Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models [1 ed.] 1607614464, 9781607614463, 9781607614470

Citation preview

METHODS

IN

MOLECULAR BIOLOGY™

Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK

For other titles published in this series, go to www.springer.com/series/7651

Liposomes Methods and Protocols Volume 2: Biological Membrane Models

Edited by

Volkmar Weissig Department of Pharmaceutical Sciences, Midwestern University College of Pharmacy Glendale, Glendale, AZ, USA

Editor Volkmar Weissig Department of Pharmaceutical Sciences Midwestern University College of Pharmacy Glendale Glendale, AZ USA [email protected]

ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-60761-446-3 e-ISBN 978-1-60761-447-0 DOI 10.1007/978-1-60761-447-0 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2009933261 © Humana Press, a part of Springer Science+Business Media, LLC 2010 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Cover illustration: Background art is derived from Figure 6c in Chapter 21 Printed on acid-free paper Humana Press is a part of Springer Science+Business Media (www.springer.com)

Preface Efforts to describe and model the molecular structure of biological membranes go back to the beginning of the last century. In 1917, Langmuir described membranes as a layer of lipids one molecule thick [1]. Eight years later, Gorter and Grendel concluded from their studies that “the phospholipid molecules that formed the cell membrane were arranged in two layers to form a lipid bilayer” [2]. Danielli and Robertson proposed, in 1935, a model in which the bilayer of lipids is sequestered between two monolayers of unfolded proteins [3], and the currently still accepted fluid mosaic model was proposed by Singer and Nicolson in 1972 [4]. Among those landmarks of biomembrane history, a serendipitous observation made by Alex Bangham during the early 1960s deserves undoubtedly a special place. His finding that exposure of dry phospholipids to an excess of water gives rise to lamellar structures [5] has opened versatile experimental access to studying the biophysics and biochemistry of biological phospholipid membranes. Although during the following 4 decades biological membrane models have grown in complexity and functionality [6], liposomes are, besides supported bilayers, membrane nanodiscs, and hybrid membranes, still an indisputably important tool for membrane biophysicists and biochemists. In vol. II of this book, the reader will find detailed methods for the use of liposomes in studying a variety of biochemical and biophysical membrane phenomena concomitant with chapters describing a great palette of state-of-the-art analytical technologies. Moreover, besides providing membrane biophysicists and biochemists with an immeasurably valuable experimental tool, Alex Bangham’s discovery has triggered the launch of an entirely new subdiscipline in pharmaceutical science and technology. His observation that the lamellar structures formed by phospholipids exposed to aqueous buffers are able to sequester small molecules has lead to the development of the colloidal drug delivery concept. Following initial studies of enzyme encapsulation in liposomes as an approach towards the treatment of storage diseases [7, 8], a few years later in two New England Journal of Medicine landmark papers, Gregory Gregoriadis outlined the huge carrier potential of liposomes in biology and medicine [9, 10]. The following 2 decades saw immense efforts in academia and in soon-to-be-founded start-up companies to turn Gregoriadis’ vision into clinical reality. These 20 years of intense work in liposome laboratories around the world finally culminated with the FDA (USA) approval of the first injectable liposomal drug, Doxil, in February of 1995. Today, liposomes present the prototype of all nanoscale drug delivery vectors currently under development. Lessons learned in the history of over 40 years of Liposome Technology should be heeded by new investigators in the emerging field of pharmaceutical and biomedical nanotechnology. Volume I of this book is dedicated to state-of-the-art aspects of developing liposome-based pharmaceutical nanocarriers. All chapters were written by leading experts in their particular fields, and I am extremely grateful to them for having spent parts of their valuable time to contribute to this book. It is my hope that together we have succeeded in providing an essential source of practical

v

vi

Preface

know-how for every investigator, young and seasoned ones alike, whose research area involves in one way or another phospholipids, glycolipids, and cholesterol. Last but not least, I would like to thank John Walker, the series editor of “Methods in Molecular Biology,” for having invited me to assemble this book and above all for his unlimited guidance and help throughout the whole process. Glendale, AZ

Volkmar Weissig

References 1. Bangham AD, Standish MM, Watkins JC (1965) Diffusion of univalent ions across the lamellae of swollen phospholipids. J Mol Biol 13(1):238–252 2. Chan YH, Boxer SG (2007) Model membrane systems and their applications. Curr Opin Chem Biol 11(6):581–587 3. Danielli JF, Davson H (1935) A contribution to the theory of permeability of thin films. J Cell Comp Physiol 5:495–508 4. Gorter E, Grendel F (1925) On bimolecular layers of lipoids on the chromocytes of the blood. J Exp Med 41:439–443 5. Gregoriadis G (1976) The carrier potential of liposomes in biology and medicine (second of two parts). N Engl J Med 295(14):765–770

6. Gregoriadis G (1976) The carrier potential of liposomes in biology and medicine (first of two parts). N Engl J Med 295(13):704–710 7. Gregoriadis G, Ryman BE (1971) Liposomes as carriers of enzymes or drugs: a new approach to the treatment of storage diseases. Biochem J 124(5):58P 8. Gregoriadis G, Leathwood PD, Ryman BE (1971) Enzyme entrapment in liposomes, FEBS Lett 14(2):95–99 9. Langmuir I (1917) The constitution and structural properties of solids and liquids. II. Liquids. J Am Chem Soc 39:1848–1906 10. Singer SJ, Nicolson GL (1972) The fluid mosaic model of the structure of cell membranes. Science 175(23):720–731

Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v xi

1 Utilization of Liposomes for Studying Drug Transfer and Uptake . . . . . . . . . . . . . Alfred Fahr and Xiangli Liu 2 The Use of Liposomes in the Study of Drug Metabolism: A Method to Incorporate the Enzymes of the Cytochrome P450 Monooxygenase System into Phospholipid, Bilayer Vesicles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . James R. Reed 3 Use of Liposomes to Study Cellular Osmosensors . . . . . . . . . . . . . . . . . . . . . . . . . Reinhard Krämer, Sascha Nicklisch, and Vera Ott 4 Studying Mechanosensitive Ion Channels Using Liposomes . . . . . . . . . . . . . . . . . Boris Martinac, Paul R. Rohde, Andrew R. Battle, Evgeny Petrov, Prithwish Pal, Alexander Fook Weng Foo, Valeria Vásquez, Thuan Huynh, and Anna Kloda 5 Studying Amino Acid Transport Using Liposomes . . . . . . . . . . . . . . . . . . . . . . . . Cesare Indiveri 6 Use of Liposomes for Studying Interactions of Soluble Proteins with Cellular Membranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chris T. Höfer, Andreas Herrmann, and Peter Müller 7 Liposomal Reconstitution of Monotopic Integral Membrane Proteins. . . . . . . . . . Zahra MirAfzali and David L. DeWitt 8 The Reconstitution of Actin Polymerization on Liposomes . . . . . . . . . . . . . . . . . . Mark Stamnes and Weidong Xu 9 Electroformation of Giant Unilamellar Vesicles from Native Membranes and Organic Lipid Mixtures for the Study of Lipid Domains under Physiological Ionic-Strength Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . L.-Ruth Montes, Hasna Ahyayauch, Maitane Ibarguren, Jesus Sot, Alicia Alonso, Luis A. Bagatolli, and Felix M. Goñi 10 Visualization of Lipid Domain-Specific Protein Sorting in Giant Unilamellar Vesicles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Martin Stöckl, Jörg Nikolaus, and Andreas Herrmann 11 Biosynthesis of Proteins Inside Liposomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Pasquale Stano, Yutetsu Kuruma, Tereza Pereira de Souza, and Pier Luigi Luisi 12 Study of Respiratory Cytochromes in Liposomes. . . . . . . . . . . . . . . . . . . . . . . . . . Iseli L. Nantes, Cintia Kawai, Felipe S. Pessoto, and Katia C.U. Mugnol 13 Use of Liposomes to Evaluate the Role of Membrane Interactions on Antioxidant Activity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Salette Reis, Marlene Lúcio, Marcela Segundo, and José L.F.C. Lima

1

vii

11 21 31

55

69 83 95

105

115 127

147

167

viii

Contents

14 Studying Colloidal Aggregation Using Liposomes. . . . . . . . . . . . . . . . . . . . . . . . . Juan Sabín, Gerardo Prieto, and Félix Sarmiento 15 Assessment of Liposome–Cell Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jan A.A.M. Kamps 16 Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nejat Düzgünes¸, Henrique Faneca, and Maria C. Pedroso de Lima 17 The Use of Isothermal Titration Calorimetry to Study Multidrug Transport Proteins in Liposomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . David Miller and Paula J. Booth 18 Studying Lipid Organization in Biological Membranes Using Liposomes and EPR Spin Labeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Witold K. Subczynski, Marija Raguz, and Justyna Widomska 19 Membrane Translocation Assayed by Fluorescence Spectroscopy . . . . . . . . . . . . . . J. Broecker and S. Keller 20 Interaction of Lipids and Ligands with Nicotinic Acetylcholine Receptor Vesicles Assessed by Electron Paramagnetic Resonance Spectroscopy. . . . . . . . . . . Hugo Rubén Arias 21 Environmental Scanning Electron Microscope Imaging of Vesicle Systems . . . . . . Yvonne Perrie, Habib Ali, Daniel J. Kirby, Afzal U.R. Mohammed, Sarah E. McNeil, and Anil Vangala 22 Freeze-Fracture Electron Microscopy on Domains in Lipid Mono- and Bilayer on Nano-Resolution Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . Brigitte Papahadjopoulos-Sternberg 23 Atomic Force Microscopy for the Characterization of Proteoliposomes . . . . . . . . . Johannes Sitterberg, Maria Manuela Gaspar, Carsten Ehrhardt, and Udo Bakowsky 24 Method of Simultaneous Analysis of Liposome Components Using HPTLC/FID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sophia Hatziantoniou and Costas Demetzos 25 Viscometric Analysis of DNA-Lipid Complexes. . . . . . . . . . . . . . . . . . . . . . . . . . . Sadao Hirota and Nejat Düzgünes¸ 26 Fluorometric Analysis of Individual Cationic Lipid-DNA Complexes. . . . . . . . . . . Edwin Pozharski 27 Fluorescence Resonance Energy Transfer-Based Analysis of Lipoplexes . . . . . . . . . Edwin Pozharski 28 Analysis of Lipoplex Structure and Lipid Phase Changes . . . . . . . . . . . . . . . . . . . . Rumiana Koynova 29 Fluorescence Methods for Evaluating Lipoplex-Mediated Gene Delivery. . . . . . . . Henrique Faneca, Nejat Düzgünes‚ , and Maria C. Pedroso de Lima 30 FRET Imaging of Cells Transfected with siRNA/Liposome Complexes . . . . . . . . Il-Han Kim, Anne Järve, Markus Hirsch, Roger Fischer, Michael F. Trendelenburg, Ulrich Massing, Karl Rohr, and Mark Helm 31 Spectral Bio-Imaging and Confocal Imaging of the Intracellular Distribution of Lipoplexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sebastian Schneider and Regine Süss

189 199

209

233

247 271

291 319

333 351

363 369 385 393 399 425 439

457

Contents

32 Techniques for Loading Technetium-99m and Rhenium-186/188 Radionuclides into Pre-formed Liposomes for Diagnostic Imaging and Radionuclide Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Beth Goins, Ande Bao, and William T. Phillips 33 Fluorescence Correlation Spectroscopy for the Study of Membrane Dynamics and Organization in Giant Unilamellar Vesicles . . . . . . . . . . . . . . . . . . . Ana J. García-Sáez, Dolores C. Carrer, and Petra Schwille 34 Liposome Biodistribution via Europium Complexes . . . . . . . . . . . . . . . . . . . . . . . Nathalie Mignet and Daniel Scherman 35 Biosensor-Based Evaluation of Liposomal Binding Behavior . . . . . . . . . . . . . . . . . Gerd Bendas 36 Use of Liposomes to Study Vesicular Transport. . . . . . . . . . . . . . . . . . . . . . . . . . . Kohji Takei, Hiroshi Yamada, and Tadashi Abe Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ix

469

493 509 519 531 543

Contributors TADASHI ABE • Department of Neuroscience, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama, Japan HASNA AHYAYAUCH • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain HABIB ALI • School of Life and Health Sciences, Aston University, Birmingham, UK ALICIA ALONSO • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain HUGO RUBÉN ARIAS • Department of Pharmaceutical Sciences, College of Pharmacy, Midwestern University, Glendale, AZ, USA LUIS A. BAGATOLLI • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain UDO BAKOWSKY • Department of Pharmaceutical Technology and Biopharmacy, Philipps-Universität Marburg, Marburg, Germany ANDE BAO • Department of Radiology, University of Texas Health Science Center, San Antonio, TX, USA ANDREW R. BATTLE • Molecular Biophysics Laboratory, School of Biomedical Sciences and Institute for Molecular Bioscience, The University of Queensland, Brisbane, QLD, Australia GERD BENDAS • Department of Pharmacy, Rheinische Friedrich Wilhelms University Bonn, Bonn, Germany PAULA J. BOOTH • Department of Biochemistry, University of Bristol, Bristol, UK JANA BROECKER • Leibniz Institute of Molecular Pharmacology FMP, Berlin, Germany DOLORES C. CARRER • BIOTEC, Technische Universität Dresden, Dresden, Germany COSTAS DEMETZOS • Department of Pharmaceutical Technology, School of Pharmacy, University of Athens, Athens, Greece DAVID L. DEWITT • Department of Biochemistry and Molecular Biology, Michigan State University, East Lansing, MI, USA NEJAT DÜZGÜNES¸ • Department of Microbiology, Arthur A. Dugoni School of Dentistry, University of the Pacific, San Francisco, CA, USA CARSTEN EHRHARDT • School of Pharmacy and Pharmaceutical Sciences, University of Dublin, Trinity College Dublin, Dublin, Ireland ALFRED FAHR • Department of Pharmaceutics, Friedrich-Schiller-University, Jena, Germany HENRIQUE FANECA • Faculty of Science and Technology, Center for Neuroscience and Cell Biology, University of Coimbra, Coimbra, Portugal ROGER FISCHER • German Cancer Research Center (DKFZ), Heidelberg, Germany ALEXANDER FOOK WENG FOO • Molecular Biophysics Laboratory, School of Biomedical Sciences and Institute for Molecular Bioscience, The University of Queensland, Brisbane, QLD, Australia ANA J. GARCÍA-SÁEZ • BIOTEC, Technische Universität Dresden, Dresden, Germany MARIA MANUELA GASPAR • Unidade Novas Formas de Agentes Bioactivos, iMed, Faculdade de Farmácia, Universidade de Lisboa, Lisboa, Portugal

xi

xii

Contributors

BETH GOINS • Department of Radiology, University of Texas Health Science Center, San Antonio, TX, USA FELIX M. GOÑI • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain SOPHIA HATZIANTONIOU • Department of Pharmaceutical Technology, School of Pharmacy, University of Athens, Athens, Greece MARK HELM • Department of Chemistry, Institute of Pharmacy and Molecular Biotechnology, University of Heidelberg, Heidelberg, Germany ANDREAS HERRMANN • Mathematisch-Naturwissenschaftliche Fakultät I, Institut für Biologie/Biophysik, Humboldt Universität zu Berlin, Berlin, Germany SADAO HIROTA • Department of Material Science, School of Engineering, Tokyo Denki University, Tokyo, Japan MARKUS HIRSCH • Department of Chemistry, Institute of Pharmacy and Molecular Biotechnology, University of Heidelberg, Heidelberg, Germany CHRIS HÖFER • Institut für Biologie/Biophysik, Humboldt Universität zu Berlin, Berlin, Germany THUAN HUYNH • Molecular Biophysics Laboratory, School of Biomedical Sciences, The University of Queensland, Brisbane, QLD, Australia MAITANE IBARGUREN • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain CESARE INDIVERI • Dipartimento di Biologia Cellulare, Università della Calabria, Arcavacata di Rende, CS, Italy ANNE JÄRVE • Department of Chemistry, Institute of Pharmacy and Molecular Biotechnology, University of Heidelberg, Heidelberg, Germany JAN A.A.M. KAMPS • Laboratory for Endothelial Biomedicine & Vascular Drug Targeting Research, Medical Biology Section, Department Pathology & Medical Biology, University Medical Center Groningen, Groningen, The Netherlands CINTIA KAWAI • Centro Interdisciplinar de Investigação Bioquímica CIIB, Universidade de Mogi das Cruzes, Mogi das Cruzes, S.P., Brazil SANDRO KELLER • Leibniz Institute of Molecular Pharmacology FMP, Berlin, Germany IL-HAN KIM • Department of Bioinformatics and Functional Genomics, German Cancer Research Center (DKFZ), Institute of Pharmacy and Molecular Biotechnology, University of Heidelberg, Heidelberg, Germany DANIEL J. KIRBY • School of Life and Health Sciences, Aston University, Birmingham, UK ANNA KLODA • Molecular Biophysics Laboratory, School of Biomedical Sciences, The University of Queensland, Brisbane, QLD, Australia RUMIANA KOYNOVA • Northwestern University, Evanston, IL, USA REINHARD KRÄMER • Institute of Biochemistry, University of Cologne, Cologne, Germany YUTETSU KURUMA • “Enrico Fermi” Study and Research Center, Rome, Italy JOSÉ L.F.C. LIMA • REQUIMTE, Faculdade de Farmácia, Universidade do Porto, Porto, Portugal XIANGLI LIU • Department of Pharmaceutics, Friedrich-Schiller-University, Jena, Germany MARLENE LÚCIO • REQUIMTE, Faculdade de Farmácia, Universidade do Porto, Porto, Portugal PIER LUIGI LUISI • Biology Department, University of RomaTre, Rome, Italy

Contributors

xiii

BORIS MARTINAC • Molecular Biophysics Laboratory, School of Biomedical Sciences, The University of Queensland, Brisbane, QLD, Australia ULRICH MASSING • Department of Clinical Research, Tumor Biology Center, Freiburg, Germany SARAH E. MCNEIL • School of Life and Health Sciences, Aston University, Birmingham, UK NATHALIE MIGNET • Unité de Pharmacologie Chimique et Génétique; CNRS, UMR 8151, Paris, France; Inserm, U 640, Paris, France; Faculté des Sciences Pharmaceutiques et Biologiques, Université Paris Descartes, Paris, France; ENSCP, Paris, France DAVID MILLER • Department of Biochemistry, University of Bristol, Bristol, UK ZAHRA MIRAFZALI • Encapsula NanoSciences LLC, 441 Donelson, Pike, Suite 345, Nashville, TN 37214, USA AFZAL U. R. MOHAMMED • School of Life and Health Sciences, Aston University, Birmingham, UK L.-RUTH MONTES • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain KATIA C.U. MUGNOL • Centro Interdisciplinar de Investigação Bioquímica CIIB, Universidade de Mogi das Cruzes, Mogi das Cruzes, S.P., Brazil PETER MÜLLER • Institut für Biologie/Biophysik, Humboldt Universität zu Berlin, Berlin, Germany ISELI L. NANTES • Centro Interdisciplinar de Investigação Bioquímica CIIB, Universidade de Mogi das Cruzes, Mogi das Cruzes, S.P., Brazil SASCHA NICKLISCH • Institute of Biochemistry, University of Cologne, Cologne, Germany JÖRG NIKOLAUS • Mathematisch-Naturwissenschaftliche Fakultät I, Institut für Biologie/Biophysik, Humboldt-Universität zu Berlin, Berlin, Germany VERA OTT • Institute of Biochemistry, University of Cologne, Cologne, Germany PRITHWISH PAL • Molecular Biophysics Laboratory, School of Biomedical Sciences, The University of Queensland, Brisbane, QLD, Australia BRIGITTE PAPAHADJOPOULOS-STERNBERG • NanoAnalytical Laboratory, San Francisco, CA, USA MARIA C. PEDROSO DE LIMA • Department of Biochemistry, Faculty of Science and Technology, Center for Neuroscience and Cell Biology, University of Coimbra, Coimbra, Portugal TEREZA PEREIRA DE SOUZA • Biology Department, University of RomaTre, Rome, Italy YVONNE PERRIE • School of Life and Health Sciences, Aston University, Birmingham, UK REGINE SÜSS • Department of Pharmaceutical Technology and Biopharmacy, AlbertLudwigs University, Freiburg, Germany FELIPE S. PESSOTO • Centro Interdisciplinar de Investigação Bioquímica CIIB, Universidade de Mogi das Cruzes, Mogi das Cruzes, S.P., Brazil EVGENY PETROV • Molecular Biophysics Laboratory, School of Biomedical Sciences, The University of Queensland, Brisbane, QLD, Australia WILLIAM T. PHILLIPS • Department of Radiology, University of Texas Health Science Center, San Antonio, TX, USA EDWIN POZHARSKI • Department of Pharmaceutical Sciences, University of Maryland, Baltimore, MD, USA

xiv

Contributors

GERARDO PRIETO • Biophysics and Interfaces Group, Department of Applied Physics, Faculty of Physics, University of Santiago de Compostela, Santiago de Compostela, Spain MARIJA RAGUZ • Department of Biophysics, Medical College of Wisconsin, Milwaukee, WI, USA JAMES R. REED • Department of Pharmacology, Louisiana State University Health Science Center, New Orleans, LA, USA SALETTE REIS • REQUIMTE, Faculdade de Farmácia, Universidade do Porto, Porto, Portugal PAUL R. ROHDE • Molecular Biophysics Laboratory, School of Biomedical Sciences, The University of Queensland, Brisbane, QLD, Australia KARL ROHR • Department of Bioinformatics and Functional Genomics, German Cancer Research Center (DKFZ), Institute of Pharmacy and Molecular Biotechnology, University of Heidelberg, Heidelberg, Germany JUAN SABÍN • Biophysics and Interfaces Group, Department of Applied Physics, Faculty of Physics, University of Santiago de Compostela, Santiago de Compostela, Spain FÉLIX SARMIENTO • Biophysics and Interfaces Group, Department of Applied Physics, Faculty of Physics, University of Santiago de Compostela, Santiago de Compostela, Spain DANIEL SCHERMAN • Unité de Pharmacologie Chimique et Génétique; CNRS, UMR 8151, Paris, France; Inserm, U 640, Paris, France; Faculté des Sciences Pharmaceutiques et Biologiques, Université Paris Descartes, Paris, France; ENSCP, Paris, France SEBASTIAN SCHNEIDER • Department of Pharmaceutical Technology and Biopharmacy, Albert-Ludwigs University, Freiburg, Germany PETRA SCHWILLE • BIOTEC, Technische Universität Dresden, Dresden, Germany MARCELA SEGUNDO • REQUIMTE, Faculdade de Farmácia, Universidade do Porto, Porto, Portugal JOHANNES SITTERBERG • Department of Pharmaceutical Technology and Biopharmacy, Philipps-Universität Marburg, Marburg, Germany JESUS SOT • Unidad de Biofisica (CSIC-UPV/EHU), Leioa, Spain MARK STAMNES • Department of Molecular Physiology & Biophysics, Roy J. and Lucille A. Carver College of Medicine, University of Iowa, Iowa City, IA, USA PASQUALE STANO • Biology Department, University of RomaTre, Rome, Italy MARTIN STÖCKL • Mathematisch-Naturwissenschaftliche Fakultät I, Institut für Biologie/Biophysik, Humboldt-Universität zu Berlin, Berlin, Germany WITOLD K. SUBCZYNSKI • Department of Biophysics, Medical College of Wisconsin, Milwaukee, WI, USA KOHJI TAKEI • Department of Neuroscience, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama, Japan MICHAEL F. TRENDELENBURG • German Cancer Research Center (DKFZ), Heidelberg, Germany

Contributors

xv

ANIL VANGALA • School of Pharmacy and Chemistry, Kingston University, London, UK VALERIA VÁSQUEZ • Biochemistry Department, Gordon Center for Integrative Science, The University of Chicago, Chicago, IL, USA JUSTYNA WIDOMSKA • Department of Plant Physiology and Biochemistry, Faculty of Biochemistry, Biophysics and Biotechnology, Jagiellonian University, Krakow, Poland WEIDONG XU • Department of Molecular Physiology & Biophysics, Roy J. and Lucille A. Carver College of Medicine, University of Iowa, Iowa City, IA, USA HIROSHI YAMADA • Department of Neuroscience, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama, Japan

Chapter 1 Utilization of Liposomes for Studying Drug Transfer and Uptake Alfred Fahr and Xiangli Liu Abstract On entry into the body of the patient, drugs have to overcome many barriers in order to reach the target. The knowledge of the ability of drugs to cross these barriers, which mostly consist of lipid membranes, is of utmost interest in pharmacy. High values of lipophilicity of a drug might be a good pre-requisite for crossing these barriers. It also led liposomologists to think that highly lipophilic drugs may “stick” in the lipophilic interior of liposomal phospholipid membranes and therefore these liposomes may act as a retard formulation of the lipophilic drug. The presented method here estimates the transfer time of lipophilic drugs between liposomal lipid bilayers. This may help to judge the presumed retardation function of a specific liposomal delivery system for a chosen lipophilic drug. Key words: Membrane transfer, Liposome drug delivery system, Lipophilic drug, Retardation, Mini column method

1. Introduction In the pharmaceutical field, liposomes are commonly known as drug carrier systems, but appeared at first in the scientific community as models of biological membranes (1). Similar to the early use in membrane biophysics, liposomes can also be very useful in pharmaceutical sciences, as the passive diffusion of drugs through physiological barrier membranes (e.g. epithelial cells in the gut) can be estimated with their help by different kinds of experimental setups (2). Another drug-delivery related problem can also be assessed by applying membrane biophysics methods to pharmaceutical V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_1, © Humana Press, a part of Springer Science+Business Media, LLC 2010

1

2

Fahr and Liu

liposomology: the often not verified, but very often cited assumption that a lipophilic drug should bind to a high degree to the lipophilic interior of the liposomal bilayer simply because of its similarities in lipophilicity. Rather, as we had found for several lipophilic compounds, these compounds are transferred quite rapidly to other lipophilic binding places (3), which can also be reasoned by theoretical considerations (4). Methods like dialysis (5) have their value for estimating the thermodynamics of the transfer processes, but for analysing the kinetics in the minute time-range with absolute mass transfer values, other methods should be used. Not for all drugs the elegant method of spin-label measurements (6) can be used to get an estimate of drug retardation. Here we provide a protocol for measuring the transfer of lipophilic drugs between liposomal membranes. The drug is transferred during a mixing process from donor liposomes to acceptor liposomes; the donor liposomes are removed from the sample using a micro ion exchange column method. We describe this method under the assumption, that the drugs are available with a radioactive marker. This very comfortable situation regarding the transfer analysis is typically present in big pharmaceutical companies with their own radiochemistry labs ready for labelling the drugs of interest. A chromophoric group in the drug moiety helps as well in making the analysis simple, especially if this leads to a fluorescence drug. In the author’s experience, a sensitive HPLC-method can also be a sufficient method for analysis, but it may be time-consuming for reliable results. The obtained data can in most cases be easily analysed. By incorporating an un-exchangeable marker, one can not only determine the retention of donor liposomes on the column, but can also validate the whole method.

2. Materials 2.1. Substances

1. Egg phosphatidylcholine (EPC) (Lipoid KG, Ludwigshafen, Germany). Synthetic phosphatidylcholines (e.g. from Avanti Polar Lipids, Alabaster, AL, USA.) are highly recommended. 2. Dicetylphosphate (DCP) (Sigma-Aldrich, Taufkirchen, Germany). 3. Cholesterol and 1-palmitoyl-1,2-oleoyl phosphatidylcholine (POPC) (Genzyme, Liestal, Switzerland or Avanti Polar Lipids). 4.

C-labelled cholesteryl-oleoyl-ether and 3H-labelled cholesteryl-oleate (GE Healthcare UK Ltd, Buckinghamshire, UK). The lipids were stored in –20°C before use. Typically small aliquots of a delivered batch are stored in the freezer to avoid degradation caused by repeated thaw and freeze.

14

Utilization of Liposomes for Studying Drug Transfer and Uptake

3

5. The 14C labelled cyclosporin A (CyA) was synthesised and kindly provided by the radiochemistry group of Sandoz Pharm Ltd., now Novartis Inc. (Basel, Switzerland). 1–5 are either bought as chloroformic solution or the obtained powder is dissolved in chloroform in order to make stock solutions. These stock solutions should be kept at −20 to −80°C and should not be used after three months of storage. Always allow the frozen stock solution to reach ambient temperature in order to avoid water contamination of the solution which leads to lipid degradation. 6. DEAE Sepharose CL-6B (GE Healthcare Bio-Sciences AB, Uppsala, Sweden) is supplied pre-swollen in 20% ethanol. 7. Tris, saccharose, sodium chloride, and sodium azide (all in the highest purity available) (e.g. Sigma-Aldrich, Taufkirchen, Germany). 8. The two buffers used throughout the experiments are named Buffer A: 145 mM NaCl, 10 mM Tris pH 7.4, and Buffer B: 290 mM Saccharose, 10 mM Trizma pH 7.4, 0.02 % sodium azide. 2.2. Devices

1. Minicolumns were made in the workshop of the University of Perspex® and should preferably have the dimensions as indicated in Fig. 1. 2. All other devices should be available in a standard laboratory for liposomal research and are mentioned in the following text for reference.

3. Methods 3.1. Preparation of Donor Liposomes

1. Egg phosphatidylcholine (EPC), Dicetylphosphate (DCP), and cholesterol are weighed into individual test tubes and dissolved in chloroform to get the concentrations 20 mg/ml for EPC, 10 mg/ml for DCP, and 10 mg/ml for cholesterol. Appropriate volumes of each lipid with a mole ratio of each lipid component 7:1:2 (EPC:DCP:Cholesterol = 7:1:2) and 14 C-CyA (phospholipid:CyA = 300:1) are transferred to a single tube (see Note 1). 2. The mixture is evaporated using a Rotavapor device rotating at slow speed and heated (30°C) and is then subjected to vacuum to form a thin layer of lipid on the wall of the flask (see Note 2). This cyclosporin-lipid layer is hydrated with the appropriate amount of Buffer A (to get a phospholipid concentration of 10 mg/ml) for 2 h using the same Rotavapor

4

Fahr and Liu

Fig. 1. Schematic drawing of the device for measuring the transfer of substances between liposomes by the microcolumn technique

device after degassing the buffer and purging it with nitrogen gas. The final milky suspension is vortexed for 30 s at moderate speed, until all lipids have been removed from the glass vessel. This can be easily checked by optical inspection of the glass vessel from the outside. 3. This suspension is extruded 21 times (or more, but always use an uneven number!) through polycarbonate membranes with pore size 100 nm using a commercially available extruder

Utilization of Liposomes for Studying Drug Transfer and Uptake

5

(LiposoFast, Avestin Inc., Canada) in order to get homogenously sized liposomes. 4. The resulting liposomes are assessed for size distribution and zeta-potential and stored at +4°C before use. Liposomes formed by this method typically display a diameter range of 110–130 nm and a zeta-potential of −40 to −50 mV. 3.2. Preparation of Donor Blank Liposomes

1. Donor blank liposomes (without any transferrable drug included) are prepared using the same procedures as in Subheading 3.1. 2. The donor blank liposomes are labelled by 14C-cholesteryloleoyl-ether at 1 mCi/ml (see Note 3) as a non-exchangeable lipid marker.

3.3. Preparation of Acceptor Liposomes

1. 1-palmitoyl-1,2-oleoyl phosphatidylcholine (POPC) and cholesterol are weighed into individual test tubes and dissolved in chloroform to get the concentrations 20 mg/ml for POPC and 10 mg/ml for cholesterol. Appropriate volumes of each lipid with a mole ratio of the two lipids (POPC:cholesterol = 8:2) are transferred to a single tube. 3H-labelled cholesteryl-oleate is incorporated into the formulation at 1 mCi/ml. 2. The mixture is evaporated, hydrated with an appropriate amount of buffer A (to get a phospholipid concentration of 10 mg/ml), and extruded following the procedures as described in Subheading 3.1 to get the homogenously sized liposomal suspension. 3. Liposomes display a diameter range of 130–150 nm and zetapotential of ~0 mV.

3.4. Preparation of Liposomes for Saturation

1. 1-palmitoyl-1,2-oleoyl phosphatidylcholine (POPC), and cholesterol are weighed and mixed at a mole ratio of 8:2 following the procedures as described in Subheading 3.3. 2. The mixture is evaporated and hydrated (to get a phospholipid concentration of 10 mg/ml) following the procedures as described in Subheading 3.1 to get the liposomal suspension. 3. 2 ml of the liposomal suspension is transferred to a glass vial. This vial is mounted in a MSE-Soniprep 150 (Zivy AG, Oberwil, Switzerland) ultrasonic disintegrator with a titanium probe. The tip of the titanium probe is positioned 4 mm below the surface of the liposomal suspension. The vial is kept in an ice/water bath during the sonication, and nitrogen gas is used for purging the suspension to avoid oxidation of lipids during sonication. 4. The suspension is sonicated for 60 min (2 times 30 cycles, 1 cycle = 30 s sonicating and 30 s non-sonicating) with a 50% duty cycle using a MSE process timer (Zivy AG, Oberwil, Switzerland). The amplitude of the titanium probe is adjusted to 12 mm. The temperature of the suspension is checked to ensure that it does not rise significantly during sonication.

6

Fahr and Liu

5. This method allows the preparation of very small liposomes by introducing high amounts of energy into the phospholipid suspension. The average particle size of liposomes by sonication method is around 30–60 nm. 3.5. Size Determination of Liposomes

1. The size of the liposomes is determined by Dynamic laser light scattering using a Zetasizer Nano ZES3600 (Malvern Instruments Inc., UK). 2. Data are analysed by the Dispersion Technology software version 5.02 (Malvern Instruments Inc., UK).

3.6. Preparation of the Column Filling Gel DEAE-Sepharose CL-6B

1. The necessary amount of the DEAE-Sepharose CL-6B (depending on the number of mini-columns to be filled) is poured in an Erlenmeyer flask of appropriate volume (usual 200 mL size). 2. The gel settles down and the ethanol is carefully and slowly decanted by means of a Pasteur pipette connected via a tube to a water jet vacuum. 3. The gel is washed twice with buffer A in a ratio of 75% settled gel to 25% buffer A and kept in a last washing step with buffer B (1:1, v/v) at 4°C (see Note 4).

3.7. Preparation of the Ion-Exchange Microcolumn

1. The ion-exchange microcolumns were manufactured from Perspex® in the mechanical workshop of the university. The total bed volume of the microcolumn is about 0.5 ml (not very critical). 2. Some glass wool was placed at the bottom of the inner side of the column in order to ease the retaining of the gel in the column. 1.0 ml DEAE-sepharose CL-6B suspension pretreated as described in Subheading 3.6 is filled in the column (see Note 5). 3. The column is eluted with 2 ml buffer B at a speed of 30 drops in 45 s by using a pump (for example LKB Pump P-1, Pharmacia (Uppsala, Sweden)) and packed at the same time. 4. The microcolumns are saturated by applying 20 ml of the saturation liposomes prepared as described in Subheading 3.4 and eluted with 1.5 ml buffer B (see Note 6).

3.8. Validation of the Separation Efficiency of the Microcolumns 3.8.1. Donor Blank Liposomes

1. 10 ml donor blank liposomes prepared as described in Subheading 3.2 are placed on the top surfaces of the saturated microcolumns prepared as described in Subheading 3.7 and eluted with 1.5 ml buffer B at a speed of 30 drops in 45 s by using a pump. 2. The quantity of donor liposomes in the eluate is measured by liquid scintillation counting and the capturing capacity of the microcolumns for donor liposomes is estimated.

Utilization of Liposomes for Studying Drug Transfer and Uptake

7

3. The experiment is repeated nine times by using the same columns to check the reproducibility of the experiments. 3.8.2. Acceptor Liposomes

1. 10 ml acceptor liposomes prepared as described in Subheading 3.3 are placed on the top surfaces of the saturated microcolumns prepared as described in Subheading 3.7. 2. The same procedures are performed as described in Subheading 3.8.1 to measure the recovery of the acceptor liposomes in the eluate by liquid scintillation counting and the reproducibility of the experiments.

3.8.3. Mixture of Donor and Acceptor Liposomes

1. Donor blank liposomes prepared as described in Subheading 3.2 and acceptor liposomes prepared as described in 3.3 are mixed in the ratio of 1:10 (v/v). 10 ml of the mixture is placed on the surface of the saturated microcolumns and is eluted with 1.5 ml buffer B (see Note 7). 2. The same procedures are performed as described in Subheading 3.8.1 to measure the capturing capacity of the microcolumns for donor liposomes and the recovery of the acceptor liposomes in the eluate and the reproducibility of the experiments.

3.9. Example: Cyclosporin A (CyA) Transfer Between Liposomes

1. Transfer of CyA between liposomal membranes was studied using the microcolumn prepared as described in Subheading 3.7. Donor liposomes prepared as described in Subheading 3.1 and acceptor liposomes prepared as described in Subheading 3.3 are mixed in a ratio of 1:10 (v/v) at a temperature of 37°C. 2. At certain time points, 10 ml of the mixture is placed on the top surface of the saturated microcolumns and eluted with 1.5 ml buffer B as shown in Fig. 1. 3. The CyA quantity in the eluates was measured by liquid scintillation counting. The transfer kinetics of CyA, an additional model drug (cholesterol) and 3H-cholesteryl-oleoyl-ether as non-transferrable liposome-marker is shown in Fig. 2.

3.10. Liquid Scintillation Counting

The quantity of liposomes and CyA is determined by liquid scintillation counting on a Liquid Scintillation Analyzer Tri-Carb 2800 TR (Perkin Elmer, Herrenberg, Germany).

3.11. Data Analysis

Data analysis is mostly done by applying a simple mono-exponential model to the obtained data: y = y0 × (1 – e–k × t). This describes the data in all simple transfer cases quite well and delivers an understandable value k, which can easily be converted into half times. Of course, this is only a phenomenological description of the transfer processes, but elaborate models studying the mechanisms involved in the transfer from one liposome to the other are out of the scope of this chapter.

8

Fahr and Liu

Fig. 2. Transfer of CyA between liposomes. Donor liposomes (EPC:DCP:Chol = 7:1:2) were loaded with 14C-CyA (phospholipid:CyA = 300:1) Acceptor liposomes, composed of POPC:Chol = 8:2, were mixed with donor liposomes in a ratio of 10:1. For comparison, 14C-cholesterol and 3H-cholesteryl-oleoyl-ether transfer was also studied. Curve fits were done using the equation described in Subheading 3.11

4. Notes 1. The lipids and drug used here for liposome preparation are only an example; different lipids and drugs can be used for liposome preparation for different purposes, as long as the donor liposomes are strongly negatively charged and the acceptor liposomes neutral. 2. The experimental conditions (e.g. temperature, organic solvent) depend on lipids and drugs used in the study. The vacuum should be applied gradually in order to form a homogeneous lipid layer. After formation of the lipid layer, it should be kept under vacuum for more time to remove the organic solvent as much as possible. A very good control is the experimenter’s nose. If you are suffering from a cold at the time of the experiment, ask a colleague. 3. Linear calibration between scintillation counting and concentration of the radiolabelled lipid marker should be done to choose the suitable concentration of the lipid marker incorporated into liposome. After successfully establishing the whole procedure, the amount of radioactive tracer used can

Utilization of Liposomes for Studying Drug Transfer and Uptake

9

be reduced to the absolute minimum necessary for counting. But do these measurements regularly, especially, if you have not done it for a while. 4. In the washing process, any strong or even violent stirring or shaking of the gel must be avoided. We are not exactly sure, why this influences the quality of the results, but it does! Just treat the gel as a sensitive flower at Valentine’s Day. 5. Great care should be taken to avoid air bubbles to be induced or included in the microcolumn, which will influence the experimental results by delivering erratic results. This can be easily optically checked through the Perspex® material used for making the microcolumns. 6. The optimal amount of saturation liposomes depends on the type of liposomes, so this amount should be checked in different cases. This can be done by applying small amounts of saturation liposomes to the column and observing the eluate by turbidity measurements. A steep increase in turbidity means, that sufficient saturation liposomes have been applied. The saturation of the minicolumns is necessary, as the resin (gel) contains many lipophilic binding places, which could adsorb complete or part of liposomes during elution. There will be a small loss of drug from the acceptor liposomes to the pre-saturated minicolumn gel, but as the donor liposomes are also adsorbed and stay in equilibrium with the acceptor liposomes, the total net transfer is negligible for this type of measurement for the chosen set-up. 7. The buffer used can be pre-cooled to 4°C in order to minimise transfer processes that might occur (see Note 6) during the elution.

Acknowledgment We thank Rene Schaufelberger (Novartis Pharma Inc., Basel) for excellent technical assistance in setting up these procedures.

References 1. Papahadjopoulos D, Bangham AD (1996) Biophysical properties of phospholipids. II. Permeability of phosphatidyl liquid crystal to univalent ions. Biochim Biophys Acta 126:185–188 2. Flaten GE et al (2006) Drug permeability across a phospholipid vesicle-based barrier 2. Characterization of barrier structure, storage stability and stability towards pH changes. Eur J Pharm Sci 28(4):336–43

3. Shabbits JA, Chiu GN, Mayer LD (2002) Development of an in vitro drug release assay that accurately predicts in vivo drug retention for liposome-based delivery systems. J Control Release 84(3):161–70 4. Fahr A et al (2005) Transfer of lipophilic drugs between liposomal membranes and biological interfaces: consequences for drug delivery. Eur J Pharm Sci 26:251–265

10

Fahr and Liu

5. Joguparthi V, Xiang TX, Anderson BD (2008) Liposome transport of hydrophobic drugs: gel phase lipid bilayer permeability and partitioning of the lactone form of a hydrophobic camptothecin, DB-67. J Pharm Sci 97(1):400–20

6. Bienveue A et al (1985) Kinetics of phospholipid transfer between liposomes (neutral or negatively charged) and high-density lipoproteins: a spin-label study of early events. Biochim Biophys Acta 835:557–566

Chapter 2 The Use of Liposomes in the Study of Drug Metabolism: A Method to Incorporate the Enzymes of the Cytochrome P450 Monooxygenase System into Phospholipid, Bilayer Vesicles James R. Reed Abstract Although lipids are essential for the optimal activity of the cytochromes P450 monooxygenase system, relatively little is known about the membrane environment in which these enzymes function. One approach used to mimic the structural arrangement of lipids and enzymes within the endoplasmic reticulum is to physically incorporate the cytochromes P450 and their redox partners in a vesicle bilayer of phospholipids. Several methods have been devised for this purpose. This chapter describes a method in which the P450 monooxygenase system is incorporated by first, solubilizing the enzymes and lipid with sodium glycocholate. After the protein and lipid aggregates are dispersed, the detergent is removed by adsorption using BioBeads SM-2 resin which leads to the formation of bilayer vesicles of phospholipid containing incorporated cytochrome P450 and NADPH cytochrome P450 reductase. This procedure requires relatively a short preparation time, provides concentrated reconstituted systems that can be used in a wide range of applications, allows for several enzyme samples to be prepared simultaneously so that different conditions can be compared, and results in minimal loss of active enzyme. Key words: Phospholipid vesicles, Cytochromes P450, Reconstituted systems, Drug metabolism

1. Introduction The cytochromes P450 (P450) represent a ubiquitous gene superfamily comprising a diversity of isoforms that are expressed virtually in every organism in a species-specific pattern and are responsible for most xenobiotic metabolism in vivo (1). Thus the cytochromes P450 play a key role in the oxidation and clearance of most drugs and, in some instances, bioactivate toxins and promutagens to reactive intermediates that bind to cellular macromolecules V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_2, © Humana Press, a part of Springer Science+Business Media, LLC 2010

11

12

Reed

which, in turn, may lead to toxicity and/or the initiation of carcinogenesis (2–5). Because of these attributes, metabolism by the cytochromes P450 is relevant to a range of interests, including those of pharmacologists, toxicologists, and cancer researchers. The monooxygenase reactions catalyzed by these enzymes require electrons that can be delivered by various redox partners (6). The NADPH-cytochrome P450 reductase (reductase) is the primary redox partner in vivo and is capable of delivering both electrons needed in the catalytic cycle of the P450. Thus, the minimum enzyme assemblage needed to study monooxygenase reactions by the P450 system includes the reductase and the P450 isoform of interest. Much of the early information regarding the substrate specificity of the individual isoforms has come from studies in which the metabolites are identified after incubating the compound of interest with the active P450 monooxygenase enzyme assemblage and a source of NADPH (7–9). The P450 enzymes involved in xenobiotic metabolism and their redox partners are embedded in the membrane of the endoplasmic reticulum. As discussed in more depth later on, the membrane environment is essential for the functioning of the P450 monooxygenase system. Early attempts, over 30 years ago, to purify and characterize the various isoforms were limited by the dependence of enzyme activity on an unknown, heat-stable factor that was lost during the purification of the proteins. It was later found that the heat-stable factor was microsomal lipid (8, 10). Thus, it was determined that a lipid milieu was essential for the reconstitution of the catalytic activity of the purified enzymes. Subsequent studies have proposed that the lipid serves both as a “scaffold” to properly orient the P450 enzyme and the reductase for functional interaction (11, 12) and an effector that influences catalysis by the P450 enzyme (13). The effector role of phospholipid is evidenced by the modulation of P450 enzymatic activity at lipid:P450 enzyme ratios that are too low to facilitate the formation of liposomes. The “scaffold function” of lipid is ascribed to an additional level of modulation of enzyme activity observed at lipid concentrations at which liposomes form. Most enzymatic studies with the purified P450 enzymes use a short-chain (C-12), non-physiologic lipid, dilaurylphosphatidylcholine (DLPC). The reasons for this choice are the following: (1) the ease of preparation of the reconstituted systems with this lipid and (2) the lipid stimulates metabolism with most of the commonly studied P450 enzymes (14). However, studies have shown that the enzyme-lipid assemblages in the reconstituted systems with DLPC bear little structural similarity to the monolamellar, bilayer arrangement of lipid in the endoplasmic reticulum (15–17). Thus, it is clear that in order to truly appreciate the significance of the “scaffold” effect of the lipid on P450 metabolism, the enzymes must be physically incorporated in a vesicular bilayer of lipid.

Use of Liposomes in the Study of Drug Metabolism

13

Several methods used to generate these P450 vesicular reconstituted systems (VRS) have been published (18–20). Unfortunately, the methods require detergent, and this makes the procedures considerably more difficult and time-consuming than those using DLPC. However, the methods have been applied extensively and have generated interesting findings regarding the scaffold effect of the lipid bilayer on P450 metabolism (21–25). The approach common to all the methods for the preparation of VRS involves the following sequence of steps: (a) drying of the lipid in order to remove organic solvents that would inactivate or denature the enzymes, (b) detergent treatment to solubilize and disperse the enzymes and the lipids, and (c) a detergent removal procedure which causes the lipids to coalesce and form lipid bilayer vesicles, and in the process, the enzymes are incorporated in the vesicles. The thermodynamics associated with the detergent solubilization of a membrane complicate the ability of the VRS preparation methods to study the relationship between P450 activity and lipid concentrations. As mentioned above, this is an important aspect when evaluating the scaffold effect of lipids on enzymatic activity. The complexity of membrane solubilization has been reviewed in detail previously (26). At low concentrations, detergent binds to lipid and partitions between the lipid and aqueous phases (27, 28). Thus, in the simplest terms at low concentrations, two populations of detergent can be assigned – that bound to lipid and that partitioned into solution as monomers. The ratio of detergent concentrations in the two forms is dependent on the partition coefficient of the detergent for the aqueous and lipid phases. The initial stage of lipid solubilization occurs when the detergent concentration is increased to a point at which the bilayers containing a mixture of detergent and lipid are lysed into mixed micelles (26). Furthermore, it has been shown that this process occurs when the monomeric detergent concentration, in equilibrium with that bound to lipid, approaches or reaches the critical micelle concentration (CMC) of the pure detergent in aqueous solution (26, 27). As the detergent concentration is increased above the threshold required for membrane lysis, the excess detergent binds primarily to the mixed micelles of lipid and detergent, and the monomeric detergent concentration remains relatively constant at the CMC of the detergent. In the process of increasing the detergent concentration above the lysis threshold, the sizes of these micelles are reduced. Upon complete solubilization of the lipid/enzyme assemblage, each component lipid and enzyme is contained in an individual micelle of detergent. Thus, the total concentration of detergent needed to lyse the membrane will depend on both the concentration of the lipid and the extent to which the detergent partitions to the lipid phase. In applying the VRS methods at different concentrations of lipid, it is necessary to first identify the “effective” detergent to

14

Reed

lipid ratio (which is the ratio of the lipid-bound detergent concentration to the total lipid concentration) under a given set of optimized conditions that result in a desired degree of lipid/enzyme solubilization. The effective detergent to lipid ratio can then be used to calculate the detergent concentration needed to achieve comparable degrees of solubilization at different lipid concentrations. If the total detergent concentration (and not the lipid-bound detergent concentration) is adjusted in proportion to the lipid concentration in the methods for VRS preparation, the final VRS will have an excess detergent concentration (because both the monomeric and the lipid bound detergent concentrations will be scaled up). This in turn, likely will result in a high proportion of inactive enzyme because of the destructive effects of the excess detergent. Alternatively, if the detergent concentration is not adjusted when the lipid concentration is increased dramatically, the lipid will not be sufficiently solubilized to allow for the vesicular incorporation of the enzymes. We have tested two of the published methods, cholate gel filtration and cholate dialysis (17) and have found the active P450 enzyme tends to be extremely labile to the detergent, whereas the reductase tends not to incorporate into the vesicular fraction of the enzyme/lipid assemblage. Furthermore, preparations made using gel filtration are diluted and can be prepared only one at a time, limiting the opportunity to compare different experimental conditions, whereas, the dialysis procedure used to remove detergent is labor intensive and extremely time-consuming (3 × 12 h incubations against 3 L of dialysis buffer). Because of the limitations associated with these common methods for VRS preparation, we developed an improved technique for the preparation of P450 VRS (29). In the course of this work, we found a detergent (sodium glycocholate) that was less destructive to the P450 enzyme and utilized a more rapid way to remove the detergent by adsorption to BioBeads SM2 resin. We found this method to be superior to the two most commonly applied methods because of the following characteristics: (1) samples prepared using this method were also found to contain a higher proportion of the vesicular-incorporated reductase, (2) the detergent removal step could be carried out much more rapidly than the dialysis (2 h vs. 3 days), (3) the generated VRS samples could be easily prepared at relatively high enzyme concentrations (³5 mM), and (4) several samples could be prepared simultaneously thus allowing for the comparison of different conditions used in the reconstitution of the enzymes. We found that the VRS prepared under these conditions had very high catalytic activity relative to the reconstituted systems made using sonicated DLPC (17). The later sections of this chapter describe in detail the VRS preparation method. In addition, this chapter shows an approach that can be used to estimate the

Use of Liposomes in the Study of Drug Metabolism

15

effective detergent to lipid ratio in order to obtain VRS preparations with comparable levels of enzyme incorporation and activity over a range of lipid concentrations. This VRS method can be used to address general questions about the scaffold effect of the lipid. However, this type of reconstituted system also provides a more favorable structural framework with which to study the interaction of P450 enzymes with the various redox partners. Furthermore, the potential for future interest in the methods to prepare VRS with reductase and P450 is tremendous given the fact that lipids are known to separate into functional domains identified as “rafts” in studies with the plasma membrane (30). Evidence is just starting to accumulate that functional, lipid microdomains may also be present in the endoplasmic reticulum (31) and thus, may be significant in regulating P450 metabolism.

2. Materials 1. 0.5 M Hepes (pH 7.5) 2. 1 M MgCl2 3. 10% (w/v) aqueous sodium glycocholate (Calbiochem La Jolla, CA); membrane phospholipid of choice in chloroform at a concentration of 10 mg/mL. We have routinely used phosphatidylcholine from bovine liver (Avanti Polar Lipids Alabaster, AL). The lipid is both light and air sensitive with a tendency to oxidation of unsaturated acyl chains. The chloroform solutions are stored at –20°C. 4. Concentrated enzyme stock solutions (>10°mM reductase or P450 enzyme, respectively) in 100°mM potassium phosphate (pH 7.4) with 20 % glycerol (enzyme stock solutions are frozen at −80°C). 5. BioBeads SM-2 (Bio-Rad Hercules, CA). 6. 5 mm syringe filter (GE Osmonics, Minnetonka, MN).

3. Methods The general method characterized and described previously (29) results in a 0.5 mL solution of VRS containing 5 mM each of reductase and P450 enzyme and a 500:1 ratio of lipid:P450 (see Note 1). An adaptation is described below the general method which allows for adjustments in the lipid:P450 ratio of the VRS.

16

Reed

3.1. Preparation of the BioBeads SM-2 for Detergent Removal

This needs to be done at least two days before preparing the VRS. The use of BioBeads SM-2 to remove detergent from biological preparations has been described in depth previously (32). 1. Add 200 mL methanol to 30 g of BioBeads SM-2. 2. Stir for 15 min and collect beads by filtration with a Buchner funnel fitted with standard filter paper. 3. Immediately wash beads with another 500 mL of methanol; then repeat filtration. 4. Immediately wash beads with 1000 mL of water and repeat filtration. 5. Transfer beads to a chromatographic column and slowly wash beads with 2000 mL of water. 6. The moist beads are stored in ultrapure water at 4°C until required. When used, the beads are filtered as described above and added to the preparation containing detergent, lipid, and enzyme (supernatant derived from Step 5 of Subheading 3.3). Beads have been stored up to 3 months, periodically changing the water without noticeable problems in preparing the VRS.

3.2. General Method for Preparation of VRS: Drying the Lipid

1. Dry 1 mg of phospholipid (from a 10 mg/mL solution in chloroform) in a 1.5 mL microfuge tube overnight in a lyophilizer (see Note 2). 2. Release the vacuum on the lyophilizer by filling the chamber with N2 (see Note 3). 3. Add 50 mL of 0.5 M of Hepes (pH 7.5) and 50 mL of 10% sodium glycocholate to the tube containing dried lipid (see Notes 4 and 5).

3.3. Detergent Solubilization of Lipid and enzyme

1. Blow nitrogen over the tube opening of the solution from step 3.2.3 before capping, then bath-sonicate, and periodically vortex the tube until the solution is clear (usually 5 min). 2. While the lipid is being solubilized (step 3.3.1.), add P450 enzyme and reductase (2.5 nmol of each) to a second 1.5 mL microfuge tube and dilute to 482.5 mL with ultrapure water. 3. Add 7.5 mL of 1 M MgCl2 to the tube containing the P450 enzyme and the reductase. 4. Add the solution derived from step B.1 in approximate aliquots of 25 mL (approximately 1/4th the total volume of the solution) to the mixture of the P450 enzyme and the reductase (tube from step 3.3.3.). (See Note 6). Nitrogen is blown over the tube after the addition of each aliquot, and then the tube is capped and gently inverted to mix the enzymes with the detergent/lipid. 5. After the final addition of the solution from step B.1, blow nitrogen over the tube opening, cap the tube, invert several times, and incubate at 4°C for 1 h.

Use of Liposomes in the Study of Drug Metabolism

3.4. Removal of Detergent

17

1. Add 0.25 g of BioBeads SM-2 to the solution from step 5 of Subheading 3.3, blow nitrogen over the tube opening, cap the tube, and rock at 4°C for 2 h. 2. Draw off the solution from the Bio-Beads with a 26.5 gauge needle and filter through a 5 µm syringe filter. 3. The beads are rinsed twice with 0.2 mL aliquots of 50 mM Hepes (pH 7.4) and 15 mM MgCl2. These bead washes are filtered and added to the original bead filtrate. The final filtrate (containing the original filtrate in addition to the two bead washes) is typically around 0.65 mL in volume. 4. The sample is ready for use. An aliquot (0.1 mL) of the solution is routinely taken to determine the recovery of active P450 by determining the amount of enzyme capable of forming a ferrous CO-complex (33). A second 0.1–0.2 mL aliquot is run through a Superose 6 size exclusion column (MWCO 5,000 kDa) to determine the efficiency of enzyme incorporation. In this chromatographic step, it is assumed that the protein and lipid eluting in the void volume of the column represent the components that are incorporated into the bilayer, lipid vesicles. Reductase incorporation and activity is determined in the final preparation and in the fractions from the column by measuring the rate of the reductase-mediated reduction of cytochrome c (34). In addition, the concentration of phospholipid can also be determined in each fraction (35) in order to more accurately determine the lipid:protein ratio of the vesicular fraction.

3.5. Calculation Used to Adjust the Detergent Concentration Needed for Preparation of VRS with Different Lipid Concentrations

In adjusting our VRS method to changing concentrations of lipid, we assume that at the optimized conditions, the lipid bilayer is sufficiently solubilized to allow for the physical incorporation of the enzymes into the PC bilayer vesicles that form as the detergent is removed. If the monomeric concentration of sodium glycocholate in our optimized VRS preparation is equal to the CMC (as predicted by the studies that have examined the solubilization of membranes by detergents (discussed in the Introduction)) we can approximate the concentration of lipid-bound detergent from the total concentration used in the optimized conditions (those described in the protocol). More specifically, with a molecular weight of 488 g/mol, a concentration of 1% sodium glycocholate solution corresponds to 20.5 mM. The CMC of this detergent is 7.1 mM (36). Thus, we assume that the concentration of lipidbound detergent in the mixed detergent-lipid micelles is 13.4 mM (20.5 mM–7.1 mM). This value can be used to calculate the effective detergent to lipid ratio, and at any concentration of lipid, the concentration of detergent needed to achieve the same level of solubilization is then determined by first using the effective detergent to lipid ratio to calculate the amount of lipid-bound detergent and adding this amount to the CMC.

18

Reed

An example is shown to better explain the calculation involved: Amount of lipid in the defined, optimized condition (those described in the protocol) = 1.25 mmol Amount of lipid-bound detergent = 6.7 mmol (13,400 mM * 0.0005 L) Thus, the effective detergent:lipid ratio = 5.36 mole of detergent/ mole of PC Amount of monomeric detergent in the reconstitution = 3.6 mmol (7,100 mM * 0.0005 L) If one wants to make a reconstituted system with five times the lipid concentration as that in the defined, optimized condition, the calculation is as follows: Amount of PC = 6.25 mmol Amount of bound detergent = 5.36 detergent/PC × 6.25 mmol = 33.5 mmol detergent Amount of monomeric detergent = CMC*volume = 3.6 mmol Sum of monomeric and bound detergent in the 0.5 mL reconstitution = 37.1 mmol Detergent concentration = 74.2 mM = 3.62%. Thus, one should add 3.62% (w/v) sodium glycocholate to solubilize the lipid in a VRS containing 6.25 mmol of phospholipid (see Note 7).

4. Notes 1. We have found that when the lipid concentration is £1250 mM, the bilayer vesicles of lipid do not readily form. Thus, in order to make vesicles with P450: lipid ratios £250:1, the P450 concentration should be lowered (and not the concentration of lipid) from those stated in the general method. 2. We have also prepared the VRS by drying the lipid solution under a stream of N2. The lipid must be dried slowly and for a minimum of 2 h. If the flow of the nitrogen stream is too high, residual chloroform may be “trapped” under a film of lipid. This is apparent if the detergent-solubilized lipid (Step B.1) is cloudy. If this is observed, the enzymes will not incorporate properly in the lipid vesicles of the final preparation. In general, it has been found that the reductase does not incorporate into the vesicle bilayer as readily if the lipid is dried under N2 as compared to the results obtained with lyophilized lipid. 3. Our lab flushes the lyophilizer with anaerobic nitrogen that has run through a heated column of BASF palladium catalyst (BASF RO-20). 4. Solutions are bubbled with N2 (or preferably Argon) for 2–5 min before adding to the lipid.

Use of Liposomes in the Study of Drug Metabolism

19

5. We have also prepared the VRS using 0.5 M potassium phosphate (pH 7.25) without the MgCl2 and have not observed any significant differences in the quality of the final preparations. 6. Initially, we would add the enzymes from the concentrated stock solutions to the sonicated lipid and detergent solution after the latter was diluted with water. However, it was found that this method resulted in a much higher loss of the active P450. It seems the incremental addition of the detergent is less harmful to the P450. 7. The amount of BioBeads should not be adjusted to remove the excess detergent. The adsorptive capacity of BioBeads has been studied extensively (37), and it has been shown that the beads will also adsorb lipids. Thus, when the beads are scaled up with detergent, the excess lipid is more readily adsorbed and the desired increase in lipid concentration in the VRS is not attained. References 1. Porter TD, Coon MJ (1991) Cytochrome P-450. Multiplicity of isoforms, substrates, and catalytic and regulatory mechanisms. J Biol Chem 266:13469–13472 2. Guengerich FP (2001) Common and uncommon cytochrome P450 reactions related to metabolism and chemical toxicity. Chem Res Toxicol 14:611–650 3. Weng Y, Fang C, Turesky RJ, Behr M, Kaminsky LS, Ding X (2007) Determination of the role of target tissue metabolism in lung carcinogenesis using conditional cytochrome P450 reductasenull mice. Cancer Res 67:7825–7832 4. Iyanagi T (2007) Molecular mechanism of phase I and phase II drug-metabolizing enzymes: implications for detoxification. Int Rev Cytol 260:35–112 5. Rooney PH, Telfer C, McFadyen MC, Melvin WT, Murray GI (2004) The role of cytochrome P450 in cytotoxic bioactivation: future therapeutic directions. Curr Cancer Drug Targets 4:257–265 6. Hannemann F, Bichet A, Ewen KM, Bernhardt R (2007) Cytochrome P450 systems–biological variations of electron transport chains. Biochim Biophys Acta 1770:330–344 7. Guengerich FP (1989) Characterization of human microsomal cytochrome P-450 enzymes. Annu Rev Pharmacol Toxicol 29:241–264 8. West SB, Lu AYH (1972) Reconstituted liver microsomal enzyme system that hydroxylates drugs, other foreign compounds and endogenous substrates. V. Competition between cytochromes P-450 and P-448 for reductase in 3,

9.

10.

11.

12.

13.

14.

15.

4-benzpyrene hydroxylation. Arch Biochem Biophys 153:298–303 Saine SE, Strobel HW (1976) Drug metabolism in liver tumors. Resolution of components and reconstitution of activity. Mol Pharmacol 12:649–657 Strobel HW, Lu AYH, Heidema J, Coon MJ (1970) Phosphatidylcholine requirement in the enzymatic reduction of hemoprotein P-450 and in fatty acid, hydrocarbon, and drug hydroxylation. J Biol Chem 245:4851–4854 Ingelman-Sundberg M (1977) Phospholipids and detergents as effectors in the liver microsomal hydroxylase system. Biochim Biophys Acta 488:225–234 Taniguchi H, Pyerin W (1988) Phospholipid bilayer membranes play decisive roles in the cytochrome P-450-dependent monooxygenase system. J Cancer Res Clin Oncol 114:335–340 Causey KM, Eyer CS, Backes WL (1990) Dual role of phospholipid in the reconstitution of cytochrome P- 450 LM2-dependent activities. Mol Pharmacol 38:134–142 Balvers WG, Boersma MG, Veeger C, Rietjens IM (1993) Kinetics of cytochromes P-450 IA1 and IIB1 in reconstituted systems with dilauroyl- and distearoyl-glycerophosphocholine. Eur J Biochem 215:373–381 Autor AP, Kaschnitz RM, Heidema JK, Coon MJ (1973) Sedimentation and other properties of the reconstituted liver microsomal mixedfunction oxidase system containing cytochrome P-450, reduced triphosphopyridine nucleotidecytochrome P-450 reductase, and phosphatidylcholine. Mol Pharmacol 9:93–104

20

Reed

16. French JS, Guengerich FP, Coon MJ (1980) Interactions of cytochrome P-450, NADPHcytochrome P-450 reductase, phospholipid, and substrate in the reconstituted liver microsomal enzyme system. J Biol Chem 255:4112–4119 17. Reed JR, Kelley RW, Backes WL (2006) An evaluation of methods for the reconstitution of cytochromes P450 and NADPH P450 reductase into lipid vesicles. Drug Metab Dispos 34:660–666 18. Taniguchi H, Imai Y, Iyanagi T, Sato R (1979) Interaction between NADPH-cytochrome P-450 reductase and cytochrome P-450 in the membrane of phosphatidylcholine vesicles. Biochim Biophys Acta 550:341–356 19. Ingelman-Sundberg M, Glaumann H (1980) Incorporation of purified components of the rabbit liver microsomal hydroxylase system into phospholipid vesicles. Biochim Biophys Acta 599:417–435 20. Schwarz D, Gast K, Meyer HW, Lachmann U, Coon MJ, Ruckpaul K (1984) Incorporation of the cytochrome P-450 monooxygenase system into large unilamellar liposomes using octylglucoside, especially for measurements of protein diffusion in membranes. Biochem Biophys Res Commun 121:118–125 21. Ingelman-Sundberg M, Blanck J, Smettan G, Ruckpaul K (1983) Reduction of cytochrome P-450 LM2 by NADPH in reconstituted phospholipid vesicles is dependent on membrane charge. Eur J Biochem 134: 157–162 22. Bosterling B, Trudell JR, Galla HJ (1981) Phospholipid interactions with cytochrome P-450 in reconstituted vesicles. Preference for negatively-charged phosphatidic acid. Biochim Biophys Acta 643:547–556 23. Kawato S, Gut J, Cherry RJ, Winterhalter KH, Richter C (1982) Rotation of cytochrome P-450. I. Investigations of protein-protein interactions of cytochrome P-450 in phospholipid vesicles and liver microsomes. J Biol Chem 257:7023–7029 24. Schwarz D, Pirrwitz J, Ruckpaul K (1982) Rotational diffusion of cytochrome P-450 in the microsomal membrane-evidence for a clusterlike organization from saturation transfer electron paramagnetic resonance spectroscopy. Arch Biochem Biophys 216: 322–328 25. Taniguchi H, Imai Y, Sato R (1987) Proteinprotein and lipid-protein interactions in a reconstituted cytochrome P-450 dependent

26. 27.

28.

29.

30. 31.

32. 33.

34.

35. 36.

37.

microsomal monooxygenase. Biochem 26: 7084–7090 Hjelmeland LM (1990) Solubilization of native membrane proteins. Meth Enzymol 182: 253–264 Jackson ML, Schmidt CF, Lichtenberg D, Litman BJ, Albert AD (1982) Solubilization of phosphatidylcholine bilayers by octyl glucoside. Biochem 21:4576–4582 Bayerl TM, Werner G-D, Sackmann E (1989) Solubilization of DMPC and DPPC vesicles by detergents below their critical midellization concentration: high-sensitivity differential scanning calorimetry, Fourier transform infared spectroscopy and freeze-fracture electron microscopy reveal two interaction sites of detergents in vesicles. Biochim Biophys Acta 984:214–224 Reed JR, Brignac-Huber LM, Backes WL (2008) Physical incorporation of NADPHcytochrome P450 reductase and cytochrome P450 into phospholipid vesicles using glycocholate and Bio-Beads. Drug Metab Dispos 36:582–588 Pike LJ (2004) Lipid rafts: heterogeneity on the high seas. Biochem J 378:281–292 Browman DT, Resek ME, Zajchowski LD, Robbins SM (2006) Erlin-1 and erlin-2 are novel members of the prohibitin family of proteins that define lipid-raft-like domains of the ER. J Cell Sci 119:3149–3160 Holloway PW (1973) A simple procedure for removal of Triton X-100 from protein samples. Anal Biochem 53:304–308 Omura T, Sato R (1964) The carbon monoxidebinding pigment of liver microsomes. I. Evidence for its hemoprotein nature. J Biol Chem 239: 2370–2378 Phillips AH, Langdon RG (1962) Hepatic triphosphopyridine nucleotide-cytochrome c reductase: isolation, characterization, and kinetic studies. J Biol Chem 237:2652–2660 Stewart JC (1980) Colorimetric determination of phospholipids with ammonium ferrothiocyanate. Anal Biochem 104:10–14 Antonian L, Deb S, Spivak W (1990) Critical self-association of bile lipids studied by infrared spectroscopy and viscometry. J Lipid Res 31:947–951 Levy D, Bluzat A, Seigneuret M, Rigaud JL (1990) A systematic study of liposome and proteoliposome reconstitution involving BioBead-mediated Triton×-100 removal. Biochim Biophys Acta 1025:179–190

Chapter 3 Use of Liposomes to Study Cellular Osmosensors Reinhard Krämer, Sascha Nicklisch, and Vera Ott Abstract When cells are exposed to changes in the osmotic pressure of the external medium, they respond with mechanisms of osmoregulation. An increase of the extracellular osmolality leads to the accumulation of internal solutes by biosynthesis or uptake. Particular bacterial transporters act as osmosensors and respond to increased osmotic pressure by catalyzing uptake of compatible solutes. The functions of osmosensing, osmoregulation , and solute transport of these transporters can be analyzed in molecular detail after solubilization, isolation, and reconstitution into phospholipid vesicles. Using this approach, intrinsic functions of osmosensing transporters are studied in a defined hydrophilic (access to both sides of the membrane) and hydrophobic surrounding (phospholipid membrane), and free of putative interacting cofactors and regulatory proteins. Key words: Osmotic stress, Osmosensing, Transport, Reconstitution, Liposome, Proteoliposome, Membrane protein, Phospholipid, Signal transduction

1. Introduction Under steady state conditions, cells maintain a certain ratio of internal versus external osmotic pressure which results in a particular cell turgor. A change in the external osmolality leads to a change in the turgor, and cells have developed sophisticated mechanisms of osmoregulation to respond to this challenge. Changing osmolality is one of the most frequent types of environmental stress for many cells, in particular microbial cells. In order to properly respond to this stress, cells harbor sensory mechanisms which perceive stimuli related to osmotic stress, transduce these stimuli into appropriate intracellular signals, or directly respond to these stimuli by appropriate actions.

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_3, © Humana Press, a part of Springer Science+Business Media, LLC 2010

21

22

Krämer, Nicklisch, and Ott

In the case of hypo-osmotic stress, the major elements of stimulus response are mechanosensitive channels, which are dealt with in another chapter of this book. For the response to hyperosmotic stress, a major reaction is the activation of transport systems leading to the accumulation of compatible solutes. On the other hand, membrane bound sensory systems for hyperosmotic stress must be available, leading to appropriate responses in the cell on the level of transcription. In vitro systems for the study of membrane bound osmosensing transporters for compatible solutes require the presence of topologically closed vesicles and thus the use of liposomes. Functional reconstitution of transporter proteins in proteoliposomes allows the study of these systems in the absence of any other interfering mechanisms, cofactors or proteins, facilitating experimental access to both sides of the membrane, as well as to the composition of the phospholipid membrane in which the proteins are embedded. On the other hand, proteoliposomes have a number of drawbacks for the study of osmosensing in comparison to intact cells, mainly due to the inherent lack of a cell wall. As a consequence, proteoliposomes are in general more fragile than cells, and they lack turgor pressure. Consequently, osmosensory events related to turgor cannot be studied using proteoliposomes. As a further consequence, the fact that the morphological response of liposomes to hyperosmotic stress is different from cells has to be taken into consideration. Since liposomes behave as osmometers, i.e., they change their volume by water efflux according to the changing external osmotic pressure, and membranes are rigid in terms of their surface dimension; on the other, liposomes do not shrink in size but just invaginate under conditions of hyperosmotic stress to adapt the internal volume. The physical nature of stimuli, perceived by cellular osmosensors, is difficult to define, and in most cases not known (1). There are, however, a number of transport systems (1–3) in which these stimuli have been defined to a significant extent, and one of these model systems, the betaine transporter BetP from the grampositive soil bacterium Corynebacterium glutamicum, is used as an example here. In order to study osmosensing by a transport system, solute transport has to be measured. This necessarily requires some knowledge of transport kinetics and energetics, which, at least at a basic level, can be treated exactly as enzyme kinetics and energetics, and will not be further discussed here. Another basic requirement for the issue dealt with here is the proper handling of membrane proteins, their solubilization, isolation, purification, and, in particular, reconstitution. In this article, the relevant procedures of functional reconstitution of BetP from C. glutamicum will be described; however, no detailed reference to appropriate

Use of Liposomes to Study Cellular Osmosensors

23

procedures for obtaining solubilized membrane proteins in a functional active state will be made.

2. Materials 2.1. Preparation of Preformed Liposomes for Reconstitution

1. E. coli polar lipid extract (Avanti, Alabaster, USA).

2.2. Reconstitution of BetP Into Preformed Liposomes

1. Reconstitution buffer: 100 mM phosphate buffer, pH 7.5.

2. Liposome buffer: 100 mM phosphate buffer, pH 7.5.

2. Solubilized and purified BetP protein (minimal protein concentration should be around 0.3 mg/ml, detergent concentration about 0.1% dodecyl maltoside (DDM solgrade, Anatrace, Maumee, USA). The protein concentration has to be measured with an appropriate method, e.g., using amido black (4). 3. 20% v/v Triton X-100 solution in water. 4. Spectrophotometer for monitoring liposome titration with detergent. 5. Biobeads as absorbent for detergents (Biobeads SM2, Biorad, Munich, Germany; washing of the beads in methanol/water should be carried out as described by the manufacturer). The beads are stored in water. Directly before the use for absorbing detergent, the “wet beads” are prepared by placing biobeads suspended in water on filter paper, which removes the surplus water. These “wet beads” are used for weighing. 6. Minivial ultracentrifuge for harvesting liposomes. 7. Avanti Mini Extruder System (Avanti, Alabaster, USA). 8. Polycarbonate filters, pore size 400 nm (Nucleopore, Schleicher & Schuell, Dassel, Germany).

2.3. Betaine Transport as a Response of BetP to an Osmotic Shift

1. Rapid filtration unit (FH225V, Hoefer, Holliston, USA). 2. Membrane filters (Nitrocellulose, 0.45 mm pore size; Millipore, Schwalbach, Germany). 3. Radioactively labeled 14C-betaine (synthesized from 14C-choline by the use of choline oxidase)(5). 4. Transport buffer: 20 mM phosphate buffer, pH 7.5, 25 mM NaCl (if necessary, an increased osmolality is adjusted by addition of ionic solutes (e.g., NaCl), or neutral solutes, e.g., proline, if the ionic strength is not be changed significantly), 15 mM 14 C-betaine, and 0.5 mM valinomycin. 5. Washing buffer: 0.1 M LiCl. 6. Szintillation fluid and szintillation counter.

24

Krämer, Nicklisch, and Ott

3. Methods In the following, functional reconstitution of BetP, an osmosensing membrane protein, and the quantification of its response to osmotic stress is described. BetP is a secondary active glycine betaine transporter from the grampositive soil bacterium C. glutamicum (3). This carrier is energetically coupled to the co-transport of two sodium ions and thus driven by the electrochemical sodium potential. The transport activity of BetP is regulated by the actual osmotic stress, and it has been shown that the carrier protein is able to sense the extent of osmotic stress without any additional factors or components (3, 6). Isolation and purification of membrane proteins in a functionally active state requiring optimization of the kind and concentration of detergent used, of the solubilization conditions (ionic strength, type of buffer, pH, temperature, etc.), on the one hand, as well as optimization of the purification procedure (type of tag used, variation of isolation conditions, etc.), on the other. The protocol described here starts with membrane protein(s) in solubilized and functionally active form. The stability of solubilized membrane proteins in terms of functionality is in general a serious problem and needs attention (see Note 1). In order to obtain an appropriate in vitro test system, the solubilized osmosensing membrane protein is reconstituted into liposomes, often prepared from E. coli lipids. There are a large number of different methods for membrane protein reconstitution; the method of choice used in most of the successful procedures recently published is the integration into preformed liposomes developed by Rigaud and colleagues (7, 8). A number of different aspects have to be tested and optimized when trying to obtain an appropriate in vitro system for testing functional properties of a reconstituted membrane protein. The preformed liposomes may differ in the type of lipids used depending on the requirement of the particular membrane protein to be reconstituted (see Note 2). Furthermore, lipid quality is an important issue, which means purity and lipid stability. The latter critically depends on how they are handled and stored (see Note 3). 3.1. Preparation of Preformed Liposomes for Reconstitution

1. Purified lipids dissolved in organic solvent (chloroform/methanol) are mixed. The mixture is evaporated to dryness in a rotary evaporator (30°C; temperature should be above phase transition temperature of the lipids). 2. Traces of solvent are removed overnight by freeze-drying (lyophilization). 3. Liposome buffer (plus 2 mM b-mercaptoethanol) is added at a concentration of 20 mg lipid/ml and lipids are suspended by stirring at room temperature (RT) for about 2 h.

Use of Liposomes to Study Cellular Osmosensors

25

4. The lipid suspension is flushed with N2 gas, frozen in liquid N2, and stored at −80°C until use. 5. For reconstitution an aliquot of lipids is thawed gently at RT. Thereafter, liposomes are prepared by extrusion through polycarbonate filters (pore size 400 nm, multiple extrusions 17 times). 3.2. Reconstitution of BetP Into Preformed Liposomes

1. The preformed liposomes are diluted fourfold into liposome buffer. 20% v/v Triton X-100 is added stepwise (in 2–4 ml portions, thorough mixing by the use of a pipette) until detergent saturation of liposomes is achieved. Reaching the correct physical state of the liposomes for reconstitution is monitored by measuring optical density at 540 nm (7, 8). Optimal conditions for incorporation of the solubilized protein are obtained right after reaching the maximum value of light scatter at 540 nm. 2. The solubilized protein is then added drop by drop with thorough mixing. The amount of detergent added to the mixture together with the solubilized protein should be small in comparison to the amount of Triton X-100 added before. This mixture is shaken gently for 30 min at RT. 3. For removal of detergent according to the batch-procedure (7, 8), polystyrol beads (biobeads are added according to their absorption capacity for particular detergents (9), e.g., 5 mg and 10 mg of wet beads per mg of triton or DDM, respectively) are added. After shaking for 1 h at RT the same amount of beads is added and the mixture is shaken again for 1 h. The sample is then shaken overnight at 4°C with an additional twofold amount of beads. Subsequently, wet beads (same amount as for the first addition) are again added and shaken for 45 min at 4°C (see Note 4). 4. Finally, the proteoliposomes are separated from the beads with a pipette and washed twice with ice cold liposome buffer by centrifugation (Beckman Optima TLX tabletop ultracentrifuge, 350.000 g, 20 min, 4°C) before being resuspended in liposome buffer at a concentration of about 60 mg/ml lipid. The proteoliposomes are frozen in liquid N2 and stored at −80°C. 5. In order to assess the efficiency of reconstitution, the protein content of each batch of reconstituted protein has to be thoroughly measured by protein determination, e.g., using the amido black assay (4) (see Note 5).

3.3. Betaine Transport as a Response of BetP to an Osmotic Shift

1. For transport measurements, an aliquot of the proteoliposomes is thawed gently at RT, and extruded (polycarbonate filter, pore-size 400 nm) 17 times in a total volume of up to 1 ml, before being concentrated to the original volume by centrifugation (ultracentrifuge for mini vials, 350.000 g, 20 min, 20°C) and resuspended in liposome buffer. 2. In the transport assay the proteoliposomes, supplemented with internal K+ during the preparation in liposome buffer, are rapidly

26

Krämer, Nicklisch, and Ott

diluted into transport buffer containing valinomycin and the labeled substrate, betaine, and thoroughly mixed. In a standard experiment, the reaction is started by pipetting 5 ml of concentrated proteoliposomes into 1000 ml of transport buffer. According to the dilution ratio, this will lead to a K+-diffusion potential of about 140 mV. For measuring the activity of a particular transport protein, an appropriate driving force has to be provided. The driving force for BetP is the electrochemical Na+ potential, consequently, a sufficient amount of Na+ has to be present in the medium, in addition to the membrane potential, supplied here as K+-diffusion potential. 3. For each time-point in the following kinetic assay, 200 ml samples are used. These samples are withdrawn from the transport assay and instantly filtered through membrane filters in a filtration unit. The time intervals chosen depend on the velocity of the reaction under study; experienced experimenters are able to handle time points down to 5 s intervals. The filters are washed immediately twice with about 2 ml of washing buffer each (see Note 6). 4. After finishing the kinetic experiment, the filters are removed with tweezers and transferred into a scintillation vial each, which is filled with scintillation fluid, and counted. 5. The amount of betaine taken up into the proteoliposomes by the reconstituted carrier protein at the chosen time-points is calculated on the basis of the specific radioactivity applied in the transport buffer and the amount of protein present in the 200 ml sample of proteoliposomes. Protein determination should be carried out for each experiment, since variable losses during the extrusion procedure may give rise to variation. 6. Substrate uptake activity is interpreted in terms of transport kinetics, and specific transport rates are calculated (see Note 7). Beside the activity of the reconstituted protein, the result of uptake measurements may depend on a variety of other factors (see Note 8). 7. Testing a transport protein for its osmosensing function will need variation of the medium on both sides of the membrane, with respect to concentration and quality of solutes, e.g., ionic and nonionic solutes, or type of cations and anions, respectively. The solute composition in the interior of the proteoliposomes can be varied during liposome preparation (Step 1 of Subheading 3.3.). The composition of the internal compartment can also be changed after proteoliposome preparation by applying additional freezing and thawing cycles in a medium of changed composition followed by liposome generation using the extrusion procedure. The lipid composition can also be varied during the preparation of the preformed liposomes.

Use of Liposomes to Study Cellular Osmosensors

27

Additional lipids can be integrated by repeated cycles of freezing and thawing mixtures of (proteo) liposomes with different lipid compositions. 8. It should be noted, that there are two different types of experiments testing the dependence of a membrane inserted protein on variation of osmotic properties. On the one hand, the external osmolality (or solute composition) can be changed at the start of the transport experiment, leading to an osmotic gradient and a subsequent volume change of the proteoliposomes. A variation of solute concentration or composition on both sides of the membrane at the same time, i.e., during proteoliposome preparation, on the other hand, changes the osmotic pressure without introducing an osmotic gradient and without concomitant volume changes (6).

4. Notes 1. The stability of solubilized membrane (carrier) proteins is a serious issue for the success of this type of experiment. The conditions of solubilization, isolation, and storage, if necessary, have to be optimized carefully. Most transport proteins are notoriously unstable in the solubilized state, but, in general, rather stable once properly inserted into a phospholipid membrane. Consequently, they should be purified quickly and reconstituted into proteoliposomes right after purification. 2. Often, selected lipids have to be used for the reconstitution of a particular membrane protein, in order to reach optimum reconstitution efficiency or functionality.. Different membrane proteins have different preference for lipids, with respect to the phospholipid headgroup, the headgroup charge, and the fatty acid composition. Typical lipids to start with are lipids from E. coli membranes, asolectin (soy bean lipids), or egg yolk lipids. Once, a basic activity of the reconstituted protein is observed, variation of lipids will normally improve the results. An obvious strategy to overcome some of these problems is the use of synthetic phospholipids; however, for this, a balanced composition of unsaturated (essential for most membrane proteins) and saturated fatty acids is required. For unknown reasons, effective reconstitution into synthetic lipid mixtures seems to be more difficult compared with natural lipids, e.g., from E. coli. In any case, the fact that proteins in general prefer mixtures of phospholipids, both with respect to the headgroup and the fatty acid composition over the membranes prepared from single compounds has to be taken into account.

28

Krämer, Nicklisch, and Ott

3. The same as stated for handling of the protein is also true for handling of lipids. Always the best lipids available should be used. In general, Avanti is a good choice as a supplier for lipids, however, be careful for variation of lipid quality from batch to batch. After having established an optimized protocol for reconstitution, try to examine different batches of the lipids of choice and then buy a stock of lipids of the optimal batch. In general, if the proteoliposome test system, after having been successfully established, fails to work for unknown reasons, the quality of lipids is a good candidate to think about. For this reason, it is recommended always that part of a lipid batch which was proven to function is stored away, to serve as a control for later experiments. Other reasons for failure, of course, are the functionality of the solubilized protein, and, rather often, the efficiency of detergent removal by adsorbance to biobeads, which directly affects the efficiency of reconstitution and/or the permeability properties of the proteoliposomes. 4. The stability of the lipids used is a serious issue, too. The manufacturer supplies lipids in general sealed under nitrogen gas; it is recommended that the same condition is kept when storing lipids in the lab, preferably at −80°C. Useful recommendations for lipid handling and storage are found in (10–12). 5. Detergent removal is an important step during reconstitution and is prone to a number of difficulties. Complete removal of detergent from the proteoliposomal sample can simply be tested by the absence of foam after shaking. It has to be taken into account that the beads not only absorb detergents, but all amphiphilic (and even hydrophilic) compounds, in general in the order detergent > lipid > protein. It is thus important to limit the amount of biobeads used for absorption of the detergent in order not to lose too much protein. An alternative strategy partially avoiding this problem is to presaturate the biobeads with lipids (7, 8). 6. The efficiency of reconstitution is a complex problem. First, it refers to the amount of protein successfully integrated. This value may vary with the kind of detergent and lipid used, as well as with particular aspects of the reconstitution procedure. It has to be quantified by protein determination in the proteoliposomes, e.g., by the amido black method (4). In general, the efficiency of protein insertion decreases with increasing protein/lipid ratio during reconstitution. Second, reconstitution efficiency also means the fraction of functionally active protein in comparison to the total amount of integrated protein. Only the latter is measured by protein determination. This ratio is rather difficult to determine and in general needs additional methods, e.g., quantification of substrate binding. Third, protein orientation may be an issue. In most cases, the

Use of Liposomes to Study Cellular Osmosensors

29

orientation of the reconstituted membrane protein is not known. This, however, may be of crucial significance for interpreting its function in the in proteoliposomal in vitro system, in particular for transport proteins being dependent on oriented driving forces. There are a number of methods available for defining the orientation of a membrane-inserted protein, e.g., the use of antibodies, proteases, and impermeable labeling reagents for specific amino acids. 7. Although seeming apparently simple, fast, and efficient liposome filtration is a crucial step in the whole procedure. It should be taken into account that filtration, as used here, is to a large extent adsorption to the filter material, in view of the pore size of the filters used. Before application to the filtration unit, the membrane filters have to be presoaked in washing buffer, in order to avoid labeled substrate to be soaked into the external ring of the filter which is fixed in the filtration unit. Before running the kinetic experiment, the filters should be briefly soaked by application of vacuum. When applied to the filter during transport kinetics, the proteoliposome sample should quickly be distributed over the entire filter surface. The same holds true for the washing solution, which preferentially should be applied along the walls of the filtration chamber. 8. Transport kinetics has been elaborated in relevant text books. In general, when establishing transport kinetics of a particular carrier protein, a time course including several experimental points (five at least) has to be used in order to find out in which time window the uptake kinetics is reasonably linear with time. Once established, experiments are frequently carried out using two time points only within the linear kinetic phase, because of the requirement of rapid assays and/or frequently because of a limited range of linearity (mainly due to the small size of the proteoliposomes used). In order to provide sufficient experimental reliability, these two point-kinetics have to be carried out in multiple sets, e.g., fivefold. When changing experimental conditions, multiple time point kinetics has to be applied again to verify linearity. It should furthermore be mentioned that the linearity of a transport assay also depends on the amount of substrate taken up during the time course monitored in the kinetic experiment. For this reason, the amount of labeled substrate taken up at the end of the time course should not exceed one third of the added label. This amount, however, may sometimes be much lower, if higher substrate concentrations have to be used because of a relatively low substrate affinity of the transport protein under study. 9. It should be pointed out that the result of the kinetic analysis of a reconstituted transport protein not only depends on the functionality of the protein, and the experimental conditions

30

Krämer, Nicklisch, and Ott

applied (e.g., kind and extent of driving force, pH, ionic strength, kind of lipid provided, etc.), but to a large extent also on the quality of liposomes used. This refers, on the one hand, to the integrity and stability of liposomes, which in the presence of residual detergent, or high amounts of protein integrated, can be an issue. The integrity of the proteoliposomes used can be tested by the incorporation of calceine (13). Another reason for unexpected failure of transport experiments is partial leakiness of the proteoliposomes to small ions, e.g., protons, leading to a quick collapse of the driving force. This will in general not lead to leakiness of the larger solutes, like substrates (or calceine), and is frequently caused by residual amounts of detergent present. Furthermore, the size of the liposomes used is of critical significance. An early deviation of transport kinetics from linearity may indicate the presence of liposomes too small in size for the particular transport experiment. The impact of the liposome size on the shape of the transport kinetics observed is also a major reason for using liposomes sized by the extrusion procedure through polycarbonate filters. All other methods of proteoliposome formation lead to mixtures of vesicles with a broad size distribution. This fact will have a complex influence on the resulting transport kinetics. References 1. Wood JM (1999) Osmosensing by bacteria: signals and membrane-based sensors. Microbiol Mol Biol Rev 63:230–262 2. Poolman B, Spitzer JJ, Wood JM (2004) Bacterial osmosensing: roles of membrane structure and electrostatics in lipid-protein and protein-protein interactions. Biochim Biophys Acta 1666:88–104 3. Morbach S, Krämer R (2003) Impact of transport processes in the osmotic response of Corynebacterium glutamicum. J Biotechnol 104:69–75 4. Schaffner W, Weissmann C (1973) A rapid, sensitive and specific method for the determination of protein in dilute solution. Anal Biochem 56:502–514 5. Landfald B, Strøm AR (1986) Choline-glycine betaine pathway confers a high level of osmotic tolerance in Escherichia coli. J Bacteriol 165:849–55 6. Rübenhagen R, Morbach S, Krämer R (2001) The osmoreactive betaine carrier BetP from Corynebacterium glutamicum is a sensor for cytoplasmic K+. EMBO J 20:5412–5420

7. Rigaud J-L, Levy D (2003) Reconstitution of membrane proteins into liposomes. Methods Enzymol 372:65–86 8. Paternostre MT, Roux M, Rigaud J-L (1988) Mechanism of membrane protein insertion into liposomes during reconstitution procedures involving the use of detergents. Biochemistry 27:2668–2677 9. Rigaud J-L, Mosser G, Lacapere J-J, Olofsson A, Levy D, Ranck J-L (1997) Bio-beads: an efficient strategy for two-dimensional crystallization of membrane proteins. J Struct Biol 118:226–235 10. Zuidam NJ, Crommelin DJ (1995) Chemical hydrolysis of phospholipids. J Pharm Sci 84:1113–9 11. Hernández-Caselles T, Villalaín J, GómezFernández JC (1990) Stability of liposomes on long term storage. J Pharm Pharmacol 42:397–400 12. Torchilin VP, Weissig V (2003) Liposomes, 2nd edn. Oxford University Press, New York 13. Allen TM, Cleland LG (1980) Serum-induced leakage of liposome contents. Biochim Biophys Acta 597:418–426

Chapter 4 Studying Mechanosensitive Ion Channels Using Liposomes Boris Martinac, Paul R. Rohde, Andrew R. Battle, Evgeny Petrov, Prithwish Pal, Alexander Fook Weng Foo, Valeria Vásquez, Thuan Huynh, and Anna Kloda Abstract Mechanosensitive (MS) ion channels are the primary molecular transducers of mechanical force into electrical and/or chemical intracellular signals in living cells. They have been implicated in innumerable mechanosensory physiological processes including touch and pain sensation, hearing, blood pressure control, micturition, cell volume regulation, tissue growth, or cellular turgor control. Much of what we know about the basic physical principles underlying the conversion of mechanical force acting upon membranes of living cells into conformational changes of MS channels comes from studies of MS channels reconstituted into artificial liposomes. Using bacterial MS channels as a model, we have shown by reconstituting these channels into liposomes that there is a close relationship between the physico-chemical properties of the lipid bilayer and structural dynamics bringing about the function of these channels. Key words: MscL, MscS, NMDA, Liposome reconstitution, Patch clamp, EPR spectroscopy, FRET spectroscopy, Confocal microscopy

1. Introduction MS channels present a classical example of ion channels for which the composition and properties of the surrounding lipid matrix are crucial for their function. In a nutshell, the MS channels’ leitmotif is “force from lipids” (1, 2). Studies on prokaryotic (MS) channels over many years have demonstrated that the lipid bilayer transmits the mechanical force directly to this type of MS channels enabling them to detect changes in osmotic forces acting upon these microorganisms (3, 4). Among prokaryotic MS channels, the best characterized channels are MscL and MscS of

V. Weissig, (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_4, © Humana Press, a part of Springer Science+Business Media, LLC 2010

31

32

Martinac et al.

Escherichia coli. Bacterial MS channels were the first channels shown to sense directly the membrane tension in the lipid bilayer caused by the external mechanical force applied to bacterial cell membrane (5,6). Recent evidence indicates that, similar to prokaryotic MS channels, some MS ion channels found in membranes of a variety of eukaryotic cells, such as TREK-1, TRAAK and TRPC1, can also be gated by the bilayer mechanism (7–10). The glutamate-gated NMDA receptor channel has also been shown to exhibit mechanosensitivity upon reconstitution into artificial liposomes (11). The question of how MS channel proteins detect mechanical stresses in the lipid bilayer of cellular membranes and what makes these channels undergo conformational changes in response to membrane tension has been addressed in several studies examining the physical principles underlying gating of MscL channels by bilayer deformation forces (12–14). In one study, a combination of functional patch-clamp experiments with structural electronparamagnetic resonance (EPR) spectroscopic experiments allowed the drawing of the following conclusions about forces from lipids that affect both MS channel structure and function. First, the importance of the hydrophobic surface match for MscL mechanosensitivity was shown to result from a membrane-tension-induced bilayer thinning, which stabilizes intermediate conformations of MscL leading to channel opening due to a better hydrophobic match of the thinner bilayer with the open channel compared with that of the closed conformation. Second, geometric shape inequalities between lipid molecules asymmetrically distributed between the two leaflets of the lipid bilayer were shown to be critical in triggering the opening of the MscL pore of >25 Å in diameter (12, 13). These results have been confirmed by a FRET spectroscopic study in which asymmetric distribution of lipid molecules between the two leaflets of the lipid bilayer was shown to open MscL (14). Both studies thus indicate that the mechanism of mechanotransduction in MS channels is defined by both local and global asymmetries in the transbilayer pressure profile at the lipid protein interface. The role of bilayer thickness and the concept of hydrophobic mismatch in MscL functionality have further been investigated in studies exploring the effect of static magnetic fields (SMF) of moderate intensity (~400 mT) on the open probability of MscL reconstituted in phospholipid bilayers (15, 16). Results of these studies suggest that due to co-operative superdiamagnetism of phospholipid molecules, SMF most likely cause a rearrangement of the lipids resulting in a change of bilayer thickness. Given that not only ion channels but also other signaling membrane proteins may experience local change in bilayer thickness and/or curvature as a consequence of association with lipid microdomains or chemical modification of the lipid bilayer, the findings from studies of bacterial MS channels may also have implications

Studying Mechanosensitive Ion Channels Using Liposomes

33

for the functioning of all kinds of signaling membrane proteins including ion channels. The production of the MscL or MscS channel entails the protein expression of its subunit which then readily self-assimilates into the homo-multimeric functional channel within the bacterial membrane, without any post-translational requirements. Subunit expression with a fused purification tag allows simplified isolation of the channel (subunits) from bacterial membrane preparations using commonly available commercial purification supplies. Suitable tag expression vectors are described further below. Two affinity expression tag systems commonly used include charged affinity and specific enzymatic affinity. Charged affinity relies on a gene construct to produce a fusion with six histidine amino acids to either the amino or carboxyl terminal of the protein. This allows the fusion protein to bind to nickel or cobalt chelated to sepharose beads. Capture to beads allows further washing (purification) and final elution with a charged competitor (imidazole) solution. Alternatively for higher purity at lower yields and for a product with no artificial charges, enzymatic affinity may be used. This involves a fusion product with (typically) glutathione-S-transferase (GST) which allows specific tight binding to sepharose beads coupled to the GST target, glutathione. The protein of interest is cleaved off GST with (typically) thrombin via a target site that has been designed between the two fusion products from the onset. It is also possible to design affinity histidine tags that are also cleavable (not described). It should be noted that thrombin cleavage will leave part of its recognition sequence on the final product, and other amino acids from cloning vectors are typically also attached to the final protein, including any (uncleaved) 6 × His sequence.

2. Materials 2.1. MscL Production with GST or 6 × His Affinity Fusion Tags

1. Luria Broth (LB) (10 g/l bacto-tryptone, 5 g/l yeast extract, 10 g/l NaCl) autoclave sterilized.

2.1.1. Materials Common for Both MscL Production Methods

3. Ampicillin dihydrate antibiotic, 100 mg/ml in water, 0.22 mm filter sterilized.

2. Autoclave sterilized 100 ml and ~2,800 ml culture flasks.

4. 1 M Isopropyl b-D-1-thiogalactopyranoside (IPTG) in water, 0.22 mm filter sterilized. 5. Phosphate buffered saline (PBS) (10×): 80 g NaCl, 2 g KCl, 14.4 g Na2HPO4 2H2O and 2.4 g KH2PO4 into 1 L of water. On diluting adjust pH with HCl or NaOH.

34

Martinac et al.

6. Flat bristle brushes (in the order of approximately 5 mm and 8 mm width). 7. Phenylmethanesulphonylfluoride (PMSF), 20 mg/ml in isopropanol. 8. Deoxyribonuclease I (DNase) (e.g., Sigma DN25). 9. n-Dodecyl-b-D-Maltopyranoside (DDM), 200 mM in water (Anatrace D310). 10. Bradford Reagent (Coomassie dye binding protein assay, Protein dye reagent for quantitation). 2.1.2. MscL Production with GST Affinity Fusion Tag

1. AW737KO MscL knock-out strain transformed with N-terminus gluthione-S-transferase tagged MscL (pGEX-2T plasmid, GE) expression construct. 2. GST Wash Buffer: 1× PBS pH 7.4 with 1 mM DDM 3. 10 ml chromatography columns (e.g., Biorad 731–1550) 4. Glutathione sepharose (Bioworld or GE), washed following manufacture’s protocol, but with 1 mM DDM detergent. 5. Thrombin. (GE or Sigma) Dissolved and stored as per supplier’s instructions.

2.1.3. MscL Production with 6×His Affinity Fusion Tag

1. M15 E.coli [with pREP4 plasmid (Qiagen)] and with N-terminus 6×His-tagged MscL (pQE30 plasmid, Qiagen) or C-terminus 6×His-tagged MscL (pQE70 plasmid, Qiagen) expression construct. 2. Kanamycin sulphate antibiotic, 25 mg/ml ml in water, 0.22 mm filter sterilized. 3. Talon metal affinity resin (Clonetech 635502), washed following manufacture’s protocol, but with 1 mM DDM detergent. 4. 20 ml chromatography columns (e.g., Biorad 732–1010) 5. His Wash Buffer, 1× PBS pH 6.0, 1 mM DDM, 5 mM imidazole. 6. His Elution Buffer, 1× PBS pH 6.0, 1 mM DDM, 500 mM imidazole. 7. Millipore Amicon Ultra-15 Ultracel 10 k centrifugal filter device. Washed and used to manufacturer’s instructions.

2.2. MscS Production 6×His Affinity Fusion Tags

1. M15 E. coli [with pREP4 plasmid (Qiagen)] and transformed with pRARE plasmid (Merck), made competent ready for transformation. 2. N-terminus 6 × His-tagged MscS (pQE30 plasmid, Qiagen), or C-terminus 6 × His-tagged MscS (pQE70 plasmid, Qiagen) expression construct. The C-terminus construct gives greater yield.

35

Studying Mechanosensitive Ion Channels Using Liposomes

3. Luria Broth (LB): 10 g/l bacto-tryptone, 5 g/l yeast extract, 10 g/l NaCl, autoclave sterilized. 4. Autoclave sterilized 100 ml and ~2800 ml culture flasks. 5. Ampicillin dihydrate antibiotic, 100 mg/ml in water, 0.22 µm filter sterilized. 6. Kanamycin sulphate antibiotic, 25 mg/ml ml in water, 0.22 µm filter sterilized. 7. Chloramphenicol antibiotic 25 mg/ml in ethanol. 8. Glycerol. 9. Isopropyl b-D-1-thiogalactopyranoside (IPTG) 1 M in water, 0.22 mm filter sterilized. 10. Flat bristle brushes (in the order of approximately 5 mm and 8 mm width). 11. Phenylmethanesulphonylfluoride (~115 mM) in isopropanol.

(PMSF),

20

mg/ml

12. Deoxyribonuclease I (DNase) (e.g., Sigma DN25). 13. n-Dodecyl-b-D-Maltopyranoside (DDM), 200 mM in water (Anatrace D310). 14. Talon metal affinity resin (Clonetech 635502), washed following manufacture’s protocol, but with 1 mM DDM. 15. 20 ml chromatography columns (e.g., Biorad 732–1010). 16. Phosphate buffered saline (PBS), (10×): 80 g NaCl, 2 g KCl, 14.4 g Na2HPO4 2H2O and 2.4 g KH2PO4 into 1 L of water. On diluting to make 1× adjust pH with HCl or NaOH. 17. Solubilization Buffer: 1× PBS pH 7.5, 1 mM PMSF, 10% v/v glycerol. 18. MscS Wash Buffer 1: 1× PBS pH 7.5, 1 mM DDM, 10% v/v glycerol. 19. MscS Wash Buffer 2: 1× PBS pH 6.0, 1 mM DDM, 10% v/v glycerol, 5 mM imidazole. 20. MscS Elution Buffer: 1× PBS pH 6.0, 1 mM DDM, 10% v/v glycerol, 300 mM imidazole. 21. Millipore Amicon Ultra-15 Ultracel 10 k centrifugal filter device. Washed and used to manufacturer’s instructions. 2.3. NMDA Protein Expression and Purification 2.3.1. Cell Culture and NMDA Receptor Channel Constructs

1. Cell culture used for expression of NMDA receptor NR1 and NR2 subunits. (a) SF9 Spodoptera frugiperda (insect cell culture) (1 or 2 L). (b) XL99, a suspension CHO-K1 (Chinese Hamster Ovary cells) cell line adapted to growth in suspension in serum-free media (1 L).

36

Martinac et al.

2. Growth media (a) Sf-900II (Invitrogen) serum-free media for insect cell culture. (b) CHO-S-SFMII (Invitrogen) media for CHO mammalian cell culture. 3. NMDA receptor channel constructs. (a) 6 × His-tagged NR1a subunit (C-terminus, pENTR/ D-TOPO or pcDNA 3.3-TOPO plasmids). (b) 6 × His-tagged NR2A subunit (C-terminus, pENTR/ D-TOPO or pcDNA 3.3-TOPO plasmids). 2.3.2. Buffers for Purification of NMDA Receptor Channel Proteins

1. Cell Lysis and Solubilization Buffer (100 ml, pH 7.3): 20 mM sodium phosphate 500 mM NaCl, 10 mM imidazole, 3% octyl glucoside (OG), 20 mg/ml DNase protease inhibitor tablet (Roche), 1.25 ml of TALON resin (Clontech) added after solubilization. 2. Wash Buffer (50 ml, pH 7.3): 20 mM sodium phosphate, 500 mM NaCl, 20 mM imidazole, 1.5% OG, protease inhibitor tablet (Roche). 3. Elution Buffer (4 ml, pH 7.3): 20 mM sodium phosphate, 500 mM NaCl, 150 mM imidazole, 1.5% OG, protease inhibitor tablet (Roche). 4. Gel Filtration Buffer (50 ml, pH 7.5): 20 mM sodium phosphate, 300 mM NaCl.

2.4. SDS-Polyacrylamide Gel Electrophoresis (SDS-PAGE)

1. An electrophoresis vertical mini electrophoresis rig system and if desired compatible pre-cast 12% polyacrylamide denaturing gels of 1 mm thickness, ten well for protein analysis, as well as associated running buffers as per manufacture’s specifications. For manually cast SDS-PAGE gels, the following materials are required. 2. Distilled water (not from a source utilizing ultraviolet light sterilization.) 3. Tris(hydroxymethyl)aminomethane (TRIS), electrophoresis grade. 4. 1.5 M Tris-HCl pH 8.8. 5. 0.5 M Tris-HCl pH 6.8. 6. 10% Sodium Dodecyl Sulfate (Sodium lauryl sulfate) (SDS) 7. 30% Acrylamide/Bis Solution, 37.5:1 (e.g., Biorad 161–0158). The non-polymerized acrylamide monomer solution is a potent neurotoxin hazard and must be handled appropriately. Nonvolatile when supplied as a solution. 8. 10% (in water) Ammonium persulphate (APS), stored in ~500 ml aliquots stored at −20°C. 9. Tetramethylethylenediamine (TEMED).

Studying Mechanosensitive Ion Channels Using Liposomes

37

10. 5× Elpho Buffer: 15.1 g Tris, 94 g Glycine, double distilled water to 900 ml, mix, add 50 ml 10% SDS, top up to 1 L. 11. 4× Sample Buffer: 1.52 g Tris-HCl, 20 ml glycerol, 2 g SDS, 2 ml 2-mercaptoethanol, 1 mg bromophenol blue. 12. Protein molecular weight marker (suitable is Fermentas #SM0431). 13. Coomassie Stain: 1.25 g (Coomassie) Brilliant Blue R-250, 225 ml methanol, 50 ml glacial acetic acid, double distilled water to 500 ml. 14. Destain (300 ml methanol, 10% glacial acetic acid, water to 500 ml). 2.5. Liposome Reconstitution 2.5.1. Dehydration/ Rehydration (D/R) Method

1. Chloroform, AR grade. 2. Lipid. Azolectin (L-a-Phosphatidylcholine from soybean, Type II-S, Sigma P5638, or Type IV-S, Sigma P3644) or pure lipids (e.g., Phosphatidylcholine (PC) with acyl lengths of 16 to 20, or PC mixture with phosphatidylglycerol (PG) and/or Phosphatidylethanolamine (PE) pure lipids from Avanti Polar Lipids, Inc, Alabaster) 10 mg/ml in chloroform, stored at −20°C. 3. Nitrogen (N2) Industrial Grade Dry Gas supply, fed through a glass pipette for a nitrogen stream. 4. Purified mechanosensitive channel protein, as described within this chapter. 5. Dehydration/Rehydration Buffer Solution (D/R Buffer): 200 mM KCl, 5 mM HEPES (pH 7.2 adjusted with KOH). 6. Bio-Beads SM-2 Adsorbent (Biorad #152–8920) prepared by washing thrice (on shaker/rocker) in methanol for 30 min each. Remove methanol. Rinse briefly in distilled water by shaking three times, followed by two 30 min washes in distilled water. Decant and cover beads with fresh water. Degas by vacuum for 3 h to overnight. Store at 4°C. Before use, blot the required amount of beads with tissue paper to remove excess moisture. 7. Ethanol, AR grade. 8. Glass microscope slides, cleaned with ethanol, and dried. 9. Filter paper or similar absorbent paper/tissue type.

2.5.2. Sucrose Method

2.6. Patch-Clamp Recording From MS Channels Reconstituted into Liposomes

In addition to materials 1 to 6 listed above in Subheading 2.5.1, sucrose is also required. 1. Borosilicate glass calibrated microcapillary tube (e.g Drummond Scientific Co., Broomall, PA, 100 ml Calibrated Pipets [sic] cat: 2-000-100).

38

Martinac et al.

2. Recording Solution: 200 mM KCl, 40 mM MgCl2, 5 mM HEPES (pH 7.2 adjusted with KOH). 3. 0.22 mm syringe filter. 2.7. Confocal Microscopy of Liposome Reconstituted MS Channels

1. Single cysteine mutant expression construct and the resulting protein purified using deoxygenated Wash Buffer and Elution Buffer (see text). 2. Labeling Buffer: Deoxygenated PBS pH 7.5, with 1 mM DDM (n-Dodecyl-b-D-Maltopyranoside (DDM) (Anatrace D310)) 3. 40 mM tris(2-carboxyethyl)phosphine (TCEP) in distilled water. 4. Alexa Fluor® 488 C5-maleimide (AF488) (Invitrogen Molecular Probes A10254). Resuspended in water, separated in 100 mg aliquots which are then lyophilized/desiccated and stored at −20°C in a light sealed container. 5. Optional b-mercaptoethanol. 6. Dialysis cassette, 10 K molecular weight cut off, 3 ml capacity (e.g., Pierce, Thermo Fisher Scientific Inc, # 66380). 7. Dialysis Buffer: PBS pH 7.5 with 0.5% Triton X-100, chilled to 4°C for use. A 10× stock may be preferentially made, adjusted to pH 7.5 on dilution. 8. Confocal microscope with 60× objective lens, and appropriate excitation laser. 9. Cover slips and glass slides.

3. Methods Different constructs of MscL and MscS (GST fusion proteins or 6×His-tagged proteins) have been used for liposome reconstitution. The former is more suitable for testing channel function by the patch-clamp technique while the latter is appropriate for structural studies using EPR or FRET spectroscopy. In the case of NMDA receptor channels a 6×His-tagged protein has been used. The amount of protein required for reconstitution depends on the protein-to-lipid ratio (w/w for patch clamp experiments or mol/mol for structural EPR and FRET experiments), reconstitution method, and the type of experiment. For MscL reconstitution we have also found that the lipid-to-protein ratio depends on which construct is being used. Particularly, for MscS we have found that the lipids commonly used for electrophysiological measurements, such as DOPC or E. coli polar lipids cannot be used for MscS structural determination through spectroscopy, since they promote two-dimensional aggregation, regardless of the expression and purification conditions (20). This has not

Studying Mechanosensitive Ion Channels Using Liposomes

39

been tested in detail for experiments with MscL or NMDA. Assuming that dehydration/rehydration (D/R) method is used (see further) the number of active channels per liposome patch varies with the protein-to-lipid ratio (e.g., 1:1000 w/w) and is higher when the GST-MscL construct is reconstituted compared to 6 × His-tagged MscL constructs. If a sucrose reconstitution method is used (see further) less protein is required and higher percentage of liposome patches with active channels become available for experiments. Using the Vesicle PrepPro (Nanion Technologies GmbH) (see further) MscL (1/900 w/w protein-to-lipid ratio) can be reconstituted into liposomes in less than three hours total preparation time. Usually 5 micrograms of any of the MS channels described here is sufficient for several liposome preparations using the D/R method. If protein is scarce, to minimize the amount of protein required for reconstitution smaller amounts of lipids (ca. 2 mg) can be used per single liposome prep, which usually gives enough material for two days of patch clamping. 3.1. MscL- GST or His Tag Fusion Protein Purification

This procedure takes several days, thus it is advisable to start at the beginning of a week. Days 2 and 3 may be combined into a single long day.

3.1.1. Day 1: Overnight Starter Culture

1. Add 10 ml glycerol stock of bacterial strain (with expression construct) into a 100 ml culture flask with 12 ml LB and 12 ml 100 mg/ml Ampicillin. For pQE His constructs in M15 strain, also add 12 ml 25 mg/ml kanamycin sulphate. 2. Incubate at 37°C at 240 rpm (for ¾ inch orbit) in an orbital shaker incubator.

3.1.2. Day 2: Growth, Expression, and Cell Collection

1. Add 10 ml of overnight culture into a ~2,800 ml culture flask with 1000 ml LB and 1 ml 100 mg/ml Ampicillin. For pQE 6 × His constructs in M15 strain, also add 1000 ml 25 mg/ml kanamycin sulphate. 2. For aeration, seal the flasks with tissue paper (such as low lint fine task wipers) and incubate at 37°C at 150 rpm (for 2 inch orbit) in an orbital shaker incubator. 3. When optical density of the culture (OD) A600 = 0.8 (occurs in ~2–2.5 h) induce expression with the addition of 1000 ml 1 M IPTG. Also add 500 ml 100 mg/ml ampicillin. (Hint: For multiple samples, hold a considerably faster culture(s) at 4°C until others catch up before adding IPTG etc.). 4. Incubate with shaking for a further 4 h. 5. Centrifuge culture at around 8,000 ´ g for 15 min at 4°C. Discard supernatant. 6. For every 10 g of cell pellet resuspend in 10 ml PBS (with a ~8 mm wide brush). 7. Store the resuspension at −20°C or continue with day two.

40

Martinac et al.

3.1.3. Day 3: Cell Disruption and Membrane Processing

1. Thaw sample and keep on ice. 2. Add roughly a “match head” amount of DNase, and 1/100 volume of 20 mg/ml PMSF. 3. French press (Thermo Electron Corporation) twice at 16,000 PSI. 4. Remove debris, unbroken cells, or inclusion bodies by high speed centrifugation at 8,000 ´ g (k-Factor ~3,600) for 30 min at 4°C. Collect supernatant. 5. Pellet membranes by ultracentrifugation at 235,000 × g (k-Factor ~133) for 3 h at 4°C. Discard supernatant. 6. Suspend pellet in 5 ml PBS with a 5 mm width brush. 7. Add a further 17 ml PBS and 880 ml 200 mM DDM. 8. Solubilize on a roller wheel overnight at 4°C.

3.1.4. Day 4: Clarification and Column Binding and Purification

1. Clarify the solubilization by ultracentrifugation at 10,000 for 20 min at 4°C. Recover supernatant to use. 2. For 6×His fusion constructs add 2 ml prepared cobalt-resin beads. For GST fusion constructs add 1 ml prepared glutathione sepharose beads. 3. Incubate on a rocker or roller for 2 h at room temperature. 4. Pour the cobalt resin suspension into 20 ml chromatography columns. Pour the GST resin suspension into 10 ml chromatography columns. Release bottom caps when resin has settled somewhat after several minutes, and allow all liquid to flow out as discard. 5. Wash the cobalt resin twice with 25 ml His Wash Buffer. Wash GST resin twice with 10 ml GST Wash Buffer. For GST fusion protein, recap column and mix beads by pipetting with 1 ml GST Wash Buffer and transfer to a 2 ml centrifugation tube (see Note 1). Add 20 units of thrombin, and incubate overnight at room temperature. (B) For His tag protein elute with 15 ml His Elution Buffer into/or then transfer to a centrifugal filter device for concentration down to or just under 1 ml. Centrifuge to concentrate. If desired, imidazole can be removed, or reduced, by careful layering (to avoid mixing) with ~10 ml Wash Buffer and concentrated again. Repeat the “reverse decanting” process with another 10 ml Wash Buffer. Protein is ready for analysis and / or for use. Ensure protein is at less concentration than 500 µg / ml to avoid precipitation. Store at 4 °C. 6. Protein concentration can be determined by the Bradford assay at an absorbance of 595 nm, (typically measure 33.3 ml in 1 ml reagent) following the supplier’s instructions, referenced to a standard. Imidazole does not affect the assay.

Studying Mechanosensitive Ion Channels Using Liposomes 3.1.5. Day 5: Protein Collection and Determination of Protein Concentration

41

1. For GST fusion protein, centrifuge the beads at 1,000 × g (with slow down ramp) for 5 min. Collect supernatant as purified protein. Store at 4°C. 2. Protein concentration can be determined by the Bradford assay at an absorbance of 595 nm, (typically measure 33.3 ml in 1 ml reagent) following the supplier’s instructions, referenced to a standard. 3. For GST fusion protein, (optional), if poor thrombin cutting is suspected, analyze uncleaved protein present by loading ~15 ml of leftover beads by polyacrylamide gel electrophoresis (PAGE), or by first eluting protein off 15 ml leftover beads with 15 ml of 20 mM glutathione. Collect supernatant for PAGE analysis.

3.2. MscS- His Tag Fusion Protein Purification

This procedure takes several days, thus it is advisable to start at the beginning of a week. Membrane processing is required to be the same day as cell disruption, thus day 2 is a longer than usual work day.

3.2.1. Day 1: Overnight Starter Culture

1. Using sterile technique, on ice, to 50 ml competent M15 cells (with pRARE and pREP4) add 1 ml expression construct, mix, and let stand for 20 min. Heat shock at 42°C for 45 s, and place on ice for 2 min. 2. Add the transformation mixture into a 100 ml culture flask with 12 ml of LB and 12 ml Ampicillin (100 mg/ml- to maintain pQE plasmid), 12 ml kanamycin sulphate (25 mg/ml- to maintain pREP4 plasmid), and 12 ml chloramphenicol (25 mg/ ml-to maintain pRARE plasmid). 3. Incubate at 37°C at 240 rpm (for ¾ inch orbit) in an orbital shaker incubator.

3.2.2. Day 2: Growth, Expression, Cell Disruption, and Membrane Processing

1. Add 10 ml of overnight culture into a ~2,800 ml culture flask containing 1000 ml LB, 1 ml 100 mg/ml ampicillin, 1 ml 25 mg/ ml kanamycin sulphate, and 1 ml 25 mg/ml chloramphenicol. 2. For aeration, seal the flasks with tissue paper (such as low lint fine task wipers) and incubate at 37°C at 150 rpm (for 2 inch orbit) in an orbital shaker incubator. 3. When optical density of the culture (OD) A600 = 1.0 to 1.2 (occurs in ~2–2.5 h), let the culture stand at room temperature without shaking for an hour. 4. Induce expression with the addition of 800 ml 1 M IPTG and 4 ml glycerol (0.4% final). Incubate within shaker as previously, but at 25 °C for 4 h. 5. Centrifuge culture at around 8,000 × g for 15 min at 4°C. Discard supernatant.

42

Martinac et al.

6. For every 10 g of cell pellet resuspend (with a ~8 mm wide brush) in 10 ml PBS with 145 ml 20 mg/ml PMSF. Continue with cell disruption on the same day. 7. Add roughly a “match head” amount of DNase, mix. 8. French press (Thermo Electron Corporation) twice at 16,000 PSI. 9. Remove debris, unbroken cells, or inclusion bodies by high speed centrifugation at 8,000 × g (k-Factor ~3,600) for 22 min at 4°C. Collect supernatant for next step. 10. Pellet membranes by ultracentrifugation at 235,00 × g (k-Factor ~133) for 3 h at 4°C. Discard supernatant. 11. Suspend pellet in 5 ml Solubilization Buffer with a ~5 mm width brush. 12. Transfer suspended pellet to a 50 ml tube. Add 2.24 ml 200 mM DDM. Make up to final volume 50 ml using Solubilization Buffer (8 mM final DDM). 13. Solubilize on a roller wheel overnight at 4°C. 3.2.3. Day 3: Clarification and Column Binding, Protein Purification and Determination of Protein Concentration

1. Clarify the solubilization by ultracentrifugation at 100,000 × g for 20 min at 4°C. Recover supernatant to use. 2. Add 1.5 ml prepared cobalt-resin beads to the supernatant. Incubate on a rocker or roller for 3 h at 4°C. 3. Pour the cobalt resin suspension into a 20 ml chromatography column. Release bottom caps when resin has settled somewhat after several minutes, and allow all liquid to flow out and discard. 4. Wash the cobalt resin twice with 7.5 ml MscS Wash Buffer 1. 5. Wash the cobalt resin with 20 ml MscS Wash Buffer 2. 6. Elute with 4.5 ml MscS Elution Buffer into/or then transfer to a centrifugal filter device for concentration down to or just under 1 ml. Centrifuge to concentrate. Avoid over concentration that will result in precipitation. 7. If desired, imidazole can be removed, or reduced, by careful layering (to avoid mixing) with ~10 ml MscS Wash Buffer 1 and concentrate again as per step 6. Repeat the “reverse decanting” process with another 10 ml of MscS Wash Buffer 1. Protein is ready for analysis and/or for use. 8. Protein concentration can be determined by the Bradford assay measuring absorbance at 595 nm, (typically measure 33.3 ml in 1 ml reagent) following the supplier’s instructions, referenced to a standard. 9. When expressing from E. coli MS-channel knockout cells, ensure the protein concentration is lower than 500 µg/ml, to avoid precipitation. Expression in M15 (pQE32 or pQE70) or BL21-

Studying Mechanosensitive Ion Channels Using Liposomes

43

Rosetta cells (pET28) appears to allow further concentration of wild type MscS up to 10–15 mg/ml without precipitation. However, for some mutants, it is not recommeded to go above 4–5 mg/ml. Store at 4°C, do not freeze. 3.3. NMDA-6 × His-Tagged Protein Purification 3.3.1. Protein Expression in Sf9 cells

1. Grow Sf9 cells in Sf-900II media with 120 rpm shaking at 27°C. 2. For co-expression of NR1A and NR2A subunits, the cells are co-infected at the same MOI (multiplicity of infection) with both NR1A and NR2A recombinant baculoviruses at 3 × 106 cells/ml, using a MOI of 2–5 plaques forming unit per cell. The optimal time of harvest for the co-expression of NR1A and NR2A subunits is 72 h post infection. 3. Centrifuge the culture at 8,000 × g for 15 min at 4°C. Store the cell pellet at −80°C until required.

3.3.2. Protein Expression in CHO Cells

1. CHO cells are planted in 500 ml flasks at a density 1 × 106 cells/ml in CHO-S-SFMII (serum free, low protein media). 2. Cell culture is incubated at 37°C in a humidified atmosphere of 37°C, 7.5% CO2 in air. Samples are taken daily for determination of viable cell density using tryptan blue exclusion. 3. For transfection, cells are cultured to 80–90% confluence. 4. DNA constructs (1.6 mg/ml of transfection volume) and lipofectamine (1.6 ml) are diluted in 40 ml CHO-S-SFMII media and mixed gently. 5. After 5 min of incubation at room temperature the diluted DNA constructs and lipofectamine are combined and incubated for 20 min at room temperature to allow complex formation to occur. 6. The DNA-Lipofectomine complexes are slowly added to the flask containing CHO-S-SFMII media and gently mixed by swirling the flask. 7. Cells are incubated at 37°C, 7.5% CO2 on shaker at 150 rpm for 24 or 48 h.

3.3.3. Protein Purification

1. For cell lysis, the cell pellet from a 1 L expression is resuspended in 100 ml cell lysis and solubilization buffer without the detergent. 2. Cells are disrupted by sonication using a probe at 75 W for 15 s with 10–20 s cooling periods on ice (3×). 3. The OG concentration is adjusted to 3% after sonication and a sample is placed on rotating wheel for 20 min. 4. The lysed sample is clarified by centrifugation at 18,000 × g for 15 min at 4°C.

44

Martinac et al.

5. The clarified lysate is applied to Talon resin at a ratio 10:1 (e.g., 12.5 ml of lysate is applied to 1.25 ml talon) and placed on a rotating wheel for 20–30 min. 6. Unbound proteins are removed by centrifugation at 700 × g for 2 min at 4°C. 7. After binding, each 1.25 ml talon is washed for 10 min using 50 ml wash buffer. 8. Elution of the protein is performed by transferring 2.5 ml talon to 10 ml disposable column (Biorad, Hercules, CA) and protein is eluted with 4 × 1 ml of Elution Buffer. 9. For gel filtration, 500 ml of Talon purified protein sample is loaded onto a Superdex 75 10/300 GL column (GE Healthcare, Rydalmere NSW) at 0.25 ml/min using gel filtration buffer, and fractions (25 × 0.5 ml) are collected for analysis by Western blot and reconstitution into liposomes. 3.4. SDS-PAGE

Typically, a 12 % SDS PAGE gel is used for gel analysis of both MscL and MscS. 1. Assemble the electrophoresis vertical mini electrophoresis rig system to manufacturer’s instructions and use precast gels as prescribed. If manual gels are to be cast, the following gel recipe is generally suitable for a 1 mm thick mini protein gel: 2. In a disposable tube, make 10 ml 12% Running Gel by mixing together (without undue aeration): 3.35 ml distilled water, 2.5 ml 1.5 M Tris-HCl pH 8.8, 100 ml 10% SDS, 4 ml 30% acrylamide/Bis solution, 50 ml 10% APS, mix, 5 ml TEMED, mix. 3. With a pipette fill the bottom ~3/4 of constructed glass plate structure with the freshly prepared Running Gel before it sets. Usually two gels are prepare with the 10 ml solution, as usually two gels are required for electrophoresis within the rig. 4. Without mixing, carefully drop about 500 ml of water to the top to act as an air barrier. Allow the gel to set, which occurs within about half an hour. 5. In a disposable tube, make 10 ml 12% Stacking Gel by mixing together (without undue aeration): 3 ml distilled water, 1.25 ml 0.5 M Tris-HCl pH 6.8, 50 ml 10% SDS, 650 ml 30% acrylamide/Bis solution, 25 ml 10% APS, mix, 5 ml TEMED mix. 6. Turn gel setting rig upside down to remove water, use lint free tissues to remove remaining water if required. Return rig to normal position. 7. With a pipette, add the Stacking Gel on top of the Running Gel, to the top of the constructed glass plate structure (see Note 2). Place the well-combs into the top to allow sample

Studying Mechanosensitive Ion Channels Using Liposomes

45

well moulds to set. Place combs without introducing air bubbles. 8. Once the Stacking Gel layer has set, place the gel(s) within the running rig (as per manufacturer’s instructions) and add 1 × Elpho Buffer (at the same time to stop the gel from drying) to the top/full level, so that buffer is above the well layer. The apparatus will have inner and outer buffer reservoirs – both should be full to the same level to halt buffer from possible leak below the gel top and to allow uniform dissipation of heat. Remove well-combs. 3.4.1. Preparation of Protein Samples for PAGE and Electrophoresis

1. To ~20 ml protein sample mix in ¼ volume 4 × Sample Buffer in a microcentrifuge tube. Cap. 2. Heat (denature) sample using a hot block at 95°C for 5 min. Denature protein molecular weight marker only if instructed by manufacturer. (Fermentas #SM0431 requires denaturing.). 3. Spin any condensation down, and load sample(s) and molecular weight marker into separate gel wells (see Note 3). 4. To manufacturer’s instructions, cover rig and run electrophoresis at 100 V (or as appropriate) until the bromophenol blue marker dye is near the end of the gel. 5. Remove gel from glass plate and incubate, covered, with rocking in Coomassie Stain for several hrs to overnight. 6. Remove Stain solution (can be reused), rinse briefly with Destain and incubate covered with rocking in Destain for several hrs to overnight or until sample bands are distinct, and background is down to a desired level. Refresh Destain if saturated with stain. Adding lint free tissues to the Destain solution will sequester excess brilliant blue diffusing from the gel.

3.4.2. Gel Analysis: MscL

Wild-type MscL will appear at a molecular weight position of around 15 kDa, or theoretically higher at 16 kDa depending on how many further extra amino acids are added due to the cloning construct used. For the Fermentas #SM0431 marker, MscL will migrate in between the two smallest marker bands at 18.4 and 14.4 kDa. For reasons not yet defined, MscL often appears as a doublet at its size position. His-tag reliant purifications may yield faint contamination of multiple bands, whilst GST-tag reliant purifications may yield specific contaminants potentially identified as: 37 kDa bovine thrombin 29 kDa of the GST (Glutathione S-transferase) cleaved protein 58.5 kDa of the GST dimer 44 kDa MscL-GST fusion protein

46

Martinac et al.

3.4.3. Gel Analysis: MscS

The MscS subunit is about 31 kDa and will migrate at this size. Despite the denaturing conditions of SDS PAGE, an intense dimer band may exist. If denaturing is suboptimal, trimer and tetramer bands may also be detected. Of notable importance, degraded MscS will yield a band (a sub-band) just slightly smaller than the 31 kDa band, which may be reflected in slightly smaller multimers. Although the sub-band would normally not be present in fresh preparations, its presence, which may become the predominant product, will be reflective of compromised MscS function, and is a detrimental indicator. MscS, which can be stored for several months or to a year or so, should be analyzed on a gel periodically (or by other characterization methods) to detect the presence of the sub-band.

3.4.4. Gel Analysis: NMDA Receptor Channel Proteins

The metal affinity chromatography resulted in several low molecular weight mass bands visible on SDS/PAGE gel indicating nonspecific bindings and/or breakdown products of NR1a and NR2A subunits. Gel filtration chromatography purification is required to separate two main protein bands corresponding to 170 kDa for NR2A subunit and 100 kDa for the NR1a subunit from the contaminants.

3.5. Liposome Reconstitution

We have used several methods for reconstitution of MS channels into artificial liposomes (Fig. 1). Details are described below.

3.5.1. Dehydration/ Rehydration (D/R) Method

This method was the first developed for studies of MS channels by reconstitution into liposomes (17) and has since been the most frequently used method for characterization of biophysical and pharmacological properties of MS channels (7, 11, 13, 18) (Fig. 1a).

3.5.1.1. Preparation of Liposome Vesicles Using Bio-Beads SM2 (BioRad)

1. Rinse glass test-tube with chloroform, discard, and dry with nitrogen stream. 2. Dissolve 20 mg lipid into 1 ml chloroform in the glass tube. 3. Obtain an even lipid film on the surface of the tube by swirling the tube rapidly under a nitrogen stream until all the chloroform has evaporated. 4. Completely remove all chloroform by placing the tube under a high nitrogen stream for at least 15 min. 5. Resuspend lipid(s) in 2 ml of D/R Buffer and vortex for 60 s until cloudy. 6. Bath sonicate within a standard sonicator cleaner for 10–20 min. 7. Within a sealable tube, mix 200 ml (2 mg lipid) of liposome vesicles (from the sonication) with an amount of purified mechanosensitive protein to obtain the desired protein-tolipid wt/wt ratio (see Note 4). Fill to 2 ml with D/R Buffer.

47

Studying Mechanosensitive Ion Channels Using Liposomes

a D/R฀buffer฀

Purified฀MscL protein฀ Rehydrated lipids฀in฀D/R฀ buffer฀

Sonicate

3฀hr฀rocking/Remove฀ detergent฀(Biobeads), centrifuge

Dried฀lipid film

b

Dehydration/ Rehydration

Patch clamp pipette Blister฀

Pellet

MscL฀or฀MscS

0.4฀M฀Sucrose

Oven incubation,฀3฀ hrs฀at฀45฀ºC

Reconstituted proteoliposome

3฀hr฀rocking฀ (or฀overnight฀ rocking)฀

Patch clamp pipette Lipid cloud

Dried฀lipid฀ film

c

Reconstituted proteoliposome

O-ring

Purified MscL protein*฀

NMDAR protein*฀

Blister฀

Reconstituted proteoliposome

AC,฀3฀V,฀5฀Hz,฀ ฀ 2฀hours฀ Patch clamp pipette

Electroformation Dried฀lipid฀film฀

GUVs

Blister฀

Fig. 1. Diagram showing different liposome reconstitution methods used for studies of MS channels (Subheading 3.5.3). (a) Dehydration/rehydration method. (b) Sucrose method. (c) Electric field method (Vesicle PrepPro, Nanion Technologies). (* Please note MscL is added before formation of GUVs, whereas NMDAR protein is added after GUV formation.)

8. Roller mix at room temperature for 1 h. 9. Add 320 mg of tissue blotted Bio-Beads per ml of protein solubilized in 1 mM DDM. (The amount of Bio-Beads used is determined by total detergent content.) The mixture is gently stirred for 3 h. 10. Allow beads to settle to the bottom of the tube, and use the supernatant which contains detergent-free proteoliposomes. 11. Micro ultracentrifuge the supernatant at 250,000 × g for 45 min at 4°C. 12. Remove supernatant completely. Resuspend pellet in 80 ml D/R Buffer by pipetting.

48

Martinac et al.

13. Place 2–4 spots of 20 ml of proteoliposome suspension on a cleaned microscope slide. 14. Place slide(s) in vacuum desiccator for at least 6 h at 4°C. (Dehydration step.). 15. Add 20 ml D/R Buffer to each spot and leave overnight at 4°C within a Petri dish containing filter paper moistened with distilled water. (Rehydration step.) Sample is now ready to transfer into a patch recording chamber. 3.5.2. Sucrose Method

3.5.2.1. Day 1

Our laboratory recently developed this reconstitution method (19), which has several advantages over the classical D/R method (Fig 1b). Requiring preparation times of 6 h or less, this method significantly saves time compared with the D/R procedure, which on average requires two days preparation time. It also represents the first highly reproducible method for incorporation of the MscS channel protein, which does not incorporate readily into azolectin liposomes or liposomes made of mixtures of pure lipids. Therefore, this new method has the potential to be used for studies of ion channels that may be difficult to study by liposome reconstitution using D/R method. 1. Dissolve 10 mg of PC lipid (azolectin) in 1 ml of chloroform. 2. Take 200 mL aliquots and place into small test tubes (12 × 75 mm is ideal). 3. Remove solvent under a stream of N2 whilst swirling tube. 4. After continued drying of the lipid film under a stronger stream of N2 add 2 mL of H2O as a prehydration step (5 min)place drop of H2O directly onto the top of the lipid film; after prehydration the clear film develops opaqueness. 5. To the prehydrated lipids add 1 ml of 0.4 M sucrose and place in an oven or water bath at 45°C for 3 h. After this time the lipid will peel off the surface of the glass. 6. Add required protein amount to achieve a desired protein/lipid ratio. Important: add to bottom of tube, and do not shake. 7. Place on an orbital mixer at approximately 100 rpm and shake for 3 h. 8. Protein has now incorporated and can be used for experiments. However to ensure complete incorporation it is recommended to mix overnight.

3.5.2.2. Day 2

1. The lipids should have formed a cloud floating in the middle of the solution. Add approximately a cm³ of prepared BioBeads, and return to the mixer for 3 h. 2. Add 2 ml of the lipid cloud to recording bath, and they will form very large blisters.

Studying Mechanosensitive Ion Channels Using Liposomes

49

3.5.3. Electric Field Method (Vesicle PrepPro, Nanion Technologies)

20 ml of 5 mM lipid dissolved in chloroform was dried on the conductive side of an ITO-slide. A rubber O-ring acts as a spacer in between the two slides (Fig 1c).

3.5.3.1. MscL Protein Reconstitution

1. Once dried, add 250 ml of 210 mM D-sorbitol into the confines of the O-ring. 2. Add required protein/lipid ratio (typically 1:900 w/w) 3. The settings used are: Temperature = 23°C Frequency = 5 Hz Amplitude = 3 V 4. Allow to run for 2 h 5. After process is completed, extract GUVs with a 1 ml pipette.

3.6. Patch-Clamp Recording From MS Channels Reconstituted Into Liposomes

The results from studies of MS channels reconstituted into artificial liposomes have been described in numerous publications. Several reviews and original research papers (2, 3, 7, 11, 13, 17, 18, 20–22) highlight major findings of the structural and functional properties of MS channels using liposome reconstitution techniques.

3.6.1. Patch-Clamp Pipettes

1. Recording micropipettes are formed from borosilicate glass microcapillaries by using a pipette puller (e.g., P-87 Flaming/ Brown, Sutter Instrument Co., Novato, CA). Pulled recording pipettes should be ~1 mm in diameter corresponding to a pipette resistance in the range of 3.0–6.0 MW in Recording Solution. 2. To reduce electrical noise, pipette tips can be coated using Sylgard 184 (23) or transparent nail enamel (24) but is not required if single channel current is more than around 15 pA.

3.6.2. Recording

1. A 2–5 ml aliquot of liposomes made using any of the above methods is introduced into the recording chamber filled with the Recording Solution at 22°C. Recording pipettes are back filled with the same recording solution cleaned through a 0.22 mm filter. (Symmetrical solutions are preferentially used for determining single channel conductance of MS channels.). 2. Unilamellar blisters will be visible from several minutes up to 30 min (17) by which time the liposomes will have settled to the bottom of the chamber. The most common causes of “blistering” failure are incorrect pH or temperature of the Recording Solution. 3. Lower recording pipette to a blister and form a gigaohm seal (>1 GW) by applying suction to the patch pipette with a syringe. Suction is halted when a sudden decrease in pipette current occurs. MscL and MscS activities are usually recorded in excised liposomes patches (18, 19) which are obtained by next exposing the pipette tip to air for an instant only. (Fig. 2b.).

50

Martinac et al.

Fig. 2. (a) Confocal imaging of liposomes containing AlexaFluor488-labeled MscS proteins (Subheading 3.7) reconstituted by the Sucrose method at 1:100 (w/w) ratio (Subheading 3.5.2). Images were recorded about 10 min after addition of liposomes to the imaging buffer (same as buffer used for patch-clamp recording) and were obtained by an Olympus FluoView1400 Laser Scanning Microscope (Olympus) using a Nikon 60× water-immersion objective, NA 1.00 (Nikon, Japan), with excitation at 488 nm by an Argon laser (Melles–Griot). (b) Patch-clamp recording of the AF488-MscS liposomes prepared by the same method showing that the channel activity (recorded at pipette voltage VP = +30 mV) is not compromised by labeling. Patch-clamping was performed as per protocols described in Subheading 3.6

NMDA receptor channel activity is recorded after inclusion of agonists in the pipette solution or adding agonists directly to the bath solution (11). 4. Ion currents arising from activation of MS channels using negative pipette pressure are recorded with a patch clamp amplifier (e.g., Axon 1D or Axopatch 200B, Axon Instruments). The suction is monitored with a piezoelectric pressure transducer (Omega Engineering, Stamford, USA). Currents are usually filtered at 2 kHz and digitized at 5 kHz for offline analysis. 5. Single channel recordings can be analyzed using software such as pCLAMP (Axon Instruments) or in-house applications. 6. For NMDA receptor channels the protein is mixed with GUVs at required ratio depending on protein concentration (usually 1:10 or 1:20). 3.7. Confocal Microscopy of Liposome Reconstituted MS Channels

Confocal imaging of fluorescent-tagged MS proteins reconstituted into liposomes is often useful to observe protein expression, clustering phenomena, or for analytical studies such as Föster Resonance Energy Transfer (FRET) studies. MS proteins can be fluorescently tagged either genetically or by generating a single cysteine mutant for site-specific attachment of thiol-reactive probes (Fig. 2a). While an MscL protein tagged genetically with Green Fluorescent Protein has been expressed and characterized (25), this approach is more problematic for MscS due to its large cytoplasmic domain. Here we describe a protocol for confocal imaging of liposome reconstituted MS proteins tagged with the fluorescent dye, Alexa Fluor 488 C5-maleimide (AF488).

Studying Mechanosensitive Ion Channels Using Liposomes 3.7.1. Protein Labeling

51

1. Generate expression constructs of single cysteine mutants using site directed mutagenesis. Suitable kit systems include QuickChange (Stratagene) or Phusion (Finnzymes). 2. Express and purify the six histidine-tagged single cysteine mutants using protocols similar to wild-type proteins described in Subheading 3.1, except that the Wash and Elution buffers are previously deoxygenated thoroughly (before addition of DDM to avoid frothing) and supplemented with 10 mM tris(2-carboxyethyl)phosphine (TCEP) to reduce any protein disulfide bonds that may have formed. A suitable method of deoxygenation employs percolating nitrogen gas for 15 min through a fish-tank aerator placed at the bottom a buffer. 3. Set up a labeling reaction consisting of 10 nmole of protein (usually ~300 mg), to a tenfold excess of AF488 (i.e., use a 100 mg prepared aliquot) made up to a final volume of 2 ml with Labeling Buffer (see Notes 5–7). Rock for 2 h at room temperature or overnight at 4°C. 4. Optional: stop reaction using 5 mM b-mercaptoethanol final concentration. 5. Transfer reaction to a dialysis cassette and place in 2 L chilled Dialysis Buffer with stirring at 4°C. Replace the buffer twice, after 2 and 4 h, and then dialyse overnight. 6. Collect protein from dialysis cassette and concentrate if necessary. 7. Reconstitute labeled proteins into azolectin lipids by either the D/R or Sucrose Method (Subheading 3.5) at desired concentrations. For imaging of reconstituted liposomes, protein to lipid ratios of 1:100 to 1:250 (w/w) usually suffices. However, for applications such as FRET, it is advisable to use ratios of 1:20 to 1:50 (w/w). 8. Add liposomes to the buffer used for patch-clamp recordings (Subheading 3.6) and wait about 10 min for blisters to form. 9. Image liposomes using a 60× lens and appropriate laser excitation (488 nm for AF488) and power settings (See Fig. 2a).

4. Notes 1. End of pipette may be cut off for a bigger hole to avoid clogging. Also do not place tip to very bottom of column. Allow room to enable beads to enter the tip. 2. Avoid introducing air bubbles because bubbles would cause mould disruption. 3. Long narrow gel load micropipette tips are available.

52

Martinac et al.

4. Usually, a 1:4000 to 1:6000 protein-to-lipid ratio is used for the purpose of single channel recordings. To obtain a membrane fraction containing many mechanosensitive proteins a 1:100 to 1:200 protein-to-lipid ratio may be used. 5. Handling of Alexa Fluor dye should be done in subdued light. 6. In case the protein is not labeled immediately after purification, it should be pre-treated with a tenfold excess of TCEP for 1 h at room temperature. 7. The pH of the labeling reaction solution should be between 7.2 and 7.5; therefore, it is advisable to ensure that the added protein (previously eluted at pH 6.0) be at an adequate concentration.

Acknowledgments We wish to thank Dr Stephen Hughes for his contribution to studies of magnetic field effects on the MscL channels using liposome reconstitution technique. This research has been supported by grants of the Australian Research Council and the National Health and Medical Research Council of Australia to B. Martinac and A. Kloda. References 1. Kung CA (2005) A possible unifying principle for mechanosensation. Nature 436:647–54 2. Martinac B (2005) Force from lipids: physical principles of gating mechanosensitive channels by mechanical force revealed by chemical manipulation of cellular membranes. The Chemical Educator 10(2):107–14 3. Hamill OP, Martinac B (2001) Molecular basis of mechanotransduction in living cells. Physiol Rev 81:685–740 4. Martinac B (2001) Mechanosensitive channels in prokaryotes. Cell Physiol Biochem 11:61–76 5. Markin VS, Martinac B (1991) Mechanosensitive ion channels as reporters of bilayer expansion. A theoretical model. Biophys J 60:1120–7 6. Martinac B, Adler J, Kung C (1990) Mechanosensitive ion channels of E. coli activated by amphipaths. Nature 348:261–3 7. Maroto R, Raso A, Wood TG, Kurosky A, Martinac B, Hamill OP (2005) TRPC1 forms the stretch-activated cation channel in vertebrate cells. Nat Cell Biol 7:179–85 8. Maingret F, Patel AJ, Lesage F, Lazdunski M, Honoré E (2000) Lysophospholipids open the two-pore domain mechano-gated K+

9.

10.

11.

12.

13.

channels TREK-1 and TRAAK. J Biol Chem 275:10128–33 Zhang Y, Gao F, Popov VL, Wen JW, Hamill OP (2000) Mechanically gated channel activity in cytoskeleton-deficient plasma membrane blebs and vesicles from Xenopus oocytes. J Physiol 523(Pt 1):117–30 Zhou XL, Batiza AF, Loukin SH, Palmer CP, Kung C, Saimi Y (2003) The transient receptor potential channel on the yeast vacuole is mechanosensitive. Proc Natl Acad Sci USA 100:7105–10 Kloda A, Lua L, Hall R, Adams DJ, Martinac B (2007) Liposome reconstitution and modulation of recombinant N-methyl-D-aspartate receptor channels by membrane stretch. Proc Natl Acad Sci USA 104:1540–5 Perozo E, Cortes DM, Sompornpisut P, Kloda A, Martinac B (2002) Open channel structure of MscL and the gating mechanism of mechanosensitive channels. Nature 418:942–8 Perozo E, Kloda A, Cortes DM, Martinac B (2002) Physical principles underlying the transduction of bilayer deformation forces during mechanosensitive channel gating. Nat Struct Biol 9:696–703

Studying Mechanosensitive Ion Channels Using Liposomes 14. Corry B, Rigby P, Liu ZW, Martinac B (2005) Conformational changes involved in MscL channel gating measured using FRET spectroscopy. Biophys J 89:L49–51 15. Hughes S, El Haj AJ, Dobson J, Martinac B (2005) The influence of static magnetic fields on mechanosensitive ion channel activity in artificial liposomes. Eur Biophys J 34:461–8 16. Petrov E, Martinac B (2007) Modulation of channel activity and gadolinium block of MscL by static magnetic fields. Eur Biophys J 36:95–105 17. Delcour AH, Martinac B, Adler J, Kung C (1989) Modified reconstitution method used in patch-clamp studies of Escherichia coli ion channels. Biophys J 56:631–6 18. Häse CC, Le Dain AC, Martinac B (1995) Purification and functional reconstitution of the recombinant large mechanosensitive ion channel (MscL) of Escherichia coli. J Biol Chem 270:18329–34 19. Battle AR, Petrov E, Pal P, Martinac B (2009) Rapid and improved reconstitution of bacterial mechanosensitive ion channel proteins MscS and MscL into liposomes using

20.

21.

22. 23. 24.

25.

53

a modified sucrose method. FEBS Letters 583: 407–412 Vasquez V, Cortes DM, Furukawa H, Perozo E (2007) An optimized purification and reconstitution method for the MscS channel: strategies for spectroscopical analysis. Biochemistry 46:6766–73 Sukharev S (2002) Purification of the small mechanosensitive channel of Escherichia coli (MscS): the subunit structure, conduction, and gating characteristics in liposomes. Biophys J 83:290–8 Martinac B, Kloda A (2003) Evolutionary origins of mechanosensitive ion channels. Prog Biophys Mol Biol 82:11–24 Sakmann B, Neher E (1995) Single-channel recording. Plenum Press, New York and London Martinac B, Buechner M, Delcour AH, Adler J, Kung C (1987) Pressure-sensitive ion channel in Escherichia coli. Proc Natl Acad Sci USA 84:2297–301 Norman C, Liu ZW, Rigby P, Raso A, Petrov Y, Martinac B (2005) Visualisation of the mechanosensitive channel of large conductance in bacteria using confocal microscopy. Eur Biophys J 34:396–402

Chapter 5 Studying Amino Acid Transport Using Liposomes Cesare Indiveri Abstract The transport of amino acid across the membranes has great importance in cell metabolism. Specific experimental methodologies are required for measuring the vectorial reactions catalyzed by the membrane transporters. So far, the most widely used technique to study amino acid transport was the measure of amino acid flux in intact cell systems expressing a specific transporter. Some limitations in this procedure are caused by the presence of endogenous transporters and intracellular enzymes and by the inaccessibility of the intracellular compartment. Alternative experimental strategies which allow to reducing the interferences and improving the handling of the internal compartment would be useful to the amino acid transport knowledge. An experimental protocol, which makes use of liposomes to study the transport of amino acid mediated by the glutamine/amino acid (ASCT2) transporter, solubilized from rat kidney brush borders, is described. The procedure is based on the reconstitution of the transporter in liposomes by removal of detergent from mixed micelles of detergent, solubilized protein, and phospholipid. The transport is assayed in the formed proteoliposomes measuring the Na+ dependent uptake of L-[3H]glutamine in antiport with internal L-glutamine. This method allows measuring the transport activity under well controlled experimental conditions and permits performing experiments which cannot be realized in intact cell systems. Key words: Liposomes, Reconstitution, Transport, Membrane, Amino acids, Glutamine

1. Introduction The interest in the study of amino acids transport across the cell membrane increased exponentially in the last decade. This was also due to the relevance of this research field in human physiology and pathology. In mammalian cells the transport of amino acids is carried out by a large number of proteins which are different in primary structure and function. Most of these transport systems have been functionally characterized using intact eukaryotic cell systems in which specific transporters were expressed. Other V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_5, © Humana Press, a part of Springer Science+Business Media, LLC 2010

55

56

Indiveri

transporters, which have been predicted on the basis of gene sequences, are still orphan with respect to the function. The knowledge of the state of the art in this field is clearly and systematically described in a number of reviews (1–8). Even though intact cell experimental models allowed classification of a large number of amino acid transporters, technical limitations of the models made several aspects of the structure, function, and regulation of the transporters still unclear or unknown. Thus, the development of alternative experimental strategies would give benefits to the knowledge of amino acid transport. A suitable model for transport studies is the liposome system, which has often been helpful in defining and clarifying functional properties of membrane transporters of different types. Thanks to this model it is possible to perform experiments which are forbidden in intact cells. An important advantage of reconstituting membrane transporters in liposomes with respect to the study in cell systems is the control of experimental conditions in the internal compartment, which help to determine internal parameters like the Km for substrates or the influence of internal effectors on the transport. In this respect, a reconstitution method which allows the insertion of the protein in the liposomal membrane in the same orientation of the cell membrane is particularly reliable since it mimics the physiological situation, i.e., the correspondence of the intraliposomal and extraliposomal with the intracellular and extracellular sides. Other advantages are the reduction of interferences due to the absence of enzymes which could modify the substrates and the absence of different transporters in the same vesicles that is guaranteed by the higher phospholipids/protein ratio (one transport protein per liposome) with respect to the cell membrane (several transport proteins per cell); this feature has also the effect of prolonging the time course of the substrate uptake, leading to better resolution of the initial transport rate and the kinetics. The liposome model provides the possibility of modifying the lipid composition of the membrane to study the influence of specific lipids on the transport. To get further insights into the transport of amino acids, we have pointed out a procedure of reconstitution of the glutamine/amino acid transporter ASCT2 extracted from apical plasma membranes (brush-borders) of rat kidney where the transport of glutamine is very active being involved in essential physiological functions. The previous studies in intact cells expressing the ASCT2 protein (9–14) showed that it catalyzes antiport of glutamine with neutral amino acids which is dependent on extra cellular Na+ in kidney, intestine, lung, muscle,

Studying Amino Acid Transport Using Liposomes

57

testis, as well as in nervous tissue. The external Km’s for glutamine, other amino acids, and Na+ were determined and it was proposed that, differently from other neutral amino acid transporters, ASCT2 could transport also the anionic amino acid glutamate. The reconstitution of the glutamine/amino acid transporter in liposomes, besides confirming the previous data, permitted a deeper understanding of the ASCT2 transporter (15, 16). The reconstituted transporter, which is inserted in the liposomal membrane with the same orientation of the cell membrane, is inhibited by externally added Cys and Lys specific reagents. The internal Km’s for the amino acids were determined; their values are 20 times higher than the external ones. The mechanism of the complex transport reaction was found to be random simultaneous. The transporter is regulated by internal (intracellular) ATP. It also catalyzes a glutamate/glutamine antiport mode in which, glutamate, not the zwitterion glutamic acid, is transported with lower Km at acidic pH. The protocol for performing the reconstitution and the transport assay of the glutamine/amino acid transporter ASCT2 from rat kidney is described here. This protocol requires, as starting materials, brush border membranes from rat kidney which are prepared according to a previously described method (17). The transporter is solubilized from the brush borders with the non ionic detergent C12E8 and reconstituted in liposomes. The reconstitution procedure derives from a method which was originally pointed out for mitochondrial transport proteins (18, 19). It is based on the slow removal of detergent from mixed micelles of detergent phospholipid and protein by repeated passages through chromatography columns filled with hydrophobic resin. This leads to the formation of unilamellar phospholipid vesicles with the protein inserted in the membrane. The method has been adapted for the reconstitution of plasma membrane transporters. This goal has been achieved by changing the type of hydrophobic resin used and the critical parameters of the reconstitution like the protein concentration, the detergent/phospholipid ratio,and the number of repeated passages through the chromatography columns (15, 20). The transport is finally assayed by measuring the uptake of radioactive L-glutamine externally added to the proteoliposomes. The transport is active only in the presence of external Na+ and internal unlabeled L-glutamine or other neutral amino acids, since ASCT2 catalyzes an obligatory antiport of neutral amino acids which is dependent on extraliposomal Na+. Using this protocol, several types of experiments can be performed, by changing or adding some components or modifying some steps of the procedure.

58

Indiveri

2. Materials All the solutions and suspensions must be prepared in bi-distilled water. 2.1. Preparation of Brush Border Membranes

1. Buffer 1: 0.3 M sucrose, 5 mM EGTA 12 mM Tris-HCl pH 7.4. 2. Buffer 2: 0.15 M sucrose, 2.5 mM EGTA 6 mM Tris-HCl pH 7.4. 3. 50 mM MgCl2.

2.2. Solubilization of the Transporter

1. Brush borders. Aliquots of 50 µl brush borders prepared as described in Subheading 3.1 (containing from 150 to 300 µg protein depending on the preparations) stored at −20°C. 2. Solubilization buffer: 1.95 % C12E8 (octaethylene glycol monododecyl ether from Fluka). The solution can be stored at 0–4°C for one week.

2.3. Reconstitution of the Brush Border Extract

1. Stock buffer pH 7.0: 500 mM Hepes-Tris pH 7.0. Prepare the buffer dissolving Hepes in a volume of water about 70% of that final; then add Tris powder up to pH 7.0 and water up to the final volume. Use this solution to prepare all the diluted solutions. The stock buffer pH 7.0 can be stored at 0–4°C for 4–6 weeks. 2. 100 mM L-glutamine: Prepare fresh and use this solution to prepare all the diluted solutions. 3. Amberlite resin. Swell Amberlite XAD-4 resin (20-50 mesh from Fluka) adding 2 volumes of methanol to about 50 g resin. Mix gently using a glass stick; then leave for about 30 min at room temperature. Remove the supernatant methanol and repeat the procedure from 5 to 8 times until the supernatant methanol remains limpid. Then wash the resin with excess water 5 times (leave the resin in water 15 min among the washings). Store the resin at 0–4°C in 2 volumes of water. The resin can be stored for 4 weeks without adding conservatives. 4. Amberlite columns. Prepare 4 columns of about 0.5 cm internal diameter. Pasteur pipettes whose lower outlets are closed by cotton can be used (see Note 1). Fill the columns with the Amberlite resin up to 2.5 cm height. The Amberlite columns cannot be reused. 5. Amberlite equilibrating buffer. Freshly prepare 15 ml of a solution of 30 mM glutamine, 20 mM Hepes/Tris at pH 7.0 using the 100 mM L-glutamine, and the stock buffer pH 7.0. The concentration of glutamine may vary depending on the experiments (see Subheading 3.4, item 12 and related Notes).

Studying Amino Acid Transport Using Liposomes

59

6. Solution of 10% C12E8 (see also Subheading 2.2, item 2). The solution can bnume stored at 0–4°C for 4 weeks. 7. Preformed liposomes. Prepare the liposomes dissolving 1 g phospholipid (egg yolk phospholipids from Fluka) in a final volume of 10 ml water. Vortex to homogeneity, i.e., until no solid particles are present. Then sonicate 2 ml of the suspension in a glass tube for 2 min in pulse, i.e., 1 s sonication, 1 s intermission using a sonicator (at least 100 W of maximum power equipped with a 3 mm diameter micro tip) set at 40 W output power. The tube must be maintained at low temperature in an ice/water bath during sonication (see Note 2). The preformed liposomes can be stored at 0–4°C for 3–4 days. 2.4. Transport Assay

1. Sephadex resin. Swell about 2 g of Sephadex G-75 (40-120 µm dry bead diameter from Sigma-Aldrich) in excess water overnight. Then, eliminate excess air by vacuum for 15 min. 2. Columns A: Fill 4 columns of 0.7 cm internal diameter up to 15 cm height (Econo-Columns from Bio-Rad are suitable for this purpose) with Sephadex resin. The columns can be reused (see Notes 3 and 4). Store the columns at 0–4°C. 3. Buffer A: 30 mM sucrose and 20 mM Hepes-Tris pH 7.0 (use the stock buffer pH 7.0 solution). Prepare fresh. 4. Columns B: Fill 16 columns of 0.6 cm internal diameter up to 8 cm height (Glass Columns-Reusable from Pierce are suitable for this purpose) with Sephadex resin. The columns can be reused (see Notes 3 and 4). Store the columns at 0–4°C. 5. Buffer B: 50 mM NaCl. The solution can be stored for 1–2 weeks. 6. 100 mM L-glutamine (see Subheading 2.3, item 2). 7. 2 M NaCl can be stored at room temperature for several weeks. 8. Stock buffer pH 7.0 (see Subheading 2.3, item 1) 9. Labeled glutamine. Prepare freshly the solution with 2 µl of 100 mM L-glutamine, 50 µl of 2 M NaCl, 4 µl of stock buffer pH 7.0, 2 µl radioactive L-[3H]glutamine (L-[G-3H]glutamine from GE Healthcare, 1 mCi/ml), and 42 µl water (see Note 5). The concentrations of labeled glutamine may vary depending on the experiment (see Subheading 3.4, item 12 and related Notes). 10. Inhibitor. Freshly prepared solution of 0.8 mM mersalyl (mersalyl acid from Sigma-Aldrich) in 20 mM Hepes/Tris pH 7.0 (use the stock buffer pH 7.0).

60

Indiveri

3. Methods The method for the study of the transport mediated by the glutamine/amino acid transporter ASCT2 in proteoliposomes consists of three main steps: the solubilization of brush borders, the reconstitution, and the transport assay. The brush borders are prepared using a procedure previously described by other authors (17) with few modifications: details to carry out this procedure are described in Subheading 3.1. The brush border extract is inherently labile. To obtain reliable results it is important to prepare all the materials needed for the whole procedure before thawing the brush borders for the solubilization step. Initial reconstitution mixtures are then prepared mixing the brush border extract with phospholipids and further detergent to obtain mixed micelles. The mixtures are repeatedly chromatographed on Amberlite XAD-4 columns. After repeated passages through the resin, the detergent is removed; thus, phosholipids and proteins form the proteoliposomes. These vesicles contain, in the internal and external compartments, the solutes added in the initial reconstitution mixture and in the Amberlite equilibrating buffer, i.e., the substrate glutamine and the buffer. The reconstituted proteoliposomes are stable up to 4 h at 0–4°C. The external substrate is removed by size exclusion chromatography on Sephadex G-75, before starting the transport assay. This procedure permits the starting of the transport with a concentration of external labeled glutamine (0.1 mM) lower than the concentration of the internal unlabeled glutamine (30 mM). In the case of antiport systems like the glutamine/amino acid transporter ASCT2, the low external/internal concentration ratio leads to the accumulation of the radioactive substrate (added outside) in the internal compartment of the proteoliposomes. 50 mM Na+ (as NaCl) is added together with 0.1 mM L-[3H]glutamine. The transport reaction is stopped by the inhibitor mersalyl (20 µM) added at different times. In control samples the inhibitor is added together with the L-[3H]glutamine and Na+. The L-[3H]glutamine taken up by these controls represents the unspecific permeability of the proteoliposomes that will be subtracted from the L-[3H] glutamine taken up by the samples. After stopping the transport at various times, the external radioactivity is removed by Sephadex G-75 chromatography and the radioactivity entrapped inside the vesicles is counted. The transport activity is calculated from the entrapped radioactivity data. The method described allows obtaining reliable time courses of the transport process which represent the basis for every type of experiment (see Subheading 3.4, step 12) - to study the kinetics or the dependence of the transport on different types of inhibitors and effectors (specific reagents, nucleotides, ions, phospholipids).

Studying Amino Acid Transport Using Liposomes

3.1. Preparation of the Brush-Border Membranes

61

1. Cut thin slices (1–2 mm thickness) from kidney cortex (16–20 kidneys from male wistar rats of about 200 g) upto collect about 10 g of material. Put the slices in cold buffer 1 (previously kept at 0°C for 1 h) and wash 3 times to eliminate the blood. 2. Add 60 ml of cold buffer 1 and homogenize in a blender (good quality kitchen blender of about 500 ml) at maximal speed for 3 min. 3. Add cold MgCl2 solution up to reach a final concentration of about 12 mM (see Note 6), mix and keep 15 min at 0°C. 4. Centrifuge the homogenate in 50 ml tubes at 2,600 g in a preparative centrifuge for 15 min at 0°C. Collect together the supernatants. 5. Centrifuge the supernatant in 50 ml tubes at 18,500 g in a preparative centrifuge for 30 min at 0°C. Collect the sediments and resuspend together with 30 ml of cold buffer 2. 6. Homogenize with a 30 ml potter and then add cold MgCl2 solution up to reach a final concentration of about 12 mM (see Note 6), mix and keep 15 min at 0°C. 7. Centrifuge the suspension in 50 ml tubes at 2,600 g in a preparative centrifuge for 15 min at 0°C. Collect together the supernatants. 8. Centrifuge the supernatant in 50 ml tubes at 18,500 g in a preparative centrifuge for 30 min at 0°C. Collect the sediments and resuspend together in 30 ml of cold buffer 2. 9. Homogenize with a 30 ml potter and centrifuge in 50 ml tubes at 18,500 g in a preparative centrifuge for 30 min at 0°C. Collect the sediments and resuspend together in 2 ml of cold buffer 2. 10. Store the preparation at −20°C, in aliquots of 50 µl (150– 300 µg protein) in disposable tubes of 1 or 1.5 ml. The preparation can be stored without loss of activity up to 2 months.

3.2. Solubilization of the Brush-Border Membranes

1. Thaw 50 µl of brush border in ice/water bath. When the membranes are liquid, vortex the tube and add 100 µl of the solubilization buffer. Vortex for 30 s and keep at 0°C for 2 min. 2. Centrifuge the solubilized brush border at about 13,000 g for 4 min at 0°C. Collect the supernatant (brush border extract) in a disposable tube and keep for not more than 10 min, at 0°C until the preparation of the initial reconstitution mixture (Subheading 3.3, item 2). 3. After use (see Subheading 3.3, item 2) store the remaining extract at 0°C for the determination of the protein concentration. 4. The protein concentration in the extract is determined by a standard procedure like the BioRad or the Lowry method (stock solutions from Bio-Rad). This can be done after the transport assay.

62

Indiveri

Fig. 1. Columns holders. The photo shows examples of Plexiglas home- made columns holders for Amberlite (left), A (middle) and B (right) columns. Holes of appropriate sizes permit the insertion of the columns and the tubes. In the case of the holder for columns B, a movable support for vials is inserted below the columns. This support can be removed and replaced by a container that will be marked “radioactive” (see Subheading 3.4, step 10). Movable containers for equilibrating or washing the columns are present in the holders for Amberlite and A columns. Note the black sheet beyond the column tips in the left and middle holders to facilitate the view of the turbid drops

3.3. Reconstitution of the Brush Border Extract

1. Equilibrate 4 Amberlite columns applying 3 ml of the Amberlite equilibrating buffer. Columns holders are useful in this and the following sections (Fig. 1 and see Note 4). 2. Prepare 4 initial reconstitution mixtures (see Note 7) in 1.5 ml disposable tube adding the different solutions in the following order: 25 µl of brush border extract, 75 µl of the 10% C12E8, 100 µl of preformed liposomes, 210 µl of 100 mM L-glutamine (final concentration 30 mM), 28 µl of the stock buffer pH 7.0 (final concentration 20 mM), 262 µl water (to reach a final volume of 700 µl); vortex for 30 s. These operations must be performed at 0–4°C, for example in ice/water bath. 3. Apply each reconstitution mixture onto an equilibrated Amberlite column. Discard the first 5 drops and then collect the eluate. 4. Apply the eluate onto the same Amberlite column. Collect the eluate without discarding drops and repeat this passage further 14 times on the same column (16 total passages). These operations must be performed at room temperature. After these passages, active proteoliposomes are formed which are stable at 0°C up to 4 h.

Studying Amino Acid Transport Using Liposomes

3.4. Transport Assay (8 + 8 Samples Time Course Experiment)

63

1. Prepare a sample tube with 44 µl of labeled glutamine and a control tube with 44 µl of labeled glutamine and 22 µl of inhibitor. 2. Prepare 8 inhibition tubes with 2.5 µl of inhibitor. 3. Load 550 µl of each of the 4 proteoliposome samples on each of the 4 columns A and wait for the absorption of the samples into the resin. Gently, but quickly, apply small volumes (200– 400 µl) of the Buffer A on the columns, till the proteoliposomes are in the middle of the column. Then add 10 ml of buffer A. Collect 600 µl of proteoliposomes into disposable 1.5 ml tubes starting from the second turbid droplet (see Note 8 and Fig. 1). 4. Collect 500 µl of proteoliposomes eluted from each of the 4 columns A and mix together in a 2 ml disposable tube. 5. Transfer 814 µl proteoliposomes into both the sample and control tubes and mix quickly with the pipette tip. This represents the start of the transport reaction for both the samples (S) and the controls (C). A delay of 15–30 s can be practised between the two transfers (see also Note 9). 6. After 5, 10, 20, 30, 40, 60, 80, and 100 min, rapidly transfer (in such a way as to mix the proteoliposomes with the inhibitor) 100 µl from the (S) to an inhibition tube obtaining the inhibited sample (IS). At each time quickly apply 100 µl of the (IS) and 100 µl of the (C) on two columns B (see Note 9). Wait for adsorption of the suspension into the resin and elute adding 100 µl, 200 µl, 200 µl, 400 µl, 500 µl, 500 µl of 50 mM NaCl, waiting for the complete adsorption of each aliquot of the solution before adding the following. Discard the first 900 µl eluate and collect the following 1,000 µl into vials of the appropriate size to be analyzed in a b-counter. Also in this step a column holder will be very useful (Fig. 1). 7. In a separate vial (total radioactivity) add 5 µl of the labeled glutamine for the determination of the total radioactivity added to each sample. 8. After addition of at least 4 ml of scintillation cocktail (PicoFluor 40 from PerkinElmer) vortex each vial and count the radioactivity. 9. Regenerate the columns A with 10 ml buffer A and then 50 ml water. Equilibrate the columns applying 20 ml 0.1 % NaN3 (see Note 3). 10. Regenerate the columns B, with 10 ml buffer B and then 50 ml water. Equilibrate the columns applying 10 ml 0.1 % NaN3 (see Note 3). Very important: the solution eluted from the columns B contains radioactivity (3H); it must be collected

64

Indiveri

in special containers and treated according to the local regulation concerning the use of radioisotopes (see also Note 5). 11. Calculate the specific transport using the following equation: transportt (nmol/mg protein) = (cpmsample– cpmcontrol)/(SR × mg protein) (see Note 10) where transportt is the specific transport at time t, cpmsample and cpmcontrol are the cpm measured in the (IS) and in the respective (C), SR is the specific radioactivity (cpm/nmol) calculated as: total radioactivity/nmoles of glutamine per sample (see Note 11), and mg protein is the amount of proteins per sample derived from the concentration of the protein added to the initial reconstitution mixture (see Note 12). 12. Experimental data expressed as nmol/mg protein obtained above are interpolated by non linear regression analysis using the first order rate equation: y = A(1 – e– kt) where y are the specific transport values at the different times, A are the nmoles of glutamine taken up at infinite time (extrapolated from the fitting), k is the first order rate constant of the uptake process. The initial rate of the transport process is calculated as k · A (nmol/min/mg protein).

Fig. 2. Time course of L-[3H]glutamine uptake in the proteoliposomes. 0.1 mM L-[3H] glutamine is added at time zero to proteoliposomes containing 30 mM internal glutamine in the presence of 50 mM external NaCl. The transport reaction is stopped at the indicated times using 20 µM mersalyl. In the control samples the inhibitor is added at time zero (see Subheading 3.4). The curve is derived from the interpolation of the data points in a first order rate equation (see Subheadings 3.4, item 11 and 3.4, item 12)

Studying Amino Acid Transport Using Liposomes

65

A non linear regression analysis software should be used (Grafit from Sigma-Aldrich is suitable). From the interpolation of the experimental data in the described equation a time course is obtained like the example reported in Fig. 2. The time course experiment described can be used, with some modifications, to study the effect of inhibitors or activators (see Note 13) and the transport kinetics (see Note 14).

4. Notes 1. Cut the Pasteur pipettes leaving a short tip of about 2 cm. Put the cotton in the upper part of the tip. The elution rate of the column depends mainly on the compression of the cotton. Use medium compression to obtain elution rate of 30–50 µl/s (about 1 drop/s). 2. The glass tube should have a rounded bottom. It is very important to use a Pb free glass tube. If you are not sure about the feature of the tube, then use a plastic 2 ml disposable tube; this may lead to some variability in the liposome preparations. Ensure that the temperature of the suspension in the tube does not exceed 10°C during sonication. 3. The filled columns A and B can be reused, normally for 10 to 20 times. In any case, the columns must be refilled with new swollen resin when the elution rate has evidently decreased. Before refilling with the new resin, the glass columns must be cleaned with common detergents used for dishes and then washed with bidistilled water to completely remove any trace of detergent. 4. The number of Amberlite columns, columns A and columns B depends on the type of the experiment. In the protocol for the time course described here the requirement is 4 Amberlite columns and columns A; 16 columns B. The operator has to consider that: (1) a column A is needed for each Amberlite column; (2) the proteoliposomes eluted from column A are normally sufficient to perform 4–5 transport assay samples; (3) a column B is necessary for each transport assay sample. Column holders are of great help in the experimental procedures requiring the simultaneous elution of several columns (see Fig. 1). 5. The 3H isotope is not particularly hazardous; follow the local regulation concerning the use of radioisotopes. For safety measures, it is sufficient to use disposable gloves and take care not to directly touch the solutions containing radioactive glutamine. To this aim, centrifuge by short spin the tubes which contain radioactive solutions in order to avoid some radioactive drops that may stick to the tube edge or closure.

66

Indiveri

6. The volume of the MgCl2 solution (50 mM) required to be added must be calculated on the basis of the volume of the suspended sediment, which may vary in different preparations. The final concentration of MgCl2 (12 mM) is not very critical. 7. In this protocol four identical initial reconstitution mixtures are prepared. However, the number or the composition of initial reconstitution mixtures may change depending on the type of experiment. Two general types can be distinguished: (1) experiments in which compounds or concentrations in the external liposomal space are changing. In this case identical initial reconstitution mixtures are prepared; (2) experiments in which compounds or concentrations in the intraliposomal space are changing. In this case different reconstitution mixtures are required. The Amberlite columns must be equilibrated with different solutions containing the same compounds and buffer of the corresponding reconstitution mixture. 8. To clearly distinguish the turbid droplets place a sheet of black plastic or other material beyond the outlet of the columns (see Fig. 1). 9. To be faster use two different pipettes, one for the samples (Subheading 3.4, item 5) or inhibited samples (Subheading 3.4, item 6),and the other for the controls. A constant delay of 15–30 s can be maintained, after the Subheading 3.4, item 5 procedure (start of the transport reaction), between the two different (S) and (C) samples. The delay among different samples will be very useful in the case of more complex experiments requiring large number of samples (see Notes 13 and 14). 10. Depending on the type and settings, ß-counters will give the radioactivity measurements as cpm or dpm units. Both units can be used equally if the radioactivity of the samples and the specific radioactivity are expressed in the same units. In a well performed experiment, cpmcontrol must be lower than 10% of the respective cpmsample. 11. Calculate the nmoles from the concentration of externally added L-[3H]glutamine in the100 µl sample (each sample in the present experiment contains 10 nmoles of glutamine). 12. The protein content in each sample of 100 µl should correspond to 1/7 of the amount of protein in an initial reconstitution mixture (measured in the extract). However, a dilution of the samples occurs during the chromatography on column A. We have empirically calculated that the amount of protein in the 100 µl sample actually corresponds to 1/12 of that in the initial reconstitution mixture. It has to be stressed that the protein cannot be directly assayed in the proteoliposome samples since lipids interfere with the assay methods. 13. To study the influence of effectors on the transporter on the external side, these molecules must be added to the

Studying Amino Acid Transport Using Liposomes

67

proteoliposomes before the transport assay together with the labeled glutamine in the sample and in the control tubes. To study the influence of effectors on the internal side, the molecules must be included in the initial reconstitution mixture and in the Amerlite equilibrating buffer at the same concentration. In this case, several different initial reconstitution mixtures are needed, each with a different compound or different internal concentration of the effector (see also Note 7). The concentration of each effector should be chosen according to previous data (15, 16). In the case of investigation with novel molecules, a preliminary experiment should be performed in advance with few concentrations chosen in a wide range (for example 0.1, 1, 10, 100 µM). 14. To obtain kinetic constants like Km or Vmax several time courses must be performed at different substrate concentrations and the initial transport rate calculated. In experiments of this type which require several data points, the time course can be reduced to 4 points (5, 10, 30 and 60 min). For reliable constant determinations at least five different glutamine concentrations within an appropriate range must be used and each experiment should be repeated at least three times. In particular, for the determination of external Km, the L-glutamine in the labeled glutamine solution must be varied from 0.05 to 5 mM (final concentration in the proteoliposome samples; the other components do not vary) at fixed internal glutamine concentration of 30 mM. Since the transporter catalyzes an obligatory antiport of glutamine and neutral amino acids, the internal Km can be investigated by following the time course of the uptake of fixed 1 mM L-[3H]glutamine (see above for the labeled glutamine solution) in the presence of various intraliposomal glutamine or other amino acids which are accepted by the ASCT2 transporter (15, 16). The intraliposomal concentrations of the amino acid must be varied from 1 to 30 mM; the concentration of internal glutamine or other amino acids are given by the concentration in the initial reconstitution mixture and in the Amberlite equilibrating buffer (see also Note 7).

Acknowledgments This work was supported by the PRIN (Progetti di Ricerca Scientifica di Rilevante Interesse Nazionale) 2006 grant n. 2006054479 from MiUR (Ministero dell’Università e della Ricerca). The author is indebted to Dr. Francesca Oppedisano, Dr. Lorena Pochini, and Dr. Michele Galluccio for help in preparing the manuscript.

68

Indiveri

References 1. Palacin M, Estevez R, Bertran J, Zorzano A (1998) Molecular biology of mammalian plasma membrane amino acid transporters. Physiol Rev 78:969–1054 2. Bode BP (2001) Recent molecular advances in mammalian glutamine transport. J Nutr 131:2475S–2485S 3. Broer S (2002) Adaptation of plasma membrane amino acid transport mechanisms to physiological demands. Pflugers Arch 444: 457–466 4. Mackenzie B, Erickson JD (2004) Sodiumcoupled neutral amino acid (System N/A) transporters of the SLC38 gene family. Pflugers Arch 447:784–795 5. Kanai Y, Hediger MA (2004) The glutamate/ neutral amino acid transporter family SLC1: molecular, physiological and pharmacological aspects. Pflugers Arch 447:469–479 6. Verrey F, Ristic Z, Romeo E, Ramadan T, Makrides V, Dave MH, Wagner CA, Camargo SM (2005) Novel renal amino acid transporters. Annu Rev Physiol 67:557–572 7. McGivan JD, Bungard CI (2007) The transport of glutamine into mammalian cells. Front Biosci 12:874–882 8. Broer S (2008) Amino acid transport across mammalian intestinal and renal epithelia. Physiol Rev 88:249–286 9. Utsunomiya-Tate N, Endou H, Kanai Y (1996) Cloning and functional characterization of a system ASC-like Na+-dependent neutral amino acid transporter. J Biol Chem 271:14883–14890 10. Torres-Zamorano V, Leibach FH, Ganapathy V (1998) Sodium-dependent homo- and hetero-exchange of neutral amino acids mediated by the amino acid transporter ATB. Biochem Biophys Res Commun 245:824–829 11. Broer A, Brookes N, Ganapathy V, Dimmer KS, Wagner CA, Lang F, Broer S (1999) The astroglial ASCT2 amino acid transporter as a mediator of glutamine efflux. J Neurochem 73:2184–2194

12. Dolinska M, Dybel A, Zablocka B, Albrecht J (2003) Glutamine transport in C6 glioma cells shows ASCT2 system characteristics. Neurochem Int 43:501–507 13. Dolinska M, Zablocka B, Sonnewald U, Albrecht J (2004) Glutamine uptake and expression of mRNA’s of glutamine transporting proteins in mouse cerebellar and cerebral cortical astrocytes and neurons. Neurochem Int 44:75–81 14. Lim J, Lorentzen KA, Kistler J, Donaldson PJ (2006) Molecular identification and characterisation of the glycine transporter (GLYT1) and the glutamine/glutamate transporter (ASCT2) in the rat lens. Exp Eye Res 83:447–455 15. Oppedisano F, Pochini L, Galluccio M, Cavarelli M, Indiveri C (2004) Reconstitution into liposomes of the glutamine/amino acid transporter from renal cell plasma membrane: functional characterization, kinetics and activation by nucleotides. Biochim Biophys Acta 1667:122–131 16. Oppedisano F, Pochini L, Galluccio M, Indiveri C (2007) The glutamine/amino acid transporter (ASCT2) reconstituted in liposomes: transport mechanism, regulation by ATP and characterization of the glutamine/glutamate antiport. Biochim Biophys Acta 1768:291–298 17. Biber J, Stieger B, Haase W, Murer H (1981) A high yield preparation for rat kidney brush border membranes. Different behaviour of lysosomal markers. Biochim Biophys Acta 647:169–176 18. Krämer R, Heberger C (1986) Functional reconstitution of carrier proteins by removal of detergent with a hydrophobic ion exchange column. Biochim Biophys Acta 863:289–296 19. Palmieri F, Indiveri C, Bisaccia F, Iacobazzi V (1995) Mitochondrial metabolite carrier proteins: purification, reconstitution, and transport studies. Methods Enzymol 260:349–369 20. Pochini L, Oppedisano F, Indiveri C (2004) Reconstitution into liposomes and functional characterization of the carnitine transporter from renal cell plasma membrane. Biochim Biophys Acta 1661:78–86

Chapter 6 Use of Liposomes for Studying Interactions of Soluble Proteins with Cellular Membranes Chris T. Höfer, Andreas Herrmann, and Peter Müller Abstract Methods are described that have been used for characterizing the interaction of the soluble bovine seminal plasma protein PDC-109 with liposomes. PDC-109 binds to bull sperm cells upon ejaculation and is an important modulating factor of sperm cell maturation. The binding of the protein to sperm cells is mediated via lipids of the sperm plasma membrane. Most of our current knowledge about the molecular mechanisms of PDC-109–membrane interaction has been obtained by studies employing lipid vesicles. The experimental strategy described here can be applied to investigate the interaction of soluble proteins with membranes in order to understand their physiological role. Key words: Protein–membrane interaction, Membrane integrity, Phospholipid, Seminal plasma protein, PDC-109, Fluorescence spectroscopy, ESR spectroscopy

1. Introduction The interaction of soluble proteins with membranes is crucial for the regulation of biological activities. Proteins become recruited to the leaflets of plasma or intracellular membranes by binding to protein receptors or to lipids which results in adhesion and/or insertion of the protein to the membrane. Thereby, the protein and/or membrane structure and dynamics are influenced, finally resulting in a modified biological activity. Some examples for soluble proteins that modulate physiological functions upon binding to membranes are (1) phospholipases influencing phospholipid metabolism (1), (2) annexins which are supposed to play a role in membrane trafficking events such as exocytosis, endocytosis, and cell–cell adhesion (2), (3) bacterial toxins having deleterious impact on cells (3, 4), (4) mammalian seminal plasma proteins V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_6, © Humana Press, a part of Springer Science+Business Media, LLC 2010

69

70

Höfer, Herrmann, and Müller

which interact with the plasma membrane of sperm cells modulating the genesis of the cells, and (5) amyloid proteins like a-synuclein which is probably involved in the genesis of Parkinson’s disease (5, 6). With regard to proteins of seminal plasma, a protein family which is characterized by an Fn type II domain has gained special importance. This domain was originally described for the extracellular matrix protein fibronectin, in which it constitutes part of the collagen binding region and is implicated in a variety of extracellular binding events (7). Bull seminal plasma contains various Fn type II proteins, the most prominent representative being PDC-109 (also named BSP-A1/A2) (8). PDC-109 binds to sperm cells upon ejaculation and influences the capacitation process of the cells (9). To understand the physiological role of PDC109 on a molecular level, numerous studies, in particular using liposomes, have been perfomed (for reviews see (10, 11)). It has been found that the interaction with spermatozoa is realized via binding to membrane phospholipids carrying a choline head group, which are phosphatidylcholine (PC) and sphingomyelin (SM). The lipid-binding sites have been characterized from the protein crystal structure (12). Upon binding to vesicles, the protein influences membrane structure and dynamics, in that e.g., (1) the mobility of lipids is reduced, (2) membrane integrity is impaired, and (3) lipids are extracted from the membrane. Here, the application of various methods is described to identify and characterize important aspects of the interaction of PDC-109 with membranes by using liposomes.

2. Materials 2.1. Isolation of PDC-109

1. PDC-109 is purified from the seminal plasma of reproductively active Holstein bulls by a combination of affinity chromatography on heparin-Sepharose and DEAE-Sephadex chromatography as described by Calvete et al. (13). 2. The purity of the protein is verified by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) as well as reverse-phase HPLC, N-terminal sequence amino acid, and mass spectrometric analyses. 3. For the experiments, a 1 mM stock solution of the protein is prepared in Hepes-buffered salt solution (HBS) containing 145 mM NaCl and 5 mM HEPES, pH 7.4.

2.2. Lipids

1. Lipids are obtained from Avanti Polar Lipids (Alabaster, AL), if not stated otherwise and used without further purification. Stock solutions of lipids are prepared in chloroform or chloroform/ methanol (1:1) and stored at −80°C (see Notes 1 and 2).

Protein-Membrane Interaction

71

2.3. Fluorescence Measurements

1. All fluorescence measurements are done at 37°C using an Aminco Bowman Series 2 spectrofluorometer (SLM-AMINCO, Rochester, NY) with 4 nm slit width for both excitation and emission.

2.4. Electron Spin Resonance (ESR) Measurements

1. The spin-labeled analog of PC, 1-palmitoyl-2-(4-doxylpentanoyl)sn-glycero-3-phosphocholine (SL-PC), is synthesized according to (14). 2. ESR spectra are recorded at 4°C using a Bruker EMX spectrometer (Bruker, Karlsruhe, Germany) with the following parameters: modulation amplitude 1 G, power 20 mW, scan widths 100 G, accumulation nine times.

3. Methods 3.1. Preparation of Large Unilamellar Vesicles (LUVs)

1. Appropriate quantities of the lipid stock solutions (including when necessary fluorescent or spin-labeled lipids) are transferred into a glass tube, and the solvent is removed under a stream of nitrogen and subsequently under vacuum for 1 h. For labeled vesicles, one thereby obtains symmetrically labeled membranes, i.e., the analog is localized on both leaflets. 2. Lipids are resuspended in a small volume of ethanol (giving a final ethanol concentration below 1% (v/v)). HBS is added and the mixture is vortexed to induce the formation of multilamellar vesicles (see Note 3). 3. For the preparation of LUVs, the lipid suspension is subjected to five freeze–thaw cycles and extruded 10 times at 40°C through a 100-nm-diameter polycarbonate filter (Nucleopore GmbH, Tübingen, Germany) using either an extruder (Extruder, Lipex Biomembranes Inc., Vancouver, Canada) or a mini-extruder (Avanti Polar Lipids, Alabaster, AL) (15) (see Note 4).

3.2. Binding of PDC-109 to Liposomes Measured by Flotation Assay

1. As a first approach to characterize membrane binding of a protein, ultracentrifugation techniques can be used to separate the liposome-bound fraction of the protein from the free unbound protein. A variety of methods based upon this principle exists, and the appropriate one has to be chosen according to individual requirements. Either liposomes and bound protein can be pelleted using multilamellar vesicles (16) or sucrose-loaded vesicles (17), or liposomes and bound protein are floated on a sucrose density gradient (18). The flotation technique allows a clear separation of membrane-bound floating proteins especially for studying proteins that tend to aggregate or form large complexes that might be pelleted independently of the binding to liposomes. The assay can

72

Höfer, Herrmann, and Müller

serve for qualitative as well as quantitative analysis and can be utilized to investigate binding conditions such as pH, ionic strength, and lipid specificity. 2. LUVs are prepared consisting of 2 mM 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) and 0.5 mol% 1,2-dipalmitoylsn-glycero-3-phosphoethanolamine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) (N-NBD-DPPE) (see Notes 3 and 5). 3. Sucrose solutions (75% and 25% w/v) are prepared in HBS. 4. PDC-109 (12 µM) is mixed with different amounts of LUVs in a total volume of 150 µl in 1.5-ml Polyallomer Microfuge tubes for ultracentrifugation (Beckman Instruments Inc., Palo Alto, CA). 5. The solution is mixed thoroughly with 100 µl of 75% sucrose giving a final sucrose concentration of 30%. This bottom fraction is carefully overlayered with 200 µl of 25% sucrose and 50 µl of HBS. 6. The tubes are centrifuged for 1 h at 240,000 g and 4°C using an ultracentrifuge TL-100 and rotor TLA-100.3 (Beckman Instruments Inc., Palo Alto, CA). 7. Immediately after centrifugation, the fractions of the gradient are collected manually (see Note 6). 8. Aliquots of all fractions are analyzed by SDS-PAGE followed by silver staining. 9. The fluorescence of N-NBD-DPPE can be detected in the front part of the gel by scanning the gel before or after silverstaining on a fluorescent image analyzer FLA-3000 (FUJIFILM, Düsseldorf, Germany) with excitation at 473 nm. 10. Flotation of PDC-109 with DOPC-LUVs shows an increase in unbound protein with decreasing lipid concentration (Fig. 1). Above a protein-to-lipid ratio of 1:40, all protein is membrane-bound, whereas at 1:20 part of the protein is unbound. Since only the lipids of the outer leaflet of LUVs are accessible to protein binding, these data indicate a binding stoichiometry of about 10–20 lipid molecules per one protein molecule, which is in agreement with other studies (19). 3.3. Binding of PDC-109 to Liposomes Measured by Intrinsic Protein Fluorescence

1. Upon binding to membranes, protein structure is often modified. Those changes might be followed by measuring the intrinsic protein fluorescence if the protein contains fluorescent amino acids. Among these, tryptophan is most useful because of its comparatively high quantum yield. Since fluorescence is dependent on the local environment of the fluorophore, conformational changes can be monitored at least for those parts of the protein containing fluorescent amino acids. 2. PDC-109 is diluted from the stock solution to a concentration of 2.5 µM into a fluorescence cuvette containing lipid vesicles and

Protein-Membrane Interaction sucrose gradient

0% 25 %

= LUVs

NBD-labeled LUVs + protein

73

NBD fluorescence 1h

Collecting

240 000 g

the fractions

30 % 0% 25% 30%

Fig. 1. Top: Schematic diagram of the flotation assay. Bottom: Flotation assay analysis of PDC-109 binding to lipid vesicles. PDC-109 was floated with DOPC-LUVs in a sucrose step gradient at different protein-to-lipid ratios. After centrifugation, the fractions of the gradient with 30% sucrose (bottom fraction), 25% sucrose (middle fraction), and 0% sucrose (floating liposome fraction) were analyzed by SDS-PAGE, and subsequent silver staining showed a decrease of floating liposomebound protein with decreasing lipid concentration. The fluorescence of the membrane-anchored liposome label N-NBD-DPPE in the gel front was detected on a fluorescent image analyzer, and indicated the presence of LUVs in 0% and 25% fractions if the overall lipid concentration was sufficient. The bottom fraction was devoid of lipids 0.6

fluorescence intensity [a.u.]

3 0.5

2 0.4

1 0.3

0.2

0.1

0.0 300

320

340

360

380

400

420

wave length [nm] Fig. 2. Influence of lipid vesicles on the intrinsic fluorescence of PDC-109. The fluorescence spectra (excitation at 280 nm) of 2.5 µM PDC-109 were recorded at 37°C in the absence of liposomes (curve 1) and in the presence of 25 µM eggPC/ eggPE(2:1)-LUVs (curve 2 ) or eggPC-LUVs (curve 3). Note the shift of the wavelength of the maximum fluorescence intensity upon addition of LUVs

fluorescence spectra are recorded in the range of 300–400 nm (excitation wave length 280 nm) (Note 7). 3. Upon interaction with eggPC-LUVs, the intrinsic fluorescence of PDC-109 is increased and shifted to shorter wave lengths (Fig. 2, curves 1 and 3) (see Note 8). These data indicate a localization of tryptophan residues in a more hydrophobic

74

Höfer, Herrmann, and Müller

environment in the presence of vesicles which is probably caused by an intercalation of (part of) the protein into the bilayer (see Notes 9 and 10). 4. Mixing of PDC-109 with eggPC/eggPE(2:1)-LUVs results in lower changes of protein fluorescence compared to those of pure PC liposomes (Fig. 2, curve 2). This indicates that PDC-109 has a specificity for phosphorylcholine-containing phospholipids (19, 20). 3.4. Influence of PDC-109 on Membrane Integrity Measured by Calcein Leakage

1. LUVs are loaded with a high concentration of the nonpermeable fluorophore calcein, resulting in low fluorescence intensity due to self-quenching. A perturbation of membrane integrity results in a release of the fluorophore from the vesicles, which can be measured by the increase of fluorescence. 2. LUVs are prepared in HBS containing additionally 70 mM of calcein (Fluka Feinchemikalien, Neu-Ulm, Germany) (see Note 11). 3. Calcein-filled vesicles are separated from bulk calcein using NAP-5 columns (GE Healthcare, Freiburg, Germany) at room temperature and HBS as elution buffer (see Note 12). 4. Calcein-filled LUVs are diluted with HBS in a fluorescence cuvette (giving a final lipid concentration of 10–20 µM, see Note 13) while continuously stirring the solution. The timedependent fluorescence is monitored at 515 nm (excitation wave length 490 nm). After about 30 s, 5 µM PDC-109 is added and maximal leakage is determined by addition of 0.5% (w/v) Triton X-100 after about 230 s. 5. The addition of PDC-109 to eggPC-LUVs induces a release of calcein, indicating that the protein is able to disturb the membrane integrity of these vesicles (Fig. 3) (see Note 14). In contrast, the perturbation of eggPC/eggPE membranes in the presence of PDC-109 is significantly lower as seen from the low calcein leakage, a result again supporting the specificity of PDC-109 for phosphorylcholine-containing lipids.

3.5. Verification of PDC-109 Intercalation into Membrane Bilayer Measured by Förster Resonance Energy Transfer (FRET)

1. In order to see whether PDC-109 intercalates into the hydrophobic lipid phase upon binding to LUVs, FRET was recorded from tryptophan residues of the protein (donor) to the fluorescent lipid 1-hexadecanoyl-2-(1-pyrenedecanoyl)-sn-glycero-3phosphocholine (pyrPC, Invitrogen Ltd, Karlsruhe, Germany). This analog bears the pyrene group (acceptor) in the hydrophobic part of the membrane. A FRET signal strongly indicates a penetration of tryptophan residues into the lipid bilayer since for such signal both fluorophores have to approach below 10 nm. 2. EggPC-LUVs are labeled with 1 mol% pyrPC (see Subheading 3.1). Fluorescence spectra of labeled LUVs (final lipid concentration 20 µM) are recorded between 300 and

Protein-Membrane Interaction

75

normalized fluorescence

1.0

Triton X-100 0.8

0.6

PDC-109 1

0.4

0.2

2 0.0 0

50

100

150

200

250

time [s] Fig. 3. Influence of PDC-109 on the leakage of calcein from lipid vesicles. The time-dependent fluorescence of calcein was measured at 37°C for LUVs composed of eggPC (curve 1) or eggPC/eggPE (2:1) (curve 2). Leakage was induced by adding 5 µM PDC-109 to the vesicles at time zero, giving a final lipid-to-protein ratio of about 4. After 230 s, Triton X-100 was added (final concentration 0.5 % (w/v)) to obtain complete leakage of calcein, which was set to 1

fluorescence intensity [a.u.]

2.5

8 6

2.0

4 2

1.5

2

0 300

350

400

450

3

1.0

4

1 3

0.5

0.0 300

350

400

450

500

550

wave length [nm] Fig. 4. Förster resonance energy transfer (FRET) from tryptophan residues of PDC-109 to pyrene-labeled PC. eggPC-LUVs were labeled with 1 mol% pyrPC, and the fluorescence spectra of PDC-109 and pyrPC were recorded at 37°C in the absence (curve 1) and in the presence of unlabeled (L/P = 10, curve 2) and labeled PC-LUVs (L/P = 10, curve 3) with excitation at 280 nm as well as in the presence of labeled LUVs (L/P = 10, curve 4) with an excitation at 345 nm, i.e., direct excitation of pyrene. The inset shows the spectra with the entire range of the y-axes

500 nm in the absence and in the presence of 2 µM PDC-109 with excitation at 280 nm (Fig. 4). For comparison, fluorescence spectra of PDC-109 without vesicles and with unlabeled PC-LUVs were also recorded (see Notes 15 and 16).

76

Höfer, Herrmann, and Müller

3. The spectra reveal that part of the PDC-109 excitation energy is transferred to the pyrene moiety as seen from a decrease in tryptophan fluorescence at around 340 nm and an increase of pyrene (monomer) fluorescence between 380 and 450 nm (Fig. 4, compare unlabeled and labeled LUVs in the presence of protein, curves 2 and 3, respectively). 4. From these data, it can be concluded that upon binding of PDC-109 to lipid membranes (part of) the protein bearing tryptophan residues intercalates into the membrane phase (see Notes 9 and 17). 3.6. Interaction of PDC-109 with Spin-Labeled Lipids

1. ESR spectra of spin-labeled lipids localized in membranes reflect the lipid mobility and are sensitive to lipid–protein interactions (21). 2. EggPC-LUVs are labeled with 2.5 mol% of SL-PC (see Subheading 3.1). 3. Labeled LUVs (2 mM) are mixed with 0.2 mM PDC-109 and, after a 5 min incubation on ice, ESR spectra are recorded at 4°C. 4. Figure 5 (spectrum 1) shows a typical membrane spectrum of SL-PC in eggPC-LUVs, reflecting a partially restricted motion of the analog within the membrane. 5. In the presence of PDC-109, one observes an additionally superimposed spectrum which is especially visible in the region of the low-field peak (Fig. 5, spectrum 2, see arrow).

10 G 1

2

3

Fig. 5. Influence of PDC-109 on the ESR spectrum of spin-labeled membranes. 2 mM eggPC-LUVs were labeled with 0.05 mM SL-PC, and the ESR spectra were recorded at 4°C in the absence (spectrum 1) and in the presence (spectrum 2) of 0.2 mM PDC-109. The immobilized component caused by PDC-109 (see arrow) was extracted by spectra subtraction, giving spectrum 3

Protein-Membrane Interaction

77

6. The spectral component that originated in the presence of PDC-109 can be obtained by subtracting the ESR spectrum in the absence of protein from that in the presence of protein (Fig. 5, spectrum 3) (see Note 18). This spectrum reflects a decreased mobility of the analog as seen from the peak broadening, indicating that PDC-109 causes an effective restriction of the lipid mobility in PC membranes (22) (see Note 19). 3.7. Summary

The experiments described here allow characterization of the interaction of a soluble protein with membranes. Other approaches partly using liposomes are described in the literature. To study PDC-109, the following methods have been also applied: Fouriertransform infrared (FTIR) spectroscopy, CD spectroscopy, and calorimetric methods (differential scanning calorimetry, isothermal titration calorimetry) (see (10, 11)). The results of those experiments, in combination with in vitro and in vivo studies on the biologically relevant cellular system, allow the elucidation of the physiological role(s) of the respective protein (11, 12).

4. Notes 1. In case of storage for longer times, we prefer to store the lipids in the dry form, i.e., after evaporation of the solvent, at −80°C in order to prevent the lipids from any decomposition. 2. During handling with the organic solutions of lipids, the usage of any plastics should be prevented and solely glassware (tubes, pipette tips etc.) should be used since the solvents may dissolve substances (e.g., softeners) from the plastics. 3. When preparing liposomes, the detaching of lipids from the glass wall and their dissolution in the buffer is, in our opinion, an important point to be considered. Certain lipid species may be incompletely dissolved, resulting in a deviation of the desired lipid composition of liposomes. In general, unsaturated PC species or the often-used eggPC mixture can be easily dissolved by the addition of buffer and subsequent vortexing. However, problems may arise when using other lipid such as PE, SM, or cholesterol (also when used in a mixture with PC species) due to their distinct physicochemical parameters. For example, it is extremely difficult to dissolve pure cholesterol in an aqueous buffer. The dissolution of lipids could be facilitated by either pre-resuspending the dried lipids in a small volume of ethanol and/or by performing all steps of liposome production at higher temperatures. For example, when using SM species (which are characterized by

78

Höfer, Herrmann, and Müller

long fatty acid chains) we work at 65°C. After preparation of liposomes, lipid concentration of vesicles can be checked by measuring the phospholipid and, if necessary, the cholesterol content (see [23] and references therein). For measuring the lipid composition, the use of thin-layer chromatography or mass spectrometric methods (ESI, MALDI-TOF) is recommended. 4. Basically, different kinds of lipid vesicles can be used: small unilamellar vesicles (SUVs), large unilamellar vesicles (LUVs), or multilamellar vesicles (MLVs). SUVs are prepared by sonication of aqueous lipid dispersion. These vesicles are rapidly prepared. However, owing to their small size (diameter of about 30 nm), they have a large surface curvature questioning their use as a model for a biological membrane. Moreover, SUVs are comparatively unstable and fuse. Therefore, those vesicles should be used shortly after preparation. MLVs can also be prepared easily (see Subheading 3.1). However, due to the presence of more than one bilayer, these vesicles have an undefined accessible surface area. Therefore, use of MLVs could be unfavorable for certain problems. LUVs, although their preparation takes comparatively more effort, are appropriate for investigating numerous aspects of protein–membrane interaction, e.g., their membranes mimick biological membranes with regard to membrane curvature. After preparation, LUVs are supposed to be stable for several days; however, this might be checked. 5. The addition of the fluorescent lipid N-NBD-PE allows (1) visualization of the liposome fractions on a fluorescent image analyzer and (2) estimation of the loss of lipids during the preparation of LUVs by comparing the fluorescence of the lipid solution before and after vesicle preparation. For an accurate quantitative analysis, one has to determine the phospholipid and, if necessary, cholesterol content of LUVs (see Note 3). 6. There are two strategies for collecting the fractions. The three fractions of 250, 150, and 50 µl are collected from the bottom to the top with a Hamilton syringe, permitting a very clean separation of the fractions and restricting the liposomes mainly to the 0% sucrose floating fraction which is suitable for a quantitative analysis. But, if the protein adsorbs to the surface of the tubes as membrane-binding proteins often do, it is preferable to collect the fractions from the top to the bottom, as otherwise the floating fraction can become contaminated with protein sticking to the bottom of the tube. This can be checked by rinsing and boiling the emptied tube with 50 µl sample buffer and subsequent analysis by SDS-PAGE. After collecting the 0% and 25% fractions from the top, the 30% bottom fraction is taken from the bottom of the tube leaving

Protein-Membrane Interaction

79

a rest of 20 µl in the tube to prevent an incompletely removed lipid film from mixing with the bottom fraction. Therefore, the latter technique is not very suitable for quantitative analysis, but, nevertheless, can qualitatively show whether a protein binds to liposomes. If the utilized lipid concentration is high enough, one can visualize the localization of liposomes in the tube under UV light due to the fluorescence of N-NBDDPPE and estimate whether the floating of liposomes was succesful (Fig. 1). 7. In order to correct fluorescence spectra for the influence of light scattering caused by liposomes, the spectra of lipid vesicles in the absence of protein are also recorded and subtracted from those in the presence of protein. 8. For quantification of spectra, the fluorescence intensity at 333 nm is determined and normalized to the protein fluorescence in the absence of vesicles. By measuring the fluorescence changes at different lipid-to-protein ratios, the stoichiometry of protein lipid interaction can be determined which is about 10 for PDC-109 (19). Moreover, one can also estimate from the spectra the wavelength of the fluorescence maximum which is shifted to shorter wavelength (blue shift) upon change of tryptophan residues to a more hydrophobic environment. 9. If the protein to be investigated contains more than one tryptophan residue (for PDC-109 five residues), the exact determination which residue(s) intercalate(s) into the membrane bilayer is hampered since the measured fluorescence spectrum is the superposition of the spectra of each residue. The stepwise replacement of tryptophan residues by nonfluorescent amino acids using molecular biological approaches and measuring fluorescence spectra of those mutants in the presence of liposomes could allow the determination of the region(s) of the protein that are involved in membrane intercalation. Moreover, measurement of the life-time or quenching of fluorescence may allow distinguishing several tryptophan residues (usually 2–3). 10. In principle, an increase of fluorescence intensity and a blue shift of fluorescence reflect a change of the fluorophore to a more hydrophobic environment. However, changes of fluorescence with opposite tendencies in the presence of liposomes do not argue inevitably against a protein–membrane interaction since tryptophan fluorescence could also be influenced by other interactions, e.g., leading to fluorescence quenching. Moreover, if the protein solely adheres to the membrane surface, it is possible that the intrinsic fluorescence remains unchanged.

80

Höfer, Herrmann, and Müller

11. For preparing the calcein solution, first add only half of the necessary volume of buffer to calcein, then adjust the pH of the solution with NaOH to pH = 7.5, and finally fill up with buffer to the desired volume. The high calcein concentration (70,000 times the concentration of general fluorescence measurements) causes an extreme contamination of all labware that comes into contact with this solution. Therefore, all labware has to be cleaned thoroughly after contact with calcein solutions protecting other colleagues from undesirable calcein measurements. This especially holds for the extruder. To prevent its expensive cleaning, we used to reserve one mini-extruder in the lab solely for calcein measurements. 12. NAP-5 columns calibrated for the addition of 500 µl solution are used according to the manufacturer’s instructions. If the volume of liposome solution is smaller and has to remain undiluted, one can load the original solution onto the column and allow it to enter the gel bed completely. Subsequently, the eluting buffer is added. The running front containing the liposomes can be easily followed by eye because of the intense color of calcein and the eluant is collected by dropwise sampling. 13. In case LUVs are diluted after running on NAP-5 columns, the final lipid concentration varies resulting in different molar lipid-to-protein ratios in different experiments. Therefore, in order to estimate the exact ratio, the lipid concentration of calcein-loaded LUVs has to be measured after column chromatography (see [23]). 14. The degree of leakage mediated by PDC-109 increases with rising protein concentrations (23). The leakage kinetics is quantified by determining the amount of leakage (L) according to

L=

Ft − F0 Fmax − F0

where F0 and Fmax refer to the initial fluorescence intensity of calcein-filled LUVs before the addition of PDC-109 and the fluorescence intensity after addition of Triton X-100, respectively (see Fig. 3). Ft denotes to the fluorescence intensity reached in the plateau phase after addition of PDC-109 (in the experiment of Fig. 3 after 200 s). Intensities are corrected for dilution due to the addition of PDC-109 and Triton X-100. 15. Since pyrene is sensitive to quenching by oxygen, it is recommended to degas buffers with nitrogen before using. 16. As in our experiments, one may observe some pyrene fluorescence by excitation of a control sample (i.e., labeled LUVs in the absence of protein) at 280 nm. Therefore, this spectrum is subtracted from that measured in the presence of protein to extract the component solely caused by FRET (see Fig. 4).

Protein-Membrane Interaction

81

17. As a negative control, in order to exclude that FRET occurs from protein dissolved in the bulk buffer onto pyrPC in the membrane, a search was made for FRET between the soluble, tryptophan-containing protein bovine serum albumin and pyrPC labeled LUVs. Upon mixing bovine serum albumin with labeled vesicles, no FRET-dependent pyrene fluorescence was detected at all (24). 18. The extent of immobilization, i.e., the amount of spin-labeled lipids affected by the protein, can be estimated by quantifying the subtraction (22). However, this quantification presumes that (1) solely two fractions of lipids exist in the presence of the protein (i.e., a part of molecules influenced by the protein and the other not influenced) and (2) the exchange rate between these states is negligible in the timescale of ESR. 19. The spin-labeled PC used here is characterized by a short fatty acyl chain bearing the spin moiety at the sn2 position of the glycerol backbone. This feature allows incorporation of the analog also into preformed membranes labeling solely the outer leaflet of LUVs membranes (22). However, those experiments can also be performed by using other spin-labeled phospholipid analogs having different head groups and/or two long fatty acyl chains (long-chain PC analogs available from Avanti Polar Lipids, Alabaster, AL) or spin-labeled steroids (partly available from Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany) in order to characterize the lipid specificity of the protein (22, 25). Adding these analogs during liposome preparation results in symmetrically labeled vesicles (see also Subheading 3.1); however, the labeling of preformed membranes might be difficult.

Acknowledgments The work was supported by the Deutsche Forschungsgemeinschaft (Mu 1017/2). The fruitful cooperation during this project with Edda Töpfer-Petersen is kindly acknowledged. We thank Anja Arbuzova for critical reading of the manuscript.

References 1. Winget JM, Pan YH, Bahnson BJ (2006) The interfacial binding surface of phospholipase A2s: PLA2. Biochim Biophys Acta 1761:1260–1269 2. Waisman DM (1995) Annexin II tetramer: structure and function. Mol Cell Biochem 150:301–322

3. Cabiaux V, Wolff C, Ruysschaert JM (1997) Interaction with a lipid membrane: a key step in bacterial toxins virulence. Int J Biol Macromol 21:285–298 4. Palmer M (2004) Cholesterol and the activity of bacterial toxins. FEMS Microbiol Lett 238:281–289

82

Höfer, Herrmann, and Müller

5. Davidson WS, Jonas A, Clayton DF, George JM (1998) Stabilization of alpha-synuclein secondary structure upon binding to synthetic membranes. J Biol Chem 273:9443–9449 6. Beyer K (2007) Mechanistic aspects of Parkinson’s disease: alpha-synuclein and the biomembrane. Cell Biochem Biophys 47:285–299 7. Skorstengaard K, Thogersen HC, Petersen TE (1984) Complete primary structure of the collagen-binding domain of bovine fibronectin. Eur J Biochem 140:235–243 8. Manjunath P, Sairam MR (1987) Purification and biochemical characterization of three major acidic proteins (BSP-A1, BSP-A2 and BSP-A3) from bovine seminal plasma. Biochem J 241:685–692 9. Therien I, Bleau G, Manjunath P (1995) Phosphatidylcholine-binding proteins of bovine seminal plasma modulate capacitation of spermatozoa by heparin. Biol Reprod 52: 1372–1379 10. Wah DA, Fernández-Tornero C, Sanz L, Romero A, Calvete JJ (2002) Sperm coating mechanism from the 1.8 A crystal structure of PDC-109-phosphorylcholine complex. Structure 10:505–514 11. Manjunath P, Therien I (2002) Role of seminal plasma phospholipid-binding proteins in sperm membrane lipid modification that occurs during capacitation. J Reprod Immunol 53:109–119 12. Ekhlasi-Hundrieser M, Müller P, TöpferPetersen E (2008) Male secretory proteins sperm tools for fertilisation. In: Glander HJ, Paasch U (eds) Biology of male germ cells. Shaker Publisher GmbH, Aachen, Germany 13. Calvete JJ, Varela PF, Sanz L, Romero A, Mann K, Töpfer-Petersen E (1996) A procedure for the large-scale isolation of bovine seminal plasma proteins. Protein Expr Purif 8:48–56 14. Fellmann P, Zachowski A, Devaux PF (1994) Synthesis and use of spin-labeled lipids for studies of the transmembrane movement of phospholipids. Methods Mol Biol 27:161–175 15. Mayer LD, Hope MJ, Cullis RP, Janoff AS (1985) Solute distributions and trapping efficiencies observed in freeze-thawed multilamellar vesicles. Biochim Biophys Acta 817:193–196

16. Heuer K, Arbuzova A, Strauss H, Kofler M, Freund C (2005) The helically extended SH3 domain of the T cell adaptor protein ADAP is a novel lipid interaction domain. J Mol Biol 348:1025–1035 17. Buser CA, McLaughlin S (1998) Ultracentrifugation technique for measuring the binding of peptides and proteins to sucrose-loaded phospholipid vesicles. Methods Mol Biol 84:267–281 18. Bigay J, Casella JF, Drin G, Mesmin B, Antonny B (2005) ArfGAP1 responds to membrane curvature through the folding of a lipid packing sensor motif. EMBO J 24: 2244–2253 19. Desnoyers L, Manjunath P (1992) Major proteins of bovine seminal plasma exhibit novel interactions with phospholipid. J Biol Chem 267:10149–10155 20. Müller P, Erlemann KR, Müller K, Calvete JJ, Töpfer-Petersen E, Marienfeld K, Herrmann A (1998) Biophysical characterization of the interaction of bovine seminal plasma protein PDC-109 with phospholipid vesicles. Eur Biophys J 27:33–41 21. Marsh D, Horvath LI (1998) Structure, dynamics and composition of the lipidprotein interface. Perspectives from spinlabelling. Biochim Biophys Acta 1376: 267–296 22. Greube A, Müller K, Töpfer-Petersen E, Herrmann A, Müller P (2001) Influence of the bovine seminal plasma protein PDC-109 on the physical state of membranes. Biochemistry 40:8326–8334 23. Tannert A, Töpfer-Petersen E, Herrmann A, Müller K, Müller P (2007) The lipid composition modulates the influence of the bovine seminal plasma protein PDC-109 on membrane stability. Biochemistry 46:11621–11629 24. Greube A, Müller K, Töpfer-Petersen E, Herrmann A, Müller P (2004) Interaction of fibronectin type II proteins with membranes: the stallion seminal plasma protein SP-1/2. Biochemistry 43:464–472 25. Müller P, Greube A, Töpfer-Petersen E, Herrmann A (2002) Influence of the bovine seminal plasma protein PDC-109 on cholesterol in the presence of phospholipids. Eur Biophys J 31:438–447

Chapter 7 Liposomal Reconstitution of Monotopic Integral Membrane Proteins Zahra MirAfzali and David L. DeWitt Abstract In spite of considerable progress in the methodology for reconstitution of membrane proteins into the liposomes, a successful reconstitution still appears to be more an art than a science. Reconstitution of membrane proteins into bilayers is required for establishing several aspects of the functions of membrane proteins and lipids and for elaborating models of naturally occurring membranes. Cyclooxygenase (COX)-1 and -2 (also prostaglandin endoperoxide H2 synthase, PGHS-1 and -2) belong to the class of monotopic membrane proteins. Membrane-binding domains of both COX-1 and -2 contain four short, consecutive, amphipathic a-helices (A, B, C, and D). Crystal structures of the COXs indicate that basic, hydrophobic, and aromatic residues in the membrane-binding domain are oriented away from the protein core and form a surface on the enzyme, which has been proposed to interact with the lipid bilayer (1). In this chapter, we describe a fast and efficient method for direct incorporation of COX-1 and -2 isozymes – as models for monotopic integral membrane proteins – into preformed liposomes containing fatty acids without loss of activity. Key words: Monotopic membrane protein, Proteoliposomes, Liposome reconstitution, Direct incorporation, Membrane defect, Incorporating impurities into membranes, Cyclooxygenase enzyme

1. Introduction Monotopic membrane proteins such as cyclooxygenase (COX)-1 and -2 interact only with one leaflet of the lipid bilayer. Detergent removal is a classical method that is used for incorporation of integral membrane proteins into liposomes. However, this technique does not work with monotopic membrane proteins such as COX isozymes. Thus, it seems that a more generally applicable method is needed for this class of proteins. Medium-length-chain

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_7, © Humana Press, a part of Springer Science+Business Media, LLC 2010

83

84

MirAfzali and DeWitt

(C12–28) fatty acids, cholates, and cholesterol have been shown to promote the reconstitution of integral membrane proteins into liposomes. Therefore, one of the key features for successful incorporation of solubilized membrane proteins into preformed liposomes appears to be the state of organization of the lipid bilayer. Bilayers conducive to direct incorporation of large membrane proteins are achieved by incorporating “impurities”, such as fatty acids, lysophospholipids, mixtures of structurally different phospholipids, and cholesterol. The organizational defects in a bilayer decreases the activation energy for insertion of proteins and increases the free energy change associated with incorporation of proteins [2]. In this chapter, we explain the step by step process of incorporating a monotopic membrane protein into preformed liposomes using direct incorporation method.

2. Materials 2.1. Expression of Recombinant COX into SF21 Insect Cells

1. Spodoptera frugiperda (SF21) insect cells (Invitrogen Corporation, Carlsbad, CA). 2. Recombinant baculovirus expressing His-tagged COX-1 or COX-2 isozymes. 3. Media: HyQ-SFX-Insect serum-free insect cell media (Hyclone, Logan, UT) is supplemented with 0.1% pluronic F-68, 1× lipid concentrate, and 0.2% glucose. 4. Fernbach flasks (Bellso Biotechnology Inc., Vineland, NJ). 5. Innova model 4000 benchtop gyrotory incubator shaker (New Brunswick Scientific Co. Inc, Edison, NJ). 6. Hausser Nageotte Bright-Line Hemacytometer (Thermo Fisher Scientific, Waltham, MA). 7. Zeiss light microscope (Thermo Fisher Scientific, Waltham, MA). 8. Beckman refrigerated floor centrifuge (Beckman-Coulter, Fullerton, CA). 9. DuPont-Sorvall HS-4 centrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 10. Nalgene centrifuge bottles for Sovall HS-4 centrifuge rotor (Thermo Fisher Scientific, Waltham, MA).

2.2. Purification of His-tagged COX from Baculovirus-Infected SF21 Insect Cells

1. Bioneb cell disruption system (Glas-Col, Terre Haute, Indiana). 2. Nitrogen cylinder. 3. Nickel-homogenization buffer: 25 mM NaPO4, 100 mM NaCl, 20 mM imidazole, pH = 7.4.

Liposomal Reconstitution of Monotopic Integral Membrane Proteins

85

4. Nickel-wash buffer: 25 mM NaPO4, 300 mM NaCl, 20 mM imidazole, pH = 7.4. 5. Nickel-elution buffer: 25 mM NaPO4, 100 mM NaCl, 200 mM imidazole, pH = 7.4. 6. Dialysis buffer: 25 mM Tris-HCl, 50 mM KCl, pH 7.4. 7. Fast-flow Ni-NTA Sepharose resin (Qiagen Inc, Valencia, CA). 8. Polypropylene centrifuge tubes 50 ml with caps (Thermo Fisher Scientific, Waltham, MA). 9. Beckman refrigerated floor ultracentrifuge (BeckmanCoulter, Fullerton, CA). 10. Sorvall T865 ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 11. Ultracentrifuge tubes for Sorvall T865 ultracentrifuge rotor (Beckman-Coulter, Fullerton, CA). 12. Slide-A-Lyzer dialysis cassette with 10,000 MWCO (Thermo Fisher Scientific, Waltham, MA). 13. Amcon ultracentrifuge filter device with 3,000 MWCO (Millipore, Billerica, MA). 2.3. Preparation of Unilamellar Liposomes

1. Buchi rotavapor with temperature-controlled water bath connected to a vacuum pump (Buchi Corporation, New Castle, DE). 2. Round-bottom flask (250 ml). 3. 1,2-Dioleoyl-sn-glycero-3-phosphocholine (DOPC) (Avanti Polar Lipids, Alabaster, AL). 4. 1,2-Dioleoyl-sn-Glycero-3-[Phospho-L-Serine] (Sodium Salt) (DOPS) (Avanti Polar Lipids, Alabaster, AL). 5. Avanti mini liposome extruder with heating block (Avanti Polar Lipids, Alabaster, AL). 6. Two gas-tight Hamilton syringes (1 ml) (Hamilton Company, Reno, NV). 7. Nuclepore polycarbonate membrane filter (100 nm pore size, 19 mm) (Thermo Fisher Scientific, Waltham, MA). 8. Liposome buffer: 25 mM Tris–HCl, 50 mM KCl, pH 7.4.

2.4. Incorporation of COX Enzyme into Preformed Liposomes 2.5. Separation of Proteoliposomes from Free Proteins

1. Microfuge tubes (2 ml) (Thermo Fisher Scientific, Waltham, MA). 2. Dry block temperature controlled incubator (Thermo Fisher Scientific, Waltham, MA). 1. Beckman refrigerated floor ultracentrifuge (BeckmanCoulter, Fullerton, CA). 2. Sorvall SW 55 Ti ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA).

86

MirAfzali and DeWitt

3. Ultracentrifuge tubes for Sorvall SW 55 Ti ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 4. Syringe and needle (Thermo Fisher Scientific, Waltham, MA). 2.6. Protein Quantification Using Bicinchoninic Acid(BCA)

1. Bicinchoninic acid (BCA) protein assay reagent (Thermo Fisher Scientific, Waltham, MA). 2. Albumin standard (2 µg/µl) (Thermo Fisher Scientific, Waltham, MA). 3. Thermo Fisher UV/vis micro-plate reader (Thermo Fisher Scientific, Waltham, MA). 4. Disposable plates (96-wells) (Thermo Fisher Scientific, Waltham, MA). 5. Incubator oven (37°C) (Thermo Fisher Scientific, Waltham, MA).

2.7. Phosphorous Quantification

1. Required solutions a. 8.9 N H2SO4: 123.5 ml of concentrated H2SO4 is added to 376.5 ml of deionized water. b. Ascorbic acid (10%): 5 g of ascorbic acid is put into a 50-ml volumetric flask. Fifty milliliters of deionized water is added. c. Ammonium molybdate(VI) tetrahydrate (2.5%): 1.25 g of ammonium molybdate(VI) tetrahydrate is put into a 50-ml volumetric flask. Fifty milliliters of deionized water is added. d. Phosphorus standard solution (0.64 mM): 0.1 ml of 32 mM phosphorus standard solution is added to 4.9 ml of deionized water. 2. Metal rack for test tubes (Thermo Fisher Scientific, Waltham, MA). 3. Glass test tubes (Thermo Fisher Scientific, Waltham, MA). 4. Thermo Fisher laboratory oven (Thermo Fisher Scientific, Waltham, MA). 5. Thermo Fisher UV/vis spectrophotometer (Thermo Fisher Scientific, Waltham, MA). 6. Spectrophotometer cuvettes (Thermo Fisher Scientific, Waltham, MA).

2.8. Cyclooxygenase Assay

1. Assay buffer: 0.1 M Tris–HCl, pH 8.0. 2. Required solutions a. Heme in DMSO: 1 ml solution of 8 mM of heme in DMSO is made and kept in the freezer. b. Arachidonic acid buffered solution: 1 ml solution of 100 µM arachidonic acid in assay buffer is made and kept in the refrigerator.

Liposomal Reconstitution of Monotopic Integral Membrane Proteins

87

3. YSI 5300A biological oxygen monitor (YSI, Yellow Springs, OH) 4. Hamilton syringe (10 µl) connected to a long needle (Hamilton Company, Reno, NV). 2.9. Gold Labeling Proteoliposomes for Electron Microscopy

1. Beckman refrigerated floor ultracentrifuge (Beckman-Coulter, Fullerton, CA). 2. Sorvall SW 55 Ti ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 3. Ultracentrifuge tubes for Sorvall SW 55 Ti ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 4. Syringe and needle (Thermo Fisher Scientific, Waltham, MA).

2.10. Transmission Electron Microscopy on Gold Labeled Proteoliposomes

1. Affinity-purified rabbit anti-COX antibody (custom made). 2. Gold immunoprobe NANO-GOLD-goat anti-rabbit IGGNRF (1.4-nm particles) (Immunoprobe, Yaphank, NY). 3. Beckman refrigerated floor ultracentrifuge (Beckman-Coulter, Fullerton, CA). 4. Sorvall SW 55 Ti ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 5. Ultracentrifuge tubes for Sorvall SW 55 Ti ultracentrifuge rotor (Thermo Fisher Scientific, Waltham, MA). 6. Syringe and needle (Thermo Fisher Scientific, Waltham, MA). 7. Formvar (polyvinylformaldehyde)- carbon coated copper grids (G-300 mesh) (Electron Microscopy Sciences Inc, Fort Washington, PA). 8. Whatman filter paper #1 (Thermo Fisher Scientific, Waltham, MA). 9. JOEL transmission electron microscope (JOEL, Tokyo, Japan).

3. Methods Large quantities of COX enzyme are expressed in insect cells using baculovirus system containing the DNA of His-tagged COX enzyme. The purity of COX enzyme that is used for incorporation studies is more than 95%. The enzyme is purified using nickel-affinity chromatography method. In order to use this purification method, the protein is tagged with six histidines. In vitro mutagenesis is used to introduce a six-residue histidine sequence (His-tag) near the amino terminal end of the human COX-1 and -2. These isozymes are expressed using the baculovirus system. The His-tags are located one and two amino acids beyond the signal peptide cleavage sites of COX-1 and COX-2, respectively,

88

MirAfzali and DeWitt

positions that do not affect their activities or sensitivities to nonsteroidal anti-inflammatory drugs. When expressed in sf-21 cells, the His-tagged enzymes have Km values for arachidonate, and IC50 values for inhibition by nonsteroidal anti-inflammatory drugs that are similar to values reported for the nontagged enzymes. The His-tags allowed purification of the COX isozymes by a simplified protocol involving nickel-affinity chromatography [3]. Unilamellar liposomes containing impurities such as oleic acid are made in buffer and sized to 100 nm by extrusion using polycarbonate membrane filters. Various amounts of enzyme are added to fixed amount of liposomes in order to find the optimum protein-to-lipid ratio. Proteoliposomes (protein-incorporated liposomes) are separated from free liposomes (unincorporated proteins) using ficoll density gradient and ultracentrifugation methods. The incorporated proteins can be observed on the surface of the liposomes using immunogold labeling in combination with high-resolution transmission electron microscopy [4]. 3.1. Expression of Recombinant COX into SF21 Insect Cells

1. Spodoptera frugiperda (SF21) insect cells are added to sterile media in 1 L cultures at 27°C in 2.8-L Frenchback flasks shaken at 120 rpm using Innova model 4000 benchtop gyrotory incubator shaker. 2. The insect cells are counted every 8 h using Hausser Nageotte Bright-Line Hemacytometer and light microscope. 3. When cells reach a density of 1.5–2.0 × 106 cells/ml, baculovirus containing the DNA of COX enzyme is added at a multiplicity of infection of 1.0, and the infection is allowed to proceed for 72–96 h. 4. During the 72–96 h wait period, the cells are counted and lysed every 8 h and assayed for COX activity in order to monitor the progress of protein expression (see Note 1). 5. To harvest the cells, the cells and media are poured into Nalgene centrifuge bottles for Sorvall HS-4 centrifuge rotor. 6. The cells are pelleted at the bottom of the centrifuges tube using the Beckman refrigerated floor centrifuge operating at 10,000× g for 30 min. 7. The cells are washed using phosphate buffered saline (PBS) buffer and pelleted again using centrifugation. 8. The buffer is separated from the pelleted cells. 9. The tubes containing the cells are flash-frozen using liquid nitrogen and stored in a −80°C freezer.

3.2. Purification of His-tagged COX from Baculovirus-Infected SF21 Insect Cells

1. The tubes containing frozen cells are thawed at room temperature. 2. SF21 cell pellets are resuspended in 3 ml/g wet weight of nickel-homogenization buffer containing (1% v/v) of Tween 20 detergent.

Liposomal Reconstitution of Monotopic Integral Membrane Proteins

89

3. The cells are disrupted by Bioneb cell disruption system. The Bioneb is connected to high pressure nitrogen at 150 psi. The cells are passed through the system at least three times in order to achieve 100% disruption of all cells. 4. The cells are checked under the light microscope to make sure that they are completely lysed. 5. After low-speed (10,000× g for 1 h) centrifugation to pellet cell debris, the supernant which contains the solubilized COX is poured into cramp-seal ultracentrifuge tubes. 6. The insoluble material is removed by ultracentrifugation at 100,000× g for 2 h. 7. The solubilized extract is next incubated with fast-flow Ni-NTA resin overnight at 4°C with rocking. 8. The resin is poured into a column and washed with five volumes of nickel-homogenization buffer containing 0.1% Tween 20 detergent. 9. Next the resin is washed with three volumes of nickel-wash buffer containing 0.1% Tween 20. 10. The protein is eluted with nickel-elution buffer containing 0.2% Tween 20. 11. Fractions with high specific COX activity are pooled and concentrated in the Amicon ultracentrifugal filter device with 30,000 MWCO. 12. The purified COX enzyme is injected into a Slide-A-Lyzer (10,000 MWCO) dialysis cassette using a syringe. 13. The dialysis cassette is dialyzed against 2 L dialysis buffer. 14. The dialysis buffer is changed every 6 h for 24 h (see Note 2). 3.3. Preparation of Unilamellar Liposomes

1. Unilamellar liposomes are prepared using a DOPC: DOPS molar ratio of 3:7 and 9.2% (w/w) oleic acid. DOPC, DOPS, and oleic acid are dissolved in 1:1 (v/v) mixture of chloroform and methanol in a round bottom flask. 2. The round bottom flask is dried using a Buchi rotary evaporator connected to a vacuum pump for 4 h. 3. The resulting film is hydrated in a liposome buffer for 2 h. The total concentration of lipid is 38 mM. 4. The liposome dispersion is extruded 20 times through a polycarbonate membrane with 100 nm pore diameter using a handheld mini-extruder (see Note 3). Extruded liposomes form a transparent milky suspension.

3.4. Incorporation of COX Enzyme into Preformed Liposomes

1. The COX protein is added to the preformed liposomes. The concentration of the lipids in the liposome solution is 38 mM. A 1:500 protein: lipid molar ratio is used. 2. The mixture is incubated for 20 min at 37°C.

90

MirAfzali and DeWitt

3.5. Separation of Proteoliposomes from Free Proteins

1. A 0.5-ml liposome suspension is mixed with 1 ml of 30% (w/v) Ficoll in liposome buffer to give a final concentration of 20% (w/v) Ficoll. 2. The liposome suspension is transferred to a 5-ml ultracentrifuge tube. 3. Three milliliters of 10% (w/v) Ficoll is gently layered on top of the liposome suspension. 4. The swing-out rotor (Beckman SW 50.1) is used for 30 min at 100,000× g at 4°C. 5. The liposomes and proteoliposomes create a band at the interface between the 0% Ficoll and 10% Ficoll layers. A 1-ml syringe is used to isolate the proteoliposomes band from the buffer solution. The activity and concentration of the enzyme in the isolated proteoliposomes band are measured in order to calculate the incorporation efficiency of COX enzyme. 6. The unincorporated protein is retained in the 30% Ficoll layer. The activity and the concentration of the free protein are measured in order to calculate the incorporation efficiency of the COX enzyme.

3.6. Protein Quantification Using Bicinchoninic Acid (BCA)

1. The oven is turned on at 37°C. 2. Seven 1.5-ml tubes 0, 0.2, 0.4, 0.6, 0.8, 1, and 1.2 are labeled for the standard curve. 3. For the standard curve, 0, 10, 20, 30, 40, 50, and 60 µl of (2 µg/µl of albumin stock solution) is added to 100, 90, 80, 70, 60, 50, and 40 µl of water. 4. Samples for standard curve are added in triplicates at a volume of 10 µl to the wells in the 96-well plate. 5. Ten microliters of diluted aliquots of COX samples is added in triplicates to the wells in the 96-well plate. 6. Two hundred microliters of the BCA working reagent is added to each of the wells (some of the wells may turn purple after addition of the working reagent). 7. The plate that is covered by a lid is incubated for 30 min in the 37°C oven. 8. The plate is read at 562 nm. 9. The concentration of COX enzyme is calculated by linear regression of the standard curve.

3.7. Phosphorous Quantification

1. Liposome samples (~ 0.1 µmoles phosphorus) are placed in glass tubes. 2. Samples for standard curve are prepared by adding 0 (0 µmoles), 50 (0.032 µmoles), 100 (0.064 µl), 175 (0.112 µmoles), 250

Liposomal Reconstitution of Monotopic Integral Membrane Proteins

91

(0.160 µmoles), and 350 µl (0.224 µmoles) of phosphorus standard solution to each glass tube. 3. The sample and standard tubes are placed in a metal test tube rack. 4. To each of the tubes, 0.45 ml 8.9 N H2SO4 is added. Each tube is capped with a glass marble. 5. The samples are kept in oven at 200–215°C for 25 min. Temperature must be above 200°C. 6. Tubes are removed from the oven and cooled for 5 min. 7. To each tube, 150 µl H2O2 is added. 8. The tubes are returned to the oven and heated for an additional 30 min. The samples are colorless at this point (see Note 4). Tubes are cooled to ambient temperature. 9. To each tube, 3.9 ml deionized water is added. 10. To each tube, 0.5 ml of ammonium molybdate(VI) tetrahydrate solution is added. 11. The tubes are vortexed for 1 min. 12. To each tube, 0.5 ml ascorbic acid solution is added . 13. The tubes are vortexed for 1 min. 14. The tubes are heated at100°C for 7 min. Each tube is capped with a glass marble. 15. The tubes are cooled to room temperature. 16. The absorbance of each standard and sample is read at 820 nm. 17. The total amount of phosphorous in sample is calculated by linear regression of the standard curve. 3.8. COX Assay

1. To each chamber of YSI 5300A biological oxygen monitor device, 3 ml of assay buffer is added. The buffer is warmed up in to 37°C using the temperature control unit on the device. 2. The buffer inside the chamber is stirred using the magnet inside the chamber. 3. To the chamber, 50 µl of COX protein in added. 4. To the chamber ,5 µl of heme solution in DMSO is added (see Note 5). 5. The oxygen probe is inserted into the chamber. 6. Using a Hamilton syringe, 10 µl of arachidonic acid solution in buffer is added to the chamber. 7. The recorder that is connected to the oxygen probe records the rate of oxygen consumption. 8. The activity of the COX enzyme is calculated on the basis of the oxygen consumption rate.

92

MirAfzali and DeWitt

3.9. Gold Labeling Proteoliposomes for Electron Microscopy

1. About 100 ml of proteoliposomes is placed into an Eppendorf tube and 10 ml of primary antibody is added. The molar ratio of COX to antibody is 1:4. 2. The mixture is incubated for 30 min in room temperature with occasional shaking. 3. The nonattached primary antibodies are separated from the complex of proteoliposomes–primary antibodies using the Ficoll gradient technique, which was described previously. 4. The liposomes and the complex of proteoliposomes–primary antibodies are separated at the interface between the 0% Ficoll and 10% Ficoll layers. The unattached antibody will remain in the 30% Ficoll layer. 5. The complex of proteoliposomes–primary antibodies is placed into an Eppendorf tube and about 100 ml of gold immunoprobe NANOGOLD-anti rabbit (NRF) (1.4 nm particle attached to affinity-purified Fab fragment, raised in goat, against rabbit IgG (whole molecule)) is added to the tube and incubated for 30 min in room temperature with occasional shaking. 6. The nonattached gold-labeled secondary antibodies are separated from the complex of proteoliposomes–primary antibodies– gold labeled secondary antibodies using the Ficoll gradient method. 7. The liposomes and the complex of proteoliposomes–primary antibodies–gold labeled secondary antibodies are separated at the interface between the 0% Ficoll and 10% Ficoll layers. The unattached gold-labeled secondary antibody will remain in the 30% Ficoll layer.

3.10. Transmission Electron Microscopy on Gold Labeled Proteoliposomes

1. A drop of liposome solution (concentration about 5 mM of lipid) is placed on Formvar (polyvinylformaldehyde)-carbon coated copper grid G-300 mesh, and after 1 min the excess is removed with a Whatman filter paper #1. 2. A thin film is left on the grid and allowed to air dry. One drop of 0.1% solution of uranyl acetate is placed the grid (see Note 6). 3. After 1 min, this drop is again removed with the filter paper, and the resulting stained film is dried in a dust-free place. 4. A JEOL transmission electron microscope operating at 100 kV under vacuum is used to observe the liposomes (Figs. 1 and 2). Fixation and dehydration of liposome for electron microscopy flattens the liposomes and allows one to observe gold labeling on both sides of the liposomes. The average number of gold spots observed in each proteoliposomes is 123 ± 33 (n = 3). Assuming that average proteoliposomes are spheres of 100 nm, and that the two leaflets essentially double the lipid surface area, the total

Liposomal Reconstitution of Monotopic Integral Membrane Proteins

93

Fig. 1. Negative stain transmission electron micrograph of liposomes reconstituted with COX-2 and immunogold labeled with anti-COX-2 antisera (bar = 100 nm, magnification = 200,000). 0.1% uranyl acetate stain was used. Reproduced with permission from (4)

Fig. 2. Negative stain transmission electron micrograph of liposomes reconstituted with COX-2 (bar = 200,000 nm, magnification = 67,000). 1% uranyl acetate stain was used. Reproduced with permission from (4)

surface area of the proteoliposomes is (4p × 500 Å 2) × 2 = 6,280,000 Å2 of phospholipid. DOPC has been calculated to have a surface area of 59 Å2 per molecule. To simplify these calculations, we assumed that DOPS has the same surface area as DOPC. Using these assumptions, it calculated that there are 106,440 molecules of phospholipid per liposome. With our estimated incorporation ratio of 1,000:1 for phospholipid to COX per liposome, this would predict 106 molecules of COX per liposome, close to the 123 average spots observed.

94

MirAfzali and DeWitt

4. Notes 1. COX-2 is expressed in significantly higher level than COX-1 using the baculovirus system. 2. The buffer is exchanged to a nonphosphate-based buffer. The concentration of lipid in proteoliposomes is measured by phosphorous assay and therefore the buffer solution should be phosphate-free. 3. Liposomes should be extruded at a temperature that is above the liquid-to-gel phase transition temperature (Tc) of the lipid mixture. Attempts to extrude below the Tc will be unsuccessful, as the membrane has a tendency to foul with rigid membranes that cannot pass through the pores. 4. H2O2 is added in order to bleach the brown solution. If any brown color persists, 50 µl H2O2 is added to all cooled tubes and the tubes are heated again for 15 min until they become colorless. 5. During the purification process, the heme is stripped off from COX enzyme. Therefore, heme should be added to the enzyme before assaying the enzyme. 6. Transmission electron microscopy on gold-labeled liposome is carried out using a diluted stain (0.1–0.2 % solution of uranyl acetate). Because the NANOGOLD particles are small, overstaining with uranyl acetate may tend to obscure direct visualization of individual NANOGOLD particles. Therefore a diluted stain is used.

Acknowledgments The authors would like to thank Dr. Alicia Pastor and Mr. Robert Pcionek from Michigan State University Center for Advanced Microscopy for their help with the electron microscopy work.

References 1. Picot D, Loll PJ, Garavito RM (1994) The X-ray crystal structure of the membrane protein prostaglandin H2 synthase-1. Nature 367: 243–249 2. Jain MK, Zakim D (1987) The spontaneous incorporation of proteins into preformed bilayers. Biochim Biophys Acta 906: 33–67

3. Smith T, Leipprandt J, DeWitt D (2000) Purification and characterization of the human recombinant histidine-tagged prostaglandin endoperoxide H synthases-1 and -2. Arch Biochem Biophys 375:195–200 4. Mirafzali Z, Leipprandt J, DeWitt D (2005) Fast, efficient reconstitution of cyclooxygenases into proteoliposomes. Arch Biochem Biophys 443:60–65

Chapter 8 The Reconstitution of Actin Polymerization on Liposomes Mark Stamnes and Weidong Xu Abstract Membrane-associated actin polymerization is of considerable interest due to its role in cell migration and the motility of intracellular organelles. Intensive research efforts are underway to investigate the physiological role of membrane-associated actin as well as the regulation and mechanics of actin assembly. Branched actin polymerization on membranes is catalyzed by the Arp2/3 complex. Signaling events leading to the activation of the guanosine triphosphate (GTP)-binding protein Cdc42 stimulate Arp2/3dependent actin polymerization. We have studied the role of Cdc42 at the Golgi apparatus in part by reconstituting actin polymerization on isolated Golgi membranes and on liposomes. In this manner, we showed that cytosolic proteins are sufficient for actin assembly on a phospholipid bilayer. Here we describe methods for the cell-free reconstitution of membrane-associated actin polymerization using liposomes and brain cytosol. Key words: Liposome, Actin, Arp2/3, Wiscott–Aldrich syndrome protein (WASP), Cdc42

1. Introduction The actin cytoskeleton plays two basic roles connected to cell migration and the motility of intracellular organelles. First, actin microfilaments serve as tracks for myosin-motor-based motility. The best characterized role for an actomyosin complex is in muscle contraction (1). Myosin-based movement along actin is also utilized for vesicle targeting, organelle positioning, and the retraction of the cell’s trailing edge during migration (2). A second role for actin is to provide a propulsive force directly through Arp2/3catalyzed polymerization at a membrane (3). The most dramatic examples of this are the comet-tail type of motility used by endosomes that is also usurped by pathogenic bacteria such as Listeria. Similarly, actin polymerization is used directly to force the extension

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_8, © Humana Press, a part of Springer Science+Business Media, LLC 2010

95

96

Stamnes and Xu

of the leading edge membrane during cell migration. The ability to utilize actin polymerization to direct motility requires precise temporal and spatial regulation. A key regulatory mechanism for Arp2/3-dependent actin polymerization involves Cdc42, a member of the Rho-family of guanosine triphosphate (GTP)-binding proteins, and its effectors N-WASP or WAVE (4). GTP-bound Cdc42 causes a conformation change in N-WASP that exposes a C-terminal Arp2/3-binding domain. The actin polymerization activity of Arp2/3 is stimulated when bound to N-WASP. Thus, Cdc42 activation can stimulate actin polymerization at a distinct time and place. In many cell types, Cdc42 is localized to the Golgi apparatus through a binding interaction with the vesicle-coat protein, coatomer (5). There, Cdc42 serves to regulate both actin and microtubule-dependent trafficking events. We have investigated the role of Cdc42 at the Golgi complex, in part through the cell-free reconstitution of Cdc42-regulated Arp2/3dependent actin polymerization on isolated membranes and liposomes (6–10). The ability to reconstitute actin polymerization in vitro has served as a powerful tool for dissecting the mechanisms and regulation of this process at the molecular level. Indeed, the primary catalyst of actin polymerization, the Arp2/3 complex, was discovered through a biochemical purification using cell-free actin polymerization as an assay (11). The reconstitution of actin polymerization on biological and liposomal membranes has revealed details about the regulation and role of actin in the secretory pathway (5). It was recently demonstrated that the vesicle-associated GTP-binding protein ARF1 causes comet-tail-like motility of liposomes (12). We have exploited ARF1 and Cdc42 dependent actin polymerization on liposomes to demonstrate that this process involves the recruitment of Arp2/3 to the membrane (8). Here we provide detailed methods for reconstituting actin polymerization on membranes using brain cytosol and liposomes.

2. Materials 2.1. Preparation of Bovine-Brain Cytosol

1. Brain buffer: 25 mM Tris-HCl, pH 7.4, 320 mM sucrose. 2. Protease inhibitor stock solutions: 10 mg/ml aprotinin in H2O, 10 mg/ml leupeptin in H2O, 1 mM pepstatin A in dimethyl sulfoxide, 50 mM 1,10-phenanthroline in H2O, pH 5.0, 100 mM phenylmethanesulfonyl fluoride (PMSF) in 2-propanol.

Actin Polymerization on Liposomes

97

3. Breaking buffer: 25 mM Tris-HCL, pH 8.0, 500 mM KCl, 250 mM sucrose, 2 mM EGTA. Add dithiothreitol (DTT) and protease inhibitors (from the stock solutions) immediately before use to the following final concentrations, 1 mM DTT, 2 mg/ml aprotinin, 0.5 mg/ml leupeptin, 2 mM pepstatin A, 500 mM 1,10-phenanthroline, 1 mM PMSF. 4. Dialysis buffer: 25 mM Tris-HCl, pH 8.0, 50 mM KCl, add 1 mM DTT just prior to use. (see Note 1). 5. Dialysis tubing with 12,00–14,000 molecular weight cut off and 2.8 cm diameter (Membra-Cel®). 2.2. Preparation of Liposomal Membranes

1. Phospholipids: 10 mg/ml phospholipid stock solutions are prepared in chloroform. A bovine liver lipid extract (Avanti Polar Lipids) is an inexpensive source to provide a complex phospholipid bilayer. (see Note 2). 2. Resuspension buffer: 20 mM HEPES, pH 7.2, 150 mM potassium acetate, 250 mM sucrose.

2.3. Actin Polymerization Reactions

1. 5× Reaction buffer: 125 mM HEPES, pH 7.2, 12.5 mM magnesium acetate, 75 mM potassium chloride. 2. 100× ATP regenerating stock solution: 100 mM creatine phosphate, 25 mM UTP, 5 mM ATP. 3. Creatine phosphokinase stock solution: 1,000 units/ml in H 2O. 4. GTPgS solution: 10 mM in H2O. 5. 45% Sucrose/Tris (ST) solution: 10 mM Tris-HCl, pH 7.4, 45% sucrose (weight/weight, i.e. 45 g sucrose plus 55 ml water). 6. 35% ST solution: 10 mM Tris-HCl, pH 7.4, 35% sucrose (weight/weight). 7. 15% ST solution: 10 mM Tris-HCl, pH 7.4, 15% sucrose (weight/weight). 8. TCA stock solution: 100% trichloroacetic acid (weight/volume) in water.

2.4. Characterizing Actin Polymerization by SDS-PAGE

1. Laemmli sample buffer: 100 mM Tris-HCl, pH 6.8, 3% (weight/volume) sodium dodecyl sulfate (SDS), 10% (volume/volume) glycerol, 715 mM-mercaptoethanol, 0.03% (weight/volume) bromophenol blue.

2.5. Characterizing Actin Polymerization by Western Blotting

1. TBS-Tween solution: 20 mM Tris-HCl, pH 7.6, 137 mM NaCl, 0.5% Tween-20 (volume/volume). 2. Blocking solution: 5% non-fat milk powder (weight/volume) in TBS-Tween solution.

98

Stamnes and Xu

3. Methods The general procedure for the reconstitution of actin polymerization involves incubating the membranes with cytosol under conditions where GTP-binding proteins are active. This may be done by including the non-hydrolyzable GTP analog GTPgS in the reaction. Following the incubation, the liposomes are isolated from the reaction by centrifugation and analyzed by SDS-PAGE and Western blotting. A key to these experiments is obtaining good-quality cytosol to provide a source of actin, Arp2/3, and regulatory proteins. 3.1. Preparation of Bovine-Brain Cytosol

1. Obtain fresh bovine brains from a slaughter house. Two to three brains provide a 100–300 ml of cytosol. The brains should be placed in ice-cold brain buffer as soon as possible and stored in the buffer on ice while in transit. All of the following steps should be performed on ice or in a 4°C cold room. 2. Remove and discard the brain stem and cerebellum. Carefully peel the meninges and blood vessels from the cerebrum using forceps. Weigh the remaining cerebral tissue. 3. Chop the cerebrum into approximately 1 cm pieces using scissors. Immediately place the brain tissue into a small volume of breaking buffer (with protease inhibitors). Once the tissue is fully chopped, add additional breaking buffer such that the final volume of brain tissue plus buffer equals 1.25 L per 500 g tissue. 4. The tissue is homogenized using a blender. Blend the sample 2 times at a high setting for 30 s each. 5. Pour the homogenate into 500 ml centrifuge bottles. Centrifuge for 1 h at 14,000× g in a Sorvall SLA-3000 rotor (Thermo Scientific) or equivalent. 6. Decant the supernatant and pour into polycarbonate or quickseal tubes for Type 45 Ti rotor (Beckman-Coulter). Spin for 90 min at 140,000× g. 7. Collect the supernatant and distribute to several 1 m sections of dialysis tubing. Seal the tubing by tying knots or using clips. Be certain to leave extra space in the tubing as the retentate volume will expand considerably during the dialysis. Dialyze the sample for 4 h in 30 L of dialysis buffer. Switch the tubing to fresh dialysis buffer and dialyze for an additional 4–8 h (see Note 3). 8. Record the volume of the retentate and spin the sample for 90 min at 140,000× g using the Type 45 Ti rotor. Collect the supernatant. 9. The cytosol is concentrated by precipitation with ammonium sulfate. Slowly add 36.5 g of ammonium sulfate per 100 ml

Actin Polymerization on Liposomes

99

cytosol while stirring to obtain a final concentration of 60% saturation. Allow the sample to stir for an additional 30 min. 10. Recover the precipitate by centrifugation in 500 ml bottles in the SLA-3000 rotor 30 min. at 14,000× g. Discard the supernatant. Resuspend the pellet in dialysis buffer using one-sixth the volume recorded in step 8 (see Note 4). 11. Dialyze the sample against 30 L dialysis buffer for 4 h. Transfer the dialysis tubing to 30 L fresh dialysis buffer without DTT for an additional 4–8 h (see Note 3). 12. Collect the retentate and centrifuge for 90 min at 140,000× g using the Type 45Ti rotor. Recover the supernatant. 13. Mix the sample and determine the protein concentration of the cytosol using bicinchoninic acid (BCA) protein assay kit (Pierce). A typical preparation has a concentration of ~10 mg/ml. 14. Freeze the cytosol in aliquots using liquid nitrogen. Store the cytosol at −80°C. 3.2. Preparation of Liposomal Membranes

1. Dispense the appropriate volumes of the phospholipid stock solutions into the bottom of a glass round-bottom tube (see Note 5). 2. Evaporate the chloroform by blowing air across the tube or using an Evap-O-Rac (Cole-Parmer). 3. The phospholipids are hydrated by adding resuspension buffer and mixing using a vortex mixer (see Note 6). The final phospholipid concentration should be 2.3 mg/ml. 4. The lipid resuspension is subjected to 10 freeze-thaw cycles using a dry-ice/methanol bath. 5. Unilamellar vesicles with a uniform size can be generated by extrusion through a polycarbonate membrane. Nine passages through a LiposoFast membrane extrusion device (Avestin) containing a membrane with pore diameter of 400 nm works well. 6. The liposomes are best if used immediately, but we have successfully stored them as aliquots at −80°C for several months.

3.3. Actin Polymerization Reactions

1. The reactions are set up in 1.5 ml plastic tubes with a final volume of 200 ml (see Note 7). The final reaction conditions are as follows: 1× reaction buffer, 1× ATP regenerating solution, 8 units/ml creatine phosphokinase, 0.2 mg/ml bovine brain cytosol, 0.23 mg/ml liposomes. Actin polymerization may be stimulated by adding 20 mM GTPgS to activate GTP-binding proteins. 2. Incubate the reactions at 37°C for 20 min. 3. At the end of the incubation, isolate the liposomes by microcentrifugation for 30 min at 15,000 × g.

100

Stamnes and Xu

4. The liposomes may be separated from non-membrane-associated F-actin by flotation through a sucrose cushion. Begin by resuspending the liposome pellet in 50 ml of 45% ST solution (see Note 7). The sample is placed at the bottom of a 7 × 20 pollyallomer ultracentrifugation tube (Beckman-Coulter). The sample is overlayed with 130 ml 35% ST solution followed by 20 ml 15% ST solution. The sample is subjected to centrifugation for 45 min at 386,000× g using a TLA-100 rotor (Beckman Coulter). 5. After centrifugation, the sample tubes are frozen by careful immersion in liquid nitrogen. Slice the tube with a guillotine or heated razor blade to recover the top 100 ml of the gradient. 6. Precipitate the liposome-associated proteins by addition of trichloroacetic acid to a final concentration of 10%. Recover the precipitate by spinning in a microcentrifuge for 15 min. 7. Wash the precipitate with ice-cold acetone. 3.4. Characterizing Actin Polymerization by SDS-PAGE

1. Resuspend the precipitate using Laemmli sample buffer. 2. Load sample onto a 12% SDS-PAGE Gel (BioRad) and install the gel into an electrophoresis apparatus containing running buffer (BioRad). 3. Carry out electrophoresis at 200 V until the bromophenol blue dye front reaches the bottom of the gel. 4. Liposome-bound proteins can be visualized using Coomassie blue or silver stain as shown in Fig. 1.

Fig. 1. Shown is a Coomassie-blue-stained SDS PAGE gel of liposomes isolated by flotation from incubations carried out with brain cytosol. When the non-hydrolyzable GTP analog GTPgS is included in the reaction, multiple proteins bind to the liposomes. Actin is observed as a prominent band at 42 kD. The actin plus end-binding toxin cytochalasin D partly inhibits actin polymerization. Brefeldin A (BFA), an inhibitor of the GTP-binding protein ARF1, is a potent inhbitor of actin assembly on the liposomes. A peptide p23 that blocks Cdc42 recruitment to the membrane also inhibits actin polymerization on liposomes

Actin Polymerization on Liposomes

101

Fig. 2. Shown is a Western blot probed with antibodies against actin and tubulin. Tubulin binding to the liposomes is not affected by the activation of GTP-binding proteins. By contrast, actin levels are greatly increased in the presence of GTPgS. Actin polymerization is inhibited by cytochalasin D (CytoD), brefeldin A (BFA), and the p23 peptide as described in the legend to Fig. 1

3.5. Characterizing Actin Polymerization by Western Blotting

1. Instead of staining the SDS gel, the presence or absence of specific proteins can be assessed by Western blot analysis. 2. Transfer proteins onto a nitrocellulose membrane using an electroblot apparatus containing transfer buffer (BioRad). 3. Block the nitrocellulose by incubating in blocking solution for 1 h at room temperature. 4. Wash the membrane two times with TBS–Tween solution. 5. To detect actin, incubate the membrane in rabbit anti-actin antibody (Sigma) diluted 1/500 in TBS–Tween for 1 h at room temperature (see Note 8). 6. Wash the membrane five times with TBS–Tween and incubate with peroxidase-conjugated anti-rabbit secondary antibodies (BioRad) diluted 1/4000 in TBS–Tween for 30 min. 7. Wash the membrane five times with TBS–Tween and incubate with a chemiluminescence reagent (Pierce). 8. Expose blot to film as shown in Fig. 2.

4. Notes 1. Chill approximately 100 L of water ahead of time. 50 L cylindrical tanks with lids (Nalgene) are convenient for chilling water and carrying out the dialyses. 2. Alternatively, synthetic phospholipids may be mixed to obtain a bilayer with a defined phospholipid composition. 3. The second dialysis tank can be reused for the first dialysis in Subheading 3.1, step 11. 4. The ammonium sulfate pellets may be dislodged from the centrifuge bottles with a pipette and easily resuspended using a 100-ml dounce homogenizer.

102

Stamnes and Xu

5. A trace amount of fluorescent phospholipid may be added to the preparation so that the recovery of liposomes may be quantified using a fluorescence spectrometer. For this 10 mg/ml NBD-phosphatidylcholine (Avanti Polar Lipids) works well. 6. A bath sonicator may also be used to resuspend the phospholipids. 7. The reaction can easily be scaled up to characterize the properties of less abundant cytokeletal proteins. In this case, controls should be carried out to confirm that liposomes are successfully recovered from larger volume sucrose gradients. 8. Duplicate blots may be probed with other antibodies to characterize the properties of multiple cytoskeleton-related proteins.

Acknowledgments This work was supported by NIH grant RO1 GM068674 (M.S.). The methods for preparing cytosol and liposomes have been adapted from (13, 14). We thank Heidi Hehnly for reading the manuscript.

References 1. Cooke R (2004) The sliding filament model: 1972–2004. J Gen Physiol 123:643–656 2. Wu X, Jung G, Hammer JA III (2000) Functions of unconventional myosins. Curr Opin Cell Biol 12:42–51 3. Theriot JA (2000) The polymerization motor. Traffic 1:19–28 4. Ridley AJ (2006) Rho GTPases and actin dynamics in membrane protrusions and vesicle trafficking. Trends Cell Biol 16:522–529 5. Hehnly H, Stamnes M (2007) Regulating cytoskeleton-based vesicle motility. FEBS Lett 581:2112–2118 6. Fucini RV, Chen JL, Sharma C, Kessels MM, Stamnes M (2002) Golgi vesicle proteins are linked to the assembly of an actin complex defined by mAbp1. Mol Biol Cell 13:621–631 7. Chen JL, Fucini RV, Lacomis L, ErdjumentBromage H, Tempst P, Stamnes M (2005) Coatomer-bound Cdc42 regulates dynein recruitment to COPI vesicles. J Cell Biol 169:383–389

8. Chen JL, Lacomis L, Erdjument-Bromage H, Tempst P, Stamnes M (2004) Cytosol-derived proteins are sufficient for Arp2/3 recruitment and ARF/coatomer-dependent actin polymerization on golgi membranes. FEBS Lett 566:281–286 9. Fucini RV, Navarrete A, Vadakkan C, Lacomis L, Erdjument-Bromage H, Tempst P, Stamnes M (2000) Activated ADP-ribosylation factor assembles distinct pools of actin on golgi membranes. J Biol Chem 275:18824–18829 10. Xu W, Stamnes M (2006) The actin-depolymerizing factor homology and charged/ helical domains of drebrin and mAbp1 direct membrane binding and localization via distinct interactions with actin. J Biol Chem 281: 11826–11833 11. Welch MD, Iwamatsu A, Mitchison TJ (1997) Actin polymerization is induced by Arp2/3 protein complex at the surface of Listeria monocytogenes. Nature 385: 265–269

Actin Polymerization on Liposomes 12. Heuvingh J, Franco M, Chavrier P, Sykes C (2007) ARF1-mediated actin polymerization produces movement of artificial vesicles. Proc Natl Acad Sci U S A 104:16928–16933 13. MacDonald RC, MacDonald RI, Menco BP, Takeshita K, Subbarao NK, Hu LR (1991) Smallvolume extrusion apparatus for preparation of

103

large, unilamellar vesicles. Biochim Biophys Acta 1061:297–303 14. Wattenberg BW, Rothman JE (1986) Multiple cytosolic components promote intra-Golgi protein transport. Resolution of a protein acting at a late stage, prior to membrane fusion. J Biol Chem 261:2208–2213

Chapter 9 Electroformation of Giant Unilamellar Vesicles from Native Membranes and Organic Lipid Mixtures for the Study of Lipid Domains under Physiological Ionic-Strength Conditions L.-Ruth Montes, Hasna Ahyayauch, Maitane Ibarguren, Jesus Sot, Alicia Alonso, Luis A. Bagatolli, and Felix M. Goñi Abstract Giant unilamellar vesicles (GUVs) constitute a cell-sized model membrane system that allows direct visualization of particular membrane-related phenomena, such as domain formation, at the level of single vesicles using fluorescence microscopy-related techniques. Currently available protocols for the preparation of GUVs work only at very low salt concentrations, thus precluding experimentation under physiological conditions. In addition, the GUVs thus obtained lack membrane compositional asymmetry. Here we show how to prepare GUVs using a new protocol based on the electroformation method either from native membranes or organic lipid mixtures at physiological ionic strength. Additionally, we describe methods to test whether membrane proteins and glycosphingolipids preserve their natural orientation after electroformation of GUVs composed of native membranes. Key words: Giant unilamellar vesicles (GUVs), Electroformation, Physiological conditions, Biological membranes, Lipid domains

1. Introduction Extensive documentation has appeared in recent years describing the use of giant unilamellar vesicles (GUVs, 20 µm mean diameter) as model systems to study different physical aspects of membranes (lateral structure, mechanical properties), particularly considering the effect of not only lipid–lipid but also lipid–DNA, lipid–peptide, and lipid–protein interactions (1–6). Owing to their size, single vesicles can be directly observed using light microscopy techniques (1, 2). V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_9, © Humana Press, a part of Springer Science+Business Media, LLC 2010

105

106

Montes et al.

However, two main drawbacks regarding GUV preparation are still unsolved: (1) low salt concentration (530 nm) cut-off filter (e.g. Corning 3–68).

2.6.2. Release of the Tb/ DPA Complex

1. Solution to be encapsulated in large unilamellar “Tb/DPA liposomes”: 1.25 mM TbCl3, 25 mM Na citrate, 25 mM Na dipicolinate, 10 mM NaCl, 10 mM TES, pH 7.4. 2. Solution to be encapsulated in small unilamellar “Tb/DPA liposomes”: 7.5 mM TbCl3, 75 mM Na citrate, 75 mM Na dipicolinate, 10 mM TES, pH 7.4. 3. Buffer A: 100 mM NaCl, 1 mM EDTA, 10 mM TES, pH 7.4. Solutions should be stored at 4˚C. 4. Buffer B: 100 mM NaCl, 10 mM TES, pH 7.4.

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

217

5. Sephadex G-75 or G-50 column equilibrated with Buffer A. 6. Triton X-100 stock solution (100×) at 10% (w/v) or C12E8 (20×) at 16 mM. Detergent solutions are kept at room temperature. 7. Fluorometer, as above. 2.6.3. The Release of ANTS/DPX

1. Solution to be encapsulated in large unilamellar “ANTS/ DPX-liposomes”: 12.5 mM ANTS, 45 mM DPX, 20 mM NaCl, 10 mM TES, pH 7.4. Solutions of fluorescent probes should be kept in the dark by wrapping aluminum foil over the tube, and stored at 4˚C. 2. Buffer C: 100 mM NaCl, 0.1 mM EDTA, 10 mM TES, pH 7.4. Solutions should be stored at 4˚C. 3. Sephadex G-75 or G-50 column equilibrated with Buffer C. 4. Triton X-100 stock solution (100×) at 10% (w/v) or C12E8 (20×) at 16 mM. Detergent solutions are kept at room temperature. 5. Fluorometer, as above.

3. Methods

3.1. Liposome Preparation 3.1.1 Large Unilamellar Liposomes Prepared by Reverse Phase Evaporation

1. Pure or mixed phospholipids in chloroform are measured using gastight Hamilton syringes under a chemical hood, and dispensed into a glass tube. The tube is covered with teflon tape, and then placed inside a larger tube that fits on a rotary evaporator until the lipid dries into a thin film in vacuum. The total lipid is usually 10–20 µmol. 2. A few milliliters of diethyl ether is washed with a similar volume of distilled or purified water in a tightly capped glass tube by gentle shaking, and the mixture is allowed to separate. One milliliter of the ether (the top layer) is removed by means of a glass pipette or syringe and added to the dried phospholipid film, ensuring that the lipid dissolves completely. 3. The buffer to be encapsulated (0.34 ml) is added to the phospholipid solution in ether. A gentle stream of argon gas is flushed over the mixture using a Pasteur pipette, and the tube is sealed with teflon tape and a screw-cap. The mixture is sonicated for 2–5 min, resulting in a stable emulsion. If an emulsion is not formed it is likely that the sonicator is not at an optimal setting. In this case, either the level of the water in the bath or the power supply needs to be adjusted.

218

Düzgünes¸, Faneca and Lima

4. The cap is opened in the chemical hood, the tube is sealed again with teflon tape, and placed inside the larger glass tube that fits onto the rotary evaporator. About 1 ml of water is added to the outer tube, both to facilitate thermal contact and to minimize the evaporation of the aqueous solution in the inner tube. The outer tube is immersed in the water bath of the rotary evaporator, kept at 30°C. 5. The ether is evaporated gently in controlled vacuum, at about 350 mm Hg, under constant supervision. The tube is purged occasionally with argon gas attached to the top of the evaporator to maintain the vacuum level and to prevent excessive bubbling. The vacuum is allowed to build up when a stable gel is formed. 6. To break up the gel, the inner glass tube is removed and vortexed vigorously for 5–10 s. The tube is placed again in the outer tube and controlled rotary evaporation is resumed. This step is repeated once or twice, until an aqueous opalescent suspension is formed. An additional 0.66 ml of the encapsulation buffer is added to the suspension and rotary evaporation is continued for an additional 20 min to remove any residual ether (see Note 1). 7. To achieve a uniform size distribution, the liposome suspension is passed several times through polycarbonate membranes of 100 nm pore diameter (or other desired diameter), using a high-pressure or syringe extruder. 8. The average size and the size distribution of the liposomes are assessed by dynamic light scattering in a Beckman Coulter N4 Plus Submicron Particle Sizer (or equivalent instrument). 9. Liposomes are stored at 4˚C after gently flushing the top part of the tube with a stream of argon, and sealing the cap with parafilm. 3.1.2. Large Unilamellar Liposomes Prepared by Extrusion of Multilamellar Liposomes

1. Pure or mixed phospholipids (usually 10–20 µmol) in chloroform are measured using gastight Hamilton syringes under a chemical hood, and dispensed into a glass tube. The tube is covered with teflon tape, and placed inside a larger tube that fits on a rotary evaporator until the lipid dries into a thin film in vacuum. The evaporator is purged with argon to reduce the vacuum, and the inner tube is placed in a vacuum jar or vacuum oven, and kept under high vacuum for at least 2 h to eliminate any residual chloroform. 2. The phospholipid film is hydrated with the buffer to be encapsulated (e.g., the Tb citrate solution), the tube is purged with argon gas and covered with teflon tape, and the cap is closed tightly. The mixture is vortexed for 10 min at room temperature to form multilamellar liposomes. If the lipid has a

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

219

high gel to liquid-crystalline phase transition temperature, it is hydrated with buffer at a higher temperature. The vortex mixing is interrupted every 30 s to place the tube in a water bath at this high temperature for about 15 s. 3. To achieve a uniform size distribution, the liposome suspension is passed 21 times through a polycarbonate membrane of 100 nm pore diameter (or other desired diameter), using a syringe extruder. If the extrusion requires too much force, the multilamellar liposomes should first be passed through a membrane of larger pore diameter (e.g. 800 or 400 nm) (see Note 2). 4. The average size and the size distribution of the liposomes are assessed by dynamic light scattering in a Beckman Coulter N4 Plus Submicron Particle Sizer (or equivalent instrument). 5. Liposomes are stored at 4˚C after gently flushing the top part of the tube with a stream of argon, and sealing the cap with parafilm. 3.1.3. Small Unilamellar Liposomes Prepared by Sonication

1. Ten micromoles of a phospholipid mixture in chloroform are dispensed into the glass tube that is then covered with teflon tape over the top of the tube. The tube is placed in the larger glass tube that fits on the rotary evaporator, and the chloroform is allowed to evaporate, using condensation precautions to trap the solvent. When the lipid forms a dry film, the evaporator is purged with argon to reduce the vacuum. The inner tube is then placed in a vacuum jar or vacuum oven, and is kept under high vacuum for at least 2 h to eliminate any residual chloroform. 2. The dried phospholipid film is hydrated with the buffer to be encapsulated, the tube is purged with argon gas and covered with Teflon tape, and the screw-cap is closed tightly. The mixture is vortexed for 10 min at room temperature. If a lipid with a high gel to liquid-crystalline phase transition temperature is used, the lipid is hydrated with buffer at a temperature above the transition temperature, and the tube is placed in a water bath at this high temperature in between short periods of vortexing. 3. The multilamellar suspension is sonicated in a bath-type sonicator for 0.5–1 h. The level of water in the bath is adjusted such that the water surface breaks up into small droplets under sonication. The top of the liposome suspension is aligned with the level of the water such that an aerosol forms occasionally in the tube. Overheating of the bath should be avoided, by circulating water through the bath or by adding some ice and re-adjusting the level of the water. For lipids with high transition temperatures, the bath should be maintained at a few degrees above this temperature.

220

Düzgünes¸, Faneca and Lima

4. To remove any remaining large liposomes, the resulting opalescent suspension is centrifuged at 100,000g for 1 h at 4˚C, preferably in a swinging bucket rotor. The supernatant is removed without disturbing the pellet, and used as the small unilamellar liposome preparation (see Note 3). 5. The average size and the size distribution of the liposomes are assessed by dynamic light scattering in a Beckman Coulter N4 Plus Submicron Particle Sizer (or equivalent instrument). 6. Liposomes are stored at 4˚C after gently flushing the top part of the tube with a stream of argon, and sealing the cap with parafilm. 3.1.4. Measurement of the Phospholipid Concentration of Liposomes

1. Liposome samples are placed in triplicate tubes at an estimated amount of less than 0.1 µmol inorganic phosphate. Phosphate standards are pipetted in triplicate (for example 0.01, 0.05, 0.075, 0.1 µmol inorganic phosphate). Four hundred microliters of 10 N H2SO4 are added to each tube that are then heated for 30 min on the heating block. 2. The tubes are cooled at room temperature and 100 µl H2O2 are added using a pipettor or repeater pipette (e.g. Pipetman or Eppendorf). The tubes are returned to the heating block for 30 min on the heating block. The fumes are tested for the absence of H2O2, using the indicator strips. 3. The tubes are placed in a round metal rack (that will fit eventually into an electric water boiler), 4.6 ml of 0.22% ammonium molybdate reagent are added to the tubes and mixed on a vortex mixer. Two hundred microliters of ANSA (or Fiske) reagent are added to each tube and then mixed on a vortex mixer. Alternatively, 100 µl ascorbate can also be used for this step. 4. The metal rack is lowered into a boiling water bath for 7–10 min, and then cooled. The contents of the tubes are transferred to spectrophotometer cuvettes. Protective latex or vinyl gloves are recommended for this procedure. A spectrophotometer with a sipper accessory is preferable to avoid handling of acid-containing tubes. 5. The absorbance of the solutions in each cuvette is measured at 812 nm or at 660 nm (if the solution is too concentrated). The phosphate content of the sample is assessed from the standard curve (see Note 4).

3.2. Liposome Fusion: Intermixing of Aqueous Contents 3.2.1. The Tb/DPA Acid Assay

1. For large unilamellar liposomes, the osmolality of the solutions inside the liposomes should match that of the medium, and if lower, should be adjusted with the addition of small amounts of NaCl. The Tb-, DPA- and Tb/DPA-liposomes are prepared as described in Subheadings 3.1.1 or 3.1.2. (see Note 5).

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

221

2. The 3 Sephadex G-75 or G-50 columns are equilibrated with Buffer A. The liposomes are passed through separate columns to eliminate the solution outside the liposomes, using Buffer A for elution. It is necessary to include the 1 mM EDTA in Buffer A to prevent the binding of Tb ions to the phospholipids. The first 4 ml after the liposomes are loaded on the gel are discarded, and the next 3 ml are collected in the 15-ml culture tube. Argon gas is flushed over the liposome suspensions for 10 s without causing any significant disruption of the surface of the liquid; then the tubes are capped tightly, sealed with parafilm, and stored in ice. 3. One milliliter of the collected Tb-liposomes is placed on the third Sephadex column equilibrated with Buffer B, and collected as in step 2. Since EDTA would interfere with the formation of the Tb/DPA complex when the vesicles are lysed to obtain 100% fluorescence, this chromatography step replaces the EDTA in the medium, producing the “Tb minus EDTA liposomes.” 4. The lipid concentration of the liposomes is measured in each of the culture tubes by inorganic phosphate analysis (see Subheading 3.1.3). 5. For a fusion assay utilizing 50 µM total lipid concentration, 25 µM Tb-liposomes are mixed with 25 µM DPA-liposomes. To calibrate the fluorescence to “100%” fluorescence; i.e. the maximal fluorescene that can be obtained if all the liposomes formed a mega-liposome, 25 µM of the “Tb minus EDTA liposomes” are placed in a fluorometer cuvette containing the appropriate volume of Buffer B. The amount of buffer should be adjusted to allow for the volumes of the subsequent ingredients. Ten microliters of the 2 mM DPA stock are added to the cuvette (final DPA concentration 20 µM). To lyse the liposomes, 50 µl of either of the detergent stocks are added and the fluorescence is allowed to equilibrate. The final concentrations of the detergents should be 0.5% (w/v) for Na cholate, or 0.8 mM for C12E8. 6. The excitation wavelength of the fluorometer is set at 276– 278 nm. The emission wavelength is set to 545 nm. Intermediate slit widths are used to optimize the intensity, but to minimize scattering artifacts. The high-pass cut-off filter is placed before the emission monochromator, to minimize any contributions from light scattering. Crossed polarizers can also be used to minimize lightscattering contributions. For our experiments we have used SLM 4000, SLM 8000, and Spex Fluorolog fluorometers. Other instruments with similar light intensity and sensitivity can be used for this assay.

222

Düzgünes¸, Faneca and Lima

7. A 1:1 mixture of the Tb-liposomes and the DPA-liposomes is prepared as a stock solution in a 15-ml polystyrene culture tube, at a final lipid concentration that is tenfold higher than that used for the actual assay. For example, a suspension of 500 µM lipid is prepared to achieve a final concentration of 50 µM in the assay. The stock solution (0.1 ml) is diluted into 0.9 ml of Buffer B; this process also dilutes the final EDTA concentration to 0.1 mM. 8. For a fluorometer with a strip-chart recorder, the low fluorescence of this suspension is set to 0% Fmax, using the offset function of the fluorometer. Using the gain function, necessary adjustments are made to the 100% level with the cuvette containing the liposomes lysed in the presence of the DPA. The calibration procedure is repeated after a set of measurements to ensure that the lamp intensity has not changed in the course of the experiment. 9. For a fluorometer with computerized data acquisition, the assay can be calibrated by subtracting the initial level of fluorescence [I(0)] from the data set, and dividing the resulting data set by the numerical difference between the fluorescence intensity of the calibration vesicles [I (•)] and the new 0% level. If the resulting value is multiplied by 100, the percentage value is obtained. Thus, the extent of fusion, F(t), as a percentage of maximal fluorescence, is given by F (t ) = 100 x [I (t ) - I(0)]/[I (∞) - I(0)] where I(t) is the fluorescence intensity at time t. 10. The assay may be calibrated in an alternative manner also. Here, 50 µM of the Tb/DPA liposomes are considered to have the maximal, 100%, fluorescence. These liposomes represent the fusion product of all the Tb-liposomes and DPA-liposomes in the fusion assay. Of course, these calibration liposomes should have the same size distribution and should be used at the same lipid concentration as the total of the Tb- and DPAliposome populations (25 µM of each). One slight complication of the use of Tb/DPA liposomes is that after they are transferred to room temperature from the storage temperature (0˚C), it takes about 0.5 h for the fluorescence to reach a steady state. 3.2.2. The ANTS/DPX Assay

1. Three batches of large unilamellar liposomes are prepared with the ANTS, DPX, and ANTS/DPX solutions. Small unilamellar liposomes have not produced reliable results with this assay (see Note 6). 2. The liposomes are chromatographed on separate Sephadex G-75 or G-50 columns equilibrated with Buffer C. The first 4 ml, measured from the point where the liposomes are

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

223

allowed to partition into the gel, are discarded as the void volume. The next 3 ml are collected in the polystyrene culture tube. The tube is flushed gently with argon, capped tightly, sealed with parafilm, and stored in ice until use. 3. The phosphate concentration of each of the liposome preparations is measured using the assay described in Subheading 3.1.4. 4. The excitation monochromator of the fluorometer is set to 360 nm, and the emission monochromator to 530 nm, with relatively wide slit widths (10–20 nm). A high-pass filter (e.g. Corning 3–68) is placed in the emission channel to eliminate light scattering artifacts. 5. ANTS-liposomes (25 µM lipid) are placed in the appropriate amount of buffer (1–2 ml, depending on the fluorometer, with “flea” or “castle” stir-bars), and the gain of the fluorometer is adjusted to an arbitrary unit of 100%. The fluorescence of 50 µM ANTS/DPX-liposomes is taken as 0%. The latter liposomes represent the theoretical fusion product of all the ANTS vesicles and all the DPX vesicles (also 25 µM) used in the assay. 6. Putative fusogens, such as divalent cations, protons, or fusogenic peptides are added to the cuvette from a concentrated stock solution under constant stirring. Liposome fusion results in the decrease of fluorescence due to quenching of ANTS by the DPX. 3.3. Liposome Fusion: Intermixing of Lipids 3.3.1. The NBD/Rhodamine Assay

1. “Labeled liposomes” are prepared as described in Subheadings 3.1.1, 3.1.2, or 3.1.3, with 0.8 mole% each of NBD-PE and Rh-PE in the initial chloroform mixture in addition to other phospholipids (see Note 7). 2. “Unlabeled liposomes” of the same lipid composition are prepared without the fluorophores present. 3. “Calibration liposomes”, or “mock fused liposomes,” are made as in 1, but with 0.08 mole% of each fluorophore (see Note 8). 4. The phosphate concentration of each of the liposome preparations is measured (Subheading 3.1.4), and aliquots from the labeled and unlabeled liposomes are mixed at a ratio of 1:9 in 2 mL of Buffer C. For a total lipid concentration of 50 µM, this would be 5 µM-labeled liposomes and 45 µM-unlabeled liposomes. 5. The excitation and emission monochromators are set to 460 nm and 530 nm, respectively. The maximal (100%) fluorescence is set with 50 µM of the calibration liposomes. The residual fluorescence of the mixture of labeled and unlabeled liposomes is set to 0% fluorescence. The percentage of lipid mixing as a function of time is given by M(t) = 100 × [I(t) − I(0)]/ [I(•) − I(0)], where I(t) is the fluorescence intensity at time t,

224

Düzgünes¸, Faneca and Lima

I(0) is the residual fluorescence, and I(•) is the maximal fluorescence. 6. Fusion of the liposomes results in an increase in NBD fluorescence as the probes are diluted into the unlabeled liposomes. 7. Maximal fluorescence can also be set by lysing the labeled liposomes at the same concentration to be used in the assay, with a detergent that does not affect the quantum yield of NBD. Commercial preparations of Triton X-100 require a correction factor of 1.4–1.5, while some purified preparations of this detergent may have no inhibitory effect on the fluorescence. Alternative detergents are C12E8 or C12E9 (Calbiochem) 3.3.2. Inner Monolayer Mixing Assay

1. Liposomes containing 0.75 mole% each of N-NBD-PS and either N-Rh-PE or diI(5)C18 (20 mM total phospholipid) are prepared in Buffer C, as described in Subheadings 3.1.1 or 3.1.2 (see Note 9). 2. To a 10 mM liposome suspension on ice, dithionate is added to a final concentration of 80–100 mM, and the mixture incubated for 30–45 min. 3. The dithionate is removed by centrifuging100 µl aliquots through the spin columns. 4. The remainder of the assay is as described in Subheading 3.3.1 to measure the intermixing of the inner monolayers.

3.4. Fusion of Liposomes and Lipoplexes with Cultured Cells

1. THP-1 human monocytic leukemia cells are grown in the growth medium described in Subheading 2.4, at 37˚C and 5% CO2. 2. The cells are harvested by centrifugation at 180 g for 7 min at room temperature, and the pellet is resuspended in phenol red-free RPMI-1640 with HEPES buffer. The cells are centrifuged again. This procedure is then repeated twice. The final pellet of cells is suspended at a concentration of 20 × 106 cells/ml in a polypropylene culture tube and kept on ice until use the same day. Cell viability is determined by mixing a small sample of cells 1:1 with the Trypan blue solution, while counting the cells in a hemacytometer. 3. Large unilamellar liposomes containing the cationic lipid DOTAP alone, or mixed with PE or PC, are prepared first by hydration of a dried lipid film in 150 mM NaCl, 10 mM HEPES, pH 7.4, at a final concentration of 5 mM lipid (see Subheading 3.1.2). This is followed by vortexing under argon to form multilamellar liposomes, and by extrusion through polycarbonate membranes of 100 nm pore diameter. Rh-PE is included in the initial chloroform mixture of lipids at 5 mole%, at which concentration it is self-quenched.

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

225

4. Lipid mixing between the liposomes and cell membrane is monitored continuously in a magnetically stirred fluorometer cuvette maintained at 37˚C or other desired temperatures. Cells (100 µl) are added to the cuvette containing 1.9 ml of 150 mM NaCl, 10 mM HEPES, pH 7.4, and allowed to equilibrate to the set temperature. Liposomes are added to a final concentration of 20 µM. 5. The fluorescence is measured with the excitation wavelength at 568 nm and the emission wavelength at 586 nm. In a Spex fluorometer, the excitation and emission slits are set to 0.5 and 1 mm, respectively. The fluorescence scale is set with the initial fluorescence of liposomes and cells being 0%, and the maximal fluorescence (100%) is obtained by dissolving the liposomes with 0.5% (v/v) Triton X-100. 6. To prevent endocytosis, and hence to measure fusion at the plasma membrane only, the cells are pre-incubated with 1 µg/ml antimycin A, 10 mM NaF and 0.1% (w/v) NaN3 for 30 min at 37˚C. Fusion is monitored in the presence of these inhibitors. 7. To examine the fusion of cationic liposome-DNA complexes with cells, the complexes are prepared at varying lipid nitrogen/DNA phosphate (+/−) ratios right before they are added to the cells in the cuvette. The effect of the maturation of the complexes on fusion can also be investigated by preincubating the complexes for 1–2 h before addition to the cells (see Note 10). 3.5. Intracellular Delivery of Liposome Contents 3.5.1. Flow Cytometric Analysis

1. Large unilamellar liposomes encapsulating 80 mM calcein and containing 1 mole% Rh-PE in their membrane are prepared by extrusion of multilamellar liposomes (Subheading 3.1.2). 2. Monocytic human THP-1 cells are cultured in RPMI 1640 medium supplemented with 10% FBS. The cells are differentiated to macrophage-like cells by adding 160 nM phorbol 12-myristate 13-acetate in 24-well culture plates (106 cells per well). The cells are incubated for 5–6 days, and the culture medium is replaced with fresh medium. 3. The liposomes are added to the cells at a final phospholipid concentration of 100 µM, and incubated for various times at 37°C. The cells are then washed twice with phosphate-buffered saline (PBS) without calcium or magnesium ions. 4. The cells are detached from the plastic by adding 0.5 ml of dissociation buffer and mixed with 0.5 ml of PBS containing divalent cations, 2% FBS and 1 mg/ml propidium iodide. Rhodamine and calcein fluorescence is detected with a Becton Dickinson FACStar Plus flow cytometer (or equivalent),

226

Düzgünes¸, Faneca and Lima

controlled by a Hewlett-Packard computer with Lysis II software (Becton Dickinson, San Jose, CA). The samples are analyzed for rhodamine using excitation at 528 nm and emission at 575 nm. Calcein is excited at 488 nm and the fluorescence emission is detected at 520 nm with a 0.1 neutral density filter. For each sample, 10,000 events are recorded. Forward scatter and propidium iodide fluorescence signals are used to gate the cell subset of interest and eliminate debris, dead cells, and cell aggregates. 5. Mean rhodamine fluorescence values reflect the binding and uptake of liposomes by the cells. The mean calcein fluorescence indicates the intracellular dequenching of the dye. The calculated ratio of calcein to rhodamine fluorescence is taken to measure the amount of aqueous marker released intracellularly per cell-associated liposome. The initial calcein to rhodamine fluorescence ratio of liposomes bound to the cells, in the absence of endocytosis, is obtained by incubating the liposomes with the cells at 4°C (see Note 11). 3.5.2. Image Analysis

1. THP-1 cells (105/well) are differentiated for 5–6 days in Lab-Tek chambered coverglasses for tissue culture, obtained from Nunc (Naperville, IL), by adding 160 nM phorbol 12-myristate 13-acetate to the culture medium. Cells are washed with cold RPMI media without phenol red, containing 20 mM HEPES buffer, pH 7.4. 2. Liposomes are added to cells at a final phospholipid concentration of 200 µM and incubated for 1 h at 4°C in phenol red-free RPMI medium with 20 mM HEPES buffer. After this pre-binding step, the cells are washed with cold media and the initial calcein and rhodamine fluorescence images are recorded using the Photon Technology International (PTI) ratio imaging system. To evaluate the kinetics of calcein dequenching, the cold medium is removed, medium at 37°C is added, and the cells are incubated for various times. 3. The cells are washed with cold medium, and the calcein and rhodamine fluorescence images are recorded. Cells are observed in a Nikon Diaphot epi-fluorescence microscope (Melville, NY) using a 100× objective and filters for FITC/ TXRD obtained from Chroma (Rockingham, VT). Averages of 16 snapshots are taken to reduce the background. Ratio images are produced using PTI (or equivalent) software. 4. Histograms of calcein to rhodamine ratio for each cell are determined using a square of 100 × 100 pixels, representing the approximate area of one cell at the magnification used. This provides a measure of calcein dequenching per cell. Average histograms for each liposome composition and time points are then calculated in Excel.

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

227

5. Medians for each average histogram are calculated from cumulative curves. To calculate the growth of the area occupied by the calcein that is dequenched, the highest ratio in the control experiment (incubation at 4°C) is taken as the cut-off. The sum of the pixels with ratios higher than the cutoff ratio, are taken to estimate the dequenched area at various time points (see Note 12). 3.5.3. Fluorometric Measurements

1. THP-1 cells (5 × 106) are differentiated by incubation for 5–6 days in 5–ml of medium containing 160 nM phorbol 12-myristate 13-acetate in 6-well culture plates. The medium is removed and the cells are washed with phenol red-free RPMI medium with 20 mM HEPES buffer. 2. Cells are incubated with liposomes at a final phospholipid concentration of 150–200 µM for different times at 37°C in phenol red-free RPMI medium/20 mM HEPES buffer, and then washed. 3. Two milliliters of dissociation buffer are added, and the cells are incubated for 10 min at 37°C. The cells are detached from plastic with disposable scrapers (Costar Corporation, Cambridge, MA) and transferred into disposable fluorometer cuvettes (Hughes & Hughes Limited, Tonedale, Wellington). 4. Fluorescence measurements are performed in a SPEX Fluorolog 2 fluorometer (SPEX Industries, Inc. Edison, NJ), Perkin Elmer LS50 fluorometer or equivalent instrument. Calcein fluorescence is read at excitation and emission wavelengths of 490 and 520 nm, respectively, using 0.5 mm excitation and 1.0 mm emission slits. Rhodamine fluorescence is measured at excitation and emission wavelengths of 568 and 600 nm, respectively, using 0.5 mm excitation and 1.5 mm emission slits. 5. The sample chamber is adjusted to front-face configuration to prevent cell-scattering effects and is equipped with a magnetic stirrer. The temperature is maintained at 20°C with a thermostatic water circulator (see Note 12).

3.6. Liposome Permeability 3.6.1. Release of Carboxyfluorescein or Calcein

1. Large unilamellar liposomes are prepared using 50 mM carboxyfluorescein or calcein (Na salt) with 10 mM TES, pH 7.4. This concentration is iso-osmotic with Buffer C. 2. Small unilamellar liposomes are prepared using 100 mM carboxyfluorescein or calcein, 10 mM TES, pH 7.4. Larger concentrations of encapsulated material can be used in this case, since these liposomes are not active osmotically. 3. The unencapsulated material is separated by gel filtration on Sephadex G-75, using Buffer C as the elution buffer. 4. The phosphate concentration of the liposomes is measured as in Subheading 3.1.4.

228

Düzgünes¸, Faneca and Lima

5. The excitation wavelength is set to 490 nm and the emission wavelength to 530 nm, using 0.5 mm excitation and 1.0 mm emission slits in the Spex fluorometer, and 5 nm and 10 nm slits in the Perkin-Elmer fluorometer. 6. The maximal fluorescence (100%) of a certain concentration of liposome is set by lysing the vesicles with 0.1% (w/v) Triton X-100 or 0.8 mM C12E8. 7. The extent of release, R(t) may also be calculated from R(t) = 100 × [I(t) − I(0)]/[I(•) − I(0)], where I(t) is the fluorescence intensity at time t, I(0) is the initial residual fluorescence of the liposomes before any membrane-active agents are added, and I(•) is the fluorescence obtained when the liposomes are lysed with detergent (see Note 13). 3.6.2. Release of the Tb/DPA Complex

1. Liposomes are prepared using the Tb/DPA solution appropriate for large or small unilamellar liposomes (Subheading 3.1). 2. The lipsomes are passed through the column to eliminate unencapsulated material, using Buffer A for elution. 3. The phosphate concentration of the liposomes is measured as in Subheading 3.1.4. 4. A stock solution is prepared in Buffer B at a lipid concentration tenfold higher than that to be used in the final assay. The stock solution (0.1 ml) is diluted into 0.9 ml of Buffer B, diluting the final EDTA concentration to 0.1 mM. 5. If the liposomes are not affected by the presence of Ca2+ (i.e. they do not contain a high-mole fraction of anionic phospholipids), this cation should be included in the external buffer to facilitate the dissociation of the Tb/DPA complex. 6. The initial fluorescence (excitation at 278 nm, emission at 545 nm, with the 3–68 cutoff filter) of the liposomes is set to 100%. After placing the liposomes into buffer, it is necessary to allow some time for the fluorescence to stabilize. The minimum fluorescence, corresponding to maximal leakage, can be set by lysing the liposomes with detergent (0.1% (w/v) Triton X-100 or 0.8 mM C12E8 (final concentrations)). 7. The extent of release, R(t), may be calculated from R(t) = 100 × [I(0) − I(t)]/[I(0) − I(•)], where I (t) is the fluorescence intensity at time t, I(0) is the initial fluorescence of the Tb/DPA liposomes before any membrane-active agents are added, and I(•) is the fluorescence obtained when all the liposome contents leak. The liposomes are lysed with detergent (see Note 14).

3.6.3. The Release of ANTS/DPX

1. Large unilamellar liposomes are prepared to encapsulate the ANTS/DPX solution. This assay is not suitable for use with small unilamellar liposomes.

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

229

2. The liposomes are passed through a Sephadex column to eliminate unencapsulated material, using Buffer C for elution. 3. The phosphate concentration of the liposomes is measured as in Subheading 2.1.4. 4. The fluorometer is set as follows: Excitation wavelength at 360 nm, the emission wavelength at 530 nm, using relatively wide slit widths (10–20 nm), and a high-pass filter (Corning 3–68) in the emission channel to eliminate light-scattering artifacts. 5. The assay is calibrated to maximal fluorescence by lysing the vesicles with 0.1% (w/v) Triton X-100 or 0.8 mM C12E8 (final concentrations). The extent of release as a function of time can be calculated according to the formula in Subheading 3.6.1. (see Note 15).

4. Notes 1. The reverse-phase evaporation method was first reported by Szoka and Papahadjopoulos [1] and further developed by Düzgünes¸ et al. [2]. 2. Extrusion of multilamellar liposomes to obtain large unilamellar liposomes was developed by Olson et al. [3] and later by Mayer et al. [4]. 3. Small unilamellar liposomes prepared by sonication were first reported by Papahadjopoulos and Miller [5]. The method described here is adapted from Düzgünes¸ et al. [2]. 4. The method for determination of inorganic phosphate described here was first described by Bartlett [6]. The details described here were developed in the Papahadjopoulos laboratory at UCSF by T. Heath and D. Alford. 5. The interaction of Tb and DPA results in the formation of the fluorescent [Tb(DPA)3]3- chelation complex. Internal energy transfer from DPA to Tb results in an increase in the fluorescence intensity of the latter by four orders of magnitude. The reactants are encapsulated in different populations of liposomes. Membrane fusion results in an increase in fluorescence [7, 8]. Contents that are released into the external medium are prevented from interacting by the presence of EDTA in the medium. Divalent cations used to induce fusion in same cases further inhibit the interaction. The presence of citrate in the Tb-liposomes is essential to prevent the interaction of Tb3+ with anionic phospholipids in the liposome

230

Düzgünes¸, Faneca and Lima

membrane. If phosphatidic acid is used as an anionic lipid, it is necessary to use a stronger chelator of Tb3+, such as nitrilotriacetic acid. 6. This assay is based on the collisional quenching of ANTS fluorescence by DPX. These reagents are encapsulated in separate populations of liposomes [9]. Membrane fusion results in the intermixing of ANTS and DPX and the quenching of ANTS fluorescence. If the contents are released and diluted into the medium, fluorescence quenching does not occur, since high concentrations of DPX are required for quenching. 7. The lipid-mixing assay measures the dilution of a fluorescence energy donor/acceptor pair from “labeled” liposomes to “unlabeled” liposomes as a result of membrane fusion. The efficiency of resonance energy transfer is then decreased, since the rate and efficiency of the transfer depend on the inverse sixth power of the distance between the two fluorophores (as well as the overlap of the emission spectrum of the donor and the absorption spectrum of the acceptor). Low concentrations of probes, usually less than 1 mole% of total lipid, are used for the assay, minimizing the extent of membrane perturbation. We describe a resonance energy transfer pair that has been used extensively by a number of laboratories [10, 11]. The fluorophores are attached to the headgroups of phosphatidylethanolamine and thus do not cause any appreciable perturbation of bilayer packing. 8. To avoid the use of detergents that may affect NBD fluorescence, liposomes corresponding to the fusion product of the labeled and unlabeled liposomes can be used to calibrate the assay to 100% fusion [12]. 9. The interaction of ions, proteins and other fusogens with the outer monolayer of liposomes containing fluorescent probes may alter the fluorescence intensity or lateral distribution of the probes. If the fluorophores were located only in the inner monolayer of liposomes, the fluorescence would not be affected by ion or protein binding to the liposome surface. In this method, the fluorophores exposed on the outer monolayer of the liposome are reduced with dithionate, thereby eliminating the fluorescence of outer monolayer fluorophores. N-NBD-phosphatidylserine (PS) was found to be more suitable for these experiments than N-NBD-PE, since it is less prone to transbilayer movement following reduction [13]. 10. Results obtained by this method have been described by Pires et al. [14].

Methods to Monitor Liposome Fusion, Permeability, and Interaction with Cells

231

11. This method was first described by Slepushkin et al. [15] to evaluate the ability of sterically stabilized pH-sensitive liposomes to deliver their contents into cultured cells. 12. Results obtained with this method may be found in Simões et al. [16]. 13. The use of liposomal carboxyfluorescein and calcein was described by Weinstein et al. [17] and Allen et al. [18], respectively. The effect of bacterial toxins on liposome membrane permeability, using these fluorophores, was reported by Kayalar and Düzgüneş [19]. 14. The dissociation of the Tb/DPA complex was initially used to measure liposome permeability during membrane fusion [20]. 15. The ANTS/DPX release assay was first developed by Ellens et al. [21].

References 1. Szoka F Jr, 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 2. Düzgüneş N, Wilschut J, Hong K, Fraley R, Perry C, Friend DS, James TL, Papahadjopoulos D (1983) Physicochemical characterization of large unilamellar vesicles prepared by reverse-phase evaporation. Biochim Biophys Acta 732:289–299 3. Olson F, Hunt CA, Szoka FC, Vail WJ, Papahadjopoulos D (1979) Preparation of liposomes of defined size distribution by extrusion through polycarbonate membranes. Biochim Biophys Acta 557:9–23 4. Mayer LD, Hope MJ, Cullis PR (1985) Vesicles of variable sizes produced by a rapid extrusion procedure. Biochim Biophys Acta 858:161–168 5. Papahadjopoulos D, Miller N (1967) Phospholipid model membranes. I. Structural characteristics of hydrated liquid crystals. Biochim Biophys Acta 135:624–638 6. Bartlett GR (1959) Phosphorus assay in column chromatography. J Biol Chem 234:466–468 7. Wilschut J, Düzgüneş N, Fraley R, Papahadjopoulos D (1980) Studies on the mechanism of membrane fusion: Kinetics of Ca2+induced fusion of phosphatidylserine vesicles followed by a new assay for mixing of aqueous vesicle contents. Biochemistry 19:6011–6021

8. Düzgüneş N, Wilschut J (1993) Fusion assays monitoring intermixing of aqueous contents. Methods Enzymol 220:3–15 9. Ellens H, Bentz J, Szoka FC (1985) H+- and Ca2+-induced fusion and destabilization of liposomes. Biochemistry 24:3099–3106 10. Struck DK, Hoekstra D, Pagano RE (1981) Use of resonance energy transfer to monitor membrane fusion. Biochemistry 20:4093–4099 11. Hoekstra D, Düzgüneş N (1993) Lipid mixing assays to determine fusion in liposome systems. Methods Enzymol 220:15–32 12. Düzgüneş N, Allen TM, Fedor J, Papahadjopoulos D (1987) Lipid mixing during membrane aggregation and fusion. Why fusion assays disagree. Biochemistry 26:8435–8442 13. Meers P, Ali S, Erukulla R, Janoff AS (2000) Novel inner monolayer fusion assays reveal differential monolayer mixing associated with cation-dependent membrane fusion. Biochim Biophys Acta 1467:227–243 14. Pires P, Simões S, Nir S, Gaspar R, Düzgüneş N, Pedroso de Lima MC (1999) Interaction of cationic liposomes and their DNA complexes with monocytic leukemia cells. Biochim Biophys Acta 1418:71–84 15. Slepushkin VA, Simões S, Dazin P, Newman MS, Guo LS, Pedroso de Lima MC, Düzgüneş N (1997) Sterically stabilized pH-sensitive liposomes: intracellular delivery of aqueous contents and prolonged circulation in vivo. J Biol Chem 272:2382–2388

232

Düzgünes¸, Faneca and Lima

16. Simões S, Slepushkin V, Düzgüneş N, Pedroso de Lima MC (2001) On the mechanisms of internalization and intracellular delivery mediated by pH-sensitive liposomes. Biochim Biophys Acta 1515:23–37 17. Weinstein JN, Yoshikami S, Henkart P, Blumenthal R, Hagins WA (1977) Liposomecell interaction: transfer and intracellular release of a trapped fluorescent marker. Science 195:489–492 18. Allen TM, Cleland LG (1980) Serum-induced leakage of liposome contents. Biochim Biophys Acta 597:418–426

19. Kayalar C, Düzgüneş N (1986) Membrane action of colicin E1: detection by the release of carboxyfluorescein and calcein from liposomes. Biochim Biophys Acta 860:51–56 20. Bentz J, Düzgüneş N, Nir S (1983) Kinetics of divalent cation-induced fusion of phosphatidylserine vesicles: correlation between fusogenic capacities and binding affinities. Biochemistry 22:3320–3330 21. Ellens H, Bentz J, Szoka FC (1984) pHinduced destabilization of phosphatidylethanolamine-containing liposomes: role of bilayer contact. Biochemistry 23:1532–1538

Chapter 17 The Use of Isothermal Titration Calorimetry to Study Multidrug Transport Proteins in Liposomes David Miller and Paula J. Booth Abstract Biophysical measurements of multidrug transporters in vitro can often be of limited relevance to the natural in vivo behavior. In particular, the properties of transporters when removed from their native bilayer and solubilized in detergents or lipids can differ significantly from their in vivo properties, reducing the value of in vitro measurements for the design of antagonists to the transporters. This problem can be addressed by studying the transport protein in liposomes in which the properties of the liposome bilayer are altered through systematic changes in lipid composition. Isothermal titration calorimetry can be used to determine the properties of the lipid-reconstituted protein in bilayers of different lipid compositions as well as to quantify the percentage recovery of functional protein in different lipids. Both these measurements lead to an accurate analysis of substrate binding activity. The approach is illustrated here for the small multidrug transport protein, EmrE from Escherichia coli. The percentage of functional EmrE successfully reconstituted into liposome depends on lipid composition. Differences in ligand binding and subtle differences in the secondary structure also occur in different lipid compositions. Key words: Small multidrug transporters, Ligand binding activity, Dissociation constant, Isothermal titration calorimetry, Reconstitution

1. Introduction Multidrug exporters are of great pharmaceutical interest because of their role in drug resistance of bacterial pathogens and their overexpression in a large number of cancerous cells resulting in resistance to anti-cancer pharmaceuticals. The study of multidrug transporters in vitro is essential for a complete understanding of the mechanism and for the successful design of antagonists to the transporters. The small multidrug transporter EmrE is capable

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_17, © Humana Press, a part of Springer Science+Business Media, LLC 2010

233

234

Miller and Booth

of recognising and exporting a broad range of hydrophobic cationic substrates from Escherichia Coli (E. coli) (1) and has homologues in a number of bacterial species. The behavior of multidrug transporters in synthetic lipid bilayers is influenced by the bilayer composition and the quantity of functionally incorporated protein. The heterogenous nature of membrane bilayers complicates the detailed biophysical studies of integral membrane proteins. Interactions between the integral membrane proteins and the different regions of the bilayer are complex, with hydrophobic interactions occurring between the protein and the central hydrophobic region of the lipid bilayer, whilst polar and charge interactions exist with the lipid headgroups and aqueous solvent on either side of the bilayer. Furthermore, different lipid compositions, varying in headgroup and aliphatic chain composition, have significant effects on the protein function within the membrane, and there are variations in composition between different cellular compartments and different species (2). Altering the bilayer composition affects the properties of a number of integral membrane proteins, such as changes in the activity of enzymes involved in membrane lipid metabolism (Reviewed in (3, 4)), mechanosensitive channel opening properties (5–7), and the stability of the potassium channel KscA (8, 9). The consequences of lipid composition are not always considered in the structural and functional studies of membrane protein, which are often determined in detergent micelles, or in lipid environments that differ greatly from the native bilayer. The amount of functional transporter incorporated into the lipid bilayers will affect the kinetic parameters determined for transport or activity. We have used an isothermal titration calorimetry (ITC) based assay to quantify the amount of functional protein in synthetic liposomes, reconstituted from a ligand-binding competent, detergent-solubilized state in vitro. The ITC assay accurately quantifies the amount of functional protein from the ligand-binding activity and provides information on the ligandbinding properties in different lipid compositions. Specifically, the dissociation constant, Kd, is determined. In the ITC experiment, ligand is added to the transporter protein reconstituted into liposomes. The amount of heat released in the exothermic binding reaction between ligand and protein is measured for increasing ligand concentrations. This generates a binding isotherm from which Kd can be found by fitting the binding data to a particular binding model. The assay presented here is likely to be applicable for the study of other classes of transporters and receptors, provided a suitable ligand is available.

Use of Isothermal Titration Calorimetry

235

2. Materials 2.1. Protein Purification

All chemicals are purchased from Sigma unless otherwise stated. 1. Solubilization buffer: 40 mM Tris HCl, pH 8.2, 100 mM NaCl, 10 mM 2-mercaptoethanol (b-ME), and 4% (w/v) n-Dodecyl-b-D-Malopyranoside (DDM) from Anatrace. All buffers containing b-ME are freshly prepared before use. 2. Ni2+-NTA agarose (QIAGEN) 3. Ni2+ wash buffer 1: 20 mM Tris-HCl, pH 8.3, 400 mM NaCl, 15 mM imidazole, 0.1% w/v DDM, and 5 mM b-ME. 4. Ni2+ wash buffer 2: 20 mM Tris-HCl, pH 8.3, 20 mM NaCl, 15 mM imidazole, 0.1% w/v DDM, and 5 mM 2-mercaptoethanol. 5. Ni2+ elution buffer: 20 mM Tris-HCl, pH 8.3, 25 mM NaCl, 200 mM imidazole, 0.1% w/v DDM, and 5 mM 2-ME. 6. NH4DDM buffer: 15 mM Tris-HCl, pH 7.5, 190 mM NH4Cl, 0.08% (w/v) DDM, and 5 mM 2ME. 7. Denaturing wash buffer: 10 M urea and 5% w/v SDS (173 mM, CMC ~8 mM or ~0.2% w/v) in 20 mM Tris-HCl, pH 8.3. 8. Denaturing elution buffer: 10 M urea, 5% w/v SDS, 20 mM Na Acetate, pH 4.0. 9. Amicon Ultra 50,000 MWCO centrifugal concentrator (Millipore). 10. HiLoad 16/160 Superdex 200 gel filtration column (Amersham).

2.2. EmrE Reconstitution

1. 1,2-Dioleoyl-sn-Glycero-3-Phosphocholine (DOPC), 1,2-Dioleoyl-sn-Glycero-3-[Phospho-rac-(1-glycerol)](DOPG), and 1,2-Dioleoyl-sn-Glycero-3-Phosphoethanolamine (DOPE) (Avanti polar lipids, as powder) prepared as 50 mg/ml stock solutions in either 1:1 chloroform:methanol or cylclohexane. 2. n-Dodecyl-b-D-Malopyranoside (DDM) and Octyl-b-DGlucoside (OG) were purchased from Anatrace and prepared fresh as a 20% (w/v) stock in water. 3. Reconstitution buffer: 15 mM TRIS HCl, pH 7.5 or 8.5, and 190 mM NH4Cl. 4. 1 ml HisTrap columns (Amersham).

2.3. Circular Dichroism

1. Resuspension buffer: 5 mM TRIS HCl, pH 8.0, and 190 mM NH4Cl. 2. Dilution buffer: 20 mM TRIS HCl at pH 7.5 or pH 8.5.

236

Miller and Booth

3. TPP+: tetraphenyphosphonium bromide in 20 mM TRIS HCl, pH 7.5, freshly prepared from a 20 mg/ml stock in water. 4. Ethidium: Ethidium bromide in 20 mM TRIS HCl, pH 8.5, freshly prepared from a 10 mg/ml stock and stored in dark. 2.4. Protein Quantification

1. 1:1 chloroform:methanol solution prepared fresh and stored, wrapped in foil to prevent exposure to light.

3. Methods Efficient reconstitution of membrane protein is essential for ITC and circular dichroism (CD) measurements as large amounts of protein are typically required, and yields of overexpressed membrane protein are often low. Subsequent quantification of reconstituted transporter in lipid bilayers is a prerequisite for accurate activity measurements. The choice of detergent for purification and as a prelude to reconstitution is important as detergents such as DDM, which are very good at stabilising membrane proteins, are very difficult to remove from proteolipid samples, and may interfere with functional assays. Detergents such as OG, although easy to remove from proteolipid samples, do not suitably stabilize EmrE. Detergent choice is particularly important where activity assays involve a pH gradient across the bilayer as this cannot be maintained when residual DDM is present. A pH gradient is required in transport assays of EmrE. However, measurement of ligand binding is not required for the maintenance of a pH gradient, enabling the use of DDM. ITC binding assays provide a quantitative measure of ligand binding through the measurement of Kd. The stoichiometry of binding (n) can also be calculated from the binding isotherm. The binding of two ligands to EmrE is presented: the tight binding TPP+ (small Kd) and the weaker binding ethidium (larger Kd). ITC also has the advantage of determining the quantity of functional protein. Ligand binding to the small multidrug exporters occurs within the transmembrane region, and as ligand binding is a prerequisite for transport, it can be used as a quantitative measure of the amount of functional protein present. A prior knowledge of n enables the amount of functional reconstituted transporter to be determined by ITC (through non-linear curve fitting of the binding data). This is the case for EmrE, where n is known to be 0.5; one ligand binds to an EmrE dimer in DDM (10). The amount of protein in the reconstituted sample can be determined by various other methods. However, these often report on the amount of protein present and not whether it is also functional. A simple method for quantification

Use of Isothermal Titration Calorimetry

237

of the amount of protein involving solubilization of reconstituted protein in chloroform:methanol and quantification by A280 using standards prepared in the same solvent is presented. CD is also used to assess the overall secondary structure content of EmrE reconstituted into different lipid compositions. The use of a high intensity UV source with a detector geometry, suitable for collecting scattered light, is essential for accurate CD measurements of liposome containing samples. A synchrotron radiation source with correct detector geometry is the preferred method for collection of CD data of proteolipid samples (11). Scattering of light from samples and failure to collect a significant proportion of the scattered light will result in significant artefacts in data, most notably at shorter wavelengths. For an accurate comparison between different lipid compositions, concentration of inserted protein must be accurately determined (i.e., by chloroform/methanol solubilization). However, a qualitative comparison of a single lipid composition in the presence and absence of a ligand is possible without an accurate protein concentration, provided all the proteins present are functional. 3.1. Sample Preparation: EmrE Overexpression and Purification

1. EmrE-His (cloned into pT7-7) is overexpressed and purified as previously described (12, 13). Briefly, after protein expression, cultures are harvested and disrupted using a cell disrupter. Membranes are isolated by centrifugation, solubilized at 4°C for at least 2 h in solubilising buffer. Solubilized material is diluted 1:1 with water, and insoluble material removed by centrifugation at 35,000× g for 30 min at 4°C. NaCl and imidazole are added to final concentrations of 350 mM and 15 mM, respectively. 2. DDM solubilized protein is subsequently incubated with 1.5 ml of Ni-NTA agarose for 1.5 h at 4°C. Non-specifically bound protein is removed by washing with 20 columns of Ni2+ wash buffer 1, followed by 20 column volumes of Ni2+ wash buffer 2. 3. EmrE is eluted with 10 column volumes of Ni2+ elution buffer. Eluted protein is concentrated to 2 ml using an Amicon Ultra-15 Ultracel 50 k centrifugal concentrator (Millipore) and applied to a HiLoad Superdex 200 gel filtration column (Amersham), equilibrated with NH4DDM buffer. 4. A280 of eluted protein is monitored and fractions collected. Fractions which contain a symmetric protein elution peak, determined to contain EmrE, are collected; fractions outside of this range are discarded, including those which contributed to small shoulders on the EmrE elution peak. Eluted EmrE is concentrated to approximately 1–2 mM as determined by A280 (A280 of 1 mg/ml = 2.56) (10), snap frozen in liquid N2 and stored at −80°C until required.

238

Miller and Booth

3.2. Reconstitution of EmrE in Liposomes

1. Fresh stock solutions of the lipid: DOPC, DOPE, and DOPG at 50 mg/ml are prepared in cyclohexane and mixed to the desired molar ratios, evaporated to a thin film at room temperature and freeze-dried overnight. Dried lipid mixtures are rehydrated to 6.25 mg/ml (for ITC TPP+ binding studies) or 12.5 mg/ml (for ethidium binding) in reconstitution buffer at pH 7.5 or 8.5 for TPP+ or ethidium binding experiments, respectively (the higher pH being required for the weaker binding ethidium substrate). 2. Lipids are stirred for 15 min at room temperature (~22°C) and OG added from a 20% (w/v) stock solution to 0.65% or 1% (w/v) for TPP+ and ethidium binding experiments, respectively, and stirred for an additional hour. Lipid vesicles are extruded to 200 nm diameter by passaging 10 times through a polycarbonate filter (Nucleopore Track-Etch Membrane from Whatman) using a LIPEXTM extruder (Northern Lipids Inc). 3. EmrE is reconstituted to a calculated maximum of 10 µM (for TPP+ experiments) or 50 µM (for ethidium) by rapid mixing of 2.2 ml (for TPP+) or 4.4 ml (for ethidium) of lipid vesicle preparation to purified EmrE, giving a final EmrE concentration of 10 or 25 µM for TPP+ and ethidium, respectively, followed by incubation at room temperature for 30 min. Higher concentrations of EmrE are required for weaker binding ligands (Note 1) 4. Lipid vesicles are diluted 90-fold into the appropriate pH reconstitution buffer, stirred at room temperature for 1 h. 5. Proteoliposomes are recovered by centrifugation at 370,000× g for 50 min, followed by resuspension in 2.2 ml of reconstitution buffer. Reconstitution efficiency varies depending on the lipid composition. Therefore, quantification of the amount of functional protein present in the bilayer is essential for activity measurements of either transport or binding assays.

3.3. Isothermal Titration Calorimetry (ITC)

1. All ITC experiments are performed with a VP-ITC microcalorimeter (MicroCal), using supplied Origin 5.0 software (MicroCal) for data analysis. 2. ITC experiments provide a large amount of information, but are susceptible to a number of artifacts that are not normally observed in other binding assays. It is essential that buffers used in ITC experiments are identical (Notes 2–4). 3. All samples are thoroughly degassed and equilibrated to 5°C below the experimental temperature before use with a Thermovac (MicroCal). 4. Experiments are performed at 25°C, monitoring the change in power to the measurement cell compared to the reference cell.

Use of Isothermal Titration Calorimetry

239

The integral over time of the differential power is equal to the thermal energy (∆H), and is dependent on the binding reaction monitored in the particular buffer conditions (Note 5). Typically, experiments are performed using 1 initial injection of 7.5 µl of ligand, followed by 17 injections of 15 µl or using 1 initial injection of 10 µl of ligand, followed by 14 injections of 20 µl for TPP+ and ethidium binding experiments, respectively. Ligand concentrations between 20 and 40 µM TPP+ and 100–200 µM ethidium are used depending on the lipid composition and the percentage reconstitution of EmrE. Higher concentrations of ligand are required in experiments with a greater percentage of reconstituted EmrE to ensure saturation of binding sites after addition of all ligands (Note 6). 5. Control experiments to determine the heats of dilution of ligand are performed using lipid vesicles in the absence of EmrE (for TPP+ and ethidium binding experiments) and lipid vesicles containing a non-functional EmrE mutant, E14C (for ethidium binding experiments). Control heats of dilution for ligand-binding experiments were subtracted from either heats of dilution at saturation of EmrE or controls performed with ligand and lipid vesicles containing no protein; these two approaches are often identical. Controls are difficult to subtract from ethidium binding data because of differences in the background ethidium binding properties of liposomes. Control heats of dilution are determined to be linear for ethidium plus liposomes with no protein, and ethidium plus lipid vesicles containing inactive E14C EmrE. Heats of dilution for ethidium binding experiments were determined by subtracting a linear background determined from binding data after saturation. 6. The first data point in ITC binding experiments is always excluded from analysis because of premixing of the ligand solution in the injection syringe tip. After subtraction of a linear correction for control heats of TPP+ and ethidium dilution, binding data are fitted to a single-site binding model using the equations (14): 2    4X t  nM t ∆HV o  Xt Xt 1 1  1+ + − 1 + + − Q =  nM t nKM t 2 nM t nKM t  nM t    

Where ∆H is the molar heat of ligand binding, Vo is the working volume of the calorimeter cell, Q is the total heat content of Vo, Xt is the bulk concentration of ligand, Mt is the bulk concentration of receptor, and n is the number of binding sites. The heat evolved from the ith injection can be calculated as:

240

Miller and Booth

∆Q (i) = Q (i) +

dVi  Q (i) + Q (i − 1)   − Q (i − 1) Vo  2 

Where ∆Q(i) is the heat released from the ith injection and Vi is the injection volume. Fitting of the experimental data to the above equations is handled by Origin 5.0 (MicroCal) to determine the values for n, K, and ∆H. Kd is calculated as 1/K. Representative data of TPP+ and ethidium binding to EmrE in liposomes consisting of 100% DOPC is shown in Fig. 1. 3.4. Circular Dichroism Spectroscopy

1. Circular Dichroism (CD) measurements are performed at a Synchrotron Radiation Source (those shown were collected at Daresbury, UK). 2. Preparation of samples for CD measurements is essentially the same as required for ITC. EmrE for ligand-binding experiments was reconstituted as described above, with the exception of EmrE at 20 µM and lipid at 6.25 mg/ml.

Fig. 1. ITC data of EmrE binding the ligand (a) and (b): TPP+ in DOPC lipids, (c) and (d): EmrE binding ethidium in DOPC. Raw data is shown in (a) and (c) and corrected and integrated data shown in (b) and (d). Values of nexp calculated from the non-linear fitting of the binding isotherm (and resulting values for percentage reconstitution) of functional protein are 0.26 (52%) and 0.23 (46%) for TPP+ and ethidium binding experiments shown in (b) and (d), respectively

Use of Isothermal Titration Calorimetry

241

3. After recovery of liposomes by ultra-centrifugation, samples are resuspended in 5 mM TRIS HCl, pH 8.0, and 190 mM NH4Cl. Samples are diluted in 20 mM TRIS HCl at pH 7.5 with ligand at 50 µM or buffer only. Assuming maximum recovery of sample, the final EmrE and lipid concentrations is 30 µM and 9.4 mg/ml, respectively. Any insoluble material is removed by centrifugation at 13,000× g for 10 min, prior to measurement. 4. All CD measurements are performed in a 0.1 mm cuvette (Helma) at 25°C using a 1 nm wavelength increment and slit width. Duplicate samples are prepared and the resulting scans averaged. The amount of time required for measurement at each wavelength is dependent on the UV source used. A nonsynchrotron source will require significantly longer measurement times. CD data determined for EmrE in different lipid conditions in the presence and absence of the ligand TPP+ is shown in Fig. 2.

Fig. 2. Circular dichroism of EmrE, reconstituted in different lipid compositions in the presence and absence of the ligand TPP+. Data are shown for ligand binding in (a): 1 mole fraction (100%) DOPC, (b): 1 mole fraction (100%) DOPG, (c): 0.5 mole fraction DOPE+ 0.5 mole fraction DOPC (50% DOPE, 50% DOPC), and (d): 0.5 mole fraction DOPE + 0.5 mole fraction DOPG (50% DOPE, 50% DOPG). Solid lines: EmrE in the absence of ligand, dashed lines: EmrE in the presence of saturating concentrations of TPP+. Samples + and − ligand are prepared from identical proteolipid stock solutions and are directly comparable within each lipid composition

242

Miller and Booth

3.5. Protein Quantification by Solubilization of Lipid Vesicles with Organic Solvent

1. Proteolipid samples are solubilized in 1:1 chloroform:methanol, resulting in a final mix of 43:43:14 chloroform:methanol:sample (v/v). Other ratios may be used, as long as the resulting mixture does not phase separate. 2. The absorbance spectrum is measured between 250 and 310 nm to determine the protein absorbance at 280 nM (Note 7). Although only absorbance at 280 nM is required for protein concentration determination, it is useful to check the absorbance around this region to determine the contribution of other components to the measured absorbance. 3. The absorbance of EmrE standards of known concentration in DDM is determined in the region from 250 to 310 nm. 4. A linear baseline is subtracted from the absorbance spectra between 255 and 305 nm to eliminate non-protein absorbance and the UV absorbance band integrated. The band intensity can also be used for calibration, but integration of the area will provide a more accurate concentration. 5. Non-linear regression is used to determine the sample protein concentration. An example of representative data is shown in Fig. 3.

3.6. Protein Quantification by Non-linear Curve Fitting of ITC Binding Data

1. An ITC binding isotherm of the protein of interest is obtained and a value of nexp from the non-linear curve fitting obtained. To increase the accuracy of the determined nexp, EmrE can be reconstituted to a higher concentration of 50 mM in 12.5 mg/ml, and ITC experiments performed using TPP+ at 200 mM using 5 ml injections.

Fig.3.Protein quantification by solubilization of proteolipid samples in chloroform:methanol. Raw data are shown for standards (black lines) and an EmrE sample in DOPC (grey line). Standards are prepared from accurately determined EmrE samples in DDM solubilized in chlorofrm:methanol. Calibration curve derived from the integration of protein absorbance is shown as an inset. ITC data of TPP+ binding to the sample is shown in Fig. 1

Use of Isothermal Titration Calorimetry

243

2. Percentage reconstitution is calculated from: % reconstitution = nexp/nbinding × 100, where nexp is the experimentally determined value for n and nbinding is the known stoichiometry of the interaction (Notes 8 and 9).

4. Notes 1. The range of binding affinities that can be obtained by ITC is limited at the lower end (high affinity) by the sensitivity of the instrument, and at the upper end (low affinity) by the solubility and available concentrations of receptor. Realistically, for membrane proteins in liposomes, this limits the range from low nanomolar to low micromolar affinities. The lower limits of detection for ITC measurement require receptor at a concentration of ~ 5 µM; if ligand-binding affinity is weak then concentrations will have to be significantly higher. Although it is possible to work with low concentrations of protein to determine a dissociation constant, Kd, for weak binding, this will not produce enough data for fitting all parameters and values for the ligand:protein binding stoichiometry, n, will typically need to be defined. This is not possible if accurate protein concentrations are unknown or when reconstitution is inefficient. 2. Identical buffers for ITC experiments can be obtained either by dialysis of the protein in the required buffer, followed by the preparation of the ligand in dialysate, or ideally by gel filtration of the protein and preparation of the ligand in the gel filtration buffer. This has an advantage over dialysis of a final polishing purification of the protein and complete buffer exchange into the required buffer for ITC in less time than required by dialysis. 3. Ligand pH should be checked after solubilsation to ensure that there is no significant deviation in pH between ligand and protein solutions. Any pH differences must be adjusted, or significant buffer dilution effects will be observed. 4. Care should be taken with reducing agents in ITC as these can result in a significant deviation of the baseline during an experiment. Reducing agents such as DTT or TCEP should be avoided and concentrations of b-ME should be reduced. 5. ITC measurements may not be possible due to a small enthalpy change of the binding reaction under the observed conditions. Altering the temperature of a given reaction will result in a change in the observed enthalpy. If the binding reaction involves a protonation or deprotonation event, a buffer such as TRIS which has a large enthalpy of protonation can be used

244

Miller and Booth

to increase the signal. Care should be taken in the analysis of the resulting enthalpy of binding in such cases, as this will be a sum of all enthalpy changes in the cell including changes observed from the buffer. 6. The concentration of ligand required for binding experiments is dependent on the percentage reconstitution of functional protein. An approximate value for this can be determined from an initial binding experiment using an estimated concentration of ligand. Ideally upon completion of an ITC binding experiment, ligand should be at approximately two-fold excess to the available binding sites to ensure that a complete binding isotherm is obtained. 7. Evaluation of the protein absorbance band is necessary to ensure that the measured sample concentrations are accurate. After solubilization in organic solvent, the protein absorbance from approximately 250--300 nm should be determined and compared to control samples. 8. Accurate measurements of binding stoichiometry and quantification of functional protein are dependent on an accurate protein concentration – errors in protein concentration will be observed in any calculated stoichiometry. Such errors are typically obvious if the binding stoichiometry of fully functional protein in detergent is determined to be a non-integer value. Colorimetric assays for protein concentration may not provide an accurate value for membrane proteins, even when accounting for the presence of detergents. An accurate extinction coefficient for protein absorbance at 280 nm is ideal which can be determined by amino acid analysis. 9. Care must be taken with the assumption that the amount of ligand binding is proportional to the amount of functional protein. This may not be valid for integral membrane proteins in which ligands bind to a soluble domain that may function independently of the membrane-spanning region such as some classes of receptors. In the case of small multidrug transporters, the ligand-binding region is within the membrane and very little of the structure extends beyond the bilayer.

Acknowledgments This work was supported by the BBSRC (B19845), Leverhulme Trust (F/00182/AW) and The Royal Society. PJB holds a Royal Society Wolfson Research Merit Award. We thank Kalypso Charalambous, Paul Curnow, Mark Lorch, and other members of PJB’s research group for discussions of experimental work and Kath Moreton for excellent technical assistance.

Use of Isothermal Titration Calorimetry

245

References 1. Rotem D, Schuldiner S (2004) EmrE, a multidrug transporter from Escherichia coli, transports monovalent and divalent substrates with the same stoichiometry. J Biol Chem 279(47):48787–48793 2. Morein S, Andersson A, Rilfors L, Lindblom G (1999) Wild type Escherichia coli cells regulate the membrane lipid composition in a “window” between gel and non-lamellar structures. J Biol Chem 271:6801–6809 3. Johnson AE, van Waes MA (1999) The translocon: a dynamic gateway at the ER membrane. Annu Rev Cell Dev Biol 15:799–842 4. Cornell R, Arnold R (1996) Modulation of the activities of enzymes of membrane lipid metabolism by non-bilayer-forming lipids. Chem Phys Lipids 81(2):215–227 5. Perozo E, Kloda A, Cortes DM, Martinac B (2002) Physical principles underlying the transduction of bilayer deformation forces during mechanosensitive channel gating. Nat Struct Biol 9(9):696–703 6. Moe P, Blount P (2005) Assessment of potential stimuli for mechano-dependent gating of MscL: effects of pressure, tension, and lipid headgroups. Biochemistry 44(36): 12239–12244 7. Perozo E, Cortes DM, Sompornpisut P, Kloda A, Martinac B (2002) Open channel structure of MscL and the gating mechanism

8.

9.

10.

11.

12.

13.

14.

of mechanosensitive channels. Nature 418(6901):942–948 van den Brink-van der Laan E, Chupin V, Killian JA, de Kruijff B (2004) Stability of KcsA tetramer depends on membrane lateral pressure. Biochemistry 43(14):4240–4250 van den Brink-van der Laan E, Chupin V, Killian JA, de Kruijff B (2004) Small alcohols destabilize the KcsA tetramer via their effect on the membrane lateral pressure. Biochemistry 43(20):5937–5942 Butler PJ, Ubarretxena-Belandia I, Warne T, Tate CG (2004) The Escherichia coli multidrug transporter EmrE is a dimer in the detergentsolubilised state. J Mol Biol 340(4):797–808 Wallace BA (2000) Synchrotron radiation circular-dichroism spectroscopy as a tool for investigating protein structures. J Synchrotron Radiat 7(Pt 5):289–295 Tate CG, Kunji ER, Lebendiker M, Schuldiner S (2001) The projection structure of EmrE, a proton-linked multidrug transporter from Escherichia coli, at 7 A resolution. Embo J 20(1–2):77–81 Muth TR, Schuldiner SA (2000) membraneembedded glutamate is required for ligand binding to the multidrug transporter EmrE. Embo J 19(2):234–240 MicroCal (1998) ITC Data Analysis in Origin. MicroCal

Chapter 18 Studying Lipid Organization in Biological Membranes Using Liposomes and EPR Spin Labeling Witold K. Subczynski, Marija Raguz, and Justyna Widomska Abstract Electron paramagnetic resonance (EPR) spin-labeling methods provide a unique opportunity to determine the lateral organization of lipid bilayer membranes by discrimination of coexisting membrane domains or coexisting membrane phases. In some cases, coexisting membrane domains can be characterized without the need for their physical separation by profiles of alkyl chain order, fluidity, hydrophobicity, and oxygen diffusion-concentration product in situ. This chapter briefly explains how EPR spin-labeling methods can be used to obtain the above-mentioned profiles across lipid bilayer membranes (liposomes). These procedures will be illustrated by EPR measurements performed on multilamellar liposomes made of lipid extracts from cortical and nuclear fractions of the fiber cell plasma membrane of a cow-eye lens. To better elucidate the major factors that determine membrane properties, results for eye lens lipid membranes and simple model membranes that resemble the basic lipid composition of biological membranes will be compared. Key words: Liposomes, Lipid bilayer, Membrane domain, Cholesterol, Lens lipid, Hydrophobic barrier, Fluidity, Order, Oxygen permeation, Spin label, EPR

1. Introduction Electron paramagnetic resonance (EPR) spin-labeling methods provide information about the lateral organization of lipid membranes and also about the molecular dynamics and structure in the direction of the membrane depth. Using lipid spin labels, with EPR monitoring groups (free radical nitroxide moieties) located at different depths in the membrane, profiles of different membrane properties across the lipid bilayer can be obtained. Because of the overall similarity of the molecular structures of

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_18, © Humana Press, a part of Springer Science+Business Media, LLC 2010

247

248

Subczynski, Raguz, and Widomska

Fig. 1. Chemical structures of lipid spin labels including phospholipid analogs: n-PC, T-PC, and n-SASL and cholesterol analogs: CSL and ASL. Chemical structures of POPC and cholesterol molecules are included to illustrate approximate localizations of these molecules and nitroxide moieties of spin labels across the membrane

these spin labels with phospholipids and cholesterol (Fig. 1), they should – to a certain degree – approximate the distribution of phospholipid and cholesterol molecules between membrane domains, as well as cholesterol–phospholipid and cholesterol– cholesterol interactions in the membrane. These spin labels can be distributed between different membrane domains or membrane phases, which makes it possible not only to discriminate these domains and phases but also to characterize them without the need for their physical separation by profiles of certain membrane properties obtained in coexisting domains and phases. This is the case for the raft domain that coexists within bulk lipids (1–3) or the liquid-ordered phase that coexists with the liquid-disordered or solid-ordered phases (4, 5). Information concerning coexisting membrane domains is practically missing from membrane research. For some membrane compositions, and membrane lateral organization, the distribution of the lipid spin label is unique; it allows

Lipid Organization in Biological Membranes

249

Fig. 2. Schematic drawing of the raft domain in bulk lipids (a) and the pure cholesterol crystalline domain in the bulk phospholipid-cholesterol bilayer (b). The distribution and approximate localization of lipid spin labels in these domains are also shown. Phospholipid spin labels 5-, 10-, 16-, T-PC, and 9-SASL are distributed between the raft domain and bulk lipids (a) and are only located in the bulk phospholipid-cholesterol domain when it coexists with the pure cholesterol crystalline domain (b). Spin-labeled cholesterol analogues ASL and CSL are distributed between both domains for coexisting raft domain and bulk lipids (a) and for coexisting cholesterol crystalline and bulk phospholipid-cholesterol domains (b). The nitroxide moieties of spin labels are indicated by black dots

additional information about the structure and dynamics of coexisting membrane domains to be obtained. This is the case for membranes overloaded with cholesterol in which pure cholesterol crystalline domains are formed (6). Figure 2 is a schematic drawing illustrating the cases mentioned above. Both phospholipid-type and cholesterol analogue spin labels are distributed between raft and bulk domains (Fig. 2a), which allows these domains to be identified using the discrimination by oxygen transport (DOT) method, and profiles of the oxygen transport parameter (oxygen diffusion-concentration product) across each domain

250

Subczynski, Raguz, and Widomska

can be obtained (7). In membranes with the cholesterol crystalline domain, phospholipid-type spin labels should partition only into the bulk phospholipid-cholesterol domain (Fig. 2b). Thus, in addition to the profile of the oxygen transport parameter, profiles of the order parameter, fluidity, and hydrophobicity can be obtained using these spin labels (which should only describe properties of the bulk phospholipid-cholesterol domain without “contamination” from the cholesterol crystalline domain) (6). Cholesterol analogues should distinguish between both domains (Fig. 2b), and thus only ASL and CSL can detect and discriminate both the coexisting domains and allow information about the pure cholesterol crystalline domain to be obtained (6). The DOT method has already been described in Methods in Molecular Biology (7). This method has been successfully applied to discriminate the domains in reconstituted membranes crowded with integral membrane proteins (8), as well as in influenza virus envelope membranes, which contain cholesterol- and protein-rich raft domains (1). In model membranes, made from binary mixtures of phosphatidylcholine and cholesterol or sphingomyelin and cholesterol, liquid-ordered, liquid-disordered, and solidordered phases were distinguished and characterized in different regions of a phase diagram when they formed a single phase or when two phases coexisted (4, 5). In membranes made from a ternary raft-forming mixture, the raft domain was also distinguished from bulk lipids by the DOT method (2, 3). Similarly, the DOT method can be used to study the lipid organization in lipid bilayer membranes (liposomes), derived from the lipid extract of certain biological membranes. The main focus of this chapter will be on new applications of EPR spin-labeling methods to study membranes overloaded with cholesterol, like those of eye-lens fiber cell plasma membranes (6, 9, 10). The presence of the pure cholesterol crystalline domain in these membranes broadens the possibilities of the DOT method, which, in combination with conventional EPR spin-labeling approaches, enables rather complete information about membrane structure and its physical properties to be obtained (6). 1.1. EPR Spin-labeling Approaches for Profiles of Membrane Properties

To obtain detailed profiles across lipid bilayers, a variety of lipid spin labels are incorporated into the membrane for probing at specific depths and specific membrane domains (Fig. 1). EPR spin-labeling methods apply conventional EPR and saturationrecovery EPR techniques. Information obtained from conventional EPR spectra includes profiles of the order parameter (11) and hydrophobicity (12). Saturation-recovery EPR signals primarily provide information about collisions between paramagnetic molecules, including nitroxide–nitroxide (13), nitroxide–oxygen (14, 15), and nitroxide–paramagnetic metal ions (12). These collisions can be extracted from conventional EPR measurements

Lipid Organization in Biological Membranes

251

using the line-broadening approach (16) or continuous wave power saturation measurements (17), however, with less detail and less precision. Profiles of spin-lattice relaxation time, obtained directly from saturation-recovery EPR signals, also contain useful information about the dynamics (fluidity) of the membrane interior. All of the profiles presented here were obtained for liquid-crystalline phase membranes, except for hydrophobicity profiles. Hydrophobicity profiles were drawn based on measurements performed for a frozen suspension of membranes (liposomes) at −150 to −165°C. This is necessary to distinguish between motional and solvent effects on the EPR spectra (12). 1.2. Detailed Profiles Across Lipid Bilayers

To determine the correct profile across the lipid bilayer, knowledge of the position (depths) at which the nitroxide moiety is located in the membrane is very important. However, vertical fluctuations of the nitroxide moiety of stearic acid spin labels (n-SASL) and phospholipid spin labels (n-PC) toward the polar surface of the lipid bilayer have been reported (13, 18). It can be presumed from these studies that distribution of the vertical positions of the nitroxide moiety of n-SASL and n-PC in the membrane exists, with the mean value of each distribution shifting toward the center as the quantum n increases. Positions of carbon atoms in the alkyl chain have been determined by neutron diffraction (19), which shows that the mean positions of these carbons (or nitroxide moieties attached to these carbons) can be defined with an accuracy of ±1 Å, even in the liquid-crystalline state. Assuming a Gaussian distribution of the labeled segments in the projection on the bilayer normal, these authors reported that time-averaged positional fluctuations increase from 1.5 Å for the C4 position to 3.4 Å for the C12 position. It can be concluded that a nitroxide moiety stays at the position determined by neutron diffraction for most of the time (see Note 1).

1.3. Necessity of Measurements for Simple Membrane Models

It is strongly recommended that the results obtained for membranes made of lipids extracted from biological membranes and those obtained for simple two- or three-component membranes, made of commercially available lipids and that resemble the basic lipid composition of biological membranes, are compared. This comparison makes it possible to better elucidate the major factors (to indicate the major membrane components) that determine certain membrane properties. In the presented example of coweye lens lipid membranes, the simple models are membranes made of an equimolar binary mixture of POPC and cholesterol and of pure POPC (9, 10). This comparison allows us to conclude that the high cholesterol content in the lipid extracts from fiber cell plasma membranes is responsible for the unique membrane properties observed with EPR spin-labeling methods (see Subheading 3.5).

252

Subczynski, Raguz, and Widomska

2. Materials 1. Phospholipid spin labels (1-palmitoyl-2-(n-doxylstearoyl) phosphatidylcholine (n-PC, where n = 5, 7, 10, 12, 14, or 16), or tempocholine-1-palmitoyl-2-oleoylphosphatidic acid ester (T-PC)) can be purchased from Avanti Polar Lipids, Inc. (Alabaster, AL). n-Doxylstearic acid spin labels (n-SASL, where n = 5, 7, 9, 10, 12, or 16), cholestane spin label (CSL), and androstane spin label (ASL) can be purchased from Sigma (St. Louis, MO). Spin labels are dissolved in chloroform at 1 mM and stored in a freezer at −70°C. 2. Chloroform solutions of the total lipids extracted from a biological material (usually 5–20 mg/mL) are kept in a freezer at −70°C (see Note 2). 3. Stock solutions of commercially available lipids (phospholipids and cholesterol) from Avanti Polar Lipids, Inc. (Alabaster, AL) in chloroform (usually, 20–50 mg/mL) are kept in a freezer at −70°C. These lipids are used to form simple two- or threecomponent membrane models (see Subheading 1.3). 4. Buffers: Typically, 10 mM PIPES and 150 mM NaCl; pH 7.0 is used as a buffer. For samples with n-SASL, 0.1 M borate at pH 9.5 is used as a buffer. In this case, a rather high pH is chosen to ensure that all SASL probe carboxyl groups are ionized in lipid bilayer membranes (20, 21) (see Note 3).

3. Methods 3.1. Spin Labeling

1. Chloroform solutions of extracted lipids and an appropriate spin label are mixed to obtain a final concentration of 0.5 or 1 mol% of spin labels in the lipid bilayer (see Note 4). 2. n-PC, T-PC, and n-SASL are spin labels that allow hydrophobicity profiles and profiles of the oxygen transport parameter across the lipid bilayer to be obtained (see Fig. 1 for their structures and approximate localization in the lipid bilayer). 3. n-PC and n-SASL allow profiles of the alkyl chain order parameter across the lipid bilayer to be obtained. 4. n-PC, T-PC, and n-SASL allow discrimination of raft and bulk domains (1–3) or liquid-ordered, liquid-disordered, and solid-ordered phases (4, 5) (see Fig. 2a). 5. Only the spin-labeled cholesterol analogue ASL allows discrimination of the cholesterol crystalline domain from the bulk phospholipid-cholesterol domain (6) (see Fig. 2b and Subheading 3.5.4).

Lipid Organization in Biological Membranes

3.2. Preparation of Multilamellar Liposomes

253

Membranes used for EPR measurements are usually multilamellar dispersions of lipids (multilamellar liposomes (see Note 5)) prepared in the following way: 1. A chloroform solution of lipids and spin label (usually, 2–4 mg of total lipid per sample) (see Subheading 3.1, step 1) is placed in a glass test tube, and chloroform is evaporated with a stream of nitrogen. The lipid film on the bottom of the test tube is thoroughly dried under reduced pressure (about 0.1 mmHg) for about 12 h. 2. A buffer solution (usually, 0.5–1.0 mL) is added to the dried film at a temperature above the phase transition temperature of investigated membranes (for lipids extracted from biological materials, 40°C is sufficient) and vortexed vigorously. 3. The lipid dispersion is centrifuged briefly (15 min at 4°C with an Eppendorf bench centrifuge at 16,000× g), and the loose pellet (about 20% lipid, w/w) is used for EPR measurements.

3.3. Conventional EPR

1. For all EPR measurements, the loose pellet is transferred to a capillary made of gas-permeable methyl-pentene polymer known as TPX (see Note 6), and the end of the capillary is sealed with Baxter Miniseal wax B4425.1 (see Notes 7 and 8). This plastic is permeable to oxygen, nitrogen, and other gases, and is substantially impermeable to water. 2. The TPX capillary is fixed inside the EPR dewar insert in the resonator of the X-band EPR spectrometer with a special Teflon holder (see Note 9) and equilibrated with nitrogen gas, which is used for temperature control. 3. The sample is thoroughly deoxygenated, yielding correct EPR line shape (see Note 10). 4. To obtain profiles of the order parameter, EPR spectra are recorded for spin labels with the nitroxide moiety at different depths in the membrane (see Subheading 3.1, step 3). Only one type of spin-label molecule is present in each sample. Recording conditions include modulation amplitude of 0.5–1.0 G and an incident microwave power of about 5 mW. 5. The order parameter S is calculated using the equation (22) S = 0.5407(A¢II – A¢⊥)/a0, where a0 = (AII¢ + 2A⊥¢)/3

(1)

The values used for the calculation of the hydrocarbon chain order parameter, AII¢ and A¢⊥, are measured directly from EPR spectra as indicated in Fig. 3. 6. To obtain hydrophobicity profiles across the membrane, the z-component of the hyperfine interaction tensor, AZ, for spin labels with the nitroxide moiety at different depths in the membrane is determined directly from EPR spectra for samples frozen at about –165°C as indicated in

254

Subczynski, Raguz, and Widomska

Fig. 3. Panel of EPR spectra of 5-, 10-, and 16-PC in membranes made of cortical and nuclear cow-lens lipids. Spectra were recorded at 25°C. Measured values for evaluating the order parameter are indicated. The positions of certain peaks were evaluated with a high level of accuracy by monitoring them at 10 times higher receiver gain and, when necessary, at higher modulation amplitude

Fig. 4. EPR spectra of 16-PC in membranes made of cortical and nuclear cow-lens lipids. Spectra were recorded at −163°C to cancel motional effects. The measured 2AZ value is indicated

Fig. 4 (see Subheading 3.1, step 2). Only one type of spin-label molecule is present in each sample. Recording conditions include modulation amplitude of 2 G and an incident microwave power of 2 mW (12). In hydrophobicity profiles, 2AZ is plotted as a function of the approximate position of the nitroxide moiety in the lipid bilayer (see Note 11). 3.4. SaturationRecovery EPR (see Note 12)

The saturation-recovery EPR method for measuring electron spin-lattice relaxation time (T1) is a technique in which the recovery of the EPR signal is measured at a low level microwave field

Lipid Organization in Biological Membranes

255

(weak observing microwave power) after the end of the saturating microwave pulse. The time scale of this recovery is characterized by the spin-lattice relaxation time, T1. 1. The sample, prepared as described in Subheading 3.2, is transferred to the TPX capillary, positioned in the loop-gap resonator of the saturation-recovery EPR spectrometer, and deoxygenated by blowing nitrogen gas around the TPX capillary (see Notes 8 and 10). Thorough deoxygenation allows correct measurement of the spin-lattice relaxation time T1. The saturation-recovery signal is recorded at the required temperature for the spin label, with the nitroxide moiety at a fixed depth in the membrane. Only one type of spin-label molecule is present in each sample. 2. To obtain values of the oxygen transport parameter, the saturation-recovery signal is also recorded for the same sample, which is equilibrated with the required partial pressure of oxygen at the required temperature (see Note 13). 3. The same procedure is repeated for other spin labels, with the nitroxide moieties at different depths in the membrane (see Subheading 3.1, step 2). 4. T1s of spin labels in the absence and presence of molecular oxygen are determined by analyzing the saturation-recovery signal of the central line obtained by short-pulse saturationrecovery EPR (see Note 14). 5. The pulse length for short-pulse experiments is in the range of 0.1–0.5 ms. Pump power is selected to maximize the amplitude of the saturation-recovery signal and is typically in the range of 2–3.5 G. Observing power is selected to be as high as possible without affecting the time constant of the recovery. The minimum time between the end of the pulse and the beginning of observation of the recovery is determined by the ring-down time of the resonator and the switching transients, and is usually longer than 0.35 ms. Typically, 105–106 decays are acquired with 2,048 data points on each decay. Sampling intervals are from 1 to 32 ns, depending on the sample, temperature, and oxygen tension. The total accumulation time is typically 2–5 min. 6. Saturation-recovery signals are fitted by single and double exponentials and compared (see Fig. 5). If there is no substantial improvement in the fitting when the number of exponentials is increased from one, recovery curves are analyzed as single exponentials (see Fig. 5a–e, and Note 15). This is often the case for samples equilibrated with nitrogen (Fig. 5a, c, e). For samples equilibrated with a different partial pressure of oxygen, the saturation-recovery signal can often be fitted successfully with the double-exponential curve (as shown in Fig. 5f), indicating the presence of two coexisting domains or two coexisting phases (see Note 16).

256

Subczynski, Raguz, and Widomska

Fig. 5. Representative saturation-recovery signals with fitted curves and the residuals (the experimental signal minus the fitted curve) for 7-PC (a, b), CSL (c, d), and ASL (e, f) in membranes made of lens lipids isolated from the nuclear fraction of fiber cells of cow eyes. Signals were recorded at 25°C for samples equilibrated with 100% nitrogen gas (a, c, e) and with a gas mixture of 50% air and 50% nitrogen (b, d, f). Saturation-recovery signals for 7-PC and CSL were satisfactorily fitted to a single-exponential function in both the absence and presence of molecular oxygen with time constants of 4.37 ± 0.01 ms (a), 2.05 ± 0.01 ms (b), 2.95 ± 0.01 ms (c), and 1.91 ± 0.01 ms (d). For ASL, the saturation-recovery signal in the presence of molecular oxygen can be fit satisfactorily only with a double-exponential curve with time constants of 1.50 ± 0.28 ms and 0.55 ± 0.04 ms (compare the upper residual for single and lower residual for double-exponential fit in f), whereas single-exponential fit with a time constant of 2.71 ± 0.01 ms was satisfactory in the absence of molecular oxygen (e). Additional criteria for the goodness of single- and double-exponential fits are explained in Note 16

7. Calculation of the oxygen transport parameter from single-exponential decays is shown in Fig. 5a–e. The oxygen transport parameter is calculated using the equation (14) W(x) = T1–1(Air, x) – T1 –1 (N2, x) = AD (x)C (x).

(2)

T1(Air, x) and T1(N2, x) are spin-lattice relaxation times of nitroxides in samples equilibrated with atmospheric air and nitrogen, respectively. Note that W(x) is normalized to the sample equilibrated with atmospheric air. W(x) is proportional to the product of the local translational diffusion coefficient D(x) and the local concentration C(x) of oxygen at a depth x

Lipid Organization in Biological Membranes

257

in the membrane, which is in equilibrium with atmospheric air (see Note 17). 8. Calculation of the oxygen transport parameter from doubleexponential decays is shown in Fig. 5f. In membranes consisting of two lipid environments with different oxygen transport rates – the fast oxygen transport (FOT) domain and the slow oxygen transport (SLOT) domain – the saturation-recovery signals recorded for samples equilibrated with air and nitrogen are simple double-exponential curves with time constants of T1–1(Air, FOT), T1–1(Air, SLOT), and T1−1(N2, FOT), T1–1(N2, SLOT), respectively (1, 4, 5, 7, 8) (see Note 18). Thus, values of the oxygen transport parameter in each domain can be calculated using the equations: W(FOT) = T1–1 (Air, FOT) – T1–1(N2, FOT)

(3)

W(SLOT) = T1–1(Air, SLOT) – T1–1(N2, SLOT)

(4)

Here, “x” from (2) is changed to the two-membrane domains FOT and SLOT, and the depth fixed (the same spin label is distributed between the FOT and SLOT domains). 3.5. Profiles of Membrane Properties Across Homogeneous Membranes and Coexisting Membrane Domains

The final goal in the study of lipid organization in biological membranes using liposomes and EPR spin labeling is not only to characterize membranes by single (at one depth) spectral parameters but also to obtain detailed profiles of these parameters across membranes. These detailed profiles contain unique information about membrane structure and dynamics. Additionally, these profiles can often be obtained in coexisting membrane domains without the need for their physical separation, which provides unique opportunities in studies of physical properties of domains in situ. Using various spin-labeling techniques as well as conventional and saturation-recovery EPR spectroscopy (covering a time scale of 100 ps–10 ms), membrane molecular organization and dynamics can be investigated in the ps-to-ms range. The profiles of four parameters that were obtained with EPR spin-labeling methods and that describe the different properties of biological and model membranes are presented in the following sections, together with a short explanation of the information that can be extracted from these profiles.

3.5.1. Order Parameter

In the membrane, the alkyl chain of n-PC or n-SASL with the nitroxide moiety attached at the Cn position (see Fig. 1) undergoes a rapid anisotropic motion about the long axis of the spin label and a wobbling motion of the long axis within the confines of a cone imposed by the membrane environment. The order parameter (Eq. 1) is a measure of the amplitude of the wobbling motion. Increase in the order parameter indicates that the angle

258

Subczynski, Raguz, and Widomska

of the cone, responsible for the wobbling motion of the alkyl chain, decreases. Moreover, moving from the bilayer surface to the membrane interior, deviations in the alkyl chain segment direction from the bilayer normal accumulate. Consequently, ordering of the alkyl chain induced by stearic contact with the plate-like portion of cholesterol also causes ordering of the distal fragment of the alkyl chain, even though the rate of wobbling fluctuations can be higher (9, 13) (see Subheading 3.5.2). Although the order parameter indicates the static property of the lipid bilayer, for brevity, the change in the order parameter is most often described as a change in spin-label mobility, and thus as a change in membrane fluidity. Profiles of the molecular order parameter obtained at 25°C for the bulk phospholipid-cholesterol domain of cortical and nuclear cow-lens lipid membranes are displayed in Fig. 6a. In both membranes, values of the order parameter measured at the same depths are practically the same and are close to those measured for membranes made of the equimolar POPC/cholesterol mixture (Fig. 6b). They are, however, significantly greater than those measured for the pure POPC membrane (Fig. 6b), indicating that a saturating amount of cholesterol is responsible for the rigidity of the lensmembrane. In all membranes, profiles have an inverted bell shape and alkyl chain order that gradually decreases with an increase in membrane depth.

Fig. 6. Profiles of the molecular order parameter at 25°C obtained with n-PC and n-SASL across membranes made of cortical and nuclear cow-lens lipids are presented in (a) and across membranes made of the POPC/Chol equimolar mixture and of pure POPC are presented in (b). Approximate localizations of nitroxide moieties of the spin labels are indicated by arrows. For POPC and POPC/Chol membranes, data were taken from Refs. (9) and (11)

Lipid Organization in Biological Membranes 3.5.2. Fluidity

259

There are no easily obtainable EPR spectral parameters for lipid spin labels that can describe the profiles of membrane fluidity related to dynamic membrane properties. Only in the membrane center can membrane fluidity be evaluated directly from conventional EPR spectra of 16-PC (or 16-SASL) by measuring the effective rotational correlation time of the nitroxide moiety of the lipid spin label and assuming its isotropic rotational motion (23). However, already, 14-PC shows anisotropic rotational motion. Also, in the presence of 30 mol% cholesterol, motion of 16-PC moves into the slow-tumbling regime, where no conventional parameterization has been established (see Note 19). As was indicated in the subheading 3.5.1, the order parameter, which is most often used as a measure of membrane fluidity, describes, in principle, the static membrane properties – namely, the amplitude of the wobbling motion. Fortunately, the spin-lattice relaxation time (T1) is a spectral parameter that can be obtained from saturationrecovery EPR measurements with lipid spin labels (see Subheading 3.4). This parameter depends primarily on the rate of motion of the nitroxide moiety within the lipid bilayer, and thus describes the dynamics of the membrane environment at a depth at which the nitroxide fragment is located. It should be mentioned here that both the rotational motion (24) and the Brownian translational motion (25) are mechanisms involved in the spin-lattice relaxation process of nitroxide spin labels. Thus, T1 can be used as a conventional quantitative measure of membrane fluidity that indicates the rate of motion of phospholipid alkyl chains (or nitroxide free radical moieties attached to those chains). If T1 is measured for n-PC or n-SASL spin labels, a fluidity profile across the lipid bilayer can be obtained that reflects the membrane dynamics. In principle, these fluidity profiles can be obtained in coexisting domains and coexisting phases without the need for their physical separation (see Note 18). Fluidity profiles (T1 versus depth in the membrane) for the bulk phospholipidcholesterol domain of the cortical and nuclear cow-lens lipid membranes obtained at 25°C are presented in Fig. 7a. As expected, membrane fluidity (membrane dynamics) increases toward the membrane center, and profiles in both membranes are inverted bell-shaped, and are similar to profiles obtained for membranes made of the equimolar POPC/cholesterol mixture and pure POPC (Fig. 7b). As shown by comparing the profiles for the pure POPC bilayer and the POPC/cholesterol bilayer (Fig. 7b), cholesterol decreases the membrane fluidity close to the membrane surface and increases it at the membrane center. This confirms the earlier results (13, 15, 26) (see Note 20). The order parameter (the static membrane property) cannot differentiate the effects of cholesterol at different depths (see Subheading 3.5.1), while another dynamic parameter – namely, the oxygen transport parameter – clearly shows the differences between the membrane

260

Subczynski, Raguz, and Widomska

Fig. 7. Profiles of the electron spin-lattice relaxation time T1 for n-PC spin labels at 25°C across membranes made of cortical and nuclear cow-lens lipids are presented in (a) and across membranes made of the POPC/Chol equimolar mixture and of pure POPC are presented in (b). Approximate localizations of nitroxide moieties of the spin labels are indicated by arrows. For POPC membranes, data were taken from Ref. (40)

Fig. 8. Hydrophobicity profiles (2AZ) across membranes made of cortical and nuclear cow-lens lipids are presented in (a) and across membranes made of the POPC/Chol equimolar mixture and of pure POPC are presented in (b). Upward changes indicate increases in hydrophobicity. Because T-PC contains a different nitroxide moiety than n-PC and n-SASL, its points are not connected with other points. However, the relative changes of the hydrophobicity in the polar headgroup region can be evaluated (see Note 13). Broken lines indicate hydrophobicity in the aqueous phase (see Note 13). Approximate localizations of nitroxide moieties of the spin labels are indicated by arrows. For POPC and POPC/Chol membranes, data were taken from Refs. (9) and (11)

region where the cholesterol ring structure is located and the dipper region where the isooctyl chain of cholesterol is located (see Subheading 3.5.4). 3.5.3. Hydrophobicity

Figure 8a shows hydrophobicity profiles across the bulk phospholipid-cholesterol domain of cortical and nuclear cow-lens lipid membranes. Here, 2AZ data, obtained as described in

Lipid Organization in Biological Membranes

261

Subheading 3.3, step 5, are presented as a function of the approximate position of the nitroxide moiety of the spin label within the lipid bilayer. Smaller 2AZ values (upward changes in the profiles) indicate higher hydrophobicity. In both membranes, hydrophobicity profiles show a similar rectangular shape, with an abrupt increase of hydrophobicity between the C9 and C10 positions. 2AZ values in the center of both membranes (positions 10-, 12-, 14-, and 16-PC) indicate that the hydrophobicity in this region is only slightly lower than for pure hexane (e = 2) and can be compared to that of dipropylamine (see Note 21). It can be seen from Fig. 8a that the center of the bulk phospholipid-cholesterol domain of the nuclear membrane is less hydrophobic than that of the cortical membrane. A similar difference is also observed close to the membrane surface (5- and 7-PC positions). These interesting findings are in good agreement with an earlier observation (12) that cholesterol causes a significant increase in hydrophobicity of the PC lipid bilayer center when its concentration increases up to ~30 mol%; this is followed by a moderate decrease in hydrophobicity when the cholesterol concentration increases further, up to 50 mol%. The addition of cholesterol (from 0 to 50 mol%) monotonically decreases the hydrophobicity in the region close to the membrane surface. This confirms that the cortical bulk phospholipid-cholesterol domain has not yet reached saturation with cholesterol, while the nuclear bulk phospholipidcholesterol bilayer is saturated with cholesterol. Profiles presented in Fig. 8a are similar to those measured for the membrane made of the equimolar POPC/cholesterol mixture (Fig. 8b). These profiles also show a rectangular shape and an abrupt increase in hydrophobicity between the C9 and C10 positions, but differ from the typical bell-shaped profile of the pure POPC membrane (Fig. 8b), with a gradual increase in hydrophobicity toward the bilayer center. Also, the change in hydrophobicity between the membrane surface and the center is significantly lower for the pure POPC membrane than for the cortical and nuclear membranes. It can be concluded that the rectangular shape of the hydrophobicity profiles is characteristic of membranes with high cholesterol content. 3.5.4. Oxygen Transport Parameter

The oxygen transport parameter was introduced as a conventional quantitative measure of the rate of collision between spin label and molecular oxygen (Eq. 2). Kusumi et al. (14) conclude that the oxygen transport parameter is a useful monitor of membrane fluidity that reports on translational diffusion of small molecules. The profiles of the oxygen transport parameter for the bulk phospholipid-cholesterol domain of the cortical and nuclear cow-lens lipid membranes obtained at 25°C are presented in Fig. 9a (see Note 22). All profiles have a rectangular shape with an abrupt increase in the oxygen transport parameter between the C9 and C10 positions. This abrupt increase is as large as ~3 times, and the overall

262

Subczynski, Raguz, and Widomska

Fig. 9. Profiles of the oxygen transport parameter (oxygen diffusion-concentration product) at 25°C across membranes made of cortical and nuclear cow-lens lipids are presented in (a) and across membranes made of the POPC/Chol equimolar mixture and of pure POPC are presented in (b). Broken lines indicate the oxygen transport parameter in the aqueous phase. Approximate localizations of nitroxide moieties of spin labels are indicated by arrows. For POPC and POPC/Chol membranes, data were taken from Refs. (9) and (40). (PCD: phospholipid-cholesterol domain; CCD: cholesterol crystalline domain)

change of the oxygen transport parameter across the membrane becomes as large as ~6 times. The oxygen transport parameter from the membrane surface to the depth of the ninth carbon is as low as in gel-phase PC membranes, and at locations deeper than the ninth carbon, as high as in the fluid-phase membranes (4, 9, 11, 15, 26). These profiles are practically identical if we take into account the accuracy of the measurements (evaluated as 10%). Profiles are also very similar to those for the membrane made of the equimolar mixture of POPC and cholesterol (Fig. 9b). These profiles also show a rectangular shape and an abrupt increase in the oxygen transport parameter between the C9 and C10 positions. However, profiles for cortical and nuclear membranes differ from the bell-shaped profile across the pure POPC bilayer (Fig. 9b). This additionally confirms that high cholesterol content is responsible for the unique properties of lens lipid membranes. It should also be indicated that the abrupt change in hydrophobicity (Fig. 8) and oxygen transport parameter profiles (Fig. 9) is observed between the C9 and C10 positions, which is approximately where the steroid-ring structure of cholesterol reaches into the membrane. ASL shows two exponential saturation-recovery signals in nuclear membranes equilibrated with air/nitrogen mixture (Fig. 5f), indicating the presence of two membrane environments around ASL (see (Eq. 3) and (Eq. 4), and Notes 18 and 22). Comparing values of the oxygen transport parameter obtained with ASL and obtained with phospholipid-type spin labels in the bulk phospholipid-cholesterol domain allows the conclusion that ASL molecules, which give a greater oxygen transport

Lipid Organization in Biological Membranes

263

parameter, are located in the bulk phospholipid-cholesterol domain (see profile of the oxygen transport parameter for the nuclear lens lipid membrane with added points for the cholesterol analogue spin labels in Fig. 9a). The greater oxygen transport parameter value monitored with ASL is practically the same as the value obtained with 10-PC. This confirms that in the bulk phospholipid-cholesterol bilayer the nitroxide moiety of ASL is located close to the C10 position (6). The second value of the oxygen transport parameter monitored by ASL is about 5 times smaller than the value monitored by ASL in the bulk phospholipid-cholesterol domain and can be considered as that characterizing the pure cholesterol crystalline domain. In cortical membranes, ASL in the presence and absence of oxygen shows single-exponential saturation-recovery signals, indicating a homogenous environment (Fig. 5a). The oxygen transport parameter value detected by ASL is practically the same as that detected by 10-PC. It allows the assumption that cortical membranes consist of the bulk phospholipid-cholesterol domain without a detectable cholesterol crystalline domain. CSL in both cortical and nuclear membranes detects only homogenous environments, showing single-exponential saturation-recovery signals. Thus, the oxygen transport parameter values detected in the polar headgroup region of the nuclear lens lipid membrane, with CSL located in the bulk phospholipid-cholesterol domain and in the cholesterol crystalline domain, are very similar and cannot be distinguished by the DOT method (see Note 23). These results are in agreement with results obtained with ASL and CSL in piglens lipid membranes in which the cholesterol crystalline domain was induced by adding excess cholesterol (6). Values of the oxygen transport parameter measured with CSL and ASL in the nuclear lens lipid membrane are indicated in Fig 18.9a, and the profile of the oxygen transport parameter across the cholesterol crystalline domain, which coexists with the bulk phospholipidcholesterol domain, is shown. We would like to direct readers to Ref. (41), which was recently accepted for publication in Biochimica et Biophysica Acta. This paper describes in detail our investigation of the physical properties of membranes derived from the total lipid extract from the lens cortex and nucleus of a two-year-old cow using EPR spin-labeling methods.

4. Notes 1. The thickness of the lens lipid membrane is assumed to be the same as the thickness of the POPC/Chol = 1/1 membrane. It is also assumed that the location of the alkyl chain carbon atom in the membrane changes linearly with the position of

264

Subczynski, Raguz, and Widomska

the alkyl chain. The nitroxide moieties of n-PC and n-SASL are located at the same depth as the carbon atoms of the 2-chain of phospholipid. For details, see (10, 15, 26). 2. The commonly used procedure for extracting lipids from animal tissues was described by Folch and co-authors in 1957 (27). This procedure was also used with minor modification to extract lipids from cow-eye lenses (9, 10). The choice of extraction procedure should depend on the nature of the tissue matrix as well as other factors, such as specific lipid classes. The principles and good practice of tissue handling and lipid separation has been described in many specialist articles, books, and reviews (28–30). 3. At a pH of 7.0, a mixture consisting of two forms of SASLs can be presented in the lipid bilayer (with protonated and ionized carboxyl groups), and nitroxide moieties can be positioned at two different depths in the membrane. 4. A concentration of spin labels in the lipid bilayer that is too high affects the measurement of spectral parameters, especially spinlattice relaxation time. If the composition of the lipid extract (see Note 2) is known, the concentration of spin labels can be easily calculated. Otherwise, the average molecular weight of phospholipids in the total lipid extract has to be assumed and used for the calculation of spin-label concentration (see (9)). 5. Membranes in samples made from multilamellar liposomes are tightly packed, giving much better signal-to-noise ratio than in samples made from unilamellar liposomes. 6. For measurements at X-band, sample tubes are machined from TPX with dimensions of 0.6 mm ID, 0.1 mm wall thickness, and 25 mm length. Capillaries are machined from TPX rods, which can be purchased from Midland Plastic (Madison, WI). 7. This sealant is no longer commercially available but can be found in many laboratories. Other tube sealants can be used, including Critoseal (Fisher Scientific) and X-Sealant (Bruker Biospin). 8. It is often desirable to additionally concentrate the sample inside the TPX capillary by centrifugation in order to improve the signal-to-noise ratio (31). 9. TPX capillaries, together with the Teflon holder, can be obtained from Molecular Specialties (Milwaukee, WI). 10. Molecular oxygen is paramagnetic, having a triplet ground state, and bimolecular collisions of molecular oxygen with spin labels affect the EPR spectral parameters including line width and spin-lattice relaxation time (32). 11. Hydrophobicity profiles are constructed based on EPR measurements of 2AZ for n-PC and n-SASL spin labels. For SASLs in the aqueous phase, 2AZ was calculated as described

Lipid Organization in Biological Membranes

265

in (12). In addition, T-PC is used to probe the membrane polar headgroup region. Because the chemical structure of the nitroxide moiety of T-PC differs from that for n-PC and n-SASL, 2AZ values measured with T-PC cannot be directly compared with those for n-PC and n-SASL. However, T-PC shows relative hydrophobicity changes in polar headgroup regions after the addition of certain membrane modifiers (like cholesterol). 2AZ values measured with T-PC can also be used to compare the hydrophobicity of polar headgroup regions in different membranes. 12. The state-of-the-art X- and Q-band saturation-recovery EPR spectrometers were built and are available at the National Biomedical EPR Center, Medical College of Wisconsin, Milwaukee, WI, USA. The mission of the Center is to make advanced EPR research resources available to investigators nationally, regionally, and locally (see the link for Center use: http://www.mcw.edu/display/router.asp?docid=3211). Another X-band saturation-recovery spectrometer was built and is located at the Department of Biophysics, Faculty of Biotechnology, Jagiellonian University, Krakow, Poland. Presently, Bruker produces EPR spectrometers that are capable of saturation-recovery measurements at X-band. Pulse saturation recovery is possible on an E-580 FT/EPR system equipped with DC-AFC and LCW (low power CW arm) options and combined with an AmpX CW microwave power amplifier. Saturation recovery is treated as an accessory to the E-580 and is not usually a stand-alone configuration. 13. Switching the gas around the TPX capillary from nitrogen to the air/nitrogen mixture allows one to equilibrate the sample with the required partial pressure of oxygen for oximetry measurements and for obtaining T1s in the presence of molecular oxygen. Because the same gas mixture is used for temperature control, samples are equilibrated with oxygen at the required temperature. The mixture of air and nitrogen is adjusted with flow meters (Matheson Gas Products, Montgomeryville, PA, Model 7631 H-604). 14. The short-pulse method is favorable for multiexponential decays in oximetry measurements (33, 34). For a short pulse, only populations of the irradiated transition are affected; for a long pulse, all populations are altered because of transverse relaxations. Ref. (35) addresses the long- and short-pulse saturation-recovery methods in more detail. 15. Additional criteria for a good single-exponential fit are the negligible preexponential coefficient for the second component, the large standard deviation of T1 for the second component, and the repetition of the fit for different recording conditions such as the number of points and time increment.

266

Subczynski, Raguz, and Widomska

16. Although saturation-recovery signals in the absence of molecular oxygen cannot differentiate between these two domains, in the presence of oxygen the recovery curves are very different in each domain, indicating that the collision rate of molecular oxygen, or the oxygen transport parameter, is quite different. This is a good illustration of why the method is named DOT – because different domains can be clearly discriminated and characterized only in the presence of molecular oxygen. 17. A is remarkably independent of the hydrophobicity and viscosity of the solvent and of spin-label species (36–38). 18. When located in two different membrane domains, the spin label alone most often cannot differentiate between the two, giving very similar conventional EPR spectra and similar T1 values (T1−1(N2, FOT) » T1−1(N2, SLOT)). However, even small differences in lipid packing in these domains will affect oxygen partitioning and oxygen diffusion, which can be easily detected by observing the different T1s from spin labels in these two locations in the presence of oxygen. 19. In principle, the rotational motion of the 16-PC molecule as a whole, which is anisotropic, has to be distinguished from segmental motion, which comes from gauche-trans isomerization of the alkyl chain. For carbon atoms near the terminal methyl group (16-PC position), it is assumed that the segmental motion is not restricted, and as a result, motion of the nitroxide moiety is approximately isotropic. At higher temperatures, the motion of all molecules becomes so great that segmental motion dominates. At lower temperatures, the segmental motion is diminished and the motion of 16-PC becomes highly anisotropic. 20. To the best of our knowledge, the profile of the spin-lattice relaxation time is used here for the first time as a monitor of membrane dynamics and fluidity. It should be indicated that T1 is sensitive to the molar concentration of spin labels in the lipid bilayer. Therefore, the concentration of all spin labels in the lipid bilayer should be the same, as well as in membranes that are compared. This is not an easy task, especially for lipid bilayer membranes (liposomes) derived from the lipid extract of certain biological membranes. This may be the reason that profiles for cortical and nuclear lens lipid membranes presented in Fig. 7a are shifted relative to each other. Other EPR spectral parameters (molecular order, hydrophobicity, and oxygen transport parameter) are affected much less by spin-label concentration. Profiles for POPC and POPC/Chol presented in Fig. 7b are based on data obtained in 2003 (40) and 2007 (9), respectively. The effects of cholesterol on T1 of the lipid spin labels located in the membrane center and close

Lipid Organization in Biological Membranes

267

to the membrane surface are also clearly shown in Refs. (13, 15, 26). 21. For brevity, we refer to Fig. 2 in Ref. (12) to show the relation of the local hydrophobicity as observed by 2AZ to the hydrophobicity (or e) of the bulk organic solvent. 2AZ in bulk solvent provides a convenient yardstick for describing the local hydrophobicity in the membrane, and comparison of two 2AZ values may help to develop a “feel” for local hydrophobicity. Such a comparison is only semi-quantitative and could be made operationally because small changes in 2AZ correspond to large changes in e and the mechanism by which the presence of water affects 2AZ is not well understood (12, 39). 22. Restrictions on the distribution of lipid spin labels in membranes containing the cholesterol crystalline domain indicate that only spin-labeled cholesterol analogues can discriminate this domain (6). These analogues should approximate the distribution of cholesterol molecules in the membrane because of the overall similarity of CSL, ASL, and cholesterol molecular structures. Phospholipid spin labels, which should not partition into the cholesterol crystalline domain, cannot discriminate these domains. Indeed, in both cortical and nuclear membranes, saturation-recovery signals for n-PC, 9-SASL, and T-PC are single-exponential signals and are assigned as signals characterizing the bulk phospholipidcholesterol bilayer. 23. Results presented in Figs. 5 and 9 are an excellent illustration of the advantages and limitations of the DOT method. To detect membrane domains, lipid spin labels have to be distributed between these domains (like CSL and ASL, but unlike n-PC (see Fig. 2b)). Even when located in two different membrane domains, the spin label alone most often cannot differentiate between the two, giving very similar T1 values (in the absence of oxygen, saturation-recovery signals for both CSL and ASL were single-exponential signals (Fig. 5), indicating that T1 values in both environments are very close). In membranes equilibrated with air and consisting of two lipid environments with different oxygen transport rates – the fast oxygen transport (FOT) domain and the slow oxygen transport (SLOT) domain – the saturation-recovery signal should be a double-exponential curve with time constants of T1(air, FOT) and T1(air, SLOT) (see Subheading 3.5 and (Eq. 3) and (Eq. 4)). This is the case with ASL, which shows doubleexponential saturation-recovery signals for the sample equilibrated with the air/nitrogen mixture (Fig. 5f). CSL cannot distinguish these domains (Fig. 5c, d), probably because collision rates between oxygen and the nitroxide moiety of CSL located in the polar headgroup region are similar.

268

Subczynski, Raguz, and Widomska

Acknowledgments This work was supported by grants EY015526, EB002052, and EB001980 of the National Institutes of Health. References 1. Kawasaki K, Yin J-J, Subczynski WK, Hyde JS, Kusumi A (2001) Pulse EPR detection of lipid exchange between protein-rich raft and bulk domains in the membrane: methodology development and its application to studies of influenza viral membrane. Biophys J 80:738–748 2. Wisniewska A, Subczynski WK (2006) Accumulation of macular xanthophylls in unsaturated membrane domains. Free Radic Biol Med 40:1820–1826 3. Wisniewska A, Subczynski WK (2006) Distribution of macular xanthophylls between domains in a model of photoreceptor outer segment membranes. Free Radic Biol Med 41:1257–1265 4. Subczynski WK, Wisniewska A, Hyde JS, Kusumi A (2007) Three-dimensional dynamic structure of the liquid-ordered domain as examined by a pulse-EPR oxygen probing. Biophys J 92:1573–1584 5. Wisniewska A, Subczynski WK (2008) The liquid-ordered phase in sphingomyelincholesterol membranes as detected by the discrimination by oxygen transport (DOT) method. Cell Mol Biol Lett 13:430–451 6. Raguz M, Widomska J, Dillon J, Gailard ER, Subczynski WK (2008) Characterization of lipid domains in reconstituted porcine lens membranes using EPR spin-labeling approaches. Biochim Biophys Acta 1778:1079–1090 7. Subczynski WK, Widomska J, Wisniewska A, Kusumi A (2007) Saturation-recovery electron paramagnetic resonance discrimination by oxygen transport (DOT) method for characterizing membrane domains. In: McIntosh TJ (ed) Methods in molecular biology. Lipid rafts, vol 398. Humana Press, Totowa, pp 143–157 8. Ashikawa I, Yin J-J, Subczynski WK, Kouyama T, Hyde JS, Kusumi A (1994) Molecular organization and dynamics in bacteriorhodopsin-rich reconstituted membranes: discrimination of lipid environments by the oxygen transport parameter using a pulse ESR spin-labeling technique. Biochemistry 33:4947–4952

9. Widomska J, Raguz M, Dillon J, Gaillard ER, Subczynski WK (2007) Physical properties of the lipid bilayer membrane made of calf lens lipids: EPR spin labeling studies. Biochim Biophys Acta 1768:1454–1465 10. Widomska J, Raguz M, Subczynski WK (2007) Oxygen permeability of the lipid bilayer membrane made of calf lens lipids. Biochim Biophys Acta 1768:2636–2645 11. Subczynski WK, Lewis RNAH, McElhaney RN, Hodges RS, Hyde JS, Kusumi A (1998) Molecular organization and dynamics of 1-palmitoyl-2-oleoylphosphatidylcholine bilayers containing a transmembrane a-helical peptide. Biochemistry 37:3156–3164 12. Subczynski WK, Wisniewska A, Yin J-J, Hyde JS, Kusumi A (1994) Hydrophobic barriers of lipid bilayer membranes formed by reduction of water penetration by alkyl chain unsaturation and cholesterol. Biochemistry 33:7670–7681 13. Yin J-J, Subczynski WK (1996) Effects of lutein and cholesterol on alkyl chain bending in lipid bilayers: a pulse electron spin resonance spin labeling study. Biophys J 71:832–839 14. Kusumi A, Subczynski WK, Hyde JS (1982) Oxygen transport parameter in membranes as deduced by saturation recovery measurements of spin-lattice relaxation times of spin labels. Proc Natl Acad Sci U S A 79: 1854–1858 15. Subczynski WK, Hyde JS, Kusumi A (1989) Oxygen permeability of phosphatidylcholinecholesterol membranes. Proc Natl Acad Sci U S A 86:4474–4478 16. Smirnov AI, Clarkson RB, Belford RL (1996) EPR linewidth (T2) method to measure oxygen permeability of phospholipids bilayer and its use to study the effect of low ethanol concentration. J Magn Reson B 111:149–157 17. Altenbach C, Froncisz W, Hyde JS, Hubbell WL (1989) Conformation of spin-labeled melittin at membrane surface investigated by pulse saturation recovery and continuous wave power saturation electron paramagnetic resonance. Biophys J 56:1183–1191

Lipid Organization in Biological Membranes 18. Merkle H, Subczynski WK, Kusumi A (1987) Dynamic fluorescence quenching studies on lipid mobilities in phosphatidylcholine-cholesterol membranes. Biochim Biophys Acta 897: 238–248 19. Zaccai G, Büldt G, Seelig A, Seelig J (1979) Neutron diffraction studies on phosphatidylcholine model membranes II. Chain conformation and segmental disorder. J Mol Biol 134:693–706 20. Egreet-Charlier M, Sanson A, Ptak M, Bouloussa O (1978) Ionization of fatty acids at lipid–water interface. FEBS Lett 89:313–316 21. Kusumi A, Subczynski WK, Hyde JS (1982) Effects of pH on ESR spectra of stearic acid spin labels in membranes: probing the membrane surface. Fed Proc 41:1394 22. Marsh D (1981) Electron spin resonance: spin labels. In: Grell E (ed) Membrane Spectroscopy. Springer-Verlag, Berlin, pp 51–142 23. Berliner LJ (1978) Spin labeling in enzymology: spin-labeled enzymes and proteins. Rotational correlation times calculation. Methods Enzymol 49:466–470 24. Atkins PW, Kivelson D (1966) ESR linewidth in solution. II. Analysis of spin-rotational relaxation data. J Chem Phys 44:169–174 25. Robinson BH, Hass DA, Mailer C (1994) Molecular dynamics in lipid spin lattice relaxation of nitroxide spin labels. Science 263:490–493 26. Subczynski WK, Hyde JS, Kusumi A (1991) Effect of alkyl chain unsaturation and cholesterol intercalation on oxygen transport in membranes: a pulse ESR spin labeling study. Biochemistry 30:8578–8590 27. Folch J, Lees M, Sloane Stanley GH (1957) A simple method for the isolation and purification of total lipids from animal tissues. J Biol Chem 226:497–509 28. Markham JE, Li J, Cahoon EB, Jaworski JG (2006) Separation and identification of major plant sphingolipid classes from leaves. J Biol Chem 281:22684–22694 29. Bodennec J, Pelled D, Futerman AH (2003) Aminopropyl solid phase extraction and 2 D TLC of neutral glycosphingolipids and neutral lysoglycosphingolipids. J Lipid Res 44:218–226 30. Christie WW (2003) Lipid analysis: isolation, separation, identification and structural analysis of lipids, 3rd edn. Oily, Bridgwater

269

31. Subczynski WK, Felix CC, Klug CS, Hyde JS (2005) Concentration by centrifugation for gas exchange EPR oximetry measurements with loop-gap resonators. J Magn Reson 176:244–248 32. Subczynski WK, Swartz HM (2005) EPR oximetry in biological and model samples. In: Eaton SS, Eaton GR, Berliner LJ (eds) Biological magnetic resonance, biomedical epr-part a: free radicals, metals, medicine, and physiology, vol 23. Kluwer/Plenum, New York, pp 229–282 33. Yin J-J, Hyde JS (1987) Spin-label saturationrecovery electron spin resonance measurements of oxygen transport in membranes. Z Phys Chem (Munich) 153:57–65 34. Hyde JS, Yin J-J, Feix JB, Hubbell WL (1990) Advances in spin label oximetry. Pure Appl Chem 62:255–260 35. Yin J-J, Hyde JS (1989) Use of high observing power in electron spin resonance saturation-recovery experiments in spin-labeled membranes. J Chem Phys 91:6029–6035 36. Hyde JS, Subczynski WK (1984) Simulation of ESR spectra of the oxygen-sensitive spin-label probe CTPO. J Magn Reson 56:125–130 37. Hyde JS, Subczynski WK (1989) Spin-label oximetry. In: Berliner LJ, Reuben J (eds) Biological magnetic resonance, vol 8. Plenum, New York, pp 399–425 38. Subczynski WK, Hyde JS (1984) Diffusion of oxygen in water and hydrocarbons using an electron spin resonance spin-label technique. Biophys J 45:743–748 39. Griffith OH, Dehlinger PJ, Van SP (1974) Shape of the hydrophobic barrier of phospholipids bilayers (evidence for water penetration into biological membranes). J Membr Biol 15:159–192 40. Subczynski WK, Pasenkiewicz-Gierula M, McElhaney RN, Hyde JS, Kusumi A (2003) Molecular dynamics of 1-palmitoyl-2oleoylphosphatidylcholine membranes containing transmembrane a-helical peptides with alternating leucine and alanine residues. Biochemistry 42:3939–3948 41. Raguz M, Widomska J, Dillon J, Gaillard ER, Subczynski WK (2009) Physical properties of the lipid bilayer membrane made of cortical and nuclear bovine lens lipids: EPR spinlabeling studies. Biochim Biophys Acta, doi:10.1016/j.bbamem.2009.09.005

Chapter 19 Membrane Translocation Assayed by Fluorescence Spectroscopy Jana Broecker and Sandro Keller Abstract Assessing the ability of biomolecules or drugs to overcome lipid membranes in a receptor-independent way is of great importance in both basic research and applications involving the use of liposomes. A combination of uptake, release, and dilution experiments performed by steady-state fluorescence spectroscopy provides a powerful, straightforward, and inexpensive way of monitoring membrane translocation of fluorescent compounds. This is particularly true for peptides and proteins carrying intrinsic tryptophan residues, which eliminates the need for attaching extrinsic labeling moieties to the compound of interest. The approach encompasses three different kinds of fluorescence titrations and some simple calculations that can be carried out in a spreadsheet program. A complete set of experiments and data analyses can typically be completed within two days. Key words: Membrane binding, Membrane permeability, Membrane permeation, Flip–flop, Transbilayer movement, Uptake, Release, Dilution, Tryptophan fluorescence, Vesicles

1. Introduction Determining the permeability of lipid membranes to peptides, proteins, small molecules, and other biologically relevant compounds is a frequent task in many fields of liposome research. Heerklotz et al. (1–3) established a combination of so-called uptake and release experiments as a widely applicable solution to this problem on the basis of isothermal titration calorimetry (ITC). This approach has since been exploited to monitor membrane translocation (also referred to as membrane permeation or flip–flop) of such diverse compounds as photoactivatable nucleotide precursors (4), small molecules developed for conditional gene expression (5) or ion channel gating (6), surfactants (7–10),

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_19, © Humana Press, a part of Springer Science+Business Media, LLC 2010

271

272

Broecker and Keller

and peptides (11). A detailed step-by-step guide to microcalorimetric uptake and release experiments can be found in the literature (12). We have recently adapted (11) this approach to assess membrane translocation by spectroscopic methods. Fluorescence spectroscopy lends itself particularly well to this purpose, because it is available in many laboratories, generates datasets that can be interpreted in a straightforward way, and sensitively monitors membrane binding over a broad range of fluorophore concentrations. If, in addition, the molecule of interest is a peptide (or a protein) whose association with lipid membranes is accompanied by changes in the emission spectrum of intrinsic fluorescent amino acid residues, there is no need for attaching extrinsic fluorescent labels, which might drastically affect interactions with lipids. For the sake of simplicity, we will henceforth refer to the fluorophore at hand as peptide; however, the approach can easily be extended to any other membrane-interacting molecule whose fluorescence properties change upon membrane binding. Even if this is not the case, dialysis may be used to adapt the assay to the needs of other spectroscopic methods such as simple absorbance readings (13). The rationale underlying the fluorescence-spectroscopic approach is simple, involving three different titrations referred to as uptake, release, and dilution experiments (Fig. 1): In the uptake experiment, lipid vesicles are titrated into a peptide solution. a

uptake

+

–

or

release

b

–

+

or

c

+

dilution

– or

Fig. 1. Schematic representation of (a) uptake, (b) release, and (c) dilution experiments to assess the translocation of a peptide (black) across lipid vesicle membranes (grey). The two extreme cases of one-sided membrane binding (−) and transbilayer equilibration (+) are depicted. (Reproduced with permission from ref. 11. Copyright Wiley-VCH Verlag GmbH & Co. KGaA.)

Membrane Translocation Assayed by Fluorescence Spectroscopy

273

Uptake of the peptide occurs into the outer membrane leaflet only or into both leaflets depending on the peptide’s ability to translocate across the membrane. In the release experiment, lipid vesicles are first homogeneously preloaded with peptide on both leaflets and then titrated into pure buffer, causing release of the peptide from the membrane into the aqueous phase. Importantly, complete desorption can be achieved only if the peptide is membrane-permeant; otherwise, the fraction trapped inside the vesicles will remain adsorbed to the inner leaflet. In the dilution experiment, peptide is first applied externally to preformed lipid vesicles, and this mixture is then injected into buffer. Irrespective of membrane translocation, the peptide should thus completely desorb from the membrane upon strong dilution. This means that the release and dilution experiments will yield identical results only if the peptide can equilibrate across lipid bilayers, whereas differences between the two setups indicate one-sided membrane binding or incomplete translocation on the experimental time scale. A quantitative comparison between the release and dilution experiments is best accomplished by calculating the fraction of membrane-bound peptide, J, according to

J =

∑ (F (l ) − F

aq

∑ (F

b

)(

(l ) F b (l ) − F aq (l )

(l ) − F aq (l )

)

2

)

(1)

Here, F(l) is the concentration-normalized fluorescence intensity at wavelength l; F aq(l) is the corresponding value of free peptide in aqueous (aq) solution, which can be measured in the absence of lipid vesicles; and F b(l) is the corresponding value of completely membrane-bound (b) peptide, which is obtained in the presence of excess lipid (11). This chapter provides a recipe for performing uptake, release, and dilution experiments with the aid of intrinsic tryptophan fluorescence, and for analyzing such data on the basis of Eq. (1). The approach is exemplified using the cationic cell-penetrating peptide (CPP), penetratin (14), which carries two intrinsic tryptophan residues and avidly interacts with negatively charged phospholipid membranes (11, 13). The mechanisms of cellular internalization of penetratin and other CPPs have been the subject of controversy for several years (15, 16), and passive transbilayer diffusion has been invoked as a possible route of entry (17, 18). Using the fluorescence-spectroscopic approach presented here, however, we found no evidence of translocation across pure lipid membranes (11), which is in agreement with the results obtained by a number of other methods (13). The protocol, as presented here, turned out to be optimal for the specific case of penetratin. Other peptides, and even more

274

Broecker and Keller

non-peptidic fluorophores, will most likely require adaptations with respect to lipid and fluorophore concentrations. We, therefore, recommend to initially perform the uptake experiment to get an idea of the range of lipid concentrations required to observe near-complete membrane binding. The release and dilution experiments should then be carried out using lipid concentrations that lead to virtually complete membrane binding in the titrant solutions while allowing for desorption of >80% of the bound peptide after the first injection into buffer. At the same time, the peptide concentration has to be chosen so as to afford a reliably measurable fluorescence signal over the entire concentration range.

2. Materials 2.1. Stock Solutions

1. Lyophilized peptide (~100 nmol) or peptide stock solution (~1 mL, 100 µM). Peptide purity should be >95% by analytical high-performance liquid chromatography (HPLC). Most lyophilized peptides can be stored at −20°C for several months. The stability of a peptide solution depends on the peptide being used and has to be checked before starting any experiment. 2. Phospholipid (~45 mg, assuming an effective molar mass of 780 g/mol for a “typical” phospholipid (12)) dissolved in organic solvent (typically, a chloroform/methanol mixture) or in powder form. Lipids can be purchased from several suppliers, for instance, Avanti Polar Lipids (Alabaster, USA), Biosynth (Staad, Switzerland), or Genzyme Pharmaceuticals (Liestal, Switzerland). A zwitterionic phospholipid frequently used to mimic uncharged outer leaflets of mammalian plasma membranes is 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), whereas negatively charged bacterial membranes or inner leaflets of mammalian membranes can be imitated by using a mixture of POPC and either of the anionic phospholipids 1-palmitoyl-2-oleoyl-sn-glycero-3[phosphorac-(1-glycerol)] (POPG) or 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoserine (POPS) at a molar ratio of 3:1. Lipid purity should be >99%. Phospholipid dissolved in organic solvent or in powder form can be stored at −20°C for several months when overlaid with argon or nitrogen to prevent oxidation and hydrolysis. Phospholipid powder is potentially harmful if ingested or inhaled; avoid contact with eyes, skin, or clothing. 3. Buffer of choice (~20 mL; see Note 1). We routinely use 10 mM phosphate (NaH2PO4 and Na2HPO4) buffer containing 154 mM NaF and adjusted to pH 7.4. Buffer should be

Membrane Translocation Assayed by Fluorescence Spectroscopy

275

sterile-filtered and stored at 4°C to prevent microbial contamination. 4. Plastic vials (~2 mL, ~20×) and pipette tips; if necessary siliconized (see Note 2). 5. Round-bottomed glass vials (~5 mL, 2×) with tightly sealed lids. 6. If lipids are purchased in powder form, chloroform (~3 mL) of highest available purity (see Note 3). Store in a cool place away from light and ignition sources. Chloroform is dangerous to health; do not ingest, inhale, or get into contact with eyes, skin, or clothing. Always work under a hood and wear laboratory coat, goggles, and appropriate gloves that do not allow fast permeation of chloroform (made from, e.g., polyvinyl alcohol (PVA) or Viton). Waste material should be handled according to your institution’s waste disposal guidelines. 7. Disposable glass Pasteur pipette. 8. Nitrogen or argon gas source. 9. Exsiccator connected to high vacuum (~10−2–10−3 mbar). 2.2. Vesicle Preparation

1. Titanium-tip ultrasonicator (e.g., Labsonic L from B. Braun Biotech, Melsungen, Germany; Sonopuls HD 2070 from Bandelin electronic, Berlin, Germany) with clamp, ring stand, and 200-mL beaker. 2. Eppendorf or similar tubes (2 mL, 5×) suitable for centrifugation.

2.3. Fluorescence Spectroscopy

1. Fluorescence spectrometer (fluorometer). Light sources such as xenon lamps emit highly intense visible and ultraviolet (UV) radiation, which can seriously damage eyes. Never look directly into the light source and always wear safety glasses. 2. Quartz cuvette (1 × 1 cm; Hellma, Mühlheim, Germany). 3. Magnetic stir bar (Tc, but with a slower cooling rate than 104 K/s (Fig. 2B in (8)). At a similar cooling rate but quenched at temperature below Tc , the liposomal fracture planes display two kinds of ridges, termed l/2 (zigzag) and l (wave-like) ridges (Fig. 2C in (8)). The l ridges indicate molecular ordering characteristics of the Pb’ phase an intermediate bilayer phase between gel (Lb’) and liquid-crystalline state (La). The l/2 ridges are characteristic for the Lb’ gel phase (24). In similar proteoliposomes, made from 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) and the intrinsic protein bacteriorhodopsin, the protein particles decorate the lipid ridges or are localized in structural defects of these lipid ridges (Fig. 4a–f in (14)).

338

Papahadjopoulos-Sternberg

3.2. Fracturing and Shadowing of the Fracture Faces

1. Setting up freeze-fracture device Jeol JED-9000 Freeze-EtchingEquipment by servicing the gun chamber with fresh Pt/C and C electrodes and fresh filaments, if needed, and positioning the electrodes neatly in the upper center of the filament. 2. Add liquid nitrogen to the diffusion pump Dewar when a total pressure of 99% and were stored at −20°C. 2. Cholesterol (Chol) (also stored at −20°C), human holotransferrin (Tf) (stored at 4°C) and HEPES (stored at room temperature) (Sigma-Aldrich,Dublin, Ireland). 3. N-(3-Dimethylaminopropyl)-N¢-ethylcarbodiimide hydrochloride (EDC) (Fluka, Dublin, Irleand); the EDC working solution (2 mg EDC per 1 µmol of lipid in PBS needs to be prepared freshly before use. 4. N-Hydroxysulfosuccinimide (Sulfo-NHS) (Pierce, Rockford, IL). 5. BCA protein assay kit (Pierce, Rockford, IL).

2.3. Atomic Force Microscopy

1. As substrate for AFM-sample preparation silicon wafers with a natural silicon oxide surface layer (thickness 3.8 nm) and a surface roughness of 0.3 nm were used (Wacker Chemie AG, Munich, Germany). The wafers were split into small pieces of about 1 × 1 cm. The pieces were cleaned in a bath sonicator for 20 min in chloroform:methanol (2:1 v/v), then dried in a dry air stream. For better handling and to avoid artifacts (e.g., fingerprints), one side of the wafers was labeled as “bottom” surface. Store in a dust free atmosphere. 2. For AFM imaging in ambient conditions, NSC 16/Cr–Au intermittent contact cantilevers from Anfatec Instruments AG (Oelsnitz, Germany) with a nominal force constant of 45 N/m, a resonance frequency of 170 kHz and a length of 230 µm were used. The tip of the radius was smaller than 10 nm (manufacturer’s datasheet). 3. For AFM imaging under water, CSC 21 AlBS cantilevers from Anfatec Instruments AG were used. Cantilevers have a nominal force constant of 2 N/m and a resonance frequency

354

Sitterberg et al.

of 105 kHz (in air). The tip of the radius was smaller than 10 nm (manufacturer’s datasheet).

3. Methods 3.1. Preparation of Proteoliposomes

3.2. Preparation of P-selectin-Modified Proteoliposomes

To study membrane proteins in a proteoliposomal bilayer, several factors have to be taken into consideration optimizing the protein reconstitution: the homogeneity of the protein, the molar lipid-toprotein ratio, the final orientation of the protein and the size and permeability of the liposomal membrane. For most proteins, the direction of the protein within the lipid bilayer is of great importance and has to be taken into account when optimizing the reconstitution method. There are several methods for the insertion of membrane proteins into artificial membranes using mechanical treatment, freeze-thawing, organic solvents, or detergents (16). Since, due to their hydrophobic character, solubilization of membrane proteins generally requires the use of detergents, it is very convenient to use an insertion method which also involves the use of detergents to prepare proteoliposomes. P-selectin has been chosen as an example for this chapter, but other transmembrane proteins can be incorporated into liposomal bilayers following the same procedures. Other types of proteoliposomes have the protein attached to linker lipids that are part of the phospholipids mixture comprising the membrane. The proteins can be conjugated either directly to the lipid, i.e., residing at the membrane, or be coupled to the end of spacers (e.g., polyethylene glycol chains). Proteoliposomes of the latter two types are commonly used in drug targeting and to investigate ligand-receptor interactions (17). We have chosen a modification with the serum protein, transferrin, as an example for this chapter. 1. Disperse the purified P-selectin in TBS containing octyl-betaD-gluco-pyranoside to achieve a final protein concentration of 50 µg/ml (see Fig. 1). 2. Add 1 ml of this micellar dispersion to the flask containing the lipid film. 3. Ultrasonicate the flask for 5 min at 25°C. 4. Incubate this dispersion inside a dialysis membrane against 5 L distilled water to remove the octylglycoside. Change the water every 12 h for a total duration of 2 days.

3.3. Preparation of Transferrin-Modified Proteoliposomes

1. Liposomes are prepared by the thin film hydration method. The lipid composition is DSPC, Chol, DSPE-PEG and DSPE-PEG-COOH at the molar ratio of 1.85:1:0.132:0.018 (see Fig. 1).

AFM for the Characterization of Proteoliposomes

355

Fig. 1. Scheme of the investigated proteoliposomes

2. Prior to weighing, lipids should be brought to room temperature and weighed using an analytical balance, e.g., 9.7 mg DSPC, 2.6 mg Chol. 2.5 mg DSPE-PEG and 0.34 mg DSPE-PEG-COOH for preparing to 2 ml of liposomes at a phospholipid concentration of 10 µmol/ml. 3. All lipids are then dissolved in an organic solvent, e.g., chloroform. Once the lipid has been fully dissolved, the solvent must be removed completely creating a lipid film using a rotary evaporator (see Note 2). 4. The dried lipid film is then hydrated with the desired buffer to form multilamellar vesicles. The hydration of the lipid film has to be performed at a temperature at approximately 5°C above the temperature of the highest melting lipid component at least for 30 min. In this case hydration was performed at 60°C. 5. After the hydration step, and in order to produce a homogenous liposomal suspension, liposomes are submitted to extrusion using a Lipex extruder (Northern Lipids Vancouver, BC) at moderate pressures of 200–500 psi, by sequentially filtering the suspension through polycarbonate membranes until an average vesicle size of 0.1 µm was achieved. This step has to be also performed at 60°C. 6. The size of the liposomes in suspension can be determined by dynamic light scattering, e.g., in a ZetaSizer, Nano Series (Malvern Instruments, Malvern, UK). As a measure of particle size distribution of the dispersion, the equipment reports the polydispersity index ranging from 0 for an entirely monodisperse sample up to 1.0 for a polydisperse suspension. 7. The coupling of transferrin to liposomes is performed through an amide bound between the carboxylic groups of

356

Sitterberg et al.

the linker lipid, DSPE-PEG-COOH, and free amino groups of the transferrin in the presence of a water-soluble carbodiimide, EDC. The reaction of carboxyl group in the presence of EDC forms an amine-intermediate, which after reacting with Sulfo-NHS will form a semi-stable amine reactive NHS ester that afterwards will react with amino groups of Tf leading to a stable amide bond (17). 8. After the phospholipid quantification of the liposomal suspension, 1 ml of the buffer selected for liposomes preparation is added to 1 ml of liposomes at a lipid concentration of 10 µmol/ ml, 180 µl of Sulpho-NHS (0.25 M) and 180 µl of EDC (0.25 M) freshly prepared using the selected buffer (see Note 3). This mixture is allowed to incubate for 10 min at room temperature. After this incubation period, 125 µg of Tf/µmol of lipid are added and gently agitated overnight (see Note 4). 9. The unbound protein is separated from the liposomes after dilution of the suspension in the work buffer, followed by an ultracentrifugation step at 250,000×g for 3 h. The pelleted liposomes are re-suspended and their physicochemical properties are characterized in terms of mean size, surface charge, phospholipid concentration and Tf binding efficiency. 3.4. Atomic Force Microscopy

The methods are described for a Digital Nanoscope IV Bioscope (Veeco Instruments, Santa Barbara, CA) but can easily be adapted to other makes and models. The atomic force microscope was vibration- and acoustically damped (for detailed description see Oberle et al. (18). All measurements were performed in tapping™ mode (see Note 5). The applied force to the sample surface was adjusted to a minimum to avoid damage to the sample (see Note 6). The specimen was investigated and scanned under a constant force. The scan speed was proportional to the scan size and the scan frequency was between 0.5 and 1.5 Hz. Images were obtained by displaying the amplitude signal of the cantilever in the trace direction, and the height signal in the retrace direction, both signals being simultaneously recorded. The results were visualized either in height (the measured height of the sample in a resolution of 0.3 nm) or in amplitude mode (the damping of the cantilever oscillation due to tip-sample-interactions) (see Note 7).

3.4.1. AFM of P-SelectinModified Proteoliposomes

Various methods are available for the sample preparation (see Note 8). Proteoliposomes containing P-selectin are very fragile and have a tendency to spread across interfaces (sample support/water interface). Furthermore, these proteoliposomes show a loss of stability in diluted media. 1. The liposomal dispersion is directly transferred to a glass chamber with a piece of silicon wafer as a sample support immobilized at the bottom. During 1 h of incubation, liposomes are

AFM for the Characterization of Proteoliposomes

357

Fig. 2. Visualisation of P-selectin modified liposomes. The AFM measurements were performed under water. (a) Height mode overview. (b) higher resolution. Visualisation of the surface of a single P-selectin modified liposome. The small P-selectin molecules are visible as bright dots. The protein raises the lipid bilayer about 0.6 nm. The diameter of the protein is about 6.5 nm. (c) surface of a plain liposome. The surface is smooth (d) Section analysis of the surface in b

allowed to adsorb to the support. AFM imaging is conducted under water in the liposomal dispersion (see Fig. 2). 2. The cantilever is brought into contact with the sample surface in contact mode. 3. The resonance frequency of the cantilever is determined in the liposomal dispersion near the silicon support (see Note 9). The drive frequency of the piezo crystal is set slightly lower than the resonance frequency (see Note 10). 4. The AFM cantilever is approached to the sample surface until it contacts the surface. 5. Scanning of the sample is performed. The scanning speed is set proportional to the scan size to achieve constant speed of the AFM tip across the surface. The ideal speed for scanning, highly depends on the height of the sample and found to be 1 µm/s or slower. 6. The “speed” of the feedback loop (the so called integral gain or I-gain), which keeps the cantilever sensor system at a constant height relative to the sample, is adjusted. It should be as high as possible without visible “noise” in the AFM picture. This is monitored best using a topography oscilloscope visualizing each line that is scanned.

358

Sitterberg et al.

7. When the ideal settings for the sample are found, the imaging is restarted to avoid artifacts of sub-ideal settings. 8. After an overview of the sample is retrieved by scanning of at least three independent spots on the sample of an area of 10 × 10 µm², single proteoliposomes are investigated at higher resolution of 1.5 × 1.5 µm² up to 100 × 100 nm to visualize the liposome surface (Perform steps 3–6.). 3.4.2. AFM of TransferrinModified Liposomes

A very convenient procedure for the preparation of liposomes is the self-assembly technique (19). Small pieces of silicon wafers (about 1 × 1cm) as sample support material are immersed in the liposomal dispersion for 20 min at room temperature. During this time, the liposomes adsorb on the support surface under equilibrium conditions. After 20 min, the silicon supports are removed from the dispersions. The samples are dried in a dry air flow at room temperature and investigated within 2 h. 1. The resonance frequency of the AFM cantilever is determined by the manufacturer’s software. The driving frequency is set slightly smaller than resonance frequency (see Note 9). As preliminary settings, the standard settings of the AFM software can be used (see Fig. 3).

Fig. 3. Visualisation of Tf modified liposomes. The AFM measurements were performed in air (60% r.h.). (a) Height mode overview. The liposomes are visible as small bright colored objects. (b), higher resolution, but no surface features are visible in this height mode image. (c) 3-D image of a single liposome, some Tf molecule could be visualized at the surface (marked with arrows, height image), (d, e) Amplitude mode images. The Tf molecules are clearly visible. (f) Section analysis of the surface. The diameter of the Tf molecules could be measured and is between 3.0 and 3.5 nm

AFM for the Characterization of Proteoliposomes

359

2. The AFM cantilever is approached to the sample surface until it contacts the surface. 3. Then, adjusting and scanning is performed as in steps 3–8 in Subheading 3.4.1. 4. For imaging of Transferrin - liposomes, a scan speed of 1.5 µm/s is found to be ideal.

4. Notes 1. High performance liquid chromatography (HPLC) grade solvents should be used to avoid contamination by impurities. 2. Special care should be taken during evaporation of the solvent in order to prevent certain lipid components from crystallizing, and thus leading to a less homogenous lipid mixture. 3. The pH and composition of the re-suspending buffer is of utmost importance. When preparing Tf-modified proteoliposomes, PBS (pH 5.9 and 7.4) and 10 mM sodium citrate containing 140 mM NaCl (pH 5.9) showed the best results at coupling efficiencies of 90–98%. For comparison, 10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) containing 140 mM NaCl (pH 7.4) resulted in a coupling efficiency of only ~56%. 4. Two different incubation temperatures are generally used for the Tf coupling reaction, room temperature and 4°C. In a side-by-side comparison, we could find slightly better results when performing this reaction at 4°C. 5. Tapping mode™ or intermittent contact mode is a key advance in AFM of soft, adhesive or fragile samples, i.e., biological materials. In this mode, the cantilever is oscillating near its resonance frequency and touches the sample only at its lower amplitude. During sample scanning, this contact of the tip with the sample causes a reduction in the oscillation amplitude. This change in oscillation amplitude is used to identify and measure surface features. Unlike in contact mode, where the cantilever is scanning across the surface and is in touch with the surface all the time, tapping mode minimizes the lateral forces applied to the sample. This allows imaging of soft material with high resolution without damage or alteration of the sample. This technique can be used for imaging in air and also in liquid environments. 6. Although lateral forces are minimized in the tapping mode, a force perpendicular to the sample surface is still applied to the sample. This is because at each of the lower amplitude

360

Sitterberg et al.

of cantilever oscillation, the tip contacts the sample. As the dampening of the lower amplitude due to sample tip interaction is used to determine sample height, the value of this dampening has to be high enough to clearly identify tip sample interactions. On the other hand, a higher dampening is proportional to a higher force applied onto the sample, which may lead to structural alterations or destruction of the sample. Therefore, the force should be adjusted to a point where both the force is high enough for high resolution scanning and the force is still as low as possible to avoid sample destruction. 7. AFM allows visualizing of a sample in different ways: The one which is easiest to understand is the “height mode” where the color scale of the AFM image gives information on the actual height of the sample at the specific spot, like a landscape map. Another mode, which allows visualizing of smaller surface features, is the “amplitude mode”, where the damping of the cantilever oscillation is indicated. This gives information about the steepness of the sample at the specific spot. Pictures visualizing amplitude mode always look as if illuminated from either left or right and show surface details more prominently. 8. Sample preparation is crucial for reliable and reproducible AFM images. Especially when working with labile structures like liposomes, several factors have to be considered. One of them is the tendency of lipids to form self assembled lipid bilayers at interfaces. This spreading may be avoided by adjustment of the surface chemistry of the support (e.g., chemical modification via silanization or by working in a liquid environment. As rehydration of samples may cause artifacts due to structural changes of the sample, keeping the sample in the liquid all time until imaging in fluids, is highly recommended. Special fluid chambers are commercially available for most atomic force microscopes. 9. The oscillation of the cantilever is not only dependent on the cantilever properties (its dimensions and material), but also on the medium it is surrounded by. In water, a damping of the oscillations due to the viscosity of water causes a shift of the resonance to lower frequencies. This is further affected by a higher damping near a surface. To determine the driving frequency which may be used for scanning under water, it is best to determine it while in the same medium and distance (e.g., 300 nm) from the surface as in scanning. 10. The oscillation of the cantilever is damped, when it is scanning across the sample. This causes the resonance frequency to decrease slightly. This phenomenon is taken into account by setting the driving frequency of the piezo crystal to a frequency lower than the free resonance frequency of the cantilever.

AFM for the Characterization of Proteoliposomes

361

Acknowledgments This work was supported in parts by DFG Forschergruppe 627 Nanohale and JPK Instruments Berlin (Germany) (UB and JS). This work was supported in part by grants from Enterprise Ireland under the National Development Plan co-funded by EU Structural Funds and Science Foundation Ireland (CE and MMG). References 1. Lukas K, Zhifeng S (1998) The application of AFM to biomembranes Biomembrane Structures 2. Schneider S, Lärmer J, Henderson R, Oberleithner H (1998) Molecular weights of individual proteins correlate with molecular volumes measured by atomic force microscopy. Pflügers Arch 435:362–7 3. Rasch P, Wiedemann U, Wienberg J, Heckl W (1993) Analysis of banded human chromosomes and in situ hybridization patterns by scanning force microscopy. Proc Natl Acad Sci USA 90:2509 4. Hansma HG, Kasuya K, Oroudjev E (2004) Atomic force microscopy imaging and pulling of nucleic acids. Curr Opin Struct Biol 14:380–5 5. Kamruzzahan A, Kienberger F, Stroh C et al (2004) Imaging morphological details and pathological differences of red blood cells using tapping-mode. AFM Biol Chem 385:955–60 6. Engel A, Müller D (2000) Observing single biomolecules at work with the atomic force microscope. Nat Struct Biol 7:715–8 7. Müller DJ, Engel A (1999) Voltage and pHinduced channel closure of porin OmpF visualized by atomic force microscopy. J Mol Biol 285:1347–51 8. Zeidel M, Nielsen S, Smith B, Ambudkar S, Maunsbach A, Agre P (1994) Ultrastructure, pharmacologic inhibition, and transport selectivity of aquaporin channel-forming integral protein in proteoliposomes. Biochemistry 33:1606–15 9. Awasthi S, Singhal S, Pikula S et al (1998) ATP-Dependent human erythrocyte glutathioneconjugate transporter. II. Functional reconsti tution of transport activity. Biochemistry 37:5239–48 10. Dass C (2008) Drug delivery in cancer using liposomes. Methods Mol Biol 437:177–82

11. Lian T, Ho R (2001) Trends and developments in liposome drug delivery systems Journal of Pharmaceutical. Sciences 90:667–80 12. Opinion E (2008) Antibody-targeted liposomes in cancer therapy and imaging. Expert Opin Drug Deliv 5:189–204 13. Bendas G, Krause A, Bakowsky U, Vogel J, Rothe U (1999) Targetability of novel immunoliposomes prepared by a new antibody conjugation technique. Int J Pharm 181:79–93 14. Pignataro B, Steinem C, Galla H, Fuchs H, Janshoff A (2000) Specific adhesion of vesicles monitored by scanning force microscopy and quartz crystal microbalance. Biophys J 78:487–98 15. Moore K (1991) GMP-140 binds to a glycoprotein receptor on human neutrophils: evidence for a lectin-like interaction. J Cell Biol 112:491–9 16. Rigaud J (2002) Membrane proteins: functional and structural studies using reconstituted proteoliposomes and 2-D crystals Brazilian. J Med Biol Res 35:753–66 17. Anabousi S, Laue M, Lehr C, Bakowsky U, Ehrhardt C (2005) Assessing transferrin modification of liposomes by atomic force microscopy and transmission electron microscopy. Eur J Pharm Biopharm 60:295–303 18. Oberle V, Bakowsky U, Zuhorn I, Hoekstra D (2000) Lipoplex formation under equilibrium conditions reveals a three-step mechanism. Biophys J 79:1447–54 19. Kneuer C, Ehrhardt C, Bakowsky H et al (2006) The influence of physicochemical parameters on the efficacy of non-viral DNA transfection complexes: a comparative study. Journal of Nanoscience and Nanotechnology 6:2776–82

Chapter 24 Method of Simultaneous Analysis of Liposome Components Using HPTLC/FID Sophia Hatziantoniou and Costas Demetzos Abstract Liposomes are composed of different kind of lipids or lipophilic substances and are used as carriers of bioactive molecules. The characterization of the prepared liposomes consists of the calculation of the drug to lipid molar ratio by measuring the lipids and the encapsulated molecule. The present work describes an analytical methodology on simultaneous determination of all the lipid ingredients of the liposome formulation, using Thin Layer Chromatography coupled with a Flame Ionization Detector (TLC/FID), using the least possible sample quantity. The method consists of a chromatographic separation of the liposomal ingredients on silica gel scintillated on quartz rods and subsequent detection of the ingredients by scanning the rods by a hydrogen flame. The produced ions are detected by a Flame Ionization Detector and the signal is converted to a chromatogram. This method may be applied on every step of the liposome preparation for examining the quality of the raw materials, tracking possible errors of the preparation procedure and finally analyzing the content of the final liposomal composition. Key words: Liposome, Lipid analysis, Drug/lipid ratio, Bioactive molecule, HPTLC/FID

1. Introduction Liposome technology is widely applied to both pharmaceutical and cosmetic formulations. Liposomes are used as a suitable vehicle for bioactive molecules in order to overcome their poor water solubility or their possible non desired side effects on normal cells (1, 2). The main component of the lipid bilayers of the liposomes are acyl-phosphatidylcholines of natural or synthetic origin. Other lipid substances such as cholesterol or charged

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_24, © Humana Press, a part of Springer Science+Business Media, LLC 2010

363

364

Hatziantoniou and Demetzos

lipids may be used in order to give desired performance to the liposomal formulation (3). The characterization of liposomes requires the calculation of the drug to lipid ratio, measuring the amount of the encapsulated active ingredient, as well as all the used components (4). In the present work we describe a method of simultaneous determination of all the liposomal ingredients using Thin Layer Chromatography coupled with a Flame Ionization Detector (TLC/FID) (2, 5, 6). The sample ingredients are separated on re-usable silica coated quartz rods using the classical Thin Layer Chromatography techniques and subsequently analyzed by passing through a Flame Ionization Detector. The flame of the burner is generated by an external Hydrogen supply and atmospheric oxygen supplied by the air pump that is incorporated into the instrument. The procedure requires only one measurement per sample and it can be applied even in very small or much diluted samples. The speed of analysis and the ability to assess many samples at the same time makes this method suitable for routine assessment. This method may find application on the assessment of liposomal formulations, on the quality control of the raw materials and preparation procedure.

2. Materials 2.1. Liposomal Sample Lyophilization

1. Screwed cup glass vials of 5 ml capacity

2.2. Sample Preparation

1. Pasteur pipettes

2.3. Sample Spotting

1. Chromarods-III

2. Parafilm

2. Cotton wool

2. Glass syringe of 1ml capacity 2.4. Sample Development

1. Development tank 2. The following solvent mixture of analytical grade have been used for liposome component separation: Chloroform/methanol/d-Water 45:25:5 (v/v) (7, 8) (see Note 1). Iatroscan newMK-5 (Iatron Laboratories, INC. Tokyo, Japan). 1. Lipids and bioactive components of analytical grade for preparation of standard solutions at concentrations similar to that of the samples.

Analysis of Liposome Components by HPTLC/FID

365

3. Methods 3.1. Liposomal Sample Lyophilization

1. Prepare 500 ml aliquots of each liposome sample in screwed cup glass vials of 5 ml capacity. 2. Cover the vials with parafilm and pierce it with a needle, to create thin holes. 3. Freeze the samples (see Note 2) and place them in the freezedrier overnight (9). 4. Place the cups on the vials and store at 4°C until use.

3.2. Sample Preparation

1. Dilute the freeze-dried residue in chloroform or other suitable solvent mixture. 2. Filter the samples through cotton filters in order to remove the sugars used as cryoprotectants, or the salts of the buffers used (1) (see Note 3). 3. Wash the residue on the cotton filter with 1 ml chloroform twice and add the filtrates. 4. Remove the chloroform under nitrogen stream 5. Weigh the residue and add a proper volume of chloroform (see Note 4) to a final concentration of about 20 mg/ml.

3.3. Sample Spotting

1. Run a blank scan to ensure that the Chromarods-III are clean. 2. Using a glass syringe of 1 ml capacity spot 1 ml of the sample solutions on the zero point of the Chromarods-III (see Note 5). Apply the sample on two to three Chromarods-III to calculate the mean area of three measurements. 3. Use the last Chromarod-III for corresponding standards at concentrations similar to the samples.

3.4. Sample Development

1. Line the rear side of the development tank with a piece of filter paper. 2. Pour the mobile phase (60–70 ml of solvent mixture) in the development tank and cover it with its glass lid (see Note 1). 3. Gently move the tank to wet the filter paper lining and allow the tank to saturate with solvent vapor (see Note 6). 4. Place the rod holder in the development tank and leave them until the solvent mixture front reaches the desired height (see Note 7). 5. Remove the rod holder from the development tank and allow the excessive solvent to drain. 6. Hold the rod holder under a stream of hot air for 1 min to completely remove the solvent residue (see Note 8).

366

Hatziantoniou and Demetzos

3.5. Sample Scanning

1. Place the rod holder in the scanning frame 2. Scan the Chromarods-III at the following conditions: scanning speed 30 s/scan, airflow 2000 L/h, and hydrogen flow 160 ml/min.

3.6. Calibration Curve

1. Prepare standard solutions for each component of the sample having twice the concentration of the expected in the sample. 2. Spot gradually increasing volumes of the standard solution from 0.2–1 ml using two rods per volume. 3. Place the rod holder into the development chamber and allow the development with the same conditions to the sample (see subheading 3.4). 4. Scan the rods and obtain the chromatogram. 5. Calculate the calibration curve plotting the area under the peak against ingredient’s quantity.

3.7. Qualitative and Quantitative Analysis

1. After obtaining the chromatogram of the sample, identify each peak, comparing the retention time to that of the corresponding standard. An example of the chromatogram is shown in Fig. 1. 2. Calculate the content of each component using the peak area (Area Under Curve, AUC) of the unknown and the corresponding calibration curve.

Fig. 1. Chromatogram of a liposomal formulation containing phosphatidylcholine [1] and bioactive molecule [2]. The separation of the liposomal ingredients was achieved using the multiple development technique. The two subsequent mobile phases were: (a) CHCl3/CH3OH/d-H2O 45:25:5 (v/v) up to 5 cm (50% of the Chromarod-III), (b) Hexane/Diethyl ether 40:60 (v/v) up to 10 cm (100% of the Chromarod-III). After the separation of the ingredients and the elimination of the solvent vapor the Chromarods-III were scanned at the following conditions: scanning speed 30 s/scan, airflow 2000 L/h, and hydrogen flow 160 ml/min

Analysis of Liposome Components by HPTLC/FID

367

4. Notes 1. For enhancement of the component separation several solvent mixtures should be tried. The best technique suggested is the multiple development method in which two successive mobile phases are used: At first the samples are allowed to develop until 5 cm from zero point (50% of the rods) in Chloroform/ Methanol/d-Water 45:25:5 (v/v) in order to separate the phospholipids (2, 5, 6). Subsequently, the solvent is removed under hot air stream and the samples are placed in the second development tank containing either Hexane/Diethyl ether/ glacial acetic acid 80:20:2 (v/v) or Hexane / Diethyl ether 40:60 (v/v) until 10 cm (100% of the rods) for the separation of non-polar components from the phospholipids. 2. Place the samples in deep freezer for 20 min or immerse them in dry ice/iso-butanol bath until frozen. 3. Place cotton wool in Pasteur pipettes creating cotton filters of 1 cm height. Wash them with 1 ml chloroform twice. 4. The appropriate solvent selection is based on the sufficient solubility of the sample components. The solvent with lowest boiling point and polarity is preferred in order to produce narrow spots. 5. Hold the syringe in such a way to avoid scratching of the silica surface, allowing the sample drop to touch the surface of the Chromarod-III. The sample volume should be placed in small aliquots to produce a narrow spot. The sample spot should be as narrow as possible (less than 3 mm) to ensure good separation performance. 6. Wet the filter paper with the solvent immediately before starting the development in order to ensure complete solvent vapor saturation. 7. Do not allow the solvent front to exceed 100 cm from zero point because some separated components may be out of the scanning area. 8. If the solvent is not completely removed noise peaks will appear on the baseline signal and the results will not be reproducible.

References 1. Hatziantoniou S, Dimas K, Georgopoulos A, Sotiriadou N, Demetzos C (2006) Cytotoxic and antitumor activity of liposome-incorporated sclareol against cancer cell lines and human

colon cancer xenografts. Pharmacol Res 53(1): 80–7 2. Goniotaki M, Hatziantoniou S, Dimas K, Wagner M, Demetzos C (2004) Encapsulation

368

Hatziantoniou and Demetzos

of naturally occurring flavonoids into liposomes: physicochemical properties and biological activity against human cancer cell lines. J Pharm Pharmacol Oct 56(10):1217–24 3. Gabizon A, Papahadjopoulos D (1988) Liposomes formulations with prolonged circulation time in blood and enhanced uptake by tumors. Proc Natl Acad Sci USA 85:6949–6953 4. New RRC (1992) Characterization of liposomes. In: New RRC (ed) Liposomes: a practical approach. Oxford University Press, Oxford, pp 105–108 5. Hatziantoniou S, Demetzos C (2006) Qualitative and quantitative one step analysis of lipids and encapsulated bioactive molecules in liposome preparations by HPTLC/FID (Iatroscan). Journal of Liposome Research 16(4):321–330

6. Kaourma E, Hatziantoniou S, Georgopoulos A, Kolocouris A, Demetzos C (2005) Development of simple thiol-reactive liposome formulations, one-step analysis and physicochemical characterization. J Pharm Pharmacol 57(4):527–31 7. Henderson JR, Tocher DR (1992) Thin-layer chromatography. In: Hamilton RJ, Hamilton S (eds) Lipid Analysis: a practical approach. Oxford University Press, Oxford, pp 100–108 8. De Schriijver R, Vermeleulen D (1991) Separation and quantitation of phospholipids in animal tissues Iatroscan TLC/FID. Lipids 26(1):74–76 9. Madden TD, Boman N (1999) Lyophilization of liposomes. In: Janoff S (ed) Liposomes Rational Design. Marcel Dekker, Inc, N. York, pp 261–282

Chapter 25 Viscometric Analysis of DNA-Lipid Complexes Sadao Hirota and Nejat Düzgünes¸ Abstract DNA-cationic lipid complexes, “lipoplexes”, are used as gene carriers for molecular biology and gene therapy applications. Colloidal properties of lipoplexes can be determined by viscometric analysis. (1) The shape parameter of lipoplexes can be one of the factors affecting transfection efficiency; (2) the volume fraction of free liposomes remaining after lipoplex formation can be used as an index of purity of the lipoplex product; (3) the shear dependence of the viscosity of a diluted lipoplex suspension can be used as a macroscopic shape factor: (4) the attraction force parameter between particles can be a colloidal stability factor. These properties should be characterized and specified for process control of lipoplex production and quality control of lipoplex products. We describe an automated mini-capillary viscometer for a sample volume of 0.5 ml, and its application to the characterizations of lipoplexes. We show a procedure for viscosity measurements and provide a calculation using complexes of plant DNA-distearyldimethylammonium chloride (DDAC) at a charged ratio of 1:4 (−/+), in which the amount of DNA is less than 250 µg. The prolate/ellipsoidal axial ratio, a/b, was found to be 70. Determination of the shape parameter with a/b is found to be better than that with other shape parameters, e.g., a of the Sakurada equation, because fractionation of the particle size is not necessary. By the proposed method, colloidal parameters of lipoplexes and bioactive polymer complexes are characterized quantitatively. Key words: Shape parameter, Ellipsoid, Liposomes, Lipoplexes, Viscosity, Capillary viscometer, Quality control in large scale production

Symbols a b c L k¢ k k0k P r R

Longer semi-diameter of ellipsoid (cm) Shorter semi-diameter of ellipsoid (cm) Concentration in molality (mol/kg) Length of capillary (cm) Huggins coefficient Attracting force parameter between particles at zero shear Pressure difference between both ends of capillary (dyn/cm2) Distance from axis of capillary (cm) Radius of capillary (cm)

V. Weissig (ed.), Liposomes: Methods and Protocols, Volume 2: Biological Membrane Models, Methods in Molecular Biology, vol. 606, DOI 10.1007/978-1-60761-447-0_25, © Humana Press, a part of Springer Science+Business Media, LLC 2010

369

370

Re S SR S0 t u U a j jav ji jnet jc mred [m] h h0 hrel hsp hred [h]

Hirota and Düzgünes¸

Reynolds number Shear stress (dyn/cm2) Shear stress at inner wall of capillary (dyn/cm2) Yield stress (dyn/cm2) Time (s) Linear velocity (cm/s) at distance r from axis of capillary (cm/s) Volume velocity, V/t, (ml/s) Shape parameter of the Sakurada equation Volume fraction, volume of particles / volume of suspension Average volume fraction Volume fraction of inner aqueous phase Net volume fraction without inner aqueous phase Volume fraction of cationic liposomes Reduced viscosity, mred = hsp/c (L/g or L/mol) Intrinsic viscosity, reduced viscosity at infinite dilution (L/g or L/mol) Viscosity of sample liquid (poise) Viscosity of solvent or suspending medium (poise) Relative viscosity, hrel =h/ho Specific viscosity, hsp =(hrel–1) Non-dimensional reduced viscosity, hred = hsp/j Non-dimensional intrinsic viscosity, [h] = lim(j→0)hred

1. Introduction A capillary viscometer can provide an accurate measure of viscosity, and could be an excellent tool for quality control of liposomes and lipoplexes during large-scale production for clinical or other uses. However, it has seldom been used for this purpose. The reasons are that it requires a large amount of sample and that it necessitates long strenuous attention of a researcher at the flow time measurements. To address this problem, we have developed a mini-capillary viscometer for sample volumes of 0.5 ml with automated measurements (1). 1.1. Shape Parameter

Lipoplexes have different shapes (2) that contribute to the efficiency of gene delivery (3, 4). It is therefore important to characterize the shape of particular lipoplexes before preclinical or clinical studies are undertaken. The shape of lipoplex particles has been studied by electron microscopy, but this technique does not provide an average value of all the particles in the suspension. Moreover, sample preparation for electron microscopy can alter the shape of the particles. A viscometric analysis, however, gives an average value of the shape parameter of lipoplexes.

Viscometric Analysis of DNA-Lipid Complexes

371

Einstein (5) derived an equation for the viscosity of a dilute suspension of spherical, non-attracting particles: h = h0(1 + 2.5j)

(1)

where h is the viscosity of the suspension, h0 is the viscosity of the suspending medium, and j the volume fraction of the suspended particles. Note that Eq. 1 does not include the particle size, and holds irrespective of the particle size or size distribution. Simha (6) has extended the Einstein equation to ellipsoidal, non-atracting particles as h = h0 (1 + nj)

(2)

Where n is called the Simha factor, or shape parameter, which is 2.5 for spherical particles in the Einstein Eq. 1, and is invariably larger than 2.5 for ellipsoidal particles. Note that Eq. 2 also does not include particle size. It holds independent of the particle size or size distribution. Non-attracting conditions are attained when j approaches zero. From Eq. 2, n = l im(j →0)(hrel -1)/j = lim (j →1)hsp / j = lim(j →1)hred = [h]

(2.1)

where hrel = h/h0: relative viscosity, hsp = hrel−1: specific viscosity, hred = hsp/j: non-dimensional reduced viscosity, [h]: non-dimensional intrinsic viscosity Here we see that the Simha factor or shape parameter, n, is equivalent to the non-dimensional intrinsic viscosity. Hirota (7) related n to the ellipsoidal axial ratio, a/b, in the range 1 < a/b < 100, for prolate (rod shaped) ellipsoids as, n = 0.057(a/b)2 + 0.61a/b + 1.83

(3)

and for oblate (disc shaped) ellipsoids as n = 0.001(a/b)2 + 0.59a/b + 1.90

(4)

The non-dimensional intrinsic viscosity as a function of the ellipsoidal axial ratio a/b is illustrated in Fig. 1. 1.2. Free Cationic Liposomes Remaining After Complexation

Oberle et al. (2) reported that “part of the initially formed lipoplexes are/remain unstable and eventually aggregate and/or merge further into larger complexes upon prolonged incubation.” They suggested that “this fraction may actually originate from vesicles that in the early phase of preparation still display surface-bound plasmid, or at least some uncovered part of DNA and that this fraction will be excluded from involvement of cellular transfection, as such a size will preclude cellular uptake. Thus, the shape parameter determination described above should

372

Hirota and Düzgünes¸

Nondimensional intrinsic viscosity

300

Prolate 200

100 80 60 40

Oblate

20 10

20

30

40

50

60

70

Ellipsoidal axial ratio, a/b Fig. 1. Non-dimensional intrinsic viscosity as a function of the ellipsoidal axial ratio, a/b

ideally be performed on properly formed lipoplexes and not extremely large complexes. It is also important to eliminate the uncomplexed cationic liposomes or to assess their volume fraction. During lipoplex production, a process test can be performed to confirm the absence of plain cationic liposomes in the final product. With ordinary cationic lipid materials, the density of lipoplexes is in the range 1.1–1.2, whereas the density of cationic liposomes is less than 1.0. Thus, it should be possible to separate the free liposomes from lipoplexes by differential centrifugation or density gradient centrifugation. As cationic liposomes are spherical, the volume fraction of free liposomes, jc, can be determined using Eq. 1 by measuring the relative viscosity, hrel = h/h0, of the supernatant after the centrifugation (see Note 1). jc = (hrel – 1)/2.5

(1.1)

With an ideal product, jc should be zero. 1.3. Shear Dependence of the Viscosity of a Diluted Lipoplex Suspension

In a dilute uniform lipoplex suspension at about j