Heterologous Expression of Membrane Proteins: Methods and Protocols (Methods in Molecular Biology, Vol 601)

Methods

in

Molecular Biology™

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



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

Heterologous Expression of Membrane Proteins Methods and Protocols

Edited by

Isabelle Mus-Veteau Institut of Developmental Biology and Cancer, UMR CNRS 6543 Université de Nice-Sophia Antipolis, Parc Valrose, Nice, France

Editor Isabelle Mus-Veteau, Ph.D. Institut of Developmental Biology and Cancer, UMR CNRS 6543 Université de Nice-Sophia Antipolis Parc Valrose, Nice France [email protected]

ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-60761-343-5 e-ISBN 978-1-60761-344-2 DOI 10.1007/978-1-60761-344-2 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2009932113 © Humana Press, a part of Springer Science+Business Media, LLC 2010 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or ­dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, ­neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Cover illustration: The background art is derived from Figure 5 in Chapter 12. Printed on acid-free paper Humana Press is a part of Springer Science+Business Media (www.springer.com)

Preface Membrane proteins account for roughly 30% of all open reading frames in fully sequenced genomes. These proteins are of central importance to living cells. They are required for transport processes, sensing changes in the cellular environment, transmission of signals, and control of cell–cell contacts. These proteins are implicated in numerous pathologies, like cancer, cystic fibrosis, epilepsy, hyperinsulinism, heart failure, hypertension and Alzheimer disease, but studies of these and other disorders are hampered by a lack of information about the proteins involved. Knowing the structure of membrane proteins is an essential prerequisite for understanding how these proteins function and, further, how their functions can be modified by small molecules. This is of paramount importance in the pharmaceutical industry, which produces many drugs that bind to membrane proteins (50% of all drug targets are G protein-coupled receptors [GPCRs]), and recognizes the potential of many recently identified GPCRs, ion channels and transporters as targets for future drugs. However, whereas high-resolution structures are available for myriad soluble proteins (more than 42,000 in the Protein Data Bank), three-dimensional (3D) structures have so far been obtained for only 204 membrane proteins, the majority of which are from prokaryotic organisms, with only 25 from mammalian membrane proteins (see http:// blanco.biomol.uci.edu/Membrane_Proteins_xtal.html). The first membrane proteins were crystallized owing to their natural abundance, circumventing all the difficulties associated with overexpression. However, the majority of medically and pharmaceutically relevant membrane proteins are present in tissues at very low concentration, making overexpression of recombinant membrane proteins in heterologous cells suitable for largescale production a prerequisite for structural studies. The development of heterologous expression systems capable of delivering high-quality recombinant protein for structural studies remains an essential goal. In 2005, the two first atomic structures of recombinant mammalian membrane proteins expressed in yeast were resolved. Since that time, extensive optimization of heterologous expression systems has begun to bear fruit, and 11 eukaryotic membrane protein structures have been solved to high resolution using recombinant material. This volume proposes an overview of different heterologous expression systems that produce an adequate amount of membrane proteins for structural analysis. Methods and protocols are described for each heterologous expression system proposed. Some chapters of this volume treat membrane protein solubilization, purification and instability in solution and comment on the strategies that allowed the determination of the structure of the first heterologously expressed mammalian membrane proteins. Isabelle Mus-Veteau Nice, France

v

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

v ix

  1  Heterologous Expression of Membrane Proteins for Structural Analysis . . . . . . . . Isabelle Mus-Veteau   2  Production of Membrane Proteins in Escherichia coli and Lactococcus lactis . . . . . . Eric R. Geertsma and Bert Poolman

1

  3  Mammalian Membrane Receptors Expression as Inclusion Bodies in Escherichia coli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bernard Mouillac and Jean-Louis Banères   4  Expression of Membrane Proteins at the Escherichia coli Membrane for Structural Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Manuela Zoonens and Bruno Miroux   5  Membrane Protein Expression in Lactococcus lactis . . . . . . . . . . . . . . . . . . . . . . . . . Annie Frelet-Barrand, Sylvain Boutigny, Edmund R.S. Kunji, and Norbert Rolland   6  Heterologous Expression of Human Membrane Receptors in the Yeast Saccharomyces cerevisiae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Olivier Joubert, Rony Nehmé, Michel Bidet, and Isabelle Mus-Veteau   7  Mammalian Membrane Protein Expression in Baculovirus-Infected Insect Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Céline Trometer and Pierre Falson   8  Expression of membrane proteins in Drosophila melanogaster S2 Cells: Production and Analysis of a EGFP-Fused G Protein-Coupled Receptor as a Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Karl Brillet, Carlos A. Pereira, and Renaud Wagner   9  Membrane Protein Expression in the Eyes of Transgenic Flies . . . . . . . . . . . . . . . Valérie Panneels and Irmgard Sinning 10  Expression of Mammalian Membrane Proteins in Mammalian Cells Using Semliki Forest Virus Vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kenneth Lundstrom 11  Membrane Protein Expression in Cell-Free Systems . . . . . . . . . . . . . . . . . . . . . . . Birgit Schneider, Friederike Junge, Vladimir A. Shirokov, Florian Durst, Daniel Schwarz, Volker Dötsch, and Frank Bernhard 12  Practical Considerations of Membrane Protein Instability during Purification and Crystallisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Christopher G. Tate 13  Membrane Protein Solubilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Katia Duquesne and James N. Sturgis

vii

17

39

49 67

87

105

119 135

149 165

187 205

viii

Contents

14  Amphipols and Fluorinated Surfactants: Two Alternatives to Detergents for Studying Membrane Proteins In vitro . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Cécile Breyton, Bernard Pucci, and Jean-Luc Popot 15  Heterologous Expression and Affinity Purification of Eukaryotic Membrane Proteins in View of Functional and Structural Studies: The Example of the Sarcoplasmic Reticulum Ca2+-ATPase . . . . . . . . . . . . . . . . . . 247 Delphine Cardi, Cédric Montigny, Bertrand Arnou, Marie Jidenko, Estelle Marchal, Marc Le Maire, and Christine Jaxel Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269

Contributors Bertrand Arnou  •  CEA, iBiTecS – Institut de Biologie et Technologies de Saclay, University of Paris-Sud, Gif-sur-Yvette, France Jean-Louis Banères  •  Institut des Biomolécules Max Mousseron, Université Montpellier I, Université Montpellier II, Faculté de Pharmacie, Montpellier, France Frank Bernhard  •  Centre for Biomolecular Magnetic Resonance, University of Frankfurt/Main, Institute for Biophysical Chemistry, Frankfurt/Main, Germany Michel Bidet  •  Institut of Developmental Biology and Cancer, UMR CNRS 6543, Université de Nice-Sophia Antipolis, Parc Valrose, Nice, France Sylvain Boutigny  •  Laboratoire de Physiologie Cellulaire Végétale, Grenoble, France; and CEA, DSV, iRTSV, Grenoble, France; INRA, Grenoble, France; Université Joseph Fourier, Grenoble, France Cécile Breyton  •  Institut de Biologie Structurale, UMR 5075 CNRS/CEA/UJF, 41, rue Jules Horowitz, 38027 Grenoble France Karl Brillet  •  Dpt Récepteurs et des Protéines Membranaires, Illkirch, France Delphine Cardi  •  CEA, iBiTecS – Institut de Biologie et Technologies de Saclay, University of Paris-Sud, Gif-sur-Yvette, France Volker Dötsch  •  Centre for Biomolecular Magnetic Resonance, University of Frankfurt/Main, Institute for Biophysical Chemistry, Frankfurt/Main, Germany Katia Duquesne  •  Laboratoire d’Ingenierie des Systèmes Maromoléculaire, Institut de Biologie Structurale et Microbiologie, CNRS and Aix Marseille Université, Marseille, France Florian Durst  •  Centre for Biomolecular Magnetic Resonance, University of Frankfurt/Main, Institute for Biophysical Chemistry, Frankfurt/Main, Germany Pierre Falson  •  Laboratoire des Protéines de Résistance aux Agents Chimiothérapeutiques (LPRAC), Institut de Biologie et Chimie des Protéines (IBCP), Unité Mixte du Centre National de la Recherche Scientifique (CNRS) et de l’Université Lyon I, Institut Fédératif de Recherches Lyon, France Annie Frellet-Barrand  •  Institut de Biologie Structurale, UMR 5075 CNRS/CEA/ UJF, 41, rue Jules Horowitz, 38027 Grenoble France Eric R. Geertsma  •  Department of Biochemistry, Groningen Biomolecular Sciences and Biotechnology Institute (GBB), Netherlands Proteomics Centre (NPC), and Zernike Institute for Advanced Materials, University of Groningen, Groningen, The Netherlands Christine Jaxel  •  CEA, iBiTecS – Institut de Biologie et Technologies de Saclay, University of Paris-Sud, Gif-sur-Yvette, France Marie Jidenko  •  CEA, iBiTecS – Institut de Biologie et Technologies de Saclay, University of Paris-Sud, Gif-sur-Yvette, France Olivier Joubert  •  Université Henri Poincaré- Nancy 1, Faculté de Pharmacie, 5 rue Albert Lebrun, BP 80403, F-54001 NANCY CEDEX Friederike Junge  •  Centre for Biomolecular Magnetic Resonance, University of Frankfurt/Main, Institute for Biophysical Chemistry, Frankfurt/Main, Germany ix

x

Contributors

Edmund R.S. Kunji  •  The Medical Research Council, Dunn Human Nutrition Unit, Cambridge, UK Marc le Maire  •  CEA, iBiTecS – Institut de Biologie et Technologies de Saclay, University of Paris-Sud, Gif-sur-Yvette, France Kenneth Lundstrom  •  PanTherapeutics, Lutry, Switzerland Estelle Marchal  •  CEA, iBiTecS – Institut de Biologie et Technologies de Saclay, University of Paris-Sud, Gif-sur-Yvette, France Bruno Miroux  •  Université Paris-7, Laboratoire de Biologie Physico-Chimique des Proteines Membranaires, Institut de Biologie Physico-Chimique, Paris, France Cédric Montigny  •  CEA, iBiTecS – Institut de Biologie et Technologies de Saclay, University of Paris-Sud, Gif-sur-Yvette, France Bernard Mouillac  •  Institut de Génomique Fonctionnelle, Montpellier, France; INSERM U661, Montpellier, France; Université Montpellier I, Montpellier, France; Université Montpellier II, Montpellier, France Isabelle Mus-Veteau  •  Institut of Developmental Biology and Cancer, UMR CNRS 6543, Université de Nice-Sophia Antipolis, Parc Valrose, Nice, France Rony Nehmé  •  Institut of Developmental Biology and Cancer, UMR CNRS 6543, Université de Nice-Sophia Antipolis, Parc Valrose, Nice, France Carlos A. Pereira  •  Laboratório de Imunologia Viral, Instituto Butantan, São Paulo, Brasil Valérie Panneels  •  Heidelberg University Biochemistry Center (BZH), Heidelberg, Germany Bert Poolman  •  Department of Biochemistry, Groningen Biomolecular Sciences and Biotechnology Institute (GBB), Netherlands Proteomics Centre (NPC), and Zernike Institute for Advanced Materials, University of Groningen, Groningen, The Netherlands Jean-Luc Popot  •  Laboratoire de Physico-Chimie Moléculaire des Membranes Biologiques, CNRS and Université Paris-7, Institut de Biologie Physico-Chimique, Paris, France Bernard Pucci  •  Laboratoire de Chimie Bioorganique et des Systèmes Moléculaires Vectoriels, Université d’Avignon et des Pays de Vaucluse, Faculté des Sciences, Avignon, France Daniel Schwarz  •  Centre for Biomolecular Magnetic Resonance, University of Frankfurt/Main, Institute for Biophysical Chemistry, Frankfurt/Main, Germany Birgit Schneider  •  Centre for Biomolecular Magnetic Resonance, University of Frankfurt/Main, Institute for Biophysical Chemistry, Frankfurt/Main, Germany Vladimir A. Shirokov  •  Institute of Protein Research, Russian Academy of Sciences, Pushchino, Moscow Region, Russia Irmgard Sinning  •  Heidelberg University Biochemistry Center (BZH), Heidelberg, Germany James N. Sturgis  •  Laboratoire d’Ingenierie des Systèmes Maromoléculaire, Institut de Biologie Structurale et Microbiologie, CNRS and Aix Marseille Université, Marseille, France Christopher G. Tate  •  MRC Laboratory of Molecular Biology, Cambridge, UK Céline Trometer  •  Laboratoire des Protéines de Résistance aux Agents Chimiothérapeutiques (LPRAC), Institut de Biologie et Chimie des Protéies

Contributors

xi

(IBCP), Unité Mixte du Centre National de la Recherche Scientifique (CNRS) et de l’Université Lyon, Institut Fédératif de Recherches, Lyon, France Norbert Rolland  •  Laboratoire de Physiologie Cellulaire Végétale, CNRS, Grenoble, France; and CEA, DSV, iRTSVGrenoble, France; INRA, Grenoble, France; Université Joseph Fourier, Grenoble, France Renaud Wagner  •  Dpt Récepteurs et des Protéines Membranaires, Illkirc, France Manuela Zoonens  •  Université Paris-7, Laboratoire de Biologie Physico-Chimique des Proteines Membranaires, Institut de Biologie Physico-Chimique, Paris, France

Chapter 1 Heterologous Expression of Membrane Proteins for Structural Analysis Isabelle Mus-Veteau Abstract  Membrane proteins (MPs) are responsible for the interface between the exterior and the interior of the cell. These proteins are involved in numerous diseases, like cancer, cystic fibrosis, epilepsy, hyperinsulinism, heart failure, hypertension and Alzheimer disease. However, studies of these disorders are hampered by a lack of structural information about the proteins involved. Structural analysis requires large quantities of pure and active proteins. The majority of medically and pharmaceutically relevant MPs are present in tissues at low concentration, which makes heterologous expression in large-scale productionadapted cells a prerequisite for structural studies. Obtaining mammalian MP structural data depends on the development of methods that allow the production of large quantities of MPs. This review focuses on the heterologous expression systems now available to produce large amounts of MPs for structural proteomics, and describes the strategies that allowed the determination of the structure of the first heterologously expressed mammalian MPs. Key words:  Integral membrane proteins, heterologous expression systems, solubilization, stabilization, crystallization, 3D structure

1. Introduction Integral membrane proteins (MPs) account for roughly 30% of all open reading frames in fully sequenced genomes (1). These proteins are of central importance to living cells. They are required for transport processes, sensing changes in the cellular environment, transmission of signals, and control of cell–cell contacts. MPs are implicated in numerous pathologies like cancer, cystic fibrosis, epilepsy, hyperinsulinism, heart failure, hypertension and Alzheimer disease, but studies of these and other disorders are hampered by a lack of information about the proteins involved. Knowing the structure of MPs is an essential prerequisite for understanding how MPs function and, further, how their functions can be modified by small molecules. This is of paramount

I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_1, © Humana Press, a part of Springer Science + Business Media, LLC 2010

1

2

Mus-Veteau

importance in the pharmaceutical industry, which produces many drugs that bind to MPs (50% of all drug targets are G proteincoupled receptors (GPCRs) (2)), and recognizes the potential of many recently identified GPCRs, ion channels and transporters as targets for future drugs. However, whereas high-resolution structures are available for myriad soluble proteins (>42,000 in the NCBI database Protein Data Bank), three-dimensional (3D) structures have so far been obtained for only 204 integral MPs, the majority of which are from prokaryotic organisms, with only 25 from mammalian MPs (see http://blanco.biomol.uci.edu/ Membrane_Proteins_xtal.html). The first MPs were crystallized thanks to their natural abundance, circumventing all the difficulties associated with overexpression (3–6)). The majority of medically and pharmaceutically relevant MPs are present in tissues at low concentration, making overexpression of recombinant MPs in heterologous cells suitable for large-scale production a prerequisite for structural studies. The development of heterologous expression systems capable of delivering high-quality recombinant protein for structural studies remains an essential goal. In 2005, the two first atomic structures of recombinant mammalian MPs expressed in yeast were resolved (7, 8)). Extensive optimization of heterologous expression systems (see (9) for review) has begun to bear fruit, and 11 eukaryotic MPs structures have been solved to high resolution using recombinant material (10–21). This review introduces the different heterologous expression systems available to produce amounts of MPs adequate for structural analysis, for which detailed methodologies are described in the following chapters of this volume, and comments on the strategies that allowed the determination of the structure of the first heterologously expressed eukaryotic MPs.

2. Heterologous Expression Systems 2.1. Bacteria 2.1.1. Prokaryotic Membrane Protein Expression in Bacteria

Due to the development of efficient genetic tools, simplicity and low cultivation cost, Escherichia coli is considered the most popular host for the overproduction of MPs. However, E. coli often has difficulties with the expression, folding, stability and assembly of eukaryotic MPs. A convenient strategy to obtain structural and functional information on a certain MP that is not abundant in any natural source involves the screening of prokaryotic homologues, which can often be expressed in bacteria in large quantities. To rapidly establish an over-expression system, several homologues, types and location of affinity tags necessary for purification and expression strains need to be screened. For this, high-throughput procedures have been developed. The large number of plasmids required for this strategy implicates the development of efficient

Heterologous Expression of Membrane Proteins for Structural Analysis

3

cloning procedures such as ligation-independent cloning (LIC) (22). Following initial screening for expression, systematic optimization of the production conditions to increase the amount of well-folded protein inserted into the membrane is often required. This process, which involves time-consuming steps such as the isolation of membrane vesicles and activity assays, has been greatly accelerated by the use of green fluorescent protein (GFP) fused to the C-terminal end of the target protein. Since proper maturation of GFP in E. coli depends on the correct folding of its fusion partner, only correct folding of the target protein leads to a fluorescent fusion protein (23). Geertsma and Poolman detail in Chapter 2 of this volume methodology to rapidly establish and optimize prokaryotic MP expression in E. coli and Lactococcus lactis. This method cannot easily be extended to mammalian MPs because these proteins are mostly expressed in inclusion bodies (IBs) in E. coli, from which they are usually impossible to purify under non-denaturing conditions. Therefore, other bacteria like L. lactis, or derivatives of E. coli BL21 (DE3) selected to grow to high-saturation cell density and to overproduce proteins without toxic effect for the host cell, have been successfully used for production of some eukaryotic MPs. 2.1.2. Production of Eucaryotic Membrane Proteins in E. coli Inclusion Bodies

One phenomenon frequently observed in E. coli is that many heterologous proteins become incorrectly folded and accumulate in the cytoplasm as insoluble aggregates called inclusion bodies (IBs). As described by Margreiter and co-workers (24), IBs are observed as refractive particles either in the cytoplasm or in the periplasmic space if the heterologous protein is engineered for secretion. They can grow larger than 1 mm in diameter, a large proportion of a single bacterial cell, and thus are clearly visible under a microscope. IBs contain mostly recombinant protein (up to 99%) but also chaperones and membrane fragments. Because IBs are usually mechanically stable, they can be isolated from cells and separated by centrifugation. In many cases, IB formation strongly boosts MP production up to several thousand-fold without toxicity for bacteria, which is particularly interesting for structural approaches. MPs accumulated in IBs are protected against degradation by proteolysis. However, to recover the entrapped recombinant proteins, IBs must be dissolved by highly concentrated chaotropic agents like 8M urea or 6M guanidine hydrochloride. The bottleneck of this strategy is to establish efficient refolding protocols after purification to obtain pure functional MPs. Several MPs have been successfully refolded after purification from IBs: the light-harvesting complex (LHC) (25), diacylglycerol kinase (DAGK) (26), and the leukotriene B4 receptor BLT1 (27, 28). In Chapter 3 of this volume, Mouillac and Banères describe general trends for expression and purification of membrane receptors from E. coli IBs, as well as a general screening assay for refolding conditions.

4

Mus-Veteau

2.1.3. E. coli Strains for Functional Expression of Eukaryotic Membrane Proteins

Aggregation of foreign MPs in IBs can be reduced and targeting to the membrane favoured by using low-copy-number plasmids with weak promoters and fusion proteins between the MP of interest and a partner known to target the membrane, optimizing the cell growth conditions and selecting host mutants resistant to toxicity. MP expression at the E. coli membrane can be increased in specific mutant strains. C41 and C43 (DE3) are mutant hosts that allow synthesis of some MPs at high levels (29). Expression of certain MPs in the C43 (DE3) strain appears to be accompanied by proliferation of intracellular membrane, providing more space for MP insertion (30). Proteolysis of MPs can be decreased using BL21 derivative strains with low proteolytic activity. Zoonens and Miroux describe in Chapter 4 of this volume a protocol for mammalian MP expression at the E. coli membrane.

2.1.4. Eukaryotic Membrane Protein Expression in L. lactis

Lactococcus lactis is a gram-positive lactic bacterium traditionally used for the production of lactic acid in fermented food products. Genetic engineering tools and molecular characterization have been developed on Lactococcus (31–33), and now this bacterium is widely used in biotechnology for large-scale overproduction of heterologously expressed proteins. As an expression host, L. lactis presents advantages over E. coli: It is a food-grade organism, does not produce endotoxins, possesses low protease activity, does not form IBs, and has only one membrane. In 2005, Kunji and coworkers reported the overproduction of 11 mitochondrial transport proteins from yeast in a functional state in the L. lactis cytoplasmic membrane (34). A. Frelet-Barrand and collaborators report in Chapter 5 of this volume detailed protocol for the expression of MPs in L. lactis using the nisin-inducible controlled gene expression (NICE) system, which allows fine control of gene  expression based on the auto-regulation mechanism of the ­bacteriocin nisin. Despite the advantages offered by bacteria, success rates of eukaryotic MP expression for structural proteomics in E. coli are between 10% and 30% compared with 40–70% in eukaryotic systems (2, 35).

2.2. Yeast

This simple eukaryotic cell is a multi-purpose host, performing many of the post-translational modifications seen in higher eukaryotic cells (glycosylation, disulphide bond formation and proteolytic processing), combined with the ease of growing a large volume of cells in a short period of time. Cloning, functional expression, identification of interacting partners, mutagenesis of the protein or its partners and overexpression for biophysical analysis can all be achieved in yeast (36). However, yeast glycosylation patterns and lipid composition of membranes (no endogenous cholesterol) do differ from those of mammalian cells, and this can have important consequences on activity, folding, trafficking,

Heterologous Expression of Membrane Proteins for Structural Analysis

5

stability and structural studies of MPs (8, 19, 37, 38). Exciting possibilities also exist for the exploitation of yeast genetics to tailor yeast strains so that they are optimized for the expression of mammalian MPs, like yeast with humanized glycosylation machineries that have become available recently (39). Two yeast species have so far been used successfully for heterologous expression of mammalian MPs: Saccharomyces cerevisiae and Pichia pastoris. From the 11 atomic structures of mammalian MPs derived from a heterologous overproduction source, 5 are derived from yeast: 1 from S. cerevisiae, the rabbit sarcoplasmic reticulum Ca2+-ATPase (SERCA1A) (7); 3 from P. pastoris, the rat potassium channel (8), the chimera potassium channel Kv1.2/Kv2.1 (15) and the human LTC4 synthase (16). The structure of the last protein was also obtained from its expression in S. Pombe (10). 2.2.1. Membrane Protein Expression in the Yeast S. cerevisiae

Saccharomyces cerevisiae has been successfully used to functionally express and purify several recombinant mammalian MPs, like the human P-glycoprotein (40), the rat vesicular monoamine transporter (41), the human receptor of the Hedgehog pathway (42), the Aquaporin AQP0 (43) and the rabbit SERCA1a Ca2+-ATPase (adenosine triphosphatase) that gave the first atomic structure of a heterologously expressed mammalian MP (7). A lot of work has been done on S. cerevisiae to understand the parameters involved in MP expression, and many strains are available. The expression vectors used are 2 m multicopy plasmids with the inducible GAL1 promoter or the strong constitutive promoter from the yeast plasma membrane ATPase (PMA1). Depending on the conditions, the human P-glycoprotein expression reaches 8% of total MPs using a proteasedeficient strain of S. cerevisiae, with a level of expression comparable to that of mammalian cells (COS cells and Human Embryonic Kidney 293 cells) at only a fraction of the cost, time and effort (40). It has been shown for several MPs that the functional expression in S. cerevisiae can be enhanced by modification of certain codons in the translation initiation region to codons preferred by S. cerevisiae and by reducing the temperature to 16°C (42, 44, 45). In Chapter 6 of this volume, Joubert and co-workers describe a protocol to functionally express human MPs in the yeast S. cerevisiae.

2.2.2. Membrane Protein Expression in the Yeast Pichia pastoris

The methylotrophic yeast P. pastoris has been used successfully as a tool for large-scale recombinant soluble protein production (46). Pichia pastoris is a poor fermenter, a major advantage relative to S. cerevisiae. In cultures with high cell density, ethanol (a product of S. cerevisiae fermentation) rapidly builds to toxic levels, limiting further growth and foreign protein production. With its preference for respiratory growth, P. pastoris can be cultured at extremely high densities (500 OD600 U/mL) in the controlled environment of a fermenter. Also, P. pastoris is less likely to hyperglycosylate recombinant MPs than S. cerevisiae and foreign genes are stably ­integrated

6

Mus-Veteau

in single or multiple copies behind the AOX1 (alcohol oxidase 1) promoter, one of the strongest and most tighly regulated promoters known. Pichia pastoris has been used to produce many eukaryotic MPs in high yield (47), including the rat potassium channel and the human LTC4 synthase used for atomic structure determination (8, 16), and even several GPCRs (42, 48). 2.3. Insect Cells

Although more expensive to culture than yeast, insect cells remain simpler to maintain in large-scale culture than mammalian cells. Furthermore, even if insect cells glycosylate poorly and essentially with mannose (49) and their membrane presents a low level of cholesterol, they offer a protein-processing machinery and membrane composition closer to those of mammalian cells than yeast (50).

2.3.1. Mammalian Membrane Protein Expression in Baculovirus Infected Insect Cells

The baculovirus expression vector system (BEVS) is based on replacement of a late, non-essential, viral gene coding for the polyhedron with a gene of interest. BEVS allows cloning and expression of recombinant proteins in insect cells of Lepidopterans such as Spodoptera frugiperda (Sf9, Sf21) and Trichoplusia ni (Hi5). Recombinant baculovirus and insect cells have been essayed for expression of several GPCRs (51), and were used to produce the b2- and the b1-adrenergic receptors employed to determine the X-ray structures (11, 18, 21). This system was also used to produce the aquaporin AQP4 protein employed in high-resolution electron crystallographic studies (13) and the acid-sensing ion channel 1 that gave the X-ray atomic structure (14). Trometer and Falson detail in Chapter 7 of this volume the experimental procedure they developed to produce a human MP, the breast cancer resistance protein (BCRP), by using the Bac-to-Bac baculovirus-based expression system.

2.3.2. Mammalian Membrane Protein Expression in Transiently or Stably Transfected Drosophila Schneider 2 Cells

The Drosophila Schneider 2 (S2) cells are immortalized nontumorigenic cells isolated from primary cultures of Drosophila melanogaster embryos (52). One specific characteristic of S2 cells is their ability to grow to high cell densities, as high as 3 × 107 cells/mL, which is tenfold higher than what can be attained with other insect cells, like Sf9 (53). Moreover, S2 cells can be transiently or stably transfected using a non-lytic plasmid-based system that provides additional benefits over virus-infected insect systems. S2 cells possess all the mammalian-like cell machineries for gene expression, protein processing and trafficking, including posttranslational modifications that may be critical for proper maturation, localization and function of target protein (54). Although not yet fully explored and poorly used, S2 cells have been shown to be suitable hosts for the expression of large amounts of several recombinant mammalian MPs like GPCRs (55–59). In Chapter 8 of this volume, Brillet and co-workers present a series of protocols allowing handling S2 cell expression system through the example of a human GPCR fused to the enhanced GFP.

Heterologous Expression of Membrane Proteins for Structural Analysis

7

2.4. Drosophila Eyes

The Sinning lab developed the expression of recombinant MPs in the eyes of transgenic Drosophila melanogaster as a new, alternative method for overexpression. This idea was based on the observation that the rhodopsin, the first GPCR from which an X-ray structure was obtained in 2000 (3), is densely packed in retina membranes, where it is targeted to the specialized membrane stacks in the photoreceptor cells, termed rhabdomeres in Drosophila (60). Using developmental drivers (promoters) that activate the expression of specific genes at certain stages of development, it is possible to express MPs that localise to these specialized membranes. Expression in photoreceptor cells provides a number of advantages compared to conventional expression systems. In these membranes, MPs are protected from proteases. Internal membranes are reduced to a minimum, and about 90% of the membranes are plasma membrane. This avoids or minimizes the problem of proteolysis and of mistargeted or incompletely matured protein. Several GPCRs and transporters not only from Drosophila but also from mammals were successfully expressed in the fly eyes demonstrating the great potential of this method (61). In Chapter 9 of this volume, Panneels and Sinning present a protocol of expression of MPs in Drosophila eyes using GFP fusion constructs that allow monitoring expression, purification and localization of the proteins.

2.5. Mammalian Cells

Mammalian cells clearly offer the most native cellular environment for the expression of mammalian MPs. There are some mammalian MPs for which heterologous expression of functional protein is limited to mammalian cells in culture, as is the case with many GPCRs (62). For at least some recombinant mammalian MPs, mammalian cells provide the greatest amount of functional protein (2, 63, 64). But, generation of milligram quantities required for crystallization trials with these cells is expensive and labour-intensive and post-translational modifications like glycosylation can be highly heterogeneous. However, the first structure of a recombinantly produced GPCR was obtained from a rhodopsin mutant expressed in mammalian cells (19). Mammalian expression systems have relied on transiently transfected cells, the isolation of stable transformants or the use of viral vectors.

2.5.1. Production of Membrane Proteins in Stable or Transient Transformant Mammalian Cells

Efforts have focused on careful optimization of culture conditions and media components with respect to transfection efficiency and protein expression levels. As a result, together with the development of the serum-free media and of relatively cheap transfection reagents, large-scale transient transfection (LSTT) of mammalian cells has achieved growing acceptance as an alternative expression method (65). The transient expression of the recombinant rhodopsin in COS-1 cells allowed production of enough stable protein to determine the first atomic structure of a recombinantly produced GPCR (19).

8

Mus-Veteau

2.5.2. Production of Membrane Proteins in Mammalian Cells Infected by the Semliki Forest Virus

Among the most widely viral vectors used for MP expression in mammalian cells is the Semliki Forest virus (SFV). The broad host range of SFV allows transduction of a large number of mammalian cell lines and primary cell cultures. Expression of GPCRs and ion channels from SFV vectors has provided pharmacologically functional MPs, has demonstrated coupling of G proteins to GPCRs (66) and has shown electrophysiological responses to the ion channel (67). Efficient infection of both adherent and suspension cultures of mammalian cells has allowed scale-up of recombinant protein production in spinner and roller flasks as well as in bioreactors, which has allowed the production of hundreds of milligrams of recombinant MPs (2, 68, 69). In Chapter 10 of this volume, Lundstrom describes a protocol for expression of recombinant mammalian MPs in mammalian host cells applying SFV vectors.

2.6. Cell-Free System

The cell-free expression system mimics the natural cytoplasmic cell environment yet it remains independent from the requirements and sensitivity of living cells. The tolerance of the cell-free expression system for a wide variety of supplements offers a wealth of perspectives for the high-level production of proteins that would be difficult or even impossible to obtain in cellular hosts, like some mammalian integral MPs. In principle, almost any compounds that might be beneficial for the stabilization or the folding of a recombinant protein like protease inhibitors, cofactors, substrates or any kind of ligands can be added directly into the cell-free system. Problems of toxicity, instability or protein folding can therefore be specifically addressed in many cases. In particular, the addition of detergents into the cell-free system offers the possibility to synthesize integral MPs by providing an artificial hydrophobic environment that prevents the aggregation of freshly translated MPs and supports their functional folding in the detergent micelles (70–74). This strategy was successful for several integral MPs, including GPCRs (75–78). Liposomes can also be added in the cell-free reaction allowing the synthesized MPs to directly integrate into the lipidic bilayer (79–81). In Chapter 11 of this volume, Schneider and co-workers describe the basic protocols for the general production of MPs in a cell-free expression system using E. coli extracts.

3. Toward Mammalian Membrane Protein Crystallization

Crystallization requires pure MP that is chemically and conformationally homogeneous, stable at relatively high concentrations for a long period and available in milligram amounts.

Heterologous Expression of Membrane Proteins for Structural Analysis

9

3.1. Membrane Protein Solubilization and Stability

Purification of MPs requires the use of detergents to extract them from the membrane. MPs are then in aqueous solution in complex with detergents and lipids. Finding a suitable detergent and conditions that ensure protein homogeneity, functionality, stability and crystallization is a limiting and crucial step. Strategies have been developed to facilitate the purification of MPs in a functional form leading to new possibilities for structure determination. These strategies allowed the resolution of the three first atomic structures derived from recombinant GPCRs. Tate reviews these strategies in Chapter 12 of this volume. Scaffolds of the crystal lattice are found predominantly between the exposed, polar surfaces of proteins, while the transmembrane parts remain buried from the detergent micelle. Proteins with large extra-membranous domains are favoured, and detergents that assemble into small micelles, such as octyl-b-d-glucopyranoside (OG) or dimetyldodecylamineoxide (LDAO) are preferred in crystal trials (82). However, dodecyl-b-d-maltoside (DDM), a detergent that forms large micelles, has given rise to some structures, such as those of the acid-sensing ion channel 1 (14) or the human LTC4 synthase (16). In Chapter 13 of this volume, Duquesne and Sturgis give a general protocol for optimizing MP solubilization and indicate the major parameters that are important for reproducibility.

3.2. New Surfactant Molecules to Stabilize Membrane Proteins in Solution

A detergent efficient for the solubilization of a MP that ensures functionality and stability but is not suitable for crystallization can be exchanged during or after purification by another more suitable for crystallization. Surfactant molecules that would be more suitable than classical detergents for MP integrity, functionality and stability, like non-ionic fluorinated surfactants or amphipols, are currently under development by Pucci and co-workers (83) and Popot and co-workers (84), respectively. These molecules do not solubilize biological membranes but are able to maintain MPs that have been extracted from membrane using classical detergents in solution after exchange on an affinity column, for example. These new surfactants have been efficient for MP stability in solution (85, 86), refolding (28) and MP expression in a cell-free system (78), but as yet no MP has been crystallized in these surfactants. Breyton and co-workers describe in Chapter 14 of this volume the protocols they developed to obtain stable MPs in fluorinated surfactants and in amphipols.

3.3. Membrane Protein Purification

The purification strategy and protocol are crucial for the suitability of MPs for structural studies. The choice and the position of the affinity tags; the presence of the MP ligand, agonist or antagonist; the choice of the detergent that must be efficient for solubilization, must preserve MP integrity and must be suitable for crystallization; the possibility of detergent exchanges and the addition

10

Mus-Veteau

of lipids are central determinants. The purification protocol must preserve the MP integrity and homogeneity. Affinity purification allows efficient and fast purification in a few steps. This can prevent the complete removal of native lipids bound to the protein and preserve MP integrity. However, this requires expressing the target recombinant protein with one or several purification tags in its N- or C- terminus. The presence of these tags may have an effect on expression levels and on the conformational state of the protein and may affect protein crystallization (87). The incorporation of proteolytic cleavage sites that permit downstream removal of the tag can be important. The availability of a test that allows following the conformational state of the target MP during purification, such as ligand-binding measurements, is essential to establish the best purification conditions. Cardi and co-workers describe in the last chapter of this volume (Chapter 15) the protocol used to express and purify the recombinant sarcoplasmic reticulum Ca2+-ATPase from which the atomic structure was resolved.

4. The Atomic Structures of Mammalian Membrane Proteins Derived from Heterologous Expression Systems

The rabbit sarcoplasmic reticulum Ca2+-ATPase (SERCA1a) is the first successful crystallization of mammalian MP derived from a heterologous expression system. The recombinant protein was expressed in the yeast S. cerevisiae, with a Biotin Acceptor Domain (BAD) in its C-terminus. The recombinant SERCA1a was solubilized in DDM and purified by affinity chromatography using a streptavidin-sepharose resin followed by size exclusion chromatography on which the detergent DDM was exchanged with C12E8 more suitable for crystallization. The recombinant SERCA1a crystallized in a form that is isomorphous to the native SERCA1a protein from rabbit, and the diffraction properties were similar (7). This result is important since it demonstrates that a recombinant MP produced in a heterologous expression system can adopt the same structure as the native one. At the same time, the crystal structure of a recombinant mammalian voltage-dependent potassium channel (Kv1.2) in complex with an oxidoreductase b2 regulatory subunit was solved at 2.9-Å resolution (8). Kv1.2 was mutated to eliminate the glycosylation site in the S1–S2 linker. Both proteins were modified in their terminal extremity with a polyhistidine tag and co-expressed in the yeast P. pastoris. Proteins were purified in a mixture of lipids and detergent DDM using metal affinity chromatography followed by gel filtration. The crystal structure revealed that the pore of Kv1.2 is similar to that of prokaryotic K+ channels. In 2007, Long and co-workers crystallized a chimaeric voltage-dependent K+ channel in complex with lipids. The chimera

Heterologous Expression of Membrane Proteins for Structural Analysis

11

was expressed in the yeast P. pastoris with a ten-histidine tag in its N-terminus and purified in a mixture of detergents and lipids using metal affinity chromatography followed by gel filtration. The complete structure at 2.4-Å resolution revealed the pore and voltage sensors embedded in a membrane-like arrangement of lipid molecules (15). The crystal structure of the human leukotriene C4 synthase in its apo and glutathione-complexed forms to 2.00- and 2.15-Å resolution, respectively, was also derived from heterologous expression in the yeast P. pastoris (16). The recombinant protein was purified using two affinity chromatography steps followed by a gel filtration in which the detergents used for membrane solubilization (Triton X-100 and sodium deoxycholate) were exchanged for DDM. The N-terminus hexahistidine tag used for purification has not been eliminated and is clearly visible in the structure. The same protein with a hexahistidine tag in the C-terminus was produced in the fission yeast S. pombe and provided a structure in complex with glutathione at 3.3-Å resolution (10). Structures derived from both expression systems are comparable. The structure of a MP from the same family, the human 5-lipoxygenase-activating protein (FLAP) modified with a polyhistidine tag in the C-terminus, was derived from the recombinant protein expressed in E. coli (12). This is the only structure of human PM derived from bacteria. Schertler and collaborators determined in 2007 the first structure of a recombinant GPCR transiently expressed in mammalian COS cells: the rhodopsin mutant N2C/D282C. The mutant was designed to form a disulfide bond between the N terminus and loop E3, which allows handling of opsin in detergent solution and increases the thermal stability of rhodopsin. Recombinant rhodopsin was solubilized in DDM and purified using antibody resin. DDM was exchanged for C8E4 on gel filtration, and the recombinant protein was further purified using an ion-exchange column (19). The second structure of a recombinant GPCR, the human b2-adrenergic receptor, derives from recombinant proteins expressed in baculovirus-infected insect cells. In the first strategy, the human b2-adrenergic receptor was produced in a complex with an antibody fragment, solubilized in DDM and purified by sequential antibody and ligand affinity chromatography (18). A higher-resolution structure (2.4 Å) was obtained from an engineered human b2-adrenergic receptor fused in the middle of the third cytoplasmic loop to T4 lysozyme, solubilized in DDM and purified by sequential antibody and ligand affinity chromatography (11). Together with Tate, Schertler reported the 2.7-Å resolution structure of a mutant b1-adrenergic receptor in complex with the high-affinity antagonist cyanopindolol. The turkey receptor was modified by limited mutagenesis to improve thermostability and

12

Mus-Veteau

expressed in baculovirus-infected insect cells. After solubilization in decylmaltoside, the recombinant protein was purified by affinity chromatography thanks to the hexahistidine tag added at its C-terminus, and decylmaltoside was exchanged for octylglucoside for crystallization (21). To obtain crystals allowing the structure determination at 1.9-Å resolution of acid-sensing ion channel 1 (ASIC), Jasti and collaborators (14) produced in baculovirus-infected Sf9 cells a mutant recombinant protein that possesses a deletion of 25 N-terminal and 64 C-terminal residues. Recombinant ASIC was purified in DDM using metal and size exclusion chromatography.

5. Conclusion This review presents an overview of the heterologous expression systems that allow overproduction of MPs for structural analysis. The two first structures of mammalian MPs derived from heterologous expression were solved in 2005. Since that time, nine new structures have been obtained from recombinant mammalian MPs heterologously expressed. From the 11 recombinant mammalian MP atomic structures obtained since 2005, 5 were derived from MPs expressed in yeast, 4 from MPs expressed in insect cells, 1 from mammalian cells and 1 from bacteria. These recent structures of mammalian MPs were obtained thanks to the use of recombinant proteins that enabled the exploration of mutants designed to improve crystal quality.

Acknowledgments I thank M. Bidet, O. Joubert and R. Nehmé for critical reading of the manuscript and the European Commission through the FP6 specific targeted research project Innovative Tools for Membrane Structural Proteomics (LSH-2003-1.1.0-1) that financially supported several authors of this volume for their research on heterologous expression systems, new surfactant molecules and MP stabilization. References 1. Liu J, Rost B (2001) Comparing function and structure between entire proteomes. Protein Sci 10:1970–1979 2. Lundstrom K (2006) Structural genomics for membrane proteins. Cell Mol Life Sci 63:2597–2607

3. Palczewski K, Kumasaka T, Hori T et al (2000) Crystal structure of rhodopsin: a G proteincoupled receptor. Science 289:739–745 4. Stock D, Leslie AG, Walker JE (1999) Molecular architecture of the rotary motor in ATP synthase. Science 286:1700–1705

Heterologous Expression of Membrane Proteins for Structural Analysis 5. Toyoshima C, Nakasako M, Nomura H, Ogawa H (2000) Crystal structure of the calcium pump of sarcoplasmic reticulum at 2.6 A resolution. Nature 405:647–655 6. Unwin N (2005) Refined structure of the nicotinic acetylcholine receptor at 4A resolution. J Mol Biol 346:967–989 7. Jidenko M, Nielsen RC, Sorensen TL et  al (2005) Crystallization of a mammalian membrane protein overexpressed in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A 102: 11687–11691 8. Long SB, Campbell EB, Mackinnon R (2005) Crystal structure of a mammalian voltagedependent Shaker family K+ channel. Science 309:897–903 9. Junge F, Schneider B, Reckel S, Schwarz D, Dotsch V, Bernhard F (2008) Large-scale production of functional membrane proteins. Cell Mol Life Sci 65:1729–1755 10. Ago H, Kanaoka Y, Irikura D et  al (2007) Crystal structure of a human membrane protein involved in cysteinyl leukotriene biosynthesis. Nature 448:609–612 11. Cherezov V, Rosenbaum DM, Hanson MA et al (2007) High-resolution crystal structure of an engineered human beta2-adrenergic G protein-coupled receptor. Science 318: 1258–1265 12. Ferguson AD, McKeever BM, Xu S et  al (2007) Crystal structure of inhibitor-bound human 5-lipoxygenase-activating protein. Science 317:510–512 13. Hiroaki Y, Tani K, Kamegawa A et al (2006) Implications of the aquaporin-4 structure on array formation and cell adhesion. J Mol Biol 355:628–639 14. Jasti J, Furukawa H, Gonzales EB, Gouaux E (2007) Structure of acid-sensing ion channel 1 at 1.9 A resolution and low pH. Nature 449:316–323 15. Long SB, Tao X, Campbell EB, MacKinnon R (2007) Atomic structure of a voltage-dependent K+ channel in a lipid membrane-like environment. Nature 450:376–382 16. Martinez Molina D, Wetterholm A, Kohl A et  al (2007) Structural basis for synthesis of inflammatory mediators by human leukotriene C4 synthase. Nature 448:613–616 17. Pedersen BP, Buch-Pedersen MJ, Morth JP, Palmgren MG, Nissen P (2007) Crystal structure of the plasma membrane proton pump. Nature 450:1111–1114 18. Rasmussen SG, Choi HJ, Rosenbaum DM et  al (2007) Crystal structure of the human beta2 adrenergic G-protein-coupled receptor. Nature 450:383–387

13

19. Standfuss J, Xie G, Edwards PC, Burghammer M, Oprian DD, Schertler GF (2007) Crystal structure of a thermally stable rhodopsin mutant. J Mol Biol 372:1179–1188 20. Tornroth-Horsefield S, Gourdon P, Horsefield R et  al (2007) Crystal structure of AcrB in complex with a single transmembrane subunit reveals another twist. Structure 15:1663–1673 21. Warne T, Serrano-Vega MJ, Baker JG et al (2008) Structure of a beta1-adrenergic G-protein-coupled receptor. Nature 454: 486–491 22. Aslanidis C, de Jong PJ (1990) Ligationindependent cloning of PCR products (LICPCR). Nucleic Acids Res 18:6069–6074 23. Geertsma ER, Groeneveld M, Slotboom DJ, Poolman B (2008) Quality control of overexpressed membrane proteins. Proc Natl Acad Sci U S A 105:5722–5727 24. Margreiter G, Schwanninger M, Bayer K, Obinger C (2008) Impact of different cultivation and induction regimes on the structure of cytosolic inclusion bodies of TEM1-betalactamase. Biotechnol J 3:1245–1255 25. Rogl H, Kosemund K, Kuhlbrandt W, Collinson I (1998) Refolding of Escherichia coli produced membrane protein inclusion bodies immobilised by nickel chelating chromatography. FEBS Lett 432:21–26 26. Gorzelle BM, Nagy JK, Oxenoid K, Lonzer WL, Cafiso DS, Sanders CR (1999) Reconstitutive refolding of diacylglycerol kinase, an integral membrane protein. Biochemistry 38: 16373–16382 27. Baneres JL, Martin A, Hullot P, Girard JP, Rossi JC, Parello J (2003) Structure-based analysis of GPCR function: conformational adaptation of both agonist and receptor upon leukotriene B4 binding to recombinant BLT1. J Mol Biol 329:801–814 28. Damian M, Perino S, Polidori A et al (2007) New tensio-active molecules stabilize a human G protein-coupled receptor in solution. FEBS Lett 581:1944–1950 29. Miroux B, Walker JE (1996) Over-production of proteins in Escherichia coli: mutant hosts that allow synthesis of some membrane proteins and globular proteins at high levels. J Mol Biol 260:289–298 30. Arechaga I, Miroux B, Karrasch S et al (2000) Characterisation of new intracellular membranes in Escherichia coli accompanying large scale over-production of the b subunit of F(1) F(o) ATP synthase. FEBS Lett 482:215–219 31. Gasson MJ, de Vos WM (ed.) (1994) Genetic and Biotechnology of Lactic acid ­bacteria. Blackie Academic and Professional, London, UK.

14

Mus-Veteau

32. Teuber M, Geis A (2006) The genus Lactococcus, in the prokaryotes (Dworkin M, ed.), Springer-Verlag, New York 4:205–228 33. Wood BJB, Warner PJ (eds.) (2003) Genetics of lactic acid bacteria. Springer, New York 34. Monne M, Chan KW, Slotboom DJ, Kunji ER (2005) Functional expression of eukaryotic membrane proteins in Lactococcus lactis. Protein Sci 14:3048–3056 35. Aricescu AR, Assenberg R, Bill RM et  al (2006) Eukaryotic expression: developments for structural proteomics. Acta Crystallogr D Biol Crystallogr 62:1114–1124 36. Bill RM (2001) Yeast—a panacea for the structure-function analysis of membrane proteins? Curr Genet 40:157–171 37. Helenius A, Aebi M (2004) Roles of N-linked glycans in the endoplasmic reticulum. Annu Rev Biochem 73:1019–1049 38. Lee AG (2004) How lipids affect the activities of integral membrane proteins. Biochim Biophys Acta 1666:62–87 39. Hamilton SR, Davidson RC, Sethuraman N et al (2006) Humanization of yeast to produce complex terminally sialylated glycoproteins. Science 313:1441–1443 40. Figler RA, Omote H, Nakamoto RK, Al-Shawi MK (2000) Use of chemical chaperones in the yeast Saccharomyces cerevisiae to enhance heterologous membrane protein expression: high-yield expression and purification of human P-glycoprotein. Arch Biochem Biophys 376:34–46 41. Yelin R, Schuldiner S (2001) Vesicular monoamine transporters heterologously expressed in the yeast Saccharomyces cerevisiae display high-affinity tetrabenazine binding. Biochim Biophys Acta 1510:426–441 42. De Rivoyre M, Bonino F, Ruel L, Bidet M, Therond P, Mus-Veteau I (2005) Human receptor Smoothened, a mediator of Hedgehog signalling, expressed in its native conformation in yeast. FEBS Lett 579:1529–1533 43. Palanivelu DV, Kozono DE, Engel A et  al (2006) Co-axial association of recombinant eye lens aquaporin-0 observed in loosely packed 3D crystals. J Mol Biol 355:605–611 44. Lenoir G, Menguy T, Corre F et  al (2002) Overproduction in yeast and rapid and efficient purification of the rabbit SERCA1a Ca(2+)ATPase. Biochim Biophys Acta 1560:67–83 45. Pisani DF, Rivoyre MD, Ruel L et al (2005) Mouse myodulin, a new potential angiogenic factor, functionally expressed in yeast. Biochem Biophys Res Commun 331:552–556 46. Cereghino JL, Cregg JM (2000) Heterologous protein expression in the methylotrophic yeast

47.

48.

49.

50.

51.

52. 53. 54. 55.

56.

57.

58.

Pichia pastoris. FEMS Microbiol Rev 24: 45–66 Nyblom M, Oberg F, Lindkvist-Petersson K et al (2007) Exceptional overproduction of a functional human membrane protein. Protein Expr Purif 56:110–120 Andre N, Cherouati N, Prual C et al (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 Kost TA, Condreay JP, Jarvis DL (2005) Baculovirus as versatile vectors for protein expression in insect and mammalian cells. Nat Biotechnol 23:567–575 Luckow VA, Summers MD (1988) Signals important for high-level expression of foreign genes in Autographa californica nuclear polyhedrosis virus expression vectors. Virology 167:56–71 Akermoun M, Koglin M, Zvalova-Iooss D, Folschweiller N, Dowell SJ, Gearing KL (2005) Characterization of 16 human G protein-coupled receptors expressed in baculovirus-infected insect cells. Protein Expr Purif 44:65–74 Schneider I (1972) Cell lines derived from late embryonic stages of Drosophila melanogaster. J Embryol Exp Morphol 27:353–365 Sondergaard L (1996) Efficiency of different lipofection agents in Drosophila S-2 cells. In Vitro Cell Dev Biol Anim 32:386–387 Benting J, Lecat S, Zacchetti D, Simons K (2000) Protein expression in Drosophila Schneider cells. Anal Biochem 278:59–68 Brillet K, da Conceicao MM, Pattus F, Pereira CA (2006) Bioprocess parameters of cell growth and human mu opioid receptor expression in recombinant Drosophila S2 cell cultures in a bioreactor. Bioprocess Biosyst Eng 28:291–293 Brillet K, Perret BG, Klein V, Pattus F, Wagner R (2008) Using EGFP fusions to monitor the functional expression of GPCRs in the Drosophila Schneider 2 cells. Cytotechnology 57:101–109 De Rivoyre M, Ruel L, Varjosalo M et  al (2006) Human receptors patched and smoothened partially transduce hedgehog signal when expressed in Drosophila Cells. J Biol Chem 281:28584–28595 Perret BG, Wagner R, Lecat S et  al (2003) Expression of EGFP-amino-tagged human mu opioid receptor in Drosophila Schneider 2 cells: a potential expression system for large-scale production of G-protein coupled receptors. Protein Expr Purif 31:123–132

Heterologous Expression of Membrane Proteins for Structural Analysis 59. Schetz JA, Kim OJ, Sibley DR (2003) Pharmacological characterization of mammalian D1 and D2 dopamine receptors expressed in Drosophila Schneider-2 cells. J Recept Signal Transduct Res 23:99–109 60. Zuker CS (1996) The biology of vision of Drosophila. Proc Natl Acad Sci U S A 93:571–576 61. Eroglu C, Cronet P, Panneels V, Beaufils P, Sinning I (2002) Functional reconstitution of purified metabotropic glutamate receptor expressed in the fly eye. EMBO Rep 3:491–496 62. Sarramegna V, Muller I, Milon A, Talmont F (2006) Recombinant G protein-coupled receptors from expression to renaturation: a challenge towards structure. Cell Mol Life Sci 63:1149–1164 63. Eifler N, Duckely M, Sumanovski LT et  al (2007) Functional expression of mammalian receptors and membrane channels in different cells. J Struct Biol 159:179–193 64. Tate CG, Haase J, Baker C et  al (2003) Comparison of seven different heterologous protein expression systems for the production of the serotonin transporter. Biochim Biophys Acta 1610:141–153 65. Baldi L, Hacker DL, Adam M, Wurm FM (2007) Recombinant protein production by large-scale transient gene expression in mammalian cells: state of the art and future perspectives. Biotechnol Lett 29:677–684 66. Lundstrom K, Mills A, Buell G, Allet E, Adami N, Liljestrom P (1994) High-level expression of the human neurokinin-1 receptor in mammalian cell lines using the Semliki Forest virus expression system. Eur J Biochem 224:917–921 67. Hovius R, Tairi AP, Blasey H, Bernard A, Lundstrom K, Vogel H (1998) Characterization of a mouse serotonin 5-HT3 receptor purified from mammalian cells. J Neurochem 70:824–834 68. Hassaine G, Wagner R, Kempf J et al (2006) Semliki Forest virus vectors for overexpression of 101 G protein-coupled receptors in mammalian host cells. Protein Expr Purif 45: 343–351 69. Lundstrom K, Michel A, Blasey H et al (1997) Expression of ligand-gated ion channels with the Semliki forest virus expression system. J Recept Signal Transduct Res 17:115–126 70. Berrier C, Park KH, Abes S, Bibonne A, Betton JM, Ghazi A (2004) Cell-free synthesis of a functional ion channel in the absence of a membrane and in the presence of detergent. Biochemistry 43:12585–12591

15

71. Elbaz Y, Steiner-Mordoch S, Danieli T, Schuldiner S (2004) In vitro synthesis of fully functional EmrE, a multidrug transporter, and study of its oligomeric state. Proc Natl Acad Sci U S A 101:1519–1524 72. Ishihara G, Goto M, Saeki M et  al (2005) Expression of G protein coupled receptors in a cell-free translational system using detergents and thioredoxin-fusion vectors. Protein Expr Purif 41:27–37 73. Klammt C, Schwarz D, Fendler K, Haase W, Dotsch V, Bernhard F (2005) Evaluation of detergents for the soluble expression of alphahelical and beta-barrel-type integral membrane proteins by a preparative scale individual cell-free expression system. FEBS J 272:6024–6038 74. Klammt C, Schwarz D, Lohr F, Schneider B, Dotsch V, Bernhard F (2006) Cell-free expression as an emerging technique for the large scale production of integral membrane protein. FEBS J 273:4141–4153 75. Keller T, Schwarz D, Bernhard F et  al (2008) Cell free expression and functional reconstitution of eukaryotic drug transporters. Biochemistry 47:4552–4564 76. Klammt C, Schwarz D, Eifler N et al (2007) Cell-free production of G protein-coupled receptors for functional and structural studies. J Struct Biol 158:482–493 77. Liguori L, Marques B, Villegas-Mendez A, Rothe R, Lenormand JL (2007) Production of membrane proteins using cell-free expression systems. Expert Rev Proteomics 4:79–90 78. Park KH, Berrier C, Lebaupain F et al (2007) Fluorinated and hemifluorinated surfactants as alternatives to detergents for membrane protein cell-free synthesis. Biochem J 403:183–187 79. Kalmbach R, Chizhov I, Schumacher MC, Friedrich T, Bamberg E, Engelhard M (2007) Functional cell-free synthesis of a seven helix membrane protein: in situ insertion of bacteriorhodopsin into liposomes. J Mol Biol 371:639–648 80. Liguori L, Marques B, Villegas-Mendez A, Rothe R, Lenormand JL (2008) Liposomesmediated delivery of pro-apoptotic therapeutic membrane proteins. J Control Rel 126: 217–227 81. Nozawa A, Nanamiya H, Miyata T et  al (2007) A cell-free translation and proteoliposome reconstitution system for functional analysis of plant solute transporters. Plant Cell Physiol 48:1815–1820 82. Prive GG (2007) Detergents for the stabilization and crystallization of membrane proteins. Methods 41:388–397

16

Mus-Veteau

83. Polidori A, Presset M, Lebaupain F et al (2006) Fluorinated and hemifluorinated surfactants derived from maltose: synthesis and application to handling membrane proteins in aqueous solution. Bioorg Med Chem Lett 16:5827–5831 84. Tribet C, Audebert R, Popot JL (1996) Amphipols: polymers that keep membrane proteins soluble in aqueous solutions. Proc Natl Acad Sci U S A 93:15047–15050 85. Breyton C, Chabaud E, Chaudier Y, Pucci B, Popot JL (2004) Hemifluorinated surfactants:

a non-dissociating environment for handling membrane proteins in aqueous solutions? FEBS Lett 564:312–318 86. Chabaud E, Barthelemy P, Mora N, Popot JL, Pucci B (1998) Stabilization of integral membrane proteins in aqueous solution using fluorinated surfactants. Biochimie 80: 515–530 87. Tornroth-Horsefield S, Wang Y, Hedfalk K et  al (2006) Structural mechanism of plant aquaporin gating. Nature 439:688–694

Chapter 2 Production of Membrane Proteins in Escherichia coli and Lactococcus lactis Eric R. Geertsma and Bert Poolman Abstract As the equivalent to gatekeepers of the cell, membrane transport proteins perform a variety of critical functions. Progress on the functional and structural characterization of membrane proteins is slowed due to problems associated with their (heterologous) overexpression. Often, overexpression fails or leads to aggregated material from which the production of functionally refolded protein is challenging. It is still difficult to predict whether a given membrane protein can be overproduced in a functional competent state. As a result, the most straightforward strategy to set up an overexpression system is to screen a multitude of conditions, including the comparison of homologues, type and location of (affinity) tags, and distinct expression hosts. Here, we detail methodology to rapidly establish and optimize (membrane) protein expression in Escherichia coli and Lactococcus lactis. Key words: Arabinose promoter, Escherichia coli, folding indicator, GFP, Lactococcus lactis, membrane proteins, nisin-controlled expression, overexpression, PBAD

1. Introduction Although for the elementary steps in the biogenesis of membrane proteins, which are membrane targeting, membrane insertion, and assembly and folding, the crucial components are known (reviewed in (1)), our current understanding of the overall process and the factors involved are far from complete. Predicting whether a given membrane protein can be overproduced in a functional competent state is currently not possible, and bottlenecks have not been identified that could guide a rational troubleshooting strategy should overexpression fail. To date, membrane protein biogenesis has been mostly studied from the perspective of the known components involved in targeting, membrane insertion, and folding. In general, this is done using a limited number of

I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_2, © Humana Press, a part of Springer Science + Business Media, LLC 2010

17

18

Geertsma and Poolman

small and simple model proteins, while systematic genomewide studies have indicated that overexpression of complex membrane proteins with multiple transmembrane segments is most challenging (2–4). Only recently efforts have been made to understand the complete process in the context of a cell by determining the physiological responses underlying (un)successful overproduction of (complex) membrane proteins (5,6). As a consequence, if one wishes to obtain structural and functional information on a certain membrane protein that is not abundant in any natural source, a convenient strategy is to screen a wide range of conditions. Established initial variables involve the screening of homologues (7,8), varying the type and location of affinity tags and domains (9,10), and screening multiple-expression hosts (11,12). In practice, for prokaryotes screening is restricted to several Escherichia coli strains, which offers far less variation than the screening of different species. The gram-positive bacterium Lactococcus lactis is an attractive alternative to E. coli and yeast-based membrane protein expression systems (13,14). Overexpression of membrane proteins in L. lactis leads mostly to well-folded, membrane-inserted material; inclusion bodies of aggregated material are rare. The lactococcal membrane is easily solubilized with a wide range of detergents. Growth of L. lactis proceeds rapidly at 30°C and does not require aeration. The nisin-inducible controlled expression (NICE) system (15) allows reproducible and modulatable expression from low to high levels. Previous difficulties associated with direct cloning in L. lactis (10) have been overcome by the high-throughput compatible vector backbone exchange (VBEx) procedure (16). As L. lactis is auxotrophic for multiple amino acids, adjustment of the lactococcal chemically defined medium (17) allows incorporation of labeled amino acid analogues into proteins (44, 45). Concerning membrane protein expression, there are five important differences between E. coli and L. lactis. First, the composition of the machinery involved in membrane protein insertion: L. lactis contains two paralogs of the Oxa/YidC/Alb family of membrane-inserted chaperones (18), whereas E. coli contains one (YidC). Lactococcus lactis does not contain homologues of the Sec translocon accessory proteins SecDF (19) and SecM (20). Finally, substantial differences in structure are observed between lactococcal and E. coli homologues of YidC and SecE (20,21). Second, the lipid composition of the membrane is very different: The major lipid component of the E. coli cytoplasmic membrane is phosphatidylethanolamine (PE), which is absent in L. lactis. Instead, anionic glycolipids and phosphoglycolipids are the major constituents in L. lactis (22). The third difference involves codon usage (23); L. lactis DNA is relatively AT rich (64% AT, compared to 50% AT in E. coli). The fourth difference is the absence of disulfide isomerase ­homologues in L. lactis (24), albeit that disulfides

Production of Membrane Proteins in Escherichia coli and Lactococcus lactis

19

are readily formed either spontaneously or via an alternative route in secreted proteins (25–27). The final difference is that some proteins toxic to one host do not affect the physiology of the other host (C. Mulligan and E.R. Geertsma, March 2007). In addition, although fewer attempts have been made than in E. coli, overexpression of eukaryotic membrane proteins in L. lactis is relatively successful (28). 1.1. High-Throughput Cloning in E. coli and L. lactis

To establish an overexpression system rapidly, several homologues, types and location of affinity tags, and expression hosts need to be screened. Due to the large number of plasmids required for this strategy, it will be advantageous to avoid low-throughput techniques and implement efficient cloning procedures such as ligationindependent cloning (LIC) (29). LIC is less restricted in the design of the sequences flanking the genes than other high-throughput cloning techniques, like Gateway (30) or the univector plasmidfusion system (31). Therefore, the cloning-related sequences attached to the recombinant protein can be minimized, which might be advantageous for structural and functional analysis. The LIC procedure described here, outlined in Fig.  2.1a, involves linearization of the vector by a unique SwaI site in the middle of the LIC cassette. Using the 3¢ to 5¢ exonuclease activity of T4 DNA polymerase, single-stranded overhangs are generated. By performing the exonuclease treatment in the presence of dCTP (deoxycytidine triphosphate), a nucleoside not present in the 3¢ sequences adjacent to the SwaI site, the removal of bases halts once the first dCMP (deoxycytidine monophosphate) is encountered, thereby creating overhangs of a defined length. Complementary single-stranded overhangs are created in a similar way in the polymerase chain reaction (PCR) product to be cloned (i.e., exonuclease treatment in the presence of dGTP (deoxy­ guanosine triphosphate)). On mixing the T4 DNA polymerasetreated vector and PCR product, a stable heteroduplex is formed, and the mixture can be transformed with high efficiency to E. coli. The efficiencies for direct cloning in L. lactis are low, and LIC is virtually not possible. This bottleneck in cloning has been overcome by the VBEx procedure (16). Using VBEx, the initial LIC step can be performed with high efficiency in E. coli while the final expression vector that is generated can be kept devoid of alien E. coli elements. The VBEx procedure, outlined in Fig. 2.1b, relies on the bisection of a bona fide plasmid of the expression host into two parts, thereby separating the selection marker from the origin of replication. For L. lactis, pNZxLIC, a LICcompatible derivative of the well-established expression vector pNZ8048 (15) was accordingly bisected. The segment containing the origin of replication was fused to a sequence coding for an erythromycin selection marker, yielding plasmid pERL, which can be maintained in L. lactis. The other segment of pNZxLIC,

20

Geertsma and Poolman

Fig. 2.1. High-throughput cloning in recalcitrant bacteria using LIC and VBEx. (a) Outline of the LIC procedure. Gene X is amplified using primers holding LIC-specific overhangs. The plasmid is linearized by SwaI restriction in the LIC cassette. Singlestranded overhangs on the PCR product and vector are generated using T4 DNA polymerase. The complementary overhangs of PCR product and vector anneal on mixing. The resulting heteroduplex is transformed efficiently to Escherichia coli. (b) Outline of the VBEx strategy. After introduction of two distinct SfiI sites, the Lactococcus lactis expression vector pNZxLIC was bisected. Plasmid pERL received the pSH71 replicon, which was fused to the erythromycin marker. Plasmid pRExLIC received the chloramphenicol marker and LIC sequence, which were fused to the E. coli pBR322 replicon and b-lactamase marker. As these vectors are available, in practice the cloning involves the submission of vector pRExLIC to the LIC procedure (depicted in a). Subsequently, the VBEx procedure is applied: The pNZxLIC vector is restored by mixing pERL and pRExLICgene X, digestion with SfiI, ligation, and selection on the ability to replicate in L. lactis (the presence of the pSH71 replicon) in the presence of chloramphenicol. (c) The SfiI sites flanking both segments of the vectors yield different, nonpalindromic overhangs, thereby greatly reducing the number of possible ligation products. (Adapted from (16)).

containing the chloramphenicol resistance gene and LIC sequence, was fused to the backbone of an E. coli vector (containing an E. coli origin of replication and b-lactamase resistance gene). The resulting plasmid pRExLIC, allows the LIC manipulation to take place in E. coli.

Production of Membrane Proteins in Escherichia coli and Lactococcus lactis

21

Rapid and facile regeneration of the L. lactis expression vector is enabled as the relevant segments of the pRExLIC and pERL vectors are flanked with distinct SfiI sites. DNA cleavage by SfiI generates a 3¢ overhang that can be composed of any combination of three nucleotides. The two SfiI sites used yield different, incompatible overhangs (Fig.  2.1c). Each segment of the pNZxLIC vector has unique selectable properties: (1) the ability to replicate a plasmid in L. lactis and (2) the resistance to chloramphenicol. Thus, on mixing a pRExLIC derivative and pERL, digestion by SfiI, followed by ligation and transformation to L. lactis in combination with selection of clones using chloramphenicol, the pNZxLIC derivative can be exclusively recovered with high efficiencies. DNA sequences from virtually all sequenced genomes are compatible with VBEx as SfiI sites are rare (16). Furthermore, genes containing internal SfiI sites are not necessarily excluded as 64 different 3¢ overhangs can be generated after SfiI digestion. Internal SfiI sites with 3¢ overhangs not matching those of the vector will not form a bottleneck in the procedure. 1.2. Optimization of Membrane Protein Overexpression

Following initial screening for expression, systematic optimization of the production conditions to increase the amount of well-folded protein inserted into the membrane is often required. As indicated, membrane protein biogenesis requires several additional steps and components beyond translation. Exceeding the capacity of the cell to process the nascent membrane protein correctly may reduce the final yield of well-folded material. This is especially true for E. coli as membrane protein production in this host is regularly accompanied by inclusion body formation. Adjustment of the expression rate is most easily done by varying the inducer concentration for well-tunable promoters, such as the E. coli PBAD (32) or the lactococcal PnisA (15). In addition, variation of the growth temperature during induction is an important parameter to affect the expression rate (33,34). While traditionally optimization of functional expression involved time-consuming steps such as the isolation of membrane vesicles and activity assays, this process has been greatly accelerated by the use of green fluorescent protein (GFP), fused to the C-terminus of the target protein, as folding reporter. Proper maturation of GFP in E. coli, leading to a fluorescent species that maintains its folded state during sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), depends highly on the correct folding of its fusion partner (35–38). Consequently, only correctly folded target yields a fluorescent fusion protein that has a higher electrophoretic mobility than the misfolded fusion protein. Plain SDS-PAGE and subsequent immunoblotting allow simultaneous quantification of both the well-folded and aggregated protein produced (Fig. 2.2) (38). This facilitates the assignment of bottlenecks limiting the functional overexpression. For instance, the absence of the upper, nonfluorescent band suggests that transcription or

22

Geertsma and Poolman

Fig. 2.2. SDS-PAGE analysis of membrane protein GFP fusions expressed to different levels. Expression of the GFP fusion proteins was controlled by the PBAD, and Escherichia coli MC1061 cultures were induced with the percentage l-arabinose (w/v) indicated above the lanes. Upper panels represent the images obtained by in gel GFP fluorescence. Lower panels represent immunoblots of the same gels decorated with anti-His tag antibody. Black and white arrows indicate the positions of the aggregated nonfluorescent and well-folded fluorescent species of the GFP fusion proteins, respectively. (Adapted from (38)).

t­ ranslation limits the functional expression, which can be resolved by increasing the inducer concentration. In contrast, the strong presence of the upper, nonfluorescent band suggests that the trajectory beyond transcription/translation is limiting. In addition to this application, the use of GFP fusions has additional advantages at the stages beyond membrane protein overexpression, such as the rapid screening of detergents for solubilization and stabilization (37,39).

2. Materials 2.1. LigationIndependent Cloning Procedure 2.1.1. Preparation of the DNA Insert

1. Autoclaved TlowE: 10 mM Tris-HCl, pH 7.5, 0.2 mM Na2EDTA (ethylenediaminetetraacetic acid). 2. Adhesive plate seals (e.g., plate seal AB-0580 from Thermo Scientific). 3. Phusion DNA polymerase and dedicated buffers (Finnzymes). 4. DNA gel extraction kit (e.g., the High Pure PCR Product Purification kit from Roche).

2.1.2. Preparation of the LIC Vector

1. LB-amp: Dissolve 10 g tryptone, 5 g yeast extract, and 10 g NaCl in 1 L of demineralized water. Sterilize by autoclaving. Once the medium is cooled to room temperature, add 1 mL of a filter-sterilized solution of 100 mg/mL ampicillin dissolved in water. 2. Plasmid isolation kit (e.g., the Qiaprep Spin Miniprep kit from Qiagen). 3. 10% (w/v) SDS solution. 4. Proteinase K solution: Dissolve 20 mg/mL proteinase K in 10 mM Tris-HCl, pH 8.0, 1 mM Na2-EDTA. Store aliquots at −20°C. 5. Phenol:chloroform:isoamyl alcohol (25:24:1) mixture. Phenol is extremely toxic; wear protective clothing and work in a fume hood.

Production of Membrane Proteins in Escherichia coli and Lactococcus lactis

23

6. Chloroform, analytical grade. Chloroform is irritating and a carcinogen; wear protective clothing and work in a fume hood. 7. 96% EtOH, analytical grade, precooled at −20°C. 8. 3M Na-acetate, pH 5.3; adjust the pH with concentrated acetic acid. 9. 70% EtOH, analytical grade, precooled at −20°C. 10. SpeedVac (Savant). 11. SwaI and dedicated buffer. 12. Chemically competent E. coli MC1061. Protocols for the preparation of chemically competent E. coli can be found in (40). 13. LB-amp agar plates: Dissolve 10 g tryptone, 5 g yeast extract, and 10 g NaCl in 1 L of demineralized water and add 15 g agar. Sterilize by autoclaving. Once the medium is cooled to 50–60°C, add 1 mL of a filter-sterilized solution of 100 mg/ mL ampicillin dissolved in water; mix well and pour the medium in sterile Petri dishes. 2.1.3. T4 DNA Polymerase Treatment of Vector and Insert

1. T4 DNA polymerase and dedicated buffer. 2. 25 mM dCTP: Dilute a 100 mM dCTP stock in sterile Milli-Q. Store aliquots at −20°C. 3. 25 mM dGTP: Dilute a 100 mM dGTP stock in sterile Milli-Q. Store aliquots at −20°C. 4. Heat block.

2.1.4. Verifying Clones by Colony PCR

1. Autoclaved Milli-Q. 2. Sterile toothpicks. 3. 96-well PCR plates. 4. Aluminum sealing films (e.g., PCR-AS-200 from Axygen). 5. Taq DNA polymerase and dedicated buffer. 6. dNTPs: Make a stock solution containing 10 mM of each nucleotide. Store aliquots at −20°C. 7. Generic primers (see Note 1).

2.2. Vector Backbone Exchange Procedure 2.2.1. Preparation of pERL

1. GM17-ery: Dissolve 37.25 g of M17 broth (Oxoid) (41) in 1 L of demineralized water. Sterilize by autoclaving. Once the medium is cooled to room temperature, add 25 mL sterile 20% (w/v) glucose and 1 mL of a filter-sterilized solution of 5 mg/mL erythromycin dissolved in water. 2. Resuspension buffer: 10 mM Tris-HCl, pH 8.1, 10 mM Na2EDTA, 50 mM NaCl, and 20% (w/v) sucrose. Autoclave and store at 4°C. Supplement with 20 mg/mL lysozyme prior to use.

24

Geertsma and Poolman

2.2.2. Vector Backbone Exchange

1. SfiI and dedicated buffer. 2. 50 mM ATP stock: 50 mM Na2-ATP, 50 mM MgSO4, and 50 mM phosphate buffer, pH 7.0. Adjust the pH to 6.5–7.0 with NaOH. Store aliquots at −20°C. 3. T4 DNA ligase and dedicated buffer.

2.2.3. Electrotransformation of L. lactis

1. Sterile electroporation cuvets (2 mm gap between the electrodes). The cuvets can be reused several times. Clean the cuvets after use with 70% EtOH and wash extensively with Milli-Q. Allow the cuvets to dry and sterilize the cuvets by 20 min exposure to an ultraviolet (UV) source. Immediately close the cuvets with the supplied cap. 2. Electroporation device. 3. Rescue medium: Mix 1 volume of sterile M17 medium with 1 volume of sterile 1M sucrose, 1% (w/v) glucose, 40 mM MgCl2, 4 mM CaCl2. 4. SGM17-cam agar plates: Mix 37.25 g of M17 broth (Oxoid) and 15 g agar in 750 mL of demineralized water. Sterilize by autoclaving. Once the medium is cooled to about 60°C, add 250 mL sterile 2M sucrose, 25 mL 20% (w/v) glucose, and 1 mL of a filter-sterilized solution of 5 mg/mL chloramphenicol dissolved in EtOH. Mix well and pour the medium in sterile Petri dishes. 5. Parafilm.

2.3. Initial Characterization and Optimization of Protein Production 2.3.1. Cultivation and Induction of MP Overexpression in E. coli

1. Sterile 87% (w/v) glycerol. 2. Cryovials. 3. TB-amp: Dissolve 24 g tryptone, 48 g yeast extract, and 17.2 mL 87% (w/v) glycerol in 0.9 L of demineralized water and autoclave. In addition, autoclave 100 mL of 170 mM KH2PO4, 720 mM K2HPO4. Mix the solutions prior to use and add 1 mL of 100 mg/mL ampicillin. 4.

l-Arabinose:

Filter-sterilize a 20% (w/v) l-arabinose stock.

5. Screw-cap tubes compatible with the FastPrep-24 (e.g., screw-cap microtube 72.693 from Sarstedt). 2.3.2. Cultivation and Induction of MP Overexpression in L. lactis

1. GM17-cam: M17 medium supplemented with 0.5% (w/v) glucose and 5 mg/mL chloramphenicol after sterilization. 2. Nisin-containing supernatant of a L. lactis NZ9700 culture: Use a 1% (v/v) inoculum of a preculture of L. lactis NZ9700 on GM17 to inoculate 200 mL GM17 and incubate overnight at 30°C. Next day, pellet the cells, filtrate the supernatant using a 0.2-mm filter, and store the aliquots at −20°C.

Production of Membrane Proteins in Escherichia coli and Lactococcus lactis 2.3.3. Cell Disruption and Sample Preparation

25

1. Glass beads with a diameter of about 0.1 mm (SigmaAldrich). 2. Disruption buffer: 50 mM phosphate buffer, pH 7.5, 10% (w/v) glycerol, and 1 mM MgSO4. Precool the solution on ice. Add 1 mM PMSF (phenylmethylsulfonyl fluoride) (from a 200 mM stock in EtOH (ethanol)) and 0.1 mg/mL DNase (deoxyribonuclease) prior to use. 3. FastPrep-24 device (MP Biomedicals). 4. 5X protein sample buffer: 120 mM Tris-HCl, pH 6.8, 50% (w/v) glycerol, 100 mM dithiothreitol (DTT), 2% (w/v) SDS, and 0.1% (w/v) bromophenol blue (see Note 2). Store aliquots at −20°C.

2.3.4. SDS-PAGE, In Gel GFP Fluorescence, and Western Blotting

1. SDS-PA gel: For a 10% running gel, mix 3.38 mL 30% acrylamide/bis solution (cross-linker ratio 37.5:1), 2.54 mL 1.5M Tris-HCl, pH 8.8, 4.15 mL Milli-Q, 100 mL 10% (w/v) SDS, 40 mL 10% (w/v) APS (ammonium persulfate), and 10 mL TEMED. For the stacking gel, mix 0.68 mL 30% acrylamide/bis solution (cross-linker ratio 37.5:1), 1.13 mL 0.5M Tris-HCl, pH 6.8, 2.25 mL Milli-Q, 50 mL 10% (w/v) SDS, 30 mL 10% (w/v) APS, and 5 mL TEMED (see Note 2). 2. Precision Plus Protein Dual Color standard (Bio-Rad). 3. LAS-3000 imaging system (Fujifilm). 4. Electroblotting and immunodetection system with primary antibody directed against a His-tag.

3. Methods 3.1. LigationIndependent Cloning Procedure 3.1.1. Preparation of the DNA Insert

1. Design two sets of LIC primers for each gene to be cloned. The nLIC and cLIC primer sets will finally produce proteins with an N- or C-terminal affinity tag, respectively. The LIC strategy requires the presence of dedicated 5¢ extensions on the primers (indicated in Table 2.1). In addition, the genespecific part of the primer should be designed so that it (1) does not contain the endogenous start or stop codon of the target gene; (2) is sufficiently long to anneal in a stable and selective way with the target DNA (in general, 20–25 nucleotides suffice; see Note 3). The primer should not contain strong secondary structure elements that could interfere with the PCR (see Note 4). Keep the total length of the

26

Geertsma and Poolman

Table 2.1 Primer extensions needed for ligation-independent cloning Primer type

5¢ Extensions of primers

nLIC forward

5¢ AT GGT GAG AAT TTA TAT TTT CAA GGT

nLIC reverse

5¢ T GGG AGG GTG GGA TTT TCA TTA

cLIC forward

5¢ ATG GGT GGT GGA TTT GCT

cLIC reverse

5¢ TTG GAA GTA TAA ATT TTC

primers below 50 residues to ensure sufficient quality and cost-efficiency. 2. For ordering, select the lowest synthesis scale offered by the supplier and mere desalting as the purification step. For large sets (>48 primers), order the primers in a 96-well plate. 3. On delivery, spin down the material and dissolve the primers to 100 mM in TlowE and make a substock of 5 mM. Use a transparent sticker to seal the plates. The 100 mM stock can be stored for several months at −20°C. Store the 5 mM stock at 4°C. Spin down the material before use. 4. PCR the desired open reading frames using the Phusion DNA polymerase, which combines high fidelity with high production (see Notes 5 and 6). Prepare a 50 mL reaction mix according to the manufacturer’s protocol and add the polymerase just prior to the start of the reaction. Place the sample in a PCR machine preheated to 98°C and start the reaction. A good starting point for a touchdown PCR program (see Note 7) is a. 60 s at 98°C. b. 10 s at 98°C. c. 15 s at 58.5°C (decrease 0.5°C per cycle). d. XX s at 72°C (15–30 s/kb). Repeat steps 4a–4d 14 times. e. 10 s at 98°C. f. 15 s at 51°C. g. XX s at 72°C (15–30 s/kb). Repeat steps 4e–4g 14 times. h. 300 s at 72°C. i. Forever at 10°C. Once the PCR is finished, add DNA loading buffer to the sample and analyze all the material by TAE (Tris-acetate-EDTA) gel electrophoresis (see Note 8). Gel purify the band using a

Production of Membrane Proteins in Escherichia coli and Lactococcus lactis

27

Table 2.2 Basic and additional LIC vectors Vector name

Protein sequence

Protein sequence after TEV protease cleavage

Expression host

pBADnLIC

M-His10-G-TEV site-protein

G protein

Escherichia coli

pBADcLIC

MGGGFA-protein-TEV site-His10

MGGGFA-proteinENLYFQ

E. coli

pBADcLIC-GFP

MGGGFA-protein-TEV site-GFP-His10

MGGGFA-proteinENLYFQ

E. coli

pBAD-OmpA-nLIC

M-ssOmpAa-His10-G-TEV site-protein

G-protein

E. coli

pREnLIC

M-His10-G-TEV site-protein

G-protein

Lactococcus lactis

pREcLIC

MGGGFA-protein-TEV site-His10

MGGGFA-proteinENLYFQ

L. lactis

pREcLIC-GFP

MGGGFA-protein-TEV site-GFP-His10

MGGGFA-proteinENLYFQ

L. lactis

pRE-USP45-nLIC

M-ssUSP45b-His10-G-TEV site-protein

G-protein

L. lactis

a ssOmpA indicates the signal sequence of the E. coli OmpA protein. The pBAD-OmpA-nLIC vector is used if the N-terminus of the membrane protein of interest is predicted to be on the outside of the cytoplasmic membrane or to replace the signal sequence of the protein of interest with the ssOmpA b ssUSP45 indicates the signal sequence of the L. lactis USP45 protein. The pRE-USP45 -nLIC vector is used if the N-terminus of the membrane protein of interest is predicted to be on the outside of the cytoplasmic membrane or to replace the signal sequence of the protein of interest with the ssUSP4

commercial gel extraction kit according to the manufacturer’s protocol. Determine the DNA concentration spectrophotometrically (see Note 9). Store the material at −20°C. 3.1.2. Preparation of the LIC Vector

1. Inoculate approximately 20 mL LB-amp with E. coli MC1061 containing one of the pBAD- or pRE-derived LIC vectors (Table  2.2). Cultivate overnight at 37°C with vigorous shaking. 2. Isolate the plasmid using a commercial plasmid isolation kit and elute the plasmid with TlowE. Remove remaining proteins by combining a proteinase K treatment with a phenol:chloroform extraction as described in steps 3–5. 3. Combine several fractions and adjust the volume to 300 mL. Supplement the sample with SDS to a final concentration of 0.5% and add proteinase K to a final concentration of 40 mg/mL. Incubate for 30 min at 60°C to digest remaining proteins.

28

Geertsma and Poolman

4. In a fume hood, add 1 volume of phenol:chloroform and vortex shortly. After 2 min, the sample is vortexed again and centrifuged for 2 min at full speed in a table centrifuge. Collect most of the top phase without disturbing the (white) interface and transfer to a new cup. Add 1 volume of chloroform and mix by vortexing. Centrifuge for 2 min at full speed in a table centrifuge and transfer the top phase without disturbing the interface to a new cup. 5. Add 2.5 volume of ice-cold 96% EtOH and 0.1 volume of 3M Na-acetate, pH 5.3. Vortex the sample and precipitate the DNA for 30 min at −80°C. Next, centrifuge the sample at 4°C for 15 min at full speed in a table centrifuge. Remove the supernatant and add 0.9 mL of ice-cold 70% EtOH. Centrifuge 5 min at full speed at room temperature in a table centrifuge. Remove the supernatant and dry the pellets shortly in a SpeedVac. Dissolve the pellet in 20 mL TlowE, determine the DNA concentration, and verify the integrity of the plasmid by restriction analysis. 6. Take 5 mg plasmid and digest the plasmid in a total volume of 40 mL containing 20 U SwaI according to the manufacturer’s protocol (see Note 10). After 3 h, supplement the reaction with 10 U SwaI and extend the reaction for another 3 h. Analyze the full sample on 0.8% (w/v) agarose TAE gel using combs with wide wells. Isolate the linearized vector using a gel purification kit and determine the concentration. Store the material at −20°C. 7. To test the completeness of the digestion, transform 15 ng of the linearized plasmid to chemically competent E. coli MC1061 according to standard procedures. After the transformation procedure and 1 h incubation at 37°C with LB medium, sediment the cells by centrifugation (7,600g, 5 min), remove most of the supernatant, and plate the resuspended material on LB-amp agar plates. After overnight incubation, the colony count should be low (<100 CFU (colony-forming units)). 3.1.3. T4 DNA Polymerase Treatment of Vector and Insert

1. Take 200 ng of the SwaI-digested and gel-purified vector and adjust the volume to 10 mL. Add 3 mL of 5X T4 DNA polymerase buffer, 1.5 mL of 25 mM dCTP, and 0.5 mL of T4 DNA polymerase (0.5 U) (see Note 11). Mix and incubate for 30 min at room temperature. Immediately heat inactivate the T4 DNA polymerase by incubating for 20 min at 75°C and store the T4 DNA polymerase-treated vector at −20°C. 2. Take an amount of gel-purified insert equimolar to 200 ng vector (see Note 12) and adjust the volume to 10 mL. Add 3 mL of 5X T4 DNA polymerase buffer, 1.5 mL of 25 mM dGTP,

Production of Membrane Proteins in Escherichia coli and Lactococcus lactis

29

and 0.5 mL of T4 DNA polymerase (0.5 U) (see Note 11). Mix and incubate for 30 min at room temperature. Immediately heat inactivate the T4 DNA polymerase by incubating for 20 min at 75°C and store the T4 DNA polymerasetreated vector at −20°C. 3.1.4. LigationIndependent Cloning

Mix 3 mL of the treated insert with 1 mL of the vector and incubate for 5 min at room temperature. Place the cup on ice and after 5 min add chemically competent E. coli MC1061. Transform the cells using standard procedures. Plate 10% and 90% of the transformed cells on LB-amp (see Note 13). Incubate overnight at 37°C.

3.1.5. Verifying Clones by Colony PCR

1. Divide the bottom plate of an empty LB-amp agar plate in 32 numbered squares with a marker. Fill each well in a 96-well PCR plate with 12.5 mL sterile Milli-Q. 2. Use a sterile toothpick to pick a colony and set a small streak on the agar plate (see Note 14). This plate will be temporarily used for archiving the clones. Subsequently, dip the toothpick in the appropriate well of the 96-well PCR plate and discard it. A set of four clones per construct usually suffices to finally obtain at least three positive clones. 3. Once the clones have been picked, seal the 96-well plate thoroughly with an aluminum sealing film (see Note 15) and lyse the cells for 15 min at 98°C in a PCR machine. 4. Allow the plate to cool to room temperature and supplement each well with 12.5 mL of a twofold concentrated PCR mix containing double the amount of Taq DNA polymerase, buffer, dNTPs, primers (see Note 1), and other additives as detailed in the manufacturer’s protocol. After sealing the plate with a fresh aluminum sticker, continue the PCR according to a. 30 s at 94°C. b. 30 s at 94°C. c. 60 s at 54°C. d. XX s at 72°C (~45 s/kb). Repeat steps 4a–4d 30 times. e. 300 s at 72°C. f. Forever at 10°C. Add DNA loading buffer to the sample and analyze all the material by TAE gel electrophoresis. Positive clones of the pBADxLIC vectors can be used for the expression analysis described in Section  2.3.3. immediately. Positive clones of the pRExLIC vectors first need to be converted into pNZxLIC vectors using the VBEx procedure described next. If positive clones are not sequenced at this stage, continue with three positive clones to spread the risk of PCR errors.

30

Geertsma and Poolman

3.2. Vector Backbone Exchange Procedure 3.2.1 Isolation of pRExLIC Plasmids Containing Inserts

3.2.2. Preparation of pERL

1. Inoculate 5 mL LB-amp with a positive streak from the agar plate containing the streak archive from the colony PCR. Incubate overnight at 37°C with vigorous aeration. 2. Isolate the plasmid using a commercial miniprep kit and elute the DNA in TlowE buffer. Determine the concentration (see Note 16) and store the sample at −20°C. 1. Inoculate 20 mL of GM17-ery with NZ9000/pERL. Incubate overnight at 30°C (see Note 17). 2. Pellet the cells and resuspend the cell pellet in resuspension buffer. Incubate the cells at 55°C for 15 min to disrupt the cell wall (see Note 18). 3. Proceed with the plasmid isolation using a commercial miniprep kit, starting at the addition of the cell lysis buffer. Elute the plasmid in TlowE. 4. Remove remaining proteins by proteinase K digestion and Phenol:chloroform extraction as described in Section 2.3.1.2., steps 3–5. Dissolve the pellet of the EtOH precipitation in 20 mL TlowE, determine the DNA concentration. and verify the integrity of the plasmid by restriction analysis. Store at −20°C.

3.2.3. Vector Backbone Exchange

1. In a PCR cup, mix about 125 ng of pERL, about 125 ng of pRExLIC vector containing the insert, 1 mL of 10X SfiI buffer, sufficient Milli-Q to adjust the volume to 9.5 mL, and 0.5 mL of SfiI (5 U; see Note 19). 2. Digest the plasmids for 1 h at 50°C in a PCR machine with heated lid to avoid excessive evaporation. Subsequently, inactivate SfiI by heating for 20 min at 80°C (see Note 19). 3. Allow the sample to reach room temperature and add 1.5 mL of 8 mM ATP and 0.5 mL of T4 DNA ligase (0.5 U; see Note 20). Incubate for 1 h at room temperature and inactivate the ligase in a PCR machine by heating for 20 min at 65°C. Store the sample at −20°C.

3.2.4. Electrotrans­ formation of L. lactis

1. Thaw electrocompetent L. lactis NZ9000 (see Note 21) on ice. 2. Transfer 50 mL cells to an ice-cold electroporation cuvet and add 2 mL of the sample resulting from the VBEx procedure (Section 2.3.2.3, step 3). Mix by pipetting up and down. 3. Expose the cells to a single electrical pulse with a field strength of 2.0 kV/2 mm, 25 mF capacitance, and 200 W resistance. Immediately following discharge, add 1 mL ice-cold rescue medium and leave the cells on ice for 10 min. 4. Incubate the cells, kept in the electroporation cuvet, for 2 h at 30°C.

Production of Membrane Proteins in Escherichia coli and Lactococcus lactis

31

5. Plate 10% and 90% of the transformed cells on SGM17-cam agar (see Note 13). Seal the plates with Parafilm (see Note 22) and incubate overnight at 30°C. 6. The colonies of L. lactis NZ9000/pNZxLIC holding the insert obtained in this step can be used for the expression analysis described in Section 2.3.3. 3.3. Initial Characterization and Optimization of Protein Production 3.3.1. Cultivation and Induction of MP Overexpression in E. coli

1. Inoculate 3 mL LB-amp from a verified streak obtained in Section  2.3.1.5. and incubate overnight at 37°C with vigorous aeration. 2. Mix 770 mL of culture with 230 mL of 87% (w/v) glycerol in a cryovial. Snap-freeze the vial in liquid nitrogen and archive the frozen stock at −80°C. 3. Use 60 mL of the remaining culture to inoculate a culture tube containing 6 mL TB-amp (see Note 23). Incubate the cells at 37°C and 200 rpm and follow the optical density of the culture (see Note 24). 4. At OD660 = ~0.2, place the tubes in a water bath at room temperature for 15 min and subsequently continue growth at 25°C. 5. After 30 min, induce the cultures with 60 mL of 5.10−1% (w/v) l-arabinose stock (100-fold concentrated) to obtain a final l-arabinose concentration of 5.10−3% (w/v). Incubate an additional 4 h. 6. Place the tubes in an ice-water bath and determine the OD660 of all cultures. Assuming that OD660 = 1 corresponds to a total protein concentration of 0.3 mg/mL, pellet the amount of culture equivalent to 1 mg protein in a FastPrep-compatible tube. Completely remove the supernatant, snap-freeze the pellet in liquid nitrogen, and store the sample at −20°C.

3.3.2. Cultivation and Induction of MP Overexpression in L. lactis

1. Inoculate 3 mL of GM17-cam with a single colony obtained in Section 2.3.2.4. and incubate overnight at 30°C. 2. Mix 770 mL of culture with 230 mL of 87% (w/v) glycerol in a cryovial. Snap-freeze the vial in liquid nitrogen and archive the frozen stock at −80°C. 3. Use 80 mL of the remaining culture to inoculate a culture tube containing 8 mL of GM17-cam (see Note 23). Incubate the cells at 30°C and follow the optical density of the culture (see Note 24). 4. The nisin A present in the supernatant of an overnight culture of the nisin-producing strain L. lactis NZ9700 is used to induce the cells. Dilute an aliquot of this supernatant 50-fold in GM17-cam to obtain a 100X concentrated stock. At OD660 = ~0.5, induce the cells with 80 mL of the 50-fold diluted

32

Geertsma and Poolman

supernatant to obtain a final dilution of 5,000-fold. Incubate an additional 2.5 h. 5. Place the tubes in an ice-water bath and determine the OD660 of all cultures. Assuming that OD660 = 1 corresponds to a total protein concentration of 0.2 mg/mL, pellet the amount of culture equivalent to 1 mg protein in a FastPrepcompatible tube. Completely remove the supernatant, snapfreeze the pellet in liquid nitrogen, and store the sample at −20°C. 3.3.3. Cell Disruption and Sample Preparation

1. Add about 300 mg of glass beads and 400 mL of ice-cold disruption buffer to the frozen cell pellets. 2. Disrupt the cells by two rounds of bead-beating in the FastPrep-24 device at force 6 for 20 s (for E. coli) or 30 s (for L. lactis). Place the cups in an ice-water bath for 5 min after each round (see Note 25). 3. Allow the glass beads to sediment for about 1 min and mix a 100 mL fraction from the supernatant in a new cup with 25 mL of 5X sample buffer (see Note 2). Do not boil the sample as this will lead to aggregation of the membrane protein and denaturation of GFP fusion proteins. Store the sample at −20°C or proceed immediately with the SDS-PAGE.

3.3.4. SDS-PAGE, In Gel GFP Fluorescence, and Western Blotting

1. Load 10 mL of each sample per lane on an SDS-polyacrylamide gel (see Note 2). Include the Precision Plus Protein dualcolor prestained molecular weight marker and run the gel at 150 V. Once the dye front has reached the bottom of the gel, disassemble the setup and soak the gel in Milli-Q. 2. For GFP fusion proteins, determine the in gel fluorescence. Place the gel on the black imaging tray of the LAS-3000 imaging system. Adjust the focus of lens to the gel and expose the gel to 460 nm light from the blue LED (light-emitting diode) and use a 515 nm emission filter (see Note 26). The fluorescent bands of the marker at 25 and 75 kDa should be clearly visible. 3. Submit the gel to immunoblotting to detect the His-tagged proteins. Use the LAS-3000 in chemiluminescence mode to measure the signal (see Note 26). The 75 kDa band of the marker should give a clear signal. Once a satisfactory exposure has been obtained, without moving the gel, use the top white light mode of the LAS-3000 to obtain an image of the membrane to locate the prestained marker bands. 4. Use the marker bands detected during the in gel fluorescence, chemiluminescence, and the white light image to estimate the intensity and the molecular weight of the detected bands and correlate the different images (see Note 27).

Production of Membrane Proteins in Escherichia coli and Lactococcus lactis 3.3.5. Expression Optimization

33

1. Optimize the expression by varying the protein expression rate over a wide range. For the PBAD system, use inducer concentrations between 1.10−5% and 1.10−1% (w/v) of l-arabinose; vary the induction temperature between 37°C, 25°C, and 17°C; and induce for 1, 2, 4, and 16 h. Determine the effect on expression using small cultures as described in Sections 2.3.3.1.– 2.3.3.4. If possible, make use of the GFP fusion protein to guide the optimization procedure in E. coli. The presence of a strong, nonfluorescent band, indicative for aggregated protein, suggests too high an expression rate; the absence of a nonfluorescent band suggests that the capacity of the targeting, insertion, and folding pipeline is not yet saturated. 2. For the lactococcal PnisA system, use dilutions of the nisincontaining supernatant ranging from 100,000- to 500-fold; vary the induction temperature between 30°C, 25°C, and 17°C; and induce for 1, 2, 4, and 16 h. Determine the effect on expression using small cultures as described in Sections 2.3.3.1.–2.3.3.4.

4. Notes 1. Use generic primers instead of gene-specific primers. This facilitates the analysis of several genes and provides a positive control for the PCR because for empty vectors a band will be produced as well. Generic primers for the pBADxLIC derivatives are 5¢ctctactgtttctccatacccg and 5¢gctgaaaatcttctctcatccg. Generic primers for the pRExLIC derivatives are 5¢cgcgagcataataaacggc and 5¢ctgctgctttttggctatcaatc. 2. Deviations in the composition from the here-described buffers for sample preparation and gel electrophoresis might result in a decrease in the intensity of the in gel GFP fluorescence. 3. The initial steps of primer design can be easily automated using a spreadsheet program. This will speed up the design process and reduce the number of errors. A spreadsheet that combines the terminal 20 nucleotides of a set of DNA sequences with the nLIC and cLIC extensions to produce a series of initial primers that might require additional corrections can be obtained from the authors’ Web site (http://www.rug.nl/gbb/research/ researchgroups/enzymology/index). 4. Removal of secondary structure elements is often best done by making silent mutations in the gene-specific part of the primer while preventing the use of rare codons as much as possible. Do not make any changes in the primer extensions required for LIC.

34

Geertsma and Poolman

5. Genomic DNA is easily compromised by repetitive freeze/ thawing, vortexing, or vigorous pipetting and thus requires careful handling; storage in sterile tubes at 4°C is preferred. 6. Combine PCRs for targets of similar length and template type (plasmid or genomic) as much as possible instead of performing a PCR for each gene separately. This will speed up the whole process and allows more time for optimizing PCRs of “difficult genes.” 7. Touchdown PCR ensures a more specific product formation (42), and as a range of annealing temperatures is used, it is a good alternative to optimizing the reaction conditions for each target on an individual basis. 8. Should the DNA template that was used for the PCR reaction be a plasmid that confers resistance to b-lactam antibiotics such as ampicillin and carbenicillin, then the PCR mix should be incubated for 30 min at 37°C with 1 mL DpnI prior to the gel electrophoresis to remove the template. As the pBADxLIC and pRExLIC vectors used for the LIC procedure carry the b-lactamase marker, traces of template plasmids that carry the same marker will lead to a high number of background colonies. 9. Should the PCR not yield the desired product in sufficient amounts, consider the use of alternative buffers supplied by the manufacturer, the addition of dimethyl sulfoxide (DMSO), or replacement of the template. More detailed suggestions can be found in the Molecular Cloning protocol series (40). 10. Depending on the producer of SwaI, the suggested incubation temperature is 25°C or 30°C. Condensation at the lid, leading to suboptimal reaction conditions, can be prevented by performing the reaction in an incubator or PCR machine with heated lid. 11. The quality of the dNTP stocks and the T4 DNA polymerase is a critical parameter for the success of the procedure. Prevent repetitive freezing/thawing of the nucleotides by preparing small aliquots and label the aliquots appropriately to prevent mixing up dCTP (used for the vector) with dGTP (used for the insert). Transport the T4 DNA polymerase in a cooling tray and keep the enzyme cold under all conditions. 12. The use of equimolar amounts of vector and insert ensures that both are treated with T4 DNA polymerase to a similar extent. The amount of the purified PCR product needed can be calculated using the following: Amount insert needed (ng) = ((200 ng * Insert size (bp))/Vector size (bp)). 13. Especially for large numbers of samples, spreading the transformed cells over the plate is most easily done by making use

Production of Membrane Proteins in Escherichia coli and Lactococcus lactis

35

of glass beads (diameter 5 mm). In addition, this prevents cross contamination. 14. Use of toothpicks over an inoculation needle prevents cross contamination when analyzing multiple samples at one time. A cross-locking tweezer facilitates handling the toothpicks for prolonged periods. 15. Take care to properly seal the plate, especially around the sides. Incomplete sealing will lead to evaporation. 16. Especially for large numbers of plasmid samples, determining the DNA concentration can be time consuming. As the VBEx reaction only requires an approximate estimation of the DNA concentration, it suffices to determine the concentration of a small subset (~5 samples) and use the average DNA concentration as representative for this batch. 17. As L. lactis does not require aeration, small cultures are not shaken and can be maintained in closed glass or plastic containers. 18. The extent of the degradation of the cell wall depends on the incubation time and the amount of cells and lysozyme used; insufficient degradation leads to lower plasmid yields. Whether the lysozyme treatment was sufficient can be easily determined by mixing a 10 mL aliquot of the cell suspension with 10 mL of the supplied lysis buffer. Adequate disruption of the cell wall will lead to complete cell lysis within 5 min. If needed, extend the incubation time or use a higher lysozyme concentration. 19. For the overall procedure, it is critical that the SfiI enzyme can be heat inactivated. For unknown reasons, this property seems to depend on the supplier. We have had a good experience with SfiI obtained from Fermentas. 20. For a large number of samples, prepare a premix containing 6 mM ATP, 0.25 U/mL T4 DNA ligase, and 1X T4 DNA ligase buffer. Use ice-cold Milli-Q and prepare the mix just prior to use. 21. Preparation of electrocompetent L. lactis NZ9000 was essentially done as described (43), but with minor modifications. Briefly, cells were grown in M17 supplemented with 0.5% glucose, 0.5M sucrose, and 2% glycine at 30°C to OD600 = ~0.5. Cells were harvested by centrifugation at 5,000g for 15 min at 4°C. Following washes with 1 volume of ice-cold solution A (0.5M sucrose and 10% glycerol, prepared in Milli-Q), 0.5 volume of solution A supplemented with 50 mM Na-EDTA, pH 7.5, and 0.25 volume of solution A, cells were resuspended in 0.01 volume of solution A. Aliquots of 50 mL were snap-frozen in liquid nitrogen and stored at −80°C until use. 22. Sealing the agar plates with Parafilm is not essential but allows faster appearance of L. lactis colonies.

36

Geertsma and Poolman

23. Preheat the medium without the antibiotics overnight at 37°C or 30°C for E. coli and L. lactis, respectively. This prevents a slow start of the cultivation due to a small temperature shock. 24. When handling a large set of samples, use an additional dedicated set of approximately six reference cultures to follow the OD660 and use their average OD660 to time the moment of cooling or induction. 25. Alternative methodologies for cell disruption, such as sonication and vortexing in the presence of glass beads, can be used as well but might be less efficient. This is especially true for L. lactis, whose thicker and stable cell wall makes it more difficult to disrupt than E. coli. French press disruption is comparable to cell disruption by a FastPrep but requires larger volumes and is therefore less suited. For L. lactis, treatment with lysozyme to digest the peptidoglycan layer prior to disruption greatly enhances the efficiency. This step is not required for FastPrep-mediated cell disruption. 26. The use of the same tray height for in gel fluorescence and chemiluminescence imaging prevents rescaling of images, thus facilitating the correlation of the images. 27. The transfer efficiencies from the gel to the PVDF (poly­ vinylidene difluoride) membrane will differ for different membrane proteins. This complicates selection of promising clones. In contrast, the in gel fluorescence of GFP fusion proteins can be compared directly.

Acknowledgments We thank G.K. Schuurman-Wolters, M.B. Tol, M. Groeneveld, D.J. Slotboom, and S. Henstra for their contribution in establishing the methodologies discussed. This work was funded by the European Membrane Protein Consortium (grant 504601), a FEBS fellowship (to E.R.G.), and the Netherlands Proteomics Centre.

References 1. Driessen AJ, Nouwen N (2008) Protein translocation across the bacterial cytoplasmic membrane. Annu Rev Biochem 77:643–667 2. Korepanova A, Gao FP, Hua Y, Qin H, Nakamoto RK, Cross TA (2005) Cloning and expression of multiple integral membrane proteins from Mycobacterium tuberculosis in Escherichia coli. Protein Sci 14:148–158

3. Dobrovetsky E, Lu ML, Andorn-Broza R, Khutoreskaya G, Bray JE, Savchenko A, Arrowsmith CH, Edwards AM, Koth CM (2005) High-throughput production of prokaryotic membrane proteins. J Struct Funct Genom 6:33–50 4. White MA, Clark KM, Grayhack EJ, Dumont ME (2007) Characteristics affecting expression

Production of Membrane Proteins in Escherichia coli and Lactococcus lactis

5.

6.

7.

8.

9.

10.

11.

12. 13.

14.

15.

16.

and solubilization of yeast membrane proteins. J Mol Biol 365:621–636 Wagner S, Baars L, Ytterberg AJ, Klussmeier A, Wagner CS, Nord O, Nygren PA, van Wijk KJ, de Gier JW (2007) Consequences of membrane protein overexpression in Escherichia coli. Mol Cell Proteom 6:1527–1550 Bonander N, Hedfalk K, Larsson C, Mostad P, Chang C, Gustafsson L, Bill RM (2005) Design of improved membrane protein production experiments: quantitation of the host response. Protein Sci 14:1729–1740 Savchenko A, Yee A, Khachatryan A, Skarina T, Evdokimova E, Pavlova M, Semesi A, Northey J, Beasley S, Lan N, Das R, Gerstein M, Arrowmith CH, Edwards AM (2003) Strategies for structural proteomics of prokaryotes: quantifying the advantages of studying orthologous proteins and of using both NMR and X-ray crystallography approaches. Proteins 50:392–399 Locher KP, Lee AT, Rees DC (2002) The E. coli BtuCD structure: a framework for ABC transporter architecture and mechanism. Science 296:1091–1098 Terpe K (2003) Overview of tag protein fusions: from molecular and biochemical fundamentals to commercial systems. Appl Microbiol Biotechnol 60:523–533 Surade S, Klein M, Stolt-Bergner PC, Muenke C, Roy A, Michel H (2006) Comparative analysis and “expression space” coverage of the production of prokaryotic membrane proteins for structural genomics. Protein Sci 15:2178–2189 Terpe K (2006) Overview of bacterial expression systems for heterologous protein production: from molecular and biochemical fundamentals to commercial systems. Appl Microbiol Biotechnol 72:211–222 Grisshammer R, Tate CG (1995) Overexpres­ sion of integral membrane proteins for structural studies. Q Rev Biophys 28:315–422 Kunji ER, Slotboom DJ, Poolman B (2003) Lactococcus lactis as host for overproduction of functional membrane proteins. Biochim Biophys Acta 1610:97–108 Quick M, Javitch JA (2007) Monitoring the function of membrane transport proteins in detergent-solubilized form. Proc Natl Acad Sci USA 104:3603–3608 de Ruyter PG, Kuipers OP, de Vos WM (1996) Controlled gene expression systems for Lactococcus lactis with the food-grade inducer nisin. Appl Environ Microbiol 62:3662–3667 Geertsma ER, Poolman B (2007) Highthroughput cloning and expression in recalcitrant bacteria. Nat Methods 4:705–707

37

17. Poolman B, Konings WN (1988) Relation of growth of Streptococcus lactis and Streptococcus cremoris to amino acid transport. J Bacteriol 170:700–707 18. Wegmann U, O’Connell-Motherway M, Zomer A, Buist G, Shearman C, Canchaya C, Ventura M, Goesmann A, Gasson MJ, Kuipers OP, van Sinderen D, Kok J (2007) Complete genome sequence of the prototype lactic acid bacterium Lactococcus lactis subsp. cremoris MG1363. J Bacteriol 189:3256–3270 19. Nouaille S, Morello E, Cortez-Peres N, Le Loir Y, Commissaire J, Gratadoux JJ, Poumerol E, Gruss A, Langella P (2006) Complementation of the Lactococcus lactis secretion machinery with Bacillus subtilis SecDF improves secretion of staphylococcal nuclease. Appl Environ Microbiol 72:2272–2279 20. van der Sluis EO, Driessen AJ (2006) Stepwise evolution of the Sec machinery in Proteobacteria. Trends Microbiol 14:105–108 21. Saaf A, Monne M, de Gier JW, von Heijne G (1998) Membrane topology of the 60-kDa Oxa1p homologue from Escherichia coli. J Biol Chem 273:30415–30418 22. Driessen AJ, Zheng T, In’t Veld G, Op den Kamp JA, Konings WN (1988) Lipid requirement of the branched-chain amino acid ­transport system of Streptococcus cremoris. Biochemistry 27:865–872 23. Nakamura Y, Gojobori T, Ikemura T (2000) Codon usage tabulated from international DNA sequence databases: status for the year 2000. Nucleic Acids Res 28:292 24. Kouwen TR, van der Goot A, Dorenbos R, Winter T, Antelmann H, Plaisier MC, Quax WJ, van Dijl JM, Dubois JY (2007) Thiol-disulphide oxidoreductase modules in the low-GC Grampositive bacteria. Mol Microbiol 64:984–999 25. Steidler L, Hans W, Schotte L, Neirynck S, Obermeier F, Falk W, Fiers W, Remaut E (2000) Treatment of murine colitis by Lactococcus lactis secreting interleukin-10. Science 289:1352–1355 26. Steidler L, Neirynck S, Huyghebaert N, Snoeck V, Vermeire A, Goddeeris B, Cox E, Remon JP, Remaut E (2003) Biological containment of genetically modified Lactococcus lactis for intestinal delivery of human interleukin 10. Nat Biotechnol 21:785–789 27. Vandenbroucke K, Hans W, Van Huysse J, Neirynck S, Demetter P, Remaut E, Rottiers P, Steidler L (2004) Active delivery of trefoil factors by genetically modified Lactococcus lactis prevents and heals acute colitis in mice. Gastroenterology 127:502–513 28. Kunji ER, Chan KW, Slotboom DJ, Floyd S, O’Connor R, Monne M (2005) Eukaryotic

38

29. 30.

31.

32.

33.

34.

35.

36.

Geertsma and Poolman membrane protein overproduction in Lacto­ coccus lactis. Curr Opin Biotechnol 16: 546–551 Aslanidis C, de Jong PJ (1990) Ligationindependent cloning of PCR products (LICPCR). Nucleic Acids Res 18:6069–6074 Walhout AJ, Temple GF, Brasch MA, Hartley JL, Lorson MA, van den Heuvel S, Vidal M (2000) GATEWAY recombinational cloning: application to the cloning of large numbers of open reading frames or ORFeomes. Methods Enzymol 328:575–592 Liu Q, Li MZ, Leibham D, Cortez D, Elledge SJ (1998) The univector plasmid-fusion system, a method for rapid construction of recombinant DNA without restriction enzymes. Curr Biol 8:1300–1309 Guzman LM, Belin D, Carson MJ, Beckwith J (1995) Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J Bacteriol 177:4121–4130 Wang DN, Safferling M, Lemieux MJ, Griffith H, Chen Y, Li XD (2003) Practical aspects of overexpressing bacterial secondary membrane transporters for structural studies. Biochim Biophys Acta 1610:23–36 Chen Y, Song J, Sui SF, Wang DN (2003) DnaK and DnaJ facilitated the folding process and reduced inclusion body formation of magnesium transporter CorA overexpressed in Escherichia coli. Protein Exp Purif 32:221–231 Waldo GS, Standish BM, Berendzen J, Terwilliger TC (1999) Rapid protein-folding assay using green fluorescent protein. Nat Biotechnol 17:691–695 Drew DE, von Heijne G, Nordlund P, de Gier JW (2001) Green fluorescent protein as an indicator to monitor membrane protein overexpression in Escherichia coli. FEBS Lett 507:220–224

37. Drew D, Lerch M, Kunji E, Slotboom DJ, de Gier JW (2006) Optimization of membrane protein overexpression and purification using GFP fusions. Nat Methods 3:303–313 38. Geertsma ER, Groeneveld M, Slotboom DJ, Poolman B (2008) Quality control of overexpressed membrane proteins. Proc Natl Acad Sci USA 105:5722–5727 39. Kawate T, Gouaux E (2006) Fluorescencedetection size-exclusion chromatography for precrystallization screening of integral membrane proteins. Structure 14:673–681 40. Sambrook J, Russell DW (2001) Molecular cloning, a laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY 41. Terzaghi BE, Sandine WE (1975) Improved medium for lactic Streptococci and their bacteriophages. Appl Microbiol 29:807–813 42. Don RH, Cox PT, Wainwright BJ, Baker K, Mattick JS (1991) “Touchdown” PCR to circumvent spurious priming during gene amplification. Nucleic Acids Res 19:4008 43. Holo H, Nes IF (1989) High-frequency transformation, by electroporation, of Lacto­c occus lactis subsp. cremoris grown with glycine in osmotically stabilized media. Appl Environ Microbiol 55: 3119–3123 44. Berntsson RP, Alia Oktaviani N, Fusetti F, Thunnissen AM, Poolman B, Slotboom DJ (2009) Selenomethionine incorporation in proteins expressed in Lactococcus lactis. Protein Sci 18(5):1121–1127 45. El Khattabi M, van Roosmalen ML, Jager D, Metselaar H, Permentier H, Leenhouts K, Broos J (2008) Lactococcus lactis as expression host for the biosynthetic incorporation of tryptophan analogues into recombinant proteins. Biochem J 409(1): 193–198

Chapter 3 Mammalian Membrane Receptors Expression as Inclusion Bodies in Escherichia coli Bernard Mouillac and Jean-Louis Banères Abstract Integral membrane proteins, in particular receptors, regulate numerous physiological functions. The primary difficulty presented by their study in vitro is to obtain them in sufficient amounts in a functional state. Escherichia coli is a host of choice for producing recombinant proteins for structural studies. However, insertion of G-protein coupled receptors into its plasma membrane usually results in bacterial death. An alternative approach consists of targeting recombinant receptors to inclusion bodies, where they accumulate without affecting bacterial growth, and then fold them in vitro. We describe here a general approach that consists of accumulating the receptor in bacterial inclusion bodies, then purifying it under denaturing conditions. A simple assay is then described to screen for refolding conditions of the protein. Key words: Escherichia coli, GPCR, inclusion bodies, membrane protein, receptor, refolding

1. Introduction Integral membrane proteins constitute 20–30% of all the proteins encoded by the human and other eukaryotic genomes and are key players in cell communication. In this membrane protein family, receptors, particularly G protein-coupled receptors (GPCRs), are responsible for the majority of cellular responses to hormones and neurotransmitters as well as for the senses of sight, smell, and taste (1, 2). Several crystal structures of GPCRs have been published so far (3–6). Although these structures represent an important first step in this way, information is still needed to assess the molecular mechanisms governing the functioning of this membrane protein family. Crystallization of the different adrenergic receptors involved overcoming several technical barriers, including the production of large amounts of functional proteins and the stabilization of the

I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_3, © Humana Press, a part of Springer Science + Business Media, LLC 2010

39

40

Mouillac and Banères

native fold out of a membrane environment (7, 8). Overexpression of GPCRs has been carried out in different expression systems: yeast, insect cells, mammalian cells, or the retina rod cells of transgenic Xenopus laevis (for a review, see (9) and the following chapters of this volume). Among all the expression systems developed so far, Escherichia coli is considered as a host of choice for producing recombinant proteins for structural studies. However, toxicity and bacterial death is usually associated with GPCR expression in E. coli, resulting in poor expression yields (10). This is the reason for the limited number of receptors that have been produced as functional proteins inserted in the membrane of E. coli. Functional expression in E. coli with yields compatible with structural studies is essentially restricted to the rat neurotensin receptor (11) and the human adenosine A2a receptor (12). An alternative approach is to target recombinant receptors to inclusion bodies, where they can accumulate in milligram amounts without affecting bacterial growth, and then refold them in vitro. This approach has been successfully applied to a few receptors, including the olfactory receptor OR5 (13), the leukotriene B4 receptor BLT1 (14), the serotonin receptor 5-HT4(a) (15), and the chemokine receptor CXCR1 (16). The procedures described in the following outline the general trends for expression, purification of receptors from E. coli inclusion bodies, and a general screening assay for refolding conditions.

2. Materials 2.1. Protein Expression 2.1.1. Expression Vectors

On a routine basis, we use the pET21a expression vector series (Novagen). The vector has been modified by introducing the sequence encoding the fusion partner followed by a thrombin cleavage site between the NdeI and BamH1 sites (Fig. 3.1). This vector contains a C-terminal 6x His-tag. The fusion partners employed are intended to increase the amount of protein in the

Fusion partner

gene

thrombin site

6x His

Fig. 3.1. pET-derived expression vector harboring various fusion proteins. The fusion partner including the thrombin cleavage site is inserted in the pET21a vector between the NdeI and Bam HI sites. The restriction sites downstream of the Bam HI site that are part of the original vector are available for cloning of the required gene. This leaves the 6xHis encoding sequence at the C-terminus of the expressed fusion protein.

Mammalian Membrane Receptors Expression as Inclusion Bodies in Escherichia coli

41

inclusion bodies (see Note 1). Ampicillin is used as a selection marker for all the expression vectors. 2.1.2. Strains

Competent cells are needed for chemical transformation. For protein expression, the Rosetta (DE3) strain (Novagen) is used (see Note 2).

Protein Expression

1. Luna broth (LB) medium supplemented with ampicillin (100 mg/mL) and chloramphenicol (34 mg/mL). 2. Terrific broth (TB) medium supplemented with ampicillin (100 mg/mL) and chloramphenicol (34 mg/mL). 3. 1M isopropyl-b-d-galactopyranoside (IPTG). Make fresh solutions in water. 4. 20% (w/v) glucose solution. 5. 15-mL sterile tubes (Falcon or equivalent). 6. 2-L flasks. 7. Temperature-controlled shaker.

2.2. Protein Analysis

1. Protein gel electrophoresis system (SE250 minvertical unit, GE Healthcare or equivalent). 2. 4X protein sample buffer (120 mM Tris-HCl, pH 6.8, 4% (w/v) sodium dodecyl sulfate (SDS), 16% (v/v) glycerol, 0.08% (w/v) bromophenol blue, 2.96M 2-mercaptoethanol). 3. Separating buffer: 1.5M Tris-HCl, pH 8.8, 0.4% SDS. 4. Stacking buffer: 0.5M Tris-HCl, pH 6.8, 0.4% SDS. 5. 5X running buffer: 125 mM Tris, 1.25M glycine, 0.5% (w/v) SDS. 6. SDS polyacrylamide gel electrophoresis (PAGE) molecular weight standards (Bio-Rad or equivalent). 7. Gel staining solution (Coomassie blue R250 staining solution, Bio-Rad or equivalent). 8. Destaining solution (40% (v/v) ethanol, 10% (v/v) acetic acid).

2.3. Cell Disruption

1. Lysis buffer: 20 mM Tris-HCl, 300 mM NaCl, 0.5 mM phenylmethanesulfonylfluoride (PMSF), 10 mg/mL benzamidine, 5 mg/mL leupeptine, pH 8. 2. Washing buffer A: 20 mM Tris-Cl, 300 mM NaCl, 0.5 mM PMSF, 2M urea, 10 mg/mL benzamidine, 5 mg/mL leupeptine, pH 8. 3. Washing buffer B: 50 mM Tris-Cl, 0.5% Triton X-100, 0.5 mM PMSF, 10 mg/mL benzamidine, 5 mg/mL leupeptine, pH 8. 4. Sonicator with microprobe and large probe.

42

Mouillac and Banères

2.4. Solubilization, Thrombin Cleavage, and Protein Purification

1. Solubilization buffer: 100 mM NaH2PO4, 10 mM Tris-HCl, 6M urea, 0.8% SDS, 4 mM 2-mercaptoethanol, 10% glycerol, 5 mM imidazole, pH 8 (see Note 3). 2. Thrombin cleavage buffer: 50 mM Tris-HCl, 150 mM NaCl, 0.2% n-lauroyl-sarcosine. 3. Ni-NTA superflow resin (Qiagen). 4. Superose 6 resin (GE Healthcare). 5. 16 × 20 and 16 × 100 cm columns with adaptors (XK16/20 and XK16/100, GE Healthcare or equivalent). 6. Chromatographic system (Akta purifier or equivalent). 7. Thrombin (restriction-grade thrombin from Novagen or equivalent). 8. 12,000–14,000 MWCO (molecular weight cutoff) dialysis bags (Spectrapor or equivalent).

2.5. Refolding Assay

1. Solubilization buffer: 20 mM Tris-HCl, 50 mM NaCl, 0.8% SDS, pH 8. 2. Refolding buffer: 20 mM Tris-Cl, 50 mM NaCl, containing the required detergent or detergent/lipid mixture (see Note 4). 3. 1-mL Ni-NTA columns (1-mL His GraviTrap columns, GE Healthcare or equivalent). 4. Detergents (Anatrace) (see Note 4). 5. Asolectin (Fluka).

3. Methods 3.1. Protein Expression

1. Transform Rosetta- (DE3-) competent cells with the expression plasmid (see Note 5). Plate cells on LB agar plates containing appropriate antibiotics and incubate at 37°C. 2. Inoculate a single colony in 5 mL LB medium containing the appropriate antibiotic in a 15-mL tube, followed by growth for 16 h at 37°C with shaking. 3. Following growth, dilute the culture 1:100 into 500 mL of fresh TB medium (see Note 6) containing the appropriate antibiotic in a 2-L flask. Incubate at 37°C with shaking until the OD600 reaches 1 to 1.1 (see Note 7). 4. Add glucose 20% to reach a 0.2% final concentration (see Note 8). 5. Induce protein expression with 500 mL of 1M IPTG. 6. Incubate at 37°C with shaking for 4 h additional. 7. Harvest cells by centrifugation (4,000g for 15 min at 4°C).

Mammalian Membrane Receptors Expression as Inclusion Bodies in Escherichia coli

3.2. Cell Disruption

43

1. Resuspend cells in 20–30 mL lysis buffer (see Note 9). 2. Sonicate sample 5 min with large probe (see Note 10). 3. Centrifuge for 30 min at 27,000g at 4°C. Discard supernatant. 4. Resuspend the pellet in 10 mL lysis buffer. 5. Sonicate sample 5 min with microprobe. 6. Centrifuge for 30 min at 27,000g at 4°C. Discard supernatant. 7. Resuspend the resulting pellet in wash buffer A and incubate for 30 min at 4°C. 8. Centrifuge at 27,000g for 30 min. Recover the pellet. 9. Resuspend the pellet in wash buffer B and incubate for 30 min at 4°C (see Note 11). 10. Centrifuge at 27,000g for 30 min at 4°C. Recover the pellet.

3.3. Protein Solubilization and Purification

1. Resuspend the pellet in solubilization buffer. Let stand overnight at room temperature with slow stirring (see Note 12). 2. Centrifuge at 27,000g for 30 min; collect the supernatant. 3. Resuspend the pellet in solubilization buffer. Incubate for 2 h at room temperature with slow stirring. 4. Centrifuge for 30 min at 27,000g at 4°C. Collect the supernatant. 5. Mix both supernatants (see Note 13) and incubate overnight at room temperature with 10 mL of 50% Ni-NTA superflow slurry previously equilibrated in solubilization buffer (see Note 14). 6. Pour the resin in the 16/20 mm column, let the resin settle, discard the flowthrough and connect to the chromatographic system. 7. Wash with solubilization buffer at a 0.5-mL/min flow rate (see Note 15) while collecting 2-mL fractions. Wash until the absorbance of the eluted fractions reaches a zero value. 8. Wash with 10 volumes of a 10–60 mM imidazole gradient in solubilization buffer at a 0.5-mL/min flow rate (see Note 16). Collect 2-mL fractions. 9. Elute the fusion protein with 5 column volumes of solubilization buffer containing 200 mM imidazole. Collect 1-mL fractions. 10. Analyze the protein content of the fractions in the different elution peaks by SDS-PAGE on a 12% SDS-polyacrylamide gel (see Note 17). 11. Pour the fractions containing the receptor protein.

44

Mouillac and Banères

3.4. Thrombin Cleavage of the Fusion Protein

1. Dialyze the fusion protein extensively (see Note 18) in the thrombin cleavage buffer. 2. Determine the optimal cleavage conditions by carrying out a time-dependent digestion at 20°C of an aliquot of the fusion protein using increasing concentrations in thrombin (0.1–1 units of thrombin per milligram of fusion protein) after addition of CaCl2 up to 2.5 mM (see Note 19). 3. Digest the protein batch under the same conditions using the time and enzyme concentration determined in the timecourse experiment. 4. Stop the reaction by adding urea up to 6M and SDS up to 0.8% (see Note 20).

3.5. Purification of Isolated Receptor or Membrane Protein

1. Equilibrate 10 mL of Ni-NTA resin in solubilization buffer. 2. Incubate the digestion reaction with the resin slurry overnight at room temperature with slow stirring. 3. Mix the resin slurry and the thrombin-digested protein and incubate overnight at room temperature with slow mixing (see Note 14). 4. Pack the sample–resin suspension into the 16/20 column. Let the resin settle and discard the flowthrough. Connect to the chromatographic system. 5. Wash the resin (0.3-mL/min flow rate) with the solubilization buffer. Collect 2-mL fractions. Wash until the absorbance of the eluted fractions reaches a zero value. 6. Elute the protein with the solubilization buffer containing 200 mM imidazole. Collect 1-mL fractions (see Note 21). 7. Analyze the protein content of the fractions in the different elution peaks by SDS-PAGE on a 12% SDS-polyacrylamide gel. 8. Pour the fractions containing the receptor protein and dialyze (12,000–14,000 MWCO) into a 50 mM Tris-HCl, 0.8% SDS, pH 8 buffer (see Note 22). 9. Recover the dialyzed protein and, eventually, concentrate so that the final volume does not exceed 3–4 mL (see Note 23). 10. Equilibrate a Superose 6 column (16/100 cm) with a 50 mM Tris-HCl, 0.8% SDS, pH 8 buffer. 11. Load the protein on the column and elute with the 50 mM Tris-HCl, 0.8% SDS, pH 8 buffer (flow rate 0.2 mL/mL; 1-mL fractions). 12. Analyze the protein content of the fractions in the different elution peaks by SDS-PAGE on a 15% SDS-polyacrylamide gel. 13. Discard the aggregated fractions and pour the fractions containing the nonaggregated receptor protein.

Mammalian Membrane Receptors Expression as Inclusion Bodies in Escherichia coli

3.6. Refolding Assays

45

1. Equilibrate the required number of His GraviTrap columns in a 50 mM Tris-HCl, 0.8% SDS, pH 8 buffer. 2. Load 0.5 mg of purified protein into each of the column and wash with the 50 mM Tris-HCl, 0.8% SDS, pH 8 buffer. 3. Refold the protein by washing the column with 10 volumes of the refolding buffer. 4. Elute refolded protein with the refolding buffer containing 200 mM imidazole. 5. Analyze the protein content of the fractions in the elution peak by SDS-PAGE on a 15% SDS-polyacrylamide gel. 6. Dialyze into the refolding buffer to remove imidazole (see Note 24). 7. Assay for activity to determine the refolding efficiency (see Note 25).

4. Notes 1. Usually, membrane proteins, particularly receptors, do not accumulate in inclusion bodies when expressed in E. coli. Only a few examples have been reported (17). Different fusion partners have been developed to address membrane proteins to bacterial inclusion bodies. These include commercial ones such as the T7 peptide (14, 17), GST (15), or KSI (14, 18). In particular, a great deal of work has been done to systemize the expression of receptors as fusions with GST (19). Noncommercial fusion partners such as Bcl-X (20) or a fragment of the a-subunit of the a5b1 integrin (B. Mouillac and J.L. Banères, in preparation) have also been described that help addressing membrane proteins to inclusion bodies. 2. Usually, the Rosetta (DE3) strain is used to increase the expression yields. Depending on the protein, however, other strains such as the BL21 (DE3) one can be favorably used. 3. Although SDS and reducing agents can interfere with the purification of the nickel column, we have never observed increased leakage or resin damage at these 2-mercaptoethanol and SDS concentrations. 4. The refolding assay consists of screening different buffer conditions that will help in refolding the membrane protein and stabilizing its native fold. For this, different buffers will be tested that will contain different nonionic detergents, in the absence or the presence of lipids (we usually use natural lipid mixtures such as asolectin) and different salt concentrations.

46

Mouillac and Banères

Other parameters, such as the presence of different concentrations in reducing agents, can be tested. Due to the fact that the refolding is carried out with the protein immobilized on an Ni-NTA matrix, acidic pH cannot be tested as a parameter for protein refolding. 5. We use chemically competent cells from Novagen for heat shock transformation. Other methods, such as electroporation, can nevertheless be used. 6. A rich medium such as TB is used at this step to increase bacterial growth. TB can be replaced by 2YT or any other rich medium. 7. Protein synthesis is induced at high ODs to circumvent any toxicity problem. If no toxicity is observed (i.e., if protein induction does not stop cell growth), induction can be carried out at lower OD values. 8. Glucose is added at this step in case of toxicity of the expressed protein. 9. There is no need to fully suspend the E. coli at this step because sonication in the following step will take care of cell suspension. 10. The sonicator produces some heat in the sample. Sonication is therefore carried out while maintaining the protein in an ice bath. Sonication is usually carried out with repeated sequences corresponding to 30-s sonication followed by a 30-s lag period. 11. This step helps remove some residual membrane proteins. If the purity of the protein after the Ni-NTA chromatographic step is sufficient, it can nevertheless be omitted. 12. Incubating at room temperature is not a problem since 6M urea and 0.8% SDS will protect from any protease activity. Do not incubate at 4°C since this will lead to SDS crystallization. Unless otherwise indicated, all steps for which the buffer contains SDS are carried out at room temperature. 13. The protein content in the supernatant after solubilization can be assessed by SDS-PAGE. If the second solubilization supernatant still contains a high protein content, a third solubilization step can be carried out under the same conditions. 14. Overnight contact at room temperature is carried out to ensure full immobilization of the protein on the nickel matrix. However, the contact time can be reduced if necessary to a few hours without a significant decrease in the amount of immobilized protein. 15. If some leakage is observed at this step, the flow rate can be reduced.

Mammalian Membrane Receptors Expression as Inclusion Bodies in Escherichia coli

47

16. Depending on the protein considered, some elution can be observed during this washing step. In this case, the concentration of imidazole must be adjusted to prevent such an early elution of the target protein. 17. Do not boil the samples before loading on the gel. 18. Extensive dialysis is required to remove all the SDS that will inhibit thrombin. For this, several changes of the dialysis buffer are carried out. The two first dialyses are carried out at room temperature to prevent SDS crystallization, and then the dialysis is transferred at 4°C to prevent protein degradation. 19. Usually, the time-course digestion is carried out for 4–5 h at 20°C. If necessary, urea up to 1M can be added to the digestion buffer to limit nonspecific proteolysis at secondary sites. 20. If preferred, the digestion reaction can be stopped by adding a protease inhibitor such as PMSF. However, the addition of SDS at this stage is usually required to ensure full unfolding of the protein before the chromatographic step. 21. No washing at an intermediate imidazole concentration is required since this purification step is only intended for the purpose of separation of the fusion partner from the receptor protein. If contaminant proteins are still present, however, one can introduce washing steps at intermediate imidazole concentrations (20 mM, 40 mM). 22. To achieve good resolution, the volume of the loaded sample should not exceed 5% of the total column volume. 23. If some protein precipitation is observed, one can increase the salt or detergent concentration. Another possibility is to add some reducing agent, such as 2-mercaptoethanol. 24. If the imidazole does not interfere with the activity assay, this dialysis step can be omitted. 25. The activity assay will depend on the protein considered. The easiest way when dealing with a protein that displays well-defined ligand-binding properties is to monitor ligand binding. Otherwise, one can monitor indirect ligandbinding effects such as ligand-induced conformational changes or ligand-dependent recruitment or activation of associated proteins. Solubility cannot be considered as a stringent criterion since partially folded nonactive intermediates can be soluble in detergent solutions. In the same way, structural criteria such as circular dichroism or fluorescence spectra can hardly be considered in the absence of a reference for the native fold of the protein.

48

Mouillac and Banères

References 1. Bockaert J, Pin JP (1999) Molecular tinkering of G protein-coupled receptors: an evolutionary success. EMBO J 18:1723–1729 2. Bockaert J, Claeysen S, Becamel C, Pinloche S, Dumuis A (2002) G protein-coupled receptors: dominant players in cell-cell communication. Int Rev Cytol 212:63–132 3. Palczewski K, Kumasaka T, Hori T, Behnke CA, Motoshima H, Fox BA, Le Trong I, Teller DC, Okada T, Stenkamp RE, Yamamoto M, Miyano M (2000) Crystal structure of rhodopsin: a G protein-coupled receptor. Science 289:739–745 4. Cherezov V, Rosenbaum DM, Hanson MA, Rasmussen SG, Thian FS, Kobilka TS, Choi HJ, Kuhn P, Weis WI, Kobilka BK, Stevens RC (2007) High-resolution crystal structure of an engineered human beta2adrenergic G protein-coupled receptor. Science 318:1258–1265 5. Rasmussen SG, Choi HJ, Rosenbaum DM, Kobilka TS, Thian FS, Edwards PC, Burghammer M, Ratnala VR, Sanishvili R, Fischetti RF, Schertler GF, Weis WI, Kobilka BK (2007) Crystal structure of the human beta2 adrenergic G-protein-coupled receptor. Nature 450:383–387 6. Warne T, Serrano-Vega MJ, Baker JG, Moukhametzianov R, Edwards PC, Henderson R, Leslie AG, Tate CG, Schertler GF (2008) Structure of a beta1-adrenergic G-proteincoupled receptor. Nature 454:486–491 7. Kobilka B, Schertler GF (2008) New G-proteincoupled receptor crystal structures: insights and limitations. Trends Pharmacol Sci 29: 79–83 8. Serrano-Vega MJ, Magnani F, Shibata Y, Tate CG (2008) Conformational thermostabilization of the beta1-adrenergic receptor in a detergent-resistant form. Proc Natl Acad Sci USA 105:877–882 9. Sarramegna V, Talmont F, Demange P, Milon A (2003) Heterologous expression of G-protein-coupled receptors: comparison of expression systems from the standpoint of large-scale production and purification. Cell Mol Life Sci 60:1529–1546

10. Mancia F, Hendrickson WA (2007) Expression of recombinant G-protein coupled receptors for structural biology. Mol Biosyst 3:723–734 11. Tucker J, Grisshammer R (1996) Purification of a rat neurotensin receptor expressed in Escherichia coli. Biochem J 317:891–899 12. Weiss HM, Grisshammer R (2002) Purification and characterization of the human adenosine A(2a) receptor functionally expressed in Escherichia coli. Eur J Biochem 269:82–92 13. Kiefer H, Krieger J, Olszewski JD, Von Heijne G, Prestwich GD, Breer H (1996) Expression of an olfactory receptor in Escherichia coli: purification, reconstitution, and ligand binding. Biochemistry 35:16077–16084 14. Banères JL, Martin A, Hullot P, Girard JP, Rossi JC, Parello J (2003) Structure-based analysis of GPCR function: conformational adaptation of both agonist and receptor upon leukotriene B4 binding to recombinant BLT1. J Mol Biol 329:801–814 15. Banères JL, Mesnier D, Martin A, Joubert L, Dumuis A, Bockaert J (2005) Molecular characterization of a purified 5-HT4 receptor: a structural basis for drug efficacy. J Biol Chem 280:20253–20260 16. Park SH, Prytulla S, De Angelis AA, Brown JM, Kiefer H, Opella SJ (2006) Highresolution NMR spectroscopy of a GPCR in aligned bicelles. J Am Chem Soc 128: 7402–7403 17. Bane SE, Velasquez JE, Robinson AS (2007) Expression and purification of milligram levels of inactive G-protein coupled receptors in E. coli. Protein Expr Purif 52:348–355 18. Damian M, Martin A, Mesnier D, Pin JP, Baneres JL (2006) Asymmetric conformational changes in a GPCR dimer controlled by G-proteins. EMBO J 25:5693–5702 19. Kiefer H, Vogel R, Maier K (2000) Bacterial expression of G-protein-coupled receptors: prediction of expression levels from sequence. Receptors Channels 7:109–119 20. Thai K, Choi J, Franzin CM, Marassi FM (2005) Bcl-XL as a fusion protein for the high-level expression of membrane-associated proteins. Protein Sci 14:948–955

Chapter 4 Expression of Membrane Proteins at the Escherichia coli Membrane for Structural Studies Manuela Zoonens and Bruno Miroux Abstract Structural biology of membrane proteins is often limited by the first steps in obtaining sufficient yields of proteins because native sources are seldom. Heterologous systems like bacteria are then commonly employed for membrane protein over-expression. Escherichia coli is the main bacterial host used. However, overproduction of a foreign membrane protein at a non-physiological level is usually toxic for cells or leads to inclusion body formation. Those effects can be reduced by optimizing the cell growth conditions, choosing the suitable bacterial strain and expression vector, and finally co-expressing the target protein and the b-subunit of E. coli adenosine triphosphate (ATP)-synthase, which triggers the proliferation of intracytoplasmic membranes. This chapter is devoted to help the experimenter in choosing the appropriate plasmid/bacterial host combination for optimizing the amount of the target membrane protein produced in its correct folded state. Key words: Bacterial mutant, b-subunit of F-ATPase, co-expression, internal membrane

1. Introduction Among all the proteins encoded in the genome of eukaryotic cells, nearly 30% correspond to integral membrane proteins (MPs). Those proteins are involved in essential functions, such as import and export of nutriments, intercellular signalling and energy transduction. They also represent the majority of the targets of currently marketed drugs. Detailed knowledge of their atomic structures is therefore essential for understanding their function and dysfunction as well as for a wide range of biomedical and biotechnological applications. The total number of high-resolution MP structures reflects only about 1% of currently available structures in the protein data bank. The Web site

I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_4, © Humana Press, a part of Springer Science + Business Media, LLC 2010

49

50

Zoonens and Miroux

of Stephen White’s laboratory (http://blanco.biomol.uci. edu/Membrane_Proteins_xtal.html) lists the MP structures known so far, and this percentage represents 203 unique MPs after excluding the redundancy (two MPs of the same type but from different species are considered unique). Despite this scarcity, the number of MP structures progresses each year with the same rhythm as the number of soluble protein structures, but 25 years later (1). The reasons for this delay are accounted for by two main factors: a low level of naturally abundant MPs and poor biochemical stability in the presence of surfactants, which are used to extract MPs from biological membranes and keep them water soluble since the transmembrane surface of MPs is highly hydrophobic. To address the first difficulty, strategies are under development to overexpress MPs in recombinant systems; the second difficulty could be solved using mild surfactants for MP extraction or purification (cf. Chapter 14, this volume). With progress in synchrotron radiation, informatics power, and crystallographic techniques with, for instance, nanodrop technology, the amount of MPs required to obtain a threedimensional (3D) structure has been reduced by a factor of ten. However, the necessity to overexpress MPs is still essential, especially for MPs weakly expressed in the cells. Several strategies of MP overproduction have been elaborated, but recombinant expression in Escherichia coli remains the most useful and convenient system because of the simplicity of genetic manipulations associated with this bacteria. In spite of that, this protein source is limited by two complications: the MP targeting toward the membranes and the competition between the high-level expression of the desired MP and the production of other vital host MPs, leading to toxicity and cell death. To circumvent these hindrances to its growth, the bacterial host is able to accumulate the MP of interest as insoluble aggregates in the cytoplasm. These inclusion bodies are abundant but require a refolding step, which is a process with limited success rates. To produce low levels of proteins and then prevent inclusion body formation and toxicity, low-copy-number plasmids with weak promoters are usually used. Also, fusion proteins between the MP of interest and a partner known to target the membrane could aid the traffic of MPs. Successful attempts for MP expression in E. coli have also been obtained using toxicity as a genetic target for selecting resistant host mutants (2). Culminating this kind of mutants in the presence of the F-ATPase (adenosine triphosphatase) b-subunit has shown that bacteria were able to synthesize lipids and create a large amount of internal membranes where MPs might be inserted (3). A detailed protocol for this high-level expression strategy for MPs is presented in this chapter.

Expression of Membrane Proteins at the Escherichia coli Membrane for Structural Studies

51

2. Materials 2.1. Chemicals

1. 1X LB medium: 10 g Bacto tryptone, 5 g Bacto yeast extract, 5 g NaCl, sterile water to 1 L. Adjust to pH 7.2–7.5 with NaOH (a few pellets of the solid) and autoclave. 2. 2X TY medium: 16 g Bacto tryptone, 10 g Bacto yeast extract, 5 g NaCl, 2 g glucose, sterile water to 1 L. Adjust to pH 7.0 with NaOH (a few pellets of the solid) and autoclave. 3. IPTG (isopropyl-b-d-galactoside): A 1M stock can be prepared in water and stored at −20°C. IPTG can be obtained from Sigma-Aldrich. 4. Antibiotics: Working concentrations of some antibiotics: Antibiotic

Concentration (mg/mL)

Amp

Ampicillin

100

Kan

Kanamycin

25

A 1M stock of phenylmethanesulphonylfluoride (PMSF) can be prepared in methanol and stored at −20°C. This protease inhibitor is highly toxic. 2.2. Buffer Composition

TEP buffer: 10 mM Tris-HCl, pH8; 1 mM ethylenediaminetetraacetic acid (EDTA), and 0.001% (w/v) PMSF.

2.3. Web Resources

1. SMART (protein domain database): http://smart.embl-heidelberg.de/. 2. EMBL Protein Production: http://www.embl-hamburg.de/services/protein/production/ index.htmlPlease provide current URL; this one is invalid.. 3. ExPASy: http://www.expasy.ch/. 4. Membrane Proteins of Known 3D Structure: http://blanco.biomol.uci.edu/Membrane_Proteins_xtal. html.

52

Zoonens and Miroux

5. The QIAexpressionist handbook from Quiagen: http://www1.qiagen.com/literature/handbooks/literature.aspx?id = 1000068.

3. Methods 3.1. Designing Constructs for Expression

Over-expression of the b-subunit of the E. coli ATP synthase (UncF gene) has been shown to induce massive internal membrane synthesis in C43 (DE3) bacterial strain (3). Co-expressing your target MP with the b-subunit might create an ideal situation in which both MPs would insert correctly in those membranes. As a starting point, you should express your target MP alone and see whether inclusion body formation occurs or if part of it is targeted to the membrane. If your MP is partially targeted to the membrane, then before making any fusion or co-expression, select the appropriate host–vector couple that optimises the amount of the target protein produced while keeping the cells viable (see Section  4.3.3.). By doing so, you will increase the amount of protein targeted to the membrane, and the co-expression with the b-subunit might then be useful. If your target gene is produced as inclusion bodies only, then it is worth trying to fuse your target protein with an MP that is readily inserted to the membrane or that crosses the membrane to go to either the periplasm or the outer membrane. Maltosebinding protein (MBP also called MalE), Mistic protein, and the b-subunit of the ATPase are good candidates that help target heterologous MPs at the E. coli membrane. The final level of expression of the target gene and of the kinetics of induction of the gene is drive not only by the expression system but also by the intrinsic stability of the target messenger RNA (mRNA) and protein sequence (see Note 1). Therefore, it might be wise before starting any experiment to screen the literature for existing mutants that increase mRNA or protein stability or even increase the protein activity. Predictive programmes also allow the determination of secondary structure susceptible to RNA degradation (4) and of the half-life of your protein in various expression systems (see ExPASy Web site). Finally, Tate and colleagues reported and patented methods that increase thermostability of the mutated MP and the consecutive advantages for 3D crystallization of the proteins (see Note 2).

3.2. Choosing the Origin of Replication of the Expression Vectors

Table  4.1 lists the origin of replication currently used for overexpression of proteins in E. coli. Most plasmids are based on an origin of replication that derives from the colicin E and therefore are not inheritable. They require selective pressure to be maintained

Expression of Membrane Proteins at the Escherichia coli Membrane for Structural Studies

53

Table 4.1 Origins of replication used in expression vectors Origin of replication

Origin

Copy number

Classification

Type of plasmid

Source/ reference

ColE1

Colicin E1

30

Low copy

pQE

Quiagen

pMB1

Colicin E1

500–700

High copy

pUC, pRSET

Promega, (5)

par region of pSC101

pSC101

5

Very low but inheritable

pKVpP

(28, 29)

par mutants

pSC101

30–250

Low to high, inheritable

pJP serie

(9)

p15A

Compatible 15–60 with ColE1 or pSC101

Low copy

pACYC184

(6)

at a constant copy number into the cell; therefore, those plasmids carry genes that provide resistance toward an antibiotic such as ampicillin, chloramphenicol, kanamycin, or tetracycline. Plasmids that are derived from pBR322, such as all pET vectors, contain a low copy number ColE origin of replication and maintain around 20–40 copies per cell. In contrast, pUC-deriving vectors like pRSET expression vectors (5) carry an pMB1 origin of replication (ColE mutant), which increases the number of copies per cell up to 700. Compatible origins of replication have also been developed to maintain several plasmids into the strain. For instance, the p15A origin of replication (Table 4.1, (6)) has been used for the co-expression of lysosyme in BL21 (DE3) during overexpression of proteins (7). Finally, expression vectors with an inheritable origin of replication like pSC101 or pE1B1 from Vibrio anguillarium MVM425 strain (8) have been constructed. They are maintained into the cell at a high copy number in the absence of antibiotic and are also compatible with ColE-replicating plasmid (9). Choose the appropriate plasmids for your project. To express your protein alone or fused to a protein partner (b or MBP), select a vector with a high and a low copy number. To co-express your target gene with the b-subunit of the ATP synthase, use the high-copy-number vector pMW7 (2, 10) to fully induce membrane proliferation associated with the b-protein expression and clone your target gene in a low- and a high-copy-number vector.

54

Zoonens and Miroux

With such a large variety of vectors and eventually mutants to test, it is not difficult to become overloaded, so keep the number of constructs well within a limit you can realistically manage. 3.3. Choosing the Appropriate Promoter for the Expression Vector

Massive induction of internal membranes only occurs with high levels of the b-subunit. Lower levels of the b-subunit might not lead to internal membrane proliferation, while a higher level per cell triggers misfolding of the b-subunit and is toxic to the cell (3). As mentioned in Section 4.3.2., use the pMW7-b plasmid in combination with the C43 (DE3) bacterial strain to fully induce internal membrane synthesis. To co-express your target gene, you could use a T7 promoter-based plasmid, but you may also select an alternative promoter such as the arabinose- or rhamnoseinducible promoters (Table 4.2; see Note 3) so that your target gene is expressed after the metabolic adaptation of the bacteria and internal membrane proliferation (Fig. 4.1). When expressing your target protein as a fusion protein with the b-subunit, mistic or MBP, it is wise to use a strong but repressible system like the T7 expression system.

3.4. The Bacterial Host

Table 4.3 lists the main bacterial strains commercially available. The BL21 host is one of the most widely used bacterial strains essentially because it is deficient for the lon protease-encoding gene. Most of the bacterial hosts are RecA+, which could increase plasmid instability when the expression of heterologous proteins is toxic. In principle, the methods described in Section  4.3.2. remove the toxicity of the expression plasmid, but in case of persistent multimerization or instability of the plasmid you should move to an RecA− bacterial host such as BL21 BLR (see Note 4). Another parameter to consider before starting is the eventual presence of rare codon in the coding sequence of the target gene. If so, then use bacterial strains that provide transfer RNA (tRNA)

3.4.1. Identifying a Suitable Host Strain

Table 4.2 Promoters used for overproduction of membrane proteins Promoter

Origin

Repressor

Strength

Leak

Reference

lacUV5

lac mutant

lacI

Strong

High

(20)

Arabinose

Arabinose operon

Moderate

Very low

(30, 31)

rhaT

Rhamnose operon

Glucose

Strong

Very low

(7, 32)

T7

Bacteriophage T7

no

Very strong

High

(33)

T7lac

T7/lac hybrid

lacI

Very Strong

Moderate

(34)

Expression of Membrane Proteins at the Escherichia coli Membrane for Structural Studies

55

Fig. 4.1. Strategy for co-expressing a membrane protein with the b-subunit of the ATP synthase. The Escherichia coli C43 (DE3) bacterial strain is transformed with two plasmids, a high-copy pMW7-b vector containing the UncF gene encoding for the ATP synthase b-subunit and a second vector containing the target gene encoding for the membrane protein of interest. The cells from a single colony present on an LB agar plate are grown overnight in a small volume (50 mL) of 2X TY medium plus antibiotic at 37°C with shaking. The following day, the cells are inoculated 1/100 in a fresh 2X TY medium without antibiotic to promote protein production. The cell culture is grown at 37°C with shaking, and when OD600nm reaches about 0.4, IPTG is added to the flask. After overnight induction at 25°C, the bacteria have produced an internal membrane for insertion of the highly expressed b-subunit. The second inducer is then added to the medium, and cells are harvested by centrifugation 3 h after induction of the target gene.

for rare codons, such as BL21 Rosetta strain. If you encountered mRNA instability, then move to BL21 Star strain as the host is deficient for RNAseE activity (11). If you believe that disulphide bond formation is essential for the folding and the activity of your target gene, then consider using bacterial mutants (BL21 Origami, AD494) for which cytoplasmic disulphide bond formation is enhanced. Expression strains such as BL21(DE3)-pLys-S already carry a plasmid that is required for the regulation of the expression

56

Zoonens and Miroux

or that provide additional functions. Make sure it is compatible with your expression vectors, especially when co-expressing two MPs in two separate plasmids. 3.4.2. Selecting an Appropriate Bacterial Host to Your Expression Vectors

It is necessary to find a balance between the level of expression of your target MP and the toxicity associated with the expression of your protein. We and others found that when transcription of the target gene is too fast or too strong, the bacteria growth is impaired (2, 12). Low expression of the b-subunit or of your target MP will not trigger membrane proliferation. To select a bacterial host that fulfils both high-level expression and reduced toxicity conditions, proceed as follows: 1. Prepare a frozen stock of competent cells of all the bacterial host you have previously identified as suitable for your target gene. The basic CaCl2 procedure is sufficient (13). If the expression plasmids do not transform the low competent bacterial host, then the combination of host–vector is not appropriate. 2. Prepare LB or TY agar plates with ampicillin and with 0.7 mM IPTG. If you decided to use the Lemo21 (DE3) (7) or if your target promoter is non-IPTG inducible, then prepare a set of plates with the inducer(s) of your expression system. 3. Transform your plasmids into your set of bacterial hosts and plate half of the transformation on a plate without inducer and the other half on a plate containing the inducer of the expression system. The method relies on the fact that when the expression of the target MP is toxic, the bacteria are unable to form a colony on plates containing the inducer of the system. By analysing the presence, number and size of the colonies, you can estimate the degree of tolerance of the bacterial host during the production of the target gene (see Note 5, (2)). 4. When bacterial growth is not impaired by the expression of the target MP, there is a risk that the expression system is too weak either because the copy number of the plasmid is low or because the combination of the plasmid promoter with the bacterial host triggers weak induction of the expression. The target MP may be produced but at low concentration. When the presence of your target gene prevents the growth of the bacterial host or gives a pinhead colony phenotype in the presence of the inducer, then optimal expression of the target gene is compromised, and metabolic adaptation that is required to stimulate internal membrane proliferation will not occur. Therefore, choose the plasmid–host that produces normal-size colonies in the absence of inducer and colonies of 30–50% of the normal size in the presence of inducer.

Expression of Membrane Proteins at the Escherichia coli Membrane for Structural Studies

57

Table 4.3 Bacterial strains used for overexpression of membrane proteins Name

Genotype

Phenotype

B834

F-ompT hsdSB (rB-mB-)gal dcm met

Parental strain of BL21, methionine auxotroph

BL21

F-ompT hsdSB (rB-mB-)gal dcm

Protease deficient

Novagene

(37)

BL21 BLR

F-ompT hsdSB (rB-mB-)gal dcm,RecA

Low recombinase activity

Novagene

(38)

BL21 Origami, AD494

F-ompT hsdSB (rB-mB-)gal dcm, trxB/gor)

Enhanced cytoplasmic disulfide bond formation

Novagene

(39)

BL21 Tuner series

F-ompT hsdSB (rB-mB-)gal dcm LacY mutants

Impaired IPTG import

Novagene

BL21 Star

F-ompT hsdSB (rB-mB-)gal dcm,rne131 mutation

RnaseE deficiency

Invitrogene

BL21 Rosetta (2)

F-ompT hsdSB (rB-mB-)gal dcm pRARE(CamR)

pRARE supply rare tRNA

Novagene

BL21(DE3)

Source

Reference (35, 36)

(11)

Novagene

BL21(DE3) pLysS/E

Lysozyme inhibits the T7 RNA polymerase

Novagene

(40)

Lemo21(DE3)

PrhaBADT7LysY

Tunable lysozyme expression

Xbrane Bioscience AB

(7)

C41(DE3)

BL21(DE3) mutant

Reduced amount of lysozyme

Eurogenetec

(2, 7)

C43(DE3)

C41(DE3) mutant

Delayed induction of LacY

Eurogenetec

(2)

M15(pREP4)a

K12 derivative

T5/lac suitable strain pREP4 provide LacI

Quiagen

(41)

SG13009(pREP4)a

K12 derivative

T5/lac suitable strain pREP4 provide LacI

Quiagen

(42)

See the QIAexpressionist handbook, page 17, from Quiagen (http://www1.qiagen.com/literature/handbooks/ literature.aspx?id = 1000068) for more detailed information on those bacterial hosts a

58

Zoonens and Miroux

5. If none of the plasmid–host combinations is able to maintain the growth during the production of the target MP, then select the plasmid–host combination that prevents cell growth during expression and proceed to Section 4.3.4.3. 3.4.3. Obtaining Mutant Strains by Host Selection

The best method to fit the expression system to the target gene production is to select the bacterial host based on its ability to form colonies on the plate in the presence of inducer. The size of the colonies on IPTG-containing plates can be considered a signature for each vector–host couple (14). The protocol given next uses, as an example, the green fluorescent protein (GFP) from Aequora victoria (15). When inserted into the high-copy-number pMW7 T7-based vector (10), the expression of the GFP protein is toxic for the bacterial host BL21(DE3) (16). The method uses this toxicity to select the mutant on IPTG-containing plates. The steps are as follows: 1. Use a freshly transformed colony to inoculate 50 mL of LB medium. Monitor cell growth by measuring the optical density (OD) at 600 nm. When the OD reaches 0.4–0.6, induce expression of the GFP reporter gene by adding IPTG to 0.7 mM final concentration. 2. The growth behaviour of the bacteria in liquid or solid medium is the same. When the expression of the target gene is toxic, the bacteria will not form a colony on solid medium; in liquid medium, the bacteria will stop dividing as soon the expression of GFP starts. As a consequence, the OD of the culture will remain stable around 0.6. In severe cases, like MP over-expression, bacterial cells will eventually lyse shortly after induction, and the OD will decrease. If the optical density continues to increase 2 h after induction of the expression, then stop the experiment (see Note 6). 3. Make serial dilutions of the culture between 1:10 and 1:1,000 with LB medium or sterile water. Cells are then immediately plated on LB agar plates containing the appropriate inducer (e.g., 0.7 mM IPTG) and antibiotic. 4. After overnight incubation at 37°C, you should have mainly two populations of colonies on the plates. The frequency of appearance of the mutants is around 1/10,000. In the case of GFP, the interesting mutants are immediately revealed under ultraviolet (UV) light. As shown in Fig. 4.2, the intensity of GFP florescence is inversely correlated with the size of the colony. Most of the normal-size colonies are not fluorescent simply because the bacteria managed to lose the expression of the target gene while keeping the resistance to the antibiotic. Few colonies exhibiting a normal size phenotype are poorly fluorescent (see arrow, Fig. 4.2b). On the other hand, almost all small-size colonies are highly fluorescent, which exemplifies

Expression of Membrane Proteins at the Escherichia coli Membrane for Structural Studies

59

Fig. 4.2. Selection of bacterial hosts using the GFP as gene reporter. (a) size of the colonies after serial dilution of the BL21 (DE3) culture harbouring the pMW7-GFP plasmid on IPTG-containing plates. (b) Revelation of the GFP fluorescence by illuminating the Petri dishes with UV light.

that a reduction of 30–50% in the colony size is the best compromise between the level of expression of the target gene and the viability of the cell. To select bacterial hosts with expression of MPs, follow the same rules. 5. Select ten small-size colonies and one normal-size colony as a negative control and inoculate 1 mL culture in LB medium plus antibiotic. After overnight culture, save the mutant on a plate without IPTG and add 100 mL of the pre-culture to 1 mL of pre-warmed LB medium containing the antibiotic. Two hours later, add IPTG (0.7 mM) and mix 50 mL of the culture with 50 mL of SDS-PAGE (sodium dodecyl sulphate polyacrylamide gel electrophoresis) solubilisation solution at 3 and 16 h after induction (see Note 7). Load 10 mL of each culture on the appropriate acrylamide gel and check the level of expression of all mutants after Coomassie blue staining of the gel. Keep the mutant host–vector that expresses the highest level of the target gene. Repeat step 4 three to five times to isolate several independent mutants. 6. Once you have selected several mutants, you need to know whether the mutation is in the bacterial host or into the vector. For this, the cells must be cured of the plasmid. The simplest method consists of maintaining the selected strain for a week in LB medium without antibiotic. Make a serial dilution of the culture every day, inoculate a new LB medium culture and

60

Zoonens and Miroux

plate the cells on LB agar plates containing IPTG but without antibiotic. In the presence of IPTG, cells that still carry the expression vector will form small-size colonies, while cells that have spontaneously lost the plasmid will form normal-size colonies. Pick up two of those colonies and prepare competent cells following the CaCl 2 procedure. Retransform the original plasmid into the isolated host and verify that the small-size phenotype on an IPTG plate is restored. If you do not recover the phenotype on IPTG plates, then the mutation might be in the expression vector. To check this, prepare DNA plasmid from a 1-mL culture of your isolated clone and transform the original bacterial host. If the recovered plasmid restores the growth on an IPTG-containing plate then sequence the complete target gene to make sure that there is no mutation. 3.5. Optimisation of Growth Conditions Allowing Internal Membrane Proliferation

The best condition for triggering membrane proliferation on the induction of the b-subunit of the ATP synthase when using C43 (DE3) host strain with the pMW7-b plasmid is to grow cells in 2X TY medium to induce the cells at OD600nm 0.4–0.6 and to switch the temperature of the incubator to 25°C for an overnight induction. In these conditions, 100% of the cells survive the expression and keep the expression vector. However, if you coexpress your target gene with the b-subunit either in the same plasmid or in two different plasmids, you may have to optimise the growth condition again. A detailed protocol was given by Shaw and Miroux (16). Although internal membrane vesicle formation has been observed on low expression of the complete F1Fo ATP synthase, it is wise to keep high levels of the b-subunit to ensure high yield of membrane synthesis. 1. Make a 50-mL culture in a sterile 250-mL conical flask by inoculating 2X TY medium plus antibiotic with cells from a single colony. Grow overnight at 37°C with shaking at about 150–200 rotations per minute. 2. Pre-warm six 2-L Erlenmeyer flasks with 500 mL of 2X TY medium but without antibiotic (see Note 8). 3. Inoculate each flask with 5 mL of the pre-culture. Grow the cells at 37°C with shaking. Monitor cell growth by measuring the OD600nm every 30 min. 4. Induce cells at 0.4 OD600nm with 0.7 mM IPTG. Switch the temperature of the incubator to 25°C and leave the culture under agitation overnight. 5. Record the final OD600nm. When membrane proliferation is optimal, the final OD is about 9.

Expression of Membrane Proteins at the Escherichia coli Membrane for Structural Studies

3.6. Large-Scale Harvesting of Internal Membranes 3.6.1. Protocol for the Purification of the Internal Membrane After Overnight Induction at 25°C

61

The purification of the b-associated membranes is unusual when cells have been induced at 25°C. For some unclear reasons, the density of those membranes is high, probably because the membranes are associated with DNA and cell debris. Therefore, it is possible to collect them at 2,500g. However, after washing the pellet, the membrane density decreases, and ultracentrifugation is required as usual for cellular membranes. 1. Harvest the 3 L culture by centrifugation (2,000g, 10 min at 4°C). Pre-cool the cell of the French press in ice. 2. Resuspend the pellets in a final volume of 20 mL TEP buffer. 3. Disrupt the bacteria by passing the suspension twice through a French pressure cell at 4°C. 4. Collect the internal membranes by centrifugation at 2,500g, 10 min at 4°C. 5. Centrifuge the supernatant at 100,000g to collect the other membranes. 6. Wash the 2,500g pellet with 20 mL TEP buffer and centrifuge at 600g to remove the unbroken cells and debris. Centrifuge the supernatant again at 100,000g to pellet the internal membranes (see Note 9). 7. Resuspend the pellet in 5 mL, assay the protein concentration and load 10 mg of protein on an SDS-PAGE gel. The final yield of the b-subunit is around 30 mg/L of culture, and all of the b-subunit is inserted into the internal membranes.

3.6.2. Alternate Protocol for E. coli Membrane Purification

When the induction of the b-subunit in C43 (DE3) occurs at 37°C, the density of the internal membrane after passing through the French press is normal, and membranes do not pellet at 2,500g. Also, by co-expressing your target MP with the b-subunit, you may encounter partial inclusion body formation. Making additional membrane does not imply necessarily that your target MP will insert and fold correctly into those membranes. The following method allows you to separate inclusion bodies from E. coli membranes: 1. After disruption of the cells through the French press, pellet the unbroken cells at 600g for 10 min. 2. Take out the supernatant and collect the putative inclusion bodies at 10,000g for 15 min at 4°C. 3. Resuspend the pellet in 10 mL TEP buffer. 4. Centrifuge the supernatant at 100,000g for 2 h to collect the membrane fraction. 5. Assay the protein concentration of all fractions and load 20 mg on SDS-PAGE.

62

Zoonens and Miroux

4. Notes 1. An example of gene regulation by degradation of the mRNA was given in Arechaga et al.’s study (17). We observed that the a-subunit of the E. coli F1Fo (UncB gene) could not be overexpressed in E. coli. Fusion or co-expression of the a-subunit with GFP completely abolished the expression of GFP. Analysis of the mRNA levels of the UncB gene and gene fusion demonstrated that the middle region of the gene, encoding amino acids 92–171, was highly susceptible to endonucleolytic degradation. Fragments of the a-subunit other than 92–171 could be highly produced in fusion with GFP. 2. Detergent-solubilized GPCRs are conformationally unstable, shifting in a cyclic way between inactive and active states. Due to this intrinsic behaviour, a 3D crystal structure of those receptors is unattainable. Mutagenesis could increase the stability of GPCRs in one favourable conformational form. Tate and co-workers used this approach and obtained an increase in the thermostability of the b1-adrenergic receptor by introducing six point mutations (18). Those mutations were determined in two steps. First, all amino acid residues were individually replaced by an analine residue (alanines from the native sequence were replaced by leucine residues). The alanine scanning mutagenesis was used to define amino acid residues that altered the thermostability of the protein. Among a total of 318 single mutants made and expressed in E. coli and then tested for stability, 18 mutants showed higher stability than the wild type. Additional mutation experiments showed that the alanine scanning gave two-thirds of the best amino acid residues involved in thermostability of the protein. The second step was to combine mutations to make an optimally stable receptor. Polymerase chain reactions (PCRs) were performed by using various mixes of primers to combine up to five different mutations in a random manner. The mutants were then tested for thermostability and the most thermostable mutant produced by this approach increased by 21°C its apparent melting temperature, which is the temperature that gives a 50% decrease in functional ligand binding after heating the receptor for 30 min. This mutant was used for structural studies, and a 2.7-Å resolution crystal structure was successfully obtained (19). Such an approach could be applied to make increase MP amenability to crystallography. 3. The lacUV5 promoter is three times stronger than the natural lac promoter that regulates the lactose operon but is not so easy to repress (20). Tac promoter, an hybrid promoter deriving from the lac and trp (tryptophane operon) promoters, was

Expression of Membrane Proteins at the Escherichia coli Membrane for Structural Studies

63

designed to increase the strength of the expression system (21). The most powerful and widely used expression system is based on the T7 bacteriophage promoter (22). When infecting cells, the T7 phage uses the E. coli RNA polymerase to synthesize its own RNA polymerase (T7RNAPol), which is ten times more active than the E. coli one. Once produced, the T7RNAPol will specifically transcribe at high level only genes that are under the control of a T7 promoter. The activity of the T7RNAPol is so high that it does overload the transcription/translation machinery of the bacteria that ultimately lead to cell death (12). 4. In response to physical or chemical stress leading to inhibition of DNA replication, DNA breaks into the bacterial genome, or replication fork blockage, the cell initiates an SOS response to escape to death. This response includes the expression of 20 genes involved in DNA repair, recombination, and inhibition of cell division. The initial event consists of the activation RecA protein followed by the cleavage of the LexA repressor, leading to the activation of the SOS operon. Several authors have showed that introduction of high-copy-number plasmids containing a strong promoter could induce the SOS response. For instance, Aris et al. reported the induction of expression of recA and sfiA after induction of the strong lambda lytic promoter (23). Similarly, Benito et  al. observed an immediate arrest of cell DNA synthesis and cell growth as well as a stimulation of the SOS response on the expression of foot-and-mouse disease virus antigen (24). In some rare cases, there was a clear correlation between the activity of the recombinant protein and the toxicity to the cell (25, 26), and point mutations of the recombinant gene eventually removed the toxicity and increased the yield (27). However, in most of the cases, the toxicity to the cell was irrespective of the biological activity of the foreign gene but rather related to the expression of heterologous genes. The BL21 BRL host strain is the only one T7 expression host that is RecA deficient, which might compromise the induction of the SOS response during the overexpression of the target gene. 5. We have frequently observed that some expression systems are so leaky (i.e., overexpression occurs before induction of the expression system) that even in the absence of inducer on the plates, cells are unable to form regular-size colonies. In severe cases, such as the mitochondrial adenosine diphosphate (ADP)/ATP carrier or the b-subunit of the E. coli ATP synthase expressed in BL21 (DE3) with the pMW7 vector, cells form pinhead-sized colonies, which do not grow in liquid medium. In such cases, a new combination of vector and host needs to be found.

64

Zoonens and Miroux

6. When the expression of the target gene is toxic, expression plasmids are unstable, especially those based on the resistance toward ampicillin. We have observed that after an overnight pre-culture most of the cells had already lost the plasmid. In this case, the OD of the culture will continue to increase after induction of the expression of the target gene simply because the culture is overgrown by cells that have lost the expression of the target gene. To test the stability of the plasmid, it is recommended to make serial dilution of the pre-culture and, for the same dilution, compare the number of colonies on plates with or without antibiotic. If you find more colonies on the plate without antibiotic, then the plasmid is unstable, and the best strategy is to choose a new combination of vector and bacterial host. 7. Given that the colony has been selected on an IPTG-containing plate, it is also possible to speed up the protocol by inoculating the selected colony in 1 mL LB medium in the morning. At noon, an aliquot of the culture can be saved and spread out on an LB agar plate; then, IPTG is added for overnight induction, which is likely to give the best expression level of the target gene. 8. When you find a non-toxic combination of vector–bacterial host, then the expression plasmid can be stable. In these cases, we obtained better yield of protein production by removing the antibiotic in the large culture. 9. Although the internal membrane can be recovered at 2,500g after the disruption of the cells through the French press, once washed in TEP buffer, the membrane density is again low and needs 100,000g centrifugation to be collected.

Acknowledgments This work was supported by INSERM, CNRS and the “Agence Nationale de la Recherche” (grant 05-JCJC-0092-01 to B.M.).

References 1. White SH (2004) The progress of membrane protein structure determination. Protein Sci 13:1948–1949 2. Miroux B, Walker JE (1996) Over-production of proteins in Escherichia coli: mutant hosts that allow synthesis of some membrane proteins and globular proteins at high levels. J Mol Biol 260:289–298

3. Arechaga I, Miroux B, Karrasch S, Huijbregts R, de Kruijff B, Runswick MJ, Walker JE (2000) Characterisation of new intracellular membranes in Escherichia coli accompanying large scale over-production of the b subunit of F(1) F(o) ATP synthase. FEBS Lett 482:215–219 4. Lundberg U, Kaberdin V, von Gabain A (1999) The mechanism of mRNA degradation in

Expression of Membrane Proteins at the Escherichia coli Membrane for Structural Studies

5. 6. 7.

8.

9. 10.

11.

12.

13.

14.

15.

16.

bacteria and their implication for stabilization of heterologous transcripts. In: Demain AL, Davies JE (eds) Manual of industrial microbiology and biotechnology. ASM Press, Washington, DC, pp 585–596 Schoepfer R (1993) The pRSET family of T7 promoter expression vectors for Escherichia coli. Gene 124:83–85 Hiszczynska-Sawicka E, Kur J (1997) Effect of Escherichia coli IHF mutations on plasmid p15A copy number. Plasmid 38:174–179 Wagner S, Klepsch MM, Schlegel S, Appel A, Draheim R, Tarry M, Hogbom M, van Wijk KJ, Slotboom DJ, Persson JO, de Gier JW (2008) Tuning Escherichia coli for membrane protein overexpression. Proc Natl Acad Sci U S A 105:14371–14376 Zhang H, Wu H (2007) A novel high-copy plasmid, pEC, compatible with commonly used Escherichia coli cloning and expression vectors. Biotechnol Lett 29:431–437 Peterson J, Phillips GJ (2008) New pSC101derivative cloning vectors with elevated copy numbers. Plasmid 59:193–201 Way M, Pope B, Gooch J, Hawkins M, Weeds AG (1990) Identification of a region in segment 1 of gelsolin critical for actin binding. EMBO J 9:4103–4109 Lopez PJ, Marchand I, Joyce SA, Dreyfus M (1999) The C-terminal half of RNase E, which organizes the Escherichia coli degradosome, participates in mRNA degradation but not rRNA processing in vivo. Mol Microbiol 33:188–199 Dong H, Nilsson L, Kurland CG (1995) Gratuitous overexpression of genes in Escherichia coli leads to growth inhibition and ribosome destruction. J Bacteriol 177:1497–1504 Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning a laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY Walker JE, Miroux B (1999) Selection of Escherichia coli host that are optimized for the overexpression of proteins. In: Demain AL, Davies JE (eds) Manual of industrial microbiology and biotechnology. ASM Press, Washington, DC, pp 575–584 Chalfie M, Tu Y, Euskirchen G, Ward WW, Prasher DC (1994) Green fluorescent protein as a marker for gene expression. Science 263:802–805 Shaw AZ, Miroux B (2003) A general approach for heterologous membrane protein expression in Escherichia coli: the uncoupling protein, UCP1, as an example. Methods Mol Biol 228:23–35

65

17. Arechaga I, Miroux B, Runswick MJ, Walker JE (2003) Over-expression of Escherichia coli F1F(o)-ATPase subunit a is inhibited by instability of the uncB gene transcript. FEBS Lett 547:97–100 18. Serrano-Vega MJ, Magnani F, Shibata Y, Tate CG (2008) Conformational thermostabilization of the beta1-adrenergic receptor in a detergent-resistant form. Proc Natl Acad Sci U S A 105:877–882 19. Warne T, Serrano-Vega MJ, Baker JG, Moukhametzianov R, Edwards PC, Henderson R, Leslie AG, Tate CG, Schertler GF (2008) Structure of a beta1-adrenergic G-proteincoupled receptor. Nature 454:486–491 20. Arditti RR, Scaife JG, Beckwith JR (1968) The nature of mutants in the lac promoter region. J Mol Biol 38:421–426 21. de Boer HA, Comstock LJ, Vasser M (1983) The tac promoter: a functional hybrid derived from the trp and lac promoters. Proc Natl Acad Sci U S A 80:21–25 22. 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 23. Aris A, Corchero JL, Benito A, Carbonell X, Viaplana E, Villaverde A (1998) The expression of recombinant genes from bacteriophage lambda strong promoters triggers the SOS response in Escherichia coli. Biotechnol Bioeng 60:551–559 24. Benito A, Viaplana E, Corchero JL, Carbonell X, Villaverde A (1995) A recombinant footand-mouth disease virus antigen inhibits DNA replication and triggers the SOS response in Escherichia coli. FEMS Microbiol Lett 129:157–162 25. Laity JH, Shimotakahara S, Scheraga HA (1993) Expression of wild-type and mutant bovine pancreatic ribonuclease A in Escherichia coli. Proc Natl Acad Sci U S A 90:615–619 26. Bedouelle H, Guez V, Vidal-Cros A, Hermann M (1990) Overproduction of tyrosyl-tRNA synthetase is toxic to Escherichia coli: a genetic analysis. J Bacteriol 172:3940–3945 27. Sisk WP, Bradley JD, Kingsley D, Patterson TA (1992) Deletion of hydrophobic domains of viral glycoproteins increases the level of their production in Escherichia coli. Gene 112:157–162 28. Cohen SN, Chang AC, Boyer HW, Helling RB (1973) Construction of biologically functional bacterial plasmids in  vitro. Proc Natl Acad Sci U S A 70:3240–3244 29. Liu T, Chen JY, Zheng Z, Wang TH, Chen GQ (2005) Construction of highly efficient

66

Zoonens and Miroux

E. coli expression systems containing low oxygen induced promoter and partition region. Appl Microbiol Biotechnol 68:346–354 30. Cronan JE (2006) A family of arabinoseinducible Escherichia coli expression vectors having pBR322 copy control. Plasmid 55: 152–157 31. Guzman LM, Belin D, Carson MJ, Beckwith J (1995) Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J Bacteriol 177:4121–4130 32. Giacalone MJ, Gentile AM, Lovitt BT, Berkley NL, Gunderson CW, Surber MW (2006) Toxic protein expression in Escherichia coli using a rhamnose-based tightly regulated and tunable promoter system. Biotechniques 40:355–364 3 3. Studier FW, Moffatt BA (1986) Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J Mol Biol 189:113–130 34. Lama J, Carrasco L (1992) Inducible expression of a toxic poliovirus membrane protein in Escherichia coli: comparative studies using different expression systems based on T7 promoters. Biochem Biophys Res Commun 188: 972–981 35. Wood WB (1966) Host specificity of DNA produced by Escherichia coli: bacterial mutations affecting the restriction and modification of DNA. J Mol Biol 16:118–133 36. Leahy DJ, Hendrickson WA, Aukhil I, Erickson HP (1992) Structure of a fibronectin type III

domain from tenascin phased by MAD analysis of the selenomethionyl protein. Science 258: 987–991 37. Phillips TA, VanBogelen RA, Neidhardt FC (1984) lon gene product of Escherichia coli is a heat-shock protein. J Bacteriol 159: 283–287 38. Zhao JB, Wei DZ, Tong WY (2007) Identification of Escherichia coli host cell for high plasmid stability and improved production of antihuman ovarian carcinoma x antihuman CD3 single-chain bispecific antibody. Appl Microbiol Biotechnol 76:795–800 39. Prinz WA, Aslund F, Holmgren A, Beckwith J (1997) The role of the thioredoxin and glutaredoxin pathways in reducing protein disulfide bonds in the Escherichia coli cytoplasm. J Biol Chem 272:15661–15667 40. Studier FW (1991) Use of bacteriophage T7 lysozyme to improve an inducible T7 expression system. J Mol Biol 219:37–44 41. Reichel C, Brugger R, Bang H, Geisslinger G, Brune K (1997) Molecular cloning and expression of a 2-arylpropionyl-coenzyme A epimerase: a key enzyme in the inversion metabolism of ibuprofen. Mol Pharmacol 51:576–582 42. Gupta SK, Sharma M, Behera AK, Bisht R, Kaul R (1997) Sequence of complementary deoxyribonucleic acid encoding bonnet monkey (Macaca radiata) zona pellucida glycoprotein-ZP1 and its high-level expression in Escherichia coli. Biol Reprod 57: 532–538

Chapter 5 Membrane Protein Expression in Lactococcus lactis Annie Frelet-Barrand, Sylvain Boutigny, Edmund R.S. Kunji, and Norbert Rolland Abstract Membrane proteins play key roles in cellular physiology, and they are important drug targets. Approximately 25% of all genes identified in sequenced genomes are known to encode membrane proteins; however, the majority have no assigned function. Although the resolution of soluble protein structure has entered the high-throughput stage, only 100 high-resolution structures of membrane proteins have been described until now. Lactococcus lactis is a gram-positive lactic bacterium that has been used traditionally in food fermentations, but it is now used widely in biotechnology for large-scale overproduction of heterologously expressed proteins. Various expression vectors based on either constitutive or inducible promoters exist. The nisin-inducible controlled gene expression (NICE) system is the most suitable for recombinant membrane protein expression allowing for fine control of gene expression based on the autoregulation mechanism of the bacteriocin nisin. Recombinant membrane proteins can be produced with affinity tags for efficient detection and purification from crude membrane protein extracts. The purpose of this chapter is to provide a detailed protocol for the expression of membrane proteins and their detection using the Strep-tag II affinity tag in L. lactis. Key words: Cell disruptor, French press, Lactococcus lactis, membrane proteins, Strep-tag II

1. Introduction Lactococcus lactis is a homofermentative bacterium that is traditionally used for the production of lactic acid in fermented food products. In the last 20 years, remarkable progress has been made toward the development of genetic engineering tools and the molecular characterization of Lactococcus (1–3). This toolbox now includes transformation, gene integration, gene knockout, conjugation, and constitutive and regulated gene expression systems (for a review, see (4)). The availability of an easy-to-handle and strictly nisin-controlled gene expression system (NICE) has been essential

I. Mus-Veteau (ed.), Heterologous Expression of Membrane Proteins, Methods in Molecular Biology, vol. 601, DOI:10.1007/978-1-60761-344-2_5, © Humana Press, a part of Springer Science + Business Media, LLC 2010

67

68

Frelet-Barrand et al.

for the development of many applications, such as the expression of bacterial and viral antigens for safe vaccination, the production of human cytokines and other therapeutic agents for in situ treatments, or the production of specific compounds and pharmaceutical products (5–16). To take advantage of the nisin autoinduction mechanism for gene induction, the genes of the signal transduction system (nisK and nisR) have been isolated from the nisin operon and inserted into the chromosome of L. lactis subsp. cremoris MG1363 (nisin negative), generating the strain NZ9000 (Table 5.1; (17)). When a gene of interest is subsequently placed under the control of the inducible promoter PnisA in a plasmid such as pNZ8148 (Table 5.1; (14, 17)), the addition of subinhibitory amounts of nisin (0.1–5 ng/mL) to the culture medium induces the expression and the production of the corresponding protein (18–20) (Fig. 5.1). The advantages of L. lactis over Escherichia coli as an expression host are that it is a food-grade organism (including the selection marker), does not produce endotoxins, and does not form inclusion bodies. The organism possesses low protease activity, does not sporulate, is nonmotile, and has only one membrane. At present, the NICE system has become an essential tool for expressing and studying prokaryotic and eukaryotic membrane proteins (13–15, 21–23).

Table 5.1 Commonly used host strains and plasmids of the NICE system (all strains belong to the species Lactococcus lactis subsp. cremoris) Name

Characteristics

Reference

NZ9700

Progeny of the conjugation between nisin producer strain NIZO B8 with MG1614 (RifR StrpR derivative of MG1363). Carries nisin–sucrose transposon Tn5276. Often used as nisin producer strain for induction

(17, 18, 21)

NZ9000

Integration of nisRK into the pepN gene of strain MG1363 using plasmid pNZ9573. Most commonly used host of the NICE system

(17)

pNZ8048

NcoI site used for translational fusions, standard vector; CmR

(17)

pNZ8148

NcoI site used for translational fusions, deletion of 60-bp residual DNA from Bacillus subtilis, standard vector; CmR

(14)

Strains

Plasmids

Membrane Protein Expression in Lactococcus lactis

69

Fig. 5.1. Nisin-controlled gene expression system. After binding of nisin, the specific receptor NisK (histidine-protein kinase) autophosphorylates and transfers the phosphate group to NisR (response regulator). Activated NisR induces transcription from the target gene placed under the control of the promoter PnisA (Gene X). Depending on the presence or absence of the corresponding targeting signals, the protein is expressed in the cytoplasm or the membrane or secreted into the medium.

Once proteins have been produced, the next step is to purify them to recover amounts compatible with further functional (biochemical) and structural studies. Strep-tag II is an acceptable compromise between efficiency of purification and sufficient yield of membrane proteins (24–26). Various proteins have been successfully crystallized with Strep-tag II (27–31). Moreover, this tag has already been used for expression in gram-positive bacteria (32). Previous reports have detailed amplified gene expression in gram-positive bacteria and the reconstitution of purified membrane transport proteins in liposomes (21, 33). The purpose of this chapter is to provide detailed methodologies describing the subcloning of genes encoding membrane proteins tagged with Strep-tag II into NICE expression vectors, transformation and recombinant protein expression in L. lactis, and the detection of these proteins by Western blotting. Protocols include tips we have found to optimize the yield of recombinant membrane proteins to levels sufficient for functional and structural studies.

2. Materials 2.1. Cloning 2.1.1. Insert Preparation

  1. High-fidelity Taq DNA polymerase (e.g., Phusion HighFidelity DNA Polymerase; Finnzymes, Espoo, Finland) compatible with high GC-containing primers.

70

Frelet-Barrand et al.

2. Purification kit for extraction of DNA fragment from agarose gels (e.g., Nucleospin Extract II; Macherey Nagel). 3. Restriction endonucleases. 2.1.2. Vector Preparation

  1.  Plasmid pNZ8148 (NIZO, The Netherlands). 2. M17 medium: M17B (broth) or M17A (agar) (see Note 1). 3. Glucose: Prepare 20% (w/v) solution in Milli-Q water and sterilize by filtration through a 0.2-mm membrane (Minisart, Sartorius) under a hood bench. Store at 4°C. 4. Chloramphenicol: 34 mg/mL in absolute ethanol. Store at −20°C. 5. M17BG (M17 broth, 0.5% [w/v] glucose) medium or M17BGChl (M17BG, 10 mg/mL chloramphenicol) medium. 6. Plasmid DNA purification kit (e.g., Nucleospin Plasmid Kit; Macherey Nagel) that contains buffers used to remove lysozyme (see Sections 5.3.1.2. and 5.3.3.). 7. Lysozyme (Fluka, Sigma). 8. Kit for the purification of DNA fragments from agarose gels (e.g., Nucleospin Extract II Kit; Macherey Nagel). 9. Restriction endonucleases.

2.1.3. Ligation

For ligation, use T4 DNA ligase and the provided T4 DNA ligase buffer 10X (NEB).

2.2. Cell Culture and Transformation

  1.  Lactococcus lactis NZ9000 strain (NIZO). 2. G-SGM17B medium: M17 broth, 0.5M sucrose, 0.33M glycine. Sterilize the solution by autoclaving. Add 20% (w/v) glucose to a final concentration of 0.5%. 3. Medium A: 0.5M sucrose, glycerol 10% in Milli-Q water. Sterilize the solution by autoclaving. 4. Medium B: 0.5M sucrose, 10% glycerol, and 0.05M ethylenediaminetetraacetic acid (EDTA), pH 8.0, in Milli-Q water. Sterilize the solution by autoclaving. 5. M17BG medium. 6. M17AGChl medium: M17 agar, 0.5% (w/v) glucose, 10 mg/ mL chloramphenicol.

2.3. Plasmid Isolation and Verification

  1.  M17BGChl medium. 2. Plasmid DNA purification kit (e.g., NucleoSpin Plasmid, Macherey Nagel). 3. Lysozyme (Fluka, Sigma-Aldrich). 4. Restriction endonucleases. 5. Agarose gels.

Membrane Protein Expression in Lactococcus lactis

2.4. Glycerol Stock

For glycerol stock, use glycerol-RPE (Carlo Erba Reagents).

2.5. Expression and Induction

  1.  Lactococcus lactis NZ9700 strain (NIZO).

71

2. Laboratory glassware bottles (1-L Schott bottles). 3. M17G1%Chl medium: M17 medium containing 1% (w/v) glucose and 10 mg/mL chloramphenicol. 4. Incubator for cell growth (Innova 4230, New Brunswick Scientific).

2.6. Preparation of Membranes and Extraction of Proteins

  1. Phosphate-buffered saline (PBS) 20X, pH 7.4: For 1 L, dissolve 160 g NaCl (2.75M), 4 g KCl (0.05M), 4 g KH2PO4 (0.03M), 29 g Na2HPO4 (0.04M) in 900 mL Milli-Q water. Adjust to pH 7.4 with 10M NaOH and add Milli-Q water to make 1 L. Store at room temperature after sterilization. 2. K-EDTA 0.5M, pH 7.0, solution: Dissolve 29 g EDTA without Na (molecular weight (PM) = 292.24 g/mol, MP Bio­ chemicals) in 150 mL Milli-Q water. Adjust to pH 7.0 with KOH pellets and add Milli-Q water to bring to 200 mL. 3. PBS/glycerol: Add sterile 98% glycerol into PBS to a final concentration of 10% (v/v). Store at 4°C until use.

2.6.1. French Press

  1. Lysozyme (Fluka, Sigma-Aldrich) is dissolved at 10 mg/mL in PBS buffer. 2. Cold Milli-Q water. 3. French press: French® Pressure Cell Press (SIM-AMINCO, Spectronic Instruments).

2.6.2. Cell Disruptor

2.7. Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis

For the cell disruption system, use One Shot (Constant Cell Disruption Systems, Northants, UK) (see Note 2). 1. Acrylamide-bis ready-to-use solution, 30% (w/v) (37.5:1; Merck). 2. 8X Laemmli resolving gel buffer (3M Tris-HCl, pH 8.8): 60.6 g Tris-HCl, Milli-Q water to make 900 mL; adjust to pH 8.8 at 25°C with 12N HCl and add enough Milli-Q water to make 1 L. Store at room temperature. 3. 4X Laemmli stacking gel buffer (0.5M Tris-HCl, pH 6.8): 363 g Tris-HCl, water to a volume of 900 mL; adjust to pH 6.8 at 25°C with 12N HCl and add enough Milli-Q water to make 1 L. Store at room temperature. 4. Aqueous solution 20% (w/v) sodium dodecyl sulfate (SDS). 5. Ammonium persulfate: Prepare 10% (w/v) solution in water and immediately freeze in single-use (200-mL) aliquots at −20°C. 6. TEMED.

72

Frelet-Barrand et al.

7. Resolving gels (10% acrylamide): 3.3 mL of 30% (w/v) acrylamide solution, 1.25 mL of 8X Laemmli resolving gel buffer, 50 mL of 20% (w/v) SDS, 5.3 mL of water, 4 mL of TEMED, and 100 mL of 10% (w/v) ammonium persulfate. 8. Stacking gels (5% [w/v] acrylamide): 2.8 mL of 30% (w/v) acrylamide solution, 1.25 mL of 4X Laemmli stacking gel buffer, 25 mL of SDS 20% (w/v), 2.8 mL of water, 5 mL of TEMED, and 50 mL of 10% (w/v) ammonium persulfate. 9. Laemmli running buffer (10X): For 1 L, 144.2 g of glycine (192 mM), 30.3 g of Tris-HCl (25 mM); add 50 mL of 20% (w/v) SDS (0.1% final concentration) and enough Milli-Q water to make 1 L. Store at room temperature. 10. Molecular weight marker: Prestained Protein MW Marker (Fermentas). 11. Reducing sample buffer (4X): 0.08M Tris-HCl, pH 6.8, 40% (v/v) glycerol, 1% (w/v) SDS, 0.1 mM bromophenol blue, 10 mM DTT (dithiothreitol). Store at −20°C. 12. Sample buffer: 100 mL reducing buffer and 20 mL of 20% (w/v) SDS. 13. Control protein: Recombinant Strep-tag II fusion protein, MW about 28 kDa (0.1 mg/mL) (IBA, Goettingen, Germany). 2.8. Western Blotting

  1. System for protein transfer to nitrocellulose membranes (central core assembly, holder cassette, nitrocellulose filter paper, fibber pads, and cooling unit). 2. Protein transfer buffer: Running buffer 1X diluted with ethanol to obtain 20% (v/v) final ethanol concentration. Store at room temperature. 3. Protran nitrocellulose membrane from Schleicher and Schuell (Whatman) and 3 MM paper from Whatman. 4. Tris-buffered saline (TBS) 20X: For 1 L, dissolve 121.1 g Tris-HCl (1M), 175.3 g NaCl (3M); add 950 mL Milli-Q water and adjust to pH 7.4 with 12N HCl. 5. TBS-T: Dilute TBS 20X to obtain TBS 1X. Add 500 mL/L Triton X-100 (0.05% v/v). Store at room temperature. 6. Blocking buffer: 3% (w/v) bovine serum albine fraction V (Roche) in TBS-T. 7. Antibody: Strep-Tactin horseradish peroxidase (HRP) conjugate (IBA). 8. 100 mM Tris-HCl, pH 8. 9. Solution A: 90 mM P-coumaric acid (14 mg/mL in dimethyl sulfoxide [DMSO]). Store at 4°C and protect from light. 10. Hydrogen peroxide (H2O2) 30% (v/v) (Carlo Erba Reagents).

Membrane Protein Expression in Lactococcus lactis

73

11. Solution B: 250 mM luminol (3-aminophalhydrazin) (44 mg/mL in DMSO). Store at 4°C and protect from light. 12. Developer and fixer solutions (Kodak GBX). 13. Chemiluminescence-adapted films (Hyperfilm ECL, Amersham).

3. Methods 3.1. Cloning

Different cloning strategies have been developed to facilitate the insertion of cDNA into Lactococcus vectors (see Note 3). Here, we describe the simplest method, which consists of (1) polymerase chain reaction (PCR) amplification of the DNA fragment with two primers containing an NcoI site in the forward primer (see Note 4) and another restriction site from the multicloning site in the reverse primer (Table  5.2); (2) a digestion of the obtained fragment with these two restriction enzymes; and (3) the insertion of the DNA fragment into the plasmid previously digested with the same restriction endonucleases (Fig. 5.2).

Fig. 5.2. Map of the pNZ8148 vector for insertion of the gene of interest. The pNZ8148 vector contains an origin of replication (ORI), the gene for the resistance to chloramphenicol, two genes for the replication proteins repA and repC, the nisin-inducible promoter (PnisA), and the transcription terminator (T). The gene, tagged with a Strep-tag II (STREP) followed by a stop codon (*), can be inserted between the necessary NcoI site and another endonuclease site from the multicloning site (MCS) such as PstI, SphI, KpnI, SpeI, XbaI, SacI and HindIII.

74

Frelet-Barrand et al.

3.1.1. Insert Preparation

  1. Amplify the DNA fragment from either plasmid miniprep or complementary DNA (cDNA) library using the high-fidelity Taq polymerase (Phusion enzyme): 1 mL DNA (200 ng), GC Phusion buffer 1X, 0.8 mM forward primer containing the NcoI site (see Table 5.2 and Note 4), 0.8 mM reverse primer containing the Hind III site (see Note 5 and Table 5.2), 0.1 mM dNTP, 0.7 mL Phusion (1.4 U), and 1 mL DMSO (see Note 6). PCR program: 98°C, 30 s followed by 30 cycles of DNA amplification (98°C, 10 s; 58°C, 30 s; and 72°C, 30 s/kb) and one final incubation at 72°C for 10 min (see Note 7). 2. Load 2 mL of the 50-mL PCR reaction on an agarose gel to estimate the amount of amplified DNA and validate the size of the obtained DNA fragment. 3. Purify the amplified fragment using the NucleoSpin Extract II Kit (see Note 8). Elute the purified DNA in 20 mL. 4. Digest 1 mg of the amplified fragment with the corresponding enzyme (20 U NcoI and 20 U HindIII) for 2 h at 37°C (see Note 9). 5. Clean the digested amplified fragment using the NucleoSpin Extract II Kit. Elute the purified DNA in 20 mL (see Note 10). 6. Load 1 mL on an agarose gel for DNA quantification.

Table 5.2 Design of primers. Two types of forward primers have to be designed depending on the sequence of the gene of interest. They both contain the NcoI site (in bold) and four upstream nucleotides (CATG) that are required for higher efficiency of DNA restriction. If the gene sequence contains a G residue after the start codon, then let the sequence follow by the first 18 nucleotides of the coding region. If this is not the case, add two cytosines to produce an extra Ala codon behind the start codon ATG, followed by the first 18 nucleotides of the coding sequence. The reverse primer contains (1) a HindIII site (in bold) preceded by three nucleotides (CCC), which are required for higher efficiency of DNA restriction; (2) the stop codon (underlined) and (3) the Strep-tag II sequence (in bold and italic) followed by the 15 final nucleotides of the coding sequence of the reverse strand Primer name

Primer sequence (5¢-3¢)

fwd primer 1

CATGCCATGG + first 18 nucleotides of the gene of interest

fwd primer 2

CATGCCATGGCCATG + first 18 nucleotides of the gene of interest

rev primer

CCCAAGCTTCTACTTTTCGAACTGCGGGTGGCTCCA + last 15 nucleotides of the gene of interest of the reverse strand

Membrane Protein Expression in Lactococcus lactis 3.1.2. Vector Preparation

75

  1. Prepare 10 tubes (15-mL Falcon tubes) containing 5 mL of M17GChl medium inoculated with one clone (scrape into glycerol stock; see Section 5.3.4.) and incubate overnight at 30°C without shaking (see Notes 11 and 12). 2. Pellet the bacteria by centrifugation at 4,350g for 15 min at 4°C. 3. Resuspend the pellet in 250 mL of A1 buffer from NucleoSpin Plasmid Kit. 4. Add 12 mL of 50 mg/mL lysozyme (not provided in the plasmid DNA purification kit) (see Note 13). 5. Incubate the tubes at 37°C for 30 min for efficient digestion by lysozyme. 6. Return to the manufacturer’s instructions: Add A2 and A3 buffers and centrifuge at 16,100g for 10 min at room temperature. Transfer the supernatant to the provided column. 7. Wash the DNA with AW buffer to remove the lysozyme and with A4 buffer. 8. Elute the plasmid DNA in 20 mL of elution buffer into each Eppendorf tube. 9. Pool the plasmid DNA from the ten tubes and clean it using the NucleoSpin Extract II Kit. Elute in one tube in 20 mL of elution buffer. These two steps lead to sufficient amounts of pNZ8148 (see Note 11). 10. Digest 20 mg of pNZ8148 with 30 U NcoI and 30 U HindIII overnight at 37°C (see Note 14). 11. Purify the digested plasmid DNA using the NucleoSpin Extract II Kit. Elute the purified digested plasmid in 20 mL of elution buffer (see Note 10). 12. Load 1 mL of plasmid DNA on an agarose gel for quantification.

3.1.3. Ligation

  1. Mix together 50 ng of digested plasmid DNA and 250–300 ng of digested insert (see Note 15). 2. Add sterile water to 11 mL. 3. Incubate for 10 min at 45°C (see Note 16). 4. Place on ice for 3 min. 5. Add 3 mL of T4 DNA ligase buffer 10X and 3 mL of T4 DNA ligase (see Note 17). 6. Incubate overnight at room temperature (see Note 18). 7. Add sterile water in a quantity sufficient for 100 mL for the ligation, clean the ligation with the NucleoSpin Extract II Kit, and elute in 20 mL of sterile water (see Note 19).

76

Frelet-Barrand et al.

3.2. Transformation of Bacteria 3.2.1. Preparation of Electrocompetent Cells

  1. Inoculate 5 mL of G-SGM17B medium (in 15-mL Falcon tubes) with the glycerol stock at 30°C without shaking. 2. After 24 h, inoculate 50 mL of G-SGM17B medium with the 5 mL preculture overnight at 30°C without shaking. 3. The next morning, inoculate 400 mL of G-SGM17B medium with the 50 mL preculture. 4. Grow until OD600 reaches 0.2–0.3 (approximately 3 h). 5. Centrifuge for 20 min at 5,000g at 4°C and keep the pellet. 6. Wash the bacteria with 400 mL of medium A. 7. Centrifuge for 20 min at 5,000g at 4°C and keep the pellet. 8. Keep on ice for 15 min in 200 mL of medium B. 9. Centrifuge for 20 min at 5,000g at 4°C and keep the pellet. 10. Wash the pellet with 100 mL of medium A. 11. Centrifuge for 20 min at 5,000g at 4°C and keep the pellet. 12. Resuspend the bacteria in 4 mL of medium A. 13. Keep the bacteria in small aliquots (40 mL) at −80°C.

3.2.2. Electroporation

  1. Add 1 mL of plasmid DNA or 5 mL of ligation reaction to 40 mL of electrocompetent cells. 2. Store on ice for 1 min and transfer to the electroporation cuvette. 3. Use the following parameters: 2,000 V, 25 mF, 200 W. Press pulse. The time should be between 4.5 and 5 ms. 4. Add 1 mL of M17BG medium. 5. Store the cuvette on ice for 10 min and incubate at 30°C for 3 h. 6. Spread bacteria on two independent M17AGChl Petri dishes (1/10e and 9/10e of the volume). 7. Incubate for 1–2 days at 30°C.

3.3. Plasmid Isolation and Verification

  1. Prepare 10 tubes (15-mL Falcon tubes) containing 5 mL of M17GChl medium inoculated with independent clones and incubate overnight at 30°C without shaking. 2. To purify the plasmid DNA, follow the procedure described in Section 5.3.1.2. (steps 2–8). 3. Confirm the presence of the insert using these three methods: a. Single digestion to validate the presence of the insert. b. Double digestion to confirm the correct insert orientation. c. PCR amplification with a forward primer annealing on the vector sequence upstream of the NcoI site and a reverse primer specific for the gene sequence.

Membrane Protein Expression in Lactococcus lactis

3.4. Glycerol Stock (Normal/Concentrated)

77

  1.  For the normal glycerol stock, add 850 mL of the small 5 mL culture corresponding to a positive recombinant clone to 150 mL of sterile 100% glycerol and store at −80°C. 2. To prepare the concentrated glycerol stock, prepare 25 mL of M17G1%Chl medium and scrape the normal glycerol stock with tip. 3. Incubate overnight at 30°C (90 rpm). 4. Pellet the bacteria by centrifugation at 4350g for 15 min at 4°C and add sterile 100% glycerol to a final concentration of 15% (v/v) (see Note 20).

3.5. Expression and Induction

  1. Inoculate 10 mL of M17G1% medium with the concentrated glycerol stock of the Lactococcus NZ9700 strain.

3.5.1. Nisin Preparation

2. Incubate at 30°C with gentle agitation (90 rpm in our incubator) for 6 h. 3. Inoculate 500 mL of M17G1% medium in a Schott bottle with the 10 mL preculture. 4. Incubate at 30°C, 90 rpm, for 24 h. 5. Pellet the bacteria by centrifugation at 6,300g for 5 min at 4°C and transfer the supernatant into 50-mL Falcon tubes. Store theses tubes at −80°C until use (see Note 21).

3.5.2. Induction of Membrane Protein Expression

  1. Add 300 mL of the concentrated glycerol stock of bacteria to 25 mL of M17G1%Chl medium (see Note 22). 2. Incubate overnight at 30°C, 90 rpm. 3. Inoculate 1 L of M17G1%Chl medium in a Schott bottle with the 25 mL preculture (see Note 23). 4. Incubate at 30°C, 90 rpm, until OD600 reaches 0.8 (see Note 24) and induce with 5 mL of nisin (see Note 25). 5. Grow for an additional 4 h at 30°C; measure OD600 every hour (see Note 26). 6. Harvest the cells in buckets. Centrifuge at 5,000g for 20 min at 4°C. 7. Resuspend the pellet in 40 mL of PBS solution and transfer to a 50-mL Falcon tube. 8. Centrifuge at 5,000g for 15 min at 4°C. 9. Store the pellet at −20°C (see Note 27).

3.6. Preparation of Membranes and Extraction of Proteins 3.6.1. French Press

  1.  Resuspend the pellet in PBS-lysozyme solution and incubate at 30°C for 30 min (see Note 28). 2. Transfer the Falcon tube on ice. 3. Perform three cell lyses through French press with pressure at 18,000 psi (1.15 kbar) for each sample. Keep the lysate on ice. Wash the cylinder of the French press with cold Milli-Q water between each different recombinant bacterium.

78

Frelet-Barrand et al.

4. Centrifuge at 10,000g for 12 min at 4°C (see Note 29). 5. Transfer the supernatant into ultracentrifuge tubes. 6. Centrifuge at 150,000g for 1 h at 4°C and resuspend the pellet in 2 mL of PBS-glycerol solution. The supernatant contains the soluble proteins, while the membrane proteins are recovered in the pellet. 7. Take an aliquot of the membrane protein preparation to measure the protein concentration (see Note 30), add 5 mL of sample buffer and 5 mL of sterile water, and analyze the proteins by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) (see Section 5.3.7.). 8. Store the membrane proteins at −80°C. 3.6.2. Cell Disruptor

  1. Take out the bacterial pellets (see Section  5.3.5.2.) and let them thaw on ice (shake them from time to time to resuspend homogeneously). 2. Add 40 mL of PBS solution and 1.2 mL of 0.5M K-EDTA, pH 7.0 (see Note 31). 3. Disrupt cells at 35,000 psi (2.3 kbar) (see Note 32) and keep the lysate on ice. 4. Centrifuge at 10,000g for 12 min at 4°C and transfer the supernatant to ultracentrifuge tubes. 5. Resuspend the pellets in 10 mL of PBS solution and add 300 mL of 0.5M K-EDTA, pH 7.0 (see Note 33). 6. Disrupt cells at 35,000 psi (2.3 kbar) and keep the lysate on ice. 7. Centrifuge at 10,000g for 12 min at 4°C and transfer the supernatant to the corresponding ultracentrifuge tubes. 8. Centrifuge at 150,000g for 1 h at 4°C and resuspend the pellet with 2 mL of PBS-glycerol solution. 9. Take an aliquot of the membrane protein preparation to measure the protein concentration; add 5 mL of sample buffer and 5 mL of sterile water and analyze the proteins by SDS-PAGE (see Section 5.3.7.). 10. Store the membrane proteins at −80°C.

3.7. SDS-PAGE

  1. Prior to the experiment, prepare slab gels for protein electrophoresis (see Note 34). a. Prepare the gel apparatus according to the manufacturer’s specifications (see Note 35). b. Prepare the different gel solutions (stacking gel, 10% acrylamide separating gel). The volumes to be used are determined by gel dimensions and therefore by the specifications of the apparatus.

Membrane Protein Expression in Lactococcus lactis

79

2. Heat the protein samples at 95°C for 5 min to solubilize the proteins (see Note 36). Place protein samples (20 mg) into gel slots by means of a pipet. Load 6 mL of the molecular weight marker in another well and in three consecutive gel wells, respectively, 2 mg, 200 ng, and 20 ng of the control protein (Fig. 5.3). 3. Set the conditions for the electrophoresis at 150 V with constant voltage. Run gels for 1 h at room temperature (until the bromophenol blue dye reaches the lower part of the gel). 4. After electrophoresis, remove the gels for transfer and Western blotting analysis. 3.8. Western Blotting

  1. Prepare the cassette as follows: Add successively one fiber pad, two nitrocellulose filter papers, a nitrocellulose membrane, the gel, two nitrocellulose filter papers, and one fiber pad and then insert the sandwich into the holder cassette (the membrane should be placed beside the positive electrode). 2. Insert the cassette into the central core assembly unit (together with the cooling unit). 3. Perform the transfer for 1 h 30 min at 100 V in protein transfer medium. 4. Recover the nitrocellulose membrane. The following incubation and washing steps require agitation on a rocking plate at room temperature. 5. Rinse the membrane with water.

Fig. 5.3. Western blot analysis of six membrane proteins expressed in Lactococcus lactis. The six membrane proteins are of different origins and have various topologies: 1 Rhodobacter sphaeroides, 1 TM helix; 2 Rhodopseudomonas palustris, anchored to the membrane; 3 Pneumococcus, 1 TM helix; 4 Arabidopsis thaliana, 1 TM helix; 5 Arabidopsis thaliana, electrostatic interactions with membrane; and 6 Arabidopsis thaliana, 6 TM helices. 20 mg of total membrane proteins were loaded on a 10% SDS-PAGE gel and analyzed by Western blot using HRP conjugate specific to the Strep-tag II.

80

Frelet-Barrand et al.

6. Wash the membrane twice for 5 min with TBS-T. 7. Wash the membrane twice for 5 min with TBS (see Note 37). 8. Saturate the membrane with BSA-containing TBS-T for 1 h (see Note 38). 9. Wash the membrane three times for 5 min with TBS-T (see Note 39). 10. Incubate with antibody diluted at 1/10,000 in TBS-T for 1 h (see Note 40). 11. Wash the membrane twice for 5 min with TBS-T. 12. Wash the membrane twice for 5 min with TBS (see Note 41). 13. Mix 9 mL of 100 mM Tris-HCl, pH 8.0, with 40 mL of solution A and 5 mL of H2O2. 14. Mix 9 mL of 100 mM Tris-HCl, pH 8.0, with 90 mL of solution B. 15. Mix the two solutions in a dark room. 16. Incubate the nitrocellulose membrane for 1 min in the previous mixture. 17. Place the film successively in contact with the membrane in a Kodak Biomax Cassette for increasing times depending on the signal intensity. 18. Incubate the film in the developer solution until bands appear. 19. Wash the film in water and incubate in the fixer solution for 5 min.

4. Notes 1. M17 medium has been adapted for lactic streptococci (34). 2. One-shot (Constant Cell Disruption System, Northants, UK) disruption system is the most suitable system to disrupt the Lactococcus cell wall with the yield of crude membranes improved more than fivefold when compared to that obtained by lysozyme and French press treatment. 3. Geertsma and Poolman (2007) described a generic method for high-throughput cloning in recalcitrant bacteria (35). This method involves ligation-independent cloning in an intermediary Escherichia coli vector, which is rapidly converted via vector backbone exchange (VBEx) into an organism-specific plasmid ready for high-efficiency transformation. 4. In pNZ8148, NcoI is the unique restriction site placed immediately downstream of the nisin-inducible promoter for translational fusion. 5. The reverse primer contains one of the restriction sites located in the pNZ8148 multicloning site. Here, a Strep-Tag II tag

Membrane Protein Expression in Lactococcus lactis

81

was added for detection of the protein, but other affinity tags, such as His-tag, may be used. 6. Adding 1 or 2 mL DMSO to the PCR reaction will prevent the formation of secondary structure by the primers. 7. This type of PCR program is recommended to optimize the amplification with the Phusion polymerase enzyme. 8. This step allows purifying the PCR fragment and avoids the negative impact of the compounds present in the PCR solution on the efficiency of the restriction enzymes. 9. A partial digestion could also be performed when one or more restriction sites present in the multicloning sites of pNZ8148 are also present in the gene sequence (36). 10. This step allows purifying the digested PCR fragment and limits the impact of the compounds of the digestion on the efficiency of the ligation reaction. 11. pNZ8148 expression vector acts as a lower-copy-number plasmid when compared to standard vectors in E. coli. 12. Lactococcus lactis is aerotolerant and able to grow in anaerobic conditions and without shaking. 13. Lysozyme is an enzyme that digests the peptidoglycan in cell walls of gram-positive bacteria. 14. An overnight digestion allows for complete digestion of the plasmid (ready for 40 reactions). In these conditions, we have not observed star activities. 15. Usually, a 3:1 ratio (insert to vector) is recommended for a ligation (36). For L. lactis vectors, this ratio can be increased from 5:1 to 10:1 for optimal ligation rates. Large amounts of DNA (~1 mg) are required to obtain higher transformation yields. 16. The goal of this step performed at 45°C for 10 min and followed by 3 min incubation on ice is to melt any annealed cohesive ends. 17. The T4 DNA ligase is used in high proportions to obtain successful ligation. 18. The ligation temperature is optimum at room temperature instead of 16°C often used for ligation (36). 19. The ligation product has to be cleaned, and elution of the purified DNA has to be performed with sterile Milli-Q water to avoid problems during electroporation. 20. To obtain higher OD for precultures, the normal glycerol stock from preculture is replaced by a concentrated stock. 21. Nisin is produced in large amounts for reproducibility of experiments. It can be stored at −20°C for prolonged periods of time and even at −80°C. The Falcon tube is thawed on ice the morning before the induction.

82

Frelet-Barrand et al.

22. The culture medium used for induction and expression contains 1% glucose to allow optimum growth of Lactococcus (21). 23. Lactococcus lactis is able to grow in anaerobic conditions; the Schott bottles are thus suitable for this use and can be filled to the top. 24. Several values (from 0.5 to 0.8) for OD600 have been reported for induction of protein expression. In our hands, an OD600 of 0.8 is the optimum OD for induction (21). It is, however, recommended to test for optimum OD for the production of other proteins. 25. The volume of nisin to add for induction has been optimized to produce higher amounts of proteins and corresponds to 5 mL of nisin produced (see Section 5.3.5.1.). This volume has to be determined for every new preparation of nisin (12). 26. The induction time can be optimized depending on the expressed protein. For the proteins analyzed here, an induction time of 4 h allowed production of sufficient amounts of protein for our analyses. 27. The pellets containing intact bacteria can be stored at −20°C overnight. These pellets are used for the extraction of membrane proteins the following day. 28. The lysozyme that lyses peptidoglycans of the cell wall is essential for further breakage of the bacteria with the French press. The main problem with lysozyme is that the protein is hard to remove in subsequent purification steps. 29. This low-speed centrifugation allows the removal of remaining intact bacteria and cell wall components. 30. Protein content of membrane protein preparations can be estimated using the Bio-Rad protein assay reagent (37), but better results are obtained with the BCA assay (38). 31. K-EDTA prevents aggregation and thereby loss of membrane vesicles during low-speed centrifugation. 32. A pressure of 35,000 psi (2.3 kbar) is required to break Lactococcus, as recommended by the manufacturer. 33. This second step helps recover all bacteria and increases the yield. 34. We routinely use the procedure described by Chua (39) to separate membrane by SDS-PAGE. This chapter describes in detail all stock solutions and media for stacking and separating gels. 35. We use a Bio-Rad apparatus with 7 cm long gels.

Membrane Protein Expression in Lactococcus lactis

83

36. Heating of the samples for 5 min at 95°C is not necessary. Solubilization also works with an incubation of the samples with the buffer for 5 min at room temperature. 37. These two washing steps have been added to remove the remaining transfer solution, on the membrane, before incubation in the TBS-T solution. 38. The saturation of the membrane is performed for 1 h instead of overnight. This incubation is sufficient to completely block the membrane. 39. Washing the membrane three times for 5 min with TBS-T solution removes excess BSA and avoids nonspecific interaction with the antibody. 40. The antibody dilution adapted for Lactococcus has been reduced because of the presence of nonspecific bands when analyzing samples isolated from Lactococcus. 41. These two last washing steps have been increased to 5 min to properly wash the membrane to remove background and to enhance specific binding.

Acknowledgments This work has been supported by the Commissariat à l’Energie Atomique (CEA-PM project, A.F., S.B., and N.R.). We thank Shaun Peters and Daphné Seigneurin-Berny for critical reading of the manuscript. We thank Igor Mierau (NIZO, The Netherlands) for kindly allowing us to illustrate this chapter with the figure on the NICE system. References 1. Gasson MJ, de Vos WM (eds) (1994) Genetics and biotechnology of lactic acid bacteria. Blackie Academic and Professional, London 2. Wood BJB, Warner PJ (eds) (2003) Genetics of lactic acid bacteria. Kluwer Academic/ Plenum Publishers, New York 3. Teuber M, Geis A (2006) The genus Lactococcus. In: Dworkin M et al. (eds) The Prokaryotes, vol 4. Springer Verlag, New York, pp 205–228 4. Morello E, Bermúdez-Humarán LG, Llull D, Solé V, Miraglio N, Langella P, Poquet I (2008) Lactococcus lactis, an efficient cell factory for recombinant protein production and secretion. J Mol Microbiol Biotechnol 14:48–58

5. Hols P, Kleerebezem M, Schanck AN, Ferain T, Hugenholtz J, Delcour J, de Vos WM (1999) Conversion of Lactococcus lactis from homolactic to homoalanine fermentation through metabolic engineering. Nat Biotechnol 17:588–592 6. Hugenholtz J, Kleerebezem M, Starrenburg M, Delcour J, de Vos WM, Hols P (2000) Lactococcus lactis as a cell factory for highlevel diacetyl production. Appl Environ Microbiol 66:4112–4114 7. Hugenholtz J, Sybesma W, Groot MN, Wisselink W, Ladero V, Burgess K, van Sinderen D, Piard JC, Eggink G, Smid EJ, Savoy G, Sesma F, Jansen T, Hols P, Kleerebezem M (2002) Metabolic engineering

84

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

Frelet-Barrand et al. of lactic acid bacteria for the production of nutraceuticals. Antonie Van Leeuwenhoek 82:217–235 Nouaille S, Ribeiro LA, Miyoshi A, Pontes D, Le Loir Y, Oliveira SC, Langella P, Azevedo V (2003) Heterologous protein production and delivery systems for Lactococcus lactis. Genet Mol Res 2:102–111 Hanniffy S, Wiedermann U, Repa A, Mercenier A, Daniel C, Fioramonti J, Tlaskolova H, Kozakova H, Israelsen H, Madsen S, Vrang A, Hols P, Delcour J, Bron P, Kleerebezem M, Wells J (2004) Potential and opportunities for use of recombinant lactic acid bacteria in human health. Adv Appl Microbiol 56:1–64 Leroy F, Devuyst L (2004) Lactic acid bacteria as functional starter cultures for the food fermentation industry. Trends Food Sci Technol 15:67–78 Le Loir Y, Azevedo V, Oliveira SC, Freitas DA, Miyoshi A, Bermùdez-Humaran LG, Nouaille S, Ribeiro LA, Leclercq S, Gabriel JE, Guimaraes VD, Oliveira MN, Charlier C, Gautier M, Langella P (2005) Protein secretion in Lactococcus lactis: an efficient way to increase the overall heterologous protein production. Microb Cell Fact 4:2 Mierau I, Leij P, van Swam I, Blommestein B, Floris E, Mond J, Smid EJ (2005) Industrialscale production and purification of a heterologous protein in Lactococcus lactis using the nisin controlled gene expression system NICE: the case of lysostaphin. Microb Cell Fact 4:15 Mierau I, Olieman K, Mond J, Smid EJ (2005) Optimization of the Lactococcus lactis nisincontrolled gene expression system NICE for industrial applications. Microb Cell Fact 4:16 Mierau I, Kleerebeem M (2005) 10 years of the nisin-controlled gene expression system (NICE) in Lactococcus lactis. Appl Microbiol Biotechnol 68:705–717 Zhou XX, Li WF, Ma GX, Pan YJ (2006) The nisin-controlled gene expression system: construction, application and improvements. Biotechnol Adv 24:285–295 Wells JM, Mercenier A (2008) Mucosal delivery of therapeutic and prophylactic molecules using lactic acid bacteria. Nat Rev Microbiol 6:349–362 Kuipers OP, de Ruyter PGGA, Kleerebezem M, de Vos WM (1998) Quorum sensing-controlled gene expression in lactic acid bacteria. J Biotechnol 64:15–21 Kuipers OP, Beerthuyzen MM, Siezen RJ, de Vos WM (1993) Characterization of the nisin

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

gene cluster nisABTCIPR of Lactococcus lactis. Requirement of expression of the nisA and nisI genes for development of immunity. Eur J Biochem 216:281–291 de Ruyter PG, Kuipers OP, Beerthuyzen MM, Alen-Boerrigter I, de Vos WM (1996) Functional analysis of promoters in the nisin gene cluster of Lactococcus lactis. J Bacteriol 178:3434–3439 de Ruyter PG, Kuipers OP, de Vos WM (1996) Controlled gene expression systems for Lactococcus lactis with the food-grade inducer nisin. Appl Environ Microbiol 62:3662–3667 Kunji ERS, Slotboom DJ, Poolman B (2003) Lactococcus lactis as host for overproduction of functional membrane proteins. Biochim Biophys Acta 1610:97–108 Kunji ERS, Chan KW, Slotboom DJ, Floyd S, O’Connor R, Monné M (2005) Eukaryotic membrane protein overproduction in Lactococcus lactis. Curr Opin Biotechnol 16:546–551 Monné M, Chan KW, Slotboom DJ, Kunji ERS (2005) Functional expression of eukaryotic membrane proteins in Lactococcus lactis. Protein Sci 14:3048–3056 Lichty JJ, Malecki JL, Agnew HD, MichelsonHorowitz DJ, Tan S (2005) Comparison of affinity tags for protein purification. Protein Exp Purif 41:98–105 Rahman M, Ismat F, McPherson MJJ, Baldwin SA (2007) Topology-informed strategies for the overexpression of membrane proteins. Mol Memb Biol 24:407–418 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 Roberts SA, Wildner GF, Grass G, Weichsel A, Ambrus A, Rensing C, Montfort WR (2003) A labile regulatory copper ion lies near the T1 copper site in the multicopper oxidase CueO. J Biol Chem 278:31958–31963 Korndorfer IP, Schlehuber S, Skerra A (2003) Structural mechanism of specific ligand recognition by a lipocalin tailored for the complexation of digoxigenin. J Mol Biol 330:385–396 Korndorfer IP, Dommel MK, Skerra A (2004) Structure of the periplasmic chaperone Skp suggests functional similarity with cytosolic chaperones despite differing architecture. Nat Struct Mol Biol 11:1015–1020 Breustedt DA, Korndorfer IP, Redl B, Skerra A (2005) The 1.8-Å crystal structure of human tear lipocalin reveals an extended branched cavity with capacity for multiple ligands. J Biol Chem 280:484–493

Membrane Protein Expression in Lactococcus lactis 31. Tang J, Ebner A, Ilk N, Lichtblau H, Huber C, Zhu R, Pum D, Leitner M, Pastushenko V, Gruber HJ, Sleytr UB, Hinterdorfer P (2008) High-affinity tags fused to s-layer proteins probed by atomic force microscopy. Langmuir 24:1324–1329 32. Nguyen HD, Phuong Phan TT, Schumann W (2007) Expression vectors for the rapid purification of recombinant proteins in Bacillus subtilis. Curr Microbiol 55:89–93 33. Poolman B, Doeven MK, Geertsma ER, Biemans-Oldehinkel E, Konings WN, Rees DC (2005) Functional analysis of detergentsolubilized and membrane-reconstituted ABC Transporters. Meth Enzymol 400: 429–459 34. Terzaghi BE, Sandine WE (1975) Improved medium for lactic streptococci and their bacteriophages. Appl Microbiol 29:807–813

85

35. Geertsma ER, Poolman B (2007) Highthroughput cloning and expression in recalcitrant bacteria. Nat Methods 4:705–707 36. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory Press, New York 37. Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye bonding. Anal Biochem 72: 248–254 38. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC (1985) Measurement of protein using bicinchoninic acid. Anal Biochem 150:76–85 39. Chua NH (1980) Electrophoretic analysis of chloroplast proteins. Methods Enzymol 69: 434–436

Chapter 6 Heterologous Expression of Human Membrane Receptors in the Yeast Saccharomyces cerevisiae Olivier Joubert, Rony Nehmé, Michel Bidet, and Isabelle Mus-Veteau Abstract Due to their implication in numerous diseases like cancer, cystic fibrosis, epilepsy, hyperinsulinism, heart failure, hypertension, and Alzheimer disease, membrane proteins (MPs) represent around 50% of drug targets. However, only 204 crystal structures of MPs have been solved. Structural analysis requires large quantities of pure and active proteins. The majority of medically and pharmaceutically relevant MPs are present in tissues at low concentration, which makes heterologous expression in large-scale productionadapted cells a prerequisite for structural studies. The yeast Saccharomyces cerevisiae is a convenient host for the production of mammalian MPs for functional and structural studies. Like bacteria, they are straightforward to manipulate genetically, are well characterized, can be easily cultured, and can be grown inexpensively in large quantities. The advantage of yeast compared to bacteria is that they have proteinprocessing and posttranslational modification mechanisms related to those found in mammalian cells. The recombinant rabbit muscle Ca2+-ATPase (adenosine triphosphatase), the first heterologously expressed mammalian MP for which the crystal structure was resolved, has been produced in S. cerevisiae. In this chapter, the focus is on expression of recombinant human integral MPs in a functional state at the plasma membrane of the yeast S. cerevisiae. Optimization of yeast culture and of MP preparations is detailed for two human receptors of the Hedgehog pathway: Patched and Smoothened. Key words: GPCR, membrane protein, yeast, Saccharomyces cerevisiae, Hedgehog, heterologous expression, Patched, Smoothened

1. Introduction Membrane proteins (MPs) have a main role in cell homeostasis and signal transducing. They are involved in several pathologies, such as cancers and neurodegenerative diseases. Surprisingly, the structural data available on MPs are scarce. MPs represent around 50% of drug targets; however, we know only 204 unique structures of those proteins (http://blanco.biomol.uci.edu/membrane_ proteins_xtal.html), less than 1% of the known protein structures.

I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_6, © Humana Press, a part of Springer Science + Business Media, LLC 2010

87

88

Joubert et al.

The main bottleneck is that their natural expression is weak, making heterologous expression in host cells compatible with largescale production a prerequisite to obtain the amount of proteins necessary for structural studies. The yeast Saccharomyces cerevisiae is a convenient host for the production of mammalian MPs for functional and structural studies. Like bacteria, the yeast are straightforward to manipulate genetically, are well characterized, can be easily cultured, and can be grown inexpensively in large quantities. The advantage of yeast compared to bacteria is that they have protein-processing and posttranslational modification mechanisms related to those found in mammalian cells. However, yeast glycosylation patterns do differ somewhat from those of mammalian cells; in particular, S. cerevisiae has been shown to hyperglycosylate proteins (1). Yeast also clearly differ from mammalian cells in the lipid composition of their membranes, which can affect the functionality of recombinantly expressed MPs (2). However, the first structure of a heterologously expressed mammalian MP, the rabbit sarcoendoplasmic reticulum Ca2+-ATPase (adenosine triphosphatase) SERCA1A (Sarcoendoplasmic reticulum calcium adenosine triphosphatase (3)), was obtained after heterologous expression in the yeast S. cerevisiae. Moreover, this structure is identical to that obtained after purification of the native protein from rabbit muscle (4). This result is encouraging for the development of the heterologous expression of other mammalian MPs in the yeast S. cerevisiae to achieve additional structures. Our team succeeded in expressing the two human membrane partners of the Hedgehog pathway Patched (Ptc) (5) and Smoothened (Smo) (6). We detail here the different optimizations carried out for the functional production of these human MPs in S. cerevisiae.

2. Materials 2.1. Strains

1. Escherichia coli DH5a for vector preparation. 2. Saccharomyces cerevisiae strain K699 (Mata, ura3, and leu 2–3), generously given by R. Arkowitz for heterologous expression.

2.2. Vector Preparation

1. Yeast Episomal Plasmid (YEP) PMA vector containing the plasma membrane proton ATPase (PMA) promoter, generously given by Dr. Al-Shawi (7). 2. Complementary DNAs (cDNAs) from target MPs. 3. Proofstart polymerase (Qiagen). 4. A polymerase chain reaction (PCR). 5. pCR™2.1 plasmid (Invitrogen).

Heterologous Expression of Human Membrane Receptors in the Yeast

89

6. Restriction enzymes. 7. Ligase. 2.3. Yeast Transformation

1. Plate mixture: 1M lithium acetate, 90% (w/v) polyethylene glycol (PEG) 4000, 1M Ultra-Pure™ Tris-HCl, pH 7.5, 0.5M ethylenediaminetetraacetic acid (EDTA) disodium salt dihydrate. 2. 10 mg/mL salmon sperm DNA. 3. 1M 1,4-dithio-D-L threitol-threitol (DTT) (Euromedex). 4. Huber Ministat water bath set at 42°C. 5. Plasmid. 6. Sterile flasks.

2.4. Yeast Culture 2.4.1. YNB Medium (Minimal Medium) (Composition for 1 L)

1. Dissolve 8 g of yeast nitrogen base (YNB) without amino acids, 55 mg tyrosin, 55 mg uracil, and 55 mg adenosine hemisulfate in 900 mL of Milli-Q water. 2. For solid plates, add 20% (w/v) agar. 3. Autoclave at 120°C. Store at room temperature.

2.4.2. 10X D: (20% d-(+)-Glucose)

1. Dissolve progressively 200 g of D-glucose (minimum 99.5%) in 800 mL of Milli-Q water and then adjust to 1 L with Milli-Q water. 2. Filter sterilize on 0.22-mm filter and store at +4°C. The shelf life of this solution is approximately 1 year.

2.4.3. 100X AA (Drop Out) (0.5% of Each Amino Acid)

1. Dissolve 500 mg each of l-arginine, l-histidine, l-iso-leucine, l-leucine, l-lysine, l-methionine, l-phenylalanine, l-threonine, and l-tryptophane in 100 mL of water. 2. Filter sterilize on 0.22-mm filter and store in aliquots at −20°C. The shelf life of this solution is approximately 1 year.

2.4.4. YEP Medium (Rich Medium): (Composition for 1 L)

1. Dissolve 11 g yeast extract, 22 g bacteriological peptone, 55 mg adenine hemisulfate in 900 mL of Milli-Q water.

2.4.5. Other Materials

1. Erlenmeyer flasks.

2. Autoclave at 120°C. Store at room temperature.

2. Spectrophotometer (Eppendorf) and cuvettes. 3. INNOVA™ 44 (New Brunswick Scientific) incubator shaker.

2.5. Yeast Membrane Fraction Preparation

1. 50 mM Ultra-Pure™ Tris-HCl (pH 7.4). 2. 150 mM sodium chloride (NaCl). 3. 2.5 mM EDTA disodium salt dihydrate, pH 8.3 (NaOH).

90

Joubert et al.

4. 4 mM benzamidine hydrochloride hydrate minimum 98%. 5. 1 mM phenylmethylsulfonylfluoride (PMSF). 6. Glass beads, acid washed, 425–600 mm. 7. 10% glycerol. 8. Heidolph™ multireax vortexer. 2.5.1. For solution preparation:

1. Buffer A (2X): Dissolve 12.114 g Tris-HCl, 17.53 g NaCl in 900 mL Milli-Q water. Adjust to pH 7.4 by adding HCl. Adjust to 1 L with Milli-Q water. This is the grinding buffer. 2. Buffer B (2X): Dissolve 12.114 g Tris-HCl, 17.53 g NaCl in 700 mL Milli-Q water. Add 200 mL glycerol. Adjust to pH 7.4 by adding HCl. Adjust volume to 1 L with Milli-Q water. 3. PMSF (200X): Inhibits serine proteases. Dissolve 0.35 g in 10 mL pure ethanol. The 200 mM stock solutions are stable for months at +4°C. Do not try to dissolve in water. The halflife of PMSF in water is 1 h. PMSF is poisonous and should be handled carefully. 4. Benzamidine hydrochloride (200X): Also inhibits serine proteases. Dissolve 1.25 g in 10 mL Milli-Q water. Store at +4°C. 5. EDTA (200X): Used to inactivate metalloproteases by depleting ions. Dissolve 8.18 g in 50 mL of water. EDTA solution will not go into solution until the solution is adjusted to approximately pH 8.0 by adding NaOH pellets. Store at +4°C.

2.6. Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis

1. Separating buffer: 375 mM Ultra-Pure Tris-HCl (pH 8.7). 2. 30% acrylamide/bis-acrylamide solution (37.5:1). 3. Stacking buffer: 125 mM Tris-HCl (pH 6.8). 4. 0.1% (w/v) Ultra-Pure sodium dodecyl sulfate (SDS). 5. 0.1% (w/v) ammonium persulfate (APS). 6. N,N,N,N¢-Tetramethyl-ethylenediamine (TEMED). 7. Running buffer (5X): 125 mM Ultra-Pure Tris-HCl, 960 mM glycine, 0.5% (w/v) SDS. Store at room temperature. 8. Loading buffer (4X): 250mM Ultra-Pure Tris-HCl, pH 6.8, 8% (w/v) SDS, 0.04% (w/v) bromophenol blue, 40% (w/v) glycerol, 20% (v/v) 2-b-mercaptoethanol. 9. PageRuler™ prestained protein ladder (Fermentas). 10. Bio-Rad Mini-Protean® 3 cell.

2.7. Western Blot for Hemagglutinin-Tagged Proteins

1. 25 mM Ultra-Pure Tris-HCl. 2. 192 mM glycine, molecular biology grade. 3. 20% ethanol. 4. 0.05% (w/v) Ultra-Pure SDS. 5. Hybond™-C Extra, nitrocellulose, supported 45 mm, RPN303E, Amersham Biosciences.

Heterologous Expression of Human Membrane Receptors in the Yeast

91

6. Tween®-20, polyoxyethylene sorbitan monolaurate polysorbate 20, cell culture and bacteriology grade. 7. Tris-buffered saline with Tween-20 (TBS-T): 20 mM UltraPure Tris-HCl (pH 7.4), 450 mM NaCl, and 0.1% (v/v) Tween-20. 8. 4% or 0.2% nonfat dry milk. 9. Primary antibody: Monoclonal mouse anti-HA (hemagglutinin; supernatant of hybridoma culture) (see Note 1). 10. Secondary polyclonal goat antimouse immunoglobulins conjugated to horseradish peroxidase (HRP; Dako). 11. Immobilon™ Western, chemiluminescent HRP substrate (HRP substrate peroxide solution and HRP substrate Luminol reagent) (Millipore). 12. 0.1% (w/v) Amidoblack. 13. 10% (v/v) isopropanol. 14. 25% (v/v) acetic acid. 15. Las 3000. 16. Plastic bags and plastic films. 17. Rotating wheel. 2.7.1. For the solution preparation:

1. Transfer buffer (10X): Dissolve 58 g Ultra-Pure Tris-HCl (do not adjust pH), 29 g glycine, and 3.7 g SDS in 1 L of water. Store at room temperature (with cooling during use). 2. TBS-T: Dissolve 2.42 g Ultra-Pure Tris-HCl and 26.3 g NaCl in 900 mL of Milli-Q water. Adjust to pH 7.4. Add 1 mL Tween-20 and complete to 1 L with Milli-Q water. Store at room temperature. 3. Blocking buffer: 4% (w/v) nonfat dry milk in TBS-T. Dissolve 40 g powder milk in 1 L TBS-T. 4. Amidoblack 0.1% solution: Dissolve 0.1 g of amidoblack in 25 mL acetic acid, 10 mL isopropanol, and 65 mL of Milli-Q water.

3. Methods 3.1. Construction of Expression Vectors

1. The multitag affinity purification (MAP) sequence encoding (a) a factor Xa, a TEV (Tobbaco Etch Virus), and a thrombin cleavage sites to eliminate the MAP sequence; (b) a calmodulin-binding domain (CBD), a streptavidin tag, and an hexahistidine tag for affinity chromatography; and (c) a HA peptide for anti-HA Western blot analysis (6) was inserted into the BamHI and XhoI restriction sites of the YEpPMA polylinker.

92

Joubert et al.

2. PCR with the Proofstart polymerase was carried out on the target cDNA to introduce at the 5¢ end two restriction sites (XbaI, SpeI) and a sequence of six adenosines and at the 3¢ end an NheI restriction site using the adequate primers. 3. The PCR product was subcloned in the pCR2.1 plasmid and sequenced. 4. Target cDNA was then digested by SpeI and NheI and subcloned in MAP XbaI/NheI sites, giving YEpPMA–target cDNA–MAP expression vector. 3.2. Yeast Transformation

Day 0 1. Start an overnight culture of S. cerevisiae K699 cells to be transformed. In a 50-mL sterile flask, inoculate 5 mL of rich medium (see Section 6.3.3) with yeast cells. Incubate overnight at 30°C in an incubator shaker at 200 rpm. Day 1 1. Four hours before transformation, dilute the saturated overnight culture to a third into the same medium. Incubate for 2–4 h at 30°C at 200 rpm in the incubator shaker. 2. In a 1.5-mL Eppendorf tube, spin down 0.5 mL of the culture (see Note 2). 3. Discard supernatant. This leaves 50–100 mL of medium. 4. Add 10 mL of 10 mg/mL carrier salmon sperm DNA, resuspending the cells with the pipet. 5. Add 1 mg of your plasmid (see Note 3). Vortex. 6. Add 0.5 mL of filter-sterilized plate mixture. Vortex 7. Add 20 mL of 1M DTT (see Note 4). Vortex. 8. Incubate overnight at room temperature on your benchtop. Day 2 1. Cells will have settled to the bottom of the tube. 2. Heat shock the cells by incubating for 10 min at 42°C. 3. Use a pipet to suck up the bottom 50 mL of cells and plate mixture (see Note 5). 4. Plate cells on selective plate. In our case, it is the YNB + 10X D + 100X AA + agar (see Section 6.3.3). 5. Incubate 2 days at 30°C.

3.3. Yeast Culture

Day 0: Material preparation 1. Autoclave thirteen 2-L Erlenmeyer flasks and one 500-mL measuring cylinder. 2. Prepare all solutions and media as indicated.

Heterologous Expression of Human Membrane Receptors in the Yeast

93

Day 1: Preculture: 1. Inoculate 10 mL of minimal medium (9 mL YNB + 1 mL 10X D + 0.1 mL 100X AA) in a 50-mL shake flask with an isolated colony of transformed S. cerevisiae from the YNB plate. 2. Grow overnight at 30°C in the incubator under vigorous shaking at 200 rpm. 3. Make three precultures in three different flasks. Day 2: Culture Morning 1. Check optical density (OD) at 600 nm (OD600) making a 1:10 dilution in the spectrophotometer cuvette. Yeast should have reached the stationary phase (OD600 > 2). 2. Dilute the 30-mL preculture in 300 mL minimal medium (270 mL YNB + 30 mL 10X D + 3 mL 100X AA) in a 2-L autoclaved Erlenmeyer flask. 3. Keep at 30°C under shaking at 200 rpm. Let yeast grow until 2 OD. Afternoon 1. Put 500 mL of rich medium (450 mL YEP + 50 mL of 10X D solution) in each of the twelve 2-L Erlenmeyer flasks prepared for culture. 2. Inoculate each Erlenmeyer flask with 25 mL of your preculture (check your culture for any bacterial contamination by microscopic observation). Thus, your culture starts at 0.1 OD. 3. Change your incubator’s temperature to 18°C. Day 3: 1. Check OD600 in the morning. Yeast should have reached an OD600 around 2. 2. Check microscopically to determine if your culture has been contaminated. 3. At 24 h postincubation, yeasts should have reached an OD600 about 6–7. 4. Harvest the cells by centrifugation for 10 min at 3000 g at +4°C. 5. The wet pellet in these conditions should weigh around 60 g (which indicates 10 g/L culture). 6. Resuspend the pellet first with cold water. 7. Centrifuge 10 min at 3000 g at 4°C. Discard the supernatant. 8. Wash with ice-cold buffer A (1X) containing 50 mM Tris-HCl, 500 mM NaCl, pH 7.4, supplemented with protease inhibitors: 1 mM PMSF, 4 mM benzamidine hydrochloride, 2.5 mM EDTA, pH 8.

94

Joubert et al.

9. Centrifuge again 10 min at 3000 g at 4°C. Discard the supernatant. 10. Freeze the pellet at −80°C until next use. 3.4. Yeast Membrane Fraction Preparation

1. Make 200 mL of buffer A (1X) by mixing 100 mL of ice-cold buffer A (2X) with 1 mL PMSF (200X), 1 mL EDTA (200X), 1 mL benzamidine hydrochloride (200X) and complete to 200 mL with ice-cold Milli-Q water. 2. Thaw yeast pellet by adding 3 volumes of this freshly prepared ice-cold buffer A supplemented with protease inhibitors. 3. In a 50-mL Falcon tube, put 17.5 mL of yeast solution and 2.5 mL of glass beads (see Note 6). 4. Grind yeast by vortexing at 2,000 rpm at 4°C for 15 min. This is the optimal time, as shown in Fig. 6.1 using a Heidolph multireax device. 5. With a pipet, separate the supernatant from the beads and centrifuge at 3,000 g for 10 min to eliminate unbroken cells, debris, and nuclei. Here you, can estimate the grinding’s efficiency by measuring the ratio between the starting and the remaining volume of pellet (see Note 7). 6. The supernatant is then centrifuged at 18,000g for 1 h at 4°C. 7. The supernatant is discarded.

Fig. 6.1. Saccharomyces cerevisiae grinding with glass beads. Each 50-mL Falcon tube corresponding to 1 L of culture was vortex shaken at maximum speed (2000 rpm) ­during the indicated time (1–60 min). Membrane proteins were prepared as indicated in the protocol, and the total amount of material was estimated by the Bradford method.

Heterologous Expression of Human Membrane Receptors in the Yeast

95

8. This pellet containing yeast plasma membrane is washed twice in the same buffer without EDTA. Again, centrifuge at 18,000g for 1 h at +4°C. 9. Make 200 mL of buffer B 1X by mixing 100 mL of ice-cold buffer B (2X) with 1 mL PMSF (200X) and 1 mL benzamidine hydrochloride (200X) and complete to 200 mL with ice-cold water. 10. Finally, the pellet is resuspended in buffer B plus protease inhibitors without EDTA. It is necessary to remove EDTA before solubilizing membrane fraction for purification on a Ni-NTA column. 11. Using DC Bio-Rad assay reagents and bovine serum albumin (BSA) at concentrations ranging from 2 to 20 mg/mL as a standard, quantify total MPs. 3.5. SDS Polyacrylamide Gel Electrophoresis

Tables to help you make separating and stacking gels with different acrylamide concentrations are easily available (see Tables 6.1 and 6.2). These instructions assume the use of the Bio-Rad MiniProtean gel system. 1. Glass plates to be used should be well cleaned and extensively rinsed with distilled water and ethanol. 2. Prepare a 0.75-mm thick, 8% separating gel. See Table 6.1 for gel composition. Wear gloves because acrylamide is neurotoxic and carcinogenic when unpolymerized. 3. Pour the gel, leaving space (1 cm below comb teeth) for a stacking gel, and overlay with isobutanol or isopropanol (30–50 mL). Allow the gel to polymerize for at least 10 min. Polymerization time depends on room temperature. 4. Pour off the isobutanol or isopropanol and rinse twice with water. 5. Prepare the stacking gel. See Table 6.2 for composition. Pour the gel and immediately insert the comb. Wait few minutes to polymerize before removing the comb. 6. Prepare the running buffer by diluting 200 mL of the 5X buffer with 800 mL of water. Mix before use. 7. Once the stacking gel has set, carefully remove the comb and wash wells with water using pipet (see Note 8). 8. Put the gel in the electrophoresis unit. 9. Mix 15 mL of your sample with 5 mL of the 4X sample buffer. 10. Screen heating time necessary to denature your protein without aggregating it. We tested 0, 2, 5, 10, and 15 min. This step depends on your protein (see Fig. 6.2).

96

Joubert et al.

Table 6.1 Solutions for preparing resolving gels for Tris-glycine SDS-PAGE (Modified from Harlow and Lane 1988) Component volumes (mL) per gel mold volume (mL) of Solution components

5

10

15

20

25

30

40

50

H2O

2.6

5.3

7.9

10.6

13.2

15.9

21.2

26.5

30% acrylamide mix

1.0

2.0

3.0

4.0

5.0

6.0

8.0

10.0

1.5M Tris-HCl (pH 8.8)

1.3

2.5

3.8

5.0

6.3

7.5

10.0

12.5

10% SDS

0.05

0.1

0.15

0.2

0.25

0.3

0.4

0.5

10% ammonium persulfate

0.05

0.1

0.15

0.2

0.25

0.3

0.4

0.5

TEMED

0.004

0.008

0.012

0.016

0.02

0.024

0.032

0.04

H2O

2.3

4.6

6.9

9.3

11.5

13.9

18.5

23.2

30% Acrylamide mix

1.3

2.7

4.0

5.3

6.7

8.0

10.7

13.3

1.5M Tris-HCl (pH 8.8)

1.3

2.5

3.8

5.0

6.3

7.5

10.0

12.5

10% SDS

0.05

0.1

0.15

0.2

0.25

0.3

0.4

0.5

10% ammonium persulfate

0.05

0.1

0.15

0.2

0.25

0.3

0.4

0.5

TEMED

0.003

0.006

0.009

0.012

0.015

0.018

0.024

0.03

H2O

1.9

4.0

5.9

7.9

9.9

11.9

15.9

19.8

30% acrylamide mix

1.7

3.3

5.0

6.7

8.3

10.0

13.3

16.7

1.5M Tris-HCl (pH 8.8)

1.3

2.5

3.8

5.0

6.3

7.5

10.0

12.5

10% SDS

0.05

0.1

0.15

0.2

0.25

0.3

0.4

0.5

10% ammonium persulfate

0.05

0.1

0.15

0.2

0.25

0.3

0.4

0.5

TEMED

0.002

0.004

0.006

0.008

0.01

0.012

0.016

0.02

H2O

1.6

3.3

4.9

6.6

8.2

9.9

13.2

16.5

30% acrylamide mix

2.0

4.0

6.0

8.0

10.0

12.0

16.0

20.0

1.5M Tris-HCl (pH 8.8)

1.3

2.5

3.8

5.0

6.3

7.5

10.0

12.5

10% SDS

0.05

0.1

0.15

0.2

0.25

0.3

0.4

0.5

10% ammonium persulfate

005

0.1

0.15

0.2

0.25

0.3

0.4

0.5

TEMED

0.002

0.004

0.006

0.008

0.01

0.012

0.016

0.02

6%

8%

10%

12%

15% (continued)

97

Table 6.1 (continued) Component volumes (mL) per gel mold volume (mL) of Solution components

5

10

15

20

25

30

40

50

H2O

1.1

2.3

3.4

4.6

5.7

6.9

9.2

11.5

30% acrylamide mix

2.5

5.0

7.5

10.0

12.5

15.0

20.0

25.0

1.5M Tris-HCl (pH 8.8)

1.3

2.5

3.8

5.0

6.3

7.5

10.0

12.5

10% SDS

0.05

0.1

0.15

0.2

0.25

0.3

0.4

0.5

10% ammonium persulfate

0.05

0.1

0.15

0.2

0.25

0.3

0.4

0.5

TEMED

0.002

0.004

0.006

0.008

0.01

0.012

0.016

0.02

Table 6.2 Solutions for preparing 5% stacking gels for Tris-glycine SDS-PAGE (Modified from Harlow and Lane 1988) Component volumes (mL) per gel mold volume (mL) of Solution components

1

2

3

4

5

6

8

10

H2O

0.68

1.4

2.1

2.7

3.4

4.1

5.5

6.8

30% Acrylamide mix

0.17

0.33

0.5

0.67

0.83

1.0

1.3

1.7

1.0M Tris-HCl (pH 6.8)

0.13

0.25

0.38

0.5

0.63

0.75

1.0

1.25

10% SDS

0.01

0.02

0.03

0.04

0.05

0.06

0.08

0.1

TEMED

0.001

0.002

0.003

0.004

0.005

0.006

0.008

0.01

Fig.  6.2. Anti-HA Western blot. Each sample contains an equivalent amount of membrane proteins (expressing human receptor Smoothened, hSmo). Loading buffer (4X) was added, and then each sample was boiled for the indicated time (lanes 1–4; 0–15 min). The band corresponding to the hSmo disappears when the sample is boiled, probably due to aggregation.

98

Joubert et al.

11. Load the samples into the wells with a Hamilton syringe or a pipet using gel loading tips. Load also 3 mL of prestained molecular weight markers. 12. If using Bio-Rad’s patented sample loading guide, place it between the two gels in the electrode assembly. 13. Use the sample loading guide to locate the sample wells. Insert the Hamilton syringe or pipet tip into the slots of the guide and fill the corresponding wells (see Note 9). 14. Remove the sample loading guide. 15. Place the lid on the minitank. Make sure to align the color-coded banana plugs and jacks. The correct orientation is made by matching the jacks on the lid with the banana plugs on the electrode assembly. A stop on the lid prevents incorrect orientation. 16. Insert the electrical leads into a suitable power supply with the proper polarity. 17. Constant 100 V is recommended for SDS polyacrylamide gel electrophoresis (SDS-PAGE). Run time is around 1 h 30 min. Time depends on the molecular weight of your protein. 18. After electrophoresis is complete, turn off the power supply and disconnect the electrical leads. 19. Remove the tank lid and carefully lift out the inner chamber assembly. Pour off and discard the running buffer (see Note 10). 20. Open the cams of the clamping frame. Pull the electrode assembly out of the clamping frame and remove the gel cassette sandwiches. 21. Remove the gels from the gel cassette sandwich by gently separating the two plates of the gel cassette. The green, wedge-shaped, plastic gel releaser may be used to help pry the glass plates apart. 22. Remove the gel by floating it off the glass plate by inverting the gel and plate under fixative (for silver staining, for instance) or transfer solution (proceed to Section  6.3.6), agitating gently until the gel separates from the plate. 3.6. Western Blot

These directions for the Western blot assume the use of the BioRad Mini Trans-Blot electrophoretic transfer cell. 1. Prepare the 1X transfer buffer just before use by diluting 100 mL of the 10X transfer buffer in 700 mL Milli-Q water and add 200 mL ethanol (to have a final concentration of 20%). 2. Cut the membrane and the filter paper to the dimensions of the gel. Always wear gloves when handling membranes to prevent contamination. Soak the membrane, filter paper, and fiber pads in transfer buffer 15 min before blotting. 3. Prepare the gel sandwich.

Heterologous Expression of Human Membrane Receptors in the Yeast

99

4. Place the cassette, with the black side down, in a recipient containing 1X transfer buffer. 5. Place one prewetted fiber pad on the gray side of the cassette. 6. Place a sheet of filter paper on the fiber pad. 7. Place the gel on the filter paper. 8. Place the prewetted membrane on the gel. Remove any air bubbles between gel and membrane by rolling a 5-mL tube on the membrane. 9. Complete the sandwich by placing a piece of filter paper on the membrane. 10. Add the last fiber pad. 11. Close the cassette firmly, being careful not to move the gel and filter paper sandwich. 12. Lock the cassette closed with the white latch. 13. Place the cassette into the transfer tank such that the nitrocellulose membrane is between the gel and the anode. It is vitally important to ensure this orientation or the proteins will be lost from the gel into the buffer rather than transferred to the nitrocellulose. 14. Add the frozen Bio-Ice cooling unit. Place in tank and completely fill the tank with buffer. 15. Put on the lid, plug the cables into the power supply, and run the blot under constant voltage of 100 V for 1 h (high-intensity field ~ 350 mA). 16. On completion of the run, disassemble the blotting sandwich and remove the membrane for development. Clean the cell, fiber pads, and cassettes with laboratory detergent and rinse well with deionized water. 17. The colored molecular weight markers should be clearly visible on the membrane. 18. To check whether your transfer has been correctly done, rinse the blot for 2 min with 0.1% of amidoblack solution to stain the transferred bands. 19. Wash twice with TBS-T buffer to remove amidoblack. 3.6.1. Classical Antibody Incubation Method (see Fig. 6.3, Lane 2)

1. The nitrocellulose is then incubated in 50 mL blocking buffer for 1 h at room temperature on a rocking platform. 2. The blocking buffer is discarded, and the membrane is placed in a plastic bag containing 5 mL of blocking buffer supplemented with the primary antibody (here 1:20 dilution of anti-HA antibodies) overnight at 4°C on a rotating wheel. 3. The primary antibody is then removed, and the membrane is washed three times for 10 min each with 50 mL blocking buffer.

100

Joubert et al.

Fig. 6.3. Antibody incubation methods. After transfer, the nitrocellulose membrane was cut; the two methods were tested: With the classical method, exposing the blot for 20 s was sufficient to have the signal corresponding to hSmo, whereas with SNAP method, time exposure was longer (2 min) to obtain the same signal. Lane 1: membrane fraction of untransformed yeast. Lane 2: membrane fraction expressing hSmo detected using classical method. Lane 3: membrane fraction expressing hSmo detected with SNAP i.d method.

4. The membrane is placed in a plastic bag containing 5 mL of blocking buffer supplemented with the convenient secondary antibody at the dilution given by the manufacturer (here, it is the secondary polyclonal antimouse immunoglobulin antibody used as a 1:5,000-fold dilution). Place the bag on a rotating wheel for 2 h at 4°C (see Note 11). 5. The secondary antibody is discarded, and the membrane is washed twice for 10 min each with blocking buffer and a third time with TBS-T buffer (without milk). 3.6.2. Rapid Antibody Incubation Method (see Fig. 6.3, Lane 3)

3.7. Expression Level as a Function of OD600

Millipore has introduced a new method to optimize blotting conditions in record time for maximum results. Unlike conventional Western blotting, for which diffusion is the primary means of reagent transport, the SNAP i.d Protein Detection System applies a vacuum to actively drive reagents through the membrane. This novel method, compatible with all standard membranes, reduces incubation time and optimizes your blot in 30 min from blocking to revelation steps. This method requires a small amount of nonfat dry milk (as low as 0.2% w/v in our case for blocking, antibody dilution, or washing steps). In this method, time exposure is longer than in the classical method. On the following, we analyzed the expression level of hSmo as a function of yeast cell density. 1. Measure yeast OD by diluting culture to 1:10 in a 2-mL spectrophotometer cuvette and reading the absorbance at 600 nm. 2. Stop culture at OD600 of 3, 5, 10, 15, and 20. 3. Prepare membrane fraction of each culture condition.

Heterologous Expression of Human Membrane Receptors in the Yeast

101

4. Quantify protein using the Bradford method using a BioRad kit. 5. Protein expression in the membrane fraction increases and reaches a maximum around OD 7; therefore, we choose to cultivate to 7 OD (see Fig. 6.4). 3.8. Expression Level as a Function of Air-to-Medium Ratio

We studied the effect of culture Erlenmeyer air-to-medium ratio on the protein expression level. 1. In a 2 L-Erlenmeyer flask, pour 200, 300, 400, or 500 mL YEP medium. The air-to-medium ratio is therefore 10:1, 7:1, 5:1, and 4:1, respectively. 2. Culture is performed at 18°C under shaking at 200 rpm. 3. Let yeast grow until about 7 OD600. 4. Prepare membrane fraction for each culture condition. 5. Quantify protein using the Bradford method using a Bio-Rad kit. 6. As shown on the Western blot in Fig. 6.5, the expression level was higher when yeasts were grown with the highest oxygenation condition.

Fig. 6.4. Constitutive expression of hPtc in Saccharomyces cerevisiae plasma membrane fraction as a function of the yeast optical density (OD). Western blot revealed with an ­anti-HA antibody; 40 mg of protein were loaded in each lane.

Fig. 6.5. Constitutive expression of hSmo in Saccharomyces cerevisiae plasma membrane fraction as a function of the air-to-medium ratio in the culture flask. Western blot revealed with anti-HA antibodies. Expression yield was better when air volume in the culture flask was higher.

102

Joubert et al.

4. Notes 1. Anti-HA antibody is stored at +4°C by addition of 0.01% final concentration sodium azide, conveniently done by dilution from a 10% stock solution. Azide is highly toxic; therefore, it should be cautiously handled. 2. To pellet the cells, 5 s are sufficient. 3. Usually, 1-3 μL should be sufficient. 4. DTT is instable in solution. It should be prepared, immediately aliquoted, and stored at −20°C. 5. Try to get as many cells as possible but do not worry about keeping every last cell. You should easily get 80–90% of cells. 6. This repartition seems to be optimal. Exceeding these volumes may alter vortexing; therefore, grinding will not be efficient. 7. To estimate grinding’s efficiency, you can compare the volume of the yeast pellet before and after grinding if you are using the 50-mL Falcon tubes. If not, weigh the pellet before and after grinding. Generally, this ratio is around 50–60%. 8. Remove the comb slowly and vertically to avoid damaging the wells. 9. Load samples slowly to allow them to settle evenly on the bottom of the well. Be careful not to puncture the bottom of the well with the syringe needle or pipet tip. 10. Always pour off the buffer before opening the cams to avoid spilling the buffer. 11. You can try to reduce time by incubating the membrane with the secondary antibody for 1 h at room temperature.

Acknowledgment This work was supported by the European Community Specific Target Research Project grant FP6-2003-LifeSciHealth, “Innovative Tools for Membrane Structural Proteomics.” References 1. Orlean P (1997) Biogenesis of yeast wall and surface components, in The molecular and cellular biology of the yeast Saccharomyces cerevisiae: cell cycle and cell biology (Pringle, J.R., Broach, J.R., Jones, E.W. eds.), Cold Spring Harbor Laboratories, NY, pp 229–362

2. Lee AG (2004) How lipids affect the activities of integral membrane proteins. Biochim Biophys Acta 1666:62–87 3. Jidenko M, Nielsen RC, Sorensen TL, Moller JV, le Maire M, Nissen P, Jaxel C (2005) Crystallization of a mammalian membrane

Heterologous Expression of Human Membrane Receptors in the Yeast protein overexpressed in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 102: 11687–11691 4. Toyoshima C, Nakasako M, Nomura H, Ogawa H (2000) Crystal structure of the calcium pump of sarcoplasmic reticulum at 2.6 A resolution. Nature 405:647–655 5. Joubert O, De Rivoyre M, Mus-Veteau I, Nehmé R, Bidet M (2008) A strain of S. cerevisiae able to produce the human form of Patched, receptor of Sonic Hedgehog. (CNRS, ed) Patent number 08/06125, France 6. De Rivoyre M, Bonino F, Ruel L, Bidet M, Therond P, Mus-Veteau I (2005) Human

103

receptor Smoothened, a mediator of Hedgehog signalling, expressed in its native conformation in yeast. FEBS Lett 579: 1529–1533 7. Figler RA, Omote H, Nakamoto RK, Al-Shawi MK (2000) Use of chemical chaperones in the yeast Saccharomyces cerevisiae to enhance heterologous membrane protein expression: high-yield expression and purification of human P-glycoprotein. Arch Biochem Biophys 376: 34–46 8. Sambrook J, Fritsch EF, Maniatis T (ed.) (1989) Molecular Cloning. Cold Spring Harbor Laboratory Press, NY.

Chapter 7 Mammalian Membrane Protein Expression in Baculovirus-Infected Insect Cells Céline Trometer and Pierre Falson Abstract A system of expression of mammalian membrane proteins in baculovirus-infected insect cells is described, allowing analytical or preparative production in the milligram range of such type of proteins. This is illustrated by the setup of the expression system of the human breast cancer resistance protein (BCRP), which is a homodimeric multidrug ABC (adenosine triphosphate-binding cassette) transporter. The system used is Bac to Bac™, which allows generation of a viral genome that includes the protein of interest by using a shuttle plasmid and a bacterial host carrying the bacmid. In the present case, the use of a molecular chaperon, ninaA, is illustrated, co-expressed with BCRP by using a plasmid allowing the expression of two proteins, pFastBac Dual. The method is detailed to allow the expression of such proteins and membrane protein in general. Key words: ABC transporters, baculovirus, heterologous expression, insect cells, mammalian membrane proteins

1. Introduction Producing membrane proteins from higher eukaryotes (e.g., mammalian) in the milligram range and at the same time ensuring correct folding and full activity of the expressed protein remains the main bottleneck in obtaining structural information at high resolution of such membrane proteins. One main reason for this is that most of these membrane proteins require a specific maturation process to be fully active: They are produced in the endoplasmic reticulum and are then targeted to a given membrane through a process involving glycosylation (1) and association with given lipids (2), both steps achieved in the endoplasmic reticulum and Golgi apparatus. In that quest, to date the baculovirus-infected cells has been one of the most versatile expression systems, ensuring quantity and quality of membrane protein production while I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_7, © Humana Press, a part of Springer Science + Business Media, LLC 2010

105

106

Trometer and Falson

remaining easy to handle and to scale up. Indeed, a large number of soluble and several membrane-bound proteins have been successfully produced using this system (3). Baculoviruses have been known for a long time, with early reports describing the infection of Chinese silkworms; these viruses infect only larval lepidopterans. Baculoviruses are enveloped, double-stranded DNA viruses of 80–180 kbp with rodshaped nucleocapsids. They exist as budded and occluded forms, the latter consisting of multiple nucleocapsids embedded in a polyhedrin (PH) matrix. The most extensively studied baculovirus strain is Autographa californica multiple nuclear polyhedrosis virus (AcMNPV). Baculovirus gene expression is divided into two main phases, which precede (early phase) or follow (late phase) viral DNA replication. During the late phase, some hyperexpressed genes, such as those coding for p10 and polyhedron, remain transcribed at high levels. Such a type of promoters is used for protein expression, although in some cases those of early genes are used (4). The baculovirus expression vector system (BEVS) was initiated by Smith and Summers in 1982 (5). It is based on replacement of a late, non-essential, viral gene coding for the PH with a gene of interest. Most of the transfer vectors use either early (Ie1) or late (p10, pPH) promoters. Based on linearized vectors, the BEVS allows rapid cloning (1 week) and expression (2 weeks) of recombinant proteins in insect cells of lepidopterans such as the fall armyworm (Spodoptera frugiperda: Sf9, Sf21) or the cabbage looper (Trichoplusia ni, Hi5). The most popular BEVS are Bac to Bac™ and BaculoDirect™ (Invitrogen); flashBAC™/BacMagic (EMD/ OET/Nextgen); and BacPAK6/BaculoGold (BD Biosciences/ Clonetech). In each case, the system involves a bacterial-type plasmid including sequences of recombination; between them, the gene of interest is inserted and placed under the control of an early or late promoter. Insertion of the gene into the viral genome is carried out by recombination using bacteria containing the viral genome. After insertion, the viral genome is prepared in large extent and used to transfect the insect cells. Optimization of expression involves the time of expression (24–96 h postinfection, hpi), temperature (20–28°C) and type of culture medium. The overall process of expression is rather simple to set up and to scale up from plates to large volumes of liquid culture. One limitation concerns the glycosylation of proteins, which is distinct from that of mammalian cells; while the latter tends to glycosylate to a large extent, with complex N-glycans including mannose, N-acetyl glucosamine, galactose and at the end, sialic acid, insect cells glycosylate poorly and essentially with mannose (3). This problem can be partially solved by using genetically adapted cells, such as Mimic cells (Invitrogen), in which human glycosyltransferases have been introduced (6). A second limitation

Mammalian Membrane Protein Expression in Baculovirus-Infected Insect Cells

107

in the use of insect cells concerns their lipid composition, which differs significantly from that of mammalian cells; notably, the level of cholesterol is lower (7). Again, a solution can be found by adding cholesterol to the culture media (8). The present chapter details the experimental procedure to produce a membrane protein, here the human breast cancer resistance protein (BCRP), by using the Bac to Bac baculovirus-based expression system. BCRP belongs to the adenosine triphosphate (ATP)-binding cassette transporter family (9). It was initially identified for the resistance it confers to anticancer drugs (10, 11). The protein is addressed to the plasma membrane and organized as a homodimer. Each monomer is made of 655 residues that form an intracellular nucleotide-binding domain, a six-span membrane domain and a short extracellular domain. Expression of BCRP has been achieved in baculovirus-infected cells using Sf9 (12) or Hi5 (13, 14) cell lines, the latter producing a higher level of protein, but heterogeneously. We show that the use of a biological chaperon, NinaA (15, 16), can prevent this problem and can help produce more membrane protein.

2. Materials The material used for human BCRP expression via baculovirusinfected cells described here is provided from Invitrogen as included in the Bac to Bac expression kit. 2.1. Proteins

1. The gene coding for NinaA (membrane protein chaperon, UniProtKB/Swiss-Prot:P15425) was provided by Dr. Bertrand Mollereau, ENS Lyon. 2. The gene coding for human ABCG2 (UniProtKB/SwissProt:Q9UNQ0) was from a pTriEx construct described in (14).

2.2. Lab Equipment

1. All the material needed for molecular biology experiments. 2. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and Western blots are carried out using the MiniProtean 3 apparatus and related devices from Bio-Rad. 3. Transfection and (solid and liquid) culture of insect cells are achieved in a class I type room (see Note 1) equipped with a Steril Bio Ban 48, an incubator Heraeus BK6160 with a H + P Biomag Biomodule 40B, a microscope Olympus CKX31 and a low-speed centrifuge handling 15-/30-mL Falcon-type tubes. 4. Six-well, T-25, T-75 and T-225 plates (Falcon) are used for adherent cell cultures. Liquid cultures of 250, 500, and 1,000 mL are carried out in spinners (Bellco).

108

Trometer and Falson

2.3. Cells

1. Insect cell line Sf9, High Five™, or Mimic™ Sf9 (not used here but useful to achieve full glycosylation if necessary). 2. MAX Efficiency® DH10Bac™ chemically competent Escherichia coli (Invitrogen) that contains a baculovirus shuttle vector (bacmid, bMON14272, 136 kb) and a helper plasmid, coding for a transposase, allowing recombination between the bacmid and the pFastBac™ Dual construct. 3. One Shot® TOP10 chemically competent E. coli or equivalent to generate the recombinant plasmid of interest.

2.4. Media

1. Insect cells are grown in Sf-900 II SFM or Express Five® SFM (both from Invitrogen) as appropriate. Both media contain 0.1% of the surfactant F68 pluronic acid and are supplemented with 10% fetal bovine serum (FBS), 1% penicillin and streptomycin. Express Five is supplemented with 2 mM glutamic acid. Media are stored for long term at 4°C. 2. Bacteria are grown in Luria-Bertani medium: 10 g/L Bactotryptone, 5 g/L Bacto-yeast extract, 10 g/L NaCl, supplemented with 15 g/L agar for plates. Autoclave 20 min at 120°C. Before use add antibiotics depending on the strain used, either 100 mg/L ampicillin (see Note 2) for TOP 10 or 50 mg/L kanamycin, 7 mg/L gentamicin, 10 mg/L tetracycline, 40 mg/mL IPTG (isopropyl-b-d-galactopyranoside) and 100 mg/ mL X-gal (5-bromo-4-chloro-3-indolyl-b-d-galactopyranoside) for DH10Bac transformed with pFastBac Dual plasmid. Store without antibiotics at room temperature.

2.5. Transfection

1. Reagent: Cellfectin® reagent (Invitrogen, included in the Bac to Bac kit) is a 1:1.5 (M/M) liposome formulation of the cationic lipid N,NI,NII,NIII-tetramethyl-N,NI,NII,NIIItetrapalmitylspermine (TM-TPS) and dioleoyl phosphatidylethanolamine (DOPE) in membrane-filtered water. Store at 4°C. 2. Medium: Unsupplemented Grace’s insect cell culture medium (Invitrogen). Store for long term at −20°C and short term (1 week) at 4°C. Note that Grace’s insect cell culture medium should not contain supplements or FBS as their protein content will interfere with the Cellfectin Reagent, inhibiting the transfection.

2.6. Molecular Biology

1. The plasmid used here is the pFastBac Dual vector (Invitrogen, included in the Bac to Bac kit), which contains two multiple cloning sites, allowing expression of two genes, controlled by (late) promoters of PH and p10.

Mammalian Membrane Protein Expression in Baculovirus-Infected Insect Cells

109

2. The kits for small- (3–10 mg) and medium- (50–100 mg) scale plasmid (5–10 kbp) DNA preparations are NucleoSpin™ Plasmid and Nucleobond® PC100 from Macherey-Nagel. 3. The kit for large DNA preparation (Bacmid, 150 kbp) is PureLink™ HiPure Plasmid Midiprep from Invitrogen. 4. Accuprime Taq DNA polymerase used for polymerase chain reaction (PCR) was from Invitrogen. Other restriction and modifying enzymes were from New England Biolabs. 2.7. SDS-PAGE

1. Prepare stock solutions (room temperature) of separating buffer (1.5M Tris-HCl, pH 8.8), stacking buffer (1M Tris-HCl, pH 6.8), SDS 10%, acrylamide/bis (see Note 3) solution (40%, 37.5:1 with 2.6% C) and ammonium persulphate 10% (stored at 4°C for 15 days max). N,N,N,N¢-Tetramethylethylenediamine (TEMED) is used pure. All products come from Bio-Rad. 2. For each electrophoresis, prepare 0.1 L of 10X running buffer (125 mM Tris buffered with 960 mM glycine, 0.5% w/v SDS), which can be stored at room temperature. 3. Laemmli-type loading buffer “2XU”: (2x) 100 mM Tris-HCl, pH 8.0, 8M urea, 4% SDS, 1.4M b-mercaptoethanol, 0.0025% bromophenol blue. The solution is stored at −20°C and aliquoted to freeze/thaw ten times maximum. 4. Prestained molecular weight markers: Kaleidoscope markers (Bio-Rad).

2.8. Western Blotting

1. Transfer buffer: 10 mM CAPS (N-cyclohexyl-3-aminopro­ panesulfonic acid), pH 11.1, 10% methanol (see Note 4). Prepare fresh and use cold, with a cooling ice bag during transfer. 2. Polyvinyl fluoride membrane (PVDF (Polyvinylidene fluoride), Millipore) and 3 MM chromatography paper (Whatman) or nitrocellulose membrane. 3. Tris-buffered saline with Tween (TBS-T): Prepare a 10X stock with 1.37M NaCl, 27 mM KCl, 250 mM Tris-HCl, pH 7.4, 1% Tween-20. Dilute with water for use. 4. Primary antibody dilution buffer: TBS-T supplemented with 2% (w/v) fraction V bovine serum albumin (BSA). Primary antibody BXP21 (Millipore/Chemicon) is used 1/250 diluted. 5. Secondary antibody antimouse immunoglobulin G (IgG) rabbit conjugated to horseradish peroxidase (DakoCytomation) is used 1/3,000 diluted. 6. Enhanced chemiluminescent (ECL) reagents (GE healthcare) and Bio-Max ML film (Kodak) are used for revelation.

110

Trometer and Falson

3. Methods 3.1. Molecular Biology

Cloning of Drosophila NinaA and human BCRP into the pFastBac Dual plasmid (see restriction map in Fig.  7.1, panel A) is achieved using classical methods of molecular biology described in the booklet included in the Bac to Bac kit and in (17). NinaA was cloned between the BamH I and EcoR I restriction sites inside the polylinker region controlled by the PPH (Polyhedrin promoter) promoter from a DNA fragment produced by PCR using the wild-type gene as template, amplified with the 5¢ and 3¢ primers

kbp 6 4 3 2.5 5

SV40 pA pFastBac Dual

Tn7L

Ampicillin/ Gentamicin resistance

Tretacycline resistance

M 1

– – – – –

2 3 4 5341 4627 3274

2560

Kanamycine resistance

PCR

helper

helper

lacZ

HSV tk

Cloning

recombinant

E.coli (Lac Z+)

Tn7R

B

A

bacmid

bacmid

pfastbacdual

E.coli (Lac Z-)

C

Bacmid extraction

Virus amplification, Titration, Protein production

F

E

D

Fig. 7.1. Scheme of protein expression based on the Bac to Bac™ system. (a) The Drosophila protein chaperone NinaA (714 bp) and human BCRP (2,067 bp, including additional codons at the 5¢ end to insert affinity and detection tags) — or the membrane protein of interest — are cloned into the pFastBac™ Dual shuttle plasmid, into each polylinker region under the control of the Pp10 and PPH promoters, respectively. Dotted region corresponds to the fragment DNA inserted into the bacmid by recombination, between Tn7R and Tn7L regions, indicated in the scheme by a cross. (b) Competent DH10Bac™ Escherichia coli bacteria are then transformed with the plasmid, and positive transformants are selected by ampicillin/gentamicin resistance, in addition to tetracycline and kanamycin resistance, which allows maintaining the bacmid and helper plasmid, respectively, already present in this strain. (c) Transposition occurs by the transposase encoded in the helper plasmid. Bacmid having incorporated the foreign genes are screened by PCR using primers flanking the region of transposition: lane M molecular weight marker, lane 1 unmodified bacmid giving a PCR product of 2,560 bp, lane 2 bacmid including human BCRP (4,627 bp), lane 3 bacmid including NinaA (3,724 bp), lane 4 bacmid integrating both human BCRP and NinaA (5,341 bp). (d) Bacmid of each construct is prepared using a method dedicated to large DNAs (see Section 7.3.). (e, f  ) Sf 9 insect cells are then transfected and checked for expression and virus titration as described in Section 7.3. (Scheme adapted from the Bac to Bac™ booklet. With permission).

Mammalian Membrane Protein Expression in Baculovirus-Infected Insect Cells

111

5¢-ACCGGATCCCCGCAAAATCATGAAGTCATTGC­ TCAATCGGATAAT and 5¢-AATGGAATTCAGCAGTA­CATGT­ TGAGCT (coding sequence is underlined, stop is double underlined; bold characters indicate restriction sites). Human BCRP was subcloned from a pTriEx construct in which the gene is already under the control of the Pp10 promoter, using the Pac I (5¢ end, cutting inside the Pp10 region) and Xho I (3¢ end) restriction sites. The following plasmids including the different constructs were generated: (1) pFastBac Dual (original plasmid), (2) pFastBac Dual–BCRP, (3) pFastBac Dual–NinaA, (4) pFastBac Dual–NinaA + BCRP. Once constructed, each plasmid was checked by sequencing and used to transform competent DH10BAC E. coli bacteria (Fig. 7.1, panel B). Positive clones, in which transposition between pFastBac Dual and the Bacmid has occurred, are selected by the white/ blue screen, with transposition leading to an insertion inside the gene coding for the b-galactosidase. Then, they are screened by PCR as follows: 1. Pick clones resistant to antibiotics (ampicillin, kanamycin and tetracycline) on a new LB (plus antibiotics) plate with a toothpick wetted into 10 mL water into a PCR tube. 2. Boil the solution for 10 min and withdraw 1 mL used as template for the PCR. 3. Carry out the PCR with M13 forward (−40) and M13 reverse primers (included in the Bac to Bac kit), located 128 bp upstream and 145 bp downstream from the locus of insertion, respectively. 4. Each construct is characterized by a given size PCR product as shown in Fig.  7.1, panel C: (1) unmodified bacmid giving a 2,560-bp PCR product, (2) bacmid including human BCRP (4,627 bp), (3) bacmid including NinaA (3,724 bp), and (4) bacmid integrating both human BCRP and NinaA (5,341 bp). 5. Extract each bacmid from positive clones using the kit for large DNA preparation PureLink HiPure Plasmid Midiprep (Fig. 7.1, panel D) and used to transfect Sf9 insect cells. 3.2. Sf9 Cell Transfection with Bacmids

1. In a T-25 flask (25 cm2, about 7.106 cells at confluence), dilute 500 mL of (liquid N2 frozen) stock cell suspension into 5 mL of Sf-900 II SFM medium supplemented with 10% FBS and 1% penicillin and streptomycin (both optional but recommended) and incubate for 2 days at 27°C. 2. Cells are then suspended in the liquid medium by tapping the flask, and 2.5 mL are withdrawn and mixed with 7.5 mL of fresh medium and incubated for 2 additional days in the same conditions.

112

Trometer and Falson

3. This initial culture is amplified in T-75 (75 cm2, about 15 × 106 cells at confluence) or a T-225 (225 cm2, about 50 × 106 cells at confluence) flask, seeded at 10 and 30 × 106 cells, to obtain enough cells for achieving transfection. Cells are counted with a microscope. 4. Spread 106 fresh cells in each well of a six-well culture plate (9 × 6 cm2/well, about 2.5 × 106 cells at confluence), one well for each assay of transfection plus a negative control (C−) in which no bacmid will be added. 5. Incubate the cells 1 h at 27°C to allow adhesion. 6. During this time, prepare for each trial of transfection the following mix: a. In a sterile Eppendorf tube A, add 1 mg recombinant bacmid/100 mL unsupplemented 1X Grace’s insect medium. b. In a sterile Eppendorf tube B, add 6 mL of cellfectin/100 mL unsupplemented 1X Grace’s insect medium. c. Mix each solution gently (three times by inversion) and pour the contents of tube B into tube A. Mix gently and incubate for 30 min at room temperature. 7. At the end of 1 h incubation of the cells, withdraw the medium from each of the wells and wash two times with 2 mL of unsupplemented 1X Grace’s insect medium. 8. Add 800 mL of unsupplemented 1X Grace’s insect medium to each DNA/lipofectin mix; the total volume will be 1 mL. Mix gently and add the mix gently to the corresponding well; the cells are then incubated at 27°C for 5 h, a time necessary to allow penetration of the bacmid. 9. The mix is then discarded and replaced by 2 mL of Sf-900 II SFM medium supplemented with 10% FBS and 1% penicillin and streptomycin, followed by a 72 h-incubation at 27°C. 10. The virus population generated from each assay is withdrawn as follows: In each well, cells are suspended by gentle pipetting and then pelleted by centrifugation at 500g for 10 min at room temperature. The 2 mL supernatant containing the baculovirus is transferred to a new sterile tube, constituting the generation called P1; it is stored at 4°C in the dark. 11. The cell pellet harvested at this stage from each 9.6-cm2 well of should give at 72 h postinfection about 2 × 5.106 cells at confluence. This is enough to check the expression of the protein of interest by SDS-PAGE and Western blot since about ten times less (2 × 105 cells, about 10 mg protein) are loaded onto the gels (see section 2.7 and 2.8). Protein expression is illustrated in Fig. 7.2a and detailed below.

Mammalian Membrane Protein Expression in Baculovirus-Infected Insect Cells

113

a

b

c

Fig. 7.2. Expression of BCRP in insect cells as a function of time, molecular chaperone and type of cell strain. Endpoint dilution. (a) Expression of BCRP in insect Sf 9 cells is carried out as described in Section 7.3.4. Cells were infected with a bacmid from the P3 population bearing pFastBacDual, alone (C−), with the gene coding for BCRP (BCRP) or with the genes coding for BCRP and NinaA (BCRP + NinaA) and grown for 24, 48 and 72 h postinfection in a six-well plate. BCRP expression was then analysed by SDS-PAGE and Western blot loading 10 mg of proteins onto a 10% SDS-PAGE (see Section 7.3.5.). BCRP is revealed with the antibody BXP21 diluted to 1/250, and the film was exposed for 5 min before revelation. The arrow indicates the position of migration of BCRP. (b) Compared expression of BCRP in High Five and Sf 9 insect cells. The expression is carried out as in A for 72 h and revealed as above. (c) Endpoint dilution. Dilution of the P3 stock suspension is carried out as indicated on the panel a and used to infect Sf 9 insect cells. Expression is checked after 72 h incubation as above.

3.3. Virus Amplification



The initial stock P1 is used to amplify the virus. Roughly, one can estimate that this solution corresponds to 5.106 plaque-forming units (PFU)/mL, with this value increasing tenfold at each generation. To generate the P2 stock solution (e.g., 30 mL using a T-225 flask seeded with 30 × 106 cells), the volume of P1 to add can be calculated using Equation 7.1: Volume of virus solution (mL) = (MOI × Number of cells)/ Solution virus titre, PFU/mL (Equation 7.1) where MOI is the multiplicity of infection estimated to 0.1.



⇔ 0.1 × 30.106 (cells)/5.106 (PFU/mL) = 0.6 mL P1 The culture is carried out as for P1 and harvested 72 h after infection. For large cultures, it is necessary to generate a P3 stock, infecting 5 T-225 flasks seeded at 30 × 106 cells with 60 mL of the P2 solution (thus ten times less than for generating P2 from P1), harvesting about 150 mL 72 h after infection. Each stock is stored at 4°C in the dark.

114

Trometer and Falson

3.4. Analysis of Protein Expression by SDS-PAGE and Western Blot

At each step, P1 to P3, the expression of the protein is checked by immunoblot after migration on SDS-PAGE and transfer on PVDF membrane. 1. Generate the separating gel (3.4 mL) of a 10% SDS-PAGE by mixing 1.92 mL of water, 1 mL of 40% acrylamide bisacrylamide solution, 1 mL of 1.5M Tris-HCl pH 8.8, 40 mL of 10% SDS, 40 mL 10% ammonium persulphate and 1.6 mL TEMED. Pour the Bio-Rad Mini-Protean 3 device to be 8 mm under the bottom of the wells. Add 200 mL of water at the surface of the gel for preventing the formation of waves. Do not use organic solvent for this step as membrane proteins have a tendency to interact with it. Polymerization occurs in 30 min at room temperature (22°C). 2. Generate the 5% stacking gel by mixing 1.46 mL of water, 0.25 mL of 40% acrylamide bisacrylamide solution, 0.25 mL of 1.5M Tris-HCl pH 6.8, 20 mL of 10% SDS, 20 mL 10% ammonium persulphate and 2 mL TEMED. 3. During polymerization of the stacking gel, prepare the protein samples as follows: a. Usually, 10 mg protein (estimated by the bicinchoninic acid method [18]) are loaded onto the gel, corresponding to a protein content of 2 × 105 cells. b. Pellet 4.105 cells and suspend them into 10 mL water and 10 mL 2XU Laemmli-type buffer. Allow the cells to be lysed and membrane proteins fully solubilised by 1-h incubation at room temperature. Do not heat the proteins, which tends to aggregate them, especially glycosylated ones. 4. Load 10 mg protein/10 mL onto the stacking gel and run for about 1.5 h at 25 mA and 120 V at room temperature. 5. During migration, a. Prepare the CAPS buffer for transfer and cool it at 4°C. b. Cut a 5 × 8 cm piece of PVDF membrane and six pieces of the same dimension of 3 MM chromatography paper. 6. After electrophoresis, transfer the protein from the SDS-PAGE to the PVDF membrane as follows: a. Incubate the gel in 5 mL of cold transfer buffer for 5 min. b. Wet the PVDF membrane with methanol briefly and then for 5 min in the cold transfer buffer. c. Prepare the transfer sandwich, built by superposing successively three paper sheets briefly wet in the transfer buffer, the acrylamide gel, the PVDF membrane (load the PVDF membrane onto the gel only one time since some proteins transfer quickly), and again three wet paper sheets.

Mammalian Membrane Protein Expression in Baculovirus-Infected Insect Cells

115

d. Add the ice cube to the Mini-Protean 3 transfer device and a magnetic fly and transfer for 1 h at 250 mA and 100 V under agitation to optimize cooling. e. After transfer, block the membrane for 1 h into 20 mL of TBS-T buffer containing 2% BSA. f. Add the primary antibody to the solution and incubate for an additional 1 h. g. Wash three times with 20 mL of TBS-T buffer. h. Incubate with 20 mL of TBS-T buffer containing 2% BSA and the second antibody for 1 h. i. Wash three times with 20 mL of TBS-T buffer. j. Withdraw the buffer and incubate with a 1:1 mix of 2 mL ECL solutions A and B for 5 min and expose onto a sensitive Kodak film for 1–10 min depending on the antibodies. A typical result is illustrated in Fig. 7.2. 3.5. Optimization of BCRP Expression in Insect Cells

Once the stock of P3 is obtained, it is useful to optimize the time of infection, check the effect of the molecular chaperon NinaA, check other insect cells and set up the dilution level of the virus stock for further expressions, here done by the method of endpoint dilution.

3.5.1. Optimization of the Time of Expression

1. In a six-well plate, seed in triplicate 2 mL of Sf-900 II SFM medium supplemented with 10% of FBS and 1% penicillin and streptomycin with 1.106 Sf9 cells and incubate for 2 h at 27°C before infection with 2 mL (1/100 dilution) with the P3 (pFastBac Dual alone or with BCRP or BCRP and NinaA), followed by a further incubation of 24, 48 and 72 h postinfection. 2. At the given time, cells are pelleted, solubilised and checked for BCRP expression by SDS-PAGE and Western Blot as described. The result is illustrated in Fig. 7.2a. It shows that the expression increases with time. Note that longer times of expression can be tested, such as 96 h.

3.5.2. Effect of Chaperon NinaA on BCRP Expression

As observed in Fig.  7.2a, NinaA plays the role of molecular chaperone for BCRP, influencing the level of expression and the homogeneity of BCRP. In absence of the chaperone, BCRP appears as two bands on the gel (lane 72 h), the upper being the most abundant, while when NinaA is co-expressed, BCRP appears as only one band corresponding to the upper one. It has been established that the lower band of BCRP produced in insect cells corresponds to an inactive form that comes from neither proteolysis nor differential glycosylation (19). The effect of NinaA on BCRP expression is thus positive since it tends to favour the folding

116

Trometer and Falson

of the protein. NinaA has a prolyl cis-trans isomerase activity, which could explain the observed effect, considering that BCRP has two proline residues located in the third span of the membrane domain for which a cis-trans isomerisation could have an impact on the folding of the protein. Note that in addition the co-expression of NinaA with BCRP leads to a two- to threefold increase of expression of BCRP. 3.5.3. Expression of BCRP in Sf 9 and Hi5 Insect Cells

One can use different strains for checking expression of the target protein. Initial transfection is carried out with Sf9 insect cells, but Hi5 cells are described to get a higher level of protein expression. Figure 7.2b illustrates the comparison of BCRP expression in the presence of NinaA in both types of cells. As expected and shown, Hi5 cells produce a higher amount of BCRP than Sf9 cells. However, most of the protein produced corresponds to the lower band species, although the co-expression of NinaA. The level of the upper band of BCRP is the same in both types of cells, indicating that Sf 9-type cells used with NinaA constitute the best choice for expressing BCRP.

3.5.4. Endpoint Dilution

Once the P3 solution is generated, check the dilution at which the stock of virus will be used. 1. Starting with P3, dilute the viral suspension (Bacmid pFastBac Dual - BCRP + NinaA) to 4, 10, 50, 102, 104, 106 and 108 times into 2 mL Sf-900 II SFM medium supplemented with 10% of FBS and 1% penicillin and streptomycin and pour a six-well plate loaded with 1.106 Sf 9 cells. Incubate for 72 h at 27°C. 2. At the end of the culture, withdraw 4.105 cells and check BCRP expression by SDS-PAGE and Western Blot as described. The result is illustrated in Fig. 7.2c, which shows that a 1/100 dilution is enough to maintain a high level of expression.

3.5.5. Scale-up in Spinners

Liquid cultures are carried out in 1-L spinners seeded with 5.109 Sf9 cells infected with 10 mL of P3 stock (10−2 dilution) and incubated for 72 h at 27°C and 60 rpm on a H + P Biomag Biomodule 40B. This leads routinely to 2 mg BCRP/L.

4. Notes 1. Baculovirus is not infectious for humans; its culture does not require class II-type equipment. However, depending on the gene expressed, it can be necessary to use this type of classified room. 2. Ampicilline should be prepared fresh to a maximal efficiency.

Mammalian Membrane Protein Expression in Baculovirus-Infected Insect Cells

117

3. Acrylamide is neurotoxic when unpolymerised, thus handle with gloves. 4. Methanol is neurotoxic. References 1. Jarvis DL, Finn EE (1996) Modifying the insect cell N-glycosylation pathway with immediate early baculovirus expression vectors. Nat Biotechnol 14:1288–92 2. van Meer G, Voelker DR, Feigenson GW (2008) Membrane lipids: where they are and how they behave. Nat Rev Mol Cell Biol 9:112–124 3. Kost TA, Condreay JP, Jarvis DL (2005) Baculovirus as versatile vectors for protein expression in insect and mammalian cells. Nat Biotech 23:567–575 4. Blissard GW, Rohrmann GF (1990) Baculovirus diversity and molecular biology. Annu Rev Entomol 35:127–155 5. Smith GE, Vlak JM, Summers MD (1982) In vitro translation of Autographa californica nuclear polyhedrosis virus early and late mRNAs. J Virol 44:199–208 6. Jarvis DL, Kawar ZS, Hollister JR (1998) Engineering N-glycosylation pathways in the baculovirus-insect cell system. Curr Opin Biotechnol 9:528–533 7. Marheineke K, Grunewald S, Christie W, Reilander H (1998) Lipid composition of Spodoptera frugiperda (Sf9) and Trichoplusia ni (Tn) insect cells used for baculovirus infection. FEBS Lett 441:49–52 8. Gilbert RS, Nagano Y, Yokota T, Hwan SF, Fletcher T, Lydersen K (1996) Effect of lipids on insect cell growth and expression of recombinant proteins in serum-free medium. Cytotechnology 22:211–216 9. Dean M, Hamon Y, Chimini G (2001) The human ATP-binding cassette (ABC) transporter superfamily. J Lipid Res 42:1007–1017 10. Ross DD, Yang W, Abruzzo LV, Dalton WS, Schneider E, Lage H, Dietel M, Greenberger L, Cole SP, Doyle LA (1999) Atypical multidrug resistance: breast cancer resistance protein messenger RNA expression in mitoxantrone-selected cell lines. J Natl Cancer Inst 91:429–433 11. Litman T, Brangi M, Hudson E, Fetsch P, Abati A, Ross DD, Miyake K, Resau JH, Bates

12.

13.

14.

15.

16.

17.

18.

19.

SE (2000) The multidrug-resistant phenotype associated with overexpression of the new ABC half-transporter, MXR (ABCG2). J Cell Sci 113(Pt 11):2011–2021 Ozvegy C, Varadi A, Sarkadi B (2002) Characterization of drug transport, ATP hydrolysis, and nucleotide trapping by the human ABCG2 multidrug transporter. Modulation of substrate specificity by a point mutation. J Biol Chem 277:47980–47990 McDevitt CA, Collins RF, Conway M, Modok S, Storm J, Kerr ID, Ford RC, Callaghan R (2006) Purification and 3D structural analysis of oligomeric human multidrug transporter ABCG2. Structure 14:1623–1632 Pozza A, Perez-Victoria JM, Sardo A, AhmedBelkacem A, Di Pietro A (2006) Direct interaction with purified breast cancer resistance protein ABCG2 indicates arginine-482 involvement in drug transport, not in binding. Cell Mol life Sci 63:1912–1922 Baker EK, Colley NJ, Zuker CS (1994) The cyclophilin homolog NinaA functions as a chaperone, forming a stable complex in  vivo with its protein target rhodopsin. EMBO J 13:4886–4895 Lenhard T, Reilander H (1997) Engineering the folding pathway of insect cells: generation of a stably transformed insect cell line showing improved folding of a recombinant membrane protein. Biochem Biophys Res Commun 238:823–830 Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning, a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, New York Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC (1985) Measurement of protein using bicinchoninic acid. Anal Biochem 150:76–85 Pozza A, Perez-Victoria JM, Di Pietro A (2009) Overexpession of homogeneous and active ABCG2 in insect cells. Protein Expr Purif 63(2):75–83

Chapter 8 Expression of Membrane Proteins in Drosophila Melanogaster S2 Cells: Production and Analysis of a EGFP-Fused G Protein-Coupled Receptor as a Model Karl Brillet, Carlos A. Pereira, and Renaud Wagner Abstract In the process of selecting an appropriate host for the heterologous expression of functional eukaryotic membrane proteins, Drosophila S2 cells, although not yet fully explored, appear as a valuable alternative to mammalian cell lines or other virus-infected insect cell systems. This nonlytic, plasmid-based system actually combines several major physiological and bioprocess advantages that make it a highly potential and scalable cellular tool for the production of membrane proteins in a variety of applications, including functional characterization, pharmacological profiling, molecular simulations, structural analyses, or generation of vaccines. We present here a series of protocols and hints that would serve the successful expression of membrane proteins in S2 cells, using an enhanced green fluorescent protein (EGFP)/G protein-coupled receptor (EGFP-GPCR) as a model. Key words: Drosophila S2 cells, heterologous expression, membrane proteins, recombinant GPCR

1. Introduction “A relatively recently developed cell technology for heterologous gene expression has been developed with Schneider 2 (S2) Drosophila cells. Although still lacking further detailed studies, this system offers, as a premise, some interesting advantageous possibilities such as, the high cell density easily attained (50–100 times higher when compared to mammalian cells), the growth of cells in suspension (circumventing the use of microcarriers), the use of inducible promoters (concentrating the protein production to a period of time and avoiding degradation) and the establishment of a continuous bioprocess (in contrast to bioprocess involving cell lysis induced by viral vectors).” These introductory statements were extracted from the foreword for a special issue of

I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_8, © Humana Press, a part of Springer Science + Business Media, LLC 2010

119

120

Brillet, Pereira, and Wagner

Cytotechnology on heterologous gene expression in Drosophila melanogaster cells (1) published in May 2008; it brings a representative illustration of how this system is increasingly used for the synthesis of a great variety of proteins. In particular, S2 cells are regularly proving to be an efficient and adequate system for the expression of demanding candidates. These cells represent a highly versatile expression system useful for both the functional analysis of genes that are naturally expressed at low levels and the production of significant amounts of recombinant proteins that are then amenable to biochemical or structural studies. This typically meets the requirements needed for the expression of functional membrane proteins, as exemplified in several studies focusing not only on monotopic membrane proteins such as lectins (2), enzymes (3), or viral membrane glycoproteins (4–6), but also on more complex candidates, including transporters (7, 8), ion channels (9–14), G protein-coupled receptors (GPCRs) (15–28), or even membrane protein complexes (13, 29). From all these studies conducted for functional characterization, pharmacological profiling, molecular simulations, structural analyses, or vaccinology purposes, the S2 cells constantly appeared as an efficient and valuable expression system due to a number of appealing properties. 1.1. The Drosophila S2 Cell Line

The S2 cells were isolated more than 30 years ago from primary cultures of Drosophila melanogaster embryos (30) as immortalized nontumorgenic cells, which may be derived from blood cells as recently suggested (31). S2 cells usually grow in a loose monolayer manner with a slight tendency to pile up, and they present epithelial-like characteristics. In contrast to mammalian cells, S2 cells are advantageously grown at room temperature using simple media and do not require any CO2 supplementation. One specific characteristic of S2 cells over other cell lines is their ability to grow to high cell densities, routinely reaching 3 × 107 cells/mL, which is about tenfold higher than what can be attained with other insect cells. Using slightly modified media, S2 cells can also be grown in suspension and can conveniently accommodate largescale culturing using bioreactors (16, 32). As a higher eukaryotic host, S2 cells possess all the mammalianlike cell machineries for gene expression, protein processing, and trafficking, including finicky posttranslational modifications that may be critical for proper maturation, localization, and function of demanding proteins, even if N-glycosylations are not strictly similar (2). In addition, S2 cells have been shown to lack endogenous surface proteins such as voltage-gated channels (33), which makes them particularly appropriate for the expression and functional studies of ectopic channels or neurotransmitter receptors.

Expression of Membrane Proteins in Drosophila Melanogaster S2 Cells

1.2. S2 Cells as an Expression System

121

Expression of recombinant proteins in S2 cells relies on a nonlytic plasmid-based system that thus provides additional benefits over virus-infected insect systems. S2 cells can actually be transiently or stably transfected using a series of plasmids bearing constitutive or inducible promoters, offering a significant panel of flexible expression conditions to be tested. The most widely used vectors contain either the constitutive promoter from the actine 5C gene (pAc5) (34) or the copper-inducible metallothionein promoter (pMT) (35), the last by far preferred for membrane protein expression. Stable cell lines are usually obtained after 3–4 weeks by cotransfecting the expression vector together with a selection vector bearing a drug-resistance marker, with the most common one the hygromycin B-phosphotransferase gene allowing for hygromycin B selection. A remarkable characteristic of S2 stable cell lines is the presence of multicopy insertions of the two vectors (several hundreds of copies) (36), a substantial gene amplification that results in an increased expression of the protein of interest. While this system is now well established and commercially available (Drosophila Expression System, DES®, from Invitrogen), several alternative expression strategies utilizing S2 cells have been explored, some of them successfully expressing proteins in nonlytic, single or multiple, baculovirus infection procedures (37, 38), some other achieving stable S2 cell lines with a single vector (39, 40). In this chapter, we propose to use a human mu opioid receptor (hMOR) N-terminally fused to the enhanced green fluorescent protein (EGFP) as a model membrane protein to illustrate a series of protocols that allow handling of a regular S2 cell expression system. In addition to the description of generic procedures for S2 cell culturing, this EGFP-fused expressed GPCR offers a good opportunity to exemplify some membrane protein-specific experiments that can be performed for assessing both the protein yield and functionality.

2. Materials 2.1. Cell Culture Procedures

1. The Drosophila melanogaster S2 cells and the pCoHYGRO and pMT/BiP vectors were obtained from Invitrogen.

2.1.1. S2 Cell Line and Vectors

2. The coding sequence of the mu opioid receptor was cloned into a pcDNA3.1 vector; the sequence of the EGFP was from a pEGFP-C3 vector (Clontech). 3. Use common materials and enzymes for standard cloning methods.

2.1.2. Culturing and Freezing S2 Cells

1. Laminar flow bench, incubator. 2. Hemacytometer and trypan blue.

122

Brillet, Pereira, and Wagner

3. InsectXpress complete medium: InsectXpress (Cambrex), 10% fetal bovine serum (FBS) heat inactivated 30 min at 65°C, 0.2% Pluronic F-68, 50 mg/mL gentamicin. 4. Dimethyl sulfoxide (DMSO). 5. Sterile 25-, 75-, and 175-cm2 flasks, six-well plates, and pipets. 6. Sterile 15-mL conical tubes and cryovials. 2.1.3. Suspension and Bioreactor Cell Culturing

1. Sterile T75-cm2 flasks and pipets. 2. InsectXpress complete medium. 3. Heater/shaker incubator. 4. Hemacytometer and trypan blue. 5. Sterile 500-mL shake flasks (Schott-type bottles fit well). 6. Sterile 1-L Schott-type bottle with double tubing through the septum of an open-top cap, one equipped with a 0.22-mm filter and one that can be connected directly to an inlet pipe of the reactor. 7. We use a 3-L reactor from Applikon® with a working volume of 2.7 L; other suitable reactors can be used. The reactor bears three marine impellers and is sparging aerated (1.8 L/min) through filters with a pore size of 0.22 mm. DO, pH, stirrer speed, and temperature parameters are controlled and recorded using dedicated software (Bioxpert, Applikon).

2.2. Procedures for Membrane Protein Expression and Analysis

1. Laminar flow bench, incubator.

2.2.1. Cell Transfection, Clone Selection, and Induction of Expression

5. Hygromycin.

2.2.2. Cell Sorting

2. Sterile six-well plate, combitips, and pipets. 3. Sterile 150 mM NaCl. 4. JetPEI™ DNA transfection reagent (Polyplus Transfection). 6. CuSO4. 1. Laminar flow bench, incubator. 2. Sterile and degassed PEB buffer: PBS supplemented with 5 mM ethylenediaminetetraacetic acid (EDTA) and 1% bovine serum albumin (BSA). 3. Anti-EGFP monoclonal antibody (see Note 1). 4. Goat antimouse immunoglobulin G (IgG) magnetic beads from Miltenyi Biotec (Bergisch Gladbach, Germany). 5. MACS® column from Miltenyi Biotec with the magnet support. 6. Sterile 12-well plates.

Expression of Membrane Proteins in Drosophila Melanogaster S2 Cells 2.2.3. Cell Lysis and Membrane Preparation

123

1. Lysis buffer: 50 mM Tris-HCl, pH 8, 1 mM EDTA. 2. Potter homogenizer. 3. Benchtop ultracentrifuge. 4. Membrane buffer: 50 mM Tris-HCl, pH 7.4, 320 mM sucrose. 5. 26-gauge needle. 6. BCA Protein Assay kit (Pierce).

2.2.4. Western Blot Analysis

1. Equipment for polyacrylamide gel electrophoresis (PAGE) and blotting device (from Bio-Rad or another supplier). 2. 10% sodium dodecyl sulfate (SDS) polyacrylamide gel. 3. Molecular weight marker. 4. MPBST buffer: 3% nonfat dried milk in PBS supplemented with 0.5% Tween-20. 5. 1/1,000 anti-EGFP monoclonal (Boehringer Ingelheim) in MPBST. 6. 1/2,000 antimouse IgG horseradish peroxidase (HRP)linked whole antibody from sheep (GE Healthcare) in MPBST. 7. Supersignal West Pico chemiluminescent substrate (Pierce).

2.2.5. Fluorescence Measurements

1. Insect Elliot buffer: 130 mM NaCl, 5.5 mM KCl, 1.2 mM MgCl2, 4.2 mM NaHCO3, 7.3 mM NaH2PO4, and 20 mM HEPES, pH 6.2. 2. Hemacytometer and trypan blue. 3. A 1-mL cuvette with magnetic stirring (Hellma). 4. Spectrofluorimeter (PTI).

2.2.6. Ligand-Binding Assays

1. PBS buffer supplemented with 320 mM sucrose. 2. 50 mM Tris-HCl, pH 7.4, 320 mM sucrose. 3. Polyethylenimine. 4. Nunc® minisorp tubes. 5. [3H]-Diprenorphine (PerkinElmer). 6. Naloxone (Sigma Aldrich). 7. GF/B filters. 8. Brandel® cell harvester (Alpha Biotech). 9. Ultima Gold scintillation cocktail (PerkinElmer). 10. Hinge-cap vials. 11. Tricarb® scintillation counter (PerkinElmer).

124

Brillet, Pereira, and Wagner

3. Methods 3.1. Cell Culture Procedures

This section describes the different methods required for culturing S2 cells, from T-flask culturing (initiating, maintaining, and freezing cells) to suspension culturing in shake flasks or in a bioreactor.

3.1.1. Culturing S2 Cells

This method describes how to start and maintain the S2 culture from 1 mL of liquid nitrogen-frozen stock containing 1 × 107 cells (see Note 2). 1. Remove the vial of cells from liquid nitrogen and thaw quickly in your hand. 2. Decontaminate the outside of the vial with 70% ethanol and transfer quickly in 5 mL of room temperature InsectXpress complete medium. 3. Centrifuge the cells at 100g for 2–3 min and remove the medium. Resuspend the cells with 5 mL of InsectXpress complete medium and plate them in a T25-cm2 flask. 4. Incubate at 27°C under normal atmosphere for 24 h. 5. The next day, remove the medium from the flask and carefully add 5 mL of InsectXpress complete medium. Incubate at 27°C until the cells reach up to 20 × 106 cells/mL. This may take 3 or 4 days. 6. To maintain the culture, S2 cells are usually split twice a week in InsectXpress complete medium at a 1:5 to 1:10 dilution into new culture vessels. Cells can be split at a final density of 2 × 106 cells/mL. Never start a new culture at a density below 0.5 × 106 cells/mL.

3.1.2. Freezing S2 Cells

1. Remove the cells from the flask; tap several times to remove the cells attached to the surface. Briefly pipet the cell suspension up and down to break up clumps of cells. 2. Count a sample of the cell suspension in a hemacytometer using a standard trypan blue exclusion assay to determine the number of cells per milliliter and the viability (95–99%). 3. Pellet the cells by centrifuging at 100g for 2–3 min in a tabletop centrifuge at 4°C. Remove the supernatant from the cell pellet. 4. Resuspend in 90% FBS and 10% DMSO at a density of 1 × 107 cells/mL. 5. Aliquot 1 mL of the cell suspension per cryovial. Place the vials in a polystyrene box and allow the cells to freeze in a −80°C freezer for 24 h before transferring the vials to liquid nitrogen for long-term storage.

Expression of Membrane Proteins in Drosophila Melanogaster S2 Cells 3.1.3. Suspension Culture

125

1. Prepare a T75-cm2 flask S2 cells. When the culture reaches confluence, almost 20 × 106 cells/mL, inoculate a 500-mL bottle with 10 mL of the cell suspension and add 90 mL of InsectXpress complete medium (see Note 3). 2. Incubate the bottle in a heater/shaker incubator at 110 rpm and 27°C. 3. After 72 h, count a sample of the cell suspension in a hemacytometer with trypan blue to determine the number of cells per milliliter and the viability. The cell suspension is supposed to reach a density of almost 30 × 106 cells/mL with 95–99% viability. 4. The suspension culture can be then amplified and tested for production in larger volumes.

3.1.4. Bioreactor Cell Culture

1. Autoclave the empty Applikon bioreactor vessel, as well as a 1-L bottle equipped with the needed tubing (filtered vent hole and connection to the bioreactor). 2. Prepare 100 mL of a S2 cell suspension culture in InsectXpress complete medium as described. 3. Count a sample of the cell suspension in a hemacytometer with trypan blue to determine the number of cells per milliliter and the viability. 4. In the bottle prepared at step 1, dilute the cell in fresh InsectXpress complete medium to end with 1 L at a final concentration of 2.5 × 106 cells/mL and sterilely connect the bottle to the bioreactor. 5. Add the fresh cell suspension to the bioreactor by pressuring the bottle connected to the bioreactor (see Note 4). 6. Start the bioreactor culture using the following settings: 20% dissolved oxygen (DO) (see Note 5), 120 rpm stirrer speed, and a temperature controlled at 27°C. The DO can be controlled by sparging air or oxygen. 7. During the whole run, samples can be sterilely collected via a pipe in the culture fluid and transferred to sample vessels by pressuring the bioreactor for analysis (see Note 6).

3.2. Membrane Protein Expression

In this section, we describe how to transfect S2 cells and generate stable recombinant cell lines for the heterologous expression of membrane proteins.

3.2.1. Expression Vector

Due to space limitation reasons, we will not go into detail for the construction of the expression vector. Briefly, the BiP signal sequence, the EGFP, and hMOR open reading frames (ORFs) were cloned in frame under the control of the pMT promoter in the pMT/BiP vector using standard cloning procedures

126

Brillet, Pereira, and Wagner

(see Note 7). The resulting construct was sequence checked and, as was done for the pCoHYGRO vector, was Escherichia coli amplified and purified using a standard DNA miniprep kit. 3.2.2. Transfection for the Generation of Stable Cell Lines

For S2 cell transfection, several reagents can be successfully used (lipofectin, lipofectamine, cellfectin, etc.) [41). We prefer to use the jetPEI reagent, which proved the most efficient in our hands (see Note 8). 1. One day before transfection, plate 1.5 × 106 S2 cells on a sixwell plate with InsectXpress complete medium (see Note 9). 2. Dilute 5–10 mg of the respective pMT construct and 0.25– 0.5 mg pCoHYGRO (1:20 ratio) into a final volume of 100 mL of 150 mM NaCl. Vortex gently and spin down briefly. 3. Dilute 10–20 mL of jetPEI solution into a final volume of 100 mL of 150 mM NaCl. Vortex gently and spin down briefly. 4. Add the 100 mL of jetPEI solution to the 100 mL of DNA solution all at once (important: do not mix the solution in the reverse order). 5. Vortex the solution immediately and incubate for 30 min at room temperature. 6. Remove the cell medium from the six-well plate and add 800 mL of fresh complete medium. 7. Gently add 200 mL of jetPEI/DNA mixture per well and incubate 24 h at 27°C.

3.2.3. Selection of Stable Cell Lines

1. At 24 h after transfection, remove the medium and start selection by adding hygromycin at a final concentration of 300 mg/mL. 2. Change medium containing hygromycin every 2 or 3 days. 3. Two weeks after starting the selection, progressively decrease the antibiotic concentration to 0. 4. Two weeks later, you can proceed to protein expression analysis after CuSO4 induction for each of your stable cell lines. Select the best clone and start the amplification.

3.2.4. Induction of Recombinant Protein Expression

Induction of expression is easily achieved by simply adding sterile copper sulfate to the medium to a final concentration of 700 mM, whatever the culture format used (T-flask, bottle or bioreactor). It is important to spend time in optimizing this step since major variations can occur, resulting in dramatic improvements of the yield and activity of the expressed membrane protein. Several parameters can be tested, such as time course postinduction (from 12 to 72 h, for instance), temperature of induction (i.e., 22–28°C), or concentration of CuSO4 (250 mM to 1 mM). In the particular

Expression of Membrane Proteins in Drosophila Melanogaster S2 Cells

127

case of GPCR expression, addition of chemical additives such as 0.5–4% DMSO or specific antagonist ligands during the induction phase often proves helpful (15) (see Note 10). 3.2.5. Cell Sorting

As frequently observed with the S2 system, stable cell lines obtained after clone selection are often considered polyclonal and not homogeneously expressing the desired protein. Separative techniques such as FACS experiments could then be helpful in enriching the cell lines in receptor-expressing subpopulations. We describe here a robust alternative immunogenic method that can be applied to most membrane proteins to be expressed and that does not require expensive equipment (see Note 11). 1. After CuSO4 induction of an S2 culture, pellet about 107 cells at 900 rpm for 3 min and resuspend the cells in 25 mL PEB buffer. 2. Incubate at 4°C for 10 min with PEB buffer supplemented with 1/200 anti-EGFP monoclonal antibody (see Note 1). 3. Wash twice with PEB buffer and resuspend the cells in 25 mL PEB buffer supplemented with 1/5 volume goat antimouse IgG magnetic microbeads. 4. Incubate at 4°C for 15 min. 5. Wash twice with PEB buffer, centrifuge the sample at 100g for 10 min, and resuspend in 500 mL PEB buffer. 6. Fix the MACS column onto its magnet, apply 500 mL of degassed PEB buffer on top of the column, and let the buffer run through. 7. Pipet the magnetically labeled cell suspension onto the column and collect effluent as the negative fraction. 8. Wash the column three times with 500 mL PEB buffer and collect total effluent as the negative fraction. 9. Remove column from its magnet and place the column on a new sterile collection tube. 10. Apply 1 mL of PEB buffer to the reservoir of the column and flush the cells using the plunger supplied. 11. Wash the sorted cells with InsectXpress complete medium and split in 12-well plates. 12. After amplification of the sorted cells, test protein expression by CuSO4 induction.

3.3. Analyzing Yield and Activity of the Expressed Membrane Protein

Several techniques can be employed for assessing the activity and the quantity of an expressed membrane protein, starting either from whole S2 cell material or from membrane preparation samples. The most common approaches to evaluate membrane protein expression include the enzyme-linked immunosorbent assay

128

Brillet, Pereira, and Wagner

(ELISA) or ligand-binding assay for protein quantification, flow cytometry for evaluation of cells expressing the protein, confocal microscopy for analyzing the protein localization, spectrofluorimetry for the protein evaluation through fluorescence measurement, and Western blotting for initial qualitative analysis of the recombinant protein. At this step, the most helpful way, in terms of bioprocess optimization, is to express the amount of recombinant protein as evaluated by ELISA per number of cells in culture. After describing an adapted protocol for preparing total S2 membranes, we present in this section three methods suited to our EGFP–hMOR fusion that can be easily adapted to similar constructs. 3.3.1. S2 Cells Lysis and Membrane Preparation (See Note 12)

1. Centrifuge approximately 8 × 108 induced S2 cells at 100g for 3 min at 4°C. 2. Discard the supernatant and freeze the sample at −80°C. 3. Thaw the cells in 1 mL 50 mM Tris-HCl, pH 8, 1 mM EDTA. 4. Lyse the cells with 20 strokes in a Potter homogenizer. 5. Pellet the nuclei and the unbroken cells by centrifugation at 2,000g for 10 min at 4°C. Keep the supernatant. 6. Resuspend the pellet in 1 mL Tris-HCl, pH 8, 1 mM EDTA, lyse, and centrifuge again at 1,000g for 5 min at 4°C. 7. Pool both supernatants from steps 5 and 6 and ultracentrifuge at 100,000g for 30 min at 4°C. 8. Resuspend the membrane pellet in 500 mL of 50 mM TrisHCl, pH 7.4, 320 mM sucrose. 9. Homogenize through a 26-gauge needle and proceed to a protein concentration assay using the BCA kit. 10. Directly use samples for further analysis or store as aliquots at −80°C.

3.3.2. Western Blot Analysis

1. Mix the sample with Laemmli buffer for 30 min at room temperature before loading onto a 10% SDS-polyacrylamide gel and proceed to SDS-PAGE. 2. Electrotransfer the proteins onto a PVDF membrane at 60 V for 2 h. 3. Block the membrane in 3% nonfat dried milk, PBS, 0.5% Tween-20 (MPBST) for 30 min. 4. Incubate the membrane for 1 h in MPBST supplemented with 1/1,000 anti-EGFP or anti-hMOR monoclonal antibodies. 5. Wash the membrane three times for 5 min each in MPBST and incubate for 1 h in MPBST supplemented with 1/2,000 antimouse IgG HRP-linked whole antibody from sheep.

Expression of Membrane Proteins in Drosophila Melanogaster S2 Cells

129

6. Wash four times for 5 min each with MPBST and then detect the chemiluminescence using Supersignal West Pico chemiluminescent substrate according to the manufacturer’s instructions. 3.3.3. Fluorescence Measurements

As the protein is N-terminally fused to EGFP, fluorescence methods can be used directly on whole cells or subcellular fractions to visualize the expressed protein for different purposes (cytolocalization, quantification, etc.). We present here a method allowing the evaluation of the expression level in whole cells by directly recording the EGFP fluorescence. 1. In parallel cultures, induce recombinant and wild-type S2 cells. 2. Collect the cells from the flask and tap several times to remove the cells attached to the surface. Briefly pipet the cell suspension up and down to break up clumps of cells. 3. Count a sample of the cell suspension in a hemacytometer with trypan blue to determine the number of cells per milliliter and the viability. 4. Pellet the cells by centrifuging at 100g for 2–3 min in a tabletop centrifuge at 4°C. Remove the supernatant from the cell pellet. 5. Resuspend 107 cells in 1 mL of Insect Elliot Buffer. 6. Place the cell suspensions in a 1-mL cuvette with magnetic stirring and maintain at 21°C in the thermostated cell handler of a dedicated spectrofluorimeter, with slits sets to yield a bandwidth of 2 nm at excitation and emission. 7. After excitation at 470 nm, record the EGFP fluorescence from 490 to 750 nm. 8. Calculate the specific fluorescence of the sample by withdrawing the background of the wild-type cells to the fluorescence of the sample at 510 nm.

3.3.4. Ligand-Binding Assay

The typical GPCR activity ligand-binding assay can be carried out either on whole cells (present protocol) or on membrane preparations (see Note 13). In the latter case, 5–20 mg of membrane preparations can be used in each assay in place of whole cells. 1. After induction of recombinant S2 cells, prepare the cells as described in steps 2 to 4 in the previous protocol for fluorescence assay (see Section 3.3.3). 2. Wash the obtained cell pellet twice in PBS buffer supplemented with 320 mM sucrose. 3. Resuspend the cells in 50 mM Tris-HCl, pH 7.4, 320 mM sucrose at a final concentration of 5 × 106 cells/mL.

130

Brillet, Pereira, and Wagner

4. Distribute 400 mL of the cell suspension in each immunosorp tube. 5. In tubes prepared for the determination of total binding activity, add 100 mL of 50 mM Tris-HCl, pH 7.4, 320 mM sucrose supplemented with a concentration range of [3H]-diprenorphine, typically from 0.05 to 3.2 nM final concentration (see Note 14). 6. In parallel tubes prepared for the determination of nonspecific binding, add the same solution as in step 8, supplemented with 2 mM Naloxone (see Note 14). 7. Vortex each tube and incubate for 40 min at room temperature. 8. Presoak GF/B filters with 0.1% polyethylenimine. 9. Filter the samples and wash twice with ice-cold 50 mM TrisHCl, pH 7.4, using a Brandel cell harvester. 10. Transfer the filters in hinge-cap vials and add 5 mL of Ultima gold scintillation cocktail. 11. Incubate 1 h in a dark place before counting the sample in a suitable scintillation counter. 12. Analyze the data to determine the specific ligand-binding values. This can be done using appropriate software such as GraphPad Prism.

4. Notes 1. Any other monoclonal antibody directed against the membrane protein of interest can work as well provided the detected epitope is exposed at the cell surface. 2. New stocks can be thawed every 4 months. 3. For an optimal suspension culture, we recommend using a working volume of 1:5 in shake flasks or bottles to provide suitable oxygen transfer through the medium surface to the cell culture during agitation (42). Also, whereas the Pluronic F-68 reagent is optional for classical culture in T-flasks, it is necessary to keep this reagent in suspension culture as it plays a crucial role for maintaining cell integrity. 4. If necessary, an antifoaming agent can be used, such as Contraspum 210 or Contraspum A 4,060 (Zschimmer and Schwarz) or any other antifoaming agents previously tested on an S2 cell culture.

Expression of Membrane Proteins in Drosophila Melanogaster S2 Cells

131

5. It is usually recommended to set the DO at 20%. We found, however, that lower DO (down to 5%) was also suitable and generated less foam. 6. Analyses of bioreactor samples typically include not only cell density and viability evaluations and determination of yields and activity in case of recombinant proteins but also enzymatic assays or nutrient consumption/metabolite production measurements to obtain information on the physiology of the cultured cells for bioprocess optimization. 7. For membrane proteins bearing an extracellular N-terminus, such as GPCRs, we recommend employing a BiP signal sequence fusion for better expression yields and protein activity. Alternatively, use expression vectors without the BiP sequence combined with a constitutive pAc promoter. Also, the selection gene (hygromycin) can be cloned into the expression vector, allowing a single transfection instead of a cotransfection as described here. 8. According to the manufacturer’s recommendations, use an optimal jetPEI ratio of 2 mL of jetPEI for 1 mg of plasmid DNA. 9. Contrary to other reagents, jetPEI is not affected by the presence of serum during transfection, so the InsectXpress complete medium can be used. 10. We recommend testing the different mentioned parameters that may affect the productivity and activity of the expressed protein separately and then combining in a second optimization round those exhibiting a positive effect (e.g., DMSO is believed to potentiate the effect of ligand supplementation). 11. A limiting dilution method for cell cloning can be laborious for S2 cells since they do not easily grow at densities lower than 103 cells/well, and a feeder cell methodology is not yet well developed for insect cells. 12. This protocol is designed for the preparation of total membranes. Further membrane fractionations may also be useful as they may allow the separation of active from nonactive membrane protein subpopulations, as illustrated in (14). 13. The present assay is designed for the mu opioid receptor but can be adapted for other receptors using appropriate ligands. Note also that the binding assay on whole cells will mainly detect receptors exposed at the cell surface, with the accessibility of intracellularly located receptors to ligand dependent on the hydrophobicity of the ligands used. 14. For each concentration of radioligand tested, we recommend carrying out the assay in triplicate.

132

Brillet, Pereira, and Wagner

References 1. Pereira CA (2008) Foreword for the special issue: heterologous gene expression in Drosophila melanogaster cells. Cytotechnology 57:9 2. Benting J, Lecat S, Zacchetti D, Simons K (2000) Protein expression in Drosophila Schneider cells. Anal Biochem 278:59–68 3. Percival MD, Bastien L, Griffin PR, Kargman S, Ouellet M, O’Neill GP (1997) Investigation of human cyclooxygenase-2 glycosylation heterogeneity and protein expression in insect and mammalian cell expression systems. Protein Exp Purif 9:388–398 4. Astray RM, Augusto E, Yokomizo AY, Pereira CA (2008) Analytical approach for the extraction of recombinant membrane viral glycoprotein from stably transfected Drosophila melanogaster cells. Biotechnol J 3:98–103 5. Yokomizo AY, Jorge SA, Astray RM, Fernandes I, Ribeiro OG, Horton DS, Tonso A, Tordo N, Pereira CA (2007) Rabies virus glycoprotein expression in Drosophila S2 cells I. Functional recombinant protein in stable co-transfected cell line. Biotechnol J 2:102–109 6. Zhang F, Ma W, Zhang L, Aasa-Chapman M, Zhang H (2007) Expression of particulateform of Japanese encephalitis virus envelope protein in a stably transfected Drosophila cell line. Virol J 26:4–17 7. Ohara T, Ohashi-Kobayashi A, Maeda M (2008) Biochemical characterization of transporter associated with antigen processing (TAP)-like (ABCB9) expressed in insect cells. Biol Pharm Bull 31:1–5 8. Aumiller JJ, Jarvis DL (2002) Expression and functional characterization of a nucleotide sugar transporter from Drosophila melanogaster: relevance to protein glycosylation in insect cell expression systems. Protein Exp Purif 26:438–448 9. Pantazis A, Segaran A, Liu CH, Nikolaev A, Rister J, Thum AS, Roeder T, Semenov E, Juusola M, Hardie RC (2008) Distinct roles for two histamine receptors (hclA and hclB) at the Drosophila photoreceptor synapse. J Neurosci 28:7250–7259 10. Bass C, Lansdell SJ, Millar NS, Schroeder I, Turberg A, Field LM, Williamson MS (2006) Molecular characterisation of nicotinic acetylcholine receptor subunits from the cat flea, Ctenocephalides felis (Siphonaptera: Pulicidae). Insect Biochem Mol Biol 36:86–96 11. Huang Y, Williamson MS, Devonshire AL, Windass JD, Lansdell SJ, Millar NS (1999) Molecular characterization and imidacloprid selectivity of nicotinic acetylcholine receptor

subunits from the peach-potato aphid Myzus persicae. J Neurochem 73:380–389 12. Huang Y, Williamson MS, Devonshire AL, Windass JD, Lansdell SJ, Millar NS (2000) Cloning, heterologous expression and coassembly of Mpbeta1, a nicotinic acetylcholine receptor subunit from the aphid Myzus persicae. Neurosci Lett 284:116–120 13. Yagodin S, Hardie RC, Lansdell SJ, Millar NS, Mason WT, Sattelle DB (1998) Thapsigargin and receptor-mediated activation of Drosophila TRPL channels stably expressed in a Drosophila S2 cell line. Cell Calcium 23:219–228 14. Buckingham SD, Matsuda K, Hosie AM, Baylis HA, Squire MD, Lansdell SJ, Millar NS, Sattelle B (1996) Wild-type and insecticideresistant homo-oligomeric GABA receptors of Drosophila melanogaster stably expressed in a Drosophila cell line. Neuropharmacology 35:1393–1401 15. Brillet K, Perret BG, Klein V, Pattus F, Wagner R (2008) Using EGFP fusions to monitor the functional expression of GPCRs in the Drosophila Schneider 2 cells. Cytotechnology 57:101–109 16. Brillet K, da Conceição MM, Pattus F, Pereira CA (2006) Bioprocess parameters of cell growth and human mu opioid receptor expression in recombinant Drosophila S2 cell cultures in a bioreactor. Bioprocess Biosyst Eng 28:291–293 17. De Rivoyre M, Ruel L, Varjosalo M, Loubat A, Bidet M, Thérond P, Mus-Veteau I (2006) Human receptors patched and smoothened partially transduce hedgehog signal when expressed in Drosophila cells. J Biol Chem 281:28584–28595 18. Perret BG, Wagner R, Lecat S, Brillet K, Rabut G, Bucher B, Pattus F (2003) Expression of EGFP-amino-tagged human mu opioid receptor in Drosophila Schneider 2 cells: a potential expression system for largescale production of G-protein coupled receptors. Protein Exp Purif 31:123–132 19. Schetz JA, Kim OJ, Sibley DR (2003) Pharmacological characterization of mammalian D1 and D2 dopamine receptors expressed in Drosophila Schneider-2 cells. J Recept Signal Transduct Res 23:99–109 20. Cordova D, Delpech VR, Sattelle DB, Rauh JJ (2003) Spatiotemporal calcium signaling in a Drosophila melanogaster cell line stably expressing a Drosophila muscarinic acetylcholine receptor. Invert Neurosci 5: 19–28

Expression of Membrane Proteins in Drosophila Melanogaster S2 Cells 21. Aldecoa A, Gujer R, Fischer JA, Born W (2000) Mammalian calcitonin receptor-like receptor/receptor activity modifying protein complexes define calcitonin gene-related peptide and adrenomedullin receptors in Drosophila Schneider 2 cells. FEBS Lett 471:156–160 22. Torfs H, Oonk HB, Broeck JV, Poels J, Van Poyer W, De Loof A, Guerrero F, Meloen RH, Akerman K, Nachman RJ (2001) Pharmacological characterization of STKR, an insect G protein-coupled receptor for tachykinin-like peptides. Arch Insect Biochem Physiol 48:39–49 23. Torfs H, Shariatmadari R, Guerrero F, Parmentier M, Poels J, Van Poyer W, Swinnen E, De Loof A, Akerman K, Vanden Broeck J (2000) Characterization of a receptor for insect tachykinin-like peptide agonists by functional expression in a stable Drosophila Schneider 2 cell line. J Neurochem 74:2182–2189 24. Wang YK, Samos CH, Peoples R, PérezJurado LA, Nusse R, Francke U (1997) A novel human homologue of the Drosophila frizzled wnt receptor gene binds wingless protein and is in the Williams syndrome deletion at 7q11.23. Hum Mol Genet 6:465–472 25. Vanden Broeck J, Vulsteke V, Huybrechts R, De Loof A (1995) Characterization of a cloned locust tyramine receptor cDNA by functional expression in permanently transformed Drosophila S2 cells. J Neurochem 64:2387–2395 26. Millar NS, Baylis HA, Reaper C, Bunting R, Mason WT, Sattelle DB (1995) Functional expression of a cloned Drosophila muscarinic acetylcholine receptor in a stable Drosophila cell line. J Exp Biol 198:1843–1850 27. Tota MR, Xu L, Sirotina A, Strader CD, Graziano MP (1995) Interaction of [fluoresceinTrp25]glucagon with the human glucagon receptor expressed in Drosophila Schneider 2 cells. J Biol Chem 270:26466–26472 28. Bernard AR, Kost TA, Overton L, Cavegn C, Young J, Bertrand M, Yahia-Cherif Z, Chabert C, Mills A (1994) Recombinant protein expression in a Drosophila cell line: comparison with the baculovirus system. Cytotechnology 15:139–144 29. Wossning T, Reth M (2004) B cell antigen receptor assembly and Syk activation in the S2 cell reconstitution system. Immunol Lett 92:67–73 30. Schneider I (1972) Cell lines derived from late embryonic stages of drosophila melanogaster. J Embryol Exp Morphol 27:363–365 31. Rämet M, Manfruelli P, Pearson A, MatheyPrevot B, Ezekowitz RA (2002) Functional

32. 33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

133

genomic analysis of phagocytosis and identification of a Drosophila receptor for E. coli. Nature 416:644–648 Sondergaard L (1996) Drosophila cells grow to high densities in a bioreactor. Biotechnol Tech 10:161–166 Yagodin S, Hoyland J, Mason WT, Miyake T, Sattelle DB (1997) Imaging of intracellular Ca2+, pH and Cl-1 transients in Drosophila cell lines. Bioimages 5:111–118 Angelichio ML, Beck JA, Johansen H, IveyHoyle M (1991) Comparison of several promoters and polyadenylation signals for use in heterologous gene expression in cultured Drosophila cells. Nucleic Acids Res 19:5037–5043 Bunch TA, Grinblat Y, Goldstein LS (1988) Characterization and use of the Drosophila metallothionein promoter in cultured Drosophila melanogaster cells. Nucleic Acids Res 16:1043–1061 Kirkpatrick RB, Matico RE, McNulty DE, Strickler JE, Rosenberg M (1995) An abundantly secreted glycoprotein from Drosophila melanogaster is related to mammalian secretory proteins produced in rheumatoid tissues and by activated macrophages. Gene 153:147–154 Lee DF, Chen CC, Hsu TA, Juang JL (2000) A baculovirus superinfection system: efficient vehicle for gene transfer into Drosophila S2 cells. J Virol 74:11873–11880 Kim KR, Kim YK, Cha HJ (2008) Recombinant baculovirus-based multiple protein expression platform for Drosophila S2 cell culture. J Biotechnol 133:116–122 Iwaki T, Castellino FJ (2008) A single plasmid transfection that offers a significant advantage associated with puromycin selection in Drosophila Schneider S2 cells expressing heterologous proteins. Cytotechnology 57:45–49 Jorge SAC, Santos AS, Spina A, Pereira CA (2008) Expression of the Hepatitis B virus surface antigen in Drosophila S2 cells. Cytotechnology 57:51–59 Park JH, Kim HY, Han KH, Chung IS (1999) Optimization of transfection conditions for expression of green fluorescent protein in Drosophila melanogaster S2 cells. Enzyme Microb Technol 25:558–563 Rodas VM, Marques FH, Honda MT, Soares DM, Jorge SAC, Antoniazzi MM, Medugno C, Castro ME, Ribeiro BM, Souza ML, Tonso A, Pereira CA (2005) Cell culture derived AgMNPV bioinsecticide. Biological constraints and bioprocess issues. Cytotechnology 48:27–39

Chapter 9 Membrane Protein Expression in the Eyes of Transgenic Flies Valérie Panneels and Irmgard Sinning Abstract Eukaryotic membrane proteins are often difficult to obtain in sufficient amounts for structural studies using classical cell culture overexpression systems. This could be due to incomplete protein maturation and insufficient trafficking to the plasma membrane or to a cytotoxic effect of the recombinant membrane protein changing the overall metabolism. The method presented here takes advantage of the membrane stacks in the retina, where rhodopsin resides. The idea is to direct G protein-coupled receptors (GPCRs) and transporters to these membrane stacks and to express the target proteins in the retina of transgenic Drosophila melanogaster. Drosophila was chosen since fly genetics are well established and rather easily accessible. Metabotropic glutamate receptor mGluRa from Drosophila was among the first examples for GPCRs expressed in this system. It showed high expression yield, functionality and high homogeneity. When the same protein was expressed in Sf 9 cells, however, contamination by immature protein was a problem. Encouraged by this success, the fly system is now successfully used for a larger variety of membrane proteins, including mammalian GPCRs and transporters. Key words: Drosophila, EAAT2, fly genetics, glutamate transporter, GPCR, membrane protein, overexpression, photoreceptor cells, retina, rhodopsin, transporter

1. Introduction Many membrane proteins (MPs) are present in their native membrane only in minute amounts. Therefore, in most cases the amounts needed for biochemical and structural studies can only be obtained by recombinant expression using different established techniques (1–4). Rhodopsin is a remarkable exception as it is present in high amounts in the retina, where it is involved in vision. Besides this important function, it is the most abundant G proteincoupled receptor (GPCR) and has served as a prototype for ­structural and functional studies of this family. The structure of rhodopsin was the first X-ray structure of a GPCR (5) and could be I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_9, © Humana Press, a part of Springer Science + Business Media, LLC 2010

135

136

Panneels and Sinning

determined starting from membranes of the retina, where rhodopsin is densely packed (6, 7). How rhodopsin is targeted to the specialized membrane stacks in the photoreceptor cells — termed rod outer segment (ROS) in mammals or rhabdomeres in Drosophila — is not completely understood (8). Membrane insertion and targeting of rhodopsin must be efficient since rhodopsin — in addition to its abundance —shows a high turnover. Encouraged by these observations, we developed the expression of recombinant MPs in the eyes of transgenic animals as a new, alternative method for overexpression. Drosophila melanogaster was chosen as it offers the advantage of well-established fly genetics. Expression levels, kinetics and location of the target protein can be modulated and controlled using different promoters and a transactivating Gal4/ upstream activating sequence (UAS) system (9, 10). First, a stable fly strain containing a UAS for the ectopic yeast protein Gal4 followed by the target protein to be expressed is generated. Next, this fly is crossed with another strain named the driver strain and containing the “promoter-Gal4” transposon to generate a population expressing the gene under the control of Gal4. The glutamate receptor from D. melanogaster (mGluRa) was successfully expressed as the first GPCR using this method (11). It was shown to be functional, and its capacity to bind glutamate was shown to depend on cholesterol (12). Since this first study, the method was developed further using green fluorescent protein (GFP) fusion constructs that allow monitoring of expression, purification and localization of the protein. A larger number of GPCRs and transporters, not only from Drosophila but also mammals, were successfully expressed in the fly eyes, demonstrating the great potential of this method (described elsewhere). Here, we give a practical introduction to this new approach for the expression of MPs. Briefly, the MP constructs are cloned into the transposon cassette of a vector, which is injected into Drosophila embryos. Transformant flies obtained from random transposition into the genome are selected and stabilised by crossing with balancer flies. A driver fly for eye-specific induction of expression is crossed with the stable strain, and cultures are scaled-up. The heads of frozen flies are collected, and membranes are prepared for further functional or structural studies.

2. Materials 2.1. Membrane Protein Cloning

1. pUAST vector (gift from C. Desplan, New York, USA). 2. Human glutamate transporter EAAT2 gene construct (gift from M. T. Besson and S. Birman, Marseille, France). 3. Enhanced GFP (EGFP) gene construct (BioCat, Germany).

Membrane Protein Expression

137

4. Topo® TA Cloning® kit (Invitrogen). 5. Qiaprep® Spin Miniprep and QIAquick Gel Extraction kits (Qiagen). 6. T4-DNA ligase (Biolabs). 7. SOC medium: 0.5% Yeast Extract, 2% Tryptone, 10 mM NaCl, 2.5 mM KCl, 10 mM MgCl2, 10 mM MgSO4, 20 mM Glucose. 8. Luria-Bertani (LB) medium: 10 g/L Bactotryptone, 5 g/L Bacto-yeast extract, 10 g/L NaCl, supplemented with 15 g/L agar for plates. 2.2. Transfection in Schneider Cells

1. Schneider’s medium (Gibco). 2. Actin-Gal4 construct (gift from J. Grosshans, Heidelberg, Germany). 3. FuGENE6® transfection reagent (Roche). 4. Rabbit polyclonal antibody against GFP (BioCat).

2.3. Generation of Transgenic Drosophila

1. Incubator for fly cultures with a humidity controller WB750KHF (Mytron, Germany). The humidity was set to 70%. 2. Balancer and driver fly strains (Bloomington Drosophila Stock Center at Indiana University, http://flystocks.bio.indiana. edu/). Balancer flies If/CyO and If/CyO; Sb/TM3Ser (gift from A. M. Voie, Heidelberg, Germany). Driver strains Rh1Gal4 and GMR-Gal4 (gift from C. Desplan, New York, USA). 3. CO2 bottle connected through two different tubings to a glass filter (Neolab 25 209 04) and an injecting needle, respectively, for anaesthetising the flies. 4. Classical stereomicroscope (10× objective) for the observation and sorting of the flies.

2.4. Analysis of Overexpression by Western Blot and Fluorescence Microscopy 2.4.1. Western Blot 2.4.2. Fluorescence Stereomicroscopy

1. Eppendorf homogeniser (Neolab). 2. Classical sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) gel (8% separating gel and 3% stacking gel). 3. PVDF membrane (Immobilon-Millipore). 4. Rabbit polyclonal antibody against GFP (BioCat). 1. Stereomicroscope (10× objective) mounted with an HBO mercury lamp and a GFP filter (illumination path BP 480/40 nm, dichroic mirror/reflector 505 nm, observation path LP 510 nm). 2. Fluorescence was recorded with a digital camera (DC200, Leica).

2.5. Flies and Head Harvesting 2.5.1. Fly Food

1. Fly food (see Note 1): Agar (8 g/L), yellow cornmeal (80 g/L), dry yeast (18 g/L), light corn syrup (22 g/L), malt (80 g/L), soybean meal (8 g/L) and the fungicide nipagin

138

Panneels and Sinning

(0.5 g/L) in tap water (13). All components are boiled for 10 min except nipagin, which has to be added when the mixture is cooled to 60°C. Add propionic acid (7 mL per litre food). 2. Pour the food in small and large vials provided by Greiner BioOne (205101 and 960177, respectively) with, respectively, 10 and 35 mL warm food. 3. After solidification, plugs antimites (25- or 50-mm diameter, respectively; Buddeberg) are set on the tubes. 4. The food is stored at 18°C (see Note 1). 2.5.2. Fly Head Harvesting

1. Microsieve kit for fly head collection (Neolab). 2. Glass beads, 4 mm diameter (Biospec Products).

2.6. Membrane Preparation

1. Glass-to-glass hand homogeniser (20 mL) (Sartorius). 2. Complete® protease inhibitor cocktail (Roche), benzamidine and phenylmethylsulfonylfluoride (PMSF). 3. Sucrose buffer: 50 mM Tris-HCl, 150 mM NaCl, 2 mM MgCl2, 1 mM ethylene glycol-bis(2-aminoethylether)-N, N, N′, N′-tetraacetic acid (EGTA), 250 mM sucrose, pH 7.4.

3. Methods Despite the large body of knowledge in the field (14), fly genetics are mostly restricted to specialised laboratories. The generation of transgenic fly strains overexpressing MPs in the eyes for structural studies requires the combination of technical knowledge in two different areas of research: MP engineering and Drosophila genetics. We describe here the overexpression of the human glutamate transporter EAAT2 in the eyes of transgenic flies. The description includes the molecular cloning of the MP into a transposon, the generation of the transgenic fly and the collection of heads for preparation of membranes as a starting material for functional and structural studies. 3.1. Membrane Protein Cloning

pUAST from the pCaSpeR family is a plasmid that is often used for transposition in the Drosophila genome (9). It contains a multiple cloning site (EcoRI-BglII-NotI-EagI-XhoI-KpnIXbaI) for insertion of the gene of interest under the regulation of a Gal4 UAS, followed by the white plus gene (w+), a marker coding for an ABC transporter restoring the natural red colour of the eyes, the P-elements flanking both cistrons and recognised for transposition into the genome and a gene for ampicillin

Membrane Protein Expression

139

Fig. 9.1. (a) Generation of transgenic flies. The gene construct for the membrane protein (MP) to be expressed in fusion with a histidine-tagged GFP (tev: tev-protease site) was cloned in the multiple cloning site (MCS) of the pUAST plasmid vector, downstream from the regulating element UAS-Gal4. This cistron, together with another one coding for the white gene (w+), a red-eye marker, are included in the transposon delimited by the P-element recognition sites (P ). The pUAST construct was co-injected with a helper plasmid coding for a translocase into the posterior pole of Drosophila embryos. After the usual 12-day cycle of the fly, the survivors were crossed with a white-eye w1118 fly. The red/orange-eye transformants were balanced by successive crossings with If/CyO and If/CyO;Sb/TM3Ser flies. The stable, balanced fly was finally crossed with a driver fly (here GMR-Gal4) for expression into the eyes. (b) The crossing of the UAS-MP-EGFP fly (here UAS-EAAT2-EGFP coding for a fluorescent glutamate transporter) with the driver GMR-Gal4 generated a population of fluorescent flies that could be easily selected using a fluorescence stereomicroscope (10× objective). Left panel a fluorescent eye of a transgenic fly expressing the human EAAT2 in fusion with EGFP. Right panel shows, as a negative control, a transgenic fly that does not have the EGFP transgene.

resistance (Fig.  9.1a) (see Note 2 for sequence and map of pUAST). 1. Amplify the MP gene, in our case the human glutamate transporter EAAT2 fused to EGFP, by classical polymerase chain reaction (PCR) with a high-fidelity polymerase using, for

140

Panneels and Sinning

instance, 200 ng template, 1 mL each forward and reverse primers (10 mM) (see Note 3), 5 mL polymerase buffer (10X), 1 mL polymerase Expand (3.4 U) and a dNTP (deoxynucleotide 5¢-triphosphate) mixture (25 mM of each four bases) in 50 mL final volume. PCR program: One cycle of 2 min at 95°C, 18 cycles of amplification (95°C for 30 s, 55°C for 30 s, 72°C for 1 min/kbp) and a last cycle of 10 min at 72°C. 2. Analyse the PCR product the same day by electrophoresis in an agarose gel. 3. Ligate 4 mL of PCR product with 1 mL of the pCR2.1 vector provided by the Topo TA Cloning kit. 4. Transform 50 mL chemical-competent Escherichia coli with 5 mL of the ligation mixture by incubation of the mixture 15 min on ice, heat shock for 30 s at 42°C, quick cooling for 2 min on ice and addition of 250 mL LB medium and incubation for 1 h at 37°C in a rolling wheel. 5. Plate the bacteria and select positive clones on kanamycin or ampicillin plates if the template used for the PCR gives ampicillin or kanamycin resistance, respectively. Let grow overnight at 37°C. 6. Inoculate the bacteria clones to test, usually six colonies, in 2 mL LB medium containing kanamycin (3 mg/mL) or ampicillin (10 mg/mL) and let grow overnight at 37°C. 7. Extract and purify the new construct using the classical Miniprep kit (Qiagen) and test the positive clones by digestion of one tenth of the sample by EcoRI. Analyse the restriction products on an agarose gel. 8. Send positive clones for sequencing with the M13(-21) and M13 reverse universal primers annealing on pCR2.1 and internal primers if necessary. 9. The gene is then subcloned into pUAST: Digest 1 mg pUAST vector and 2 mg gene construct in pCR2.1, respectively, with, for instance, EcoRI (20 U) and XbaI (20 U) (see Note 4) in buffer 2 in 30 mL. 10. Purify the linearised vector and the gene construct from an agarose gel using the gel extraction kit. 11. Ligate overnight 10 ng linearised pUAST with the purified gene construct in a stoichiometric ratio of 1:5 (see Note 5) with the T4-DNA ligase (1 mL, 400 U) in a final volume of 30 mL at 18°C. 12. Transform 50 mL chemical-competent E. coli with 10 mL of the ligation mixture by incubation of the mixture for 15 min on ice, heat shock for 30 s at 42°C, quick cooling for 2 min on ice, addition of 250 mL SOC medium and incubation for 1 h at 37°C in a rolling wheel.

Membrane Protein Expression

141

13. Plate the bacteria and select positive clones on an ampicillin plate. Let grow overnight at 37°C. 14. Inoculate the bacteria clones to test, usually 24 colonies, in 5 mL LB medium containing ampicillin (10 mg/mL) and let grow overnight at 37°C. 15. Extract and purify the new construct using the classical Miniprep kit and test the positive clones by double digestion of one fifth of the sample by EcoRI and XbaI. Analyse the restriction products on an agarose gel. 16. Positive constructs are verified for expression in Schneider cells and sequenced (see Note 6) before the injection into fly embryos. 3.2. Transfection in Schneider-2 Cells

The new construct was first tested in a quick transfection experiment. Because the pUAST plasmid has UAS (UAS-Gal4) for regulating the MP expression, a co-transfection with a plasmid offering Gal4 expression (i.e., under the actin promoter) is necessary. 1. Seed S2 Schneider cells at a density of 106 cells per 3-cm Petri dish in 1 mL Schneider cell medium. 2. Transfect the cells after 4 h with a mixture of 0.4 mg of purified actin-Gal4 vector and 0.4 mg of pUAST construct following the instructions for transfection with the FuGENE6 reagent. 3. Analyse protein expression after 48 h by Western blot using, for instance, a commercial antibody directed against the tag or a GFP antibody (see Note 7). 4. If the MP was cloned in frame with the GFP, analyse expression in the cells in vivo by fluorescence microscopy. Typical yields of co-transfection are around 30%.

3.3. Generation of Transgenic Drosophila 3.3.1. Injection of Embryos

3.3.2. Establishing Stable Fly Strains Using Balancer Chromosomes and Chromosome Mapping

The pUAST construct is co-injected with a helper plasmid coding for a transposase into the posterior pole of the embryos of white-eye flies w1118 (Fig.  9.1a). Because this step requires sophisticated material like a microinjection system and some routine of handling (for technical description, see (15)), it is recommended to send the purified pUAST construct (around 30 mg DNA; see Note 8) to a company offering a Drosophila injection service. Why Balance the Flies? In the first transformants (UAS-EAAT2eGFP), meiotic recombination could still occur in females, generating unstable genotypes and possibly different chromosomal location of the transposon. To reduce this effect, the transformants have to be crossed with flies having balancer chromosomes that are rearranged with multiple inversions and cannot undergo transposition events (for more details, see (16)). In the progeny,

142

Panneels and Sinning

heterozygous flies having one copy of the transposed chromosome and one copy of the balancer chromosome are kept as a stable “balanced stock.” Chromosome Mapping. Chromosome mapping is necessary to know which chromosome received the transposon and needs to be balanced. The chromosome number of Drosophila is 4: a sex chromosome of small size (also called chromosome 1), two large chromosomes 2 and 3, and a small chromosome 4. The probability to obtain a transposition in chromosomes 2 and 3 is higher than for the two other chromosomes. Chromosome mapping is deduced by phenotype exclusion, using the markers carried by the balancer flies. The strain used for balancing over chromosome 2 was If/CyO (see Note 9) (for the nomenclature, see Note 10). The presence of the If marker gives the flies rough eyes, and the Curly O marker CyO gives a curly wings phenotype (14). If in the second offspring of the crossing of the transformant fly with the balancer If/CyO, a fly has all three phenotypes (red eyes, curly wings (CyO phenotype) and rough eyes (If phenotype), the UAS element cannot be on chromosome 2. Therefore, the probability that the element is on chromosome 3 is high. Choice of the Balancer Chromosome. Most of our driver strains carry the Gal4 transposon on chromosome 2. It is therefore ideal to select an UAS-EAAT2-eGFP strain having the UAS element on chromosome 3. The transposed chromosome was then balanced over the TM3 Serrate chromosome (TM3Ser, chromosome 3) using the balancer fly If/CyO; Sb/TM3Ser. The Serrate marker is easy to detect and exhibits a typical wing phenotype (scalloping of the wing margin). The final stable fly stocks have the UAS-EAAT2eGFP/TM3Ser genotype and are ready for further crossing with the driver fly providing Gal4 and induction of expression. Practically (see Fig. 9.1a), 1. Collect If/CyO virgin flies in the early morning plus twice a day from a classical culture of flies grown in large vials. The flies are sorted under a stereomicroscope while anaesthetised under a CO2 stream. A virgin fly can be recognised through different criteria, including the larger size, the black dot visible through the skin of the abdomen, the green meconium presence, pupae-like folded wing (see (16)). The peak of female fertility is between 4 and 15 days after hatching. 2. Select orange- or red-eye transformants (here UAS-EAAT2eGFP) among the male flies delivered by the Drosophila injection service (see Note 11). 3. Cross each male with three virgins of the balancer strain If/ CyO at 25°C. 4. Let flies lay eggs and develop through all stages of development: embryos, first- until third-instar larvae, pupae (life cycle of 12 days).

Membrane Protein Expression

143

5. It is important to remove the parents from the vial before hatching. 6. Collect virgin flies from the balancer strain If/CyO;Sb/ TM3Ser. 7. Collect one male from the progeny of each cross, having red/ orange eyes and the “If” or “CyO” phenotype; cross each with three virgins from the balancer strain If/CyO;Sb/TM3Ser. 8. Remove the parents before hatching. 9. Let the offspring of the first crossing with If/CyO mate again for the chromosome mapping. 10. From the progeny of the crossing with balancer If/CyO;Sb/ TM3Ser, cross a red-eye virgin of phenotype CyO, TM3Ser, not If because of the eye phenotype, with a red-eye male of the same phenotype. The progeny of this crossing are balanced and stable and can be stored at 18°C. The establishment of a new stable fly strain takes around 2 months (see Note 12). If the UAS element was mapped on chromosome 3, the expected genotype is most likely:

+;

CyO UAS ; ;+ + TM3

If the UAS element was mapped on chromosome 2, the expected genotype is most likely: +;

3.3.3. Choice of the Driver

UAS TM3 ; ;+ CyO +

The balanced fly is still not expressing the recombinant MP because of the lack of the transcription factor Gal4. It requires crossing with a so-called driver, a fly expressing Gal4 under the control of a specific promoter. There are a number of different eye drivers that can be used that are available in the Bloomington Drosophila Stock Center. The “glass multiple reporter” GMR was used in this study for its high Gal4 induction and therefore high MP expression. 1. Collect virgin flies from the driver strain GMR-Gal4. 2. Cross a male of the new stable UAS strain balanced over TM3Ser (chromosome 3) with three virgin fly homozygotes for GMR-Gal4 on chromosome 2 (see Note 13). 3. Remove the parents before hatching. 4. From the new TM3Ser progeny, recross flies having the phenotype CyO, not Ser/TM3, and select the best fluorescent offspring for obtaining a homozygous population (see Note 14).

144

Panneels and Sinning

3.4. Analysis of Overexpression by Western Blot and In Vivo Fluorescence Microscopy

The transgenic flies expressing an MP fused with EGFP can be analysed in Western blot and fluorescence microscopy. For the Western blot analysis, five heads are dissected and homogenised using an Eppendorf homogeniser in 30 mL loading buffer. The samples are analysed as described for the Schneider cell expression. To test the overall fluorescence of the eyes (Fig. 9.1b) and selecting the best-expressing fly, a mercury fluorescence lamp and a GFP filter simply adapted on a classical stereomicroscope are required, which should be affordable for most laboratories.

3.5. Flies and Heads Harvesting

Flies are reared in small vials at 18°C for stocks and 25°C for crossings. Large vials are used for scaling-up the cultures, and flies are grown at room temperature (22°C). For large cultures (milligram amounts of target MP), 100 large vials distributed in three racks (43 × 30cm) give 50 mL flies per week. Flies are put to sleep by CO2 anaesthesia and collected in liquid nitrogen. The vials can be reused two or three times. New crossings were performed every 3 months to avoid overgrowth by contaminating flies.

3.5.1. Fly Cultures

3.5.2. Harvesting the Heads

1. Cool in liquid nitrogen a 5-mL volume of glass beads, the set of sieves, a funnel small enough to be adapted on a 15-mL Falcon™ tube and a 15-mL Falcon tube. 2. Transfer the fly stock from the −80°C freezer into liquid nitrogen. 3. Mix the 5 mL frozen beads with a 10-mL volume of flies in a cold 50-mL Falcon tube and shake vigorously three times to get rid of the wings that remain sticking to the Falcon tube. 4. Mount rapidly the sieve kit on the bench. Take the three smallest of the five sieves of the kit and mount a triple-sieve stack arranged in downward decreasing mesh diameters. 5. Transfer rapidly the mixture of flies and beads on the top of the cold triple-sieve stack. 6. Shake the sieve stack to break the fragile neck of the frozen flies. 7. Mount quickly the 15-mL Falcon tube on ice with the cold funnel on top. 8. Harvest the heads retained on the middle sieve by pouring them in the funnel (see Note 15). 9. Transfer the triple-sieve stack, the funnel and the 15-mL Falcon tube containing the heads back into liquid nitrogen. 10. Start again from the beginning with the next 10 mL of frozen flies (see Note 16). 11. Store the heads at −80°C.

3.6. Membrane Preparation

1. Pour a volume of 1 mL frozen heads into a 20-mL glass-toglass hand homogeniser (see Note 16).

Membrane Protein Expression

145

2. Because fly heads float in liquid, first give a small volume of 200 mL ice-cold sucrose buffer with protease inhibitors (Complete, 1 mM benzamidine, 1 mM PMSF) and remove air bubbles by simple pressure of the piston. 3. Add 1 mL ice-cold buffer to the compact tissue and homogenise the heads. 4. Centrifuge the homogenised material for 10 min at 2,000g, 4°C, to pellet cuticles and store the supernatant on ice. 5. Resuspend the pellet in 500 mL sucrose buffer and rehomo­ genise. 6. Repeat the procedure until the supernatant of centrifugation does not contain membranes, around three times. 7. Pool supernatants and centrifuge membranes for 60 min at 70,000g, 4°C. 8. Resuspend the resulting pellet in sucrose buffer at a concentration of 10 mg protein/millilitre and store in aliquots at −80°C. The membranes are ready for protein purification or activity tests (the MP analysis and purification are described in (17)).

4. Notes 1. The quality of the fly food is important, especially optimising the water content: If the food is too wet, the flies will stick and die; if the food is too dry, the culture will give low yields of larvae. The fly food must preferably be consumed in less than 3 weeks to avoid fungi growing and dryness. 2. The correct sequence of pUAST is available from Brand’s group (http://www.gurdon.cam.ac.uk/~brandlab/). The sequence from other Web sites may be incorrect. 3. The forward primer is designed with a Kozak-like sequence “cacaag” or “cggagc” directly preceding the start codon (18). Depending on the MP, different tags will be designed in the forward or reverse primer. For histidine, depending on the construct, the tag can be elongated from 6 to 8 or 10 histidines if necessary. 4. Unmethylated pUAST is necessary for optimal restriction with XbaI. 5. Different stoichiometric ratios of the gene construct and pUAST vector should be tried in case of tricky ligation. 6. The following sequencing primers matching with pUAST were routinely used: forward 5¢-CGTCAATTCAATTCAAA

146

Panneels and Sinning

CAAGC (−196 upstream EcoRI) and the reverse 5¢-GAGCTTT AAATCTCTGTAGG (−53 downstream XbaI). 7. For Western blot, MPs are not boiled but prepared 30 min at 45°C in a classical loading buffer. The antibody against GFP gave better results if incubated overnight. 8. pUAST is a low -copy plasmid and a midi-preparation (100 mL of bacteria culture) will be necessary for Schneider cell tests and fly generation. 9. Balancer fly strains are available in the Bloomington Drosophila Stock Center at Indiana University (http://flystocks.bio. indiana.edu/). 10. The standard nomenclature in Fly Genetics utilises a “+” for wild type, the four chromosomes are separated by a “;” (i.e., a wild-type fly would be +;+;+;+). The two alleles of sister chromosomes are separated by a “/” or “_” (i.e., If/CyO is the name of a fly having a balancer chromosome 2 “If” and a sister balancer chromosome 2 “CyO”). For ease of writing, wild-type chromosomes 1 and 4 of the strain If/CyO;Sb/ TM3Ser (+;If/CyO;Sb/TM3Ser;+) have been omitted. 11. To obtain additional fly strains per construct, the females can also be used and crossed with males of the balancer fly. 12. Already balanced fly strains can be ordered in from Drosophila injection services to avoid this time-consuming task that demands some experience in fly genetics. 13. Sometimes, it is advisable to use a weaker promoter (e.g., when protein expression is toxic). The rhodopsin promoter Rh1-Gal4 was successfully tested as a driver (e.g., for the human vasopressin receptor V1aR) (11). 14. The homozygous CyO genotype is lethal. 15. The top sieve with 25 mesh retained both the thorax-abdomen parts and the glass beads but not the heads, and the sieve with 35 mesh retained the heads but not the legs and wings. 16. A 50-mL volume of flies represents around 10,000 flies and gives 1 g of heads or 20 mg of membranes.

Acknowledgments We thank Ann Mari Voie (EMBL, Heidelberg) for technical advices and Silke Adrian for excellent technical assistance. This work was supported by the European Community Specific Targeted Research Project grant IMPS (Innovative Tools for Membrane Structural Proteomics, FP6-2003-LifeSciHealth 513770).

Membrane Protein Expression

147

References 1. Grisshammer R, Tate CG (1995) Overexpression of integral membrane proteins for structural studies. Q Rev Biophys 28:315–422 2. Tate CG (2001) Overexpression of mammalian integral membrane proteins for structural studies. FEBS Lett 504:94–98 3. Grisshammer R (2006) Understanding recombinant expression of membrane proteins. Curr Opin Biotechnol 17:337–340 4. Junge F, Schneider B, Reckel S, Schwarz D, Dotsch V, Bernhard F (2008) Large-scale production of functional membrane proteins. Cell Mol Life Sci 65:1729–1755 5. Palczewski K, Kumasaka T, Hori T, Behnke CA, Motoshima H, Fox BA, Le Trong I, Teller DC, Okada T, Stenkamp RE et  al (2000) Crystal structure of rhodopsin: a G proteincoupled receptor. Science 289:739–745 6. Zuker CS (1996) The biology of vision of Drosophila. Proc Natl Acad Sci U S A 93: 571–576 7. Fotiadis D, Jastrzebska B, Philippsen A, Muller DJ, Palczewski K, Engel A (2006) Structure of the rhodopsin dimer: a working model for G-protein-coupled receptors. Curr Opin Struct Biol 16:252–259 8. Kock I, Bulgakova NA, Knust E, Sinning I, Panneels V (2009) Targeting of Drosophila rhodopsin requires helix 8 but not the distal C-terminus. PLoS One 4(7):e6101 9. Brand AH, Perrimon N (1993) Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118:401–415

10. Elliott DA, Brand AH (2008) The GAL4 system: a versatile system for the expression of genes. Methods Mol Biol 420:79–95 11. Eroglu C, Cronet P, Panneels V, Beaufils P, Sinning I (2002) Functional reconstitution of purified metabotropic glutamate receptor expressed in the fly eye. EMBO Rep 3: 491–496 12. Eroglu C, Brugger B, Wieland F, Sinning I (2003) Glutamate-binding affinity of Droso­ phila metabotropic glutamate receptor is modulated by association with lipid rafts. Proc Natl Acad Sci U S A 100:10219–10224 13. Roberts DB (1998) Drosophila: a practical approach. Oxford University Press, Oxford 14. Lindsley DL, Zimm GG (1992) The genome of Drosophila melanogaster. Academic, San Diego, CA 15. Voie AM, Cohen S (1998) Cell biology: a laboratory handbook, vol. 3. Academic, New York, pp 510–517 16. Ashburner M (1989) Drosophila: a laboratory handbook and manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY 17. Panneels V, Eroglu C, Cronet P, Sinning I (2003) Pharmacological characterization and immunoaffinity purification of metabotropic glutamate receptor from Drosophila overexpressed in Sf9 cells. Protein Exp Purif 30: 275–282 18. Cavener DR (1987) Comparison of the consensus sequence flanking translational start sites in Drosophila and vertebrates. Nucleic Acids Res 15:1353–1361

Chapter 10 Expression of Mammalian Membrane Proteins in Mammalian Cells Using Semliki Forest Virus Vectors Kenneth Lundstrom Abstract One of the major bottlenecks in drug screening and structural biology on membrane proteins has for a long time been the expression of recombinant protein in sufficient quality and quantity. The expression has been evaluated in all existing expression systems, from cell-free translation and bacterial systems to expression in animal cells. In contrast to soluble proteins, the expression levels have been relatively low due to the following reasons: The topology of membrane proteins requires special, posttranslational processing, folding, and insertion into membranes, which often are mammalian cell specific. Despite these strict demands, functional membrane proteins (G protein-coupled receptors, ion channels, and transporters) have been successfully expressed in bacterial, yeast, and insect cells. A general drawback observed in prokaryotic cells is that accumulation of foreign protein in membranes is toxic and results in growth arrest and therefore low yields of recombinant protein. In this chapter, the focus is on expression of recombinant mammalian membrane proteins in mammalian host cells, particularly applying Semliki Forest virus (SFV) vectors. Replication-deficient SFV vectors are rapidly generated at high titers in BHK-21 (Baby Hamster Kidney) cells, which then are applied for a broad range of mammalian and nonmammalian cells. The SFV system has provided high expression levels of topologically different proteins, especially for membrane proteins. Robust ligand-binding assays and functional coupling to G proteins and electrophysiological recordings have made the SFV system an attractive tool in drug discovery. Furthermore, the high susceptibility of SFV vectors to primary neurons has allowed various applications in neuroscience. Establishment of large-scale production in mammalian adherent and suspension cultures has allowed production of hundreds of milligrams of membrane proteins that has allowed their submission to serious structural biology approaches. In this context, a structural genomics program for SFVbased overexpression of 100 GPCRs was established. Key words: Large-scale production, mammalian cells, Semliki forest virus, structural biology

1. Introduction Semliki Forest virus (SFV) is a single-stranded RNA (ss-RNA) alphavirus with an envelope structure belonging to the Togaviridae family (1). The SFV system exists in different variations, namely, I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_10, © Humana Press, a part of Springer Science + Business Media, LLC 2010

149

150

Lundstrom

replication-deficient, replication-competent, and DNA-based vectors (2, 3). Here, the focus is only on replication-deficient vectors as these are best suited for recombinant protein expression in mammalian cell lines. The gene of interest is cloned into the SFV expression vector downstream of the highly efficient subgenomic SFV 26S promoter (Fig. 10.1). Next, RNA is in vitro transcribed from the expression vector and cotransfected into BHK-21 (Baby Hamster Kidney cells) together with RNA coding for the structural proteins (capsid and envelope proteins). Due to the presence of the packaging signal only located on the recombinant RNA originating from the expression vector, the viral progeny will only contain this type of RNA (Fig. 10.2). The lack of RNA coding for the structural proteins renders the recombinant SFV particles replication deficient. However, the introduction of the recombinant RNA including the SFV nonstructural replicase genes on infection results in highly efficient RNA replication (estimated to 200,000 copies per cell), which is the basis for the high level of heterologous gene expression. Moreover, the SFV infection results in an almost-complete shutdown of endogenous gene expression, which further contributes to the almost-unique production of recombinant protein. The SFV system has found its solid place due to the following reasons: The SFV expression vector can accommodate at least 6-kb inserts (4); virus stocks can be prepared in less than 2 days, resulting in titers of 109 to 1010 infectious particles per milliliter. The broad host range of SFV allows transduction of a large number of mammalian cell lines and primary cell cultures (5). Expression of G protein-coupled receptors (GPCRs) and ion channels from SFV vectors has provided pharmacologically functional membrane proteins (6) and has demonstrated coupling of G proteins to GPCRs (7) and electrophysiological responses of ion channels (8). When coexpression of several genes is requested, for instance, G proteins and GPCRs, multiple SFV vectors can be coinfected (9). Efficient infection of both adherent and suspension cultures of mammalian cells has allowed scale-up of recombinant protein production in spinner and roller flasks as well as in bioreactors, which has allowed the production of hundreds of milligrams of recombinant membrane proteins (10). This development has allowed the application of SFV for structural genomics initiatives on membrane proteins (11, 12). The disadvantage of applying the SFV system is that, especially for large-scale production, the virus stock production is relatively expensive. The high costs are related to the in vitro transcription process, which requires expensive reagents such as (m7G(5') ppp(5')G) CAP analogue and SP6 RNA polymerase. Attempts to facilitate and reduce the costs for virus stock preparation have been introduced by the development of alphavirus packaging cell lines (13). Another concern has been the safety aspects of using

Expression of Mammalian Membrane Proteins in Mammalian Cells

151

SFV, especially for large-scale production, which obviously requires large volumes of virus. This aspect has been thoroughly investigated (14), and there is strong confidence in the safe use of SFV vectors. In this chapter, the subcloning into SFV vectors, recombinant virus production, and application of SFV particles for expression of membrane proteins in mammalian cell lines are described.

2. Materials 2.1. Cell Cultures

1. BHK-21 (Baby Hamster Kidney) cells.

2.1.1. Adherent Cells

2. CHO-K1 (Chinese Hamster Ovary) cells. 3. HEK293 (Human Embryonic Kidney) cells. BHK-21 cells are commonly used host cells for the packaging of recombinant SFV particles. They can be cultured in various media, but a 1:1 mixture of Dulbecco’s modified F-12 medium (Gibco BRL) and Iscove’s modified Dulbecco’s medium (IMDM) (Gibco BRL) supplemented with 4 mM glutamate and 10% fetal calf serum seemed to give robust cell growth and high titer virus stock production. CHO-K1 and HEK293 cells were cultured in the same medium. Application of alternative cell lines may require special media.

2.2. Suspension Cultures

2.3. SFV and Other Plasmid Vectors

1. DHI is a combination of DME, HamF12 and IMDM media 2. HL medium. Suspension cultures of BHK-21 and CHO-K1 cells for large-scale protein production were established in DHI medium: a mixture of DME, Ham F-12 and IMDM (Iscore’s Modified Dublecco’s Medium) media (1:1:2) supplemented with 5 mM glutamine, 5 g/L glucose, 5 mg/L insulin, 6 mg/L transferrin, 0.25% (v/v) Primatone RL (Sheffield Products/Quest International), 0.01% (v/v) Synperonic F68 (Serva), 20 mM selenite, 20 mM ethanol amine, and 2.5 mM b-mercaptoethanol (15). HEK293 suspension cultures were grown in HL medium: A 2:1 (by weight) mixture of DHI and RPMI-1640 supplemented with 5 mM glutamine and 5 g/L glucose (16). A higher concentration of 15 mg/L of transferrin was used for HEK293 cells. Otherwise, the components were identical to those used for BHK-21 and CHO-K1 cells. For SFV-based expression, a two-vector system for recombinant particle production was applied (Fig. 10.1). The most commonly used SFV vectors are pSFV1 and pSFV2gen (also called pSFV4.2).

152

Lundstrom 276 Eco RV 513 Bsi WI

a

1106 Bcl I 1425 Bpl I 1463 Stu I 1636 Ssi I 1799 Ecl 136II 1799 Sac I

Xmn I 9507 Pvu I 9279

pSFVgen2 Bsp LU11I 8019 Sap I 7896 Nru I 7830 Bsp 120I Apa I Avr II Bst BI Not I Spe I Xho I Sci I Xma I Sma I Bss HII Rsr II Bam HI

7484 7484 7469 7463 7454 7441 7432 7432 7424 7424 7415 7406 7400 Bgl II 6713 Xba I 6638

10610 base pairs Unique Sites

4916 Bsu 36I

Age I 5372

Fig. 10.1. Maps of SFV expression and helper vectors. (a) pSFV2gen; (b) pSFV-Helper2.

These vectors are linearized by SpeI and NruI, respectively. Both vectors can also be cut with SapI. SpeI is used for the linearization of the pSFV-helper2 vector. Polymase chain reaction (PCR) frag-

Expression of Mammalian Membrane Proteins in Mammalian Cells

153

ments are preferentially cloned into the pCR4Blunt-TOPO vector (Invitrogen) prior to subcloning into SFV vectors. 2.4. Reagents and Equipment

1. Restriction enzymes: NruI, SapI, and SpeI. 2. Agarose gel: 0.8%. 3. Gel electrophoresis apparatus. 4. Phenol/chloroform/isoamylalcohol: 25:24:1 (v/v/v). 5. 3M sodium actetate, pH 4.8. 6. Ethanol, 70% and 95% (v/v). 7. MicroSpin™ S-200 HR columns (Amersham). 8. SP6 buffer (10X): 400 mM HEPES, pH 7.4, 60 mM magnesium acetate, 20 mM spermidine. 9. 50 mM dithiotheritol (DTT). 10. rNTP mix: 10 mM rATP, 10 mM rCTP, 10 mM rUTP, 5 mM rGTP. 11. m7G(5¢)ppp(5¢)G CAP analogue. 12. RNase (ribonuclease) inhibitor: 10–50 U/mL. 13. SP6 RNA polymerase: 10–20 U/mL. 14. Trypsin–ethylenediaminetetraacetic acid (EDTA): 0.25% trypsin, 1 mM EDTA × 4 Na. 15. Phosphate-buffered saline (PBS). 16. Microcentrifuge, 1.5-mL microcentrifuge tubes. 17. Heating blocks. 18. Water baths. 19. Electroporator. 20. Electroporation cuvettes: 0.2 and 0.4 cm. 21. Tissue culture flasks: T25, T75, and T175. 22. Microwell plates: 6-, 12-, and 24-well plates. 23. Falcon tubes: 15 and 50 mL. 24. Sterile syringes: 1, 10, and 50 mL. 25. Sterile filters: 0.22 mm. 26. Opti-MEM I reduced-serum medium (Gibco BRL). 27. X-gal stock solutions: 50 mM K ferricyanide (+4°C); 50 mM K ferrocyanide (+4°C); 1M MgCl (room temperature); 2% X-gal in dimethylformamide (DMF) or dimethyl sulfoxide (DMSO) (−20°C). 28. X-gal staining solution: 1X PBS (1/10 of stock), 5 mM K ferricyanide (1/10 of stock), 5 mM K ferrocyanide (1/10 of stock), 2 mM MgCl (1/500 of stock), 1 mg/ml X-gal (1/20 of stock).

154

Lundstrom

29. Moviol 4–88 containing 2.5% DABCO (1,4-diazobicyclo(2.2.2)-octane). 30. Lysis buffer (50 mM Tris-HCl, pH 7.6, 150 mM NaCl, 2 mM EDTA, 1% (v/v) Nonidet P-40 (NP40)). 31. Hybond ECL nitrocellulose filter (Amersham). 32. TBST (TBS (Tris-based saline) buffer with 0.1% Tween-20). 33. ECL Chemiluminescence kit (Amersham). 34. Starvation medium (methionine-free MEM (Minimal Essential Medium), 2 mM glutamine, 20 mM HEPES). 35. Chase medium (E-MEM, 2 mM glutamine, 20 mM HEPES, 150 mg/mL unlabeled methionine).

3. Methods 3.1. Subcloning into SFV Vectors

General cloning procedures are performed for the subcloning of genes of interest into the multiple cloning sites (MCSs) of the SFV expression vectors. Due to the large size of the SFV vectors, it is advisable initially to clone PCR fragments into other types of cloning vectors, such as the pCR4Blunt-TOPO vector (Invitrogen). Please note that the linearization sites (SpeI, SapI, and NruI) in SFV cannot be used as cloning sites as the region between the MCS and the linearization sites contains the RNA replication and polyA+ signals. The unique XmnI site (at position 9929) in the pSFV1 vector cannot be used as the RNA transcript becomes too long to function properly. The presence of inserts and their correct orientation can be analyzed by restriction endonuclease digestions and nucleotide sequencing. DNA templates for in  vitro transcription reactions should preferentially be of DNA Midiprep or Maxiprep quality. However, initial tests can be carried out with Miniprep DNA (see Note 1).

3.2. DNA Linearization

SFV plasmid vectors are linearized by SpeI, SapI, or NruI under standard restriction digestion conditions. The completion of linearization should be verified by agarose gel electrophoresis (0.8%). 1. Linearize at least 5–10 mg plasmid DNA (larger quantities can be stored at −20°C) 2. Confirm complete digestion by agarose gel electrophoresis in comparison to uncut plasmid. 3. Purify linearized DNA by phenol/chloroform extraction followed by ethanol precipitation (overnight at −20°C or 15 min at −80°C) or on MicroSpin S-200 HR columns according to the manufacturer’s instructions.

Expression of Mammalian Membrane Proteins in Mammalian Cells

155

4. In case of phenol/chloroform extraction, centrifuge ethanol precipitates for 15 min at 18,000g at +4°C and wash with 70% ethanol.

3.3. In Vitro Transcription

Efficient in vitro RNA transcription is essential for the production of high-titer virus stocks (Fig. 10.2). Preferentially, fresh RNA preparations are made for each electroporation, although RNA transcripts can be stored for shorter periods (weeks) at −80°C. In case large quantities of virus stocks are produced, the process can be scaled up by multiplying the volumes for the in vitro transcription reactions and performing multiple electroporations in parallel. Prepare the in vitro transcription reactions at room temperature as the SP6 buffer contains spermidine, which might lead to precipitation at lower temperatures. Add the enzyme components last. Set up separate in vitro transcription reactions for expression and helper vectors in sterile 1.5-mL microcentrifuge tubes (see Note 2).

3.3.1. SFV In Vitro Transcription Reaction

5 mL (2.5 mg) linearized plasmid DNA 5 mL 10X SP6 buffer 5 mL 10 mM m7G(5¢)ppp(5¢)G

In vitr o transcription

Electroporation

BHK cells

rSFV parlticles

recombinant protein

Fig. 10.2. Schematic presentation of SFV particle production.

B & W IN PRINT

5. Repeat centrifugation for 5 min, air dry or lyophilize the DNA pellet, and resuspend in RNase-free H2O at a final concentration of 0.5 mg/mL.

156

Lundstrom

5 mL 50 mM DTT 5 mL rNTP mix (10 mM rATP, 10 mM rCTP, 10 mM rUTP, 5 mM rGTP) x mL RNase-free H2O to reach a final volume of 50 mL 1.5 mL (50 U/mL) RNase Inhibitor mL (20 U/mL) SP6 RNA polymerase 1. Mix all reaction components and spin briefly in a micro­ centrifuge. 2. Incubate 1 h at 37°C (see Note 3). 3. Load 1- to 4-mL aliquots on a 0.8% agarose gel for RNA quality control and continue the incubation for the rest of the samples. High-quality RNA generates relatively thick bands and no smearing. The size of ss-RNA molecules cannot be directly compared to DNA markers (l DNA or DNA ladders) as ss-RNA has four times faster mobility, but RNA from expression vectors has an approximate (depending on insert size) mobility of 8 kb, whereas helper RNA runs faster. 4. RNA molecules can be directly subjected to electroporations or stored at −80°C for later use. In case of storage, it is essential to reevaluate the RNA quality on thawing. Each transcription reaction should generate approximately 20–50 mg of RNA. 3.4. Electroporation of RNA

BHK-21 cells are good producers of high-titer SFV stocks, although other host cell can be used. The cells should not be passaged for more than 3 months as they lose their viability. Moreover, cells should not be cultured for more than 48 h prior to electroporation and should not exceed more than 80% confluency as old and too dense cell cultures generate significantly lower virus titers (see Note 4). 1. Plate  1:3 (overnight) or 1:5 (over two nights) sufficient quantities of BHK-21 cells in T175 flask(s). 2. Confirm cell growth and morphology by microscopy. 3. Wash cells once with PBS and trypsinize with 6 mL trypsinEDTA per T175 flask for 5 min at 37°C. 4. Resuspend cells to remove clumps and add cell culture medium to 25 mL. 5. Centrifuge for 5 min at 800g. 6. Resuspend cell pellet in a small volume (<5 mL) of PBS, add PBS to 25 mL, and centrifuge for 5 min at 800g. 7. Resuspend cells well in approximately 2.5 mL PBS per T175 flask, equivalent to 1–2 × 107 cells per milliliter. Use cells immediately for electroporation, although shorter storage (<1 h) on ice is acceptable.

Expression of Mammalian Membrane Proteins in Mammalian Cells

157

8. Transfer 0.4 mL cell suspension to 0.2-cm cuvettes or 0.8 mL to 0.4-cm cuvettes. 9. Add 20–45 mL of recombinant RNA and 20 mL of helper RNA from the in vitro RNA transcription reactions to the cell suspension. 10. Place cuvettes in the electroporator holder and apply two pulses. When using the Bio-Rad Gene Pulser, the following settings should be applied: 0.2-cm Cuvette

0.4-cm Cuvette

Capacitance extender (mF)

960

960

Voltage (V)

1,500

850

Capacitor (mF)

25

25

Resistance (pulse controller)

∝W

Disconnected

Expected time constant (tc)

0.8 s

0.4 s

The settings for the Bio-Rad Gene Pulser II require the following modifications: ●● The pulse controller should be set to “high range” and “∝.” ●●

●●

The capacitance rotary switch should be set to “high capacitance.” The following settings should be applied: 360 V and 75 mF.

The obtained resistance for 0.2-cm cuvettes is 10 W, and the time constant is 0.7–0.8 s. Please consult the manufacturer for appropriate conditions when using any other electroporator.

●●

11. Dilute the cells immediately after the electroporation 25-fold in cell culture medium and transfer cells to T flasks or plates. 12. Incubate cells at 37°C overnight in an incubator with 5% CO2. 3.5. Lipid-Mediated Transfection of RNA

Transfection reagents can be applied alternatively to electroporation. DMRIE-C and other transfection reagents have been applied for BHK-21 cells with relatively good results. Reagents have also been tested for other cell lines, such as COS7 and CHO-K1 cells. 1. Plate  1.5–3 × 105 BHK-21 cells in 35-mm dishes or on six-well plates in 2 mL medium and culture the cells to approximately 80% confluency. 2. Wash cells with 2 mL Opti-MEM I reduced-serum medium. 3. Prepare cationic lipid–RNA complexes as follows: Add 1 mL Opti-MEM I reduced-serum medium at room temperature to six 1.5-mL microcentrifuge tubes. Add 0, 3, 6, 9, 12, and 15 mL of DMRIE-C to the six tubes and vortex briefly. Mix 10 mL (~5 mg) in vitro transcribed recombinant RNA and 5

158

Lundstrom

mL (~2.5 mg) helper RNA, add the mixture to each tube, and vortex briefly. 4. Add the lipid–RNA complexes immediately to the washed cells and incubate at 37°C for 4 h. 5. Replace the transfection medium with prewarmed (37°C) complete BHK medium. 6. Incubate the BHK cells at 37°C overnight in an incubator with 5% CO2. 3.6. Harvest of Recombinant Viral Particles

Efficient production of recombinant SFV particles occurs within the first 24 h. Generally, titers in the range of 108 to 109 (occasionally 1010) infectious particles per milliliter are obtained. Slightly improved titers may be obtained by extension of the incubation time to 48 h (see Note 5). 1. Harvest virus by carefully removing the medium from the BHK-21 cells. 2. Remove cell debris and possible contaminants from the in vitro transcription reaction by filter sterilization through a 0.22-mm filter (see Note 6). 3. Aliquot virus stocks to avoid repeated cycles of freezing and thawing, which can reduce the titers significantly. 4. Store virus at −20°C (for weeks) and at −80°C (for years).

3.7. Activation of Recombinant SFV Particles

Virus production from conventional SFV helper vectors generates directly infectious particles. However, application of the pSFVHelper2 vector generates only conditionally infectious particles (17). In this case, the following additional activation steps are required: 1. Prepare stock solutions of a-chymotrypsin (20 mg/mL) and aprotinin (trypsin inhibitor) (10 mg/mL). 2. Add a-chymotrypsin to the virus stocks at a final concentration of 500 mg/mL and incubate at room temperature for 20 min (see Note 7). 3. Stop the reaction with aprotinin at a final concentration of 250 mg/mL. 4. The activated SFV particles are now ready for infections.

3.8. Verification of Virus Titers

Before proceeding with any experiments, it is essential to determine the titers of the generated SFV particles. As replication-deficient SFV does not produce plaques, the titers cannot be determined by conventional methods. Reporter genes such as green fluorescent protein (GFP) or b-galactosidase allow titer estimation by determination of the number of infected cells (i.e., cells expressing the respective reporter gene). Indirect titer determinations can also be performed by applying immunofluorescence methods.

Expression of Mammalian Membrane Proteins in Mammalian Cells

159

Approximate titer determination can also be done by cell morphology characterization by light microscopy. SFV-infected cells show a dramatic decrease in growth and can be easily distinguished from noninfected control cells through their changed morphology (round up). 3.9. Visualization of Reporter Gene Expression

1. Plate BHK-21 (or other) cells and culture them to a defined concentration on 6- or 12-well plates or on coverslips. 2. Infect cells with serial dilutions (e.g., twofold dilutions in the range expected to give 20–50 positive cells per microscope field) of virus stock expressing GFP or b-galactosidase. 3. Culture cells over night at 37°C. 4. Estimate the titers no later than 48 h postinfection. At earlier time points, the signal, especially for some mutant vectors, may be suboptimal, and at later time points the cytopathic effects result in an increasing number of detached cells (see Note 8).

3.9.1. GFP Detection

1. Apply fluorescence microscopy to count the number of GFPpositive cells. 2. Estimate the titer as infectious particles per milliliter based on the number of GFP-positive cells per well and taking into account the virus dilution.

3.9.2. X-Gal Staining

1. Wash cells with PBS and fix in cold methanol at −20°C for 5 min. 2. Wash three times with PBS. 3. Stain cells for at least 2 h in staining solution at 37°C or room temperature. 4. Count the X-gal-positive (blue) cells and estimate the titers as for GFP detection.

3.9.3. Immuno­fluorescence

1. Rinse coverslips twice with PBS and fix for 6 min at −20°C in methanol. 2. Wash coverslips three times in PBS. 3. Incubate for 30 min at room temperature in PBS containing 0.5% gelatin and 0.25% bovine serum albumin (BSA) to block unspecific binding. 4. Replace the blocking buffer with primary antibody in the same buffer for 30 min at room temperature. 5. Wash three times with PBS and incubate with secondary antibody for 30 min at room temperature.

160

Lundstrom

6. Wash three times with PBS and once with water and air dry coverslips. 7. Mount on glass slides using 10 mL Moviol 4–88 containing 2.5% DABCO. 8. Count the positive cells and estimate the titers as for GFP detection. 3.10. Evaluation of Gene Expression

Before any additional detailed and large-scale expression is conducted, it is good to rapidly evaluate the transgene expression from generated virus stocks. This approach also allows for determining the best expression conditions. Previous experience with alphaviruses has shown that virus concentration, host cell line, and time of cell harvest, among other factors, are important in expression optimization. Expression levels and size of the gene product can be relatively easily obtained by Western blotting. This, however, requires available antibodies against the target protein or introduction of tags in the vector constructs. If neither of these criteria is feasible, cells can be metabolically labeled with (35S) methionine as a means to verify transgene expression (Fig. 10.3).

3.10.1. Western Blotting

1. Infect appropriate cells (BHK-21, CHO-K1, HEK293) cultured on 6-, 12-, or 24-well plates with serial dilutions of virus stocks. 2. Add 250 mL lysis buffer per six-well plates at various times postinfection and incubate 10 min on ice. 3. Detach the cells by thorough resuspension and load samples on 10–12% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). 4. Electrotransfer the samples to a Hybond ECL nitrocellulose filter. 5. Treat the filter with 5% milk in TBST at +4°C for 30 min. 6. Treat the filter with the primary antibody at room temperature for 30 min. 7. Treat the filter with the secondary antibody at room temperature for 30 min. 8. Visualize specific bands with the ECL Chemiluminescence kit.

3.10.2. Metabolic Labeling

1. Infect appropriate cells (BHK-21, CHO-K1, HEK293) cultured on 6-, 12-, or 24-well plates with serial dilutions of virus stocks. 2. Remove medium from cells and wash once with PBS. 3. Add starvation medium and incubate at 37°C for 30 min.

Expression of Mammalian Membrane Proteins in Mammalian Cells

161

Fig. 10.3. Metabolic labeling of electroporated and SFV-infected cells. Lane 1 BHK-21 cells electroporated with pSFV1-NK1R and pSFV-Helper2 RNA. Lane 2 BHK-21 cells infected with recombinant SFV-NK1R particles.

4. Replace medium with 50–100 mCi/mL of (35S) methionine (in starvation medium) and incubate 20 min at 37°C. 5. Remove medium and wash twice with PBS. 6. Add chase medium for appropriate time (15 min to 3 h). 7. Remove chase medium and wash once with PBS. 8. Add 250 mL lysis buffer per six-well plate and incubate 10 min on ice. 9. Detach the cells by thorough resuspension, load samples on 10–12% SDS-PAGE, and run gel under standard conditions. 10. Fix the gel in 10% acetic acid, 30% methanol for 30 min at room temperature. 11. Replace the fixation solution with Amplify and incubate 30 min at room temperature.

162

Lundstrom

12. Dry the gel and expose on Hyperfilm-MP for 2–24 h (depending on signal) at room temperature or at −80°C, applying radioactivity-intensifying screens. 13. Develop film to visualize radioactive bands.

4. Notes 1. The quality of the DNA is essential for obtaining high-titer virus stocks. It is highly recommended to use commercial Midiprep or Maxiprep DNA kits for plasmid DNA preparation. 2. The composition of the in  vitro transcription reaction is important, especially the concentration of the CAP analogue m7G(5¢)ppp(5¢)G. Particularly, commercial buffers should be carefully evaluated as although RNA yields might be superior, the SFV titers might be inferior. 3. Obviously, the quality of the in  vitro transcribed RNA is crucial for efficient virus production. If the insert introduced into the expression vector exceeds 4 kb, the RNA yields might be improved by extending the incubation time for the in vitro transcription reaction. 4. The transfection/electroporation efficacy can be monitored for constructs with reporter genes by fluorescence (GFP) and X-gal staining (b-gal). 5. The efficiency of virus particle packaging can be indirectly validated by metabolic labeling of transfected/electroporated cells. High expression levels of the viral structural proteins (capsid, E1, and p62) suggest potentially good virus production (Fig. 10.3). Follow-up infection studies are, however, mandatory. 6. Generally, it is sufficient to filter sterilize the SFV stocks, and no further purification or concentration is necessary. However, the medium from BHK-21 cell has demonstrated some toxicity on primary cells in culture. For instance, for studies in primary neurons, it is recommended to subject the SFV stocks to ultracentrifugation or affinity column purification. 7. Only use fresh batches (or aliquots) of a-chymotrypsin for the SFV particle activation. Also, notice that there is a difference in potency of different batches of a-chymotrypsin. It is advisable to include reporter gene SFV stocks with known titers as a control. 8. In case low transduction efficiency is achieved, make the following verifications: Make sure that the host cell density is not too high in relation to the virus concentration. Validate that the virus activation is appropriate and that the virus stock is fresh or stored accordingly.

Expression of Mammalian Membrane Proteins in Mammalian Cells

163

9. It is essential that the mammalian cells (adherent or suspension cultures) used for large-scale production are physiologically healthy. Make sure to passage the cells frequently. 10. Avoid using any antibiotics for the cell cultures. Although the presence of antibiotics reduces contaminations, it does not prevent the growth of antibiotic-resistant bacteria, which certainly will affect both host cell functions and recombinant protein production.

References 1. Strauss JH, Strauss EG (1994) The alphaviruses: gene expression, replication, and evolution. Microbiol Rev 58:491–562 2. Liljeström P, Garoff H (1991) A new generation of animal cell expression vectors based on the Semliki Forest virus replicon. Bio/Technology 9:1356–1360 3. Lundstrom K (2003) Semliki Forest virus vectors for rapid and high-level expression of integral membrane proteins. Biochim Biophys Acta 1610:90–96 4. Ehrengruber MU, Lundstrom K, Schweitzer C, Heuss C, Schlesinger S, Gähwiler BH (1999) Recombinant Semliki Forest virus and Sindbis virus efficiently infect neurons in hippocampal slice cultures. Proc Natl Acad Sci U S A 96:7041–7046 5. Lundstrom K, Michel AD, Blasey HD, Bernard AD, Hovius R, Vogel H, Surprenant A (1997) Expression of ligand-gated ion channels with the Semliki Forest virus expression system. J Receptor Signal Transduct Res 17:115–128 6. Hassaine G, Wagner R, Kempf J, Cherouati N, Hassaine N, Prual C, André N, Reinhart C, Pattus F, Lundstrom K (2005) Semliki forest virus vectors for overexpression of 101 G protein-coupled receptors in mammalian host cells. Protein Exp Purif 45:343–351 7. Lundstrom K, Wagner R, Reinhart C, Desmyter A, Cherouati N, Magnin T, Zeder-Lutz G, Courtot M, Prual C, André N, Hassaine G, Michel H, Cambillau C, Pattus F (2006) Structural genomics on membrane proteins - comparison of more than 100 GPCRs in 3 expression systems. J Struct Funct Genomics 7:77–91 8. Lundstrom K (2005) Biology and application of alphaviruses in gene therapy. Gene Ther 12: S92–S97 9. Lundstrom K, Ehrengruber MU (2003) Semliki forest virus (SFV) vectors in neurobiology and gene therapy. Methods Mol Med 76:503–523

10. Lundstrom K, Mills A, Buell G, Allet E, Adami N, Liljestrom P (1994) High-level expression of the human neurokinin-1 receptor in mammalian cell lines using the Semliki Forest virus expression system. Eur J Biochem 224: 917–921 11. Hovius R, Tairi A-P, Blasey H, Bernard A, Lundstrom K, Vogel H (1998) Characterization of a mouse 5-HT3 receptor purified from mammalian cells. J Neurochem 70:824–834 12. Scheer A, Björklöf K, Cotecchia S, Lundstrom K (1999) Expression of the a1B-adrenergic receptor and G protein subunits in mammalian cells using the Semliki Forest virus expression system. J Receptor Signal Transduct Res 19: 369–378 13. Polo JM, Belli BA, Driver DA, Frolov I, Sherrill S, Hariharan MJ, Townsend K, Perri S, Mento SJ, Jolly DJ, Chang SM, Schlesinger S, Dubensky TW Jr (1999) Stable alphavirus packaging cell lines for Sindbis virus and Semliki Forest virus-derived vectors. Proc Natl Acad Sci U S A 96:4598–4603 14. Lundstrom K, Hermann D, Rotmann D, Schlaeger EJ (2001) Semliki Forest virus vectors in large-scale recombinant protein production: safety aspects and achievements. Cytotechnology 35:213–221 15. Schlaeger E-J (1996) The protein hydrosylate, Primatone RL, is a cost-effective multiple growth promoter of mammalian cell culture in serum-containing and serum-free media and displays anti-apoptosis properties. J Immunol Meth 194:191–199 16. Schumpp B, Schlaeger E-J (1990) Optimization of cell culture conditions for high density proliferation of HL60 human promyelocytic leukemia cells. J Cell Sci 97: 639–647 17. Berglund P, Sjöberg M, Garoff H, Atkins GJ, Sheahan BJ, Liljeström P (1993) Semliki Forest virus expression system: production of conditionally infectious recombinant particles. Bio/Technology 11:916–920

Chapter 11 Membrane Protein Expression in Cell-Free Systems Birgit Schneider, Friederike Junge, Vladimir A. Shirokov, Florian Durst, Daniel Schwarz, Volker Dötsch, and Frank Bernhard Abstract Cell-free expression has emerged as a promising tool for the fast and efficient production of membrane proteins. The rapidly growing number of successfully produced targets in combination with the continuous development of new applications significantly promotes the distribution of this technology. Membrane protein synthesis by cell-free expression does not appear to be restricted by origin, size or topology of the target, and its global application is therefore a highly valuable characteristic. The technology is relatively fast to establish in standard biochemical labs, and it does not require expensive equipment. Moreover, it enables the production of membrane proteins in completely new modes, like the direct translation into detergent micelles, which is not possible with any other expression system. In this protocol, we focus on the currently most efficient cell-free expression system for membrane proteins based on Escherichia coli extracts. Key words: Cell extracts, cell-free expression system, coupled transcription/translation, detergents, integral membrane protein, solubilization

1. Introduction Cell-free (CF) expression systems have emerged as efficient and versatile tools for the production of even complex integral membrane proteins (MPs) (1–3). The CF approach mimics the natural cytoplasmic cell environment yet remains independent from the requirements and sensitivity of living cells. It thus overcomes general limitations known from conventional cellular expression systems, like toxic effects caused by membrane incorporation of recombinant MP, by incompatibility of MP function with cell physiology or by overloading of essential MP-targeting machineries. In addition, the option to synthesize MPs directly into artificial hydrophobic environments like detergent micelles makes CF

I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_11, © Humana Press, a part of Springer Science + Business Media, LLC 2010

165

166

Schneider et al.

expression strategies independent from complex and inefficient transport and translocation systems necessary for the localization of MPs into lipid membranes (4). Depending on the selected CF protocol and reaction setup, incorporation efficiencies of supplied amino acids into the synthesised protein of higher than 10% can be obtained, resulting in the production of several milligrams of protein per 1 mL of reaction (5, 6). Further beneficial characteristics of CF expression systems are small handling volumes of only few millilitres, short reaction times that usually finish within a day, high tolerance of the CF systems for a wide variety of substances like detergents or lipids and high efficiency in label incorporation combined with almost no background protein expression. In contrast to conventional in vivo expression systems are the translation machineries of CF systems: open, accessible and not enclosed by a membrane. This feature primarily determines the high versatility of CF expression systems, allowing the supplementation of compounds at any point of the reaction. Such target-specific additives could be bioactive substances like ligands, substrates or inhibitors, but even protein components serving as chaperones, subunits or coenzymes could be considered. It should therefore be emphasised that reaction protocols can easily be modified by the supplementation of potentially beneficial additives according to specific requirements of a given MP target. Furthermore, efficient labelling of an expressed MP with, for example, seleno-methionine, 15N- or 13C-labelled amino acids or fluorescence-enhanced amino acids can simply be done by replacement of the desired amino acids in the reaction mixture (RM) with their labelled derivatives (7–9). Different CF expression modes, extract sources and reaction setups for the CF expression of MPs have been established, and the optimization and modification of these protocols is still a highly dynamic field. MPs can be CF produced as precipitate (P-CF) and solubilised in different detergents after expression (see Fig. 11.1a) (1, 10). They can further be expressed directly in soluble modes by providing hydrophobic environments with detergents (D-CF), lipids (L-CF) or mixtures thereof (6, 11–14). The most frequently used extract sources are currently Escherichia coli cells, wheat germs and rabbit reticulocytes. Post-translational modifications other than disulphide bridge formation have not been reported from either E. coli or wheat germ systems (15). Some evidence of glycosylations in the less-productive rabbit reticulocyte extracts exists (16). Reaction setups could be onecompartment batch formats or the more complicated two-compartment continuous exchange cell-free (CECF) formats (see Fig. 11.1b). The more efficient CECF configuration contains a reservoir of low molecular weight precursors in a feeding mixture (FM) that is separated from the RM by a semi-permeable membrane (17, 18). This ensures the exchange of substrates and

Membrane Protein Expression in Cell-Free Systems

167

Fig. 11.1. Modes and systems for the CF expression of MPs. (a) Expression modes for the CF production of MPs. In the precipitate mode (P-CF), no hydrophobic environments are provided, and the MPs precipitate after translation. In the detergent-based mode (D-CF), the synthesised MPs can stay soluble by insertion into the provided detergent micelles. In the lipid-based mode (L-CF), the synthesised MPs have the possibility to integrate into supplemented liposomes. (b) CF expression in the CECF (left) and in the batch system (right). The two-compartment CECF system consists of a RM and an FM. Both are separated by a semi-permeable membrane, which provides optimal exchange of low molecular weight compounds. In the batch system, only a RM compartment is present.

by-products between the two compartments over a certain time period. resulting in higher production efficiency that can yield several milligrams of MP per 1 mL of RM. CECF reactions can be

168

Schneider et al.

scaled up to RM volumes of several millilitres without significant loss of efficiency. This chapter describes the basic protocols for the general CF production of MPs in the CECF configuration using E. coli extracts.

2. Materials 2.1. Preparation of T7 Polymerase

1. Bacterial strain, like E. coli BL21 (DE3) (Merck, Darmstadt, Germany). 2. Bactotryptone. 3. Chloroform. 4. Dialysis buffer: 10 mM K2HPO4/KH2PO4, pH 8.0, 10 mM NaCl, 0.5 mM ethylenediaminetetraacetic acid (EDTA), 5% glycerol and 1 mM 1,4-dithiotreitol (DTT). 5. Dialysis membrane, molecular weight cutoff (MWCO) 30 kDa. 6. Equilibration buffer: 30 mM Tris-HCl, pH 8.0, 1 mM EDTA, 50 mM NaCl, 5% glycerol and 10 mM b-mercaptoethanol. 7. 1M isopropyl b-d-1-thiogalactopyranoside (IPTG). 8. Phenol. 9. Q-Sepharose column. 10. Plasmid pAR1219 or another plasmid carrying the gene for T7 RNA polymerase (T7RNAP). 11. Resuspension buffer: 30 mM Tris-HCl, pH 8.0, 10 mM EDTA, 50 mM NaCl, 5% glycerol and 10 mM b-mercaptoethanol. 12. Sample concentrators, MWCO 30,000. 13. Sonicator or French press device. 14. 30% (w/v) streptomycin sulphate. 15. 5X transcription buffer: 200 mM Tris-HCl, pH 8.0, 60 mM MgCl2, 25 mM DTT and 5 mM spermidine. 16. Yeast extract.

2.2. Preparation of E. coli S30 Extract

1. 10 L fermenter. 2. Centrifuge. 3. Photometer. 4. Dialysis membrane, MWCO 12–14 kDa. 5. 2M DTT. 6. 2xYTPG medium (per litre): 2.99 g KH2PO4, 6.97 g K2HPO4, 19.82 g glucose, 16 g tryptone, 10 g yeast extract, 5 g NaCl.

Membrane Protein Expression in Cell-Free Systems

169

7. b-Mercaptoethanol. 8. Bacterial strains like E. coli A19 (E. coli Genetic Stock Center, New Haven, CT) or E. coli BL21 DE3 (Merck) or E. coli Rosetta (Merck) or E. coli BL21-Star (DE3) (Invitrogen, Carlsbad, CA) (see Note 1). 9. Bactotryptone. 10. French press cell disruption device or high-pressure cell homogenizer. 11. Liquid nitrogen. 12. Phenylmethanesulfonylfluoride (PMSF). 13. Tris-(hydroxymethyl)-aminomethan (Tris). 14. S30-A buffer: 10 mM Tris-acetate, pH 8.2, 14 mM magnesium acetate tetrahydrate [Mg(OAc)2], 0.6 mM KCl, 6 mM b-mercaptoethanol (see Note 2). 15. S30-B buffer: 10 mM Tris-acetate, pH 8.2, 14 mM Mg(OAc)2, 0.6 mM KCl, 1 mM DTT and 0.1 mM PMSF (see Note 2). 16. S30-C buffer: 10 mM Tris-acetate, pH 8.2, 14 mM Mg(OAc)2, 0.6 mM potassium acetate (KOAc), 0.5 mM DTT (see Note 2). 17. Yeast extract. 2.3. E. coli CECF Expression

All solutions are prepared with Milli-Q water and kept at −20°C if not stated otherwise. 1. Reaction container (see Section 11.2.4. and Fig. 11.2). 2. Dialysis membrane, MWCO 12–14 kDa. 3. Slide-A-Lyzer (Pierce, Rockford, IL, USA), 10-kDa MWCO (see Note 3). 4. 50-mL glass vials. 5. Rolling device (Fröbel Labortechnik, Lindau, Germany) or temperature-controlled shaking incubator. 6. 500 mM DTT. 7. 1M acetyl phosphate lithium potassium salt (AcP), pH 7.0 adjusted with KOH. 8. 4 mM amino acid stocks (see Note 4). 9. 16.7 mM amino acid mix R, C, W, M, D and E (see Note 5). 10. 50-fold stock Complete® protease inhibitors cocktail with EDTA (Roche Diagnostics). 11. Detergents: Polyoxyethylene-(23)-laurylether (Brij-35), polyoxyethylene-(20)-cetyl-ether (Brij-58), polyoxyethylene-(20)-stearylether (Brij-78), digitonin (Sigma-Aldrich); n-dodecyl-b-d-maltoside (DDM) (Applichem SigmaAldrich); 1-myristoyl-2-hydroxy-sn-glycero-3-[phospho-rac(1-glycerol)] (LMPG) (Avanti lipids); Fos-Choline 12 (DPC) (Anatrace) (see Note 6).

170

Schneider et al.

Fig. 11.2. Reaction containers suitable for CECF expression. (a) Dimensional sketch of CECF mini-reactors as described (19). The RM is kept in a flat chamber for optimal volume-to-surface ratio, which is highlighted in dark in the cross section. Volumes of 50–70 mL are commonly used. These containers are suitable for incubation in 24-well plates with the wells holding the appropriate volume of FM. (b) Dimensional sketch of CECF maxi-reactors. Commercially available SlideA-Lyzers are used as containers for the RM; these are placed in a Plexiglas box that is specifically designed for a 1: 17 ratio of RM to FM (see Note 24). (c1) Standard 0.5-mL plastic reaction tube. The inner part of the cap holds the RM; the membrane is fixed between the cap and the tube. The bottom part of the tube is cut off, and the FM can then be filled into the tube. The tube is then sealed with Parafilm to prevent evaporation. Substance exchange between RM and FM can be improved by placing an appropriate magnetic bar in the RM to allow stirring during incubation. (c2) D-Tubes Mini and Midi. The RM is inside the reaction vial, which can be closed by a screw cap. For incubation, the D-tube has to be placed into a selected container of suitable size holding the FM. (c3) DispoDialyzer. The assembly is the same as explained for the analytical-scale D-tube devices. (c4) D-Tubes Maxi or Mega. 1 RM, 2 FM, 3 dialysis membrane.

12. 10 mg/mL folinic acid calcium salt. 13. 2.4M HEPES, pH 8.0, adjusted with KOH. 14. 1M Mg(OAc)2. 15. 75-fold NTP mix: 90 mM adenosine 5¢-triphosphate (ATP), 60 mM each guanosine 5¢-triphosphate (GTP), cytosine 5¢-triphosphate (CTP) and uridine 5¢-triphosphate (UTP) (Roche Diagnostics), pH 7.0 adjusted with NaOH. 16. 1M phospho(enol)pyruvic acid monopotassium salt (PEP) (Sigma-Aldrich), pH 7.0 adjusted with KOH.

Membrane Protein Expression in Cell-Free Systems

171

17. Plasmid vectors containing T7 regulatory elements, like pET (Merck) or pIVEX (Roche Diagnostics). 18. Plasmid DNA purification kits (e.g., Macherey and Nagel). 19. 40% (w/v) polyethylene glycol 8000 (PEG8000). 20. 4M KOAc. 21. 10 mg/mL pyruvate kinase (Roche Diagnostics). 22. 40 U/mL RiboLock™ RNase (ribonuclease) inhibitor (Fermentas) (see Note 7). 23. S30-C buffer without DTT (see Note 2). 24. S30 extract (see Section 11.3.1.3.). 25. 10% (w/v) sodium azide (NaN3). Store at room temperature (RT). 26. 400 U/mL T7-RNA polymerase (see Section 11.3.1.2.) (see Note 8). 27. 40 mg/mL total transfer RNA (tRNA) E. coli (Roche Diagnostics) (see Note 9). 2.4. Reaction Container

1. Home-made container: CECF mini-reactor for 50–70 mL RM, composed of Plexiglas with a Teflon ring (see Fig. 11.2a) (19). CECF maxi-reactor for 1–3 mL RM, composed of Plexiglas (see Fig. 11.2b) (14). 2. Commercial container (see Fig. 11.2c): D-Tubes 12- to 14-kDa MWCO (Merck) for analytical and preparative scale. DispoDialyzer (Spectrum, Rancho Dominguez, CA, USA) with a MWCO of 25 kDa composed of regenerated cellulose for preparative scale.

3. Methods 3.1. Preparation of Basic Reaction Components 3.1.1.Template Design for E. coli Extract-Based CF Expression

The T7RNAP can be used for DNA template transcription in E. coli systems if specific regulatory sequences have been attached to the target sequences (see Fig. 11.3). Key elements for E. coli CF systems are T7-promotor, Epsilon-enhancer of T7 phage gene g10, the ribosome -binding site and a T7-terminator region. It is also recommended to attach small purification or detection tags like, for example, a poly(His)n-tag, Strep-tag or T7-tag to the Nor C-terminus of the MP. In particular, codon-optimized N-terminal tags could generally enhance the expression levels (3, 11, 14). Standard expression vectors like pIVEX or pET

172

Schneider et al.

Fig. 11.3. DNA template construction for the CF expression of MPs with the T7RNAP in E. coli extracts. Optimal regulatory elements controlling the transcription of the target sequence comprise the T7-promotor, Epsilon-enhancer, a ribosombinding site and the T7-terminator. PCR DNA templates can be generated by, for example, the split-primer PCR. This approach requires four primers as illustrated. First, a PCR product is generated with equal amounts of primers 3 and 4 (10 nM each). This product is then used as template in a second PCR reaction with a 100-fold excess of primers 1 and 4 with respect to primer 2 (100 to 1 nM ).

derivatives can be used. DNA templates can be provided either as plasmids or as polymerase chain reaction (PCR) products. Linear PCR templates containing all required T7-promoter sequences can be generated by the split primer method using several overlapping primers (see Fig. 11.3) (20). The quality of DNA templates is crucial for the success of CF reactions. The template should be free of salts and residual RNA. If RNase was used during plasmid preparation, phenol/chloroform deproteinisation should be performed. In addition, relatively high concentrations of approximately 0.2 mg/mL are necessary, and the DNA should be resolved in water (see Note 10).

Membrane Protein Expression in Cell-Free Systems 3.1.2. T7-RNA Polymerase Preparation

173

T7RNAP is needed in high amounts and in high concentrations, and it can efficiently be prepared by heterologous overproduction in BL21 (DE3) cells. 1. Take a fresh overnight culture of BL21 (DE3) × pAR1219, carrying the gene for the T7RNA polymerase (21) and inoculate 1 L LB medium in a ratio of 1:100 in a culture flask. 2. Incubate on a shaker at 37°C until an OD600 of 0.6–0.8 and induce expression with 1 mM IPTG. 3. After 5 h of incubation at 37°C, harvest the cells for 15 min at 4,500g and 4°C. 4. Resuspend the pellet in 30 mL resuspension buffer (see Section 11.2.1.). 5. Disrupt the cells with a French press cell-disrupting device at 20,000 psi or by sonication. 6. Centrifuge for 30 min at 20,000g and 4°C. 7. Precipitate nucleic acids with 4% streptomycin sulphate for 5 min on ice and centrifuge for 30 min at 20,000g and 4°C. 8. Apply the supernatant on a 40-mL Q-Sepharose column equilibrated with 2 CV equilibration buffer (see Section 11.2.1.). 9. Remove unbound proteins by washing with equilibration buffer (see Section 11.2.1.) at a flow rate of 4 mL/min. 10. Elute bound proteins with a gradient from 50 to 500 mM NaCl at 3 mL/min. 11. Pool the purest fractions containing T7RNAP and dialyse overnight against the dialysis buffer (see Section 11.2.1.). 12. Concentrate T7RNAP to 1–4 mg/mL by ultrafiltration. Adjust to 10% glycerol to avoid precipitation of the T7RNAP during the concentration procedure. 13. Store T7RNAP in aliquots in the dialysis buffer adjusted to 50% glycerol at −70°C up to 1 year. The catalytic activity of purified T7RNAP can be analysed by in vitro transcription and be compared to the activity of a commercially obtained polymerase. In that case, the protocol has to be continued with the following steps: As template, a plasmid containing a reading frame under control of the T7-promoter has to be used. A template messenger RNA (mRNA) of defined size is produced by runoff transcription after linearising the plasmid by restriction at the 3¢ end of the reading frame. 14. Purify the linearised plasmid from restriction enzymes by phenol/chloroform extraction. 15. Combine 50 ng of linearised plasmid with 0.1 U/mL RiboLock, 2 mM dNTPs and transcription buffer (see Section 11.2.1.). 16. Add different concentrations of the purified T7RNAP samples.

174

Schneider et al.

17. Incubate the reactions for 2 h at 37°C and stop by adding 2 mM EDTA. 18. Analyse and compare the intensity of the synthesized RNAs on agarose gels. 3.1.3. Preparation of E. coli Extracts

The protocol described is principally based on Zubay’s procedure for preparation of S30 extracts (22), but it utilizes a modified runoff step. First day 1. Inoculate 100 mL LB media with the selected strain and incubate with shaking at 37°C overnight. 2. Prepare the 2xYTPG media in the 10 L fermenter (see Note 11). 3. Chill 15 L of distilled water. 4. Prepare 50-fold stock solutions for S30-A, -B and -C buffer, without b-mercaptoethanol, DTT and PMSF, sterile filter and store them at 4°C. Prepare also a 4M NaCl solution for the dissociation step. 5. Autoclave a funnel for filling the components into the fermenter. Second day 6. Inoculate the 10 L fermenter with 2xYTPG medium in a ratio of 1:100 with the preculture (see Note 12). 7. Grow the cells at 37°C with 500 rpm stirring and high aeration until they reach mid-logarithmic phase (see Note 13). 8. Chill the cells within the fermenter rapidly to a temperature below 12°C. This step should not exceed 45 min. After this step, the cells have to be kept at 4°C. 9. Harvest the cells by centrifugation at 5,000g for 15 min at 4°C. 10. Resuspend the pellet with a glass stick in 300 mL precooled S30-A buffer and centrifuge for 15 min at 5,000g and 4°C. Repeat this step three times with the same buffer but extend the centrifugation time at the final step for 30 min. 11. Weigh the pellet and resuspend it in 110% (v/w) of precooled S30-B buffer. A 10-L fermenter should produce a pellet with a wet weight of approximately 100–150 g depending on the fermentation conditions (see Note 14). 12. Disrupt the cells once with a precooled French press device using a constant pressure of 20,000 psi or with a high-pressure homogenizer at 17,500 psi and a flow rate of 1–3 mL/ min (see Note 15). 13. Centrifuge the disrupted cells at 30,000g for 30 min at 4°C. Take the upper non-turbid two-thirds of the supernatant and transfer it to fresh centrifuge tubes. Repeat the centrifugation

Membrane Protein Expression in Cell-Free Systems

175

step with the supernatant and again take the upper two-thirds of the supernatant. 14. The supernatant has to be adjusted to a final concentration of 400 mM NaCl followed by incubation at 42°C for 45 min. This step is used to get rid of endogenous mRNA. The solution will become turbid after incubation (see Note 16). 15. Dialyze the turbid extract first against 100 volumes of precooled S30-C buffer for 2 h at 4°C with gentle stirring using a dialysis membrane with a 12- to 14-kDa MWCO. Change buffer once and dialyse a second time overnight at 4°C. Third day 16. Centrifuge the extract at 30,000g for 30 min at 4°C. 17. Aliquot the extract into suitable volumes and freeze immediately in liquid nitrogen. A 10-L fermenter should yield approximately 100–150 mL of final S30 extract. The frozen extract can be stored at −80°C for several months. 3.2. CECF Reaction Setup

CF reactions are performed in analytical scales for optimization studies and in preparative scales for high-level production (see Note 17). For the reaction setup, it is necessary to develop first a pipetting scheme (e.g., in Excel). Table 11.1 gives an example for the stock concentrations and final volumes in a reaction of 70 mL using the E. coli system.

3.2.1. Reaction Container Setup

Choice of appropriate reaction containers is a critical issue that can affect yield and reproducibility of a reaction. A variety of home-made or commercially available containers can be considered for analytical- or preparative-scale CECF reactions (see Fig. 11.2). The ratio between RM and FM should be between 1:10 and 1:30 for the best compromise of yields and costs (see Note 18). D-Tube dialyzers are available in different sizes, starting from 10 mL up to several millilitres and can be used for small analytical-scale reactions as well as for preparative-scale reactions. DispoDialyzers are recommended for RM volumes of 0.5–5 mL. The commercial devices serve as the RM container and have to be placed into suitable containers (e.g., plastic tubes) that hold the desired FM volume. The general handling of the commercial devices should be done according to the manufacturer’s recommendations.

A nalytical-Scale CECF Reactions in Microcentrifuge Tubes (18)

1. Remove the upper tip from the bottom part of a standard 0.5-mL microcentrifuge tube (e.g., Eppendorf) (see Note 19). Cut off the hinge from the lid and use the inner part of the lid as the RM chamber and the tube for the FM. 2. Completely fill the lid with approximately 100 mL of RM. 3. Cover the lid with a disk of dialysis membrane approximately 8 mm in diameter and with the desired MWCO. Ensure that no air bubbles are trapped.

176

Schneider et al.

Table 11.1 Protocol for P-CF expression in CECF mini-reactors Stock Finala concentration concentration

Added volume (mL)

10

0.05

PEG 8000 (%)

40

KOAc (mM)

Component (unit)

Volumes allocated for RM

FM

5.4

0.4

5

2

53.5

3.5

50

4,000

150.8

40.3

2.6

37.7

Mg(OAc)2 (mM)b

1,000

10.1

10.8

0.7

10.1

HEPES (M)

2.4

0.1

44.6

2.9

41.7

Complete® (-fold)

50

1

21.4

1.4

20

Folinic acid (mg/mL)

10

0.1

10.7

0.7

10

DTT (mM)

500

2

4.3

0.3

4

NTP-mix (-fold)

75

1

14.3

0.9

13.4

PEP (mM)

1,000

20

21.4

1.4

20

AcP (mM)

1,000

20

21.4

1.4

20

AA-mix (mM)

4

0.5

133.8

8.8

125b

RCWMDE-mix (mM)

16.7

1

64.2

4.2

60

446.1

29.2

416.9

Master mix NaN3 (%)

b

Final volumes Feeding mix Master mix

–

416.9

AA-mix (mM)

4

1c

125

S30-C buffer (%)

100

35

350

Milli-Q water

Fill up to 1,000 mL

Final volume FM

1,000 mL

Reaction mix Master mix

-

29.2

Pyruvate kinase (mg/mL)

10

0.04

0.3

Plasmid DNA/PCR fragment (mg/mL)

0.2

0.015

5.3

RiboLock (U/mL)

40

0.3

0.5

T7RNAP (U/mL)

400

6

1.1 (continued)

Membrane Protein Expression in Cell-Free Systems

177

Table 11.1 (continued) Stock Finala concentration concentration

Added volume (mL)

tRNA Escherichia coli (mg/mL)

40

0.5

0.9

S30 extract (%)

100

35

24.5

Component (unit)

Volumes allocated for RM

Milli-Q water

Fill up to 70 mL

Final volume RM

70 mL

FM

Final concentrations of low molecular weight compounds are equal in the RM and FM, except AA-mix The protocol is given for a final concentration of K+ of 290 mM and Mg2+ of 15 mM, but both compounds are subject to optimization. Additional sources of K+ and Mg2+ ions are HEPES, PEP, AcP and the S30 extract c FM is supplemented with additional 0.5 mM of the AA-mix a

b

4. Fix the dialysis membrane by carefully pressing down the tube on the lid. Prevent sliding or damage of the dialysis membrane (see Note 20). 5. Fill the FM into the tube through the bottom hole and cover the hole and the lid rim with Parafilm to prevent evaporation. 6. Place the tube, lid facing down, in a 48-well microplate and incubate on a shaker at 30°C (see Note 21).  eactions in CECF R Mini-Reactors (see Note 22)

1. Prepare pieces of dialysis membranes (2.5 × 2.5 cm) with the appropriate MWCO. 2. Fix the membranes with the Teflon ring to the reactor chamber (see Note 23). 3. Fill 50–70 mL of RM from the top into the reaction chamber. 4. Fill a cavity of a 24-well plate with 800–1,000 mL FM to give a final ratio of 1:14. 5. Place the reaction container into the cavity with the FM. Prevent air bubbles between the dialysis membrane and the FM. 6. Seal the reaction setup with Parafilm to prevent evaporation and incubate on a 30°C shaker.

 eactions in CECF R Maxi-Reactors (see Note 22) 

1. Home-made CECF maxi-reactors are combined with commercially available Slide-A-Lyzers, MWCO of 10 kDa, holding the RM. Fill the Slide-A-Lyzer with 1–3 mL of RM (see Note 3).

178

Schneider et al.

2. Add the FM in a RM-to-FM ratio of 1:17 in the container (see Note 24). 3. Put the Slide-A-Lyzer into the reactor. Close the container either with Parafilm or with a lid and incubate it in a 30°C shaker. 3.2.2. E. coli CECF Reaction Setup

1. Thaw all stock solutions; enzymes should be thawed on ice. Mix every stock thoroughly except extract and enzymes (see Note 25). 2. Prepare a master mix of all components that have identical molarities in FM and RM (see Table 11.1). 3. Mix and remove the required RM volume from the master mix to a separate tube and store it on ice. 4. Complete the FM with onefold S30-C buffer without DTT, additional 0.5 mM amino acid mixture and water and preincubate at 30°C. 5. Complete the RM with the enzymes, the DNA template and the S30 extract on ice. Carefully mix all components (see Note 25). 6. Fill the RM and the FM in a ratio of 1:14 or 1:17 in the selected reaction container and incubate overnight at 30°C with shaking or rolling for 16–24 h. 7. For MP analysis, centrifuge the RM after the reaction for 15 min at 18,000g. Analyse supernatant and pellet fractions separately by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting if appropriate. Resuspend the pellet in water or buffer equivalent to the volume of the RM and analyse 1–3 mL of the sample by SDS-PAGE (see Fig. 11.4). 8. To optimise the yield of expressed MPs, it is recommended to perform an initial Mg2+ and K+ screen (see Note 17) (14).

3.2.3. CF Expression Modes for MP Production

Different CF expression modes for the synthesis of MPs are possible (see Fig. 11.1a). In the absence of any hydrophobic environment in the precipitate CF expression mode (P-CF) MPs will become insoluble and form a precipitate. These precipitates appear to contain at least partly structured protein as they often can be solubilized in relatively mild detergents without denaturation/renaturation procedures (1, 10). If detergents or amphipols are added into the RM, the MPs can be inserted into micelles directly after or during translation. This detergent-based CF expression mode (D-CF) therefore produces soluble proteomicelles (6, 11, 12, 23). Alternatively, the synthesized MPs could become directly inserted into provided lipid environments in the lipid-based CF expression mode (L-CF). The lipids could be added in the form of defined liposomes (13), as complex mem-

Membrane Protein Expression in Cell-Free Systems

179

Fig. 11.4. Eukaryotic MPs expressed in the P-CF mode. Proteins were expressed in CECF mini-reactors, the RM was centrifuged at 18,000g for 10 min at 4°C, and the pellet was resuspended in a buffer volume equivalent to the RM volume. The following aliquots of these suspensions were mixed with loading buffer and separated by SDS-PAGE on a 12% gel and stained with Coomassie blue: ETRB, NPY2R, NPY5R, MTNR1A, MTNR1B, SS1R, SS2R, V1BR, HRH1: 3 mL; V2R, CRF: 2.5 mL; PS1-CTF: 4 mL. G protein-coupled receptors: ETRB (human endothelin B receptor); NPY2R and NPY5R (human neuropeptide Y2 and Y5 receptors, respectively); MTNR1A and MTNR1B (human melatonin 1A and 1B receptors); SS1R and SS2R (human somatostatin 1 and 2 receptors); V1BR (human vasopressin 1B receptor); HRH1 (human histamine 1 receptor); V2R (human vasopressin 2 receptor); CRF (rat corticotropin-releasing factor receptor); PS1CTF: C-terminal fragment of human presenilin-1. The marker is given in kilodaltons; arrows indicate the expressed MPs.

brane vesicles (24), as giant liposomes out of egg yolk phosphatidylcholine (PC) and 1,2-dioleyl-sn-glycero-3-phosphocholine (DOPC) (25) or even as nanolipoprotein particles consisting of planar bilayers surrounded by a membrane scaffold protein (26). Mixtures of lipids and detergents provided as bicelles or mixed micelles might further facilitate the solubilization of expressed target MPs. Some residual E. coli lipids in amounts of approximately 100 mg/mL are furthermore still present in the S30 extract. The currently most established hydrophobic compounds that are suitable for the solubilization of MPs in the different CF expression modes are listed in Table 11.2. MP expression in the different L-CF modes is an emerging approach, and it is still too preliminary to recommend specific guidelines. Only details for the more established P-CF and D-CF modes are presented. Solubilization of P-CF-Expressed MPs

P-CF mode reactions are in particular recommended for initial expression screens of new MP targets to evaluate the best individual reaction conditions. Yields of MPs are often significantly higher in the P-CF mode if compared with corresponding D-CF mode expressions, thus facilitating the expression monitoring by simple SDS-PAGE analysis (see Fig. 11.4). The basic reaction protocols established by P-CF expression screens can exactly be transferred as basic conditions for the expression in the D-CF mode.

180

Schneider et al.

1. Expression of MPs in the P-CF mode is done as described in Section  11.3.2.2. The RM should become turbid by the synthesised and precipitated MPs. 2. Harvest the MPs by centrifugation of the RM at 18,000g for 10 min at 4°C. 3. To reduce co-precipitated E. coli proteins, discard the supernatant and resuspend the pellet in appropriate solubilization buffer without detergent (e.g., 20 mM Na2HPO4, pH 7.5, 1 mM DTT), then centrifuge again for 10 min at 18,000g at 4°C. Repeat this washing step three times. More persistent co-precipitated proteins could be removed by washing in the presence of some detergents that are not effective in the solubilization of the target MP (e.g., Brij derivatives). Elaborated washing procedures could already result in relatively pure MP samples. 4. Resuspend the pellet in solubilization buffer with the appropriate detergent (e.g., 2% [w/v] LMPG) by pipetting up and down (see Note 26). Incubate for at least 30 min at 30–37°C with shaking (see Note 27). 5. Centrifuge once more for 10 min at 18,000g to pellet residual insoluble protein. Be careful about the centrifugation temperature because some detergents, like SDS and LMPG, will precipitate at lower temperatures. Take the supernatant for further studies. 6. Once solubilized, the detergent could be exchanged against more favourable ones after immobilizing the MP on a Ni2+ column by virtue of a terminal poly(His)n-tag. D-CF Expression of MPs

1. Prepare stock solutions of the selected detergents in distilled water in a concentration of at least ten times higher than the working concentration required for D-CF expression according to Table 11.2 (see Note 28). First, a screen with a variety of selected detergents or detergent combinations should be performed in analytical-scale reactions to identify the most suitable compound for the solubilization of the desired target MP. 2. Prepare a master mix like for the P-CF mode expression as given in Section 11.3.2.2. and add suitable amounts of detergent to reach the final concentrations given in Table  11.2. Subtract the added detergent volume from the water volume. 3. Continue as given in Section 11.3.2.2. 4. Successfully D-CF-expressed MPs will preferentially stay in the supernatant after centrifugation of the RM for 10 min at 18,000g. For the effective soluble expression, the MP should be found mainly in the soluble fraction in the SDS-PAGE. Since there are also many E. coli proteins in the soluble

Membrane Protein Expression in Cell-Free Systems

181

Table 11.2 Additives suitable for the solubilization of CF-expressed MPsa Supplement (mode)

CMC (mM)

Final conc. (%)

X CMC

Reference

Triton X-100

0.23

0.1 (0.2)

6.7 (13.4)

(6, 11)

DDM

0.12

0.1

15.0

(6)

DM

1.8

0.2

2.3

(6)

Digitonin

0.73

0.4

4.5

(6, 11)

Brij-35 (C12/23)

0.08

0.1

10.4

(6, 11)

Brij-58 (C16/20)

0.08

1.5

178.1

(6)

Brij-78 (C18/20)

0.05

1.0

189.0

(6)

Brij-98 (C18-1/20)

0.03

0.2

69.6

(6)

diC8PC

0.22

0.1

8.9

(6)

C6F-TAC

0.3

0.32–2.88

6.7–60

(23)

C8F-TAC

0.03

0.34–3.06

67–600

(23)

HF-TAC

0.44

0.36–3.2

4.5–40

(23)

LMPG

0.05

1–2

420–840

(6)

LPPG

n.a.

1–2

n.a.

(6)

DPC

0.9–1.5

1–2

19–38

(6)

SDS

9

1–2

4.2–8.4

(6)

Detergents (D-CF)

Amphipols (D-CF)

Detergents (P-CF)

Lipids (L-CF) Escherichia coli lipids

0.4

(1, 24)

DMPC

0.4

(13)

DOPC

0.4

(13)

The table lists a selection of detergents and lipids that typically provide good results for MP production in the D-CF, P-CF or L-CF modes. The indicated concentrations are recommended for initial expression experiments and might be subject to optimization CMC critical micellar concentration; Triton X-100 polyethylene glycol P-1,1,3,3-tetra-methyl-butylphenyl ether; DDM n-dodecyl-b-d-maltoside; DM n-decyl-b-D-maltoside; Brij-35 polyoxyethylene-(23)-lauryl-ether; Brij-58 polyoxyethylene-(20)-cetyl-ether; Brij-78 polyoxyethylene-(20)-stearyl-ether; Brij-98 polyoxyethylene-(20)-oleyl-ether; C6FTAC C6F13C2H4-S-poly[tris(hydroxymethyl)aminomethane]; C8F-TAC C8F17C2H4-S-poly[tris-(hydroxylmethyl) aminomethane]; HF-TAC C2H5C6F12C2H4-S-poly[tris (hydroxyl-methyl)-aminomethane]; diC8PC 1,2-dioctanoyl-snglycero-3-phosphocholine; LMPG 1-myristoyl-2-hydroxy-sn-glycerol-3-[phosphor-rac-(1-glycerol)]; LPPG 1-palmitoyl2-hydroxy-sn-glycerol-3-[phosphor-rac-(1-glycerol)]; DPC dodecylphosphocholine; SDS sodium dodecyl sulphate; DMPC 1,2-dimyristoyl-sn-glycero-3-phosphocholine; DOPC 1,2-dioleyl-sn-glycero-3-phosphocholine a

182

Schneider et al.

fraction, it depends on the expression level whether the target MP can be detected. It is recommended to verify protein expression by Western blotting or to analyse samples after purification (e.g., by affinity chromatography) (see Note 29). 5. Identified optimal expression conditions can be scaled up to 1- to 3-mL preparative-scale reactions in the CECF maxireactors (see Note 18).

4. Notes 1. In our hands, E. coli K-12 strain A19 [rna19 gdh A2 his95 relA1 spoT1 metB1] (27) produces the most efficient extracts. Its ribonuclease I deficiency helps to stabilize mRNA during transcription/translation. However, other E. coli strains like BL21 (DE3) (Merck) and derivatives could also produce an effective extract (28, 29). The additional overexpression of rare codon tRNAs could be advantageous for the expression of heterologous MPs. 2. Prepare a 50-fold stock solution of the S30 buffers without DTT, b-mercaptoethanol or PMSF, which can be stored at 4°C up to several months. 3. The Slide-A-Lyzers can be reused several times for expression of the same protein until the membrane starts to become clogged. Each corner of the Slide-A-Lyzers has one gasket. Always use the same gasket for filling or removing the RM and place the Slide-A-Lyzer into the CECF maxi-reactor with the side containing the used gasket upward to prevent leakage of the RM. It is recommended to vent the Slide-A-Lyzer during the filling procedure with a needle inserted into the gasket at the other edge of the same side. 4. Prepare for tyrosine a 20 mM stock solution and for the other 19 amino acids 100 mM stock solutions. l-Tryptophan has to be dissolved in 100 mM HEPES, pH 8.0, l-aspartic acid, l-cysteine, l-glutamic acid and l-methionine have to be dissolved in 100 mM HEPES, pH 7.4. All other amino acids are dissolved in distilled water. Sonication or heating to 60°C could improve solubility. However, some stocks (in particular W, D, E, N, C, Y) will not dissolve completely and have to be handled as suspensions. For the amino acid mixture, take 2 mL of each of the 100 mM amino acid stocks and 10 mL of the 20 mM l-tyrosine stock and fill up with distilled water to 50 mL. This results in a 4 mM stock concentration of each amino acid. The standard final amino acid concentration in

Membrane Protein Expression in Cell-Free Systems

183

the CF reaction is 1 mM, while higher concentrations might increase the final protein yields in some cases. 5. The amino acids R, C, W, M, D and E are subjected to metabolic degradation during incubation. Increased amounts enhance MP production. 6. Only few frequently used detergents are listed. CF systems are highly tolerant of a wide variety of detergents. For more complete lists, please consult Table 11.2 or (6, 11, 12). 7. Some commercially available RNase inhibitors may contain energy consumptive impurities. Evaluation of the quality is therefore recommended when using new batches or suppliers. 8. It is also possible to use commercially available T7RNAP (e.g., Roche Diagnostics), but besides the high costs, the concentrations are usually too low, and additional concentration steps would be required. 9. Increased final tRNA concentrations could reduce premature termination products of targets with larger size, heterologous origin and codon usages very different from that of the E. coli expression background. 10. The volume of the RM is limited, and higher concentrations of the DNA template save space for further additives. Relatively high amounts of DNA template are required, and vectors with increased copy numbers like the pIVEX might be preferred, if possible. 11. Autoclave 2xYT medium and phosphate buffer separately to avoid precipitation. Sterilize glucose by filtration. Combine the solutions and add 1–2 mL antifoam before starting the fermentation. 12. Stir the medium vigorously for 5 min before adding the preculture to enrich it with oxygen. 13. Harvesting of the cells in the mid-logarithmic phase is important because rapidly growing cells contain more active ribosomes, and extracts are therefore more productive. Rapid cooling of the cells after fermentation will conserve them in this active state. 14. After this step, the pellet could be frozen in thin layers at −70°C and stored for few weeks. However, it is recommended to continue with the extract preparation without storage to achieve highest activities. 15. Cell disruption by sonication might reduce extract activity as a result of ribosome disintegration. 1 6. This procedure causes the dissociation of ribosomes from mRNA, which subsequently will become degraded by

184

Schneider et al.

endogenous RNases. Incubation time and temperature are important to ensure the optimal extract efficiency while excessive inactivation of essential proteins has to be prevented. 17. For every new MP target expression optimization screens in the analytical scale for key compounds like Mg2+ and K+ ions are highly recommended to achieve best results. With E. coli extracts and with the described protocols, Mg2+ concentrations between 13 and 17 mM and K+ concentrations between 270 and 310 mM are usually optimal. 18. The exchange area of the membrane becomes more and more limiting with increased RM volumes, resulting in increasingly lower yields. Dialysis membranes with a relatively wide range of pore sizes between 10 and 50 kDa can be used. The choice of the MWCO also depends on the size of the expressed MP, which preferentially should remain in the RM. Membranes consisting of regenerated cellulose should be preferred. 19. The inner part of the lid fits approximately 100 mL of RM, and the tube fits 0.5 mL of FM, resulting in a somehow ­sub-optimal RM: FM ratio of 1:5. Alternatively, the RM volume could be decreased to 50 mL by addition of few glass beads into the lid chamber. 20. Damage of the membrane, as often observed when using larger 1.5-mL microcentrifuge tubes, will cause leakage. 21. Alternatively, a small magnetic bar could be included in the RM chamber and the tube be incubated on a magnetic stirrer. 22. All reactors should be cleaned with distilled water after use, and they must be completely dry before use. 23. The reaction chamber is formed between the dialysis membrane and a 1-mm cavity at the bottom part of the CECF mini-reactor, which is formed by a corresponding rim as illustrated in Fig. 11.2a. Every time, use new dialysis membranes for the CECF mini-reactors. Make sure that the Teflon ring is correctly adjusted on the same level as the reaction chamber and that the dialysis membrane is completely planar without any crinkles at the bottom part, which serves as exchange area between RM and FM. First, place the Teflon ring on a piece of Parafilm, cover the ring with the membrane and then press the reactor over the ring until it stops. The dialysis membrane should be large enough that its ends extend over the Teflon ring to ensure that the device is tight and no leakage of the RM occurs. The Teflon ring should fit tightly to the reactor without membrane. Correct adjusting and absence of leakage could preliminary be monitored with, for example, dextran blue. 24. For the CECF maxi-reactors, it is important that the levels of RM and FM are identical when the Slide-A-Lyzer is inside

Membrane Protein Expression in Cell-Free Systems

185

the container. Adjust the levels by removing some air from the RM compartment or by addition of extra FM. 25. Too vigorous mixing might denature proteins or ribosomes in the extract. 26. Sometimes, the solution becomes clear during the pipetting process, in particular in the presence of SDS and LMPG. If the sample stays turbid, the target MP could still be present in the soluble fraction while other impurities might remain in the insoluble fraction. Choice of buffer and pH value are subject of optimization and can be crucial for the resolubilisation efficiencies. 27. The solubilization efficiency of different detergents strongly depends on each particular MP. Different types of detergents at different concentrations and at different temperatures have to be tested. LMPG, LPPG, SDS and DPC are usually efficient and among the first choices. 28. Detergent stocks should be as highly concentrated as possible. Higher concentrated stocks of long-chain Brij derivatives become soluble by incubation at 65°C, and they can be stored frozen at −20°C. The stocks might become turbid again after thawing or extended storage for hours in non-frozen conditions. It is recommended always to screen different detergent concentrations by expression in the D-CF mode as the detergent tolerance is also somehow target protein dependent. 29. If using a Ni2+ purification column, first dilute your sample with buffer to avoid Ni2+ reduction by the 2 mM DTT in the RM. Purification directly after expression is recommended. References 1. Klammt C, Löhr F, Schäfer B, Haase W, Dötsch V, Rüterjans H, Glaubitz C, Bernhard F (2004) High level cell-free expression and specific labeling of integral membrane proteins. Eur J Biochem 271:568–580 2. Liguori L, Marques B, Villegas-Mendez A, Rohte R, Lenormand JL (2007) Production of membrane proteins using cell-free expression systems. Expert Rev Proteomics 4:79–90 3. Klammt C, Schwarz D, Eifler N, Engel A, Piehler J, Haase W, Hahn S, Dötsch V, Bernhard F (2007) Cell-free production of G protein-coupled receptors for functional and structural studies. J Struct Biol 158:482–493 4. Wagner S, Bader ML, Drew D, de Gier JW (2006) Rationalizing membrane protein overexpression. Trends Biotechnol 24:364–371 5. Kim DM, Swartz JR (1999) Prolonging cell-free protein synthesis with a novel ATP regeneration system. Biotechnol Bioeng 66:180–188

6. Klammt C, Schwarz D, Fendler K, Haase W, Dötsch V, Bernhard F (2005) Evaluation of detergents for the soluble expression of a-helical and b-barrel-type integral membrane proteins by a preparative scale individual cell-free expression system. FEBS J 272:6024–6038 7. Ozawa K, Wu PS, Dixon NE, Otting G (2006) 15 N-labelled proteins by cell free protein synthesis: Strategies for high-throughput NMR for proteins and protein-ligand complexes. FEBS J 273:4154–4159 8. Reckel S, Sobhanifar S, Schneider B, Junge F, Schwarz D, Durst F, Löhr F, Güntert P, Bernhard F, Dötsch V (2008) Transmembrane segment enhanced labeling as a tool for the backbone assignment of alpha-helical membrane proteins. Proc Natl Acad Sci U S A 105: 8262–8267 9. Trbovic N, Klammt C, Koglin A, Löhr F, Bernhard F, Dötsch V (2005) Efficient

186

Schneider et al.

strategy for the rapid backbone assignment of membrane proteins. Am Chem Soc 127: 13504–13505 10. Keller T, Schwarz D, Bernhard F, Dötsch V, Hunte C, Gorboulev V, Koepsell H (2008) Cell-free expression and functional reconstitution of eukaryotic drug transporters. Biochemistry 47:4552–4564 11. Ishihara G, Goto M, Saeki M, Ito K, Hori T, Kigawa T, Shirouzu M, Yokoyama S (2005) Expression of G protein coupled receptors in a cell-free translational system using detergents and thioredoxin-fusion vectors. Protein Expr Purif 41:27–37 12. Nozawa A, Nanamiya H, Miyata T, Linka N, Endo Y, Weber APM, Tozawa Y (2007) A cell-free translation and proteoliposome reconstitution system for functional analysis of plant solute transporters. Plant Cell Phys 48: 1815–1820 13. Kalmbach R, Chizhov I, Schumacher MC, Friedrich T, Bamberg E, Engelhard M (2007) Functional cell-free synthesis of a seven helix membrane protein: in situ insertion of bacteriorhodopsin into liposomes. J Mol Biol 371: 639–648 14. Schwarz D, Junge F, Durst F, Frölich N, Schneider B, Reckel S, Sobhanifar S, Dötsch V, Bernhard F (2007) Preparative scale expression of membrane proteins in E. coli based continuous exchange cell-free systems. Nat Protocols 2:2945–2957 15. Goerke AR, Swartz JR (2007) Development of cell-free protein synthesis platforms for disulfide bonded proteins. Biotechnol Bioeng 99:351–367 16. Arduengo M, Schenborn E, Hurst R (2007) The role of cell-free rabbit reticulocyte expression systems in functional proteomics. In: Kudlicki W, Katzen F, Bennett R (eds) Cell-free expression. Landes Bioscience, Austin, TX, pp 1–18 17. Spirin AS, Baranov VI, Ryabova LA, Ovodov SY, Alakhov YB (1988) A continuous cellfree translation system capable of producing polypeptides in high yield. Science 242: 1162–1164 18. Shirokov VA, Kommer A, Kolb VA, Spirin AS (2007) Continuous-exchange protein-synthesizing systems. In: Grandi G (ed) Methods in molecular biology, vol 375: In vitro transcription

19.

20.

21.

22. 23.

24.

25.

26.

27. 28.

29.

and translation protocols, 2nd edn. Humana Press, Totowa, NJ, pp 19–55 Kopeina GS, Afonina ZA, Gromova KV, Shirokov VA, Vasiliev VD, Spirin AS (2008) Step-wise formation of eukaryotic doublerow polyribosomes and circular translation of polysomal mRNA. Nucleic Acids Res 36: 2476–2488 Sawasaki T, Ogasawara T, Morishita R, Endo Y (2002) A cell-free protein synthesis system for high-throughput proteomics. Proc Natl Acad Sci U S A 99:14652–14657 Davanloo P, Rosenberg AH, Dunn JJ, Studier FW (1984) Cloning and expression of the gene for bacteriophage T7 RNA Polymerase. Proc Natl Acad Sci U S A 81:2035–2039 Zubay G (1973) In vitro synthesis of protein in microbial systems. Annu Rev Genet 7:267–287 Park KH, Berrier C, Lebaupain F, Pucci B, Popot JL, Ghazi A, Zito F (2007) Fluorinated and hemifluorinated surfactants as alternatives to detergents for membrane protein cell-free synthesis. Biochem J 403:183–187 Wuu JJ, Swartz JR (2008) High yield cell-free production of integral membrane proteins without refolding or detergents. Biochim Biophys Acta 1778:1237–1250 Nomura S, Kondoh S, Asayama W, Asada A, Nishikawa S, Akiyoshi K (2007) Direct preparation of giant proteo-liposomes by in-vitro membrane protein synthesis. J Biotechnol 133:190–195 Katzen F, Fletcher JE, Yang JP, Kang D, Peterson TC, Cappuccio JA, Blanchette CD, Sulchek T, Chromy BA, Hoeprich PD, Coleman MA, Kudlicki W (2008) Insertion of membrane proteins into discoidal membranes using a cell-free protein expression approach. J Proteome Res . doi:10.1021/pr800265f Gesteland RF (1966) Isolation and characterization of ribonuclease 1 mutants of Escherichia coli. J Mol Biol 16:67–84 Kang SH, Kim DM, Kim HJ, Jun SY, Lee KY, Kim HJ (2005) Cell-free production of aggregation-prone proteins in soluble and active forms. Biotechnol Prog 21:1412–1419 Kim TW, Keum JW, Oh IS, Choi CY, Park CG, Kim DM (2006) Simple procedure for the construction of a robust and cost-effective cell-free protein synthesis system. J Biotechnol 126:554–561

Chapter 12 Practical Considerations of Membrane Protein Instability during Purification and Crystallisation Christopher G. Tate Abstract Crystallisation of integral membranes requires milligrams of purified protein in a homogeneous, monodisperse state, and crucially, the membrane protein must also be fully functional and stable. The stability of membrane proteins in solution is dependent on the type of detergents used, but unfortunately the use of the most stabilising detergent can often decrease the probability of obtaining crystals that diffract to high resolution, especially of small membrane proteins. A number of strategies have been developed to facilitate the purification of membrane proteins in a functional form, which have led to new possibilities for structure determination. Key words: Detergents, G protein-coupled receptors, thermostability

1. Introduction A necessary prerequisite for the purification of any integral membrane protein is the dissolution of the biological membrane in which the protein resides so that the membrane protein exists in solution in a monodisperse state. Membrane protein solubilisation is usually accomplished using a variety of small amphipathic molecules called detergents. Detergents play a critical role in membrane protein purification because they attempt to mimic the amphipathic environment of a lipid bilayer whilst maintaining the structure of the membrane protein in a physiologically relevant state. The success or failure of a purification strategy, and whether crystals can eventually be grown, is largely dependent on the type of detergent that is used, both for the initial solubilisation and the subsequent steps of column chromatography. This chapter first describes some of the properties of detergents and then how to choose the best detergent for a particular membrane protein

I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_12, © Humana Press, a part of Springer Science + Business Media, LLC 2010

187

188

Tate

based on assays for membrane protein stability. A variety of strategies that have stabilised detergent-solubilised membrane proteins is also discussed.

2. Types of Detergents Detergents are a structurally diverse group of molecules that share a common amphipathic character (1–3) along with the ability to exist in solution either as a monomer or, as their concentration increases, as micelles (Fig. 12.1). The concentration of detergent where micelles start to form is called the critical micellar concentration (CMC), which varies depending on the type of detergent and the temperature, and for charged detergents, it can also vary significantly with the ionic strength or pH of the solution. It is essential to know the CMC of a detergent being used to purify a membrane protein because the membrane protein will only stay in solution if the detergent concentration remains at or above the CMC. The structure of detergents varies enormously (Fig. 12.2). In addition, the ability of a detergent to inactivate a membrane protein varies widely, but there are a few guidelines that help to classify the detergents into mild or harsh depending on their structure. In this context, harsh means that there is a high probability of a membrane protein being denatured by the detergent, and mild

Fig. 12.1. Behaviour of detergent molecules in solution. At low concentrations, detergents are monomeric (blue line), but above the CMC (dashed line) micelles are formed. Membrane protein solubilisation and purification require concentrations above the CMC (green area).

Practical Considerations of Membrane Protein Instability during Purification

189

Fig. 12.2. Structures of some commonly used mild and harsh detergents. Digitonin is a mild detergent but is unsuitable for crystallography. All the other detergents have the same size hydrophobic portions but differ in their mildness due to the different head groups (red ). Both DDM and C12E8 are mild detergents, but as the head group size decreases and becomes charged, the detergent becomes harsher. Only a few membrane proteins are stable in SDS, but the structure of many membrane proteins is preserved on purification in DDM.

means that there is a good chance that a physiologically relevant structure is maintained. Precisely how it can be determined whether a membrane protein in detergent solution is ‘active’ or ‘inactive’ is described in another section. When considering whether a detergent is mild or harsh, two factors are considered. Firstly, as the head group of a detergent is varied from large, small, zwitterionic or singly charged, the harshness of the detergent increases, providing that the hydrophobic tail structure remains constant. Secondly, as the hydrophobic tail decreases in size from polyaromatic, long alkyl chain or short alkyl chain, then the harshness of the detergent also increases. Thus, when confronted by a choice of over 50 different detergents, it is possible to make an estimate of their harshness by using this rule of thumb (Fig. 12.2). However, this rule of thumb only considers the bulk property of the detergent and how the micelle can effectively shield the hydrophobic portion of the membrane protein. In addition, there is often a specific ‘chemical effect’ between the detergent and how it interacts with a membrane protein; for example, the thermostabilised b1-adrenergic receptor is remarkably stable in decylmaltoside (DM) (4), but it is extremely

190

Tate

unstable in even mild polyoxyethylene detergents like C12E8, even though the rule of thumb would suggest that C12E8 should be as mild, or even milder, than DM. In contrast, the sarcoplasmic reticulum Ca2+-ATPase (adenosine triphosphatase) strongly prefers being purified in C12E8, and all the crystals of this protein have been grown in this detergent (5). In fact, it is often clear from highresolution structures why a particular detergent is favoured because sometimes they are highly ordered in a specific location on the hydrophobic surface of a membrane protein in a similar fashion to how some lipids can also be ordered (6). Probably the most useful series of detergents are the alkyl maltosides and alkyl glucosides. These are all commercially available in highly purified forms and have been used extensively for the purification and crystallisation of a wide variety of integral membrane proteins (7). Dodecylmaltoside (DDM; also called laurylmaltoside, LM) is considered to be one of the best detergents because it is mild, it solubilises membranes efficiently, and many membrane proteins have been crystallised using it. The increasing mildness of the detergent series dodecyl-, tetradecyland hexadecyl- maltosides is apparent when solubilising rhodopsin as the longer-chain detergents are better at preserving the oligomeric state of rhodopsin on solubilisation (8). Digitonin is milder than DDM and can preserve weak interactions between integral membrane proteins, making it ideal for studying protein– protein interactions in  vivo using co-immunoprecipitation (9). However, it is also a poor solubilising agent and often only solubilises about 50% of the membrane proteins. Even milder detergents, such as Brij-35 or Brij-70 are proving extremely useful for the production of functional membrane proteins in in vitro transcription–translation systems (10) but are of little utility for membrane protein solubilisation and purification. Although DDM has proven to be an excellent detergent for purifying bacterial membrane proteins in a functional form, it is not mild enough to maintain the activity of some mammalian membrane proteins, which therefore have to be stabilised by the addition of additives (see below).

3. Choice of Detergents for Structural Studies

From this discussion, it might be considered that all you have to do is purify the membrane protein in the mildest possible detergent (e.g., digitonin), crystallise it, and solve the structure. Unfortunately, this is not the case, and no structures have ever been solved using digitonin. What also has to be taken into consideration in the choice of detergent for structural studies is the size of the detergent micelle. If it is too large in relation to

Practical Considerations of Membrane Protein Instability during Purification

191

the size of the protein, then it may inhibit crystallisation. Most membrane protein crystals grown by vapour diffusion methodology, as either sitting drops or hanging drops, have crystal contacts formed through the interaction of the hydrophilic domains of the protein. If the membrane protein has large hydrophilic domains, then the size of the detergent micelle is less critical than if the membrane protein is composed almost entirely of transmembrane domains and the detergent occludes the hydrophilic faces (Fig. 12.3). For example, the multidrug transporter EmrE contains 110 amino acid residues that form four transmembrane domains; the EmrE dimer (30 kDa) when purified in DDM is surrounded by 107 kDa of specifically bound detergent (11), which is equivalent to 1 mg of EmrE binding 3.5 mg of detergent (molar ratio of 1:110). Crystals were only grown in nonylglucoside, which has a considerably smaller micelle (12). Thus, the choice of detergent for crystallisation is usually the smallest detergent in which the membrane protein is sufficiently stable. If other biophysical techniques are to be used, then the size of the micelle may not matter, and a milder detergent can be used. One potential way to minimise the loss of activity of membrane proteins during purification in a harsh detergent is to solubilise and purify them in a mild detergent and, on the final purification step, exchange them into the harsher detergent. It is usually best to perform this exchange with the protein bound to a column matrix as this maximises the removal of the mild detergent from the membrane protein. In contrast, dialysis can be considerably slower, and if the membrane protein is on the cusp of denaturation, this could prevent obtaining crystals. If two detergents are used during purification, the choice of both detergents could be crucial for the success of crystallisation. Firstly, some mild detergents, such as DDM, are extremely difficult to displace

Fig. 12.3. Size of the detergent micelle surrounding a membrane protein. The structure of the bovine adenosine 5'-triphosphate/adenosine 5¢-diphosphate (ATP/ADP) translocase (AAC1) is shown as a surface representation, with hydrophobic amino acid residues in orange, and the detergents are shown as space-filling models. The size of the detergent–AAC1 complex determined by SEC is depicted by the coloured elipse (57); 8M, octylmaltoside (red); 10M, decylmaltoside (yellow); 13M, tridecylmaltoside (purple). As the detergent becomes smaller, the more hydrophilic area on AAC1 is exposed to solvent and will therefore be available for making crystal contacts during crystallisation. (Reproduced from (58). With permission).

192

Tate

with a second detergent, so it may be far better starting with DM, which can be efficiently replaced (12, 13). Secondly, sometimes crystals require one detergent to make specific interactions and the second detergent to allow crystal contacts to form. For example, the best crystals of rhodopsin were obtained if the purification was in LDAO (lauryldimethylamine-oxide) followed by an exchange into C8E4 immediately before crystallisation (14); an ordered molecule of LDAO was seen between crystallographic dimers, but the C8E4 allowed better crystals to form than if just LDAO was used (15). Another potential affect of using different detergents during purification is perhaps their differing abilities in removing lipids from membrane proteins. Although systematic studies have not been performed, it is well known that removal of lipids during membrane protein purification is a primary cause for destabilisation and the prevention of crystal formation (16–18).

4. Assessing the Stability of Membrane Proteins in Detergent Solution

There is a wide variety of methods for determining the stability of a membrane protein in detergent solution, and they can be conveniently divided into those that can be performed on unpurified samples or those that require the protein to be pure. The first category is probably the most useful at the start of a project when it has to be decided whether it will even be possible to purify a particular protein in a functional form or whether a homologue should be used instead. Probably the most useful of these techniques is the use of green fluorescent protein (GFP) fusions coupled to size exclusion chromatography (SEC). The principle of this technique is that if a membrane protein is solubilised in a detergent that is denaturing, then the protein will invariably aggregate, which will be observed by SEC as a peak at the void volume and an absence of a peak where the protein elutes in mild detergents (19). Even detergents that slightly perturb the structure can be identified through the broadening of the SEC peak or when it becomes distinctly asymmetric. The advantage of using GFP coupled to either the N- or C-terminus of a membrane protein is that the experiment can be performed directly on detergent-solubilised membranes (20). As the fluorescence of the GFP tag can be directly monitored, so the behaviour of the GFP fusion protein on SEC can be observed in an unpurified sample. In addition, the tag facilitates the improvement of expression levels as well as the optimisation of downstream purification steps (21–23). A more specific approach to test the stability of a membrane protein is to use radioligands that specifically bind to the membrane protein. It is normally assumed that if a substrate or inhibitor binds to a solubilised membrane protein with the same affinity

Practical Considerations of Membrane Protein Instability during Purification

193

as for the native protein in the membrane, then it is likely that the protein conformation is unperturbed by the detergent. To determine the thermostability of a membrane protein, such as a receptor or transporter, in a given detergent, membranes are solubilised and then heated for a given time over a range of temperatures (4, 24) or at a fixed temperature for a range of times. After quenching on ice, the amount of receptor remaining is determined by performing a binding assay using the radioligand (Fig. 12.4). The temperature at which half of the receptor remains is the apparent Tm, or a half-life can be determined from the rate of decrease in activity at a single temperature. This method can be used to optimise the stability of a particular membrane protein with different detergents, buffers and salt concentrations, or it can be used to compare the stabilities of different membrane proteins, provided that the same conditions are used for the heating step. Using this methodology, it has been shown that binding ligands can often improve the stability of a detergent-solubilised receptor (4, 24). The advantages of using a ligand-binding assay are that it can be performed on unpurified samples containing less than a picomole of receptor, it is reproducible and easy to perform, and it is an excellent way to define the amount of active versus inactive membrane protein being purified. When working with a protein containing prosthetic groups or ligands that absorb in the visible spectrum, a simple spectrum will give the same information regarding the conformation of the detergent-solubilised protein as a radioligand-binding assay. For example, rhodopsin has a defined spectrum due to covalently bound retinal, which can be used during the purification of the protein to determine how much functional rhodopsin is present or during thermostability assays (25, 26). Once a membrane protein has been purified, it is possible to determine the thermostability using a variety of other methods, such as, for example, blue native polyacrylamide gel electrophoresis (PAGE) and fluorescence (27), sodium dodecyl sulphate (SDS) PAGE and SEC (28), absorbance (29, 30), circular dichroism (CD) (27, 31) and differential scanning calorimetry (DSC) (32, 33). In CD and DSC, the temperature of the sample is gradually increased, and the measurements of the structural content of the membrane protein are recorded. The maximal change in structure will occur around the transition temperature. One of the problems associated with these biophysical approaches is that they require reasonable quantities of purified protein and have to be performed in series if multiple samples are to be analysed. A fluorescent method has been devised that can be performed in parallel on 96 samples in a microtitre plate that requires on the order of about 30–50 mg of purified membrane protein per condition (34). The fluorescence assay uses the thiol-specific fluorochrome CPM (N-(4-(7-diethylamino-4-methyl-3-coumarinyl)

194

Tate

Fig. 12.4. Stability of membrane proteins in detergent. (a) The stability of purified EmrE was determined in decylmaltoside (blue squares) or octylglucoside (red triangles) in the absence of the substrate tetraphenylphosphonium (TPP+). At pH 8.0, the substrate has an extremely slow off rate, so the stability was tested in the presence of TPP+ after purification in DM (purple circles) or OG (black diamonds). Samples were left at 4°C (blue background) and transferred to 18°C (red background) after 6 days. The calculated half-life for EmrE in OG in the absence of ligand was 8.2 days at 4°C and 2.8 days at 18°C (C.G.T., unpublished data). (b) Stability of the thermostabilised b1-adrenergic receptor mutant b1AR-m23 in different detergents. Samples were solubilised in DDM, bound to Ni2+-NTA beads, washed in the new detergent, and eluted with imidazole. Stabilities were determined by heating the receptors at the relevant temperature for 30 min, quenching on ice, and the amount of receptor remaining determined by a radioligand-binding assay. The apparent Tms, the point at which half of the receptor remained, are as follows: DDM, 52°C (pale blue squares); DM, 48°C (dark blue triangles); NG, 25°C (green circles); LDAO, 23°C (brown diamonds); OG, 17°C (red inverted triangles). (Reproduced from (4)).

Practical Considerations of Membrane Protein Instability during Purification

195

phenyl) maleimide), which fluoresces only after it reacts with a Cys residue. In membrane proteins, most Cys residues are buried inside the protein, so there is only a low background of fluorescence; as the sample is heated, the protein gradually denatures, exposing the buried Cys residues, which then react with CPM, resulting in an increase in fluorescence. This assay is ideal for testing multiple conditions and ligands to find the best combination for crystallisation trials.

5. Purifying Unstable Membrane Proteins

Once an ideal detergent for solubilisation has been found, it is often the case that the membrane protein can then be purified with little difficulty. However, some membrane proteins are rapidly inactivated during purification, and some can be purified but then aggregate badly during concentration. Both of these phenomena indicate that a factor, probably lipidic in nature, has been removed from the solubilised protein during purification. For example, purified lactose permease, LacY, usually aggregates badly during concentration, but the addition of small amounts of lipid during the purification prevented this and allowed the protein to be crystallised (17). In a more extreme case, the DDM-solubilised adenosine A2a receptor is rapidly inactivated when it is bound to an Ni2+-NTA (nitrilotriacetic acid) resin via its C-terminal His10 tag when washed with buffers containing only DDM; this is despite the receptor being stable in the unpurified DDM-solubilisate. However, inclusion of small amounts of cholesterol hemisuccinate (CHS) to the DDM allowed the receptor to be purified in a fully functional form (35). It appears that CHS dramatically stabilises the protein, probably through binding to a specific site found at the periphery of the receptor (36). Small amphiphiles can also stabilise membrane proteins in a similar manner to the stabilisation of soluble proteins. For example, glycerol (10–40%) has marked stabilisation of detergent-solubilised receptors (24, 37). There are also continuing efforts to develop novel detergents and detergent mimetics (38). The current successes in membrane protein structure determination are largely dependent on the development of detergents 20–30 years ago, and it is likely that there is always room for further improvements. Probably the best characterised of the detergent mimetics are amphipols, which have been shown to substantially stabilise membrane proteins when the surrounding detergent molecules are replaced with amphipols or when the protein (e.g., bacteriorhodopsin) is refolded directly into amphipols (39, 40). Peptidebased detergents have also been shown to thermally stabilise bacteriorhodopsin (41), rhodopsin (42) and glycerol-3-phosphate

196

Tate

dehydrogenase (43). However, although all these detergent mimetics undoubtedly stabilise membrane proteins and are useful in many biophysical studies, they have not yet provided a more efficient route toward the determination of membrane protein structures.

6. Why Are Integral Membrane Proteins Unstable in Detergent?

The difficulty encountered in purifying membrane proteins in a functional conformation arises because membrane proteins have not evolved to be stable in detergent, but they have evolved to be sufficiently stable in the membrane to confer a selective advantage for the survival of the organism in which they are found (38). Early successes in determining the structures of bacterial reaction centres and bacterial photosynthetic light-harvesting complexes were in part not only due to their abundance and ease of purification, but also because these proteins had evolved to be stable to provide precise alignment of their prosthetic groups for optimal energy transfer. Therefore, these proteins are stable in many detergents; in addition, perturbation of their structure was easily observed through a change in colour of the preparation. Similarly, members of the bacteriorhodopsin family are also robust proteins whose structures have been determined to high resolution. Successes in crystallising integral membrane proteins have been largely driven by identifying those membrane proteins that are sufficiently stable for their structures to remain unperturbed by purification and crystallisation in short-chain detergents. On the whole, bacterial membrane proteins have proven to be more robust than their eukaryotic counterparts. This may be related to the harsh environments in which some of these bacteria grow (e.g. thermophilic and halophilic bacteria), or it may be related to a strong selective pressure for the rapid synthesis of the membrane protein, for which there could be a relationship between stability and ease of folding; clearly, there is little value to a bacterium, which is dividing every 20 min and in competition for nutrients with other bacteria, to have nutrient uptake systems that take hours to synthesise. In contrast, in mammalian cells, each individual cell does not have to compete for resources and takes roughly 24 h to divide, so the fact that membrane proteins can take hours to reach the cell surface is irrelevant. What is relevant is that sufficient molecules are at the plasma membrane to confer a selective phenotype on the whole organism. Another possible factor for the poor stability of many eukaryotic membrane proteins is that their activity is often regulated by posttranslational modification, such as phosphorylation, and this, by necessity, requires the protein to be flexible and dynamic.

Practical Considerations of Membrane Protein Instability during Purification

7. Stabilising Membrane Proteins by Genetic Engineering

197

From the above discussion, it is clear that many membrane ­proteins have not evolved to be stable in the membrane and certainly not stable enough in short-chain detergents for crystallisation and structure determination. This of course suggests that evolutionary methods with a suitable selective pressure can be used to evolve membrane proteins into more stable forms ideal for crystallography. The intentional evolution in stability of a membrane protein was first shown for Escherichia coli diacylglycerol kinase (DGK), which contains three transmembrane domains and forms a trimer in the membrane. A limited Cys scan of one helix in DGK identified a few thermostable mutants (31), so a more extensive mutational analysis was subsequently performed by random mutagenesis (44). The most stable mutants were identified by colorimetric assays for enzyme function; these assays were performed after heating the solubilised DGK mutants for a defined time. Twelve thermostabilising mutations were identified, four of which were combined to produce an optimally stable enzyme with a half-life at 80°C of 15 min, compared to a half-life of only 6 min at 55°C for the native protein (44). This remarkable improvement in thermostability was probably due mainly to the improved stability of the trimer, as the thermostabilised mutant was far less sensitive to disruption with SDS than the native trimer. Despite the improvement in thermostability, the structure of the mutant has yet to be solved by X-ray crystallography, although the mutant has proven extremely robust in nuclear magnetic resonance (NMR) experiments (45). The work from Bowie’s group on both DGK (31, 44) and on helix B in bacteriorhodopsin (29) clearly demonstrated that point mutations in a membrane protein can have a dramatic effect on the thermostability of a membrane protein. The utility of this methodology for the stabilisation of monomeric eukaryotic membrane proteins and for crystallisation was subsequently shown for G protein-coupled receptors (GPCRs). Extensive efforts during the 1990s and the first half of the following decade to crystallise heterologously expressed GPCRs did not result in any crystals of sufficient quality to solve a structure. We thought that the lack of crystals was largely attributed to the poor stability of the GPCRs in detergent solution, so three receptors in our group were chosen for stabilisation (4, 24, 46): the turkey b1-adrenergic receptor (b1AR), the adenosine A2a receptor (A2aR) and the neurotensin receptor (NTR). Each of the receptors was stabilised by performing alanine-scanning mutagenesis followed by the identification of thermostabilising mutations using a ligand-binding assay performed on the detergent-solubilised receptor after a heating step. Alanine was chosen for the scanning mutagenesis because Ala is

198

Tate

tolerated in both hydrophilic and hydrophobic portions of the structure and because it has the greatest helix-forming propensity of any amino acid (47), which could result in the stabilisation of some transmembrane a-helices. Individual thermostabilising mutations were combined to produce optimally stable receptors. In the case of b1AR (4), the mutant b1AR-m23 contained six point mutations and was 21°C more thermostable than the native protein when solubilised in alkylmaltoside detergents; this is equivalent to about a 350-fold improvement in stability. There were two other dramatic effects of the thermostabilisation procedure on the properties of b1AR-m23. Firstly, the receptor was considerably more stable in short-chained detergents like nonylglucoside and octylglucoside than the native receptor (Fig. 12.4); in fact, the thermostability of the native receptor in these detergents could not be determined because the protein irreversibly aggregated as soon as these detergents were used. Secondly, b1AR-m23 was preferentially in the antagonist-bound conformation even in the absence of an antagonist. There is an equilibrium between the G protein-coupling states of the receptor, that is, between either the inactive antagonist-bound state and the active agonist-bound state, which occurs even in the unliganded receptor. This will inevitably result in conformational heterogeneity in the purified receptor during crystallisation and is therefore undesirable. In addition, it is known that the active conformation of GPCRs is less stable than the inactive state (48), so anything that prevents the formation of the active state by altering the equilibrium (e.g., either point mutations or the binding of an inverse agonist) will also improve the stability of the receptor. b1AR-m23 was subsequently crystallised in octylthioglucoside and the structure determined at 2.7-Å resolution with cyanopindolol in the ligand-binding site (49). Stabilisation of the A2aR was achieved in two different conformations (24). If the screen was performed using an antagonist, the thermostabilised mutant A2aR-rant21 was preferentially in the antagonist state, whereas if the selection was performed using an agonist, the thermostabilised mutant A2aR-rag23 was preferentially in the agonist-binding state. We have called this strategy conformational thermostabilisation in recognition of the simultaneous thermostabilisation coupled to selection for a single conformation of the receptor. A2aR-rant21 contained four point mutations, and A2aR-rag23 also contained four; the mutants showed an improvement in apparent Tm of 17°C and 9°C, respectively. NTR has also been stabilised in an agonist-binding conformation by four point mutations, but in this instance the thermostable mutant does not couple to G proteins, suggesting that the mutant receptor is preferentially in a state similar to the inactive conformation despite binding an agonist (46). Thus, by exerting different selective pressures, it will be possible to trap

Practical Considerations of Membrane Protein Instability during Purification

199

specific states of each receptor, which should eventually allow the structure determination of a whole range of conformations of the same receptor. A different evolutionary strategy has also been employed to stabilise NTR (50). Here, an assumption was made that receptor mutants that exhibit higher levels of expression in the cell are also more likely to be thermostable than the native receptor. A strategy was therefore devised based on iterative cycles of mutagenesis by error-prone polymerase chain reaction (PCR) and multiple rounds fluorescence activated cell sorting (FACS), where the cells expressing the highest levels of NTR were selected using a fluorescent NTR analogue. The final evolutionary product was mutant D03, which contained nine point mutations and five silent mutations and was approximately three to four times more stable than the native receptor, although expression levels had increased by tenfold. The relatively modest improvement in thermostability, despite the presence of nine point mutations, is probably because improved expression and thermostability are only weakly correlated (4, 24, 46). Many of the most thermostabilising mutations from our work do not improve expression levels, and some actually decrease expression. In addition, the selection based on increased expression levels is necessarily performed with the receptor in the membrane, whereas the conformational thermostabilisation strategy evolves the detergent-solubilised receptor.

8. Engineering Membrane Proteins for Improved Crystallisability

Although this chapter focuses on membrane protein stability, improving the ability of membrane proteins to crystallise is extremely pertinent, particularly if it allows the use of detergents and additives that normally preclude crystallisation because they form large micelles. It is usual practice in crystallisation to remove any portions of the protein that may be flexible because this aids the formation of crystals; with membrane proteins, this could mean deletion of parts of the N- or C-termini and parts of large flexible loops (49, 51, 52). However, from the previous discussion, it is clear that decreasing the size of hydrophilic areas is more likely to hinder rather than facilitate crystallisation because smaller, more denaturing, detergents will be needed to prevent occlusion of the hydrophilic surfaces by the detergent micelle. This was circumvented in the crystallisation of the b1 receptor by making it more thermostable, so despite the removal of large portions of hydrophilic sequences, crystals were grown because the receptor tolerated very-short-chain detergents (4, 49). An alternative strategy to the crystallisation of the b1 receptor is to increase the hydrophilic surface area of the membrane protein

200

Tate

Fig. 12.5. Comparison in size of crystallised b receptors and associated proteins. Structures of the b2AR are shown, with the co-crystallised Fab or T4 lysozyme fusion shown in grey; b1AR-m23 was crystallised using short-chain detergents. Receptors are shown in rainbow colouration with the N-terminus in blue and C-terminus in red. Models were made in Pymol from co-ordinates with PDB accession numbers 2VT4 (b1AR), 2R4R (b2AR:Fab) and 2RH1 (b2AR:T4).

by binding antibody fragments to the membrane protein surface or by making fusion proteins; in both cases, the increase in size of the hydrophilic domains will allow the use of larger, milder, detergents for crystallisation (Fig. 12.5). The b2AR is a good example for which the receptor is not particularly stable, and both the successful strategies to crystallise and solve its structure relied on increasing the hydrophilic area. The first structure of b2AR was solved by co-crystallising the receptor bound to an Fab fragment (53), which allowed the receptor to be purified in bicelles that are large, lipid-rich, micelles known to stabilise membrane proteins (54). The second structure was solved of a b2AR:T4 lysozyme sandwich fusion (55, 56), where the T4 lysozyme was inserted into the third cytoplasmic loop (CL3) of the receptor. The CL3 loop, between transmembrane helices 5 and 6, is extremely variable in size in GPCRs, it is not directly involved in ligand binding, and it is thought to be flexible because it makes essential contacts with the G protein on receptor activation with an agonist. The b2AR:T4 fusion was more resistant to proteases, but the thermostability has not yet been compared to the native receptor. From the structure of b2AR:T4, it is apparent that the T4 lysozyme makes virtually all the crystal contacts. This suggests that the fusion methodology could be helpful in crystallising other GPCRs, which has been substantiated by the structure determination of the A2aR:T4 fusion (36), although in this instance the space group and crystal contacts are different from the b2AR:T4 crystal. In the case of A2aR, inclusion of CHS during purification is essential to maintain the stability of the receptor (35), but this also has the

Practical Considerations of Membrane Protein Instability during Purification

201

undesirable effect of increasing the detergent micelle and preventing crystallisation using vapour phase diffusion. It is thus likely that the lipid mesophase acts as a reservoir that can remove excess CHS and detergent, allowing crystallisation to proceed driven by the T4 lysozyme interactions.

9. Conclusions There have been many developments in technology that, when combined, place us on the cusp of rapid expansion in the number of membrane protein structures determined. Compared with 15 years ago, we have a large resource of chemically distinct, highly purified, detergents that are essential for the purification of membrane proteins. There is also far more understanding of how detergents can affect the stability of membrane proteins and how they also affect crystallisation. In addition, there are now strategies to genetically engineer improved stability in membrane proteins and to improve their ability to form well-ordered crystals. The challenge for researchers is to choose the optimal strategy for their particular membrane protein, to go from overexpressed protein to crystals. References 1. le Maire M, Champeil P, Moller JV (2000) Interaction of membrane proteins and lipids with solubilizing detergents. Biochim Biophys Acta 1508:86–111 2. Helenius A, Simons K (1975) Solubilization of membranes by detergents. Biochim Biophys Acta 415:29–79 3. Garavito RM, Ferguson-Miller S (2001) Detergents as tools in membrane biochemistry. J Biol Chem 276:32403–32406 4. Serrano-Vega MJ, Magnani F, Shibata Y, Tate CG (2008) Conformational thermostabilization of the beta1-adrenergic receptor in a detergent-resistant form. Proc Natl Acad Sci U S A 105:877–882 5. Toyoshima C, Nakasako M, Nomura H, Ogawa H (2000) Crystal structure of the calcium pump of sarcoplasmic reticulum at 2.6 A resolution. Nature 405:647–655 6. Palsdottir H, Hunte C (2004) Lipids in membrane protein structures. Biochim Biophys Acta 1666:2–18 7. Newstead S, Ferrandon S, Iwata S (2008) Rationalizing alpha-helical membrane protein crystallization. Protein Sci 17:466–472

8. Jastrzebska B, Fotiadis D, Jang GF, Stenkamp RE, Engel A, Palczewski K (2006) Functional and structural characterization of rhodopsin oligomers. J Biol Chem 281:11917–11922 9. Tate CG, Whiteley E, Betenbaugh MJ (1999) Molecular chaperones stimulate the functional expression of the cocaine-sensitive serotonin transporter. J Biol Chem 274:1 7551–17558 10. Ishihara G, Goto M, Saeki M, Ito K, Hori T, Kigawa T, Shirouzu M, Yokoyama S (2005) Expression of G protein coupled receptors in a cell-free translational system using detergents and thioredoxin-fusion vectors Protein. Expr Purif 41:27–37 11. Butler PJ, Ubarretxena-Belandia I, Warne T, Tate CG (2004) The Escherichia coli multidrug transporter EmrE is a dimer in the detergentsolubilised state. J Mol Biol 340:797–808 12. Chen YJ, Pornillos O, Lieu S, Ma C, Chen AP, Chang G (2007) X-ray structure of EmrE supports dual topology model. Proc Natl Acad Sci U S A 104:18999–19004 13. Warne T, Serrano-Vega MJ, Tate CG, Schertler GF (2009) Development and crystallization of

202

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

Tate a minimal thermostabilized G protein-coupled receptor. Protein Exp Purif 65:204–213 Edwards PC, Li J, Burghammer M, McDowell JH, Villa C, Hargrave PA, Schertler GF (2004) Crystals of native and modified bovine rhodopsins and their heavy atom derivatives. J Mol Biol 343:1439–1450 Li J, Edwards PC, Burghammer M, Villa C, Schertler GF (2004) Structure of bovine rhodopsin in a trigonal crystal form. J Mol Biol 343:1409–1438 Lemieux MJ, Reithmeier RA, Wang DN (2002) Importance of detergent and phospholipid in the crystallization of the human erythrocyte anion-exchanger membrane domain. J Struct Biol 137:322–332 Guan L, Smirnova IN, Verner G, Nagamori S, Kaback HR (2006) Manipulating phospholipids for crystallization of a membrane transport protein. Proc Natl Acad Sci U S A 103:1723–1726 Lee SY, Lee A, Chen J, MacKinnon R (2005) Structure of the KvAP voltage-dependent K+ channel and its dependence on the lipid membrane. Proc Natl Acad Sci U S A 102: 15441–15446 Auer M, Kim MJ, Lemieux MJ, Villa A, Song J, Li XD, Wang DN (2001) High-yield expression and functional analysis of Escherichia coli glycerol-3-phosphate transporter. Biochemistry 40:6628–6635 Kawate T, Gouaux E (2006) Fluorescencedetection size-exclusion chromatography for precrystallization screening of integral membrane proteins. Structure 14:673–681 Drew D, Newstead S, Sonoda Y, Kim H, von Heijne G, Iwata S (2008) GFP-based optimization scheme for the overexpression and purification of eukaryotic membrane proteins in Saccharomyces cerevisiae. Nat Protoc 3:784–798 Newstead S, Kim H, von Heijne G, Iwata S, Drew D (2007) High-throughput fluorescent-based optimization of eukaryotic membrane protein overexpression and purification in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A 104:13936–13941 Drew DE, von Heijne G, Nordlund P, de Gier JW (2001) Green fluorescent protein as an indicator to monitor membrane protein overexpression in Escherichia coli. FEBS Lett 507:220–224 Magnani F, Shibata Y, Serrano-Vega MJ, Tate CG (2008) Co-evolving stability and conformational homogeneity of the human adenosine A2a receptor. Proc Natl Acad Sci U S A 105:10744–10749

25. Standfuss J, Xie G, Edwards PC, Burghammer M, Oprian DD, Schertler GF (2007) Crystal structure of a thermally stable rhodopsin mutant. J Mol Biol 372:1179–1188 26. De Grip WJ (1982) Thermal stability of rhodopsin and opsin in some novel detergents. Methods Enzymol 81:256–265 27. Galka JJ, Baturin SJ, Manley DM, Kehler AJ, O’Neil JD (2008) Stability of the glycerol facilitator in detergent solutions. Biochemistry 47:3513–3524 28. Engel CK, Chen L, Prive GG (2002) Stability of the lactose permease in detergent solutions. Biochim Biophys Acta 1564:47–56 29. Faham S, Yang D, Bare E, Yohannan S, Whitelegge JP, Bowie JU (2004) Side-chain contributions to membrane protein structure and stability. J Mol Biol 335:297–305 30. Hughes AV, Rees P, Heathcote P, Jones MR (2006) Kinetic analysis of the thermal stability of the photosynthetic reaction center from Rhodobacter sphaeroides. Biophys J 90: 4155–4166 31. Lau FW, Nauli S, Zhou Y, Bowie JU (1999) Changing single side-chains can greatly enhance the resistance of a membrane protein to irreversible inactivation. J Mol Biol 290:559–564 32. Maneri LR, Low PS (1988) Structural stability of the erythrocyte anion transporter, band 3, in different lipid environments. A differential scanning calorimetric study. J Biol Chem 263:16170–16178 33. Grinberg AV, Gevondyan NM, Grinberg NV, Grinberg VY (2001) The thermal unfolding and domain structure of Na+/K+-exchanging ATPase. A scanning calorimetry study. Eur J Biochem 268:5027–5036 34. Alexandrov AI, Mileni M, Chien EY, Hanson MA, Stevens RC (2008) Microscale fluorescent thermal stability assay for membrane proteins. Structure 16:351–359 35. Weiss HM, Grisshammer R (2002) Purification and characterization of the human adenosine A(2a) receptor functionally expressed in Escherichia coli. Eur J Biochem 269:82–92 36. Jaakola VP, Griffith MT, Hanson MA, Cherezov V, Chien EY, Lane JR, Ijzerman AP, Stevens RC (2008) The 2.6 angstrom crystal structure of a human A2A adenosine receptor bound to an antagonist. Science 322: 1211–1217 37. Tucker J, Grisshammer R (1996) Purification of a rat neurotensin receptor expressed in Escherichia coli. Biochem J 317(Pt 3):891–899 38. Bowie JU (2001) Stabilizing membrane proteins. Curr Opin Struct Biol 11:397–402

Practical Considerations of Membrane Protein Instability during Purification 39. Gohon Y, Dahmane T, Ruigrok RW, Schuck P, Charvolin D, Rappaport F, Timmins P, Engelman DM, Tribet C, Popot JL, Ebel C (2008) Bacteriorhodopsin/amphipol complexes: structural and functional properties. Biophys J 94:3523–3537 40. Picard M, Dahmane T, Garrigos M, Gauron C, Giusti F, le Maire M, Popot JL, Champeil P (2006) Protective and inhibitory effects of various types of amphipols on the Ca2+ATPase from sarcoplasmic reticulum: a comparative study. Biochemistry 45:1861–1869 41. McGregor CL, Chen L, Pomroy NC, Hwang P, Go S, Chakrabartty A, Prive GG (2003) Lipopeptide detergents designed for the structural study of membrane proteins. Nat Biotechnol 21:171–176 42. Zhao X, Nagai Y, Reeves PJ, Kiley P, Khorana HG, Zhang S (2006) Designer short peptide surfactants stabilize G protein-coupled receptor bovine rhodopsin. Proc Natl Acad Sci U S A 103:17707–17712 43. Yeh JI, Du S, Tortajada A, Paulo J, Zhang S (2005) Peptergents: peptide detergents that improve stability and functionality of a membrane protein, glycerol-3-phosphate dehydrogenase. Biochemistry 44:16912–16919 44. Zhou Y, Bowie JU (2000) Building a thermostable membrane protein. J Biol Chem 275:6975–6979 45. Oxenoid K, Kim HJ, Jacob J, Sonnichsen FD, Sanders CR (2004) NMR assignments for a helical 40 kDa membrane protein. J Am Chem Soc 126:5048–5049 46. Shibata Y, White J, Magnani F, Serrano-Vega MJ, Grisshammer R, Tate CG (2009) Thermostabilisation of the neurotensin receptor in a defined agonist-binding ­conformation. J Mol Biol 390:262–277 47. Chou PY, Fasman GD (1978) Empirical predictions of protein conformation. Annu Rev Biochem 47:251–276 48. Gether U, Ballesteros JA, Seifert R, SandersBush E, Weinstein H, Kobilka BK (1997) Structural instability of a constitutively active G protein-coupled receptor. Agonistindependent activation due to conformational flexibility. J Biol Chem 272:2587–2590 49. Warne T, Serrano-Vega MJ, Baker JG, Moukhametzianov R, Edwards PC, Henderson R, Leslie AG, Tate CG, Schertler GF (2008)

50.

51.

52.

53.

54.

55.

56.

57.

58.

203

Structure of a beta1-adrenergic G-proteincoupled receptor. Nature 454:486–491 Sarkar CA, Dodevski I, Kenig M, Dudli S, Mohr A, Hermans E, Pluckthun A (2008) Directed evolution of a G protein-coupled receptor for expression, stability, and binding selectivity. Proc Natl Acad Sci U S A 105:14808–14813 Doyle DA, Morais Cabral J, Pfuetzner RA, Kuo A, Gulbis JM, Cohen SL, Chait BT, MacKinnon R (1998) The structure of the potassium channel: molecular basis of K+conduction and selectivity. Science 280:69–77 Lemieux MJ, Song J, Kim MJ, Huang Y, Villa A, Auer M, Li XD, Wang DN (2003) Threedimensional crystallization of the Escherichia coli glycerol-3-phosphate transporter: a member of the major facilitator superfamily. Protein Sci 12:2748–2756 Rasmussen SG, Choi HJ, Rosenbaum DM, Kobilka TS, Thian FS, Edwards PC, Burghammer M, Ratnala VR, Sanishvili R, Fischetti RF, Schertler GF, Weis WI, Kobilka BK (2007) Crystal structure of the human beta2 adrenergic G-protein-coupled receptor. Nature 450:383–387 McKibbin C, Farmer NA, Jeans C, Reeves PJ, Khorana HG, Wallace BA, Edwards PC, Villa C, Booth PJ (2007) Opsin stability and folding: modulation by phospholipid bicelles. J Mol Biol 374:1319–1332 Cherezov V, Rosenbaum DM, Hanson MA, Rasmussen SG, Thian FS, Kobilka TS, Choi HJ, Kuhn P, Weis WI, Kobilka BK, Stevens RC (2007) High-resolution crystal structure of an engineered human beta2-adrenergic G proteincoupled receptor. Science 318:1258–1265 Rosenbaum DM, Cherezov V, Hanson MA, Rasmussen SG, Thian FS, Kobilka TS, Choi HJ, Yao XJ, Weis WI, Stevens RC, Kobilka BK (2007) GPCR engineering yields high-resolution structural insights into beta2-adrenergic receptor function. Science 318:1266–1273 Bamber L, Harding M, Butler PJ, Kunji ER (2006) Yeast mitochondrial ADP/ATP carriers are monomeric in detergents. Proc Natl Acad Sci U S A 103:16224–16229 Kunji ERS, Harding M, Butler PJG, Akamine P (2008) Determination of the molecular mass and dimensions of membrane proteins by size exclusion chromatography. Methods 46:62–72

Chapter 13 Membrane Protein Solubilization Katia Duquesne and James N. Sturgis Abstract A critical step in any in vitro analysis of membrane proteins is the solubilization of the membrane to extract the protein of interest in an active form to obtain an aqueous solution containing the membrane protein complexed with detergents and lipids in a form suitable for purification and further analysis. This process is particularly delicate as the aim is to maximally disrupt the lipid components of the membrane while putting the protein components in an un-natural detergent environment without perturbing them. Looked at this way, it is remarkable that it ever works. Although the process is difficult and hard to master, an increasing number of membrane proteins have been successfully solubilized in active forms, allowing some general principles to be established that we illustrate in the method developed in this chapter. Key words: Cytoplasmic membrane, detergent, membrane proteins, purification, titration

1. Introduction The solubilization of a membrane protein from a biological sample often involves a series of related steps. It is hard to establish a method without examining the various steps involved and the use to which the solubilized protein will be put. Here, we describe an approach to optimising the solubilization of your membrane protein of interest; we presume that you have some method of specifically detecting the presence of this protein and some measure of its activity or function. While the main theme of this chapter is the solubilization of membrane proteins, we also treat the closely associated aspect of the purification of the membrane fragments. We give a general protocol for optimising membrane protein solubilization and indicate the major parameters that are important for reproducibility. In the Notes section, we give several indications of tests that might be useful.

I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_13, © Humana Press, a part of Springer Science + Business Media, LLC 2010

205

206

Duquesne and Sturgis

2. Materials 2.1. Preparation of Bacterial Cytoplasmic Membranes

1. Bacterial cells. 2. Centrifuge and rotors. 3. Ultracentrifuge and rotors. 4. French press. 5. Isolation buffer: 10 mM Tris-HCl, pH 8.0, 250 mM sucrose, 100 mg/mL deoxyribonucleaseI (DNaseI), 20 mg/mL ribonuclease (RNase), anti-protease cocktail. 6. Lysozyme. 7. 70% sucrose solution in 10 mM Tris-HCl, pH 7.5, 5 mM ethylenediaminetetraacetic acid (EDTA). 8. 50% sucrose solution in 10 mM Tris-HCl, pH 7.5, 5 mM EDTA. 9. 20% sucrose solution in 10 mM Tris-HCl, pH 7.5, 5 mM EDTA.

2.2. Solubilization of Membranes

1. Membranes. 2. Ultracentrifuge suitable for small volumes. 3. Spectrophotometer for absorption measurements. 4. 10 mM phosphate buffer, pH 7.0 (see Note 1). 5. Concentrated detergent solution. In this example, 1M octylglucoside in 10 mM phosphate buffer, pH 7.0.

2.3. Glycosidic Detergent Analysis

1. Standards: Glucose 1 mg/mL stock solution. 2. 80% (w/w) phenol solution. 3. Concentrated sulfuric acid. 4. Spectrophotometer (490 nm).

3. Methods 3.1. Preparing Membrane Fragments

Before considering solubilizing a membrane protein, it is necessary to prepare and purify membrane fragments. Several different methods can be used for opening cells and fragmenting the membranes. The choice often depends on the possible contaminants. The example that we show here involves the purification of bacterial cytoplasmic membranes from Escherichia coli. While the procedure is specific for E. coli cytoplasmic membranes, the protocol is easily adapted to other membrane systems and illustrates

Membrane Protein Solubilization

207

clearly the general approach usually used for preparing purified membrane fragments suitable for subsequent solubilization. Why have we chosen this protocol? This protocol includes examples of the different procedures that are frequently found in membrane preparations, notably fragmentation by sheering, differential ultracentrifugation and density gradient centrifugation. It yields membranes that are reasonably pure rapidly; this is often important as various membrane-associated proteases exist and can degrade your sample if the preparation protocol is too long. Finally, E. coli is often the first choice for heterologous membrane protein over-expression and thus perhaps the most appropriate choice for this chapter. The protocol is loosely based on that published online by the Hancock laboratory for outer membrane purification (1) and is illustrated in Fig. 13.1. The protocol we describe includes a differential centrifugation to separate membrane fragments essentially on the basis of size, coupled with density gradient centrifugation to separate membrane fragments on the basis of density.

Fig. 13.1. Illustration of the Escherichia coli cytoplasmic membrane purification protocol. Left panel: a flow diagram for the purification, OM outer membrane, IM inner (cytoplasmic) membrane. Right panel: 12.5% SDS-PAGE analysis of the soluble inner membrane and outer membrane fractions. Samples were heated at 95°C for 2 min, markers are at relative molecular masses of 250,000, 98,000, 64,000, 50,000, 36,000, 30,000, 16,000 and 6,000.

208

Duquesne and Sturgis

1. The washed bacteria are resuspended by gentle pipetting in an isolation medium at a final concentration of 250 UOD cells/mL. (UOD, Cell concentration is measured in OD units at 600 nm.) 2. Add lysozyme to a final concentration of 1 mg/mL. 3. Break open the cells by two passages through the French pressure cell at 1,200 psi (8,300 kPa). 4. Remove debris by centrifuging at 10,000g for 15 min. 5. Collect the large outer membrane fragments by ultracentrifugation (e.g., 40,000 rpm for 3 min in TLA55 rotor (72,000g)). 6. Continue separation by layering the supernatant onto a discontinuous sucrose gradient containing 2 mL of 70% sucrose, 7 mL of 50% sucrose. 7. Centrifuge 16 h (overnight) at 39,000 rpm in a SW40 swinging bucket rotor (192,000g). 8. Fractionate the gradient and identify inner-membrane fragments (typically at or just below the 20–50% sucrose interface. Inner membranes can be identified by immuno-detection of inner-membrane protein or assay of reduced nicotinamide adenine dinucleotide (NADH) oxidase activity. 9. Pool the selected fractions, dilute three times with sucrosefree buffer, and collect the inner membrane fragments by ultracentrifugation (e.g., 40,000 rpm for 60 min in 70Ti rotor (120,000g)). 10. Resuspend the pellet in 1 mL buffer. Why spend the time purifying the membrane fragments if we can purify the protein afterward? There are essentially two different and important reasons why it is worth spending the effort to obtain a purified membrane fraction before isolation of a membrane protein. First, this step often represents considerable purification, increasing the specific activity of the protein of interest by removing many un-related membrane proteins that are found in other membranes or insoluble proteins and ribosomes that often co-purify with membranes in simpler protocols. These unrelated proteins often cause problems during the subsequent purification steps and are hard to separate. Second, the reproducibility of membrane solubilization protocols depends critically on the quality and reproducibility of the membrane preparation. Thus, while a particular membrane sample can be purified using a certain method, the presence of impurities can drastically modify the solubilization step to the point that any modification of the membrane purification can result in the need to re-evaluate the solubilization step. For this reason, it is not worth trying to perfect a solubilization protocol if you do not have a reproducible membrane preparation protocol.

Membrane Protein Solubilization

3.2. Membrane Solubilization

3.2.1. Basic Solubilization Procedure

209

The development of a membrane solubilization protocol is a difficult and repetitive process. Often, it is necessary to restart the work because some aspect of the final preparation is inadequate for the desired use or some critical aspect of the starting material changes. Nevertheless, the route we trace is relatively straightforward and allows the rapid establishment of reliable, reproducible routine conditions for solubilization. The process is inherently multivariant, with many parameters playing an important role. It is usual to select a protein and membrane first, usually determined by the project, and then chose a detergent that is promising for solubilization and finally to try to optimise solubilization conditions by modifying the concentrations, temperature and solubilization time. 1. Incubate a known amount of membrane sample (mg) in a known volume (mL) of solution with a known final concentration of your selected detergent (mM or mg/mL) at a given temperature (°C) for a given time (minutes). As can be immediately appreciated, there are a large number of variables that need to be optimised, and these are treated here in slightly more detail. It cannot be overemphasised enough that the amount of membrane, the volume and the final concentration of detergent are three independent variables (see Note 2). In the example experiments (Fig. 13.2), extractions were for 5 min at 20°C in a total volume of 1 mL, the membrane quantities used were 2–10 mg protein and the detergent octyl-glucoside concentrations ranged from 0 to 50 mM. 2. Separate the solubilized protein from the un-solubilized “detergent-resistant” membranes. This step most usually

Fig. 13.2. Illustration of membrane solubilization. Bacterial cytoplasmic membranes were solubilized with the detergent octyl-glucoside by titrating the detergent from 0 to 50 mM. As the detergent was added, the light-scattering changes were measured (panel A) in a spectrophotometer at 600 nm, a wavelength at which the sample does not absorb strongly. Selected points in the curve, for example, the end of the decline corresponding to clarification, correspond to a given level of solubilization and vary as a function of the membrane concentration. This is shown in panel B for complete (upper line) and 50% (lower line) solubilization.

210

Duquesne and Sturgis

involves an ultracentrifugation necessary to sediment the remaining membrane fragments. For this purpose, a small benchtop ultracentrifuge (type airfuge or TL100) is particularly useful (see Notes 3 and 4). In the experiments in Fig. 13.2, this separation was made by centrifugation at 40,000 rpm for 60 min in a TLA45 rotor (120,000g at 20°C). 3. Separate the supernatant of solubilized material from the pellet, and if necessary for analysis resuspend the pellet. It is usually easiest to resuspend the pellet in about the same volume of buffer as it was separated from (see Note 5). 4. Analyse the solubilized and unsolubilized fractions to determine the yield, purity, activity and stability of the protein of interest. When considering the purity of membrane protein fractions, it is important to look beyond other proteins and consider the presence of other nonprotein contaminants (e.g., lipids). Several variations on this general scheme can be found in the literature; perhaps the most interesting involves repeated extractions. Often, it is possible to extract certain types of impurities selectively and by using two successive solubilizations obtain considerable purification (2, 3). 3.2.2. Choice of Detergent

The basic protocol described requires choosing a detergent for extraction. The number of different detergents available is considerable, and these have different properties and prices. It is usually not feasible to test more than a dozen or two in the extraction of the membrane protein of interest. The choice of detergent is primarily dictated by the sensitivity of the protein of interest to the detergent; it is thus imperative to find a detergent in which the protein of interest is stable, at least for the time necessary for isolation (see Note 6). Unfortunately, this most important criterion for detergent selection is the hardest to predict; worse, there is no guarantee that a detergent exists in which your protein is stable. The choice of detergent also depends to some extent on the subsequent use envisaged for the solubilized protein; for example, some detergents are useful for membrane protein crystallisation, and others are not (see Note 7). In Table 13.1, we have collected some of the most commonly used detergents, and a few that are less commonly used, together with some information useful when considering their use for solubilizing membrane proteins. Different laboratories often have different cultures in testing detergents and different favourites. Some laboratories swear by Triton X-100, while others refuse to use it. The wide variety of detergents available and the relative paucity of systematic studies (4) make this aspect particularly daunting for the novice.

Membrane Protein Solubilization

211

Table 13.1 Table of detergent properties. For a selection of detergents, the CMC, Nagg (aggregation number) and several notes on their uses are shown. Values are from the Anatrace and Sigma literature (5, 6), in which information on many other detergents can also be found, with the exception of di hexanoyl phosphatidyl choline (DHPC) (7). Detergent name

cmc mM

Nagg

Notes

Dodecyl-b-d-maltoside

0.17

78–149

a,b

Decyl-b-d-maltoside

1.8

69

a

Octyl-b-d-glucoside

18.0

78

a,b

C8E5

7.1

41

a,b

C10E5

0.81

73

C12E8

0.09

90–180

Triton X-100

0.23

75–165

c,d

Tween-20

0.059(3)

Brij-58

0.08

70

c

Cholate

9.5

2–3

Deoxycholate

6.0

22

CHAPS

8.0

10

a

Big-Chap

2.9

10

a

Digitonin

<0.5

60

a

SDS

2.6

62—101

e

Fos-choline-12

1.5

50—60

Used for NMR

Lauryl-DAO

1.0

76

b

ANZERGENT 3 12

2.8

55–87

DHPC

10.0

19

Glucidic group

Oxyethylene group

Polycyclic group

Ionic group Anionic

Zwitterionic, cheap

Phospholipid, can make bicelles

a

SDS sodium dodecyl sulfate; DAO dimethylamine Noxide a Reputedly mild b Used in crystallography c Polydisperse d Strong ultraviolet (UV) absorption e Reputed strong and denaturing

As a general rule, when developing purification protocols it is necessary to try 5–30 different detergents in the search for one in which the protein is active and stable over a few hours and the

212

Duquesne and Sturgis

time necessary for the solubilization. To help in the selection of these detergents, I would suggest the following criteria be applied: Use a variety of detergent chemistries. For example: 1. Include detergents with polyoxyethylene, glucosidic and zwitterionic head groups 2. Include aliphatic and sterol-based detergents 3. Include detergents with low and high critical micellar concentration (CMC) Avoid detergents that are incompatible with the anticipated tests. It is disheartening after having established a protocol to restart because a change in detergent is necessary. After initial tests, variants on the promising chemistries can be tested. To test detergents for their potential in solubilizing your protein of interest without completely immediately and irreversibly destroying the protein, it is often possible to rule out some detergents by testing their effects on activity, at a series of different concentrations, and then checking the solubilization in the most promising detergents. It must be stressed that the choice of detergent still can be considered as an art, and currently there are no hard and fast rules, which perhaps reflects our lack of understanding of protein–detergent interactions. 3.2.3. Detergent Titration

In this section, we describe and illustrate the titration of bacterial membrane fragments with the detergent n-octyl-bd -glucopyranoside (octyl-glucoside). This experiment is shown in Fig.  13.2. The experiment that we describe allows us to determine the solubilization profiles for different proteins and establish the amounts of detergent and membrane necessary for reproducible solubilization. It is important to insist that reproducible solubilization depends also on reproducible membrane preparations, and a change in the membrane preparation can result in changes in the solubilization. 1. Prepare a set of samples containing several different concentrations of membrane. In the example, shown we used samples containing 2, 5, 8, and 10 mg/mL of membrane. (This experiment requires about 30 mg of membrane total.) 2. In a spectrophotometer, measure the optical density due to turbidity at an appropriate wavelength (600 nm in Fig. 13.2). For these measurements, the cuvette should be thermostated at the desired temperature (20°C in Fig. 13.2). 3. Repeat the measurements of turbidity as a function of detergent concentration, adding detergent stepwise and allowing the solution to equilibrate for an appropriate time between additions (5 min in Fig.  13.2). A typical graph of optical density as a function of detergent concentration is

Membrane Protein Solubilization

213

shown in Fig.  13.2a. Note that clarification of the solution reduces the optical density but not necessarily to zero if there is some absorption (as is the case in the membranes used to construct the figure). 4. Repeat the measurements as a function of detergent concentration for each of the different membrane concentrations and in each case determine the detergent concentrations that correspond to the points on the curves. In the experiment shown in Fig.  13.2a, we chose the solubilization endpoint and the point at which the absorption had diminished by 50% of the total decline. 5. Prepare a graphic along the lines of Fig. 13.2b in which the detergent concentrations of the different points on the titration curves are shown as a function of the membrane concentration. These points should show a linear dependence on the membrane concentration, and a linear regression thus allows the calculation of the free detergent concentration (y-axis intercept) and the detergent/membrane ratio (slope). For example, the upper line in Fig.  13.2b (complete clarification of the membrane suspension) corresponds to a free detergent concentration of 20.4 mM (about the CMC) and a detergent/protein ratio of 1.7 mmol detergent/ mg protein. 6. It is worth pointing out that even at concentrations above that necessary to clarify the solution, changing the detergent concentration can modify the association of the different components in the solution (8); however, the free detergent concentration will remain more or less constant at the CMC. 7. Repeat a final titration, selecting several key points in the titration curve (Fig. 13.2a) at which to remove samples for comparison of the solubilized and non-solubilized proteins. 8. Separate the solubilized and non-solubilized fractions as described and then analyse each sample for composition, for example, by enzymatic assay or sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) (see Note 8). This experiment provides a wealth of information on the solubilization process and allows the evaluation of conditions for reproducible solubilization with different amounts of membrane. The graphic in Fig. 13.2b illustrates clearly that the total concentration of detergent necessary to arrive at a particular place on the solubilization profile corresponds not to a given concentration of detergent but to a given concentration (the y intercept) plus a given ratio (the slope) times the membrane concentration. This needs to be remembered during scale-up from analytical to preparative solubilization.

214

Duquesne and Sturgis

3 .2.4. Time and Temperature

In the experiment shown in Fig. 13.2, the time of solubilization and temperature of solubilization were arbitrarily set to 5 min and 20°C, respectively. However, both these parameters can have a profound effect on the solubilization. In particular, low temperatures and short times can between them offer kinetic selection of fast-solubilizing proteins. Beyond this, temperature can change dramatically the way in which the lipid and detergents mix, possibly resulting in completely different extraction profiles and “detergent-resistant” membranes with different characteristics and composition. In general, the range of reasonable values is relatively limited since often we do not wish to subject the membranes to excessively high temperatures or long incubations for fear of protease activity and limitations of protein stability. As a general rule, solubilization is most reproducible away from any lipid-phase transitions, and typical temperatures to test are 4, 20, and 37°C. Solubilization time sometimes has a more dramatic effect, with certain combinations of detergent and membrane solubilizing only slowly, particularly at low temperature.

3.3. Analytical Techniques

When analysing protein and membrane solubilization, it is necessary to have several analytical tools to follow the extraction of the protein of interest and the dissolution of the membrane by the lipids. It is important also to keep in mind that the solubilized protein is almost always a particle containing protein, lipid and detergent. Spectroscopic methods such as infra-red absorption can be used to determine the concentrations of lipids, detergents and proteins (9).

3.3.1. Detergent Analysis

No technique is universally appropriate for measuring the detergent concentration of a sample as detergents are chemically too variable. Certain spectroscopic techniques can be used; however, not all laboratories are equipped with an appropriate spectrophotometer, and the analysis of the spectra is often time consuming and is beyond the scope of this chapter. In some simple cases, specific sensitive chemical tests are available for determining the concentration of certain detergents. An example is the phenol-sulphuric acid method (10), which can be adapted to determine the concentration of glycosidic detergents such as octyl-glucoside (in the absence of glyco-lipids and glyco-proteins). The protocol is relatively straightforward. 1. Prepare standards and samples containing 0–50 mg carbohydrate in 50 mL of solution. Typically, use as standards 0, 5, 10, … , 50 mL of a 1 mg/mL glucose solution for the standard curve. 2. Add 25 mL of 80% phenol solution to each sample and mix well.

Membrane Protein Solubilization

215

3. Add 1.0 mL of concentrated sulphuric acid in a stream to each sample. This addition results in considerable heating and probably hydrolyses many less-stable anhydride linkages. 4. Let stand 10 min at room temperature for the colour to develop. 5. Read absorbance at 490 nm. 6. Construct a standard curve using the glucose solution, with which you can estimate the carbohydrate composition of the sample. 3.3.2. Lipid Analysis

As with detergents, there is no universally appropriate method for measuring the concentration of lipids in a sample, as they also are chemically variable. However, again there are a number of methods that can be used to assay certain lipids. A convenient measure for phospholipids is based on the iron-molybdate phosphate assay.

4. Notes There are many places where the protocol indicated can go wrong or can appear inconsistent or hard to reproduce. Here, we indicate several types of problems and guides to resolving them. 1. In adapting this method to the system of interest, the buffer type, pH and ionic strength and selected detergent can be changed to suit the system. 2. It is usually best to add detergent slowly to the sample while mixing the sample to minimise sudden local changes in detergent concentration that can lead to poorly reproducible results. 3. Do not dilute the sample to make up the volume necessary for ultracentrifugation; use a tube appropriate for the volume you are solubilizing. Such a change in volume will modify the parameters. 4. Do not unnecessarily change the sample temperature; membranes solubilize most easily at higher temperatures, but proteases are more active. 5. The detergent-resistant unsolubilized membranes contain a considerable amount of detergent and thus behave quite differently from un-treated membranes. Furthermore, on addition of buffer some of the detergent will move from the membrane to the buffer, possibly carrying membrane proteins with it. Try to avoid foaming during resuspension.

216

Duquesne and Sturgis

6. It is sometimes beneficial for the stability of proteins to add extra lipids during solubilization. 7. Several surfactants that are not good for membrane solubilization offer promise for stabilising membrane proteins; these include amphipols (11) and fluorinated surfactants (12). Such surfactants can be used for detergent exchange immediately after extraction. 8. To cook or not to cook? Membrane proteins often behave (very) badly on SDS gels, and one of the known difficulties is solubilization in SDS. The situation can be simply summarised; if membrane proteins are heated, even in SDS they tend to form insoluble aggregates that do not enter the gel. If on the other hand the sample is not heated, membrane proteins are often found in large aggregates that do not dissolve easily even in SDS and thus do not enter the gel. As a consequence, if you heat the sample you run the risk of not seeing proteins because they are denatured and aggregated, while if you do not heat the sample you run the risk of not seeing proteins because they have not been denatured. Sometimes this heat sensitivity of aggregation can be used to advantage in determining the oligomeric state of the isolated protein. This has been used for identifying trimeric porins from bacterial outer membranes (13) and for the investigation of potassium channel assembly (14). This situation results in another domain in which different laboratories have rather different protocols. In our laboratory, we tend to cook the samples only for short periods of time (2 min), but even so some proteins can become aggregated and disappear. So, you should be aware that detecting your protein of interest on a protein gel might be more difficult than anticipated.

Acknowledgements This work, and K.D., were supported by the CNRS and ANR (grant number Blanc-2007-1_183779). References 1. Hancock REW (2000) Outer membrane preparation. http://www.cmdr.ubc.ca/bobh/ methods/ 2. Pucheu N, Kerber N, Garcia A (1976) Isolation and purification of reaction center from Rhodopseudomonas viridis NHTC 133 by means of LDAO. Arch Microbiol 109:301–305

3. Garavito R, Rosenbusch J (1986) Isolation and crystallization of bacterial porin. Methods Enzymol 125:309–328 4. le Maire M, Champeil P, Moller J (2000) Interaction of membrane proteins and lipids with solubilizing detergents. Biochim Biophys Acta 1508:86–111

Membrane Protein Solubilization 5. Sigma. Biological detergents. http://www. sigmaaldrich.com/ 6. Anatrace. Detergent technical data. http:// www.anatrace.com/ 7. Lin T-L, Chen S-H, Roberts MF (1987) Thermodynamic analysis of the structure and growth of asymmetric linear short-chain lecithin micelles based on small-angle neutron scattering data. J Am Chem Soc 109:2321–2328 8. Fisher LE, Engelman DM, Sturgis JN (2003) Effect of detergents on the association of the glycophorin a transmembrane helix. Biophys J 85:3097–3105 9. daCosta C, Baenziger J (2003) A rapid method for assessing lipid:protein and detergent: protein ratios in membrane-protein crystallization. Acta Crystallogr D Biol Crystallogr 59:77–83 10. DuBois M, Gilles KA, Hamilton JK, Rebers PA, Smith F (1956) Colorimetric method for determination of sugars and related substances. Anal Chem 28:350–356

217

11. Pocanschi CL, Dahmane T, Gohon Y, Rappaport F, Apell HJ, Kleinschmidt JH, Popot J (2006) Amphipathic polymers: tools to fold integral membrane proteins to their active form. Biochemistry 45:13954–13961 12. Breyton C, Chabaud E, Chaudier Y, Pucci B, Popot J-L (2004) Hemifluorinated surfactants: a non-dissociating environment for handling membrane proteins in aqueous solutions? FEBS Lett 564:312–318 13. Rosenbusch J (1974) Characterization of the major envelope protein from Escherichia coli. Regular arrangement on the peptidoglycan and unusual dodecyl sulfate binding. J Biol Chem 249:8019–8029 14. Molina M, Encinar J, Barrera F, FernandezBallester G, Riquelme G, Gonzalez-Ros J (2004) Influence of C-terminal protein domains and protein-lipid interactions on tetramerization and stability of the potassium channel KcsA. Biochemistry 43:14924–14931.

Chapter 14 Amphipols and Fluorinated Surfactants: Two Alternatives to Detergents for Studying Membrane Proteins In vitro Cécile Breyton, Bernard Pucci, and Jean-Luc Popot Abstract Handling integral membrane proteins in aqueous solutions traditionally relies on the use of detergents, which are surfactants capable of dispersing the components of biological membranes into mixed micelles. The dissociating character of detergents, however, most often causes solubilized membrane proteins to be unstable. This has prompted the development of alternative, less-aggressive surfactants designed to keep membrane proteins soluble, after they have been solubilized, under milder conditions. A short overview is presented of the structure, properties, and uses of two families of such surfactants: amphiphilic polymers (“amphipols”) and fluorinated surfactants. Key words: Amphipols, detergents, fluorinated surfactants, membrane proteins

1. Introduction Why do we bother designing, synthesizing, and testing new surfactants to handle membrane proteins (MPs)? But for a handful of them, MPs, as a rule, need to be extracted from their original membrane to be purified and studied (see Chapters 12 and 13, this volume). Unfortunately, whereas most of them are reasonably stable in their natural environment, they tend to become unstable once transferred to detergent solutions. Choosing the appropriate detergent (not necessarily the same one for every step!) and determining the concentrations to be used to solubilize the membrane and to purify, characterize, reconstitute, and crystallize the protein of interest without inactivating it are the most critical tasks the MP biochemist has to face. C.G. Tate, in Chapter 12, details his views of protein instability in detergent solution. This point therefore is not extensively discussed here. Let us just recall some important considerations I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_14, © Humana Press, a part of Springer Science + Business Media, LLC 2010

219

220

Breyton, Pucci, and Popot

that have guided us in the design of new surfactants. Detergents need to be dissociating to solubilize biological membranes. In other words, they need to efficiently compete with lipid–lipid and lipid–protein interactions to disrupt them and to disperse their components under the form of detergent-solubilized particles. Once the target protein has been extracted, though, the dissociating character of the detergent is difficult to control and frequently entails the destabilization and irreversible inactivation of the protein (see, e.g., 1–6). Part of this problem originates from the fact that detergent molecules are in a rapid equilibrium between soluble monomers, micelles, and the detergent layer that covers the transmembrane surface of the protein. To prevent its desorption from this surface, which leads to protein aggregation, the detergent must be present at a concentration higher than its critical micellar concentration (CMC). Two mechanisms of destabilization of MPs by detergents appear particularly frequent: (1) detergent micelles act as a hydrophobic phase into which stabilizing hydrophobic cofactors, lipids, or subunits can partition; (2) detergents may interfere with intramolecular protein–protein interactions (e.g., those between transmembrane a-helices) that stabilize the three-dimensional structure of proteins. These two mechanisms are not exclusive, and they are supported by many observations (see, e.g., 7–13), and references therein). Another hypothesis that has been proposed relates to the dynamics of MPs, some of which (e.g., among receptors and transporters) are in equilibrium between different conformational states. They are not rigid entities, but rather wobbling buildings, relatively stable in their natural environment. The detergent belt however does not share all of the physical-chemical characteristics of a lipid bilayer, which may lead to instability, collapse, unfolding, or aggregation of the protein (3, 14, 15) (see Chapter 12, this volume). Here, we focus on the design of new amphipathic molecules as alternatives to detergents. Most of these molecules do not belong to the class of detergents in the sense that they do not solubilize biological membranes or do so inefficiently. They are surfactants, however, that partition at the air–water interface and associate with the transmembrane surface of MPs.

2. Alternatives to Classical Detergents

Various innovative approaches have been developed to go beyond the level of stability that can be achieved using classical practices. This can be done by using proteins from thermophilic bacteria or by selecting stable mutants. Multiple stabilizing mutations with a

Amphipols and fluorinated surfactants

221

cumulative effect can be deliberately accumulated, as first demonstrated with diacylglycerol kinase (16). The latter approach has been successfully applied to stabilizing the b1-adrenergic receptor (14), which has been one of the keys to solving the crystallographic structure of this G protein-coupled receptor (GPCR) (17). The same strategy (see Chapter 12, this volume) has been applied to stabilizing the agonist-binding conformation of the adenosine A(2a) receptor (15). Alternatively, one can attempt to solve the problem by developing surfactants that are able to keep MPs water soluble without presenting the dissociating character of detergents. This has potentially two advantages: (1) The protein itself does not have to be modified; (2) once an improved surfactant has been developed and validated, it may be used for a wide number of MPs, if not most, without having to start from scratch, for every new target, the work-intensive procedure of selecting stable variants. Needless to say, the genetic and physical-chemical approaches could conceivably be combined to tame particularly intractable proteins. One approach to developing milder surfactants is to devise detergent-looking molecules whose hydrophobic moiety has been designed to be, it is hoped, less intrusive, either because it is more rigid (e.g., because it contains rings) or because it is comprised of several short chains rather than a single long one. The most developed molecules of the latter kind are the so-called tripod amphiphiles, whose hydrophobic moiety is made up of three short chains, one of them cyclic (see 18, 19, and references therein). Some of them appear to be particularly mild to bacteriorhodopsin (BR), rhodopsin, and purple bacteria photosynthetic complexes, and one of them has yielded crystals of BR (20). As for most of the molecules discussed in this chapter, tripod amphiphiles are not strongly dissociating and are probably mostly useful as a medium into which to transfer MPs once they have been solubilized by a classical detergent. Another approach that has been studied consists of transferring MPs to small patches of lipid bilayers whose rim is stabilized either by small surfactants such as bile salts or short-chain lipids or by lipoproteins. The first type of structure is called a bicelle, the second one a nanodisk or reconstituted high-density lipoprotein (rHDL) phospholipid bilayer particles (see, e.g., 21–24, and references therein). No systematic study of the stability of MPs transferred to these environments seems to have been carried out, but the fact that the protein is primarily or entirely in contact with lipids rather than detergent is a priori a highly favorable factor. Bicelles have several interesting features, including that they can orient in the high magnetic field of nuclear magnetic resonance (NMR) spectrometers, taking the MP along (see, e.g., 21, 22). Nanodiscs/rHDL particles have the unique property of encompassing lipid patches of definite sizes, which are defined by that of the lipoprotein chosen. This makes it possible,

222

Breyton, Pucci, and Popot

for instance, to selectively trap GPCR monomers, but not dimers, which has led to their use in functional studies aiming at determining what is the minimal signaling unit that can activate G proteins (see, e.g., 23, 24). The size of currently available variants of lipoproteins imposes an upper limit on that of the transmembrane domains that can be trapped, the largest one reported to date being that of the trimer of BR (21 transmembrane helices) (25). Many other types of molecules, such as polyenic antibiotics or amphipathic peptides, can also form the rim of bilayer patches. This is true, in particular, of lytic peptides that fold into laterally amphipathic a-helices, the hydrophobic face of which can match the lipids while the hydrophilic one is in contact with water, such as melittin and many other toxins. The concept can be carried further: Why not use such peptides to provide the interface between the solution and the hydrophobic transmembrane surface of a MP? This idea led to the development of “peptitergents,” a-helix-forming amphipathic peptides that were shown to keep soluble some, but not all, of several model MPs (26). Improvements on the concept have been made by linking a fatty acyl chain to each extremity of the amphipathic a­-helix, with the view of providing a softer interface between the peptide and the surface of the protein. Complexes formed by such “lipopeptide detergents” and PagP (a b-barrel MP) are well defined and small enough to yield good solution NMR spectra (27). As with all of the approaches mentioned in this section, it is to be hoped that a larger number of MPs will be trapped with these surfactants, and their properties studied, to lead to more general conclusions about the prospects they offer. It is, indeed, an underlying tenet of this chapter that none of the novel approaches it discusses can be considered a panacea, but that, rather, they are complementary in diversifying the toolbox of the biochemist and biophysicist. The two approaches whose discussion forms the core of this chapter are based on different premises. In one case, that of fluorinated surfactants (FSs), the original idea was to modify classical detergents to make them nonintrusive and their micelles poor solvents for lipids and hydrophobic cofactors. In the second case, that of amphipols, the initial predicament was to get rid of micelles altogether.

3. Fluorinated Surfactants 3.1. Rationale

A direction our laboratories has been investigating is the use of FSs. FSs have the same general structure as classical detergents

Amphipols and fluorinated surfactants

223

Fig. 14.1. Chemical structures of fluorinated and hemifluorinated surfactants: F-TAC (38) and HF-TAC (41), HF-AO (43), HF-Lac (37), HF-Malt (44), and HF-Mono-, Di- or Triglu (45). See text.

(i.e., a hydrophilic head group and a hydrophobic tail), but the latter, rather than being a hydrogenated aliphatic chain, comprises fluorine atoms (Fig. 14.1). They were designed based on the observation that alkanes and perfluorinated alkanes, while both hydrophobic, are poorly miscible (28–31). Surfactants with fluorinated alkyl chains could be expected to be less aggressive toward MPs than detergents are for the following two reasons: (1) Their micellar phase is expected to be a poor solvent for lipids and other stabilizing hydrophobic cofactors; (2) their hydrophobic moieties, more bulky and more rigid than their hydrogenated counterparts and having little affinity for the hydrogenated surfaces of transmembrane protein segments, could be expected to compete less efficiently with protein–protein interactions and therefore to intrude less easily into the protein structure itself. However, by the same token, perfluorinated chains might have little affinity for the hydrogenated transmembrane surface of MPs and be inefficient at keeping them from aggregating. To improve interactions with the protein, a hydrogenated tip can be grafted at the end of the fluorinated tail, yielding “hemifluorinated” surfactants (HFSs; Fig.  14.1). As both FSs and HFSs are lyophobic (both hydrophobic and oleophobic), they do not partition well into lipids (32); as a result, they have little cytolytic effect (33–35). Membrane proteins therefore have to be extracted with a detergent before being transferred to

224

Breyton, Pucci, and Popot

either FSs or HFSs, hereafter discussed collectively as (H)FSs. Transfer from detergent to (H)FSs can be performed using classical detergent exchange methods: dilution under the CMC of the detergent in the presence of the (H)FS, centrifugation in sucrose gradients containing the (H)FS (see, e.g., 36), specific removal of the detergent from a mixture of detergent and (H)FS using BioBeads—as the (H)FSs have a far lower affinity for BioBeads than detergents (see, e.g., 37), and so on. The first FS to be successfully tested was C6F13C2H4S-poly-Tris-(hydroxymethyl) aminomethane (F-TAC in Fig. 14.1) (36, 38). A slightly longer version, C8F17C2H4-S-poly­ Tris-(hydroxymethyl)aminomethane (F8-TAC in Fig. 14.1) (38) has also been used in several studies (32, 39, 40). As expected, it was found that F-TAC was unable to solubilize photosynthetic thylakoid membranes. However, cytochrome b6 f, a complex extracted from these membranes and purified in detergent solution, remained soluble, in its native form, and enzymatically active after transfer to F-TAC (36). In the experimental conditions of that time, a tendency to aggregation was nevertheless observed (possibly due to the presence of extraneous lipids). This led us to synthesize a hemifluorinated version of the ­molecule, C 2H 5C 6F 12C 2H 4-S-poly-Tris-(hydroxymethyl)aminomethane (HF-TAC in Fig. 14.1) (41). HF-TAC also displayed no detergent activity, but cytochrome b6 f, the transmembrane b-barrel of Escherichia coli outer membrane protein A (tOmpA) and BR all migrated as single sharp bands during centrifugation in HF-TACcontaining sucrose gradients. Furthermore, the b6 f complex and BR, both of which are detergent sensitive, were kept in their native and active state (42). The synthesis of the polar head of HF-TAC involves radical polymerization, which inevitably leads to polydispersity and batch-to-batch variations, two properties that are far from ideal from the biochemist’s point of view. We thus developed the synthesis and investigated the potentialities of (H)FSs bearing chemically defined polar heads deriving from aminoxide (43), from lactose ((H)F-Lac) (37), from maltose ((H)F-Malt) (44), from polyoxyethylene ((H)F-E8, unpublished), or from a Tris motif grafted with one, two, or three glucose molecules ((H)F-Mono-, Di-, and Triglu, respectively; (45, 45a)) (Fig. 14.1). Before discussing the applications of (H)FSs, we describe some of their unusual physical-chemical properties, which directly affect their use in biochemistry. 3.2. Physical Chemistry of Fluorinated Surfactants

Because fluoroalkanes are more hydrophobic than their hydrogenated counterparts, FSs are less soluble in water than detergents with the same polar head and hydrophobic chain length, and their CMC is much lower (see, e.g., 36, 38). As a general rule, the hydrophobic character of a CF2 group within a fluorinated

Amphipols and fluorinated surfactants

225

aliphatic chain is equivalent to that of about 1.7 CH2 groups. For a given CMC, FSs therefore have shorter chains than detergents. However, at variance with what is observed for detergents, the addition of an ethyl group at the tip of a fluorocarbon chain does not lower the CMC, even though the chain is elongated by two carbons (see, e.g., 37, 41, 44, 45). At least part of this effect probably arises from unfavorable van der Waals interactions between fluorinated and hydrogenated groups or from steric hindrance and thus poor packing in the core of hemifluorinated micelles. Electron withdrawing by fluorine atoms confers an acidic character to the methylene group vicinal to the fluorocarbon core, which should favor hydrogen bonding with the aqueous phase and may also work against micellization (37, 45). Another striking feature of (H)FSs is the type of micelles they form. Whereas F-TAC and HF-TAC form well-defined, small, globular micelles (Stokes radius 2.5–3 nm), as most usual detergents do, (H)F-Lac, (H)F-E8, and (H)F-Malt, even in dilute solutions, form large, heterogeneous, probably cylindrical aggregates (Stokes radius 9.5–16 nm) according to dynamic light scattering (DLS) and analytical ultracentrifugation (AUC) measurements, in keeping with molecular dynamics simulations of F-Lac (37). The type of aggregates formed affects the nature of MP–surfactant complexes: Surfactants that form large and heterogeneous aggregates lead to heterogeneous complexes, because the amount of surfactant bound to the protein varies from one complex to the next ((44), unpublished data). Increasing the bulk of the polar head group of a surfactant is known to increase the interfacial curvature of the aggregates it forms and thus to favor the formation of small, globular micelles (46). The application of this principle to (H)FSs was tested by synthesizing surfactants whose head group was comprised of a Tris motif grafted with one to three glucose molecules. Indeed, whereas F-Monoglu again formed large aggregates (mean Stokes radius 12–13 nm) (H)F-Di- and -Triglu formed small, homogeneous, globular micelles (Stokes radius 2.5–3 nm), as measured by DLS, small-angle neutron scattering (SANS), and AUC (45, 45a)). 3.3. Stability of Membrane Proteins in Fluorinated Surfactants

When transferred to HF-TAC, not only are both BR and the b6 f complex in their native state and active, but they also exhibit increased biochemical stability on extended storage in the presence of a high concentration of surfactant, as compared to incubation with a similar excess of detergent micelles, including those of the hydrogenated analogue of HF-TAC (H-TAC) (42) (Fig.  14.2). Irrespective of the homo- or heterogeneity of protein–surfactant complexes, this trend has been confirmed with most (H)FSs tested to date ((37, 44, 45a), unpublished data). Thus, the hypotheses that led to the design of (H)FSs appear

226

Breyton, Pucci, and Popot

v­ indicated: Proteins that demand special care when handled in detergent solutions (e.g., because, sensitive to delipidation, they are stable only at detergent concentrations close to the CMC) or in the presence of added lipids (11) are as stable or markedly more stable, depending on the concentration, in (H)FSs than in an equivalent concentration of detergent (Fig. 14.2). Another interesting consequence of the lyophobic character of FSs emerged upon analyzing preparations of (H)FS-solubilized BR. In its native purple membrane, BR is a trimer (47); upon solubilization in detergent solutions, it monomerizes (48). When BR is transferred from detergent to (H)FS solutions under specific conditions, however, dimeric and trimeric forms can be isolated (Fig. 14.3) that have an absorption spectrum similar to that of BR in the purple membrane (49). Upon centrifugation in sucrose density gradients and AUC, no higher oligomeric forms could be

Fig. 14.2. Compared stability of the cytochrome b6f complex in dodecylmaltoside (DDM) and in HF-TAC at low (0.2 and 1 mM, respectively) or high (5 mM) surfactant concentration. (a) Sedimentation analysis of the b6f complex on 10–30% (w/w) sucrose gradients. The gradients were centrifuged for 4 h at 200,000g. D dimer, M monomer. (b) Evolution over time of the enzymatic activity of cytochrome b6f. Samples were collected from the gradients and kept at 4°C in the dark. Activity was monitored immediately after dilution of catalytic amounts of b6f into 0.3 mM DDM by following spectroscopically the reduction of plastocyanin (the electron acceptor) in the presence of plastoquinol (the electron donor) (83).

Fig. 14.3. Sedimentation analysis of bacteriorhodopsin on 10–30% sucrose gradients, in either 1 mM DDM or 2 mM F-TAC. While only the monomer can be observed in DDM, three bands are observed in F-TAC, most likely the monomer, dimer, and trimer of BR (49). The gradients were centrifuged for 4 h at 200,000g. M monomer.

Amphipols and fluorinated surfactants

227

detected. This observation tends to confirm that (H)FSs, rather than dissociating MP complexes, as most detergents are prone to do, may help preserve — or reform — protein–protein interactions that stabilize oligomers. A similar protective effect has been noted in the cases of the yeast adenosine 5¢-triphosphate (ATP)-synthase (40) and of the dimers of a GPCR, BLT1 (J.-L. Banères 2004, personal communication). If the environment experienced by MPs differs between the membrane and a detergent belt, it is even more different in a membrane–protein–FS complex, where the transmembrane surface of the protein is in contact with fluorocarbon. Both cytochrome b6 f and BR are colored proteins, whose visible spectra (due to the presence of chlorophyll and of retinal, respectively) report on the native state of the protein. Whereas no spectral alteration could be detected for the b6 f in FSs, such is not the case for monomeric BR: BR is purple whether it is detergent-solubilized or in its native membrane environment, but monomeric BR transferred to any FS turns blue, a spectral shift that is reversible on back-transfer to detergent solutions (49). The reasons for this phenomenon are under investigation. Our working hypotheses are either the stabilization of an intermediate state of the photocycle of BR or the occurrence of small conformational changes that would modify the environment of the retinal. As mentioned, both the dimer and the trimer of BR exhibit, in FSs, dark-adapted spectra similar to that of the purple membrane. F-TAC and HF-TAC are now being used in several laboratories. As more proteins are handled in these surfactants, more data will become available, and it should become clearer which phenomena underlies their effects on BR and whether those are related to the observed stabilization of MPs in general. 3.4. Applications

Besides handling detergent-sensitive MPs in aqueous solutions, there are a number of applications for which (H)FSs appear particularly suitable.

3.4.1. In Vitro Synthesis

Overexpressing MPs under their functional form in vivo is a difficult task, and yields are most often limited. Cell-free synthesis is developing as an alternative method. Membrane proteins can be expressed in vitro in the presence of various detergents or left to precipitate and later solubilized (see Chapter 11, this volume). It is to be expected, however, that many proteins that are unstable in detergent solution will not stand either procedure. (H)FSs may provide a particularly mild medium in which to allow them to fold without precipitating. Using the mechanosensitive channel of E. coli MscL as a test protein, we have shown that HF- and F-TAC and F- and HF-Diglu do not interfere with in vitro protein synthesis, and that these two surfactants provide a suitable environment for folding and oligomerization of this pentameric channel (45a, 50).

228

Breyton, Pucci, and Popot

3.4.2. Inserting Membrane Proteins into Preformed Membranes

When working with an MP, in particular with channels or transporters, reinsertion into a lipid bilayer (a liposome, a black film, the membrane of a living cell, etc.) may be an important issue (e.g., for activity measurements). Using conventional detergents above their CMC, which is necessary to keep MPs soluble, is incompatible with maintaining the integrity of the target membrane. Reconstitution is usually carried out either by mixing the protein and lipids in a detergent solution and removing the detergent or by injecting the detergent-solubilized protein in the detergent-free buffer bathing the target membrane, which initiates a kinetic competition between protein aggregation and insertion. (H)FSs, being lyophobic, do not partition well into membranes, so that supramicellar concentrations of these surfactants can coexist with intact membranes: both F- and HF-TAC, at concentrations in excess of their CMC, cause no or little permeabilization of lipid vesicles, respectively, and neither of them partitions efficiently into lipid bilayers (32). They can therefore be used to deliver a MP to a preexisting membrane. Thus, the low-pH, membrane-active form of the diphtheria toxin T domain can be presented to preformed lipid vesicles as an (H)FS-stabilized, nonaggregating form (51). Such an unusual experimental configuration has opened the way to thermodynamic studies of MP insertion and the determination of the associated free energy of transfer (32, 39). Similarly, MscL solubilized in (H)F-TAC can be inserted into preformed liposomes for the sake of patch-clamp measurements (50).

3.4.3. Using Monolayers of (H)FSs as Surfaces onto Which to Organize Detergent-Solubilized Membrane Proteins

Conventional detergents do not penetrate or solubilize air–water monolayers of (H)FSs, which are stronger surfactants and with which they are poorly miscible. This made it possible to insert tOmpA, kept soluble by C8E4, into the aqueous space separating the two F-TAC monolayers of a Newton black film and to study the structure of this highly organized three-layer system by X-ray reflectivity measurements (52). The adsorption of detergentsolubilized MPs under monolayers of specially designed HFSs spread at the air–water interface may hold interesting promises for forming two-dimensional MP crystals (53).

3.4.4. Folding Membrane Proteins

The potential of (H)FSs as a mild environment in which to (re) fold MPs was investigated using BR as a convenient test protein (54). Preliminary tests showed that, after purple membrane had been denatured by sodium dodecyl sulfate (SDS), more than 50% refolding could be achieved in the presence of either F- or HF-TAC, following dodecyl sulfate precipitation by KCl and extensive dialysis (49). In the case of BR, a limiting factor may be the inefficient insertion of the retinal in the refolded apoprotein, which could be slowed by (H)FSs. tOmpA could be refolded into (H)FSs as well, albeit to a lesser extent than using conventional detergents (including H-TAC), with HF-TAC more efficient than

Amphipols and fluorinated surfactants

229

F-TAC (49). The usefulness of (H)FSs for MP folding is clearly worthy of further investigations. 3.5. Which Molecules to Choose?

The HFSs are tedious and expensive to synthesize and purify. Indeed, the chemical synthesis of the hemifluorinated tail is relatively difficult, requiring selective functionalization of 1,10-diiodo-1H,1H,2H,2H,9H,9H,10H,10H-perfluorodecane, which is used as the starting material. The limited availability and high cost of HFSs is a serious constraint for the biochemist since, for instance, it is difficult, if not impossible, to prepare the large volumes of solution containing the surfactant above its CMC that are required for chromatographic experiments. The proof of concept having been made, HFSs that can be obtained through a simpler route have been designed, and their synthesis is currently under development (Abla, Durand and Pucci). In the meanwhile, the question arises whether to favor the use of the more readily available FSs. Ten years ago, FSs did not seem that promising, given the prominent protein aggregation that had been observed in the first experiments (36). More recent data, however, indicate that this phenomenon is not general, and that it can be slow (37, 44, 45a). For applications for which long storage times are not necessary, FSs, which are easier and faster to synthesize than HFSs and therefore can be obtained in much larger amounts, can probably often be given the preference. Experiments requiring long-term monodispersity of protein–surfactant complexes, on the other hand, should probably resort to HFSs. As illustrated by observations with lipid bilayers (32) and with BR (as discussed in this chapter), the two classes of molecules are not always interchangeable, and one may easily imagine circumstances (e.g., for the sake of protein functionality or of membrane integrity) for which using one over the other will be preferable. Regarding the nature of the polar head, most data until now have been obtained with (H)F-TAC, whose properties and usefulness therefore are better characterized. In the longer term, however, it will probably become advantageous to turn to (H)FSs with a defined polar head, such as (H)F-Diglu, which do not present the potential drawbacks of polydispersity or the risks of batch-to-batch variations(88).

4. Amphipols 4.1. Rationale

As recalled, detergents and FSs are in a constant equilibrium between monomers, micelles, and the surfactant layer that covers the transmembrane region of the protein and makes it hydrophilic.

230

Breyton, Pucci, and Popot

To prevent MPs from aggregating, they must be handled above the CMC of the detergent (i.e., in the presence of free micelles). The idea behind the concept of amphipols was to devise molecules that would have such a high affinity for the surface of the protein that they would never dissociate or, at least, that their equilibrium distribution would be so strongly displaced in favor of the surface of the protein that traces of free surfactant in the solution would suffice to keep the protein soluble. As a result, the micellar phase, which in detergent solutions acts as a sink for stabilizing lipids and cofactors, would essentially vanish. An MP transferred to this new medium would therefore face no difficulty retaining its associated lipids, cofactors, and subunits, which would have, so to speak, nowhere else to go, and therefore would be strongly stabilized. Such a desirable result can a priori be achieved by resorting to surfactants that would carry not one or a few but a large number of hydrophobic chains. This was the initial idea that led us to devise “amphipols,” polymers that were so dubbed to distinguish them from the many other types of amphiphilic polymers used by physicists and in the industry. In the following, amphipol (APol) is used to refer to an amphipathic polymer that is able to keep individual MPs soluble under the form of small complexes. 4.2. Design, Synthesis, and Properties of A8-35

For an amphipathic polymer to form compact and stable complexes with an MP, it must be small (typically in the 5- to 20-kDa range), the distribution of its hydrophobic chains has to be dense (to endow it with high enough an affinity for the protein’s transmembrane surface and to prevent the formation of extended loops), its solubility in water must be high (to keep MPs soluble up to tens of grams per liter), and it must be highly flexible (to be able to adapt to the irregularities and small radius of curvature of MP transmembrane regions). Finally, it is also desirable that the synthesis of tens of grams be reasonably simple and not too costly. These considerations led to the design, in 1994, of a first series of APols (55). One of the members of this series, A8-35, has become by far the most extensively studied APol. A8-35 (55–57) is comprised of a relatively short polyacrylate chain (~70 acrylate residues), some of the carboxylates of which have been grafted with octylamine (~17 of them) or isopropylamine (~28 units) (Fig. 14.4). The approximately 25 acid groups that have remained free are charged in aqueous solutions (56), which makes the polymer highly water soluble, while the octylamide moieties, which are spaced, statistically, about every nanometer, render it highly amphipathic. Grafting with isopropylamide groups is not essential for making A8-35 an APol (the sister structure, A8-75, which does not comprise any such groups, is just as good when it comes to keeping MPs water soluble (55)), but it reduces the charge density along the chain, which seems to have a favorable

Amphipols and fluorinated surfactants

231

Fig.  14.4. Chemical structure of amphipol A8-35; x » 0.35, y » 0.25, and z » 0.4. The average number of acrylate units is about 70 and the average molecular mass about 9 kDa (From 55).

effect on the stability of the trapped proteins. The average molecular mass of a molecule of A8-35 is 9–10 kDa. Note that batches of A8-35 comprise a highly complex mixture of molecules, which differ from each other (1) in their overall length (because the polyacrylic acid that forms the backbone is itself polydisperse) and (2) in the distribution of lateral chains (which are grafted at random). Whether this heterogeneity is a favorable or an unfavorable feature from the biochemist’s and biophysicist’s points of view is a tangled issue whose discussion is beyond the scope of the present synopsis. The solubility of A8-35 in water is extremely high (>200 g/L). In a process that is rather unusual for amphipathic polymers, it forms, in aqueous solutions, well-defined, small, globular particles (57). Each A8-35 particle has a mass of about 40 kDa. It comprises, therefore, an average of about 4 individual molecules, for a total of about 80 octyl chains. The last number is close to that of hydrophobic chains in a typical detergent micelle. The critical aggregation concentration (that above which individual molecules start self-assembling) is less than 2 mg/L (58), meaning that, under most usual conditions, nearly all of the free polymer is assembled into particles. Because the solubility of A8-35 depends on the presence of charges on its carboxylate groups, it is affected by every factor that will affect the latter, such as lowering the pH or adding multivalent cations. If aggregation is to be avoided, it should not be used, therefore, below pH 7.0 (57, 59) or in the presence of millimolar concentrations of calcium (60). This can create difficulties for some applications, for which better-suited APols would be desirable (see section 4.6). On the other hand, these limitations seldom turn out to be a serious hindrance. Since it has proven quite difficult and lengthy to develop, validate, and characterize alternative chemical structures presenting as satisfying properties as those of A8-35, most of the detailed studies of APols and MP–APol complexes to date have been carried out on the latter. This provides the putative user with a considerable body of information about what can and what cannot be attempted using these particular molecules (for earlier reviews, see 5, 61).

232

Breyton, Pucci, and Popot

4.3. Trapping Membrane Proteins with A8-35 and Properties of the Resulting Complexes

Protein/detergent complex

As a rule, A8-35 by itself will not extract a MP from its native environment. Some exceptions have been encountered (reviewed in 61) and, when starting to work on a new MP, it may be worth checking whether it does not happen to be one of them. The standard procedure, however, is to solubilize the target protein in a classical detergent and then to transfer it, whether before or after purification, to the polymer. This is achieved by forming first a tertiary complex, which happens spontaneously in solutions containing both types of surfactants (see 62, 63, and references therein). The ratio of polymer to protein to be used varies depending on the molecular mass of the protein and the extension of its transmembrane surface. For small, deeply membrane-inserted proteins like BR or tOmpA, a good trapping ratio is typically around 4–5 g APol per gram protein (59, 64), while for a large complex with extended extramembrane domains like cytochrome bc1, it can drop to 1.5 g per gram (65). The protein will bind, ultimately, only part of the polymer added. The detergent is then removed by any convenient means (Fig. 14.5). A good trapping procedure is to start with a preparation with a concentration of detergent not too much above its CMC and, after adding the APol, to dilute the tertiary mixture below it. The monomerized detergent can then be removed (e.g. by adsorption onto polystyrene

Ternary complex

Protein/amphipol complex

Fig. 14.5. Transfer of a membrane protein from detergent solution to amphipols. A protein in detergent solution  is supplemented with amphipols , leading to the formation of ternary protein–amphipol–detergent complexes  and mixed amphipol–detergent particles. Polystyrene beads are added , onto which the detergent adsorbs. The protein– amphipol complexes formed  will retain any lipids that had remained associated with the protein in detergent solutions (see, e.g., 59, 70). The excess amphipol can be removed, if need be (e.g., by SEC (72) or by affinity chromatography (62)) (Adapted from 64).

Amphipols and fluorinated surfactants

233

beads, by dialysis, by cycles of dilution/concentration using ultrafiltration devices, by ultracentrifugation, by molecular sieving, or by any combination thereof (see, e.g., 59, 64, 65). An alternative way of forming an MP–A8-35 complex is to start from the protein in a denaturing environment, typically urea for b-barrel MPs and SDS for a-helical ones, and to (re)fold the protein by adding the APol and removing the denaturant (66, 67). In the final, detergent-free complexes, the polymer associates specifically with the transmembrane surface of the protein (64, 64a), where it forms a compact, relatively thin layer, 1.5- to 2-nm thick (59). This is slightly thicker than a detergent layer (68). For various technical reasons, measuring precisely the amount of protein-bound APol is difficult, particularly for small proteins (see 59,  62). Estimates vary from about 2 g per gram protein for a small, mainly transmembrane protein like BR (59) to about 0.13 g per gram for a large complex like cytochrome bc1 (65). This usually corresponds to a number of bound octyl chains that is significantly lower (by up to two to three times) than the number of classical detergent molecules that binds to the same protein in detergent solutions, suggesting a different arrangement of the hydrophobic chains relative to the surface of the protein in the two types of complexes (61). In most experiments, such as small-angle X-ray and neutron scattering or AUC, MP–A8-35 complexes behave like globular proteins (59, 65). On size exclusion chromatography, however, they seem to migrate as if they were somewhat bigger than they actually are (see 59, 64). When properly prepared, MP–A8-35 complexes are essentially monodisperse (59, 64, 65). It is, however, difficult to totally avoid the presence of minor fractions of small oligomers, which can seriously complicate, in particular, radiation-scattering experiments (59). Among the factors that can lead to aggregation are the use of too little APol at the trapping step (62), lowering the pH at or below neutrality (59), the presence of calcium ions (60), drifts from the nominal composition of the polymer (59, 60), and removal of the extra, free polymer that coexists with MP–APol complexes at the end of a trapping experiment (62). The latter phenomenon, which is reversible, is probably due to the poor dispersive power of APols: The presence of free polymer favors the formation of MP monomers, while its removal shifts the equilibrium toward protein autoassociation. Although few proteins have been tested, it seems that freezing MP–APol complexes can be not detrimental, while lyophilizing them is ((59), and unpublished data). Complexation by A8-35 will, as a rule, stabilize biochemically an MP as compared to detergent solutions (see, e.g., 55, 59–61, 67, 69, and references therein). Space limitations prevent a detailed discussion of the underlying mechanisms. Suffice it to say that they are multiple and include the following factors: limitation

234

Breyton, Pucci, and Popot

of the “hydrophobic sink”; retention of MP-associated lipids, cofactors, and subunits (which is likely to be a consequence not only of the former, but also of the poorly dissociating character of A8-35); and, most likely, damping of the dynamics of conformational excursions of transmembrane a-helix bundles, which limits opportunities for unfolding or aggregation (for discussions, see 60, 61). The functionality of APol-trapped MPs is generally preserved (see, e.g., 59, 61, 70, 71), as is their ability to bind ligands (reviewed in 72). However, MPs whose functional cycle involves large rearrangements of the surface of their transmembrane region, like the sarcoplasmic calcium adenosine triphosphatase (ATPase), may see their activity reversibly slowed or blocked, presumably because the adsorbed polymer damps such transconformations (for a discussion, see 60, 61). As mentioned, the last phenomenon (“Gulliver effect”) probably contributes to slowing MP denaturation (60). 4.4. Labeled or Functionalized Derivatives of A8-35

Radiolabeled APols have been synthesized to follow the distribution of the polymers in biochemistry experiments and to help assess the amount of APol bound per trapped MP. Experiments were carried out first with a 14C-labeled version of A8-75 (73) and later with 3H-labeled A8-35 (59). APols labeled with stable isotopes are precious for many biophysical experiments. A deuterated version of A8-35 has been synthesized (56) in which most of the nonexchangeable hydrogens, namely, those located in the octyl and isopropyl chains, have been replaced by deuterium. Deuterated A8-35 has been most useful in SANS experiments (it moves the contrast-matching point of the polymer far away from that of the protein) (56, 57, 59), in NMR studies (to remove most of the contribution of the surfactant from proton spectra, as well as to map the distribution of the polymer at the surface of the protein) (64, 74), and in AUC experiments (57, 59). At variance with detergents, APols are large molecules. A single functional group attached to the polymer is not expected to strongly affect its physical-chemical properties. This has been checked in comparative experiments using two versions of A8-35, one of which carried a fluorescent NitroBenzoxadiazolylDerythro NBD group (62). Other functionalized APols carry different fluorescent labels, such as fluorescein or rhodamine (58), or other functional groups, such as a biotin (72). Some of their uses are briefly described next.

4.5. Applications

The original aim behind the design of APols was to make it possible to handle and study in aqueous solutions MPs that are destabilized or whose functional properties are perturbed by detergents. While stabilization indeed seems fairly general, it may be acquired at the cost of making some experimental approaches more difficult, as is the case for NMR solution studies and, most

Amphipols and fluorinated surfactants

235

conspicuously, for crystallization. Before embarking on a given study, one has, therefore, to weigh the pros and the cons of resorting to APols. A detailed discussion of the advantages and constraints of using APols in different experimental circumstances is beyond the scope of the present chapter, but a quick overview is provided in Table 14.1, along with some key references. We mention here, briefly, only three examples of successful applications: ●● The nicotinic acetylcholine receptor (nAChR) of cholinergic synapses undergoes complex allosteric transitions on binding of the neurotransmitter, which result in the transitory opening of a transmembrane cation-specific channel. On solubilization with detergents, allosteric equilibria are perturbed. It has long been controversial whether this is due to the loss of the membrane environment or to the detergent binding to sites at the surface of the protein. Substitution of the detergent with APols has shown the latter to be true (70). ●●

●●

Folding MPs to their native state in vitro is both of considerable practical importance and precious to understand folding mechanisms. From a fundamental perspective, it provides insights into the respective roles of the amino acid sequence, the insertion machinery, and the membrane environment in determining the three-dimensional fold adopted by the polypeptide. From a practical point of view, overexpressing MPs as inclusion bodies and folding them in vitro may give access to producing large amounts of naturally rare MPs, which is difficult to achieve when trying to express them directly under their functional state. We have shown that APol A8-35 provides a highly efficient medium for folding MPs (66), including several GPCRs (67) (Fig.  14.6). Generalization of this strategy could constitute a major breakthrough for structural and functional studies of many MPs. From a fundamental point of view, these observations are consistent with all of the chemical information needed for these proteins to correctly fold being stored in their sequences (54). They also imply that, while shielding the folding protein from the aqueous solution is of course necessary, the highly specific and anisotropic environment provided by a lipid bilayer is in no way essential to decoding this information (67). Surface-based in vitro assays of ligand binding to proteins are appealing because they require minimum amounts of reagents, and they are suitable for multiplexing and high-throughput screening. An extremely valuable property of APols is that, as long as they are not displaced by another surfactant, their association with MPs, albeit noncovalent, is irreversible and resists extensive washing with surfactant-free buffer (61, 62, 72, 73). Trapping an MP with a functionalized APol therefore results in a permanently functionalized complex (62).

Rationale

A complex issue. Includes limitation of hydrophobic sink, preservation of MP–lipid interactions, and damping of transmembrane helix movements (see text).

Avoid perturbations or destabilization by detergents.

The mildness of APols, along with other factors, makes them an excellent environment to fold or refold denatured MPs, such as those produced as inclusion bodies.

APols ought to be particularly useful to study fragile MPs and MP supercomplexes.

Properly functionalized APols will simultaneously make an MP water soluble, stabilize it biochemically, and anchor it onto a solid support.

APols do not lyse target membranes (lipid vesicles or black films, cell plasma membrane) and can therefore be used to deliver to them hydrophobic cargoes such as MPs.

Type of application

Stabilization

Functional studies

Folding/refolding

Electron microscopy (EM)

MP immobilization onto solid supports

Delivery of MPs to preexisting membranes

MPs have been delivered both to lipid vesicles and to black films. Caveats: insertion process expected to be traumatic for fragile MPs; carrier APols will remain associated to the target membrane.

Hitherto applied to detecting antibody binding by surface plasmon resonance (on chips) and binding of a neurotoxin by fluorescence microscopy (on beads).

Better spread of particles in cryo-EM, a particularly promising application that still has to be fully explored.

To date applied to two porins, BR, and four GPCRs.

Ligand binding generally unperturbed. No perturbation of BR or nAChR functions. Slowing of calcium ATPase enzymatic cycle, probably due to damping of large-scale transmembrane conformational changes.

Most MPs tested to date are more stable in APols than in detergent solutions.

Observations

(66, 75)

(72) See Fig. 14.7

(59, 86, 87)

(66, 67) See Fig. 14.6

(59–61, 70–72)

(55, 59–61, 67, 69)

References

Table 14.1 A schematic overview of various validated or foreseeable applications of APols to MP studies and their current state of development. Examples of the rationales for using APols for the application considered, and a brief summary of current observations are given, and, if applicable, caveats are provided.

236 Breyton, Pucci, and Popot

Usual conditions for solution NMR of MPs are extremely aggressive (high detergent concentrations, high temperature); APols may permit study of MPs that do not stand them. A8-35 is easier to deuterate than most detergents.

Analysis of APol-trapped MPs.

Improving yields over that in detergent solutions?

Observing and studying MP assemblies that are too labile to resist exposure to detergents.

Letting MPs synthesized in vitro fold in a mild environment.

Stabilizing MPs under crystallization conditions.

Those go from following the distribution of the APol and its interaction with MPs or with target membranes or cells to FRET (Förster resonance energy transfer) experiments.

NMR

Mass spectrometry

Isoelectrofocusing

Trapping labile supercomplexes, assembly intermediates, etc.

Cell-free synthesis

X-ray crystallography

Applications of fluorescent APols

Used to study the irreversibility of the MP–APol association and its displacement by unlabeled APols or detergents, the critical association concentration of APols, etc.

Chemical structure of A8-35 far from ideal (charges). Current data limited to low-resolution diffraction by tertiary MP-APoldetergent complexes.

Under study.

Perhaps one of the most interesting applications of APols, but still underdeveloped. In progress.

Requires strictly neutral APols. Study in progress.

Doable. Study in progress.

MP–APol particles are slightly larger than the smallest MP–detergent complexes, which lowers the resolution. The greatest current limitation, however, is that A8-35 is usable only above pH 7.0. Full-fledged application of APols to solution NMR awaits the coming of age of pH-insensitive APols.

(58, 62)

(65)

(78a)

–

–

(74)

(64, 64a, 74, 78)

Amphipols and fluorinated surfactants 237

238

Breyton, Pucci, and Popot

Fig. 14.6. Amphipol-assisted folding of four G protein-coupled receptors. Folding in amphipol A8-35 (AP) or in a mixture of A8-35 and lipids (AP + L) is compared to folding in lipid–detergent mixtures (D + L). The receptors were obtained from inclusion bodies and solubilized in SDS under an inactive form. Folding in detergent was achieved by immobilizing on an affinity colum the SDS-solubilized receptors and substituting SDS with the detergent–lipid mixture (84). Folding in amphipols was achieved by supplementing the receptors in SDS solution with either A8-35 or a 5:1 (w/w) A8-35–lipid mixture and removing the SDS by precipitation with KCl followed by dialysis (66, 85). The extent of folding is given as the ratio between the number of binding sites for specific ligands and the number of receptor molecules initially present in the SDS solution, based on total protein content. (a) Folding of the leukotriene BLT1 receptor. The BLT1–A8-35 mass ratio was varied from 1:1 to 1:20, as indicated. (b) Folding of the serotonin 5-HT4(a), the leukotriene BLT2, and the cannabinoid CB1 receptors (Adapted from (67)).

This has innumerable possible applications. In particular, appropriately tagged APols can be used to immobilize the MPs with which they associate onto solid supports in a particularly mild and versatile manner. The ligand-binding properties of the immobilized proteins can then be studied, in detergent-free solutions, by any convenient observation method, opening the way to a host of applications in diagnostics, drug discovery, or the search for natural biological partners (72) (Fig. 14.7). A schematic overview of those applications of APols, proven or hypothetical, for which there exist published or unpublished data is given in Table 14.1. 4.6. Other Types of Amphipols

Putative APol users can certainly benefit, whenever circumstances permit, from using A8-35, whose properties and applications have been studied in most detail. For reasons that have been alluded to here, there are uses for which A8-35 is either suboptimal or plainly inadequate. To take just two examples: First, A8-35 is not the most convenient surfactant for NMR studies involving amide protons because MP–A8-35 complexes cannot be brought to the acidic pH that is usually optimal for such studies; second, APols with a net charge like A8-35 cannot be used for isoelectric focusing (IEF). There is therefore an incentive to develop alternative structures with, for such uses, more convenient properties. A word of caution is in order: While it is easy to draw dozens of

Amphipols and fluorinated surfactants

239

Fig. 14.7. Amphipol-mediated immobilization of membrane proteins onto solid supports and ligand-binding measurements. (a) Principle. A membrane protein in detergent  is trapped with an amphipol carrying an appropriate function . The complex is then immobilized, via the amphipol, onto a complementary surface . The support is washed and the binding of ligands to the protein  assayed in surfactant-free buffers by any convenient technique. (b) Binding of a fluorescently labeled snake venom toxin to the nicotinic acetylcholine receptor from Torpedo immobilized onto streptavidincovered beads via a biotinylated amphipol. Top, left to right: a field of beads seen in fluorescent light, a zoom on a region of the field, and the same region seen in visible light. Bottom: control experiments: left: toxin-binding sites on the immobilized receptor were blocked with unlabeled toxin prior to adding the fluorescent one; right: beads carrying only the biotinylated amphipol. Scale bars: 25 mm. (c) Quantitative analysis of the fluorescence of beads exposed to the fluorescent toxin: dark gray: beads carrying only the biotinylated amphipol; light gray: beads carrying the receptor. (d) Kinetics of toxin binding (Adapted from (72)).

structures of polymers expected to behave as APols (i.e., to trap MPs into small water-soluble complexes), identifying structures with the correct properties, characterizing these properties and those of MP–APol complexes, standardizing APol synthesis and purification, recognizing pitfalls, and developing new applications are extremely time-consuming undertakings. In the present section, we therefore list a number of interesting (or lessinteresting) attempts, while warning the potential user that, notwithstanding how much interest some of these new APols seem to present, how well they will perform in the field and what they are good for remain, to date, largely uncharted territories. In the initial stage of developing A8-35, it was shown that some changes could be brought to its chemical structure without compromising the ability of the resulting polymers to keep MPs soluble in the absence of detergents. This included longer polymers,

240

Breyton, Pucci, and Popot

with a molecular weight in the 35-kDa range (A34-35, A34-75) (55), shorter ones (unpublished data), polymers with a higher density of charges (A8-75, A34-75) (55), or polymers carrying dodecyl rather than octyl hydrophobic chains (unpublished data). Polymers similar to A8-75 but differing from it, mainly, by the distribution of the alkyl chains have also been proposed (75). One of the difficulties intrinsic to creating efficient APols is to combine a high density of hydrophobic chains with a high solubility, which imposes that strongly hydrophilic groups be part of the structure. The presence of charged moieties is an obvious answer to this constraint, but a net charge is not necessary and not necessarily optimal. Thus, polymers have been proposed that comprise zwitterionic groups and carry no net charge at basic pH (but, because the anionic group is a carboxylate, carry positive charges in acidic solutions) (71). Similarly, phosphorylcholine-based APols owe their high solubility to their globally neutral but highly charged polar heads as well as to the presence of a secondary amine function (63, 76, 77). Sulfonated APols are being developed with the view of mimicking the properties of A8-35 while remaining soluble at acidic pH, mainly for NMR purposes (60, 78). The search for totally nonionic APols has been a protracted endeavor due to the difficulty in making these molecules soluble enough without driving the formation of huge particles and complexes, which would be of no use for most applications. The feasibility of the concept was established relatively early, using the hydroxyl moieties of Tris groups to confer hydrophilicity (79), but the solubility of this family of molecules turned out to be too difficult to control for convenient use. Promising results have been obtained with glycosylated nonionic APols (80, 80a), which are currently under study. Finally, molecules hailed as “biocompatible” APols have been derived from a natural polymer, pullulane (81). On closer examination, they turned out not to be efficient at to keep MPs soluble (82). A possible reason for this failure could be the excessive persistence length (rigidity) of the polymer chosen as a scaffold. In most cases, however, the ability of these new APols to keep MPs soluble in the absence of detergents has been demonstrated, usually by simple centrifugation experiments. Few studies have been reported of the size and physical and biochemical stability of the complexes thus produced (see, e.g., 60, 77, 79). Some of these new molecules may become important tools, once their use has been validated, for specific applications for which A8-35 is suboptimal, such as MP solution NMR studies or crystallization, or cannot be used at all, such as IEF. Parallel studies comparing the properties of various families of APols and their ability to stabilize panels of MPs (see, e.g., 60, 77) are, of course, extremely desirable to enable the potential user to pick his or her choice in this widening family of molecules.

Amphipols and fluorinated surfactants

241

5. Conclusion It is a natural reaction, when investigating complex biological phenomena, to hesitate to resort to novel, unfamiliar chemical tools. True, they may help solve problems that are recurrently encountered when working with detergents. However, one may justifiably feel nervous at the prospect of exploring unknown depths using research tools that have not been validated by over a generation of physical-chemical and biochemical studies, as is the case for detergents. We feel, however, that with close to 15 years of accumulated experience — both projects were launched in the early 1990s — time is probably ripe for amphipols and FSs to progressively become what they were designed to be, not amusing curiosities, but routine tools at the service of all membrane biochemists and biophysicists.

Acknowledgments Particular thanks are due to J.-L. Banères, L.J. Catoire, D. Charvolin, T. Dahmane, F. Giusti, Y. Gohon, F. Lebaupain, K.L. Martinez, and M. Zoonens for suggestions about the manuscript and help with the figures. Thanks are also due to A. Polidori and G. Durand, from LCBOSMV, who were greatly involved in the synthesis and physical-chemical analysis of various FSs or non-ionic APols. This work has benefited, in particular, from funding by the EC BIO4-CT98-0269; STREP LSHG-CT-2005-513770 IMPS (Innovative Tools for Membrane Protein Structural Proteomics), the Human Frontier Science Program Organization (grant RG00223/2000-M), the Agence Nationale pour la Recherche PCV07 186241, and the CNRS (Interdisciplinary program Physique et Chimie du Vivant; DRITT grant ST 83747-04). References 1. Garavito RM, Ferguson-Miller S (2001) Detergents as tools in membrane biochemistry. J Biol Chem 276:32403–32406 2. Bowie JU (2001) Stabilizing membrane proteins. Curr Opin Struct Biol 11:397–402 3. Rosenbusch JP (2001) Stability of membrane proteins: relevance for the selection of appropriate methods for high-resolution structure determinations. J Struct Biol 136: 144–157

4. Gohon Y, Popot J-L (2003) Membrane proteinsurfactant complexes. Curr Opin Colloid Interface Sci 8:15–22 5. Sanders CR, Hoffmann AK, Gray DN, Keyes MH, Ellis CD (2004) French swimwear for membrane proteins. ChemBioChem 5: 423–426 6. Privé GG (2007) Detergents for the stabilization and crystallization of membrane proteins. Methods 41:388–397

242

Breyton, Pucci, and Popot

7. Brotherus JR, Jost PC, Griffith OH, Hokin LE (1979) Detergent inactivation of sodiumand potassium-activated adenosinetriphosphatase of the electric eel. Biochemistry 18: 5043–50 8. McIntosh RV, Cohen BB, Steel CM (1984) The use of detergents in velocity sedimentation of cell culture IgM. J Immunol Methods 74:59–64 9. Page MG, Rosenbusch JP, Yamato I (1988) The effects of pH on proton sugar symport activity of the lactose permease purified from Escherichia coli. J Biol Chem 263: 15897–15905 10. Lund S, Orlowski S, de Foresta B, Champeil P, le Maire M, Møller JV (1989) Detergent structure and associated lipid as determinants in the stabilization of solubilized Ca2+-ATPase from sarcoplasmic reticulum. J Biol Chem 264:4907–4915 11. Breyton C, Tribet C, Olive J, Dubacq J-P, Popot J-L (1997) Dimer to monomer transition of the cytochrome b6 f complex: causes and consequences. J Biol Chem 272: 21892–21900 12. Musatov A, Ortega-Lopez J, Robinson NC (2000) Detergent-solubilized bovine cytochrome c oxidase: dimerization depends on the amphiphilic environment. Biochemistry 39:12996–13004 13. Asmar-Rovira GA, Asseo-García AM, Quesada O, Hanson MA, Cheng A, Nogueras C, Lasalde-Dominicci JA, Stevens RC (2008) Biophysical and ion channel functional characterization of the Torpedo californica nicotinic acetylcholine receptor in varying detergent-lipid environments. J Membr Biol 223:13–26 14. Serrano-Vega MJ, Magnani F, Shibata Y, Tate CG (2008) Conformational thermostabilization of the b1-adrenergic receptor in a detergentresistant form. Proc Natl Acad Sci U S A 105: 877–882 15. Magnani F, Shibata Y, Serrano-Vega MJ, Tate CG (2008) Co-evolving stability and conformational homogeneity of the human adenosine A2a receptor. Proc Natl Acad Sci U S A 105:10744–49 16. Zhou Y, Bowie JU (2000) Building a thermostable membrane protein. J Biol Chem 275: 6975–6979 17. Warne T, Serrano-Vega MJ, Baker JG, Moukhametzianov R, Edwards PC, Henderson R, Leslie AGW, Tate CG, Schertler GFX (2008) Structure of a b1-adrenergic G proteincoupled receptor. Nature 454:486–491 18. Yu SM, McQuade DT, Quinn MA, Hackenberger CP, Krebs MP, Polans AS, Gellman SH (2000)

An improved tripod amphiphile for membrane protein solubilization. Protein Sci 9:2518–2527 19. Chae PS, Wander MJ, Bowling AP, Laible PD, Gellman SH (2008) Glycotripod amphiphiles for solubilization and stabilization of a membrane-protein superassembly: importance of branching in the hydrophilic portion. ChemBioChem 9:1706–1709 20. Theisen MJ, Potocky TB, McQuade DT, Gellman SH, Chiu ML (2005) Crystallization of bacteriorhodopsin solubilized by a tripod amphiphile. Biochim Biophys Acta 1751: 213–216 21. Sanders CR, Prosser RS (1998) Bicelles: a model membrane system for all seasons? Structure 6:1227–1234 22. Prosser RS, Evanics F, Kitevski JL, Al-AbdulWahid MS (2006) Current applications of bicelles in NMR studies of membrane-associated amphiphiles and proteins. Biochemistry 45:8453–8465 23. Nath A, Atkins WM, Sligar SG (2007) Applications of phospholipid bilayer nanodiscs in the study of membranes and membrane proteins. Biochemistry 46:2059–2069 24. Whorton MR, Bokoch MP, Rasmussen SGF, Huang B, Zare RN, Kobilka B, Sunahara RK (2007) A monomeric G protein-coupled receptor isolated in a high-density lipoprotein particle efficiently activates its G protein. Proc Natl Acad Sci U S A 104:7682–7687 25. Bayburt TH, Grinkova YV, Sligar SG (2006) Assembly of single bacteriorhodopsin trimers in bilayer nanodiscs. Arch Biochem Biophys 450:215–222 26. Schafmeister CE, Miercke LJW, Stroud RA (1993) Structure at 2.5 Å of a designed peptide that maintains solubility of membrane proteins. Science 262:734–738 27. McGregor C-L, Chen L, Pomroy NC, Hwang P, Go S, Chakrabartty A, Privé GG (2003) Lipopeptide detergents designed for the structural study of membrane proteins. Nat Biotechnol 21:171–176 28. Kissa E (1994) Fluorinated surfactants: synthesis, properties, applications. Marcel Dekker, New York 29. Mukerjee P (1994) Fluorocarbon – hydrocarbon interactions in micelles and other lipid assemblies, at interfaces, and in solutions. Colloid Surf A 84:1–10 30. Nakano TY, Sugihara G, Nakashima T, Yu SC (2002) Thermodynamic study of mixed hydrocarbon/fluorocarbon surfactant system by conductometric and fluorimetric techniques. Langmuir 18:8777–8785 31. Barthélémy P, Tomao V, Selb J, Chaudier Y, Pucci B (2002) Fluorocarbon-hydrocarbon

Amphipols and fluorinated surfactants

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

non-ionic surfactant mixtures: a study of their miscibility. Langmuir 18:2557–2563 Rodnin MV, Posokhov YO, Contino-Pépin C, Brettmann J, Kyrychenko A, Palchevskyy SS, Pucci B, Ladokhin AS (2008) Interactions of fluorinated surfactants with diphtheria toxin T-domain: testing new media for studies of membrane proteins. Biophys J 94:4348–4357 Zarif L, Riess JG, Pucci B, Pavia AA (1993) Biocompatibility of alkyl and perfluoroalkyl telomeric surfactants derived from THAM. Biomater Artif Cells Immobilization Biotechnol 21:597–608 Maurizis JC, Pavia AA, Pucci B (1993) Efficiency of non-ionic telomeric surfactants for the solubilization of subcellular fractions proteins. Bioorg Med Chem Lett 3:161–164 Polidori A, Pucci B, Pavia AA (1994) Nonionic glycosidic surfactants derived from tris(hydroxymethyl) amidomethane : synthesis, physical and biocompatibility data. New J Chem 18:839–848 Chabaud E, Barthélémy P, Mora N, Popot J-L, Pucci B (1998) Stabilization of integral membrane proteins in aqueous solution using fluorinated surfactants. Biochimie 80: 515–530 Lebaupain F, Salvay AG, Olivier B, Durand G, Fabiano A-S, Michel N, Popot J-L, Ebel C, Breyton C, Pucci B (2006) Lactobionamide surfactants with hydrogenated, hemifluorinated or perfluorinated tails: physical-chemical and biochemical characterization. Langmuir 22:8881–8890 Pavia AA, Pucci B, Riess JG, Zarif L (1992) New perfluoroalkyl telomeric non-ionic surfactants: synthesis, physicochemical and biological properties. Makromol Chem 193: 2505–2517 Posokhov YO, Rodnin MV, Das SK, Pucci B, Ladokhin AS (2008) FCS study of the thermodynamics of membrane protein insertion into the lipid bilayer chaperoned by fluorinated surfactants. Biophys J 95:L54–L56 Talbot J-C, Dautant A, Polidori A, Pucci B, Cohen-Bouhacina T, Maali A, Salin B, Brèthes D, Velours J, Giraud, M-F (2009) Hydrogenated and fluorinated surfactants derived from tris(hydroxymethyl)-acrylamidomethane allow the purification of a highly active yeast F1F0 ATP synthase with an enhanced stability. J Bioenerg Biomemb, in the press (DOI: 10.1007/s10863-009-9235-5) Barthélémy P, Améduri B, Chabaud E, Popot J-L, Pucci B (1999) Synthesis and preliminary assessment of ethyl-terminated perfluoroalkyl non-ionic surfactants derived from Tris(hydroxymethyl)acrylamidomethane. Org Lett 1:1689–1692

243

42. Breyton C, Chabaud E, Chaudier Y, Pucci B, Popot J-L (2004) Hemifluorinated surfactants: a non-dissociating environment for handling membrane proteins in aqueous solutions? FEBS Lett 564:312–318 43. Chaudier Y, Zito F, Barthélémy P, Stroebel D, Améduri B, Popot J-L, Pucci B (2002) Synthesis and preliminary biochemical assessment of ethylterminated perfluoroalkylamine oxide surfactants. Bioorg Med Chem Lett 12:1587–1590 44. Polidori A, Presset M, Lebaupain F, Améduri B, Popot J-L, Breyton C, Pucci B (2006) Fluorinated and hemifluorinated surfactants derived from maltose: synthesis and application to handling membrane proteins in aqueous solution. Bioorg Med Chem Lett 16:5827–5831 45. Abla M, Durand G, Pucci B (2008) Glucosebased surfactants with hydrogenated, fluorinated, or hemifluorinated tails: synthesis and comparative physical-chemical characterization. J Org Chem 73:8142–8153 45a. Breyton C, Gabel F, Abla M, Pierre Y, Lebaupain F, Durand G, Popot J-L, Ebel C, Pucci B (2009) Micellar and biochemical properties of (hemi)fluorinated surfactants are controlled by the size of the polar head. Biophys J 97:1077–1086 46. Israelachvili JN, Mitchell DJ, Ninham BW (1977) Theory of self-assembly of lipid bilayers and vesicles. Biochim Biophys Acta 470: 185–201 47. Henderson R (1975) The structure of purple membrane from Halobacterium halobium: analysis of the X-ray diffraction pattern. J Mol Biol 93:123–138 48. Dencher NA, Heyn MP (1978) Formation and properties of bacteriorhodopsin monomers in the non-ionic detergents octyl-b-d-glucoside and Triton X-100 FEBS Lett 96:322–326 49. Lebaupain F (2007) Développement de l’utilisation des tensioactifs fluorés pour la biochimie des protéines membranaires, Doctorat de l’Université Paris-7, Paris 50. Park K-H, Berrier C, Lebaupain F, Pucci B, Popot J-L, Ghazi A, Zito F (2007) Fluorinated and hemifluorinated surfactants as alternatives to detergents for membrane protein cell-free synthesis. Biochem J 403:183–187 51. Palchevskyy SS, Posokhov YO, Olivier B, Popot J-L, Pucci B, Ladokhin AS (2006) Chaperoning of membrane protein insertion into lipid bilayers by hemifluorinated surfactants: application to diphtheria toxin. Biochemistry 45:2629–2635 52. Petkova V, Benattar J-J, Zoonens M, Zito F, Popot J-L, Polidori A, Jasseron S, Pucci B (2007) Free-standing films of fluorinated

244

Breyton, Pucci, and Popot

surfactants as 2D matrices for organizing detergent-solubilized membrane proteins. Langmuir 23:4303–4309 53. Dauvergne J, Polidori A, Vénien-Bryan C, Pucci B (2008) Synthesis of a hemifluorinated amphiphile designed for self-assembly and two-dimensional crystallization of membrane proteins. Tetrahedron Lett 49:2247–2250 54. Huang K-S, Bayley H, Liao M-J, London E, Khorana HG (1981) Refolding of an integral membrane protein. Denaturation, renaturation, and reconstitution of intact bacteriorhodopsin and two proteolytic fragments. J Biol Chem 256:3802–3809 55. Tribet C, Audebert R, Popot J-L (1996) Amphipols: polymers that keep membrane proteins soluble in aqueous solutions. Proc Natl Acad Sci U S A 93:15047–15050 56. Gohon Y, Pavlov G, Timmins P, Tribet C, Popot J-L, Ebel C (2004) Partial specific volume and solvent interactions of amphipol A8–35. Anal Biochem 334:318–334 57. Gohon Y, Giusti F, Prata C, Charvolin D, Timmins P, Ebel C, Tribet C, Popot J-L (2006) Well-defined nanoparticles formed by hydrophobic assembly of a short and polydisperse random terpolymer, amphipol A8-35. Langmuir 22:1281–1290 58. Giusti F, Popot J-L, Tribet C (2010) Förster resonance energy transfer analysis of the assembly of amphipol particles (in preparation) 59. Gohon Y, Dahmane T, Ruigrok R, Schuck P, Charvolin D, Rappaport F, Timmins P, Engelman DM, Tribet C, Popot J-L, Ebel C (2008) Bacteriorhodopsin/amphipol complexes: structural and functional properties. Biophys J 94:3523–3537 60. Picard M, Dahmane T, Garrigos M, Gauron C, Giusti F, le Maire M, Popot J-L, Champeil P (2006) Protective and inhibitory effects of various types of amphipols on the Ca2+-ATPase from sarcoplasmic reticulum: a comparative study. Biochemistry 45:1861–1869 61. Popot J-L, Berry EA, Charvolin D, Creuzenet C, Ebel C, Engelman DM, Flötenmeyer M, Giusti F, Gohon Y, Hervé P, Hong Q, Lakey JH, Leonard K, Shuman HA, Timmins P, Warschawski DE, Zito F, Zoonens M, Pucci B, Tribet C (2003) Amphipols: polymeric surfactants for membrane biology research. Cell Mol Life Sci 60:1559–1574 62. Zoonens M, Giusti F, Zito F, Popot J-L (2007) Dynamics of membrane protein/ amphipol association studied by Förster resonance energy transfer. Implications for in vitro

studies of amphipol-stabilized membrane proteins. Biochemistry 46:10392–10404 63. Tribet C, Diab C, Dahmane T, Zoonens M, Popot J-L, Winnik, FM (2009) Thermodynamic characterization of the exchange of detergents and amphipols at the surfaces of integral membrane proteins. Langmuir, in the press, PMID: 19594168 64. Zoonens M, Catoire LJ, Giusti F, Popot J-L (2005) NMR study of a membrane protein in detergent-free aqueous solution. Proc Natl Acad Sci U S A 102:8893–8898 64a. Catoire LJ, Zoonens M, van Heijenoort C, Giusti F, Guittet E, Popot J-L (2009) Solution NMR mapping of water-accessible residues in the transmembrane β-barrel of OmpX. Eur Biophys J, in the press, PMID: 19639312. 65. Charvolin D, Picard M, Huang L-S, Courant J-C, Berry EA, Popot J-L (2010) Solution behavior and crystallization of cytochrome bc1 in the presence of amphipols (in preparation) 66. Pocanschi CL, Dahmane T, Gohon Y, Rappaport F, Apell H-J, Kleinschmidt JH, Popot J-L (2006) Amphipathic polymers: tools to fold integral membrane proteins to their active form. Biochemistry 45: 13954–13961 67. Dahmane T, Damian M, Mary S, Popot, J-L, Banères J-L (2009) Amphipol-assisted in vitro folding of G protein-coupled receptors. Biochemistry 48:6516–6521 68. le Maire M, Champeil P, Møller JV (2000) Interaction of membrane proteins and lipids with solubilizing detergents. Biochim Biophys Acta 1508:86–111 69. Champeil P, Menguy T, Tribet C, Popot J-L, le Maire M (2000) Interaction of amphipols with the sarcoplasmic Ca2+-ATPase. J Biol Chem 275:18623–18637 70. Martinez KL, Gohon Y, Corringer P-J, Tribet C, Mérola F, Changeux J-P, Popot J-L (2002) Allosteric transitions of Torpedo acetylcholine receptor in lipids, detergent and amphipols: molecular interactions vs physical constraints. FEBS Lett 528:251–256 71. Gorzelle BM, Hoffman AK, Keyes MH, Gray DN, Ray DG, Sanders CR II (2002) Amphipols can support the activity of a membrane enzyme. J Am Chem Soc 124: 11594–11595 72. Charvolin D, Perez J-B, Rouvière F, Giusti F, Bazzacco P, Abdine A, Rappaport F, Martinez KL, Popot J-L (2009) The use of amphipols as universal molecular adapters to immobilize membrane proteins onto solid supports. Proc Natl Acad Sci U S A 106:405–410

Amphipols and fluorinated surfactants 73. Tribet C, Audebert R, Popot J-L (1997) Stabilisation of hydrophobic colloidal dispersions in water with amphiphilic polymers: application to integral membrane proteins. Langmuir 13:5570–5576 74. Catoire LJ, Zoonens M, van Heijenoort C, Giusti F, Popot J-L, Guittet E (2009) Interand intramolecular contacts in a membrane protein/surfactant complex observed by heteronuclear dipole-to-dipole cross-relaxation. J Magn Res 197:91–95 75. Nagy JK, Kuhn Hoffmann A, Keyes MH, Gray DN, Oxenoid K, Sanders CR (2001) Use of amphipathic polymers to deliver a membrane protein to lipid bilayers. FEBS Lett 501:115–120 76. Diab C, Winnik FM, Tribet C (2007) Enthalpy of interaction and binding isotherms of nonionic surfactants onto micellar amphiphilic polymers (amphipols). Langmuir 23:3025–3035 77. Diab C, Tribet C, Gohon Y, Popot J-L, Winnik FM (2007) Complexation of integral membrane proteins by phosphorylcholinebased amphipols. Biochim Biophys Acta 1768:2737–2747 78. Dahmane T, Giusti F, Catoire LJ, Popot J-L 2010 Sulfonated amphipols: synthesis, properties and applications. In preparation 78a. Bazzacco P (2009) Non-ionic amphipols: new tools for in vitro studies of membrane proteins. Validation and development of biochemical and biophysical applications. Doctorat de l’Université Paris-7-Denis Diderot, Paris. 79. Prata C, Giusti F, Gohon Y, Pucci B, Popot J-L, Tribet C (2001) Non-ionic amphiphilic polymers derived from Tris(hydroxymethyl)acrylamidomethane keep membrane proteins soluble and native in the absence of detergent. Biopolymers 56:77–84 80. Sharma KS, Durand G, Giusti F, Olivier B, Fabiano A-S, Bazzacco P, Dahmane T, Ebel C, Popot J-L, Pucci B (2008) Glucose-based amphiphilic telomers designed to keep mem-

245

brane proteins soluble in aqueous solutions: synthesis and physicochemical characterization. Langmuir 24:13581–13590 80a. Bazzacco P, Sharma KS, Durand G, Giusti F, Ebel C, Popot J-L, Pucci B (2009) Grafted glucose-based amphipols: Synthesis, solution behavior, and biochemical properties. Submitted for publication 81. Duval-Terrié C, Cosette P, Molle G, Muller G, Dé E (2003) Amphiphilic biopolymers (amphibiopols) as new surfactants for membrane protein solubilization. Protein Sci 12: 681–689 82. Picard M, Duval-Terrié C, Dé E, Champeil P (2004) Stabilization of membranes upon interaction of amphipathic polymers with membrane proteins. Protein Sci 13:3056–3058 83. Pierre Y, Breyton C, Kramer D, Popot J-L (1995) Purification and characterization of the cytochrome b6 f complex from Chlamydomonas reinhardtii. J Biol Chem 270:29342–29349 84. Banères J-L, Martin A, Hullot P, Girard J-P, Rossi J-C, Parello J (2003) Structure-based analysis of GPCR function. Conformational adaptation of both agonist and receptor upon leukotriene B4 binding to recombinant BLT1. J Mol Biol 329:801–814 85. Popot J-L, Gerchman S-E, Engelman DM (1987) Refolding of bacteriorhodopsin in lipid bilayers: a thermodynamically controlled two-stage process. J Mol Biol 198: 655–676 86. Tribet C, Mills D, Haider M, Popot J-L (1998) Scanning transmission electron microscopy study of the molecular mass of amphipol/cytochrome b6 f complexes. Biochimie 80:475–482 87. Flötenmeyer M, Weiss H, Tribet C, Popot J-L, Leonard K (2007) The use of amphipathic polymers for cryo-electron microscopy of NADH:ubiquinone oxidoreductase (Complex I). J Microsc 227:229–235

Chapter 15 Heterologous Expression and Affinity Purification of Eukaryotic Membrane Proteins in View of Functional and Structural Studies: The Example of the Sarcoplasmic Reticulum Ca2+-ATPase Delphine Cardi, Cédric Montigny, Bertrand Arnou, Marie Jidenko, Estelle Marchal, Marc le Maire, and Christine Jaxel Abstract Heterologous SERCA1a Ca2+-ATPase (sarco-endoplasmic reticulum Ca2+-adenosine triphosphatase isoform 1a) from rabbit was expressed in yeast Saccharomyces cerevisiae as a fusion protein, with a biotin acceptor domain (BAD) linked to the SERCA C-terminus by a thrombin cleavage site. Thanks to the pYeDP60 vector, the recombinant protein was expressed under the control of a galactose-inducible promoter. Biotinylation of the protein occurred directly in yeast. Optimizing the number of galactose induction steps and increasing the amount of Gal4p transcription factor both improved expression. Lowering the temperature from 28 to 18°C during expression enhanced the recovery of detergent-extractible active protein. In the “light membrane fraction,” thought to mainly contain internal membranes, we are able to recover about 14–18 mg Ca2+-ATPase per liter of yeast culture in a bioreactor. Solubilization of this membrane fraction by n-dodecyl b-D-maltopyranoside (DDM) allowed us to recover the largest amount of active protein. The in vivo biotinylated recombinant protein was then bound to a streptavidin-Sepharose resin. Selective elution of the biotinylated SERCA1a was carried out after thrombin action on the resinbound protein. We were able to obtain 200–500 mg/L of yeast culture of a 50% pure SERCA1a that displays an ATPase activity similar to that of the native rabbit Ca2+-ATPase. To succeed in crystallization, an additional size exclusion chromatography step was necessary. This step increases purity to 70%, removes aggregated protein and exchanges DDM for C12E8. Key words: Biotin–streptavidin affinity, Ca2+-ATPase, heterologous expression, membrane protein, yeast, bioreactor.1 These three authors contributed equally to this work

1. Introduction We have worked out a procedure for affinity purification of rabbit SERCA1a (sarco-endoplasmic reticulum Ca2+-ATPase 1a) heterologously expressed in yeast cells, producing sufficient I. Mus-Veteau (ed.), heterologous Expression of Membrane Proteins, Methods in molecular Biology, vol. 601 DOI 10.1007/978-1-60761-344-2_15, © Humana Press, a part of Springer Science + Business Media, LLC 2010

247

248

Cardi et al.

amounts for ­crystallization and biophysical studies. The wildtype (WT) form gives a crystal that is isomorphous to the native SERCA1a protein from rabbit (1). The crystal structures of two mutant forms, D351A and P312A, as well as some of their functional characterization (2) have been described. Our procedure for heterologous expression and affinity purification (3) has also been applied to Plasmodium falciparum PfATP6, a SERCA-type protein (4). In fact, a few other eukaryotic membrane proteins have yielded high-resolution X-ray or electron diffraction data after recombinant expression (one in 2005, two in 2006, five in 2007; see review (5) and (6–9) for more recent data). The expression systems that were used in these other successful cases are diverse: Escherichia coli (one successful case); yeast, Pichia pastoris, or Saccharomyces cerevisiae (four cases); insect cells (three cases) or even mammalian cells (one case). The type of detergents or lipids used for solubilization or crystallization does not give obvious clues for best choices, except maybe that a mixture of lipid with the detergent is often present in the final crystallization steps. An emerging concept in the field is that it is important to aim at purifying a native protein together with improving the expression yield and the purity assessed on sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). In line with this idea of looking for an active protein, we believe that there was a key factor for the success in our case: In vivo yeast biotinylation of the biotin acceptor domain (BAD) added as a tag to SERCA1a was essential to purification of a small subfraction of fully active proteins. This procedure implies a positive selection of properly folded enzymes, a procedure similar to the one developed by Drew and coworkers (10), also successful. Although many parameters must always be optimized (choice of detergent, expression temperature, etc.) this type of approach could open the way for easier crystallization of this important class of protein. In our case, the optimization has been a long process, starting in 1994 with the expression of the protein (11), followed by its purification, first using a poly-His tag (12), then more recently using the BAD domain mentioned (1, 3). Here, we present our recent advances in the scale-up of the expression protocol using a 20-L bioreactor to follow and to control precisely the yeast culture parameters. This new protocol made it possible to increase both the yeast density and the proportion of expressed protein compared to results obtained in glass vessels. In fact, the yields in glass vessels are more variable than in the bioreactor and depend on various parameters such as the volume of yeast cultures. In general, the new protocol results in an increase of the total expression per liter of culture by a factor of 4.

Affinity Purification of P-type ATPases for Structural Studies

249

2. Materials 2.1. Yeast Culture

All our culture media were from Becton Dickinson Biosciences (Gibco BRL, Fisher Scientific Bioblock). All chemical products were purchased from Sigma-Aldrich except if noted otherwise. 1. YPD (A) medium: 1% (w/v) bactopeptone, 1% (w/v) yeast extract, 1% (w/v) glucose, 20 mg/mL adenine. A 10% overconcentrated medium containing only bactopeptone and yeast extract is prepared. It is sterilized for 20 min at 120°C. Glucose and adenine are added after cooling from, respectively, 20% (w/v, see 15.2.1.6) and 10 mg/mL (see 15.2.1.8) stock solutions. 2. S1 buffer: 100 mM Lithium acetate, 40% polyethylene glycol (PEG) 4000, 100 mM dithiothreitol (DTT). 3. S6 (A) minimum medium: 0.1% (w/v) bactocasamino acids, 0.7% (w/v) yeast nitrogen base, 2% (w/v) glucose, 20 mg/ mL adenine: A 10% overconcentrated medium containing only bactocasamino acids and yeast nitrogen base is prepared. It is sterilized for 20 min at 120°C. Glucose and adenine are added after cooling from, respectively, 20% (w/v, see 15.2.1.6) and 10 mg/mL (see 15.2.1.8) stock solutions. 4. S6 (A) agar minimum medium: 0.1% (w/v) bactocasamino acids, 0.7% (w/v) yeast nitrogen base, 2% (w/v) glucose, 20 mg/mL adenine, 1.5% agar. Medium is prepared as described for S6 (A) minimum medium with agar present from the beginning of the preparation. 5. S5 (A) minimum medium: 0.1% (w/v) bactocasamino acids, 0.7% (w/v) yeast nitrogen base, 2% (w/v) galactose, 20 mg/ mL adenine. The sterilization for 20 min at 120°C is done before the addition of galactose and adenine. Medium is prepared as described for S6 (A) minimum medium. 20% (w/v) galactose is prepared as described in 15.2.1.7. 6. 20% (w/v) glucose stock solution is sterilized by filtration onto a 0.22-mm Steritop filter (Millipore S.A.) and is added in the various media after sterilization and cooling at 40°C. 7. 20% (w/v) galactose stock solution is sterilized by filtration onto a 0.22-mm Steritop filter (Millipore S.A.) and is added in the various media after sterilization and cooling at 40°C. 8. A solution of 10 mg/mL adenine is prepared in 0.1N HCl and filtrated onto a 0.22-mm Steritop filter (Millipore S.A.) and is added in the various media after sterilization and cooling at 40°C. 9. A 50% (w/v) trichloroacetic acid (TCA) stock solution is prepared from powder in Milli-Q water. It can be stored for some months

250

Cardi et al.

at 4°C. A working solution of 2% (w/v) TCA is prepared just before use by dilution of the 50% stock solution in water. 10. Glass beads (0.5-mm diameter, Biospec Products) are stored at 4°C. They are reusable after decontamination with fungicidal solution, intensive rinsing with water, and sterilization with ethanol. Beads are dried by incubation in a Pasteur oven before storage at 4°C. 11. The incubator is a Multitron incubator shaker system (INFORS HT), and the bioreactor is a Techfors-S apparatus (INFORS HT) coupled to a recirculating cryostat (FE1100, Julabo). 12. YPGE2X rich medium: 2% (w/v) Bacto™ peptone, 2% (w/v) yeast extract, 1% (w/v) glucose, and 2.7% (v/v) ethanol. The medium without glucose and ethanol is sterilized directly in the bioreactor. The final volume takes into account the 10% water evaporation occurring during the sterilization process. 13. A highly concentrated solution of galactose at 50% (w/v) is prepared by heating the solution in a microwave. 14. Antifoam A concentrate is used. 2.2. Membrane Preparation

  1. Yeast are centrifuged in bottles (polycarbonate bottles with screw-cap assemblies, Beckman Coulter), each inwardly covered with a liner (harvestLine™ system liners, Beckman Coulter). JLA8.1000, JLA10.500, 45Ti, and Type 19 rotors used are from Beckman Coulter. 2. Milli-Q water: 4 L are stored at 4°C. 3. TEKS buffer: 50 mM Tris-HCl, pH 7.5, 1 mM ethylenediaminetetraacetic acid (EDTA), 0.6M sorbitol, 0.1M KCl; 1 L is prepared beforehand and stored at 4°C. 4. TES buffer: 50 mM Tris-HCl, pH 7.5, 1 mM EDTA, 0.6M sorbitol; 2 L are prepared beforehand and stored at 4°C. 5. Phenylmethanesulfonylfluoride (PMSF) is dissolved to 0.2M in ethanol; 10 mL are prepared beforehand and could be stored at 4°C for some weeks. 6. HEPES-sucrose buffer: 20 mM HEPES-Tris, pH 7.5, 0.3M sucrose (Labosi, Elancourt, France), 0.1 mM CaCl2. 7. Reagents for protein assay: Bicinchoninic acid (BCA commercial stock solution) and Copper(II) sulfate (CuSO4 ­commercial stock solution at 4% (w/v)).

2.3. Solubilization of SERCA1a

  1. Presolubilization buffer: 50 mM MOPS-Tris, pH 7.0, 100 mM KCl, 20% glycerol (w/w), 1 mM CaCl2, 1 mM PMSF, and complete EDTA-free antiprotease cocktail (in our hands, it was sufficient to use half of the recommended amount, i.e., 1 tablet for 100 mL) (Roche Diagnostics) plus possibly 1 mM b-mercaptoethanol or 1 mM DTT.

Affinity Purification of P-type ATPases for Structural Studies

2.4. Purification of SERCA1a by Affinity Chromatography

251

  1. Streptavidin-Sepharose™ high-performance resin from G.E. Healthcare (Courtaboeuf, France). 2. Solubilization buffer: 50 mM MOPS-Tris, pH 7.0, 100 mM KCl, 20% glycerol (w/w), 1 mM CaCl2 and 15 mg/mL DDM (Anatrace) (the DDM amount depends on protein concentration; see Section 15.3.4). 3. High-salt (HS) washing buffer: 50 mM MOPS-Tris, pH 7.0, 1M KCl, 20% glycerol (w/w), 1 mM CaCl2 and 0.5 mg/mL DDM. 4. Low-salt (LS) washing buffer: 50 mM MOPS-Tris, pH 7.0, 100 mM KCl, 20% glycerol (w/w), 2.5 mM CaCl2 and 0.5 mg/mL DDM. 5. Thrombin Bovine High Activity (VWR International). 6. Filter tube after thrombin reaction is a Handee™ centrifuge column (Pierce).

2.5. Purification of SERCA1a by Size Exclusion Chromatography

  1.  Centricon Ultracel-YM 30000 Da are from Amicon, Millipore. 2. Resin is a 0.78 × 30 cm TosoHaas TSK-gel G3000SWXL column mounted on a Gold HPLC System (Beckman Coulter). 3. Size exclusion chromatography (SEC) buffer: 100 mM MOPS-Tris, pH 6.8, 80 mM KCl, 1 mM CaCl2,1 mM MgCl2, 15% sucrose, and 0.5 mg/mL C12E8 (Nikko Chemicals). 4. DOPC (1,2-dioleoyl-sn-glycero-3-phosphocholine, Avanti Polar Lipids) is dissolved at 40 mg/mL in chloroform; then, chloroform is evaporated, and a thin layer of N2-dried DOPC is formed in a tube.

2.6. SDS-PAGE

  1.  Mini-Protean 3 cell system is used (Bio-Rad). 2. Separating gel solution: 375 mM Tris-HCl, pH 8.8; 8% acrylamide (w/v); 0.1% Sodium Dodécyl Sulfate (SDS, w/v); 0.1% ammonium persulfate (w/v); 0.06% N,N,N¢,N¢Tetramethylethylenediamine (TEMED, v/v). 3. Stacking gel solution: 125 mM Tris-HCl, pH 6.8; 5% acrylamide (w/v); 0.1% SDS; 0.1% ammonium persulfate; 0.1% TEMED (v/v). 4. Denaturing buffer 2X-concentrated: 100 mM Tris-HCl, pH 8, 1.4M b-mercaptoethanol, 4% SDS (w/v), 1 mM EDTA, 8M urea, 0.05% bromophenol blue (w/v). 5. Running buffer: 25 mM Tris-HCl, 250 mM glycine, 0.1% SDS. 6. Coomassie blue staining solution: 0.05% Coomassie blue R, 40% methanol, 10% acetic acid.

252

Cardi et al.

2.7. Western Blotting for SERCA1a

  1. Mini-Trans-Blot Mini-Protean 3 cell system cell is used (Bio-Rad). 2. CAPS buffer: 10 mM 3-(Cyclohexylamino)-1-propanesulfonic acid (CAPS), 10% methanol, pH 11.1 adjusted with NaOH. Prepare 1 L per electrophoresis tank (for one or two blots, Mini-Protean 2 or 3 systems). It is stored at 4°C with stirring for 2–3 h for cooling. 3. The polyvinylidene-difluoride (PVDF) membranes are from  Millipore and chromatography paper from ATGC Biotechnologies. 4. Precision Plus Protein all blue Standards (PPPS) is a marker solution (Bio-Rad). 5. PBS-T buffer: Phosphate buffer, pH 7.4 (10 mM KH2PO4, 90 mM K2HPO4), 100 mM NaCl, 0.2% Tween-20. 6. The anti-SERCA1a primary antibody is either (79B), a guinea pig antibody, a gift from Dr. A.-M. Lompré (INSERM U 621, Paris, France), or a mouse monoclonal anti-SERCA1a (IIH11 clone). 7. The secondary antibody is either an anti-guinea pig goat antibody conjugated to horseradish peroxidase (HRP) purchased from TEBU Bio, coupled to the 79B antibody, or a goat antimouse immunoglobulin G1 (IgG1) (g) (Caltag Laboratories), coupled to the monoclonal anti-SERCA1a antibody. 8. Avidin peroxidase conjugate is a commercial solution (SigmaAldrich) frozen for single-use (5-mL) aliquots at −70°C. 9. Enhanced chemiluminescent (ECL) reagents are from G.E. Healthcare, and detection is made by using a G:BOX camera (Syngene distributed by Ozyme).

3. Methods To express SERCA1a in yeast, we use plasmid pYeDP60 initially given by Denis Pompon (13) and described after modification in (1, 3, 12) (see Fig. 15.1). It contains a bacterial part to amplify the DNA in bacteria and a yeast part for vector replication and for expression of the recombinant protein. For the selection of transformed yeast, the plasmid contains two auxotrophy markers to allow the synthesis of uracil and adenine. The gene of interest is under the transcriptional control of a hybrid promoter containing the inducible part of GAL10 (promoter of the galactose catabolism in the baker yeast) and the strong promoter of CYC1 (strong promoter of iso-1-cytochrome c). The S. cerevisiae strain used for the expression is W303.1b Gal4-2 (a, leu2, his3, trp1::TRP1-GAL10-GAL4, ura3, ade2-1, canr, cir+) and is auxotrophous to uracil and adenine. This strain has a low constitutive expression level of GAL4p transactivation

Affinity Purification of P-type ATPases for Structural Studies

253

Fig. 15.1. Plasmid used for the heterologous expression of SERCA1a-BAD. Cloning and amplification in Escherichia coli (light grey): Ori bact. bacterial replication origin; AmpR gene coding for b-lactamase to allow resistance to ampicillin (selection marker). Amplification in yeast (dark grey): ADE2 auxotrophy selection marker for adenine; URA3 auxotrophy selection marker for uracil; 2m yeast replication origin. Expression (white): GAL10-CYC1 fusion promoter of the inducible part of GAL10 and RNA polymerase binding part of CYC1; PGK phosphoglycerate kinase terminator sequence; SERCA1a, coding sequence for Ca2+-ATPase; Thr sequence coding for thrombin cleavage site; BAD biotin acceptor domain. Serca1a, Thr, and BAD sequences were cloned in the same coding frame to allow expression of the fusion protein SERCA1a-BAD.

factor (one to three copies of molecules per cell). It has been genetically modified to induce expression of that transactivator, under the control of a GAL10 promoter. In presence of glucose, the system is repressed by constitutive binding of an inhibitor to GAL4p. When glucose is entirely consumed and galactose is added to the medium, Gal4p constitutive copies are released, and it can activate the GAL10 promoter. This event first allows high synthesis of transactivator from the yeast genome and second induction of the expression of the protein of interest from the plasmid. This expression is induced when a sufficient concentration of yeast has been obtained and for a duration defined to avoid toxicity (see 15.3.1.3.6 and 15.3.1.3.8). Finally, the gene coding SERCA1a is cloned in the same open reading frame of a sequence coding a C-terminal BAD separated by a sequence coding a thrombin cleavage site. This way, the fusion protein, which is expressed, can be biotinylated directly in yeast to allow its purification by affinity chromatography onto a streptavidin-Sepharose resin (14, 15). The direct cleavage by thrombin on the resin allows simultaneous elution of SERCA1a and removal of the tag. For an improvement of the purity of the protein preparation, a second step of purification by SEC is added. For all these expression and purification steps, the protein is detected on polyacrylamide gel revealed by Coomassie blue staining

254

Cardi et al.

(purity) or by Western blot stained either by a specific antibody (that recognizes SERCA1a) or by an avidin probe coupled to HRP, which binds to the biotin already linked to BAD. In our case, the enzymatic activity of SERCA1a cannot be easily detected directly on yeast membrane but is nicely estimated using the purified protein. To broaden this procedure, several DNA vectors have been prepared. The aim is to allow the purification of other membrane proteins. For some of them, the position of the tag on the N-terminal side may be desirable. 3.1. Yeast Culture 3.1.1. Yeast Transformate

3.1.2. Selection of Individual Clones in Minimum Medium

First, intact yeast cells are transformed with pYeDP60-modified plasmids for the expression of SERCA1a by the protocol developed by Chen et al. (16) using the LiAc/single-stranded DNA/ PEG method (17). After an overnight culture of yeast at 28°C, in a shaking incubator at 200 rpm, in YPD (A) medium, 5.107 cells, in stationary phase, are pelleted 5 min at 4°C and 3,000gav. The pellet is gently resuspended on ice in cold mix containing 90 mL of S1 buffer, 5 mL of denatured single-stranded carrier DNA (10 mg/mL), and about 1 mg of plasmid DNA. Yeasts were then heat shocked at 45°C for 45 min in a water bath to help DNA to enter into the competent cells and transferred onto S6 (A) agar plates and incubated at 28°C for 48 h. This minimum medium is deprived of uracil. As S. cerevisiae W303.1b Gal4-2 is auxotrophous for uracil, only those transformed with vectors derived from pYeDP60 will have the ability to grow on this minimum medium since this plasmid will complement the ura3 genotype of this yeast.   1. A colony streaked onto an S6 (A) agar minimum medium storage plate (see Note 1) is toothpicked into 5 mL of S6 (A) minimum medium and grown at 28°C for 24 h in a shaking incubator at 200 rpm (see Note 2). 2. For each assay, 500 mL of the previous S6 (A) precultures are centrifuged 5 min at 4°C and 1000gav and resuspended in 5 mL of S5 (A) minimum medium to induce expression. These cultures are incubated at 28°C for 15–18 h in a shaking incubator at 200 rpm. 3. For each culture, 5 OD600nm are centrifuged 5 min at 4°C and 8,000gav in 2-mL tubes. The pellets are washed with water, resuspended, and centrifuged 5 min at 4°C and 1,000gav. 4. The supernatants are removed, and the pellets are resuspended in 400 mL of cooled 2% TCA. Glass beads are added to reach the meniscus. The suspensions are mixed with a vortex at maximal speed for 8 min at room temperature to break the cells. 5. The tubes are then placed on ice, and glass beads are sedimented. The supernatant is removed by pipetting, and that crude extract is kept on ice.

Affinity Purification of P-type ATPases for Structural Studies

255

6. Glass beads are washed three times with 400 mL of 2% TCA, and these three collected supernatants are gathered with the crude extract. 7. The resulting solution is kept 15 min on ice for protein precipitation. Then, the samples are centrifuged 15 min at 4°C and 30,000gav. 8. The pellets are resuspended in 100 mL of 100 mM Tris-HCl, pH 7.5 (see Note 3). These samples are analyzed by Western blotting to choose the best clones. These may be stored, starting from the S6 (A) minimum medium precultures, either on storage plates or in 20% glycerol samples stored at −70°C. 3.1.3. Large-Scale Expression of SERCA1a

The procedure is described for 6 L of yeast culture in a bioreactor, but that protocol may be easily adapted for 20 L. 1. For large-scale expression, a colony streaked onto an S6 (A) agar minimum medium storage plate is toothpicked into 5 mL of S6 (A) minimum medium and is grown at 28°C for 24 h in a shaking incubator at 200 rpm (the cell density reaches saturation: OD600nm = 3, corresponding to 9–15.107 cells/mL). 2. This culture is used to inoculate 40 mL of S6 (A) minimum medium and is grown at 28°C for 24 h in a shaking incubator at 200 rpm (density also reaches saturation; see Note 4). 3. This second culture is used to inoculate 400 mL of S6 (A) minimum medium and grown at 28°C for about 10 h in a shaking incubator at 200 rpm. In the meantime, 6 L of YPGE2X medium without ethanol are preequilibrated at 28°C and saturated with air under high aeration (1 volume of air per volume of medium per minute, i.e., 1 vvm) and agitation at 300 rpm for 12 h. 4. The bioreactor is inoculated with the culture of 400 mL at exponential phase (OD600nm = 0.6 or 2.107 cells/mL) (see Note 5); the final density inside the fermenter must be OD600nm = 0.03 (corresponding to 106 cells/mL). Add 0.0001% of foam suppresser (Antifoam A), and the culture is performed at 28°C, 300 rpm, and 1 vvm. 5. At 15–18 h after inoculation, ethanol is added (see Note 6). A regulation sequence between the rate of agitation and the aeration is activated to maintain the dissolved dioxygen up to 20%. This fixed parameter allows maintaining yeast in aerobiosis. In that condition, continuous measurement of the dissolved dioxygen, using a probe, induces automatic regulation of stirring, from 300 to 1,000 rpm and air flow between 6 and 18 L/min (i.e., 1 to 3 vvm in the case of a 6-L culture). 6. At 34 h after inoculation (OD600nm = 30 corresponding to 9–15.108 cells/mL), the aeration is reduced to 2 L/min, and

256

Cardi et al.

the agitation is lowered to 300 rpm (see Note 7) to maintain cells in anaerobiosis. At the same time, the temperature of the medium is lowered to 18°C to slow the yeast metabolism and consequently favor the folding of expressed membrane proteins. We wait for one generation time (~2 h) so that the culture can adapt to the new conditions. 7. At 36 h after inoculation, 2% galactose is added from a fresh 50% stock solution to induce the expression of recombinant proteins. 8. At 49 h after inoculation, a second induction is triggered by addition of 2% galactose, and the culture is pursued for 5 h under the same conditions. 9. After 54 h, the culture is stopped by decreasing the temperature to 6°C. The stirring is maintained to 300 rpm to avoid yeast sedimentation in the bioreactor before harvesting (see Fig. 15.2). The procedure is described for 6 L of yeast culture (about 350–400 g of yeast).

dO2 > 20%

50 Optical density at 600 nm

Expression at 18°C

6°C

no O2 input

40

+Gal

+Gal

9

100

8

80

30

7

20

6

40

10

5

20

4

0

0 0

10

20

30

40

50

60

dO2, %

Growing phase at 28°C

pH

3.2. Membrane Preparation

60

Time, h

Fig. 15.2. Follow-up of culture parameters during expression process. 6 L of YPGE2X medium were inoculated with yeast strain containing vector for SERCA1a-BAD overexpression as described in Methods. Cell density is estimated by optical density measurements at 600 nm using a Novaspec spectrophotometer (closed circles and continuous line). Dissolved dioxygen concentration (dO2, dark and short dashed line), pH (long dashed line), and temperature are measured continuously thanks to a probe connected to the recording system. dO2 and pH are used as real-time reporter of yeast metabolism. First phase going from inoculation to 34 h: Yeast growth is optimized by maintaining dissolved dioxygen up to 20% by automatic control of air flow rate and stirring. Second phase corresponding to expression phase: Temperature is lowered at 18°C to slow cell division. Expression is induced by two successive galactose additions (grey arrows at 36 and 49 h). Last phase after 54 h: Growth and expression are suddenly stopped by decreasing temperature to 6°C and stirring at 300 rpm.

Affinity Purification of P-type ATPases for Structural Studies

257

1. At the end of the fermentation, yeast are recovered directly in three 1-L centrifugation bottles using the sampling/harvesting valve of the bioreactor. They are centrifuged for 9 min at 4°C and 4,500gav, and the supernatants are removed. The bottles are filled and centrifuged again with the remaining yeast until the bioreactor is empty. The same liners are used during all steps 1–4; rubber tubing is used to pour out each centrifugation supernatant without the need to cut it, and the resuspensions of the pellets are easily carried out directly in the liners by hand kneading. 2. The yeast pellets are washed with resuspension in 2.1 L (700 mL per liner, the volume equivalent to five times the mass of yeast) of cooled Milli-Q water and centrifuged 9 min at 4°C and 4,500gav in a JLA8.1000 rotor. After removing the supernatants, the pellets are weighed. Usually, about 50 g of yeast/L are obtained. 3. The pellets are resuspended in 700 mL (in fact, 240 mL per liner, corresponding to about 2 mL/g of yeast) of TEKS buffer. The solution is incubated 15 min at 4°C and then centrifuged 9 min at 4°C and 4,500gav in a JLA8.1000 rotor. 4. The pellets are resuspended in the same volume of TES buffer (700 mL, corresponding to about 2 mL/g of yeast) and supplemented with 1 mM PMSF and 1 tablet of complete EDTA-free antiprotease cocktail per 100 mL (for an alternate protocol, stop membrane preparation here; see Note 8). The pH is controlled; it must be at 7.0–7.5 (see Note 9). 5. The yeast suspension is distributed in several 500-mL centrifugation plastic bottles (about 80 mL per bottle). Cold glass beads are added until they reach the meniscus (about 450 g of beads per bottle; bottles should not be more than half full to improve breaking). The bottles are tightly closed with screw caps and Parafilm to avoid leaks. Beads are released from the bottom side of the bottles by manual agitation, and the bottles are placed horizontally on the tray of an INFORS Multitron incubator and fixed with adhesive tape. Agitation is then performed for 20 min at 350 rpm (orbital agitation) and the room temperature set at 17°C to break yeast membranes. The culture will warm from 4°C, and the temperature has to be less than 12°C at the end of shaking to prevent protein degradation. Otherwise, the breaking step can be done in successive phases of shaking-cooling. All subsequent steps are performed at 4°C in a cold room. 6. For recovering the crude extract from the bottles, a vacuum pump is connected to a filtering flask as a trap (3-L Erlenmeyer), which is itself connected to a 25-mL pipet (without cotton plug). This home-made pump system allows the efficient sucking of the crude extract from the beads into the trap.

258

Cardi et al.

7. The glass beads are washed in three successive steps with 1.1 L (three times 40 mL per bottle [and 10 bottles], corresponding to about three times the mass of yeast) of the TES buffer supplemented with 1 mM PMSF and complete EDTA-free antiproteases cocktail. The last supernatant should be clear, indicating that the beads are properly washed. The pH of the total crude extract (the first extract plus three washes will make up about 1.5 L) is controlled; it must be at 7.0–7.5. Otherwise, pH is adjusted to 7.5 by adding 10N NaOH. 8. The total crude extract is centrifuged in 500-mL bottles for 15 min at 4°C and 1,000gav in a JLA10.500 rotor. The pellet (P1), containing non- and partially broken cells, is discarded. 9. The first supernatant (S1) is centrifuged in 500-mL bottles for 15 min at 4°C and 18,000gav in a JLA10.500 rotor. The second supernatant (S2) is removed with great care, and the pellet (P2), containing heavy membranes, is resuspended in about 15 mL of the TES buffer (with 1 mM PMSF and complete EDTA free) per bottle and saved for further analyses. 10. This S2 supernatant is then centrifuged in 250-mL bottles for 5 h at 10°C and 29,000gav in a type 19 rotor. The third supernatant (S3) is removed. 11. The pellet (P3), containing light membranes, is resuspended in 70 mL of the HEPES-sucrose buffer (i.e., 0.2 mL/g of yeast). The membrane suspension is homogenized using a Potter-type mixing system and is conserved at −70°C after fast cooling in liquid nitrogen (see Fig. 15.3). 3.3. Estimation of Total Membrane Protein Concentration

Estimation of the membrane protein concentration is performed by BCA assay (18), which has been adapted in a microplate format. The active reagent is prepared just before use to avoid premature oxidation by mixing BCA and 4% CuSO4 with a 50/1 (v/v) ratio. Each sample (20 mL of the sample to be quantified, 12 mL of water and 8 mL of 10% SDS) is compared to a reference curve obtained by using a concentration range of bovine serum albumin (BSA) samples (from 1 to 8 mg/ mL) in a total volume of 40 mL. To be valuable, the BSA has to be diluted in the same buffer as the sample to be quantified (i.e., TES or HEPES-sucrose buffer) and has to contain a final concentration of 2% (w/v) SDS. After adding 200 mL of BCA active reagent, each sample is incubated for 30 min at 37°C, and absorbance at 540 nm is measured. This method is preferentially used for the estimation of membrane proteins because it is more robust in the presence of detergent; in fact, presence of SDS is strictly necessary to allow complete removal and denaturation of the membranous parts of proteins. At the

Affinity Purification of P-type ATPases for Structural Studies

259

Fig. 15.3. Membrane preparation. (a) Principle of the fractionation. TCE total crude extract; S1 and P1 supernatant and pellet from first centrifugation, respectively; S2 and P2, supernatant and pellet from second centrifugation, respectively; S3 and P3, supernatant and pellet from third centrifugation, respectively. (b) Western blot analysis of various aliquots from the membrane preparation: Samples (1 mg of total proteins) were analyzed by SDS-PAGE. For the SR samples, they were added to 1 mg of P3 membrane prepared from yeast transformed with the empty expression vector. Proteins were transferred onto a PVDF membrane and immunostained with a specific antibody directed against SERCA1a (79B) or revealed with avidin peroxidase probe (AP), which can bound to all the biotinylated proteins. SR SERCA1a extracted from rabbit sarcoplasmic reticulum; AcoA. carb. Acetyl-CoA carboxylase; pyr. carb. pyruvate carboxylase.

end, the protein concentration of the light membranes (P3) is about 40 mg/mL with a content of about 3% recombinant Ca2+-ATPase. This ratio is estimated from Western blots. 3.4. Washing of Light Membranes and Solubilization of Membrane Proteins (See Fig. 15.4)

  1. Light membranes P3 are thawed using a water bath at 20°C and diluted at 5 mg/mL of total proteins in the presolubilization buffer (Fig. 15.4, lane Tw). To prevent lipids and proteins from oxidation, 1 mM b-mercaptoethanol or DTT can be added (see Note 10). This washing step is optional. However, we observe that the level of expression of yeast endogenous biotinylated proteins (acetyl-CoA [coenzyme A] carboxylase, pyruvate carboxylase and Arc1p) was remarkably increased during the expression phase, as can be expected from yeast metabolic pathway adaptation in anaerobiosis. Those soluble or extrinsic proteins were dragged with the P3 pellet and represent the major contaminants for affinity chromatography. This washing step by simple dilution enables removing most of those proteins (see Note 11).

Cardi et al.

a

kDa

MWM

260

SR SR T w S w T DM S DM FT HS LS R 0 R 30 ’ R 60 ’ Res E 1 E 2 Co 0.1 0.5

250 150 -

_ S1a -BAD

100 -

¯ S1a

75 50 -

b 150 -

_ S1a -BAD

100 -

¯ S1a

c

- ACoA Carb. 150 100 -

_ Pyr . Carb. ¯ S1a -BAD

Fig. 15.4. Purification of SERCA1a by streptavidin affinity chromatography. In all cases, 20 mL of the sample to be tested are mixed with 20 mL of denaturing buffer 2X-concentrated, and 20 mL of this mix are loaded onto the gel. (a) Coomassie blue stained SDS-PAGE. MWM molecular weight marker; TW diluted membrane fraction in presolubilization buffer; SW supernatant of membrane washing containing eliminated proteins; TDM membrane pellet resuspended in presolubilization buffer supplemented with DDM; SDM solubilized fraction; FT flowthrough; HS washing at 1M KCl; LS washing at 50 mM KCl; R0 – R30’ – R60’, resin samples taken at time 0, 30, and 60 min, respectively, during thrombin cleavage step; Res resin after elution of SERCA1a; E1 and E2, first and second elutions, respectively; Co elution pool concentrated on YM-30 membranes (diluted 20-fold); SR0.1 and SR0.5, 0.1 and 0.5 mg native rabbit sarcoplasmic reticulum Ca2+-ATPase (SR), respectively, used as standards. (b) Western blot and immunodetection using anti-SERCA1a antibody. Samples from TW to LS fractions were tenfold, resin and eluted fractions samples 50-fold, and “Co” fraction 1,000-fold diluted in LS buffer, respectively. After electrophoresis, proteins were transferred to PVDF membranes, and SERCA1a was detected thanks to the use of mouse monoclonal anti-SERCA1a (IIH11 clone) as a primary antibody and goat antimouse IgG1 (g) as a secondary antibody. Intensities of bands are compared to standard amount of 10 and 50 ng of SR for quantitative estimation. (c) Western blot and detection using avidin-HRP probe. Previous membranes are stripped by incubation in a 2% SDS, 1 mM EDTA, 0.1M b-mercaptoethanol, and 50 mM Tris-Cl pH 6.8 buffer for 30 min. After extensive washing with PBS-T buffer, avidin-peroxidase probe is added. In this case, only biotinylated proteins are detected; the purified SERCA1a protein is not visible. SR extracted from rabbit sarcoplasmic reticulum; S1a-BAD SERCA1a fused to BAD; S1a SERCA1a; ACoA Carb acetyl-CoA carboxylase; Pyr.Carb pyruvate carboxylase.

2. Diluted membranes are centrifuged for 1 h at 10°C and 100,000gav in a 45Ti rotor. The supernatant is discarded (Fig. 15.4, lane Sw). 3. Solubilization is performed at the final ratio DDM/total proteins of 3:1 (w/w). The membrane pellet is resuspended using a Potter-type homogenizer in the presolubilization buffer at a

Affinity Purification of P-type ATPases for Structural Studies

261

final concentration of 10 mg/mL. At the same time, detergent is dissolved in the presolubilization buffer at a final concentration of 30 mg/mL and equilibrated for 1 h at 20°C before use. Then, both solutions (proteins and detergent) are mixed and kept at 20°C under gentle agitation for 5 min. Resulting concentrations are about 5 and 15 mg/mL, respectively, for total proteins and DDM. This solution is homogenized using a Potter-type homogenizer (~10 up and down). The solution becomes clearer (Fig. 15.4, lane TDM). 4. Finally, the solution is centrifuged for 1 h at 10°C and 100,000gav in a 45Ti rotor. The obtained supernatant contains the solubilized proteins (Fig. 15.4, lane SDM). 3.5. Batch Purification of SERCA1a by Affinity Chromatography (See Fig. 15.4)

The purification is carried out with 1 mL of resin for 4 mg of total Ca2+-ATPase contained in the P3 fraction. Note that it might be useful to adjust this ratio according to the relative amount of the other biotinylated proteins compared to the level of expression of the protein of interest. Indeed, those proteins will compete for binding onto the streptavidin resin (see Note 12). In our hands, this ratio seems to be optimal for wild-type (WT) SERCA1a, but it was not the case for some mutants and other isoforms. We use a batch procedure throughout; each step consists of gently mixing the resin with the appropriate solution (5 min), centrifuging for 15 min at 4°C and 1,000gav in a swinging bucket rotor (11133 from Sigma). 1. The resin, stored in ethanol by the provider, has to be washed with 10 resin volumes of Milli-Q water and equilibrated twice with 10 resin volumes of solubilization buffer. 2. This equilibrated resin is then added to the solution of solubilized proteins, and the whole solution is gently stirred at 4°C overnight. However, thanks to the high affinity of streptavidin for biotin, this step may also be carried out for 2 h at 20°C. All subsequent steps are done in 50-mL centrifuge tubes. 3. The resin is collected in 50-mL centrifuge tubes by successive centrifugations; each tube is a.  Filled with the suspension from step 15.3.5.2 b.  Centrifuge 5 min at 4°C and 1,000gav c. The supernatant is removed, and these three steps are repeated until each tube contains about a maximum volume of 5 mL resin. All supernatants are pooled, and this fraction is named the flowthrough (Fig. 15.4, lane FT). 4. The resin is washed twice with 10 resin volumes of HS washing buffer to remove unspecific bound proteins (Fig.  15.4, lane HS).

262

Cardi et al.

5. The resin is washed twice with 10 resin volumes of LS washing buffer to reduce the KCl concentration before elution of SERCA1a and to increase the calcium concentration required for the thrombin cleavage (Fig. 15.4, lane LS). 6. Add 1 volume of LS buffer (kept at 20°C) to the resin; this suspension is stirred 5 min on a wheel to equilibrate the temperature at 20°C (Fig. 15.4, lane R0). 7. Add 25 units of thrombin per milliliter of resin to the solution. They are stirred slowly to prevent foaming on the wheel for 30 min at 20°C (Fig. 15.4, lane R30). 8. Add 25 more units of thrombin per milliliter of resin, and the incubation is performed in the same conditions (Fig.  15.4, lane R60). 9. The thrombin cleavage is stopped by adding 1 mM PMSF and a complete EDTA-free antiprotease cocktail. The solution is then stirred 5 min at 20°C on the wheel and, at the end, is transferred into a filter tube. 10. SERCA1a is finally removed from the resin by centrifugation for 2 min at 10°C and 1,000gav (Fig. 15.4, lanes E1 and E2). This protocol allows us routinely to produce milligrams of pure Ca2+-ATPase depending on the level of expression for mutants and obviously on the amount of membranes used at the start of the purification. Note that this eluted fraction can be directly used for functional assays like ATPase activity measurements. For storage, it is frozen in liquid nitrogen and conserved at −70°C after adjusting the glycerol concentration to 40%. 3.6. Detergent Exchange by Size Exclusion Chromatography

DDM is exchanged with C12E8 by SEC, an additional step that also improves the purity of the Ca2+-ATPase preparation and the delipidation of the sample. This step is optional for functional analysis but is essential for structural studies such as crystallogenesis trials. 1. SERCA1a eluted from affinity chromatography resin (typically 30 mL) is first concentrated on Centricon Ultracel-YM 30 until the volume is smaller than 500 mL. 2. This concentrate is applied at 0.5 mL/min into a silica gel filtration column (0.78 × 30 cm TosoHaas TSK-gel G3000SWXL column), mounted on a Gold HPLC System, and equilibrated with 100 mM MOPS-Tris pH 6.8, 80 mM KCl, 1 mM CaCl2,1 mM MgCl2, 15% sucrose, and 0.5 mg/ mL C12E8. 3. The eluted fractions containing the Ca2+-ATPase peak are pooled (about 1.5 mL total volume) and concentrated on the Centricon-30 to a final volume of about 50 mL. The final protein concentration is about 10 mg/mL, and C12E8 is estimated

Affinity Purification of P-type ATPases for Structural Studies

263

to be about 17–20 mg/mL on the basis of previous C12E8 binding data (19). 4. The concentrated, delipidated, and C12E8-solubilized Ca2+ATPase is supplemented with phospholipid DOPC to reach a C12E8/DOPC ratio of 3:1 (w/w). This is done by adding the protein sample in a centrifuge tube already deposited with a thin layer of N2-dried DOPC, gently vortexing the tube, and storing it overnight at 4°C for equilibration. 5. Finally, the sample is subjected to centrifugation at 120,000gav in a TLA-100 rotor for 20 min to remove aggregated (inactive) proteins just before the crystallization assays (see Fig. 15.5). 3.7. SDS-PAGE

  1. Gels are prepared according to the Laemmli procedure (20). The 0.75-mm gels at 8% acrylamide are prepared by mixing 1 mL of 1.5M Tris-HCl pH 8.8 with 1.07 mL of 30% acrylamide/bis solution (29/1), 1.85 mL of water, 40 mL of 10% SDS, 40 mL of 10% ammonium persulfate, and 2.4 mL of TEMED (ammonium persulfate and TEMED are added just before pouring) for a total of 4 mL corresponding to one gel. The gel (3.2 mL) is poured, leaving space for a stacking gel, and overlaid with water. The gel should polymerize in about 30 min. 2. The stacking gel is prepared by mixing 250 mL of 1M TrisHCl pH 6.8 with 330 mL of 30% acrylamide/bis solution, 1.38 mL of water, 20 mL of 10% SDS, 20 mL of 10% ammonium persulfate, and 2 mL of TEMED (ammonium persulfate and TEMED are added just before pouring) for a total of 2 mL

Fig. 15.5. SEC-HPLC purification step for crystallization attempts. The figure depicts the elution profile at 280 nm (in arbitrary units, Arb. U.) from size exclusion chromatography of SERCA1a after affinity chromatography. The absorption peaks corresponding to aggregates, monomeric SERCA1a, or thrombin are indicated by arrows. The inset shows (A) the SDS-PAGE profile of SERCA1a after SEC-HPLC (high-performance liquid chromatographic) purification (SERCA1a is indicated by an arrow) and (B) crystals obtained using the protein analyzed in (A). Crystals were obtained by the vapor diffusion method in hanging drops using PEG 2000 as the precipitant and ter-butanol as additive.

264

Cardi et al.

corresponding to one gel. The water is removed, the amount of stacking gel allowing at least 1 cm below the wells is poured, and the comb is inserted. The gel should polymerize in about 30 min at 20°C. 3. The running buffer is added to the upper and lower chambers of the gel unit. 4. For SDS-PAGE, samples from membrane preparation and purification are diluted in the corresponding LS washing buffer, mixed with an equal volume of denaturing buffer 2X concentrated, and immediately heated at 100°C for 1 min to denature proteases, which could be harmful for our proteins in the presence of SDS. Note that the presence of urea is crucial for avoiding the aggregation of membrane proteins (21). The samples are then cooled and loaded onto the gels. The gel is run at 120 V and 25 mA for 1 h. 5. Proteins are stained with Coomassie brilliant blue; the gel is incubated 10 min in the Coomassie blue staining solution and then rinsed several times (three to four times for some seconds) with water. After removing the excess of the staining solution, the gel in water is destained by using a microwave oven (about 1 min at 900 W; stop it just before boiling). Destaining is improved by adding a piece of absorbent paper in the hot water and by shaking gently the gel until complete disappearance of the background. 3.8. Western Blotting for SERCA1a

  1. Samples that have been separated by SDS-PAGE can be transferred to PVDF membrane. Prestained standard (PPPS) can be used to check transfer efficiency. The transfer is done in CAPS buffer. This buffer allows better electrophoretic mobility of the proteins. In those conditions, we are able to transfer all the proteins (MW about 20–200 kDa) from an 8% Laemmli type gel in less than 1 h at 250 mA and about 80–90 V instead of many hours in classical Tris-glycine buffer. Nevertheless, one has to be careful to proceed in presence of an ice cooling unit in the tank to prevent warming, which can modify the mobility of proteins. 2. In the case of an immunodetection, the membranes are blocked for 10 min at 20°C in presence of PBS-T supplemented with 5% dry milk. 79B anti-SERCA1a guinea pig antibody (which recognizes the N-terminal part of SERCA1a) is added in the blocking media at 1/20,000 dilution for 1 h. The membrane is rinsed once using PBS-T. Anti-guinea pig antibody coupled to HRP is then added at 1/10,000 in PBS-T and incubated for 1 h. In the case of the other couple of antibodies, the mouse monoclonal anti-SERCA1a (IIH11 clone) is diluted 1/2,500, and the goat antimouse IgG1 (g) is diluted 1/2,000. In case of low protein expression, a better

Affinity Purification of P-type ATPases for Structural Studies

265

signal-to-noise ratio can be obtained by heating the PVDF membrane in SDS/b-mercaptoethanol (22). 3. In the case of detection of biotinylated proteins by an avidinperoxidase probe, the procedure is slightly different; it is indispensable to avoid incubation of the probe with dry milk because of the presence of biotinylated proteins in it. So, after the first step consisting of blocking membrane with PBS-T supplemented with 5% dry milk, we wash it three times for 10 min with PBS-T to remove milk. The probe is added at a 1/50,000 dilution in PBS-T and incubated for 1 h at 20°C. 4. For the two systems of detection, membranes are washed three times for 10 min with PBS-T buffer. The bound antibodies are revealed with ECL reagents, and emitted light is detected using a CCD (charge coupled device) cooled camera. To quantify the amount of SERCA proteins, the intensities of the bands are compared to a range of native Ca2+-ATPase isolated from the rabbit sarcoplasmic reticulum (SR; SR membranes provided by Dr. Philippe Champeil). In the case of membranous solutions, the comparison may suffer from the presence of many other membrane proteins. To be more relevant, a similar amount of P3 membrane prepared from yeast transformed with the empty expression vector is added to the SR range. This addition is particularly important when the detection system is not sensitive and when the relative amount of expressed proteins is low.

4. Notes 1. The plates have to be prepared less than 1 week before. Otherwise, an additional preculture must be done. 2. For each clone, several colonies are checked to choose the best one. 3. These samples may be stored as aliquots (15 mL) at −70°C. 4. Both cultures are performed in Erlenmeyer-type flasks containing a volume of at least ten times the medium volume to allow an optimal aeration of the yeast cultures. 5. This procedure allows obtaining cells in a synchronous state and at exponential phase for the following inoculation in the bioreactor. 6. A chiller is connected to cool the effluents and thus limit the ethanol evaporation. 7. The anaerobiosis stimulates the production of networks of internal membranes, which is convenient for storing overexpressed membrane proteins.

266

Cardi et al.

8. At this step, the pellets may be resuspended in only one volume V of TES buffer with 1 mM PMSF and complete EDTAfree antiprotease cocktail corresponding to 1 time the weight of yeast (350 mL for 6 initial liters of yeast culture) and stored at −70°C after fast freezing in liquid nitrogen. In this case, the solution is rapidly thawed in water in a bath water at 20°C, and one additional volume V of TES buffer with 1 mM PMSF and complete EDTA-free antiprotease cocktail is added before the breaking of the yeast. 9. The pH may be lower if yeast cells are altered by an efflux of lysosomal contents. 10. The membrane protein solution can be diluted in the range of 2–10 mg/mL. Nevertheless, note that you will need to be careful to adjust the subsequent centrifugation time as a function of the final concentration. 11. Note that KCl concentration may be increased to 600 mM to improve the removal of extrinsic proteins from the membranes. However, this treatment may somewhat induce protein proteolysis (12]. 12. Note that it might be of interest to verify the biotinylation level of the protein beforehand.

Acknowledgments We would like to thank Dr. Philippe Champeil for the gift of SR membranes and for discussions. This work was supported by grants from the Commissariat à l’Energie Atomique (CEA) program Signalisation et transport membranaires and by the Agence Nationale de la Recherche (grant ANR-06-BLAN-0239-01). References 1. Jidenko M, Nielsen RC, Sørensen TL-M, Møller JV, le Maire M, Nissen P, Jaxel C (2005) Crystallization of a mammalian membrane protein overexpressed in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A 102:11687–11691 2. Marchand A, Winther AML, Holm PJ, Olesen C, Montigny C, Arnou B, Champeil P, Clausen JD, Vilsen B, Andersen JP, Nissen P, Jaxel C, Møller JV, le Maire M (2008) Crystal structure of D351A and P312A mutant forms of the mammalian sarcoplasmic reticulum Ca2+ATPase reveals key events in phosphorylation and Ca2+ Release. J Biol Chem 283: 14867–14882

3. Jidenko M, Lenoir G, Fuentes JM, le Maire M, Jaxel C (2006) Expression in yeast and purification of a membrane protein, SERCA1a, using a biotinylated acceptor domain. Protein Expr Purif 48:32–42 4. Cardi D. et al. In preparation 5. Midgett CR, Madden DR (2007) Breaking the bottleneck: eukaryotic membrane protein expression for high-resolution structural studies. J Struct Biol 160:265–274 6. Ferguson AD, McKeever BM, Xu S, Wisniewski D, Miller DK, Yamin TT, Spencer  RH, Chu L, Ujjainwalla F, Cunningham BR, Evans JF, Becker JW (2007) Crystal ­structure of inhibitor-bound

Affinity Purification of P-type ATPases for Structural Studies

7.

8. 9.

10.

11.

12.

13.

14.

human 5-­lipoxygenase-activating protein. Science 317:510–512 Jasti J, Furukawa H, Gonzales EB, Gouaux E (2007) Structure of acid-sensing ion channel 1 at 1.9 Å resolution and low pH. Nature 449:316–323 Sprang SR (2007) Structural Biology: A receptor unlocked. Nature 450:355–356 Pedersen BP, Buch-Pedersen MJ, Morth JP, Palmgren MG, Nissen P (2007) Crystal structure of the plasma membrane proton pump. Nature 450:1111–1114 Drew D, Newstead S, Sonoda1 Y, Kim H, von Heijne G, Iwata S (2008) GFP-based optimization scheme for the overexpression and purification of eukaryotic membrane proteins in Saccharomyces cerevisiae. Nat Protocol 3:784–798 Centeno P, Deschamps S, Lompré A-M, Anger M, Moutin MJ, Dupont Y, Palmgren MG, Møller JV, Falson P, le Maire M (1994) Expression of the sarcoplasmic reticulum Ca2+ATPase in yeast. FEBS Lett 354:117–122 Lenoir G, Menguy T, Corre F, Montigny C, Pedersen PA, Thinès D, le Maire M, Falson P (2002) Over-production in yeast and rapid and efficient purification of the rabbit SERCA1a Ca2+-ATPase. Biochim Biophys Acta Biomembranes 1560:67–83 Pompon D, Louerat B, Bronine A, Urban P (1996) Yeast expression of animal and plants P450s in optimized redox environments. Methods Enzymol 272:51–64 Cronan JE Jr (1990) Biotination of proteins in  vivo. A post-translational modification to label, purify, and study proteins. J Biol Chem 265:10327–10333

267

15. Pouny Y, Weitzman C, Kaback HR (1998). In vitro biotinylation provides quantitative recovery of highly purified active lactose permease in a single step. Biochemistry 37:15713–15719 16. Chen DC, Yang BC, Kuo TT (1992) Onestep transformation of yeast in stationary phase. Curr Genet 21:83–84 17. Gietz D, St Jean A, Woods RA, Schiestl RH (1992) Improved method for high efficiency transformation of intact yeast cells. Nucleic Acids Res 20:1425 18. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner MD, Provenzano MD, Fugimoto EK, Goeke NM, Olson BJ, Klenk DC (1985) Measurement of protein using bicinchoninic acid. Anal Biochem 150: 76–85 19. Møller JV, le Maire M (1993) Detergent binding as a measure of hydrophobic surface of integral membrane proteins. J Biol Chem 268:18659–18672 20. Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680–685 21. Soulié S, Møller JV, Falson P, le Maire M (1996) Urea reduces the aggregation of membrane proteins on sodium dodecyl sulfatepolyacrylamide gel electrophoresis. Anal Biochem 236:363–364 22. Fuentes JM, Lompré A-M, Møller JV, Falson P, le Maire M (2000) Clean Western blots of membrane proteins after yeast heterologous expression following a shortened version of the method of Perini et  al. Anal Biochem 285:276–278

Index A Amidoblack 0.1% solution......................................... 91, 99 Amphipols design, synthesis and properties of A8-35........................................230–231, 240 labeled or functionalized derivatives of A8-35....................................................... 234 Analyzing yield and activity of the expressed membrane protein................................. 126–130

reactions in Maxi-CECF-Reactors.........177–178, 182, 184 reactions in Mini-CECF-Reactors...................170, 176, 177, 179, 184 template design for E. coli extract based CF expression....................................... 171–172 T7-RNA polymerase preparation..............171, 173–174 Co-expression with the b-subunit of the ATP synthase..................... 52, 53, 55, 60, 61, 63 Cytochrome b6   f.......................................................224, 226

D

B Bacteria E.coli BL21 Origami.................................. AD494, 55, 57 BL21(DE3)-pLys-S....................................... 55, 57 BL21 Rosetta (DE3)...............................55, 57, 169 BL21 Star................................................55, 57, 169 C43(DE3)...........................................52, 54, 55, 57 MC1061..............................................22, 23, 27–29 Lactococcus lactis nisin-producing strain L. lactis NZ9700...............................................24, 31, 68 NZ9000 strain................................................ 68, 70 Bac to Bac baculovirus-based expression system....................................................... 6, 107 Baculovirus expression vector system (BEVS)..................................................... 6, 106 Baculovirus infected insect cells...............6, 11, 12, 105–117 Blocking buffer...................................... 72, 91, 99, 100, 159

C Cell disruption................................................25, 32, 36, 41, 43, 71, 80, 169, 183 Cell free expression modes for membrane protein production............................ 8, 165–185 D-CF expression of membrane proteins.......... 180–182 solubilization of P-CF expressed membrane proteins............................... 179–180 Cell free system analytical scale CECF reactions in microcentrifuge tubes....................... 175–177 E. coli CECF reaction set-up.................................... 178 preparation of E. coli extracts.............168–169, 174–175 reaction container set-up...................169–171, 175–178

Detergent analysis...........................................206, 214–215 Detergent choice for structural studies................... 190–192 Detergent exchange by size exclusion chromatography.................................... 262–263 Detergents C8E4.......................................................................... 192 C12E8..........................................................189, 190, 221 decylmaltoside...........................................190, 191, 194 dimetyldodecylamineoxide (LDAO).............9, 192, 194 dodecyl-º-D-maltoside (DDM)........................... 9, 169 laurylmaltoside......................................................... 190 octyl-º-D-glucopyranoside (OG)..................9, 211, 212 Detergent titration................................................. 212–214 Drosophila Schneider 2 (S2) cells bioreactor S2 cell culturing................120, 122, 124, 125 culturing S2 cells...................................................... 124 freezing S2 cells................................................ 121, 124 induction of recombinant protein expression in S2 cells.............................................. 126–127 S2 cells lysis and membrane preparation.................. 128 selection of stable S2 cell lines...................121, 126–128 suspension culture of S2 cells............................ 125, 151 transfection for the generation of stable S2 cell lines................................................... 126

E E. coli inclusion bodies.................................................. 3, 40 Electroporation of RNA......................................... 155–157 Electrotransformation of L. lactis.......................................24 Engineering membrane proteins for improved crystallisability...................................... 199–201 Establishing stable fly strains.................................. 141–143 Evaluation of gene expression......................................... 160 Expression in fly eyes.......................................... 7, 135–146

269

Heterologous Expression of Membrane Proteins 270  Index

  

Expression in insect cells effect of chaperon NinaA on BCRP expression............................................. 115–116 expression of BCRP in Sf9 and to Hi5 insect cells..................................................... 116 optimization of BCRP expression in insect cells......................................... 115–116 scale up in spinners................................................... 116

F Fluorescence measurements................................... 123, 129 Fluorinated surfactants C6F13C2H4-S-poly-Tris-(hydroxymethyl) aminomethane)............................................. 224 C8F17C2H4-S-poly-Tris-(hydroxymethyl) aminomethane)............................................. 224 C2H5C6F12C2H4-S-poly-Tris-(hydroxymethyl) aminomethane)............................................. 224 Fly cultures..................................................................... 144 Fly food.................................................................. 137–138 Fly heads harvesting........................................137–138, 144

G Generation of transgenic fly choice of the driver................................................... 143 establishing stable fly strains............................. 141–143 Glycosidic detergent analysis.................................. 206, 214 Green fluorescent protein (GFP)...........................3, 21, 58, 121, 136, 137, 158, 162, 192 detection........................................................... 159, 160 fluorescence in gel................................ 22, 25, 32, 33, 36

H High-throughput cloning in E. coli and L. lactis..................................................18–21, 80 Human breast cancer resistance protein (BCRP).............18, 107, 110, 111, 113, 115, 116 Human glutamate transporter.........................136, 138, 139 Human Hedgehog receptors........................................ 5, 88 Human mu opioid receptor (hMOR)..............121, 125, 128

I Immunofluorescence.............................................. 158–160 Induction of membrane proteins expression....................................................... 77 Induction of membrane proteins overexpression in E. coli..................................................................24, 31 in L. lactis...............................................................24, 31 Induction of recombinant protein expression in S2 cells.............................................. 126, 129 Insect cells..................6, 11, 12, 40, 105–117, 120, 131, 248 Drosophila Schneider 2 cells.................................. 6, 119 scale up in spinners................................................... 116

Spodoptera frugiperda (Sf9, Sf21).............................6, 106 Trichoplusia ni (Hi5)..............................6, 106, 107, 116

L Lactococcus lactis....................................3, 4, 17–36, 67–83 Laemmli running buffer................................................... 72 Large scale expression of SERCA1a...................... 255–256 Large scale harvesting of internal membranes...................................................... 61 Large scale transient transfection (LSTT) of mammalian cells........................................... 7 Ligand binding assays............................................. 123, 235 Ligation................................................................20, 21, 70, 75, 81, 140, 145 Ligation independent cloning (LIC), vector.......................................21, 23–24, 27–30 Lipid analysis.................................................................. 215 Lipid-mediated transfection of RNA..................... 157–158

M Mammalian cells BHK-21 (Baby Hamster kidney) cells..............150, 151, 156–160, 162 CHO-K1 (Chinese Hamster ovary)........................ 151, 157, 160 COS-1 cells.................................................................. 7 HEK293 (human embryonic kidney) cells................................................152, 153, 160 infected by the Semliki Forest virus.............................. 8 Membrane preparation from drosophila Schneider 2 (S2) cells........................... 6 from fly heads................................................... 138, 144 from yeast............................................................. 89–90 Membrane protein chaperon NinaA..........107, 110, 115–116 Membrane protein crystallization acid-sensing ion channel 1...................................... 9, 12 human β1-adrenergic receptor...................................... 6 human β2 adrenergic receptor.................................... 11 human leukotriene C4 synthase (LTC4)........................................................... 11 human 5-lipoxygenase-activating protein (FLAP)............................................... 11 rabbit sarcoplasmic reticulum Ca2+ -ATPase (SERCA 1a)................................................... 10 recombinant rhodopsin............................................... 11 voltage-dependent potassium channel (Kv1.2)............................................................ 10 Membrane proteins in complex with A8-35..................................... 232–234 concentration........................................ 2, 8, 21, 56, 185, 193, 213, 258–259 inserted into preformed membranes......................... 228 preparation from E. coli......................................174–175 solubilization.................................9, 205–216, 259–261

stability in detergent solution................................... 192–195 in fluorinated surfactants............................ 225–227 Metabolic labeling.................................................. 160–162

Heterologous Expression of Membrane Proteins 271 Index     

Nisin inducible controlled gene expression (NICE) system................................4, 18, 67–69 Nisin preparation.............................................................. 77

harvest of recombinant viral particles....................... 158 lipid-mediated transfection of RNA................. 157–158 SFV in vitro transcription..........................150, 155–156 subcloning into SFV vectors............................. 152–155 verification of virus titers.................................. 158–159 Stabilising membrane proteins by genetic engineering........................................... 197–199 Stacking gels............................................. 25, 71, 72, 78, 95, 97, 114, 137, 251, 263, 264

O

T

Obtaining mutant strains by host selection................ 58–60 Origin of replication of the expression vectors........... 52–54

Thrombin cleavage of fusion protein................................ 44 Transfection........................................ 7, 107, 108, 111–113, 116, 122, 126, 131, 137, 141, 157, 158, 162 for the generation of stable S2 cell lines................... 126 of Sf9 cell with bacmids.................................... 111–113 Transfer buffer................................. 72, 91, 98, 99, 109, 114 Transformation of bacteria............................................... 76 Tris-buffered saline (TBS)..................................72, 80, 154 Tris-buffered saline with tween (TBS-T).................. 72, 80, 83, 91, 99, 100, 109, 115

N

P Plasmid isolation and verification............................... 70, 76 Preparation of E. coli extracts.................................. 174–175 Preparation of electrocompetent bacteria................... 35, 76 Purification E. coli membrane........................................... 61 Purification of membrane proteins biotin acceptor domain (BAD)..................248, 253, 254 biotin-streptavidin affinity........................................ 261 calmodulin binding domain........................................ 91 his GraviTrap columns......................................... 42, 45 polyhistidine tag / Ni-NTA superflow resin...................................................10, 44, 248 streptavidin tag II....................................................... 69 Purification of SERCA by affinity chromatography................251, 253, 261–262 by size exclusion chromatography (SEC)......... 251, 253 Purification of the internal membrane............................. 61 Purifying unstable membrane proteins................... 195–196

R Reducing sample buffer.................................................... 72 Refolding assay........................................................... 42, 45 Resolving gels..................................................71, 72, 96–97 Rhabdomeres..........................................................................7

S Sarcoplasmic reticulum Ca2+ -ATPase (SERCA 1a)................................5, 10, 247–266 SDS-PAGE laemmli running buffer............................................... 72 reducing sample buffer............................................... 72 resolving gels.............................................71, 72, 96–97 stacking gels........................................ 25, 71, 72, 78, 95, 97, 114, 137, 251, 263, 264 Selection of stable S2 cell lines....................................... 126 Selection of yeast clones on minimal medium........ 254–255 Semliki Forest virus (SFV) infected mammalian cells activation of recombinant SFV particles........... 158, 162 electroporation of RNA.................................... 156–157

V Vector backbone exchange procedure........18, 23–24, 30–31 Vectors LIC vector.......................................................19, 20, 27 pAc5 containing the constitutive promoter from the actine 5C gene........................................ 121 pCoHYGRO.................................................... 121, 126 pEGFP-C3 vector.................................................... 121 pERL.........................................................20, 21, 23, 30 pET21a...................................................................... 40 pFastBac dual vector................................108, 110, 111 pMT containing the copper-inducible metallothionein promoter............................. 121 pMW7-b plasmid........................................... 53–55, 60 preparation.................................................70, 75, 88–89 pUAST...................................... 136, 138–141, 145, 146 pYeDP60.......................................................... 252, 254 pYEpPMA........................................................... 91, 92 Verifying clones by colony PCR................................. 23, 29 Virus amplification......................................................... 113 Visualization of reporter gene expression............... 159–160

W Web resources EMBL protein production......................................... 51 ExPasy 51, 52 membrane proteins of known 3D structures........................................................ 51 SMART (protein domain database)........................... 51 Western blotting amidoblack 0.1% solution..................................... 91, 99 blocking buffer.........................................72, 91, 99, 100

Heterologous Expression of Membrane Proteins 272  Index

  

Western blotting (Continued) transfer buffer...................................................... 72, 91, 98, 99, 109, 114 tris-buffered saline (TBS)........................72, 80, 91, 109 tris-buffered saline with tween (TBS-T)..................72, 80, 91, 99, 100, 109, 115

X X-gal staining..................................................154, 159, 162

Y Yeast Pichia pastoris................................................5–6, 11, 248 Saccharomyces cerevisiae K699............................................................... 88, 92 W303.1b Gal4-2........................................ 252, 254 transformation...................................................... 89, 92 Yeast culture YEP medium.............................................................. 89 YNB medium............................................................. 89