, B. J., 218 Babailov, S. P., 199 Baginski, M., 100 Baglioni, P., 249–276 Baldelli Bombelli, F., 249–276 Banchelli, M., 249–276 Banerjee, R., 119 Bangham, A. D., 20, 33, 36, 82, 155 Barber, R. F., 169 Ba¨rmann, M., 32 Barrow, D. A., 258 Bartlett, G. R., 61 Battaglia, L., 119 Batzri, S., 155

Bauer, B., 316, 317 Bayburt, T. H., 211–228 Beales, P. A., 250, 275 Beatty, W. L., 81 Beckman, E. V., 80 Benadie, Y., 100, 102 Berti, D., 249–276 Betagiri, J. V., 169, 174 Bhaskar, K., 119, 120 Bi, R., 119 Bisgaard, H., 169 Bjo¨rck, L., 162 Bohren, C. F., 287 Boldog, T., 218, 220, 224 Bolinger, P. Y., 317 Bondi, M. L., 119 Borch, J., 218 Boroske, E., 195 Bosquillona, C., 172 Boyer, C., 295, 296, 299 Brazhnik, K. P., 328, 329 Brenner, H., 320 Bretscher, A., 34 Briggs, J. A., 56 Brioschi, A., 113, 119, 120 Briscoe, P., 172 Brumm, T., 195 Bryan, J., 33 Bucher, K., 223 Bugelski, P. J., 59, 61 Bunjes, H., 108 Burda, C., 132, 135 Butler, W. R., 82 Byers, H. R., 32 Byrdwell, W. C., 61 Byron, P. R., 169, 187 C Cabuil, V., 195 Caffrey, M., 59 Callaghan, R., 224 Campbell, R. B., 345 Cant, K., 33 Cardoso, A., 155 Carr, R. L., 174 Caruthers, M. H., 245 Casadei, M. A., 119 Castanotto, D., 343 Castelli, D. D., 193–208

355

356

Author Index

Castillo, J. A., 327–339 Castro, G. A., 118 Cengiz, E., 118 Cevc, G., 252 Chan, E. D., 80 Chan, Y.-H. M., 234, 250 Chandra, R. A., 234 Chattopadhyay, N., 118 Chen, H. B., 115, 118 Chen, H. H., 119 Chen, J., 281–283 Chen, P. S., 220 Chen, Y., 11 Chien, Y.W, 179 Chifflet, S., 227 Chilkoti, A., 132 Chiruvolu, S., 297 Chithrani, B. D., 140 Chiu, D. T., 283 Cho, W. G., 344 Choppin, P. W., 57 Chougnet, A., 212 Chougule, M., 167–187 Chrambach, A., 218 Christy, R. W., 287 Chu, S., 200 Chung, J. W., 93 Church, G. M., 20 Civjan, N. R., 218 Claessens, M. M. A. E., 35 Clarke, D. M., 225 Clarke, S. W., 169 Cloutier, T. E., 234, 237 Collins, K., 34, 37 Coluccio, L. M., 34 Condeelis, J. S., 32 Coorssen, J. R., 58 Corsi, D. M., 200 Cortese, J. D., 32, 33 Courrier, H. M., 168, 169 Cramer, L. P., 32 Criado, M., 315 Crowell, K. J., 195 Cruz, F., 218 Cullis, P. R., 280, 282, 328 Czolkos, I., 313 D Dalal, K., 218, 224, 225 Daniel, M. C., 132, 135 Das, A., 212, 218 Davidsen, J., 281 Davidson, M., 316, 320 Davidson, N., 262 de Fougerolles, A., 343 de Kruijff, B., 328 de Ven, H. V., 119 Deamer, D., 20

Denisov, I. G., 211–228 Dennison, C., 82 Dere´nyi, I., 328 des Rieux, A., 116 Dianzani, C., 113 Dickson, D., 328 Dimitrov, D. S., 36, 66 Discher, B. M., 282 Dischino, D. D., 195 Doktorovova´, S., 105–121 Dommersnes, P. G., 318, 320 Dong, X., 119 Dowben, R. M., 59 Driver, C.H. S., 79–102 Duan, H., 218 Dulkeith, E., 303 Du¨zgu¨nes¸, N., 60, 61, 155, 220, 328 E Eastoe, J., 92 Ebashi, F., 37 Ebashi, S., 37 Edmondson, D. E., 218 Edwards, R. A., 33 El Assawy, N., 119 El-Sayed, M. A., 132 Elbashir, S. M., 344 Elbert, K. J., 183 Ellis, C., 92 Erickson, H. P., 3–16 Esposito, E., 113, 119 Evans, E., 310 Ewert, K. K., 328 Eytan, G. D., 315 F Fahmy, R. H., 119 Faisal, S., 119 Fang, J. Y., 114, 115, 118, 328 Fang, X., 183 Fenske, D. B., 328 Fernig, D. G., 79–102 Fiegel, J., 184, 185 Fielding, R. M., 179 Fischer, A., 20, 21 Forster, A. C., 20 Fraley, R., 61 Frederick, P. M., 336 Frederik, P. M., 297 Frusawa, H., 328 Fry, D. W., 257 Fuchs, S., 183 Fujikawa, K., 32 Funakoshi, K., 35, 36 Fundaro, A., 111 Fygenson, D. K., 32

357

Author Index G Gallarate, M., 119 Gao, Y., 263 Garcia-Sastre, A., 57 Gaydos, L. J., 320 Geller, D. E., 168 Genzel, L., 286 Gibrat, G., 50 Glebe, D., 152 Goel, B. K., 174, 175 Goh, Y., 147–164 Gokce, E. H., 117 Goluch, E. D., 212 Gompf, B., 293, 301, 304 Goodrum, M. A., 80, 81 Gosse, C., 254 Govorov, A. O., 132 Grabar, K. C., 134 Grazi, E., 34 Greenwald, R. B., 143 Gregoriadis, G., 21 Grinkova, Y. V., 211–228 Grua, P., 292 Guedeau-Boudeville, M. B., 72 H Haba, Y., 135 Ha¨ckl, W., 33 Hama, S., 149 Hamada, T., 35, 36, 41 Han, C., 118 Han, F., 119, 120 Han, J., 89 Hanafy, A., 116 Hanczyc, M. M., 20 Hao, E., 283, 284 Hartley, G. S., 87 Hartwig, J. H., 34 Hase, M., 35 Hayashi, H., 136 Hayes, M. A., 327–339 Heikal, A., 225 Heinrich, V., 329, 337 Helfrich, M. R., 318 Hendrickson, R. C., 100 Henry, M. R., 265 Hentschel, A., 118 Hersey, J. A., 169 Hickey, A. J., 168, 169, 173 Higashi-Fujime, S., 37 Higgins, C. F., 223 Hinnen, A., 153 Hirano, Y., 131–143 Hirota, S., 195 Hirsch, L. R., 281, 282, 286 Hjelmeland, L. M., 218 Hock, R. S., 34

Honda, M., 31–50 Hong, K., 132 Hong, R. L., 149 Ho¨o¨k, F., 234, 238, 250, 254, 275 Horne, R. W., 20, 155 Hosoda, K., 19–29 Hotani, H., 32, 33, 37, 38, 45 Hou, L., 34 Houghton, R. L., 80 Hsu, L., 174 Hsu, M. H., 119 Hu, C., 168 Hu, F.-Q., 112, 118 Huang, G., 109–111, 113 Huang, L., 155 Huang, S.-L., 280, 282, 294 Huang, X., 281, 282 Huang, Z.-r., 119 Hubert, D. H., 297 Huffman, D. R., 287 Hyung, J. L., 174 I Ichikawa, S., 20, 21 Iijima, M., 147–164 Imamura, M., 34 Immordino, M. L., 132, 347 Iscan, Y., 114, 118 Ishikawa, K., 22 Iwasaki, Y., 150 J Jain, D., 119 Jain, P. K., 132 Jain, S. K., 120 Jakobsen, U., 233–247, 275 Janson, L. W., 35, 49 Janssen, A. P., 347 Jayakar, H. R., 56 Jee, J. P., 118 Jesorka, A., 309–323, 337 Johansson, H. O., 318 John, G., 328 Johnson, P. B., 287 Johnsson, M., 108 Jones, B., 329 Jontes, J. D., 37 Joshi, M. R., 115, 118, 119, 173, 174, 182 Jung, G. H., 328 Jung, H. T., 297 Jung, J., 147–164 K Kadimi, U. S., 169 Kageyama, Y., 25 Kalyanasundaram, K., 254

358

Author Index

Kaneko, T., 20, 37 Karlsson, A., 310, 317, 318 Karlsson, M., 310, 311, 314, 315, 328, 329, 337, 338 Karlsson, R., 318, 320, 328 Karp, E. S., 195 Kasuya, T., 147–164 Kaszuba, M., 174, 175 Katayama, H., 333 Keller, B. U., 315 Kelly, M. W., 282, 294 Kent, J., 174 Khant, H. A., 279–304 Kijac, A. Z., 218 Kikuchi, A., 156 Kikuchi, H., 21 Kilburn, J. O., 82 Kim, D., 132 Kim, H. D., 89–91 Kim, I. W., 224, 225 Kim, S. I., 346 Kimelberg, H. K., 168 Kimura, Y., 224 Kinoshita, R., 147–164 Kirby, C., 21 Kirschner, M., 32 Kisak, E. T., 281, 295, 296 Kita, H., 19–29 Kobayashi, M., 153 Kojima, C., 131–143 Kompella, U., 169, 172 Konduri, K., 168 Konkoli, Z., 318, 320, 321, 323 Kono, K., 131–143 Koppel, D. E., 259 Korf, J. E., 81, 82 Korn, E. D., 155 Kostarelos, K., 132, 343–353 Kotaidis, V., 293 Koushik, K., 169 Kreibig, U., 286, 289, 291 Kronvall, G., 162 Kuchler, S., 118 Kulkarni, V., 328, 329 Kumar, V. V., 113, 119 Kuntsche, J., 108 Kuo, Y. C., 119 Kurata, N., 152 Kuroda, S., 147–164 Kwiatkowski, D. H., 34 Kyle Vanderlick, T. J., 250, 275 L Lancaster, R. M., 32 Lapotko, D. O., 282, 302 Lasic, D. D., 33, 140, 149, 328 Lee, G. S., 118

Lee, M. K., 119 Lee, V. H. L., 179 Lee, W. J., 89 Leitz, A. J., 218 Lemmer, Y., 79–102 Lenghaus, K., 312 Lentz, B. R., 258 Levene, M. J., 310 Lewis, D. L., 344 Li, H. L., 116, 119 Li, S. D., 155 Li, X., 115, 117, 132 Li, Y., 216, 218 Li, Z., 236 Lian, J., 114, 119 Liang, H. P., 283, 284 Libchaber, A., 4, 20, 35, 36 Limozin, L., 33 Lin, C. P., 282, 294 Link, S., 132, 282, 283, 291–293 Lipowsky, R., 33 Liu, H., 115, 116, 119 Liu, J., 115, 118 Liu, K., 119 Liu, R., 225 Liu, W., 119 Lizana, L., 310, 318, 320, 322, 323 Lobovkina, T., 315 Lombardi Borgia, S., 112 Long, M. S., 317 Loo, T. W., 225 Lorusso, V., 148 Lovrein, R., 82 Lu, B., 119, 120 Lu, C., 6 Lu, D., 173 Lu, P., 225 Lu, X., 118 Luisi, P. L., 20 Luo, Y., 116 Lutkenhaus, J., 4, 5 Lv, Q., 115 M Macdonald, P. M., 195 MacDonald, R. C., 280, 282 Maeda, H., 148 Maemichi, H., 45 Mahajan, N., 328 Mandawgade, S. D., 118 Manjunath, K., 109–111, 119 Marcotte, I., 195 Maria, L. I., 174 Marin, V. L., 212, 218 Markman, M., 148, 159 Markstrom, M., 317 Markus, M., 174

359

Author Index

Marshall, E., 149 Martin, F. J., 149, 328 Martin, J. T., 168 Martins, S., 106, 116, 119 Matsumura, F., 33 Matsuura, T, 19–29 Matsuzaki, T., 147–164 Mayer, L. D., 59 Mayhew, E. G., 168 Mc Mahon, M. T., 194 McCreedy, B. J. Jr., 57 McDevitt, C. A., 224 Meister, G., 344 Menager, C., 195 Messersmith, P. B., 328 Meyer, R. K., 34 Mezzena, M., 118 Mi, L. Z., 218 Mikhailovsky, A., 279–304 Milligan, R. A., 37 Mimura, N., 34 Mina, A., 174 Mintzer, M. A., 344 Mirkin, C. A., 236, 240, 275 Miroux, B., 5 Mishra, B. K., 328 Misra, A., 167–187 Mitchison, T. J., 32 Mitragotri, S., 295 Miyamoto, H., 37 Miyata, H., 33, 37, 38, 42, 45, 49 Moghimi, S. M., 148, 149 Molony, L., 37 Monnard, P. A., 21 Morrissey, D. V., 344 Morrissey, J., 212, 218 Mueller, A., 280 Muguruma, M., 37 Mu¨ller, R. H., 106, 109, 111–116, 118, 119 Murtas, G., 20, 21 N Nagai, T., 5 Nagaoka, T., 151 Naito, M., 225 Namba, Y., 169 Nath, A., 212, 218 Negishi, M., 31–50 New, R. R. C., 58, 59, 169, 174 Newman, S. P., 168, 169 Newton, J. M., 168 Niidome, T., 286 Nir, S., 328 Niven, R. W., 169 Noble, C. O., 281 Noireaux, V., 4, 20, 21, 35, 36, 50 Nolkrantz, K., 329 Nomura, S. M., 20, 21

Norman, R. S., 281, 282 Nykypanchuk, D., 275 O Oberholzer, T., 20 Oku, N., 169 Olbrich, C., 109 Oldenburg, S. J., 283 Onyebujoh, P., 80 Orlowski, S., 223 Orwar, O., 309–323, 337 Osawa, M., 3–16 Otake, K., 170 Otto, J. J., 33 Ozoemena, K. I., 79–102 P Paasonen, L., 132 Palade, G., 56 Palade, G. E., 56 Paliwal, R., 119 Paliwal, S., 295, 301 Pan, J., 81 Pang, Y., 183, 184 Paolicelli, P., 119 Papahadjopoulos, D., 155, 328, 331 Papermaster, D. S., 223 Pardeike, J., 114 Park, S. H., 132, 195 Parmar, M., 172 Parr, M. J., 149 Parsegian, V. A., 59 Patel, G., 167–187 Patravale, V., 115, 118 Patrick, 174 Patton, J. S., 169 Pautot, S., 4, 35, 36 Pecha, R., 293, 301, 304 Peer, D., 281 Peter, B. J., 59 Petrova, H., 292 Pfeiffer, I., 234, 238, 250, 254, 275 Phayre, A. N., 329 Pichoff, S., 4, 5 Pilcher, L. A., 89 Pissuwan, D., 132 Pitsillides, C. M., 302, 304 Podesta, J. E., 343–353 Pojarova, M., 118 Pollard, T. D., 35, 48 Polli, J. W., 225, 226 Polovyanenko, D. N., 118 Ponce, A. M., 281, 302 Pontani, L.-L., 33, 35, 41, 50 Popinet, S., 293 Pople, P. V., 118 Portner, A., 56

360

Author Index

Prasad, V., 281, 291, 293, 294, 300, 304 Prevo, B. G., 281–284, 288, 291, 293, 294, 300, 302, 304 Priano, L., 119, 120 Prosser, R. S., 195, 197, 205 Puglia, C., 115, 118 Pupo, E., 21 Pysher, M. D., 329, 331 Q Qu, Q., 225 R Ramesh, R., 346 Rathman, J. F., 118 Reddy, L. H., 114 Reeves, J. P., 59 Regine, 174 Reischl, U., 80 Rezzani, R., 119 Rhodes, D. G., 168 Richardson, H. H., 132 Richardson, M., 148 Rigaud, J. L., 221, 222, 227 Ritchie, T. K., 211–228 Robinson, J. R., 179 Rodriguez, O. C., 32 Rohr, K., 236, 240, 242 Rongen, H. A. H., 328 Roper, D. K., 293 Rosenberg, M. F., 225 Rosenblatt, C., 328 Rosenthal, E., 148 Rosi, N. L., 240 Rossi, J. J., 343 Rossier, O., 338 Roux, A., 337, 338 Ruckmani, K., 114, 119 Ruktanonchai, U., 118 Runnicles, D. F., 87 Rustom, A., 328 S Sackmann, E., 33 Sagrera, A., 57 Sakagami, M., 183, 184, 186, 187 Salter, R., 328 Samanich, K. M., 80 Sanjula, B., 111, 119 Sarmento, B., 119 Savulescu, J., 149 Scalia, S., 118 Scheid, A., 57 Schleicher, G. K., 81, 95, 98, 100 Schroeder, T. E., 32 Schwartzberg, A. M., 283

Sengupta, S., 280, 282 Serpe, L., 119 Shah, K. A., 118 Shah, S. P., 171, 173, 174, 182 Shah, V. P., 179 Shahgaldian, P., 118 Shahiwala, A., 173, 182 Shankaran, D. R., 89 Sharma, P., 119 Shaw, A. W., 218 Shek, P. N., 169 Shi, W., 286 Shimizu, Y., 21, 22 Shnyrova, A. V., 55–73 Shum, H. C., 35, 36 Shum, P., 280 Shyianovskaya, I. V., 195 Silva, A. C., 118 Simanek, E. E., 344 Simo˜es, S., 281, 328 Singh, A., 328 Singh, K. K., 118 Singh, M., 167–187 Singh, S., 175 Sligar, S. G., 211–228 Solon, J., 56 Sott, K., 317, 320, 323, 329, 337 Sou, K., 297 Souto, E. B., 105–121 Spector, M. S., 328 Stachowiak, J. C., 35, 36 Stecova, J., 114, 118 Steinberg-Yfrach, G., 328 Stengel, G., 234, 250, 275 Stephenson, S. M., 347 Stewart, J. C. M., 253 Stoltz, A. C., 79–102 Stromberg, A., 314 Su, F. L., 119 Su, Y. C., 119 Sugawara, T., 20 Sun, Y. G., 281–284 Sunami, T., 19–29 Sungawara, T., 20 Sunkara, G., 169, 172 Suresh, G., 116, 120 Suzuki, H., 19–29 Svenson, S., 328 Swai, H. S., 79–102 Sweeney, L., 172, 173 Szebeni, J., 149 Szeto, T. H., 5 Szoka, F. Jr., 155, 331 Szostak, J. W., 20 T Takakura, K., 20 Takano, M., 225

361

Author Index

Takayama, M., 159 Takiguchi, K., 31–50 Takimoto, T., 56 Taly, V., 25 Tanaka-Takiguchi, Y., 31–50 Tanizawa, K., 147–164 Taton, T. A., 240 Tatulian, S., 86 Tausch, C., 114 Taylor, A. M., 224–227 Taylor, D. L., 32 Taylor, K. M. G., 168 Tecilla, P., 89 Teeranachaideekul, V, 115, 118 ten Bokum, A. M. C., 79–102 Terreno, E., 193–208 Thanyani, S. T., 79–102 Thomas, B. N., 328 Thomas, J. K., 254 Thurston, G., 345 Tian, J., 119 Tokarz, M., 317, 318, 320 Tong, L., 281, 282, 301, 303 Torchilin, V., 20, 21 Triplett, M. D., 118 Tronde, A., 183, 184 Tsumoto, K., 20, 21 Tsuruo, T., 225 Tsutsui, Y., 150, 152, 153 Tumpane, J., 266 Turkevich, J., 283, 284 Tuschl, T., 344 U Unruh, T., 108 Urban, S., 152 V Vaccaro, A. M., 59 van Wyngaardt, S., 79–102 van Ziji, P. C. M., 194 Venkateswarlu, V., 109–111, 119 Venter, L., 79–102 Verschoor, J. A., 79–102 Videira, M. A., 115 Vivek, K., 119 Vogel, S., 233–247 Volodkin, D. V., 132, 133 Vrey, P. J., 79–102 Vyas, S. P., 174 W Wachsstock, D. H., 34, 48 Walde, P., 20, 21 Walker, J. E., 5

Wang, D., 119 Wang, E. J., 225 Wang, J. J., 119 Wang, Y., 111, 118 Watkins, S. C., 328 Waugh, R., 337 Weissig, V., 20 Weissleder, R., 281 Welsch, S., 56 West, J., 328, 329, 337 Westwell, A. E., 87 Wetmur, J. G., 262 Whitehead, K. A., 344 Widom, J., 234, 237 Wiley, B., 283 Wissing, S., 115 Wolff, J. A., 344 Woods, M., 194 Woolfe, A. J., 171 Wu, G., 132, 279–304 Wu, W., 111, 118 X Xia, Q., 118 Xia, Y. N., 283, 284 Xiang, Q. Y., 118 Xie, X. S., 310 Xu, J., 34, 48 Xu, Z., 113, 119 Y Yamada, A., 31–50 Yamada, T., 149, 150, 152–154 Yamashiro, S., 34 Yamashiro-Matsumura, S., 33, 37 Yang, B., 328 Yang, L., 119 Yang, S. C., 111 Yao, C. P., 302 Ye, J., 114, 118 Ying, X. Y., 119 Yomo, T, 19–29 Yoshikawa, K., 31–50 Yoshimoto, N., 147–164 Yoshina-Ishii, C., 234 Yu, D., 149 Yu, W., 20–22 Yuan, H., 116, 118, 119 Z Zaleski, S., 293 Zalipsky, S., 297 Zana, R., 255 Zara, G. P., 110, 111 Zasadzinski, J. A., 279–304 Zhang, G. R., 234

362 Zhang, J. Q., 119 Zhang, S., 89, 90 Zhang, W., 111 Zhang, X., 119 Zhang, Y., 89, 346 Zhao, J., 212

Author Index

Zhou, J., 194 Zhou, S. F., 223 Zhu, R. R., 118 Zimmerberg, J., 55–73 Zimmermann, H., 328 Zolnerciks, J. K., 211–228

Subject Index

A Actin-cross-linking proteins, cell-sized liposome F-actin bundles, model of, 33 methods for, 36 natural swelling method, 34–35 coencapsulation of, 37–38 F-actin localization, 38 liposome membrane visualization, 38 preparation of, 37 spontaneous transfer method, actoHMM experimental setup, 38–39 F-actin visualization, 41 observation chamber, 39–40 oil containing phospholipids, 40 oil/water interface, 40–41 W/O droplets preparation and transfer, 41 a-Actinin, 33–34, 37–38 Administration routes, SLN formulations, 117 ophthalmic application, 115 pulmonary application, 115–116 residence time, 115 skin, 113–114 Air-interfaced culture (AIC), 184 Allophycocyanin (APC), 27 Amiloride hydrochloride (AH), NLDPFs differential scanning calorimetric graphs, 181 intratracheal instillation, 182–183 in vitro release profile, 182 particle size distribution, 176 solid-state characterization and residual water content, 175 spray-drying method, 172 Amphotericin B, 100–101 Antimycolic acid antibodies detection, liposomal biosensors. See Mycolic acid (MA) B Bacteriorhodopsin (bR), 218, 221–222 Bio-nanocapsule (BNC) anticancer drug delivery, 148–149 BNC-LP conjugates antibodies, 161–162 biotin, 163 retargeting, 152–153 DOX conjugates in vitro cytotoxic effects, 159–160 in vivo therapeutic effects, 161–162

preparation, 158–159 first-generation, 149–150 lipoplex conjugates DNA, 156–157 in vitro transfection, 156–157 in vivo transfection, 158 overexpression, S. cerevisiae column chromatography, 154–155 spheroplast method, 153 ultracentrifugation, 154 ZZ-BNC purification, column chromatography, 155 PEGylated LPs, 148–149 reticuloendothelial system (RES), 148 second-generation BNC-LP conjugates, preparation, 151 electroporation, 150–151 fusogenic activity, 151 human liver cells, 152 viral vectors, 149 Bionanotubule formation apparatus for, electric field application, 333 cell-size liposomes, gangliosides, 337–338 electric fields role, 329–330 features, 328 hydrodynamic flows, 338–339 in solution, electric fields fluorescence image, 334, 336–337 liposome preparation, 336 scanning electron microscopy (SEM), 334–336 liposomes preparation buffer preparation, 331 NBD-labeled liposomes, 331 Oregon green-labeled liposomes, 331–332 mechanical fission, vesicles, 337–338 pipette-aspiration protocols, 337 surface attachment experimental apparatus, 332 liposome attachment, 333 sample wells, 332 substrate coating, 333 techniques, 328–329 tubular extension, 333 tubular growth monitoring, 334 tubular quantitation, 334 Biotin, BNC-LP conjugates, 163 Bleaney’s constants, 199 BNC. See Bio-nanocapsule (BNC)

363

364

Subject Index

Brush border (BBMI), 34, 37, 45 Bulk magnetic susceptibility (BMS), 199–200 C 6-Carboxy fluorescein (CF), 295–297 Cationic liposome. See Short interfering RNA (siRNA), cationic liposome Cell-sized liposome, actin. See also Actincross-linking proteins, cell-sized liposome cross-linking proteins F-actin bundles, model of, 33 methods for, 36 natural swelling method, 34–35 preparation of, 37 spontaneous transfer method, 35 functions, 32 morphogenesis a-actinin, 43–44 actoHMM, 45, 47 BBMI, 45 F-actin, 45–46 fascin, 43 filamin, 43–45 G-actin, 42–43 mechanism, 46, 48–49 S-1, 46 CF. See 6-Carboxy fluorescein (CF) Chemical exchange saturation transfer (CEST), 194, 207 Column chromatography, BNCs purification, 154–155 D Dapsone, SEM image analysis, 180 Differential scanning calorimetry (DSC), NLDPFs, 176, 181 1,2-Dioleoyl-sn-glycero-3-[phospho-rac(1-glycerol)] (DOPG), 16, 59, 68 DIONEXÒ Acclaim C18, 246 Dithiothreitol (DTT), 317 DLS. See Dynamic light scattering (DLS) DNA-controlled assembly of liposomes chain length dependence, probe design lipid-modified macrocycles, 238–240 single palmitoyl-chain membrane anchor, 239 UV-monitored thermal denaturation data, 238 materials and techniques experimental procedure, 244–245 HPLC purification, 246–247 lipid-modified DNA conjugates synthesis, 245–246 POPC liposomes preparation, 245 transition temperatures measurement, 244 probe design chain length dependence, 237–240

double membrane anchor single DNA-probe design, 235–237 light scattering, 242–243 single membrane anchor dual-probe design, 236–237 solid nanoparticles, 234 thermal denaturation experiments label-free setup, 241 mismatch discrimination data, 242 ultraviolet (UV) spectroscopy, 240–241 Doxorubicin (DOX) conjugates, BNC in vitro cytotoxic effects, 159–160 in vivo therapeutic effects, 161–162 preparation, 158–159 Drug delivery systems (DDS), 132 Dry powder formulations (DPF), 168–169. See also Nanoliposomal DPFs (NLDPFs) Dry powder inhalers (DPIs), 168 DSC. See Differential scanning calorimetry (DSC), NLDPFs Dynamic light scattering (DLS) Au NPs, 137–138 soft hybrid nanostructures autocorrelation functions, 259–260 hydrodynamic thickness (Ho), 261 E Egg yolk phosphatidylcholine (EYPC) liposomes, 134, 140–141 ESPRIT SPR biosensor, MA cholesteroid nature, 100–101 cuvette and needles cleaning, 98 degassed buffers, 92–93 gold surface, 91 liposome preincubation, second serum exposure inhibition values, 97–98 P129 (TB-positive) serum, 96–97 MARTI-assay ELISA analysis, 99 sensorgrams, 98–99 octadecanethiol (ODT), 90–91 ODT-coated gold disks, regeneration, 98 PBS/AE, first serum exposure, 93–96 principle, 89 saponin concentration, 92–93 self-assembled monolayers (SAMs), 89–90 F F-actin, 33, 38, 41 Fascin, 33, 37–38, 43 Filamin, 34, 37–38, 43–45 Fluorescein diphosphate (FDP), 322 Fluorescence-activated cell sorting (FACS), 25–28 Fluorescence intensity, calcein release, 139

365

Subject Index

Fluorescent recovery after photobleaching (FRAP), 9–10 Formaldehyde, 317 Freeze-dried empty liposomes (FDEL) method, 21–23 FtsZ rings reconstitution. See Tubular liposomes, FtsZ rings reconstitution G G-actin, 42–43 Gel exclusion chromatography dextran Sephadex, 257 light scattering curve, 258 relative values, 258, 260 UV absorption spectra, 258–259 Giant liposomes, protein synthesis and RNA replication artificial cells, 20 FDEL method, 21–23 fluorescence-activated cell sorting (FACS) APC/R-PE, 27 data analysis, 27–28 green fluorescent products, 25 internal reaction, time course of, 25–27 PFB-fluorescein and CM-fluorescein, 27 b-glucuronidase, 22, 24 self-encoded replicase, 24–25 Giant unilamellar vesicles (GUVs) electroformation, 67 M protein interaction budding activity, 71 buffers and negative protein controls, 72 electrode configuration, 69–70 patch-clamp type experiment, 72 pipettes for, 70–71 preparation, 66–69 b-Glucuronidase, 22, 24 Gold nanoparticles (Au NPs), liposomes complexes calcein release, 137–139 DLS analysis, 137–138 drug carriers, 132 EYPC liposomes, 134 lipid components, optimization EYPC structures, 140–141 time-dependent UV-Vis spectra, 141–143 particle numbers, estimation, 139–140 schematic image, 133 TEM analysis, 136–137 time-dependent SPR, 134–135 types, 132–133 Gold nanoshells. See Hollow gold nanoshells (HGNs) Good manufacturing practice (GMP), 151–152 GUVs. See Giant unilamellar vesicles (GUVs)

H HccA dye, multilamellar liposomes, 9–10 Heavy meromyosin (HMM), 34, 45, 47. See also Actin-cross-linking proteins, cell-sized liposome HeLa cells, 352 Hepatitis B virus (HBV) vaccine, 149–150 High-pressure homogenization (HPH) process, 107 Hollow gold nanoshells (HGNs) coupling method, liposomes continuous-wave laser irradiation, 299 Cryo-TEM tilt series, tethering, 297–298 DPPC, 295 encapsulation of, 6-carboxy fluorescein (CF), 295–297 external HGNs, 298 pulsed laser optics, 298–299 dimensions, optimization of absorption efficiency, 287, 289 diameter, 288, 290 extinction cross section, 287, 289 extinction spectra, 289, 291 Mie scattering formalism, 286–288 scattering efficiency, 287 shell thickness, 287–288, 290 drug release, 280 femtosecond NIR laser pulses acoustic signal amplitude, 293, 295 extinction spectra, 291–292 heat dissipation, 293 hydrophone, 293 nonradiative relaxation processes, 292 SPR, 291 TEM images, collapse of, 294 heat-transfer analysis, 282 NIR, advantage of, 281–282 proximity effect, 303 synthesis of polyethylene glycol (PEG), 285–286 replacement reaction, 284 schematic illustration, 283 silver nanoparticles, 283–284 TEM images, 285 tetrachloroauric acid, 284 triggered liposome release comparison of, 300 energy density effect, 300–301 in situ fluorescence, kinetics of, 302 permeability of, 302 I IAsys biosensor cuvettes, MA disadvantages, 88 optimal concentration of, 87 PBS/AE, 87–88

366

Subject Index

Intermediate sized unilamellar vesicles (IUVs), M protein aqueous dye leakage, 65–66 bound protein, calculation, 64–65 ficoll gradient flotation/sedimentation methods, 61–62 formation and characterization, 58–59 multistep gradients, 62 preparation hydration method, extrusion procedure, 59–60 lipid loss, 61 protein binding, assay steps, 63 Rh fluorescence, 63 two-step flotation experiments, 62 Isolated perfused lung (IPL), 184–187 Isolated perfused rat lung (IPRL), 186 L Lanthanide-loaded paramagnetic liposomes chemical exchange saturation transfer (CEST), 194 experiments cryo-TEM image, 202–203 DINTRALIPO, 204–205 1 H NMR spectra, 200–201 integral values, intraliposomal water protons, 201–202 in vivo MR-CEST experiments, 207 physicochemical variables correlations, 203 Tm(III), 205–206 magnetic resonance imaging (MRI), 194 NMR characterization, nonspherical LPs BMS, 199–200 magnetic moments and Bleaney’s constants, 199 water protons, chemical shift, 198–199 osmotic shrinkage, 197–198 phospholipidic nanotubes, 195 shift reagents, 195–197 Leuprolide acetate (LA), 173 LIPEXTM, 245 Lipid-based drug delivery systems (LDDS), 106 Lipoplex conjugates, BNC DNA, 156–157 in vitro transfection, 156–157 in vivo transfection, 158 Liquid-covered culture (LCC), 184 M MA. See Mycolic acid (MA) Magnetic resonance imaging (MRI), 194, 207 Mean residence time (MRT), 111–112 Membrane budding reconstitution, unilamellar vesicles

GUVs electroformation, 67 M protein interaction, 69–72 preparation, 66–69 IUVs aqueous dyes leakage, 65–66 bound protein, calculation, 64–65 ficoll gradient flotation/sedimentation methods, 61–62 formation and characterization, 58–59 multistep gradients, 62 preparation, 59–61 protein binding, assay steps, 63 Rh fluorescence, 63 two-step flotation experiments, 62 matrix (M) protein purification, 57–58 viral budding, 56 Membrane scaffold protein (MSP), nanodiscs constructs, 212–214 expression, 216–217 purification, 217–218 MicroFilTM, 71 Mycolic acid (MA) ELISA, 81 ESPRIT SPR biosensor cholesteroid nature, 100–101 cuvette and needles cleaning, 98 degassed buffers, 92–93 gold surface, 91 liposome preincubation, second serum exposure, 96–98 MARTI-assay, 98–100 octadecanethiol (ODT), 90–91, 98 PBS/AE, first serum exposure, 93–96 principle, 89 saponin concentration, 92–94 self-assembled monolayers (SAMs), 89–90 fluorescent labeling, 82 IAsys biosensor cuvettes disadvantages, 88 optimal concentration of, 87 PBS/AE, 87–88 liposome capacity, determination of egg phosphatidylcholine, 83 triphase partition method, 82 liposome size cholesterol concentration, 84 fluorescence, 84–85 pH, 86 purification of, 81–82 tuberculosis (TB), 80 Mycolic acid antibody real-time inhibition (MARTI-assay) ELISA analysis, 99 sensorgrams, 98–99 Myosins, 48

367

Subject Index N Nanodiscs, membrane proteins reconstitution apolipoprotein A-1, 212 assembly mixture, 220 bacteriorhodopsin (bR) trimer, 218, 221–222 detergent, 218 elution profile, 222 empty disks, 219 membrane scaffold protein (MSP) constructs, 212–214 expression, 216–217 purification, 217–218 monomeric rhodopsin, assembly, 223 P-glycoprotein (P-gp) crystal structure, 223–224 functional activity, 226–228 incorporation, 225 N-Dodecyl-b-D-maltoside (DDM), 224 reconstitution mixture preparation, 220–221 structure and properties, 213, 215–216 Nanoliposomal DPFs (NLDPFs) in vivo studies animal models, 181–183 lung epithelial cell models, 183–184 lung tissue models, 184–187 physicochemical characterization animal models, 179, 182 assay, 174 differential scanning calorimetry (DSC), 176, 181 drug retention and stability studies, 175, 177–178 flow properties, 174 in vitro drug release study, 179, 182 in vitro lung deposition studies, 176, 179, 181 moisture content determination, 174–175 particle size, 175–176 reconstitution time and volume, 175 scanning electron microscopy (SEM), 176, 179–180 preparation characteristics, 169 flow chart, 171 freeze-drying, 172–173 spray-drying, 171–172 spray freeze-drying, 173 supercritical fluid technique (SCF), 169 ternary mixture, 170–171 Nanotube-vesicle networks (NVN) complexity and topology, 314–315 enzymatic reactions fluorescein diphosphate (FDP), 322 transport, 320–321 fabrication protocols experimental setup, 312 giant vesicles, 313 schematic representation, 310–311

internal and membrane functionalization dithiothreitol (DTT) and formaldehyde, 317 hydrogels, 317–318 red blood cells, 316 three vesicle network, 317 transport phenomena and mixing procedures diffusional transport, 319–320 electrophoresis, 320 Marangoni transport, 319–320 nanotube-mediated fusion, 318 Near infrared (NIR), 281–282. See also Hollow gold nanoshells (HGNs) Newcastle disease virus (NDV), 56 NLDPFs. See Nanoliposomal DPFs (NLDPFs) N-4-nitrobenz-2-oxa-1,3-diazole phosphatidic acid (NBD-PA), 331 Nuclear magnetic resonance (NMR), Ln(III) paramagnetic liposomes BMS, 199–200 magnetic moments and Bleaney’s constants, 199 water protons, chemical shift, 198–199 NVN. See Nanotube-vesicle networks (NVN) O Octadecanethiol (ODT), 90–91 Oregon GreenÒ , 331 P 1-Palmitoyl-2-oleoyl-sn-glycero-3phosphocholine (POPC), 59, 68, 213, 245, 252, 256, 263–264, 271, 274 PermatexÒ , 332 P-glycoprotein (P-gp), nanodiscs crystal structure, 223–224 functional activity ATPase activity, 226–227 MSP1E3D1 Nanodiscs, 227 incorporation, 225 N-Dodecyl-b-D-maltoside (DDM), 224 reconstitution, 225–226 Phosphate-buffered saline (PBS), 87–88, 93–96, 135–136, 138 Phosphatidylcholine (PC), 16 Poly(ethylene glycol) (PEG), 285–286, 346–348 Poly(dimethylsiloxane) (PDMS) sheet, 39–40 Protein synthesis. See Giant liposomes, protein synthesis and RNA replication R Replicase, 24–25 Reticuloendothelial system (RES), 111, 113, 148, 344 Rhodamine (Rh) fluorescence, 61 R-phycoerythrin (R-PE), 27

368

Subject Index S

Saccharomyces cerevisiae, BNC overexpression column chromatography, 154–155 spheroplast method, 153 ultracentrifugation, 154 ZZ-BNC purification, column chromatography, 155 Scanning electron microscopy (SEM) bionanotubule formation, 334–336 NLDPFs, 176, 179–180 Self-assembled monolayers (SAMs), 89–90 Short interfering RNA (siRNA), cationic liposome DOTAP/cholesterol, 345–346 DOTAP:cholesterol:DSPE-PEG2000, 346–347 experimental methods, 347 complexation studies, 350 encapsulation of, 349–350 encapsulation protocol, 351 gene silencing assessment, 352 in vitro gene silencing, 350 lipid film hydration method, 348 liposome/siRNA complex formation, 349 materials, 348 mixing protocol, 351 PEGylated liposome preparation, 348 size and surface charge of, 351 side effects, 344 therapeutic agent, criteria for, 344 troubleshooting, 353 Skin, drug administration, 113–114 SLN. See Solid lipid nanoparticles (SLN) formulations Soft hybrid nanostructures, oligonucleotides, 276 characterization dynamic light scattering (DLS) measurements, 256–261 size exclusion chromatography, 256–260 complementary oligonucleotides, in solution kinetics of, 263 noncovalent anchorage, 265 nucleation process, 262 rate constants for, 264 ssDNA concentration, 263 UV stopped-flow experiments, 262 DNA-directed self-assembly, 250 kinetics, 272–275 liposome preparation and lipid content determination, 252–253 materials, 251 oligonucleotides incorporation grafting density, 254–256 lipid anchor, 253–254 single-cholesterol and multiple-cholesterol oligonucleotide, 254–255 preparation procedures

single-step strategy, 266, 268–270 stepwise strategy, 270–272 self-assembled DNA nanostructures, 265–267 tetraethylene glycol (TEG), 250–251 uses, 250 Solid lipid nanoparticles (SLN) administration routes, biopharmaceutical aspects, 117 ophthalmic application, 115 pulmonary application, 115–116 residence time, 115 skin, 113–114 clinical pharmacology concentration fluctuations and therapeutic range, 121 dose size, 120 local effects and systemic distribution, 117–120 toxicological concentration, 121 degradation velocity, 106–107 modified release profile short/relatively long half-life drugs, 113 theoretical models, 112–113 tissue targeting, implantation, 113–114 pharmacokinetics and pharmacodynamics mean residence time (MRT), 111–112 modified release, 107 peroral administration route, 110 plasma concentration vs. time profiles, 110 polymorphism, 108 presystemic loss of, 112 TAGs crystallization, 108 temozolomide, 109 terminal half-life, 111 xenobiotics removal, 111 production of, 107 surfactants, 106 Spray-drying method, 172 Spray freeze-drying, 173 Surface plasmon resonance (SPR), 134–135. See also ESPRIT SPR biosensor, MA T Tacrolimus, NLDPFs in vitro aerosol deposition data, 181 stability data, 177–178 Temozolomide, 109 Tetraethylene glycol (TEG), 250–251 Transmission electron microscopy (TEM), Au NPs, 136–137 Tuberculosis (TB), 80 Tubular liposomes, FtsZ rings reconstitution bacterial expression of, 5 buffer changing, flow/perfusion, 11–12, 14

369

Subject Index

factors diameter, 15–16 FtsZ-YFP-mts behavior, 15 PC and DOPG, 16 shape of, 14–15 multilamellar liposomes permeability, 9–10 preparation, 7–9 purification of freeze-thaw method, 6 FtsZ-mts, 6 FtsZ-YFP-mts, 6–7 reagents, 4–5 renatured preparation, 7 technical problems, 4 utility of, 16 Z-ring formation, 11–13

U Ultracentrifugation, BNC purification, 154 Ultraviolet (UV) spectroscopy, 240–241 W Water-in-oil phospholipid-coated cell-sizeddroplets (W/O), 35 Y Yellow fluorescent protein (YFP), 4. See also Tubular liposomes, FtsZ rings reconstitution Z ZZ domain-displaying BNC (ZZ-BNC), 155

This page intentionally left blank

A

B

0.2

Au NP only

0.2

EYPC/Au = 1/10 1h

Abs

Abs

0h

0.1 4h 8h 12 h

Abs

C

500 600 700 Wavelength (nm)

800 D

0.2

0.1

Liposome only 0 400

1h 4h 8h 12 h

EYPC/Au = 1/1

500 600 700 Wavelength (nm)

0.4

800

EYPC/Au = 10/1

0.3 1 h→12 h Abs

0 400

4h 8h 12 h

0.1

0.2 0.1

0 400

Liposome only

Liposome only 500 600 700 Wavelength (nm)

800

0 400

500 600 700 Wavelength (nm)

800

Chie Kojima et al., Figure 7.2 Time-dependent UV–Vis spectra of the complex of EYPC liposomes (200 nm in diameter) with Au NPs at different ratios in PBS. The time-dependent UV–Vis spectra of Au NPs are only shown as a control (A). The mole ratio of EYPC to Au is 1/10 (B), 1/1 (C), and 10/1 (D). [Au] ¼ 46 mM (from Kojima et al., 2008).

A

B

0.4

0.4

Lipid/Au = 10/1 1 h→12 h

0.3 Abs

Abs

0.3

Lipid/Au = 10/1

0.2 0.1

0.2 0.1

Liposome only 0.0 400

Abs

D

0.1 0.05 0 400

800

F Lipid/Au = 1/10

0.2 0.15

4h 8h 12 h

800

Lipid/Au = 1/10

0.1

1h 4h 8h 12 h

0.05

0.05 Liposome only

0.0 400

500 600 700 Wavelength (nm)

1h

Abs

Abs

Liposome only

500 600 700 Wavelength (nm)

0.2

0.1

1h 4h 8h 12 h

0.15

Liposome only

0.15

800

Lipid/Au = 1/1

0.05

E

500 600 700 Wavelength (nm)

Lipid/Au = 1/1

1h 4h 8 h 12 h

0.1

0.0 400

Liposome only

0.2

0.2 0.15

0 400

800

Abs

C

500 600 700 Wavelength (nm)

1 h→12 h

500 600 700 Wavelength (nm)

Liposome only 800

0 400

500 600 700 Wavelength (nm)

800

Chie Kojima et al., Figure 7.7 Time-dependent UV–Vis spectra of the complex of (A, C, E) DDAB- or (B, D, F) PEG-bearing liposomes with Au NPs at different ratios. The mole ratio of lipid to Au is (A, B) 10/1, (C, D) 1/1, and (E, F) 1/10. [Au] ¼ 46 mM.

A

0.2 PEG-2000 0.15

Abs

1h 4h 0.1

8h 12 h

0.05

Liposome only 0 400

600

500

700

800

Wavelength (nm)

B 0.2

PEG-5000

Abs

0.15

1h 4h 8h 12 h

0.1

0.05 Liposome only 0 400

500

600 700 Wavelength (nm)

800

% Drug release

Chie Kojima et al., Figure 7.8 Time-dependent UV–Vis spectra of the complex of (A) PEG-2000- or (B) PEG-5000-bearing liposomes containing alkanethiol with Au NPs at the lipid/Au of 1/10. [Au] ¼ 46 mM.

100 90 80 70 60 50 40 30 20 10 0 0

2

SAHL

4

6 Time (h)

SLDPIM

8

SLDPIL

10

12

SLDPIS

Gaurang Patel et al., Figure 9.5 In vitro release profile of developed amiloride hydrochloride NLDPFs and plan drug DPFs (SAHL ¼ plain drug formulation, SLDPIM ¼ NLDPF containing mannitol, SLDPIL ¼ NLDPF containing lactose, and SLDPFs ¼ NLDPF containing sucrose). (Reprinted with permission from Chougule et al. (2006). Copyright # American Scientific Publishers, www.aspbs.com.)

Conventional liposome

ΔINTRALIPO = DHYP

(ΔBMS= 0)

B0 Non spherical liposome (magnetically orientable) ΔINTRALIPO = ΔHYP + ΔBMS

ΔχLIPO < 0

ΔχLIPO > 0

ΔBMS > 0

ΔBMS < 0

Paramagnetic amphiphilic metal complex (ΔχLnL < 0) Paramagnetic hydrophilic metal complex Paramagnetic amphiphilic metal complex (ΔχLnL > 0)

Silvio Aime et al., Scheme 10.2 Basic correlations among the physicochemical variables involved in the magnetic orientation and chemical shift of intraliposomal water protons for paramagnetic liposomes.

Liposome(s)

= Lipid membrane anchor

= DNA target

= DNA probe

Ulla Jakobsen and Stefan Vogel, Figure 12.1 Double membrane anchor single DNAprobe design. Schematic representation of liposome aggregation upon duplex formation between a lipophilic probe DNA and an unmodified target DNA. Liposomes and DNA strands are not drawn to scale.

Probe A Liposomes Target Probe B

= DNA probes

= DNA target

Ulla Jakobsen and Stefan Vogel, Figure 12.2 Single membrane anchor dual-probe design. Schematic representation of liposome aggregation upon duplex formation between a lipophilic probe DNA and an unmodified target DNA. Liposomes and DNA strands are not drawn to scale.

A 0.8

Abs

0.6

0.4

0.2

0.0 220

240

260 280 Wavelength (nm)

300

320

220

240

260 280 Wavelength (nm)

300

320

B 0.8

Abs

0.6

0.4

0.2

0.0

Martina Banchelli et al., Figure 13.2 UV absorption spectra of the different fractions from the gel separation experiment: system before the separation (black), first fraction (red), second fraction (green), third fraction (blue); (A) POPC vesicles (1.3 mM) and ON-Chol (4.1 mM) and (B) free ON-Chol as the control.

100% 90% 80% 70% 60%

Fraction 3 Fraction 2 Fraction 1

50% 40% 30% 20% 10% 0% Liposome + ON-Chol

ON-Chol

Martina Banchelli et al., Figure 13.3 Relative values of % UV absorbance at 260 nm for the filtered fractions with respect to the absorption values before filtration.

Name

Sequence

FA-2

5’-ACGAGCCTTTGACGCTTGGA-TT-TAGTGCGTAACATAGGCTAC-TTCTGAAATTATGATAAAGA-3’

E’F’

5’-ATTTACCTGGAAGCAGCCAC-TT-TCCAAGCGTCAAAGGCTCGT-3’

ED

5’-GTGGCTGCTTCCAGGTAAAT-TT-CACTATGTAACTGGTCTCTTA-3’

D’C’

5’-TAGAGACCAGTTACATAGTG-TT-TGACCTCAGTCGCAAGGCTG-3’

CB

5’-CAGCCTTGCGACTGAGGTCA-TT-TCGGGTCAACGAATGGCTGC-3’

B’A’

5’-GCAGCCATTCGTTGACCCGA-TT-GTAGCCTATGTTACGCACTA-3’

B’X

5’-GCAGCCATTCGTTGACCCGA-TT-CCCCCCCCCTTTTTTTTTTT-3’

Pseudo-hexagonal structure D

E D⬘

C

E⬘

C⬘

F⬘ B⬘

B

F

A⬘ 5⬘

A

2⬘ 3⬘ Open structure

2

A

F⬘

E⬘

D⬘

C⬘

B’

F

E

D

C

B

X

Martina Banchelli et al., Scheme 13.2 DNA nanostructure composing strands. Adapted from Baldelli Bombelli et al. (2008).

12

Efficiency

10 Scattering Absorption

8 6 4 2 0 400

600 800 Wavelength (nm)

1000

Guohui Wu et al., Figure 14.2 Theoretical absorption and scattering efficiency for HGN with a diameter of 33 nm and shell thickness of 1.75 nm in water.

14

12 300 10 200 8 100

Absorption cross-section (cm2)

Absorption efficiency

400 ⫻ 10−12

6 20

40

60 80 HGN diameter (nm)

100

Guohui Wu et al., Figure 14.4 Absorption efficiency (red) and cross section (blue) at 800 nm of HGNs with optimized geometry as a function of diameter.

1.4

Absorbance (a.u.)

1.2 1.0 0.8

Experiment Simulation

0.6 0.4 0.2 0.0 400

500

600 700 800 Wavelength (nm)

900

1000

Guohui Wu et al., Figure 14.6 Measured (red) and calculated (blue) extinction spectra of the experimental distribution of HGN diameters and shell thicknesses from Fig. 14.5.

2 NIR

Extinction (a.u.)

1.5

1

0.5

0

400

600 800 Wavelength (nm)

1000

Guohui Wu et al., Figure 14.7 Extinction spectra (normalized at 350 nm) of: (blue) HGN suspension; (red) HGN suspension laser-irradiated at a power density of 16.1 mJ/cm2 for 8 min (120 fs pulses at 1-kHz repetition rate). Below the peaks of extinction curves are photographs of the corresponding dispersions. The spectra are consistent with the collapse and annealing of the HGN into solid gold/silver alloy nanoparticles.

100

50

80

40

60

30

40

20

20

10

0

0 Au NS freely outside liposomes

80

40

60

30

40

20

20

10

0

Derivative

Dye release (%)

Au NS inside liposomes

0 0

2 4 6 8 10 12 14 Laser power density (mJ/cm2)

16

Guohui Wu et al., Figure 14.12 Effect of energy density on CR release from liposomes with HGNs encapsulated inside and suspended outside (black data points). The solid red curves are sigmoidal fits to the data; the blue curves are the derivative of the fit. The peak in the derivative gives the threshold energy for CF release, which is the same for both samples. The maximum release is quite different due to the different average distances between HGNs and the liposome bilayers.

A

B

+

C

−

Pi

Pi

VB

F = 1–10 pN

GUV

VA

MLV

ftube ~ 100 nm

fGUV ~ 10-100 mm D

Pi VC

E

F

VA VB

ΨB/A =

3rtube VB = VA 2rGUV(A)

Aldo Jesorka and Owe Orwar, Figure 15.1 Schematic representation of a fabrication strategy for complex nanofluidic networks composed of buffer filled giant unilamellar phospholipid vesicles (GUV) and nanotubes. (A) Penetration of a GUV with a glass injection pipette after electroporation with a DC pulse. The injection pipette is backfilled with a solution that can be the same as or differ from the liquid content of the vesicle, for example, a buffered enzyme solution, here indicated by a different color. Attached to the GUV is an MLV, which serves as a lipid reservoir for subsequent network building. (B) By retracting the needle from the vesicle, using a micromanipulation device, a nanotube is drawn from the GUV. (C) Injecting the content of the injection pipette with positive backpressure Pi leads to formation of a second, the so-called daughter vesicle, at the end of the nanotubes. (D) The procedure can be repeated with more differently backfilled injection pipettes, leading to a contentdifferentiated complex nanofluidic network. (E) For the creation of a new daughter vesicle, the ratio C of the two volumes involved is dependent only on the dimensions of the nanotubes connecting them and the size of the mother vesicle.

A

C

B

D

Aldo Jesorka and Owe Orwar, Figure 15.4 Examples of internal and membrane functionalization of NVNs. (A, B) Fluorescence micrograph of a four-vesicle network with reconstituted membrane proteins (fluorescently labeled AE1 ion channels). The center vesicle is also schematically represented. The lower two vesicles in this panel are additionally filled with small unilamellar vesicles, containing the same labeled protein. (C) Fluorescence micrograph of a three vesicle network with different fluorescent dyes in each container. (D) Compartmentalization of a GUV by means of a hydrogel derived from a thermoresponsive LCST polymer hydrogel. The diameter of a GUV in each panel is  25 mm. Reprinted in permission of the American Chemical Society.

+p

A

rGUV

Solvent B

Solvent A

2rtube

3 mm 5 %A 15 mm 1 %A 30 mm 0.5 %A

Membrane reservoir

B

C D1

D2

MV

2.56 1.92 1.28 0.64 0 (s)

D 10 mm

Aldo Jesorka and Owe Orwar, Figure 15.5 Mixing and diffusional transport in NVNs. (A) Schematic drawing of mixing driven by membrane flow. Solvent A (blue arrows) from the left-hand vesicle is transported by viscous drag through the interconnecting nanotube as the right-hand vesicle grows during injection of solvent B (green arrows). The red figures represent the increasing diameter of the inflated vesicle, the blue figures are the corresponding percentages of solvent A in that vesicle. Membrane material is supplied from a membrane reservoir (red arrows). (B) Fluorescence micrograph of a three vesicle system constructed by micropipette injection according to panel (A). Mixing in a chain of consecutively microinjected vesicles leads to a dilution series that is represented by the decreasing fluorescence of the fluorescein solution originating from the mother vesicle (MV), as it is increasingly diluted with the injected buffer solution in daughter vesicles D1 and D2. (C) Schematic representation of two vesicle network interconnected by a nanotube. The red enclosure is the area of the nanotube where diffusive transport of a 30 nm single latex particle has been followed over a period of 2.56 s. (D) Time-dependent positional information of the particle in the enclosed area from panel (C). The color represents the point in time the particle was detected in the particular position. Reprinted in permission of the American Chemical Society.

Josemar A. Castillo and Mark A. Hayes, Figure 16.5 Fluorescence image of a lipid tubule formed by applying an electric field (50 V/cm) to a 1:9 NBD-PA:PC liposome in solution. The approximate size of the structure is 5  50 mm (Hayes et al., 2007).