Production of Membrane Proteins E-Book

• Herausgeber: Wiley-VCH
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 775

Veröffentlichungsjahr: 2011

Beschreibung

Designed as a research-level guide to current strategies and methods of membrane protein production on the small to intermediate scale, this practice-oriented book provides detailed, step-by-step laboratory protocols as well as an explanation of the principles behind each method, together with a discussion of its relative advantages and disadvantages.
Following an introductory section on current challenges in membrane protein production, the book goes on to look at expression systems, emerging methods and approaches, and protein specific considerations.
Case studies illustrate how to select or sample the optimal production system for any desired membrane protein, saving both time and money on the laboratory as well as the technical production scale.
Unique in its coverage of "difficult" proteins with large membrane-embedded domains, proteins from extremophiles, peripheral membrane proteins, and protein fragments.

Leseprobe

Table of Contents

Cover

Related Titles

Title page

Copyright page

Preface

List of Contributors

Introduction

Expression

Solubilization and Structural Methods

Abbreviations

Part One: Expression Systems

1 Bacterial Systems

1.1 Introduction

1.2 Understanding the Problem

1.3 Vector/Promoter Types

1.4 T7 Expression System

1.5 Tunable T7 Expression Systems

1.6 Other Useful Membrane Protein Expression Strains

1.7 Clone Stability

1.8 Media Types

1.9 Fusion Partners/Membrane Targeting Peptides

1.10 Chaperone Overexpression

1.11 Cautionary Notes Related to Chaperone Overexpression

1.12 Emerging Role of Quality Control Proteases

1.13 Tag Selection

1.14 Potential Expression Yield

1.15 Strategies to Overcome Protein Instability

Acknowledgments

2 Membrane Protein Expression in Saccharomyces cerevisiae

2.1 Introduction

2.2 Getting Started

2.3 Special Considerations

2.4 Case Studies

2.5 Conclusions

3 Expression Systems: Pichia pastoris

3.1 Introduction

3.2 A (Brief) Summary on the (Long) History of P. pastoris

3.3 Introducing P. pastoris as a Biotechnological Tool: Its (Extended) Strengths and (Limited) Weaknesses

3.4 Basics of the P. pastoris Expression System

3.5 Successful Large-Scale Expression of Membrane Proteins Using P. pastoris

3.6 Guidelines for Optimizing Membrane Protein Expression in P. pastoris Using GPCRs as Models

3.7 Conclusions and Future Directions

Acknowledgments

4 Heterologous Production of Active Mammalian G-Protein-Coupled Receptors Using Baculovirus-Infected Insect Cells

4.1 Introduction

4.2 Experimental

4.3 Conclusions and Future Perspectives

5 Membrane Protein Expression in Mammalian Cells

5.1 Introduction

5.2 Mammalian Systems

5.3 Case Studies

5.4 Conclusions

6 Membrane Protein Production Using Photosynthetic Bacteria: A Practical Guide

6.1 Introduction

6.2 Preparation of Expression Constructs

6.3 Transfer of Plasmid DNA to Rhodobacter via Conjugal Mating

6.4 Small-Scale Screening for Expression and Localization of Target Protein in Rhodobacter

6.5 Large-Scale Culture

6.6 Detergent Solubilization and Chromatographic Purification of Expressed Membrane Proteins

6.7 Protein Identification and Assessment of Purity

6.8 Preparations of Specialized Rhodobacter Membranes

Appendix: Media and Buffer Formulations

Part Two: Protein-Specific Considerations

7 Peripheral Membrane Protein Production for Structural and Functional Studies

7.1 Introduction

7.2 Case Studies of Peripheral Membrane Proteins

7.3 Conclusions

Acknowledgments

8 Expression of G-Protein-Coupled Receptors

8.1 Introduction

8.2 Bacterial Expression of GPCRs

8.3 Expression of GPCRs in Inclusion Bodies, and Refolding

8.4 Expression of GPCRs in Yeast

8.5 Expression of GPCRs in Insect Cells

8.6 Expression of GPCRs in Mammalian Cell Lines

8.7 Expression of GPCRs in Retina Rod Cells

8.8 Expression of GPCRs in a Cell-Free System

8.9 Stabilization of GPCRs during Solubilization and Purification

8.10 Conclusions

Acknowledgments

9 Structural Biology of Membrane Proteins

9.1 Introduction

9.2 Folding and Structural Analysis of Membrane Proteins

9.3 Test Cases

Acknowledgments

Part Three: Emerging Methods and Approaches

10 Engineering Integral Membrane Proteins for Expression and Stability

10.1 Introduction

10.2 Engineering Higher Expression

10.3 Engineering Higher Stability

10.4 Conclusions

11 Expression and Purification of G-Protein-Coupled Receptors for Nuclear Magnetic Resonance Structural Studies

11.1 Introduction: G-Protein-Coupled Receptor Superfamily

11.2 CXCR1

11.3 GPCR Structures

11.4 NMR Studies of GPCRs

11.5 Expression Systems

11.6 Cloning of CXCR1 into pGEX2a

11.7 Expression of CXCR1

11.8 Purification

11.9 Refolding and Reconstitution

11.10 Binding and Activity Measurements

11.11 NMR Spectra

Acknowledgments

12 Solubilization, Purification, and Characterization of Integral Membrane Proteins

12.1 Introduction

12.2 Solubilization of IMPs

12.3 IMP Purification

12.4 Characterization of Solubilized IMPs

Appendix

Acknowledgments

13 Stabilizing Membrane Proteins in Detergent and Lipid Systems

13.1 Introduction

13.2 Choice of Detergent: Solubilization versus Stability

13.3 Mitigating Protein Denaturation

13.4 Making or Selecting a Stable Protein

13.5 Conclusions

14 Rapid Optimization of Membrane Protein Production Using Green Fluorescent Protein-Fusions and Lemo21(DE3)

14.1 Introduction

14.2 Main Protocol

14.3 Materials

14.4 Expression and Isolation of GFP-His8

14.5 Conclusions

Acknowledgments

Index

Related Titles

Becker, CFW

Chemical Biology of Membrane Proteins

432 pages

Hardcover

ISBN: 978-0-470-87194-2

Carta, G., Jungbauer, A.

Protein Chromatography

Process Development and Scale-Up

364 pages with 178 figures and 45 tables

2010

Hardcover

ISBN: 978-3-527-31819-3

Behme, S.

Manufacturing of Pharmaceutical Proteins

From Technology to Economy

404 pages with 136 figures and 41 tables

2009

Hardcover

ISBN: 978-3-527-32444-6

Spirin, A. S., Swartz, J. R. (eds.)

Cell-free Protein Synthesis

Methods and Protocols

262 pages with 93 figures and 14 tables

2008

Hardcover

ISBN: 978-3-527-31649-6

Walsh, G.

Pharmaceutical Biotechnology

Concepts and Applications

498 pages

2007

Softcover

ISBN: 978-0-470-01245-1

The Editor

Prof. Anne Skaja Robinson

University of Delaware

Dept. of Biochemical Eng.

150 Academy St.

Newark, DE 19716

USA

We would like to thank Dr. David Salmon and Dr. Krzysztof Palczewski (Polgenix, Inc.) for providing us with the graphic material used in the cover illustration.

All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.

Library of Congress Card No.: applied for

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at <http://dnb.d-nb.de>.

© 2011 Wiley-VCH Verlag & Co. KGaA, Boschstr. 12, 69469 Weinheim, Germany

All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form – by photoprinting, microfilm, or any other means – nor transmitted or translated into a machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law.

ISBN: 978-3-527-32729-4

ePDF ISBN: 978-3-527-63454-5

oBook ISBN: 978-3-527-63452-1

ePub ISBN: 978-3-527-63453-8

Mobi ISBN: 978-3-527-63455-2

Preface

With this volume, we have attempted to provide a guide for those interested in structural biology and biochemistry of membrane proteins. It is our hope that this text will be useful both to experts and to those new to the field. The various chapters illustrate the breadth and depth of approaches to, and the challenges associated with, membrane protein production and solubilization. Our goals were to compare and critically analyze methods that have been used successfully for the expression and isolation of membrane proteins, to identify and help disseminate emerging methods, and to provide a practical reference for those working with these proteins.

Interest in membrane proteins has grown substantially in the last 25 years, in part due to the recognition of their importance in major cellular functions, including signaling, transport, and adhesion. As a result of their roles, membrane proteins comprise 30–60% of all drug discovery efforts, but remain a challenge for structure-based discovery efforts because of the paucity of high-resolution structures.

As of the time of writing, about 260 unique high-resolution membrane protein structures have been solved, with an exponential growth of new structures (e.g., 39 unique structures were determined in 2009). Several centers and focused funding efforts, such as the Protein Structure Initiative, are aimed at accelerating the membrane structure rate, although it is not yet clear whether these approaches will increase the pace to equal that of soluble proteins (see http://blanco.biomol.uci.edu/MP_Structure_Progress.html). These initiatives by federal funding agencies and private organizations have yielded new technologies and creative approaches for the production and stabilization of membrane proteins, many of which are described in this volume.

In this collection, the Introduction gives an overview of the challenges and successes of structural biology of membrane proteins. Chapters 1–6 outline best practices for expression in both prokaryotic and eukaryotic hosts. Chapters 7–9 focus on aspects that are specific for individual protein classes and applications. Chapters 10–14 highlight new approaches to producing and working with membrane proteins.

Of course, the credit for this book goes to the outstanding array of chapter authors. On their behalf, I thank Stefanie Volk, the Project Editor at Wiley-VCH, for her support and encouragement, Frank Weinreich at Wiley-VCH for initiating this project, and Rechilda Alba at the University of Delaware for technical assistance.

Anne Skaja Robinson

December 2010

List of Contributors

Fatima Alkhalfioui

IREBS Institute

CNRS and University of Strasbourg Pain and GPCRs Research Group

Boulevard Sébastien Brant

67412 Illkirch

France

Brian J. Bahnson

University of Delaware

Department of Chemistry & Biochemistry

Newark, DE 19716

USA

Rebecca Batchelor

University of Hull

Department of Chemistry

Cottingham Road

Kingston-upon-Hull HU6 7RX

UK

Michael J. Betenbaugh

Johns Hopkins University

Department of Chemical and Biomolecular Engineering

221 Maryland Hall 3400 North Charles Street

Baltimore, MD 21218

USA

Olivier Bornert

IREBS Institute

CNRS and University of Strasbourg Pain and GPCRs Research Group

Boulevard Sébastien Brant

67412 Illkirch

France

Zachary Britton

University of Delaware

Department of Chemical Engineering, Colburn Laboratory

150 Academy Street

Newark, DE 19716

USA

Özge Can

Acibadem University

School of Medicine Gulsuyu Mahallesi

Fevzi Cakmak Caddesi Divan Sokak, No: 1 Maltepe, Istanbul

Turkey

Fabio Casagrande

University of California at San Diego

Department of Chemistry and Biochemistry

9500 Gilman Drive

La Jolla, CA 92093

USA

Mark Chiu

Department of Structural Biology

Abbott Laboratories R46Y AP10-LL08

100 Abbott Park Road

Abbott Park, IL 60064

USA

Igor Dodevski

University of Zurich

Department of Biochemistry

Winterthurerstrasse 190

8057 Zurich

Switzerland

David Drew

Imperial College

Membrane Protein Crystallography Group, Division of Molecular Biosciences

South Kensington Campus

London SW7 2AZ

UK

Timothy Esbenshade

Abbott Laboratories

Department of Neurosciences

100 Abbott Park Road

Abbott Park, IL 60064

USA

Brian Estvander

Abbott Laboratories

Department of Neurosciences

100 Abbott Park Road

Abbott Park, IL 60064

USA

Dimitra Gialama

Stockholm University

Department of Biochemistry and Biophysics, Center for Biomembrane Research

Sv. Arrheniusväg 16c

10691 Stockholm

Sweden

Jan-Willem de Gier

Stockholm University

Department of Biochemistry and Biophysics Center for Biomembrane Research

Sv. Arrheniusväg 16c

10691 Stockholm

Sweden

and

Xbrane Bioscience AB

Stureplan 15

111 45 Stockholm

Sweden

Deborah K. Hanson

Argonne National Laboratory

Biosciences Division

9700 South Cass Avenue

Lemnot, IL 60439

USA

Deniz B. Hizal

Johns Hopkins University

Department of Chemical and Biomolecular Engineering

221 Maryland Hall 3400 North Charles Street

Baltimore, MD 21218

USA

Steve Kakavas

Abbott Laboratories

Advanced Technologies

100 Abbott Park Road

Abbott Park, IL 60064

USA

Hans Kiefer

Biberach University of Applied Sceinces Karlstrasse 11

88400 Biberach

Germany

Mirjam Klepsch

Stockholm University

Department of Biochemistry and Biophysics Center for Biomembrane Research

Sv. Arrheniusväg 16c

10691 Stockholm

Sweden

Kathy Krueger

Abbott Laboratories

Department of Neurosciences

100 Abbott Park Road

Abbott Park, IL 60064

USA

Philip D. Laible

Argonne National Laboratory

Biosciences Division

9700 South Cass Avenue

Lemnot, IL 60439

USA

Marc Lake

Abbott Laboratories

Advanced Technologies

100 Abbott Park Road

Abbott Park, IL 60064

USA

Tzvetana Lazarova

Universitat Autònoma de Barcelona Unitat de Biofísica

Departament de Bioquímica i de Biologia Molecular Facultat de Medicina and Centre d’Estudis en Biofísica

08193 Bellaterra

Barcelona

Spain

Christel Logez

IREBS Institute

CNRS and University of Strasbourg Pain and GPCRs Research Group

Boulevard Sébastien Brant

67412 Illkirch

France

Patrick Loll

Drexell University College of Medicine

Department of Biochemistry and Molecular Biology

Room 10-102 New College Building

245 N. 15th Street, Mailstop 497 Philadelphia, PA 19102-1192

USA

Mark Lorch

University of Hull

Department of Chemistry

Cottingham Road

Kingston-upon-Hull HU6 7RX

UK

Víctor Lórenz-Fonfría

Universitat Autònoma de Barcelona

Unitat de Biofísica Departament de Bioquímica i de Biologia Molecular Facultat de Medicina and Centre d’Estudis en Biofísica

08193 Bellaterra

Barcelona

Spain

Sunny Mai

Johns Hopkins University

Department of Chemical and Biomolecular Engineering

221 Maryland Hall 3400 North Charles Street

Baltimore, MD 21218

USA

Klaus Maier

Membrane Receptor Technologies

9381 Judicial Drive

Suite 140 San Diego, CA 92121

USA

Patrick McNeely

University of Delaware

Department of Chemical Engineering, Colburn Laboratory

150 Academy Street

Newark, DE 19716

USA

Donna L. Mielke

Argonne National Laboratory

Biosciences Division

9700 South Cass Avenue

Lemnot, IL 60439

USA

Andrea Naranjo

University of Delaware

Department of Chemical Engineering Colburn Laboratory

150 Academy Street

Newark, DE 19716

USA

Erika Ohsfeldt

Johns Hopkins University

Department of Chemical and Biomolecular Engineering

221 Maryland Hall 3400 North Charles Street

Baltimore, MD 21218

USA

Stanley J. Opella

University of California at San Diego

Department of Chemistry and Biochemistry

9500 Gilman Drive

La Jolla, CA 92093

USA

Esteve Padrós

Universitat Autònoma de Barcelona Unitat de Biofísica

Departament de Bioquímica i de Biologia Molecular Facultat de Medicina and Centre d’Estudis en Biofísica

08193 Bellaterra

Barcelona

Spain

Krzysztof Palczewski

Case Western Reserve University

School of Medicine, Department of Pharmacology W319

10900 Euclid Avenue

Cleveland, OH 44106-4965

USA

Sang Ho Park

University of California at San Diego

Department of Chemistry and Biochemistry

9500 Gilman Drive

La Jolla, CA 92093

USA

Alex Perálvarez-Marín

Universitat Autònoma de Barcelona

Unitat de Biofísica Departament de Bioquímica i de Biologia Molecular Facultat de Medicina and Centre d’Estudis en

Biofísica 08193 Bellaterra

Bellaterra

Spain

Ana Pereda-Lopez

Abbott Laboratories

Advanced Technologies

100 Abbott Park Road

Abbott Park, IL 60064

USA

Andreas Plückthun

University of Zurich

Department of Biochemistry

Winterthurerstrasse 190

8057 Zurich

Switzerland

Anne Skaja Robinson

University of Delaware

Department of Chemical Engineering Colburn Laboratory

150 Academy Street

Newark, DE 19716

USA

David Salom

Polgenix, Inc.

Suite 260

11000 Cedar Avenue

Cleveland, OH 44106

USA

James Samuelson

New England Biolabs

Gene Expression Division

240 County Road

Ipswich, MA 01938

USA

Susan Schlegel

Stockholm University

Department of Biochemistry and Biophysics Center for Biomembrane Research

Sv. Arrheniusväg 16c

10691 Stockholm

Sweden

Krishna Vukoti

Center for Proteomics and Broinformatics Case Western Reserve University

BRB 9th Floor

10900 Euclid Avenue

Cleveland, OH 44106

USA

Renaud Wagner

IREBS Institute

CNRS and University of Strasbourg

Pain and GPCRs Research Group Boulevard Sébastien Brant

67412 Illkirch

France

David Wickström

Stockholm University

Department of Biochemistry and Biophysics Center for Biomembrane Research

Sv. Arrheniusväg 16c

10691 Stockholm

Sweden

and

Xbrane Bioscience AB

Stureplan 15

111 45 Stockholm

Sweden

Alexei Yeliseev

National Institutes of Health

National Institute on Alcohol Abuse and Alcoholism

5625 Fishers Lane, Room 3N17

Rockville, MD 20852

USA

Carissa Young

University of Delaware

Department of Chemical Engineering, Colburn Laboratory

150 Academy Street

Newark, DE 19716

USA

Introduction

Anne Skaja Robinson and Patrick J. Loll

Membrane proteins, acting as channels, receptors, and transporters, enable the cell to transport information and materials across the plasma membrane. As such, they are vital for cellular functioning and remain major drug discovery targets. Both to facilitate drug development and to understand the triggers of cellular action and reaction, structural information is required. However, although membrane proteins represent 30% of all proteins in the genome, they represent only around 1% of all high-resolution structures. One reason for this disparity is that biophysical and biochemical studies of membrane proteins require large quantities of purified protein – arguably even more than soluble proteins, because of the optimization of detergent conditions that is required. The first membrane protein structure ever determined – that of the photosynthetic reaction center of Rhodopseudomonas viridis, in 1985 – used protein purified from the native source [1]. Additional new structures have continued to appear over the last 25 years, but for many years primarily relied on protein purified from native sources. This is probably one reason why the rate of increase in membrane protein structures has been somewhat slower than the corresponding rate for soluble proteins [2]. Ten years ago, the majority of membrane protein structures were still obtained from proteins available in high natural abundance. More recently, heterologous expression has shown success, but this approach is still challenging, particularly for multispanning proteins (Figure 1) (reviewed by [2–5]).

Figure 1 Comparison of numbers of soluble and membrane protein structures determined [6, 7]; proteins from different organisms, as well as mutants of the same structure are included in total membrane structures. Unique membrane structures include membrane protein structures for proteins from different organisms, but exclude protein variants and substrate-bound forms. Data obtained from membrane proteins are further divided into those expressed in non-native hosts (heterologous), eucaryotic proteins, and GPCRs. Note: any errors are attributed to A.S.R. rather than the websites referenced herein.

The path to a high-resolution structure of a membrane protein involves many potential obstacles, including poor expression, limited extraction and purification yields, and challenging structural approaches. Successfully navigating this path requires a combination of scientific insight, creativity, opportunism, and trial-and-error empiricism. In this book, we highlight the state-of-the-art approaches for membrane protein expression, solubilization, and high-resolution structural methods that should provide guidance for those new to the field as well as established membrane protein scientists.

Expression

Results for many classes of membrane proteins suggest that in unmodified hosts, only a small subset of the desired target proteins will be expressed efficiently. For example, the MePNet consortium screened expression of 103 G-protein-coupled receptors (GPCRs) in Sf9, Escherichia coli, or Pichia expression systems, and obtained only two that expressed at modest levels in Sf9 cells, eight in Pichia, and none in E. coli [8]. The Cross and Nakamoto groups found that expression of Mycobacterium tuberculosis membrane protein targets in E. coli similarly gave low yields of well-expressed, membrane-localized membrane proteins, with no proteins larger than 15 kDa expressing well (Figure 2) [9]. In fact, the report of the Mid-Course Review Panel for the National Institutes of Health (NIH) Structural Biology Working Group states “… membrane protein production is still very challenging and much more must be learned to significantly advance the field” [10]. Successful approaches to expression of membrane proteins in E. coli (Chapter 1), Saccharomyces cerevisiae (Chapter 2), Pichia pastoris (Chapter 3), insect cells (Chapter 4), mammalian cells (Chapter 5), and photosynthetic bacteria (Chapter 6) are described in Part One of the book. In addition, each chapter contains case studies that highlight successes, and suggested hosts and expression systems to yield the best outcomes for expression.

Figure 2 Production of M. tuberculosis membrane protein targets in E. coli shows that protein production is a major bottleneck to biophysical and structural studies [9]. “In Membrane” designates those proteins determined to inserted into the membrane via detergent solubilization studies.

An interesting class of proteins is the peripheral membrane proteins, which can either exist in an aqueous environment or associated with the membrane. Nonspecific electrostatic interactions, hydrophobic association, covalent lipid anchors, or lipid-binding domains facilitate membrane association of these proteins. Special aspects of expression and purification of these molecules are discussed in Chapter 7.

Perhaps one of the most challenging classes of membrane proteins for expression, solubilization, and crystallization are the multipass membrane proteins, with GPCRs representing the largest family of these. GPCRs are expressed in virtually all human tissues and transmit a wide variety of signals in response to diverse stimuli (including light, hormones, injury, and inflammation). These signals regulate a diverse set of cellular responses via interaction with GTP-binding proteins [11, 12].

The structure of rhodopsin, purified from native tissue, was the first high-resolution GPCR structure determined by X-ray crystallography [13, 14]. In 2007, the human β2-adrenergic receptor (β2AR) structure, with either lysozyme or a Fab fragment inserted in cytoplasmic loop 3 (CL3), was determined by Kobilka et al., after many years of tweaking expression and crystallization approaches [15, 16]. A structure for an engineered version of turkey β1AR, with a truncated C-terminus and a shortened CL3, soon followed [17]. The publication of the high-resolution X-ray structure of the adenosine A2A receptor (A2AR) in late 2008 [18] has generated additional excitement in the GPCR field. To reduce structural flexibility and increase crystallizability of A2AR the C-terminal tail was removed and the CL3 was also substituted with T4 lysozyme, as was done successfully for β2AR. These structures have revealed that the GPCR family is “extremely adaptable” and can tolerate a wide range of helix distortions or reorientations that impact ligand binding [19].

The protein re-engineering required to facilitate crystallization can alter the native features and function of the receptors, however. The third intracellular loop [20] and the preceding transmembrane domain five are involved in G-protein coupling [21], and it is not surprising that the replacement of the third intracellular loop of A2A with T4 lysozyme inhibited receptor interactions with the G-protein and ligand [18]. Indeed, as impressive as the recent GPCR structural efforts have been, in silico ligand design using these structures suggests limited utility due to the inactive state nature of these structures [22]. Continued effort is needed to study this important family and strategies to express GPCRs are described in Chapter 8.

Directed evolution approaches to re-engineering a membrane protein of interest is an alternative to the specialized modifications described above. This method, as described in Chapter 9, relies only on screening and selection for membrane protein function (e.g., through ligand binding activity). One alternative to focusing on re-engineering a particular membrane protein is to screen different membrane proteins in order to identify those that are readily expressed. Most recently, this has been addressed via high-throughput approaches. For example, the University of California at San Francisco Membrane Protein group supported under the NIH Roadmap cites a strategy for “… ‘discovery-oriented’ selection of tractable targets …” based on expression level, detergent solubilization, and chromatographic behavior; notably, this approach yielded only five tractable targets out of an initial 384 candidates drawn from the S. cerevisiae membrane proteome [23]. A similar approach using Green Fluorescent Protein (GFP) tagging to screen for “tractable” eukaryotic membrane proteins has yielded notable success (Chapter 10) [24, 25]. One caveat to these approaches is that the screens pick up “soluble” proteins based on GFP fluorescence, but do not test function – a key feature for drug design. For example, membrane insertion and proper localization do not effectively distinguish active GPCRs from inactive ones [26, 27]; similarly, the MePNet consortium found no correlation between expression level and activity in a test of around 50 GPCRs expressed in Sf9 cells and Pichia [8]. Further, such screening approaches can only identify those protein targets with the innate ability to be expressed at high levels (low-hanging fruit); they offer no hope for improving poorly expressed targets.

Solubilization and Structural Methods

Once high-level expression is achieved, often the next step is to carry out biophysical characterization or high-resolution structural analysis. Unfortunately, purification and solubilization methods are still primarily trial-and-error approaches, but information in Chapters 11 and 12 (Part Three) should provide a guide for commonly used methods and strategies for any protein of interest.

Structure determination of membrane proteins via nuclear magnetic resonance (NMR) has been ongoing for decades, but has been primarily focused on peptides and small proteins, since detergent micelles, bicelles, or lipids increase the size and heterogeneity of membrane protein mixtures. However, in recent years significant progress has been made in both solution NMR methods for larger proteins and solid-state NMR, such that complete structures have been determined for several membrane proteins [28–30]. In particular, for solution NMR the TROSY (transverse relaxation optimized spectroscopy) method developed by Kurt Wüthrich enabled structure determination of several larger integral membrane proteins (e.g., several β-barrel proteins from E. coli). For solid-state NMR (SSNMR), the ability to align structures with bicelles and bilayers enables anisotropic interactions to be resolved [31]. In addition, SSNMR measurements do not limit the sample to solution phase, which enables insoluble or aggregate structure determination. The use of magic angle spinning for SSNMR has also facilitated the resolution of larger structures. Dror Warschawski has developed a database of membrane protein structures determined by NMR (http://www.drorlist.com/nmr/MPNMR.html), which includes 39 unique structures as of 1 June 2010. Although β-barrels dominate the larger structures, there are some larger α-helical structures as well, including that of diacylglycerol kinase, a trimer of 43 kDa. In addition to providing high-resolution structural information, NMR can determine protein conformational changes and flexibility under near-physiological conditions, which is a key feature to understanding function for many integral membrane proteins. This book provides information to assist in developing metabolically labeled proteins for NMR characterization (Chapter 13); readers interested in reviews of NMR methods applied to membrane proteins should examine several good references in the literature [29–32].

Protein structure determination by X-ray crystallography predates NMR methods by many years. However, just as obstacles are faced with NMR approaches, so too the production of suitable crystals has proven a significant stumbling block for crystallography. In fact, not so long ago crystallization of membrane proteins was widely held to be impossible. Thankfully, during the past three decades membrane protein crystallization stopped being considered impossible, moved through a stage during which it was thought of as somewhat heroic, and has now entered a phase in which it is generally accepted to be a practical undertaking.

The first crystals of OmpF porin, a bacterial outer membrane channel, were reported in 1980 [33]. Porin crystallization raised several interesting points, many of which remain relevant today. First, the protein was purified from an overproducing strain of E. coli [34] and was not made by the recombinant methodologies that now dominate the structural biology of soluble proteins. This reflected the difficulties associated with heterologous overexpression of membrane proteins; such difficulties continue to dog membrane protein work today and are discussed at length in this book. Second, porin binds tightly to bacterial lipopolysaccharide, and the key to producing good crystals turned out to be the patient and careful removal of as much of this lipid as possible [34]. While this precise formula has by no means proven universal – indeed, there are instances where overzealous removal of lipids compromises crystallization – it presaged the emergence of protein–lipid interactions as important contributors to both protein stability and crystallization behavior. Finally, workers studying the porin system were the first to identify the detergent cloud point as a region of the phase diagram that is fraught with unusual interest for the crystallization of detergent-solubilized membrane proteins [35, 36]; investigators probing the physical mechanisms underlying membrane protein crystallization have returned to this observation time and again in intervening years.

While porin is credited as the first membrane protein to yield well-ordered crystals, the first reported structure belongs to the photosynthetic reaction center from the bacterium R. viridis [1]. Crystals of this protein were obtained after development of a rigorous screening procedure aimed at fostering the packing of protein–detergent complexes into a crystalline lattice (so-called “Type II” crystals [37]). One key concept to emerge from this work is the notion that small amphiphiles can help fine-tune the structures and/or interactions of protein–detergent complexes. While such molecules have not proven to be magic bullets that facilitate crystallization for all membrane proteins, they have proven useful in many cases, and are extremely important insofar as they highlight the idea that micelle structure and phase behavior can be altered in rational ways. Approaches to obtaining protein for high-resolution crystallography are described in Chapter 14.

A particularly interesting advance made during the last 15 years was the development of lipidic mesophase crystallization methods [38]. Lipidic mesophases such as the cubic phase form gel-like materials in which bilayers assemble into complex three-dimensional structures. Proteins embedded in these phases bilayers encounter a native-like environment (since they are in a bilayer), while the spatial ordering of the lipid enables them to diffuse through three dimensions; presumably this facilitates crystal contacts not only within the plane of the bilayer, but also orthogonal to it [39]. This methodology has produced some notable successes, most prominently the formation of well-ordered crystals of a variety of seven-transmembrane helical proteins, including both GPCRs and their prokaryotic homologs. While it seems unlikely that “in cubo” or “in meso” methods, as they are known, will entirely supplant the direct crystallization of protein-detergent complexes – the majority of structures currently being published still rely on the latter method – the stunning successes provided by mesophase approaches ensure them a lasting place in the crystallizer’s tool kit.

Membrane protein structural biology is now well established, and in recent years the number of new crystal and NMR structures has grown exponentially [2, 6, 29]. Nonetheless, challenges remain. Efforts aimed at finding “low-hanging fruit” for high-resolution studies have generated much new structural information, but have not provided the insights necessary to develop rational approaches for more challenging targets. Eukaryotic membrane proteins are still under-represented in the structural databases, which most likely reflects difficulties with protein production, low stability, and/or inherent flexibility on the part of many of these molecules. Hence, in the coming decade proponents of membrane protein structural biology will not lack for suitable challenges.

Abbreviations

β2AR β2-adrenergic receptor

A2AR adenosine A2A receptor

CL3 cytoplasmic loop 3

GFP Green Fluorescent Protein

GPCR G-protein-coupled receptor

NIH National Institutes of Health

NMR nuclear magnetic resonance

SSNMR solid-state NMR

TROSY transverse relaxation optimized spectroscopy

References

1 Deisenhofer, J., Epp, O., Miki, K., Huber, R., and Michel, H. (1985) Structure of the protein subunits in the photosynthetic reaction centre of Rhodopseudomonas viridis at 3 Å resolution. Nature, 318, 618–624.

2 White, S.H. (2004) The progress of membrane protein structure determination. Protein Sci., 13, 1948–1949.

3 Tate, C.G. (2001) Overexpression of mammalian integral membrane proteins for structural studies. FEBS Lett., 504, 94–98.

4 Chiu, M.L., Tsang, C., Grihalde, N., and MacWilliams, M.P. (2008) Over-expression, solubilization, and purification of G protein-coupled receptors for structural biology. Comb. Chem. High Throughput Screen., 11, 439–462.

5 White, S.H. (2009) Membrane Proteins of Known 3D Structure, http://blanco. biomol.uci.edu/Membrane_Proteins_xtal.html (accessed 10 January 2009).

6 White, S.H. (2010) Progress in Membrane Protein Structure Determination, http://blanco.biomol.uci.edu/MP_Structure_Progress.html (accessed 13 July 2010).

7 Berman, H.M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T.N., Weissig, H., Shindyalov, I.N., and Bourne, P.E. (2000) The protein data bank. Nucleic Acids Res., 28, 235–242

8 Lundstrom, K., Wagner, R., Reinhart, C., Desmyter, A., Cherouati, N., Magnin, T., Zeder-Lutz, G., Courtot, M., Prual, C., Andre, N., Hassaine, G., Michel, H., Cambillau, C., and 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.

9 Korepanova, A., Gao, F.P., Hua, Y., Qin, H., Nakamoto, R.K., and Cross, T.A. (2005) Cloning and expression of multiple integral membrane proteins from Mycobacterium tuberculosis in Escherichia coli. Protein Sci., 14, 148–158.

10 Buchanan, S.K., Laughlin, M.R., Ramm, L., Rees, D.C., Stenkamp, R.E., Tamm, L., and White, S.H. (2009) Mid-Course Review – Report on the Structural Biology Working Group Membrane Proteins, http://nihroadmap.nih.gov/structuralbiology/midcoursereview/summary2008.asp (accessed 8 June 2009).

11 Lee, N.H. and Kerlavage, A.R. (1993) Molecular biology of G-protein-coupled receptors. Trends Biomed. Res., 6, 488–497.

12 Strader, C.D., Fong, T.M., Tota, M.R., Underwood, D., and Dixon, R.A. (1994) Structure and function of G protein-coupled receptors. Annu. Rev. Biochem., 63, 101–132.

13 Bourne, H.R. and Meng, E.C. (2000) Structure. Rhodopsin sees the light. Science, 289, 733–734.

14 Palczewski, K., Kumasaka, T., Hori, T., Behnke, C.A., Motoshima, H., Fox, B.A., Le Trong, I., Teller, D.C., Okada, T., Stenkamp, R.E., Yamamoto, M., and Miyano, M. (2000) Crystal structure of rhodopsin: a G protein-coupled receptor. Science, 289, 739–745.

15 Cherezov, V., Rosenbaum, D.M., Hanson, M.A., Rasmussen, S.G., Thian, F.S., Kobilka, T.S., Choi, H.J., Kuhn, P., Weis, W.I., Kobilka, B.K., and Stevens, R.C. (2007) High-resolution crystal structure of an engineered human beta2-adrenergic G protein-coupled receptor. Science, 318, 1258–1265.

16 Rasmussen, S.G., Choi, H.J., Rosenbaum, D.M., Kobilka, T.S., Thian, F.S., Edwards, P.C., Burghammer, M., Ratnala, V.R., Sanishvili, R., Fischetti, R.F., Schertler, G.F., Weis, W.I., and Kobilka, B.K. (2007) Crystal structure of the human beta2 adrenergic G-protein-coupled receptor. Nature, 450, 383–387.

17 Warne, T., Serrano-Vega, M.J., Baker, J.G., Moukhametzianov, R., Edwards, P.C., Henderson, R., Leslie, A.G., Tate, C.G., and Schertler, G.F. (2008) Structure of a beta1-adrenergic G-protein-coupled receptor. Nature, 454, 486–491.

18 Jaakola, V.P., Griffith, M.T., Hanson, M.A., Cherezov, V., Chien, E.Y., Lane, J.R., Ijzerman, A.P., and Stevens, R.C. (2008) The 2.6 angstrom crystal structure of a human A2A adenosine receptor bound to an antagonist. Science, 322, 1211–1217.

19 White, S.H. (2009) Biophysical dissection of membrane proteins. Nature, 459, 344–346.

20 Audet, M and Bouvier, M. (2008) Insights into signaling from the beta2-adrenergic receptor structure. Nat. Chem. Biol., 4, 397–403.

21 Khafizov, K., Lattanzi, G., and Carloni, P. (2009) G protein inactive and active forms investigated by simulation methods. Proteins, 75, 919–930.

22 Kolb, P., Rosenbaum, D.M., Irwin, J.J., Fung, J.J., Kobilka, B.K., and Shoichet, B.K. (2009) Structure-based discovery of beta2-adrenergic receptor ligands. Proc. Natl. Acad. Sci. USA, 106, 6843–6848.

23 Li, M., Hays, F.A., Roe-Zurz, Z., Vuong, L., Kelly, L., Ho, C.M., Robbins, R.M., Pieper, U., O’Connell, J.D., 3rd, Miercke, L.J., Giacomini, K.M., Sali, A., and Stroud, R.M. (2009) Selecting optimum eukaryotic integral membrane proteins for structure determination by rapid expression and solubilization screening. J. Mol. Biol., 385, 820–830.

24 Drew, D., Slotboom, D.J., Friso, G., Reda, T., Genevaux, P., Rapp, M., Meindl-Beinker, N.M., Lambert, W., Lerch, M., Daley, D.O., Van Wijk, K.J., Hirst, J., Kunji, E., and De Gier, J.W. (2005) A scalable, GFP-based pipeline for membrane protein overexpression screening and purification. Protein Sci., 14, 2011–2017.

25 Newstead, S., Kim, H., von Heijne, G., Iwata, S., and Drew, D. (2007) High-throughput fluorescent-based optimization of eukaryotic membrane protein overexpression and purification in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA, 104, 13936–13941.

26 Butz, J.A., Niebauer, R.T., and Robinson, A.S. (2003) Co-expression of molecular chaperones does not improve the heterologous expression of mammalian G-protein coupled receptor expression in yeast. Biotechnol. Bioeng., 84, 292–304.

27 O’Malley, M.A., Mancini, J.D., Young, C.L., McCusker, E.C., Raden, D., and Robinson, A.S. (2009) Progress toward heterologous expression of active G-protein-coupled receptors in Saccharomyces cerevisiae: linking cellular stress response with translocation and trafficking. Protein Sci., 18, 2356–2370.

28 Doreleijers, J.F., Mading, S., Maziuk, D., Sojourner, K., Yin, L., Zhu, J., Markley, J.L., and Ulrich, E.L. (2003) BioMagResBank database with sets of experimental NMR constraints corresponding to the structures of over 1400 biomolecules deposited in the Protein Data Bank. J. Biomol. NMR, 26, 139–146.

29 Kim, H.J., Howell, S.C., Van Horn, W.D., Jeon, Y.H., and Sanders, C.R. (2009) Recent advances in the application of solution NMR spectroscopy to multi-span integral membrane proteins. Prog. Nucl. Magn. Reson. Spectrosc., 55, 335–360.

30 McDermott, A. (2009) Structure and dynamics of membrane proteins by magic angle spinning solid-state NMR. Annu. Rev. Biophys., 38, 385–403.

31 Opella, S.J. and Marassi, F.M. (2004) Structure determination of membrane proteins by NMR spectroscopy. Chem. Rev., 104, 3587–3606.

32 Opella, S.J., Nevzorov, A., Mesleb, M.F., and Marassi, F.M. (2002) Structure determination of membrane proteins by NMR spectroscopy. Biochem. Cell Biol., 80, 597–604.

33 Garavito, R.M. and Rosenbusch, J.P. (1980) Three-dimensional crystals of an integral membrane protein: an initial X-ray analysis. J. Cell Biol., 86, 327–329.

34 Misra, R. and Reeves, P.R. (1987) Role of micF in the tolC-mediated regulation of OmpF, a major outer membrane protein of Escherichia coli K-12. J. Bacteriol., 169, 4722–4730.

35 Rosenbusch, J.P. (1990) The critical role of detergents in the crystallization of membrane proteins. J. Struct. Biol., 104, 134–138.

36 Zulauf, M. (1991) Detergent phenomena in membrane protein crystallization, in Crystallization of Membrane Proteins (ed. H. Michel), CRC Press, Boca Raton, FL, pp. 53–72.

37 Michel, H. (1983) Crystallization of membrane proteins. Trends Biochem. Sci., 8, 56–59.

38 Landau, E.M. and Rosenbusch, J.P. (1996) Lipidic cubic phases: a novel concept for the crystallization of membrane proteins. Proc. Natl. Acad. Sci. USA, 93, 14532–14535.

39 Nollert, P., Qiu, H., Caffrey, M., Rosenbusch, J.P., and Landau, E.M. (2001) Molecular mechanism for the crystallization of bacteriorhodopsin in lipidic cubic phases. FEBS Lett., 504, 179–186.

Part One: Expression Systems

1

Bacterial Systems

James Samuelson

1.1 Introduction

The study of membrane protein structure and function is limited by various challenges. In native cells, membrane protein copy number is often very low, so the study of individual proteins is often not feasible. Alternatively, overexpression of these hydrophobic molecules in heterologous hosts is not a routine endeavor as it is for many water-soluble proteins. Most modern bacterial expression systems have been engineered for maximal output of recombinant protein. This characteristic is ideal for well-behaved soluble proteins, but less desirable when the target protein normally resides within a lipid environment. A compounding problem in the study of membrane proteins is that the isolated target protein may exhibit polydispersity, meaning that diverse oligomeric complexes can spontaneously accumulate. This latter concern may be influenced by the expression method, but primarily depends on the detergent/lipid and buffer used for solubilization. This chapter highlights preferred strategies for membrane protein expression in bacteria that will increase the likelihood of isolating adequate amounts of homogenous target protein. Many sections will also detail the features of expression strains that are relevant to the yield and quality of expressed protein.

In this chapter, the term membrane protein will generally be used to represent α-helical membrane proteins that reside within a phospholipid bilayer environment of either eukaryotic or prokaryotic cells. Such integral membrane proteins are the most difficult to manipulate since each contains hydrophobic transmembrane (TM) regions as well as hydrophilic extramembrane regions or domains. In the case of single-spanning membrane proteins, often the catalytic domain is a water-soluble entity that may be studied by expression of a ΔTM variant. However, multispanning membrane proteins such as ion channels must be expressed without gross deletions of hydrophobic residues.

Membrane proteins with β-barrel structure such as those found in the Gram-negative bacterial outer membrane or the mitochondrial outer membrane are typically expressed at high levels as inclusion bodies within the Escherichia coli cytoplasm. Isolation and washing of these inclusion bodies often leads to a relatively pure sample of recombinant protein and the literature contains many examples of refolding of β-barrel proteins, such as Omp proteins from E. coli [1]. In contrast, refolding of α-helical integral membrane protein is quite a difficult challenge, although some successes have been reported [2–4]. The default method of expressing α-helical membrane proteins should be to direct them to the membrane fraction of the host cell and to perform purification procedures beginning with isolation of the cellular membrane fraction.

1.2 Understanding the Problem

Each recombinant membrane protein clone should be assumed to be “toxic” to the host cell. This is particularly true when bacterial hosts are employed. It is well established that uncontrolled expression of most membrane proteins in E. coli will lead to induction of cellular stress responses and occasionally cell death. In some cases, the plasmid transformation step may fail because the transformed cell cannot recover due to the uncontrolled expression of membrane protein. Therefore, the first bit of advice in designing expression clones is to use a vector that propagates at 40 copies or less per cell (pMB1+rop, oriV, p15A, pSC101 replication origins). Accordingly, a vector with a pUC-derived origin should be avoided. Secondly, the promoter driving protein expression should be controllable (inducible). Much of this chapter is allocated to describing appropriate host/vector/promoter combinations (see Table 1.1 for a summary).

Table 1.1 Recommended E. coli strains for membrane protein expression.

In bacteria, passage through the inner membrane Sec translocase [5] is recognized as the primary bottleneck during the overexpression of recombinant membrane protein. Yet, many other factors may contribute to a limited expression yield. There are reports of Sec-independent membrane translocation, but true host protein-independent membrane assembly by a heterologous protein has not been clearly substantiated in the literature. For example, membrane assembly of Mistic fusion proteins [6] may be initiated by the affinity of the Mistic protein for the cytoplasmic face of the E. coli inner membrane; however, proper membrane assembly of the fused protein of interest must still require assistance from the Sec translocase when large extracellular hydrophilic domains need to be translocated across the inner membrane.

Our lab has investigated several possible modes of Sec-independent membrane assembly without arriving at any evidence that a heterologous integral membrane protein can bypass the Sec translocase (unpublished data). Furthermore, we have attempted to increase the efficiency of membrane integration by overexpressing the endogenous YidC protein that is thought to aid the Sec translocase or act independently as a membrane insertase [7]. We specifically chose to study the effect of YidC on the membrane integration of phage M13 p8 fusion proteins, as p8 protein by itself requires YidC for inner membrane assembly [8]. To our surprise, a 10-fold increased level of YidC had no effect on the membrane translocation of p8-derived fusion proteins containing a C-terminal PhoA domain as a reporter. One conclusion of this experiment is that the activity of SecA ATPase may be the limiting factor for the translocation of the large hydrophilic PhoA domain. Recently we determined that the p8 fusion partner (p8CBDek described in Luo et al. 2009 [9]) utilizes the cotranslational signal recognition particle (SRP) pathway [10–12], the route traveled by most endogenous membrane proteins. During cotranslational membrane protein assembly, there is less opportunity for hydrophobic amino acid segments to aggregate after emerging from the ribosome tunnel. Perhaps the limiting factor in p8 fusion protein expression and the overexpression of most membrane proteins is simply the rate of protein translation (or efficiency of translation initiation) at the ribosome. With this thought in mind, we tested various ribosomal binding sites (RBSs) and found a distinct difference in the efficiency of p8CBDek-mediated polytopic membrane protein assembly. Strikingly, the clone containing the much weaker RBS (AGGACGGCCGGatg) produced a greater level of protein per cell after a 20-h expression period at 20 °C. In contrast, the stronger RBS provided more protein per cell in the first stage of expression, but also resulted in jamming of the translocation pathway and cessation of culture growth. Thus, the take-home message from our recent work is to express recombinant membrane proteins “in moderation.” This advice may seem obvious, but many expression systems do not allow for careful control of expression. The solution of genetically engineering the appropriate RBS for the protein of interest may not be a preferred method of optimization. Instead, a much simpler solution for expression optimization is to employ a promoter that allows fine control of the level of mRNA encoding the membrane protein of interest.

1.3 Vector/Promoter Types

The most-studied bacterial promoters are those controlling operons for sugar metabolism (lacZYA, araBAD, rhaBAD). Many variants of the lac promoter have been isolated but all suffer to some degree from the inability to completely shut off expression with the LacI repressor protein. The wild-type lac promoter is a good choice for membrane protein expression due to its moderate strength. However, very few expression vectors encode the unmodified lac promoter. Vectors pUC18/pUC19 carry a simple lac promoter, but again pUC derivatives are not good choices due to high copy number and overproduction of β-lactamase (AmpR) that enables the growth of cells lacking plasmid. Vectors utilizing modified lac promoters are highlighted in Table 1.2. The lacUV5 promoter has two mutations within the −10 region of the lac promoter. In addition, a mutation is present at −66 within the catabolite gene activator protein (CAP) binding site. These mutations increase the promoter strength relative to the wild-type lac promoter and expression from lacUV5 is less subject to catabolite repression [13]. The tac promoter was first described by deBoer et al. [14–15]. This strong promoter is a hybrid of the −10 region of the lacUV5 promoter and the −35 region of the trp promoter. Amann et al. reported that the tac promoter is at least 5 times more efficient than the lacUV5 promoter [16]. The trc promoter is equivalent to the tac promoter since the 1-bp difference in spacing between the −35 and −10 consensus sequences does not affect promoter strength [17]. Note that the tac and trc promoters are not subject to catabolite repression as the CAP binding site is missing. Ptac and Ptrc systems are generally well controlled by LacI repression. When employing any type of modified lac promoter, LacI should be overexpressed from a lacI or lacIq gene carried by the expression vector. Also, isopropyl-β-D-thiogalactopyranoside (IPTG) induction should be tested in the low range (e.g., 0, 10, 100 versus 400 µM). The lacIq mutant was reported by Calos in 1978 and this mutation is simply an “up” promoter mutation resulting in a 10-fold enhancement of LacI repressor expression [18].

Table 1.2 Common vectors/promoters/types of regulation (for more options, a comprehensive vector database is maintained by the EMBL Protein Expression and Purification Core Facility: http://www.pepcore.embl.de/strains_vectors/vectors/bacterial_expression.html).

The pQE vectors from Qiagen utilize the phage T5 promoter that is controlled by two lac operator sequences. The T5 promoter is recognized by the E. coli RNA polymerase and induction is accomplished by IPTG addition to release the Lac repressor from the dual operator sequence. Since pQE vectors do not carry the lacI gene, the host strain must supply an excess of Lac repressor. Two options exist for LacI supplementation: copropagation of multicopy pREP4 (QIAexpress manual) or use of a strain that carries the lacIq gene. Many K-12 strains (e.g., JM109) carry the lacIq gene, but few B strains offer LacI overexpression. One recommendation is NEB Express Iq, which is a BL21 derivative that carries a miniF-lacIq which does not require antibiotic selection (Table 1.1).

Guzman et al. characterized the araBAD promoter in exquisite detail in 1995 [19], and the resulting the pBAD vector series offers many options for gene cloning and expression using L-arabinose induction. Note that some pBAD vectors do not encode RBS sites, so the gene insert must contain an appropriate translation initiation sequence. When glucose is added to the outgrowth media, expression from araBAD is essentially shut off (Table 1 in Guzman et al. [19]). For many years, the araBAD system was a first choice for tightly regulated expression, as protein output appears to correlate very well with the amount of inducer (Figure 4 in Guzman et al. [19]) However, careful studies of the araBAD promoter by Siegele et al. [20] and Giacalone et al. [21] both agreed that at subsaturating levels of L-arabinose, protein expression cultures contain a mixed population with only some of the cells expressing protein. In addition, the potential for protein overexpression is generally lower when a pBAD vector is compared to T7-mediated expression from a pET construct.

A more recently characterized sugar promoter is derived from the rhamnose operon. The rhaBAD promoter is induced by L-rhamnose. When protein is expressed directly from PrhaBAD, the expression level within each cell falls within a range that correlates very well with the amount of inducer added to the culture [21]. In fact, Giacalone et al. presents convincing data that the pRHA-67 vector is more tunable and is capable of higher output than a high-copy vector containing the araBAD promoter. The pRHA-67 vector is commercially available from Xbrane Bioscience. Data presented by Haldimann et al. [22] indicates that expression from the rhaBAD promoter is very tightly regulated, yet this system also offers the potential for 5800-fold induction when glycerol is used as the primary carbon source.

The tetracycline inducible system is also very tightly regulated. Although we do not have experience with this system, Skerra et al. [23] reports that the pASK75 vector utilizing the tetA promoter/operator and encoding the cognate repressor gene (tetR) displays tightly regulated and high-level expression of heterologous protein in several E. coli K-12 and B strains. Induction is accomplished with low concentrations of anhydrotetracycline (aTc) and the induction potential is comparable to that of the lacUV5 promoter. Lutz and Bujard [24] described additional aTc-inducible vectors that make use of the engineered PLtetO-1 promoter, which is also controlled by TetR repression. The pZ vectors offer low, medium or high level expression from PLtetO-1 corresponding to the copy number dictated by the pSC101, p15A, or ColE1 origins of replication, respectively. The aTc-inducible pZ vectors require expression in strains overexpressing TetR (e.g., DH5αZ1). The pSC101 version offers the most strictly regulated expression with an induction/repression ratio of 5000.

1.4 T7 Expression System

Over the last 20 years, the most common vector series for bacterial protein expression is the pET series (plasmid for expression by T7 RNA polymerase). The T7 expression system was developed primarily by F. William Studier and colleagues at Brookhaven National Laboratory [25]. The T7 system is best recognized for the capacity to generate a high level of recombinant protein as the phage T7 RNA polymerase is very active and also very selective for phage T7 promoters (e.g., ϕ10). Therefore, T7 transcription within a bacterial cell can be specifically directed at a single promoter within the pET vector carrying the gene of interest. In most T7 expression strains, the chromosomal DE3 prophage carries the T7 RNA polymerase gene (T7 gene 1), which is expressed from the lacUV5 promoter. Since this promoter is not completely shut off by LacI, some molecules of T7 RNA polymerase are continuously expressed and are able to make considerable amounts of target mRNA in the absence of IPTG. With respect to membrane protein expression, this is an unacceptable situation. An early partial solution to this problem was to include the lacI repressor gene on the multicopy pET vector. Thus, LacI repressor protein is produced in large excess relative to its operator binding site present in the lacUV5 promoter driving T7 gene 1. Another partial solution to leaky T7 expression was the introduction of the T7–lac hybrid promoter to the pET vector series. In vectors beginning with pET-10, the lac operator sequence overlaps the T7 promoter so that excess LacI is able to inhibit T7-mediated transcription of the target gene. However, even with this improvement uninduced expression is observed in many experiments employing BL21(DE3). Uninduced expression of even mildly toxic gene products may be lethal to BL21(DE3) at the transformation step.

A very effective means to control T7 expression is to coexpress T7 lysozyme, the natural inhibitor of T7 RNA polymerase. Until recently, three types of lysozyme strains were available and all were designed to produce lysozyme at a relatively constant level: pLysS and pLysE express wild-type T7 lysozyme from a low-copy plasmid, and in NEB lysY strains, an amidase-negative variant of T7 lysozyme (K128Y) is expressed from a single-copy miniF plasmid. The K128Y variant does not degrade the peptidoglycan layer of the E. coli cell wall [26] and, accordingly, lysY results in greater overall culture stability when membrane proteins are targeted to the cell envelope. In constitutive lysozyme systems, the level of lysozyme is sufficient to sequester the basal level of T7 RNA polymerase by a 1 : 1 protein interaction. When IPTG is added, the level of T7 RNA polymerase is present in large excess and target protein expression proceeds. If a membrane protein expression plasmid does not yield transformants when using BL21(DE3) or other basic T7 expression strains, the first response should be to test transformation into a lysozyme strain. LysY or pLysS strains may yield normal colonies and express the protein of interest at moderate to high levels. Finally, it should be noted that the choice of lysY or pLysS should take into account downstream processing of cells. Strains expressing active lysozyme often lyse spontaneously upon one freeze–thaw cycle and the resulting cell pellets may be difficult to process.

1.5 Tunable T7 Expression Systems

A recent development in T7 expression is the ability to tune the level of expression. Tunable expression provides a means for optimizing the traffic flow into the membrane translocation pathway. Four commercial strains promote this feature: Tuner™ from Novagen, BL21-AI from Invitrogen, the KRX strain from Promega, and the Lemo21(DE3) strain from New England Biolabs.

The Tuner strain does not express lac permease (lacY) and this allows more uniform uptake of IPTG. However, T7 expression in Tuner strains may still be too robust for membrane protein expression unless the plasmid has a T7–lac promoter and lysozyme is coexpressed.BL21-AI offers greater potential for expressing toxic gene products as the araBAD promoter controls the expression of the T7 RNA polymerase. The associated pDEST expression vectors contain a plain T7 promoter (no lac operator site).In the Single Step KRX strain, T7 gene 1 expression is controlled by the rhaBAD promoter, so greater potential for toxic protein expression is expected. This K-12 strain has been designed for cloning and protein expression.The Lemo21(DE3) strain [27] is a tunable T7 expression strain derived from BL21(DE3). Lemo means “less is more” as often less expression results in more protein produced in the desired form. The Lemo strain is distinct from other T7 host strains since the fraction of functionally active T7 RNA polymerase is regulated by varying the level of T7 lysozyme (lysY). Fine-tuning is possible since the LysY inhibitor protein is expressed from the L-rhamnose inducible promoter. The wide-ranging expression potential of Lemo21(DE3) is sampled to find the appropriate level for each target membrane protein. When using Lemo21(DE3), expression media should lack glucose since this carbon source affects lysozyme expression from PrhaBAD.

1.6 Other Useful Membrane Protein Expression Strains

C41(DE3) and C43(DE3) have been employed as membrane protein expression strains since their isolation from parent strain BL21(DE3) in 1996 [28]. Recently, Wagner et al. [26] reported that these two strains carry mutations within the promoter driving expression of the T7 RNA polymerase. Therefore, the characteristic robust T7 expression of DE3 strains is attenuated in C41(DE3) and C43(DE3), and this accounts for the advantage observed in the expression of some toxic proteins.

More recently, the TOP10 strain was subjected to a genetic selection procedure that produced several mutant strains exhibiting improved expression of heterologous membrane protein. This work was completed by Elizabeth Massey-Gendel et al. under the direction of James Bowie at the University of California at Los Angeles (UCLA). Target membrane proteins were expressed with a C-terminal cytoplasmic fusion to mouse dihydrofolate reductase (DHFR) (providing resistance to trimethoprim) or to a kanamycin resistance protein. A positive hit in the selection was obtained when a mutant strain was capable of expressing both fusion proteins at a level sufficient to provide resistance to both drugs. Five of the selected strains have been characterized in some detail [29] and the genomes of two such strains have been sequenced. At the January 2010 Peptalk meeting in San Diego, Professor Bowie reported that his lab is currently investigating the relevance of the mutations identified in the TOP10 derivatives designated as EXP-Rv1337-1 and EXP-Rv1337-5. The results of this investigation are widely anticipated. The DE3 prophage has been added to the EXP strains so that T7 expression is possible.

The Single Protein Production System (SPP System™) was developed by Masayori Inouye [30] and is marketed by Takara Bio. This is a two-vector system suitable for use in most E. coli strains. The target protein is expressed from a vector with the cold-inducible E. coli cspA promoter, which is of course consistent with membrane protein expression. The unique, enabling feature of the SPP System is the inducible expression of a site-specific mRNA interferase (MazF) from a second plasmid, which degrades endogenous mRNA by acting at ACA sites. Accordingly, the gene of interest must be synthesized to lack ACA sequences. The net result is that the target mRNA persists and becomes a preferential substrate for the translation machinery. The elimination of most host-derived mRNA is reported to create a quasidormant cell where expression of the target membrane protein is sustained. If Sec translocase function is also sustained, then this system may offer an advantage, as the target protein should encounter less competition from endogenous proteins on the membrane translocation pathway.

1.7 Clone Stability

When expressing membrane proteins, clone stability should always be a concern. The first indication of clone toxicity is often realized during the initial cloning/transformation step. Poor transformation results may indicate that mutant genes are being selected during the cloning step, so sequence verification is always advised and is absolutely critical if the gene has been amplified by polymerase chain reaction. If a clone is suspected to be toxic, certain precautions should be followed. First, lower growth temperatures are often stabilizing. Also, it is beneficial to include 0.1% glucose in selection plates in many situations. Glucose will repress basal expression from Plac, PlacUV5, ParaBAD, and Prha. (Note: Ptac and Ptrc are not subject to glucose repression as the CAP binding site is absent from these promoters). Glucose containing plates are also advantageous when transforming clones into T7 Express and DE3 expression strains, as the T7 RNA polymerase gene is controlled by Plac and PlacUV5 in these strains, respectively. One exception is transformation into Lemo21(DE3) where glucose repression is not stabilizing. When transforming extremely toxic clones into Lemo21(DE3), 500 µM rhamnose addition to selection plates and starter cultures will reduce the basal expression to an undetectable level (Figure 1.1).

Figure 1.1 T7 expression is tightly regulated in Lemo21(DE3) cells. Whole-cell lysates were subjected to SDS–PAGE, and target protein was detected using anti-YidC serum that recognizes both endogenous wild-type YidC and recombinant 6His-YidC membrane protein expressed from pET28c. (1) No vector control indicating endogenous YidC level; (2) cells containing pET28-6hisyidC, no IPTG, no rhamnose; (3) cells containing pET28-6hisyidC, 500 µM rhamnose, no IPTG; (4) cells containing pET28-6hisyidC, 500 µM rhamnose, 400 µM IPTG. Arrow indicates YidC target.

During the outgrowth stage for protein expression, plasmid maintenance should be examined. This is especially critical when propagating AmpR vectors, as the resistance protein (β-lactamase) is secreted and ampicillin may be completely degraded. Plasmid maintenance is easily checked by plating cells at the point of induction onto drug containing plates versus nondrug plates. If a significantly lower number of colonies are counted on the drug plates (below 80% the number counted on nondrug plates), then modifications to the protocol or the clone may be necessary. If plasmid maintenance is an issue with AmpR constructs, increasing the level of ampicillin to 200 µg/ml is recommended. Alternatively, initiate growth with 100 µg/ml ampicillin and then spike in another dose (100 µg/ml) at mid-log stage. Carbenicillin (at 50–200 µg/ml) may be used in place of ampicillin. According to the Novagen pET system manual, pET (AmpR) clones may be stabilized by using high concentrations of carbenicillin and by changing the medium twice prior to induction. Carbenicillin is more stable than ampicillin in low pH conditions, which may be encountered after extended fermentation periods.

Vectors expressing KanR or CamR are preferred for creating membrane protein clones. One versatile KanR vector is pET28, which allows for simple construction of genes tagged at either end with the polyhistidine coding sequence. pBAD33 (CamR) is also a good choice as expression is tightly regulated. (Note: when cloning into the pBAD33 polylinker, a translation initiation signal (RBS site) must be included with the gene insert). In extreme cases plasmid maintenance systems can be incorporated. For example, the hok/sok system has been utilized by groups expressing G-protein-coupled receptors (GPCRs) in E. coli [31–32].

1.8 Media Types

The type of media is also an important consideration. Although the use of LB is commonly cited, we generally observe a greater level of membrane protein expression in Terrific Broth (TB). This conclusion was made after multiple expression trials using a tac promoter, which is insensitive to glucose repression. TB is a rich broth buffered by potassium phosphate and containing glycerol as a carbon source [33]. When the target protein is expressed from Plac, PlacUV5, ParaBAD, or Prha, a rich media containing a low-level of glucose may be more appropriate. With respect to controlling expression in BL21(DE3), Pan and Malcolm [34] found that 1% glucose addition to either TB or M9 starter cultures minimized basal expression to a level equal to pLysS-containing strains. These researchers further demonstrated that glucose addition is less important in a strain expressing lysozyme to control basal T7 expression. When target protein expression is driven directly from a sugar promoter, then glucose repression is advised. For example, pBAD constructs may be stabilized by growth in media containing 0.1% glucose, which should be metabolized by the point of induction with arabinose. Such a protocol leads to a discussion of “autoinduction” media [35] marketed as the Overnight Express™ autoinduction system for simplified T7 expression. The advantages of this system are: (i) manual IPTG induction is not required and (ii) expression trials are more reproducible as growth is carried out in a defined media containing a mix of carbon sources (generally glucose, lactose, and glycerol). When glucose is depleted, lactose serves to induce expression of the T7 RNA polymerase from the lacUV5 promoter in DE3 strains. The actual inducer molecule is allolactose that is produced by β-galactosidase (the lacZ gene product). Thus, Studier points out that autoinduction should be performed in strains encoding an intact lac operon. Note: The T7 Express line of strains (NEB) are not suitable for autoinduction protocols as the T7 RNA polymerase gene disrupts the lacZ open reading frame (ORF).

1.9 Fusion Partners/Membrane Targeting Peptides

A first step in cloning or characterizing heterologous membrane protein ORFs is the analysis of membrane topology using more than one algorithm. Four common predictors are: SPOCTOPUS [36] (octopus.cbr.su.se), TopPred (http://mobyle.pasteur.fr/cgi-bin/portal.py?form=toppred), Phobius [37] (phobius.sbc.su.se), and TargetP 1.1 [38] (http://www.cbs.dtu.dk/services/TargetP/). Nielsen et al.