Membrane protein structure and function characterization : methods and protocols 978-1-4939-7151-0, 1493971514, 978-1-4939-7149-7, 205-233-247-2

Table of contents :
Front Matter ....Pages i-xii
Recombinant Overexpression of Mammalian TSPO Isoforms 1 and 2 (Lucile Senicourt, Soria Iatmanen-Harbi, Claude Hattab, Mariano Anibal Ostuni, Marie-France Giraud, Jean-Jacques Lacapere)....Pages 1-25
Functional Assembly of Soluble and Membrane Recombinant Proteins of Mammalian NADPH Oxidase Complex (Hajer Souabni, Aymen Ezzine, Tania Bizouarn, Laura Baciou)....Pages 27-43
Direct Extraction and Purification of Recombinant Membrane Proteins from Pichia pastoris Protoplasts (Lucie Hartmann, Estelle Metzger, Noémie Ottelard, Renaud Wagner)....Pages 45-56
Cell-Free Expression for the Study of Hydrophobic Proteins: The Example of Yeast ATP-Synthase Subunits (Isabelle Larrieu, James Tolchard, Corinne Sanchez, Edmond Yazo Kone, Alexandre Barras, Claire Stines-Chaumeil et al.)....Pages 57-90
Wheat Germ Cell-Free Overexpression for the Production of Membrane Proteins (Marie-Laure Fogeron, Aurélie Badillo, François Penin, Anja Böckmann)....Pages 91-108
Methyl-Specific Isotope Labeling Strategies for NMR Studies of Membrane Proteins (Vilius Kurauskas, Paul Schanda, Remy Sounier)....Pages 109-123
Labeling of Membrane Complexes for Electron Microscopy (Francesca Gubellini, Rémi Fronzes)....Pages 125-138
Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane Proteins: Case Study of the Mouse 5-HT3 Receptor (Ghérici Hassaïne, Cédric Deluz, Luigino Grasso, Romain Wyss, Ruud Hovius, Henning Stahlberg et al.)....Pages 139-168
Characterization of New Detergents and Detergent Mimetics by Scattering Techniques for Membrane Protein Crystallization (Françoise Bonneté, Patrick J. Loll)....Pages 169-193
Secondary Structure Determination by Means of ATR-FTIR Spectroscopy (Batoul Srour, Stefan Bruechert, Susana L. A. Andrade, Petra Hellwig)....Pages 195-203
Native Mass Spectrometry for the Characterization of Structure and Interactions of Membrane Proteins (Jeroen F. van Dyck, Albert Konijnenberg, Frank Sobott)....Pages 205-232
Mass Spectrometry of Mitochondrial Membrane Protein Complexes (Luc Negroni, Michel Zivy, Claire Lemaire)....Pages 233-246
Functional Studies on Membrane Proteins by Means of H/D Exchange in Infrared: Structural Changes in Na+ NQR from V. cholerae in the Presence of Lipids (Yashvin Neehaul, Sebastien Kriegel, Blanca Barquera, Petra Hellwig)....Pages 247-257
Reconstitution of Membrane Proteins in Liposomes (Alice Verchère, Isabelle Broutin, Martin Picard)....Pages 259-282
Ion Channels as Reporters of Membrane Receptor Function: Automated Analysis in Xenopus Oocytes (Michel Vivaudou, Zlatomir Todorov, Gina Catalina Reyes-Mejia, Christophe Moreau)....Pages 283-301
The CRACAM Robot: Two-Dimensional Crystallization of Membrane Protein (Philippe Rosier, Frédéric Gélébart, Nicolas Dumesnil, Gauthier Esnot, Manuela Dezi, Marc Morand et al.)....Pages 303-316
Reconstitution of Membrane Proteins into Nanodiscs for Single-Particle Electron Microscopy (Laetitia Daury, Jean-Christophe Taveau, Dimitri Salvador, Marie Glavier, Olivier Lambert)....Pages 317-327
Solid-State NMR of Membrane Protein Reconstituted in Proteoliposomes, the Case of TSPO (Lucile Senicourt, Luminita Duma, Vassilios Papadopoulos, Jean-Jacques Lacapere)....Pages 329-344
Sample Preparation for Membrane Protein Structural Studies by Solid-State NMR (Denis Lacabanne, Britta Kunert, Carole Gardiennet, Beat H. Meier, Anja Bo¨ckmann)....Pages 345-358
Simulation of Ligand Binding to Membrane Proteins (Samuel Murail)....Pages 359-381
Molecular Modeling of Transporters: From Low Resolution Cryo-Electron Microscopy Map to Conformational Exploration. The Example of TSPO (Aurore Vaitinadapoule, Catherine Etchebest)....Pages 383-416
Back Matter ....Pages 417-419

Citation preview

Methods in Molecular Biology 1635

Jean-Jacques Lacapere Editor

Membrane Protein Structure and Function Characterization Methods and Protocols

METHODS

IN

MOLECULAR BIOLOGY

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

For further volumes: http://www.springer.com/series/7651

Membrane Protein Structure and Function Characterization Methods and Protocols

Edited by

Jean-Jacques Lacapere Sorbonne Universités—UPMC Univ Paris 06 École Normale Supérieure—PSL Research University Département de Chimie, CNRS UMR 7203 LBM, Paris, France

Editor Jean-Jacques Lacapere Sorbonne Universite´s—UPMC Univ Paris 06 , E´cole Normale Supe´rieure—PSL Research University De´partement de Chimie, CNRS UMR 7203 LBM Paris, France

ISSN 1064-3745 ISSN 1940-6029 (electronic) Methods in Molecular Biology ISBN 978-1-4939-7149-7 ISBN 978-1-4939-7151-0 (eBook) DOI 10.1007/978-1-4939-7151-0 Library of Congress Control Number: 2017943682 © Springer Science+Business Media LLC 2017 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Printed on acid-free paper This Humana Press imprint is published by Springer Nature The registered company is Springer Science+Business Media LLC The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.

Preface Membrane proteins are essential components of biological processes such as ions, metabolites or water transport, signal transduction, sensing cell environment, and control of cell-cell contact. Dysfunction of these proteins induces numerous human pathologies like cystic fibrosis which originates from a genetic disorder that is linked to the most common case to the deletion of one amino acid. Many other disorders like cancers and inflammatory diseases occur through cell-cell communications and involve membrane proteins. Although many studies have been performed to characterize membrane proteins, there is a lack of information upon this class of proteins both in terms of structures and function. One of the bottlenecks is the low abundance of these proteins in native membranes, making difficult the characterization and the study of their function. To overcome this difficulty, overexpression in various systems has gained increasing interest over the last decades, leading, in particular, to the production of sufficient amounts for structural studies. Although the number of high resolution structures of membrane proteins has increased, it remains well below that of soluble proteins. Some membrane proteins have been easily overexpressed, but it is not a common case. Expression remains challenging especially for mammalian proteins compared to bacterial ones, the former often requiring molecular chaperones to fold correctly, and posttranslational modifications to be functional. This volume addresses different approaches to produce membrane proteins, to purify them, to verify their function, to determine their structure, and to model them in membrane. Since every membrane protein behaves mostly in a unique way, knowledge of guidelines and tricks may help to increase chances to express, purify, and characterize a peculiar membrane protein. Production of correctly folded protein remains a challenge. Moreover, getting a functional and stable protein requires optimizing membrane mimicking environments that can be detergent or artificial membranes. In some cases, the finding of the correct ligand that will stabilize the desired conformation is needed. In other cases, stabilization can be obtained using specific antibodies. This volume also presents different techniques to analyze the functional status of membrane proteins. Recent progresses in structural biology of membrane proteins are described using examples for the main approaches that are electron microscopy, X-Ray diffraction, nuclear magnetic resonance (NMR), and bioinformatic computing. The new detector used for electron microscopes has increased resolution gained from either two-dimensional crystals or single particles. X-Ray free electron laser has permit to get high resolution structures using micron-size crystals. Solid-state NMR study gives de novo atomic structure of low molecular weight membrane proteins reconstituted in liposomes, but only spectral fingerprints of larger proteins. Recent developments in computing both from computer power (hardware) and from algorithms (software) have made possible molecular dynamics of larger membrane proteins at timescales compatible with conformational changes in the millisecond range. Structure-based drug design is a reality for membrane proteins in a close-to-native environment. Paris, France

Jean-Jacques Lacapere

v

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

v ix

1 Recombinant Overexpression of Mammalian TSPO Isoforms 1 and 2 . . . . . . . . . Lucile Senicourt, Soria Iatmanen-Harbi, Claude Hattab, Mariano Anibal Ostuni, Marie-France Giraud, and Jean-Jacques Lacapere 2 Functional Assembly of Soluble and Membrane Recombinant Proteins of Mammalian NADPH Oxidase Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hajer Souabni, Aymen Ezzine, Tania Bizouarn, and Laura Baciou 3 Direct Extraction and Purification of Recombinant Membrane Proteins from Pichia pastoris Protoplasts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lucie Hartmann, Estelle Metzger, Noe´mie Ottelard, and Renaud Wagner 4 Cell-Free Expression for the Study of Hydrophobic Proteins: The Example of Yeast ATP-Synthase Subunits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isabelle Larrieu, James Tolchard, Corinne Sanchez, Edmond Yazo Kone, Alexandre Barras, Claire Stines-Chaumeil, Benoıˆt Odaert, and Marie-France Giraud 5 Wheat Germ Cell-Free Overexpression for the Production of Membrane Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Marie-Laure Fogeron, Aure´lie Badillo, Franc¸ois Penin, and Anja Bo¨ckmann 6 Methyl-Specific Isotope Labeling Strategies for NMR Studies of Membrane Proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vilius Kurauskas, Paul Schanda, and Remy Sounier 7 Labeling of Membrane Complexes for Electron Microscopy. . . . . . . . . . . . . . . . . . Francesca Gubellini and Re´mi Fronzes 8 Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane Proteins: Case Study of the Mouse 5-HT3 Receptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ghe´rici Hassaı¨ne, Ce´dric Deluz, Luigino Grasso, Romain Wyss, Ruud Hovius, Henning Stahlberg, Takashi Tomizaki, Aline Desmyter, Christophe Moreau, Lucie Peclinovska, Sonja Minniberger, Lamia Mebarki, Xiao-Dan Li, Horst Vogel, and Hugues Nury 9 Characterization of New Detergents and Detergent Mimetics by Scattering Techniques for Membrane Protein Crystallization . . . . . . . . . . . . . . Franc¸oise Bonnete´ and Patrick J. Loll 10 Secondary Structure Determination by Means of ATR-FTIR Spectroscopy . . . . Batoul Srour, Stefan Bruechert, Susana L.A. Andrade, and Petra Hellwig

1

vii

27

45

57

91

109 125

139

169 195

viii

11

12 13

14 15

16

17

18

19

20 21

Contents

Native Mass Spectrometry for the Characterization of Structure and Interactions of Membrane Proteins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jeroen F. van Dyck, Albert Konijnenberg, and Frank Sobott Mass Spectrometry of Mitochondrial Membrane Protein Complexes . . . . . . . . . . Luc Negroni, Michel Zivy, and Claire Lemaire Functional Studies on Membrane Proteins by Means of H/D Exchange in Infrared: Structural Changes in Na+ NQR from V. cholerae in the Presence of Lipids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yashvin Neehaul, Sebastien Kriegel, Blanca Barquera, and Petra Hellwig Reconstitution of Membrane Proteins in Liposomes . . . . . . . . . . . . . . . . . . . . . . . . Alice Verche`re, Isabelle Broutin, and Martin Picard Ion Channels as Reporters of Membrane Receptor Function: Automated Analysis in Xenopus Oocytes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Michel Vivaudou, Zlatomir Todorov, Gina Catalina Reyes-Mejia, and Christophe Moreau The CRACAM Robot: Two-Dimensional Crystallization of Membrane Protein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Philippe Rosier, Fre´de´ric Ge´le´bart, Nicolas Dumesnil, Gauthier Esnot, Manuela Dezi, Marc Morand, and Catherine Ve´nien-Bryan Reconstitution of Membrane Proteins into Nanodiscs for Single-Particle Electron Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Laetitia Daury, Jean-Christophe Taveau, Dimitri Salvador, Marie Glavier, and Olivier Lambert Solid-State NMR of Membrane Protein Reconstituted in Proteoliposomes, the Case of TSPO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lucile Senicourt, Luminita Duma, Vassilios Papadopoulos, and Jean-Jacques Lacapere Sample Preparation for Membrane Protein Structural Studies by Solid-State NMR. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Denis Lacabanne, Britta Kunert, Carole Gardiennet, Beat H. Meier, and Anja Bo¨ckmann Simulation of Ligand Binding to Membrane Proteins . . . . . . . . . . . . . . . . . . . . . . . Samuel Murail Molecular Modeling of Transporters: From Low Resolution Cryo-Electron Microscopy Map to Conformational Exploration. The Example of TSPO. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aurore Vaitinadapoule and Catherine Etchebest

205 233

247

259

283

303

317

329

345

359

383

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 417

Contributors SUSANA L.A. ANDRADE  BIOSS Centre for Biological Signalling Studies, Freiburg, Germany; Institut f€ ur Biochemie, Albert-Ludwigs-Universit€ at Freiburg, Freiburg, Germany LAURA BACIOU  Laboratoire de Chimie Physique, UMR 8000 CNRS Universite´ Paris Sud, Universite´ Paris Saclay, Orsay cedex, France AURE´LIE BADILLO  Institut de Biologie et Chimie des Prote´ines, Molecular Microbiology and Structural Biochemistry, Labex Ecofect, UMR 5086 CNRS, Universite´ de Lyon, Lyon, France BLANCA BARQUERA  Department of Biological Sciences, Center for Biotechnology and Interdisciplinary Studies, Rensselaer Polytechnic Institute, Troy, NY, USA ALEXANDRE BARRAS  Sys TEMM, IBGC, CNRS, UMR 5095, IBGC, Bordeaux, France; CNRS, UMR 5248, CBMN, Universite´ de Bordeaux, Pessac, France TANIA BIZOUARN  Laboratoire de Chimie Physique, UMR 8000 CNRS Universite´ Paris Sud, Universite´ Paris Saclay, Orsay cedex, France ANJA BO¨CKMANN  Institut de Biologie et Chimie des Prote´ines, Molecular Microbiology and Structural Biochemistry, Labex Ecofect, UMR 5086 CNRS, Universite´ de Lyon, Lyon, France FRANC¸OISE BONNETE  IBMM UMR 5247 CNRS-UM-ENSCM, Chimie BioOrganique et Systemes Amphiphiles, Universite d’Avignon, Avignon, France ISABELLE BROUTIN  Laboratoire de Cristallographie et RMN Biologiques, Faculte´ de pharmacie, UMR 8015 Centre National de la Recherche Scientifique, Universite´ Paris Descartes, Paris, France STEFAN BRUECHERT  Institut f€ ur Biochemie, Albert-Ludwigs-Universit€ a t Freiburg, Freiburg, Germany LAETITIA DAURY  CBMN, UMR CNRS 5248, Universite´ de Bordeaux, Pessac, France CE´DRIC DELUZ  Laboratory of Physical Chemistry of Polymers and Membranes, Ecole Polytechnique Fe´de´rale de Lausanne, Lausanne, Switzerland ALINE DESMYTER  Architecture et Fonction des Macromole´cules Biologiques, Unite´ Mixte de Recherche 7257 Centre National de la Recherche Scientifique, Aix-Marseille, France MANUELA DEZI  IMPMC, UMR7590, UPMC-MNHN-IRD, Paris, France LUMINITA DUMA  CNRS Enzyme and Cell Engineering Laboratory, Sorbonne Universite´s, Universite´ de Technologie de Compie`gne, Compie`gne Cedex, France NICOLAS DUMESNIL  IMPMC, UMR7590, UPMC-MNHN-IRD, Paris, France JEROEN F. VAN DYCK  Biomolecular and Analytical Mass Spectrometry Group, Chemistry Department, University of Antwerp, Antwerp, Belgium GAUTHIER ESNOT  IMPMC, UMR7590, UPMC-MNHN-IRD, Paris, France CATHERINE ETCHEBEST  Unite´ INSERM UMRS1134, Institut National de la Transfusion Sanguine, Universite´ Paris 7 Denis Diderot, Paris, France AYMEN EZZINE  Laboratoire de Chimie Physique, UMR 8000 CNRS Universite´ Paris Sud, Universite´ Paris Saclay, Orsay cedex, France MARIE-LAURE FOGERON  Institut de Biologie et Chimie des Prote´ines, Molecular Microbiology and Structural Biochemistry, Labex Ecofect, UMR 5086 CNRS, Universite´ de Lyon, Lyon, France

ix

x

Contributors

RE´MI FRONZES  Unite´ G5 Biologie structurale de la se´cre´tion bacte´rienne, Institut Pasteur, Paris, France; UMR 3528, CNRS, Institut Pasteur, Paris, France; European Institute of Chemistry and Biology, University of Bordeaux, Pessac, France; CNRS UMR 5234 Microbiologie Fondamentale et Pathoge´nicite, Paris, France CAROLE GARDIENNET  Molecular Microbiology and Structural Biochemistry, Labex Ecofect, UMR 5086 CNRS – Universite´ de Lyon, Lyon, France; CRM2, UMR 7036, CNRS, Universite´ de Lorraine, Vandoeuvre-le`s-Nancy, France ´ FREDE´RIC GE´LE´BART  IMPMC, UMR7590, UPMC-MNHN-IRD, Paris, France MARIE-FRANCE GIRAUD  SysTEMM, IBGC, CNRS, UMR 5095, Bordeaux, France; Universite´ de Bordeaux, Bordeaux, France MARIE GLAVIER  CBMN, UMR CNRS 5248, Universite´ de Bordeaux, Pessac, France LUIGINO GRASSO  Laboratory of Physical Chemistry of Polymers and Membranes, Ecole Polytechnique Fe´de´rale de Lausanne, Lausanne, Switzerland FRANCESCA GUBELLINI  Unite´ G5 Biologie structurale de la se´cre´tion bacte´rienne, UMR 3528 CNRS, Institut Pasteur, Paris, France; UMR 3528, CNRS, Institut Pasteur, Paris, France; Unite´ de Microbiologie Structurale, Institut Pasteur, Paris, France LUCIE HARTMANN  Biotechnology and Cell Signalling, IMPReSs Protein Facility, UMR7242 CNRS-University of Strasbourg, Illkirch, France GHE´RICI HASSAI¨NE  Theranyx, Marseille, France CLAUDE HATTAB  UMR-S1134, Integrated Biology of Red Blood Cells, Inserm, Universite´ Paris Diderot, Sorbonne Paris Cite´, Institut National de la Transfusion Sanguine, Laboratoire d’Excellence GR-Ex, Paris, France PETRA HELLWIG  Laboratoire de Bioe´lectrochimie et Spectroscopie, UMR 7140 CNRS, Chimie de la Matie`re Complexe, Universite´ de Strasbourg, Strasbourg, France RUUD HOVIUS  Laboratory of Physical Chemistry of Polymers and Membranes, Ecole Polytechnique Fe´de´rale de Lausanne, Lausanne, Switzerland SORIA IATMANEN-HARBI  Sorbonne Universite´s—UPMC Univ Paris 06, E´cole Normale Supe´rieure—PSL Research University, De´partement de Chimie, CNRS UMR 7203 LBM, Paris, France ALBERT KONIJNENBERG  Biomolecular and Analytical Mass Spectrometry Group, Chemistry Department, University of Antwerp, Antwerp, Belgium SEBASTIEN KRIEGEL  Laboratoire de Bioe´lectrochimie et Spectroscopie, UMR 7140 CNRS, Chimie de la matie`re complexe, Universite´ de Strasbourg, Strasbourg, France BRITTA KUNERT  Molecular Microbiology and Structural Biochemistry, Labex Ecofect, UMR5086 CNRS, Universite´ de Lyon, Lyon, France VILIUS KURAUSKAS  Universite´ Grenoble Alpes, Institut de Biologie Structurale (IBS), Grenoble, France; CNRS, IBS, Grenoble, France; CEA, IBS, Grenoble, France DENIS LACABANNE  Molecular Microbiology and Structural Biochemistry, Labex Ecofect, UMR5086 CNRS, Universite´ de Lyon, Lyon, France JEAN-JACQUES LACAPERE  Sorbonne Universite´s—UPMC Univ Paris 06, E´cole Normale Supe´rieure—PSL Research University, De´partement de Chimie, CNRS UMR 7203 LBM, Paris, France OLIVIER LAMBERT  CBMN, UMR CNRS 5248, Universite´ de Bordeaux, Pessac, France ISABELLE LARRIEU  SysTEMM, CNRS, UMR5095, IBGC, Bordeaux, France; Universite´ de Bordeaux, Bordeaux, France CLAIRE LEMAIRE  CNRS-UMR9198, CEA-IBITECS, Universite´ Paris-Sud, I2BC, Gif- sur-Yvette, France

Contributors

xi

XIAO-DAN LI  Laboratory of Biomolecular Research, Paul Scherrer Institute, Villigen, Switzerland PATRICK J. LOLL  Department of Biochemistry and Molecular Biology, Drexel University College of Medicine, Philadelphia, PA, USA LAMIA MEBARKI  Theranyx, Marseille, France BEAT H. MEIER  Physical Chemistry, ETH Zurich, Zurich, Switzerland ESTELLE METZGER  Biotechnology and Cell Signalling, IMPReSs Protein Facility, UMR7242 CNRS-University of Strasbourg, Illkirch, France SONJA MINNIBERGER  Universite´ Grenoble Alpes, CNRS, CEA, CNRS, IBS, Grenoble, France MARC MORAND  IMPMC, UMR7590, UPMC-MNHN-IRD, Paris, France CHRISTOPHE MOREAU  Univ. Grenoble Alpes, CNRS, CEA, CNRS, IBS, Grenoble, France SAMUEL MURAIL  Laboratoire de Biochimie The´orique, CNRS, UPR9080, University Paris Diderot, Sorbonne Paris Cite´, Paris, France; Air Liquide, Centre de Recherches ParisSaclay, Jouy-en-Josas, France YASHVIN NEEHAUL  Laboratoire de Bioe´lectrochimie et Spectroscopie, UMR 7140, Chimie de la Matie`re Complexe, Universite´ de Strasbourg, Strasbourg, France; Mauritius Oceanography Institute, Albion, Mauritius LUC NEGRONI  CNRS-UMR5248, Universite´ de Bordeaux, Bordeaux, France; IGBMC, CNRS-UMR 7104, Illkirch, France HUGUES NURY  Universite´ Grenoble Alpes, CNRS, CEA, CNRS, IBS, Grenoble, France BENOIˆT ODAERT  CBMN, CNRS, UMR 5248, Pessac, France; Universite´ de Bordeaux, Pessac, France MARIANO ANIBAL OSTUNI  UMR-S1134, Integrated Biology of Red Blood Cells, Inserm, Universite´ Paris Diderot, Sorbonne Paris Cite´, Institut National de la Transfusion Sanguine, Laboratoire d’Excellence GR-Ex, Paris, France NOE´MIE OTTELARD  Biotechnology and Cell Signalling, IMPReSs Protein Facility, UMR7242 CNRS-University of Strasbourg, Illkirch, France VASSILIOS PAPADOPOULOS  The Research Institute of the McGill, University Health Center, Montreal, Quebec, Canada; Department of Medicine, McGill University, Montreal, Quebec, Canada LUCIE PECLINOVSKA  Universite´ Grenoble Alpes, CNRS, CEA, CNRS, IBS, Grenoble, France FRANC¸OIS PENIN  Institut de Biologie et Chimie des Prote´ines, Molecular Microbiology and Structural Biochemistry, Labex Ecofect, UMR 5086 CNRS, Universite´ de Lyon, Lyon, France MARTIN PICARD  Laboratoire de Biologie Physico-Chimique des Prote´ines Membranaires, UMR 7099, Institut de Biologie Physico-Chimique, FRC 550, Centre National de la Recherche Scientifique, Universite´ Paris 7, Paris, France GINA CATALINA REYES-MEJIA  Universite´ Grenoble Alpes, Institut de Biologie Structurale (IBS), Grenoble, France; CNRS, IBS, LabEx ICST, Grenoble, France; CEA, IBS, Grenoble, France PHILIPPE ROSIER  Institut de Mine´ralogie, Physique des Mate´riaux et Cosmochimie, UMR 7590 CNRS-UPMC-MNHN-IRD Case Courrier 115, Paris, France DIMITRI SALVADOR  CBMN, UMR CNRS 5248, Universite´ de Bordeaux, Pessac, France CORINNE SANCHEZ  SysTEMM, IBGC, CNRS, UMR 5095, Bordeaux, France; Universite´ de Bordeaux, Bordeaux, France

xii

Contributors

PAUL SCHANDA  Universite´ Grenoble Alpes, Institut de Biologie Structurale (IBS), Grenoble, France; CNRS, IBS, Grenoble, France; CEA, IBS, Grenoble, France LUCILE SENICOURT  Sorbonne Universite´s—UPMC Univ Paris 06, E´cole Normale Supe´rieure—PSL Research University, De´partement de Chimie, CNRS UMR 7203 LBM, Paris, France FRANK SOBOTT  Biomolecular and Analytical Mass Spectrometry Group, Chemistry Department, University of Antwerp, Antwerp, Belgium; Astbury Centre for Structural Molecular Biology, University of Leeds, Leeds, UK; School of Molecular and Cellular Biology, University of Leeds, Leeds, UK HAJER SOUABNI  Laboratoire de Chimie Physique, UMR 8000 CNRS Universite´ Paris Sud, Universite´ Paris Saclay, Orsay cedex, France REMY SOUNIER  Institut de Ge´nomique Fonctionnelle (IGF), CNRS, INSERM, Univ. Montpellier, Montpellier, France BATOUL SROUR  Laboratoire de Bioe´lectrochimie et Spectroscopie, UMR 7140 CNRS, Chimie de la Matie`re Complexe, Universite´ de Strasbourg, Strasbourg, France HENNING STAHLBERG  Center for Cellular Imaging and NanoAnalytics, Biozentrum, University of Basel, Basel, Switzerland CLAIRE STINES-CHAUMEIL  CRPP, CNRS, UPR 8641, Universite´ de Bordeaux, Pessac, France JEAN-CHRISTOPHE TAVEAU  CBMN, UMR CNRS 5248, Universite´ de Bordeaux, Pessac, France ZLATOMIR TODOROV  Universite´ Grenoble Alpes, Institut de Biologie Structurale (IBS), Grenoble, France; CNRS, IBS, LabEx ICST, Grenoble, France; CEA, IBS, Grenoble, France JAMES TOLCHARD  CBMN, CNRS, UMR 5248, Pessac, France; Universite´ de Bordeaux, Pessac, France TAKASHI TOMIZAKI  Swiss Light Source, Paul Scherrer Institute, Villigen, Switzerland AURORE VAITINADAPOULE  Unite´ INSERM UMRS1134, Laboratory of Excellence, Institut National de la Transfusion Sanguine, Universite´ Paris-Diderot, Sorbonne Paris Cite´, Universite´ de la Re´union, Paris, France CATHERINE VE´NIEN-BRYAN  IMPMC, UMR7590, UPMC-MNHN-IRD, Paris, France ALICE VERCHE`RE  Department of Biochemistry, Weill Cornell Medicine, New York, NY, USA MICHEL VIVAUDOU  Institut de Biologie Structurale (IBS), Universite´ Grenoble Alpes, Grenoble, France; CNRS, IBS, LabEx ICST, Grenoble, France; CEA, IBS, Grenoble, France HORST VOGEL  Laboratory of Physical Chemistry of Polymers and Membranes, Ecole Polytechnique Fe´de´rale de Lausanne, Lausanne, Switzerland RENAUD WAGNER  Biotechnology and Cell Signalling, IMPReSs Protein Facility, UMR7242 CNRS-University of Strasbourg, Illkirch, France; UMR7242, ESBS, Illkirch, France ROMAIN WYSS  Laboratory of Physical Chemistry of Polymers and Membranes, Ecole Polytechnique Fe´de´rale de Lausanne, Lausanne, Switzerland EDMOND YAZO-KONE  Sys TEMM, IBGC, CNRS, UMR 5095, Bordeaux, France; Universite´ de Bordeaux, Bordeaux, France MICHEL ZIVY  CNRS, PAPPSO, UMR Ge´ne´tique Quantitative et Evolution – Le Moulon, Gif-sur-Yvette, France

Chapter 1 Recombinant Overexpression of Mammalian TSPO Isoforms 1 and 2 Lucile Senicourt, Soria Iatmanen-Harbi, Claude Hattab, Mariano Anibal Ostuni, Marie-France Giraud, and Jean-Jacques Lacapere Abstract TSPO is a 18 kDa membrane protein that exists in mammalian as two isoforms 1 and 2. They are involved in different functions and are located in different membranes. TSPO1 is mainly located in outer mitochondrial membrane, whereas TSPO2 is encountered in plasma membrane of red blood cells. Determination of their structures is a milestone to understand their function. Their natural abundance is not sufficient to get large amounts usually required for structural studies. We described heterologous overexpression in both bacterial and cell-free system and purification on immobilized-metal affinity chromatography (IMAC) of both proteins. Using the same vector, TSPO1 is mostly recovered in bacterial inclusion bodies whereas TSPO2 is found in both bacterial cytosol and inclusion bodies, but in low amounts. Cell-free expression was the best system to overexpress pure TSPO2. Key words TSPO, Expression vector, Escherichia coli bacteria, Cell-free, Ni-NTA, NMR

1

Introduction Overexpression of membrane proteins in bacteria is one of the simplest and most widely used approaches to get recombinant protein in relatively large amounts. On top of this, production of labeled protein in bacteria is quite straight full since a change of culture media from rich to minimum medium supplemented with nitrogen- and/or carbon-enriched sources, or desired labeled amino acids or labeled amino acids precursors is generally efficient [1–3]. Global strategy requires a preliminary cloning step of the target gene into an expression vector [4], then insertion of the vector into desired bacterial strain and optimization of the expression conditions [5]. The purification step remains crucial since membrane proteins are not soluble and thus have either to insert into bacterial membrane alone or with the help of co-produced proteins or tags [6]. Several fusion tags have been tested to increase protein solubility in particular for soluble proteins [7]. Aggregation

Jean-Jacques Lacapere (ed.), Membrane Protein Structure and Function Characterization: Methods and Protocols, Methods in Molecular Biology, vol. 1635, DOI 10.1007/978-1-4939-7151-0_1, © Springer Science+Business Media LLC 2017

1

2

Lucile Senicourt et al.

of proteins into insoluble inclusion bodies was longer described as the main limiting factor, but it was converted in an advantage when coupled to refolding strategies [8, 9]. However, bacterial overexpression is not always successful and other strategies have to be developed. Here, we described how the overexpression of two isoforms of TSPO has been obtained combining Escherichia coli bacteria and cell-free system using the same vector.

2

Materials

2.1 Bacterial Expression

1. pET15b vector was used to insert cDNA of mouse TSPO1 and human TSPO2 between the BamH I and EcoRI restriction sites, just following His tag and Thrombin cleavage site. 2. Plasmid quality control was assessed by agarose gel electrophoressis (I-Mupid mini agarose gel electrophoresis system, Eurogentec, France). Agarose gel was obtained by mixing 0.8 g agarose in 100 mL TBE buffer (10.8 g/L Tris, 5.5 g/L Borate, 2 mM EDTA adjusted at pH 8) and heating with microwave oven until the solution becomes clear. Sample loading buffer (6): 1 mL sterile H2O, 1 mL Glycerol, and enough bromophenol blue to make the buffer deep blue (~0.05 mg). DNA ladder standards: DNA molecular weight marker III (Boehringer Mannhein). Running buffer: TBE buffer (0.5). Running conditions were 135 V. Stain and visualization was done with SYBR safe DNA gel stain (Invitrogen) and blue light transilluminator (Gentaur, France). 3. Escherichia coli bacteria strains BL21(DE3) and BL21(DE3) pLysS were used to overexpress the TSPO, whereas strains XL1-blue and DH5 alpha were used to get plasmid amplification. 4. Ampicillin stock solution (100 mg/mL): 1 g ampicillin (Euromedex, France) are dissolved in 10 mL water, solution can be kept frozen at 20  C. 5. Chloramphenicol stock solution (30 mg/mL): 300 mg chloramphenicol (Applichem, Coger-France) are dissolved in 10 mL EtOH, solution can be kept at 20  C. 6. Isopropyl-1-Thio Beta-D-galactopyranoside (IPTG) stock solution (1 M): 2.38 g IPTG are dissolved in 10 mL water, solution can be kept frozen at 20  C. 7. Agar plates were obtained dissolving 40 g of LB agar (Miller) in 1 L of distilled water, autoclave solution. When it cools at ~50  C antibiotics (ampicillin with or without chloramphenicol

Recombinant mTSPO1 and hTSPO2

3

for pLysS bacteria) were added, solution poured into plates and let solidified at room temperature. 8. Bacteria were cultured in LB broth or M9 minimum media for nonlabeled or labeled protein, respectively. LB broth contain 10 g/L Casein Peptone Plus (Organotechnie, France), 5 g/L Yeast extract (Organotechnie, France), 10 g/L NaCl (VWR, France), and distilled water. If needed pH is adjusted at 7.0 with NaOH (5 N). Basic M9 medium contains 2 g/L KH2PO4, 8 g/L Na2HPO4, 0.5 g/L NaCl, and distilled water. Enriched M9 medium contains basic M9 medium supplemented with 10 ml/L of stock solution of trace elements, 1 mL/L of solution A, 4 g/L glucose (13C6), and 1 g/L ammonium (15N). Trace elements stock solution (1 L final volume) contained: 5 g EDTA, 830 mg FeCL3, 84 mg ZnCl2, 13 mg CuCl2(2H2O), 10 mg CoCl2(6H2O), and 10 mg H3BO3. Complementation (solution A): 50 mg/L CaCl2, 10 mg/L Thiamine, 246 mg/L MgSO4. All media were sterilized before mixing with bacteria (autoclave, VS 130 Getinge, France). 9. Incubation was performed in 50 mL, 500 mL, and 2 L flasks containing 10 mL, 100 mL, and 500 mL medium, respectively, in a Sartorius Stedim Biotech (Certomat BS-T) operated at 180–200 rpm and 37  C. 2.2 Purification from Bacteria

1. Buffer A: 150 mM NaCl and 50 mM Hepes-NaOH pH 7.8. 2. Buffer B: Buffer A supplemented with 250 mM Imidazole, pH 7.8 (care has to be taken to get a lot of imidazole from a manufacturer that has a low UV absorbance). 3. Buffer C: Buffer B diluted 1/50 in Buffer A (i.e., 5 mM final concentration for Imidazole), pH 7.8. 4. Bacteria were lyzed either by sonication with Fischer Scientific (Vibra Cell 75115 500 W) operated at 40% power using a 3 mm probe or a cell disruptor (Constant System Ltd.,)in buffer A. For sonification, three cycles of 40 s were performed, the probe plunging in 50 mL pot placed in ice bucket to reduce heating effects. For cell disruptor one or two cycles at 200 bar (2900 psi) were performed. 5. Bacteria were recovered and bacterial fragments separated by centrifugation using Beckman (AVANTI J-E) and rotor JLA 9.100 for 1 L pots, rotor JA 25–50 for 30 mL tubes. 6. Membrane protein solubilization was performed using sodium docecyl sulfate (SDS, Sigma-Aldrich, Saint-Quentin Fallavier, France).

4

Lucile Senicourt et al.

7. Benzonase Nuclease HC (Novagen, France) was used to degrade nucleic acids. 8. Ni-NTA Agarose (Qiagen France SA) was either placed in 2 mL eppendorf tube for minipurification or loaded in 12 mL Polyprep columns (BioRad, France) for standard purification. 2.3 Cell-Free Expression

1. Cell-free extracts: ARN polymerase T7, S30 Extract, were prepared as previously described [10] and stored at 80  C. 2. Master mix: final concentration are NaN3 (0.05%), PEG (2%), potassium acetate (150 mM), magnesium acetate (7 mM), Hepes (0.1 M), Complete protease inhibitor cocktail (1/50), folinic acid (0.1 mg/mL), DTT (2 mM), nucleotide triphosphate mix (ATP, 1.2 mM, CTP, UTP, GTP, 0.8 mM each), phosphoenol pyruvate (20 mM), acetyl phosphate (20 mM), amino acid mix (0.5 mM), RCWMDE mix (1 mM). 3. Feeding mix (17 mL total): 8.7 mL master mix, 5.9 mL S30 buffer, 2.1 mL amino acid mix (4 mM), and 0.3 mL H2O. 4. Reaction mix (1 mL total): 512 μL master mix, 4 μL pyruvate kinase (10 mg/mL), 12 μL ARNt (40 mg/mL), 14 μL ARN polymerase T7 (420 U/μL), 9 μL RNase inhibitor (40 U/μL), 350 μL S30 Extract (100%), 75 μL DNA plasmid (0.2 mg/ mL), and 24 μL H2O.

2.4 Purification from Cell-Free Production

1. Precipitate from cell-free reaction medium is centrifuged, washed and protein solubilized with SDS.

2.5

1. Protein composition was analyzed with SDS-PAGE. Mini protean cell system (Bio-Rad, France) using precast gels (12% acrylamide, 10 wells comb of 30 μL).

Characterization

2. 0.5 mL Ni-NTA Agarose (Qiagen France SA) was loaded in 12 mL Poly-prep columns (BioRad, France) for purification.

Bacteria resuspension buffer contains 50 mM Hepes-Na pH 7.8, 150 mM NaCl, 1% SDS. Loading buffer (5) contains 200 mM Tris–HCl pH 6.8, 10% SDS (W/v), 20% glycerol (v/v), 10 mM β-mercaptoethanol, and 0.05% bromophenol blue (w/v). Running buffer contains 3 g/l Tris base, 14.4 g/L Glycine, 1 g/L SDS, and water qsp for 1 L, pH between 8.2 and 8.8. Electrophoretic migration was run at constant voltage (150 V). Staining buffer 1 g/L Coomassie blue (R250), 50% ethanol (v/v), 10% acetic acid (v/v), and 40% distilled water (v/v). Destaining buffer 10% ethanol (v/v), 7.5% acetic acid (v/v), and 82.5% distilled water (v/v). 2. Protein concentration was determined by UV absorbance recorded on UNICAM UV 300 (Fischer Scientific, Bioblock, Illkirch, France).

Recombinant mTSPO1 and hTSPO2

5

3. Mass spectroscopy was performed on MALDI-TOF: equal volumes of purified TSPO (1 mg/mL) and HCCA matrix were mixed in the presence of DDM (1 mg/mL), deposited on sample plate (AB Sciex, Canada), and observed in linear mode. 4. Western blots were performed using anti His monoclonal antibody (Protein Tech. Europe Ltd., France) or immunopurified anti-TSPO2 serum (Eurogentec, France). Anti TSPO2 serum was raised in rabbit against CPVHQPQPTEKSD peptide as immunogen; Pre-immune serum of the same rabbit was used as a control. 5. PageRuler plus prestained protein ladder (Thermo Scientific, France). 6. For Western blots SDS-PAGE was performed using 4–12% NuPAGE Bis Tris mini gels (Life Technologies, France) in NuPAGE MES-SDS running buffer (Life Technologies, France) following manufactured protocols. 7. Proteins were transferred to nitrocellulose membrane (BA45, GE Healthcare life science), using an XCell II module and NuPage transfer buffer (Life Technologies), supplemented with 10% ethanol. Transfer was performed 90 min at constant voltage (30 V) and room temperature. 8. Unspecific antigenic sites were saturated by incubation 60 min at room temperature in Tris buffer saline (TBS, pH 7.5) containing 5% fat-free milk. Saturation was followed by three times 5 min washing in TBS. After washing membranes were incubated in the presence of the primary antibody diluted in TBS/5% free fat milk. In the case of WB targeting His-tag, 0.5% Tween20 were added into buffers. Conditions for anti His antibody were: overnight at 4  C, dilution 1/10,000. Conditions for immunopurified antiTSPO2 serum or for control pre-immunized serum were: 60 min at 37  C, dilution 1/400. Primary antibody incubation was followed by three times 10 min at room temperature washing into TBS/0.5% Tween20. Secondary antibody (dilution 1/5000 into TBS/0.5% free fat milk) was incubated for 60 min at room temperature. After three times 10 min washing in TBS/Tween20 and one time in TBS, membranes were revealed using enhanced chemiluminescence (ECL) reactive (GE Healthcare life science).

3

Methods

3.1 TSPO (1 and 2) Plasmid Characterization

Whatever the way you get a plasmid containing the cDNA sequence of interest (home-made preparation, company purchase, or gift from another lab), it is generally good to regularly verify the integrity of the sequence and its quality (see Note 1). 1. A first step of plasmid amplification is needed to get amounts sufficient for characterization. Competent bacteria (XL1-blue

6

Lucile Senicourt et al.

or DH5-alpha) were transformed with plasmid using a thermal-shock procedure: 1 μL of plasmid (10 ng/μL) was mixed with 100 μL of competent bacteria and incubated for 15 min on ice. Bacteria were placed in a water bath set at 42  C for 45 s and replaced on ice. 400 μL of sterile LB broth was added and bacteria grown at 37  C for 1 h upon agitation (200 rpm). Transformed bacteria were plated onto an LB agarose plate containing ampicilin (100 μg/mL) and incubated overnight. A single colony was picked and inoculated to 10 mL sterile LB broth supplemented with ampicillin (100 μg/mL) and cultured overnight at 37  C upon agitation (200 rpm). Depending on the amount of plasmid needed a second step of bacterial culture (from 25 to 500 mL) has to be performed to increase the amount of bacteria. 2. Extraction of plasmid can done with standard Mini, Midi, or Maxi prep kit available from different companies and following the described protocol (see Note 2). 3. The ratio of absorbance at 260 nm (nucleic acids) and 280 nm (proteins) permits us to assess the purity of DNA and a value of around 1.8 is generally accepted as “pure” for DNA. 4. However, it does not prove the integrity of DNA, since absorbance methods do not distinguish between DNA and free nucleotides. Use of fluorescent probes specific for one type of molecule (DNA, RNA, proteins) and which fluorescence increases upon binding can help to solve this problem. Recently, system such as Qubit Fluorometer (Life Technology) gives a simple, but costly way to characterize purified plasmids. 5. A more classical approach such as agarose gels gives good and cheaper results (Fig. 1). 6 μL of purified plasmids were mixed with 4 μL sample loading buffer. Carefully pipette 10 μL of each sample mixture into separate wells in the gel. Pipette 3 μL of the DNA ladder standard into at least one well of each row on the gel. Run the gel in TBE applying the correct voltage and maintain the power run until the blue dye approaches the end of the gel. Stain the gel by adding SYBR safe DNA gel stain according to manufacturer, remove the stain, wash with water, and visualize the band on blue light transilluminator. 6. Sequence integrity is easily verified by commercially available sequencing using T7 promoter. 3.2 Transfection of Overexpressing Bacteria with TSPO (1 and 2) Containing Plasmid

1. Competent bacteria strains BL21(DE3) and BL21(DE3) pLysS were transformed with plasmid as described above. 2. Transformed bacteria were plated onto LB plates containing ampicilin (100 μg/mL) and incubated overnight at 37  C (see Note 3). 3. A single colony was picked and inoculated to 10 mL sterile LB broth supplemented with antibiotic.

Recombinant mTSPO1 and hTSPO2

7

Direction of migration

Relaxed circular plasmid

12000

Supercoiled DNA

5000 2000 1000

DNA ladder

pET15b-hTSPO2

pET15b-mTSPO1

DNA ladder

Fig. 1 Agarose gel of plasmid preparation. 0.8% agarose gel was loaded with DNA ladder and purified pET15b plasmids containing either mTSPO1 or hTSPO2 DNA sequence (100 ng). Gel shows a major band of supercoiled DNA and in the same lane fainter band corresponding to relaxed circular plasmid. The pET15b plasmid contains 5708Bp and TSPO sequence 527Bp giving a total of 6235Bp. Migration of the major band is at the expected kb as seen in the DNA ladder

3.3 Expression and Purification of Mouse TSPO1 (mTSPO1)

1. Preculture: Bacteria containing the plasmids were inoculated to 15 mL sterile LB broth containing 100 μg/mL ampicillin. Bacterial growth was followed by measuring optical density (OD) at 600 nm. Usually, overnight incubation at 37  C gives OD close to 5. Depending on the bacteria needed, a second preculture can be performed in a larger volume of 200 mL of LB in a 1 L flask (see Note 4). 2. Culture: adequate preculture volume was transferred into 500 mL sterile LB supplemented with ampicillin in 2 L flask in order to start with an OD above 0.2 (see Note 5). Good amounts of mTSPO1 production were obtained with four flasks, i.e., 2 L of culture. When bacteria are in the exponential growth phase, typically when optical density is between 0.7-1, 1 mM of IPTG was added to induce protein production for at least 3–4 h (Fig. 2a). Bacterial growth was followed by measuring optical density (OD) at 600 nm (Fig. 2a). In order to follow mTSPO1 expression, aliquot (1 mL) was taken during culture at various time intervals, OD was adjusted to 1 by adding LB and 1 mL samples centrifuged (5000  g) to get bacterial pellet. The supernatant was discarded in sodium hypochlorite (bleach) and pellet was resuspended with 100 μL resuspension buffer.

Lucile Senicourt et al.

A

6 5 OD at 600nm

8

IPTG

4 3 2 1 0 0

1

2

3 4 Time, hours

5

6

7

B 250 150 100 75 60 37 25 20

mTSPO1

15 10

3.0 4.0

4.5

5.0

6.0

6.5

Time,

hours

IPTG

Fig. 2 Bacterial expression of mTSPO1. (a) Time course of bacterial growth measured as optical density at 600 nm changes as a function of time. (b) SDSPAGE (12% acrylamide (w:v)) of fraction collected at different times after induction by the addition of IPTG. (first lane) Molecular weight standards. (second and last lane) Purified mTSPO1

30 μL were taken out and mixed with 7.5 μL loading buffer, 20 μL were loaded on 12% acrylamide precast gel and gel run under standard conditions (Fig. 2b). 3. Bacterial harvesting: at the end of the induction for protein production, bacteria were collected by centrifugation in 1 L bottles at 6700  g for 20 min at 4  C. The supernatant was discarded (in sodium hypochlorite, bleach) and each pellet was washed with 25 mL of fresh buffer A and centrifuged again. Pellets were resuspended in 25 mL of buffer A, pooled in 30 mL bottles, and centrifuged at 6700  g for 20 min at 4  C. Cell pellets can be stored frozen and treated later or resuspended in 15 mL buffer A.

Recombinant mTSPO1 and hTSPO2

9

4. Bacterial lysis: bacterial suspension can be treated either by sonication or by cell disruption depending on the amount of available material (see Note 6). Inclusion bodies were recovered by centrifugation (14,000  g, 4  C, 20 min). 5. Protein extraction: mTSPO1 was mostly encountered in inclusion bodies (IB) and was solubilized with sodium dodecyl sulfate (SDS). The first step was to resuspend each pellet in 19 mL of buffer A. In a second step, mTSPO1 was solubilized with 1 mL SDS 20%. Insoluble materials were removed by centrifugation in 30 mL bottles at 50,000  g for 30 min at 20  C (see Note 7). Supernatant (solubilized IB, sIB) contained a lot of DNA-RNA fragments that can be digested using benzonase. Benzonase nuclease was used to digest DNA and RNA fragments present in the supernatant. 6. Protein purification: the amount of membrane protein production was not the same depending on the culture conditions (LB or M9 for example), thus the amount of protein had to be estimated to optimize the purification step and in particular to adapt the amount of resin used to get the collected fractions with the highest concentration. Mini-purification in batch: small-scale purification is obtained by incubating 1 mL of sIB with Ni-NTA resin (200 μL agarose) pre-equilibrated in buffer A supplemented with 1% SDS for 15 min under stirring. Benchtop centrifuge was used to collect unbound materials and absorbance spectra were recorded (Fig. 3a). 1 mL of buffer C supplemented with 1% SDS was added to the Eppendorf containing Ni-NTA resin for washing. The tube was stirred and benchtop centrifuge was used to collect the wash material in the supernatant. This step was repeated three times and absorbance spectra were recorded for the three washing steps (W1, W2, and W3 in Fig. 3a). 1 mL of buffer B supplemented with 1% SDS was added to the Eppendorf containing Ni-NTA resin for elution; the tube was stirred and centrifuged to collect the eluted material. This step was repeated twice and absorbance spectra were recorded for the three elution steps (E1, E2, and E3 in Fig. 3a). mTSPO1 amount (Fig. 3b) can be estimated by converting optical densities at 280 nm using extinction coefficient of 3.88 mg/mL/cm calculated from recombinant mTSPO1 sequence (see Note 8). Purification with a column: the volume of resin stacked in the 12 mL Poly-prep columns (between 0.5 and 2 mL) was adjusted according to the amount of mTSPO1 determined with the “mini”-purification (see Note 9). Ni-NTA resin (desired volume, named one column volume, 1 V) was washed and equilibrated with 3 V of buffer A supplemented with 1% SDS. sIB volume was then layered on the top of the column (see Note 10).

Lucile Senicourt et al.

A

1.8 1.6

Optical Density

1.4 Tot dil1/10

1.2

W1 dil1/10 W2

1

W3

0.8

Buffer

0.6

e2

e1 e3

0.4 0.2 0 240

260

280

300

320

340

Wavelength, nm

B

0.4

0.3 Protein, mg/ml

10

0.2

0.1

0.0 e1

e2

e3

Fig. 3 mTSPO1 mini-purification in batch. (a) Absorbance spectra of the various steps of purification (Tot diluted 10 times in buffer A supplemented with 1%SDS), W1 diluted 10 times (same buffer), and nondiluted W2, W3. (e1, e2, and e3) Total and wash fractions have an absorption maximum at 260 nm corresponding probably to nucleotides. Elution fractions show a minimum at 250 nm and a maximum at 280 nm with a shoulder at 290 nm corresponding to protein containing tryptophan. (b) Estimated concentrations of mTSPO1 in each elution fraction

Column was first washed with 4 V of loading buffer, then with 4 V buffer C supplemented with 1% SDS to remove loosely bound contaminant proteins. Elution was performed by the addition of 0.5–1 V of buffer B supplemented with 0.1% SDS. Absorption spectra were recorded for each collected fraction (Fig. 4a). It has to be noticed that dilution of fractions is often needed to avoid saturation of the spectrophotometer. The change in maximal wavelength absorption was a characteristic of a change from nucleotide-rich fractions (260 nm) to proteinrich fractions (280 nm). The fractions having the highest 280 over 260 absorbance ratios were characteristic of protein.

Recombinant mTSPO1 and hTSPO2

B

A 1.6

Tot di1/10

1.2 Optical Density

Optical Density at 280nm

1.4 W1 dil1/10

1

W2 dil1/10 W3 dil1/4

0.8

e1 dil1/4

0.6

e2 dil1/16

0.4

e3 Buffer

6 5 4 3 2 1

0.2 0 240

11

0

260

280

300

320

340

W1

W2

W3

e1

e2

e3

Fractions

Wavelength, nm

C

D 250 150 100 75

37 25 20 15

Tot

Nb

W1 W2 W3

e1

e2

e3

5 Protein ,mg/ml

60

6

4 3 2 1

10

0 0

5

10

Fractions, mL

Fig. 4 Large-scale purification of mTSPO1. (a) Absorbance spectra of the various steps of purification (Tot diluted 10 times, W1 diluted 10 times, W2 diluted 10 times, W3 diluted 4 times, e1 diluted 4 times, e2 diluted 16 times, and nondiluted e3). Total and wash fractions have an absorption maximal at 260 nm corresponding probably to nucleotides. Elution fractions show a minimum at 250 nm and a maximum at 280 nm with a shoulder at 290 nm corresponding to protein containing tryptophan. (b) Absorbance at 280 nm for each corrected from base line and dilution. (c) SDS-PAGE (12%) of fraction collected upon purification on Ni-NTA of mTSPO1 extracted from sIB purification. 20 μL were deposited on each well. Tot, Nb, and e3 nondiluted; W3 and e1 diluted 2 times: W2, W1, and e2 diluted 3, 5, and 20 times respectively. (d) Calculated concentrations of mTSPO1 in each fraction (W and e) using extinction coefficient of mTSPO1 since SDS-PAGE has revealed that all fractions contain “only” mTSPO1

Plots of optical density at 280 nm for the various fractions give a first estimation of protein content (Fig. 4b). 7. Protein characterization: analysis of samples (20 μL) from various purification steps on SDS-PAGE revealed the protein content of each collected fraction (Fig. 4c). Fractions were highly pure, containing only mTSPO1, thus protein chromatogram can be drawn using the extinction coefficient of mTSPO1 (Fig. 4d). Mass spectrometry of mTSPO1-enriched fractions

12

Lucile Senicourt et al.

revealed that recombinant TSPO corresponded to the fulllength TSPO amino acid sequence plus a poly-histidine tag and was devoid of the N-terminal methionine [11] giving a molecular weight of 20,873 Da. This characterization of mTSPO1 was helpful since apparent molecular weight migration on gels below 20 kDa was lower than the calculated molecular weight (21,004 Da for 189 amino acids). 8. Production of labeled mTSPO1: for structural studies by NMR, 15N and 13C labeled proteins were needed. We used a well-described protocol [1, 12] with some modifications. A first step consisted of bacterial production in LB without induction to generate cell mass. The second step where the induction occurred, is primary an exchange to labeled media at high cell density followed by a short time of growth recovery before the IPTG induction (Fig. 5a). Cell harvesting and lysis, as well as IPTG

A Optical Density at 600nm

5 4 3 2 1 0 -2

-1

0 1 2 3 Induction Time, hours

B

C mTSPO [15N15C] (mg/mL)

250 150 100 75 60 37 25 20 15 10

Tot

Nb

W4 W5

e3

e4

e5

e6

4

5

20 18 16 14 12 10 8 6 4 2 0 0

2

4 6 fractions Ei

8

10

Fig. 5 Expression and purification of labeled mTSPO1. (a) Time course of bacterial growth in M9 medium measured as optical density at 600 nm changes as a function of time. (b) SDS-PAGE (12%) of fraction collected upon purification on Ni-NTA of mTSPO1 extracted from sIB. 20 μL were deposited on each well. Tot and Nb nondiluted; W3 and W4 diluted 4 times: e3 to e6 diluted 50 times. (c) Calculated concentrations of [15N, 13C] labeled mTSPO1 in each elution fraction

Recombinant mTSPO1 and hTSPO2

13

protein extraction and purification, were performed as previously described above. SDS-PAGE of main collected fractions revealed the presence of mTSPO1 (Fig. 5b). Elution fractions were smaller (0.3 mL instead of 1 mL) to gain more concentrated material (Fig. 5c). Mass spectroscopy indicated a yield of labeling above 90% that permits recording two-dimensional NMR spectra such as 1H-15N and 1H-13C Heteronuclear Single Quantum Coherence [11]. 3.4 Expression and Purification of Human TSPO2 (hTSPO2)

hTSPO2 cDNA was inserted in the same pET15b plasmid used for mTSPO1 and at the same restriction sites, expecting that the sequence homologies between mTSPO1 and hTSPO2 could give rise to similar expression profile in Escherichia coli bacteria (see Note 11). Two bacterial strains BL21(DE3) pLysS and BL21(DE3) were used, the first one for its higher specific expression since the second one is described having leaky expression of T7 polymerase that can lead to un-induced expression of potentially toxic proteins (see Note 12). 1. Precultures and cultures were made following the same procedure as described for mTSPO1, unless a second antibiotic (chloramphenicol, 30 μg/mL) was used for BL21(DE3) pLysS (see Note 3). 2. Bacterial harvesting and lysis were also performed as described for mTSPO1. After cell disruption, a first centrifugation at 14,000  g for 20 min at 4  C permitted recovering inclusion bodies in the pellet and cytosol containing membranes in the supernatant (see Note 13). 3. Purification from inclusion bodies (IB): Same protocol as for mTSPO1 was used, briefly, pellet from the first centrifugation was resuspended in buffer A containing 1% SDS, nonsoluble material was removed by centrifugation at 50,000  g for 30 min at room temperature. Benzonase nuclease was added to decrease the viscosity of the sample that was further loaded on top of Ni-NTA column. Column wash and elution steps were done as described above. Absorption spectra were recorded for each collected fraction and absorbance at 280 nm plotted (Fig. 6a). Protein characterization of the collected fraction showed on SDS-PAGE (Fig. 6b, c) that elution fractions contained high molecular weight proteins (on top of gel, see Note 14) which contributed for the large absorbance measured at 280 nm. Western blots of the sIB (Tot), nonbound fraction (Nb), as well as wash (W4) and elution (e1, e3) fractions revealed the presence of proteins containing a histidine-tag with different molecular weights (Fig. 7a) (see Note 15). A major one (in between 36 and 55 kDa) was not retained upon purification on Ni-NTA (lanes Tot and Nb), a second (close to 36 kDa) was not fully

Lucile Senicourt et al.

A

20 Optical Density at 280nm

14

15

10

5

0 W2 W3 W4 W5 e1

e2

e3

e4

e5

e6

e7

Fractions

B

Tot

Nb

Tot

e1

w1

w2

w3

w4

w5

e4

e5

e6

e4

250 150 100 75 60 37 25 20 15 10

C

e2

e3

e7

250 150 100 75 60 37 25 20 15 10

Fig. 6 Purification of hTSPO2 extracted from inclusion bodies of Escherichia coli bacteria strains BL21(DE3) pLysS. (a) 280 nm absorbance profile for wash (W) and elution (e) fractions from Ni-NTA column. (b and c) SDS-PAGE of loaded solution (Tot, sIB), flow through (Nb, not bound), wash (W), and elution (e) fractions. 20 μL were deposited on each well. (b) Tot, Nb, and W1 diluted 10 times; W2 diluted 2 times; W3 to W5 nondiluted; e4 diluted 4 times. (c) Tot and e3 diluted 10 times; e1, e5 to e7 nondiluted; e2 and e4 diluted 3 and 4 times, respectively

e3

e3

e3

e1

W4

Nb

kDa

15

B

Tot

A

Rec-hTSPO1

Recombinant mTSPO1 and hTSPO2

190 115 80 70 50

contaminant

30 25

Rec-hTSPO2 15

10

preimmune serum

Anti hTSPO2

(serum)

Anti His

Fig. 7 Western blot of selected fractions from purification of inclusion bodies containing hTSPO2 produced in Escherichia coli strains BL21(DE3) pLysS. (a) anti-histidine (b) serum anti TSPO2. 10 μL of Tot, Nb, W4 and e1 nondiluted fraction, and 0.1 μL of e3 fraction were deposited for antiHis, whereas 10 μL of e3 fraction were deposited for antiTSPO2

retained, but was concentrated in the elution. Finally, a third one (at a molecular weight close to that of hTSPO1) present in the total was highly reduced in the nonretained fraction and is slightly observed in the elution. Western blot using polyclonal anti-TSPO2 (rabbit serum anti-TSPO) revealed in elution (e3) a band in between 17 and 28 kDa corresponding to hTSPO2 (Fig. 7b). This band was not observed when using pre-immune serum (right lane in Fig. 7b). 4. Purification from membranes present in the supernatant fraction: due to the low presence of hTSPO2 in IB, it was important to check its content in both the soluble and membrane fractions. SDS 1% was added to bacterial cytosolic fraction to solubilize membranes. Benzonase nuclease was added to

16

Lucile Senicourt et al.

decrease the viscosity of the solution. Nonsoluble material was removed by centrifugation at 4000  g for 15 min at room temperature. The solution was loaded on top of Ni-NTA column pre-equilibrated with buffer A supplemented with 1% SDS. The column was first washed with 2 V of loading buffer (W1, W2), then with 2 V of loading buffer supplemented with 1% SDS and 5 mM imidazole (W3, W4) and finally with 2 V buffer C supplemented with 1% SDS (W5, W6) to remove contaminant protein loosely bound. Elution was performed by the addition of 0.5–1 V of buffer B supplemented with 0.1% SDS. Absorption spectra were recorded for each collected fraction and absorbance at 280 nm plotted as a function of collected fraction (Fig. 8a). Protein characterization of the collected fraction showed on SDS-PAGE (Fig. 8b, c) that elution fractions contained large amounts of protein in between 20 and 25 kDa but also some higher molecular weight proteins. Western blots of the cytosolic extract (Tot), nonbound fraction (Nb), as well as wash (W2) and elution (e1, e3) fractions revealed the presence of two proteins containing histidine-tags with different molecular weights (Fig. 9a). The largest molecular weight protein (below 55 kDa) was not retained on the Ni-NTA column (lanes Tot, Nb, and W2), the second (close to 36 kDa) was only observed in the elution. Western blot using specific anti-TSPO2 rabbit serum revealed in elution (e3) a band in between 15 and 25 kDa corresponding to hTSPO2 (Fig. 9b). This band was not observed when using pre-immune serum (right lane in Fig. 7b), but was also absent in western blot made with antihistidine monoclonal antibody antibody (see Note 16). 5. Due to the presence of high amounts of contaminant proteins in purified fraction from either inclusion bodies or cytosol fractions of bacterial strain BL21(DE3) pLysS, we tested classical strain BL21(DE3). Similar processes were performed with this second strain: preculture, culture, cell harvesting, lysis, inclusion bodies and cytosol fractioning, as well as purification on Ni-NTA column. Absorption spectra recorded for each collected fraction and absorbance at 280 nm measured for both purifications (sIB and Cyto) is presented (Fig. 10a). Eluted fractions from cytosol purification seem to contain higher amounts of protein according to 280 nm absorbance. However, SDS-PAGE revealed the main presence of high molecular weight proteins (between 100 and 150 kDa) far away from calculated molecular weight of hTSPO2 (Fig. 10b). Conversely, SDS-PAGE of eluted fraction from IB (Fig. 10c) showed three bands close to 25 kDa. Western blots (Fig. 11) of the most concentrated fraction eluted from purified sIB (e2) suggested the presence of hTSPO2 since a band

Recombinant mTSPO1 and hTSPO2

A

17

Optical Density at 280nm

20

15

10

5

0 W2 W3 W4 W5

e1

e2

e3

e4

e5

e6

Fractions

B

Tot

Tot cent

Nb

W1

W2

e1

e2

e3

W3 W4 IB

W4

250 150 100 75 60 37 25 20 15 10

C

W4 W5

e3 IB

e4

e5 250 150 100 75 60 37 25 20 15 10

Fig. 8 Purification of hTSPO2 from cytosolic fraction, containing membrane, of Escherichia coli bacteria strains BL21(DE3) pLysS. (a) 280 nm absorbance profile for wash and elution fractions from Ni-NTA column. (b and c) SDS-PAGE of loaded solution (Tot), flow through (Nb, not bound), wash (W), and elution (e) fractions. 20 μL were deposited on each well. Fractions from inclusion bodies purification were added for comparison (W4 and e3 labeled in square). Tot, Nb, and WI diluted 30 times; W3 to W5 as well as W4IB and e1, e2, e4 and e5 nondiluted; W2 diluted 5 times; e3 diluted 6 times; e3IB diluted 10 times

Lucile Senicourt et al.

e3

e3

e3

e1

W2

Nb

kDa

B

Tot

A

Rec-hTSPO1

18

190 115 80 70 50

30 25

Rec-hTSPO2

15

10

preimmune serum

Anti hTSPO2

(serum)

Anti Hist

Fig. 9 Western blot of selected fractions from purification of cytosolic fraction containing hTSPO2 produced in Escherichia coli strains BL21(DE3) pLysS. (a) Anti-histidine (b) serum anti TSPO2. 10 μL of Tot, Nb, W2, and e1 nondiluted fraction, and 0.3 μL of e3 fraction were deposited for antiHis, whereas 10 μL of e3 fraction were deposited for antiTSPO2

was observed at a molecular weight close to that of mTSPO1. This band was not observed in “controls” with preimmune serum and wash fraction (W5). However, no band was also observed at the same molecular weight in western blot performed with anti-histidine antibody. In conclusion, the use of Escherichia coli strain BL21(DE3) did not improve the quality of purified fractions compared with strain pLysS, SDS-PAGE and WB suggested a very low production of hTSPO2 and most important, a highly impure purification. 3.5 Cell-Free Expression of hTSPO2

One way to reduce the amount of contaminant proteins produced in bacteria is to use bacterial machinery alone with an appropriate plasmid such as in cell-free expression system [13]. Different setups have been described, mostly batch and two-compartment

Recombinant mTSPO1 and hTSPO2

A

19

8

Optical density at 280nm

7 6 5 IB

4

Cyto

3 2 1 0 W2 W3 W4 W5 W6

e1

e2

e3

e4

e5

Fractions

(Cyto)

B 250 150 100 75 60 37 25 20 15 10

Mb

Tot Nb

W1 W2 W3

W4

e3

W5 W6 e1

e2

e3

e4

e5

e2 IB

(IB)

C 250 150 100 75 60 37 25 20 15 10

Be Af

Tot

Nb

W1 W2

W3

e2

W4

W5 W6 e1

e2

e3 e4 e5

Fig. 10 Purification of hTSPO2 overexpressed in Escherichia coli strains BL21(DE3). (a) 280 nm absorbance profile for wash and elution fractions from Ni-NTA columns loaded with either inclusion bodies (IB) or cytosolic (Cyto) extracts. (b and c) SDS-PAGE of loaded and collected fractions of Cyto (b) and IB (c) extracts. Pure extracts (Mb, cytosolic fraction before sds treatment) (Be and Af bacteria before and after induction), loaded solution (Tot), flow through (Nb, not bound), wash (W), and elution (e) fractions. 20 μL were deposited on each well. Fraction from inclusion bodies purification was added for comparison (e2 IB labeled in square) in gel of cyto extract. Tot, Nb, and W1 diluted 10 times; W2 diluted 5 times, W3 to W6 as well as e1 and e5 nondiluted; e2 diluted 2 times; e3 and e4 for cyto diluted 4 times; e3IB and e4IB nondiluted

e4 cyto fraction

e2 IB fraction

W5 IB fraction

e2 IB fraction

e4 cyto fraction

W5 IB fraction

e4 cyto fraction

e2 IB fraction

kDa

W5 IB fraction

Lucile Senicourt et al.

Rec-hTSPO1

20

190 115 80 70 50

30 25

Rec-hTSPO2 15 10

Anti Hist

Anti hTSPO2 (serum)

preimmune serum

Fig. 11 Western blot of selected fractions from purification of inclusion bodies (IB) or cytosolic (Cyto) extract produced in Escherichia coli strains BL21(DE3) with antibodies directed against (left) the histidine tag (Anti His), (right) serum anti hTSPO2 or preimmune serum. 10 μL of W5 fraction from IB purification, 1.25 μL of e2 and e4 fraction from IB and cyto purifications were deposited for antiHis, whereas 10 μL of e3 and e4 fractions were deposited for antiTSPO2

continuous exchange (CECF). In the CECF configuration, a reaction mixture (RM) containing plasmid and machinery (such as ribosomes and enzymes) is separated from a feeding mixture (FM) containing precursors (such as amino acids and nucleotides) by a semi-permeable membrane. 1. Plasmid was amplified in Escherichia coli DH5 alpha bacteria strain and purified using Midi-Prep kit to get sufficient amounts (for small or large-scale cell-free expression, ~7 μg for 100 μL RM or ~15 μg for 1 mL of RM in test experiment or production, respectively). 2. 15 μg of purified plasmid (pET15b) containing hTSPO2 sequence was incubated in 1 mL reaction mix placed in 17 mL feeding mix for 20 h at 28  C (Fig. 12a). A clear white precipitate was observed. 3. RM was centrifuged in an Eppendorf tube to collect the precipitate (Fig. 12b). 4. The pellet was washed twice with 1 mL water and resuspended in 200 μL SDS 1%. 5. Absorption spectra recorded between 240 and 350 nm revealed the presence of contaminating nucleotides, but also

Recombinant mTSPO1 and hTSPO2

A

21

C Purification Pellet 1.4

Optical Density

1.2

B

1 0.8

Total e2

0.6

EluBuf f

0.4 0.2 0 240 260 280 300 320 340

190

115 80

115 80

70

70

50

50

30 25

30 25

15

15

10

10

Ni-NTA elution

Cell free pellet

Ni-NTA elution

Rec-mTSPO1

Ni-NTA elution

Ni-NTA elution

Cell free pellet

kDa

190

Rec-mTSPO1

Ni-NTA elution

Ni-NTA elution

kDa

Cell free pellet

D

Rec-mTSPO1

Wavelength, nm

Rec-hTSPO2

Anti Hist

Anti hTSPO2 (serum)

preimmune serum

Fig. 12 Cell-free production of hTSPO2. (a) Experimental setup of CECF reactor formed by a beaker containing the feeding mixture (FM) in which a dialysis membrane tube containing reaction mixture (RM) has been placed and closed by clips at both ends. (b) Protein precipitate recovered by centrifugation in an Eppendorf tube. (c) Absorption spectra of SDS resuspended pellet and elution fraction (e2) from Ni-NTA column. (d) Western blot of cell-free expression pellet and eluted fractions from purification on Ni-NTA column. (left) Anti-histidine tag and (right) anti hTSPO2 using serum or preimmune serum. Rec-hTSPO1 has been added as a control for antihistidine, whereas dotted blot of immune peptide is used as a control for hTSPO2

protein (280 nm and shoulder at 290 nm). Purification by incubation with 0.5 mL Ni-NTA resin in the presence of 2 mM imidazole and load on top 12 mL Poly-prep column enable removing nucleotides and remaining enzymes by washing (W1 and W2). Elution with buffer containing 200 mM

22

Lucile Senicourt et al.

imidazole gives proper elution fraction (e2) with spectrum corresponding to protein (Fig. 12c). 6. Western blots of the cell-free pellet, as well as elution (e2) fraction from Ni-NTA purification, revealed the presence of hTSPO2 at a molecular weight close to that of mTSPO1 using anti histidine antibody and a band at the same molecular weight using anti TSPO2 antibody (Fig. 12d). No reactivity was observed when pre-immune serum is used. 7. Mass spectrometry characterization confirmed that hTSPO2 with a molecular weight of 21,300 Da was present. Total amount of hTSPO2 produced is 0.07 mg and should be optimized to start structural studies.

4

Conclusions Overexpression of membrane proteins in bacteria might be straightforward; however, we show herein that two proteins belonging to the same family and presenting a high degree of amino acid conservation are not at all expressed similarly. Whereas mTSPO1 is highly expressed in Escherichia coli and mostly recovered in inclusion bodies, hTSPO2 is poorly expressed and recovered in both cytosol and inclusion bodies. Several points have to be reminded that (1) similarity in amino acids does not mean similarity in codons, (2) half life time of proteins can vary drastically, (3) most of membrane proteins require translocons for proper insertion into their target membrane [14]. However, it has been previously described that some membrane proteins are able to spontaneously fold and be inserted as transmembrane helix [15], suggesting that amino acid composition is the “clue.” Thus, it remains very difficult to predict what membrane protein will express “easily,” and trial and errors seem to remain the best way to succeed.

5

Notes 1. Overexpression is easily achieved by using high or medium copy plasmids. However, studies have demonstrated that plasmid-bearing cells lose their productivity fairly quickly as a result of genetic instability. Plasmids are genetically instable due to three processes: segregational instability, structural instability, and allele segregation [16]. 2. Care has to be taken upon plasmids purification and not use “overaged” kits that can lead to incorrect purification of fulllength plasmids with excess of free nucleotides. The ratio 260/ 280 suggested good purification but agarose gel revealed smeared bands.

Recombinant mTSPO1 and hTSPO2

23

3. When using Escherichia coli bacteria strains BL21(DE3) pLysS, a mixture of antibiotics is needed. 4. Size and shape of the flask are important, a large surface areato-volume ratio will permit better oxygenation important for bacteria growth. Fernbach flasks have been designed for such an issue. 5. If initial OD is too low, bacterial growth will take longer time and make working day longer! 6. Sonification with a 3 mm probe during 1 min at 40% power (Vibra cell) induces heating that can be deleterious for proteins, the cell disruptor has the advantage of using pressure but is most expensive. 7. Since SDS is solid at 4  C, all steps from centrifugation to purification have to be done at room temperature. 8. For unknown reasons, mTSPO1 amount thus calculated gives a reproducible overestimation by a factor 2. This might be due to nonspecific binding of protein that may have higher extinction coefficient than mTSPO1. 9. Most of the commercially available Ni-NTA resins have binding capacity of 10–50 mg protein per mL of resin. Less than 0.5 mL resin is difficult to manipulate and it is better to pool bacterial cultures in order to work with resin volumes between 0.5 and 2 mL. More than 2 mL resin is also difficult to manipulate, because flow at atmospheric pressure is slow and loading, washing and elution will take long time. An alternative for large-scale purification is FPLC, as previously described [12]. 10. Addition of 2–5 mM imidazole in the loading buffer may increase purification avoiding binding of contaminant proteins that have low affinity. 11. Protein having high degree of sequence homology could have large differences in base sequences leading to high variation in expression. Some of the factors involved are: (1) presence of rare codons that can be detected and modified, (2) formation of stable secondary structure of mRNA, (3) half-life time of mRNA. 12. Recovery of overexpressed protein is under control of several parameters. Different temperatures can be tested to recover more proteins either in inclusion bodies or inserted in membranes. Culture in the presence of rifampicin that has been described to increase specifically induced protein production by inhibiting Escherichia coli polymerase [17–21]. 13. A second centrifugation of the supernatant at 50,000  g for 60 min permits us to harvest enriched membranes in a second pellet. However, speed and/or time of centrifugation has to be adjusted depending on the molecular weight of the membrane protein.

24

Lucile Senicourt et al.

14. It has been described that Escherichia coli bacteria strains BL21 (DE3) pLysS overexpressed beta galactosidase which has a molecular weight of 117 kDa consistent with bands observed on the top of the gels and further characterized by mass spectroscopy. 15. It has been previously described that many native proteins from Escherichia coli were commonly co-purified by immobilized metal affinity chromatography [22]. 16. Detection of a protein on western blot is not straightforward due to cross-reaction. The presence of one or several contaminants reactive to the same antibody can reduce the intensity of the expected protein present in smaller amount. In our case (Fig. 7) the low amount of hTSPO2 is poorly detected by antiHist antibody because of the presence of contaminant (36 kDa) highly reactive to this anti body, whereas hTSPO2 is correctly detected by a more specific antibody that poorly detects the contaminant. A way to solve this problem is to deposit at least two different amounts of protein and to vary exposition time of secondary antibody.

Acknowledgment The authors would like to thank Professor V Papadopoulos for the generous gift of mTSPO1 plasmid, and F. Skouri-Panet for generous gift of some products and help in the use of equipment. References 1. Marley J et al (2001) A method for efficient isotopic labeling of recombinant proteins. J Biomol NMR 20:71–75 2. Guerrero SA et al (2001) Production of selenomethionine-labelled proteins using simplified culture conditions and generally applicable host/vector systems. Appl Microbiol Biotechnol 56:718–723 3. Verardi R et al (2012) Isotope labeling for solution and solid-state NMR spectroscopy of membrane proteins. Adv Exp Med Biol 992:35–62 4. Betton JM (2004) High throughput cloning and expression strategies for protein production. Biochimie 86:601–605 5. Schlegel S et al (2010) Revolutionizing membrane protein overexpression in bacteria. Microb Biotechnol 3(4):403–411 6. Schlegel S et al (2014) Bacterial-based membrane protein production. Biochim Biophys Acta 1843(2014):1739–1749 7. Costa S et al (2014) Fusion tags for protein solubility, purification, and immunogenicity in

Escherichia coli: the novel Fh8 system. Front Microbiol 5:63 8. Mouillac B, Baneres JL (2010) Mammalian membrane receptors expression as inclusiob bodies. Methods Mol Biol 601:39–48 9. Baneres JL et al (2011) New advances in production and functional folding of G-proteincoupled receptors. Trends Biotechnol 29 (7):314–322 10. Schwarz D et al (2007) Preparative scale expression of membrane proteins in Escherichia coli-based continuous exchange cell-free systems. Nat Protoc 2(11):2945–2957 11. Lacapere JJ et al (2014) Structural studies of TSPO, a mitochondrial membrane protein. In: Muss-Veteau I (ed) Membrane proteins production for structural studies. Springer, New York, pp 393–421 12. Robert JC, Lacapere JJ (2010) Bacterial overexpressed membrane proteins: an example: the TSPO. Methods Mol Biol 654:29–45

Recombinant mTSPO1 and hTSPO2 13. Bernhard F, Tozawa Y (2013) Cell-free expression—making a mark. Curr Opin Struct Biol 23:374–380 14. Cymer F et al (2015) Mechanisms of integral membrane protein insertion and folding. J Mol Biol 427:999–1022 15. Ulmschneider MB et al (2014) Spontaneous transmembrane helix insertion thermodynamically mimics trasnlocon-guided insertion. Nat Commun 5:4863 16. Friehs K (2004) Plasmid copy number and plasmid stability. Adv Biochem Eng Biotechnol 86:47–82 17. Lee KM et al (1995) A novel method for selective isotope labeling of bacterially expressed proteins. J Biomol NMR 5:93–96 18. Kuderova A et al (1999) Use of rifampicin in T7 RNA polymerase-driven expression of a

25

plant enzyme: rifampicin improves yield and assembly. Protein Expr Purif 16:405–409 19. Almeida FC et al (2001) Selectively labeling the heterologous protein in Escherichia coli for NMR studies: a strategy to speed up NMR spectroscopy. J Magn Reson 148:142–146 20. Cruzeiro-Silva C et al (2006) In-cell NMR spectroscopy—inhibition of autologous protein expression reduces Escherichia coli Lysis. Cell Biochem Biophys 44:497–502 21. Baker LA et al (2015) Efficient cellular solidstate NMR of membrane proteins by targeted protein labeling. J Biomol NMR 62:199–208 22. Bolanos-Garcia VM, Davies OR (2006) Structural analysis and classification of native proteins from E. coli commonly co-purified by immobilized metal affinity chromatography. Biochim Biophys Acta 1760:1304–1313

Chapter 2 Functional Assembly of Soluble and Membrane Recombinant Proteins of Mammalian NADPH Oxidase Complex Hajer Souabni, Aymen Ezzine, Tania Bizouarn, and Laura Baciou Abstract Activation of phagocyte cells from an innate immune system is associated with a massive consumption of molecular oxygen to generate highly reactive oxygen species (ROS) as microbial weapons. This is achieved by a multiprotein complex, the so-called NADPH oxidase. The activity of phagocyte NADPH oxidase relies on an assembly of more than five proteins, among them the membrane heterodimer named flavocytochrome b558 (Cytb558), constituted by the tight association of the gp91phox (also named Nox2) and p22phox proteins. The Cytb558 is the membrane catalytic core of the NADPH oxidase complex, through which the reducing equivalent provided by NADPH is transferred via the associated prosthetic groups (one flavin and two hemes) to reduce dioxygen into superoxide anion. The other major proteins (p47phox, p67phox, p40phox, Rac) requisite for the complex activity are cytosolic proteins. Thus, the NADPH oxidase functioning relies on a synergic multi-partner assembly that in vivo can be hardly studied at the molecular level due to the cell complexity. Thus, a cell-free assay method has been developed to study the NADPH oxidase activity that allows measuring and eventually quantifying the ROS generation based on optical techniques following reduction of cytochrome c. This setup is a valuable tool for the identification of protein interactions, of crucial components and additives for a functional enzyme. Recently, this method was improved by the engineering and the production of a complete recombinant NADPH oxidase complex using the combination of purified proteins expressed in bacterial and yeast host cells. The reconstitution into artificial membrane leads to a fully controllable system that permits fine functional studies. Key words Cell-free assays, NADPH oxidase, Nox, Superoxide anion, Cytochrome b558, ferricytochrome c, gp91phox, Membrane protein complex, Recombinant protein, Pichia pastoris, Arachidonic acid, Liposomes

1

Introduction The NADPH oxidase complex is well known as the source for the massive superoxide radical production in phagocytes. It is formed from the assembly of at least five cytosolic and membrane proteins which produces superoxide from the oxidation of NADPH. The key membrane-associated component of the NADPH oxidase is the flavocytochrome b558 heterodimer composed by a 91-kDa

Jean-Jacques Lacapere (ed.), Membrane Protein Structure and Function Characterization: Methods and Protocols, Methods in Molecular Biology, vol. 1635, DOI 10.1007/978-1-4939-7151-0_2, © Springer Science+Business Media LLC 2017

27

28

Hajer Souabni et al.

glycoprotein, gp91phox (also called Nox2) and a 22-kDa protein (p22phox). Exposure to microorganisms or inflammatory mediators activates these cells and through multiple signaling pathways, cytosolic subunits (p67phox, p47phox, small G protein Rac1/2, and p40phox) translocate to the membrane from cytosol and form the functional complex. The functioning of the NADPH oxidase complex in vivo is difficult to decipher due to the complexity of cell signaling pathways and the impossibility of controlling the presence of the partners in function of time. Although the comparison of results obtained in vitro and in vivo may not be straightforward, the alternative way was to develop the so-called cell-free assay that enables in vitro studies of the events occurring during NADPH oxidase activation and activity giving detailed information on each step of assembly and on the reaction that could not be easily observed in vivo. The method of cell-free activation of NADPH oxidase was discovered in Pick’s laboratory and was first described in 1984. This approach consists of purified membranes containing native Cytb558 (isolated from bovine or human neutrophils) to which the cytosolic fraction (from neutrophils) is added in the presence of suitable additives. This first system evolved nowadays into mainly three cell-free assay systems in which the major difference resides in the nature of the membrane component: (1) the semi-recombinant cell-free assay: the most frequent “canonical” cellfree system in which the membrane components are the purified membranes isolated from bovine or human neutrophils but the cytosol fraction is replaced by purified cytosolic proteins (for review see [1]); (2) the recombinant cell-free system in which the neutrophil membranes are replaced by yeast (Pichia pastoris) membranes containing the recombinant rCytb558; (3) the artificial cell-free assay in which the recombinant or native Cytb558 is reconstituted into lipid vesicles. In all cases, the cytosolic proteins are routinely produced in E. coli, purified [2] and added as desired to permit a strict quantification of each protein. NADPH is added and the measurement of the generated superoxide anions rate by the assembled NADPH oxidase complex is followed by SOD-inhibited cytochrome c reduction (see Note 1). The bottleneck to produce functional recombinant rCytb558 (gp91phox/p22phox) was overcome by using the methylotrophic yeast Pichia pastoris expression system that has been shown to be a valuable tool to produce the recombinant bovine heterodimer [3]. The Pichia pastoris expression system is capable of making the major eukaryotic posttranslational modifications in the overexpression of heterologous proteins. The maturation (glycosylation form of gp91phox), the heterodimer formation, and the prosthetic group association to the recombinant proteins produced in yeast membranes have been confirmed. Since then, many yeast clones were generated by using different vector constructions that express the heterodimer rCytb558 with different tag positions and also the gp91phox protein as a monomer.

NADH Oxidase Functional Assembly

29

The flexibility brought by the molecular biology combined to yeast expression system and the cell-free assays allowed investigations of the functioning of improbable component assembly. It has been shown, in particular, the capacity of gp91phox to produce superoxide in the absence of its membrane partner p22phox [4]. Collected information on the maturation and the activity of the recombinant gp91phox in the absence of p22phox allowed proposing a role of p22phox in the complex stabilization, its absence leading to an unconventional way of NADPH oxidase assembly. The protocols to obtain the “routine” semi-recombinant cell-free assay system have been described previously in detail (see in [5] and updated in [6]). In this chapter, we will focus on the methods to elaborate the “entirely recombinant” cell-free assay system which mix recombinant cytosolic proteins (produced in E. coli) with recombinant membrane oxidase proteins (produced in yeast) embedded into the host yeast membranes (plasma or subcellular organelle membranes from P. pastoris) or into artificial membranes.

2

Materials

2.1 Cells, Media, and Buffers

1. Restriction enzymes and other DNA-modifying enzymes (New England Biolabs, United Kingdom). 2. Reagents for DNA extraction and purification were from VWR International LLC (USA) and primers (Eurogentec, Angers, France). 3. Pichia pastoris strains X33 and SMD1168 and pAO815 and pPICZαA expression vectors (Invitrogen, Grand Island, NY, USA). 4. NEB5α bacterial strain for plasmid preparation and cloning (New England Biolabs, United Kingdom). 5. The rich medium YPD for growing yeast under nonselective conditions [1% (w/v) yeast extract, 2% (w/v) peptone, 2% dextrose and 2% agar if making solid medium]. 6. 10
YNB (13.4% Yeast Nitrogen Base with Ammonium Sulfate without amino acids): solubilize 34 g of YNB without ammonium sulfate and amino acids and 100 g of ammonium sulfate in 1000 mL of water and filter sterilize. Heat the solution to dissolve YNB completely. Store at 4  C. 7. BMGY and BMMY (Buffered Glycerol-complex Medium, Buffered Methanol-complex Medium): dissolve 10 g of yeast extract, 20 g peptone in 700 mL water. Autoclave 20 min. Cool to room temperature, add the following and mix well: 100 mL 1 M potassium phosphate buffer, pH 6.0, 100 mL 10
YNB, 2 mL 0.02% biotin, 100 mL 20% glycerol. For BMMY, add 10 mL 100% methanol (added just before use) instead of glycerol. Store media at 4  C.

30

Hajer Souabni et al.

2.2 For Heterologous Expression of Recombinant Membrane Proteins in Pichia pastoris Cells

1. P. pastoris X33 and SMD1168 host strain colonies grown on YPD agar plates (max. 2–3 days old). 2. Pichia EasyComp transformation kit from Invitrogen (Grand Island, NY, USA). 3. Shaking incubator (250 rpm). 4. 10
Dextrose solution: Dissolve 200 g D-glucose into 1 L water and filter sterilize on a 0.2 μm membrane. 5. YPD medium: dissolve 10 g yeast extract and 20 g peptone into 900 mL water. 6. 10
YNB: dissolve 134 g yeast nitrogen base (YNB) with ammonium sulfate (100 g/L) w/o amino acids into 1 L water and filter sterilize. 7. 500
Biotin solution: dissolve 20 mg biotin in 100 mL water, filter sterilized. 8. Breaking buffer: 50 mM sodium phosphate, pH 7.4, 1 mM PMSF or other protease inhibitors, 1 mM EDTA, and 5% glycerol. For 1 liter, dissolve 6 g sodium phosphate (monobasic), 372 mg EDTA, and 50 mL glycerol in 900 mL deionized water. Use NaOH to adjust pH and bring up the volume to 1 L. Store at 4  C. Right before use, add the protease inhibitors. 9. Resuspension buffer: 50 mM Tris–HCl (pH 8), 120 mM NaCl, 10% glycerol, 1 mM PMSF. For 1 L, dissolve 6 g Tris-base, 7 g NaCl, 100 mL glycerol in 900 mL deionized water. Use HCl to adjust pH and bring up the volume to 1 L. Store at 4  C. Right before use, add the protease inhibitors.

2.3 For Preparation of Plasma-Enriched Membranes from P. pastoris Cells

To prepare the sucrose gradient: make solutions of 20 mM Tris–HCl pH 7,4, 1 mM EDTA, 1 mM NaN3 supplemented with 10, 20, 40 or 60% sucrose. For 100 mL, mix 2 mL of 1 M Tris–HCl solution pH 7.4, 2 mL of 50 mM EDTA solution, 65 mg NaN3 and 10, 20, 40 or 60 g sucrose, bring up the volume to 100 mL. Store at 4  C.

2.4 For the Purification of the Recombinant Membrane Proteins

20% n-Dodecyl β-D-maltoside (DDM): dissolve 2 g DDM in 10 mL deionized water. Store at 20  C.

2.4.1 For rCytb558 (Flag-gp91phox/p22phox) Purification

1. Heparin Sepharose 6 Fast Flow column (GE healthcare, France): 20 mL of resin casted into a 20 mL column (20 mm diameter) equilibrated with the purification buffer: 20 mM Tris–HCl, pH 8, 1 mM EDTA, 0.025% DDM. 2. Sephadex 200 column (GE healthcare, France): 20 mL of resin was casted into 20 mL column (10 mm diameter) and equilibrated with the buffer for size exclusion column: 20 mM Tris–HCl, pH 8, 120 mM NaCl, and 0.025% DDM.

NADH Oxidase Functional Assembly

31

2.4.2 For His-gp91phox Purification

Buffer A (Ni-NTA equilibrating buffer): 30 mM sodium phosphate (pH 7.5), 0.5 M NaCl, 5 mM Imidazole, and 0.025% DDM. Buffer B: 30 mM sodium phosphate (pH 7.5), 0.5 M NaCl and applying a linear gradient of imidazole (100 mM). Buffer C: 20 mM Tris–HCl (pH 7.5), 50 mM NaCl and 0.025% DDM.

2.5 For the Relipidation of the Recombinant Membrane Proteins

1. Phosphate buffer solution (PBS, pH 7.4): dissolve one tablet (Fisher scientific) in 200 mL of deionized water yielding to 0.01 M phosphate buffer, 2.7 mM potassium chloride, and 0.137 M sodium chloride, pH 7.4, at 25  C. 2. A stock solution of 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) from Avanti Polar Lipids at 100 mg/mL. 3. A stock solution of 1,2-di-(9Z-octadecenoyl)-sn-glycero3-phosphocholine (DOPC) from Avanti Polar Lipids at 100 mg/mL. 4. Bio-Beads SM2 (Biorad, Life Science, Marnes-la-Coquette, France).

2.6 For Immunoblotting Test

2.7 For Cell-Free Assays

Blocking buffer: 5% Bovine Serum Albumin in Tris-buffered saline (TBS) solution (50 mM Tris–HCl (pH 7.5) and 150 mM NaCl) containing 0.1% Tween 20. Anti-His antibody conjugated with horseradish peroxidase (Clontech Inc., France), rabbit anti-gp91 (54.1) and mouse anti-p22 (FL-195) antibodies (Santa Cruz Biotechnology, Cliniscience, France); anti-rabbit NA934V and anti-mouse NA931 (GE Healthcare, France). Western Blotting Detection Kit (Amersham Biosciences, France). 1. Activity assay buffer: 20 mM Tris–HCl pH 8.0, 1 mM EDTA, and 120 mM NaCl (see Note 2). 2. Cis-unsaturated arachidonic acid (C20:4) (AA) 20 mg/mL: dissolve on ice 10 mg of AA in 500 μl absolute ethanol. Aliquot solution and stock at 80  C (see Note 3). 3. NADPH 20 mM: dissolve 4.167 mg NADPH in 250 μl activity assay buffer (see Note 2). 4. Equine cytochrome c 5 mM: dissolve 18.75 mg Cyt c in 300 μl activity assay buffer. 5. Superoxide dismutase (SOD) 2 mg/mL: dissolve 1 mg SOD in 0.5 mL activity assay buffer.

2.8

Equipment

1. Double beam spectrophotometer (e.g., Uvicon XS Secoman). 2. Refrigerated centrifuge and Ultracentrifuge (e.g., Beckman ultracentrifuges).

32

Hajer Souabni et al.

3. Swinging and fixed angle rotor (e.g., SW 32 Ti and 70 Ti rotors). 4. Orbital shaker incubators. 5. Automatic vortex Bead Beater (e.g., Biosep product) (see Note 4). ¨ kta prime system). 6. Chromatography system (e.g., A

3

Methods The classical cell-free assay is a semi-recombinant cell-free assay since the cytosolic proteins used are routinely purified recombinant proteins expressed in E. coli while Cytb558 is located in isolated membranes from neutrophils. In the unconventional system developed here, the oxidase membrane proteins are issued from heterologous expression in transgenic yeast. We describe here the preparation of membrane fractions from transgenic yeasts expressing recombinant bovine membrane proteins of the NADPH oxidase. The use of membrane fraction from transgenic yeasts in oxidase cell-free assays is possible since the superoxide anion production is catalyzed only by the heterologous NADPH oxidase proteins, no NADPH-dependent superoxide being detected in membranes from nonexpressing recombinant cytb558 yeast cells.

3.1 Expression of gp91phox or rCytb558 (gp91phox/p22phox) in Pichia pastoris Cells

To produce the membrane fraction containing the recombinant heterodimer rCytb558 (gp91phox/p22phox), P. pastoris SMD1168 strain was transformed with the pAO815 (ampicillin resistant) expression vector containing the entire expression cassette with the coding genes for both proteins gp91phox (CYBB) and p22phox (CYBBA), with a fusion tag (FLAG tag) at the N-terminus of gp91phox, as described in [3]. To produce the gp91phox monomer, P. pastoris X33 strain was transformed by electroporation with the pPICZαA (zeocin resistant) expression vector containing the gene coding for the gp91phox protein flanked with the Hisx6 fusion tag at the N-terminus [4]. The vector previously linearized is then inserted into the yeast genome by homolog recombination. Expression procedure: 1. Spread the recombinant yeast cells on a YPD plate containing 100 μg/mL of zeocin (for X33 strain harboring pPICZα/ gp91phox vector or 0.05 mg/mL ampicillin (for SMD1168 strain harboring the pAO815/gp91phox/p22phox vector) and incubate at 30  C during 3 days. 2. Inoculate BMGY medium supplemented with appropriate antibiotic with transgenic yeast clone and keep culturing overnight at 30  C and shaking at 225 rpm in incubator shaker.

NADH Oxidase Functional Assembly

33

3. Pellet cells by centrifugation (10 min at 550
g) and resuspend in fresh BMMY medium (with appropriate antibiotic) until reaching OD600nm of 1.0. Cells are grown in baffled culture flasks (1.5 L) at 30  C with shaking at 200 rpm. 4. Add 1% methanol (v/v final concentration) every 24 h during 72 h to maintain the expression induction. Harvest cells by centrifugation at 2500
g for 15 min and store them at 80  C until use. 3.2 Preparation of the Yeast Total Membrane Fractions (tMF)

All steps must be done at 4  C. 1. Thaw and resuspend cells in breaking buffer at OD600nm ¼ 50–100. 2. Add an equivalent volume of glass beads (0.5 mm diameter) and transfer the mixture into the Bead Beater. Disrupt cells by alternating 30 s vortexing followed by cooling period (1 min 30 s) on ice bath during 20 min. 3. Separate the clear supernatant from the cell debris and glass beads by centrifugation at 500
g for 10 min at 4  C. 4. Collect the total membrane fraction (tMF) by ultracentrifugation at 130,000
g for 90 min. 5. Resuspend the pellet (tMF) in a resuspension buffer using potter homogenizer (see Note 5). 6. Determine the protein concentration by BCA (bicinchoninic acid assay) with bovine serum albumin as standard (see Note 6).

3.3 Preparation of the Yeast Subcellular Membrane Fractions

Although the total membrane fractions obtained as described above are often sufficient, we also may prepare and use subcellular membrane fractions of yeast cells to investigate more precisely the enzyme activity of the oxidase components along the biosynthesis processes. 1. In SW32 centrifuge tubes, gently lay 3.7 mL cold 60, 40 then 20% sucrose solutions. 2. Load 0.7 mL of total membrane fractions on the top of the discontinuous sucrose gradient. 3. Leave the centrifuge tubes at 4  C with minimum disturbance during 4 h to stabilize the gradient. 4. Spin in swinging bucket centrifuge for 17 h (can go longer) at 110,000
g in a SW32 Ti rotor at 4  C. 5. Collect carefully fractions of 700 μl from the top to the bottom of the gradient. 6. Analyze the fractions to identify the fractions containing plasma or endoplasmic reticulum membranes with immunoblots using specific antibodies (mouse anti-yeast endoplasmic reticulum (Dpm1p) or anti-ribophorin I and mouse anti-yeast plasma membrane Pma1 anti-bodies).

34

Hajer Souabni et al.

7. For further analysis, pellet the endoplasmic reticulum (ER) and plasma membrane (Pmb) fractions at 30,000
g and resuspend in 1 mM EDTA, 10% sucrose or glycerol, 20 mM Tris–HCl pH 8.0 and store at 20  C. 3.4 Immunoblotting Assay

For western blotting assay, usually 1 μg of proteins was used. 1. Separate proteins on SDS-PAGE using 10% acrylamide gels. 2. Electro-transfer proteins to nitrocellulose membrane (0.2 μm) at 100 V for 1 h. 3. Saturate nonspecific binding with blocking buffer (see Subheading 2) overnight at 4  C. 4. For the detection of the proteins different antibodies can be used. For His-gp91phox incubate with anti-His antibody conjugated to horseradish peroxidase diluted at 1:10,000 in TBS buffer. For Flag-gp91phox/p22phox incubate with the mouse anti-Flag monoclonal antibody (1:3000). 5. For specific detection, use rabbit anti-gp91 (54.1) and mouse anti-p22 (FL-195) antibodies at a dilution of 1:1000 and incubate for 2 h at 4  C. 6. Use anti-rabbit (NA934V) and anti-mouse (NA931) IgG monoclonal antibodies at a 1:15,000 dilution as secondary antibodies to detect the anti-gp91 and anti-p22 primary antibodies, respectively. 7. Use ECL Plus Advance Western Blotting Detection Kit for the revelation of the recombinant gp91phox (Fig. 1).

Heterodimer

Monomer

Flag-Gp91phox/p22phox

His-gp91phox

tMF

(A)

ER

Pmb

tMF

(C)

kDa 120 100 -

80kDa –

60kDa –

80 -

(B) 20 -

Fig. 1 Western blots analyses of the different membrane fractions of the heterodimer rCytb558 and the monomer His-gp91phox: (a) analysis with monoclonal anti-Flag; (b) polyclonal anti-p22 of the subcellular membrane fractions containing rCytb558. (c) Total membrane fraction (tMF) of monomer His-gp91phox analyzed with monoclonal anti-His. tFM: total membrane fraction, ER endoplasmic reticulum and Pmb plasma membrane

NADH Oxidase Functional Assembly

3.5 Solubilization and Purification of the Recombinant Membrane Proteins 3.5.1 Purification of the Recombinant rCytb558

35

All purification steps were achieved at 4  C and just before using buffers, add 1 mM PMSF.

1. To solubilize the rCytb558 or gp91phox proteins, dilute tFM at 5 mg/mL of total membrane protein in resuspension buffer, add 1.5% DDM (in a detergent/protein ratio of 3:1 (w/w)). 2. Incubate for 1 h at room temperature with gentle agitation. 3. Centrifuge at 130,000
g for 90 min at 4  C. Solubilized membrane proteins are recovered in the supernatant. 4. Load solubilized protein extracts onto Heparine Sepharose column (20 mL) equilibrated with 20 mM Tris–HCl, pH 8, 1 mM EDTA, 0.025% DDM. 5. Wash with 100 mL of the same buffer. 6. Elute rCytb558 with a NaCl gradient (0-1 M NaCl). Usually, rCytb558 is eluted at about 0.4 mM NaCl. 7. Perform spectra of eluted fractions at 280 nm to evaluate the total membrane proteins and difference spectra (400–650 nm) to determine the amount of rCytb558 (see Note 7). 8. Pool fractions containing rCytb558 and concentrate using Vivaspin filters (K10). 9. Then to obtain a homogenous sample use a Superdex 200 gel filtration column equilibrated with 20 mM Tris–HCl, pH 8.0, 120 mM NaCl, and 0.025% DDM.

3.5.2 Purification of Hisgp91phox with Ni-NTA Followed by Gel Filtration Chromatography

The recombinant His-gp91phox protein was purified using a column of nickel-nitrilotriacetic acid (Ni-NTA) superflow sepharose followed by gel filtration chromatography using S200 Sephadex resin. 1. Equilibrate the Ni-NTA column using 20 mL of buffer A (see Subheading 2). 2. Load sample with a flow rate of 1.5 mL/min. 3. Elute proteins with linear imidazole gradient (5–100 mM) using Buffer B (see Subheading 2). 4. Analyze eluted fractions by spectroscopic measurement by performing difference reduced-oxidized spectra to determine the cytochrome content (see Note 7). 5. Concentrate fractions and separate proteins on gel filtration chromatography using the buffer C (see Subheading 2) with a flow rate of 1.5 mL/min, as described in Subheading 3.5.1. 6. Check the purity of the protein by sodium dodecyl sulfate polyacrylamide gel electrophoresis (Fig. 2).

36

Hajer Souabni et al.

(A)

M

1

2

3

(B)

kDa 200 – 150 – kDa 100 75 –

M

His-gp91phox

100 80 60 -

50 – 37 –

Fig. 2 Purification and identification of the recombinant gp91phox subunit. (a) SDS-PAGE analysis. The total membrane fraction (25 μg; lane 1) was solubilized in 2% DDM and the recombinant gp91phox protein was purified on Ni-NTA column (lane 2) followed by gel filtration (15 μg; lane 3). Gels were stained with Coomassie blue. (b) After SDS-PAGE separation of 15 μg of purified protein and membrane transferring, specific identification by the monoclonal anti-gp91 antibody (54.1) from Santa Cruz Biotechnology. M: Magic Mark standard protein (Invitrogen) was used for molecular weight reference 3.5.3 Production of Liposomes and Reconstitution of Purified gp91phox and rCytb558 in Liposomes

For the reconstitution of membrane oxidase protein into liposomes, a lipid/protein ratio of 5:1 (w/w) was used. Vesicles formation was followed by absorbance at 400 nm. 1. Dilute 50 μl of DOPC or DMPC solution (100 mg/mL) four times in PBS. 2. Ultrasonicate lipid by brief pulses until the solution becomes clear (~20 min) preferentially in the presence of nitrogen to avoid lipid peroxydation. The optical density at 400 nm must decrease. 3. Add slowly 200 μl proteins (1 mg/mL). Stir micellar protein– lipid–detergent mixture for 30 min at 30  C with gentle agitation. 4. Remove DDM by adding 1% Bio-Beads SM2 (1 g of Bio-Beads for 100 mg of DDM) and incubate with gentle agitation for 1 h. Remove Bio-Beads with a brief centrifugation (1 min, 10,200
g) and place the stock solution containing the proteoliposomes (PL) at 20  C, until use.

3.5.4 The Cell-Free Assays

Cell-free assays can be used in very different configurations depending on the components that are mixed together. Here in addition to using a cell-free assay fully constituted with recombinant soluble and membrane proteins, the sophistication is to use also recombinant membrane component reconstituted into artificial phospholipid membranes. The cell-free assays comprise two phases. The first one consists in the oxidase complex assembly insofar as the cytosolic proteins are dispersed in the solution and will have to translocate to the membrane to form the active NADPH oxidase complex.

NADH Oxidase Functional Assembly

37

The second phase is the proper enzyme activity leading to the NADPH-dependent superoxide anion production. Numerous anionic amphiphils act as activators of the NADPH oxidase in cell-free system mimicking the cell signaling processes implying phosphorylation (kinase pathways) and arachidonic release (phospholipaseA2 activation) facilitating protein-protein interactions leading to complex assembly. The canonical arachidonic acid is routinely used in vitro although the effects on the NADPH oxidase components are still unclear (see Note 8). The catalytic activity is then visualized by following optically at 550 nm the reduction of ferricytochrome c trapping the produced superoxide. 1. Use 1–1.5 pmol of gp91phox or rCytb558 in tFM, ER or Pmb fractions or in PL. 2. Mix with 200 nM of each purified recombinant cytosolic factors (p67phox, p47phox, and RacQ61L (see Note 9)) and arachidonic acid at different concentrations in a total volume reaction of 50 μl activity assay buffer and incubate for 5 min at 25  C to allow the assembly of the complex. 3. Add 10 μl (5 mM) of ferricytochrome c to obtain a final concentration of 100 μM, adjust volume of the reaction to 490 μL with the activity assay buffer. 4. Initiate the reaction with the addition of 10 μl NADPH (20 mM) and rapidly perform measurements of the reduction rate of cytc at 550 nm (OD versus time). 5. Add, on the same reaction mixture, 200 U of superoxide dismutase (SOD) and measure the reduction rate at 550 nm to determine the nonspecific reduction of ferricytochrome c (Fig. 3 near here).

Abs 550 nm

0.08 0.06 0.04 0.02 0

0

0.1

0.2

0.3 time (min)

0.4

0.5

0.6

Fig. 3 SOD-sensitive NADPH-oxidase activity of total membrane fractions containing rCytb558. Measure in the absence (black square) or in the presence (empty circle) of 50 μg/mL SOD. NADPH oxidase activity assay was done at 25  C with optimal arachidonic acid concentration

38

Hajer Souabni et al.

Oxidase activity (mol O2.-/s/mol Cytb558)

300

200

100

0 PL-DMPC

PL-DOPC

tFM

Fig. 4 The “artificial cell-free assay.” Anion superoxide production was measured with total membrane fraction replaced with PL-DMPC or PL-DOPC

6. Calculate the amount of superoxide generated using the molar absorption coefficient of 21.1 mM1 cm1 and normalize to 1 pmole of gp91phox used into the sample. 7. The rate of superoxide production can be compared for different experimental conditions in particular regarding the membrane environment such as different membrane lipid composition (Fig. 4). 3.5.5 NADPH Oxidase Activation Specificities Arachidonic Acid Activation: Dose-Response Curve

The final concentration of AA causing maximal activation has to be optimized. Depending on the cell-free system considered, this optimal value is highly variable (even within the same species it can be different as we often observed from human membrane fractions depending on the blood donor). The differences are correlated to species variation (human, bovine, etc.) but the fundamental reasons lie behind important parameters such as different ratio of membrane protein/lipids and off course protein and lipid composition (see Note 10). The concentration-response curve that plots the enzyme activity vs the AA concentration has commonly a bell shape. Such a curve is a characteristic of phagocyte NADPHoxidase activation, indicating that the activation process through the activator molecule is similar to native enzyme. – Realize cell-free assays as described in Subheading 3.6 with membrane fractions (tMF, PL, etc.) with increasing concentration of AA to plot AA-dependent bell-shape activation curve: start measure in the absence of AA up to 1 mM. (Fig. 5a, b).

39

B

A

300

200

Proteoliposomes

180

mole O2.-/s/mole gp91phox

Oxidase activity (mol O2.-/s/mol Cytb558)

NADH Oxidase Functional Assembly

200

100

X33/gp91phox membranes

160 140 120 100 80 60 40 20

0

0

0.2

0.4

0.6

0.8

1

Arachidonic acid concentration (mM)

1.2

0

0

0

15

0 0 0 0 0 90 75 30 45 60 Arachidonic acid concentration (µM)

Fig. 5 Activation of the NADPH-oxidase complex by arachidonic acid. (a) Activity assays were performed in cell-free system using total membrane fraction of rCytb558 with different concentrations of arachidonic acid at 25  C. (b) NADPH oxidase activity of the reconstituted gp91phox in DOPC liposomes (~2.24 nM of gp91phox) and of gp91phox total membrane fraction. Measurements of the superoxide production by PLs were performed with the standard cyt c reduction assay in the absence of any cytosolic proteins and increasing concentrations of arachidonic acid. Results are presented as the mean  SD (n ¼ 3)

Cytosolic Protein Dependences of the NADPH Oxidase Activation

Performing the same experiment as in Subheading 3.5.4. (Fig. 3) in the absence or in the presence of cytosolic proteins or AA is an interesting control of activation stage. The implication of each component in the oxidase activity of rCytb558 or gp91phox alone can be investigated using different sets of experimental condition of the cell-free assays. The absence of one of the components impedes efficient superoxide production. 1. To show the importance of incubation time for complex assembly, mix membrane fractions without or with cytosolic proteins. 2. Then add AA, cytc and NADPH, complete to 500 μl with buffer, and measure the production of superoxide anions by following absorption at 550 nm (Fig. 6a) (see Note 11). The different oxidase activities of total membrane fraction containing rCytb558 can be measured in infinite experimental conditions. Here are compared the oxidase activities with or without cytosolic proteins and with and without AA (Fig. 6b).

40

Hajer Souabni et al.

A 0.5

Abs 550 nm

0.4 0.3 0.2 0.1 0

0

0.5

1 1.5 Time (min)

2

2.5

B 120

% NADPH oxidase acvity

100 80 60 40 20 0

+

+

+

+

tFM

+

-

-

-

cytosolic proteins

+

-

+

-

AA

Fig. 6 (a) Importance of complex assembly to produce anion superoxide in cellfree assay. NADPH oxidase activity was followed with (dark square) or without pre-incubation (cross) of rCytb558 with cytosolic proteins and in the presence of optimal concentration of AA. (b) rCytb558 activation with cytosolic protein and with arachidonic acid. Cell-free assays were done in the absence or presence of AA or cytosolic protein or both. NADPH oxidase measured in the presence of optimal concentration of AA and cytosolic protein is considered 100%. The value of this reference activity is 350 mol O2./s/mol rCytb558

4

Notes 1. The measurement of the generated superoxide anions followed by cytochrome c reduction is a classical technique that can be employed to examine the catalytic parameters of all enzymes implicated into oxygenated redox reactions.

NADH Oxidase Functional Assembly

41

2. It is best to prepare this fresh just before use. Alternatively, it can be dissolved in water for better stability. 3. Small aliquots volumes must be done to avoid oxidation of arachidonic acid. 4. For reproducibility, we modified the apparatus in order to monitor automatically the cycle of vortex/pause periods. 5. If tFM is used for the separation in sucrose gradient, solubilize in 10% sucrose solution. 6. About 6 mg proteins/L of yeast culture was usually obtained. 7. The amount of gp91phox or gp91phox/p22phox is spectrophotometrically determined from the difference absorption spectra of the redox protein in its reduced minus oxidized form. The reduced form is obtained by adding few grain of Na dithionite. The Cytb558 heme amount is quantified by measuring the absorbance difference between the peak at 427 nm and the hollow at 411 nm using a millimolar extinction coefficient of Δε427-411 ¼ 200 mM1 cm1 [7]. 8. The NADPH oxidase complex can also be directly and efficiently activated by other amphiphilic molecules, such as sodium dodecyl sulfate (SDS). Arachidonic acid (AA) is a well-known second messenger in signaling pathways which potentiate the NADPH oxidase complex in vitro. Direct impact of AA on gp91phox leads to conformational changes, which might participate in the NADPH oxidase activation [8, 9]. AA interacts also directly with the cytosolic proteins [10, 11] with recently new insights of the role of cis-AA into the interaction of p67phox-Rac with gp91phox [12]. It was proposed that anionic amphiphils perturb the intramolecular bonds in p47phox between the polybasic domain and the SH3 tandem mimicking the phosphorylation events [13]. 9. Purification protocols of the cytosolic proteins used are described in [2]. In the cell-free assays, the mutant of the Rac1 protein (Rac1Q61L) which is in its active GTP bound form is often used. It presents the advantage that GTP is not necessary to be added to obtained the active Rac form. 10. The partitioning of the amphiphilic molecules depends on the membranes composition and AA will have a more diluted effect in membranes containing less target proteins (cytb558). In the literature, depending on the human semi-recombinant cell-free system considered, the optimal AA concentration values range from 30 μM [11, 14, 15], to 200 μM [16]. With the bovine cell-free systems, the maximal O2 - production occurred at 150–200 μM of AA, with neutrophil bovine membrane fractions (containing in general >0.06 mg of Cytb558/mg of total membrane proteins), while this value reaches 750–800 μM l

42

Hajer Souabni et al.

with yeast membrane fraction containing the bovine recombinant proteins (< 0.01 mg of Cytb558/mg of total membrane proteins). A more diluted target (Cytb558) protein will need higher AA concentration for activation. However, the AA interactions with the NADPH oxidase components and cellular membrane are still open questions that need to be investigated more. 11. The purified gp91phox or gp91phox/p22phox in detergent showed almost no activity. When reconstituted into PL, gp91phox/p22phox still need to be activated by the cytosolic proteins while in the absence of p22phox, gp91phox produce superoxide in the absence of cytosolic subunits but remain AA-dependent.

Acknowledgments The authors want to acknowledge Dr. M-C Dagher for providing them with the cytosolic protein containing plasmids. This work was supported by the Agence Nationale pour la Recherche (ANR 2010-blan-1536-01). References 1. Dagher MC, Pick E (2007) Opening the black box: lessons from cell-free systems on the phagocyte NADPH-oxidase. Biochimie 89:1123–1132 2. Baciou L, Erard M, Dagher MC, Bizouarn T (2009) The cytosolic subunit p67phox of the NADPH-oxidase complex does not bind NADPH. FEBS Lett 583:3225–3229 3. Ostuni MA, Lamanuzzi LB, Bizouarn T, Dagher MC, Baciou L (2010) Expression of functional mammal flavocytochrome b(558) in yeast: comparison with improved insect cell system. Biochim Biophys Acta 1798:1179–1188 4. Ezzine A, Souabni H, Bizouarn T, Baciou L (2014) Recombinant form of mammalian gp91(phox) is active in the absence of p220 (phox). Biochem J 462:337–345 5. Molshanski-Mor S, Mizrahi A, Ugolev Y, Dahan I, Berdichevsky Y, Pick E (2007) Cellfree assays: the reductionist approach to the study of NADPH oxidase assembly, or “all you wanted to know about cell-free assays but did not dare to ask”. Methods Mol Biol 412:385–428. Humana Press, Totowa, NJ 6. Pick E (2014) Cell-free NADPH oxidase activation assays: “In Vitro Veritas”. Methods Mol Biol 1124:339–403. Humana Press

7. Light DR, Walsh C, O’Callagahn A, Goetzl E, Tauber A (1981) Characteristics of the cofactor requirements for the superoxide-generating NADPH oxidase of human polymorphonuclear leukocytes. Biochemistry 17:1468–1476 8. Doussiere J, Gaillard J, Vignais PV (1996) Electron transfer across the O-2() generating flavocytochrome b of neutrophils. Evidence for a transition from a low-spin state to a high-spin state of the heme iron component. Biochemistry 35:13400–13410 9. Souabni H, Thoma V, Bizouarn T, Chatgilialoglu C, Siafaka-Kapadai A, Baciou L, Ferreri C, Houee-Levin C, Ostuni MA (2012) Trans Arachidonic acid isomers inhibit NADPHoxidase activity by direct interaction with enzyme components. BBA-Biomembranes 1818:2314–2324 10. Shiose A, Sumimoto H (2000) Arachidonic acid and phosphorylation synergistically induce a conformational change of p47(phox) to activate the phagocyte NADPH oxidase. J Biol Chem 275:13793–13801 11. Swain SD, Helgerson SL, Davis AR, Nelson LK, Quinn MT (1997) Analysis of activationinduced conformational changes in p47(phox) using tryptophan fluorescence spectroscopy. J Biol Chem 272:29502–29510

NADH Oxidase Functional Assembly 12. Matono R, Miyano K, Kiyohara T, Sumimoto H (2014) Arachidonic acid induces direct interaction of the p67(phox)-Rac complex with the phagocyte oxidase Nox2, leading to superoxide production. J Biol Chem 289:24874–24884 13. Groemping Y, Lapouge K, Smerdon SJ, Rittinger K (2003) Molecular basis of phosphorylation-induced activation of the NADPH oxidase. Cell 113:343–355 14. Curnutte JT (1985) Activation of human neutrophil nicotinamide adenine-dinucleotide phosphate, reduced (triphosphopyridine nucleotide,

43

reduced) oxidase by arachidonic-acid in a cell-free system. J Clin Investig 75:1740–1743 15. Clark RA, Leidal KG, Pearson DW, Nauseef WM (1987) Nadph oxidase of humanneutrophils - subcellular-localization and characterization of an arachidonate-activatable superoxide-generating system. J Biol Chem 262:4065–4074 16. Ligeti E, Pizon V, Wittinghofer A, Gierschik P, Jakobs KH (1993) Gtpase activity of small Gtpbinding proteins in Hl-60 membranes is stimulated by Arachidonic-acid. Eur J Biochem 216:813–820

Chapter 3 Direct Extraction and Purification of Recombinant Membrane Proteins from Pichia pastoris Protoplasts Lucie Hartmann, Estelle Metzger, Noe´mie Ottelard, and Renaud Wagner Abstract In the past decade, the methylotrophic yeast Pichia pastoris has proved to be one of the most efficient systems for mass production of recombinant eukaryotic membrane proteins (MPs), leading to the crystallization and structure determination for a variety of them. The actual overexpression of functional MPs achieved with this system is, however, often accompanied by the formation of a variable but significant proportion of misfolded and/or aggregated proteins that are co-extracted and co-purified during the purification process. In order to minimize this unwanted phenomenon, we devised a novel procedure in which MPs produced in Pichia pastoris are directly solubilized from whole cells instead of crude membrane preparation. This approach aims at favoring the extraction of correctly folded membrane proteins that have been targeted to the plasma membrane, limiting the solubilization of the misfolded proteins and protein aggregates that are stored in internal membrane compartments. The method described herewith is based on the formation of protoplasts through enzymatic treatment prior to protein solubilization. This chapter details a set of protocols going from yeast cell preparation and protein solubilization to purification using affinity and size exclusion chromatography. Key words Pichia pastoris, Yeast, Membrane protein, Protoplast, Detergent solubilization, Purification

1

Introduction When coping with membrane protein (MP) solubilization from any cellular expression system, the most common approach consists in the preliminary preparation of the membrane fraction. This membrane enrichment step can be quite valuable during the downstream extraction and purification phases by reducing the amount of soluble proteases as well as potential soluble contaminants from the sample. However, membrane fractionation necessarily involves cell lysis procedures that could be more or less drastic. When harsh mechanical conditions are used, undesired and poorly controlled events such as heat and shear forces may occur that are often detrimental to the integrity of membrane proteins.

Jean-Jacques Lacapere (ed.), Membrane Protein Structure and Function Characterization: Methods and Protocols, Methods in Molecular Biology, vol. 1635, DOI 10.1007/978-1-4939-7151-0_3, © Springer Science+Business Media LLC 2017

45

46

Lucie Hartmann et al.

In addition, standard MP protocols usually recover the whole membrane fractions issuing from the different cellular compartments and containing a mixture of properly folded and misfolded proteins. Even if this phenomenon is still poorly characterized in the context of membrane protein production, it is actually widely recognized that overexpression overwhelms the cell biosynthesis and translocation machineries that often elicits improper MP folding. Such events cause a number of stresses and responses with various outcomes including the retention of misfolded proteins within the secretory pathway and the formation of protein aggregates before their eventual degradation [1, 2]. As a result, when preparing membranes for protein purification purposes, the following solubilization step may lead to the co-extraction of various heterogeneous forms of the protein of interest that are further difficult to separate. The protocol presented here proposes to limit these phenomena by performing a short solubilization step directly on whole cells in order to favor the extraction of the undamaged, correctly folded MPs that have been targeted to the plasma membrane. This method has been developed in our lab with the yeast Pichia pastoris, a robust and versatile system that we routinely use for the recombinant expression of various MPs [3–7]. Because this cellular host possesses a very thick and robust cell wall that is detergent-resistant, a preliminary enzymatic treatment is required to obtain yeast protoplasts (i.e., yeast cells devoid of their cell wall) before the detergent solubilization and purification steps. The whole procedure has been successfully applied to a variety of eukaryotic (mainly human) MPs recombinantly expressed in P. pastoris, encompassing GPCRs, ion channels, and other integral MPs. By comparison with the standard procedure where the same proteins were extracted from membrane preparations after cell lysis, the method not only proved to be much more rapid, but, most importantly, allowed significantly reducing the presence of protein aggregates after purification in several cases. Here, we illustrate this method with the solubilization and purification of the human adenosine A2A receptor extracted from yeast protoplasts. The starting biological material consists of yeast pellets obtained after the expression of a N-Flag-10HisHUMAN_AA2A-Biotag construct, following the conditions reported in [3]. A full description of the generation and selection of Pichia pastoris clones, as well as the subsequent procedures for recombinant expression, can be found in other recent and comprehensive chapters of the Methods in Molecular Biology series [8, 9]. The present protocol reports the enzymatic treatment of yeast cells expressing recombinant protein to obtain protoplasts, which are then directly subjected to detergent solubilization (Fig. 1). A subsequent two-step purification procedure (Ion Metal Affinity

Extraction of MPs from Yeast Protoplasts FL A

G %

H

47

IS%

FLA G% H IS %

FL

AG

%

H IS%

FL

Recombinant yeast cells

A G%

FLA G % H I S%

HIS%

FLA G %

HI S%

Successive washes H2O, SED, Sorbitol CG buffer resuspension Zymolyase treatment 15 min at RT FL

AG

%

H IS

%

G F LA IS % H %

FL

AG

%

H IS%

FL

Yeast protoplasts

AG% H

F LA G % H IS%

IS

F LAG % H I S %

Detergent solubilization Solubilization buffer 30 min at RT

Ultracentrifugation 100 000 g, 30 min, 4°C Supernatant recovery FLA G

%

HIS %

Solubilized proteins to be puriƒied

Fig. 1 Protoplasting and protein solubilization procedure. Recombinant yeast cell pellet is first submitted to Zymolyase® treatment for cell wall degradation. The protoplasts formed are then directly used for protein solubilization using detergents. An ultracentrifugation step enables to recover the solubilized fraction that will undergo the purification process

Chromatography and Size Exclusion Chromatography) is also presented to illustrate the yield and purity of the MPs that can be isolated with this method.

2

Materials

2.1 Zymolyase® Treatment

1. Recombinant cell pellet. 2. Milli-Q water. 3. SED buffer: 1 M sorbitol, 25 mM ethylenediaminetetraacetic acid (EDTA), 1 M dithiothreitol (DTT) (added extemporaneously).

48

Lucie Hartmann et al.

4. 1 M sorbitol. 5. CG buffer: 20 mM trisodium citrate, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride (PMSF) (added extemporaneously). Adjust the pH to 5.8 with hydrochloric acid. 6. Zymolyase® 20 T (Amsbio, UK). It can be resuspended in water at 200 U/mL and stored at 20  C for a few months. 2.2 Protein Solubilization

1. Solubilization buffer: 50 mM Tris–HCl pH 7.4, 500 mM NaCl, 10% glycerol, 20 mM imidazole (see Note 1), detergent (see Note 2), inhibitor protease cocktail (added extemporaneously, e.g., cOmplete™ EDTA-free, Roche, France). 2. Ultracentrifuge equipped with an appropriate fixed-angle rotor and adapted polycarbonate tubes.

2.3 Purification of Solubilized Material

1. Automated protein purification system (AKTA Purifier, GE Healthcare, or equivalent). A sample pump or a superloop is required for the affinity chromatography step. 2. 1 mL prepacked nickel affinity chromatography column (e.g., HisTrap HP 1 mL, GE Healthcare). 3. 0.22 μM syringe filter. 4. Buffer A: 50 mM Tris–HCl pH 7.4, 500 mM NaCl, 20 mM imidazole, detergent (see Note 2). 5. Buffer B: 50 mM Tris–HCl pH 7.4, 500 mM NaCl, 500 mM imidazole, detergent (see Note 2). 6. Centrifugal protein concentrator (e.g., Vivaspin 6, Sartorius, cutoff adapted to the protein of interest). 7. Size Exclusion Chromatography (SEC) column (e.g., Superdex 200 Increase 10/300, GE Healthcare). 8. SEC running buffer: 50 mM Tris–HCl pH 7.4, 150 mM NaCl, detergent (see Note 2).

2.4 Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE)

1. 40% Acrylamide/Bis-acrylamide solution, 29:1. 2. 3 M Tris–HCl pH 8.45, 0.3% (w/v) SDS. 3. 80% (v/v) glycerol. 4. 10% (w/v) ammonium persulfate (APS). 5. Tetramethylethylenediamine (TEMED). 6. Gel casting stand and electrophoresis chamber (e.g., Mini-PROTEAN system, Bio-Rad). 7. Membrane preparation samples. 8. Tris-Tricine-SDS cathode running buffer: 1 M Tris–HCl pH 8.2, 1 M Tricine, 1% (w/v) SDS. 9. Tris anode running buffer: 1 M Tris–HCl pH 8.9.

Extraction of MPs from Yeast Protoplasts

49

10. 2 Tricine Sample Buffer (SB 2): 100 mM Tris–HCl pH 6.8, 25% (v/v) glycerol, 8% (w/v) SDS, 0.02% (w/v) Coomasie blue G250, 200 mM DTT. 2.5 Protein Transfer and Western Blot Immunodetection

1. Tris-Glycine transfer buffer: 25 mM Tris, 200 mM glycine, 0.02% (w/v) SDS, 20% (v/v) ethanol. 2. 0.45 μm nitrocellulose blotting membrane. 3. Whatman paper. 4. Electroblotting system (e.g., Mini Trans-Blot Cell, Bio-Rad). 5. Phosphate-buffered saline (PBS). 6. PBS containing 0.02% (v/v) Tween 80 (PBST). 7. Blocking buffer: PBST with 5% (w/v) nonfat dry milk. 8. Primary anti-tag or anti-protein antibody (e.g., monoclonal anti-FLAG antibody from mouse, Sigma). 9. Secondary anti-mouse IgG antibody linked to a reporter system (traditionally HRP-conjugated antibody, here an IRDye 800-coupled antibody). 10. Reagents and detection device adapted to the reporter system selected. 11. Orbital shaker.

3

Methods

3.1 Protoplasting: Zymolyase® Treatment of Recombinant Cell Paste

3.1.1 Cell Paste Preparation

Zymolyase® is a mixture of several carbohydrolases (i.e., ß-1,3 glucan laminaripentaohydrolase, ß-1,3 glucanase, protease and mannanase) extracted from Arthrobacter luteus. These enzymes work in conjunction to degrade efficiently the cell wall of a large spectrum of yeasts, thus leading to the formation of yeast protoplasts. A number of other enzymes for yeast cell wall degradation are commercially available but Zymolyase® has proved to be the most efficient in our hands for this protocol with Pichia pastoris. The following proportions are given for 3 g of wet cell paste, obtained from about 250 mL of Pichia pastoris culture. 1. Resuspend the cell pellet in 50 mL Milli-Q H2O. 2. Harvest the cells by centrifuging for 5 min at 5000  g, 4  C. 3. Discard the supernatant and wash the cell pellet with 50 mL SED buffer. 4. Harvest the cells by centrifuging for 5 min at 5000  g, 4  C. 5. Discard the supernatant and wash the cell pellet with 50 mL 1 M sorbitol. 6. Harvest the cells by centrifuging for 5 min at 5000  g, 4  C.

50

Lucie Hartmann et al.

7. Discard the supernatant and resuspend the cell pellet in 37.5 mL of CG buffer, in order to reach a cell paste concentration of about 80 g/L. 3.1.2 Enzymatic Cell Wall Degradation

1. Add 20 U of Zymolyase® per gram of cell paste initially prepared. 2. Incubate for 15 min (see Note 3) at room temperature under gentle agitation (see Note 4). 3. Harvest the protoplasts formed by a soft centrifugation for 5 min at 750  g, room temperature. In order to avoid protoplast lysis, reduce the acceleration and deceleration speed of the centrifuge. 4. Carefully discard the supernatant using a pipette.

3.2 Protein Solubilization

As for every MP solubilization, the successful selection of the best detergent to use is a fine-tuned combination between the extraction efficiency and the preservation of the protein fold and function. It both depends on the protein of interest and the expression system. The choice of the detergent, here DDM for AA2A, has thus to be adjusted accordingly (see Note 2). 1. Gently resuspend the pellet in 37.5 mL of solubilization buffer (no vortexing). The cell paste concentration should be close to 80 g/L. 2. Incubate for 30 min at room temperature under gentle agitation. 3. Separate the solubilized fraction by ultracentrifuging for 30 min at 100,000  g and 4  C. Store 20 μL of the suspension before ultracentrifugation at 4  C for SDS-PAGE analysis. 4. After ultracentrifugation, recover the supernatant containing the solubilized fraction. Store 20 μL at 4  C for SDS-PAGE analysis. 5. The ultracentrifuge pellet can also be analyzed by SDS-PAGE after resuspension in 37.5 mL of solubilization buffer without detergents, imidazole, and protease inhibitor cocktail. Use a glass Teflon homogenizer to obtain a homogeneous suspension.

3.3 Purification of Solubilized Material 3.3.1 Nickel Affinity Chromatography

The following protocol is designed for automated purification (e.g., using AKTA protein purification systems). It can however be easily adapted for batch/gravity flow purification. The whole purification process can be performed at a flow rate of 1 mL/min. 1. Filter the ultracentrifuge supernatant using a 0.22 μm syringe filter. 2. Add imidazole at a final concentration of 20 mM.

Extraction of MPs from Yeast Protoplasts

51

3. Equilibrate a 1 mL prepacked nickel affinity column with buffer A. 4. Inject the filtrated supernatant onto the column. Store 20 μL of the injected sample at 4  C for SDS-PAGE analysis. 5. Wash with 20 mL of buffer A, or until the 280 nm-absorbance signal returns to the baseline. 6. Wash for 10 mL with 20% buffer B (corresponding to about 100 mM imidazole) (see Note 5). 7. Elute the protein of interest either with a gradient of buffer B or with a step of 100% buffer B (see Note 6). Collect 0.5 mL elution fractions. 8. Pool the fractions representative of one elution peak and concentrate them, using a centrifugal concentrator, to a volume suited for injection onto the size exclusion chromatography column (according to the manufacturer’s recommendations, typically a maximum of 500 μL for a 24 mL column). For concentration, choose a cutoff adapted to the size of the protein of interest (typically 30 or 50 kDa MWCO). 9. Store 20 μL of the concentrated pool of interest and all other representative elution fractions (flow through and wash fractions) at 4  C for SDS-PAGE analysis. 3.3.2 Size Exclusion Chromatography (SEC)

The following protocol is designed for automated purification (e.g., using AKTA protein purification systems). Superdex 200 10/300 SEC column or equivalent is generally well suited for small-scale preparative purification of MPs between 10 kDa and 600 kDa. 1. Equilibrate the SEC column with at least 1.5 column-volume (CV) of running buffer. 2. Inject the concentrated sample. Set the flow rate according to the manufacturer’s recommendations. Collect 0.5 mL elution fractions. 3. Store 20 μL of each peak-representative fraction at 4  C for SDS-PAGE analysis. Figure 2 shows a typical SEC elution profile obtained during the purification of the adenosine A2A receptor using the procedure described above. SEC enables the separation of oligomeric objects and contaminants from monomeric receptor of interest. Oligomeric and/or aggregated objects are eluted close to the void volume of the column (corresponding to peak 1) and correspond to the inactive protein. The monomeric and ligand-binding active receptor is mainly eluted in peak 2. In typical similar experiments performed on proteins extracted from whole membrane fractions after cell lysis (standard procedure), peak 1 has systematically about

52

Lucie Hartmann et al.

mAU

2

1

25.0

20.0

3

15.0

10.0

5.0

0.0 0.0

5.0

10.0

15.0

20.0 ml

Fig. 2 Size exclusion chromatography elution profile of the adenosine A2A receptor. Profile obtained on a Superdex 200 Increase 10/300 GL SEC column with absorbance measured at 280 nm. In this case, SEC enables the separation of 3 main peaks named respectively 1, 2, and 3

3–5 times the intensity of peak 2 (data not shown). These data exemplify how this method may be highly beneficial to reduce the amount of aggregated MPs during extraction. In terms of yield, the protocol described here enables to recover about 1.5 mg of monomeric adenosine A2A receptor from 1 L of yeast culture. However, protein extraction from yeast protoplasts does not significantly improve the final amount of active purified receptors compared to more classical methods. Fractions representative of each elution peak (respectively 1, 2, and 3) have to be analyzed using biochemical methods in order to assess their protein content. In this chapter, we describe SDS-PAGE analyses, but functional tests such as radioligand binding assay may be also performed if available for the protein of interest.

Extraction of MPs from Yeast Protoplasts

3.4 Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE) 3.4.1 10% SDS Tricine Polyacrylamide Gel Preparation

53

The following proportions are given for casting two gels of 1 mm thickness in a standard Mini-Protean system from Bio-Rad. 1. Prepare the separating gel by mixing 2.5 mL of acrylamide solution, 3.3 mL of Tris–HCl SDS buffer, 1.25 mL of 80% (v/v) glycerol, 2.9 mL H2O. Add 90 μL of APS and 6 μL of TEMED, mix, and immediately cast the gel. Allow space for the stacking gel. 2. Prepare the stacking gel by mixing 0.6 mL of acrylamide solution, 1.6 mL of Tris–HCl SDS buffer, 4.1 mL H2O. Add 90 μL of APS and 6 μL of TEMED, mix, and cast carefully over the separating gel. The presence of glycerol in the separating part avoids the need to wait for its complete polymerization before casting the stacking gel. Insert a 10-well gel comb immediately without introducing air bubbles. 3. Let the gel polymerize for about 30 min.

3.4.2 Gel Electrophoresis and Coomassie Staining

1. Incubate 20 μL of the elution fractions of interest in 20 μL of SB 2 for about 10 min at room temperature (see Note 7). 2. Load in parallel two 10% SDS-polyacrylamide gels with respectively 20 μL (first gel) and 10 μL (second gel) of the sample in its sample buffer. The first gel will be used for total protein staining with Coomassie blue, the second one will be transferred on a nitrocellulose membrane for immunodetection of the protein of interest. 3. Proceed to electrophoresis using Tris-Tricine-SDS cathode running buffer and Tris anode running buffer in a tank unit. Run for about 1 h 30 min at 100 V. 4. Stain the first gel with Coomassie blue following the manufacturer’s instructions. Figure 3a presents a typical Commassie blue SDS-PAGE analysis obtained for adenosine A2A receptor purification. A number of contaminant proteins are still present in the IMAC elution pool but SEC enables to obtain highly purified proteins.

3.5 Protein Transfer and Western Blot Immunodetection

1. Transfer the proteins from the second gel to a nitrocellulose membrane by electroblotting in Tris-glycine transfer buffer for about 1 h 30 min at 100 V. An ice pack can be added into the electroblotting device to mitigate the heat produced. 2. Incubate the membrane in 50 mL blocking buffer for 1 h at room temperature on an orbital shaker. Alternatively, incubate the membrane overnight at 4  C. 3. Remove the blocking solution and incubate the membrane with the selected antibody diluted in blocking buffer (for instance a monoclonal anti-flag antibody at a final concentration of 0.1 μg/mL) for 1 h at room temperature on an orbital shaker.

54

Lucie Hartmann et al.

A. Coomassie staining P

S

FT

E

IMAC W1 W2

1

SEC 2

MW (kDa) 170 100 70

3

55 40 35 25 15

B. Western Blot P

S

FT

E

IMAC W1 W2

1

SEC 2

3

MW (kDa) 170 100 70

55 40 35 25 15

Fig. 3 SDS-PAGE analysis of representative fractions from adenosine A2A receptor purification. P: ultracentrifugation pellet; S: solubilized fraction loaded onto nickel affinity chromatography column; FT: IMAC flow through; E: IMAC elution pool; W1 and W2: IMAC washes before elution; 1, 2, and 3: SEC elution peaks. (a) Analysis of the total protein content of 10 μL of each sample on a 10% polyacrylamide Tris-Tricine gel stained with Coomassie blue. (b) Western Blot analysis of the Adenosine A2A receptor content of 5 μL of each sample. Immunoblotting realized with a M2 anti-flag antibody (0.1 μg/mL)

4. Wash the membrane three times, each time with 50 mL of PBST on an orbital shaker for 5 min at room temperature. 5. Remove the PBST and incubate the membrane with the adapted ant-IgG antibody diluted in blocking buffer (typical final concentration of 0.1 μg/mL) for 1 h at room temperature on an orbital shaker. 6. Wash the membrane three times, each time with 50 mL of PBST on an orbital shaker for 5 min at room temperature. 7. Remove the PBST and wash the membrane with 50 mL of PBS on an orbital shaker for 5 min at room temperature. 8. Store the membrane in PBS until revelation.

Extraction of MPs from Yeast Protoplasts

55

9. Proceed to membrane revelation according to the reporter system selected and following the manufacturer’s recommendations. Figure 3b shows the western blot analysis of adenosine A2A receptor purification, using an anti-flag primary antibody. Several bands corresponding to the receptor of interest can be identified. The major signal at the apparent molecular weight of 55 kDa corresponds to the complete form of the receptor. SEC peak 2 thus contains the complete and monomeric form of the adenosine A2A receptor. A radioactive ligand-binding assay further confirms the functional state of this fraction of the purified receptor (data not shown).

4

Notes 1. Imidazole can be added to the solubilized sample after the ultracentrifugation step. The use of high quality imidazole (e.g., I0250, Sigma), which shows a very limited absorption signal at 280 nm, is strongly recommended for ion immobilized metal affinity chromatography. 2. The detergent used in the buffers has to be selected according to its solubilization and stabilization properties toward your protein of interest. We recommend a preliminary solubilization screening using different detergents before large-scale purification. In terms of concentration, we recommend using about 10–20 times the critical micellar concentration (CMC) when possible during the solubilization step, and at least twice the CMC for subsequent purification steps. 3. The duration of Zymolyase® treatment and/or amount of Zymolyase® used may be critical for protein integrity and can thus be adjusted to the protein of interest. 4. Whereas yeast cells are very resistant to mechanical stress, yeast protoplasts are as fragile as mammalian cells and have thus to be handled very carefully in order to avoid mechanical cell lysis. 5. This first step enables to get rid of unspecific contaminants, and can be adapted according to the protein of interest and the cellular context. In our experience, G protein coupled receptors labeled with a decahistidine tag elute from Ni Sepharose High Performance nickel affinity columns from about 150 mM imidazole. 6. Protein elution using an imidazole gradient (typically 0–100% of B buffer spread in 25 mL) can be of great use to separate monomeric from oligomeric or aggregated MPs. Based on our experience, aggregated proteins show a higher affinity for the resin and are thus eluted later than monomeric proteins.

56

Lucie Hartmann et al.

7. It is usually not recommended to boil MP samples prior to electrophoresis. When compact and highly hydrophobic proteins such as GPCRs are boiled, they usually aggregate and stay stuck in the concentrating gel. References 1. Schlebach JP, Sanders CR (2015) The safety dance: biophysics of membrane protein folding and misfolding in a cellular context. Q Rev Biophys 48(1):1–34 2. Buck TM, Jordan R, Lyons-Weiler J, Adelman JL, Needham PG, Kleyman TR, Brodsky JL (2015) Expression of three topologically distinct membrane proteins elicits unique stress response pathways in the yeast Saccharomyces cerevisiae. Physiol Genomics 47:198–214 3. Andre´ N, Cherouati N, Prual C, Steffan T, Zeder-Lutz G, Magnin T, Pattus F, Michel H, Wagner R, Reinhart C (2006) Enhancing functional production of G protein-coupled receptors in Pichia pastoris to levels required for structural studies via a single expression screen. Protein Sci 15:1115–1126 4. Magnin T, Fiez-Vandal C, Potier N, Coquard A, Leray I, Steffan T, Logez C, Alkhalfioui F, Pattus F, Wagner R (2009) A novel, generic and effective method for the rapid purification of G protein-coupled receptors. Protein Expr Purif 64(1):1–7

5. Bornert O, Alkhalfioui F, Logez C, Wagner R (2012) Overexpression of membrane proteins using Pichia pastoris. Curr Protoc Protein Sci Chapter 29:Unit 29.2 6. Bornert O, Møller TC, Boeuf J, Candusso MP, Wagner R, Martinez KL, Simonin F (2013) Identification of a novel protein-protein interaction motif mediating interaction of GPCRassociated sorting proteins with G proteincoupled receptors. PLoS One 8(2):e56336 7. Logez C, Berger S, Legros C, Bane`res JL, Cohen W, Delagrange P, Nosjean O, Boutin JA, Ferry G, Simonin F, Wagner R (2014) Recombinant human melatonin receptor MT1 isolated in mixed detergents shows pharmacology similar to that in mammalian cell membranes. PLoS One 9(6):e100616 8. Logez C, Alkhalfioui F, Byrne B, Wagner R (2012) Preparation of expression plasmids for Pichia pastoris. Meth. Mol Biol 866:25–40 9. Hartmann L, Kugler V, Wagner R (2016) Expression of eukaryotic membrane proteins in Pichia pastoris. Methods Mol Biol 1432:143–162

Chapter 4 Cell-Free Expression for the Study of Hydrophobic Proteins: The Example of Yeast ATP-Synthase Subunits Isabelle Larrieu, James Tolchard, Corinne Sanchez, Edmond Yazo Kone, Alexandre Barras, Claire Stines-Chaumeil, Benoıˆt Odaert, and Marie-France Giraud Abstract Small hydrophobic membrane proteins or proteins with hydrophobic domains are often difficult to produce in bacteria. The cell-free expression system was found to be a very good alternative for the expression of small hydrophobic subunits of the yeast ATP-synthase, such as subunits e, g, k, i, f and the membrane domain of subunit 4, proteins that are suspected to play a role in the stability of ATP-synthase dimers. All of these proteins could be produced in milligrams amounts using the cell-free “precipitate mode” and were successfully solubilized in the presence of lysolipid 1-myristoyl-2-hydroxy-sn-glycero-3-phospho-10 rac-glycerol. Purified proteins were also found suitable for structural investigations. An example is given with the NMR backbone assignment of the isotopically labeled subunit g. Protocols are also described for raising specific polyclonal antibodies against overexpressed cell-free proteins. Key words Cell-free expression, ATP-synthase, Dimer-specific subunits, NMR, Antibodies

1

Introduction Due to their propensity to form inclusion bodies or to their toxic effect after their insertion into host membranes, membrane proteins or membrane interacting proteins are difficult to overexpress in cellular organisms. Cell-free expression (CF) consists of an open system where all machineries and substrates required for transcription, translation, and energy regeneration are added in the presence of a plasmid or a DNA fragment containing a sequence encoding the protein of interest (Fig. 1A). Several cell-free expression systems, with extract preparations from bacteria [1, 2], yeast [3], wheat germs [4, 5], insect cells [6], rabbit reticulocytes [7], CHO [8], or human cells [9, 10], have been developed and proved to be very good alternatives to recombinant expression in vivo. Each lysate has its own advantages and disadvantages, and the choice of

Jean-Jacques Lacapere (ed.), Membrane Protein Structure and Function Characterization: Methods and Protocols, Methods in Molecular Biology, vol. 1635, DOI 10.1007/978-1-4939-7151-0_4, © Springer Science+Business Media LLC 2017

57

58

Isabelle Larrieu et al.

A

B

Hole

C

Feeding Mix

Products

Reactives

kDa

P

70 55 -

13C15NAA

tRNA

40 -

1

mRNA 2 Energy Regenerating System Reaction Mix

M S

Flip Tube

dialysis membrane T7 RNAP

Plasmid

Feeding Mix

Reaction Mix 3

35 25 -

Membrane 15 -

Cap

- His6g

10 -

Fig. 1 Cell-free expression principle and setup. (A) Components involved during cell-free expression with E. coli extracts. 1: a plasmid (or a DNA fragment) with the sequence encoding the protein of interest under the dependence of a T7 RNA polymerase promoter is added in the reaction mix that contains the T7 RNA polymerase (T7 RNAP). Transcription occurs with added NTPs. A regenerating energy system consisting of pyruvate kinase/phosphoenol pyruvate and acetyl phosphate/acetate kinase maintains a high ATP content. NTPs and amino acids can diffuse from the feeding mix to the reaction mix compartment. Products like phosphate can diffuse into the larger feeding mix compartment. 2: Ribosomes present in the extract translate mRNAs using added tRNAs. Protein synthesis occurs with the incorporation of the added amino acids (labeled 13 15 C N AA or unlabeled). 3: Hydrophobic proteins aggregate in the precipitate mode (P-CF). (B) Small-scale cell-free expression in a Flip Tube®. (C) Example of the expression obtained with the pIVEX2.4a–HN-ATP20 construct leading to the production of subunit g with a his-tag at the N-terminus. Pellet(P) and supernatant(S) fractions were analyzed as indicated in Subheading 3.2.3 on a 12.5% acrylamide Tris-Tricine SDS-PAGE gel and stained with Instant-Blue® solution. M molecular weight markers

a lysate source depends on the required posttranslational modifications, the desired protein yield, experimental simplicity, and the cost of extract preparations [11]. CF extract from Escherichia coli is easy to routinely prepare in a laboratory due to well-established protocols [12] and is the system of choice for proteins that are not posttranslationally modified. Three modes of overexpression can be used for membrane proteins. In the P-CF mode, membrane proteins are produced without any detergents or lipids and synthesized proteins precipitate (Fig. 1A). In the presence of detergent (D-CF) or lipids (L-CF mode) membrane proteins can insert into membrane mimetics added to the reaction mix, such as micelles, bicelles, liposomes, or nanodiscs [13, 14]. In the P-CF mode, precipitates of hydrophobic proteins can be simply collected by centrifugation. After washes of the pellets, the overexpressed protein is usually the main entity and can be resuspended in any appropriate membrane mimetic.

In Vitro Expression of Yeast ATP-Synthase Subunits

59

F1Fo-ATP-synthases are ~600 kDa complexes that use the energy of ion gradients across membranes to synthesize ATP. The full bacterial ATP-synthase of the Caldalkalibacillus thermarum strain TA2.A1 was successfully produced and reconstituted from an operon harboring the nine genes of this enzyme in the presence of E. coli extracts [15]. However, mitochondrial ATP-synthases are more complex than their bacterial counterparts, in terms of the genetic origin of their subunits and their composition, assembly, and function. Their genes are encoded by nuclear and mitochondrial genomes which makes their assembly an intricate process that has to be assisted by chaperones [16]. In addition to their role in ATP synthesis, mitochondrial ATP-synthases are also involved in shaping the inner mitochondrial membrane by forming dimers that constitute the building blocks of large oligomers involved in mitochondrial cristae morphology [17, 18]. The yeast enzyme consists of 17 distinct subunits. Subunits e, g, and k, are not necessary for ATP-synthase function but have been found to be specific to the dimeric yeast enzyme [19]. The N-terminal region of subunit 4 has also been shown to play a major role in dimer stability [20]. 3D structures of sub-complexes of the monomeric enzyme [21–23] have helped in understanding the function of this nanomolecular machine and recent cryo-EM volumes [24, 25] have brought data on the organization of dimeric enzymes and have indicated that subunit f could also play an important role in dimer stability. However, high-resolution structural information of subunit f and the N-terminal extremity of subunit 4 (S4T) have not been established and no structural data are available on the structures of the dimer-specific subunits e, g, and k or other proteins suspected to be at the dimer periphery, such as subunit i [25–27]. The mode of assembly for subunits e, g, and k is poorly understood [28], yet in vitro reconstitution of dimers from the monomeric complex in the presence of these dimer-specific proteins could be very informative, provided these proteins can be produced. Obtaining polyclonal antibodies against membrane proteins or membrane bound proteins usually requires the purification of these proteins from organisms or the production of these proteins in Escherichia coli with all the problems inherent to the overexpression of proteins with hydrophobic domains. An alternative strategy is to chemically synthesize peptides, which in some cases may have poor immunogenic properties. Here, we show how convenient the CF system is to overexpress subunits of the yeast ATP-synthase that are not well expressed or not expressed at all in E.coli and how easy and quick this technique is to prepare samples for NMR (Nuclear Magnetic Resonance) once the best conditions for a cellular expression have been found. We also present a strategy to produce polyclonal antibodies against full or partial subunits of the yeast ATP-synthase, using the CF produced proteins.

60

2 2.1

Isabelle Larrieu et al.

Material (and Common Methods) Constructs

2.2 Plasmid Amplification and Preparation

The constructs allowing the CF expression of yeast ATP-synthase subunits are listed in Table 1. 1. LB medium: 16 g of Bacto-Tryptone, 10 g of Bacto-Yeast Extract, 5 g of NaCl. Adjust the pH at 7.5 with NaOH and the volume to 1 L with ddH2O. Autoclave for 30 min at 1 bar. 2. Incubator shaker (New Brunswick Scientific or Infors). 3. One colony of DH5α bacteria freshly transformed with the appropriate plasmid was inoculated in 200 mL of LB medium supplemented with ampicillin (100 μg/mL) or kanamycin (80 μg/mL) and cultured overnight at 37  C at 180 rpm. Vectors were purified with a Midi-Prep kit (Qiagen, France) (see Note 1). 4. The absence of mutations was confirmed by DNA sequencing with T7 primers (Eurofins Genomics, France).

2.3 Extract Preparation and CF Expression

1. Cell Washing Buffer for T7 RNA Polymerase purification: 20 mM NaCl, 2 mM EDTA 20 mM Tris–HCl pH 8.1. 2. Lysis Buffer for T7 RNA Polymerase purification: 20 mM NaCl, 2 mM EDTA, 1 mM DTT, 20 mM Tris–HCl pH 8.1. 3. Resuspension Buffer: 20 mM sodium phosphate pH 7.7, 1 mM DTT, 5% (w/v) glycerol, NaCl (with the indicated concentrations). 4. HS Pre-Equilibration Buffer: 20 mM sodium phosphate pH 7.7, 1 mM DTT, 5% (w/v) glycerol, 50 mM NaCl. 5. HS Elution buffer: 20 mM sodium phosphate pH 7.7, 1 mM DTT, 5% (w/v) glycerol, 500 mM NaCl. 6. S30 Buffer: 14 mM magnesium acetate, 0.6 mM potassium acetate, 0.5 mM DTT, 10 mM Tris-acetate pH p8.2. 7. Bacterial S30 extracts have been extensively described elsewhere and so will not be detailed here. They were prepared from BL21(DE3)Star cells (Thermo Fisher, France), according to the protocol developed by Schwarz et al. [12]. Polyethylene glycol 8000 (PEG 8000), potassium acetate, HEPES, Folinic acid (calcium salt), DL-dithiothreitol (DTT), ATP, CTP, GTP, UTP, lithium-potassium acetyl phosphate, and all amino acids (AA-Mix) were purchased from Sigma-Aldrich, Complete® EDTA free, E. coli tRNAs and pyruvate kinase from Roche, recombinant RNAsin from Promega, phosphoenol pyruvic acid from VWR, magnesium acetate from Prolabo and sodium azide from Merck. For NMR studies, CF expression was performed in the presence of labeled amino acids (15N and 13C) purchased from Cortecnet (France).

NdeI/BamH1 NcoI/SmaI

pIVEX2.3-ATP21

pET30a(þ)ATP21

pIVEX2.3-ATP19-HC

pIVEX2.3MCS-ATP18

pIVEX2.3MCS-ATP17

pET30a(þ)-ATP17

pIVEX2.3-ATP20

pIVEX2.4-HN-ATP20

e

e

kHis6

i

f

f

g

His6g

Nco1/XmaI

NcoI/XmaI

NdeI/SacI

NdeI/SacI

NdeI/SmaI

XbaI/NcoI

NcoI/SmaI

pIVEX2.3-ATP4S4T

S4T

Cloning sites

Vector

Protein name

Table 1 Vectors used for the expression of ATP-synthase subunits

1: GGGCCATGGTAAGCAGGATCCAAAATTATACCAG 2: GCGCGCCCGGGTTAGTGATGTTTATATCCC 10 : CCTACAGAAGTTTTATCGAGTTTAAAAAACATCCAAAAG 20 : CTTTTGGATGTTTTTTAAACTCGATAAAACTTCTGTAGG

1: GGGCCATGGTAAGCAGGATCCAAAATTATACCAG 2: GCGCGCCCGGGTTAGTGATGTTTATATCCC

1: GGGGCATATGGTATCTACATTGATTCCTC 2: GGGGGAGCTCGTTCTATCACTTATGTACTCC

1: GGGGCATATGGTATCTACATTGATTCCTC 2: GGGGGAGCTCGTTCTATCACTTATGTACTCC

*

1: GCGCCCATGGGTGCTGCTTATCATTTC 2: CGCGCCCCGGGCGCATCTTGCTTTTCCGAATG

1: GGGGAATTCCATATGTCGACAGTTAATGTTTTGAG 2: GGGGGGATCCGGCCTACCACCTGCAAGGG

#

(continued)

1: GGCCCCATGGCTTCCACTCCAGAAAAAC 2: CTCTCTCCCGGGTTAAGTTTCCTTGGAAACATCAAAC 10 :CATATAAAGATTTTGCCGATGCATGAATGAAGAAAGTCTCCGACG 20 : CGTCGGAGACTTTCTTCATTCATGCATCGGCAAAATCTTTATATG

Oligonucleotides used for PCR amplification or site directed mutagenesis 1: 50 Forward Primer-30 ; 2: 50 -Reward Primer-30 ; 10 and 20 : oligonucleotides used for mutagenesis

In Vitro Expression of Yeast ATP-Synthase Subunits 61

NdeI/SmaI

pIVEX2.3MCS-ATP6-HC

6His6

*

*

Oligonucleotides used for PCR amplification or site directed mutagenesis 1: 50 Forward Primer-30 ; 2: 50 -Reward Primer-30 ; 10 and 20 : oligonucleotides used for mutagenesis

Constructs are given by the name of the expression plasmid, followed by the name of the gene. When his-tags were introduced at the N- or C-termini of proteins, a HN or HC are added before or after the name of the gene, respectively. Genes were amplified by PCR from 250 ng of yeast genomic DNA. PCR fragments were digested by the indicated enzymes and ligated in pIVEX or pET plasmids digested with the same enzymes. *: For subunit 6, 8, and i synthetic genes optimized for E. coli expression were inserted in the pUC57 plasmid (Eurogentec). Constructs were digested by NdeI and SmaI and ligated in the pIVEX2.3.MCS vector, digested with the same enzymes. For S4T, a fragment encoding the first 103 residues of the protein was inserted in the pIVEX2.3. A shorter version of the protein (S4T, 83 residues) was obtained by introducing a stop codon using the 10 and 20 oligonucleotides with the Quick change mutagenesis method (ThermosFisher). Cysteine mutants of subunits e (eC28➔S) and g (gC75➔S) were produced to avoid any disulfide bridge formation during expression and purification. These mutations do not alter the ATPase activity or the ATP-synthase dimerization process in yeast. Oligonucleotides used for mutagenesis are indicated for subunit g. For subunit e, the PCR amplification was performed from a plasmid harboring the e(C28➔S) mutation [48]. #: A XbaI/BamHI fragment from the pET30a þ ATP21plasmide was inserted in the pIVEX 2.3, digested by XbaI and BamHI

NdeI/SmaI

pIVEX2.3MCS-ATP8

8

Cloning sites

Vector

Protein name

Table 1 (continued)

62 Isabelle Larrieu et al.

In Vitro Expression of Yeast ATP-Synthase Subunits

63

8. Assays were carried out in a DNAse/RNAse-free Flip Tube® (Hamilton) or in dialysis tubings with different Molecular Weight Cut-Off (MWCO) purchased from Spectrum. Spectra/Por membranes with grade 7 only require incubation of 15 min in ddH2O. Spectrum membranes with Grade 1–6 require pretreatments to remove heavy metals and contaminants. 9. The T7 RNA polymerase was purified from BL21(DE3) Star cells transformed with the pAR1219 plasmid containing the T7 RNA polymerase gene [29]. A different protocol was used than the one given in Schwarz et al. [12] for the polymerase preparation. It will be described in the Methods section. 2.4 Production of Polyclonal Antibodies

Rabbit antibodies against subunit k, subunit e or the N-terminal part of subunit 4 (S4T) were obtained from the Covalab Company (France) with the following immunization protocol. Prior to immunization, the sera from four different rabbits were tested to ensure that they would not interact with proteins from bacterial extracts or total yeast extracts. Two rabbits were chosen for the following immunization protocols: – Day 0: first boost: 0.5 mL of resuspended protein (200 μg in 20 mM sodium phophate buffer pH 7.2) and 0.5 mL of incomplete Freund’s adjuvant. – Day 21: second boost: 0.5 mL of resuspended protein and 0.5 mL of incomplete Freund’s adjuvant. – Day 42: third boost: 0.5 mL of resuspended protein and 0.5 mL of incomplete Freund’s adjuvant. – Day 53: first taking. – Day 63: fourth boost: 0.5 mL of resuspended protein and 0.5 mL of incomplete Freund’s adjuvant. – Day 74: second taking. – Day 88: last taking. Antibodies were tested during the immunization protocols. Antibodies used for the Western blots were those from the second taking.

2.5

Purification

¨ KTA purifier or BioRad). 1. FPLC system (A 2. IMAC Binding Buffer: 50 mM NaCl, 10 mM imidazole, one tablet of protease inhibitor cocktails EDTA-free from Roche®/ 50 mL, 50 mM Na2HPO4 pH 7.8. 3. IMAC Elution Buffer: 50 mM NaCl, 250 mM imidazole, one tablet of protease inhibitor cocktails EDTA-free from Roche® (Mini Complete) per 10 mL, 50 mM Na2HPO4 pH 7.4. 4. Washing Buffer 1: 20 mM Tris–HCl, pH 8.0.

64

Isabelle Larrieu et al.

5. Washing Buffer 2: 150 mM NaCl, 20 mM Tris–HCl, pH 8.0. 6. D-Buffer: 20 mM Na2HPO4 pH 7.4. 7. L-Buffer: 150 mM NaCl, 20 mM Na2HPO4 pH 6.2. 8. Superdex Buffer: 150 mM NaCl, 60 μM LMPG (cmc ¼ 50 μM), 20 mM Na2HPO4 pH 6.2. 2.6 Electrophoresis and Western Blotting

PBS-Tween 20: 0.15 M NaCl, 0.05% Tween-20, 10 mM sodium phosphate, pH 7.2. 2 Laemmli Loading Buffer: 2% (w/v) SDS, 20% Glycerol, 125 mM Tris–HCl pH 6.8. Total proteins from yeast cells were prepared as described by Egner et al. [30]. Mitochondria were prepared with the zymolyase method [31] from the strain 273–10/B/A (WT) or from mutants lacking subunits k or subunit 4. Mitochondria were dissociated with Laemmli Buffer and proteins were separated on SDS-PAGE and transferred onto 0.2 μm nitrocellulose membranes (Protran, Amersham). Membranes were saturated with 5% milk in PBSTween 20. Primary antibodies were added for 1 h at room temperature, washed with PBS-Tween, saturated with 5% milk in PBSTween 20. Secondary antibodies were added for 45 min at room temperature. Rabbit polyclonal anti-hexahistidine tag antibodies (Bethyl Laboratories, US) were used with a 1/5000 dilution and secondary peroxidase-coupled goat anti-rabbit IgGs (Jackson Immunoresearch, UK) with a 1/10,000 dilution. For antibodies against ATP-synthase subunits, dilutions ranging from 1/1000 to 1/10,000 were used. Peroxidase activity was revealed by ECL (Luminata Crescendo, Millipore) and emitted light was recorded with a digital camera (G-Box, Syngene). The antibodies directed against a peptide of subunit k (CENYLKKHSDKQDA) were kindly provided by I. Arnold. Antibodies directed against a peptide of subunit e (CISKQIYLKEGLQPPT) were obtained from Eurogentec. Unless stated, protein markers (PageRuler) were from Thermo Scientific.

2.7 NMR Data Acquisition

LMPG-Buffer: 20 mM Na2HPO4 pH 6.22 prepared by adding in 50 mL of ddH2O or D2O depending on the intended NMR investigation: 140.2 mg of NaH2PO4, 66.7 mg of Na2HPO4. All NMR experimentation was carried out at the Institut Eur´ opeen de Chimie et Biologie, Universite´ de Bordeaux (CNRS, IECB-UMS3033). Initial trials of optimal sample conditions for subunit g were carried out using a 700 MHz Bruker AVANCE III spectrometer equipped with a 5 mm PATXI triple resonance probe with Z-gradients at 293 K. All further experimentation was carried out with an 800 MHz Bruker AVANCE III spectrometer equipped with either a 5 mm CPTXI or 5 mm CPTCI cryoprobe with Z-gradients at 313 K. Double-labeled (13C-15N) samples of subunit g were prepared according to Subheading 3.3.3. Backbone

In Vitro Expression of Yeast ATP-Synthase Subunits

65

resonance assignment was carried out using the backbone walk methodology, utilizing 2D 1H-15N SOFAST-HMQC, 2D 1 H-13C HSQC [32] and 3D HNCA [33], HNCACB [34], CBCACONH [35], HNCO [36], HN(CA)CO [37], and HBHA(CO) NH [38] spectra. Spectra were processed using NMRPipe/Draw [39] and analyzed using CCPN-Analysis [40]. All indirect dimensions were indirectly referenced to TSP according to their gyromagnetic ratios [41].

3

Methods

3.1 T7 RNA Polymerase Preparation 3.1.1 T7 RNA Polymerase Production

1. Inoculate 1 L of LB broth, containing 100 μg of ampicillin/mL with a colony of BL(21)DE3Star cells freshly transformed with the pAR1219 plasmid [29]. 2. Grow cells at 37  C at 180 rpm. 3. Prepare a stock solution of 1 M IPTG (238 mg in 1 mL of ddH2O, filter the solution with a 0.2 μm filter). 4. When the OD600nm of the culture reaches 0.6, add 500 μL of IPTG solution. 5. Induce for 4 h. 6. Collect cells by centrifugation (10 min, 6500  g). 7. Wash cells twice with Cell Washing Buffer. 8. Discard the supernatants. 9. 8 g of cells are usually obtained from 1 L of LB medium (see Note 2).

3.1.2 T7 RNA Polymerase Purification

1. Resuspend cells in 24 mL of Lysis Buffer. To avoid any protein degradation, always work at 4  C. 2. Break down the bacterial cell wall by adding 6 mL of a hen egg white lysozyme solution at 1.5 mg/mL in Lysis Buffer. 3. Add a tablet of protease inhibitor cocktail. 4. Incubate for 20 min at 4  C. 5. Complete cell lysis by adding 2.5 mL of sodium deoxycholate (0.8% (w/v)). 6. Incubate for 5 min under magnetic stirring in a cold room. 7. Transfer the lysate into a glass beaker. Decrease the viscosity due to DNA by sonicating four times 3000 (Annemasse sonicator, 120 V). Between each sonication let the cell lysate cool down on ice. 8. Add 5 mL of a 2 M ammonium sulfate solution. 9. Complete to a final volume of 50 mL with Lysis Buffer.

66

Isabelle Larrieu et al.

10. Precipitate nucleic acids fragments and DNA bound polymerases with Polymin P. Prepare a 10% (w/v) Polymin P solution adjusted to pH 8.0 with HCl. 11. Add 5 mL of this solution to the lysate, drop by drop. 12. Incubate for 20 min in a cold room, under magnetic stirring. 13. Centrifuge the lysate at 39,000  g, 20 min. 14. Measure the volume of the supernatant and add 0.82 volume of a saturated ammonium sulfate solution. Add the ammonium sulfate solution drop by drop, under agitation in a cold room to avoid any local ammonium sulfate over-concentration. 15. Incubate for 15 min in the cold room, under magnetic stirring. 16. Collect the precipitate by centrifugation (10 min, 20,000  g). 17. Dissolve the pellet in 15 mL of Resuspension Buffer containing 100 mM of NaCl. 18. Dialyze twice against 1 L of Resuspension Buffer containing 100 mM of NaCl (Membrane Spectra/Por, MWCO 25 kDa prepared as indicated in chapter 3.2.2.1). 19. Centrifuge 10 min at 20,000  g. 20. To the volume of the supernatant, add an equal volume of Resuspension Buffer without NaCl to reach a final NaCl concentration of 50 mM. 21. Pre-equilibrate the column with 24 mL of HS Pre-Equilibration Buffer using a flow rate of 8 mL/min. 22. Load 5 mL of sample on a sulfopropyl column (POROS, HS/ 20, 10/100) adapted on a FPLC system. 23. Load the protein sample at 0.5 mL/min. 24. Wash the column at a flow rate of 2 mL/min with 24 mL of the HS Pre-Equilibration Buffer. 25. Elute proteins with a gradient of HS Elution Buffer in PreEquilibration Buffer at a flow rate of 2 mL/min. The linear gradient should be performed during four column volumes (40 mL). The T7 RNA polymerase is eluted between 300 and 430 mM of NaCl. 26. Dialyze pooled fractions against twice 500 mL of Resuspension Buffer without NaCl. 27. Collect the precipitate formed during dialysis by centrifugation (10 min, 20,000  g) (see Note 3). 28. Dissolve the pellet in 10 mL of Resuspension Buffer containing 100 mM NaCl and dialyze against the same buffer containing 50% glycerol to concentrate the sample and stabilize the enzyme with glycerol. Around 3 mL of protein solution (~5 mg/mL) are obtained from 1 L of cell culture. 29. Add 2 mM DTT. Store the T7 RNA polymerase by aliquots of 20 μL at 80  C (see Note 4).

In Vitro Expression of Yeast ATP-Synthase Subunits

67

CF Expression

In vitro protein expression was performed as described by Schwarz et al. [12] by mixing the volumes of the different components indicated below and in Table 2. The principle of CF expression is summarized in Fig. 1A.

3.2.1 Small-Scale CF Mixes

1. Prepare 921.1 μL of Master Mix by mixing 9 μL of 10% (w/v) sodium azide, 90 μL of 40% (w/v) PEG 8000, 67.9 μL of 4 M potassium acetate, 18.2 μL of 1 M magnesium acetate, 66 μL of 2.5 M HEPES, 36 μL of a solution containing 1 tablet of Complete®-EDTA free protease inhibitor cocktail/1 mL, 18 μL of 1% (w/v) folinic acid, 7.2 μL of 0.5 M DTT (freshly prepared), 24 μL of NTP mix (75: 60 mM GTP, 60 mM CTP, 60 mM UTP, 90 mM ATP neutralized with NaOH), 36 μL of 1 M phosphoenol pyruvate, 36 μL of 1 M acetyl phosphate, 225 μL of a stock solution containing 4 mM of each amino acid, 107.8 μL of a solution containing 16.7 mM of each of the following amino acids (R,C,W,M, D, and E) and 180 μL of ddH2O.

3.2

2. Prepare 100 μL of Reaction Mix by combining: 0.4 μL of pyruvate kinase at 2000 U/mL, 1.2 μL of E. coli tRNAs at 40 mg/mL, 1.4 μL of T7 RNA polymerase at 5 μg/μL, 0.9 μL of RNAsin at 32 U/μL, 7.5 μL of plasmid at 200 ng/μL, 35 μL of “S30” extract, and 51.2 μL of Master Mix. 3. Prepare 1700 μL of Feeding Mix: 869.9 μL of Master Mix, 595 μL of “S30 buffer” (14 mM magnesium acetate, 0.6 mM potassium acetate, 0.5 mM DTT, 10 mM Tris–acetate pH p8.2), 212 μL of a solution containing 4 mM of each amino acid, and 22.6 μL of ddH2O. 3.2.2 Expression Chamber

Membrane Preparation for Flip Tube® Experiments

Small-scale expressions are performed in a Flip Tube® (Fig. 1B) with a dialysis membrane separating the reaction mix from the feeding mix. 1. Cut dialysis tubing with the desired MWCO (3.5–10 kDa) and with a flat width of 2.4 cm. Create pieces of roughly 2  2.4 cm and cut doubled-membranes with scissors to make single membrane pieces. 2. Put the cut pieces in a beaker containing 200 mL of ddH2O and a few grains of disodium phosphate. 3. Boil 2 min in a microwave and rinse with ddH2O. 4. Add 200 mL of 0.5 mM EDTA, pH 8.0. 5. Boil 2 min in a microwave and rinse with ddH2O. 6. Store membrane pieces in 30% Ethanol, 0.5 mM EDTA. 7. Prior to use, rinse the membrane in ddH2O.

68

Isabelle Larrieu et al.

Table 2 Composition of CF mixes for expression in a Flip Tube®

Stock

Final Concentration

Master mix (μL) small scale

NaN3 (%)

10

0.05

9

PEG 8000 (%)

40

2

90

Potassium acetate (mM)

4000

150.8

67.9

Magnesium acetate (mM)

1000

7.1

18.2

HEPES (M)

2.5

0.1

66

Complete (x)

50

1

36

Folinic acid (mg/mL)

10

0.1

18

DTT (mM)

500

2

7.2

NTP Mix *(x)

75

1

24

PEP (mM)

1000

20

36

Acetyl Phosphate (mM)

1000

20

36

Amino acid Mix (mM)

4

0.5

225

RCWMDE (mM)

16.7

1

107.8

Feeding mix (μL)

Reaction mix (μL)

Compound

H2O

212,5

180 921.1 869.9

51.2

S30 buffer (v/v %)

100

35

595

Pyruvate kinase (mg/ mL)

10

0.04

0.4

tRNAs (mg/mL)

40

0.5

1.2

T7 RNA P (5 μg/μL)

5

0.07

1.4

RNAsin (U/μL)

32

0.3

0.9

DNA (ng/μL)

200

15

7.5

S30 extract (% V/V)

100

35

35

H2O

22.6

2.4

Total volume

1700

100

a: 75 corresponds to 60 mM GTP, 60 mM CTP, 60 mM UTP, 90 mM ATP

In Vitro Expression of Yeast ATP-Synthase Subunits Cell-Free Expression in Flip Tubes® (see Note 5)

69

1. Cut the bottom of the tube (4–5 mm from the bottom) with scissors. 2. Detach the lid from the top of the tube. 3. In the round compartment of the cap, add 100 μL of Reaction Mix. Avoid creating bubbles. 4. Spread the Reaction Mix so that the liquid creates a flat surface and fully fills up the compartment. 5. Wipe a square piece of membrane with a Kimwipe® (KimberlyClark). Use a flat edge tweezer to hold the membrane. 6. Cover the circular compartment with the membrane. It is essential to put the membrane at one go and to perfectly center it. 7. Carefully adapt the cut tube onto the lid. 8. Fill up the tube with the Feeding Mix. A layer of parafilm can be added on the hole to avoid evaporation. 9. Put the tube upside down in an incubator shaker. 10. Incubate at 28  C at 50–75 rpm (see Note 6). 11. After 20 h of expression, put the tube with the bottom down and aspirate the Feeding Mix with a 200 μL pipette through the hole (see Note 7). 12. With a pair of scissors, while firmly maintaining the tube with the lid on the bench, increase the size of the hole at the bottom of the tube. The hole should be large enough to allow the top of a 200 μL pipette tip to pass through. 13. Adapt a 200 μL tip onto a 200 μL pipette, set the volume to be aspirated to 100 μL, press the push button prior to quickly and firmly piercing the membrane. While slowly releasing the push button, describe quick circles underneath the broken membrane to aspirate most of the Reaction Mix. 14. Dispense the Reaction Mix in an Eppendorff tube (see Note 8).

3.2.3 Sample Analysis

To analyze the yield of expression, as well as to determine whether the protein is produced as soluble or aggregated forms, a treatment of the soluble fraction is required to avoid any perturbation during electrophoresis. Proceed as below: 1. Centrifuge the Reaction Mix 10 min at 12,000  g at 4  C. 2. Aspirate the supernatant and dispense it into a new tube. The volume should be around 90–100 μL. 3. Add 200 μL of ddH2O onto the pellet, aspirate two or three times to resuspend the pellet, and centrifuge for 10 min at 12,000  g at 4  C. Transfer the supernatant into a new tube to analyze a fraction of it on a gel, in the case the produced protein could be partly solubilized. 4. Repeat step 3.

70

Isabelle Larrieu et al.

5. Resuspend the pellet of step 4 with 100 μL of ddH2O. 6. Aspirate 7 μL of the resuspend pellet in a new tube and mix with 7 μL of a 2 Laemmli Loading Buffer. 7. Aspirate 7 μL of the supernatant of step 2 and dispense it into a new tube. Complete with 93 μL of H2O and add 11 μL of 3 M trichloroacetic acid. 8. Centrifuge at 12,000  g for 5 min at 4  C. 9. Discard the supernatant of step 7 and wash the pellet with 200 μL of cold acetone. Centrifuge at 12,000  g for 5 min at 4  C and discard the supernatant. 10. Repeat steps 8 and 9. 11. Air-dry the pellet for 10 min by leaving the tube open on the bench. 12. Add 7 μL of 5% (w/v) SDS, resuspend the pellet by aspirating up and down with a pipette. 13. Add 7 μL of 2 Laemmli Loading Buffer. If the mix appears yellow, add a few grains of Tris powder. 14. Load samples without preheating onto a mini-PROTEAN Tris–Glycine or a Tris–Tricine precasted SDS-PAGE gel with an appropriate acrylamide reticulation. 15. Run the gel using a mini-Protean electrophoresis chamber. 16. Dismount the gel and detect proteins with a Coomassie-based staining solution. An analysis of the supernatant and pellet fractions obtained after the expression in a Flip Tube® of a his-tagged version of the yeast ATP-synthase subunit g is shown Fig. 1C (see Note 9). 3.2.4 Improvement of Yield of Expression

Popular plasmids for protein expression are those from the pET series, in which the gene of the protein of interest can be placed under the dependence of a T7 RNA polymerase promoter in order to perform protein expression in bacterial strains derived from BL21(DE3) cells. If such a construct exists it can be first tested for CF expression. However, for some proteins, a swapping of the gene of the protein into a plasmid optimized for in vitro expression such as pIVEX vectors (5 PRIME) may be worthwhile as in some cases this can incredibly increase the yield of production (see Note 10). Here is an example for the yeast ATP-synthase subunit f (encoded by the ATP17 gene) that was produced in the CF expression system using the pET30a(þ) vector (Table 1). With this plasmid, low amounts of proteins were obtained. The ATP17 gene was then inserted into the pIVEX2.3 MCS plasmid. As shown in Fig. 2, a ninefold increase of subunit f expression could be obtained with the pIVEX construct compared to the pET construct.

In Vitro Expression of Yeast ATP-Synthase Subunits

71

Such an improvement could be due to differences in the sequence of these plasmids between the T7 promoter and the Shine Dalgarno region [42]. An advantage of pIVEX plasmids is that they have a high-copy-number replication origin, unlike pET vectors. Thus, the quantity of DNA required for CF expression can be easily obtained from a Midi-Prep Qiagen preparation. The small-scale expression can be used to determine the best constructs and to optimize expression according to salt concentration as indicated in Ma et al. [13] and then production can be performed in larger volumes. Usually, large expression volume experiments consist of 1 mL of Reaction Mix in dialysis tubing. Each volume indicated in Table 2 is simply multiplied by 10. The dialysis tubing is clamped and submerged into 17 mL of Feeding Mix in a 25 mL beaker covered with parafilm and expression is performed at 28  C for 20 h at 75 rpm. Table 3 summarizes the quantities of yeast ATP-synthase subunits that could be produced using the in vitro system from 1 mL of Reaction Mix (see Note 11). CF expression proved to be the system of choice for subunit f production, as the expression of this subunit with the pET30a(þ) ATP17 plasmid induced the lysis of BL21(DE3) cells after IPTG addition. Subunit g was very weakly expressed in bacteria while good production yields were obtained with the CF expression system. Subunit e was very well produced in bacteria but was part of inclusion bodies (IB), and all attempted refolding and solubilization strategies were unsuccessful. As will be shown below, this

Fig. 2 In vitro optimization of the expression of the yeast ATP-synthase subunit f. 7 μL of the supernatant (left) or pellet (right) fractions of CF experiments with the pET30a(þ)ATP17 (pET) vector or with the pIVEX2.3MCS-ATP17 (pIV) construct were analyzed on a 15% polyacrylamide Tris–Glycine SDS-PAGE gel and stained with Instant Blue®

72

Isabelle Larrieu et al.

Table 3 Expression of ATP-synthase subunits Protein name

Number MW Da of TMs

Produced in vitro

mg/mL of reaction mix

Remarks concerning Degradation E.coli expression

8734

2

+

2.7  1.3 (n ¼ 6)

NO

ND

e

10,860

1

+

1.4  0.2 (n ¼ 4)

NO

IB (a)

f

10,565

1

+

2.4  0.2 (n ¼ 3)

NO

(b)

g

12,906

1

+

1.9  0.6 (n ¼ 9)

NO

(c)

i*

6687

1

+

1.8  0.2 (n ¼ 3)

NO

ND

kHis6

8712

?

+

1.4  0.1 (n ¼ 3) #

NO

ND

28,018

5

+



YES**

ND

5822

1

 ***





ND

S4T

6His6* 8*

*: As subunits 6 and 8 are mitochondrially encoded, synthetic genes optimized for bacterial expression (Eurogentec) were inserted in pIVEX plasmids. The small gene of subunit i, ATP18, was also optimized for expression in E. coli **: subunit 6 was fully synthesized but proteolysis occurred. ***: ATP8 mRNA was synthesized but the protein was not produced (a): Subunit e was well produced from pET vectors in E. coli but formed inclusion bodies from which it was impossible to get a form of the protein that did not aggregate (b): subunit f expression induced a lysis of the cells when expressed in E. coli (c): levels of subunits g were extremely low in E. coli Averaged values are given in parentheses for the number of expression experiments TMs: Predicted transmembrane segments ?: Subunit k has a predicted hydrophobic segment but can be extracted from membranes by an alkaline treatment #: quantity estimated in the pellet fraction ND: Expression was not performed in E. coli

was not the case with precipitates obtained after CF expression of subunit e. The CF produced subunit e could be easily solubilized, indicating that in contrast to proteins in IBs, protein precipitates formed in the P-CF mode may retain partial structures. Proteins encoded by mitochondrial DNA (subunits 6 and 8) were difficult to produce using the CF system. The genes of these subunits were optimized for expression in E. coli. The ATP8 mRNA was synthesized, as confirmed by a test mRNA synthesis reaction from the pIVEX2.3MCS-ATP8 construct in the presence of the T7 RNA polymerase. However, this protein was not produced. Special factors may be required for its translation or the stabilization of its mRNA. A C-terminal his-tagged version of subunit 6 could be expressed but bands with lower molecular weights were also observed on gels. Some of these bands were recognized by antibodies directed against the N-terminal part of subunit 6 and others by the His-tag antibodies. Addition of chelating agents or protease inhibitors in supplement to the ones already present in the Complete® cocktail did not prevent the presence of these bands.

In Vitro Expression of Yeast ATP-Synthase Subunits

3.3 Sample Treatment for Further Analyses 3.3.1 Removal of Nucleic Acids

73

Whatever the expression volume, an extensive cleaning of the pellet is required for proteins produced in the P-CF mode. After two washes with ddH2O, pellets still contain large amounts of RNA and DNA fragments and bases that could interfere with further experiments like 1H–NMR or hide charges on an ion exchange chromatographic resin. The protocol below describes how to remove these molecules. 1. Recover overexpressed membrane proteins from the reaction mix by centrifugation (10 min, 12,000  g, 4  C). Discard the supernatant. 2. Wash the pellet with 1 mL of ddH2O and collect the pellet by centrifugation (10 min, 12,000  g, 4  C). 3. Repeat step 2. 4. Resuspend pellets with 1 mL of Washing Buffer 1 containing 10 μg/mL of DNAse I and 10 μg/mL of RNase A. 5. Incubate resuspended pellets for 15 min at 25  C and for 15 min at 4  C. Repeat the incubation/temperature cycles three times. 6. Centrifuge for 10 min at 12,000  g at 4  C. 7. Resuspend pellets in 1 mL of Washing Buffer 2 and incubate for 30 min at 4  C (see Note 12). 8. Centrifuge for 10 min at 12,000  g at 4  C and resuspend the pellet in 500 μL of ddH2O. Repeat step 7 until the 260 nm optical density of supernatants becomes close to zero. Hydrophobic overexpressed proteins can then be solubilized in detergent micelles. We have tested several solubilizing molecules such as dodecyl maltoside, digitonin, octyl glucoside, dodecylphosphocholine, and the lysolipid 1-myristoyl-2-hydroxy-sn-glycero3-phospho-10 -rac-glycerol (LMPG). LMPG was found to be a good solubilizing molecule for all the yeast ATP-synthase CF-produced subunits, and is also a surfactant of choice for NMR studies due to its small micelles. All experiments below were therefore performed with this molecule, at a LMPG/protein molar ratio of 110 corresponding to a molar ratio of LMPG micelles/protein of 2 (see Note 13). This condition is also appropriate for quick studies by NMR or Circular Dichroı¨sm (CD) to confirm that solubilized proteins are folded.

3.3.2 Sample Preparation for CD

1. After the cleaning of protein precipitates obtained from 1 mL of reaction mix, resuspend the pellet in 500 μL of ddH2O, vortex the tube prior to pipetting the desired volume (usually 3–5 μL) (see Note 14), and estimate the protein concentration by the Lowry method [43].

74

Isabelle Larrieu et al.

2. Vortex and aspirate the volume required to prepare 200 μL of a protein solution at 15 μM. This is adapted for proteins with a molecular weight around 10 kDa. The concentration has to be adjusted to the protein size (see Note 15). 3. Centrifuge (12,000  g, 10 min, 4  C) and discard the supernatant. 4. Resuspend in LMPG Buffer containing 1.65 mM of LMPG. 5. Incubate for 1 h at 30  C in a thermostatic bath. Vortex every 10 min (see Note 16). 6. Centrifuge for 10 min at 20,000  g. 7. Transfer the supernatant in a 1 mm path length cell (see Note 15) and acquire ten spectra from 190 to 250 nm with 1 nm intervals using a CD spectropolarimeter (Jasco J-810). Correct spectra from baseline by subtracting the LMPG buffer spectrum and average them.

θ (mdeg)

Figure 3 shows CD spectra obtained from several ATP-synthase hydrophobic subunits solubilized in LMPG and the degree of purity for each protein. Once purity has been assessed and CD spectra have been acquired to check if solubilized proteins display folded elements, structural or interactional studies can be performed. In the next chapter, we give an example of sample

40

kDa

M f

e

g

i

S4T

f

180 130

30

e

100 70

g

55 40

20

35

i

25 15

10 0 190

S4T

10

200

210

220

230

240

250

nm

-10 -20 -30

Fig. 3 Overexpression of several ATP-synthase hydrophobic subunits and CD spectra of LMPG solubilized proteins. Each subunit was produced in a 1 mL reaction volume. Pellets were washed as indicated in Subheading 3.3.1 and resuspended at 15 μM in the presence of 1.65 mM LMPG. Far-UV spectra were acquired on a Jasco J-810 CD spectrometer. The acquisition was carried out at 25  C with a 1 mm optical path length cell. Ten scans were collected from 250 to 190 nm at 1 nm intervals, baseline-corrected by subtracting the buffer spectrum and averaged

In Vitro Expression of Yeast ATP-Synthase Subunits

75

preparation to acquire NMR data on subunit g after expressing the protein in the presence of 13C and 15N labeled amino acids (see Note 17). 3.3.3 Preparation of CFProduced Proteins for NMR Studies

Sample Preparation

As a biophysical tool, NMR can be used to provide general descriptions of protein folding, stability and homogeneity as well as atomic-level information regarding protein structure, dynamics, and molecular interactions. However, to facilitate atomic-level observations, proteins should be synthesized using isotopically enriched amino acids (13C, 15N), in place of the naturally abundant forms, in both the standard amino acid stock and the additional amino acids (R, C, W, M, D, E) found in Table 2. The incorporation of specific isotopically enriched amino acids can, of course, be tailored to cost or experiment design. 1. Recover overexpressed membrane proteins from 1 mL of reaction mix by centrifugation (10 min, 12,000  g, 4  C). Discard the supernatant. 2. Remove any remaining nucleic acids according to subheading 3.3.1. 3. Resuspend the sample in 50 μL ddH2O and estimate the protein concentration using the Lowry method. 4. Prepare a stock solution of LMPG, in LMPG buffer with 10% D2O, to obtain a final lysolipid/protein ratio of 110 within a volume appropriate for the intended NMR sample tube (200–600 μL). Suitable protein concentrations for NMR analyses are typically between 0.1 and 1 mM. 5. Harvest the remaining protein resuspension by centrifugation (10 min, 12,000  g, 4  C) and aspirate all remaining supernatants. 6. Resuspend the sample in the appropriate volume of LMPG stock solution. 7. Incubate for 1 h at 30  C in a thermostatic bath. Vortex every 10 min. 8. Centrifuge for 10 min at 20,000  g to remove any remaining insoluble material. 9. Transfer the supernatant to a centrifuge tube and supplement with 100 μM Trimethylsilyl propanoic acid (TSP) for use as an internal chemical shift reference (see Note 18). Transfer sample to NMR tube for experimentation.

NMR Data

Figure 4A shows the 1H-15N SOFAST HMQC spectrum at 800 MHz of subunit g resuspended in LMPG micelles. Over 90% of the backbone amide resonances were specifically assignable using the classical backbone walk methodology. The acquisition of the

76

Isabelle Larrieu et al.

Fig. 4 NMR data for subunit g: (A) 1H-15N SOFAST HMQC spectrum of subunit g, with residue-specific assignments, at 800 MHz. (B) The Secondary structure propensity (SSP) score of subunit g determined from NMR chemical shifts (blue) compared against the in silico predicted secondary structure (PSIPRED, top) and transmembrane regions (TM HMM, black bars)

In Vitro Expression of Yeast ATP-Synthase Subunits

77

backbone chemical shifts also allowed probing the secondary structure of subunit g. NMR structural propensity scores [44] derived from the H, Hα, NH, Cα, Cβ, and CO chemical shifts describe subunit g as a predominantly α-helical protein. The positioning and length of the alpha helical regions is also in good agreement with a PSIPRED secondary structure prediction [45] and TM HMM transmembrane prediction [46] from the amino acid sequence alone (Fig. 4B). The consensus between the in silico predicted results and NMR determined secondary structure therefore suggests that the LMPG micelle environment is such that it suitably allows for typical secondary structure elements of subunit g to form successfully. 3.4 Use of CFExpressed Proteins for Immunization

Proteins produced in vitro can be used to obtain polyclonal antibodies, and below we present three such examples. The first describes the protocol used to prepare sufficient amounts of a His-tagged ATP-synthase subunit k that was found in both the pellet and supernatant fractions after CF production. The two other examples, N-terminal part of subunit 4 (S4T) and subunit e, are appropriate for untagged hydrophobic subunits.

3.4.1 Subunit k Expression

Sequence analysis of subunit k indicates that it contains a hydrophobic region but alkaline extraction experiments on mitochondria have shown that it was not strongly bound to the mitochondrial membrane [19]. The ATP19 gene encoding the yeast ATP-synthase subunit k was inserted in the pIVEX2.3 vector upstream of a His6-tag coding sequence. This construct (pIVEX2.3-ATP19-HC) allows the expression of the 68 residue long protein with a short linker (PGGGS) followed by the His6-tag (Table 1). CF expression was first assessed after expression in a Flip Tube® as described in Subheading 3.2.2, except that a membrane with a 3500 MWCO was used to avoid any partitioning of the produced protein (8.7 kDa) in the Feeding Mix compartment (see Note 19). As shown on the silver nitrate stained SDS-PAGE gel (Fig. 5A), a band corresponding to the size of the his-tagged protein (kHis6) was present in the pellet fraction obtained from the reaction mix containing the pIVEX2.3-ATP19-HC construct but absent in the pellet of the control experiment with the empty plasmid. From this gel, it was not possible to determine if subunit kHis6 was also present in the supernatant fraction as a strong band at the expected size of the kHis6 protein was observed in supernatants obtained with the empty plasmid and with the pIVEX2.3-ATP19-HC construct. Western blot analysis with antibodies directed against the His6tag revealed that a his-tagged protein, with the size of the expected kHis6 protein, was present in the supernatant and the pellet fractions (Fig. 5B). Using antibodies obtained with a subunit k peptide, it

78

Isabelle Larrieu et al.

A

kDa

M Po1 Pk1 Po3 Pk3

So

Sk

70 55 40 35 25 15 10 -

B kDa

M So

Sk Pk L1

M So

Sk Pk L2

55 40 -

35 25 -

# ** 15 -

* kHis6 k

10 -

anti-His6

anti-k-peptide

Fig. 5 CF expression of subunit kHis6. (A) Control of subunit kHis6 expression. 1/ 100th (Po1 and Pk1) or 1/33th (Po3 and Pk3) of pellet fractions from a small-scale CF expression with the empty vector (Po) or the plasmid containing the kHis6 encoding sequence (Pk) were analyzed on a 15% Tris-Tricine SDS-PAGE gel, stained with the ProteoSilver® kit and compared with 1 μL of the supernatant fractions of the control (So) or the kHis6 expression experiment (Sk). (B) Western blot analysis of kHis6 production with antibodies directed against the His-tag (titer: 1/8000) or directed against subunit k-peptide (titer: 1/1000). M: Markers, So: supernatant fraction from the experiment with the empty plasmid (1/100th of the fraction); Sk: supernatant from the subunit kHis6 CF expression (1/100th of the fraction); Pk: pellet from the subunit kHis6 CF expression (1/100th of the fraction); L1: 50 μg of a mitochondrial lysate from a strain with a His-tag on subunit i (his-tag positive control, ♭). L2: 75 μg of a mitochondrial lysate from a wild-type strain. The right membrane has been exposed a longer time to see the faint signal corresponding to the untagged subunit k. #: protein recognized by the anti-k-peptide

In Vitro Expression of Yeast ATP-Synthase Subunits

79

could be confirmed that this band was the kHis6 protein. The anti-kpeptide antibodies also recognized another protein with a size below 15 kDa (Fig. 5B, *), but as this signal was also observed in the supernatant from the control experiment (empty plasmid), it can be excluded that this additional band corresponds to an oligomeric form of the kHis6 protein. However, a band above 15 kDa (Fig. 5B, **) that was detected with the anti-his antibodies in the “Sk” supernatant but not in the control supernatant could correspond to a kHis6 dimer. As the kHis6 protein of the pellet fraction was relatively pure (Fig. 5A), this fraction was used to generate new antibodies against the full subunit k. Also no other proteins could be observed in the washed pellet fraction on the silver stained gel, an additional purification step was performed to ensure that very low amounts of contaminants were present in the sample used for immunization. The kHis6 protein was produced from 3 mL of reaction mix. The following protocol can be used for moderately hydrophobic proteins that are his-tagged and, like subunit k, are found to partition between the supernatant and pellet fractions. 1. Collect precipitated proteins by centrifugation (10 min, 12,000  g) in two 2 mL tubes. 2. Wash pellets with 1.5 mL of ddH2O per tube. 3. Centrifuge (10 min, 12,000  g), remove supernatants, and repeat steps two and three, twice. 4. Prepare Ni-NTA agarose beads (Qiagen) in two 50 mL Falcon tubes by adding in each tube 1.5 mL of slurry and wash with 2 mL of IMAC Binding Buffer. 5. Centrifuge for 5 min at 400  g at 4  C and discard the supernatant. 6. Repeat the washing of Ni-NTA agarose beads twice. 7. Resuspend each protein pellet with 1.5 mL of ddH2O and transfer each suspension in a Falcon tube containing the washed Ni-NTA agarose beads. 8. Add 13.5 mL of binding buffer per tube. 9. In each tube, add 825 μL of a 20% (w/v) SDS solution to get a final SDS-concentration around 1%. 10. Incubate for 1 h at temperatures between 13 and 18  C on a rotating wheel (10–15 rpm) (see Note 20). 11. Spin at 400  g for 5 min at 13  C, carefully remove the supernatant. 12. Wash the beads five times with: (a) 28.5 mL of binding buffer þ1.5 mL of 20% (w/v) SDS, incubate on the rotating wheel 10 min at 13–18  C, spin at 400  g for 5 min at 13  C and remove the supernatant.

80

Isabelle Larrieu et al.

(b) 29.25 mL of binding buffer þ0.75 mL of 20% (w/v) SDS, incubate on the rotating wheel 10 min at 13–18  C, spin at 400  g for 5 min at 13  C and remove the supernatant. (c) 29.62 mL of binding buffer þ0.375 mL of 20% (w/v) SDS, incubate on the rotating wheel 10 min at 13–18  C, spin at 400  g for 5 min at 13  C, and remove the supernatant. (d) 30 mL of binding buffer, incubate on the rotating wheel 10 min at 13–18  C, spin at 400  g for 5 min at 13  C, and remove the supernatant. (e) 30 mL of binding buffer, incubate on the rotating wheel for 10 min at 13–18  C, spin at 400  g for 5 min at 13  C, and discard the supernatant. 13. Add 600 μL of IMAC Elution Buffer in each Falcon tube. At this step, transfer into a 2 mL tube. Incubate on the rotating wheel (10–15 rpm) at 4  C for 5 min. Centrifuge for 10 min at 5000  g, 4  C. Collect the supernatant (see Note 21). 14. Repeat step 13 two more times. 15. Pool the three eluted fractions. 16. Dialyze the pooled sample against 500 mL of D-Buffer, in a dialysis tubing adapted to the size of the protein (a membrane with a MWCO 3500 kDa was used for subunit k) for at least 2 h at 4  C. 17. Change the dialysis buffer (500 mL of D-Buffer) and dialyze for 2–3 h at 4  C. 18. Collect the dialyzed fraction and determine the protein concentration spectrophotometrically at 280 nm. The extinction coefficient is determined from the sequence of the overexpressed protein. With this protocol 2.3 mg of subunit kHis6 could be purified. 19. Concentrate the sample with a Centricon unit (for a small protein like subunit k, a YM-3 (Amicon) with a MWCO of 3000 kDa was used). 20. Reestimate the protein concentration after the concentration step. 1.6 mg of subunit k could be recovered at this step in a final volume of 1 mL. 21. Dispatch the concentrate into Eppendorf tubes (200 μg/ tube) and dry the protein sample with a speed-vacuum at a medium drying rate. 22. Keep the dried sample at 20  C until rabbit immunization.

In Vitro Expression of Yeast ATP-Synthase Subunits 3.4.2 Examples with Hydrophobic Proteins Devoid of Histidine Tag

81

The sequences encoding the N-terminal extremity of subunit 4 (S4T) or subunit e were inserted in pIVEX plasmids (Table 1). The produced proteins had no his6-tag. They were overexpressed in 1 mL of Reaction Mix in the precipitate mode. After the treatments of the pellets to remove DNA and RNA, as indicated in Subheading 3.3.1, proteins were resuspended in 500 μL of ddH2O and their concentrations were estimated using the Lowry method. Around 2.2 mg of S4T and 1.4 mg of subunit e could be obtained. To raise antibodies, an additional step was required to obtain a high degree of purity. 1. Transfer the equivalent of 1 mg of protein into an Eppendorf tube. 2. After a centrifugation (12,000  g, 10 min, 4  C) resuspend the protein pellet in 250 μL of L-Buffer containing 30 mg/mL of LMPG. 3. Incubate in a water bath for 1 h at 30  C to allow protein solubilization. 4. Pre-equilibrate a Superdex75 10/300 column (GE Healthcare) mounted on a FPLC system with Superdex Buffer. 5. Centrifuge protein-LMPG samples for 20 min at 20,000  g to remove any insoluble materials. 6. Inject 200 μL of the supernatant onto the Superdex 75 column. 7. Elute at a flow rate of 0.5 mL/min with the Superdex buffer. 8. Collect 1.5 mL fractions. 9. Analyze 15 μL of each fraction on a SDS-PAGE gel and stain the gel with silver nitrate staining kit (commercially available or according to Ansorge et al. [47]). Proteins are usually diluted seven times after the elution step. 10. Concentrate the sample (around 1500 μL) to a final volume of 200 μL using a concentration unit (with a MWCO adapted to the protein). 11. Estimate protein concentration with the Lowry method using the Superdex Buffer as blank. 12. Dispense the protein sample in order to have 100–200 μg of protein per tube. 13. Add TCA to a final concentration of 0.3 M, vortex and centrifuge immediately for 5 min at 10,000  g. 14. Discard the supernatant. 15. Wash the pellet in 500 μL acetone. 16. Repeat acetone wash twice. 17. Air-dry the pellet.

82

Isabelle Larrieu et al.

OD280nm 0.13

M

3

0.11

BI

1

2

3

0.09 0.07 0.05 0.03

2

0.01 -0.01

1 0

10

20

30

40

50

60

Time (min)

Fig. 6 Purification of the N-terminal extremity of subunit 4 (S4T) by Size Exclusion Chromatography. S4T produced in vitro was treated with DNAse and RNase and resuspended in the presence of LMPG at a LMPG/protein molar ratio of 110 and purified on a Superdex 75 column. The insert shows the analysis of FPLC fractions on a 15% Tris-Glycine SDS-PAGE silver-stained gel. M: 1.5 μL of Markers. BI: 2 μL of the injected fraction. 1, 2 and 3: 15 μL of fraction 1, 2 and 3. Fraction 3 was used for immunization

The chromatogram corresponding to S4T purification on Superdex 75 is shown in Fig. 6 and the degree of purity of S4T is presented in the insert of this figure. Figures 7A and 7B show, respectively, the degree of purity of subunit e and subunit k used for immunization. Anti-subunit k, anti-S4T and anti-e antibodies were then tested for each taking and compared, when possible, to antibodies directed against chemically synthesized peptides. Figure 8 shows that the anti-kHis6 antibodies are very specific as they do not react with other proteins in total yeast extracts (Fig. 8A). They recognize a band corresponding to subunit k in the wildtype extracts and as expected, there is no protein detected in the extract from a strain devoid of subunit k. As anticipated, the signal for subunit k is stronger than the one observed with the antibodies directed against a k-peptide (Fig. 8B). The antibodies directed against the N-terminal part of subunit 4 recognize the subunit 4 in mitochondrial and in total cell extracts (Fig. 9). No bands are detected in the total extracts from a strain devoid of subunit 4.

In Vitro Expression of Yeast ATP-Synthase Subunits

A kDa M

83

B Po

Pe

BI

Pue

M

Po

Puk

180 130 100 70 55 40 35 25 -

15 -

10 -

Fig. 7 Purity of subunit e (A) and kHis6 (B) used for immunization. Subunit e was produced in vitro and analyzed (A) on a 15% acrylamide SDS-PAGE Tris-Tricine gel stained with silver nitrate (ProteoSilver, Sigma-Aldrich). M: Markers, Po: 1/100th of the pellet fraction obtained from a control expression with the empty pIVEX2.3. Pe: 1/100th of the pellet fraction obtained from the expression with the pIVEX2.3-ATP21. BI: 2 μL of the sample injected on Superdex75. Pue: 4.5 μg of the purified and concentrated subunit e used for immunization. B, Puk: Fraction containing the kHis6 (9 μg) after the IMAC purification

The antibodies obtained against the CF expressed subunit e recognized a band at the same size in CF pellet fraction and in yeast total extracts (Fig. 10a) and the signals observed are more intense than with anti-e-peptide antibodies previously obtained (Fig. 10b).

4

Conclusion The cell-free expression system is a very good alternative to recombinant protein production in E. coli, as shown here for yeast ATPsynthase hydrophobic subunits that were either toxic for bacteria (subunit f), poorly expressed (subunit g) or that formed inclusion bodies from which it was impossible to get a soluble or refolded form of protein (subunit e). In vitro expression in the precipitate mode has allowed protein production and purification, in milligram

84

Isabelle Larrieu et al.

A kDa

M Dk

B

WT Pk

kDa

M ME

TE

M ME TE

34 – 26 –

25 –

-

14 –

anti-k(kHis6) 1/5,000

- kHis6 -k

15 – 10 –

anti-k(peptide) 1/1,000

anti-k(kHis6) 1/1,000

Fig. 8 Analysis of anti-kHis6 antibodies and comparison with the anti-k-peptide antibody. (A) Total extracts from a strain devoid of subunit k (Δk, 100 μg) or from the wild-type strain (WT, 100 μg) were separated on a 15% Tris–Tricine gel, transferred on a nitrocellulose membrane, and detected with the anti-k-peptide antibodies. 1/100th of the pellet fraction from a 1 mL CF expression of the kHis6 protein was also analyzed on the same gel (Pk). M: Markers, (Euromedex) (B) Comparison of the signals obtained with the two anti-subunit k antibodies. Proteins from mitochondrial extracts (ME, 50 μg) or total yeast extracts (TE, 60 μg) were analyzed on a 12% Tris–Glycine SDS-PAGE gel, transferred on a nitrocellulose membrane, and detected with anti-k-peptide antibodies or antikHis6 antibodies. Sera were used at the same titer. M Markers

amounts and to a degree of purity that is compatible with biophysical analyses such as CD and NMR spectroscopy. This opens the opportunity to study the structure of these isolated proteins by NMR or X-ray crystallography. Once the structures of these proteins will be established, they will be placed in the cryo-EM density volume of dimeric species obtained in our group (unpublished data) or in other laboratories [25]. This will help us to understand the interactions at the dimeric interface. Furthermore, as subunits e, g, and k have been produced in sufficient amounts, the steps of dimer assembly can now be studied from monomeric purified enzymes and CF overexpressed dimer-specific proteins. As shown for yeast ATP-synthase subunits, proteins produced in vitro are easy, quick, and cheap to prepare and one step of purification is usually sufficient to generate very specific antibodies against moderately or strongly hydrophobic proteins.

In Vitro Expression of Yeast ATP-Synthase Subunits TEΔ4 TEWT TEWT ME P4

kDa

80

90

45 100

25

85

Po

μg

- Su 4 (23kDa)

15

- S4T (8.7kDa) 10

anti-S4T 1/5,000

Fig. 9 Analysis of anti-S4T antibodies. Proteins from total yeast extracts (TE) or from mitochondria lysates (ME) were separated on a 15% SDS-PAGE Tris-Tricine gel, transferred on a nitrocellulose membrane and revealed with antibodies directed against S4T. TEΔ4: total yeast extracts from a strain devoid of subunit 4; TEWT: total yeast extracts from a wild-type strain. ME: wild-type mitochondrial extract. P4, Po: 1/100th of the pellet fraction from CF-experiments with the pIVEX2.3-S4T construct or with the empty plasmid, respectively

5

Notes 1. Several DNA preparation kits were used but the Qiagen kit led to better and reproducible expression levels. 2. T7-expressing cells can be stored at 80  C. 3. At this step, a lot of T7 RNA polymerase is lost in the supernatant fraction but this step is necessary to obtain an enzyme devoid of ~45 kDa contaminants. 4. As T7 RNA polymerase is stabilized by DTT, 2 μL of 0.5 M DTT can be added to 20 μL T7 RNA polymerase aliquots, prior to adding the polymerase to the reaction mix. 5. Fliptubes® contain exactly 100 μL in their cap compartment. The diameter of the cap compartment is bigger than an Eppendorf tube and thus the exchange surface through the

86

Isabelle Larrieu et al.

A

B

Pe TE

Po

Pe

M

kDa

Po

Pe

M

25 -

15 -

10 Anti-eCF 1/5,000

Anti-eCF 1/5,000

Anti-e-peptide 1/5,000

Fig. 10 Analysis of anti-e antibodies. Proteins were separated on a 15% SDSPAGE Tris-Tricine gel, transferred on a nitrocellulose membrane and revealed with antibodies obtained from the CF-produced subunit e (A and left part of B) or with antibodies directed against a subunit e peptide (right part of B). Pe: 1/100th of the pellet fraction from a CF expression with the pIVEX2.3-ATP21 plasmid. TE: total extract from a wild type strain (60 μg). Po: 1/100th of the pellet fraction from a 1 mL CF expression with the empty plasmid

membrane is larger. Furthermore, the cap part can be easily detached from the tube, making their use highly practical for CF expression. 6. Higher temperatures can be used. 37  C usually improves the yield of expression but we have noticed that NMR data were better for proteins produced at 28  C compared to 37  C. For structural analyses, we recommend using a temperature of 28  C. 7. There is no need to extend the protein production up to 24 h, as there is no yield improvement between 20 and 24 h. 8. For membrane proteins in the precipitate mode, the Reaction Mix should be cloudy. 9. Controlling an expression experiment and determining whether the protein is produced in the pellet or in the supernatant can be performed in less than 24 h. 10. The swapping strategy from a pET vector to a pIVEX vector has to be tried only if low amounts of proteins are produced.

In Vitro Expression of Yeast ATP-Synthase Subunits

87

11. The cost of 1 mL CF expression with a homemade kit is around 60 € with unlabeled amino acids and 300 € with isotopically enriched amino acids (13C, 15N). 12. Samples can also be incubated overnight at 4  C in this buffer. An overnight incubation of samples often allows a better cleaning. 13. The best LMPG/protein molar ratio can be estimated as the ratio above which no more change is observed in CD spectra and the concentration at which there is no more protein in the pellet fraction after 1 h incubation at 30  C. For most of the ATP-synthase subunits, we have found that molar ratios between 42 and 110 were appropriate. We have chosen to use a ratio of 110 for NMR data. As LMPG aggregation number is 55, it corresponds to a micelle/protein molar ratio of 2. 14. At this step, due to the aggregated form of proteins, a 280 nm spectrophotometric quantification is not feasible. A colorimetric dosage is then the best way to estimate protein concentration. From the precipitate obtained from 1 mL of reaction mix, resuspend the washed pellet in 500 μL of ddH2O. It is essential to vortex the tube just prior to pipetting and to triplicate or quadruplicate the pipetting to get the best estimate of protein concentration. 15. The concentrations given here are adapted for proteins between 6 and 15 kDa with a 1 mm cell and with the Jasco810 spectropolarimeter. Concentrations have to be adapted to the protein molecular weight, the cell, and the sensitivity of the equipment. 16. The solution should be cloudy at the beginning of the incubation and very quickly become transparent. 17. If NMR studies have to be performed with LMPG solubilized proteins, it is important to record CD spectra at the temperature at which NMR spectra will be acquired to ensure that at this temperature, there is no important change in the solubilized protein structure. 18. Sodium azide, to a final concentration of 0.03% (w/v), may also be added to the NMR sample to help prevent bacterial growth. 19. With a membrane with a low molecular weight cutoff, the yields of expression are often lower than with a cutoff around 10–12 kDa. 20. Lowest temperatures would lead to SDS flocculation. 21. Subunit k was found in a soluble form after the binding of the SDS-solubilized pellet and elution of the protein in the absence of any detergent. This could be due to the presence of remaining SDS molecules that could mask hydrophobic regions, to a

88

Isabelle Larrieu et al.

proper refolding of subunit k or to an interaction of two subunits k by their hydrophobic region, allowing the kHis6 dimer to remain soluble. The protein was also soluble after step 17.

Acknowledgment We would like to thank Franck Bernhard (Institute of Biophysical Chemistry, Goethe-Universit€at Frankfurt am Main) for helpful advises concerning the CF-expression system. This work was supported by the F1Fo-Struct ANR grant (ANR-12-BSV8-024). We thank D. Bre`thes for a careful reading of the manuscript. References 1. Spirin AS, Baranov VI, Ryabova LA, Ovodov SY, Alakhov YB (1988) A continuous cell-free translation system capable of producing polypeptides in high yield. Science 242:1162–1164 2. Kigawa T, Yokoyama S (1991) A continuous cell-free protein synthesis system for coupled transcription-translation. J Biochem 110:166–168 3. Rothblatt JA, Meyer DI (1986) Secretion in yeast: reconstitution of the translocation and glycosylation of alpha-factor and invertase in a homologous cell-free system. Cell 44:619–628 4. Madin K, Sawasaki T, Ogasawara T, Endo Y (2000) A highly efficient and robust cell-free protein synthesis system prepared from wheat embryos: plants apparently contain a suicide system directed at ribosomes. Proc Natl Acad Sci U S A 97:559–564 5. Harbers M (2014) Wheat germ systems for cell-free protein expression. FEBS Lett 588:2762–2773. doi:10.1016/j.febslet.2014. 05.061 6. Stech M, Quast RB, Sachse R, Schulze C, W€ ustenhagen DA, Kubick S (2014) A continuous-exchange cell-free protein synthesis system based on extracts from cultured insect cells. PLoS One 9:e96635. doi:10.1371/jour nal.pone.0096635 7. Anastasina M, Terenin I, Butcher SJ, Kainov DE (2014) A technique to increase protein yield in a rabbit reticulocyte lysate translation system. BioTechniques 56:36–39. doi:10. 2144/000114125 8. Bro¨del AK, Sonnabend A, Kubick S (2014) Cell-free protein expression based on extracts from CHO cells. Biotechnol Bioeng 111:25–36. doi:10.1002/bit.25013

9. Bro¨del AK, W€ ustenhagen DA, Kubick S (2015) Cell-free protein synthesis systems derived from cultured mammalian cells. Methods Mol Biol 1261:129–140. doi:10.1007/ 978-1-4939-2230-7_7 10. Weber LA, Feman ER, Baglioni C (1975) A cell free system from HeLa cells active in initiation of protein synthesis. Biochemistry 14:5315–5321 11. Zemella A, Thoring L, Hoffmeister C, Kubick S (2015) Cell-free protein synthesis: pros and cons of prokaryotic and eukaryotic systems. Chembiochem 16:2420–2431. doi:10.1002/ cbic.201500340 12. Schwarz D, Junge F, Durst F, Fro¨lich N, Schneider B, Reckel S, Sobhanifar S, Do¨tsch V, Bernhard F (2007) Preparative scale expression of membrane proteins in Escherichia colibased continuous exchange cell-free systems. Nat Protoc 2:2945–2957. doi:10.1038/ nprot.2007.426 13. Ma Y, M€ unch D, Schneider T, Sahl H-G, Bouhss A, Ghoshdastider U, Wang J, Do¨tsch V, Wang X, Bernhard F (2011) Preparative scale cell-free production and quality optimization of MraY homologues in different expression modes. J Biol Chem 286:38844–38853. doi:10.1074/jbc.M111.301085 14. Henrich E, Do¨tsch V, Bernhard F (2015) Screening for lipid requirements of membrane proteins by combining cell-free expression with nanodiscs. Methods Enzymol 556:351–369. doi:10.1016/bs.mie.2014.12.016 15. Matthies D, Haberstock S, Joos F, Do¨tsch V, Vonck J, Bernhard F, Meier T (2011) Cell-free expression and assembly of ATP synthase. J Mol Biol 413:593–603. doi:10.1016/j.jmb. 2011.08.055

In Vitro Expression of Yeast ATP-Synthase Subunits 16. Rak M, Gokova S, Tzagoloff A (2011) Modular assembly of yeast mitochondrial ATP synthase. EMBO J 30:920–930. doi:10. 1038/emboj.2010.364 17. Paumard P, Vaillier J, Coulary B, Schaeffer J, Soubannier V, Mueller DM, Bre`thes D, di Rago J-P, Velours J (2002) The ATP synthase is involved in generating mitochondrial cristae morphology. EMBO J 21:221–230. doi:10. 1093/emboj/21.3.221 18. Davies KM, Daum B, Gold VAM, M€ uhleip AW, Brandt T, Blum TB, Mills DJ, K€ uhlbrandt W (2014) Visualization of ATP synthase dimers in mitochondria by electron cryotomography. J Vis Exp 91:51228. doi:10. 3791/51228 19. Arnold I, Pfeiffer K, Neupert W, Stuart RA, Sch€agger H (1998) Yeast mitochondrial F1F0-ATP synthase exists as a dimer: identification of three dimer-specific subunits. EMBO J 17:7170–7178. doi:10.1093/emboj/17.24. 7170 20. Soubannier V, Vaillier J, Paumard P, Coulary B, Schaeffer J, Velours J (2002) In the absence of the first membrane-spanning segment of subunit 4(b), the yeast ATP synthase is functional but does not dimerize or oligomerize. J Biol Chem 277:10739–10745. doi:10.1074/jbc. M111882200 21. Stock D, Leslie AG, Walker JE (1999) Molecular architecture of the rotary motor in ATP synthase. Science 286:1700–1705 22. Dautant A, Velours J, Giraud M-F (2010) Crystal structure of the Mg·ADP-inhibited state of the yeast F1c10-ATP synthase. J Biol Chem 285:29502–29510. doi:10.1074/jbc. M110.124529 23. Giraud M-F, Paumard P, Sanchez C, Bre`thes D, Velours J, Dautant A (2012) Rotor architecture in the yeast and bovine F1-c-ring complexes of F-ATP synthase. J Struct Biol 177:490–497. doi:10.1016/j.jsb.2011.10. 015 24. Allegretti M, Klusch N, Mills DJ, Vonck J, K€ uhlbrandt W, Davies KM (2015) Horizontal membrane-intrinsic α-helices in the stator asubunit of an F-type ATP synthase. Nature 521:237–240. doi:10.1038/nature14185 25. Hahn A, Parey K, Bublitz M, Mills DJ, Zickermann V, Vonck J, K€ uhlbrandt W, Meier T (2016) Structure of a complete ATP synthase dimer reveals the molecular basis of inner mitochondrial membrane morphology. Mol Cell 63:445–456. doi:10.1016/j.molcel.2016.05. 037 26. Paumard P, Arselin G, Vaillier J, Chaignepain S, Bathany K, Schmitter JM, Bre`thes D, Velours J

89

(2002) Two ATP synthases can be linked through subunits i in the inner mitochondrial membrane of Saccharomyces cerevisiae. Biochemistry 41:10390–10396 27. Giraud M-F, Paumard P, Soubannier V, Vaillier J, Arselin G, Salin B, Schaeffer J, Bre`thes D, di Rago J-P, Velours J (2002) Is there a relationship between the supramolecular organization of the mitochondrial ATP synthase and the formation of cristae? Biochim Biophys Acta 1555:174–180 28. Wagner K, Perschil I, Fichter CD, van der Laan M (2010) Stepwise assembly of dimeric F(1)F (o)-ATP synthase in mitochondria involves the small F(o)-subunits k and i. Mol Biol Cell 21:1494–1504. doi:10.1091/mbc.E09-121023 29. Studier FW, Rosenberg AH, Dunn JJ, Dubendorff JW (1990) Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol 185:60–89 30. Egner R, Mahe´ Y, Pandjaitan R, Kuchler K (1995) Endocytosis and vacuolar degradation of the plasma membrane-localized Pdr5 ATPbinding cassette multidrug transporter in Saccharomyces cerevisiae. Mol Cell Biol 15:5879–5887 31. Gue´rin B, Labbe P, Somlo M (1979) Preparation of yeast mitochondria (Saccharomyces cerevisiae) with good P/O and respiratory control ratios. Methods Enzymol 55:149–159 32. Palmer AG, Fairbrother WJ, Cavanagh J, Wright PE, Rance M (1992) Improved resolution in three-dimensional constant-time triple resonance NMR spectroscopy of proteins. J Biomol NMR 2:103–108 33. Grzesiek S, Bax A (1992) Improved 3D tripleresonance NMR techniques applied to a 31 kDa protein. J Magn Reson (1969) 96:432–440. doi:10.1016/0022-2364(92) 90099-S 34. Wittekind M, Mueller L (1993) HNCACB, a high-sensitivity 3D NMR experiment to correlate amide-proton and nitrogen resonances with the alpha- and Beta-carbon resonances in proteins. J Magn Reson 101:201–205. doi:10. 1006/jmrb.1993.1033 35. Muhandiram DR, Kay LE (1994) Gradientenhanced triple-resonance three-dimensional NMR experiments with Improved sensitivity. J Magn Reson B 103:203–216. doi:10.1006/ jmrb.1994.1032 36. Grzesiek S, Ikura M, Marius Clore G, Gronenborn AM, Bax A (1992) A 3D triple-resonance NMR technique for qualitative measurement of carbonyl-Hβ J couplings in isotopically enriched proteins. J Magn Reson (1969)

90

Isabelle Larrieu et al.

96:215–221. doi:10.1016/0022-2364(92) 90307-S 37. Clubb RT, Thanabal V, Wagner G (1992) A constant-time three-dimensional triple-resonance pulse scheme to correlate intraresidue 1HN, 15N, and 13C0 chemical shifts in 15N13C-labelled proteins. J Magn Reson (1969) 97:213–217. doi:10.1016/0022-2364(92) 90252-3 38. Grzesiek S, Bax A (1993) Amino acid type determination in the sequential assignment procedure of uniformly 13C/15N-enriched proteins. J Biomol NMR 3:185–204 39. Delaglio F, Grzesiek S, Vuister GW, Zhu G, Pfeifer J, Bax A (1995) NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J Biomol NMR 6:277–293 40. Vranken WF, Boucher W, Stevens TJ, Fogh RH, Pajon A, Llinas M, Ulrich EL, Markley JL, Ionides J, Laue ED (2005) The CCPN data model for NMR spectroscopy: development of a software pipeline. Proteins 59:687–696. doi:10.1002/prot.20449 41. Wishart DS, Bigam CG, Yao J, Abildgaard F, Dyson HJ, Oldfield E, Markley JL, Sykes BD (1995) 1H, 13C and 15N chemical shift referencing in biomolecular NMR. J Biomol NMR 6:135–140 42. Roge´ J, Betton J-M (2005) Use of pIVEX plasmids for protein overproduction in

Escherichia coli. Microb Cell Factories 4:18. doi:10.1186/1475-2859-4-18 43. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with the Folin phenol reagent. J Biol Chem 193:265–275 44. Marsh JA, Singh VK, Jia Z, Forman-Kay JD (2006) Sensitivity of secondary structure propensities to sequence differences between alpha- and gamma-synuclein: implications for fibrillation. Protein Sci 15:2795–2804. doi:10. 1110/ps.062465306 45. McGuffin LJ, Bryson K, Jones DT (2000) The PSIPRED protein structure prediction server. Bioinformatics 16:404–405 46. Sonnhammer EL, von Heijne G, Krogh A (1998) A hidden Markov model for predicting transmembrane helices in protein sequences. Proc Int Conf Intell Syst Mol Biol 6:175–182 47. Ansorge W (1985) Fast and sensitive detection of protein and DNA bands by treatment with potassium permanganate. J Biochem Biophys Methods 11:13–20 48. Arselin G, Giraud M-F, Dautant A, Vaillier J, Bre`thes D, Coulary-Salin B, Schaeffer J, Velours J (2003) The GxxxG motif of the transmembrane domain of subunit e is involved in the dimerization/oligomerization of the yeast ATP synthase complex in the mitochondrial membrane. Eur J Biochem 270:1875–1884

Chapter 5 Wheat Germ Cell-Free Overexpression for the Production of Membrane Proteins Marie-Laure Fogeron, Aure´lie Badillo, Franc¸ois Penin, and Anja Bo¨ckmann Abstract Due to their hydrophobic nature, membrane proteins are notoriously difficult to express in classical cell-based protein expression systems. Often toxic, they also undergo degradation in cells or aggregate in inclusion bodies, making delicate issues further solubilization and renaturation. These are major bottlenecks in their structural and functional analysis. The wheat germ cell-free (WGE-CF) system offers an effective alternative not only to classical cell-based protein expression systems but also to other cell-free systems for the expression of membrane proteins. The WGE-CF indeed allows the production of milligram amounts of membrane proteins in a detergent-solubilized, homogenous, and active form. Here, we describe the method to produce a viral integral membrane protein, which is the non-structural protein 2 (NS2) of hepatitis C virus, in view of structural studies by solid-state NMR in a native-like lipid environment. Key words Cell-free protein expression, Wheat germ extract, Detergents, Protein purification, Hepatitis C virus

1

Introduction The wheat germ cell-free (WGE-CF) system consists in using the translation machinery contained in wheat embryos to synthesize proteins in vitro. This system was developed in 1973, when both tobacco mosaic virus RNA and rabbit globin S9 RNA were efficiently translated using commercial wheat germ extract (WGE) [1]. However, this system has not been very popular due to the instability of WGEs [1, 2]. About 15 years ago, Yaeta Endo and colleagues from the Ehime University in Japan demonstrated that WGEs were contaminated by inhibitors during the extraction procedure, e.g., tritin (RNA N-glycosydase), thionins (small basic and cysteine-rich proteins), ribonucleases, desoxyribonucleases [3], and ribosomal RNA apurinic site-specific lyase [4], which were responsible for the WGE instability. It was shown that these inhibitors, which

Jean-Jacques Lacapere (ed.), Membrane Protein Structure and Function Characterization: Methods and Protocols, Methods in Molecular Biology, vol. 1635, DOI 10.1007/978-1-4939-7151-0_5, © Springer Science+Business Media LLC 2017

91

92

Marie-Laure Fogeron et al.

come from the endosperm, could be eliminated by extensive washing of the wheat embryos before extract preparation, leading to an efficient and stable translation system [3]. Further improvements were done few years later, especially with the use of the bilayer method [5]. Additional work showed that proteins produced in the WGE-CF system are functional [6]. The WGE-CF system presents many advantages not only over classical cell-based protein expression systems but also over other cell-free systems. Indeed, like other cell-free protein expression systems, the WGE-CF system overcomes toxicity problems that often occur in classical cell-based expression systems when working with membrane proteins. Moreover, this system does not require any cell lysis. Fewer process steps are therefore required, allowing automation of the system (see, for example, www.cfsciences.com). In addition, this system is open, offering the possibility to work with additives such as chaperones, but also detergents and lipids allowing the expression of membrane proteins directly in a solubilized form. Moreover, cell-free protein expression systems are generally very efficient to produce isotopically labeled proteins for NMR studies [7, 8]. Indeed, only the synthesized protein is isotopically labeled and metabolic scrambling is negligible in comparison to cell-based systems. In comparison to classical cell-based and prokaryotic cell-free expression systems, the WGE-CF system allows a better protein folding for eukaryotic proteins. Indeed, the rate of polypeptide growth on ribosomes differs considerably between eukaryotes and bacteria: it is five to ten times slower in eukaryotes, thus promoting the co-translational folding of proteins [9, 10]. In addition, newly synthesized proteins are stabilized by eukaryotic chaperones that also promote folding [11]. Moreover, since the WGE-CF system contains only negligible amounts of endogenous proteases, the degradation of synthesized proteins is limited. Prokaryotic cell-free lysates neither provide specific eukaryotic folding systems nor posttranslational modifications [12], some of which have been reported in the WGE-CF system, e.g., phosphorylation [13, 14]. Finally, in contrast to prokaryotic systems, the WGE-CF system shows limited codon usage preference and proteins encoded by AT-rich cDNAs can be synthesized without codon optimization. Rabbit reticulocyte lysate (RRL), another eukaryotic cell-free expression system, offers appropriate protein folding and posttranslational modifications too but is limited to analytical protein expression purpose. Codon preference is much looser in the WGE-CF system, expression yields are much higher and the scaleup is much easier as well as less expensive [11]. The major drawback of the WGE-CF system is that the extract preparation, including wheat milling and eye selection of individual germs, takes significantly longer than the preparation of E. coli lysates [12, 15]. This is, however, for complex proteins, outweighed by its high success rate in the preparation of functional protein.

Wheat Germ Cell-Free Expression

93

As mentioned above, the WGE-CF system is particularly interesting in the context of membrane proteins. There are three modes of expression for such proteins using this system, i.e., the so-called precipitate mode, the expression in the presence of detergents, and the expression in the presence of lipids [16]. Although it proved to be efficient, expression of membrane proteins in a precipitate form is not the preferred alternative. Indeed, although the precipitates can often be resolubilized in a folded form using mild detergents, irreversible aggregation or solubilization in a misfolded form could occur. While expression of membrane proteins in the presence of liposomes or nanodiscs is, in principle, very attractive, it requires high lipid-to-protein ratios, which are not suitable in the context of the solid-state NMR structural studies of interest in our lab. For these reasons, we optimized the expression of membrane proteins in the presence of detergents with the aim to reconstitute purified proteins in lipids with low lipid-to-protein ratios required by solid-state NMR. The challenge when using detergents for cellfree membrane protein expression is to find a compromise between good expression level and large extent of solubilization. Mild detergents with a low critical micelle concentration have been reported to allow the successful expression of directly solubilized membrane proteins in cell-free systems based on E. coli lysates [12, 17–19]. However, the use of some of these detergents in a WGECF system has been shown to interfere with expression yields. More detailed information on the best choice of detergent type and concentration for membrane protein expression in the WGECF system is recently becoming available [20–24]. It has been, for example, shown that Fos-choline and CHAPS detergents counteract each other’s inhibitory effects on cell-free translation activity, thus allowing the efficient production and subsequent purification of functional bacteriorhodopsin [20]. In addition, detergent representatives of a new maltose-neopentyl glycol (MNG) amphiphile family have been reported to allow the efficient production of bacteriorhodopsin and viral membrane proteins in a solubilized form [21, 22]. Here, we use the WGE-CF system for the production of the non-structural protein 2 (NS2) of hepatitis C virus (HCV), which is notoriously difficult to overexpress in cell-based systems. NS2 is a 23 kDa integral membrane protein essential for HCV polyprotein processing and virion assembly. This protein is composed of a N-terminal membrane domain believed to comprise three transmembrane segments and a cytosolic C-terminal domain displaying protease activity [25]. In the following, we describe the experimental details for the expression of this protein in a detergent-solubilized form and its purification by affinity chromatography.

94

2

Marie-Laure Fogeron et al.

Materials

2.1 Wheat Germ Extract Preparation

1. Nontreated durum wheat seeds (Sud Ce´re´ales, France). 2. Grain mill F100 (SAMAP, France). 3. Sieve (ProfiLab24, Germany). 4. Hair-drier. 5. Carbone tetrachloride (Sigma-Aldrich, France). 6. Cyclohexane (Sigma-Aldrich, France). 7. RNase AWAY (Molecular BioProducts, USA). 8. Water for molecular biology (Merck Millipore, France). 9. Sonication bath Elmasonic S30H (Elma, Germany). 10. Kitchen blender (mini blender HR2860, Philips, France). 11. Oak ridge centrifuge tubes with sealing cap (Nalgene, USA). 12. JA-20 fixed angle rotor (Beckman Coulter, France). 13. Sephadex G25 fine resin (GE Healthcare, France). 14. 50 mL syringe (Terumo, France). 15. Buffer A containing 80 mM Hepes–KOH pH 7.6, 200 mM potassium acetate, 10 mM magnesium acetate, 4 mM calcium chloride, and 8 mM DTT. 16. SUB-AMIX buffer containing 30 mM Hepes-KOH pH 7.6, 100 mM potassium acetate, 2.7 mM magnesium acetate, 16 mM creatine phosphate, 0.4 mM spermidine, 1.2 mM ATP, 0.25 mM GTP, and 4 mM DTT (CellFree Sciences, Japan). 17. 500 mL centrifuge bottle (Nalgene, USA). 18. JA-10 fixed angle rotor (Beckman Coulter, France). 19. Spin-X UF20 concentrators with a MWCO of 10,000 Da (Corning, USA).

2.2 DNA Preparation and Phenol/ Chloroform Extraction

1. NucleoBond Xtra Maxi kit (Macherey-Nagel, France). 2. Phenol/chloroform/isoamyl alcohol (25/24/1) (SigmaAldrich, France). 3. Chloroform (Sigma-Aldrich, France). 4. Water for molecular biology (Merck Millipore, France). 5. NanoDrop USA).

2.3

Transcription

2000c

spectrophotometer

(ThermoScientific,

1. NTP mix (Promega, Ref. E6000). 2. SP6 RNA Polymerase (CellFree Sciences, Japan). 3. RNase Inhibitor (CellFree Sciences, Japan).

Wheat Germ Cell-Free Expression

95

4. Transcription buffer containing 80 mM Hepes–KOH pH 7.6, 16 mM magnesium acetate, 10 mM DTT and 2 mM spermidine (CellFree Sciences, Japan). 2.4

Translation

1. Home-made wheat germ extract (can be purchased by CellFree Sciences, Japan). 2. Creatine kinase (Roche, France). 3. Feeding buffer containing 30 mM Hepes–KOH pH 7.6, 100 mM potassium acetate, 2.7 mM magnesium acetate, 16 mM creatine phosphate, 0.4 mM spermidine, 1.2 mM ATP, 0.25 mM GTP, and 4 mM DTT (SUB-AMIX NAA, CellFree Sciences, Japan). 4. Cell-free unlabeled amino acid mix (Cambridge Isotope Laboratories, England).

2.5 Protein Sample Preparation for Analysis of Small-Scale Expression Tests

1. Benzonase® Nuclease (Sigma-Aldrich, France). 2. Rolling wheel (Stuart, United Kingdom). 3. Strep-Tactin magnetic beads (IBA Lifesciences, Germany). 4. Magnetic separator (Promega, France). 5. Buffer W containing 100 mM Tris–HCl pH 8.0, 150 mM NaCl and 1 mM EDTA (IBA Lifesciences, Germany). 6. SDS-PAGE loading buffer containing 62.5 mM Tris–HCl pH 6.8, 2% SDS (w/v), 10% glycerol (v/v), 5% β-mercaptoethanol (v/v) and 0.01% bromophenol blue (w/v).

2.6 Purification by Affinity Chromatograhy

1. DDM, n-dodecyl-β-D-maltoside (Anatrace, USA). 2. Strep-Tactin gravity column (IBA Lifesciences, Germany). 3. Buffer W containing 100 mM Tris–HCl pH 8.0, 150 mM NaCl and 1 mM EDTA (IBA Lifesciences, Germany). 4. Buffer E containing 100 mM Tris–HCl pH 8.0, 150 mM NaCl, 1 mM EDTA and 2.5 mM desthiobiotin (IBA Lifesciences, Germany). 5. HABA (IBA Lifesciences, Germany).

2.7 SDS-PAGE and Western Blotting Analysis

1. SDS-PAGE loading buffer containing 62.5 mM Tris–HCl pH 6.8, 2% SDS (w/v), 10% glycerol (v/v), 5% β-mercaptoethanol (v/v) and 0.01% bromophenol blue (w/v). 2. iBlot gel transfer device (Life Technologies, USA). 3. PBS-T buffer containing 12 mM sodium phosphate pH 7.4, 137 mM NaCl, 2.7 mM KCl, 0.05% Tween 20. 4. Strep MAB Classic primary antibody (IBA Lifesciences, Germany).

96

Marie-Laure Fogeron et al.

5. Anti-mouse IgG HRP-conjugated (Promega, France).

secondary

antibody

6. ECL Prime Western Blotting Detection Reagent (GE Healthcare, France).

3

Methods The genes of interest have beforehand to be cloned in the pEUE01-MCS vector (CellFree Sciences, Japan), which is composed of a SP6 promoter, a E01 translational enhancer, and a MCS multiple cloning site (see Note 1). The proteins of interest are fused to a Strep-tag II which is an eight-residue minimal peptide sequence (WSHPQFEK) (see Note 2) [21, 26]. It is worth noting that the equipment used for all procedures described in this chapter has to be certified RNase free.

3.1 Wheat Germ Extract Preparation

There are commercially available wheat germ extracts (WGEs) but in order to significantly reduce costs we prepare home-made extracts according to the protocol developed by Yaeta Endo and coworkers at Ehime University in Japan [15] with minor modifications. 1. Grind 20–25 kg nontreated durum wheat seeds using the grain mill (see Note 3). 2. Sieve the seeds using an industrial sieve: the germs are on the upper sieve (850 μm) together with light particules. 3. Remove the light particules with cold air using a hair-drier. 4. Further isolate the germs in a solvent bath (carbone tetrachloride/cyclohexane, 600/240, v/v): the germs float on the surface whereas the waste goes down. This step is performed under the hood. 5. Harvest the germs quickly using a spatula and store them overnight under the hood till solvents are completely evaporated. 6. Although the preliminary steps described above allow an efficient isolation of wheat germs with 50–60% purity, it is necessary to sort them by eye selection in order to remove the ones with brownish particles, the ones with too much white matter coming from endosperm and the broken ones (see Note 4). 7. Place 20 g germs in a stocking for the washing steps (see Note 5) during which they are gently kneaded. 8. Perform first several washing steps in 500 mL water till the water is clear. 9. Wash the germs twice in 500 mL 0.5% Nonidet P-40 for 5 min in a sonication bath.

Wheat Germ Cell-Free Expression

97

10. Wash the germs several times in 500 mL water until there are no more bubbles present, which are caused by excess detergent anymore. 11. Wash the germs twice in 500 mL water for 2 min in a sonication bath. 12. Wash the germs five additional times in 500 mL water and dry them carefully with laboratory paper towel. 13. Crush the germs using the small bowl with four blades of a kitchen mixer in 40 mL buffer A. Run the blender five times 10 s. 14. Centrifuge the mixture for 30 min at 30,000
g, 4  C. Three phases are obtained: a precipitate at the bottom of the tube, a fatty fraction at the surface, and a supernatant (middle layer). 15. Carefully harvest the supernatant (see Note 6). 16. Further centrifuge the supernatant for 15 min at 30,000
g, 4  C. There are still a precipitate and a fatty fraction after this centrifugation step. 17. Harvest the middle layer as previously and centrifuge again for 15 min at 30,000
g, 4  C in order to obtain a purer supernatant. 18. Prepare home-made Sephadex G25 columns while packing 40 mL Sephadex G25 fine resin resuspended in RNase-free water in a 50-mL syringe with a cotton wool filter at the bottom. Two columns are necessary for each desalting step (see below), i.e., six columns for each 20 g WGE preparation. 19. Equilibrate Sephadex G25 home-made columns by gravity with 3 volumes buffer: Two columns are equilibrated with buffer A and four columns with SUB-AMIX buffer (see Note 7). 20. Hang the home-made columns in a 500 mL centrifuge bottle and capp them with aluminum foil. Centrifuge for 5 min at 750
g, 4  C. 21. Load the supernatant (refer to step 16) on a Sephadex G25 fine column equilibrated with buffer A (20 mL extract on a 40mL column). 22. Centrifuge the columns for 5 min at 750
g, 4  C. 23. Load the flow-through on a new Sephadex G25 fine column equilibrated with SUB-AMIX buffer (20 mL extract on a 40mL column). 24. Centrifuge the columns for 5 min at 750
g, 4  C. 25. Concentrate the WGE down to a final volume of 30 mL using Spin-X UF20 concentrators whose low binding polyethersulfone

98

Marie-Laure Fogeron et al.

membrane has a MWCO of 10,000 Da. The concentrators are washed twice with RNase free water before use. 26. Centrifuge at 8000
g, 4  C until the expected volume is reached. 27. Load the concentrated WGE on a new Sephadex G25 fine column equilibrated with SUB-AMIX buffer (15 mL extract on a 40-mL column). 28. Centrifuge the columns for 5 min at 750
g, 4  C. 29. Harvest the flow-through (see Note 8). 30. Aliquot the WGE and store in liquid nitrogen. Alternatively, the WGE can be stored at 80  C. 31. Control the efficiency of home-made WGEs through the expression of a standard protein such as GFP (green fluorescent protein) which can be easily followed using a fluorimeter. 3.2 DNA Preparation and Phenol/ Chloroform Extraction

Residual traces of RNAse from the buffer used to resuspend the bacterial pellet during DNA preparation can be present in DNA samples and could interfere with in vitro protein synthesis. The aim of the phenol/chloroform extraction is to remove residual RNase following DNA preparation. The phenol/chloroform extraction is indeed a liquid-liquid extraction method allowing the purification of nucleic acids and removal of proteins. The phenol/chloroform mixture is not miscible with water, which leads to the formation of two distinct phases: the upper aqueous phase containing DNA and the lower organic phase containing proteins. 1. Prepare DNA using a commercial kit according to the manufacturer’s instructions. 2. Add one volume of phenol/chloroform/isoamyl alcohol (25/24/1) to DNA solution and vortex the mixture shortly. 3. Centrifuge for 5 min at 20,800
g at room temperature (RT). 4. Transfer the supernatant in a new tube, add 1 volume of chloroform, and vortex the mixture shortly. 5. Centrifuge for 5 min at 20,800
g, RT. 6. Isolate the supernatant as previously described and add 2.5 volumes of 100% ethanol and 0.1 volume of 3 M sodium acetate pH 5.2 (see Note 9). 7. Incubate the mixture at 20  C for 10 min and centrifuge for 20 min at 20,800
g, 4  C. 8. Remove the supernatant and wash the DNA precipitate with 800 μL of 70% ethanol. 9. Centrifuge for 10 min at 20,800
g, 4  C.

Wheat Germ Cell-Free Expression

99

10. Remove the supernatant and let the pellet dry under the hood before resuspension in 500 μL of RNase-free water. 11. Measure DNA concentration and purity (A260/A280) using the NanoDrop. A A260/A280 ratio between 1.70 and 1.85 is expected (see Note 10). 3.3

Transcription

In the WGE-CF system presented here, transcription and translation are performed separately, i.e., mRNA is prepared first and then added to the WGE before the translation step (see Note 11). 1. Mix 100 μg/mL plasmid, 2.5 mM NTP mix, 1 U/μL SP6 RNA Polymerase and 1 U/μL RNase Inhibitor in transcription buffer (see Note 12). 2. Incubate for 6 h at 37  C (see Note 13). 3. Use mRNA directly for translation or store at 80  C.

3.4

Translation

Translation is performed using the so-called bilayer method [5] according to [15] in flat bottom plates, either 96-well plates for small-scale expression tests or 6-well plates for larger-scale productions. The bottom layer, called translation mix, contains the mRNA, the WGE, and creatine kinase for energy regeneration. The upper layer corresponds to the feeding buffer, which contains all substrates necessary for protein synthesis. There is no dialysis membrane between the two layers. Only the higher density of the WGE allows the formation of the bilayer. The feeding buffer is first added into the well and the reaction mix is then carefully filled at the bottom of the well. Translation is initiated in the translation mix. Byproducts are then gradually diluted in the feeding buffer whereas fresh substrates gradually diffuse to the translation mix (Fig. 1). 1. Prepare the translation mix: 20 μL per well in 96-well plates or 500 μL per well in 6-well plates. The translation mix contains 50% (v/v) mRNA, 50% (v/v) WGE (see Note 14), 40 ng/mL creatine kinase, and the appropriate detergent (see Note 15). 2. Prepare the feeding buffer: 200 μL per well in 96-well plates or 5.5 mL per well in 6-well plates. The feeding buffer (commercial SUB-AMIX, see Note 16) is supplemented with 6 mM amino acid mix (see Note 17) and the appropriate detergent. 3. Pipette the feeding buffer into the well of the plate. 4. Underlay carefully with the translation mix. 5. Cover the plate with an adhesive film (see Note 18). 6. Incubate the translation reaction at 22  C for 16 h without agitation (the bilayer would be disturbed by agitation) (see Note 19).

100

Marie-Laure Fogeron et al.

Bilayer preparation

Translation

feeding buffer (fresh substrates) translation mix (mRNA, WGE, creatine kinase)

Fig. 1 In vitro protein synthesis using the bilayer method. The color change illustrates both the gradual dilution of byproducts in the feeding buffer and the gradual diffusion of fresh substrates to the translation mix. Adapted from www. cfsciences.com/pdf/Translation_ver1-0_En_.pdf 3.5 Protein Sample Preparation for Analysis of Small-Scale Expression Tests

Small-scale expression tests are performed in order to determine the ability of detergents to allow the expression of membrane proteins in a solubilized form. On the one hand, the total cell-free samples (denoted CFS) are analyzed by SDS-PAGE followed by Coomassie staining and Western blotting to assess if the detergent has a deleterious effect on the protein expression level. On the other hand, the pellet and the supernatant obtained after centrifugation of the CFS are analyzed in the same way to assess if the detergent allows for solubilization. Note that samples are systematically treated with benzonase, an endonuclease degrading DNA and RNA independently of their shape (simple or double strand, circular or linear), with the aim of removing nucleic acids from the transcription step. Potential interaction of the protein of interest with nucleic acids could indeed lead to the sedimentation of the protein-nucleic acid complexes and the systematic detection of the protein in the pellet. As an example, the expression of full-length NS2 protein in the presence of nine different detergents showed that lauryl maltose neopentyl glycol (MNG-3) and dodecyl octaethylene glycol ether (C12E8) detergents can yield essentially solubilized membrane protein at detergent concentrations that do not inhibit the cellfree reaction (Fig. 2) [21]. 1. Incubate 150 μL of CFS with 250 units of benzonase at room temperature on a rolling wheel for 30 min. 2. Centrifuge at 20,000
g, 4  C for 30 min.

Wheat Germ Cell-Free Expression

101

Fig. 2 Production of full-length NS2 (strain JFH-1 of HCV genotype 2a) using the wheat germ cell-free expression system in the presence of nine different detergents [21]: DPC (FC-12), dodecyl-phosphocholine; DM, n-decyl-β-D-maltoside; DDM, n-dodecyl-β-D-maltoside; MNG-3, lauryl maltose neopentyl glycol; C12E8, dodecyl octaethylene glycol ether; LDAO, lauryl dimethyl amide oxide; CHAPS, 3-((3-cholamidopropyl)dimethylammonio)-1-propansulfonat; Cholate, 3, 7, 12-trihydroxy-5-cholan-24-oic acid; OG, n-octyl-D-glucopyranoside. Protein samples were analyzed by SDS-PAGE followed by Coomassie blue staining (upper panels) and Western blotting (lower panels). CFS, cell-free sample; Pellet, pellet obtained after centrifugation of CFS; SN-beads, sample enriched in protein of interest by incubation of supernatant obtained after centrifugation of CFS with Strep-Tactin magnetic beads to capture tagged protein; , negative control (no NS2 mRNA); þ, positive control (NS2 expressed in the absence of detergent). The black arrowheads indicate NS2. CFS—3.6% of total sample (220 μL reaction mix); Pellet and SN-beads—27% of total sample (220 μL reaction mix) loaded on the gel

3. Harvest the supernatant and store on ice until further processing (see step 7 below). 4. Centrifuge the pellet at 20,000
g, 4  C for 5 min. 5. Remove residual supernatant. 6. Resuspend the pellet in 25 μL 1
SDS-PAGE loading buffer. 7. Incubate the supernatant (SN) with 5 μL Strep-Tactin magnetic beads at 4  C on the rolling wheel for at least 30 min (see Note 20). 8. Wash the beads three times with 10 volumes of buffer W. 9. Remove the washing buffer on the magnet and resuspend the magnetic beads directly in 25 μL of 1
SDS-PAGE loading

102

Marie-Laure Fogeron et al.

buffer. Theses samples are further designated as SN-beads (see Note 21). 10. In addition, mix 20 μL CFS with 5 μL 5
loading buffer for SDS-PAGE and Western blotting analysis. 3.6 Purification by Affinity Chromatography

The Strep-tag II fused to the proteins of interest exhibits intrinsic affinity toward streptavidin and allows efficient purification by affinity chromatography on a matrix carrying an engineered streptavidin (Strep-Tactin). The proteins of interest are eluted by competition with D-desthiobiotin under mild buffer conditions in a biochemically active state [26]. As an example, the purification of full-length NS2 produced in the presence of 0.1% MNG-3 leads to a purity level higher than 90% in only one affinity chromatography step (Fig. 3) and a purified detergent-solubilized protein yield around 1 mg/mL WGE [27]. 1. Harvest CFS from the wells while homogenizing carefully with the pipette. 2. Add 2000 units benzonase per mL CFS. 3. For membrane proteins expressed in the presence of MNG-3 detergent, add 0.25% DDM (see Note 22) to the CFS together with benzonase. 4. Incubate at RT on a rolling wheel for 30 min. 5. Centrifuge at 20,000
g, 4  C for 30 min.

Fig. 3 Purification of full-length NS2 by affinity chromatography [27]. Full-length N-terminally Strep-tag II tagged NS2 (strain JFH-1 of HCV genotype 2a) was expressed using 4.5 mL WGE in the presence of 0.1% MNG-3 and complemented with 0.25% DDM before purification on a 5-mL Strep-Tactin column in the presence of 0.1% DDM. Protein samples were analyzed by SDSPAGE. CFS, total cell-free sample (8 μL from 108 mL loaded on the gel); P, pellet obtained after centrifugation of CFS (equivalent to 80 μL from 108 mL loaded on the gel); SN, supernatant obtained after centrifugation of CFS and loaded on the affinity column (8 μL from 108 mL loaded on the gel); E1–E6, affinity elution fractions (8 μL from 2.5 mL loaded on the gel). The black arrowhead indicates the bands corresponding to NS2

Wheat Germ Cell-Free Expression

103

6. Load the obtained supernatant on a Strep-Tactin gravity column (see Note 23) equilibrated beforehand with 2 CV (column volumes) of buffer W supplemented with 1 mM DTT and 0.1% DDM. 7. Wash the column five times with 1 CV of the buffer described above. 8. Elute the protein of interest six times with 0.5 CV of buffer E supplemented with 1 mM DTT and 0.1% DDM. 9. Regenerate Strep-Tactin columns with HABA (2-[40 -hydroxybenzeneazo]benzoic acid) according to the manufacturer’s instructions (see Note 24). 3.7 SDS-PAGE and Western Blotting Analysis

1. Resuspend samples in SDS-PAGE loading buffer. 2. Incubate at room temperature for 15 min before loading (do not heat). 3. Load 10 μL of each sample on the SDS–PAGE gels. 4. Carry out Western blotting analysis by protein transfer onto a nitrocellulose membrane. 5. Block the nitrocellulose membrane with 5% nonfat milk powder in PBS-T buffer. 6. Incubate the nitrocellulose membrane for 1 h at RT with the Strep MAB Classic primary antibody diluted 1:5000 in PBS-T. 7. Wash the nitrocellulose membrane three times for 10 min in PBS-T. 8. Incubate the nitrocellulose membrane for 1 h at RT with an anti-mouse IgG HRP-conjugated secondary antibody diluted 1:4000 in PBS-T. 9. Epitope-containing chemiluminescence.

4

bands

are

visualized

using

Notes 1. The translation enhancer E01 was optimized starting from the Ω sequence from tobacco mosaic virus, one of the most frequently used mRNA 50 leader sequences enhancing in vivo and in vitro translation [28]. Both the 50 -7mGpppG (cap) and poly (A)-tail were eliminated to increase the translation initiation and the stability of the mRNA template [11, 29]. Alternatively, a two-step PCR method called the “split-primer” PCR can be applied for the amplification of many different cDNA clones [15, 29]. 2. Some proteins from the WGEs bind unspecifically to glutathione or metal-chelating resins, hampering efficient purification.

104

Marie-Laure Fogeron et al.

There are even commercial WGEs pretreated on such resins, these pretreatments leading to a higher purity of proteins with a GST- or His-affinity tag [15, 30]. Proteins from the WGEs do not interfere with Strep-Tactin resins and no WGE pretreatment is required when proteins are expressed with a Strep-tag II [26]. 3. So far we have tested three different durum wheat strains (Dakter, Miradoux, and Clovis) and have not observed any significant difference between them. We noticed however that fresh wheat seeds allow the production of more efficient WGEs. Storage of wheat seeds over too long periods of time is therefore not recommended. 4. Eye selected germs can be used directly for the preparation of the WGE or stored at 4  C up to several months. 5. The washing steps described here are crucial because they allow the removal of residual endosperm that contains translation inhibitors such RIB (Ribosome Inactivating Proteins), thionin, or DNases and RNases [3]. 6. From there, the absorbance at 260 nm and 280 nm (A260 and A280, respectively) is measured at each step of the preparation to follow the WGE concentration. According to [15], final A260 should be at least equal to 240. We observed that WGE efficiency could also be correlated to the A260/A280 ratio. Indeed, in our hands, while WGEs with a A260/A280 ratio lower than 1.5 were poorly efficient, commercial and good home-made WGEs had a A260/A280 ratio between 1.5 and 2.25. 7. Note that both buffers do not contain any amino acids. Indeed, amino acids are added while preparing the translation reaction, which allows greater flexibility regarding isotopic labeling strategies for NMR studies. In the case of studies that do not require isotopic labeling, amino acids could be added directly during the WGE preparation. 8. After use, the Sephadex G25 columns are extensively washed with water, the resin is stored at 4  C in water containing 0.05% sodium azide and can be reused several times. 9. The addition of ethanol and salt leads to DNA precipitation. 10. If the A260/A280 ratio is not between 1.70 and 1.85, phenol/ chloroform extraction should be repeated. In addition, in order to standardize protocols, DNA preparations are diluted in RNase-free water at a concentration of 1 μg/μL. 11. Although the use of coupled transcription/translation has been described for the WGE-CF system [31], these reactions are usually done separately as uncoupling transcription and translation allows for more flexibility to work under optimal

Wheat Germ Cell-Free Expression

105

reaction conditions (e.g., temperature, Mg2+ concentration), to use additives in the translation reaction without interfering with transcription, and to better identify and solve problems when they occur [11, 15, 29]. 12. Commercial transcription buffer is used here. Alternatively, home-made solutions can be prepared using RNase-free water [15]. 13. A white precipitate (magnesium pyrophosphate) appears during transcription. In the absence of this precipitate, transcription has not been efficient and has to be repeated. 14. Home-made WGE prepared according to the protocol from [15] is used here. Alternatively, commercial WGE can be purchased from CellFree Sciences (Japan). 15. The volumes of creatine kinase and detergent added to the translation mix are negligibly small compared to the mRNA and WGE volumes. 16. Commercial feeding buffer SUB-AMIX is used here. There is also commercial buffer including amino acids available. However, the absence of amino acids allows for more flexibility, for example for isotopic labeling in view of structural studies by NMR. Alternatively, home-made solutions can be prepared using RNase-free water [15]. This alternative approach allows the experimenter to modulate the potassium and magnesium ion concentrations that have been shown to be critical and can be optimized for each protein. 17. Note that in the commercial amino acid mixture standardly used to produce unlabeled protein samples, the proportion is different for each amino acid (6% Ala, 6% Arg, 5% Asn, 8% Asp, 3% Cys, 9% Glu, 5% Gln, 5% Gly, 1% His, 3% Ile, 9% Leu, 12% Lys, 1% Met, 4% Phe, 5% Pro, 4% Ser, 4% Thr, 3% Trp, 3% Tyr and 4% Val). Therefore, depending on the amino acid composition of the protein of interest, preparing an optimized amino acid mixture using individual amino acids with the same composition than the protein of interest could help to achieve better expression yields. 18. The adhesive film avoids both evaporation and potential contamination. 19. Depending on the protein of interest, the translation reaction can be performed in a wide temperature range between 4 and 30  C, with a temperature between 15 and 26  C often giving the best results [30]. 20. The Strep-Tactin magnetic beads are beforehand washed three times with 10 volumes of buffer W. According to the manufacturer, these beads have a binding capacity up to 1 nmol protein/mg beads.

106

Marie-Laure Fogeron et al.

21. This step does not aim to purify the protein of interest that is fused to the Strep-tag II but to enrich the sample by capturing it on the Strep-Tactin coated magnetic beads. This step thus allows the analysis of comparable amounts for the pellet and the supernatant, as both pellet and magnetic beads are resuspended in the same volume of loading buffer. In addition, when detergent is used, this step allows us to assess whether the detergent interferes with the protein affinity capture. 22. The addition of 0.25% DDM before the affinity-purification step not only allows buffer exchange from MNG-3 to DDM, but also results in improved NS2 purity and yield [27]. Indeed, the purity of NS2 obtained in the presence of MNG-3 during this step is not satisfactory [21], as also observed in the presence of lower amounts of DDM. With respect to yield, while purity is good in the presence of even higher amounts of DDM (0.5–1%), the protein yield is substantially reduced under this condition since the high detergent concentration decreases NS2-Strep-tag II binding to the Strep-Tactin column [26]. 23. The Strep-Tactin column format used in a standard way is as follows: 0.2 mL for 6 mL CFS (1 well of a 6-well plate), 1 mL for 36 mL CFS (one 6-well plate), and 5 mL for 108 mL CFS (three 6-well plates). 24. Strep-Tactin columns can be used up to five times for the same protein. They are stored at 4  C in 100 mM Tris–HCl pH 8.0, 150 mM NaCl, 1 mM EDTA and 0.05% NaN3.

Acknowledgment We would like to thank Yaeta Endo and Matthias Harbers from CellFree Sciences Japan for their continuous help and interest in our work. This work was supported by the CNRS and by grants from the French ANRS (France Recherche, Nord & Sud, Sida-HIV et He´patites), an autonomous agency at INSERM, France, and the ANR (ANR-14-CE09-0024B). References 1. Roberts BE, Paterson BM (1973) Efficient translation of tobacco mosaic virus RNA and rabbit globin 9S RNA in a cell-free system from commercial wheat germ. PNAS 70:2330–2334 2. Scheele G, Blackburn P (1979) Role of mammalian RNase inhibitor in cell-free protein synthesis. PNAS 76:4898–4902 3. Madin K, Sawasaki T, Ogasawara T, Endo Y (2000) A highly efficient and robust cell-free

protein synthesis system prepared from wheat embryos: plants apparently contain a suicide system directed at ribosomes. PNAS 97:559–564 4. Ogasawara T, Sawasaki T, Morishita R, Ozawa A, Madin K, Endo Y (1999) A new class of enzyme acting on damaged ribosomes: ribosomal RNA apurinic site specific lyase found in wheat germ. EMBO J 18:6522–6531. doi:10.1093/emboj/18.22.6522

Wheat Germ Cell-Free Expression 5. Sawasaki T, Hasegawa Y, Tsuchimochi M, Kamura N, Ogasawara T, Kuroita T, Endo Y (2002) A bilayer cell-free protein synthesis system for high-throughput screening of gene products. FEBS Lett 514:102–105 6. Sawasaki T, Hasegawa Y, Morishita R, Seki M, Shinozaki K, Endo Y (2004) Genome-scale, biochemical annotation method based on the wheat germ cell-free protein synthesis system. Phytochemistry 65:1549–1555. doi:10.1016/ j.phytochem.2004.04.023 7. Kohno T, Endo Y (2007) Production of protein for nuclear magnetic resonance study using the wheat germ cell-free system. Methods Mol Biol 375:257–272. doi:10.1007/ 978-1-59745-388-2_13 8. Makino S-I, Goren MA, Fox BG, Markley JL (2009) Cell-free protein synthesis technology in NMR high-throughput structure determination. In: Weissig V (ed) Liposomes. Humana Press, Totowa, NJ, pp 127–147 9. Netzer WJ, Hartl FU (1997) Recombination of protein domains facilitated by cotranslational folding in eukaryotes. Nature 388:343–349. doi:10.1038/41024 10. Hartl FU, Hayer-Hartl M (2002) Molecular chaperones in the cytosol: from nascent chain to folded protein. Science 295:1852–1858. doi:10.1126/science.1068408 11. Endo Y, Sawasaki T (2006) Cell-free expression systems for eukaryotic protein production. Curr Opin Biotechnol 17:373–380. doi:10. 1016/j.copbio.2006.06.009 12. Schwarz D, Do¨tsch V, Bernhard F (2008) Production of membrane proteins using cell-free expression systems. Proteomics 8:3933–3946. doi:10.1002/pmic.200800171 13. Nakamura S (1993) Possible role of phosphorylation in the function of chicken MyoD1. J Biol Chem 268:11670–11677 14. Langland JO, Langland LA, Browning KS, Roth DA (1996) Phosphorylation of plant eukaryotic initiation factor-2 by the plant-encoded double-stranded RNAdependent protein kinase, pPKR, and inhibition of protein synthesis in vitro. J Biol Chem 271:4539–4544 15. Takai K, Sawasaki T, Endo Y (2010) Practical cell-free protein synthesis system using purified wheat embryos. Nat Protoc 5:227–238. doi:10.1038/nprot.2009.207 16. Junge F, Haberstock S, Roos C, Stefer S, Proverbio D, Do¨tsch V, Bernhard F (2011) Advances in cell-free protein synthesis for the functional and structural analysis of membrane proteins. New Biotechnol 28:262–271. doi:10.1016/j.nbt.2010.07.002

107

17. Klammt C, Schwarz D, Fendler K, Haase W, Do¨tsch V, Bernhard F (2005) Evaluation of detergents for the soluble expression of α-helical and β-barrel-type integral membrane proteins by a preparative scale individual cell-free expression system. FEBS J 272:6024–6038. doi:10.1111/j.1742-4658.2005.05002.x 18. Klammt C, Lo¨hr F, Schafer B, Haase W, Do¨tsch V, Ruterjans H, Glaubitz C, Bernhard F (2004) High level cell-free expression and specific labeling of integral membrane proteins. Eur J Biochem 271:568–580. doi:10.1111/j. 1432-1033.2003.03959.x 19. Schwarz D, Junge F, Durst F, Fro¨lich N, Schneider B, Reckel S, Sobhanifar S, Do¨tsch V, Bernhard F (2007) Preparative scale expression of membrane proteins in Escherichia colibased continuous exchange cell-free systems. Nat Protoc 2:2945–2957. doi:10.1038/ nprot.2007.426 20. Genji T, Nozawa A, Tozawa Y (2010) Efficient production and purification of functional bacteriorhodopsin with a wheat-germ cell-free system and a combination of Fos-choline and CHAPS detergents. Biochem Biophys Res Commun 400:638–642. doi:10.1016/j.bbrc. 2010.08.119 21. Fogeron ML, Badillo A, Jirasko V, Gouttenoire J (2015) Wheat germ cell-free expression: two detergents with a low critical micelle concentration allow for production of soluble HCV membrane proteins. Protein Expr Purif 105:39–46. doi:10.1016/j.pep.2014.10.003 22. Chae PS, Rasmussen SGF, Rana RR, Gotfryd K, Chandra R, Goren MA, Kruse AC, Nurva S, Loland CJ, Pierre Y, Drew D, Popot J-L, Picot D, Fox BG, Guan L, Gether U, Byrne B, Kobilka B, Gellman SH (2010) Maltose–neopentyl glycol (MNG) amphiphiles for solubilization, stabilization and crystallization of membrane proteins. Nat Methods 7:1003–1008. doi:10.1038/nmeth.1526 23. Nozawa A, Ogasawara T, Matsunaga S, Iwasaki T, Sawasaki T, Endo Y (2011) Production and partial purification of membrane proteins using a liposome-supplemented wheat cell-free translation system. BMC Biotechnol 11:35. doi:10.1186/1472-6750-11-35 24. Carraher C, Nazmi AR, Newcomb RD, Kralicek A (2013) Recombinant expression, detergent solubilisation and purification of insect odorant receptor subunits. Protein Expr Purif 90:160–169. doi:10.1016/j.pep.2013.06.002 25. Moradpour D, Penin F (2013) Hepatitis C virus proteins: from structure to function. In: Bartenschlager R (ed) Hepatitis C virus: from molecular virology to antiviral therapy. Springer, Berlin, Heidelberg, pp 113–142

108

Marie-Laure Fogeron et al.

26. Schmidt TG, Skerra A (2007) The Strep-tag system for one-step purification and highaffinity detection or capturing of proteins. Nat Protoc 2:1528–1535. doi:10.1038/ nprot.2007.209 27. Fogeron M-L, Paul D, Jirasko V, Montserret R, Lacabanne D, Molle J, Badillo A, Boukadida C, Georgeault S, Roingeard P, Martin A, Bartenschlager R, Penin F, Bo¨ckmann A (2015) Functional expression, purification, characterization, and membrane reconstitution of non-structural protein 2 from hepatitis C virus. Protein Expr Purif 116:1–6. doi:10.1016/j.pep.2015.08.027 28. Kamura N, Sawasaki T, Kasahara Y, Takai K, Endo Y (2005) Selection of 50 -untranslated sequences that enhance initiation of translation in a cell-free protein synthesis system from wheat embryos.

Bioorg Med Chem Lett 15:5402–5406. doi:10.1016/j.bmcl.2005.09.013 29. Sawasaki T, Ogasawara T, Morishita R, Endo Y (2002) A cell-free protein synthesis system for high-throughput proteomics. PNAS 99:14652–14657. doi:10.1073/pnas. 232580399 30. Harbers M (2014) Wheat germ systems for cell-free protein expression. FEBS Lett 588:2762–2773. doi:10.1016/j.febslet.2014. 05.061 31. Stueber D, Ibrahimi I, Cutler D, Dobberstein B, Bujard H (1984) A novel in vitro transcription-translation system: accurate and efficient synthesis of single proteins from cloned DNA sequences. EMBO J 3:3143–3148

Chapter 6 Methyl-Specific Isotope Labeling Strategies for NMR Studies of Membrane Proteins Vilius Kurauskas, Paul Schanda, and Remy Sounier Abstract Methyl groups are very useful probes of structure, dynamics, and interactions in protein NMR spectroscopy. In particular, methyl-directed experiments provide high sensitivity even in very large proteins, such as membrane proteins in a membrane-mimicking environment. In this chapter, we discuss the approach for labeling methyl groups in E. coli-based protein expression, as exemplified with the mitochondrial carrier GGC. Key words NMR spectroscopy, Methyl labeling, Refolding, Deuteration, Detergent

1

Introduction Nuclear Magnetic Resonance (NMR) spectroscopy is a versatile tool for studying the structure, dynamics, and interactions of biomolecules at atomic resolution. Over the last decades, biological NMR spectroscopy constantly progressed toward more complex and larger molecules. This evolution was triggered by developments in NMR technology in terms of spectroscopic approaches and sample preparation. In particular, the specific introduction of [1H, 13C]-labeled methyl groups in a perdeuterated environment is a key element that has allowed NMR to study large molecule systems and complexes [1]. Large molecules are inherently very difficult to study by solution-state NMR, because the slow overall reorientation leads to fast decay of NMR signals, and thus low sensitivity and resolution. Due to the rapid rotation of methyl groups (on a picosecond time scale), the methyl group is partly decoupled from the slow overall tumbling. Therefore, methyl groups have very favorable spectroscopic properties even in large molecules, in particular when the surrounding is deuterated. Sensitivity of methyl groups is also enhanced by the three-fold proton multiplicity. Methyl labeling techniques combined with appropriate

Jean-Jacques Lacapere (ed.), Membrane Protein Structure and Function Characterization: Methods and Protocols, Methods in Molecular Biology, vol. 1635, DOI 10.1007/978-1-4939-7151-0_6, © Springer Science+Business Media LLC 2017

109

110

Vilius Kurauskas et al.

spectroscopic approaches (in particular, methyl-TROSY experiments [2]) have been shown to provide insight into molecules of up to about 1 MDa in size [3]. Methyl labeling protocols are based on the incorporation of specific precursors during the expression in bacteria growth. Labeling protocols for the methyl groups of all methyl-bearing amino acids have been developed over the last two decades, ensuring that the desired methyl group is labeled at high yield and specificity without scrambling of the introduced isotopes to other sites in the protein (see Table 1). More and more laboratories are using these labeling protocols because most of the precursors can be prepared directly in the laboratory. However, to simplify the use of these protocols, precursors are now commercially available with different isotopic labeling scheme used for different types of NMR studies [4, 5]. Other precursors targeting aromatic residues (Trp, Phe, Tyr) have been introduced to label the side-chain of these residues. These labeling approaches of sites other than methyl groups are not within the scope of this review; the reader is directed toward references listed in Table 2. The six naturally methyl-bearing amino acids (Ala, Leu, Val, Ile, Thr, Met) represent about ~35% of the amino acids in soluble proteins, and up to 45% in α-helical membrane protein. Methyl groups, therefore, provide a good coverage across the structure of membrane proteins, and are in principle a good source of information of structure (through intra- and intermolecular spin-spin couplings, manifest as nuclear Overhauser effect, NOEs), dynamics, and interactions. The potential of methyl-directed NMR applied to membrane proteins is, however, somewhat compromised by the fact that methyls tend to point toward the hydrophobic membranemimicking environment (detergent, lipids). Consequently, their chemical shifts tend to be similar, resulting in narrower dispersion of methyl signals and difficulty of resonance assignment. Nonetheless, multiple NMR studies of membrane proteins show the power of the strategy, such as studies of structure and dynamics of the voltage-dependent anion channel (VDAC) [6], the prokaryotic pH-dependent potassium channel (KcsA) [7], the phototaxis receptor sensory rhodopsin II (PSRII) [8], the hexameric p7 cation channel from hepatitis C virus [9], and the mitochondrial translocator Protein (TSPO) [10]. Methyl labeling is also becoming an increasingly useful tool for proton-detected MAS solid-state NMR spectroscopy [11–13], and although it has to our knowledge so far not been applied to membrane proteins, the approach will certainly rapidly gain popularity also in this field. This chapter highlights a methyl-specific isotope-labeling scheme to study membrane proteins. We outline the approach to produce, purify, and assign the methyl groups of GTP/GDP carrier (GGC) from yeast mitochondria in detergent micelles [14]. GGC is a member of the mitochondrial carrier family [15]. These carriers

Ala

Thr Met Met Ile Ile Ile Ile Leu Leu Val Val Leu/Val Leu/Val

L-Alanine

L-Threonine

L-Methionine

4-Methyl-thio-2-ketobutyrate

2-Ketobutyrate

2-(S)-hydroxy-2-ethyl-3-ketobutyrate

2-Hydroxy-2-ethyl-3-ketobutyrate*

2-Hydroxy-2-ethyl-3-ketobutyrate*

2-Ketoisocaproate

Stereospecifically labeled L-Leucine

2-Acetolactate

Stereospecifically labeled Valine

2-Ketoisovalerate

2-Acetolactate

50 250 100 60 60 140 100 150

γ2 ε ε δ1 δ1 γ2 γ2 δ1/δ2 γ

100 100–200 300

δ1/δ2 γ1/γ2 δ/γpro-S

300

γpro-S

pro-S

pro-S

20

600

β

δ

Quantity (mg.L
1)

Methyl groups

Scrambling suppressed by co-addition of U-[D] of the different compounds

a

Amino acids

CH3 group labeling

13

Name of the precursors

Table 1 Precursors for

L-Leucine

L-Leucine

2-ketoisovalerate

2-ketoisovalerate

ketobutyrate glycine

2-ketoisovalerate isoleucine succinate glycerol

Scrambling suppression a

20

20

200

200

50 100

200 60 2500 2500

Quantity (mg.L
1)

[20]

[31] [32]

[29]

[30]

[29]

[28]

[27]

[26]

[25]

[19]

[24]

[23]

[22]

[21]

Ref.

Methyl Labelling of Membrane Proteins 111

112

Vilius Kurauskas et al.

Table 2 Precursors for Aromatic group labeling Name of the precursors

Amino acids

Aromatic groups

Quantity (mg.L
1)

Ref.

Phenylpyruvate

Phe/Tyr

δ1/δ2-ring ε1/ε2-ring

100–200

[33] [34]

Indole

Trp

indole

50

[35]

catalyze the transport of small solutes across the inner mitochondrial membrane. GGC transports external GTP into the mitochondrial matrix while exporting internal GDP out of the matrix [15]. When refolded in detergent micelle, GGC forms a highly dynamic ensemble, discussed previously [14].

2

Materials

2.1 Precursor Preparation, Bacterial Expression, and Isotope Labeling

1. Lysogeny Broth (LB) agar plates (Miller): Dissolve 40 g of LB agar (Miller) in 1 L of distilled water and in a 2 L flask. Cover with foil and autoclave. Monitor the temperature as it cools. When the temperature reaches ~50  C, add 1 mL of the antibiotic stock solution 1000 (ex 100 μg.L
1 for ampicillin). Pour ~15 mL into plate, let the LB agar solidify at room temperature, place into a plastic bag, seal with tape, and store at 4  C. 2. LB medium (Miller): Dissolve 25 g of powdered LB medium in 1 L of distilled water. Adjust the pH to 7.4 if necessary using NaOH. Autoclave and store at room temperature. Add antibiotics prior to use. 3. M9 salts: For 1 L of growth medium, 6 g of anyhydrous Na2HPO4, 3 g of anhydrous KH2PO4, 0.5 g of NaCl, 1 g of NH4Cl. The salts are first dissolved in distilled and 0.22 μm filtered H2O and autoclaved (M9H). For deuteration, the salts are dissolved in 99.8% D2O and sterile filtered through 0.22 μm filter flasks, rather than autoclaved (M9D). 4. Oligo-elements stock solution: Prepare a 1 M solution of MgSO4, a 0.1 M solution of CaCl2, a 0.1 M solution of MnSO4, a 50 mM solution of ZnCl2, a 50 mM solution of FeCl3. Sterilize on 0.22 μm filter flasks and store at 4  C. Stocks solutions should be prepared in H2O when used in M9H medium or in D2O when used in M9D medium. 5. 100 MEM Vitamin solution (Sigma-Aldrich): store 10 mL aliquots at
20  C.

Methyl Labelling of Membrane Proteins

113

6. Isopropyl-β-D-thiogalactopyranoside (IPTG): Prepare a 0.5 M solution using distilled water. Syringe filter and store 1 mL aliquots at
20  C. 7. 40% D-glucose: Dissolve 40 g of D-glucose in 100 mL distilled water. Sterilize by filtration and store at room temperature. 8. M9H medium: For 1 L of M9 salts, add 5 mL of 40% D-glucose, 1 mL of 1 M MgSO4, 1 mL of 0.1 M CaCl2, 1 mL of 0.1 M MnSO4, 1 mL of 50 mM ZnCl2, 1 mL of 50 mM FeCl3 and 10 mL of 100 MEM Vitamin solution and swirl the flask after each addition. Add antibiotics prior to use. 9. M9D medium: For 1 L of M9D salts, add 2 g of D-glucose, 1 mL of 1 M MgSO4, 1 mL of 0.1 M CaCl2, 1 mL of 0.1 M MnSO4, 1 mL of 50 mM ZnCl2, 1 mL of 50 mM FeCl3 and 10 mL of 100 MEM Vitamin solution and swirl the flask after each addition. Add antibiotics prior to use. 10. For the production of 15N-labeled protein: NH4Cl is substituted by ammonium chloride (15N  99%). 11. For the production of 13C-labeled protein: D-glucose is substituted by D-(13C)-glucose (13C  99%). For perdeuteration, 2 13 D-( H, C)-glucose (2H  98%; 13C  99%) is substituted (see Note 1). 12. For the production of methyl-specifically labeled protein: 2 12 D-glucose is substituted by D-( H, C)-glucose (2H  98%) and the addition of 13CH3-methyl specifically labeled precursors (see Note 2). 13.

13

CH3-methyl specifically labeled precursors: 2H-13CH3-Alanine (13C  99%; 2H  98%), 2H-13CH3–2-ketobutyric acid (13C  99%; 2H  98%), 2H-13CH3–2-hydroxy-2-methyl-3oxo-4-butanoic acid (13C  99%; 2H  95%), 2H–L-Isoleucine (2H  98%). Specifically labeled precursors are commercially available on a deuterated form ready for direct introduction into the culture medium (NMR-Bio) or can be prepared in the laboratory (see Note 3).

14. Innova® 43 Incubator Shaker (Eppendorf). 2.2 In Vitro Refolding and Protein Purification

1. Lysis buffer: 20 mM HEPES, pH 7.5, 150 mM NaCl, 5 mM β-mercaptoethanol (BME), 100 μg/mL phenylmethylsulfonyl fluoride (PMSF), 100 μg/mL benzamidine and 5 μg/mL leupeptine. 2. Denaturing buffer: 3 M guanidine HCl, 1% Triton X-100, 20 mM HEPES, pH 7.5, 150 mM NaCl, 5 mM BME. 3. Refolding buffer: 0.1% Triton X-100, 20 mM HEPES, pH 7.5, 150 mM NaCl, 5 mM BME.

114

Vilius Kurauskas et al.

4. Purification buffer: 0.1% n-dodecylphosphocholine (DPC, Anatrace), 20 mM HEPES, pH 7.5, 150 mM NaCl, 5 mM BME. 5. Gel-filtration buffer: 0.1% DPC, 50 mM MES, pH 6.0, 150 mM NaCl, 1 mM BME, and it is filtered on a 0.22 μm polyethersulfone membrane (Millipore). 6. NMR buffer: 0.1% DPC, 50 mM MES, pH 6.0, 1 mM BME, and 5–10% D2O, used for spectrometer lock. 7. NuPage Novex 12% Bis–Tris Gel (Invitrogen). 8. 4 NuPAGE® LDS Sample Buffer (Invitrogen). 9. NuPAGE® MES SDS running buffer (Invitrogen). 10. Ni-NTA agarose beads (Qiagen). 11. Refrigerated Akta FPLC purification system and accessories equipped with HiLoad 16/60 Superdex 200 PG (GE Healthcare). 12. Dialysis cassette: 10 kDa MWCO (Thermo Fischer). 13. Centrifugal concentrator: 10 kDa MWCO (Millipore). 14. Shigemi NMR tubes susceptibility matched to D2O (Shigemi Inc). 2.3 Solution-State NMR Spectroscopy Experiments for Resonance Assignment and Chemical Shift Perturbation Studies

3

1. All NMR experiments are conducted on a 600 MHz spectrometer equipped with cryogenic TXI probe (Bruker). 2. All spectra were processed using NMRPipe [16] and analyzed using CCPNMR [17].

Methods

3.1 Precursors Preparation, Bacterial Expression, and Isotope Labeling 3.1.1 Uniform (2H, 13C, 15N)

1. The gene of interest was inserted between the NdeI (50 ) and XhoI (30 ) sites in the pET-21a (þ) plasmid which contains a C-terminus (His)6-tag. 2. Transform the plasmid [18] into E. coli BL21(DE3) cells, plate on LB agar containing ampicillin, and incubate overnight at 37  C. 3. Pick a freshly transformed colony of BL21(DE3) cells and inoculate 5 mL LB medium in a shaking incubator (220 rpm) at 37  C during 4 h (final OD600 is about 1.0). 4. Inoculate with previous LB culture, 50 mL of M9H medium to achieve a starting OD600 of 0.1. Incubate the culture in a shaking incubator (220 rpm) at 37  C during 4 h (final OD600 is about 0.8).

Methyl Labelling of Membrane Proteins

115

5. Spin down the cells at 2000  g for 10 min at room temperature and resuspend a small amount of them in 100 mL of M9D medium to achieve a starting OD600 of 0.1. Incubate the culture in a shaking incubator (220 rpm) at 37  C during overnight (final OD600 is about 1.5) (see Note 4). 6. Inoculate with overnight culture, 900 mL of M9D medium in a 3 L baffled flask to achieve a starting OD600 of 0.1 (see Note 5). Incubate the culture in a shaking incubator (220 rpm) at 37  C until the cultures reach an OD600 of 0.6 (see Note 6). 7. Induce expression of the protein by adding 1 mL of IPTG to a final concentration of 0.5 mM at 37  C for 4 h. The final OD600 is typically between 1.2 and 1.4 (see Notes 6 and 7). 8. Harvest the cells by centrifugation at 5000  g for 30 min at 4  C. Resuspend the cells with lysis buffer in 50 mL falcon tube and freeze the cells at
80  C. 3.1.2 Methyl Specific Labeling (2H, 13C, 15N)

1. Use uniform (2H, 13C, 15N) labeling protocol described above until step 6. 2. Inoculate with overnight culture, 700 mL of M9D medium in a 3 L baffled flask to achieve a starting OD600 above 0.1 (see Note 5). Incubate the culture in a shaking incubator (220 rpm) at 37  C until the cultures reach an OD600 of 0.6. 3. Add the precursors diluted in 200 mL of M9D medium (see Note 8). (a) [U-2H], I-[13CH3]δ1-GGC: 60 mg/L of 2H-13CH3–2ketobutyric acid (13C  99%; 2H  98%). (b) [U-2H], A-[13CH3]β, L-[13CH3]proS, V-[13CH3]proSlabeled protein: 700 mg/L of 2H-13CH3-Alanine (13C  99%; 2H  98%), 300 mg/L of 2H-13CH3–2hydroxy-2-methyl-3-oxo-4-butanoic acid (13C  99%; 2 H  95%) and 60 mg/L of 2H–L-Isoleucine (2H  98%). 4. Let the culture grow for 1 h (OD600 reaches the value of 0.6) (see Note 6). 5. Induce expression of the protein by adding 1 mL of IPTG to a final concentration of 0.5 mM at 37  C for 4 h. The final OD600 is typically between 1.2 and 1.4 (see Notes 6, 7, and 9). 6. Harvest the cells by centrifugation at 5000  g for 30 min at 4  C. Resuspend the cells with lysis buffer in 50 mL falcon tube and freeze the cells at
80  C.

3.2 In Vitro Refolding and Protein Purification

1. Resuspend the cell pellets in 100 mL lysis buffer per liter growth using a Dounce homogenizer. Disrupt the cells by sonication on ice, until obtaining a clear and homogeneous suspension (3 4 min, at 50 W using 50% duty cycle). Spin

116

Vilius Kurauskas et al.

down E. coli inclusion bodies at 40,000  g for 30 min at 4  C to purify away water-soluble contaminants. 2. Solubilize the GGC-containing inclusion bodies in 50 mL of denaturing buffer and incubate under vigorous stirring at 4  C overnight. Pellet undissolved matter by centrifugation at 40,000  g for 30 min at 4  C. 3. Equilibrate 10 mL per liter of growth of Ni-NTA agarose beads on column with 3 column volumes (CVs) of denaturing buffer. 4. Add 10 mM imidazole to the supernatant and load onto equilibrated Ni-NTA column at 1–2 mL/min. 5. Wash with 3 CVs of denaturing buffer containing 10 mM imidazole to remove contaminant proteins. 6. Wash with 3 CVs of refolding buffer (see Note 10). 7. Wash with 3 CVs of purification buffer (see Note 11). 8. Elute bound GGC with 3 CVs of purification buffer containing 250 mM imidazole. 9. Pre-rinse centrifugal concentrators (10 kDa MWCO) with purification buffer at 3200  g for 5 min. Concentrate the protein to ~5 mg/mL using series of 5 min spins at 3200  g and 4  C. 10. Equilibrate the HiLoad 16/600 Superdex 200 PG column with 1.5 CVs of filtered gel-filtration buffer. 11. Load onto the gel filtration column at 1.0 mL/min, collecting 0.5 mL fractions for 1.5 CVs. Assess the purity of relevant fractions by SDS-PAGE (5 μL of each fraction) (Fig. 1). 12. Pool the pure protein fractions and dialyze for 2 h against NMR buffer using dialysis cassette (10 kDa MWCO), twice. Measure protein concentration by UV absorbance. 13. Pre-rinse centrifugal concentrators (10 kDa MWCO) with NMR buffer at 3200  g for 5 min. Concentrate the protein to 0.5–1.0 mM using series of 5 min spins at 3200  g and 4  C. Load the final sample to a Shigemi NMR tube. 3.3 Solution-State NMR Spectroscopy Experiments for Resonance Assignment and Chemical Shift Perturbation Studies

1. To achieve nearly complete sequence-specific backbone chemical shift assignment of 1HN, 15N, 13Cα, 13Cβ, and 13C’ use the TROSY versions of HNCA, HN(CO)CA, HNCACB, HN (CO)CACB, HN(CA)CO, HNCO [14]. In combination, a 3D (HN,HN)-HMQC-NOESY-TROSY with 15N, 15N and 1 N H evolution in the t1, t2, and t3 dimensions, respectively, was recorded on 0.9 mM U-[2H, 13C,15N]-GGC.

3.3.1 Sequence-Specific Assignment of Backbone

2. Process spectra, for example using the software NMRPipe [16] and assign resonances, e.g., in the software CCPNMR [17] (Fig. 2).

Methyl Labelling of Membrane Proteins

117

600

UV280 Absorption (mAU)

500 400 300 200 100 0

Fig. 1 Purification of GGC. (a) SDS-PAGE gel (12%) showing different steps of purification process. (b) SDS-PAGE gel (12%) showing the different fractions from size-exclusion chromatography. (c) Typical size exclusion chromatography profile of the eluate from the metal affinity column purification

118

Vilius Kurauskas et al.

b

a

10 Ile-δ1

Ile-δ1

15 Ile-γ2

Ala-β

20

13

Val-γ∗

C (ppm)

Met-ε

Thr-γ2

Leu-δ∗

25

c

d

10

Val-γproS

Ala-β

20

Val-γ∗

13

Ala-β

C (ppm)

15

Leu-δproS

1.5

1.0

Leu-δ∗

0.5 1

H (ppm)

0.0

1.5

25

1.0

0.5 1

0.0

H (ppm)

Fig. 2 Examples of specific methyl labeling of GGC. (a) 2D (1H,13C)-CT-HSQC of U-[13C, 15N]-GGC. Location of different methyl groups is highlighted by colored ellipses, two ellipses are shown for racemic methyl groups of Leu and Val. 2D (1H,13C)-HMQC of (b) U-[D, 15N],(Ile-δ1)-[13CH3], (c) U-[D,15N],(Ala-β, Leu/ValproS)-[13CH3] or (d) U-[D, 15N],(Ala-β)-[13CH3]- (Leu/Val)-[13CH3,12CD3]-labeled GGC

Methyl Labelling of Membrane Proteins

119

1. To assign methyl groups, one 15N–TROSY-HSQC, one 13 C–HMQC, and two 3D NOESY experiments 13C–HSQCNOESY-15 N-HSQC (CN-NOESY, 350 ms) and 15N–HSQCNOESY-13C-HSQC (NC-NOESY, 420 ms) are recorded for each labeled sample (Fig. 3).

3.3.2 Methyl Group Assignment

2. Start from a methyl group resonance (Fig. 3a) and navigate in the CN-NOESY spectrum to find all cross-peaks corresponding to magnetization transfer from the methyl group to assign (Fig. 3b). Then for each possible amide resonances (Fig. 3c), navigate in the NC-NOESY spectrum to find

13

C-HMQC

107.8 ppm 13C-HSQC-NOESY-15N-HSQC

A

24

21 120.8 ppm 13C-HSQC-NOESY-15N-HSQC

219 Val

22 24

13

23

13

C (ppm)

B 22

22

C (ppm)

20

24

123.1 ppm 13C-HSQC-NOESY-15N-HSQC

25

22 24

1.0

1.5

1

9.0

0.5

8.5

8.0 1

H (ppm)

D

C

7.5

7.0

6.5

H (ppm)

220 Gly

110

115

115

15

15

N (ppm)

N (ppm)

110

219 Val

120

120

218 Ile 125

125

22.98 ppm 15 N-HSQC-NOESY-13C-HSQC 1.5

1.0 1

H (ppm)

0.5

15

9.0

8.5

8.0 1

7.5

7.0

N-TROSY 6.5

H (ppm)

Fig. 3 Approach for methyl assignment. (a, c) 2D (1H,13C)-HMQC and (1H,15N)-TROSY of U-[D, 15N],(Ala-β, Leu/ ValproS)-[13CH3]-labeled GGC. (b) Strips extract from 3D CNH-NOESY spectrum of cross-peaks corresponding to magnetization transfer from the methyl group 219 ValproS. (d) Section of 2D slice from 3D NCH-NOESY spectrum at carbon methyl resonance of 219 ValproS

120

Vilius Kurauskas et al.

cross-peaks corresponding to the methyl group resonance (Fig. 3d). 3. Get the proton and nitrogen frequencies from CN-NOESY and NC-spectrum, respectively. 4. Assign the methyl group from neighboring amide resonances found. 5. Repeat the process for each methyl group resonance (see Note 12).

4

Notes 1. D-(2H, 13C)-glucose (2H  98%; 13C  99%) should be used when uniform 13C labeling is desired for backbone assignment or in the context of methyl assignment with magnetization transfer from methyl to backbone. Also, uniformly 13C-labeled precursors must be used. 2. D-(2H, 12C)-glucose (2H  98%) should be used when best sensitivity and resolution is desired. Such a sample can be used for chemical shift perturbation studies with 1H-13C-HMQC experiments without the use of constant time, NOESY experiments and can also be used for relaxation experiments. 3. Precursors can be purchased in protonated form and dissolved in D2O for exchange to take place. 2H-13CH3–2-ketobutyric acid (13C  99%; 2H  98%) and 2H-13CH3–2-hydroxy-2methyl-3-oxo-4-butanoic acid (13C  99%; 2H  95%) can be prepared as described [19, 20], respectively. 2H-13CH3Alanine (13C  99%; 2H  98%) can be prepared by using the tryptophan synthase enzyme, as described [21]. 4. Critical step: If cultures did not reach final OD600  1.2, abort at this step. Bacterial adaptation was not optimal. 5. Critical step: Starting OD600 in D2O should be always 0.1. 6. Analyze protein expression levels by using SDS-PAGE. Spin down cell quantities equivalent to 200 μL of OD600 ¼ 1.0 at 12000  g for 2 min at room temperature and discard supernatant. Resuspend the pellet in 15 μL of 4 NuPage® LDS Sample buffer and 45 μL of H2O, sonicate and place the sample on a 90  C heat block for 5 min. Load 20 μL sample per lane and run the gel using NuPAGE® MES SDS running buffer. 7. Optimal temperature and IPTG conditions should be explored for each construct. In some cases, lower IPTG concentration results in higher expression levels. Lower temperature and increasing induction time results in higher expression (24 h at 18  C, or 12 h at 24  C). In the case of GGC, lower the temperature leads to decrease the amount of expression.

Methyl Labelling of Membrane Proteins

121

8. To minimize isotope scrambling and maximize isotope incorporation add the labeled precursors to the culture 1 h prior to induction. To do so, dilute the cultures with a large volume of media containing the precursors (200 mL). The cultures should reach back OD600 of 0.6 after 1 h. 9. Critical step: Avoid excessively prolonged growth after induction to prevent isotope scrambling by precursor recycling. 10. In this step, we used a fast exchange by removing the guanidine HCl and decreasing the amount of detergent in one step. However, in some case, a more gentle protocol can be used by successive washing of the column with solutions of decreasing concentration. 11. In this step, different solutions of detergent and mixture can be used to optimize the sample condition. 12. For the assignment of Alanine, 2H-13CH3-15 N-Alanine (13C  99%; 2H  98%; 15 N  98%;) must be used for labeling. In case 2H-13CH3-14 N-Alanine is used, no H-N correlation peak can be observed, and consequently no methyl-backbone NOESY cross peak. In such a case, the assignment of the methyl group has to rely on NOE contacts to other backbone sites.

Acknowledgments This work was financially supported by the European Research Council (ERC-Stg-2012- 311318-ProtDyn2Function). We would like to thank Pr. J. J. Chou and Dr. K. Oxenoid for stimulating discussions, and also Dr. J. Boisbouvier and Dr. R. Kerfah for helpful insights on labeling protocols. References 1. Ruschak AM, Kay LE (2010) Methyl groups as probes of supra-molecular structure, dynamics and function. J Biomol NMR 46:75–87 2. Tugarinov V, Hwang PM, Ollerenshaw JE, Kay LE (2003) Cross-correlated relaxation enhanced 1H-13C NMR spectroscopy of methyl groups in very high molecular weight proteins and protein complexes. J Am Chem Soc 125:10420–10428 3. Sprangers R, Kay LE (2007) Quantitative dynamics and binding studies of the 20S proteasome by NMR. Nature 445:618–622 4. Tugarinov V, Kanelis V, Kay LE (2006) Isotope labeling strategies for the study of high-molecular-weight proteins by solution NMR spectroscopy. Nat Protoc 1:749–754

5. Kerfah R, Plevin MJ, Sounier R, Gans P, Boisbouvier J (2015) Methyl-specific isotopic labeling: a molecular tool box for solution NMR studies of large proteins. Curr Opin Struct Biol 32:113–122 6. Hiller S, Garces RG, Malia TJ, Orekhov VY, Colombini M, Wagner G (2008) Solution structure of the integral human membrane protein VDAC-1 in detergent micelles. Science 321:1206–1210 7. Imai S, Osawa M, Takeuchi K, Shimada I (2010) Structural basis underlying the dual gate properties of KcsA. Proc Natl Acad Sci U S A 107:6216–6221 8. Gautier A, Mott HR, Bostock MJ, Kirkpatrick JP, Nietlispach D (2010) Structure

122

Vilius Kurauskas et al.

determination of the seven-helix transmembrane receptor sensory rhodopsin II by solution NMR spectroscopy. Nat Struct Mol Biol 17:768–774 9. OuYang B, Xie S, Berardi MJ, Zhao X, Dev J, Yu W, Sun B, Chou JJ (2013) Unusual architecture of the p7 channel from hepatitis C virus. Nature 498:521–525 10. Jaremko L, Jaremko M, Giller K, Becker S, Zweckstetter M (2014) Structure of the mitochondrial translocator protein in complex with a diagnostic ligand. Science 343:1363–1366 11. Schanda P, Huber M, Boisbouvier J, Meier BH, Ernst M (2011) Solid-state NMR measurements of asymmetric dipolar couplings provide insight into protein side-chain motion. Angew Chem Int Ed 50:11005–11009 12. Agarwal V, Xue Y, Reif B, Skrynnikov NR (2008) Protein side-chain dynamics as observed by solution- and solid-state NMR spectroscopy: a similarity revealed. J Am Chem Soc 130:16611–16621 13. Andreas LB, Le Marchand T, Jaudzems K, Pintacuda G (2015) High-resolution protondetected NMR of proteins at very fast MAS. J Magn Reson 253:36–49 14. Sounier R, Bellot G, Chou JJ (2015) Mapping conformational heterogeneity of mitochondrial nucleotide transporter in uninhibited states. Angew Chem Int Ed 54:2436–2441 15. Vozza A, Blanco E, Palmieri L, Palmieri F (2004) Identification of the mitochondrial GTP/GDP transporter in Saccharomyces cerevisiae. J Biol Chem 279:20850–20857 16. Delaglio F, Grzesiek S, Vuister GW, Zhu G, Pfeifer J, Bax A (1995) NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J Biomol NMR 6:277–293 17. Vranken WF, Boucher W, Stevens TJ, Fogh RH, Pajon A, Llinas M, Ulrich EL, Markley JL, Ionides J, Laue ED (2005) The CCPN data model for NMR spectroscopy: development of a software pipeline. Proteins 59:687–696 18. Seidman CE, Struhl K, Sheen J, Jessen T (2001) Introduction of plasmid DNA into cells. Curr Protoc Mol Biol Chapter 1:Unit 1.8 19. Gardner KH, Kay LE (1997) Production and incorporation of 15N, 13C, 2H (1H-d1 methyl) isoleucine into proteins for multidimensional NMR studies. J Am Chem Soc 119:7599–7600 20. Gans P, Hamelin O, Sounier R, Ayala I, Dura MA, Amero CD, Noirclerc-Savoye M, Franzetti B, Plevin MJ, Boisbouvier J (2010) Stereospecific isotopic labeling of methyl groups for NMR spectroscopic studies of high-

molecular-weight proteins. Angew Chem Int Ed 49:1958–1962 21. Ayala I, Sounier R, Use N, Gans P, Boisbouvier J (2009) An efficient protocol for the complete incorporation of methyl-protonated alanine in perdeuterated protein. J Biomol NMR 43:111–119 22. Velyvis A, Ruschak AM, Kay LE (2012) An economical method for production of (2)H, (13)CH3-threonine for solution NMR studies of large protein complexes: application to the 670 kDa proteasome. PLoS One 7:e43725 23. Gelis I, Bonvin A, Keramisanou D, Koukaki M, Gouridis G, Karamanou S, Economou A, Kalodimos CG (2007) Structural basis for signalsequence recognition by the translocase motor SecA as determined by NMR. Cell 131:756–769 24. Fischer M, Kloiber K, Hausler J, Ledolter K, Konrat R, Schmid W (2007) Synthesis of a 13C-methyl-group-labeled methionine precursor as a useful tool for simplifying protein structural analysis by NMR spectroscopy. Chembiochem 8:610–612 25. Kerfah R, Plevin MJ, Pessey O, Hamelin O, Gans P, Boisbouvier J (2015) Scrambling free combinatorial labeling of alanine-beta, isoleucine-delta1, leucine-proS and valineproS methyl groups for the detection of long range NOEs. J Biomol NMR 61:73–82 26. Ruschak AM, Velyvis A, Kay LE (2010) A simple strategy for (1)(3)C, (1)H labeling at the Ile-gamma2 methyl position in highly deuterated proteins. J Biomol NMR 48:129–135 27. Ayala I, Hamelin O, Amero C, Pessey O, Plevin MJ, Gans P, Boisbouvier J (2012) An optimized isotopic labelling strategy of isoleucinegamma(2) methyl groups for solution NMR studies of high molecular weight proteins. Chem Commun 48:1434–1436 28. Lichtenecker RJ, Coudevylle N, Konrat R, Schmid W (2013) Selective isotope labelling of leucine residues by using alpha-ketoacid precursor compounds. Chembiochem 14:818–821 29. Miyanoiri Y, Takeda M, Okuma K, Ono AM, Terauchi T, Kainosho M (2013) Differential isotope-labeling for Leu and Val residues in a protein by E. coli cellular expression using stereo-specifically methyl labeled amino acids. J Biomol NMR 57:237–249 30. Mas G, Crublet E, Hamelin O, Gans P, Boisbouvier J (2013) Specific labeling and assignment strategies of valine methyl groups for NMR studies of high molecular weight proteins. J Biomol NMR 57:251–262 31. Tugarinov V, Kay LE (2004) An isotope labeling strategy for methyl TROSY spectroscopy. J Biomol NMR 28:165–172

Methyl Labelling of Membrane Proteins 32. Lichtenecker R, Ludwiczek ML, Schmid W, Konrat R (2004) Simplification of protein NOESY spectra using bioorganic precursor synthesis and NMR spectral editing. J Am Chem Soc 126:5348–5349 33. Lichtenecker RJ (2014) Synthesis of aromatic (13)C/(2)H-alpha-ketoacid precursors to be used in selective phenylalanine and tyrosine protein labelling. Org Biomol Chem 12:7551–7560

123

34. Lichtenecker RJ, Weinhaupl K, Schmid W, Konrat R (2013) Alpha-Ketoacids as precursors for phenylalanine and tyrosine labelling in cell-based protein overexpression. J Biomol NMR 57:327–331 35. Schorghuber J, Sara T, Bisaccia M, Schmid W, Konrat R, Lichtenecker RJ (2015) Novel approaches in selective tryptophan isotope labeling by using Escherichia coli overexpression media. Chembiochem 16:746–751

Chapter 7 Labeling of Membrane Complexes for Electron Microscopy Francesca Gubellini and Re´mi Fronzes Abstract Localization of specific subunits or domains of interest inside protein complexes can be challenging, especially for membrane machineries. The amphipatic nature of their subunits and their modular organization results in difficult genetic manipulation and instability upon purification. Here, we present different labeling approaches that have been demonstrated successful in the structural characterization of large membrane complexes. Key words Membrane complexes, Antibody, Tag, Gold, Labeling, Electron microscopy, Negative stain

1

Introduction Membrane proteins and complexes allow cells to communicate with their external environment and underlie the proper functioning of crucial processes such as secretion, host-pathogen interaction, or tissue homeostasis. Visualization of membrane proteins at a subcellular resolution commonly involves fluorescence microscopy techniques, which rely on the addition of specific probes detectable by optical microscopes, such as fluorescent proteins, self-labeling tags, nanobodies, or enzymatic methods [1–3]. Despite the recent progress in super-resolution methods, this approach is not suitable to yield structural information on individual proteins or protein complexes. The only current technique, which can achieve such detail mapping, is transmission electron microscopy (TEM). Negative stain (NS) and cryo-electron microscopy (cryo-EM) are used to directly visualize purified proteins stained with heavy metals or frozen in vitrified ice, respectively. Specific applications of TEM are cryo-electron tomography and single particle analysis. The first one is used to image macromolecular complexes in their native environment and may yield structural details at the subnanometer resolution by subtomogram averaging. Single particle analysis is increasingly used to obtain three-dimensional structures of purified

Jean-Jacques Lacapere (ed.), Membrane Protein Structure and Function Characterization: Methods and Protocols, Methods in Molecular Biology, vol. 1635, DOI 10.1007/978-1-4939-7151-0_7, © Springer Science+Business Media LLC 2017

125

126

Francesca Gubellini and Re´mi Fronzes

proteins. Its technical advantages include the small amounts of proteins required, the absence of crystal growth constraints, and the possibility to capture different conformational states. All this makes TEM the method of choice for the investigation of large membrane complexes, complementary to crystallography, NMR, SAXS, and other techniques. Moreover, thanks to recent improvements in data acquisition and analysis, the number of structural models obtained by TEM on membrane proteins has increased continuously in the last few years, as well as their resolution [4, 5]. High-resolution cryo-EM structures allow for the assignment of proteins domains or subunits positioning in protein complexes through docking or de novo modeling [6, 7]. Nevertheless, when high resolution cannot be achieved and/or atomic models are not available, interpretation of electron microscopy (EM) density maps needs introduction of labeling procedures. However, addition of epitopes on membrane proteins and complexes may have a strong impact on the system due to the amphipathic nature of their components and the constraints of protein folding. Also, additional densities may interfere with the ability of subunits to interact among them and with other proteins. Different procedures can be used to label membrane proteins, each one presenting distinct strengths and weakness. The techniques most commonly used for interpreting low- to medium-resolution EM maps (below 15 A˚) may be divided into two groups: the first one based on the addition of protein mass to the studied protein/complex, and the second one based on the addition of gold particles. In the first strategy, EM analysis will allow for the identification of an additional density compared to the original sample. These extra densities are typically placed at the N- or C-terminal extremities, but also techniques for internal labeling have been introduced recently. These densities given by inert proteins can be genetically fused to the protein of interest, or bound to their target by affinity interactions (as for affinity tags or specific antibodies). Currently used fusion proteins are: maltose-binding protein (MBP) [8], dynein interacting domain [9], actin polymers [10], and green fluorescent protein (GFP) [11]. The addition of fused terminal labels has the advantage to produce homogeneous samples, in contrast with the use of gold or antibodies. Internal labeling techniques include the recently developed domain localization by RCT sampling (DOLORS) [12] and efficient mapping by internal labeling (EMIL) [11]. In the first one, the monovalent streptavidin is added to an avi-tag sequence (15 amino acids) encoding a substrate for the enzyme BirA, which performs site-biotinylation. The insertion regions are located within the main chain of the protein of interest, so that in theory any desired domain could be localized by this technique. However, this labeling requires multiple steps, which may reduce dramatically the labeling efficiency. To overcome this difficulty, the EMIL

Protein Labeling for Electron Microscopy

127

technique proposes to form chimera introducing the compact GFP protein (27 kDa) inside the protein of interest. Although the results produced with this approach are exciting, this requires a trial-and-error process for those proteins and complexes which structures have not been previously characterized. This may be extremely time consuming, especially for membrane complexes for which even small insertions may compromise protein folding and complex assembly. Strong ligand-binding affinity interactions represent a convenient alternative to avoid cloning of large regions inside the protein of interest. Streptavidin-based resins are largely used because of their interaction with a short strep-tag peptide (eight amino acids). Also, insertion of other short peptides such as His6, FLAG, or HA tags may allow the use of antibodies to identify the region of interest with a minimal modification of the protein (six, eight, and nine amino acids respectively). Full-length antibodies against commonly used tags are inexpensive, easy to use, and easily detectable even on raw NS electron microscopy images. However, the precise localization on the protein can result difficult due to their flexibility and large size. To this aim Fab’ fragments may also be used, and have the advantage of being less flexible and closer to the region of interest. They also can be used to add density to membrane proteins previously considered too small for single particle analysis [13]. However, monoclonal Fab’ production is more expensive and time-consuming compared to the introduction of a peptide-tag combined to the use of commercially -available antibodies. In the second strategy, binding of gold particles yields a strong signal that can be visualized directly on EM images of the target proteins, provided that the diameter of the gold is large enough (see below). Usually, gold is attached to the protein via a gold-labeled antibody or by affinity interaction between a gold-conjugated molecule (i.e., Ni-NTA Nanogold) or a protein (such as streptavidingold), and the specific tag present on the protein. The distance and flexibility between the antibody and the protein of interest can limit resolution, especially if the gold is attached on the secondary antibody. Gold can also be covalently linked to the protein of interest using monomaleimide gold on an exposed Cys residue. However, this method is not suitable for unstable proteins that cannot tolerate several reaction steps and moreover, labeling yields may not be satisfactory. Recently, three-dimensional models of two different large membranes-spanning machineries involved in bacterial secretion, the Type 6 and Type 4 Secretion Systems (T6SS and T4SS, about 1.7 and 3 MDa respectively), have been obtained using NS electron microscopy and single particle analysis. These models brought important novel biological information but their resolution was limited by specific challenges including high flexibility and

128

Francesca Gubellini and Re´mi Fronzes

instability. In such cases additional structural information is needed for the interpretation of map density through identification of proteins domains and subunits. In this chapter, we will focus on three labeling strategies used for interpreting three-dimensional EM maps of large protein complexes.

2 2.1

Materials Equipment

1. Centrifuges: benchtop centrifuge capable of spinning 1.5–2 mL eppendorf tubes at 15,000  g at 4  C. Centrifuges for 1 L and 50 mL tubes. Ultracentrifuge equipped with rotors 45Ti, SW32 or equivalent. 2. PCR thermocycler. 3. Cell disruptor such as Emulsiflex (Avestin), French Press, or similar. 4. Sonicator with microtip for lysis of small volumes (we use a Branson Ultrasonic Processor Cell Disruptor). 5. Material for SDS-PAGE electrophoresis (migration cell and power supply) and western blot transfer system. 6. Material for western blotting detection such as ECL or ECF. ECL material: Kodak films and automatic developer or a digital acquisition instrument (such as: MyECL ™ Imager, Thermo Fisher Scientific). Material for ECL detection: Storm system or Typhoon FLA 7000 (both from GE Healthcare). 7. Grids glow discharger (such as ELMO by Corduan technologies or Q150T Turbo-Pumped Sputter Coater/Carbon Coater, Quorum Technologies). 8. A TEM microscope with low dose setting and possibly equipped with automatic data collection software (EPU or SerialEM). For screening we use a FEI Tecnai T12 BioTWIN LaB6 microscope operating at a voltage of 120 kV, and for data collection a FEI Tecnai F20 FEG microscope operating at a voltage of 200 kV equipped with a FEI Falcon-II detector (Gatan). 9. Software: EMAN2 [14], RELION [15], and Chimera [16] packages.

2.2 Reagents and Consumables

1. Trizma base, NaCl, Imidazole (pH 8 solution), and specific detergents or chemicals necessary to purify and stabilize the protein of choice. 2. BL21(DE3) or C43(DE3) cells. 3. Uranyl acetate 2% solution in water (CAUTION: uranyl acetate is both toxic and radioactive. It should be manipulated in

Protein Labeling for Electron Microscopy

129

restricted areas wearing appropriate protection as gloves and safety glasses). 4. 5 nm Ni-NTA- Nanogold® (Nanoprobes). 5. Antibodies anti-His6 and/or anti-Strep tags. 6. Optional: Mini-PROTEAN® TGX precast gel for fast and sensitive protein detection. 7. Chemicals for Western blot reaction with secondary antibody: enhanced chemiluminescence (ECL) reacting with Horseradish Peroxidase (or ECF for Alkaline Phosphatase conjugated antibodies). 8. Silver-stain reagents or Silver Quest kit (Invitrogen). 9. Dumont anti-capillary tweezers N5 AC. Stainless steel, superfine tips. Length 110 mm. 10. Whatman filter paper no. 1. 11. Copper EM grids, 200- or 400-mesh covered with carbon film (Cu-200-CF or Cu-400-CF at Electron Microscopy Sciences). Alternatively, home-made carbon-coated grids can be prepared following standard procedures [17].

3

Methods The methods described below represent three among the currently used techniques for protein domains localization in large membrane complexes: gold labeling (Subheading 3.1), affinity labeling (specifically antibody labeling, Subheading 3.2), and fusion protein insertion (specifically Maltose Binding Proteins, Subheading 3.3). These major procedures have been described in three subsessions each describing the corresponding protocols.

3.1 Gold Labeling on EM Grids

Different gold particles are available for labeling proteins with gold particles. Nanogold® (Nanoprobes) is in general preferred over colloidal gold because it presents much lower aspecific binding. Nanogold® particles functionalized with Ni-NTA (Ni-NTANanogold®) bind to His-tagged proteins at a short distance from the tag (about 1.5 nm), increasing the accuracy of domain localization. Two types of particles are available to date with diameters of 1.8 nm and 5 nm, respectively. Unless considering silver enhancing, the 1.8 nm ones are poorly visible and their utilization has been limited to few cases [18]. Larger, 5 nm particles can be easily observed by negative stain EM (NS-EM). On the other side, automated 2D classification analysis of the labeled images is difficult due to the strong signal of electron-dense gold particles. Therefore results analysis and interpretation need to be performed on raw images. This method is thus better suited for

130

Francesca Gubellini and Re´mi Fronzes

large membrane complexes presenting distinguishable features and it is mostly used to understand complexes orientation or position of bulky domains [8, 19, 20]. Because of the multiple sites of binding to Ni-NTA on a single 5 nm particle, in-solution labeling triggers protein precipitation. Adjusting protein-to-gold ratio and performing an additional gel-filtration step can require high amount of stable protein. The on-grid Ni-NTA-Nanogold® labeling is a one-step procedure that can be performed conveniently on large and fragile membrane complexes, using low amount of protein. However, it requires paying particular attention to stringency and controls, to avoid aspecific labeling. 1. Purify the protein complex following specific protocols for the protein of interest. Adjust the buffer in order to make it compatible with Histidine-Ni interaction (see Note 1). Typical working buffer is 50 mM Hepes pH 7.5, 150 mM NaCl, containing the detergent(s) of choice. 2. Adjust protein concentration as suitable for NS-EM analysis. Starting concentration is from 10 to 50 μg/mL but this should be determined experimentally for each protein and buffer. 3. Glow-discharge a carbon-coated EM grid for 20–30 s at 2 mA. 4. Apply the grid over a cold 10 μL sample drop placed on a clean piece of Parafilm and let it adsorb on the carbon for 2 min (see Note 2). 5. Remove the excess of liquid by blotting it on a Whatman no. 1 filter paper, and quickly place the grid on the top of 50 μL cold buffer drop including the buffer of choice and 5–50 mM Imidazole (see Note 3). 6. Blot rapidly and place the grid upside down on the same buffer containing Ni-NTA- Nanogold® 5 nM for 2 min (see Note 4). 7. Wash twice by displacing the grids on the top of two 50 μL drops of cold buffer containing imidazole (without gold beads) to eliminate unspecific binding of gold particles. 8. Wash on a 50 μL drop of buffer without imidazole (see Note 5). 9. Remove the excess of liquid as described in step 5 and quickly place the grid upside down on a 10 μL drop of uranyl acetate 2% solution at room temperature (CAUTION: wear appropriate protection for the manipulation of uranyl acetate). 10. After 20–30 s of gentle agitation, quickly blot and transfer on a second uranyl acetate 2% drop. Repeat this step if necessary to optimize the staining. 11. Carefully blot the excess of liquid from the grids by sliding the grid on a Whatman filter paper and let dry the grids before analyzing it in a transmission electron microscope.

Protein Labeling for Electron Microscopy

131

Fig. 1 Gold Labeling on Electron Microscopy grids. (a) Negative Stain EM image of labeled Type 4 Secretion Systems (T4SS3–10 as described in [8]). The 5 nm Nanogold® beads were added directly on the sample previously absorbed onto carbon-coated EM grids. Ni-NTA gold nanoparticles bind to the His-tag on subunit VirB4/TrwK forming the barrels of the inner membrane complex (IMC) [8]. Red arrows indicate the core of the T4SS complexes and blue arrows indicate the IMC. (b) Selected labeled particles from EM images aligned with the core complex on top of the IMC. (c) Schematic of the 3D model of T4SS3–10 with yellow dots illustrating the various positions observed after Nanogold® labeling

12. Screen the samples on a transmission electron microscope under low dose conditions to limit electron-beam damage [21]. Select the labeling conditions for which the control shows a clean background compatible with significant labeling efficiency of the sample. Collect images of labeled particles at the appropriate magnification using defocus 1.5 to 2.5 μm to obtain contrasted images clearly showing protein features. 13. Import images in EMAN2/BOXER [14] or equivalent programs in order to select and extract particles showing goldbound molecules (Fig. 1). 14. Repeat the procedure on a protein sample lacking the affinity tag, as a control. 3.2 Antibody Labeling

Several methods of labeling based on protein-protein interaction are available to date. Antibody labeling has the advantage of being relatively straightforward and it has been used for soluble and membrane complexes [22–24]. If the presence of detergent interferes with protein-antibody interaction in solution, precluding EM visualization of the complex, the detergent may be exchanged with Amphipols (A8-35, Anatrace) [25]. The following protocol describes labeling by binding of monoclonal anti-His antibodies to a His6-tagged protein.

132

Francesca Gubellini and Re´mi Fronzes

1. Concentrate freshly prepared protein at 0.5 mg/mL following a specific protocol for the protein of interest. 2. Prepare three samples with 20–50 μL each, final volume (see Note 6). To each sample add anti-His6 antibodies (see Note 7) at a different molar ratio. Typical protein-to-antibody ratios to test are 1:1, 2:1, and 4:1 (see Note 8). 3. Incubate for 30 min at 4  C. Spin the sample in a benchtop centrifuge at 20,000  g for 10 min at 4  C to eliminate possible precipitants. 4. Inject each sample into an analytical gel filtration column. This will eliminate the excess of unbound antibody and small aggregates (see Note 9). 5. Analyze the fractions on a SDS-PAGE gel to identify the elution of protein–antibody complex. Detect the protein bands by Silver stain or TGX labeling (see Note 10). 6. Based on the results of steps 4 and 5, choose the best proteinto-antibody ratio and repeat the steps 1–5 on the entire protein preparation sample, to obtain the labeled complex. 7. Measure and adjust protein concentration for NS-EM on carbon-coated copper grids (typically 10–50 μg/mL). 8. Perform Negative Stain of the sample on the grid as follows. Add 5 μL sample on a carbon-coated grid and incubate for 30 s to 2 min. Remove the excess of buffer by quickly blotting on Whatman no. 1 filter paper. Rinse the protein by placing it on a drop of water before stain by three 20 s incubations on three drops of uranyl acetate 2% (see Note 3). Blot completely to remove the liquid and let it dry before inserting in the TEM microscope. 9. Screen the grids to find the best labeling and staining conditions. Collect images in low-dose settings at defocuses between 0.5 and 2 μm. The magnification should be chosen depending on the particle size. 10. Evaluate and import micrographs by using EMAN2/BOXER [14] before selecting particles. 11. Analyze the particles by dedicated software such as RELION [15] to produce 2D class averages representing characteristic views of the complex by reference-free classification and averaging. Comparison of the class averages for unlabeled and labeled samples will consistently show extra-density in the region whit the bound antibody (Fig. 2).

Protein Labeling for Electron Microscopy

133

Fig. 2 Immunolabeling by monoclonal anti-His antibodies. A T4SS Core Complex [24] is labeled by addition of anti-His antibodies to the His6-tagged TraF/VirB10 subunit. (a) The antibody position on the complex is visualized as additional density on representative class averages, and colored in red in the bottom line. (b) Schematic of the 3D model of T4SS Core complex indicating the position of the anti-His antibody

3.3 MBP Fusion Labeling

The Maltose Binding Protein is part of the maltose/maltodextrin system regulating the uptake of maltodextrins in Escherichia coli. MBP-fusion has been shown to improve membrane protein solubility in E. coli. Moreover, its 42.5 kDa compact mass makes it a good candidate to identify the position of attached terminal domain. This protocol is designed for membrane proteins, but the same procedure can be adjusted for soluble proteins (i.e., using pMAL-c5X vector and proceed with the supernatant after ultracentrifugation in step 2). 1. Clone the MBP at the N-terminus of the target membrane protein by inserting the gene of interest in a commercial vector for MBP-fusion proteins as pMAL-p5X (New England Biolabs) (see Note 11). 2. Screen the MBP-fusion gene by colony PCR and control insertion by sequencing (primers for sequencing of the pMAL vector are available at www.neb.com). 3. Perform protein expression tests as follows. Transform BL21 cells and grow them in selecting media supplemented with 100 μg/mL Ampicillin. Induce 20 mL cells with 0.5 mM IPTG and grow them overnight at 16  C in the same media and antibiotic (see Note 12). Pellet and resuspend cells in 1.8 mL of lysis buffer (typically 20 mM Tris–HCl pH 8, 50 mM NaCl, 0.5 mg/mL lysozyme in the presence of DNase and 5 mM MgCl2) (see Note 13). Disrupt cells by sonication using a microtip (3 cycles of 40 s each at intensity 1–2, pulse mode). Centrifuge 10 min at 8,000  g to eliminate unbroken cells. To an aliquot of the supernatant add an

134

Francesca Gubellini and Re´mi Fronzes

equal volume of 4 SDS-PAGE Sample reducing buffer (see Note 14). Heat at 95  C for 10 min, spin down aggregated material, and apply 15 μL of supernatant on a SDS-PAGE gel. Compare expression of MBP alone, to the MBP-protein fusion and to not-induced sample (see Note 15). 4. Prepare larger-scale protein purification in the conditions established in step 1 (see Note 16). After lysis by sonication or cell disruptor (Emulsiflex, Avestin), centrifuge for 10 min at 12,000  g in a Sorvall fixed-angle rotor SS-34 to eliminate unbroken cells. Isolate bacterial membranes by ultracentrifugation (45 min at 150,000  g in a Ti45 Beckman rotor). Solubilize membranes using conditions previously optimized for the protein of interest and remove unsolubilized membranes by a second ultracentrifugation (45 min at 150,000  g in a Ti45 Beckman rotor). 5. Load the supernatant onto a pre-equilibrated amylose resin column (New England Biolabs) or an MBP-Trap column (GE Healthcare). Wash the column with buffer including the proper detergent. Elute with the same buffer supplemented with 10 mM maltose. Assess protein quality by SDS-PAGE gel. 6. If contaminants are observed, pull the fractions and concentrate before loading on a size exclusion chromatography. Analyze eluted fractions by SDS-PAGE gel. 7. To eliminate residual contamination ion exchange chromatography can be performed on column as Hitrap Q HP (GE Healthcare) (see Note 17). Dilute the sample lowering NaCl concentration to 25 mM and load it on the Hitrap Q HP column pre-equilibrated in the same buffer. Wash with 5 column volumes. Monitor the OD280 absorbance of the flowthrough to follow sample loading, wash and elution steps. Elute by continuous ionic strength gradient against the same buffer supplemented with 1 M NaCl over 10–20 column volumes. 8. Analyze the eluted protein by SDS-PAGE. 9. Prepare Negative Stain EM carbon grids with uranyl acetate 2% as described above (Subheading 3.2, step 8). 10. Perform imaging and analysis as described above (Subheading 3.2, steps 9–11). Identification of the extra density corresponding to the MBP fusion protein can be performed on class averages or on raw images (Fig. 3).

Protein Labeling for Electron Microscopy

135

Fig. 3 Labeling by localization of MBP fusion protein. (a) and (b) Negative Stain EM images of purified VirB4/ TrwK without (a) and with (b) an extra density corresponding to the MBP fusion protein. Class averages are shown in the insets corresponding to side and front views respectively (scale bar, 5 nm). (c) Fitting of the VirB4/TrwK monomer in the 3D model of T4SS3–10 [8] indicating the additional density due to MBP (light blue)

4

Notes 1. Check the buffer composition for any chemical that could interact with Ni-NTA-Nanogold® beads or its binding to the His-tag. In particular, take good care of avoiding the presence of EDTA, reducing agents or imidazole in the buffer that will be used for Ni-NTA-Nanogold® binding. If any of these reagents are present in the protein elution buffer, perform buffer exchange by desalting columns or dilute the undesired component and concentrate the protein on appropriate protein concentrators. More details about Nanogold® can be found at http://www.nanoprobes.com/pdf/Inf2082.pdf. 2. Absorption time on the EM grids is strongly dependent on the nature of the protein, its concentration and buffer. For longer incubation time, place the sample in a humidity and temperature-controlled environment to avoid evaporation or protein damage. Cold buffers should be used all along the procedure except for uranyl acetate 2% and following steps. For particularly unstable proteins work on a cold support covered with Parafilm. 3. Fast blotting procedure is required to avoid dehydration of the sample on the carbon grid. Rinsing of sample after binding on the carbon grid can be performed on a drop of buffer instead of water depending on the sample. 4. Other conditions of incubation time as well as imidazole and Ni-NTA-Nanogold® concentrations must be tested to identify the optimal binding conditions for obtaining negative background and gold-labeling conditions.

136

Francesca Gubellini and Re´mi Fronzes

5. A drop of water can be used instead of buffer with better results, depending on the detergent used. 6. Protein concentration of these aliquots should be enough for the protein to be detected (by absorbance at 280 nm) after the gel filtration of step 4. 7. Monoclonal anti-His6 antibody can be purchased from different companies. We used Sigma anti-polyhistidine produced in mouse. 8. Sub-stoichiometric amount of antibodies is used to limit precipitation, but the right concentration should be optimized for each protein and antibody. In both [20, 24] the optimal ratio was 2 moles of complex per 1 mole of antibody. 9. Analysis of small amount of samples can be conveniently performed on Tricorn 5/150 GL columns (GE Healthcare) using minimal amount of sample. Antibody binding could be observed by elution volume peak shift of 150 kDa. Protein aggregation will be reflected by lower OD280 absorption peak. 10. Because of the low protein concentration, Coomassie is usually not sensitive enough at this point. Miniprotean TGX precast gels (Biorad) present higher sensitivity than Coomassie and are faster than Silver Stain protocols. In the case of very low protein concentration Silver Stain will be needed because they present the best sensitivity. We use the Quest kit (Invitrogen), but any rapid protocol will work. 11. Alternatively, the MalE gene from E. coli can be amplified and inserted upstream the gene of interest in any plasmid of interest using the Clontech In-Fusion® HD cloning kit. The advantage of commercial vector is a tighter binding of MBP to amylose and the possibility to remove MBP domain using Factor Xa protease. 12. These are standard conditions. Optimization of protein production can require different media, temperature, or induction time that should be assessed specifically for each construct. 13. Lysis buffer as well as purification buffer should be optimized specifically for each protein for parameters such as: salt concentration, reducing conditions, and detergent used. 14. High SDS concentration will help solubilize the membranes present in the lysate. Prepare 5 mL of 4 SDS-PAGE Sample Buffer as for http://openwetware.org/wiki/SDS-PAGE_sam ple_buffer_(Morris_formulation). 15. MBP presence can also be assessed by using anti-MBP antibody and/or by protease digestion of MBP tag. 16. If the protein is thermosensitive, operate at 4  C until depositing the sample on an EM grid.

Protein Labeling for Electron Microscopy

137

17. Note that high salt concentration in the elution buffer can dissociate fragile complexes. If contaminants are present in low concentration or do not affect EM imaging this step can be avoided.

Acknowledgments The authors acknowledge Dr. Harry Low and Pr. Gabriel Waksman for their contribution on the T4SS section. This work was supported by funds from the Institut Pasteur, the CNRS, and the ERC. References 1. Guignet EG, Hovius R, Vogel H (2004) Reversible site-selective labeling of membrane proteins in live cells. Nat Biotechnol 22:440–444 2. Pleiner T, Bates M, Trakhanov S, Lee CT, Schliep JE, Chug H, Bohning M, Stark H, Urlaub H, Gorlich D (2015) Nanobodies: site-specific labeling for super-resolution imaging, rapid epitope-mapping and native protein complex isolation. Elife 4:e11349 3. Yano Y, Matsuzaki K (2009) Tag-probe labeling methods for live-cell imaging of membrane proteins. Biochim Biophys Acta 1788:2124–2131 4. Kuhlbrandt W (2014) Biochemistry. The resolution revolution. Science 343:1443–1444 5. Nogales E, Scheres SH (2015) Cryo-EM: a unique tool for the visualization of macromolecular complexity. Mol Cell 58:677–689 6. Wiedenheft B, Lander GC, Zhou K, Jore MM, Brouns SJ, van der Oost J, Doudna JA, Nogales E (2011) Structures of the RNAguided surveillance complex from a bacterial immune system. Nature 477:486–489 7. Chang L, Zhang Z, Yang J, McLaughlin SH, Barford D (2015) Atomic structure of the APC/C and its mechanism of protein ubiquitination. Nature 522:450–454 8. Low HH, Gubellini F, Rivera-Calzada A, Braun N, Connery S, Dujeancourt A, Lu F, Redzej A, Fronzes R, Orlova EV, Waksman G (2014) Structure of a type IV secretion system. Nature 508:550–553 9. Flemming D, Thierbach K, Stelter P, Bottcher B, Hurt E (2010) Precise mapping of subunits in multiprotein complexes by a versatile electron microscopy label. Nat Struct Mol Biol 17:775–778 10. Stroupe ME, Xu C, Goode BL, Grigorieff N (2009) Actin filament labels for localizing

protein components in large complexes viewed by electron microscopy. RNA 15:244–248 11. Ciferri C, Lander GC, Nogales E (2015) Protein domain mapping by internal labeling and single particle electron microscopy. J Struct Biol 192:159–162 12. Lau PW, Potter CS, Carragher B, MacRae IJ (2012) DOLORS: versatile strategy for internal labeling and domain localization in electron microscopy. Structure 20:1995–2002 13. Wu S, Avila-Sakar A, Kim J, Booth DS, Greenberg CH, Rossi A, Liao M, Li X, Alian A, Griner SL, Juge N, Yu Y, Mergel CM, Chaparro-Riggers J, Strop P, Tampe R, Edwards RH, Stroud RM, Craik CS, Cheng Y (2012) Fabs enable single particle cryoEM studies of small proteins. Structure 20: 582–592 14. Tang G, Peng L, Baldwin PR, Mann DS, Jiang W, Rees I, Ludtke SJ (2007) EMAN2: an extensible image processing suite for electron microscopy. J Struct Biol 157:38–46 15. Scheres SH (2012) RELION: implementation of a Bayesian approach to cryo-EM structure determination. J Struct Biol 180:519–530 16. Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE (2004) UCSF chimera – a visualization system for exploratory research and analysis. J Comput Chem 25:1605–1612 17. Booth DS, Avila-Sakar A, Cheng Y (2011) Visualizing proteins and macromolecular complexes by negative stain EM: from grid preparation to image acquisition. J Vis Exp pii:3227 18. Pattanayek R, Yadagiri KK, Ohi MD, Egli M (2013) Nature of KaiB-KaiC binding in the cyanobacterial circadian oscillator. Cell Cycle 12:810–817 19. Brocker C, Kuhlee A, Gatsogiannis C, Balderhaar HJ, Honscher C, Engelbrecht-Vandre S,

138

Francesca Gubellini and Re´mi Fronzes

Ungermann C, Raunser S (2012) Molecular architecture of the multisubunit homotypic fusion and vacuole protein sorting (HOPS) tethering complex. Proc Natl Acad Sci U S A 109:1991–1996 20. Durand E, Nguyen VS, Zoued A, Logger L, Pehau-Arnaudet G, Aschtgen MS, Spinelli S, Desmyter A, Bardiaux B, Dujeancourt A, Roussel A, Cambillau C, Cascales E, Fronzes R (2015) Biogenesis and structure of a type VI secretion membrane core complex. Nature 523:555–560 21. Chen JZ, Sachse C, Xu C, Mielke T, Spahn CM, Grigorieff N (2008) A dose-rate effect in single-particle electron microscopy. J Struct Biol 161:92–100 22. Dube P, Herzog F, Gieffers C, Sander B, Riedel D, Muller SA, Engel A, Peters JM, Stark H (2005) Localization of the coactivator Cdh1 and the cullin subunit Apc2 in a cryo-electron microscopy model of vertebrate APC/C. Mol Cell 20:867–879

23. Herzog F, Primorac I, Dube P, Lenart P, Sander B, Mechtler K, Stark H, Peters JM (2009) Structure of the anaphase-promoting complex/cyclosome interacting with a mitotic checkpoint complex. Science 323: 1477–1481 24. Fronzes R, Schafer E, Wang L, Saibil HR, Orlova EV, Waksman G (2009) Structure of a type IV secretion system core complex. Science 323:266–268 25. Popot JL, Althoff T, Bagnard D, Baneres JL, Bazzacco P, Billon-Denis E, Catoire LJ, Champeil P, Charvolin D, Cocco MJ, Cremel G, Dahmane T, de la Maza LM, Ebel C, Gabel F, Giusti F, Gohon Y, Goormaghtigh E, Guittet E, Kleinschmidt JH, Kuhlbrandt W, Le Bon C, Martinez KL, Picard M, Pucci B, Sachs JN, Tribet C, van Heijenoort C, Wien F, Zito F, Zoonens M (2011) Amphipols from a to Z. Annu Rev Biophys 40: 379–408

Chapter 8 Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane Proteins: Case Study of the Mouse 5-HT3 Receptor Ghe´rici Hassaı¨ne, Ce´dric Deluz, Luigino Grasso, Romain Wyss, Ruud Hovius, Henning Stahlberg, Takashi Tomizaki, Aline Desmyter, Christophe Moreau, Lucie Peclinovska, Sonja Minniberger, Lamia Mebarki, Xiao-Dan Li, Horst Vogel, and Hugues Nury Abstract There is growing interest in the use of mammalian protein expression systems, and in the use of antibodyderived chaperones, for structural studies. Here, we describe protocols ranging from the production of recombinant membrane proteins in stable inducible cell lines to biophysical characterization of purified membrane proteins in complex with llama antibody domains. These protocols were used to solve the structure of the mouse 5-HT3 serotonin receptor but are of broad applicability for crystallization or cryoelectron microscopy projects. Key words Membrane protein, Stable cell line, VHH, Lama antibody, Cys-loop receptor

1

Introduction Membrane proteins are often difficult to express and tricky to stabilize in a functional native state (or at least in a well-defined state). It is the combined progress of technologies for expression, stabilization, and the advances of structural techniques themselves that have permitted the mushrooming of membrane protein structures observed during the past decade (about ~120 structures in 2005 and ~600 in 2015). Here, we describe current protocols for the expression, the “crystallization-grade” or “EM-grade” purification, and for the stabilization using llama-derived antibody domains (VHHs), of the mouse 5-HT3 receptor. We reported the expression and purification of the 5-HT3A homopentameric receptor in 2013 [1] and then could solve its structure, in the presence of VHHs, by X-ray

Jean-Jacques Lacapere (ed.), Membrane Protein Structure and Function Characterization: Methods and Protocols, Methods in Molecular Biology, vol. 1635, DOI 10.1007/978-1-4939-7151-0_8, © Springer Science+Business Media LLC 2017

139

140

Ghe´rici Hassaı¨ne et al.

crystallography in 2014 [2]. The first part of the Methods section deals with a flavor of expression in mammalian cells well suited for proteins that do not express well, cell lines—how to obtain them, characterize them, and use them for large-scale production. The second part describes the purification of the 5-HT3 receptor. The third part presents a series of snapshots on how to select and characterize VHHs as crystallization chaperones in a structural project. While dealing with a specific case, we try to underline general principles for the reader desiring to crystallize his/her own membrane protein of interest. The case study exemplifies at least three of these principles: l

The choice of the expression system is crucial and choosing a system close to the organism of the target protein often makes sense (see Note 1).

l

The choice of the criterion for quality-control is instrumental. While it would be ideal to quantitatively assess the function of the target protein at each step, this is often impossible (e.g., for membrane transport proteins once solubilized). Alternative criteria can evaluate the binding activity (radio-ligand binding assays) or the stability of quaternary structure (fluorescence coupled size-exclusion chromatography) or merely evaluate the presence of a protein (Western blot).

l

The structure of the wild-type protein can be elusive; modifications (chopping or mutating specific amino-acids or domains) and/or external stabilization using a ligand or a crystallization chaperone are very often necessary.

The Cys-loop receptor family is a family of pentameric ligandgated ion channels, the ligand being in most cases one of the four neurotransmitters acetylcholine, serotonin, glycine, or GABA. In miniature, the evolution of what we know about Cys-loop receptor structure reflects well how technology shaped membrane protein structural studies. First, there were structures of the soluble part of the receptors (in this case a homologous protein found in marine snails [3, 4]) and one structure of a naturally abundant receptor isolated from the Torpedo fish [5]. Then came full-length structures of a couple of bacterial homologues [6, 7], the first that could be overexpressed in quantity and quality compatible with X-ray crystallography. A third period saw the successful application of insect and mammalian cell expression to obtain crystal structures of four different worm and mammalian receptors [2, 8–10]. Finally, high-resolution cryo-EM structures of a glycine receptor appeared in 2015 [11]. The move from natural, to bacterial, to mammalian receptors, the advent of cryo-EM as a valid alternative to X-ray crystallography are general tendencies in membrane protein structural studies.

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

141

In the coming years, the challenges for structural biology of Cys-loop receptors are diverse: structures of receptors in several conformations that can be unambiguously linked to functional states will help to understand the molecular mechanism of operation, and how much of it is common to the family or specific to a particular member; structures of new receptors and particularly of heteromeric receptors will hopefully provide relevant templates for rational drug design.

2 2.1

Materials Cell Culture

1. Incubation of adherent cells: standard incubator with temperature and CO2 monitoring. 2. Incubation of cells in suspension: orbital shaker with agitation, temperature, and CO2 control. Example: Eppendorf S41i. 3. Cell counting: Neubauer chamber or automatic counter. Example: Lifetech countless II FL. 4. Adherent cell culture medium: DMEM/F12 Glutamax (Thermofisher 10565-018) supplemented with 10% Newborn Calf Serum (Sigma N4637). 5. Suspension cell culture medium: Freestyle 293 (Thermofisher 12338-026). 6. Antibiotics. T-REx cell lines: 5 μg/mL of blasticidin S HCl (Thermofisher A1113902). pcDNA5/TO: 100 μg/mL of hygromycin B (Thermofisher 10687010). 7. Cell dissociation: Trypsin-EDTA 0.05% (Thermofisher 25300054). 8. Cell washing buffer: Phosphate-Buffered Saline (PBS). Supplement with 1 mM EDTA for cell dissociation. 9. Plastic cell culture flasks (75 cm2 and 175 cm2). 10. Plastic multi-well plates (12-, 24-, and 96-well). 11. Plastic roller bottles (Dutscher 680048 smooth surface 850 cm2, 681048 pleated surface 1700 cm2). 12. 8-well slides for microscopy (LabTek II separable glass slide, Dutscher 055599). 13. Plastic cloning cylinders (Sigma C1059-1EA). 14. Plastic Erlenmeyers (Dutscher 355117B 125 mL). 15. Glass baffled Erlenmeyers (Roth PK16.1250 mL, Roth PK17.1500 mL and Roth AEP1.1 1L).

142

Ghe´rici Hassaı¨ne et al.

16. T-REx-293 cell line (Thermofisher R71007) and T-REx-CHO cell line (Thermofisher R71807). 17. Mammalian expression plasmid compatible with T-REx cell lines (pcDNA5/TO Thermofisher V103320 or similar), encompassing the gene of interest. 18. Mechanical cell lysis device, example: T-25 Ultra-Turrax (IKA). 19. Lipofectamine 2000 (Thermofischer) reagent for transfection or equivalent. 2.2

Bacterial Culture

1. Media: Terrific Broth (TB) medium, Tryptone, 12 g. Yeast extract, 24 g, glycerol, 5 g, Potassium Phosphate, Dibasic, 12.5 g and Potassium Phosphate, Monobasic, 2.3 g per liter, adjusted at pH 7.2. LB medium, Tryptone, 10 g. Yeast extract, 5 g, NaCl, 10 g per liter. 2YT medium, Tryptone, 16 g. Yeast extract, 10 g, NaCl, 5 g per liter. 2. Sucrose buffer: 100 mM Tris, 500 mM sucrose, 1 mM EDTA, pH 8. 3. Ni-NTA resin. 4. Size-exclusion chromatography (SEC), superose 6 10–300 column.

2.3 Biochemistry: Purification of the 5-HT3 Receptor

1. Cell resuspension buffer: 10 mM Hepes, 1 mM EDTA pH 7.4. 2. Membrane preparation buffer: 50 mM Tris–HCl, 500 mM NaCl, 0.01% C12E9 (~2CMC), pH 8.0. 3. Purification buffer: 50 mM Tris–HCl, 125 mM NaCl, 0.01% C12E9 (~2CMC), pH 7.4. 4. Antiprotease cocktail: 5 mg of chymostatin (Sigma C7268), leupeptin (Sigma L2884), antipain (Sigma A6191), pepstatin (Sigma P5318), and 25 mg of Aprotinin (Sigma A1153) are dissolved into 50 mL of water. This stock solution is diluted 100
in buffers where antiproteases are necessary. 5. Deglycosylation enzymes: PNGase F (NEB P0704L or homemade, see Subheading 3.7, in this case the plasmid was obtained from Addgene, reference 40315). 6. Limited proteolysis enzyme: Trypsin (Euromedex TB0626). 7. Concentrators: Amicon Ultra-15 or Ultra-4100 kDa cutoff (Millipore UFC910024 or UFC810024). 8. Detergents: C12E9 10% solution in 1 mL or 5 mL bottles under argon (Anatrace APO129). Use at 1% during solubilization and 0.01% during purification. Open a fresh bottle every second day. Cymal-6 (Anatrace C326).

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

143

9. Streptactin columns: gravity-flow resin (Streptactin Sepharose, IBA 2–1201-025) or prepacked 5 mL columns (Streptactin Superflow High capacity, IBA 2-1238-001). 10. Elution of proteins from Streptactin columns: d-Destiobiotin (Sigma D1411) at 5 mM in purification buffer. 11. HPLC system. Example: Bio-Rad NGC chromatography system, equipped with a C-96 autosampler, an external fluorescence detector (RF20
, Shimadzu), and mounted with a Superose 6 Increase 10/300 column (GE Healthcare). 2.4

Selection of VHHs

2.4.1 Phage Display and ELISA

1. Streptactin-coated 96-well microplate (iba 2–1501-001). 2. 96-well microplates for bacterial cell culture with adhesive paper tops. 3. 96-well microplates with deep wells for bacterial cell culture. 4. Phage Helper M13 K07 (NEB N0315S). 5. E. coli TG1 strain for the propagation of phage. LB and 2YT media. 6. Antibiotics: Ampicillin 100 mg/mL, Kanamycin 50 mg/mL. 7. 40% (w/v) glycerol solution to prepare the bacterial glycerol stocks. 8. BSA buffer: Tris 50 mM pH 8.0, NaCl 120 mM, 0.02% C12E9 (~4CMC), 3% BSA. 9. Coating and washing buffer: Tris 50 mM pH 8.0, NaCl 120 mM, 0.02% C12E9 (~4CMC). 10. Elution buffer: Glycine 0.2 M pH 2. 11. Neutralization buffer: Tris 50 mM pH 8.0. 12. PEG buffer: 20% PEG 8000, 2.5 M NaCl. 13. TMB colorimetric substrate (Sigma T5525-50TAB). 14. Blocking solution: 1 M H2SO4. 15. Petri dishes (diameter 60 mm Sigma Z64084, 120 mm Sigma P5487). 16. Anti-M13 antibody coupled to HRP (GE Healthcare 27942101).

2.4.2 Radio-Ligand Competition Assay

1. Assay buffer: 10 mM Hepes, pH 7.4. 2. Radioligand [3H]-GR65630 (Perkin-Elmer NET1011250UC). 3. Polyethyleneimine 0.5% (w/v). 4. Ultima Gold (Perkin-Elmer 6013329). 5. Fiber filters GF/B.

144

Ghe´rici Hassaı¨ne et al.

2.4.3 Labeling

1. Streptactin Chromeo 488 conjugate (IBA 2-1542-050). 2. NTA Atto 488 BioReagent (Sigma 39625). 3. A whole range of fluorescent ligands specific of the 5-HT3 receptor have been developed by Vogel and coworkers, and Lochner and coworkers (reviewed in [12]). 4. Toxin specific to α7 nicotinic acetylcholine receptor: α-bungarotoxin-tetramethylrhodamine (Sigma T0195). 5. VHH His-tag labeling during cytometry experiments: mouse anti-His (Clonetch 631212) and anti-mouse IgG-PE (Santra Cruz 3738) antibodies. 6. Anti-His Peroxidase conjugate antibody (Sigma A7058). 7. Anti-Strep Peroxidase conjugate antibody (iba 2-1509-001).

2.4.4 Native Electrophoresis

1. Anode buffer: 50 mM BisTris, 50 mM Tricine, pH 6.8. 2. Cathode buffer: Anode buffer supplemented with 0.02% Coomassie G-250. 3. Sample buffer 4
stock solution: 200 mM BisTris, 40 mM HCl, 200 mM NaCl, 40% w/v glycerol, 0.004% Ponceau S, pH 7.2.

3

Methods

3.1 Development of Stable Inducible Cell Lines

The development of a clonal cell line takes approximately 3 months and the day numbers given in the protocol are approximates. Twelve cell lines can be conveniently prepared at the same time. Cell lines provide reproducibility and high levels of expression but their obtention is long. They are not suited for mutant screening. Alternatives for large-scale expression in mammalian cells include transient transfection [13] and baculovirus infection (BacMam) [14]. Here, we describe inducible stable cell lines in which the protein expression is repressed in the absence, and induced in the presence of tetracycline [15], which can allow levels of expression that would be toxic if the expression were constitutive. Day 1 1. Thaw an aliquot of 1 mL of T-REx-293 or T-REx-CHO cells (see Note 2).

2. Immediately pour into a 75 cm2 flask containing 20 mL of adherent cell culture medium. This is a sufficient dilution of the DMSO present in the frozen medium for cryoprotection. Day 3 3. Incubate for 2 days, then change the medium, adding a supplement of antibiotics (5 μg/mL of blasticydine selecting for the TetR repressor).

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

145

4. Maintain the cells until they reach reproducible and robust growth. This step typically takes about 10 days. Day 13 5. Prepare the transfection by passing the cells to >70% confluence in a 6-well plate.

6. Transfect the gene of interest. Typically, mix 200 μL of Optimem and 2.5 μg of DNA in a tube, 200 μL of Optimem and 4 μL of lipofectamine 2000 in a second tube, tap the tubes and wait for 5 min. Put the content of the first tube into the second one, wait for 20 min. Add 200 μL of the transfection mix per well of a 6-well plate in 2 mL of medium (see Note 3). Day 14 7. The next day, remove medium and detach the cells by pipetting up and down using 0.5 mL of trypsin-EDTA solution and 1.5 mL of medium.

8. Dilute in 3 Petri dishes containing 10 mL of medium supplemented with antibiotics (5 μg/mL of blasticydine for the Tet operator and 100 μg/mL of hygromycin for the gene of interest), at dilutions 1/50, 1/200, and 1/400, by adding 200, 50, and 25 μL of the cell solution. From this step on, antibiotics should always be present in the medium. 9. Put the Petri dishes in the incubator and let the cells grow, changing medium from time to time (typically once per week). Day 35 10. After about 3 weeks, Petri dishes should present a high number of single cell colonies. Fill two 96-well plates with 200 μL of medium with antibiotics per well.

11. With a pencil, circle about 20 well-isolated colonies on the bottom surface of the Petri dishes. 12. Remove the medium, place cloning cylinders on the chosen colonies, add 50 μL of dissociation solution, and wait for 5 min for the cells to detach. 13. Pipet up and down the cells of the first colony and put into the well A1 of the first plate. Repeat with the next colonies and put in wells B1, C1, . . . until you have picked 16 colonies. 14. Using a multichannel pipet, make 1/3 serial dilutions from left to right. Take 100 μL of solution, pour into the next well, homogenize, and repeat until your reach the last column. 15. Let it grow for 2–3 weeks. Day 63 16. Inspect the 96-well plates under a bench microscope to identify the wells with single colonies.

146

Ghe´rici Hassaı¨ne et al.

17. Remove the medium from the wells with a single colony. 18. Add 50 μL of dissociation solution to these wells and wait for 5 min. 19. Transfer the colonies to wells of a 24-well plate containing 1 mL of medium with antibiotics. 20. Let it grow for 1 week. Day 70 21. Place a cross on each well of the 24-well plate where cells are more than 40% confluent.

22. Prepare two new 24-well plates with 2 mL of medium in each well. 23. Dissociate the cells in the marked wells, using 150 μL of dissociation solution and 150 μL of medium per well. 24. Transfer 100 μL to the conservation plate (stored at 34  C). Transfer the remaining 200 μL to the test plate (stored at 37  C) that will be used for the protein expression test (see Note 4). 3.2 Big Cultures in 20 Roller Bottles

Day 1 1. Grow the cells to 90% confluence in a 150 cm2 flask.

2. Remove the medium, add 4 mL of dissociation solution and incubate for 5 min at 37  C. 3. Prepare four 150 cm2 flasks with 25 mL of medium (see Note 5). 4. Resuspend the cells by pipetting and add 1 mL of the cell suspension to each flask. 5. Grow the cells for ~3 days to reach 90% confluence. Day 4 6. Remove the medium from the four 150 cm2 flasks, add 4 mL of trypsin solution per flask, and incubate for 5 min at 37  C.

7. Prepare twenty 150 cm2 flasks with 25 mL of medium. 8. Resuspend the cells by pipetting and add 1 mL of the cells in suspension to each flask. 9. Grow the cells for ~3 days to reach 90% confluence. Day 7 10. Remove the medium from the twenty T150 flasks, add 4 mL of trypsin solution per flask.

11. Prepare 20 rollers with 450 mL of medium. 12. Resuspend the cells by pipetting. 13. Add 4 mL of resuspended cells per roller bottle.

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

147

14. Put the rollers in an incubator at 37  C (see Note 6). 15. Let the cells grow for 1 week. Day 14 16. Induce the expression by adding 4 mg/L tetracycline and 4 mM valproic acid (see Note 7). Day 16 17. After two more days of incubation during which the protein is expressed, remove the medium from the rollers, add 50 mL of PBS supplemented with 1 mM EDTA to detach the cells, and incubate for 5 min (see Note 8).

18. Resuspend the cells and centrifuge them at 2000
g for 10 min at 4  C. 19. Wash the cell pellet by resuspending them in 50 mL of DBPS and centrifuge them at 2000
g for 10 min at 4  C in 50 mL tubes. 20. Remove the PBS and weigh the cell pellets. 21. Freeze the pellets at 80  C for later use or proceed directly with the membrane preparation, Subheading 3.5. 3.3 Adaptation of Cells to Culture in Suspension

1. Grow adherent cells to confluence in a 75 cm2 flask. 2. Remove the medium, add 3 mL of dissociation solution, incubate for 5 min at 37  C. 3. Add 12 mL medium, resuspend the cells by pipetting and count them. 4. Centrifuge for 5 min at 400
g, discard the supernatant. 5. Resuspend in suspension medium supplemented with 10% serum at 106 cells/mL. 6. Transfer in cell culture 125 mL Erlenmeyers and start shaking on an orbital shaker at 37  C, 8% CO2 and 120 rpm (see Note 9). 7. Put the cells in a fresh suspension medium, at 106 cells/mL, two times per week. 8. Each week, decrease the amount of serum by approximatively a half, from 10% in the first week to 5% in the second week, to 2.5% in the third week and 1% in the fourth week (see Note 10). 9. Start to reapply the antibiotics pressure for the maintenance of the cell line once it has nicely adapted to suspension and shows normal growth and reasonable mortality (~5%). 10. Maintenance is done by diluting the cells to 0.5
106 cells/ mL intro fresh suspension medium, typically twice per week.

148

Ghe´rici Hassaı¨ne et al.

3.4 Big Cultures in Suspension

1. From a small 125 mL Erlenmeyer containing 20–50 mL of saturated culture of cells in suspension at about 3–4
106 cells/mL, start to increase the volume by diluting in a suspension medium to 0.5 cells/mL. Typical ratios for medium volume to Erlenmeyer volume are 10–40% (see Note 11). 2. When you have a sufficient volume of culture at 2
106 cells/ mL, induce the protein expression by supplementing with 4 mg/L of tetyracycline. 3. The next day, supplement with 4 mM of valproic acid. 4. Harvest the cells 48–72 h after induction by centrifugation. Proceed with membrane preparation or freeze at 80  C (see Note 12).

3.5 Membrane Preparation

1. Weigh the cell pellet. 2. Resuspend in cell resuspension buffer using 10 mL of solution per gram of cell. 3. Lyse the cells with a mechanical device. Typically with an UltraThurrax, 4-6
30 s of lysis at medium speed are enough. 4. Centrifuge at low speed, 4000
g, for 10 min to remove unbroken cells and debris, and save the supernatant for later (see Note 13). 5. Resuspend the pellet in one half of the volume used for the first resuspension, in cell resuspension buffer. 6. Homogenize again with the mechanical device. 7. Centrifuge at 4000
g for 10 min. Discard the pellet, pool this supernatant and the previous one. 8. Centrifuge at high speed, 120,000
g for 1 h. Weigh the empty tubes before filling them (see Note 14). 9. Discard the supernatant and weigh the tubes to get the mass of pelleted membranes. 10. Resuspend the membranes in membrane preparation buffer using 25 mL of buffer per gram of membrane (see Note 15). 11. Freeze for storage at 80  C.

3.6 Purification and Enzymatic Treatments

The 5-HT3 receptor purification protocol follows a classical 3-step scheme of solubilization, affinity purification, and size-exclusion chromatography. Additionally, it includes enzymatic treatments to remove carbohydrates and flexible parts of the receptor, which can be sources of heterogeneity. As enzymatic reactions will never be 100% efficient, it is recommended to treat sources of heterogeneity at the gene level when possible, by engineering smaller loops in flexible parts, mutating glycosylated asparagine residues, or using GnTI-HEK293 cells that do not produce complex N-linked glycans. Here, however, truncations or mutations resulted in receptors

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

149

that were not (or barely) expressed at the surface or aggregated receptors. Therefore, we relied on the treatment of the protein during its purification. Every step is performed at 4  C except noted otherwise. 3.6.1 Solubilization

1. Thaw the membrane suspensions in a room-temperature bath (1 g of membrane pellet has been resuspended in 25 mL of buffer; a classical preparation would start from 10 to 15 g of membrane pellet). Add 1% C12E9 detergent, use a freshly opened detergent stock solution. 2. Rotate on a wheel, or use a magnetic stirrer, for 2 h. Start the column equilibration at this step (see step 5). 3. Centrifuge to pellet the insoluble material, 100,000
g for 1 h (see Note 16). 4. Collect the supernatant. The pellet is usually small and translucid (see Note 17).

3.6.2 Streptactin Affinity Purification

5. Equilibrate the columns with the membrane preparation buffer. 6. Load the solubilized proteins on the columns. 7. Wash with 3 column volume using the purification buffer. 8. Elute with 2–4 column volumes of purification buffer supplemented with d-desthobiotin (see Note 18).

3.6.3 Deglycosylation

9. Concentrate the eluted pure protein up to 1 mg/mL on 100 kDa Amicon devices. 10. Add up to 50 units of PNGase-F per μg of protein. Gently rotate on a wheel at 37  C for 2 h (see Note 19).

3.6.4 Trypsin Treatment and Size-Exclusion Chromatography

11. Add 12 μg of trypsin per mg of protein (see Note 20). 12. Incubate at 30  C for 1 h 45 (see Note 21). 13. Block the trypsin activity by putting the tube on ice and adding 2 mM AEBSF (see Note 22). 14. Quickly concentrate the protein down to 0.3–1.0 mL using the same concentrator (as above) (see Note 23). 15. Spin on a table-top centrifuge at maximum speed for 5 min. 16. Load on a HPLC system and run the size-exclusion chromatography. We use a Superose 6 10–300 GL column equilibrated in the purification buffer. The injection loop is 1 mL and we put ~0.5 mL in it (clearly outside the manufacturer’s advised volume for a column of this size). 17. The usual peak containing the receptor-detergent complex elutes around 13.6 mL (see Note 24).

150

Ghe´rici Hassaı¨ne et al.

18. Pool the fractions of the peak, and concentrate the protein up to concentrations required for cryoEM or crystallography (1 and 2.5–10 mg/mL respectively for the 5-HT3 receptor). 3.7 Annex: PNGase F Expression and Purification

The PNGase F plasmid can be obtained from the Addgen public repository, where it was deposited by the Lott group that reported its purification in E. coli [16]. It expresses in the periplasm of bacteria. 1. Prepare an overnight preculture from a frozen glycerol stock of the PNGase F coding plasmid. 2. Inoculate 1 L of TB media supplemented with ampicillin to final O.D of 0.05 and incubate at 37  C. 3. Induce the expression of the PNGase F by adding 1 mM IPTG at OD 0.5–1. 4. Incubate at 20  C overnight. 5. Spin cells at 13,000
g for 20 min at 4  C. From this step on, use ice-cold solutions and work on ice. 6. Resuspend the pellet in 50 mL of sucrose buffer (5% of initial culture volume). The sucrose buffer contains 100 mM Tris, 500 mM sucrose, 1 mM EDTA, pH 8. 7. Centrifuge at 13,000
g for 15 min at 4  C. 8. Resuspend the pellet in 50 mL water and incubate for 10 min. 9. Add 1 mM MgCl2 and incubate on ice for 10 min. 10. Centrifuge at 13,000
g for 15 min at 4  C. Collect the supernatant and add 50 mM MOPS pH 7.0. 11. Load the protein solution on a gravity-flow column containing Ni-NTA resin equilibrated in membrane preparation buffer supplemented with 20 mM imidazaole. 12. Wash the resin twice with 5 column volumes of the same buffer. 13. Elute the PNGase F with 5 column volumes of purification buffer supplemented with 100 mM imidazole. 14. Load on a HPLC system and run the size-exclusion chromatography. We use a Superose 6 10–300 GL column equilibrated in the purification buffer with an injection loop of 1 mL. The usual peak containing the PNGase F elutes around 18.8 mL.

3.8 VHH-Receptor Complex Characterization

Crystallization chaperones in general, and llama-derived antibody fragments in particular have proved themselves extremely useful in membrane protein crystallography. In the case of the mouse 5-HT3 receptor, the VHH15 that permitted us to obtain diffracting crystals plays a double role: it stabilizes one conformation as evidenced from the functional screening (see the Electrophysiology paragraph

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

151

Subheading 3.8.7) and it extensively participates in crystal packing contacts. There are excellent and exhaustive publications describing the development of a VHH library [17] and we will focus only on the later steps of selection using phage display and then characterization of the VHHs once they are produced and purified. Of course, cocrystallization experiments are the only ones that can definitively show that one VHH is a correct crystallization chaperone, still in our experience it can be long and difficult to screen all VHHs by cocrystallization, and then even longer and more difficult to optimize the first hits into diffracting crystals. As an example, we have been trying to obtain another conformation of the 5-HT3 receptor using VHHs but have failed so far to find an exit to the maze of more than 150 crystallizations hits obtained with half a dozen VHHs. Hence, the methods described below can provide a qualitative way to rank the VHHs that are promising and can help to focus on a limited number of crystallization conditions for thorough optimization. 3.8.1 Selection of VHH Using a Phage-Display Library

The protocols in this section are minor variations of common phage-display and ELISA protocols. These minor variations are (a) the use of the purification tag of the receptor to coat it on surfaces instead of a nonspecific coating on Nunc plates and (b) the use of the mild detergent C12E9, the one of the receptor purification, instead of Tween. 1. Coating and two rounds of selection on a Streptactin-coated 96-well plate. Day 1 1. Plate TG1 bacteria on LB-Agar media from a glycerol stock. Incubate overnight at 37  C. Day 2 2. From a single colony, start a TG1 preculture in 4 mL of 2YT medium without antibiotics at 37  C and 200 rpm agitation.

3. Coat one well of a streptactin-coated plate with 100 μL of 5HT3 receptor at 50 μg/mL in coating buffer overnight at 4  C (see Note 25). Day 3 4. Wash the coated well three times with 200 μL of washing buffer (see Note 26).

5. Block the well with 100 μL of BSA buffer. Incubate for 2 h at 4  C. 6. Wash the well three times with 200 μL of washing buffer. 7. Add 100 μL of the phage library and 100 μL of coating buffer. Incubate for 2 h at 4  C.

152

Ghe´rici Hassaı¨ne et al.

8. Start a TG1 culture from the preculture at O.D. < 0.1. Use 4 mL of 2YT medium. Incubate at 37  C at 200 rpm agitation. 9. Wash ten times with 200 μL of washing buffer. For the second round of selection, wash 20 times. 10. Elute the phages bound to the protein by adding 100 μL of elution buffer. Incubate for 5 min at 4  C while shaking at 600 rpm (see Note 27). 11. Supplement the 100 μL solution containing the eluted phages with 50 μL of 100 mM Tris pH 8.5 to neutralize the glycine pH 5 buffer. 12. Use this 150 μL solution to infect the 4 mL culture of exponentially grown TG1 cells (at OD ~0.6). 13. Add the helper phage KM13 (2
1012 pfu/mL) to the culture and incubate for 30 min at 37  C without shaking. Continue the incubation for 30 min while shaking at 200 rpm. 14. For clone picking, prepare serial dilutions of the infected TG1 bacterial culture. Transfer 5 μL of the 4 mL culture into 1 mL of 2YT medium to obtain a first dilution of ~102, then continue by 1/10 dilutions to reach 103, 104, 105, 106 (for the first selection round) and down to 108 (for the second selection round). Spread 100 μL of each diluted solution on LB-Agar plates supplemented by 2% glucose and 100 μg/mL ampicillin. Incubate overnight at 30  C. 15. Pellet the rest of the 4 mL bacterial culture by centrifugation, at 4000
g for 5 min. 16. Resuspend the pellet in 1 mL of 2YT medium supplemented with 100 μg/mL ampicillin and 50 μg/mL kanamycin. The kanamycin will ensure selection of the cells infected by the helper. 17. Spread 500 μL of the infected bacteria suspension on LB-Agar big plates supplemented by 2% glucose and 100 μg/mL ampicillin. Incubate overnight at 30  C (see Note 28). 18. Incubate 500 μL of the infected bacteria suspension in 15 mL of 2YT medium supplemented with 100 μg/mL ampicillin, 50 μg/mL kanamycin, and 2% glucose, in a 50 mL culture tube. Incubate overnight at 30  C with shaking at 200 rpm. Day 3 19. Spin the overnight culture at 10,000
g for 10 min at 4  C and keep the supernatant that contains the phages.

20. For the phage precipitation, prepare a tube with 3 mL of PEG buffer and add the supernatant.

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

153

21. Spin for 10 min at 14,000
g at 4  C. Discard the supernatant. Resuspend the pellet in 1.2 mL of PBS supplemented with 15% glycerol. 22. Add 300 μL PEG buffer and incubate on ice for 30 min. 23. Spin for 10 min at 14,000
g at 4  C. Discard the supernatant. 24. Resuspend the pellet in 1 mL of PBS supplemented with 15% glycerol. 25. Spin for 5 min at 4  C at maximum speed. Collect the supernatant containing the purified library in a new tube. 26. Estimate the phage concentration by spectrometry. Measure the OD spectrum between 240 and 360 nm. [Number of phages/mL] ¼ ((Abs260-Abs320)
6
1016)/ (number of base pair of the phagemid). Use 100 μL for the second selection round and repeat all the steps from step 7 on (see Note 29). 2. ELISA binding assay. Day 1 27. Pick 95 colonies obtained at step 14 into a 96-well deep plate (the last well H12 is the negative control picked with TG1 grown in 2YT medium without antibiotics) containing 100 μL/well of 2YT medium supplemented with 100 μg/mL ampicillin and 2% glucose. Incubate at 37  C for 2 h and 500 rpm agitation.

28. Inoculate a second 96-well deep plate containing 1 mL/well of 2YT medium with 100 μg/mL ampicillin by pipetting 10 μL of each preculture in the corresponding well. The plate is incubated for 3 h at 37  C (to reach OD600 ¼ 0.6), then supplemented with 1 mM IPTG to induce VHHs expression and incubated overnight at 30  C. 29. Prepare a glycerol stock by adding glycerol to a final concentration of ~8% to the plate from step 27, mix carefully, and then freeze at 80  C. 30. Coat each well of a Streptactin-coated 96-well plate with 100 μL the 5-HT3 receptor at 20 μg/mL. Coat another plate, as a negative control, with BSA buffer. Place the two plates at 4  C overnight with 200 rpm agitation. Day 2 31. Spin culture plate from step 28 at 3000
g for 10 min at 4  C. Discard the supernatant and freeze the pelleted cells at 20  C for 30 min minimum.

32. Wash the two Streptactin-coated plates with 200 μL of washing buffer per well, repeat three times.

154

Ghe´rici Hassaı¨ne et al.

33. Block the two Streptactin-coated plates using 100 μL of BSA buffer per well. Incubate for 2 h at 4  C. 34. Thaw the plate containing the pelleted cells. Add 200 μL of PBS, incubate for 30 min at 4  C, and centrifuge the plate at 2000
g for 5 min. 35. Transfer 100 μL of the supernatant, which is the periplasmic extract, to each well of the two Streptactin-coated plates. Incubate for 2 h at 4  C with shaking at 200 rpm. 36. Wash three times with 200 μL of washing buffer. 37. Add 100 μL of anti-M13 antibody coupled to HRP at dilution 1/2000 in binding buffer and incubate for 1 h at 4  C with shaking. 38. Wash three times with 200 μL of washing buffer. 39. Add 100 μL of the TMB colorimetric substrate, some wells should turn yellow. 40. Stop the reaction by adding 100 μL of blocking solution after 2–10 min at room temperature. 41. Measure the absorption at 450 nm. All the ELISA positive clones are sent for sequencing. The different cDNA are then subcloned in an expression vector for large-scale expression, purification, and subsequent characterization. 3.8.2 Flow Cytometry

Flow cytometry can be used to quantify easily the binding of a fluorescent probe to proteins expressed at the cell surface. Because of its simplicity, we often start by this experiment, either for a first evaluation of new constructs (in this case using a probe targeted to an extracellular part or tag) or for evaluation of VHHs (in that case with a probe directed at the VHH poly-histidine tag), as depicted in Fig. 1. 1. Before the experiment. 1. Culture cells in a 25 cm2 flask. The following steps assume that the cells express the membrane protein of interest at their surface. In the case of the 5-HT3 receptor, we were inducting the expression with 4 μg/mL tetracycline 1–2 days before the experiment (see Note 30). 2. Day of the experiment. 2. CHO cells: add 5 mL PBS supplemented with 5 mM EDTA and place back in the incubator for 5 min. 3. HEK cells: add 6 mL PBS. 4. Tap the flask and pipet up and down to detach the cells. 5. Count the cells.

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

155

Cell line expressing the 7/5-HT3 chimeric receptor, incubated with a toxin against the 7 neurotransmitter binding site

100

Cell line expressing the 5-HT3 receptor incubated with VHH15

I

750

500

250 0

10e1 10e2 10e3 Fluorescence level (AU)

10e4

Induced cells

500

Frequency

Frequency

75 50 25

Frequency

NI

Induced cells

0 10e0 10e0

600

Non-induced cells

1000

Non-induced cells

400 300 200 100 0

-1

0

1

1e1

1e2

Fluorescence level (AU)

1e3

-1

0

1

1e1

1e2

1e3

Fluorescence level (AU)

Fig. 1 Example of VHH and construct screening by flow cytometry. Left panel. A VHH labeled with a fluorescent antibodies targeting its polyhistidine tag was applied on noninduced and induced cells expressing the 5-HT3 receptor. The experiment demonstrates that this particular VHH does bind to the extracellular domain of the 5HT3 receptor. Right panels. A fluorescent toxin specific of the α7 nicotinic receptor was applied on noninduced and induced cells expressing a chimeric α7/5-HT3 receptor. The toxin binds to an inter-subunit site and this probes the expression and correct folding of the pentameric receptor. This is a case of cell line characterization

6. Centrifuge for 5 min, 300
g, at room temperature. Discard the supernatant. 7. Resuspend the cell pellet in PBS to a concentration of 1–5.106 cells/mL. 8. Transfer 150 μL per sample into 1.5 mL tubes and add a VHH to a final concentration of 1 μM. Incubate for 10 min. 9. Centrifuge for 2 min, 400
g, at room temperature. Discard the supernatant. 10. Add 30 μL of FACS buffer containing the primary anti-His antibody at dilution 1/200 and incubate for 15 min at room temperature in the dark. 11. Centrifuge for 2 min, 400
g, at room temperature. Discard the supernatant. 12. Add 30 μL of FACS buffer containing the secondary antimouse fluorescent antibody at dilution 1/400 and incubate for 15 min at room temperature in the dark. 13. Add 120 μL of FACS buffer and inject on a cytometer (see Note 31). 3.8.3 Fluorescence Microscopy

1. Before the experiment. 1. Pass the cells expressing the protein of interest in an 8-well slide, using 500 μL of solution per well (see Note 32). 2. The following steps assume that the cells express the membrane protein of interest at their surface. In the case of the

156

Ghe´rici Hassaı¨ne et al.

Fig. 2 Example of cell surface staining observed by microscopy (see Note 35). (a) Scheme of the experiment. (b) and (c) Two types of staining of cells. In (b), cells expressing the 5-HT3 receptor were imaged with a confocal microscope using NTA-Atto 647 targeting the His-tag of VHHs present in the wells. The negative control was done on noninduced cells incubated with a mixture of 500 nM of each of the VHHs. In (c), cells expressing a chimeric α7/5-HT3 receptor were imaged on a wide-field microscope using a specific fluorescent toxin

5-HT3 receptor, the expression was induced with 4 μg/mL tetracycline 2 days before the experiment (see Note 33). 2. Day of the experiment. 3. Remove the cell culture medium and replace by PBS supplemented with 500 nM of VHH. Incubate for 10 min in the cell incubator. 4. Remove the solution containing the VHHs and replace by PBS supplemented with the fluorescent probe, for instance NTA Atto 488 targeting the His-tag of VHHs. Incubate for 5 min in the cell incubator. 5. Observe the fluorescence on a microscope (see Note 34). Examples are provided in Fig. 2. 3.8.4 Radio-Ligand Binding Test

Day 1 1. Pass the cells expressing the protein of interest in a 12-well plate. Place back in the incubator.

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

157

Day 2 2. Induce one well with 4 μg/mL tetracycline. One well will provide enough cells to test about 3–4 VHHs. Scale the experiment according to your requirements. Day 3 3. Wash wells with 1 mL PBS.

4. Detach the cells in the induced well by pipetting in 1 mL PBSEDTA buffer. Detach the cells of one noninduced well for the determination of the nonspecific binding. Process them in parallel. 5. Pellet the cells by centrifugation at 1000
g, 5 min at 4  C. 6. Resuspend the cell pellet in 150 μL PBS. Count the cells and dilute them in PBS to get a final concentration of 1000 cells/μL. 7. Dilute the radio-ligand [3H]-GR65630 to 4 nM in the assay buffer. Prepare 300 μL of this solution per VHH to be tested (see Note 36). 8. Dilute the VHHs to a concentration of 10 μM in the assay buffer. 9. For each VHH, mix in a 1.5 mL tube: 100 μL of VHH, 100 μL of cells, and 700 μL of assay buffer. Prepare these samples in triplicate (see Note 37). 10. Mix in a 1.5 mL tube 800 μL of assay buffer and 100 μL of cells from the induced well. Prepare this positive control (measure of total binding) in triplicate. 11. Mix in a 1.5 mL tube 800 μL of assay buffer and 100 μL of cells from the noninduced well. Prepare this negative control (measure of nonspecific binding) in triplicate (see Note 38). 12. Incubate the tubes at room temperature for 15 min. 13. Add 100 μL of the radio-ligand solution to each tube and incubate at room temperature for 45 min. 14. Simultaneously, incubate the fiber filters in the polyethyleneimine solution for 15 min. Place the filters on the vacuumdriven filtration system. The filters will retain the protein and its bound radio-ligand while the unbound radio-ligand will be washed away. 15. Gently pipet 0.8 mL of a sample on a filter. Generally, the vacuum box accommodates 12 filters simultaneously. 16. Wash the filters three times with 3 mL of ice-cold assay buffer. 17. Transfer the filters to scintillation vials, filled with 4 mL of Ultima Gold scintillation solution. 18. Incubate the vials for 1 h (or overnight in the dark).

158

Ghe´rici Hassaı¨ne et al.

19. Count the radioactivity using a liquid scintillation analyzer. 20. Compare the amounts of radio-ligand bound to each receptorVHH complex. Subtract the nonspecific binding obtained with noninduced cells. 3.8.5 Co-Elution on Size-Exclusion Chromatography

Co-elution on size-exclusion chromatography is another rapid and efficient method for preliminary screening of interacting proteins. We have used it for the 5-HT3 receptor and all VHHs that came out of the phage-display selection. Examples of profiles are depicted in Fig. 3. The experiment can be easily automated if the HPLC machine is equipped with an automatic sample changer.

Fig. 3 Example of co-elution on size-exclusion chromatography. In panels (a) and (c), examples of SEC profiles are shown respectively for the homomeric 5-HT3A receptor and the heteromeric 5-HT3(AB) receptor mixed with VHHs. The fractions corresponding to the VHH/receptor complex collected for analysis are indicated with a black bar. In panels (b) and (d), a coomassie-stained gel and a Western blot are shown demonstrating the presence of VHHs complexed with the receptor. Note that the Western blot gives an indication that VHH9 is not binding as well as the two others

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

159

1. Mix 100 μg of pure 5-HT3 receptor at a concentration of 1 mg/mL with an excess of VHH. Complete to 150 μL with buffer and incubate for 10 min (see Note 39). 2. Inject the mixture on an equilibrated Superose 6 10/300 column, running at 0.5 mL/min (see Note 40). 3. Collect the pentameric fraction and the fractions corresponding to small molecular masses (see Note 41). 4. On a Coomassie-stained gel of the fractions of interest, the band of the VHH will appear in the pentameric fraction if there is formation of a complex. To minimize the sample consumption, you can choose to reveal the gels by Western blot against the VHH tag. In this latter case, it is enough to inject 10 μg of the 5-HT3 receptor. 3.8.6 Native Gels

Native gels can qualitatively characterize the binding and the binding stoichiometry of two interacting partners. Examples are shown in Fig. 4. The following protocol follows the guidelines of the manufacturer’s manual (LifeTec Manual MAN0000557). 1. Cast 6% polyacrylamide gels. 1. Perform the following steps at 4  C. 2. Mix 10 μg of pure 5-HT3 receptor with a slight excess of VHH (typically 2 VHH per receptor subunit). Incubate for 10 min. 3. Mix with sample buffer, which is provided as a 4
stock solution. Add 1 μL of 0.4% Coomassie G-250 per 10 μL of mix. 4. Prepare the electrophoresis system. Use anode and cathode dark blue buffers prepared according to the manufacturer’s instructions. Load the wells before filling up the upper compartment with buffer, for easier visualization of deposited samples. Fill empty wells with buffer.

Fig. 4 Example of 5-HT3 receptor in complex with VHHs run on a native gel. The left lane shows the electrophoretic migration of the 5-HT3 receptor alone, while the following ones are migration patterns of the 5-HT3 receptor in complex with different VHHs. It is interesting to note that intermediate migrations are observed in between the receptor alone and the one in complex with VHH15 (that binds with a 5:1 ratio, as seen in the crystal structure)

160

Ghe´rici Hassaı¨ne et al.

5. Run the electrophoresis at 150 V for 120 min (see Note 42). 2. Perform the following steps at room temperature. 6. Put the gel in 100 mL of a fixing solution containing 40% methanol and 10% acetic acid. Microwave for 45 s. 7. Incubate on a shaker for 20 min and remove the fixing solution. 8. Put the gel in 100 mL of destaining solution containing 8% of acetic acid. Microwave for 45 s. 9. Incubate on a shaker until the background is transparent (see Note 43). 3.8.7 Electrophysiology

Qualifying the eventual functional effect/role of VHHs on the 5HT3 receptor was an additional criterion to select the most promising crystallization chaperone candidates. It is possible to use a classical two-electrode voltage clamp setup with a usual perfusion system. However, this requires a lot of purified VHH, enough to fill up 20 mL syringe(s) with a solution at the desired concentration(s) (e.g., working at 1 μM, considering a 15 kDa VHH, requires 250 μg). We have used a Hi-Clamp automate (Multi Channel Systems) that uses only 200 μL per condition (e.g., 2.5 μg in the aforementioned conditions). Moreover, once a correct protocol was devised, the automate has permitted a quick and consistent screen of 6 VHHs within a couple of days. It is beyond the scope of this chapter to describe in detail the protocols for Xenopus oocyte preparation and injection (but see Chapter 15 by Vivaudou et al., in this book) and we will focus on the specific protocol for VHH screening. Two days before the experiment, inject 2 ng of purified mRNA encoding for the m5HT3A receptor into each oocyte (about 25 oocytes per VHH gives a good safety margin). The following steps are then performed by a programmed routine on the automate, sequentially measuring oocytes until n ¼ 5 is reached (Fig. 5). l

Test the viability of the impaled oocyte and the 5-HT3 receptor expression by applying a pulse of 2.5 μM serotonin for 10 s. If the recorded current peaks are above 0.8 μA, select the oocyte for further analysis, otherwise discard (see Note 44).

l

Wash in a flow of buffer for 90 s, and then incubate for 3 min in a well containing a stirred buffer solution supplemented with the desired VHH concentration (typically 1 μM).

l

Immediately transfer in a well containing 5 μM serotonin for 10 s. The recorded current is compared to control without

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

161

Fig. 5 Scheme of the automated functional assay for VHHs. (a) One concentration and (b). Dose-response schemes showing the sequential transfers of the impaled oocyte in flowing buffer (blue) or in 200 μL wells containing serotonin (yellow) or the VHH of interest (pink). The right panels depict examples of data obtained by the two protocols with VHH15, which present a long-lasting inhibition (top) and an IC50 of 29 nM (bottom)

VHH and gives the maximal inhibition or positive modulation induced by VHH. l

Rinse in a flow of buffer for 5 min, and transfer in a well containing 5 μM serotonin for 10 s. The recorded current is a post-wash control pulse and was in most cases comparable to the initial current recorded before incubation with the VHH (a small decrease due to desensitization is often observed). If the VHH has a functional effect and a slow Koff, this current can be comparable to the current recorded just after VHH incubation. This was the case for VHH15.

If a more detailed characterization of the VHH is needed, an essentially similar protocol is used to record a dose-response curve (Fig. 5b). We have recorded the dose-inhibition curve of VHH15 and determined an IC50 of 29 nM by applying sequential serotonin pulses for 10 s after 2.5 min of incubation with the VHH, separated by 3 min of wash in a flow of buffer. Serotonin concentrations ranged from 1 to 100 nM (see Note 45). 3.8.8 Crystallization

The crystallization trials were set up using classical protocols and initial robotic screening followed by robotic and manual optimization. We benefited from the services of the HTXlab (https://embl. fr/htxlab/) crystallization facility that perform screens, proposed a

162

Ghe´rici Hassaı¨ne et al.

range of plates adapted for membrane proteins, and a web interface for crystal inspection and ranking. 1. Robotic screening. 1. Mix the pure 5-HT3 receptor at 3–7 mg/mL with a small molar excess of VHH, between 1.2 and 2 VHH per receptor subunit. 2. Add 1 CMC of Cymal-6 from a 10 CMC stock solution (see Note 46). 3. With a crystallization robot, prepare the crystallization screen using the vapor diffusion method (see Note 47). 2. Manual optimization. 4. Prepare a mix of pure 5-HT3 receptor with a small molar excess of VHH and supplement with 1 CMC of Cymal-6. Incubate for 10 min (see Note 48). 5. Set up hanging drops using siliconized covers slips and 24well Limbro screening plates. 6. The best conditions in which crystals of the 5-HT3 receptor in complex with VHH15 were obtained and diffracted are: Storage: 12  C or 20  C. Reservoir solution: 20–25% PEG 10000, 0.1 M Na2SO4 and 0.1 M Tris pH 8.5. Initial 5-HT3 receptor concentration: 2.5–10 mg/mL. Cryo-protection: brief soaking in a reservoir solution supplemented with 20% glycerol (see Note 49).

4

Notes 1. In addition to mammalian systems [18, 19], the expression and purification of the mammalian 5-HT3 receptor was also reported using baculovirus-infected insect cells [20] and even bacteria [21]. 2. We had trouble adapting the T-REx-CHO cells to suspension culture (Subheading 3.4). They are also more adherent, which lengthens the final collection of cells after roller bottles culture (step 17 in Subheading 3.2). Take these constraints in consideration when choosing the cell line. 3. The gene coding for mouse 5-HT3A receptor was modified to insert a double twin Strep-tag and a TEV cleavage site between the native signal sequence of the protein and its mature sequence. It was cloned into the pcDNA5/TO vector that contains a CMV promoter and the Tet operator that allow

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

163

strong repression in T-REx cell lines until the inducer molecule tetracycline is added. 4. Proceed during the following days with the experiment to quantify the level of expression of the different clones. A radio-ligand binding assay, such as the one described in paragraph Subheading 3.8.4 in the context of VHH selection, is a good choice. It is what was used to select the T-REx-293 cell line expressing up to 107 5-HT3 receptors per cell [1]. If working with radioactivity is not possible, flow cytometry (Subheading 3.8.2) or fluorescent SEC [22] or even Western blots are alternatives. Once the best clone is identified, freeze some cells for storage and proceed with large-scale cultures. 5. It is possible to use medium without antibiotics from this step on. We never encountered problems of a cell line losing its ability to express the protein of interest when the period without antibiotics is not too long. 6. It is possible to work in a temperature-controlled room without CO2 control at this stage. Close tightly the roller bottle lids, and unscrew them by 1/4 of turn after 2 days. 7. An alternative to valproic acid is sodium butyrate at 5 mM, which has the similar inhibitory effect of histone deacetylation. However, the smell of sodium butyrate is very unpleasant and your colleagues will be grateful for the choice of valproic acid. 8. Do not use solution that contains trypsin to detach the cells at this step, as the protein of interest is present at the cell surface and can be damaged. 9. Expect the cells to grow slowly in the beginning, and to present a high mortality (about 20–30%). They will also form clumps until the serum is removed. 10. It is also possible to be more abrupt and remove the serum in two steps (use 5% at step 5 and 2% at step 8) or even remove it completely from the start, but some cell lines will show very high mortality rate and take a long time to adapt when the adaptation protocol is harsh. 11. Square-bottom bottles can be used to reach higher densities of cells [23] and thus increase production for the same volume of medium (the medium being usually the cost limiting factor). 12. The harvesting time, as well as the temperature of incubation after induction can be optimized during a pilot experiment. Incubation at 30  C can increase the expression of well-folded protein. 13. The low-speed centrifugation steps can be skipped if in rush. Solubilization is probably a little less efficient when a fraction of unbroken cells and debris are present, but we have not noticed any measurable impact on the final purified material.

164

Ghe´rici Hassaı¨ne et al.

Note also that the membrane preparation protocol collects plasma and internal membranes, and thus all receptors present in the ER (properly folder or not). 14. We commonly use a Beckman Coulter Ti 45 rotor. It can contain six tubes of ~60 mL. Hence, the maximal volume of a membrane preparation is ~360 mL, which corresponds to ~30 g of cells. 15. Pipetting up and down is a slow method for membrane resuspension. It is also possible to vortex, use a potter or the mechanical homogenizer. The pellet can also be stored directly at 80  C. As it is not recommended to freeze ultracentrifuge tubes, one drawback is that transfer of the greasy pellet is not convenient. On the contrary, one advantage compared to membrane suspensions is the much faster thawing. 16. The rotor should be cold, and tubes shall be full and carefully equilibrated. A Beckman Coulter Ti 45 rotor filled with six full tubes contains about ~360 mL, which correspond to ~14 g of starting membranes. Keep this limit in mind when preparing the scale of the purification. 17. If you are using prepacked Streptactin columns, filtrate on 0.45 μm filters. This step can be really cumbersome and long (up to 2 h) when the membranes are derived from HEK cells. We recommend using gravity-flow resin, which does not get clogged, to avoid filtration. Pack the resin in disposable chromatography columns. 18. d-Desthobiotin is long to solubilize. Prepare it 1 h in advance under agitation. 19. Or add as much as the lab can afford. The commercial PNGase F is very expensive and we were frequently decreasing the amounts to 20 units/μg. A PNGase production and purification protocol is provided in paragraph, Subheading 3.7, we use 0.1 mg of the home-made enzyme per mg of 5-HT3 receptor. 20. Trypsin activity is heavily dependent on freeze/thaw cycle, prepare aliquots of a stock solution at 5 mg/mL and do not reuse aliquots. 21. Critical: do not over-treat with too much trypsin or too long incubation time! The protein would not survive and be chopped into small pieces. The incubation time and trypsin amount have to be optimized in a pilot experiment that typically also includes other proteases. 22. Proceed immediately with the size-exclusion chromatography, which will separate the receptor from the trypsin. It is not possible to stop at this point.

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

165

23. The volume depends on the amount of protein. It is not advised to concentrate the 5-HT3 receptor further than 5 mg/mL. Perform several injections on the SEC column if the protein is “too” abundant, but this requires proper timing of the limited proteolysis. 24. Size-exclusion chromatography is not a method to evaluate the mass of membrane proteins. However as a rule of thumb, expect your Cys-loop receptor of interest to elute with an apparent molecular mass twice its molecular mass. In other words, 1 gram of bound detergent per gram of protein is a good guess. Here for the 5-HT3 receptor, (5
50)
2 equals 500 kDa, while the elution peak corresponds to an apparent molecular mass of 540 kDa according to the calibration curve of the column. 25. The Streptactin-coated 96-well plates are made of 12 strips of 8 wells that can be used independently. The protocol is written using only one well of one strip, but can be parallelized if the same library is screened against several targets. 26. The wash solutions can be removed by pipetting or by turning the plate upside down over the sink and then over an absorbing paper to remove the last drops. This is faster when numerous washes are necessary (as in step 9). 27. An alternative solution for phage elution is to use ddesthobiotin that will detach the receptor from the surface. In principle, this should yield a more specific selection. 28. This plate is for storage or rescue if the liquid culture does not grow. It can be used to prepare a glycerol stock. 29. The phage concentration should be >1011 phages per mL. 30. Include a negative control sample (e.g., noninduced cells). The negative control will also allow you to make sure the nonspecific binding of the fluorescent probe is low enough so that no washing steps are necessary. If this binding is significant, insert washing step(s) after the incubation with the dye. 31. This volume is indicative, it corresponds to three 50 μL sample injections of a solution containing >106 cells per mL. 32. The slide can be coated with 0.2 mg/mL poly-D-lysine during 1 h for better adherence of HEK cells. It is also recommended to fill completely the wells to minimize border effects where cells accumulate on the periphery of wells. 33. For each VHH, use a well with induced cells and a well with noninduced cells for the negative control. Typically, we induce only the top line of the 8-well slide and use it to screen four different VHHs.

166

Ghe´rici Hassaı¨ne et al.

34. This protocol with no fixation of the cells and no washes is suited for a first quick qualitative characterization of the VHH binding, or of the expression of a construct. For longer observation, it is necessary to fix the cells. 35. The panels b. and c. were chosen to illustrate the diversity of surface staining (poor or good) that can be observed. 36. This radio-ligand solution will be diluted 10
in the final sample; therefore, the concentration of the radio-ligand should be ~10 KD at this step. 37. If testing the expression level of a particular construct rather than evaluating VHH, just replace the VHHs containing solution by assay buffer. 38. If testing constructs rather than VHHs, the measure of the nonspecific binding is done adding an excess of cold ligand that will compete with the radio-ligand. In this case prepare in triplicate a sample with the cells expressing each construct, as in step 9, and add quipazine to a final concentration of 1 μM. 39. We generally used fivefold molar excess per receptor subunit, which for a 15 kDa VHH is ~33 μL of pure VHH at 5 mg/mL. 40. The apo- receptor elutes around 13.6 mL and the peak is essentially unshifted when the receptor is mixed with VHHs (one -or even five- VHH(s) is (are) small compared to the protein-detergent complex). 41. A VHH alone elutes around 19.5 mL on that column. 42. When the blue color from the cathode buffer starts to come out of the gel in the transparent anode buffer, the electrophoresis is done. 43. Bands start to appear immediately after the microwave heating. 44. The m5HT3A receptor expresses robustly at the surface of oocytes. The threshold current needs to be adapted on a case-by-case basis. 45. During the protocol setup, it is mandatory to check that the wash time is long enough to completely recover from desensitization, at the maximal concentration used. 46. In screening conditions using 100 + 100 nL drops, a 96-well plate screen requires about 15 μL of protein solution. Typically, we are thus mixing 9 μL of receptor at 5 mg/mL with 6 μL of VHH at 5 mg/mL and 1.7 μL of Cymal-6 stock solution at 10 CMC. Cymal-6 was identified as an additive improving the apo 5HT3 receptor crystals long before we got crystals of the complex VHH15/5-HT3 receptor. We infer that it plays a role on the size of the detergent belt that surrounds the 5-HT3 receptor and tends to favor protein/protein contacts.

Expression, Biochemistry, and Stabilization with Camel Antibodies of Membrane. . .

167

47. Because of the low surface tension of the protein + reservoir drop, due to the presence of detergents, drops can spread on their support and form a film not suitable for crystallization. We are mostly using sitting drops on Intelli-Plate (Hampton Research) where the drops do not spread too much. 48. We have not noticed any improvement in the diffraction quality of the crystals when using a VHH-receptor complex purified by size-exclusion chromatography. We typically use a twofold molar excess of VHH over receptor subunit, but have not noticed clear effect of varying this parameter (down to 1.1-fold and up to 5-fold excess). 49. We noted a correlation between crystal size and diffraction quality. Crystals in this condition tend to be long sticks, and the bigger they were, the best diffraction was. Also, longer crystals permitted us to do helical data collection to minimize the radiation damage, which in turn resulted in an improved resolution limit.

Acknowledgments The work was supported by a grant from the European Research Council (ERC-2014-StG PentaBrain), by the Swiss National Science Foundation, by the Ecole Polytechnique Fe´de´rale de Lausanne, and by the CEA. References 1. Hassaı¨ne G, Deluz C, Li X-D, Graff A, Vogel H, Nury H (2013) Large scale expression and purification of the mouse 5-HT3 receptor. Biochim Biophys Acta Biomembr 1828 (11):2544–2552 2. Hassaı¨ne G, Deluz C, Grasso L, Wyss R, Tol MB et al (2014) X-ray structure of the mouse serotonin 5-HT3 receptor. Nature 512 (7514):276–281 3. Brejc K, van Dijk WJ, Klaassen RV, Schuurmans M, van Der Oost J et al (2001) Crystal structure of an ACh-binding protein reveals the ligand-binding domain of nicotinic receptors. Nature 411(6835):269–276 4. Nys M, Kesters D, Ulens C (2013) Structural insights into Cys-loop receptor function and ligand recognition. Biochem Pharmacol 86 (8):1042–1053 5. Unwin N (2005) Refined structure of the nicotinic acetylcholine receptor at 4A resolution. J Mol Biol 346(4):967–989 6. Hilf RJC, Dutzler R (2008) X-ray structure of a prokaryotic pentameric ligand-gated ion channel. Nature 452(7185):375–379

7. Bocquet N, Nury H, Baaden M, Le Poupon C, Changeux J-P et al (2009) X-ray structure of a pentameric ligand-gated ion channel in an apparently open conformation. Nature 457 (7225):111–114 8. Hibbs RE, Gouaux E (2011) Principles of activation and permeation in an anion-selective Cys-loop receptor. Nature 474(7349):54–60 9. Miller PS, Aricescu AR (2014) Crystal structure of a human GABAA receptor. Nature 512 (7514):270–275 10. Huang X, Chen H, Michelsen K, Schneider S, Shaffer PL (2015) Crystal structure of human glycine receptor-α3 bound to antagonist strychnine. Nature 526:277–280 11. Du J, L€ u W, Wu S, Cheng Y, Gouaux E (2015) Glycine receptor mechanism elucidated by electron cryo-microscopy. Nature 526:224–229 12. Lochner M, Thompson AJ (2015) A review of fluorescent ligands for studying 5-HT3 receptors. Neuropharmacology 98:31–40 13. Seiradake E, Zhao Y, Lu W, Aricescu AR, Jones EY (2015) Production of cell surface and

168

Ghe´rici Hassaı¨ne et al.

secreted glycoproteins in mammalian cells. Methods Mol Biol 1261:115–127 14. Goehring A, Lee C-H, Wang KH, Michel JC, Claxton DP et al (2014) Screening and largescale expression of membrane proteins in mammalian cells for structural studies. Nat Protoc 9 (11):2574–2585 15. Yao F, Svensjo¨ T, Winkler T, Lu M, Eriksson C, Eriksson E (1998) Tetracycline repressor, tetR, rather than the tetR-mammalian cell transcription factor fusion derivatives, regulates inducible gene expression in mammalian cells. Hum Gene Ther 9(13):1939–1950 16. Loo T, Patchett ML, Norris GE, Lott JS (2002) Using secretion to solve a solubility problem: high-yield expression in Escherichia coli and purification of the bacterial Glycoamidase PNGase F. Protein Expr Purif 24 (1):90–98 17. Pardon E, Laeremans T, Triest S, Rasmussen SGF, Wohlko¨nig A et al (2014) A general protocol for the generation of Nanobodies for structural biology. Nat Protoc 9(3):674–693 18. Dostalova Z, Liu A, Zhou X, Farmer SL, Krenzel ES et al (2010) High-level expression and purification of Cys-loop ligand-gated ion

channels in a tetracycline-inducible stable mammalian cell line: GABAA and serotonin receptors. Protein Sci 19(9):1728–1738 19. Wu Z-S, Cui Z-C, Cheng H, Fan C, Melcher K et al (2015) High yield and efficient expression and purification of the human 5-HT3A receptor. Acta Pharmacol Sin 36:1024–1032 20. Green T, Stauffer KA, Lummis SCR (1995) Expression of recombinant homo-oligomeric 5-Hydroxytryptamine(3) receptors provides new insights into their maturation and structure. J Biol Chem 270(11):6056–6061 21. Na J-H, Shin J, Jung Y, Lim D, Shin Y-K, Yu YG (2013) Bacterially expressed human serotonin receptor 3A is functionally reconstituted in proteoliposomes. Protein Expr Purif 88 (2):190–195 22. Hattori M, Hibbs RE, Gouaux E (2012) A fluorescence-detection size-exclusion chromatography-based thermostability assay for membrane protein precrystallization screening. Structure 20(8):1293–1299 23. Muller N, Girard P, Hacker DL, Jordan M, Wurm FM (2005) Orbital shaker technology for the cultivation of mammalian cells in suspension. Biotechnol Bioeng 89(4):400–406

Chapter 9 Characterization of New Detergents and Detergent Mimetics by Scattering Techniques for Membrane Protein Crystallization Franc¸oise Bonnete´ and Patrick J. Loll Abstract Membrane proteins are difficult to manipulate and stabilize once they have been removed from their native membranes. However, despite these difficulties, successes in membrane-protein structure determination have continued to accumulate for over two decades, thanks to advances in chemistry and technology. Many of these advances have resulted from efforts focused on protein engineering, high-throughput expression, and development of detergent screens, all with the aim of enhancing protein stability for biochemistry and biophysical studies. In contrast, considerably less work has been done to decipher the basic mechanisms that underlie the structure of protein-detergent complexes and to describe the influence of detergent structure on stabilization and crystallization. These questions can be addressed using scattering techniques (employing light, X-rays, and/or neutrons), which are suitable to describe the structure and conformation of macromolecules in solution, as well as to assess weak interactions between particles, both of which are clearly germane to crystallization. These techniques can be used either in batch modes or coupled to sizeexclusion chromatography, and offer the potential to describe the conformation of a detergent-solubilized membrane protein and to quantify and model detergent bound to the protein in order to optimize crystal packing. We will describe relevant techniques and present examples of scattering experiments, which allow one to explore interactions between micelles and between membrane protein complexes, and relate these interactions to membrane protein crystallization. Key words Membrane proteins, Detergents, Micelle behavior, Interactions, Radiation scattering, Phase diagrams

1

Introduction Despite the widely acknowledged importance of membrane proteins, we know relatively little about their structures. This knowledge gap handicaps many aspects of biomedical science, for example, structure-based drug design. The problem can be traced to an essential feature of membrane proteins: they cannot “survive” outside an amphiphilic environment mimicking the native lipid bilayer that is their natural cellular environment. Thus, to crystallize

Jean-Jacques Lacapere (ed.), Membrane Protein Structure and Function Characterization: Methods and Protocols, Methods in Molecular Biology, vol. 1635, DOI 10.1007/978-1-4939-7151-0_9, © Springer Science+Business Media LLC 2017

169

170

Franc¸oise Bonnete´ and Patrick J. Loll

membrane proteins for structural studies, one must first extract them from the cell membrane by using a detergent (i.e., an amphiphilic molecule able to dissolve the lipid membrane) and then either keep the membrane proteins in the detergent environment (i.e., crystallization in surfo phase) [1], reintroduce them in a bilayer environment that mimics the proteins’ natural lipid environment (such as lipidic cubic or sponge phase, i.e., in meso crystallization) [2, 3], or transfer them to mixed bilayer-based discs (i.e., bicelles or nanodiscs) [4, 5]. Increasing successes are being obtained using bicelles and in meso crystallization, but crystallization directly from detergent solutions remains the most frequently used method. Although fundamental approaches have been used to understand the physics behind crystal growth and hence rationalize the process [6–9], crystallization is still largely treated as an empirical problem, and is tackled with trial-and-error searches throughout a large and complex parameter space of physico-chemical variables. The inclusion of detergent expands the parameter space that must be searched, contributing to the lower success rate of structure determination for membrane proteins as compared to soluble proteins. Membrane protein solutions are composed of protein-detergent complexes, detergent micelles, and free detergent monomers, and an improved understanding of the physico-chemistry of these complex multicomponent solutions could substantially increase the success rate for membrane protein crystallization. Many different detergents are available to the membrane protein biochemist (Table 1) [10, 12], but all share a common architecture that combines a polar head, which can be ionic, nonionic, or zwitterionic, grafted onto a hydrophobic tail, which can contain linear or branched alkanes, aromatic hydrocarbons, steroids, or fluorinated moieties (Fig. 1). An ideal detergent should solubilize the target membrane protein in a non-aggregated state, maintain its structural integrity and stability, and make its crystallization possible. Unfortunately, no generally useful ideal detergent exists, and the choice of a detergent frequently requires compromises [13]. For example, in order to stabilize a membrane protein, a detergent must form micelles large enough to cover the hydrophobic transmembrane domain; however, a large micelle may also cover part of the polar portion of the protein, thereby blocking protein–protein interactions required for crystal lattice formation. On the other hand, small-micelle detergents, while leaving more of the protein molecule exposed to form lattice contacts, may also destabilize the protein, by intrusion of the detergent’s alkyl chain into the interior of the protein and/or by stripping away stabilizing lipids, cofactors, or subunits. It is frequently not clear a priori what micelle size is appropriate for a given protein. The problem of detergent choice is amplified by a proliferation of new detergents [14–19]; many new molecules have found uses in biochemical manipulations of membrane proteins, but it is less clear which, if any of them will prove

Linear PEG [8]

Linear PEG [4]

Linear PEG

Linear PEG

Polysorbate

Polysorbate

β-Glycosidic maltose

β-Glycosidic glucose

β-Glycosidic glucose

C8E4

C8POE

Triton X-100

Tween-20

Tween-80

β-Dodecylmaltoside

β-Octylglucoside

β-Octylthioglucoside

Gluconamidopropyl moieties (2
) Complex polysaccharide Linear PEG [11]

Headgroup

C12E8

Digitonin Brij-35 (C12E23)

Nonionic detergents Big CHAP

Detergent

Linear fatty acid, unsaturated (C18:1) Linear hydrocarbon alcohol (C12) Linear hydrocarbon alcohol (C8) Linear hydrocarbon thiol (C8)

Cholesterol derivative Linear hydrocarbon alcohol (C12) Linear hydrocarbon alcohol (C12) Linear hydrocarbon alcohol (C8) Linear hydrocarbon alcohol (C8) p-(2,2,4,4Tetramethylbutyl) phenol Linear fatty acid (C12)

Cholesterol derivative

Tail

9

20–25

0.15

0.012

0.059

6.6

6.5–8.5

0.11

– 0.09

2.9–3.4

CMC (mM) (at 25 oC)

–

84

98

58

–

100–150

–

–

123

60–70 40

10

Aggregation number

308.4

292.4

1230 (average) – 1310 average) 510.6

330 (average) 630 (average)

306.4

538.8

1229.3 1199.6

878.1

MW (Da)

Table 1 Typical detergents used with membrane proteins Compilation extracted from (Linke [10]) (Courtesy of D Linke)

–

25

70

76

80

–

26

66

– 48

9

Micellar weight (kDa)

7

λ/10), and also reports on weak interactions between particles. DLS focuses on the rapid fluctuations (~msec) in the intensity of scattered light that arise from the Brownian motion of the particles in solution; these fluctuations reflect how rapidly the particles are diffusing through the solution, and analysis of these fluctuations yields the distribution of particle sizes. Samples for SAXS, SLS, and DLS can be prepared in any solvents, at any temperature, and can be analyzed in a single measurement, as a function of concentration or as a function of time. The sample volumes are small (50 μL), with protein or detergent concentrations in the mg/mL range. The concentration must be measured with accuracy, since the forward scattering intensity, which makes the measurement of molar mass possible, is a sensitive function of concentration, being the sum of contributions of all components in solution as shown below: X ci Mi ∂ρ 2 e Ið0Þ ¼ ð1Þ N ∂c a i i e with Mi, ci, and ∂ρ being the molar mass, the concentration, and ∂ci the electron scattering density increment for each component, respectively.

For SAXS and SLS, the solvent without macromolecules is used for background subtraction; this can be obtained, for example, from the dialysate after a dialysis experiment. In DLS, no background subtraction occurs, but the solvent viscosity must be known accurately, since it strongly affects particle diffusion. 3.2 Light Scattering Approaches

In a typical light scattering experiment, a monochromatic laser beam (400–700 nm) illuminates a solution of particles and the intensity I(ϑ, t) of the scattered light is measured at a scattering angle θ by a photon detector. In our case studies, the particles, which are detergent micelles or protein-detergent complexes, are smaller (λ/10, the scattered intensity would depend on the scattering angle through the particle form factor P (ϑ):

X-Ray and Light Scatterings for Detergent Micelle Behavior Analysis

K :c ¼ Rθ

179



 1 1 þ 2:A 2 :c M PðθÞ

ð2Þ

with 1/(P(θ)) ¼ 1 þ (16π 2)/(3(2)RG2sin2θ and K ¼ (4π 2)/ ((4Na)(n0 ∂n/∂c)2. Rθ is the Rayleigh ratio of scattered light to incident light, P(θ) the angular dependence of sample scattering, c the particle concentration in g/mL, M the particle molar mass, A2 (in mol.mL/g2) the second virial coefficient characterizing the weak interactions between particles, K an optical constant, λ the laser wavelength, Na the Avogadro number, n0 the solvent refractive index, and ∂n ∂c the particle refractive index increment. In order to account for incident intensity, the Rayleigh ratio is measured from a reference, such as toluene [41], whose Rayleigh ratio and refractive index are known:   I sample  I solvent n20 Rθ ¼ Rtol ð3Þ I tol n2tol A linear Debye plot representing Kc/Rϑ ¼ f(c) can then be used at a constant angle to determine the particle molar mass M, from the intercept at zero concentration, and the second virial coefficient A2, from the slope (A2 contains information about the inter-particle interactions; vide infra). In this case, both solvent refractive index and particle refractive index increment have to be characterized. The concentration c is expressed either as the membrane protein complex concentration or as cdet—CMC for free-protein detergent micelle analysis. When the light is scattered by particles in solution, small intensity shifts are observed, due to particle motion. The rate at which the intensity fluctuates depends on the size of the particles; small particles give rise to more rapid fluctuations than large ones. The best way to characterize particles in motion is to calculate the intensity autocorrelation function: G ðτÞ ¼

hI ðt ÞI ðt þ τÞi hI ðt Þi2

ð4Þ

There are two different approaches to analyze this autocorrelation function, either by a single exponential fit (Cumulant analysis) [42] or by a sum of mono-exponentials (regularization methods such as Contin or NNLS) [43], depending on the polydispersity index (PdI) of the solution (i.e., the standard deviation of the mono-exponential). In the case of a solution containing monodisperse (PdI < 20%) noninteracting particles, the monomodal autocorrelation function can be written as an exponential function: G ðτÞ ¼ e q

2

Dτ

ð5Þ

180

Franc¸oise Bonnete´ and Patrick J. Loll

D being the particle diffusion coefficient (m2/s), q ¼ 4πn0sinϑ/λ the scattering vector, and τ the correlation time. The particles’ hydrodynamic radius can be obtained from D through the Stokes–Einstein relation: D¼

kB T 6πηRH

ð6Þ

with kB the Boltzmann’s constant, T the absolute temperature, and η the viscosity of the solvent. In the case of polydisperse solutions (PdI > 20%) of noninteracting particles, the autocorrelation function can be written as the sum of mono-exponentials, Ii being the normalized contribution to scattering from particles with diffusion coefficients Di. X 2 I i e Di q τ ð7Þ G ðτÞ ¼ i

In the case of concentrated solutions, two additional factors may affect the particle diffusion, namely viscosity and interactions between particles. When plotting the diffusion coefficient D versus the particle concentration c, D ¼ D0(1 þ kDc), we determine the diffusion coefficient at infinite dilution D0 and kD (in mL/g) the diffusion interaction parameter, which is related to the second virial coefficient A2 and the sedimentation interaction parameter [44] obtained from sedimentation velocity: kD ¼ 2MA2  ks 
v

ð8Þ

M is the molar mass and
v the partial specific volume of the particle. Interaction parameters determined either by SLS or DLS in semi-dilute concentration regimes are of great interest in various applications to explain aggregation, solubility, crystallization, and phase separation [45] and to predict thermodynamic behavior of particles in solution. If kD or A2 are positive, the weak interactions between particles in solution are repulsive, favoring solubility. Conversely, negative values of kD or A2 indicate attractive inter-particle interactions, favoring crystallization or amorphous precipitation. 3.2.1 SLS Analysis of Protein-Detergent Complexes

A critical advance in our understanding of protein crystallization was made in the 1990s, when Wilson and colleagues demonstrated that the second virial coefficient (A2) could be used to predict crystallization behavior [8]. The second virial coefficient is a thermodynamic parameter reporting on the interactions between particles in solution, and as such represents the degree to which a solution deviates from ideal behavior. The key insight provided by Wilson et al. was that, for solution conditions that give rise to crystals, virial coefficient values tend to cluster within a narrow range, about 1 to 8
104 mol.mL/g2. This range, which

X-Ray and Light Scatterings for Detergent Micelle Behavior Analysis

181

has been referred to as the “crystallization slot,” corresponds to weakly attractive interactions between particles. Solution conditions promoting strongly attractive inter-particle interactions give rise to amorphous precipitate, whereas conditions promoting repulsive interactions give clear solutions. This result held out hope that screening of solution conditions for appropriate second virial coefficient values might represent a rational approach to the crystallization problem [46]. The initial work mapping the crystallization slot was limited to soluble proteins, for which second virial coefficient measurements report only on protein-protein interactions. In the case of membrane proteins solubilized as protein-detergent complexes, the situation is more complicated, since detergent-detergent and protein-detergent interactions are possible as well. To address the question of whether second viral coefficient measurements might prove useful for crystallizing membrane proteins, Hitscherich et al. [27] examined the bacterial membrane protein OmpF using SLS (Fig. 4, red square). Their results showed that conditions promoting crystal growth of OmpF corresponded to slightly negative values for the second virial coefficient, implying weak attractive forces between protein-detergent complexes. Further, the second virial coefficient values fell close to the crystallization slot previously identified by Wilson and colleagues, implying that this approach might indeed be useful for the rational crystallization of membrane proteins.

Fig. 4 Second virial coefficient values for both porin-detergent complexes (squares) and protein-free detergent micelles (circles) at varying PEG concentrations in the tetragonal crystallization buffer (Hitscherich et al. [27])

182

Franc¸oise Bonnete´ and Patrick J. Loll

3.2.2 SLS Analysis of Free Protein Detergent Micelles

An important observation made by Hitscherich et al. was that the second virial coefficient behavior of empty micelles (i.e., detergent only) recapitulated that of protein-detergent complexes [27]. For example, as seen in Fig. 4 (blue dots), addition of a precipitant known to crystallize the protein-detergent complex also induces attractive micelle-micelle interactions in pure detergent solutions, suggesting that these micelle-micelle interactions might contribute strongly to (or even dominate) the interaction potentials between protein-detergent complexes. In this case, knowledge of how various precipitating agents affect a given detergent would allow the experimenter to map which conditions drive micelle-micelle interactions into the crystallization slot, such conditions would very likely represent the most fruitful conditions to screen for crystallization of protein-detergent complexes. Unfortunately, the effort required to map second virial coefficients for pure detergent solutions in a large number of conditions would be prohibitive. However, a shortcut more amenable to high-throughput methods was suggested by the insight that second virial coefficient values for detergent micelles become strongly attractive as one approaches the detergent’s cloud point, implying that the cloud point might prove to be a convenient and readily observed surrogate for the crystallization slot [47]. We can rationalize this approach by viewing the cloud point as the position in the micelle phase diagram where the second virial coefficient drops precipitously; hence, by determining the position of the cloud point and then “stepping back” slightly (e.g., by reducing precipitant concentration), one is likely to encounter slightly negative values for the second virial coefficient—i.e., be in the crystallization slot. This concept fits well with empirical observations that membrane proteins frequently crystallize near the cloud point of the detergent used [48–50]. Recently, it has also been demonstrated that screens sampling regions of the phase diagram lying near the cloud point could generate crystalline leads for over a dozen different membrane proteins [51]. Taken together, these results strongly support the notion that measurement and rational manipulation of interaction potentials between protein-detergent complexes can help to demystify the process of membrane protein crystallization.

3.2.3 DLS Analysis of Surfactant Micelle Interactions

As discussed in the previous section, a better understanding of the influence of physico-chemical parameters on interaction potentials opens the door to predicting crystallization conditions for both soluble and membrane proteins. This insight provides a strong incentive to study the influence of detergent structure on membrane protein interactions and crystallization conditions. As an example of this sort of analysis, we have studied the interactive behavior of two surfactants by DLS and then by SAXS (see the next section). These techniques are complementary to SLS, in that DLS gives information on the quality of solutions

X-Ray and Light Scatterings for Detergent Micelle Behavior Analysis

1.0

PCCM 60 mg/ml PCCM 51 mg/ml PCCM 42 mg/ml PCCM 36 mg/ml PCCM 30 mg/ml PCCM 24 mg/ml PCCM 18 mg/ml PCCM 12 mg/ml

0.9

Correlation function

0.8 0.7 0.6 0.5

b % intensity

a

183

40 30 20 10

0.4 0.3

0 0.1

0.2

1

10

100

1000

Size (r nm)

0.1 0.1

1

10

100

1000

10000

Time (ms)

Fig. 5 Autocorrelation functions (a) and size distribution in % intensity (b) of PCCM in 4% PEG 3350

(polydispersity, particle size distribution) and SAXS gives information about the shape of particles in interactions. The two surfactants under consideration are the commonly used detergent n-dodecyl-β-D-maltoside (C12M) and the newly developed mild surfactant propyl(bi)cyclohexyl α-maltoside (PCCM, Glycon) [19]; they differ in the structures of their hydrophobic groups, with C12M having a flexible aliphatic chain while PCCM contains rigid cyclohexyl groups. PCCM has been shown to stabilize three different integral membrane proteins, and to crystallize the cytochrome b6f complex from Chlamydomonas reinhardtii [19] and the RC-LH1-pufX complex from Rhodobacter blasticus [52]. In order to evaluate the influence of the hydrophobic moiety on membrane protein stabilization and crystallization, micelle structure and behavior for the two surfactants were studied in different conditions, i.e., in water and in crystallizing conditions. An example of DLS analysis for PCCM in 4% PEG 3350 is shown in Fig. 5. The autocorrelation functions measured for PCCM in various concentrations of the crystallizing agent PEG 3350 show a monomodal behavior of micelles (Fig. 5a) with a low polydispersity index (Table 3) and a good signal-to-noise ratio (intercept >0.7). Values for the hydrodynamic radius obtained from cumulant analysis (Z-average) and from CONTIN analysis (peak 1), taking into account the viscosity of the PEG solution [53], appear quite similar. The difference between the two values reflects the fact that micelles are not spherical, as assumed from the Z-average. The calculated size distributions for PCCM micelles in PEG 3350 solutions show an increase in apparent radius (peak 1) as

Franc¸oise Bonnete´ and Patrick J. Loll

184

Table 3 Parameters for PCCM in 4% PEG 3350 from DLS analysis

PdI

PdI Width (r.nm)

Diffusion Coefficient (μ2/s)

Z-Ave{ (nm)

Peak 1a % Int (RH nm)

%Pd Peak 1{

60

0.031

1.675

22.5

6.623

6.945

14.4

51

0.054

1.991

25

5.967

6.353

17.5

42

0.04

1.538

27.9

5.344

5.533

15.2

36

0.059

1.7

30.6

4.88

5.151

14.5

30

0.09

2.071

31

4.81

5.249

15.8

24

0.077

1.819

32.7

4.571

4.97

17.2

18

0.124

2.079

36.2

4.119

4.279

15.9

12

0.19

2.416

38.6

3.867

4.112

15.7

Surfactant Conc. (mg/mL)

Taking into account viscosity η of solvent

a

a

b

60

1 0.8

40 0.6 D/D0

D (m2/s)

50

30

0.4

20

0% PEG3350 2% PEG3350 4% PEG3350 6% PEG3350

0.2

10 0 0

0.02

0.04

0.06

0.08

PCCM (g/mL)

0.1

0.12

0

0

0.02

0.04

0.06

0.08

0.1

0.12

PCCM (g/mL)

Fig. 6 (a) Variations of diffusion coefficient D as a function of surfactant concentration in various % of PEG 3350; (b) Variation of normalized viscosity-corrected diffusion coefficient with PEG 3350 percentage

the concentration of PCCM increases (Fig. 5b). Two mechanisms could explain this behavior: an increase in micelle size or the presence of attractive interactions between micelles. These can be discriminated by plotting the variation of the diffusion coefficient of the surfactant as a function of its concentration in the different percentages of crystallizing agent (Fig. 6a). The normalized, viscosity-corrected diffusion coefficient (Fig. 6b) gives the hydrodynamic radius at infinite dilution of surfactant for each percentage of PEG 3350; the slopes of the lines in this plot reflect changing inter-micelle interactions, which become more and more attractive

X-Ray and Light Scatterings for Detergent Micelle Behavior Analysis

185

Table 4 Structural and interaction parameters obtained from DLS and SAXS for C12M and PCCM in different % PEG 3350 RH % PEG Viscosity D0 kD 3350 20  C PCCM PCCM PCCM (W/V) (cP) (μ2/s) (nm) (mL/g)

RG 2MA2 PCCM PCCM (nm) (mL/g)

RH D0 kD RG MA2 C12M C12M C12M C12M C12M (μ2/s) (nm) (mL/g) (nm) (mL/g)

0

1.003

57.06 3.75

1.58 3.32

1.94 61.91 3.47

0.92 3.21

3.46

2

1.172

49.34 3.71

3.83 3.23

2.90 50.16 3.65

3.74 3.07

0.60

4

1.436

42.43 3.52

7.83 3.21

3.13

8.12

5

1.572

6

1.739

3.14

15.04

36.84 3.35

11.27 3.21

13.84 39.25 3.19

8.4

19.78 39.02 3.17

9.6

as the PEG concentration increases (kD more and more negative in Table 4). We observed that the hydrodynamic radius decreases as the percentage of PEG increases, probably due to dehydration of the micelle as PEG increases. This micelle dehydration favors the increase in attraction between micelles by reducing water-mediated repulsive interactions. Compared to the interactions between C12M micelles, micelles of PCCM present more attractive interactions under the same crystallizing conditions (kD PCCM < kD C12M). This is probably due to the structure of the surfactant and/or to different micelle assemblies. To answer this question, small angle X-ray scattering (SAXS) has been used to characterize structure in solutions of PCCM micelles, i.e., molar mass, aggregation number, shape, geometry of micelles. 3.3 X-Ray Scattering Approach 3.3.1 General Information on SAXS Analysis

SAXS is highly complementary to light scattering (Fig. 3) and can be used to characterize form factors of macromolecules, colloids, and micellar assemblies in ideal or dilute solutions, and structure factors (i.e., mutual interactions) in concentrated solutions. The technique has been well described in numerous books [32, 36, 54, 55]. In a typical small angle X-ray scattering experiment, an X-ray beam (λ ~ 0.1 nm) irradiates a solution of particles and the scattered intensity is measured as a function of the particle concentration c and scattering vector q ¼ 4πsinθ/λ, where θ is one-half the scattering angle. The X-ray scattering intensity, I(c,q), can be written as the product of the form factor, I(0,q) (i.e., the scattered intensity of one particle) multiplied by the structure factor, S(c,q), which depends on the particle distribution, i.e., I(c,q) ¼ I(0,q).S(c,q). The form factor (i.e., the intensity recorded at low concentration in the limit of a meaningful signal) gives information on the size, shape, and oligomeric state of the particle. It can be analyzed at very low angle

186

Franc¸oise Bonnete´ and Patrick J. Loll

in terms of molar mass and radius of gyration of the particle by using the Guinier approximation [36]: I ð0; q Þ ¼ I ð0; 0Þe

q 2 R2 G 3

if R G :q < 1:

ð9Þ

The Guinier extrapolation, i.e., the linear regression of R2

Log I(c,q) ¼ f(q2) in the low q-range yields the slope 3G and the  2 0 ðρ -ρ Þ v . The structure factor, S(c,q), intercept I ð0, 0Þ ¼ cM p p Na obtained from the scattered intensity at high concentrations divided by the form factor can lead to the different pair potentials between particles in solution, such as electrostatic repulsion, van der Waals attraction, or hard sphere repulsion, depletion attraction, by coupling with numerical simulations and is of particular interest for protein crystallization [9, 56], but will not be detailed here. The X-ray structure factor at the q-origin, S(c,0), is related to the osmotic pressure Π of the particle solution. It gives access to the overall weak interactions between particles in solution, represented by the second virial coefficient A2. For solutions of particles in a repulsive (attractive) regime, S(c,0) is smaller (larger) than 1 and is given by Sðc; 0Þ ¼

 1 RT ∂Π M ∂c

ð10Þ

where R is the gas constant, 8.31 J.mol1.K1, T the absolute temperature in K, M the particle molar mass, and Π the osmotic ∏ ¼ M1 þ A 2 c þ terms of order c 2 [57]. pressure described by cRT A2 is positive (negative) for repulsive (attractive) interactions and is expressed in mol.mL/g2. As mentioned previously, conditions where the second virial coefficient of proteins is negative have long been known to favor crystallization [8]. Reported in the structure factor at q ¼ 0, we obtain directly: Sðc; 0 Þ ¼

Iðc; 0 Þ ¼ ð1 þ 2MA 2 c þ . . .Þ1 Ið0; 0 Þ

ð11Þ

If the term 2MA2 is small («1), the expansion in powers of c can be limited to the second virial coefficient and linearized so that S(c, 0) ¼ 1–2MA2c. SAXS experiments provide substantially more detail on micelle structure than SLS and DLS experiments. Beyond the radius of gyration obtained from Guinier approximation, the maximum particle dimensions Dmax and geometrical shapes of particles can be obtained from the pair distribution function (PDF), the inverse Fourier transformation of scattered intensity. Moreover, ab initio modeling or least-squares fitting can give additional structural

X-Ray and Light Scatterings for Detergent Micelle Behavior Analysis

187

information [58] on macromolecules or micelles, although not at atomic scale, as is obtained with single-crystal X-ray diffraction. The surfactant micelle molar mass can be estimated from SAXS data by comparison of absolute forward scattering at infinite dilution of micelles with that from a water reference [59]. The aggregation number Nagg was therefore determined by dividing the micelle molar mass by that of the surfactant monomer using the following expression: Nagg ¼ Na

dΣ

 2

Mmono :ðc  ccmc Þ:Ið0Þwater : r0 :
vp ðρmic  ρ Þ dΩ water Ið0Þmic

ð12Þ with Na the Avogadro number, r0 the classical electron radius (r0 ¼ 0.28179 1012 cm/e-),
vp the surfactant micelle specific volume (cm3/g), ρmic and ρ0, the scattering length density of

dΣ

surfactant micelle and water, respectively, and dΩ the absolute water scattering intensity of water equal to 0.01632 cm1 at 293 K. An example of SAXS experiments is presented (Fig. 7a) for PCCM at different concentrations in 4% PEG 3350. After subtraction of appropriate solvent, each individual SAXS curve has been normalized to concentration to highlight effect of structure factor at very small angle and check that the micelle form factor does not change with concentration at large angles. We observed a clear effect of PEG on structure factor at low q values and no effect on the form factor in the large q-range. The increase in scattered intensity, Iq ! 0(c, q), as the concentration of PCCM increases corresponds

3.3.2 Influence of Surfactant Structure on Micellar Assemblies and Interactions

b

100

10

1

Structure factor domain

Form factor domain

40 mg/ml 20 mg/ml 10 mg/ml 5 mg/ml 2.5 mg/ml

0.1 1

0.1 q (nm-1)

S(c,0) = I(c,0)/I(0,0)

I(c,q)/c (a.u.)

a

1.5

1 0% PEG 3,350 2% PEG 3,350 4% PEG 3,350 6% PEG 3,350 0.5

0

0.01

0.02

0.03

0.04

0.05

[PCCM] (g/mL)

Fig. 7 Normalized SAXS curves of PCCM in 4% PEG 3350 as a function of surfactant concentration (a). Variation of forward structure factors of PCCM in different % of PEG 3350 as a function of surfactant concentration (b)

Franc¸oise Bonnete´ and Patrick J. Loll 5 Interaction parameter (ml/g)

188

0

-5

C12M by SAXS PCCM by SAXS C12M by DLS PCCM by DLS

-10

-15 0

1

2

3

4

5

6

% PEG 3350 (W/V)

Fig. 8 Interaction parameters kD and A2 obtained from DLS and SAXS for C12M and PCCM in different % PEG 3350

to attractive interactions between PCCM micelles, in agreement with what we previously observed by DLS. The forward structure factor S(c,0) obtained from the linear regression I(c,0) ¼ f(c) demonstrates that the attractive interactions between micelles, characterized by 2MA2, become higher when the percentage of PEG increases as shown in Fig. 7b by the increase in the slope, in line with the results obtained with OmpF in PEG using SLS (Fig. 4). When comparing interaction parameters of C12M and PCCM obtained using DLS and SAXS, we observed that the variation in kD follows the same trend as MA2 for the two surfactants, i.e., an increase in intermicellar attraction as % PEG increases (Fig. 8). The kD values are found lower than the 2MA2 values for the two surfactants, as expected from hydrodynamic effects [44]. Overall, PCCM micelles demonstrate more attractive inter-micelle potentials than C12M; this behavior can be explained by the micelle structures and the influence of the PCCM hydrophobic chain that bears two cyclic motifs. Even though RG and RH, the respective radius of gyration and hydrodynamic radius of surfactant micelles, are slightly larger for PCCM than for C12M (Table 4), the micelle molar masses and aggregation numbers differ significantly for these two surfactants. Values of molar masses were obtained from SAXS via the absolute forward intensity measurement or from SECMALLS (size-exclusion chromatography coupled in line to multiangle laser light scattering and refractometry) via static light scattering at 90 ; the latter method is thoroughly described elsewhere [52]. Molar masses/aggregation numbers calculated by the two techniques were in good agreement, and were found to be 65,000 Da/125 and 90,000 Da/165, respectively, for C12M and PCCM. The anomeric form of PCCM and its more rigid hydrophobic group favor the formation of more densely packed micelles

X-Ray and Light Scatterings for Detergent Micelle Behavior Analysis

189

Surfactant Concentration (g/L)

60 50 Two phases

40 30 20

Single phase DDM Liq/Liq DDM Liq/Sol PCC Liq/Liq

10 0 6

7

8

9

10

% PEG 3,350 (w/v)

Fig. 9 Cloud point boundary for C12M and PCCM in PEG 3350. Inserted pictures of C12M transition phases observed by optical microscopy

due to an increase in van der Waals contacts in the interior of the micelle as compared to C12M micelles, where the alkyl chains are more flexible and present fewer van der Waals contacts. This behavior may also explain the increase in stability observed with PCCM with the cytochrome b6f complex from Chlamydomonas reinhardtii and two GPCRs [19]. As described above, weak interaction forces between detergent micelles control the phase diagram and give rise to phenomena such as the cloud point boundary, which has already been shown as a useful predictive tool for the crystallization of protein-detergent complexes [47, 60]. Accordingly, we compared the cloud point phase transitions of C12M and PCCM (Fig. 9) and found that the cloud point for PCCM is observed at lower percentage of PEG 3350 than in the case of C12M. Moreover, in the case of C12M, we observed both liquid-liquid and liquid-solid transitions, the latter of which can be deleterious for membrane protein crystallization. This result suggests that membrane proteins might crystallize at lower concentrations of precipitant when PCCM is used as a surfactant, as compared with C12M. To test this hypothesis, crystallization trials were performed with the complex RCLH1-pufX from Rb. blasticus in both PCCM and C12M, using the same range of PEG concentrations. As shown in Fig. 10, protein crystals are observed in PCCM at lower PEG concentrations than in C12M, suggesting that the protein’s solubility curve is lower in PCCM than in C12M. This result clearly demonstrates the close correlation between intermicellar interactions and protein crystallization, as postulated above. A remaining question is whether crystals grown in this new surfactant diffract better than those obtained with C12M (although initial trials fail to show an improvement in the diffraction quality). The crystals’ diffraction properties reflect both how tightly and how uniformly the molecules are packed; the

190

Franc¸oise Bonnete´ and Patrick J. Loll Crystals of RC-LH1-pufX in DDM PEG 7.0%

PEG 7.8%

Crystals of RC-LH1-pufX in PCC PEG 7.0%

PEG 7.8%

Fig. 10 Crystals of RC-LH1-pufX from Rd. blasticus purified in C12M and PPCM

nature and amount of surfactants (including detergents and lipids) bound to membrane protein complexes will clearly have an impact on both these parameters. Thus, while controlling surfactant structure and interaction behavior shows clear promise for enhancing membrane protein crystallization, future efforts will also need to consider how the detergent component of the protein-detergent complex packs into a well-ordered lattice that is compatible with high-resolution diffraction.

4

Conclusion Crystallization of membrane proteins is a challenging task due to the presence of detergent micelles, which share the solution and interchange monomers with protein-detergent complexes. The nature and concentration of the detergent both critically affect the crystallization mechanism, because membrane proteins are embedded in a detergent corona and bathed in a solution of detergent micelles. The scattering methods (SLS, DLS, SAXS) presented in this chapter provide powerful tools for probing the mechanistic roles of the detergent, allowing us to move beyond empirical approaches toward a better understanding of the physics behind detergent behavior, and ultimately yielding a deeper knowledge of the mechanisms controlling the crystallization of membrane proteins. One of these tools (SLS) has revealed a close correlation between micelle-micelle interactions and complex-complex interactions, suggesting that the design of membrane protein crystallization screens could be facilitated by an optimization of detergentdetergent interactions. This was confirmed by another tool (SAXS), which showed that not only do detergent-detergent interactions drive interactions between protein-detergent complexes, but also that the nature of the detergent (polar head and/or hydrophobic chain) may influence the membrane protein phase diagram and the crystallization conditions. Our results demonstrate that detergent behavior in solution is an important parameter not to be neglected in the search of membrane protein crystallization conditions. However, while predicting conditions that favor crystal growth is tremendously valuable, we also need to know about crystal quality,

X-Ray and Light Scatterings for Detergent Micelle Behavior Analysis

191

and specifically whether crystals are suitable for high-resolution structure determination. Such information awaits further experiments that can describe close interactions between the detergent and the protein surface. References 1. Michel H (1983) Crystallization of membrane proteins. Trends Biochem Sci 8(2):56–59. doi:10.1016/0968-0004(83)90390-0 2. Caffrey M, Cherezov V (2009) Crystallizing membrane proteins using lipidic mesophases. Nat Protoc 4(5):706–731 3. Wadsten P, Wohri AB, Snijder A, Katona G, Gardiner AT, Cogdell RJ, Neutze R, Engstrom S (2006) Lipidic sponge phase crystallization of membrane proteins. J Mol Biol 364(1): 44–53 4. Bayburt TH, Sligar SG (2010) Membrane protein assembly into Nanodiscs. FEBS Lett 584(9):1721–1727. doi:10.1016/j.febslet. 2009.10.024 5. Faham S, Boulting GL, Massey EA, Yohannan S, Yang D, Bowie JU (2005) Crystallization of bacteriorhodopsin from bicelle formulations at room temperature. Protein Sci 14(3): 836–840. doi:10.1110/ps.041167605 6. Ducruix A, Guilloteau JP, Rie`s-Kautt M, Tardieu A (1996) Protein interactions as seen by solution X-ray scattering prior to crystallogenesis. J Cryst Growth 168:28–39 7. Finet S, Vivares D, Bonnete´ F, Tardieu A (2003) Controlling biomolecular crystallization by understanding the distinct effects of PEGs and salts on solubility. In: Macromolecular crystallography, Pt C, vol 368. Methods in enzymology. pp. 105–129 8. George A, Wilson WW (1994) Predicting protein crystallization from a dilute solution property. Acta Crystallogr D50:361–365 9. Tardieu A, Le Verge A, Rie`s-Kautt M, Malfois M, Bonnete´ F, Finet S, Belloni L (1999) Proteins in solution : from X-ray scattering intensities to interaction potentials. J Cryst Growth 196:193–203 10. Linke D (2009) Chapter 34 Detergents: an overview. In: Richard RB, Murray PD (eds) Methods in enzymology, vol 463. Academic Press, pp. 603–617. doi: 10.1016/S00766879(09)63034-2 11. Lorber B, Fischer F, Bailly M, Roy H, Kern D (2012) Protein analysis by dynamic light scattering: Methods and techniques for students. Biochem Mol Biol Educ 40(6):372–382. doi:10.1002/bmb.20644

12. Arnold T, Linke D (2008) The use of detergents to purify membrane proteins. Curr Protoc Protein Sci / editorial board, John E Coligan [et al] Chapter 4:Unit 4 8 1–4 8 30. doi:10.1002/0471140864.ps0408s53 13. Tate CG (2010) Practical considerations of membrane protein instability during purification and crystallisation. Methods Mol Biol 601:187–203 14. Polidori A, Raynal S, Barret L-A, Dahani M, Barrot-Ivolot C, Jungas C, Frotscher E, Keller S, Ebel C, Breyton C, Bonnete´ F (2016) Sparingly fluorinated maltoside-based surfactants for membrane-protein stabilization. New J Chem. doi:10.1039/C5NJ03502C 15. Abla M, Unger S, Keller S, Bonnete F, Ebel C, Pucci B, Breyton C, Durand G (2015) Micellar and biochemical properties of a propyl-ended fluorinated surfactant designed for membraneprotein study. J Colloid Interface Sci 445:127–136. doi:10.1016/j.jcis.2014.12. 066 16. Lee SC, Bennett BC, Hong W-X, Fu Y, Baker KA, Marcoux J, Robinson CV, Ward AB, Halpert JR, Stevens RC, Stout CD, Yeager MJ, Zhang Q (2013) Steroid-based facial amphiphiles for stabilization and crystallization of membrane proteins. Proc Natl Acad Sci 110 (13):E1203–E1211. doi:10.1073/pnas. 1221442110 17. Chae PS, Rana RR, Gotfryd K, Rasmussen SGF, Kruse AC, Cho KH, Capaldi S, Carlsson E, Kobilka B, Loland CJ, Gether U, Banerjee S, Byrne B, Lee JK, Gellman SH (2013) Glucoseneopentyl glycol (GNG) amphiphiles for membrane protein study. Chem Commun 49(23): 2287–2289 18. Sharma KS, Durand G, Gabel F, Bazzacco P, Le Bon C, Billon-Denis E, Catoire LJ, Popot J-L, Ebel C, Pucci B (2012) Non-ionic amphiphilic homopolymers: synthesis, solution properties, and biochemical validation. Langmuir 28(10): 4625–4639 19. Hovers J, Potschies M, Polidori A, Pucci B, Raynal S, Bonnete´ F, Serrano-Vega M, Tate C, Picot D, Pierre Y, Popot J-L, Nehme´ R, Bidet M, Mus-Veteau I, Bußkamp H, Jung KH, Marx A, Timmins PA, Welte W (2011) A class of mild surfactants that keep integral

192

Franc¸oise Bonnete´ and Patrick J. Loll

membrane proteins water-soluble for functional studies and crystallization. Mol Membr Biol 28(3):171–181 20. Muschol M, Rosenberger F (1995) Interactions in under- and supersaturated lysozyme solutions. Static and dynamic light scattering results. J Chem Phys 103(24):10424–10432 21. Chae PS, Cho KH, Wander MJ, Bae HE, Gellman SH, Laible PD (2014) Hydrophobic variants of ganglio-tripod amphiphiles for membrane protein manipulation. Biochim Biophys Acta 1838(1, Part B):278–286. http:// dx.doi.org/10.1016/j.bbamem.2013.09.011 22. Hong W-X, Baker KA, Ma X, Stevens RC, Yeager M, Zhang Q (2010) Design, synthesis, and properties of branch-chained maltoside detergents for stabilization and crystallization of integral membrane proteins: human connexin 26. Langmuir 26(11):8690–8696. doi:10.1021/la904893d 23. Chae PS, Rasmussen SG, Rana RR, Gotfryd K, Chandra R, Goren MA, Kruse AC, Nurva SF, Loland C, Pierre Y, Drew D, Popot J, Picot D, Fox B, Guan L, Gether U, Byrne B, Kobilka B, Gellman S (2010) Maltose-neopentyl glycol (MNG) amphiphiles for solubilization, stabilization and crystallization of membrane proteins. Nat Methods 7:1003–1008 24. Chae PS, Gotfryd K, Pacyna J, Miercke LJW, Rasmussen SGF, Robbins RA, Rana RR, Loland CJ, Kobilka B, Stroud R, Byrne B, Gether U, Gellman SH (2010) Tandem facial Amphiphiles for membrane protein stabilization. J Am Chem Soc 132(47):16750–16752 25. Tsamaloukas AD, Beck A, Heerklotz H (2009) Modeling the micellization behavior of mixed and pure n-alkyl-maltosides. Langmuir 25 (8):4393–4401. doi:10.1021/la8033935 26. Timmins PA, Hauk J, Wacker T, Welte W (1991) The influence of heptane-1,2,3-triol on the size and shape of ldao micelles - implications for the crystallization of membrane-proteins. FEBS Lett 280(1):115–120 27. Hitscherich C, Kaplan J, Allaman M, Wiencek J, Loll PJ (2000) Static light scattering studies of OmpF porin: Implications for integral membrane protein crystallization. Protein Sci 9 (8):1559–1566 28. Pernot P, Theveneau P, Giraud T, Fernandes RN, Nurizzo D, Spruce D, Surr J, McSweeney S, Round A, Felisaz F, Foedinger L, Gobbo A, Huet J, Villard C, Cipriani F (2010) New beamline dedicated to solution scattering from biological macromolecules at the ESRF. J Phys Conf Ser 247(1):012009 29. Round A, Felisaz F, Fodinger L, Gobbo A, Huet J, Villard C, Blanchet CE, Pernot P,

McSweeney S, Roessle M, Svergun DI, Cipriani F (2015) BioSAXS Sample Changer: a robotic sample changer for rapid and reliable high-throughput X-ray solution scattering experiments. Acta Crystallogr D Biol Crystallogr 71(Pt 1):67–75. doi:10.1107/ s1399004714026959 30. Konarev PV, Volkov VV, Sokolova AV, Koch MHJ, Svergun DI (2003) PRIMUS: a Windows PC-based system for small-angle scattering data analysis. J Appl Cryst 36:1277–1282 31. Lindner P, Zemb T (2002) Neutron, X-rays and light. Scattering methods applied to soft condensed matter. Mater Today 5(11):38. doi:10.1016/S1369-7021(02)01143-4 32. Tardieu A (1994) Thermodynamics and structure–concentrated solutions–structured disorder in vision. In: Neutron and synchrotron radiation for condensed matter studies, vol III. Springer-Verlag: Les e´ditions de Physique, France, pp. 145–160 33. Boyer M, Roy MO, Jullien M (1996) Dynamic light scattering study of precrystallizing ribonuclease solutions. J Cryst Growth 167:212–220 34. Bonnete´ F, Finet S, Tardieu A (1999) Second virial coefficient: variations with lysozyme crystallization conditions. J Cryst Growth 196:403–414 35. Li S, Xing D, Li J (2004) Dynamic light scattering application to study protein interactions in electrolyte solutions. J Biol Phys 30(4): 313–324. doi:10.1007/s10867-004-0997-z 36. Guinier A, Fournet G (1955) Small angle scattering of X-rays. Wiley, New York 37. Berne BJ, Pecora R (1976) Dynamic light scattering. John Wiley, New York 38. Glatter O, Kratky O (eds) (1982) Small angle X-ray scattering. Academic Press, Cambridge, MA 39. Koch MHJ, Vachette P, Svergun DI (2003) Small-angle scattering: a view on the properties, structures and structural changes of biological macromolecules in solution. Q Rev Biophys 36(2):147–227 40. Lipfert J, Doniach S (2007) Small-angle X-ray scattering from RNA, proteins, and protein complexes. Annu Rev Biophys Biomol Struct 36:307–327. doi:10.1146/annurev.biophys. 36.040306.132655 41. Itakura M, Shimada K, Matsuyama S, Saito T, Kinugasa S (2006) A convenient method to determine the Rayleigh ratio with uniform polystyrene oligomers. J Appl Polym Sci 99 (4):1953–1959. doi:10.1002/app.22695 42. Koppel DE (1972) Analysis of macromolecular polydispersity in intensity correlation

X-Ray and Light Scatterings for Detergent Micelle Behavior Analysis spectroscopy: the method of cumulants. J Chem Phys 57(11):4814–4820. doi:10. 1063/1.1678153 43. Provencher SW (1982) CONTIN: A general purpose constrained regularization program for inverting noisy linear algebraic and integral equations. Comput Phys Commun 27(3): 229–242. doi:10.1016/0010-4655(82) 90174-6 44. Harding SE, Johnson P (1985) The concentration-dependence of macromolecular parameters. Biochem J 231(3):543–547 45. Yadav S, Scherer TM, Shire SJ, Kalonia DS (2011) Use of dynamic light scattering to determine second virial coefficient in a semidilute concentration regime. Anal Biochem 411(2): 292–296. doi:10.1016/j.ab.2010.12.014 46. Wilson WW, Delucas LJ (2014) Applications of the second virial coefficient: protein crystallization and solubility. Acta Crystallogr Sect F Struct Biol Cryst Commun 70(Pt 5): 543–554. doi:10.1107/s2053230x1400867x 47. Loll PJ, Allaman M, Wiencek J (2001) Assessing the role of detergent-detergent interactions in membrane protein crystallization. J Cryst Growth 232(1-4):432–438 48. Rosenbusch JP (1990) The critical role of detergents in the crystallization of membrane proteins. J Struct Biol 104(1–3):134–138. doi:10.1016/1047-8477(90)90068-N 49. Wiener MC, Snook CF (2001) The development of membrane protein crystallization screens based upon detergent solution properties. J Cryst Growth 232(1-4):426–431 50. Zulauf M (1991) Detergent phenomena in membrane protein crystallization. In: Michel H (ed) Crystallization of membrane proteins. CRC Press, Boca Rato, pp 53–72 51. Koszelak-Rosenblum M, Krol A, Mozumdar N, Wunsch K, Ferin A, Cook E, Veatch CK, Nagel R, Luft JR, DeTitta GT, Malkowski MG (2009) Determination and application of empirically derived detergent phase boundaries

193

to effectively crystallize membrane proteins. Protein Sci 18(9):1828–1839 52. Barret L-A, Barrot-Ivolot C, Raynal S, Jungas C, Polidori A, Bonnete F (2013) Influence of hydrophobic micelle structure on crystallization of the photosynthetic RC-LH1-PufX complex from rhodobacter blasticus. J Phys Chem B 117(29):8770–8781. doi:10.1021/ jp403483q 53. Gonzalez-Tello P, Camacho F, Blazquez G (1994) Density and viscosity of concentrated aqueous solutions of polyethylene glycol. J Chem Eng Data 39(3):611–614. doi:10. 1021/je00015a050 54. Lindner P, Zemb T (1991) Neutron, X-ray and light scattering: introduction to an investigative tool for colloidal and polymeric systems. North-Holland Delta Series, Amsterdam 55. Vachette P 1996 Small angle X-ray scattering by solutions of biological macromolecules. In: Proceedings of the International school of physics “Enrico Fermi”. pp. 269–292. 56. Vivares D, Belloni L, Tardieu A, Bonnete´ F (2002) Catching the PEG-induced attractive interaction between proteins. Eur Phys J E 9:15–25 57. Eisenberg H (1976) Biological macromolecules and polyelectrolytes in solution. Clarendon Press, Oxford 58. Pedersen JS (1997) Analysis of small-angle scattering data from colloids and polymer solutions: modeling and least-squares fitting. Adv Colloid Interface Sci 70:171–210. doi:10. 1016/S0001-8686(97)00312-6 59. Orthaber D, Bergmann A, Glatter O (2000) SAXS experiments on absolute scale with Kratky systems using water as a secondary standard. J Appl Cryst 33:218–225 60. Hitscherich C, Aseyev V, Wiencek J, Loll PJ (2001) Effects of PEG on detergent micelles: implications for the crystallization of integral membrane proteins. Acta Crystallogr D57:1020–1029

Chapter 10 Secondary Structure Determination by Means of ATR-FTIR Spectroscopy Batoul Srour, Stefan Bruechert, Susana L.A. Andrade, and Petra Hellwig Abstract Specialized infrared spectroscopic techniques have been developed that allow studying the secondary structure of membrane proteins and the influence of crucial parameters like lipid content and detergent. Here, we focus on an ATR-FTIR spectroscopic study of Af-Amt1 and the influence of LDAO/glycerol on its structural integrity. Our results clearly indicate that infrared spectroscopy can be used to identify the adapted sample conditions. Key words Secondary structure analysis, Protein-detergent interaction, Infrared spectroscopy, Ammonium transport

1

Introduction During the last few years, attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) has become one of the most powerful methods to determine the structure of biological membranes and proteins. It gives access to structural information on proteins that cannot be studied by X-ray crystallography and NMR and it is highly complementary to CD spectroscopy [1–10]. The technique is of special interest for the study of the structure, orientation, and tertiary structure changes in peptides and membrane proteins [7, 11, 12]. ATR-FTIR requires a small amount of material (1–100 μg) and spectra are recorded in a matter of minutes without limitation regarding the size of the studied molecule. The environment of the target molecules can be modulated such that conformational changes can be studied as a function of pH, salt, temperature, pressure, as well as the presence of specific substrates. Moreover, in addition to the conformational parameters that can be deduced from the shape of the infrared spectra, the orientation of various parts of the molecule can be analyzed with polarized IR [2, 12–14].

Jean-Jacques Lacapere (ed.), Membrane Protein Structure and Function Characterization: Methods and Protocols, Methods in Molecular Biology, vol. 1635, DOI 10.1007/978-1-4939-7151-0_10, © Springer Science+Business Media LLC 2017

195

196

Batoul Srour et al.

Fig. 1 As isolated Af-Amt1 trimer with individual monomers colored in white, gray, and dark red [17]. The cytoplasmic side of the protein is facing down

Here, we demonstrate the use of ATR-FTIR spectroscopy to investigate the Archaeoglobus fulgidus Af-Amt1, an integral membrane spanning ammonium transporter that is essential for nitrogen assimilation in prokaryotes and plants [15]. Figure 1 shows the high-resolution crystal structure of the Af-Amt1 trimer. The protein is predominantly alpha-helical, with each monomer displaying 11 transmembrane alpha-helices plus a smaller cytoplasmic helical segment at the C-terminus [16–18]. The aim of the study was to distinguish how different experimental parameters such as the presence of lipids, detergent, or glycerol molecules could influence the structural integrity of the protein. The protocol is focusing on the later two.

2

Materials

2.1 Sample Preparation

Af-Amt1 was prepared as described previously [17, 19] and were studied in the following buffers: Buffer A: 20 mM Tris–HCl pH 8.0, 100 mM NaCl, 10% glycerol and 0.05% LDAO. Buffer B: 20 mM Tris–HCl pH 8.0, 100 mM NaCl, 0% glycerol and 0.05% LDAO. Buffer C: 20 mM Tris–HCl pH 8.0, 100 mM NaCl, 2.5% glycerol and 0.01% LDAO. Buffer D: 20 mM Tris–HCl pH 8.0, 100 mM NaCl, 0% glycerol and 0.01% LDAO. Samples (10 mg/mL) were kept at
80  C and thawed on ice before the experiments. The infrared spectroscopic studies on the highly stable protein from archaea have been performed at room temperature. Buffer exchange was carried out using centricons from Amicon.

Infrared Spectroscopy for Secondary Structure Determination

2.2 Experimental Setup Infrared

3

197

One to two microliters of sample were dried on a diamond crystal in a HARRICK ATR cell, each. A VERTEX 70 FTIR spectrometer (Bruker), equipped with a MCT (mercury cadmium telluride) detector, was used in the spectral range from 4000 to 800/cm. Typically, 10  128 scans have been averaged at a resolution of 2/ cm. Data was analyzed with the OPUS software from Bruker and the deconvolution was processed with ORIGIN.

Methods

3.1 Sample Deposition

Samples from dried proteins have been studied. First, it was probed if the drying perturbs the stability of the sample. It is noted that spectra from fully hydrated proteins are much more challenging to obtain, since a strong signal from water overlaps the contribution of the protein backbone, but offer the advantage of more biologically relevant conditions. Infrared spectra are often measured from samples in D2O buffer, to avoid the strong H2O signal; however, these data are often biased for membrane proteins, since they do not exchange completely as was shown by HD exchange kinetics on membrane proteins (see, for example, 19). The remaining OH vibration will add to the amide I signal and the relative intensity is difficult to obtain. To yield information about the protein orientation in the membrane and analyze the protein in closer to native conditions, samples could also be monitored after reconstitution into lipid vesicles. Here, dried samples were stable and we have thus privileged this approach. In general, it was reported that monolayers lead to results that are qualitatively similar to air-dried peptide–lipid mixtures, but band frequencies are red-shifted in an orientation-dependent manner for about 2–5/cm [20].

3.2

First, a spectrum for Af-Amt1 (10 mg/mL) in buffer A was recorded and compared to the data obtained for the buffer alone (Fig. 2a, b, respectively). The data of a membrane protein can include contributions from the protein, lipids, buffer, and detergent. Table 1 summarizes the main contributions of membrane proteins together with the signal typical for the ν(C¼O) vibrational mode of phospholipids. For the analysis of the secondary structure, we focus on the amide I signal, typically present within the range around 1700–1600/cm. For each detergent type and buffer, a specific spectral signature can be expected that needs to be determined prior to the experiments. This approach can also be used to determine the amount of detergent and even lipids present in the sample [21–23]. As can be seen in Fig. 2b, the contribution of the buffer, glycerol, and eventually some remaining water is very strong, even for the dried sample.

Measurement

A

1,00

1041

Batoul Srour et al.

3308

198

1108

0,50

1650 1413

2936

Abs

0,75

0,25

0,00 4000

3500

3000

2500

2000

1500

1000

Wavenumber / cm

-1

B

1,00

1108

0,50

1656 1413

2931

Abs

0,75

0,25

0,00 4000

3500

3000

2500

2000

1500

1000

Wavenumber (cm ) -1

Fig. 2 Absorption of the buffer alone (a) and the contribution of the protein together with the buffer (b)

3.3 Deconvolution Procedures

As mentioned above, the amide I signal includes information on the secondary structure of a protein or peptide, a signal that predominantly includes the coordinates of the ν(C¼O) vibrational modes. The deconvolution is based on the idea that amide I vibrational modes can be modeled as a subset of individual harmonic oscillators that are coupled to each other but isolated from the rest of the normal modes of a peptide (see Note 1). For example, the Frenkel exciton model [24], a standard model in the treatment of vibrations in molecular crystals, was found to be a reasonable approach to theoretically describe a spectrum. By assuming that the transition dipole interaction couples the local amide I modes, qualitative agreement has been reached over the years between simulation and experimental FTIR spectra of mid-sized proteins. However,

Infrared Spectroscopy for Secondary Structure Determination

199

Table 1 Overview of the typical IR spectral feature contributions from membrane proteins [2, 6] Main contribution

Typical position (cm
1)

Involved vibrational modes

Phospholipid

1740

ν(C¼O)

Amide A

~3300

ν(N–H), ν(O–H),

Amide B

3100–3050

ν(N–H)

Amide I

1700–1600

80% ν(C¼O), ν(C–N), δ(N–H)

Amide II

1580–1480

60% δ(N–H), ν(C–N)

Amide III

1400–1200

δ(N–H), δ(C–N), ν(C¼O), δ(N–C¼O)

Amide IV

780–650

δ(C¼O) in plane, ν(C–C)

Amide V

600–540

δ(C–N) torsion

Amide VI

540–500

δ(C¼O) in plane and out of plane

Amide VII

290–250

δ(N–H) out of plane

1500 1700 1600 Wavenumber (cm–1)

1600 1700 1650 Wavenumber (cm–1)

Fig. 3 Baseline subtraction performed before amide I analysis

the transition dipole coupling model for the electrostatic, throughspace part of the interaction requires an arbitrary choice of origin and is therefore unreliable at distances that are comparable with the size of amide unit undergoing the vibration, in particular between the nearby amide units. There are several approaches accepted for the analysis of the amide I band. The first step here is the subtraction of a baseline as shown in Fig. 3. Then the second derivative is obtained in order to identify the number of contributions [25]. Alternatively, a Fourier transformation or a fine structure enhancement can be

Batoul Srour et al. 1656

200

95.4%

0.3

1622

Absorbance

0.6

4.6%

0.0 1716

1694

1672

1650

1628

1606

1584

Wavenumber [cm–1]

Fig. 4 Amide I signal from native Af-Amt-1 and its deconvolution revealing the alpha helical structure

Table 2 Characteristic secondary structure elements spectral positions [8, 24] Secondary structure

H2O (cm
1)

D2O (cm
1)

Helix α

1648–1657

1642–1660

β sheet (parallel)

1623–1641

1615–1638

β sheet (antiparallel)

1674–1695/1615–1627

1672–1694/1613–1625

β turns

1662–1686

1653–1691

Nonorganized

1642–1657

1639–1654

applied [26, 27]. Finally, the amide I signal is deconvoluted and the relative contribution of each secondary structure element calculated relative to the total contribution. Fits are mostly done using Gaussian fits or a mixture of Gaussian and Lorentzian curves. Figure 4 shows the deconvolution obtained for the protein studied here. 3.4 Assignment of the Secondary Structure

The characteristic position of each possible secondary structure element is summarized in Table 2. These assignments for H2O or D2O buffer are based on a large number of proteins and peptides studied over the years, their theoretical analysis and the comparison to data obtained via CD spectroscopy and crystallography [8, 11, 25, 28, 29]. As can be seen in the table, β-sheet type structures show quite distinct contributions, whereas alpha helical and nonorganized structures contribute at a very close position. This is the reason why the approach is often used in a complementary manner

Infrared Spectroscopy for Secondary Structure Determination

201

to CD spectroscopy, a technique where the alpha helical structures are more evident. The data obtained here on the Af-Amt1 protein clearly confirms its predominant alpha helical character (Fig. 1) that can also be corroborated with its crystallographic structure [17]. 3.5 Effect of Detergent and Glycerol

The next step was to probe the effect of different concentrations of detergent and glycerol. Figure 5 shows the evolution of the amide I band depending on the concentration of LDAO and glycerol. In the presence of 0.05% LDAO the increase of random and β-sheet structure point toward instability of the protein in this detergent in the given experimental conditions (dried sample). The variation of the glycerol concentration, however, does not seem to further affect the secondary structure. The decrease of the LDAO concentration leads to the denaturation of the sample, as can be deduced from the increased contribution of random structures to 50% and the loss of the alpha helical character of the protein. These strong changes can already be depicted from the overall change of the shape of the amide I signal and are confirmed after deconvolution of the signal. The ATR-FTIR spectroscopic data is a fast and reliable approach to identify changes in the structure of a protein, independent of its size.

Fig. 5 Absorption spectra and deconvolution of the protein in Buffer A (a), Buffer B (b), Buffer C (c), and Buffer D (d)

202

4

Batoul Srour et al.

Note 1. The study of the secondary structure of proteins and even membrane proteins via ATR-FTIR is fast and needs little sample; however, the technique needs to be applied carefully and the possible errors need to be evaluated [28, 30, 31]. The fit of several components into one large broad structure often allows several alternative results that can produce fitting errors up to 10%. The technique is much more reliable when, for example, a series of experiments is compared on the same instrument and the data is analyzed by the same processing. It often remains very difficult to distinguish alpha helical structures and random coil structures and additional investigations in D2O or by means of CD spectroscopy need to be added [29].

5

Conclusion Importantly, the potential of the technique is far beyond the protocol presented here. Samples reconstituted in membranes can be studied depending on their orientation and during interaction or coupled to electrochemistry driven, or substrate perfusion processes. In combination with infrared microscopy techniques, the identification of secondary structure elements within tissues is possible, a technique that allowed identifying the antiparallel character of toxic amyloid peptides [32].

References 1. Bandekar J (1992) Amide modes and protein conformation. Biochim Biophys Acta 1120:123–143 2. Haris PI, Chapman D (1995) The conformational analysis of peptides using Fourier transform IR spectroscopy. Biopolymers 37(4): 251–263 3. Qi XL, Holt C, McNultyn D, Clarke DT, Brownlow S, Jones GR (1997) Effect of temperature on the secondary structure of betalactoglobulin at pH 6.7, as determined by CD and IR spectroscopy: a test of the molten globule hypothesis. Biochem J 324(Pt 1):341–346 4. Tamm LK, Tatulian SA (1997) Infrared spectroscopy of proteins and peptides in lipid bilayers. Q Rev Biophys 30:365–429 5. Arora A, Tamm LK (2001) Biophysical approaches to membrane protein structure determination. Curr Opin Struct Biol 11(5):540–547 6. Barth A (2007) Infrared spectroscopy of proteins. Biochim Biophys Acta 1767:1073–1101

7. Goormaghtigh E, Cabiaux V, Ruysschaert JM (1994) Determination of soluble and membrane protein structure by Fourier transform infrared spectroscopy. I. Assignments and model compounds. In: Hilderson HJ, Raston GB (eds) Physicochemical methods in the study of biomembranes, Subcellular biochemistry, vol 23. Plenum Press, New York, pp 329–336 8. Goormaghtigh E, Cabiaux V, Ruysschaertn JM (1994) Determination of soluble and membrane protein structure by Fourier transform infrared spectroscopy. III. Secondary structures. Subcell Biochem 23:405–450 9. Kong J, Yu S (2007) Fourier transform infrared spectroscopic analysis of protein secondary structures. Acta Biochim Biophys Sin 39:549–559 10. Mercer EA, Abbott JGW, Brazier SP, Ramesh B, Haris PI, Srai SKS (1997) Synthetic putative transmembrane region of minimal potassium channel protein (minK) adopts an α-helical

Infrared Spectroscopy for Secondary Structure Determination conformation in phospholipid membranes. Biochem J 325:475–479 11. Goormaghtigh E, Gasper R, Be´nard A, Goldsztein A, Raussens V (2009) Protein secondary structure content in solution, films and tissues: redundancy and complementarity of the information content in circular dichroism, transmission and ATR FTIR spectra. Biochim Biophys Acta 1794(9):1332–1343 12. Vigano C, Manciu L, Buyse F, Goormaghtigh E, Ruysschaert JM (2000) Attenuated total reflection IR spectroscopy as a tool to investigate the structure, orientation and tertiary structure changes in peptides and membrane proteins. Biopolymers 55(5):373–380 13. Bazzi MD, Woody RW (1985) Oriented secondary structure in integral membrane proteins. I. Circular dichroism and infrared spectroscopy of cytochrome oxidase in multilamellar films. Biophys J 48(6):957–966 14. Rothschild KJ (1992) FTIR difference spectroscopy of bacteriorhodopsin: toward a molecular model. J Bioenerg Biomembr 24(2):147–167 15. Andrade SLA, Einsle O (2007) The Amt/ Mep/Rh family of ammonium transport proteins. Mol Membr Biol 24:357–365 16. Ullmann RT, Andrade SLA, Ullmann GM (2012) Thermodynamics of transport through the ammonium transporter Amt-1 investigated with free energy calculations. J Phys Chem B 116:9690–9703 17. Andrade SLA, Dickmanns A, Ficner R, Einsle O (2005) Crystal structure of the archaeal ammonium transporter Amt-1 from Archaeoglobus fulgidus. Proc Natl Acad Sci U S A 102(42):14994–14999 18. Zheng L, Kostrewa D, Berneche S, Winkler F, Li X (2004) The mechanism of ammonia transport based on the crystal structure of AmtB of Escherichia coli. Proc Natl Acad Sci U S A 101(49):17090–17095 19. Andrade SLA, Dickmanns A, Ficner R, Einsle O (2005) Expression, purification and crystallization of the ammonium transporter Amt-1 from Archaeoglobus fulgidus. Acta Cryst F61:861–863 20. Paul C, Wang J, Wimley WC, Hochstrasser RM, Axelsen PH (2004) Vibrational coupling, isotopic editing, and beta-sheet structure in a membrane-bound polypeptide. J Am Chem Soc 126(18):5843–5850

203

21. Strug I, Utzat C, Cappione A 3rd, Gutierrez S, Amara R, Lento J, Capito F, Skudas R, Chernokalskaya E, Nadler T (2014) Development of a univariate membrane-based mid-infrared method for protein quantitation and total lipid content analysis of biological samples. J Anal Methods Chem 2014:657079 22. Viana RB, da Silva ABF, Pimentel AS (2012) Infrared spectroscopy of anionic, cationic, and zwitterionic surfactants. Adv Phys Chem 2012:1–14 23. Das C, Nadler T, Strug I (2015) Detergent analysis in protein samples using mid-infrared (MIR) spectroscopy. Curr Protoc Protein Sci 81:29.12.1–29.12.15 24. Bardeen CJ (2014) The structure and dynamics of molecular excitons. Annu Rev Phys Chem 65:127–148 25. Byler M, Susi H (1986) Examination of the secondary structure of proteins by deconvoluted FTIR spectra. Biopolymers 25:469–487 26. Yang W-J, Griffiths P, Byler D, Susi H (1985) Protein conformation by infrared spectroscopy: resolution enhancement by Fourier self deconvolution. Appl Spectrosc 39:282–287 27. Barth A (2000) Fine-structure enhancement— assessment of a simple method to resolve overlapping bands in spectra. Spectrochim Acta 56:1223–1232 28. Goormaghtigh E, Ruysschaert JM, Raussens V (2006) Evaluation of the information content in infrared spectra for protein secondary structure determination. Biophys J 90(8): 2946–2957 29. Oberg KA, Ruysschaert JM, Goormaghtigh E (2004) The optimization of protein secondary structure determination with infrared and circular dichroism spectra. Eur J Biochem 271(14):2937–2948 30. Jackson M, Mantsch HH (1995) The use and misuse of FTIR spectroscopy in the determination of protein structure. Crit Rev Biochem Mol Biol 30:95–120 31. Surewicz WK, Mantsch HH, Chapman D (1993) Determination of protein secondary structure by Fourier transform infrared spectroscopy: a critical assessment. Biochemistry 389–394(32):32 32. Miller LM, Bourassa MW, Smith RJ (2013) FTIR spectroscopic imaging of protein aggregation in living cells. Biochim Biophys Acta 1828(10):2339–2346

Chapter 11 Native Mass Spectrometry for the Characterization of Structure and Interactions of Membrane Proteins Jeroen F. van Dyck, Albert Konijnenberg, and Frank Sobott Abstract Over the past years, native mass spectrometry and ion mobility have grown into techniques that are widely applicable to the study of aspects of protein structure. More recently, it has become apparent that this approach provides a very promising avenue for the investigation of integral membrane proteins in lipid or detergent environments. In this chapter, we discuss applications of native mass spectrometry and ion mobility in membrane protein research—what is important to take into consideration when working with membrane proteins, and what the requirements are for sample preparation for native mass spectrometry. Furthermore, we will discuss the types of information provided by the measurements, including the oligomeric state, subunit composition and stoichiometry, interactions with detergents or lipids, conformational transitions, and the binding and structural effect of ligands and drugs. Key words Native mass spectrometry, Ion mobility, Membrane proteins, Detergent micelles, Lipids

1

Introduction Over the past years native mass spectrometry (MS) has shown to be a promising avenue in the pursuit of structural aspects of membrane proteins and their lipid environment [1–3]. Even though native MS provides low-resolution data compared to other, more established structural biology methods, such as X-ray crystallography and nuclear magnetic resonance, MS has shown to be a powerful technique providing structural information of highly flexible, heterogeneous and polydisperse proteins and complexes, especially in cases where traditional techniques reached their limits or have failed [4–8]. Specific advantages of native MS are relatively fast measurements, low sample consumption (nano- to picomole) and that coexisting conformational and assembly states can be examined concomitantly, rather than just obtaining an average over all species present in a sample [9]. The gathered information from a measurement may comprise, for example, oligomeric states, protein subunit

Jean-Jacques Lacapere (ed.), Membrane Protein Structure and Function Characterization: Methods and Protocols, Methods in Molecular Biology, vol. 1635, DOI 10.1007/978-1-4939-7151-0_11, © Springer Science+Business Media LLC 2017

205

206

Jeroen F. van Dyck et al.

interactions, protein aggregation (amyloids) and the binding of drugs, metals, and cofactors, among others [10–13]. Native MS has also proved to be very useful for the study of protein conformational space, including phenomena of intrinsic disorder [14, 15]. Despite being important drug targets [16] and comprising more than a quarter of the proteome, integral membrane proteins (IMP) have shown to be difficult to characterize structurally, representing only 1 M) while maintaining the original ionic strength. Furthermore, it is possible to adjust the pH with either ammonia or acetic acid [42].

210

Jeroen F. van Dyck et al.

2.1 Detergent-Based Reconstitution Systems

Reconstitution systems are essential for membrane proteins, and a variety of approaches are employed for their solubilization, and in order to provide a native-like hydrophobic environment in an aqueous medium. Common systems used with other structural techniques such as X-ray crystallography, electron microscopy, and NMR can also be applied to native MS. So far the majority of structural work takes advantage of detergent micelles, though detergent-free reconstitution systems such as nanodiscs, bicelles, and amphipols [43] have also proven useful for studying membrane protein structure. As lipids often provide a specific local environment for functional IMPs or make up integral part of the protein assembly in the membrane, structural biologists are now seeking to include this important component in the reconstitution systems. Finding out what lipids are intimately involved in a membrane protein structure can be a challenge in itself though, and recently native MS has shown great promise in this regard (see below). Up to this point however, detergent-based systems are still by far the most popular reconstitution systems. Detergents form micelles when used above the critical micelle concentration (CMC) (Table 1). Regular micelles have a hydrophobic inner core made up of the tails of detergent molecules, and are hydrophilic on the outside where the detergent head groups are facing the aqueous solvent (Fig. 2). These detergent micelles can also form around the hydrophobic regions or the whole of a membrane protein, thereby preventing aggregation. While different detergents might well be effective for solubilization, it is often found that only a few of them create the appropriate native-like environment that preserves the native fold and assembly state of an IMP [44–46]. Finding this “right” detergent can be a major bottleneck in structural biology pipelines. Native MS has been used to rapidly screen the effect of different detergents on the folding and assembly state of the mechanosensitive channel of large conductance (MSCL). Of the detergents known to work with native MS, DDM showed that it kept the channel native in solution, although detergent removal in the gas phase was preceded by the dissociation of the pentamer. LDAO was found to not maintain the native state in solution, whereas triton X-100 satisfied the requirements of both the protein and the method (Fig. 3) [47, 48]. Some detergents that have been found to work for other structural biology techniques may not be well suited for native MS. The evidence so far suggests that mild and non-denaturing detergents are best, and specifically for native MS the use of nonionic (Table 1) or zwitterionic detergents is recommended, depending on the preference of the protein [49]. The most commonly used detergent is nonionic DDM, which usually also works well in combination with native MS measurements. There are however sometimes better alternatives such as C12E8 and triton X-100, as shown in a recent comparative study [50]. No study utilizing ionic detergents

Mass Spectrometry of Membrane Protein Samples

211

Table 1 List of nonionic detergents that have been found to be compatible with reconstitution of IMPs and mass spectrometry. The critical micelle concentration (CMC) is given in H2O and NaCl solution (M: molarity), as well as weight percentage (% wt.). The aggregation number is the average number of detergent molecules forming micelles. Information was gathered from the manufacturer’s website Anatrace (Maumee/Ohio, USA)

Detergent

CMC (mM) CMC (mM) CMC Aggregation Abbrev. (in H2O) (in NaCl (conc.)) (in % wt.) number

n-Decyl-β-D-Maltopyranoside

DM

1.8

1.8 (0.15 M)

0.087

69

n-Dodecyl-β-D-Maltopyranoside

DDM

0.17

0.12 (0.2 M)

0.0087

78–149

n-Undecyl-β-D-Maltopyranoside

UDM

0.59

n/a

0.029

71

n-Tridecyl-β-D-Maltopyranoside

TDM

0.033

0.24 (0.15 M)

0.017

186

CYMAL-5

Cy5

2.4–5.0

2.0 (0.15 M)

0.12

47

CYMAL-6

Cy6

0.56

n/a

0.028

91

n-Dodecyl-β-DThiomaltopyranoside

DDTM

0.05

n/a

0.0026

126

Octyl Glucose Neopentyl Glycol

OGNG

1.02

n/a

0.058

n/a

Triton X-100

TX100

0.01

n/a

0.016

75–165

C8E4

C8E4

n/a

8.0 (0.15 M)

0.25

82

C8E5

C8E5

n/a

7.1 (0.1 M)

0.25

n/a

C10E5

C10E5

0.81

n/a

0.031

73

C12E8

C12E8

n/a

0.09 (0.05 M)

0.0048

90–120

C12E9

C12E9

0.05

n/a

0.003

n/a

Anapoe-58

C16E20 0.004

n/a

0.00045

n/a

n-Octyl-β-DThioglucopyranoside

OTG

9

n/a

0.28

n/a

n-Octyl-β-D-Glucopyranoside

OG

18–20

23.4 (0.1 M)

0.53

27–100

n-Nonyl-β-D-Glucopyranoside

NG

6.5

6.0 (0.15 M)

0.20

133

in native MS has been reported to date [45]; the use of ionic detergents such as SDS can even lead to the aggregation of the membrane protein due to unfolding [17]. Another reason why ionic detergents are not useful in native MS is that they are harder to detach from the protein in the gas phase. When using detergents as reconstitution system in combination with native MS, it is advisable not to work too close to the CMC, as dilution effects can occur and the actual CMC depends somewhat on the protein and buffer present. Having a twofold excess of detergent (i.e., 2
the CMC) should ensure that micelles

212

Jeroen F. van Dyck et al.

A

C HO HO

OH

H2C

O HO CH

head group

O OH HC

H2 C OH

HO

O

O CH2

B

H 3C

CH3 CH

H2C

3

C H 3C H 3C

CH2

CH2

head group

C

H2C CH2 H2C

tail group CH2

H2C O CH2

tail group

CH2 H2C

H2C

n

CH2

OH H3C

Fig. 2 (a) A membrane protein reconstituted with the transmembrane domain (red) in detergents (light blue) and the soluble part on the outside of the micelle (dark blue). (b) Structure of triton X-100, where n is an average of 9.5 repeats. (c) Structure of DDM. Both these detergents contain a hydrophilic (head) and hydrophobic (tail) group

are formed spontaneously, with and without embedded IMP. For a wide array of detergents, the CMC will differ considerably, but for native MS it is recommended to use detergents with a CMS in the low millimolar concentration range. For example, the popular detergents DDM and triton X-100 have a CMC of 0.17 and 0.01 mM, respectively (Table 1). Another important factor for the selection of detergents for use with native MS is the amount of activation required to release the protein from the detergent in the gas phase (see below). The formation of some detergent micelles in solution is mainly driven by hydrophobic interactions, while others rely on the strength of the H-bonding network between the detergent molecules. Micelles with a low degree of H-bonding have been shown to require reduced levels of activation in order to set the protein free. Strong electrostatic interactions between the detergent molecules might however require levels of collisional activation that can already induce unwanted dissociation of the complex or collisional unfolding during

Mass Spectrometry of Membrane Protein Samples

213

Fig. 3 Native IM-MS spectra of released MscL protein. Comparison of three different detergents: (a) MscL released from DDM showed only monomers. This is due to the high collisional activation (200 V sampling cone and 150 V trap collision energy) needed to remove DDM, resulting in complex dissociation. (b) For the LDAO reconstituted MscL the drift plot revealed more extended conformations even at low activating energies, which is due to a nonnative environment in solution. (c) The pentameric MscL complex was observed with triton X-100 which needs only low collisional activation (1 V sampling cone and 10 V trap CE) and results in compact conformations in the drift plot. Beneath the drift plot is the corresponding mass spectrum with charge states in the 15þ to 20þ range retaining up to 6 triton X-100 molecules (vertical lines)

its release. If this should be a concern or not depends on the stability of the protein structure in relation to the strength of the detergent interactions, but also on the degree of possible exposed (soluble) protein structure not embedded in the micelle [49, 51]. 2.2 Detergent-Free Reconstitution Systems

Alternative, detergent-free systems, which are lipid-based or contain amphiphatic molecules (amphipols), are becoming increasingly popular for structural studies. Nanodiscs and bicelles consist of a lipid bilayer disc surrounded by either an amphipathic alpha helical membrane scaffold protein (MSP; nanodisc) or short chain lipids or detergents (bicelles), shielding the hydrophobic tails of the lipids on the sides of the lipid disc (Fig. 4). Using a lipid bilayer without

214

Jeroen F. van Dyck et al.

a

c

H2 C

H C

O

C O–

b

H2 C

H2 C

H C C O

a

NH

H C C O

b

NH

H2C

C

CH CH2

H3C

CH3

H2C CH2 H2C CH2 H2C CH3

Fig. 4 Detergent-free reconstitution systems. (a) A bicelle consisting of long-chain lipids (orange) and shortchain lipids or detergents (blue). (b) Nanodisc consisting of lipids (blue) and the amphipathic membrane scaffolding protein (MSP) which surrounds the disc as a ring (red circles). (c) Membrane protein reconstituted in amphipols (orange). The amphipathic polymer contains carboxylate groups (a) which can carry hydrophobic octylamine groups (b) or hydrophilic isopropylamine groups (c)

the size limiting “scaffold” would lead to large polydisperse mixture of liposomes. The use of lipid-based systems is one step closer to the native membrane compared to detergent micelles. Detergent micelles may well introduce a structural bias in the membrane protein [43], because of the detergent being an artificial replacement of the lipid bilayer in which the membrane proteins naturally reside, and because of the curvature of the micelle [52]. Lipid-based systems have been used successfully in structural biology, and they have also proven useful in the application of native MS [39, 43, 53]. Nanodiscs are rather monodisperse in size (Fig. 4), with different types available in the range of 9.8–17 nm outer diameter [54, 55] and corresponding lipid bilayer areas of 4400 A˚2 (MSP1D1) to 8900 A˚2 (MSP1E3D1). This somewhat limits the acceptable size of membrane proteins that can be embedded in nanodiscs though. The necessary disc size can be predicted by the number of trans-membrane helices which ˚ 2 each [56]. occupy an area of approximately 140 A Bicelles do not have a scaffold protein to define their diameter, which ranges from 20 to 40 nm [57]. Bicelle sizes are linked to the q-factor, which is the ratio between lipid and detergent, or lipid and short-chain lipids, in solution. The bicelle size increases when the relative amount of (long-tail) lipids increases compared to short chain [58]. When the concentration of bicelles in solution becomes

Mass Spectrometry of Membrane Protein Samples

215

too high, small bicelles will form, resulting in unwanted gel formation at room temperature. A very promising and recent approach for the reconstitution of native membranes uses so-called SMALPS (styrene maleic acid lipid particles), which use a specific variant of nanodisc-forming amphipathic polymer developed in order to extract pieces of bilayer membrane “cookie-cutter” style together with the incorporated membrane protein [59]. Their size distribution is however more polydisperse than for classic nanodiscs, which makes them less suitable for native MS. Amphipols are amphipathic polymers that wrap around the hydrophobic regions of the membrane protein (complex) (Fig. 4). Using MS it was shown that amphipols can solubilize IMPs in solution, while maintaining their structural integrity and providing the protection for the transition into the gas phase [43, 60]. They seem to work best however for monomeric IMPs, since complexes were not detected yet by native MS [43]. The most used amphipol is the anionic A8-35, a polyacrylate polymer that has octyl (hydrophobic) and isopropyl (hydrophilic) side chains (Fig. 4). The A8 stands for the average molecular weight of 8 kDa, and the 35 stands for the percentage of underivatized carboxylate groups. The rest of the repeating groups carry octylamine chains (25%) or are derivatized with isopropylamine (40%) [61]. These three groups form the long amphipathic poly-chain amphipol that wraps around the hydrophobic region of the IMPs. 2.3 Native MS of Membrane Proteins

Analyzing membrane proteins by native MS also makes some demands on the instrumentation and how the experiments are performed, which can be best described by following the sample through the different sections of the mass spectrometer (Fig. 5). As already mentioned in the chapter, nano-ESI is the method of choice for native MS of both soluble and membrane proteins. It has the advantage that only small amounts of sample are needed for a single measurement, as little as 1 μL [62]. For nano-ESI, an automated sample inlet system such as the Nanomate (Advion, Ithaca/ NY, USA) is used, or alternatively glass capillaries with thin metal coating or an inserted metal wire. These capillaries can be either bought or produced in-house with a capillary puller, e.g., Sutter Instrument p97 (Novato/CA, USA). The thin metal layer (typically gold) is typically applied with a sputter coater as they are also used for electron microscopy. To generate ions, a capillary voltage of typically between 1.2 and 1.6 kV is used, which causes an electric current that runs from the tip of the spray emitter through the air to the entrance aperture of the MS. Sometimes, it is also necessary to apply some backing pressure (