Cell-free Protein Synthesis: Methods and Protocols [1 ed.] 9783527316496, 3527316493

With its detailed description of membrane protein expression, high-throughput and genomic-scale expression studies, both

286 38 4MB

English Pages 255 Year 2008

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

Cell-free Protein Synthesis: Methods and Protocols [1 ed.]
 9783527316496, 3527316493

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Cell-free Protein Synthesis Edited by Alexander S. Spirin and James R. Swartz

Related Titles Budisa, N.

Engineering the Genetic Code Expanding the Amino Acid Repertoire for the Design of Novel Proteins 2006 ISBN 978-3-527-31243-6

Clare, J. J., Trezise, D. J. (eds.)

Expression and Analysis of Recombinant Ion Channels From Structural Studies to Pharmacological Screening 2006 ISBN 978-3-527-31209-2

Gellissen, G. (ed.)

Production of Recombinant Proteins Novel Microbial and Eukaryotic Expression Systems 2005 ISBN 978-3-527-31036-4

Nierhaus, K. H., Wilson, D. N. (eds.)

Protein Synthesis and Ribosome Structure Translating the Genome 2004 ISBN 978-3-527-30638-1

Fischer, R., Schillberg, S. (eds.)

Molecular Farming Plant-made Pharmaceuticals and Technical Proteins 2004 ISBN 978-3-527-30786-9

Cell-free Protein Synthesis Methods and Protocols Edited by Alexander S. Spirin and James R. Swartz

The Editors Professor Dr. Alexander S. Spirin Institute of Protein Research Russian Academy of Sciences 142290 Puschchino, Moscow Region Russia Professor Dr. James R. Swartz Department of Chemical Engineering Stauffer III, Rm 113 Stanford University Stanford, CA 94305-5025 USA

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 Die Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at . © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 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. Typesetting VTEX, Vilnius, Lithuania Printing betz-druck GmbH, Darmstadt Binding Litges & Dopf GmbH, Heppenheim Cover Design Grafik-Design Schulz, Fußgönheim Printed in the Federal Republic of Germany Printed on acid-free paper ISBN: 978-3-527-31649-6

V

Contents Preface XIII List of Contributors

1

1.1 1.1.1 1.1.2 1.1.3 1.1.4 1.1.5 1.2 1.2.1 1.2.1.1 1.2.1.2 1.2.1.3 1.2.2 1.2.2.1 1.2.2.2 1.3 1.3.1 1.3.1.1 1.3.1.2 1.3.1.3 1.3.2 1.4 1.4.1 1.4.2 1.4.3 1.4.4

XVII

Cell-free Protein Synthesis Systems: Historical Landmarks, Classification, and General Methods 1 A. S. Spirin and J. R. Swartz Introduction: Historical Landmarks 1 Discovery of Protein Synthesis in Cell Extracts 1 Translation of Exogenous Messages 1 Coupled Transcription-translation in Bacterial Extracts 2 Combined Transcription-translation Systems 3 Continuous Flow/Continuous Exchange Principle 3 Prokaryotic and Eukaryotic Types of Cell-free Expression Systems 5 Cell Extracts 5 E. coli extract (ECE) 5 Wheat Germ Extract (WGE) 6 Rabbit Reticulocyte Lysate (RRL) 6 Genetic Constructs (Expression Vectors) 7 Prokaryotic Systems 7 Eukaryotic Systems 8 Preparing Cell Extracts 11 E. coli Extracts 11 Genetics 11 Cell Growth 13 Extract Preparation 14 Wheat Germ Extracts 15 Designing Reaction Composition 16 Mg2+ and Phosphate 16 Other Salts 18 Nucleotides and Amino Acids 18 Stabilization Reagents 18

Cell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

VI

Contents

1.4.5 1.5 1.5.1 1.5.2 1.6 1.6.1 1.6.2 1.6.3 1.6.4 1.6.5

Other Factors 19 Providing Energy 19 Direct Nucleotide Regeneration 20 Indirect Nucleotide Regeneration 20 Enhancing Protein Folding 21 Temperature Effects 21 Cell Extract Concentration 23 Effects of Folding Ligands 23 Effects of Chaperones and Foldases 24 Effects of Detergents 24

2

The Constructive Approach for Cell-free Translation 35 T. Ueda Introduction 35 The Process of Protein Synthesis 36 Polypeptide Synthesis 36 Protein Maturation 38 A Constructive Approach to Protein Synthesis 40 In Vitro Reconstitution of Polypeptide Synthesis 40 Protocol of Protein Synthesis using PURE System 41 Addition of Protein Folding Machinery to the PURE System 42 Integration of a Membrane Targeting System with the PURE system 46 Protein Synthesis using the PURE System containing Molecular Chaperones 48 Conclusion 49

2.1 2.2 2.2.1 2.2.2 2.3 2.3.1 2.3.2 2.3.3 2.3.4 2.3.5 2.4

3

3.1 3.1.1 3.1.2 3.2 3.2.1 3.2.2 3.2.3 3.3 3.3.1

Functional Genomic Analysis using Sequential Cell-free Protein Synthesis 51 K. A. Woodrow and J. R. Swartz Introduction 51 The Post-genomic Era 51 Cell-free Protein Synthesis (CFPS) as a Functional Proteomic Tool 52 Developing an enabling Technology for Sequential Expression Analysis 54 Improving Linear Template Stability 55 Improving PCR Reactions for generating Genomic Linear Templates 56 Optimizing Cofactor Concentrations for Enzyme Activation 58 Demonstrating Functional Genomic Analysis with CFPS 61 Isolation and Expression of Genomic Targets 62

Contents

3.3.2 3.4

4

4.1 4.1.1 4.1.2 4.2 4.2.1 4.2.2 4.2.3 4.2.4 4.2.5 4.3 4.3.1 4.3.2 4.3.3 4.4

5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12 5.13 5.14

Effects of Sample Library on β-Lactamase Expression and Activity 62 Conclusions and Projections 64

Cell-free Technology for Rapid Production of Patient-specific Fusion Protein Vaccines 69 A. R. Goerke, J. Yang, G. Kanter, R. Levy and J. R. Swartz Introduction 69 Lymphoma and Fusion Protein Vaccine Treatments 69 Comparing Cell-free and In Vivo Production Systems 70 Developing the Fusion Protein Construct and the Cell-free Production Process 71 Fusion-protein Production in the Cell-free System 71 Oxidized Reaction Conditions and DsbC Increase Soluble Protein Yield 71 GM-CSF is more Active at the N-terminus of the Fusion Protein Vaccine 73 New Linker Improves Fusion Protein Stability 75 Expression and Purification Scale-up for Vaccine Protein Production 77 Fusion Proteins Raise Protective Antibodies 78 Design of Vaccine Constructs and Mouse Studies 78 Fusion Protein Vaccination Protects against Aggressive Tumors 79 Antibody Generation is enhanced by Fusion Partners 79 Conclusions and Projections 80

Bacterial Cell-free System for Highly Efficient Protein Synthesis 83 T. Kigawa, T. Matsuda, T. Yabuki and S. Yokoyama Overview 83 Introduction 83 Coupled Transcription–Translation System based on E. coli Extract 84 DNA Template Construction 84 Preparation of Cell Extract from E. coli 85 Batch-mode Cell-free Reaction 87 Dialysis-mode Cell-free Reaction 88 Template DNA 91 Reaction Temperature 92 Surface Area of the Dialysis Membrane 93 Stable-isotope Labeling for NMR Spectroscopy 93 Selenomethionine Incorporation for X-Ray Crystallography 94 Automation 95 Conclusion 95

VII

VIII

Contents

6

6.1 6.2 6.3 6.4 6.5 6.6 6.7

7

The Use of the Escherichia coli Cell-free Protein Synthesis for Structural Biology and Structural Proteomics 99 T. Kigawa, M. Inoue, M. Aoki, T. Matsuda, T. Yabuki, E. Seki, T. Harada, S. Watanabe and S. Yokoyama Overview 99 Introduction 100 High-throughput Expression by PCR-based Small-scale Cell-free Protein Synthesis 100 Fully Automated Protein Production using Middle-scale Cell-free Protein Synthesis 103 NMR Screening 104 Large-scale Protein Production for Structure Determination 105 Discussion 107

The Wheat Germ Cell-free Protein Synthesis System 111 T. Sawasaki and Y. Endo 7.1 Overview 111 7.2 Development of a Highly Efficient Eukaryotic Cell-free Protein Synthesis System 111 7.2.1 Preparation of a Highly Active and Robust Extract from Wheat Embryos 112 7.2.1.1 Protocol for the Preparation of Wheat Germ Extract [12] 115 115 7.2.2 mRNA 5′ and 3′ UTRs which Enhance Translation 7.2.3 Split-primer PCR for Genome-wide Generation of DNAs for Transcription 119 7.2.3.1 Protocol for “Split-primer” PCR [13] 121 First PCR 122 7.2.4 Bilayer Translation Reaction Method 122 7.2.5 Transcription and Translation in One Tube 123 7.2.5.1 Protocol for One-tube Protein Synthesis Reaction 124 7.2.6 Reaction Methods for Large-scale Protein Production 125 7.3 Completion of Protocols for the Wheat Cell-free System 126 7.3.1 Performance of the Wheat Cell-free System 127 7.3.2 Robotic Automation of the Cell-free Protein Synthesis 132 7.4 Application to High-throughput Biochemical Annotation of Genetic Information 132 7.4.1 Genome-wide Functional Analysis 132 7.4.2 Preparation of Protein for NMR Spectroscopy 134 7.5 Conclusion 136

Contents

8

8.1 8.2 8.3 8.3.1 8.3.2 8.4 8.4.1 8.4.2 8.4.3 8.5 8.5.1 8.5.2 8.5.3

9

9.1 9.2 9.3 9.3.1 9.3.2 9.3.3 9.3.4 9.4 9.4.1 9.4.2 9.4.3 9.4.4 9.5

Cell-free Expression of Integral Membrane Proteins for Structural Studies 141 C. Klammt, D. Schwarz, I. Lehner, S. Sobhanifar, F. Löhr, J. Zeelen, C. Glaubitz, V. Dötsch and F. Bernhard Overview 141 Introduction 141 Specific Characteristics for the Cell-free Expression of Membrane Proteins 143 Cell-free Expression of Membrane Proteins in the Presence of Detergents or Lipids 145 Detergents for the Efficient Resolubilization of Cell-free Produced Membrane Proteins 149 Case Studies for the High Level Cell-free Expression of Membrane Proteins 150 α-Helical Transporters 150 G-Protein Coupled Receptors 152 β-Barrel Proteins 153 Structural Characterization of Cell-free Produced Membrane Proteins 154 Crystallization of Cell-free Produced Membrane Proteins 154 Cell-free Expression as a Tool for High-resolution NMR Spectroscopy 155 Applications of Cell-free Expression for Solid-state NMR 159

Cell-free Production of Membrane Proteins in the Presence of Detergents 165 J.-M. Betton and M. Miot Introduction 165 Histidine Protein Kinases 166 Materials and Methods 168 Plasmids 168 Cell-free Protein Production 168 Protein Purification 168 Structural and Functional Protein Characterizations 169 Results and Discussion 169 169 Analytical Cell-free Production of His6 -tagged Proteins Detergents Compatible with Cell-free Synthesis 171 Fidelity of In Vitro Biosynthesis Reactions in the Presence of Brij35 173 High-level Production of Functional HPKs in CECF Technology Conclusions 177

174

IX

X

Contents

10

10.1 10.2 10.3 10.3.1 10.3.2 10.3.3 10.3.4 10.4 10.5

11 11.1 11.2 11.3 11.4 11.5 11.6 11.7 11.8 11.8.1 11.8.2 11.8.3 11.8.4 11.8.5

12 12.1 12.1.1 12.1.2 12.2 12.2.1 12.2.2 12.3 12.3.1

Novel Techniques using PCR and Cell-free Protein Synthesis Systems for Combinatorial Bioengineering 179 H. Nakano and T. Yamane Introduction 179 Improvements in the Escherichia coli Cell-free Protein Synthesis Systems 180 High-throughput Construction of a Protein Library by SIMPLEX 180 Development of SIMPLEX 180 Quality of the SIMPLEX-based Protein Library 182 Expansion of the SIMPLEX-based Library 182 Application of SIMPLEX for Combinatorial Engineering of Proteins 184 Development and Application of SICREX 186 Conclusion 188

Gene Cloning and Expression in Molecular Colonies 191 A. B. Chetverin, T. R. Samatov and H. V. Chetverina A Gap in Cell-free Biotechnology 191 Molecular Colony Technique 192 Gene Cloning in Molecular Colonies 193 Gene Expression in Molecular Colonies: Transcription 196 Gene Expression in Molecular Colonies: Translation 196 Gene Expression in Molecular Colonies: The Role of Thiol Compounds 198 Conclusions 200 Molecular Colony Protocols 201 Amplification Gels 201 Growing DNA Colonies 202 Detection of Molecular Colonies 202 Transcription in Molecular Colonies 203 Protein Synthesis in Molecular Colonies 203

Large-Scale Batch Reactions for Cell-free Protein Synthesis 207 A. M. Voloshin and J. R. Swartz Introduction 207 Cell-free Protein Synthesis 207 Comparing Cell-free Reaction Configurations; Advantages of Batch Mode 208 Challenges for Extending Batch Duration and Productivity 210 Providing Energy 210 Stabilizing the Substrates 213 Scale-up of Reactions not Requiring Oxygen in Batch Mode 216 Test-tube Scale-up Results are Disappointing 216

Contents

12.3.2 12.3.3 12.4 12.4.1 12.4.2 12.4.3 12.4.4 12.4.5 12.5 12.5.1 12.5.2

Thin-film Format Conserves Performance 216 Investigating Fundamental Influences 218 Scale-up of Reactions Requiring Oxygen 218 Test-tube Scale up is Disastrous 218 Thin-film Format Conserves Performance 222 Stirred Tank Aerated Reactor Format Requires Antifoaming Agents 222 226 Enhanced O2 Transfer Increases ATP Concentrations Protein Production in 1-liter Batch Reactions 228 Conclusions and Projections 231 Personalized Medicine 231 Large-scale Pharmaceutical Production 232 Index

237

XI

XIII

Preface We are pleased to offer the following chapters as an update on the rapidly developing field of cell-free protein synthesis (CFPS), a subset of a potentially larger field termed cell-free biology. Although this collection is far from comprehensive, it is intended to provide an overview of the field by offering examples of some of the more exciting developments. CFPS is a unique embodiment of our fascination with biology and of our desire to harness biology for societal benefit. Early on, agriculture began to harness biology and dramatically changed our social structures by providing plentiful and local food supplies. In more recent times, we industrialized biology. For example, acetone/butanol fermentations helped supply munitions for World War I, and the large scale production of antibiotics saved millions of lives during and after World War II. But the most dramatic development came with the deciphering of the genetic code and the subsequent ability to reprogram living organisms. While CFPS was instrumental as a research tool in breaking the code, it languished relative to its potential as a protein production tool. Even the breakthrough developments of continuous-flow (CFCF) and continuous-exchange (CECF) operational modes failed to rapidly launch cell-free synthesis as a wide-spread research and production technology. However, over time these developments have combined with the analysis and activation of cell-free metabolism and with a variety of other advances to open exciting new opportunities for the production of complex proteins, for convenient protein evolution, for expanding our knowledge of basic biology, and even for producing complex proteins such as membrane proteins that are very difficult to produce in vivo. In this collection, we first set the stage in Chapter 1 by describing the history of cell-free protein synthesis with emphasis on providing an overview of many of the recent advances. Chapter 2 describes a particularly interesting and powerful development, the use of purified components to catalyze robust protein synthesis. Chapter 3 describes a series of developments that enable CFPS to analyze genomes for those sequences encoding gene products that impact a broad variety of central metabolic processes. Both Chapters 2 and 3 have the potential to provide important new knowledge about protein synthesis and microbial metabolism. They also both provide unique platforms for general protein synthesis. Cell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

XIV

Preface

With Chapter 4, the focus is shifted to a particular application: patient-specific medicine. CFPS offers the speed and economy required when a new protein vaccine is needed for each patient, as is the case for Non-Hodgkins Lymphoma. This chapter describes the development of complex fusion protein vaccines that require the formation of multiple disulfide bonds, but that nevertheless can be rapidly produced by CFPS. Data are also presented to show that these vaccines elicit protective immune responses in mouse tumor models. Chapter 5 provides new advances in improving the efficiency of CFPS in CECF format using bacterial cell extracts. Chapter 6 then extends that description by showing how these advances provide significant advantages in producing the proteins required for exploring structural biology and structural proteomics. Although Chapters 2 through 6 describe work with bacterial cell extracts, this is definitely not the only viable approach. Chapter 7 describes the considerable progress made with the use of wheat germ extracts in the CECF format. This eukaryotic system has been used by a variety of laboratories to produce proteins for structural determination as well as for other research applications. An impressive recent achievement of Roche Applied Science group (Penzberg, Germany) in collaboration with GeneCopoeia (Germantown, USSA) and Fulengen (Guangzhou, China) is the expression of 12,000 genes encoding for human proteins using a wheat germ CECF system. One of the most significant challenges for CFPS has been the production of membrane proteins. Chapters 8 and 9 describe the significant progress made in this pursuit using cell-free methods to avoid the product toxicity that limits in vivo production. Chapter 10 then provides impressive examples showing how complex proteins requiring disulfide bonds, special chaperones, and cofactors can be evolved to provide new catalytic function. This is all the more significant since the evolution depends upon single molecule PCR to isolate individual genes and their products in a process termed SIMPLEX (single-molecule PCR-linked in vitro expression). With a very interesting alternative approach, Chapter 11 describes the separation, amplification, and cell-free expression of individual DNA molecules within a gel matrix. In this way, the entire process of cloning, expression, and screening of genes and gene libraries can be accomplished without living cells, and, more importantly, in hours instead of days and, potentially, with more complete evaluation of the gene library. Finally, Chapter 12 addresses technology to enable the cell-free production of proteins at industrial scales and economies. Advances are discussed that allow the relatively efficient use of inexpensive raw materials to dramatically reduce costs. Efficient energy in the form of ATP is supplied by oxidative phosphorylation, but this requires oxygen supply. Thus, methods are also described that allow the use of conventional bioreactors for the large scale, economical cell-free production of proteins. Taken together, these chapters illustrate the power and versatility of CFPS. Biological evolution has provided an incredible catalog of protein structures and functions. With CFPS, we have the opportunity to directly control the environment in which these polypeptides are produced, folded, and, when appropriate, combined

Preface

with other polypeptides. We are freed from the need to maintain cell viability, and we can also channel most, if not all, of the metabolic resources to the production only of our product. We can also design and produce a system of catalysts that will remain relatively invariant over the period of protein production. These considerations suggest that CFPS will be able to produce many proteins that are difficult to produce in vivo (as is the case with membrane proteins, for example) and may also be able to produce many proteins more economically than with in vivo approaches. For the same reasons, CFPS is developing into an even more powerful research tool. The exciting advances and applications described in this book suggest that CFPS will play an expanding role in modern biotechnology. July 2007

Alexander S. Spirin, Moscow–Pushchino, Russia James A. Swartz, Stanford, USA

XV

XVII

List of Contributors

Masaaki Aoki Protein Research Group RIKEN Genomic Sciences Center 1-7-22 Suehiro-cho Tsurumi Yokohama 230-0045 Japan

Alexander B. Chetverin Institute of Protein Research Russian Academy of Sciences Pushchino Moscow Region 142290 Russia

Frank Bernhard Institute of Biophysical Chemistry Centre of Biomolecular Magnetic Resonance Johann Wolfgang Goethe-University of Frankfurt Max-von-Laue-Str. 9 60439 Frankfurt/Main Germany

Helena V. Chetverina Institute of Protein Research Russian Academy of Sciences Pushchino Moscow Region 142290 Russia

Jean-Michel Betton Unité Biochimie Structurale Institut Pasteur URA-CNRS 2185 28, rue du Docteur Roux 75724 Paris cedex 15 France Bernd Buchberger Roche Diagnostics GmbH Roche Applied Science R&D Protein Expression Nonnenwald 2 82372 Penzberg Germany

Volker Dötsch Institute of Biophysical Chemistry Centre of Biomolecular Magnetic Resonance Johann Wolfgang Goethe-University of Frankfurt Max-von-Laue-Str. 9 60439 Frankfurt/Main Germany Yaeta Endo Cell-Free Science and Technology Research Center Ehime University Matsuyama, Ehime 790-8577 Japan

Cell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

XVIII

List of Contributors

Clemens Glaubitz Institute of Biophysical Chemistry Centre of Biomolecular Magnetic Resonance Johann Wolfgang Goethe-University of Frankfurt Max-von-Laue-Str. 9 60439 Frankfurt/Main Germany

Christian Klammt Institute of Biophysical Chemistry Centre of Biomolecular Magnetic Resonance Johann Wolfgang Goethe-University of Frankfurt Max-von-Laue-Str. 9 60439 Frankfurt/Main Germany

Aaron R. Goerke Department of Chemical Engineering Stanford University Stanford California 94305 USA

Ines Lehner Institute of Biophysical Chemistry Centre of Biomolecular Magnetic Resonance Johann Wolfgang Goethe-University of Frankfurt Max-von-Laue-Str. 9 60439 Frankfurt/Main Germany

Takushi Harada Protein Research Group RIKEN Genomic Sciences Center 1-7-22 Suehiro-cho Tsurumi Yokohama 230-0045 Japan Makoto Inoue Protein Research Group RIKEN Genomic Sciences Center 1-7-22 Suehiro-cho Tsurumi Yokohama 230-0045 Japan Gregory Kanter Department of Medicine Division of Oncology Stanford University Stanford University Medical Center Stanford California 94305 USA Takanori Kigawa Protein Research Group RIKEN Genomic Sciences Center 1-7-22 Suehiro-cho Tsurumi Yokohama 230-0045 Japan

Ronald Levy Department of Medicine Division of Oncology Stanford University Stanford University Medical Center Stanford California 94305 USA Frank Löhr Institute of Biophysical Chemistry Centre of Biomolecular Magnetic Resonance Johann Wolfgang Goethe-University of Frankfurt Max-von-Laue-Str. 9 60439 Frankfurt/Main Germany Kairat Madin Roche Diagnostics GmbH Roche Applied Science R&D Protein Expression Nonnenwald 2 D-82372 Penzberg Germany

List of Contributors

Takayoshi Matsuda Protein Research Group RIKEN Genomic Sciences Center 1-7-22 Suehiro-cho Tsurumi Yokohama 230-0045 Japan

Alexander S. Spirin Institute of Protein Research Russian Academy of Sciences Puschchino Moscow Region 142290 Russia

Marika Miot Unité Biochimie Structurale Institut Pasteur URA-CNRS 2185 28, rue du Docteur Roux 75724 Paris cedex 15 France

James R. Swartz Departments of Chemical Engineering and BioEngineering Stanford University Stanford California 94305-5025 USA

Hideo Nakano Laboratory of Molecular Biotechnology Graduate School of Bioagricultural Sciences Nagoya University Furo-cho, Chikusa-ku Nagoya 464-8601 Japan

Eiko Seki Protein Research Group RIKEN Genomic Sciences Center 1-7-22 Suehiro-cho Tsurumi Yokohama 230-0045 Japan

Timur R. Samatov Institute of Protein Research Russian Academy of Sciences Pushchino Moscow Region 142290 Russia Tatsuya Sawasaki Cell-Free Science and Technology Research Center Ehime University Matsuyama 790-8577 Japan Daniel Schwarz Institute of Biophysical Chemistry Centre of Biomolecular Magnetic Resonance Johann Wolfgang Goethe-University of Frankfurt Max-von-Laue-Str. 9 60439 Frankfurt/Main Germany

Solmaz Sobhanifar Institute of Biophysical Chemistry Centre of Biomolecular Magnetic Resonance Johann Wolfgang Goethe-University of Frankfurt Max-von-Laue-Str. 9 60439 Frankfurt/Main Germany Takuya Ueda Department of Medical Genome Sciences Graduate School of Frontier Sciences The University of Tokyo 5-1-5 Kashiwanoha, Kashiwa Chiba Prefecture 277-8562 Japan Alexei M. Voloshin Departments of Chemical Engineering Stanford University Stanford California 94305-5025 USA

XIX

XX

List of Contributors

Satoru Watanabe Protein Research Group RIKEN Genomic Sciences Center 1-7-22 Suehiro-cho Tsurumi Yokohama 230-0045 Japan Kim A. Woodrow Department of Chemical Engineering Stanford University Stanford California 94305-5025 USA Tsuneo Yamane College of Bioscience and Biotechnology Chubu University Matsumoto-cho 1200, Kasugai Aichi 487-8501 Japan Takashi Yabuki Protein Research Group RIKEN Genomic Sciences Center 1-7-22 Suehiro-cho Tsurumi Yokohama 230-0045 Japan

Junhao Yang Department of Chemical Engineering Stanford University Stanford California 94305 USA Shigeyuki Yokoyama Protein Research Group RIKEN Genomic Sciences Center 1-7-22 Suehiro-cho Tsurumi Yokohama 230-0045 Japan Johan Zeelen Max-Planck-Institute of Biophysics Department for Structural Biology Max-von-Laue-Str. 3 60438 Frankfurt am Main Germany

1

1

Cell-free Protein Synthesis Systems: Historical Landmarks, Classification, and General Methods Alexander S. Spirin and James R. Swartz

1.1 Introduction: Historical Landmarks 1.1.1 Discovery of Protein Synthesis in Cell Extracts

The demonstration of the capability of disintegrated cells and cell extracts to continue protein synthesis was among the great discoveries of the early 1950s that led to the birth of molecular biology. The original observations were made independently in several laboratories working with homogenates and homogenate fractions of animal tissues [16, 132, 154, 155, 174, 175]. Shortly afterwards, disrupted bacterial cells were also shown to be capable of synthesizing proteins [42] and the fraction of ribonucleoprotein particles called ribosomes was identified as the heart of the protein-synthesizing machinery of the cell [138]. Cell extracts freed from heavy components by centrifugation at 30000g (the so-called S30 extracts) and supplemented with amino acids, ATP and GTP, were the first cell-free proteinsynthesizing systems [65, 95, 99, 100, 114, 137, 148, 168]. In those systems, however, the ribosomes just continued to translate endogenous mRNAs and elongate the polypeptides for which synthesis had already been started. Nevertheless, a high level of globin synthesis from endogenous mRNA templates could be achieved using rabbit reticulocyte lysates, and several general molecular mechanisms of protein-synthesizing machinery were studied (see, e.g., Refs. [54, 135]). 1.1.2 Translation of Exogenous Messages

The principal breakthrough in the development of cell-free protein-synthesizing systems was made in 1961 when Nirenberg and Matthaei managed to destroy endogenous mRNA in the bacterial (E. coli) extract without damaging ribosomes and the rest of the protein-synthesizing machinery [124, 125]. The ribosome run-off and the selective destruction of endogenous mRNAs was accomplished by a simple proCell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

2

1 Cell-free Protein Synthesis Systems

cedure of pre-incubation of the extract at 30–37 ◦ C. The introduction of polyribonucleotides in such mRNA-depleted extracts resulted in effective translation of exogenous messages. The experiments with translation of synthetic polyribonucleotides, such as poly(U), poly(A), poly(C) and random copolymers poly(U,C), poly(U,C,A), etc., deciphered the genetic code. Furthermore, natural alien messages, including eukaryotic mRNAs, could be successfully expressed in such bacterial extracts. Later, addition of a Ca2+ -dependent micrococcal ribonuclease with subsequent inactivation of the enzyme by removal of Ca2+ with EGTA was used for destruction of endogenous globin mRNA in the reticulocyte lysate [130]. Such mRNA-depleted animal extracts became the basis of the most efficient eukaryotic cell-free systems [55, 57, 115]. Another type of eukaryotic cell-free system was based on wheat germ extracts [106, 136]. In this case the content of active endogenous mRNA in the extract was found to be so low that there was no need for the pre-treatment procedures described above [5, 105]. 1.1.3 Coupled Transcription-translation in Bacterial Extracts

In the case of bacterial cell-free translation systems, the addition of pre-synthesized mRNA to a cell extract violates the principle of natural prokaryotic translation. In prokaryotic cells, translation of a mRNA by ribosomes is initiated soon after the beginning of its synthesis on the DNA template. The ribosomes move along mRNA chain not far behind the RNA polymerase, and both processes proceed with synchronized rates (coupled transcription-translation) [102, 181]. When pre-synthesized, complete mRNAs are used in cell-free translation systems, the initiation of translation sometimes may be hindered by mRNA folding and tertiary structure formation, especially if ribosome-binding sites are involved. The first demonstration of DNA-dependent incorporation of amino acids into synthesized proteins as well as the first evidence for the coupled transcriptiontranslation process in bacterial cell-free systems were also made by Nirenberg’s group in the beginning of the 1960s [21, 110]. In 1967, Zubay and colleagues made significant improvements [36, 97] and introduced an efficient bacterial coupled transcription-translation system for expression of exogenous DNA [27, 189]. Their system was based on crude cell-free E. coli extract containing endogenous RNA polymerase, but devoid of endogenous DNA and mRNA due to exhaustive nuclease degradation. This improved method was broadly adopted, although somewhat modified protocols were reported by others (see, e.g., Ref. [134]). Another practical version of the bacterial coupled transcription-translation system for exogenous gene expression was proposed by Gold and Schweiger [49, 50, 149], who used the mixture of isolated E. coli ribosomes, tRNA and ribosome-free supernatant (the so-called S100 extract) purified by ion-exchange chromatography from all nucleic acids, instead of the crude extract with degraded nucleic acids.

1.1 Introduction: Historical Landmarks

1.1.4 Combined Transcription-translation Systems

The next important step in the development of cell-free gene expression was the combination of a cell extract with a specific bacteriophage RNA polymerase that used a phage-specific promoter for transcription. Either SP6 polymerase [161] or T7 polymerase [30, 123] were suggested for such cell-free systems. These polymerases direct the exclusive synthesis of the proteins encoded by genes preceded by the corresponding phage promoters. Such systems possess several advantages: (a) the phage polymerases provide a higher level of transcripts than endogenous bacterial RNA polymerase; (b) the addition of rifampicin selectively inhibits the endogenous RNA polymerase, and thus there is no need to self-digest or treat cell extract for removal of endogenous DNA; (c) due to the promoter specificity of the phage RNA polymerase, only the gene of interest is expressed; (d) the systems are convenient for expression from any plasmid constructs and PCR products where the simple phage promoters are inserted; and (e) the phage RNA polymerases and DNA templates with phage promoters can be combined both with prokaryotic [123, 161] and eukaryotic [30, 161] extracts. Strictly speaking, these systems cannot be referred to as “coupled transcriptiontranslation” systems: both spatial and temporal coupling is absent in this case because the T7 and SP6 RNA polymerases work much faster than the endogenous bacterial polymerase and translation machinery, the transcripts quickly accumulate in excess over translating ribosomes, and thus mRNA is synthesized mainly in advance in such systems. The term “combined transcription-translation” is a more appropriate designation. It should be mentioned that both purified E. coli RNA polymerase [18, 28, 98, 137, 164] and animal virus-associated RNA polymerases [7, 29, 131] were also used in the combined cell-free transcription-translation systems based on eukaryotic extracts, but the use of SP6 and T7 phage polymerases proved to be the most successfull. 1.1.5 Continuous Flow/Continuous Exchange Principle

In cell-free translation and transcription-translation systems performed in a fixed volume of a test-tube (batch format) the reaction conditions change during incubation as a result of the consumption of substrates and the accumulation of products. Translation stops as soon as any essential substrate is exhausted or any product or by-product reaches an inhibiting concentration, usually after 20–60 minutes of incubation. The limited lifetimes and, as a consequence, low yields of protein products made the early batch systems useful mainly for analytical purposes and inappropriate for preparative synthesis of polypeptides and proteins. A principal breakthrough was the invention of the so-called continuous-action or continuous-duty translation [4, 160] and transcription-translation systems [8, 10] (see also Refs. [14, 15, 153, 161, 162]). Instead of incubating the reaction mixture in a fixed volume in a test-tube, the reaction was performed under conditions of per-

3

4

1 Cell-free Protein Synthesis Systems

sistent supply of the consumable substrates (amino acids, nucleoside triphosphates and energy-regenerating compounds) and with removal of the reaction products (mainly inorganic phosphates and nucleoside monophosphates, as well as polypeptide products and by-products in some reactor versions). To achieve that, a porous (ultrafiltration or dialysis) membrane was used to retain the high-molecular-weight components of the protein-synthesizing machinery (ribosomes, mRNA, ARSases, etc.) within a defined reaction compartment. The membrane separated the reaction compartment from another compartment containing a feeding solution with a reservoir of low-molecular-weight components (substrates) for the reaction. This technique was reproduced in a number of laboratories [37, 38, 66, 90, 126, 127, 171, 172, 177]. In so-called continuous-flow cell-free (CFCF) systems [8, 10, 160] the feeding solution with substrates is continuously pumped into the chamber containing the reaction mixture, and the products are continuously removed through the ultrafiltration membrane by the outgoing flow. Reactors working in pulsating [161] (see also Ref. [162]) or “discontinuous” (see Chapter 7 of this book) modes, with alternating flow-in and flow-out instead of the direct continuous flow, can also be used. In the dialysis format of the continuous systems, designated also as continuousexchange cell-free (CECF) systems, the passive (diffusional) exchange of substrates and low-molecular-weight products through a porous barrier takes place [4, 26, 107, 153, 162]. The dialysis (CECF) format was found to be much simpler and more practical than the CFCF format, and became the most widely exploited type of the continuous-action cell-free systems in laboratories (see Refs. [67, 70, 79, 103, 145] and Chapters 5–9), as well as in commercialized technologies (see, e.g., Refs. [13, 33, 34, 108]). Both simple dialysis bags and reactors with flat dialysis membranes are used. The reactors with hollow fibers were also proposed [161, 177]. Notably, the same continuous-exchange principle can be realized also without a dialysis membrane. Reactors have been proposed where the diffusional product/substrate exchange is accomplished between gel capsules that hold the protein-synthesizing mixture and the outside feeding solution [161]. Other formats include the use of Sephadex granules to retain the feeding solution while the reaction mixture occupies the inter-granule space [17], and a configuration in which the reaction mixture and feeding solution exist in two liquid layers separated only by a phase boundary [146] (see also Ref. Chapter 7). The use of the continuous-action principle in cell-free translation and transcription-translation systems maintains more or less constant reaction conditions and prolongs the active working time of the systems up to many hours or even days. As a result, the yield of the product increased to milligrams of protein per mL of incubation mixture.

1.2 Prokaryotic and Eukaryotic Types of Cell-free Expression Systems

1.2 Prokaryotic and Eukaryotic Types of Cell-free Expression Systems

Although the general mechanisms of protein biosynthesis are considered to be universal in the biological world, there are essential differences in initiation of translation and its regulation between prokaryotic (to be more exact, eubacterial) and eukaryotic organisms [159]. Correspondingly, cell-free translation (and transcriptiontranslation) systems based on either eubacterial or eukaryotic extracts are characterized by specific features in their compositions, requirements and in their fields of application. Presently, cell-free systems employing Escherichia coli extracts, on one hand, and wheat germ and rabbit reticulocyte extracts, on the other, are almost exclusively used as standards for in vitro translation systems of the prokaryotic and eukaryotic types, respectively. 1.2.1 Cell Extracts 1.2.1.1 E. coli extract (ECE) By now the use of E. coli extracts (ECE) for cell-free translation and transcriptiontranslation seems to be the most practical and efficient for in vitro synthesis of functional proteins of both prokaryotic and eukaryotic origin. Detailed knowledge of the protein-synthesizing machinery of the E. coli cells, including structure, genetics, metabolism and regulation of its components and reactions, has enabled remarkable improvements of the bacterial cell-free systems during recent years. Currently, up to 1–1.5 mg mL−1 of protein (see Chapter 12). It can be produced in an optimized batch cell-free system, and the productivity can be further enhanced up to 10 mg mL−1 of protein by the use of continuous-action reactors (see Chapters 5 and 7). Although the issue of proper protein folding remains a substantial challenge, especially for the synthesis of multi-domain and disulfide-bonded eukaryotic proteins, virtually any genetic information can be translated into a polypeptide in the ECE systems. Protein synthesis in the ECE systems has been shown to be quite tolerant to various additives, including cofactors, metabolites, unnatural amino acids, and even detergents. This implies that one can modify synthesis and folding conditions to maximize the yield of soluble and functionally active proteins. For example, reaction conditions can be modified to provide an oxidizing environment for the synthesis and folding of proteins containing multiple disulfide bonds. Certain natural modifications of synthesized proteins, such as specific phosphorylation of serine or threonine residues, can be made by addition to the system of modifying enzymes and their substrates. Co-translational and post-translational glycosylation and other complex modifications of eukaryotic proteins, however, are problematic in ECE cell-free systems. Sometimes serious problems arise with side activities of the extracts, resulting in damage to ribosomes; degradation of plasmids, mRNA, and tRNA; uncoupled hydrolysis of NTPs; and metabolic consumption of certain amino acids. More recently, several important improvements were proposed for extract preparation. One

5

6

1 Cell-free Protein Synthesis Systems

of them is the introduction of additional amounts of some amino acids into a cellfree system (see, e.g., Ref. [73]). In particular, it was shown that arginine, cysteine, serine and tryptophan, and also methionine, aspartic acid, and glutamic acid may be subjected to especially intensive degradation in ECE, and their re-addition is sufficient to stimulate or restore the ECE system activity [72, 77]. The presence of NAD, CoA, and oxalate was shown to be useful for maintenance of energy supply sub-systems (see below) in the ECE cell-free system. The use of a condensed ECE was reported to improve the productivity of the system [67, 73]. The extracts from mutant E. coli strains with reduced degradative activities were also successfully used to increase the lifetime and productivity of the cell-free system [116]. Chapters 3, 4, and 12 describe several of these improvements. 1.2.1.2 Wheat Germ Extract (WGE) Translation systems based on WGE have been widely used for the syntheses of radioactively labeled polypeptides and proteins, mostly for analytical purposes. Only recently has substantial progress in productivity been achieved for WGE cell-free systems, mainly by extending the duration of the synthesis time (lifetime of the system) (see Chapter 7). Removal of enzymes causing damage to ribosomes, mRNAs, and other components of the protein-synthesizing machinery during extract preparation is the key to keep protein synthesis running up to several days. In this way, the productivity of the WGE systems can be enhanced up to 1 mg mL−1 of product over 24 hours. The use of continuous-action reactors with continual supply of substrates further increases the duration of the synthesis up to several days, or even 1–2 weeks. Protein synthesis in WGE can be directed by either linear or circular DNA, as well as by either capped or uncapped mRNA. The WGE systems are considered to be more adapted to eukaryotic genes and messages, as compared with ECE systems. Glycosylation and other natural eukaryotic modifications of synthesized proteins, however, do not take place in the WGE systems, but the coupling of translation with the glycosylation activity of added microsomes may be feasible in the future. At present, both the WGE and ECE cell-free systems are being adapted for high-throughput proteomics. A serious shortcoming of WGE is that the quality of the extract strongly varies, depending on the sort of wheat, the growth and ripening conditions, and the batch of seeds. Moreover, the mechanical properties of the seeds and the procedures applied for isolation of embryos from endosperm must give embryos with minimal foreign material stuck to the embryo. To compensate, candidate embryo preparation can be screened for activity before use. Endo with colleagues [103] found that careful removal of endosperm from wheat embryos prior to preparing the WGE results in high stability and productivity of the CF system, probably due to elimination of translation inhibitors, such as tritin (a ribosome-inactivating protein of wheat), thionins and RNases. 1.2.1.3 Rabbit Reticulocyte Lysate (RRL) Accumulation of reticulocytes in animals is induced by chemical treatment (with phenyl hydrazine). These cells are specialized for the massive synthesis of just a few

1.2 Prokaryotic and Eukaryotic Types of Cell-free Expression Systems

special proteins (mainly globin, and also lipoxygenase) during differentiation into red blood cells. Nevertheless, the lysates of reticulocytes preserve a high capacity for effective translation of a great variety of messages. It has been reported that the synthesis rate and functional activity of the proteins produced in RRL is at least as high as in the WGE systems. So far, however, the RRL cell-free expression system has not been developed to conveniently provide preparative amounts of synthesized proteins (though in early experiments milligram quantities of globin per mL of RRL were synthesized in the CFCF system [142]). RRL is less available in large amounts and more expensive than WGE and ECE. The unnatural treatment of the animals that provide the reticulocytes is also a concern. At the same time, the coupling of translation with possible glycosylation activity of added microsomes (such as dog pancreas microsomes) may be more promising with RRL. 1.2.2 Genetic Constructs (Expression Vectors) 1.2.2.1 Prokaryotic Systems The productivity of translation depends mainly on the translation initiation rate. In bacterial extracts, as in bacterial cells in vivo, the rate of initiation is controlled by the ribosome binding site (RBS) on mRNA, which includes the initiation codon, the polypurine track of 3 to 7 nucleotides (the so-called Shine–Dalgarno sequence) 5–12 nucleotides upstream of the initiation codon, and in some cases the U-rich enhancing sequence further upstream. The “strength” of the RBS is determined by a proper nucleotide composition of the initiation codon (AUG > GUG > UUG > AUA . . .), an optimal distance between the initiation codon and the Shine–Dalgarno sequence (6–8 nucleotides), and the size of the region of the Shine–Dalgarno sequence that complements the 3′ -section of the ribosomal 16S RNA, as well as by the presence of the upstream enhancing sequence. Neither the rest of the 5′ -untranslated region of mRNA, nor the 3′ -untranslated region plays a role (except for the cases of special regulation of translation via protein repressors or riboswitches). Thus, the construction of highly expressible mRNAs for synthesis of both prokaryotic and eukaryotic proteins in a bacterial cell-free system has a few basic requirements: the coding sequence must start with the AUG codon, be preceded by a strong Shine–Dalgarno sequence at the proper distance from the AUG and be further preceded by an unstructured enhancing sequence. The well-known enhancing sequences are the so-called epsilon motif UUUAACUUUA naturally present in highly expressed late mRNAs of bacteriophage T7 [128] or the epsilonlike upstream sequences in some bacterial mRNAs (e.g., UUUUAACU and UAAUUUAC in atp polycistronic mRNA of E. coli) [112]. In addition, in many cases, it is very important for the first six to eight codons of the structural gene to be chosen to avoid mRNA secondary structure that might block ribosome binding to the mRNA. de Smit and van Duin [35] have shown the importance of this effect for in vivo translation of the MS2 phage coat protein RNA with more than a 10-fold expression impact exerted by a single base change. This effect might be expected to be even more pronounced in a cell-free system where the

7

8

1 Cell-free Protein Synthesis Systems

approximately 20-fold lower ribosome concentrations provide a much lower mass action effect, “pushing” the ribosomes onto available ribosome binding sites. One simple approach is to design the first six to eight codons to be as AT rich as possible. However, probably a better approach is to use mRNA secondary structure prediction algorithms to guide the selection of the initial coding sequence for minimal predicted secondary structure. The nucleotide sequence approximately 30 bases upstream and downstream of the Shine–Dalgarno region should be analyzed. Maintenance of the appropriate mRNA concentration during an extended runtime is another key point, especially in the continuous-action cell-free expression systems. Various RNase activities present in cell extracts usually restrict the lifetime of mRNA. Coupled and combined transcription-translation systems where mRNA is continuously renewed by ongoing transcription of a DNA template is essentially required for the continuous-action systems based on crude bacterial extracts rich in RNase activities. Coupled replication-translation cell-free systems [140, 143] could be a promising alternative in the cases when amplifiable mRNAs are available. A stable hairpin at the 3′ -end or 5′ -end of the mRNA [113] or the hybridization of the 3′ -terminal section of mRNA with a small DNA fragment [56] may contribute to the protection of mRNA against exonucleolytic degradation. mRNA can be also stabilized by insertion of the coding sequence between mutually complementary 5′ and 3′ -terminal untranslated sequences forming a stable stem helix [169]. Circular plasmids, which are better protected against nucleolytic attacks than are linearized DNA, are commonly used as templates for transcription in the prokaryotic system. 1.2.2.2 Eukaryotic Systems The two classical eukaryotic extracts, namely WGE and RRL, have an important advantage over bacterial extracts (ECE): they have much lower endogenous nuclease activities. That is why in the combined transcription-translation systems both circular and linearized plasmid DNAs, as well as PCR fragments, serve well. However, as the excess of mRNA may inhibit translation in eukaryotic systems (the so-called self-inhibition phenomenon, see below), the rate of transcription must be adjusted to the level that prevents overproducing mRNA. In CECF transcription-translation systems, the transcription rate can be easily regulated by changing Mg2+ and NTP concentrations during the expression period via dialysis; it is particularly recommended to start the combined transcription-translation run with a higher Mg2+ concentration so that transcription is initially vigorous and mRNA accumulates, and then to continue with a lower Mg2+ concentration that provides a modest level of mRNA synthesis to compensate for mRNA degradation while favoring translation (see Ref. [15]). To prolong the life-time of the translation-only CECF system, mRNA can be periodically added into the reaction mixture, e.g., by addition of mRNA every 24 h, the wheat germ CECF translation system was maintained with a constant production rate for several days [103] (see also Chapter 7). At the same time, there are several serious practical disadvantages of eukaryotic translation and transcription-translation cell-free systems as compared with prokaryotic ones, mainly as concerns construction of genetic messages. (a) First of all, effective eukaryotic message constructs are much more complex than prokary-

1.2 Prokaryotic and Eukaryotic Types of Cell-free Expression Systems

otic messages. They must include not only a coding sequence and a preceding short RBS and an adjacent enhancing sequence, but, as a common rule, also a welldefined 5′ -end to bind ribosomal particles prior to translation initiation, a rather lengthy and properly organized 5′ -untranslated region (5′ -UTR) to be scanned by non-translating ribosomal particles, the site of translation initiation with the initiation codon and a proper sequence context [86, 87], and a sufficiently long 3′ untranslated region (3′ -UTR) to encourage translation initiation and its regulation and to discourage mRNA degradation. Moreover, the 5′ -end usually must be modified by the m7 Gppp cap structure to fortify the primary binding of ribosomal particles, and the 3′ -end must be elongated by a poly(A) sequence to enhance translation initiation. (b) In contract to the prokaryotic case, mRNA sites in eukaryotic systems are recognized not simply by ribosomes themselves, but mainly by numerous ribosome-bound and mRNA-bound proteins called eukaryotic initiation factors (eIF’s). These proteins are targets for various regulatory signals in eukaryotic cells. When working with the expression of just a single gene in a eukaryotic cell-free system, with the aim to provide maximal protein production from just this gene, the regulatory interactions in the cell extract are not needed and even can be strongly impeding. That is why the genetic constructs for cell-free expression often may require an adjustment to those used for in vivo expression, and deleting some regulatory parts from 5′ - and 3′ -UTRs of the constructs can be advantageous. (c) In contrast to the situation in vivo, the incubation of a eukaryotic cell extract during cell-free protein synthesis leads both to progressive decapping of mRNA and to its deadenylylation (shorting of the poly(A)-tail), resulting in a decrease in the rate of translation initiation in the cell-free system. This is especially problematic for extended duration systems. In the cell-free systems where translation is combined with transcription, the mRNA transcripts appear without caps. The addition of enzymatic subsystems for capping and polyadenylylation into cell extracts is practically inconvenient and expensive. A more practical solution is to search for natural RNA elements or to invent artificial RNA modules that can improve capindependent and poly(A)-independent translation of the corresponding message constructs. One of the best-translated and the least down-regulated natural messages is the capped and polyadenylylated mRNA encoding for globin in the red blood cells of vertebrates. The ca. 60 nt 5′ -leader sequences (capped 5′ -UTRs) of the β-globin mRNAs were shown to provide a high rate of initiation of translation for various foreign coding sequences both in RRL and WGE (see, e.g., Refs. [6, 39]) and thus can be considered as paradigmatic capped leaders for cell-free expression of genetic constructs in eukaryotic extracts. For long-term cell-free systems, however, the problem of decapping during incubation arises. In principle, the problem of bypassing the dependence on cap could be solved by the use of special RNA modules, the so-called internal ribosome entry sites (IRESs), typical of RNAs of some animal viruses, such as picornaviruses [12, 59]. When such an IRES immediately precedes a coding sequence, the ribosome binds to it without scanning the upstream untranslated region and starts translation from the nearest AUG codon. Animal viral IRESs, however, are rather long in sequence,

9

10

1 Cell-free Protein Synthesis Systems

(up to 500 nt), have a complicated secondary/tertiary structure and usually require additional tissue-specific protein factors for their effective action. Their structural complexity and especially tissue-specificity make them impractical for general use in cell-free translation systems. Attempts to find simpler and extract-nonspecific, but universally efficient bona fide IRESs (see Ref. [88] for review) have been so far unsuccessful. But cap-independent translation should not be equated with internal initiation via IRES [58, 88]! There exist plant viruses whose RNAs effectively initiate translation without either caps or well-defined IRESs. The paradigm of capindependent [and poly(A)-independent] initiation of translation is the RNA of satellite tobacco necrosis virus (STNV) [31, 40, 167]. Another example of cap- and poly(A)-independent translation is the RNA of barley yellow dwarf virus (BYDV) [173]. In both cases ribosomes start at short uncapped 5′ -UTRs, but 3′ -UTR sequences of these viral RNAs are principally important for the high level of the capand poly(A)-independent translation of the coding sequence. Special translationenhancing domains (TED or TE) were identified within their 3′ -UTRs. It was postulated that the relatively short 5′ -leaders of STNV and BYDV RNAs work synergistically with TED or TE. When a heterologous coding sequence is inserted between them, the 5′ -untranslated regions (5′ -UTRs) of STNV RNA and BYDV RNA serve as powerful enhancers of cap-independent translation in a eukaryotic (WGE) cellfree system [31, 40, 151, 167, 173]. Most plant virus RNAs are capped, and their capped 5′ -leaders were shown to strongly stimulate translation of chimaeric mRNAs with heterologous coding sequences both in WGE and RRL [46, 63]. Nevertheless, in some cases the effects of their uncapped 5′ -leader sequences were shown to be nearly independent of the presence of cap structure. These sequences stimulated cap-independent translation in the two classical eukaryotic cell-free systems [156, 157]. However, it is well known that RNA of many plant viruses, such as tobacco mosaic virus (TMV), alfalfa mosaic virus (AMV), brome mosaic virus (BMV), and turnip yellow mosaic virus (TYMV), are not polyadenylated and, instead, have pseudoknot/tRNA-like 3′ -UTRs [41, 53, 104]. It has been demonstrated that these 3′ -UTRs can effectively replace the poly(A)-tail in translation when attached to heterologous (non-viral) coding sequences both in plant and animal systems, in vivo and in vitro [43, 45, 144, 188]. Thus, introducing the 5′ - and 3′ -elements of plant viral RNAs in mRNA constructs circumvents the problems with both capping and polyadenylylation of mRNAs in eukaryotic cell-free translation systems. The original or modified UTRs of TMV RNA − the so-called omega () leader of TMV RNA and the pseudoknot module of the 3′ -UTR of this RNA − became the most popular viral RNA elements for genetic constructs used in eukaryotic cell-free systems [43–46, 96, 156, 157, 188], including the high-throughput CECF system based on WGE [145]. Two novel and efficient leaders for eukaryotic cell-free translation have been found. The first was based on the fact that mature strongly expressed late mRNAs of pox viruses in infected cells contain homopolymeric poly(A) sequences linked to the 5′ -terminus immediately upstream of the initiation AUG codon [129, 150, 176]. In vitro experiments confirmed that 5′ -poly(A) sequences could be responsi-

1.3 Preparing Cell Extracts

ble for a high level of translation: it was demonstrated that synthetic mRNAs with a poly(A) leader, such as (A)25 , were highly expressed in cell-free systems based on either RRL or WGE [52]. The level of expression was comparable with that provided by the paradigmatic omega leader. Productive long-term translation runs in CECF format were accomplished with mRNAs equipped with a poly(A) leader and a TMV RNA 3′ -UTR [52]. The other finding was an effective cap-independent leader of non-viral origin [151]. This 61 nt sequence was derived from the 5′ -UTR of mRNA encoding for a light-emitting protein, obelin, from the hydroid polyp Obelia longissima. The leader is unique because it does not require mRNA-binding initiation factor eIF4 for effective initiation of translation. This property of the obelin mRNA leader makes it very useful for combined transcription-translation reactions where overproduction of mRNA can inhibit translation due to sequestration of eIF4 by excess mRNA (the phenomenon of self-inhibition typical of eukaryotic systems). In other words, mRNAs with this unique leader do not manifest the self-inhibition phenomenon [151].

1.3 Preparing Cell Extracts

The heart of cell-free protein synthesis is the cell-extract that provides the collection of required catalysts and cofactors. While the so-called PURE system of Ueda et al. [152] and the predecessor “proto-pure” systems [9, 11, 47, 51, 93, 94, 184] employ individually isolated ribosomes, catalysts and other macromolecules that are then mixed together, most systems use crude cell extracts. Most of the relevant catalysts are known, but, almost certainly, many that affect cell-free protein synthesis and folding have not yet been identified. Consequently, a great deal of empiricism still influences how we prepare cell extracts. Some recent work is helping to illuminate how the overall procedure influences extract performance, but this is an area where additional research is needed. The known principles, methods, and questions are briefly discussed here, using the most common systems, E. coli and wheat germ, as examples. 1.3.1 E. coli Extracts

Fortunately, we have a huge body of literature and a rich collection of genetic tools at our disposal when we work with E. coli. Several recent advances have taken advantage of these resources in producing inexpensive and high yielding cell-free extracts. 1.3.1.1 Genetics The traditional host organism is the E. coli K-12 strain A19 with the genotype E. coli K-12 Rna-19gdhA2his-95relA1spoT1metB1 [48]. Particularly in the Swartz laboratory,

11

12

1 Cell-free Protein Synthesis Systems Table 1.1 Genetic modification of E. coli K-12 strain A19.

Genetic modification

Rationale

Ref.

1. met+ 2. endA 3. fhuA (tonA) 4. speA 5. tnaA 6. sdaA 7. sdaB 8. gshA 9. gor 10. recBCD::Plac -red-kan

Avoid methionine requirement Stabilize DNA Avoid phage infection Stabilize arginine Stabilize tryptophan Stabilize serine Stabilize serine Stabilize cysteine Eliminate glutathione reductase Stabilize linear DNA templates

116 116 116 116 116 116 116 22 81 117

this strain has been extensively modified to improve the performance of the cell extracts. Table 1.1 lists the published genetic modifications and the rationale for each. The first mutation simply restores the strain to methionine prototrophy so that methionine is not required in the growth medium. The second, endA, was intended to help stabilize DNA templates. However, later work [117] showed this mutation to be of limited utility, at least for circular DNA templates. The fhuA mutation removes a surface protein from E. coli to avoid infection by several bacteriophages [25]. The next five mutations were inserted to stabilize amino acid concentrations [22, 116]. Early work by Kim and Swartz [72] suggested that arginine, cysteine, and to a lesser extent, tryptophan were disappearing during the cell-free reactions. Later, serine depletion was also recognized. After combining mutations 4–8 shown in Table 1.1, cell extracts were produced in which no amino acids were exhausted during the course of the cell-free reactions. The gor mutation removes the gene that encodes for glutathione reductase. Kim and Swartz [69] initially recognized that extracts were capable of reducing disulfide bonds when normal cytoplasmic reactions were activated. This was problematic in the production of proteins that require disulfide bond formation for folding and activation. These S–S reducing reactions can be inactivated by pretreating the cell extract with 1 mm iodoacetamide (IAM) for 30 minutes just before use. The IAM inactivates glutathione reductase and thioredoxin reductase, the enzymes responsible for reducing the S–S bonds. This treatment stabilizes the sulfhydral redox potential so that proteins with disulfide bonds can be successfully folded. Although the IAM treatment does not seriously decrease protein synthesis when phosphoenol pyruvate is the energy source, the IAM reaction prevents the use of glucose as the energy source, apparently by inactivating glyceraldehyde 3-P dehydrogenase. However, extracts with the gor deletion require only 50 µm IAM to stabilize the –SH/S–S redox potential, a treatment sufficiently mild to allow glucose utilization by the resultant extract [81]. Finally, the recBCD::Plac -red-kan genetic modification replaced E. coli’s RecBCD recombinase system with the exo and beta genes from bacteriophage lambda [117].

1.3 Preparing Cell Extracts

This modification significantly stabilized linear DNA templates, allowing approximately equivalent production of the model protein, chloramphenicol acetyl transferase (CAT) from equimolar additions of plasmid or linear templates. The strain had already been modified with a endA deletion. The operon replacement was necessary since a endArecD double deletion was not effective and recB deletions seriously affect cell vigor [183]. The genetic replacement strain, however, grew well and provided productive extracts. These mutations represent important examples, but many additional opportunities exist for introducing mutations that (a) stabilize mRNA, proteins, and nucleotide triphosphates, (b) improve protein folding and (c) provide various other useful characteristics for the cell extracts. Alternative E. coli strains can also be used. Recently, the Kim and Choi labs in Korea have collaborated to show that commercially available E. coli B stains produce effective cell extracts [2, 3, 64, 76]. The E. coli BL21(DE3) strains have the advantage of being deficient in the cytoplasmic protease, Lon. Many also have an inactivating deletion that avoids production of the outer membrane protease, OmpT. In addition, the DE3 element allows the production of T7 RNA polymerase upon induction by IPTG [2]. Consequently, with this extract it is not necessary to separately produce the polymerase to add to the reaction. BL21(DE3) stains are also commercially available with various genetic modifications. The Kim group have evaluated several, including the Origami strain (Novagen) with gortrxB mutations to help with disulfide bond formation [64], the Rosetta strain (Novagen) that carries a plasmid to express rare tRNAs [76, 78, 91], and the BL21Star strain (Invitrogen) with a mutation to stabilize mRNA. The Rosetta strain in particular produced productive extracts [76]. While the Swartz group concentrated on stabilizing the linear DNA template to enhance expression directly from PCR products, the Kim group focused on stabilizing the mRNA. They confirmed that the commonly used T7 terminator was effective in discouraging 3′ exonuclease activity [3] and showed that for many genes the rne131 mutation was also beneficial. This mutation is present in the BL21Star(DE3) strain available from Invitrogen and avoids expression of the C-terminal half of RNase E, the major E. coli endonuclease. With this modification, high product yields were obtained from linear templates. However, at least ten-fold higher template concentrations (on a molar basis) were required for the linear templates relative to plasmid-borne templates [3]. 1.3.1.2 Cell Growth Much less attention has been paid to how the cells are grown in preparation for the cell extract, but this procedure can have important influences on the activity of the resultant extract. Most cell extracts have been produced from low density cultures grown on complex media. Yamane et al. [179] examined the growth rates and the productivity of the resultant cell extracts when the cells were grown at 42 ◦ C instead of 37 ◦ C. After supplementing the caseamino acids based medium with yeast extract and with additional asparagine, glutamine, and tryptophan, the cells grew at about twice the rate at 42 ◦ C relative to that at 37 ◦ C in unsupplemented medium.

13

14

1 Cell-free Protein Synthesis Systems

The yield of CAT then increased from about 90 µg mL−1 using the standard extract to 125 µg mL−1 with the extract from the faster growing cells. Kim and Choi [77] have examined the effect of growth medium components and showed that a yeast extract/tryptone medium (2YT) produced a more productive extract when phosphate and glucose were added to produce a medium they called 2YTPG. The resultant extract had less phosphatase activity and the CAT yields increased from 180 to 245 µg mL−1 . However, probably the most careful study of growth conditions and their effects was conducted using higher cell density fermentations. Zawada, Swartz and colleagues [185, 187] have developed a defined medium supplemented with 13 amino acids to obtain high growth rates while minimizing the depletion of individual amino acids so that transient metabolic perturbations could be avoided. A computer algorithm adjusted glucose feeding rates to control growth rates with the objective of achieving high growth rates without associated acetate accumulation. With this control system, the culture reached approximately 20 g L−1 (dry weight) in less than 7 hours with a growth rate of 0.88 h−1 [185]. This was the first time that such high growth rates at high cell density had been reported. Although oxygen supplementation was not required, the commercial B. Braun fermentor was modified to increase the cross-sectional area of the baffles to stabilize the mixing patterns and increase the oxygen transfer coefficients at high agitation rates. Rapidly growing cells have higher ribosome concentrations, and it was assumed that this was why exponentially growing cultures produced more active cell extracts. This has recently been explored more carefully using computer control of growth rate to produce the cells used for extract preparation [186]. CAT and murine granulocyte macrophage colony stimulating factor (mGMCSF) were produced using the cell extracts with the PANOx energy system described below. High cell density extracts from cells grown at µ = 0.7 h−1 produced about 750 µg mL−1 of both proteins in 4 h. In contrast, extracts from cells grown at 0.3 h−1 under otherwise identical conditions produced only about 480 and 400 µg mL−1 of CAT and mGMCSF, respectively. Sucrose gradient analysis showed that the ribosome concentrations in the extracts were approximately proportional to the relative final product titers and that there were larger polysomes (more ribosomes per message) in the cell-free reactions catalyzed by the extracts from the faster grown cells. Thus, it is indeed important to maintain high growth rates to prepare productive cell extracts. 1.3.1.3 Extract Preparation Until recently, most extracts had been prepared using the procedure described by Pratt [133] which was based on the procedures of Zubay [189] and, before that, Nirenberg [124]. In fact, the overall protocol had changed little since Nirenberg’s pioneering work in 1963. The procedure was adequate for laboratory investigations, and there was only modest motivation to simplify it. However, with the recent emphasis on larger scale and higher throughput methods, the lengthy and expensive preparation procedure became a more serious impediment. The Pratt procedure requires several cell washing steps, a high-pressure homogenization step, two 30 000g centrifugations, an incubation step with expensive additions of PEP and

1.3 Preparing Cell Extracts

ATP, and rigorous dialysis of the resultant mixture followed by a final centrifugation. In a 2005 report, Liu et al. [101] describe a methodical examination of each step with the objective of simplifying the procedure and reducing cost. It was found that a single cell washing step, a high flow rate continuous homogenization, an incubation without expensive reagents, and the omission of the dialysis steps could all be implemented. The extracts produced by this new procedure were actually more productive than those from the traditional protocol [101] when used to produce CAT with the PANOx energy system. This new method decreased extract preparation time by nearly 50% and dramatically decreased reagent costs. About a year and a half later, Kim et al. [68] demonstrated an even simpler procedure. They simply lyse the cells by high-pressure homogenization (using a French Press at 20 000 psi), perform a single 12 000g centrifugation, incubate without reagent additions for 30 min, and aliquot for storage at –80 ◦ C. Again, the BL21 derived extracts prepared in this manner and energized for protein synthesis by creatine phosphate were more productive than the extracts prepared by the traditional method of Pratt [133]. However, this new procedure was not as effective for E. coli A19 derived extracts. The Liu et al. procedure was not evaluated. Taken together, the Liu et al. and Kim et al. reports suggest that simple, inexpensive cell extract procedures can help to enable the inexpensive, large-scale cell-free production of proteins. Although the preparation procedure may require optimization for different genetic backgrounds and, possibly, also for different energy generation systems, these reports help to remove one of major barriers to large-scale commercialization. 1.3.2 Wheat Germ Extracts

In 1996, Nakano et al. [121] found that wheat germ extracts could be concentrated while retaining their activity if they were precipitated in a 20% (w/w) poly(ethylene glycol) 6000 solution. The protein concentration was increased 10fold to 140 mg mL−1 , and the resultant reactions were more productive in translation reactions fueled by creatine phosphate. Phosphatase and nuclease activities were also increased by the extract condensation, but these could be partially inhibited by 2 mm Cu(OAc)2 . Endo and his colleagues were also concerned about inhibitory and hydrolytic protein contaminants and focused on removing the deleterious activities. For example, it had been shown that tritin, a protein found in wheat seeds, is a singlechain ribosome-inactivating protein (RIP) [109]. In 2000, Madin et al. reported that significantly improved protein synthesis could be obtained using extracts prepared from wheat seed embryos subjected to extensive washing [103]. The wheat seeds were first ground in a pulverizing mill and sieved. The embryos were then floated in a cyclohexane–carbon tetrachloride solvent, and damaged embryos and contaminants were discarded. The intact embryos were then washed three times, each time with 10 volumes of water with vigorous stirring, were sonicated, and then

15

16

1 Cell-free Protein Synthesis Systems

washed again. Finally, the embryos were ground to a fine powder in liquid nitrogen, suspended in a Hepes/potassium acetate buffer, and gel-filtered by a G-25 size exclusion chromatography column to remove small molecule contaminants. These extracts produced at least three times more protein product in translation reactions fueled by creatine phosphate than extracts from unwashed embryos. They also produced more polysomes, suggesting much more efficient ribosome activation. The extracts from washed embryos also worked well in a dialysis (CECF) system, suggesting their durability [103]. For example, in the dialysis mode with repeated mRNA and reagent additions, over 4 mg mL−1 of dihydrofolate reductase (DHFR) was synthesized in 60 h. While this is an impressive accomplishment, it should also be noted that the product concentration is based on a 500 µL reaction volume that contained 300 µL of cell extract and was fed with 5 mL of feeding solution that was renewed every 24 h. Thus, the production of DHFR based on the total reagent volume is more appropriately 130 µg mL−1 . The advantage of the continuous exchange system is that synthesis duration is prolonged, and the product is more concentrated. However, reagents costs often dominate in cell-free systems, and the total reagent volume therefore becomes an important consideration.

1.4 Designing Reaction Composition

As with extract preparation methods, many of the components of the cell-free chemical mixture can be traced to research conducted by Zubay [189] and Pratt [133]. This section discusses the various reaction mixtures used for E. coli and for wheat germ based protein synthesis. 1.4.1 Mg2+ and Phosphate

Divalent magnesium cations are essential for many biological reactions, particularly those involving nucleotides. Nucleotide kinases, for example, require magnesium-nucleotide complexes for catalysis of the essential energy transfer reactions. Consequently, the total added Mg2+ concentration is one of the most influential factors in the cell-free reaction mixture. Because the free [Mg2+ ] is a complicated function of interactions with many cell-free reaction components, it is often beneficial to optimize [Mg2+ ] for each lot of cell extract and cell-free reagents to obtain maximal protein synthesis. For most prokaryotic reactions, total magnesium concentrations range from about 8 to 20 mm. Stabilizing free [Mg2+ ] is particularly challenging when high energy phosphate bond donors such as phosphoenol pyruvate and creatine phosphate are used. These compounds can complex with the Mg2+ and reduce the concentration available for the energy transfer reactions. As they are used during the course of the reaction, free phosphate is released. Table 1.2 provides the log stability constants {log([MgL]/[Mg][L]} where L = the ligand that complexes with Mg2+ ) for several

1.4 Designing Reaction Composition Table 1.2 Some compounds in cell-free reactions with an affinity for Mg2+ .

Component

Log(Mg2+ complex stability constant)

PEP Creatine phosphate Phosphate Pyrophosphate AMP ADP ATP GDP Glutamic acid Acetate Oxalic acid

2.3 1.6 2.5 4.7 1.8 3.17 4.22 3.4 1.9 0.5 2.6

common components in cell-free reactions [32]. Fortunately, all cell extracts will almost certainly have a significant pyrophosphatase activity, and thus ATP (and presumably GTP) will have the highest affinity for Mg2+ . However, other components such as the phosphorylated energy source (PEP or creatine phosphate) are added in much higher concentrations, often 30–100 mm, than the nucleotides, about 1 mm each. With PEP, it would appear that in a batch reaction the decreasing [PEP] and increasing [Pi] would have a relatively small effect on magnesium availability. However, for creatine phosphate, the Mg2+ availability should decrease as the reaction proceeds and phosphate accumulates. These interactions would also explain why additional magnesium is often added along with additional energy source in many fed-batch reaction formats [61, 72, 78]. Although the affinity of glutamic acid for Mg2+ is lower than that of phosphate, glutamic acid concentrations are often the highest of any component at 100–200 mm. Glutamate will consequently sequester a portion of the Mg2+ , but the lower affinity relative to that of ATP may allow glutamate to serve as a sort of [Mg2+ ] buffer to stabilize the available [Mg2+ ]. It is also interesting to hypothesize that the high intracellular glutamates stores in most bacteria may serve this function in addition to serving as emergency energy reserves. In short, most prokaryotic cell-free systems contain many components that interact with Mg2+ , and the availability of free Mg2+ is a complicated function of the overall mixture composition in batch reactions. This has been one of the challenges in developing long lived batch reactions and explains some of the advantages of continuous exchange and continuous flow cell-free reaction modes. In contrast, the eukaryotic translation system developed by Endo has a somewhat simpler chemical composition and typically includes 2.7 mm magnesium acetate when 16 mm creatine phosphate is the energy source [103, 145]. One final consideration that is often taken for granted is the necessity of having sufficient phosphate available. Typically, excess phosphate comes either from the energy source or the nucleotides. However, with systems using nonphosphorylated energy sources and beginning with nucleotide monophosphates

17

18

1 Cell-free Protein Synthesis Systems

(to reduce costs), 10 mm phosphate was required for maximal protein production [23]. 1.4.2 Other Salts

Additional reaction components for prokaryotic systems have included, for example, potassium and ammonium acetate [73, 133], potassium glutamate at high concentrations (>200 mm) [73], and acetyl phosphate as an energy source [141]. Wheat germ systems often contain 100 mm potassium acetate [103, 145]. One common theme is that of using potassium as the dominant cation, an approach that mimics the intracellular milieu [60]. The dominant intracellular anion for prokaryotes is glutamate, and this explains why it is used in many cell-free systems [60, 73]. 1.4.3 Nucleotides and Amino Acids

These are obviously crucial reagents since they constitute the major substrates for transcription and translation reactions. Of course, the nucleotides ATP and GTP also serve as the major energy currencies. In this regard, the wheat germ translation systems have the simplest needs and typically contain 1.2 mm ATP and 0.25 mm GTP along with 0.3 mm of each of the 20 amino acids [103]. The prokaryotic combined transcription and translation systems require all four nucleotides, usually with 1.2 mm ATP and 0.85 mm of GTP, CTP, and UTP [72, 73]. The amino acids are often added at 0.5 mm each [72, 73], but increasing to 2 mm was found to be beneficial when a longer acting energy supply was provided [74]. Even when mutations to the source cells avoided amino acid depletions [22], the higher amino acid concentrations were found to slightly improve CAT titers, although this may not be true for all proteins. Certainly it is crucial to provide the stoichiometric requirements for proteins synthesis, but higher amino acid concentrations may also be advantageous if the affinities of some of the aminoacyl-tRNA synthetases for their respective amino acids are not high. 1.4.4 Stabilization Reagents

The most obvious stabilization agent in many cell-free systems is a pH buffer. Wheat germ systems often use 40 mm Hepes buffer (pKa ≈ 7.5) [103] while many prokaryotic systems use 57 mm Hepes [73, 74]. Interestingly, a new system that mimics the cytoplasmic chemical environment does not require a pH buffer [60] when using pyruvate or glutamate as the energy source. However, when glucose is used, the pH declines dramatically and a buffer with a more appropriate pKa is required. A 90 mm bis-tris buffer (pKa ≈ 6.5) provides significantly better pH stabilization than Hepes [24]. Systems with activated metabolism tend to accumulate weak acids [24, 74]. When the energy source is also a weak acid, the pH effect is

1.5 Providing Energy

neutral and a pH buffer is not required. However, when the energy source is an uncharged molecule such as glucose, the accumulation of the weak acids causes a significant pH decrease unless the system is stabilized by a pH buffer or by base addition as required. Another, less obvious need for system stabilization concerns nucleic acids. Most early prokaryotic systems contained poly(ethylene glycol) 6000 or 8000 at significant concentrations (about 2%) [133] or 0.25 mm spermidine [160]. A more recent optimization indicated that 1.5 mm spermidine and 1 mm putrescine provided maximal stimulation in a system designed to mimic the intracellular chemical environment [60]. Finally, 1–3 mm DTT is often added to translation systems to avoid the formation of disulfide bonds that might inactivate important catalysts [78, 103]. 1.4.5 Other Factors

Various additional factors have been used in many systems. Prokaryotic systems often use 34 µg mL−1 of folinic acid or a similar molecule to stimulate the initiation of translation by providing the precursor for formyl methionine formation [74, 133]. Some systems also added cyclic AMP although this is not added to most modern systems. Many systems also add partially purified tRNA at 170 µg mL−1 to stimulate translation [24, 61]. The addition of 0.33 mm NAD+ and 0.27 mm coenzyme A stimulates energy metabolism by activating the pathway from pyruvate to acetyl phosphate [74], and 2–4 mm oxalic acid was shown to further improve energy metabolism, presumably by inhibiting PEP synthase activity [60, 75, 91]. Of course, T7 RNA polymerase is often added as well at about 33 µg mL−1 when a DNA template with a T7 promoter is used [74]. As mentioned, the use of BL21 (DE3) strains avoids that requirement. Various other factors can be added to improve protein folding, as described below.

1.5 Providing Energy

Protein synthesis in an actively growing bacterium is the most energy intensive metabolic process, and we are not freed from that burden in cell-free systems. Of course, the amino acid building blocks must also be supplied. Thus, during the protein synthesis reaction, two major types of substrates are being used: amino acids as monomers for the synthesis of polypeptide chains, and nucleoside triphosphates (NTPs) as the energy supply for translation factors, for aminoacyl-tRNA synthetases and possibly for chaperones, and also as monomers for the synthesis of mRNA in transcription-translation systems. Since amino acids are used at much slower rates than the ATP and GTP, the amino acids are typically added to batch systems only at the beginning and do not require in situ synthesis. However, in most systems, in situ NTP regeneration is highly advantageous, if not essential.

19

20

1 Cell-free Protein Synthesis Systems

The CECF and CFCF systems are exceptions since the amino acids and energy substrates are constantly supplied and the products, including Pi , are removed by diffusion exchange. In this way, the steady-state concentrations of the substrates and products are maintained during the entire synthesis reaction period. Nevertheless, even in the CECF systems, a supply and removal problem may still exist if the rate of exchange is not sufficient to keep up with both the coupled and uncoupled consumption of substrates and production of side products in the reaction mixture. Owing to NTPase and phosphatase activities in cell extracts, rapid and uncoupled hydrolysis of NTPs occurs in the incubation mixture [72, 77] in addition to their productive consumption during protein synthesis. Certain amino acids are also degraded (i.e., consumed) in many cell-free systems and special prokaryotic extracts have been developed to stabilize these critical substrates [22]. However, the most critical issue is energy supply, and this is often the main limiting factor in cell-free translation and transcription-translation systems. 1.5.1 Direct Nucleotide Regeneration

To support the energy potential of the translation system, high-energy phosphate donors are often used for regeneration of nucleoside triphosphates. In wheat germ systems, creatine phosphate is the main energy substrate currently used. In bacterial systems, phosphoenol pyruvate is commonly used, but acetyl phosphate [141] and creatine phosphate [67] are other options. According to our experience, the combination of PEP and AcP described in the present protocol (Table 1.1) ensured higher activity of the E. coli system than PEP or AcP alone. As described above, one concern with these direct systems is the accumulation of phosphate in batch reactions. 1.5.2 Indirect Nucleotide Regeneration

Several novel NTP regeneration systems have been proposed that increase the yield of protein synthesis in batch versions of bacterial cell-free systems [71, 74, 75]. These regeneration systems use the oxidation of substrates from the glycolytic pathway (pyruvate, glucose-6-phosphate, or even glucose) [23, 24] to regenerate the ATP and thus avoid the accumulation of inorganic phosphate that is inhibitory to the translation system. These new systems are described more completely in Chapter 12, and will most likely also be useful for the continuous cell-free systems as well. They activate central metabolism in the cell-free reaction to provide NTPs using the same pathways employed by living cells. For example, the PANOx system activates the pathway from pyruvate to acetyl phosphate, a compound that then phosphorylates ADP [72]. With the “cytomim” system, the metabolic activation includes oxidative phosphorylation, the most efficient natural source of ATP [60]. Employing the more natural chemical environment and growing the for cells

1.6 Enhancing Protein Folding

on a glucose and phosphate based defined medium for extract preparation allows inner membrane vesicles to be activated for respiration.

1.6 Enhancing Protein Folding

In some cases a significant portion of a polypeptide synthesized in a cell-free system accumulates in a misfolded form. Such a “non-native” (denatured) protein is functionally inactive, tends to form aggregates and often precipitates. At the same time, the proportion of the misfolded protein in the product of a cell-free translation system strongly depends on the physical and chemical conditions during incubation. In most cases, optimization of the process of co-translational folding in cell-free translation systems can result in the synthesis of correctly folded, functionally active proteins with up to 90–100% of the total polypeptide product in the correctly folded form. Among the parameters to be optimized are the following: (a) the temperature of the reaction mixture during translation (not necessarily optimal for the synthesis rate!); (b) the concentration of cell-free extract used for translation (not necessarily the higher the better!); (c) the presence in the incubation mixture during translation of low-molecular-weight ligands, such as cofactors, prosthetic groups, substrates (or their analogs) of the protein synthesized; (d) the redox potential of the incubation mixture and the presence of protein disulfide isomerases, molecular chaperones and other protein folding catalysts, when necessary, and especially for the synthesis of disulfide-bonded proteins; and (e) the addition of a non-ionic detergent to the incubation mixture. In principle, functionally active proteins of different origin can be synthesized successfully in cell-free systems using either bacterial and eukaryotic extracts (see, e.g., Ref. [82]). Unsuccessful attempts to synthesize some mammalian proteins in active form using a bacterial extract were hypothesized to result from nonoptimized folding conditions, particularly a translation elongation rate that might be too fast. However, recent measurements [170] indicate that, even in prokaryotic cell-free systems, ribosomes are only adding amino acids at a rate of 1 to 2 per second, which is significantly slower even than in living eukaryotic cells. Nonetheless, folding problems can be encountered, and the following measures are suggested to increase folding success. 1.6.1 Temperature Effects

The temperature of the reaction mixture for protein synthesis is one of the most important parameters determining the correct folding of newly synthesized polypeptides. The temperature optimum for co-translational protein folding in cell-free systems is not necessarily the optimum for the synthesis rate. For instance, the synthesis rate in E. coli systems is the highest at 37 ◦ C or slightly above, but, in many cases, the yield of a functionally active protein can be higher at lower temper-

21

22

1 Cell-free Protein Synthesis Systems

Fig. 1.1 Specific activity of luciferase as a function of temperature in a cell-free translation system. Translation in 30% E. coli S30 extract was carried out at indicated temperatures in the presence of 1.2 mm ATP, 0.8 mm GTP, 0.03 mg mL−1 folinic acid, 80 mm creatine phosphate, 0.25 mg mL−1 creatine kinase, 4% PEG 8000, 0.34 mm amino acids each except leucine, 4.0 µm [14 C]leucine, and 0.175 mg mL−1 total E. coli tRNA in a buffer containing 10 mm Mg(OAc)2 , 26 mm HEPES-KOH pH20 7.6, 100 mm KOAc and 1.7 mm dithiothreitol. The reaction mixture also contained 0.1 mm luciferin. Translation was initiated by addition of mRNA (transcribed from the plasmid pT7 luc) to the concentration of 74 nm. To determine enzymatic activity of luciferase, a

10 µL aliquot was withdrawn from the translation mixture after incubation at the given temperature and immediately added to 10 µL of a solution containing 40 µm thiostrepton, 1.2 mm ATP, 0.1 mm luciferin, 10 mm Mg(OAc)2 , 26 mm HEPES-KOH pH20 7.6, 100 mm KOAc, 1.7 mm dithiothreitol; the intensity of emitted light in the aliquots was recorded at 25 ◦ C in a luminometer. The same aliquots were then analyzed with 10% SDS-PAGE, and the radioactivity of the full-length luciferase bands was determined with Cyclone estimated as the ratio of the luminescence to the amount of the synthesized protein calculated from radioactivity (T. Kazakov, V. A. Kolb, and A. S. Spirin, unpublished).

atures. Figure 1.1 presents an example. It is seen that the specific activity of firefly (Photinus pyralis) luciferase synthesized in the E. coli translation system is the highest when synthesized at 30 ◦ C and only one-third of the total synthesized protein is properly folded at either 25 or 35 ◦ C. Another example is the synthesis of human proinsulin in the E. coli translation system. Here, decreasing the temperature of incubation to 25 ◦ C resulted in significant improvement of proinsulin solubility (up to 95%), whereas the total yield of the polypeptide diminished only slightly [85]. Thus, it seems that, to synthesize a functionally active protein with the maximal specific activity, the temperature during protein synthesis in cell-free translation systems must be optimized in each case, depending on the protein to be produced. Two plausible explanations for the temperature dependence of correct protein folding can be considered. (1) For co-translational protein folding, the elongation

1.6 Enhancing Protein Folding

of a growing polypeptide at a higher temperature might be too fast, and the local secondary structure of the chain does not have sufficient time to form and thus might interfere with following sections. (2) Irrespective of elongation rate, the fast search for correct conformations requires annealing (“sub-denaturing”) conditions where a correct conformation is still stable but intermediate and incorrect conformations are unstable and transitory; it is the “sub-denaturing” temperature that may be optimal for correct protein folding during translation. 1.6.2 Cell Extract Concentration

It is often believed that the higher is the concentration of a cell extract (that includes translation factors and other components of protein-synthesizing machinery) in a cell-free system the better is the yield of a protein to be synthesized. Indeed, in several cases, concentrated cell extracts were successfully used in effective cell-free translation systems [84, 92, 120, 178]. However, this is not a common rule. For instance, with human proinsulin synthesis the proportion of the bacterial S30 extract in the transcription-translation reaction was found to be critical for synthesis of the protein in a soluble form, and the lower concentrations were more advantageous [85]. The optimal volume fraction of the extract providing both solubility and high yield of proinsulin was approximately 18%; elevation of cell extract concentration resulted neither in the improvement of protein folding nor in a significant increase in the protein yield. Thus, highly concentrated extracts are not always beneficial for the synthesis of correctly folded proteins in cell-free systems. Generally, it can be recommended to optimize the proportion of a cell extract in a cell-free incubation mixture for each type and lot of the extract and, maybe, for each protein target if protein folding problems are encountered. 1.6.3 Effects of Folding Ligands

Specific ligands, such as cofactors, prosthetic groups, substrates and their analogs, can bind to the protein product before the completion of its synthesis during translation, i.e., co-translationally, to stabilize the active conformation (see, e.g., Refs. [83, 84]). From this, unsurprisingly, the synthesis of a protein in the presence of its specific ligand can improve folding of the protein [80, 118, 119]. For example, the presence of flavin adenine dinucleotide (FAD) during the cell-free synthesis of thioredoxin reductase resulted in a four-fold increase in the specific activity of the enzyme [80]. For firefly luciferase synthesized in the presence of its substrate, luciferin, the specific activity of the enzyme was high irrespective of whether this protein was synthesized in eukaryotic or prokaryotic translation systems [82]. In contrast, the synthesis of the enzyme in the absence of luciferin produced an enzyme product with low specific activity. Interestingly, the resultant activity could be improved by the addition of trigger factor during translation [1]. On the whole, the

23

24

1 Cell-free Protein Synthesis Systems

presence of specific ligands during cell-free synthesis is highly beneficial for the correct folding of synthesized proteins. 1.6.4 Effects of Chaperones and Foldases

As shown by Ueda and coworkers, many proteins, such as dihydrofolate reductase (DHFR), λ-lysozyme and green fluorescent protein (GFP), can be synthesized and correctly folded in the cell-free system reconstituted from purified components with no molecular chaperones being present in the incubation mixture [152]. In the Spirin laboratory, the contribution of molecular chaperones of the Hsp70 family to co-translational folding of a multidomain protein, firefly luciferase, was specially assessed in the E. coli cell-free translation system preliminarily deprived of the chaperones. The specific activity of the luciferase synthesized was high in the absence of chaperones, and the addition of chaperones of the Hsp70 family did not increase it further [165]. In some cases, however, misfolding and precipitation of the product are observed. To improve the folding success, the addition of molecular chaperones to cell-free translation systems has often been practiced (see, e.g., Refs. [89, 139], and chapters in Refs. [158, 166]). The best effects were reported for the synthesis of disulfide-bonded proteins when combinations of chaperones and a protein disulfide isomerase (technically, a foldase) were used, such as in the synthesis of manganese peroxidase [118], single-chain antibodies and Fab fragments [62, 139], soluble tissue plasminogen activator [182], granulocyte-macrophage colonystimulating factor [180] and the protease domain of murine urokinase [69]. In several of these examples, it was also important to stabilize the –SH/S–S redox potential by inactivating S–S reductase activity and adding a redox buffer [69, 182]. The activities of some other proteins (e.g., mitochondrial rhodanese [89] and the catalytic subunit of human telomerase [163]) sometimes were also strongly dependent on the presence of molecular chaperones in a cell-free system, but in most cases no or only marginal effects of chaperone addition were observed for proteins without disulfide bonds. 1.6.5 Effects of Detergents

The introduction of non-ionic detergents into a cell-free translation system can be very helpful in preventing product aggregation and improving the solubility and specific activity of the proteins synthesized (see, e.g., Refs. [19, 20, 79, 111, 122] and chapters in Refs. [158, 166]). In the Spirin laboratory, the following detergents were successfully tested in the wheat germ cell-free translation system programmed with DHFR (dihydrofolate reductase) mRNA: Brij 35, digitonin, Triton X100, Nonidet P40 and octyl glucoside. No decrease in the protein synthesizing activity was detected in the system with each of these detergents at concentrations ranging from 0.2 to 1%, except for octyl glucoside (the latter inhibited translation at a 0.2% concentration and caused a complete cessation of synthesis at 1%).

References

For human proinsulin synthesis in the bacterial S30 system, the addition of 0.2% Brij 35 to the reaction mixture produced a significant increase in solubility of the synthesized protein without serious loss of the protein yield [85]. Digitonin added to the system to the concentration of 0.5% did not change the protein yield, although only a 50% increase in solubility was observed. These effects of detergents are discussed for membrane proteins in more detail in Chapters 8 and 9. In summary, a broad variety of proteins have been successfully synthesized and folded in eukaryotic and prokaryotic systems. Several examples are presented in this book, but a comprehensive summary is beyond the scope of this chapter. The preceding guidelines suggest typical considerations and present relevant examples. However, the main guiding factor is to reproduce as completely as possible the conditions, folding factors, and rates that allow the proteins to fold properly in vivo. Since we seek to “overexpress” our products, the cell-free system gives us the freedom to add cofactors at supernatural concentrations when required. Even though the chaperone concentrations are typically lower in cell-free systems, these systems are also much less crowded in total macromolecular concentrations than in living cells. This factor often provides a decisive advantage by decreasing inhibitory interactions with other proteins or with the premature forms of the desired product. As suggested, though, some empirical experimentation often enhances results.

References 1 Agashe, V. R., Guha, S., Chang, H.-C., protein-synthesizing system from wheat Genevaux, P., Hayer-Hartl, M., Stemp, germ. Methods Emymol. 101, 635–644. M., Georgopoulos, C., Hartl, F. U., Bar- 6 Annweiler, A., Hipskind, R. A., Wirth, T. (1991) A strategy for efficient in vitro ral, J. M. (2004) Function of trigger factor and DnaK in multidomain protein folding: translation of cDNAs using the rabbit bglobin leader sequence. Nucliec Acid. Res. Increase in yield at the expense of folding 19, 3750. speed. Cell 117, 199–209. 2 Ahn, J.-H., Choi, C.-Y., Kim, D.-M. (2005) 7 Ball, L. A., White, C. N. (1976) Order of transcription of genes of vesicular stomatiEffect of energy source on the efficiency of tis virus. Proc. Natl. Acad. Sci. U.S.A. 73, translational termination during cell-free protein synthesis. Biochem. Biophys. Res. 442–446. Commun. 337, 325–329. 8 Baranov, V. I., Spirin, A. S. (1993) Gene expression in cell-free systems on prepara3 Ahn, J.-H., Chu, H.-S., Kim, T.-W., Oh, I.tive scale. Methods Enzymol. 217, 123–142. S., Choi, C.-Y., Hahn, G.-H., Park, C.-G., 9 Baranov, V. I., Belitsina, N. V., Spirin, Kim, D.-M. (2005) Cell-free synthesis of A. S. (1979) The use of columns with recombinant proteins from PCR-amplified matrix-bound polyuridylic acid for isogenes at a comparable productivity to that of plasmid-based reactions. Biochem. Biolation of translating ribosomes. Methods Emymol. 59, 382–397. phys. Res. Commun. 338, 1346–1352. 10 Baranov, V. I., Morozov, I. Y., Ortlepp, 4 Alakhov, Y. B., Baranov, V. I., Ovodov, S. A., Spirin, A. S. (1989) Gene expression S. J., Ryabova, L. A., Spirin, A. S. (1995) in a cell-free system on the preparative Method of preparing polypeptides in scale. Gene 84, 463–466. cell-free translation system. United States 11 Belitsina, N. V., Glukhova, M. A., Patent # 5,478,730. 5 Anderson, C. W., Straus, J. W., DuSpirin, A. S. (1979) Elongation factor G-promoted translocation and polypepdock, B. S. (1983) Preparation of cell-free

25

26

1 Cell-free Protein Synthesis Systems

12

13

14

15

16

17

18

19

20

21

22

tide elongation in ribosomes without GTP cleavage. Methods Enzymol. 60, 761–779. Belsham, G. J., Sonenberg, N. (2000) Picornavirus RNA translation: Roles for cellular proteins. Trends Microbiol. 8, 330– 335. Betton, J.-M. (2003) Rapid translation system (RTS): A promising alternative for recombinant protein production. Curr. Prot. Pept. Sci. 4, 73–80. Birjukov, S. V., Simonenko, P. N., Shirokov, V. A., Majorov, S. G., Spirin, A. S. (2003) Method for synthesis of polypeptides in cell-free systems. United States Patent # 6,518,058 B1. Biryukov, S. V., Simonenko, P. N., Shirokov, V. A., Spirin, A. S. (2004) Method for synthesis of polypeptides in cell-free systems. United States Patent #6,783,957 B1. Borsook, H. (1950) Protein turnover and incorporation of labeled amino acids into tissue proteins in vivo and in vitro. Physiol. Rev. 30, 206–219. Buchberger, B., Mutter, W., Röder, A. (2002) Matrix reactor: A new scalable reactor principle for cell-free protein expression. In: Cell-Free Translation Systems, ed. Spirin, A. S., Springer-Verlag, Berlin, Heidelberg, pp. 121–128. Bujard, H., Gentz, R., Lanzer, M., Stueber, D., Mueller, M., Ibrahimi, I., Haeuptle, M.-T., Dobberstein, B. (1987) A T5 promoter-based transcription-translation system for the analysis of proteins in vitro and in vivo. Methods Enzymol. 155, 416– 433. Busso, D., Kim, R., Kim, S. H. (2003) Expression of soluble recombinant proteins in a cell-free system using a 96-well format. J. Biochem. Biophys. Methods 55, 233– 240. Busso, D., Kim, R., Kim, S. H. (2004) Using an Escherichia coli cell-free extract to screen for soluble expression of recombinant proteins. J. Struct. Funct. Genomics. 5, 69–74. Byrne, R., Levin, J. G., Bladen, H. A., Nirenberg, M. W. (1964) The in vitro formation of a DNA-ribosome complex. Proc. Natl. Acad. Sci. U.S.A. 52, 140–148. Calhoun, K. A., Swartz, J. R. (2006) Total amino acid stabilization during cell-free

23

24

25

26

27

28

29

30

31

32

33

34

protein synthesis reactions. J. Biotechnol. 123, 193–203. Calhoun, K., Swartz, J. (2005) An economical method for cell-free protein synthesis using glucose and nucleoside monophosphates. Biotechnol. Prog. 21, 1146–1153. Calhoun, K., Swartz, J. (2005) Energizing cell-free protein synthesis with glucose metabolism. Biotechnol. Bioeng. 5, 606– 613. Carmel, G., Coulton, J. W. (1991) Internal deletions in the FhuA receptor of Escherichia coli K-12 define domains of ligand interactions. J. Bacteriol. 173, 4394– 4403. Chekulayeva, M. N., Kurnasov, O. V., Shirokov, V. A., Spirin, A. S. (2001) Continuous-exchange cell-free proteinsynthesizing system: Synthesis of HIV-1 antigen Nef. Biochem. Biophys. Res. Commun. 280, 914–917. Chen, H. Z., Zubay, G. (1983) Prokaryotic coupled transcription-translation. Methods Enzymol. 101, 674–690. Coen, D. M., Bedbrook, J. R., Bogorad, L., Rich, A. (1977) Maize chloroplast DNA fragment encoding the large subunit of ribulosebiphosphate carboxylase. Proc. Natl. Acad. Sci. U.S.A. 74, 5487–5491. Content, J., de Witt, L., Horisberger, M. (1977) Cell-free coupling of influenza virus RNA transcription and translation. J. Virol. 22, 247–255. Craig, D., Howell, M. T., Gibbs, C. L., Hunt, T., Jackson, R. J. (1992) Plasmid cDNA-directed synthesis in a coupled eukaryotic in vitro transcription-translation system. Nucleic Acids Res. 20, 4987–4995. Danthinne, X., Seurinck, J., Meulewaeter, F., van Montagu, M., Cornelissen, M. (1993) The 3′ untranslated region of satellite tobacco necrosis virus RNA stimulates translation in vitro. Mol. Cell Biol. 13, 3340–3349. Dawson, R. M. C., Elliott, D. C., Elliot, W. H., Jones, K. M. (1990) Data for Biochemical Research, Oxford University Press, New York. Davis, J., Thompson, D., Beckler, G. S. (1996) Large scale dialysis cell-free system. Promega Notes Mag. 56, 14–21. Demidov, V. V. (2004) Get it cell-free! Innovative technologies for high-throughput

References

35

36

37

38

39

40

41

42

43

44

protein expression in vitro. In: Modern Drug Discovery, American Chemical Society, Washington D.C., pp. 61–65. De Smit, M. H., van Duin, J. (1990) Secondary structure of the ribosome binding site determines translational efficiency: A quantitative analysis. Proc. Natl. Acad. Sci. U.S.A. 87, 7668–7672. DeVries, J. K., Zubay, G. (1967) DNAdirected peptide synthesis. II. The synthesis of the α-fragment of the enzyme βgalactosidase. Proc. Natl. Acad. Sci. U.S.A. 57, 1010–1012. Endo, Y., Oka, T., Ogata, K., Natori, Y. (1993) Production of dihydrofolate reductase by an improved continuous flow cell-free translation system using wheat germ extract. Tokushima J. Exp. Med. 40, 13–17. Endo, Y., Otsuzuki, S., Ito, K., Miura, K. (1992) Production of an enzymatic active protein using a continuous flow cell-free system. J. Biotechnol. 25, 221–230. Falcone, D., Andrews, D. W. (1991) Both 5′ untranslated region and the sequences surrounding the start site contribute to efficient initiation of translation in vitro. Mol. Cell. Biol. 11, 2656–2664. Fletcher, L., Gorbin, S. D., Browning, K. G., Ravel, J. M. (1990) The absence of a m7G cap on b.-globin mRNA and alfalfa mosaic virus RNA 4 increases the amounts of initiation factor 4F required for initiation. J. Biol. Chem. 265, 19582– 19587. Florentz, C., Giegé, R. (1995) tRNA-like structures in plant viral RNAs. In: tRNA. Structure, Biosynthesis, and Function, eds. Söll, D., RajBhandary, U. L., ASM Press, Washington, D.C., pp. 141–163. Gale, E. F., Folkes, J. P. (1954) Effect of nucleic acids on protein synthesis and amino-acid incorporation in disrupted staphylococcal cells. Nature 173, 1223– 1227. Gallie, D. R., Walbot, V. (1990) RNA pseudoknot domain of tobacco mosaic virus can functionally substitute for a poly(A) tail in plant and animal cells. Genes Dev. 4, 1149–1157. Gallie, D. R., Walbot, V. (1992) Identification of the motifs within the tobacco mosaic virus 5′ -leader responsible for en-

45

46

47

48

49

50

51

52

53

54

55

hancing translation. Nucleic Acids Res. 20, 4631–4638. Gallie, D. R., Feder, J. N., Schimke, R. T., Walbot, V. (1991) Functional analysis of the tobacco mosaic virus tRNA-like structure in cytoplasmic gene regulation. Nucleic Acids Res. 19, 5031–5036. Gallie, D. R., Sleat, D. E., Watts, J. W., Turner, P. C., Wilson, T. M. A. (1987) The 5′ -leader sequence of tobacco mosaic virus RNA enhances the expression of foreign gene transcripts in vitro and in vivo. Nucleic Acids Res. 15, 3257–3272. Ganoza, M. C., Cuningham, C., Green, R. M. (1985) Isolation and point of action of a factor from Escherichia coli required to reconstruct translation. Proc. Natl. Acad. Sci. U.S.A. 82, 1648–1652. Gesteland, R. F. (1966) Isolation and characterization of ribonuclease I mutants of Escherichia coli. J. Mol. Biol. 16(1), 67–84. Gold, L. M., Schweiger, M. (1969) Synthesis of phage-specific x- and β-glucosyl transferases directed by T-even DNA in vitro. Proc. Natl. Acad. Sci. U.S.A. 62, 892– 898. Gold, L. M., Schweiger, M. (1971) Synthesis of bacteriophage-specific enzymes directed by DNA in vitro. Methods Enzymol. 20, 537–542. Green, R. H., Glick, B. R., Ganoza, M. C. (1985) Requirements for in vitro reconstruction of protein synthesis. Biochem. Biophys. Res. Commun. 126, 792–798. Gudkov, A. T., Ozerova, M. V., Shiryaev, V. M., Spirin, A. S. (2005) 5′ -poly(A) sequence as an effective leader for translation in eukaryotic cell-free systems. Biotechnol. Bioeng. 91, 468–473. Haenni, A.-L., Joshi, S., Chapeville, F. (1982) tRNA-like structures in the genomes of RNA viruses. Prog. Nucleic Acid Res. Mol. Biol. 27, 85–104. Hardesty, B., Arlinghaus, R., Shaeffer, J., Schweet, R. (1963) Hemoglobin and polyphenylalanine synthesis with reticulocyte ribosomes. Cold Spring Harbor Symp. Quant. Biol. 28, 215–222. Henshaw, E. C., Panniers, R. (1983) Translation systems prepared from the Ehrlich ascites tumor cell. Methods Emymol. 101, 616–629.

27

28

1 Cell-free Protein Synthesis Systems 56 Hirao, I., Yoshikawa, S., Miura, K. (1993) Stabilization of mRNA in an Escherichia coli cell-free translation system. FEBS Lett. 321, 169–172. 57 Jackson, R. J., Hunt, T. (1983) Preparation and use of nuclease-treated rabbit reticulocyte lysates for the translation of eukaryotic messenger RNA. Methods Enzym. 96, 50–74. 58 Jackson, R. J. (2000) A comparative view of initiation site selection mechanisms. In: Translational Control, eds. Sonenberg, N., Hershey, J. W. B., Mathews, M. B., CSHL Press, New York, pp. 127–183. 59 Jackson, R. J., Hunt, S. L., Gibbs, C. L., Kaminski, A. (1994) Internal initiation of translation of picornavirus RNAs. Mol. Biol. Rep. 19, 147–159. 60 Jewett, M., Swartz, J. (2004) Mimicking the Escherichia coli cytoplasmic environment activates long-lived and efficient cellfree protein synthesis. Biotechnol. Bioeng. 1, 19–26. 61 Jewett, M., Swartz, J. (2004) Rapid expression and purification of 100 nmol quantities of active protein using cell-free protein synthesis. Biotechnol. Prog. 20, 102–109. 62 Jiang, X., Ookubo, Y., Fujii, I., Nakano, H., Yamane, T. (2002) Expression of Fab fragment of catalytic antibody 6D9 in an Escherichia coli in vitro coupled transcription/translation system. FEBS Lett. 514, 290–294. 63 Jobling, S. A., Gehrke, L. (1987) Enhanced translation of chimaeric messenger RNAs containing a plant viral untranslated leader sequence. Nature 325, 622–625. 64 Kang, S.-H., Kim, D.-M., Kim, H.-J., Jun, S.-Y., Lee, K.-Y., Kim, H.-J. (2005) Cell-free production of aggregation-prone proteins in soluble and active forms. Biotechnol. Prog. 21, 1412–1419. 65 Keller, E. B., Zamecnik, P. C. (1956) The effect of guanosine diphosphate and triphosphate on the incorporation of labeled amino acids in protein. J. Biol. Chem. 221, 45–59. 66 Kigawa, T., Yokoyama, S. (1991) A continuous cell-free protein synthesis system for coupled transcription-translation. J. Biochem. (Jpn.) 110, 166–168.

67 Kigawa, T., Yabuki, T., Yoshida, Y., Tsutsui, M., Ito, Y., Shibata, T., Yokoyama, S. (1999) Cell-free production and stableisotope labelling of milligram quantities of proteins. FEBS Lett. 442, 15–19. 68 Kim, T.-W., Oh, I.-S., Choi, C.-Y., Park, C.-G., Kim, D.-M. (2006) Simple procedure for the construction of a robust and costeffective cell-free protein synthesis system. J. Biotechnol. 126, 554–561. 69 Kim, D.-M., Swartz, J. R. (2004) Efficient production of a bioactive, multiple disulfide-bonded protein using modified extracts of Escherichia coli. Biotechnol. Bioeng. 85, 122–129. 70 Kim, D.-M., Choi, C.-Y. (1996) A semicontinuous prokaryotic coupled transcription/translation system using a dialysis membrane. Biotechnol. Prog. 12, 645–649. 71 Kim, D.-M., Swartz, J. R. (1999) Prolonging cell-free protein synthesis with a novel ATP regeneration system. Biotechnol. Bioeng. 66, 180–188. 72 Kim, D.-M., Swartz, J. R. (2000) Prolonging cell-free protein synthesis by selective reagent additions. Biotechnol. Prog. 16, 385–390. 73 Kim, D.-M., Kigawa, T., Choi, C.-Y., Yokoyama, S. (1996) A highly efficient cell-free protein synthesis system from Escherichia coli. Eur. J. Biochem. 239, 881– 886. 74 Kim, D.-M., Swartz, J. (2001) Regeneration of adenosine triphosphate from glycolytic intermediates for cell-free protein synthesis. Biotechnol. Bioeng. 74, 4. 75 Kim, D.-M., Swartz, J. (2000) Oxalate improves protein synthesis by enchancing ATP supply in a cell-free system derived from Escherichia coli. Biotechnol. Lett. 22, 1537–1542. 76 Kim, D.-M., Choi, C.-Y., Ahn, J.-H., Kim, T.-W., Kim, N.-Y., Oh, I.-S., Park, C.-G. (2006) Development of a rapid and productive cell-free protein synthesis system. Biotechnol. Bioprocess Eng. 11, 235–239. 77 Kim, R., Choi, C. (2000) Expressionindependent consumption of substrates in cell-free expression system from Escherichia coli. J. Biotechnol. 84, 27–32. 78 Kim, T.-W., Kim, D.-M., Choi, C.-Y. (2006) Rapid production of milligram quantities of proteins in a batch cell-free protein synthesis system. J. Biotechnol. 124, 373–280.

References 79 Klammt, C., Löhr, F., Schafer, B., Haase, W., Dötsch, V., Ruterjans, H., Glaubitz, C., Bernhard, F. (2004) High level cell-free expression and specific labeling of integral membrane proteins. Eur. J. Biochem. 271, 568–580. 80 Knapp, K. G., Swartz, J. R. (2004) Cell-free production of active E. coli thioredoxin reductase and glutathione reductase. FEBS Lett. 559, 66–70. 81 Knapp, K. G., Goerke, A. R., Swartz, J. R. (2007) Cell-free synthesis of proteins that require disulfide bonds using glucose as an energy source. Biotechnol. Bioeng. 97, 901–908. 82 Kolb, V. A., Makeyev, E. V., Spirin, A. S. (2000) Co-translational folding of an eukaryotic multidomain protein in a prokaryotic translation system. J. Biol. Chem. 275, 16597–16601. 83 Komar, A. A., Kommer, A., Krasheninnikov, I. A., Spirin, A. S. (1993) Cotranslational heme binding to nascent globin chains. FEBS Lett. 326, 261–263. 84 Komar, A. A., Kommer, A., Krasheninnikov, I. A., Spirin, A. S. (1997) Cotranslational folding of globin. J. Biol. Chem. 272, 10646–10651. 85 Kommer, A., Dashkova, I. G., Yesipov, R. S., Miroshnikov, A. I., Spirin, A. S. (2005) Synthesis of functionally active human proinsulin in a cell-free translation system. Dokl. Akad. Nauk 401, 1–5 (Dokl. Biochem. Biophys. 401, 154–158). 86 Kozak, M. (1989) Context effects and inefficient initiation at the non-AUG codons in eucaryotic cell-free translation systems. Mol. Cell. Biol. 9, 5073–5080. 87 Kozak, M. (1999) Initiation of translation in prokaryotes and eukaryotes. Gene 234, 187–208. 88 Kozak, M. (2001) New ways of initiating translation in Eukaryotes? Mol. Cell. Biol. 21, 1899–1907. 89 Kudlicki, W., Mouat, M., Walterscheid, J. P., Kramer, G., Hardesty, B. (1994) Development of a chaperone-deficient system by fractionation of a prokaryotic coupled transcription/translation system. Anal. Biochem. 217, 12–19. 90 Kudlicki, W., Kramer, G., Hardesty, B. (1992) High efficiency cell-free synthesis of proteins: Refinement of the coupled

transcription/translation system. Anal. Biochem. 206, 389–393. 91 Kuem, J.-W., Kim, T.-W., Park, C.-G., Choi, C.-Y., Kim, D.-M. (2006) Oxalate enchances protein synthesis in cell-free synthesis system utilizing 3-phosphoglycerate as energy source. J. Biosci. Bioeng. 2, 162–165. 92 Kung, H.-F., Chu, F., Caldwell, P., Spears, C., Treadwell, B. V., Eskin, B., Brot, N., Weissbach, H. (1978) The mRNA-directed synthesis of the x-peptide of β-galactosidase, ribosomal protein L12 and L10, and elongation factor Tu, using purified translational factors. Arch. Biochem. Biophys. 187, 457–463. 93 Kung, H.-F., Redfield, B., Weissbach, H. (1979) DNA-directed in vitro synthesis of β-galactosidase. J. Biol. Chem. 254, 8404–8408. 94 Kung, H.-F., Redfield, B., Treadwell, B. V., Eskin, B., Spears, C., Weissbach, H. (1977) DNA-directed in vitro synthesis of β-galactosidase. J. Biol. Chem. 252, 6889–6894. 95 Lamborg, H., Zamecnik, P. C. (1960) Amino acid incorporation into protein by extracts of E. coli. Biochim. Biophys. Acta 42, 206–211. 96 Leathers, V., Tanquay, R., Kobayashi, M., Gallie, D. R. (1993) A phylogenetically conserved sequence within viral 3′ untranslated pseudoknots regulates translation. Mol. Cell. Biol. 13, 5331–5347. 97 Lederman, M., Zubay, G. (1967) DNAdirected peptide synthesis. I. A comparison of T2 and Escherichia coli DNAdirected peptide synthesis in two cell-free systems. Biochim. Biophys. Acta 149, 253– 258. 98 Lewis, J. B., Anderson, C. W., Atkins, J. F., Gesteland, R. F. (1975) The origin and destiny of adenovirus proteins. Cold Spring Harbor Symp. Quant. Biol. 39, 581– 590. 99 Littlefield, J. W., Keller, E. B. (1957) Incorporation of C14 amino acids into ribonucleoprotein particles from the Ehrlich mouse ascites tumor. J. Biol. Chem. 224, 13–30. 100 Littlefield, J. W., Keller, E. B., Gross, J., Zamecnik, P. C. (1955) Studies on cytoplasmic ribonucleoprotein particles from the liver of the rat. J. Biol. Chem. 217, 111–123.

29

30

1 Cell-free Protein Synthesis Systems 101 Liu, D., Zawada, J., Swartz, J. (2005) Streamlining Escherichia coli S30 extract preparation for economical cell-free protein synthesis. Biotechnol. Prog. 21, 460–465. 102 Maaløe, O., Kjelgaard, N. O. (1968) Control of Macro-Molecular Synthesis. Benjamin, New York. 103 Madin, K., Sawasaki, T., Ogasawara, T., Endo, Y. (2000) A highly efficient and robust cell-free protein synthesis system prepared from wheat embryos: Plants apparently contain a suicide system directed at ribosomes. Proc. Natl. Acad. Sci. U.S.A. 97, 559–564. 104 Mans, R. M. V., Pleij, C. W. A., Bosch, L. (1991) tRNA-like structures. Structure, function and evolutionary significance. Eur. J. Biochem. 201, 303–324. 105 Marcus, A., Efron, D., Weeks, D. P. (1974) The wheat embryo cell-free system. Methods Enzymol. 30, 749–754. 106 Marcus, A., Luginbill, B., Feeley, J. (1968) Polysome formation with tobacco mosaic virus RNA. Proc. Natl. Acad. Sci. U.S.A. 59, 1243–1250. 107 Martemyanov, K. A., Shirokov, V. A., Kurnasov, O. V., Gudkov, A. T., Spirin, A. S. (2001) Cell-free production of biologically active polypeptides: Application to the synthesis of antibacterial polypeptide cecropin. Protein Expr. Purif. 21, 456–461. 108 Martin, G. A., Kawaguchi, R., Lam, Y., DeGiovanni, A., Fukushima, M., Mutter, W. (2001) High-yield, in vitro protein expression using continuous-exchange, coupled transcription/translation system. BioTechniques 31, 948–953. 109 Massiah, A., Hartley, M. (1995) Wheat ribosome-inactivating proteins: Seed and leaf forms with different specifications and cofactor requirements. Planta 197, 633–640. 110 Matthaei, J. H., Nirenberg, M. W. (1961) Characteristics and stabilization of DNAase-sensitive protein synthesis in E. coli extracts. Proc. Natl. Acad. Sci. U.S.A. 47, 1580–1588. 111 Maurer, P., Moratzky, A., FecherTrost, C., Flockerzi, V., Lenk, U., Sommer, T., Volzing, C., Zimmermann, R. (2003) Cell-free synthesis of membrane proteins on a preparative scale. In: Cell-

112

113

114

115

116

117

118

119

120

121

Free Protein Expression, ed. Swartz, J. R., Springer-Verlag, Berlin, Heidelberg, New York, pp. 133–139. McCarthy, J. E. G., Sebald, W., Gross, G., Lammers, R. (1986) Enhancement of translational efficiency by the Escherichia coli aptE translational initiation region: its fusion with two human genes. Gene 41, 201–206. McLaren, R. S., Newbury, S. F., Dance, G. S. C., Causton, H. C., Higgins, C. F. (1991) mRNA degradation by processive 3′ -5′ exoribonucleases in vitro and implications for prokaryotic mRNA decay in vivo. J. Mol. Biol. 221, 81–95. McQuillen, K., Roberts, R. B., Britten, R. J. (1959) Synthesis of nascent protein by ribosomes in E. coli. Proc. Natl. Acad. Sci. U.S.A. 45, 1437–1447. Merrick, W. C. (1983) Translation of exogenous mRNAs in reticulocyte lysates. Methods Enzymol. 101, 606–615. Michel-Reydellet, N., Calhoun, K., Swartz, J. (2004) Amino acid stabilization for cell-free protein synthesis by modification of the Escherichia coli genome. Metab Eng. 6(3), 197–203. Michel-Reydellet, N., Woodrow, K., Swartz, J. R. (2005) Increasing PCR fragment stability and protein yields in a cell-free system with genetically modified Escherichia coli extracts. J. Mol. Microbiol. Biotechnol. 9, 26–34. Miyazaki-Imamura, C., Oohira, K., Kitagawa, R., Nakano, H., Yamane, T., Takahashi, H. (2003) Improvement of H2 O2 stability of manganese peroxidase by combinatorial mutagenesis and high throughput screening using in vitro expression with protein disulfide isomerase. Protein Eng. 16, 423–428. Mouat, M. F. (2000) Dihydrofolate influences the activity of Escherichia coli dihydrofolate reductase synthesized de novo. Int. J. Biochem. Cell Biol. 32, 327– 337. Nakano, H., Tanaka, T., Kawarasaki, Y., Yamane, T. (1994) An increased rate of cell-free protein synthesis by condensing wheat-germ extract with ultrafiltration membranes. Biosci. Biotechnol. Biochem. 58, 631–634. Nakano, H., Tanaka, T., Kawarasaki, Y., Yamane, T. (1996) Highly productive

References

122

123

124

125

126

127

128

129

130

131

cell-free protein synthesis system using condensed wheat-germ extract. J. Biotechnol. 46, 275–282. Nemetz, C., Wessner, S., Krupka. S., Watzele, M., Mutter, W. (2002) Cell-free expression of soluble human erythropoietin. In: Cell-Free Translation Systems, ed. Spirin, A. S., Springer-Verlag, Heidelberg, Berlin, New-York, pp. 157–163. Nevin, D. E., Pratt, J. M. (1991) A coupled in vitro transcription-translation system for the exclusive synthesis of polypeptides expressed from the T7 promoter. FEBS Lett. 291, 259–263. Nirenberg, M. (1963) Cell-free protein synthesis directed by messenger RNA. Methods Enzymol. 6, 17–23. Nirenberg, M. W., Matthaei, J. H. (1961) The dependence of cell-free protein synthesis in E. coli upon naturally occurring or synthetic polynucleotides. Proc. Natl. Acad. Sci. U.S.A. 47, 1588– 1602. Nishimura, N., Kitaoka, Y., Niwano, M. (1995) Cell-free system derived from heat-shocked Escherichia coli: Synthesis of enzyme protein possessing higher specific activity. J. Ferment. Bioeng. 79, 131–135. Nishimura, N., Kitaoka, Y., Mimura, A., Takahara, Y. (1993) Continuous protein synthesis system with Escherichia coli S30 extract containing endogenous T7 RNA polymerase. Biotechnol. Lett. 15, 785–790. Olins, P. O., Rangwala, S. H. (1989) A novel sequence element derived from bacteriophage T7 mRNA acts as an enhancer of translation of the lacZ gene in Escherichia coli. J. Biol. Chem. 264, 16973–16976. Patel, D. D., Pickup, D. J. (1987) Messenger RNAs of a strongly-expressed late gene of cowpox virus contain 5′ -terminal poly(A) sequences. EMBO J. 6, 3787– 3794. Pelham, H. R. B., Jackson, R. J. (1976) An efficient mRNA-dependent translation system from reticulocyte lysates. Eur. J. Biochem. 67, 247–256. Pelham, H. R. B., Sykes, J. M. M., Hunt, T. (1978) Characteristics of a coupled cell-free transcription and translation

132

133

134

135

136

137

138

139

140

system directed by vaccinia cores. Eur. J. Biochem. 82, 199–209. Peterson, E. A., Greenberg, D. M. (1952) Characteristics of the amino acid-incorporating system of liver homogenates. J. Biol. Chem. 194, 359–375. Pratt, J. M., (1984) Coupled transcription – translation in prokaryotic cell-free systems. In: Transcription and Translation: A Practical Approach, eds. Hames, B. D., Higgins, S. J., IRL Press, Oxford, UK, pp. 179–209. Pratt, J. M., Boulnois, G. J., Darby, V., Holland, I. B. (1984) Identification of gene products by coupled transcriptiontranslation of DNA fragments in cell-free extracts of E. coli. In: Advanced Molecular Genetics, eds. Pühler, A., Timmis, K. N., Springer-Verlag, Berlin, Germany, pp. 239–247. Rich, A., Warner, J., Goodman, H. M. (1963) The structure and function of polyribosomes. Cold Spring Harbor Symp. Quant. Biol. 28, 269–285. Roberts, B. E., Paterson, B. M. (1973) Efficient translation of tobacco mosaic virus RNA and rabbit globin 9S RNA in a cell-free system from commercial wheat germ. Proc. Natl. Acad. Sci. U.S.A. 70, 2330–2334. Roberts, B. E., Corecki, M., Mulligan, R. C., Danna, K. J., Rozenblatt, S., Rich, A. (1975) Simian virus 40 DNA directs synthesis of authentic viral polypeptides in a linked transcription-translation cell-free system. Proc. Natl. Acad. Sci. U.S.A. 72, 1922–1926. Roberts, R. B. (ed.) (1958) Microsomal Particles and Protein Synthesis, Pergamon Press, New York. Ryabova, L. A., Desplancq, D., Spirin, A. S., Pluckthun, A. (1997) Functional antibody production using cell-free translation: Effects of protein disulfide isomerase and chaperones. Nat. Biotechnol. 15, 79–84. Ryabova, L. A., Morozov, I. Y., Spirin, A. S. (1998) Continuous-flow cell-free translation, transcription-translation, and replication–translation systems. In: Methods in Molecular Biology, Vol. 77, Protein Synthesis: Methods and Protocols, ed. Martin, R., Humana Press, Totowa, NJ, pp. 179–193.

31

32

1 Cell-free Protein Synthesis Systems 141 Ryabova, L., Vinokurov, L., Shekhovtsova, E., Alakhov, Y., Spirin, A. (1994) Acetyl phosphate as an energy source for bacterial cell-free translation systems. Anal. Biochem. 226, 184–186. 142 Ryabova, L. A., Ortlepp, S. A., Baranov, V. I. (1989) Preparative synthesis of globin in a continuous cell-free translation system from rabbit reticulocytes. Nucleic Acids Res. 17, 4412. 143 Ryabova, L., Volianik, E., Kurnasov, O., Spirin, A. S., Wu, Y., Kramer, F. R. (1994) Coupled replication–translation of amplifiable messenger RNA: A cell-free protein synthesis system that mimics viral infection. J. Biol. Chem. 269, 1501– 1505. 144 Ryabova, L. A., Torgashov, A. F., Kurnasov, O. V., Bubunenko, M. G., Spirin, A. S. (1993) The 3′ -terminal region of alfalfa mosaic virus RNA4 facilitates the RNA entry into translation in a cell-free system. FEBS Lett. 326, 264–266. 145 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. 146 Sawasaki, T., Hasegawa, Y., Tsuchimochi, M., Kamura, N., Ogasawara, T., Kuroita, T., Endo, Y. (2002) A bi-layer cell-free protein synthesis system for high-throughput screening of gene products. FEBS Lett. 514, 102–105. 147 Schachtschabel, D., Zillig, W. (1959) Untersuchungen zur biosynthese der proteine. I. Uber den einbau C14markierter aminosauren ins protein zellfreier nucleoproteid-enzyme-systeme aus E. coli B. Hoppe-Seyler’s Z. Physiol. Chem. 314, 262–275. 148 Schweet, R., Lamfrom, H., Allen, E. (1958) The synthesis of hemoglobin in a cell-free system. Proc. Natl. Acad. Sci. U.S.A. 44, 1029–1035. 149 Schweiger, M., Gold, L. M. (1969) Bacteriophage T4 DNA-dependent in vitro synthesis of lysozyme. Proc. Natl. Acad. Sci. U.S.A. 63, 1351–1358. 150 Schwer, B., Visca, P., Vos, J. C., Stunnenberg, H. G. (1987) Discontinuous transcription or RNA processing of vaccinia virus late messengers results in a 5′ poly(A) leader. Cell 50, 163–169.

151 Shaloiko, L. A., Granovsky, I. E., Ivashina, T. V., Ksenzenko, V. N., Shirokov, V. A., Spirin, A. S. (2004) Effective non-viral leader for cap-independent translation in a eukaryotic cell-free system. Biotechnol. Bioeng. 88, 730–739. 152 Shimizu, Y., Inoue, A., Tomary, Y., Suzuki, T., Yokogawa, T., Nishikawa, K., Ueda, T. (2001) Cell-free translation reconstituted with purified components. Nat. Biotechnol. 19, 751–755. 153 Shirokov, V. A., Simonenko, P. N., Biryukov, S. V., Spirin, A. S. (2002) Continuous-flow and continuousexchange translation systems and bioreactors. In: Cell-Free Translation Systems, ed. Spirin, A. S. Springer-Verlag, Berlin, Heidelberg, pp. 91–107. 154 Siekevitz, P. (1952) Uptake of radioactive alanine in vitro into proteins of rat liver fractions. J. Biol. Chem. 195, 549– 565. 155 Siekevitz, P., Zamecnik, P. C. (1951) In vitro incorporation of l-C14-DL-alanine into proteins of rat-liver granular fractions. Fed. Proc. 10, 246–247. 156 Sleat, D. E., Gallie, D. R., Jefferson, R. A., Bevan, M. W., Turner, P. C., Wilson, T. M. A. (1987) Characterization of the 5′ -leader sequence of tobacco mosaic virus RNA as a general enhancer of translation in vitro. Gene 60, 217–225. 157 Sleat, D. E., Hull, R., Turner, P. C., Wilson, T. M. A. (1988) Studies on the mechanism of translational enhancement by the 5′ -leader sequence of tobacco mosaic virus RNA. Eur. J. Biochem. 175, 75–86. 158 Spirin, A. S. (ed.) (2002) Cell-Free Translation Systems, Springer-Verlag, Heidelberg, Berlin, New York. 159 Spirin, A. S. (2000) Ribosomes, Kluwer Academic/Plenum Publishers, New York. 160 Spirin, A. S., Baranov, V. I., Ryabova, L. A., Ovodov, S. Y., Alakhov, Y. B. (1988) A continuous cell-free translation system capable of producing polypeptides in high yield. Science 242, 1162–1164. 161 Spirin, A. S. (1991) Cell-free protein synthesis bioreactor. In: Frontiers in Bioprocessing II, eds. Todd, P., Sikdar, S. K., Beer, M., American Chemical Society, Washington, D.C., pp. 31–43.

References 162 Spirin, A. S. (2004) High-throughput cell-free systems for synthesis of functionally active proteins. Trends Biotechnol. 22, 538–545. 163 Steigerwald, R., Nemetz, C., Walckhoff, B., Emrich, T. (2002) Cell-free expression of a 127 kDa protein: The catalytic subunit of human telomerase. In: Cell-Free Translation Systems, ed. Spirin, A. S., Springer-Verlag, Heidelberg, Berlin, New York, pp. 165–173. 164 Stueber, D., Ibrahimi, I., Cutler, D., Dobberstein, B., Bujard, H. (1984) A novel in vitro transcription-translation system: Accurate and efficient synthesis of single proteins from cloned DNA sequences. EMBO J. 3, 3143–3148. 165 Svetlov, M. S., Kommer, A., Kolb, V. A., Spirin, A. S. (2006) Effective cotranslational folding of firefly luciferase without chaperones of the Hsp70 family. Protein Sci. 15, 242–247. 166 Swartz, J. R. (ed.) (2003) Cell-Free Protein Expression, Springer-Verlag, Berlin, Heidelberg, New York. 167 Timmer, R. T., Benkowski, L. A., Schodin, D., Lax, S. R., Metz, A. M., Ravel, J. M., Browning, K. S. (1993) The 5′ and 3′ untranslated regions of satellite tobacco necrosis virus RNA affect translational efficiency and dependence on a 5′ cap structure. J. Biol. Chem. 268, 9504–9510. 168 Tissiéres, A., Schlessinger, D., Gros, F. (1960) Amino acid incorporation into proteins by E. coli ribosomes. Proc. Natl. Acad. Sci. U.S.A. 46, 1450–1463. 169 Ugarov, V. I., Morozov, I. Yu., Jung, G. V., Chetverin, A. B., Spirin, A. S. (1994) Expression and stability of recombinant RQ-mRNA in cell-free translation systems. FEBS Lett. 341, 131–134. 170 Underwood, K., Swartz, J. R., Puglisi, J. D. (2005) Quantitative polysome analysis identifies limitations in bacterial cellfree protein synthesis. Biotechnol. Bioeng. 91, 425–435. 171 Uzawa, T., Yamagishi, A., Ueda, T., Chikazumi, N., Watanabe, K., Oshima, T. (1993) Effects of polyamines on a continuous cell-free protein synthesis system of an extreme thermophile, Thermus thermophilus. J. Biochem. (Jpn.) 114, 732–734.

172 Volyanik, E. V., Dalley, A., McKay, I. A., Keigh, I., Williams, N. S., Bustin, S. A. (1993) Synthesis of preparative amounts of biologically active interleukin-6 using a continuous-flow cell-free translation system. Anal. Biochem. 214, 289–294. 173 Wang, S., Browning, K. S., Miller, W. A. (1997) A viral sequence in the 3′ -untranslated region mimics a 5′ cap in facilitating translation of uncapped mRNA. EMBO J. 16, 4107–4116. 174 Winnick, T. (1950) Incorporation of labeled amino acids into protein of embryonic and tumor tissue homogenates. Fed. Proc. 9, 247. 175 Winnick, T. (1950) Studies on the mechanism of protein synthesis in embryonic and tumor tissues. II. Inactivation of fetal rat liver homogenates by dialyses and reactivation by the adenylic acids system. Arch. Biochem. 28, 338–347. 176 Wright, C. F., Moss, B. (1987) In vitro synthesis of vaccinia virus late mRNA containing a 5′ poly(A) leader sequence. Proc. Natl. Acad. Sci. U.S.A. 84, 8883– 8887. 177 Yamamoto, Y. I., Nagahori, H., Yao, S., Zhang, S. T., Suzuki, E. (1996) Hollow fiber reactor for continuous flow cell-free protein production. J. Chem. Eng. Jpn. 6, 1047–1050. 178 Yamane, T., Kawarasaki, Y., Nakano, H. (1995) In vitro protein biosynthesis using ribosome and foreign mRNA: An approach to construct a protein biosynthesizer. Ann. New York Acad. Sci. 750, 146–157. 179 Yamane, T., Ikeda, Y., Nagasaka, T., Nakano, H. (2005) Enhanced cell-free protein synthesis using a S30 extract from Escherichia coli grown rapidly at 42 ◦ C in an amino acid enriched medium. Biotechnol. Prog. 21, 608–613. 180 Yang, J., Kanter, G., Voloshin, A., Levy, R., Swartz, J. R. (2004) Expression of active murine granulocyte-macrophage colony-stimulating factor in an Escherichia coli cell-free system. Biotechnol. Prog. 20, 1689–1696. 181 Yanofsky, C. (1981) Attenuation in the control of expression of bacterial operons. Nature 289, 751–758. 182 Yin, G., Swartz, J. R. (2004) Enhancing multiple disulfide bonded protein folding

33

34

1 Cell-free Protein Synthesis Systems in a cell-free system. Biotechnol. Bioeng. 86, 188–195. 183 Yu, M., Souaya, J., Julin, D. (1998) The 30-kDa C-terminal domain of the RecB protein is critical for the nuclease activity, but not the helicase activity of the RecBCD enzyme from Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 95, 981–986. 184 Zarucki-SchuIz, T., Jeres, C., Goldberg, G., Kung, H.-F., Huang, K. H., Brot, N., Weissbach, H. (1979) DNAdirected synthesis of proteins involved in bacterial transcription and translation. Proc. Natl. Acad. Sci. U.S.A. 76, 6115– 6119. 185 Zawada, J., Swartz, J. (2004) Maintaining rapid growth in moderate density Escherichia coli fermentations. Biotechnol. Bioeng. 4, 407–416.

186 Zawada, J., Swartz, J. (2005) Effects of growth rate on cell extract performance in cell-free protein synthesis. Biotechnol. Bioeng. 4, 618–625. 187 Zawada, J., Richter, B., Lodes, H., Shah, A., Swartz, J. (2003) High density, defined media culture for the production of Escherichia coli cell extracts. Fermentation Biotechnol. 862, 142–156. 188 Zeyenko, V. V., Ryabova, L. A., Gallie, D. R., Spirin, A. S. (1994) Enhancing effect of the 3′ -untranslated region of tobacco mosaic virus RNA on protein synthesis in vitro. FEBS Lett. 354, 271– 273. 189 Zubay, G. (1973) In vitro synthesis of protein in microbial systems. Annu. Rev. Genet. 7, 267–287.

35

2

The Constructive Approach for Cell-free Translation Takuya Ueda

2.1 Introduction

The cellular system for translating mRNA into the proper sequence of amino acids is a complex biological process. From an in vitro biochemical standpoint, the formation of a peptide bond between two amino acids is a simple dehydration reaction. However, in cells, the process of protein synthesis requires several additional protein factors and RNA molecules to ensure precise deciphering of the open reading frame (ORF) of the mRNA. The ribosome, which is the name given to this complex of RNAs and protein, must also sustain the efficiency of the dehydration reaction. The small RNA subunit of the ribosome provides the scaffold that ensures accurate codon–anticodon pairing, while the large ribosomal RNA subunit contains the active site that catalyzes amino acid condensation. The ribosome possesses three translational RNA (tRNA) binding sites: the A-site binds to aminoacyl-tRNA; the P-site binds to peptidyl-tRNA; and the E-site is the exit site for deacylated tRNA [1]. Aminoacyl-tRNA synthetases, which catalyze the attachment of amino acids to their cognate tRNA, are also important components of translation, although they are not considered to be part of the ribosome [2]. Translation involves a group of components called translation factors, which fall into one of three categories: initiation factor (IF), elongation factor (EF), or release factor (RF). Orchestration of such a large and diverse set of functional components is crucial for protein synthesis to progress smoothly. Here, we present an overview of the protein translation process in bacteria, and describe a cell-free system that we have developed for reconstituting translation in vitro. We also discuss the potential of our cell-free translation system in biotechnology applications.

Cell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

36

2 The Constructive Approach for Cell-free Translation

2.2 The Process of Protein Synthesis 2.2.1 Polypeptide Synthesis

In bacteria, the protein translation is divided into three steps (Fig. 2.1). Translation is initiated when an mRNA enters the ribosome through a specific interaction between the Shine–Dalgarno (SD) sequence on the mRNA [3] and the SDcomplementary sequence in the 16S ribosomal RNA (rRNA) of the 30S ribosomal. This is followed by recruitment and binding of the initiator fMet-tRNA to the AUG start codon, mediated by IF2, at the P-site of the ribosome. This complex is called the initiation complex [4, 5]. Prior to mRNA entry and fMet-tRNA association, the ribosome consists of dissociated 30S and 50S subunits, and the 30S subunit is complexed with initiation factors IF1 and IF3, and the ribosome recycling factor (RRF) [4–7]. Following dissociation of IF1, IF2 and IF3, the 50S subunit associates with the initiation complex. The transition from initiation to elongation occurs upon binding of an aminoacyl-tRNA to the A-site of the ribosome. Several protein-encoding genes in the bacteria phage genome and in Grampositive bacteria have been shown to lack an SD sequence in their 5′ untranslated regions, yet are able to initiate translation at the appropriate AUG start codon [8]. These “leaderless” mRNAs are translated by an undissociated 70S ribosome, which differs from the canonical initiation process summarized above [9]. Although subunit dissociation does not occur following translation of leaderless mRNAs, the three initiation factors, IF1, IF2, and IF3, are strictly required, indicating that there are perhaps novel functions of IF1 and IF3 yet to be characterized [9]. The elongation factor EF-Tu is a GTP-binding protein, and, in its GTP-bound form, is involved in conveying aminoacylated elongator tRNAs to the A-site of the ribosome. Upon successful and stable matching of the correct codon–anticodon pair, EF-Tu-GTP is hydrolyzed to EF-Tu-GDP, and detaches from the ribosome [10]. EF-Tu-GTP is converted into GDP by the GTP exchange factor, EF-Ts. Following correct codon–anticodon matching, the amino acid carried by the tRNA in the Asite is covalently linked to the peptidyl-moiety of the amino acid carried by the tRNA in the P-site, and the tRNA in the A-site shifts to the P-site. This tRNA movement is promoted by elongation factor EF-G and GTP hydrolysis [11]. The elongation cycle is repeated until a termination event takes place on the ribosome. When the termination codon of the ORF lines up with the A-site in the ribosome, release factor RF1 or RF2 binds to the codon, and catalyzes hydrolysis of the peptidyl-moiety in the P-site to complete peptide synthesis [12]. RF1 or RF2 dissociation from the ribosome is mediated by a third release factor, RF3, and GTP hydrolysis [13]. Disassembly of the termination complex, which now consists of the mRNA and the terminal tRNA associated with the hydrolyzed polypeptide chain, is mediated by RRF. This is followed by subunit dissociation, and binding of IF1 and IF3 to the 30S subunit [6, 7]. The translation components are then in place to efficiently repeat the process of protein synthesis.

2.2 The Process of Protein Synthesis

Fig. 2.1 Schematic representation of protein translation in bacteria.

The process consists of three steps: initiation, elongation and termination.

37

38

2 The Constructive Approach for Cell-free Translation

2.2.2 Protein Maturation

The product of translation, the polypeptide chain, must undergo a maturation process, or post-translational modification, before it is considered a mature protein. The first event associated with maturation, subsequent to the polypeptide chain exiting the peptide tunnel of the 50S ribosomal subunit, is polypeptide folding [14, 15]. According to Anfinsen’s dogma, proteins have the potential to undergo selffolding, but in cells most protein folding is facilitated by specific proteins called chaperones. Three protein folding systems exist in bacteria: the DnaK/DnaJ/GrpE system, the GroEL/ES system, and the trigger factor system. The DnaK/DnaJ/GrpE and trigger factor systems are co-translational processes, while the GroEL/ES system is a post-translational folding process (Fig. 2.2) [14, 15]. A DnaK/trigger factor double deletion mutant appears to survive and can grow at low temperature [16, 17], and we have demonstrated recently that GroEL can participate in protein folding in a co-translational manner [18]. These results suggest that all three chaperone systems in bacteria are capable of mediating folding of nascent polypeptides into the correct secondary structure. However, we have very little knowledge of chaperone-dependent folding processes, owing to the difficulty in developing appropriate experimental methods for evaluating translation-coupled protein folding. In particular, the nature of the specific substrates for each of the individual chaperone systems remains unclear. To understand the process by which functional proteins with the correct conformation are synthesized, the relationship between substrate and chaperone systems in cells should be clarified. Subsequent to protein folding, proteins may undergo various other posttranslational modifications, by such enzymes as protein kinases and glycosylases.

Fig. 2.2 Schematic of the chaperone network in E. coli. A nascent

polypeptide is recognized by trigger factor as it exits the peptide tunnel of the ribosome. DnaK and DnaJ bind the growing polypeptide on the ribosome in a co-translational maturation process. In contrast, GroEL and GroES assist protein folding in a post-translational manner.

2.2 The Process of Protein Synthesis

a

b Fig. 2.3 Schematic of the membrane targeting

and secretory processes in E. coli. (a) During translation, trigger factor binds to the nascent polypeptide chain. SRP then binds to the polypeptide, and trigger factor simultaneously detaches. The complex of SRP, nascent polypeptide and ribosome binds to membrane-associated SR, which is also associated with SecYEG. The polypeptide is then translocated into membrane with the help of YidC. (b) SecB binds to a nascent

polypeptide and acts as chaperone to prevent aggregation. Trigger factor binds to and protects the polypeptide from SRP-binding. The SecB–polypeptide complex associates with SecA, and is directed to the SecYEG complex in the plasma membrane. SecA inserts the polypeptide into the membrane, the polypeptide is transported across the membrane, and the signal sequence is digested by a membrane-associated signal peptidase.

These types of modifications occur frequently in eukaryotic protein synthesis, and regulation of post-translational modification is crucial for the synthesis of mature, functional eukaryotic proteins. Insertion of integral or trans-membrane proteins, or secreted proteins, into the cell membrane is another important aspect of polypeptide maturation. About 30%

39

40

2 The Constructive Approach for Cell-free Translation

of the proteins encoded by the bacterial genome reside in the cell membrane, or are secreted [19]. For membrane proteins (Fig. 2.3a), as the polypeptide is being translated on the ribosome it is protected from the binding of the chaperone protein SecB by trigger factor. The nascent polypeptide chain is then recognized by the Signal Recognition Particle (SRP, Ffh and 4.5S RNA complex in E. coli), which binds to a membrane-embedded SRP receptor, (SR, FtsY in E. coli) [20]. The nascent protein is translocated into the lipid-bilayer by a membrane-associated SecYEG complex [20]. With a secretory protein (Fig. 2.3b), the newly synthesized polypeptide is recognized by the chaperone protein SecB, which binds to and prevents aggregation of the growing polypeptide chain. The SecB–polypeptide complex associates with SecA, which directs the polypeptide into the translocation pore of the SecYEG complex. The signal sequence is cleaved by a signal peptidase on the membrane, and the protein is transported across the membrane and secreted [20]. Most secretory proteins form intramolecular disulfide bonds under the oxidative conditions of the periplasm. Once outside the cell, a protein disulfide isomerase mediates proper S–S bridging of secreted proteins. Through these pathways, polypeptides mature into properly folded, functional proteins with the correct localization. Thus, to faithfully and accurately synthesize proteins with native structure and activity in vitro, one must recreate the complex processes of synthesis and maturation that occur in cells.

2.3 A Constructive Approach to Protein Synthesis 2.3.1 In Vitro Reconstitution of Polypeptide Synthesis

To reconstitute the cellular process of protein synthesis in vitro, we first developed a conventional in vitro translation system that recapitulated polypeptide synthesis in the cell. The amino acid condensation reaction that occurs on the ribosome is, in principle, the same among eukaryotes, prokaryotes, and archaebacteria. Because it is one of the best understood systems, we reconstituted the protein translational system of E. coli. As mentioned above, protein synthesis in E. coli requires three initiation factors (IF1–3), three elongation factors (EF-Tu, EF-Ts, and EF-G), and four termination factors (RF1, 2, 3 and RRF). In addition, 20 aminoacyl-tRNA synthetases are required for the attachment of each of the 20 amino acids to their corresponding tRNA. The genes for these factors were amplified by PCR from E. coli genomic DNA, and cloned into an appropriate expression vectors. Through the use of specific primers, a sequence of histidine residues (His-tag) was inserted at either the N- or C-terminus of the ORF. Genes were overexpressed in E. coli, and the His-tag fusion proteins were purified by Ni-column chromatography. The purity of each protein was evaluated by polyacrylamide gel-electrophoresis, and the activity was verified using an appropriate biochemical assay. By these analyses, we determined

2.3 A Constructive Approach to Protein Synthesis

that the translation factors were purified to >95% homogeneity, and their activity was similar to that of native proteins. To obtain ribosomes, E. coli were disrupted, and ribosome fractions were prepared by sucrose-density ultracentrifugation, followed by several treatments with high-concentrations of ammonium chloride to remove non-specific binding factors. Purified translation factors and ribosome fractions obtained in this manner were combined, and polypeptide synthesis was examined by supplying a nucleic acid template, amino acids and a chemical energy source. After optimization of the concentrations of the individual components, we were able to synthesize submilligram amounts of protein in a 1 mL reaction volume, using our reconstituted system for protein translation [21]. These results indicated that our current level of understanding of the factors required for protein synthesis in cells, as described in literature and in textbooks, is sufficient to reconstitute the process of polypeptide synthesis in vitro. The system that we developed, which was named the PURE system, is completely different from canonical cell-free translation systems such as wheat germ extract, rabbit reticulocyte lysate, and E. coli S30 extract, in that the PURE system contains no components that are extraneous to the process of polypeptide synthesis. Of note, the PURE system contains neither nucleases nor proteases, both of which hinder translation efficiency by degrading DNA or mRNA templates and proteins. Because of these advantages, linear DNA consisting of only an ORF, promoter, and SD sequence can be processed using a transcription–translation coupled PURE system. Moreover, owing to the absence of ribonucleases, trans-translation reactions, in which a degradation tag is attached to incomplete polypeptides translated from cleaved mRNAs, are avoided. Notably, nucleotide triphosphate hydrolysis activity was undetectable in the PURE system, indicating that the chemical energy added to the system was utilized for peptide bond formation, and not for unrelated reactions. Despite an efficient level of conversion of chemical energy, NTP was consumed, and NDP and NMP were generated during the transcription and translation reactions, as a result of tRNA recycling by a combination of creatine phosphate, and three additional enzymes, nucleotide diphosphate kinase, adenylate kinase and creatine kinase [21, 22]. 2.3.2 Protocol of Protein Synthesis using PURE System

The template DNA corresponding to the objective protein is prepared before protein synthesis reaction of the PURE system. Both plasmid DNA and PCR fragments can be used in PURE system. In the template DNA it is necessary to encode the T7 promoter sequence, SD sequence at the upstream of the open reading frame of the gene. It is also required that the ORF initiates with an initiation codon (ATG) and terminates with a termination codon (TAG, TGA or TAA). When plasmid DNA is used, T7 terminator sequence is also necessary downstream of a termination codon. Otherwise plasmid DNA is digested by appropriate restriction nuclease downstream of the termination codon for run-off transcription by T7 RNA polymerase.

41

42

2 The Constructive Approach for Cell-free Translation

Table 2.1 shows the composition of the PURE system [23]. All the components are mixed and then the template DNA is added. After reaction mixture is incubated in the batch system at 37 ◦ C for 1–2 h without shaking, the products can be analyzed by SDS-PAGE. For the easy detection of products, [35 S]methionine or BODIPY-Lys-tRNALys (FluoroTectTM GreenLys in vitro Translation Labeling System; Promega) can be supplied to the translation reaction. The PURE system can synthesize not only prokaryotic but also eukaryotic proteins. This protein-synthesizing system enables a novel method for the purification of the synthesized protein. In the PURE system, only those essential elements for the protein synthesis are present, and all protein factors except ribosomal proteins are in His-tagged form. After protein synthesis, Ni+ -chelating resin (e.g., Ni-NTA agarose from Qiagen) added to the reaction and the mix is incubated for 4 ◦ C with shaking. Subsequently, the resin, the resin bound His-tagged proteins, and the ribosomes are simultaneously removed by ultrafiltration through a membrane with a 100-kDa MW cut-off (Microcon YM100 from Millipore). Through this procedure, the product protein, which has no tag-sequence, can be purified within 2 hours [21]. 2.3.3 Addition of Protein Folding Machinery to the PURE System

Because of the lack of chaperone proteins, the PURE system is inadequate for obtaining proteins in their native conformation. Despite this, however, we have synthesized several proteins from E. coli using the PURE system, and found that there is a higher level of protein solubility using the PURE system than for conventional E. coli S30 extracts (unpublished results). This could reflect a lower protein concentration in the PURE system, which would reduce protein aggregation, but it indicates that the absence of protein folding components in the PURE system is not necessarily disadvantageous for the production of proteins with the proper secondary and tertiary conformation. Because of the nature of PURE system, which uses well-defined, purified protein components, we can use it to begin to address the molecular specificity of chaperone-mediated protein folding. For example, analysis of mammalian chaperones involved in protein folding can be performed using PURE system, since the activity of intrinsic bacterial chaperones can be disregarded (Fig. 2.4). Thus, a made-to-order protein folding system can be functionally integrated with a system for protein translation. To adapt the PURE system for protein folding studies, we needed to know which chaperone system was involved in mediating folding of each protein, the so-called chaperone network. Unfortunately, this kind of information is not well understood, because most chaperone research has been carried out using denatured polypeptides. Thus, it was first necessary to evaluate which chaperone system was required for the proper folding of each of the proteins encoded in the bacterial cell genome. To address this issue, we examined the effect of adding chaperone proteins on the solubility of proteins translated using the PURE in vitro translation system. The entire repertoire of proteins encoded by the E. coli genome is represented in the cDNA 4000 clones of the ASKA library. These cDNAs were individually sub-

2.3 A Constructive Approach to Protein Synthesis

Table 2.1 Composition of the PURE system. Specific activities of ARS

and MTF were measured using radioactive amino acids. One unit of activity was defined as the amount of enzyme that catalyzes the formation of 1 pmol of aminoacyl-tRNA in 1 min. Part of system Buffer

Components

Hepes-KOH, pH 7.6 Potassium glutamate Magnesium acetate Spermidine DTT Energy sources ATP, GTP CTP, UTP Creatine phosphate Others 20 Amino acids 10-Formyl-5,6,7,8-tetrahydrophilic acid tRNAmix (Roche) Ribosome Translation factors IF1 IF2 IF3 EF-G EF-Tu EF-Ts RF1 RF3 RRF ARS and MTF AlaRS ArgRS AsnRS AspRS CysRS GlnRS GluRS GlyRS HisRS IleRS LeuRS LysRS MetRS PheRS ProRS SerRS ThrRS TrpRS TyrRS ValRS MTF

Concentration

Activity (U mL–1 )

50 mm 100 mm 13 mm 2 mm 1 mm 2 mm 1 mm 20 mm 0.3 mm 10 mg mL–1 56 A260 mL–1 1.2 µm 2.7 µm 0.4 µm 1.5 µm 0.26 µm 0.92 µm 0.66 µm 0.25 µm 0.17 µm 0.50 µm 1900 2500 20 mg mL–1 2500 630 1300 1900 5000 630 2500 3800 3800 6300 1300 1300 1900 1300 630 630 3100 4500

43

44

2 The Constructive Approach for Cell-free Translation Table 2.1 (continued)

Part of system

Components

Concentration

Other enzymes

Creatine kinase Myokinase Nucleoside-diphosphate kinase Pyrophosphatase T7 RNA polymerase

4.0 µg mL–1 3.0 µg mL–1 1.1 µg mL–1 2.0 units mL–1 10 µg mL–1

Activity (U mL–1 )

cloned onto an appropriate expression vector [24], then amplified by PCR using a primer that contained the T7 promoter. The amplified PCR products were subjected to transcription–translation using PURE system supplemented with T7 RNA polymerase (Fig. 2.5), in the presence of the components of one of the three bacterial chaperone systems: GroEL/ES, DnaK/DnaJ/GrpE, or trigger factors. Table 2.2 shows the concentrations of chaperones supplied to the PURE system. The translation reaction mixture was centrifuged, and the ratio of radiolabeled product in the pellet and the supernatant was assessed. Using this approach, we have been able to evaluate the effect of different protein folding components on the solubility of in vitro translated polypeptides, and we can begin to characterize the substrate requirements of each of the different protein folding systems in bacteria. As an example, we present the results of the translation of adenosylmethionine synthetase, which is encoded by the MetK gene. In the presence of GroEL and GroES, the solubility of adenosylmethionine synthetase in the PURE system translation reaction system was greatly enhanced. Addition of DnaK, DnaJ and GrpE also increased the amount of soluble protein, indicating that two different chaperone systems can recognize this protein. However, measurement of the enzymatic activity of the protein product clearly showed that only GroEL/ES mediated proper folding of the protein into a functionally competent form. This suggested that

Fig. 2.4 Schematic representation of the “constructive approach” to in vitro protein folding and membrane targeting using the PURE system in vitro translation system.

2.3 A Constructive Approach to Protein Synthesis

Fig. 2.5 Schematic of the approach for genome-wide analysis of

substrate requirements of the chaperone network using the ASKA cDNA library and the PURE system in vitro translation system. Chaperone requirements for protein folding of every protein encoded by the E. coli genome are evaluated by the addition of purified chaperone proteins.

DnaK/DnaJ/GrpE may prevent aggregation of adenosylmethionine synthetase, but fail to mediate proper folding. Thus, there appears to be a stringent requirement for GroEL/ES, but not for DnaK/DnaJ/GrpE, in adenosylmethionine synthetase maturation. In other words, in the process of MetK translation, DnaK functions as a “holder chaperone” that protects the protein from aggregating with other proteins, while GroEL functions as the “folder chaperone” that mediates proper folding [18, 25]. These results illustrate the feasibility of performing a comprehensive analysis of the specificity of the bacterial chaperone network using the chaperone depleted PURE system cell-free translation system. This type of analysis would not be possible using a conventional cell-extract system, as the presence of intrinsic chaperone proteins complicates the interpretation of the data. Interestingly, polysome analyses using PURE system revealed that GroEL was capable of binding to a growing polypeptide attached to the ribosome, indicating that GroEL functions in a cotranslational manner. This finding is inconsistent with the established model of the chaperone network, in which GroEL/ES assists in folding in a post-translational Table 2.2 Concentrations of chaperones supplied to the PURE system.

System

Components

Concentration (µm)

DnaK/DnaJ/GrpE

DnaK DnaJ GrpE GroEL ES Trigger factor

1 0.4 0.4 1 2 2.5

GroEL/ES Trigger factor system

45

46

2 The Constructive Approach for Cell-free Translation

manner. However, this data is compelling since our experimental system did not employ denatured substrates, and thus more closely mimics the process of protein folding in vivo. To generate fully functional proteins in vitro, it is important to understand the chaperone requirements for correct folding. In this regard, the PURE system provides an excellent platform for charting the appropriate chaperone system for all of the proteins encoded in the bacterial genome. 2.3.4 Integration of a Membrane Targeting System with the PURE system

About 30% of the proteins in a cell reside in the cell membrane or are secreted. Proper folding of membrane proteins does not occur in an aqueous environment but in the environment of the lipid bilayer. Secreted proteins also must pass through an hydrophobic inner leaflet of the lipid bilayer, in transit from the reductive environment of the cytoplasm to the oxidative environment of the extracellular space. Translation occurs in the aqueous, reductive environment of the cytoplasm, and newly synthesized membrane-targeted polypeptides undergo this phase transition with help of specific targeting machinery. To obtain membrane-localized and secreted proteins in vitro, we believe that the best approach is to reconstitute the translocation process that occurs naturally in cells. Our approach is to integrate components of the membrane targeting machinery with the PURE system in vitro translation system, in a manner similar to the integration of a protein folding system described above. The most important component of the membrane targeting machinery is the lipid bilayer. We prepared inverted Inner Membrane Vesicles (IMV) [26] from E. coli spheroplasts, and subjected the membranes to treatment with 6 m urea [27, 28]. Several cellular components of the membrane targeting system were expressed in and purified from E. coli: SecB and SecA, which are involved in the secretory process, and Ffh and FtsY, which are involved in the localization of integral membrane proteins. Using these purified factors and prepared IMV, an in vitro membrane protein targeting system was reconstructed (Fig. 2.4). OmpA and MtlA were used as model proteins for protein secretion and membrane integration, respectively. The feasibility of using the PURE system for reconstituting membrane localization and translocation was evaluated using resistance to proteinase K as a marker for proper localization. Because the membranes were inverted, translocation was reversed, and proteins were secreted from the “outside-in”, into the lumen of the IMV. Using this system, we were able to demonstrate successful synthesis and translocation of OmpA and MtlA into the membrane. In the cell, preOmpA binds to SecB, which acts like a chaperone and prevents protein aggregation. The nascent polypeptide is then transported and binds to SecA, which directs the polypeptide to the SecYEG complex in the membrane. The signal sequence of preOmpA is then cleaved by a membrane-associated peptidase. Our results indicate that this process is recapitulated in our in vitro PURE system

2.3 A Constructive Approach to Protein Synthesis

Fig. 2.6 In vitro synthesis and translocation of

secretory and membrane proteins using the PURE system in vitro translation system and IMV (inverted membrane vesicle). (a) For the secretory protein OmpA, purified SecA and SecB were added to the system; for the membrane protein MtlA, SRP and SR were added. Membrane targeting was evaluated using proteinase K digestion followed by gel-electrophoresis. (b) OmpA protein secretion into the lumen of the IMV. preOmpA was translated in the absence of IMV (lanes 1,

2), in the presence of untreated IMV (lanes 3, 4) and in the presence of 6 m urea-treated IMV (lanes 5, 6). After the reaction, samples were treated with proteinase K (lanes 2, 4, 6), then analyzed by gel-electrophoresis. (c) Membrane insertion of MtlA. MtlA was translated in the absence (lanes 1, 2) and in the presence (lanes 3, 4) of IMV. Samples were then treated with proteinase K (lanes 2, 4), and analyzed by gel-electrophoresis. MtlA-MPF is the membrane protected fragment of MtlA.

supplemented SecB, SecA and urea-treated IMV (Fig. 2.6b). Omission of any one of the components of the membrane targeting machinery abolished membrane integration, indicating that each of them functioned similarly to their function in the cell. Membrane integration of MtlA, which has six-transmembrane domains, was strictly dependent on the presence of SRP and SR in the translation reaction, and the proteinase K digestion pattern of MtlA indicated that it was integrated into membrane in the correct conformation and orientation (Fig. 2.6c). In conclusion, using model proteins for secretion and membrane insertion, we demonstrated that addition of the protein components of the membrane targeting machinery of the cell to the PURE system cell-free translation system generates proteins that are properly integrated into the membrane. Several groups have

47

48

2 The Constructive Approach for Cell-free Translation

shown similar results using detergents, but the effect of these detergents on the structure of the proteins is not known. To our knowledge, our system is the most accurate reflection of the cellular process of membrane targeting, and is most likely applicable to proteins other than OmpA and MtlA. For secretory proteins, the disulfide bonds may be formed by adding a peptide disulfide isomerase, or generating an oxidative environment inside the inverted membrane. Perhaps most significantly, membrane-embedded proteins could immediately be used for functional analyses or structural studies. To this end, we are in the process of synthesizing E. coli membrane proteins using the ASKA cDNA library and eukaryotic membranes and membrane proteins, in an effort to further validate our system. 2.3.5 Protein Synthesis using the PURE System containing Molecular Chaperones

SecA, SecB, Ffh and FtsY were purified as described previously [28]. IMVs were prepared from MC4100, as described previously [26]. Urea-treated IMVs were additionally prepared according to an earlier report [27]. The PURE system mixture [21, 29] was slightly altered in this analysis. Buffer conditions were changed to 157 mm potassium glutamate (pH 7.5), 18.7 mm potassium acetate, 9 mm magnesium acetate, 2 mm ATP and GTP, 1 mm UTP and CTP, 1 mm DTT, 20 mm creatine phosphate, 10 µg mL–1 5,10-methenyltetrahydrofolic acid, 100 mm 20 amino acids and 0.01 OD µL–1 tRNA mixture. Additionally, [35 S]methionine was added to detect synthesized proteins. Human placental ribonuclease inhibitor (0.4 unit µL–1 ) was also utilized. Translocation of pOmpA and integration of MtlA reaction coupled with their transcription/translation were performed in the PURE system in the presence of IMVs or Urea-treated IMVs. SecA, SecB, Ffh, FtsY, trigger factor and IMVs were present in the PURE system reaction mixture at the concentrations shown in Table 2.3. Protein synthesis was initiated by the addition of plasmid DNA. IMVs or urea-treated IMVs were added 5 min after starting the reaction. At the end, half of each reaction was treated with 0.5 mg mL–1 of proteinase K at 25 ◦ C for 30 min to assess translocation or integration followed by TCA (5%) precipitation. The other half was precipitated directly with 5% TCA. Proteins labeled with [35 S]methionine were analyzed by 12% or 15% SDS-PAGE. Table 2.3 Concentration of membrane targeting components.

Component

Concentration (µg mL–1 )

SecA SecB Ffh FtsY Trigger factor IMVs or urea-treated IMVs

96 300 5 144 96 2 mg mL–1

2.4 Conclusion

2.4 Conclusion

We have been able to integrate components of the cellular protein maturation machinery with a reconstituted protein translation system in a “constructive approach” to in vitro protein synthesis. In cells, protein maturation into a fully functional form occurs via a responding processes, such as folding with chaperones, processing, post-translation modification, membrane targeting etc. If a set of post-translational processes are available in vitro, we can achieve faithful protein synthesis, as it occurs in vivo, which is the ultimate goal of cell-free translation system.

References 1 Rheinberger, H. J., Sternbach, H., Nierhaus, K. H., Codon-anticodon interaction at the ribosomal E site. J. Biol. Chem. 1986, 261, 9140–9143. 2 Soll, D., The accuracy of aminoacylation– ensuring the fidelity of the genetic code. Experientia 1990, 46, 1089–1096. 3 Shine, J., Dalgarno, L., The 3’-terminal sequence of Escherichia coli 16S ribosomal RNA: complementarity to nonsense triplets and ribosome binding sites. Proc. Natl. Acad. Sci. U.S.A. 1974, 71, 1342– 1346. 4 Gualerzi, C. O., Brandi, L., Caserta, E., La Teana, A., Spurio, R., Tomsic, J., Pon, C. L., Translation initiation in bacteria. In: The Ribosome: Structure, Function, Antibiotics and Cellular Interactions, Grrett, R. A., Douthwaite, S. R., Liljas, A., Matheson, A. T., Moore, P. B. and Noller, H. F. (eds.). ASM Press, Washington, DC, 2000. 5 Gualerzi, C. O., Pon, C. L., Initiation of mRNA translation in prokaryotes. Biochemistry 1990, 29, 5881–5889. 6 Hirokawa, G., Demeshkina, N., Iwakura, N., Kaji, H., Kaji, A., The ribosomerecycling step: consensus or controversy? Trends Biochem. Sci. 2006, 31, 143–149. 7 Peske, F., Rodnina, M. V., Wintermeyer, W., Sequence of steps in ribosome recycling as defined by kinetic analysis. Mol. Cell 2005, 18, 403–412. 8 Van Etten, W. J., Janssen, G. R., An AUG initiation codon, not codon-anticodon complementarity, is required for the translation of unleadered mRNA in Escherichia coli. Mol. Microbiol. 1998, 27, 987–1001.

9 Udagawa, T., Shimizu, Y., Ueda, T., Evidence for the translation initiation of leaderless mRNAs by the intact 70 S ribosome without its dissociation into subunits in eubacteria. J. Biol. Chem. 2004, 279, 8539– 8546. 10 Sprinzl, M., Elongation factor Tu: a regulatory GTPase with an integrated effector. Trends Biochem. Sci. 1994, 19, 245–250. 11 Wintermeyer, W., Rodnina, M. V., Translational elongation factor G: a GTP-driven motor of the ribosome. Essays Biochem. 2000, 35, 117–129. 12 Kisselev, L. L., Buckingham, R. H., Translational termination comes of age. Trends Biochem. Sci. 2000, 25, 561–566. 13 Zavialov, A. V., Mora, L., Buckingham, R. H., Ehrenberg, M., Release of peptide promoted by the GGQ motif of class 1 release factors regulates the GTPase activity of RF3. Mol. Cell 2002, 10, 789–798. 14 Bukau, B., Deuerling, E., Pfund, C., Craig, E. A., Getting newly synthesized proteins into shape. Cell 2000, 101, 119– 122. 15 Hartl, F. U., Hayer-Hartl, M., Molecular chaperones in the cytosol: from nascent chain to folded protein. Science 2002, 295, 1852–1858. 16 Deuerling, E., Bukau, B., Chaperoneassisted folding of newly synthesized proteins in the cytosol. Crit. Rev. Biochem. Mol. Biol. 2004, 39, 261–277. 17 Genevaux, P., Keppel, F., Schwager, F., Langendijk-Genevaux, P. S., Hartl, F. U., Georgopoulos, C., In vivo analysis of the overlapping functions of DnaK

49

50

2 The Constructive Approach for Cell-free Translation

18

19

20

21

22

23

24

and trigger factor. EMBO Rep. 2004, 5, 195–200. Ying, B. W., Taguchi, H., Kondo, M., Ueda, T., Co-translational involvement of the chaperonin GroEL in the folding of newly translated polypeptides. J. Biol. Chem. 2005, 280, 12035–12040. Kihara, D., Kanehisa, M., Tandem clusters of membrane proteins in complete genome sequences. Genome Res. 2000, 10, 731–743. Muller, M., Koch, H. G., Beck, K., Schafer, U., Protein traffic in bacteria: multiple routes from the ribosome to and across the membrane. Prog. Nucl. Acid Res. Mol. Biol. 2001, 66, 107–157. Shimizu, Y., Inoue, A., Tomari, Y., Suzuki, T., Yokogawa, T., Nishikawa, K., Ueda, T., Cell-free translation reconstituted with purified components. Nat. Biotechnol. 2001, 19, 751–755. Swartz, J., A PURE approach to constructive biology. Nat. Biotechnol. 2001, 19, 732– 733. Shimizu, Y., Kanamori, T., Ueda, T., Protein synthesis by pure translation systems. Methods 2005, 36, 299–304. Saka, K., Tadenuma, M., Nakade, S., Tanaka, N., Sugawara, H., Nishikawa, K., Ichiyoshi, N., Kitagawa, M., Mori, H., Ogasawara, N., Nishimura, A., A complete set of Escherichia coli open read-

25

26

27

28

29

ing frames in mobile plasmids facilitating genetic studies. DNA Res. 2005, 12, 63–68. Ying, B. W., Taguchi, H., Ueda, T., Co-translational binding of GroEL to nascent polypeptides is followed by posttranslational encapsulation by GroES to mediate protein folding. J. Biol. Chem. 2006, 218, 21813–21819. Muller, M., Blobel, G., In vitro translocation of bacterial proteins across the plasma membrane of Escherichia coli. Proc. Natl. Acad. Sci. U.S.A. 1984, 81, 7421–7425. Helde, R., Wiesler, B., Wachter, E., Neubuser, A., Hoffschulte, H. K., Hengelage, T., Schimz, K. L., Stuart, R. A., Muller, M., Comparative characterization of SecA from the alpha-subclass purple bacterium Rhodobacter capsulatus and Escherichia coli reveals differences in membrane and precursor specificity. J. Bacteriol. 1997, 179, 4003–4012. Kuruma, Y., Nishiyama, K., Shimizu, Y., Muller, M., Ueda, T., Development of a minimal cell-free translation system for the synthesis of presecretory and integral membrane proteins. Biotechnol. Prog. 2005, 21, 1243–1251. Ying, B. W., Taguchi, H., Ueda, H., Ueda, T., Chaperone-assisted folding of a single-chain antibody in a reconstituted translation system. Biochem. Biophys. Res. Communn. 2004, 320, 1359–1364.

51

3

Functional Genomic Analysis using Sequential Cell-free Protein Synthesis Kim A. Woodrow and James R. Swartz

3.1 Introduction 3.1.1 The Post-genomic Era

The success of genome sequencing projects has produced a wealth of sequence data from several prokaryotic and eukaryotic genomes, which now present us with the challenge of determining the function of each and every gene product. Remarkable developments in high-throughput technologies have enabled gene function to be surveyed at various levels of molecular organization such as RNA, protein, and metabolite. Since proteins are often the entities that perform the diverse structural and chemical functions of the cell, many platforms have focused on identifying protein abundance (protein profiling), mapping protein interactions (interaction proteomics), and analyzing the biological activities of proteins (functional proteomics). Advances in protein-chip technology have led to the development of sophisticated platforms for studying protein interactions and biochemical activities [1, 2]. For example, protein function microarrays have been used to probe the substrate specificity of ∼98% of all the predicted yeast protein kinases [3] and to examine protein– protein and protein–lipid interactions for almost the entire yeast proteome [4]. In a novel approach that obviates the need to express, purify, and then immobilize proteins, self-assembling protein microarrays, also called nucleic acid programmable protein arrays (NAPPA), have been developed. NAPPA platforms use mammalian reticulocyte lysates for expression of epitope tagged proteins from immobilized full-length cDNAs [5], where the epitope tag mediates in situ protein immobilization. These protein microarrays have been optimized for detecting protein–protein interactions but can also be used in various other assays. Although the scope of proteomics is as broad as the tools that have been developed to probe different aspects of gene product function, the promise of the field is to unravel the mechanisms that allow a genotype to manifest into a specific phenotype. The current proteomic toolbox, however, has not yet successfully translated into a fully integrated understanding of how the function of every gene product, Cell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

52

3 Functional Genomic Analysis using Sequential Cell-free Protein Synthesis

and the complex interplay between them, leads to a specific cellular fate. As a consequence, it is still not possible to efficiently alter the genetic architecture of an organism to produce a desired phenotype. Metabolic engineering applications have scarcely benefited from the information derived from high-throughput proteomic techniques because these approaches have mainly offered only a snapshot of one stage of information transfer from gene to function [6]. Rational metabolic engineering is still slow and laborious, often being undermined by the complex regulatory mechanisms of the cell, which have evolved to maintain robust cellular performance in response to most genetic mutations. The information emerging in the post-genomic era suggests that strategies are needed to define protein function in the context of these dynamic metabolic networks that exist in vivo. 3.1.2 Cell-free Protein Synthesis (CFPS) as a Functional Proteomic Tool

In the post-genomic era, CFPS systems have proven useful for numerous applications that require simple and efficient methods for protein production. Protein microarrays based on CFPS have been used to identify and characterize protein– protein and protein–ligand interactions [7], to annotate genomes by expressing unidentified reading frames [8], and for constructing and screening protein libraries [9]. CFPS has also found many applications in structural genomics for the expression of toxic proteins, membrane proteins, and complex protein assemblies that are under-represented in genome structure databases because of challenges associated with their expression using conventional methods [10]. Incorporation of unnatural or isotope-labeled amino acids using CFPS has also assisted these structure determination initiatives [11, 12]. CFPS is also emerging as an efficient and economical platform for the production of personalized pharmaceutical therapeutics [13–15]. Most of these applications, however, have used CFPS as a production tool to express proteins rather than as a genomic tool to assess protein function in an integrated metabolic network. Traditional approaches have produced valuable information about individual proteins but they have been less successful at providing an integrating understanding of biological systems. Cell-free methods offer a more stable platform for genetic analysis since only a single gene product is evaluated and protein function can be identified within the context of a specified metabolic pathway. For example, the biochemical mechanisms of protein translation have been well characterized and involve over 100 different components [16, 17]. Probing the additional influence of other metabolic pathways and factors that affect ATP synthesis and stability, supply of nucleotide triphosphates, amino acid degradation or synthesis, mRNA stability, protein stability, and protein folding would arguably be equivalent to surveying a major fraction of the E. coli metabolome. We have developed a method that allows us to probe gene function in the context of a specified metabolic pathway. We used CFPS as a platform to identify proteins that influence in vitro transcription, translation, protein folding, and protein stability. Targets were selected for their anticipated influence on cell-free expression and

3.1 Introduction Table 3.1 Enzymatic reactions or factors expected to affect protein yields and activities.

Metabolic reactions or factors Transcription/mRNA maintenance Degradation of DNA Degradation of nucleotides Kinases Degradation of T7 RNA polymerase Translation Initiation factors Met-tRNAfMet formulation reactions Translation factors Ribosome rescue and recycling factors Aminoacyl-tRNA synthetases Amino acid supply Amino acid degradation/modification Amino acid synthesis Energy supply Kinases Energy pathway enzymes Energy pathway intermediate degradation Oxidoreductase depletion NADP transhydrogenase Protein stability Proteases Protein folding Chaperones Factors affecting translation rate

Anticipated effect

Negative Negative Positive or negative Negative Positive Positive Positive Positive Positive Negative Positive or negative Positive or negative Positive or negative Negative Negative Positive Negative Positive Positive or negative

folding (Table 3.1). For high-throughput analysis, PCR was used to generate expression templates for first round expression of effector candidate proteins (Fig. 3.1). These templates were used to direct expression of each effector protein target in reaction conditions that were conducive for protein function. The first round of protein expression was then terminated using a restriction nuclease that specifically targets unmethylated PCR templates. A methylated plasmid template for the indicator protein, β-lactamase, was then added to initiate a second round of protein synthesis. The sequential rounds of protein expression allowed us to identify effectors that influence the in vitro transcription, translation and protein folding of our reporter protein (Fig. 3.1). β-Lactamase total protein yield and specific activity varied based on the function of the target protein that was initially synthesized in the reaction chamber. These results demonstrate that increased concentrations of various gene products influence protein synthesis and/or activation of our indicator protein in a manner indicative of their in vivo function. This method was designed as a tool to identify factors that would improve in vitro transcription, translation and protein folding, and serve as a general tool to survey any metabolic process.

53

54

3 Functional Genomic Analysis using Sequential Cell-free Protein Synthesis

Fig. 3.1 Sequential CFPS expression system. PCR is used to generate a transcriptionallyactive DNA template library of genomic targets that may influence CFPS and folding. Batch CFPS reactions are conducted with a cell extract engineered to enhance linear DNA stability and using reaction conditions conducive for protein activation. The first round of expression creates an array of cell extracts that are individually enriched with a

single target protein. After the first round of expression, the linear DNA templates are specifically degraded and a second template for a reporter protein is added to initiate a subsequent round of protein expression. Protein targets that enhance or inhibit expression and folding of the reporter protein are determined by quantifying the yield and activity of the reporter protein.

3.2 Developing an enabling Technology for Sequential Expression Analysis

Despite the advantages of and promising applications for CFPS systems, the method has a reputation for being inefficient and unreliable and is further complicated by the use of low-density fermentations and expensive cell extract preparation [16]. CFPS systems suffer many of the same drawbacks seen with in vivo recombinant protein production as well as having unique challenges of their own. Limitations include short reaction lifetimes, instability of expression vectors and mRNA, inefficient translation initiation and elongation, depletion and degradation of key substrates, toxicity of small-molecule by-products or toxicity of the gene products themselves, as well as instability, heterogeneity, inappropriate folding and consequent inactivity of the protein targets [18]. Furthermore, protein yields obtained with CFPS systems fall below the 1–10 g L–1 typical of in vivo expression in E. coli high cell density cultures [15]. The productivity of in vitro systems for expressing folded and fully-active protein targets, however, rivals or is superior to in vivo sys-

3.2 Developing an enabling Technology for Sequential Expression Analysis

tems. This is because many protein targets expressed in vivo form inclusion bodies that must undergo extensive and costly in vitro refolding. Many of these disadvantages have been addressed by changes made to the configurations, energetics, and robustness of the CFPS reactions. The same analytical tools used to quantify metabolite pools and describe metabolic fluxes in vivo have been employed to describe the fate of nucleotides, amino acids, and organic acids during CFPS. This information combined with our knowledge about E. coli metabolism and genetics has enabled rational metabolic engineering of the source strain to produce cell extracts that exhibit desired performance qualities. Reaction duration has been extended to 20 h from less than 1 h [19], and the technology has evolved to more closely mimic in vivo conditions within a rapidly growing cell [20]. Large cost reductions have resulted from using glucose, glucose 6-phosphate, pyruvate, or glutamate as energy sources and nucleoside monophosphates (NMPs) as the initial nucleotide source [21, 22]. Modification of the source strain has produced more productive cell extracts [23, 24] and CFPS has demonstrated its ability to express and fold several diverse and complex proteins [25, 26]. For example, altering or deleting metabolic pathways that deplete or degrade key substrates needed for efficient protein expression has significantly improved batch CFPS yields [23, 27, 28]. Also, chromosomal deletions of genes that encode for nuclease activity have stabilized DNA and mRNA and enabled the use of expression templates generated from PCR [24]. Selective addition of small molecules has activated metabolic pathways to improve ATP regeneration [27] and protein activity [25], and the addition of chaperones has led to improved protein folding [13, 26, 29]. To design a CFPS platform for high-throughput expression of a functional protein array that could subsequently be screened for gene product activity, three separate advances were required. First, an E. coli strain was engineered to provide a cell extract with enhanced linear DNA stability. Second, methods were developed to efficiently produce transcriptionally-active linear DNA templates using PCR. Finally, the CFPS reaction conditions were optimized with factors required for the activity of some enzymes. 3.2.1 Improving Linear Template Stability

Escherichia coli cell extracts are the most commonly used lysates for CFPS because they exhibit the greatest productivity and are the easiest to prepare [10]. This platform has been well suited for many applications that require simple and efficient methods for protein production. CFPS has experienced limited success, however, in the field of high-throughput protein expression because linear DNA instability has limited the use of PCR-generated expression templates. This limitation has been attributed to nucleases present in the bacterial lysate that lead to linear DNA degradation and low protein yields. DNA is particularly susceptible to degradation by endonuclease I (encoded by endA gene), which makes double-strand breaks in DNA and is the dominant source of endonuclease activity in E. coli [30]. In addition, linear DNA can be degraded by

55

56

3 Functional Genomic Analysis using Sequential Cell-free Protein Synthesis

exonuclease V [31], which is the strongest double stranded DNA exonuclease in E. coli and is also an important enzyme in DNA recombination and repair [32]. To eliminate nuclease activity from the cell extract without impairing E. coli growth, we chose to replace the recCBD operon with the Red recombination system. The Red system, consisting of the exo and bet genes from the bacteriophage lambda, can replace the endogenous recCBD system without impairing the viability and growth rate of the strain [33]. Although the exo gene from λ-phage encodes for a nuclease, it is much less potent than the endogenous exonuclease V. We made the hypothesis that linear DNA would be more stable in the cell extract of this exonuclease mutant strain and, since the mutation does not impair the bacterial growth rate [33], it should not affect the productivity and yield of cell-free protein expression. We modified the genome of the E. coli strain A19 by removing the endA gene encoding endonuclease A and replacing the recCBD operon (in which recD encodes the exonuclease V) by the λ-phage Red recombination system to generate the nuclease deficient NMR5 strain. Replacement of the recCBD genes by the Red recombination system in the NMR5 mutant strain led to PCR fragment stability and allowed for efficient and high level expression of several proteins from PCR products (Fig. 3.2). Using the cell extract from this new strain increased the stability of PCR products amplified from a plasmid encoding for chloramphenicol acetyl transferase (CAT) (Fig. 3.2). Total CAT protein expression from PCR products was comparable to that from plasmids in a batch reaction. Our data demonstrated that DNA stability was the first limitation that must be overcome to improve cell-free protein yields from linear templates. In the absence of linear template stability we did not observe any benefit from other improvements to the strain such as amino acid stabilization. The enhanced stability of linear DNA templates in the NMR5 cell extract led to increased mRNA levels. Hence, we are strongly convinced that the NMR5 strain improved protein production from linear DNA templates at the level of transcription rather than translation. 3.2.2 Improving PCR Reactions for generating Genomic Linear Templates

A challenge facing all functional genomic platforms is synthesis of expression templates (ET) to direct transcription and translation of every gene within a genome. While not a limitation when expressing only a few genes, expressing entire genomes by conventional cloning can be time and labor intensive. The inability to tightly control the expression from many commercial vectors makes cloning of toxic gene products especially challenging. Therefore, a method was needed for making ETs rapidly and reliably without using restriction enzyme digestion, ligation, transformation, bacterial propagation, or plasmid purification. Several methods have been described to generate linear templates for protein expression but no single method has gained wide acceptance [34–40]. To date, most linear ETs used for expressing genomic targets have been amplified from cDNA libraries subcloned into expression plasmids [24, 40]. This facilitates use of generic primers homologous to the flanking regions of the vector, which greatly simplifies the PCR

3.2 Developing an enabling Technology for Sequential Expression Analysis

Fig. 3.2 Effect of extract source cell mutations

on DNA stability and protein expression. (A) DNA concentration versus time in a batch PANOx reaction using a [3 H]-labeled linear DNA template (n = 2). (B) mRNA yields as measured by TCA-precipitated [3 H]UTP at various times during a batch PANOx reaction (n = 2). (C) CAT protein yield as measured by TCA-precipitated [14 C]leucine at various times

in a batch PANOx reaction (n = 3). All reactions were initiated with 3.3 µg mL–1 of a transcriptionally-active linear DNA template encoding the cat gene (a 1.2 kb DNA fragment amplified from the pK7CAT expression vector to produce the cat coding sequence downstream of the PT7 and upstream of the T7 terminator). NMR1 (), NMR2 (), and NMR5 () cell extracts.

method but requires a prior gene cloning step. Using gene-specific primers to amplify targets from complex templates such as genomic DNA and then extending regulatory regions for transcription and translation has proven to be a greater challenge, but this approach is clearly the most advantageous for high-throughput applications. To date, there are no published examples where this approach, known as expression PCR (E-PCR), is applied in parallel to a large number and a wide variety of gene targets. The few examples have been limited to gene targets with similar size and GC-content [41], or to libraries of linear ETs representing mutants of a

57

58

3 Functional Genomic Analysis using Sequential Cell-free Protein Synthesis

single gene [9, 42–46]. In the one case where linear ETs were formed from several different gene targets, the expression elements were generated with multiple PCR steps using plasmid templates and at least one generic primer flanking the gene target [34, 40, 44, 47, 48]. The demand for high-throughput methods to generate templates for expressing genomic libraries led us to design a rapid and reproducible PCR method for making transcriptionally-active ETs. After amplifying the gene targets, these templates were constructed in two separate stages of a single PCR reaction (Fig. 3.3a). The first stage extends T7 regulatory elements onto the gene target, and the second stage uses a GC-rich single primer to amplify the extended template. In this manner, we achieved high yield and specificity for the full-length expression element while eliminating the formation and accumulation of aberrant DNA products. The technique was demonstrated by extending T7 regulatory elements onto a library of diverse gene targets obtained directly from E. coli genomic DNA. Fifty-two of the 55 gene targets were successfully extended and amplified to form their full-length template in high yield and purity (Fig. 3.3b). Only two of the 52 extended linear ETs produced a 0.4-kb contaminant that is prevalent when using other methods and has been suggested to compete for RNA polymerase binding [44, 49, 50]. These results indicated superior yield and purity for the full-length ETs compared with other methods that were examined. Separating the extension PCR from the amplification PCR and using the GC-rich SP3 single-primer to enable higher annealing temperature has collectively enhanced product yield and minimized formation of aberrant products that could inhibit transcription and translation. Furthermore, the linear expression elements can be continuously propagated by re-amplifying with the GC-rich SP3 primer. They can also be cloned into a plasmid using blunt-end ligation or, as we have shown in this work, using specific restriction sites to orient the ORF with respect to transcriptional regulatory regions in the plasmid. The plasmids can then be used for sequencing, scale-up of cell-free protein expression, or in vivo expression. 3.2.3 Optimizing Cofactor Concentrations for Enzyme Activation

Various vitamins, cofactors, and metal ions are lost during cell extract preparation due to extensive dialysis and dilution, consequently requiring that these small molecules and metal ions be reintroduced into the CFPS reaction for the correct assembly and activation of diverse protein targets [25, 51, 52]. Enhancing the availability of these components beyond their physiological concentrations is also required to mature proteins being expressed at high yields. The availability of coenzymes and metal ions is clearly a concern for developing in vitro protein arrays to effectively characterize protein function and biochemical activity [53]. The standard CFPS reaction was supplemented with a multivitamin solution that provided FAD, thiamin, riboflavin, pyridoxal 5′ -phosphate, biotin, lipoic acid, and coenzyme B12 as well as several trace metal ions (Fig. 3.4). The protein targets were expressed in parallel from linear DNA templates using batch reactions

3.2 Developing an enabling Technology for Sequential Expression Analysis

Fig. 3.3 Principle for making linear expression

templates using PCR. (a) Linear expression elements are formed using two sequential PCR reactions. In the first, gene targets are amplified using gene-specific primers that add different extensions upstream and downstream of the gene sequence. The product from this first PCR is then combined with the dsDNA for the T7 promoter and terminator elements. In the subsequent PCR, these transcriptional regulatory elements anneal to the specific extensions on the gene target and prime gene extension to form the full-length expression element. Addition of a

GC-rich end primer amplifies the full-length product. Steps 2 and 3 are conducted in the same PCR reaction tube and are separated into discrete stages by addition of the GC-rich single primer and by changing the annealing temperature after the first 10 cycles. (b) Extension and amplification of transcriptionally-competent PCR templates using a GC-rich single primer. The introduction of T7 regulatory elements adds approximately 420 bp to each ORF. Arrows help indicate extension products that are difficult to visualize or when more than one band appears.

59

60

3 Functional Genomic Analysis using Sequential Cell-free Protein Synthesis

Fig. 3.4 Addition of trace metals, vitamins, and cofactors into the CFPS reaction for optimal accumulation of chloramphenicol acetyltransferase (CAT). (A) Initial supplementation of vitamins, cofactors, and trace metal ions showed CFPS inhibition by the metal ion cocktail (B) Each trace metal was added separately to a final concentration of 250 µm to identify inhibitors of CFPS. (C) Copper(ii), manganese(ii), and zinc(ii) were titrated to final concentrations of 250 µm

(solid white), 120 µm (dots), and 60 µm (stripes) to examine CFPS inhibition. (D) The CFPS reaction was supplemented with a final trace metal formulation that provided iron (ferrous ammonium sulfate), cobalt(ii), molybdenum(vi), and boric acid at a final concentration of 250 µm each, manganese(ii), and zinc(ii) to a final concentration of 30 µm each, and copper(ii) to a final concentration of 60 µm. Vitamins and cofactors were added to a final concentration of 20 µm.

in microtiter plates. The proteins ranged in size from 10 kDa to over 100 kDa and several proteins represent subunits of much larger multienzyme complexes. Most proteins accumulated between 2–5 µm and several low molecular weight protein targets had yields much greater than 10 µm. Some of the target proteins could be visualized above the background proteins on a stained SDS-PAGE, and autoradiography demonstrated that each protein was expressed as a single product. All the enzymes that we examined exhibited activity that was measurable above the background activity associated with the cell extract. Protein targets showed increased activity when produced in CFPS reactions containing vitamin, cofactor, and metal

3.3 Demonstrating Functional Genomic Analysis with CFPS

ion supplementation. For these enzymes, the soluble product was at least 50% of the total accumulated product. Several enzymes with available colorimetric assays were tested and most showed greater than 80% of the expected activity. Several of these enzymes were nicotinamide or flavin enzymes and others required metal ions for activity. Protein purification was avoided because the activity of each enzyme was detectable significantly above the background activity associated with the cell extract. This array represents the largest and most diverse library of proteins expressed in parallel from linear DNA expression templates generated solely by PCR. This work should increase the role of CFPS systems as platforms for highthroughput evaluation of protein function.

3.3 Demonstrating Functional Genomic Analysis with CFPS

Cell-free protein synthesis is a promising platform for characterizing protein function within the context of a specified metabolic pathway. Traditional genetic approaches to metabolic engineering are usually slow and laborious. Metabolic flux modeling coupled with rDNA techniques has brought significant improvements. However, these fail when the metabolic process or the organism is poorly understood. Even with a well-studied organism, unintentional consequences often arise when an enzyme activity is changed. The process of metabolic optimization then becomes an expensive multi-step, iterative process. Often, the key enzymes that control the efficiency of a metabolic process are not known. More modern genetic approaches such as transposon mutagenesis or the expression of chromosomal libraries can then be used, but the results are often incomplete and confusing. The inactivation or overexpression of one protein often has secondary effects that produce phenotypes unrelated to the primary function of the mutated protein. Cellfree methods offer a more stable platform for genetic analysis since genome-based protein synthesis does not occur. We have developed a method that uses sequential rounds of CFPS to identify proteins that influence in vitro transcription, translation, and protein folding. The first round of CFPS creates an array of cell extracts that are individually enriched with a single target protein. The proteins are expressed from linear DNA templates constructed in parallel using a single PCR procedure wherein the overlap-extension reaction for the addition of transcription regulatory elements is separated from a GC-rich single-primer amplification of the full-length product. These expression templates are used to direct the first round of CFPS using a cell extract engineered to enhance linear DNA stability and using reaction conditions conducive for protein function. After the first round of CFPS, the linear DNA templates are specifically degraded, and a second, plasmid-based template for a reporter protein is added to initiate a subsequent round of protein expression. In this manner, the array is screened to identify protein targets that enhance or inhibit the expression and folding of the reporter protein. We report the results from screening 55 E. coli open reading frames (ORFs).

61

62

3 Functional Genomic Analysis using Sequential Cell-free Protein Synthesis

Fig. 3.5 First-round protein synthesis of cytoplasmic effector proteins. Target proteins were expressed in-parallel from linear DNA templates using batch CFPS reactions. Total () and soluble () protein expression were determined from [14 C]leucine incorporation following precipitation by TCA.

This approach should be capable of rapidly and accurately testing the effects of hundreds of enzymes on nucleotide triphosphate synthesis and stability; amino acid degradation or synthesis; mRNA stability; translation initiation, elongation and ribosome recycling; protein stability; and protein folding. 3.3.1 Isolation and Expression of Genomic Targets

Several E. coli proteins were chosen to develop the methodology and verify the concept of our functional proteomic assay. Targets were selected for their anticipated influence on cell-free expression and folding (Table 3.1). For high-throughput analysis, PCR is used to generate linear DNA templates for first round expression of each effector candidate protein. These proteins were expressed in-parallel using a 96-well plate format (Fig. 3.5). Each of our target proteins accumulated to approximately 1–3% of the total protein concentration within a CFPS reaction. This amount is almost certainly enough to exert a significant effect on system performance if the protein has the potential to do so. 3.3.2 Effects of Sample Library on β-Lactamase Expression and Activity

As indicated by Fig. 3.1, the sequential expression method requires that the DNA template used to express the first-round gene product be destroyed before the indicator protein (β-lactamase) is expressed. The first template is an unmethylated PCR product and is therefore a target for the DpnII nuclease. The target site for this nuclease is relatively common since it is a four nucleotide sequence. The procedure was optimized to allow approximately 100 µg mL–1 of the test gene product (about 1% of the total protein) to accumulate before the nuclease is added. This

3.3 Demonstrating Functional Genomic Analysis with CFPS

Fig. 3.6 Expression of pyruvate kinase from its linear ET before (—)

and after (– – –) DpnII addition; DpnII was added 30 min after synthesis is initiated.

occurs approximately 30 min after the reaction is started. At this time, there is still significant synthesis capacity remaining for expression of the indicator protein. We examined protein synthesis time courses to assess whether DpnII treatment is effective in terminating first-round protein synthesis. Figure 3.6 shows protein synthesis yields over the course of the reaction time for a representative effector protein. The general trend is a linear rate of protein synthesis during the first hour followed by a gradual slowing of expression yields over the remainder of the reaction time (Fig. 3.6, solid line). If DpnII is effective, we expect to terminate protein synthesis by degrading the PCR expression templates (Fig. 3.6, dashed line). The expected results were observed for all the effector enzymes with the exception of DHFR. The gradual cessation of protein synthesis that is observed for DHFR may be due to the stability of the mRNA, the position of the DpnII restriction sites within the ORF, or some combination. To increase the probability of effective template digestion for the full genomic survey, two DpnII sites will be placed in the upstream region of the linear ETs. Following termination of first round synthesis, a second round of protein synthesis is activated upon addition of the plasmid template used to express our indicator protein, β-lactamase, a 29 kDa monomeric protein that is well characterized in the literature and has a colorimetric assay to determine its activity. This indicator protein requires no prosthetic groups or cofactors for its activity. A single disulfide bond has been implicated in its thermal stability but is unnecessary for activity. We measured the change from baseline yields of total and soluble protein for βlactamase in the presence of each effector protein and observed that almost half of our effector enzymes have little or no effect on total or soluble protein yields. In contrast, several proteins such as the chaperone Skp, the flavoprotein thioredoxin reductase, and even pyruvate kinase enhanced the total protein yields of our indicator protein. Other proteins such as the disulfide isomerase DsbC and glutathione

63

64

3 Functional Genomic Analysis using Sequential Cell-free Protein Synthesis

reductase, a flavoprotein responsible for maintaining the reduced environment of the cytoplasm, decreased the solubility of β-lactamase. These preliminary data are promising and allow us to begin to speculate on what the effector proteins are doing to enhance or inhibit protein synthesis and folding.

3.4 Conclusions and Projections

Continued development of CFPS as a platform for rapid and economical production of therapeutic proteins will benefit from our ability to identify factors that can improve in vitro expression and folding. Insights derived from the in vitro survey promise to significantly improve the productivity, yield, and duration of cell-free protein synthesis and folding. Rational metabolic pathway engineering using the targets identified in this screen could potentially lead to faster and more significant improvements than traditional metabolic engineering approaches. Alternatively, the desired performance can be achieved by introducing purified components or removing deleterious factors. The sequential expression system can also be used with different reporter proteins to explore other metabolic pathways or activities. Many existing platforms used to describe protein function have been limited to description of protein–protein or protein–ligand interactions [4], and few systems can probe entire metabolic pathways or networks [54, 55]. Combined transcription, translation, and protein folding is arguably the most central and most complex metabolic system and is likely to be the most general assay that can be devised. This initial survey is a general primary screen for metabolic activity and, since CFPS provides a nearly constant composition of the reaction components, misleading secondary effects associated with in vivo gene expression and regulation are avoided.

References 1 LaBaer, J., N., Ramachandran, Protein microarrays as tools for functional proteomics. Curr. Opin. Chem. Biol., 2005, 9(1), 14–19. 2 Kung, L. A., M., Snyder, Proteome chips for whole-organism assays. Nat. Rev. Mol. Cell Biol., 2006, 7(8), 617–622. 3 Zhu, H. et al., Analysis of yeast protein kinases using protein chips. Nat. Genet., 2000, 26(3), 283–289. 4 Zhu, H. et al., Global analysis of protein activities using proteome chips. Science, 2001, 293(5537), 2101–2105. 5 Ramachandran, N. et al., Selfassembling protein microarrays. Science, 2004, 305(5680), 86–90.

6 Vemuri, G. N., A. A. Aristidou, Metabolic engineering in the -omics era: Elucidating and modulating regulatory networks. Microbiol. Mol. Biol. Rev., 2005, 69(2), 197– 216. 7 Kawahashi, Y. et al., In vitro protein microarrays for detecting protein–protein interactions: Application of a new method for fluorescence labeling of proteins. Proteomics, 2003, 3(7), 1236–1243. 8 Sawasaki, T. et al., Genome-scale, biochemical annotation method based on the wheat germ cell-free protein synthesis system. Phytochemistry, 2004, 65(11), 1549–1555.

References 9 Rungpragayphan, S., H. Nakano, T. Yamane, PCR-linked in vitro expression: A novel system for highthroughput construction and screening of protein libraries. FEBS Lett., 2003, 540(1–3), 147–150. 10 Katzen, F., G. Chang, W. Kudlicki, The past, present and future of cell-free protein synthesis. Trends Biotechnol., 2005, 23(3), 150–156. 11 Kigawa, T. et al., Cell-free production and stable-isotope labeling of milligram quantities of proteins. FEBS Lett., 1999, 442(1), 15–19. 12 Kigawa, T. et al., Selenomethionine incorporation into a protein by cell-free synthesis. J. Struct. Functional Genomics, 2002, 2(1), 29–35. 13 Yang, J. H. et al., Rapid expression of vaccine proteins for B-cell lymphoma in a cell-free system. Biotechnol. Bioeng., 2005, 89(5), 503–511. 14 Kanter, G. et al., A novel method for producing custom-made idiotype vaccines for lymphoma immunotherapy using a cell-free expression system. Blood, 2004, 104(11), 395A. 15 Lee, C. H., P. Leeds, J. Ross, Purification and characterization of a polysomeassociated endoribonuclease that degrades c-myc mRNA in vitro. J. Biol. Chem., 1998, 273(39), 25261–25271. 16 Swartz, J. R., Advances in Escherichia coli production of therapeutic proteins. Curr. Opin. Biotechnol., 2001, 12(2), 195–201. 17 Shimizu, Y. et al., Cell-free translation reconstituted with purified components. Nat. Biotechnol., 2001, 19(8), 751–755. 18 Altenbuchner, J., R. Mattes, Escherichia coli, in: Production of Recombinant Proteins: Novel Microbial and Eukaryotic Expression Systems, ed. G. Gellissen, 2005, WileyVCH, Weinheim, pp. 7–43. 19 Spirin, A. S. et al., A continuous cell-free translation system capable of producing polypeptides in high yields. Science, 1988, 242(4882), 1162–1164. 20 Jewett, M. C., J. R. Swartz, Mimicking the Escherichia coli cytoplasmic environment activates long-lived and efficient cellfree protein synthesis. Biotechnol. Bioeng., 2004, 86(1), 19–26.

21 Kim, D. M., J. R. Swartz, Regeneration of adenosine triphosphate from glycolytic intermediates for cell-free protein synthesis. Biotechnol. Bioeng., 2001, 74(4), 309–316. 22 Calhoun, K. A., J. R. Swartz, Energizing cell-free protein synthesis with glucose metabolism. Biotechnol. Bioeng., 2005, 90(5), 606–613. 23 Michel-Reydellet, N., K. Calhoun, J. Swartz, Amino acid stabilization for cellfree protein synthesis by modification of the Escherichia coli genome. Metabol. Eng., 2004, 6(3), 197–203. 24 Michel-Reydellet, N., K. A. Woodrow, J. R. Swartz, Increasing PCR fragments stability and protein yields in a cell-free system with genetically modified Escherichia coli extracts. J. Mol. Microbiol. Biotechnol., 2005, 9(1), 26–34. 25 Knapp, K., J. Swartz, Cell-free production of active E-coli thioredoxin reductase and glutathione reductase. FEBS Lett., 2004, 559(1–3), 66–70. 26 Yin, G., J. R. Swartz, Enhancing multiple disulfide bonded protein folding in a cell-free system. Biotechnol. Bioeng., 2004, 86(2), 188–195. 27 Kim, D. M., J. R. Swartz, Oxalate improves protein synthesis by enhancing ATP supply in a cell-free system derived from Escherichia coli. Biotechnol. Lett., 2000, 22(19), 1537–1542. 28 Kim, D. M., J. R. Swartz, Prolonging cellfree protein synthesis with a novel ATP regeneration system. Biotechnol. Bioeng., 1999, 66(3), 180–188. 29 Yang, J. H. et al., Expression of active murine granulocyte-macrophage colonystimulating factor in an Escherichia coli cell-free system. Biotechnol. Prog., 2004, 20(6), 1689–1696. 30 Lehman, I. R., E. A. Pratt, G. G. Roussos, Deoxyribonucleases of Eshcerichia coli, II. Purification and properties of a ribonucleic acid-inhibitable endonuclease. J. Biol. Chem., 1962, 237(3), 819–828. 31 Yang, H. L. et al., Cell-free coupled transcription-translation system for investigation of linear DNA segments. Proc. Natl. Acad. Sci. U.S.A.-Biol. Sci., 1980, 77(12), 7029–7033.

65

66

3 Functional Genomic Analysis using Sequential Cell-free Protein Synthesis 32 Wright, M., G. Buttin, J. Hurwitz, Isolation and characterizaiton from Escherichia coli of an adenosine triphosphate-dependent deoxyribonuclease directed by recB,C genes. J. Biol. Chem., 1971, 246(21), 6543– 6555. 33 Murphy, K. C., Use of bacteriophage lambda recombination functions to promote gene replacement in Escherichia coli. J. Bacteriol., 1998, 180(8), 2063–2071. 34 Nemetz, C., Generation of linear expression elements by PCR, in Cell-free Protein Expression, ed. J. R. Swartz, 2003, Springer, New York, pp. 5–7. 35 Olsnes, S. et al., Formation of active diptheria-toxin in vitro based on ligated fragments of cloned mutant-genes. J. Biol. Chem., 1989, 264(22), 12749–12751. 36 Lesley, S. A., M. A. D. Brow, R. R. Burgess, Use of in vitro protein-synthesis from polymerase chain-generated templates to study interaction of Escherichia coli transcription factors with core RNApolymerase and for epitope mapping of monoclonal antibodies. J. Biol. Chem., 1991, 266(4), 2632–2638. 37 Xin, W., J. Ma, D. W. Huang, Assembly of linear functional expression elements with DNA fragments digested with asymmetric restriction endonucleases. Biotechnol. Lett., 2003, 25(11), 901–904. 38 Xin, W., J. Ma, D. W. Huang, Construction of linear functional expression elements with DNA and RNA hybrid primers: A flexible and fast method for proteomics. Biotechnol. Lett., 2003, 25(3), 273–277. 39 Sykes, K. F., S. A. Johnston, Linear expression elements: A rapid, in vivo, method to screen for gene functions. Nat. Biotechnol., 1999, 17(4), 355–359. 40 Sawasaki, T. et al., A cell-free protein synthesis system for high-throughput proteomics. Proc. Natl. Acad. Sci. U.S.A., 2002, 99(23), 14652–14657. 41 Merk, H., D. Meschkat, W. Stiege, Expression-PCR: from gene pools to purified proteins within one day, in Cell-free Protein Expression, ed. J. R. Swartz, 2003, Springer, New York, pp. 15–23.

42 Betton, J. M., High throughput cloning and expression strategies for protein production. Biochimie (Paris), 2004, 86(9–10), 601–605. 43 Miyazaki-Imamura, C. et al. Improvement of H2 O2 stability of manganese peroxidase by combinatorial mutagenesis and high-throughput screening using in vitro expression with protein disulfide isomerase. Protein Eng., 2003, 16(6), 423–428. 44 Ohuchi, S., H. Nakano, T. Yamane, In vitro method for the generation of protein libraries using PCR amplification of a single DNA molecule and coupled transcription/translation. Nucleic Acids Res., 1998, 26(19), 4339–4346. 45 Rungpragayphan, S. et al., Rapid screening for affinity-improved scFvs by means of single-molecule-PCR-linked in vitro expression. J. Mol. Catal. B: Enzymatic, 2004, 28(4–6), 223–228. 46 Rungpragayphan, S. et al., Highthroughput, cloning-independent protein library construction by combining singlemolecule DNA amplification with in vitro expression. J. Mol. Biol., 2002, 318(2), 395– 405. 47 Norais, N. et al., Combined automated PCR cloning, in vitro transcription/translation and two-dimensional electrophoresis for bacterial proteome analysis. Proteomics, 2001, 1(11), 1378– 1389. 48 Resto, E. et al., Amplification of protein expression in a cell free system. Nucleic Acids Res., 1992, 20(22), 5979–5983. 49 Graentzdoerffer, A., C. Nemetz, High-throughput expression-PCR using universal plasmid-specific primers. BioTechniques, 2003, 34(2), 256–260. 50 Graentzdoerffer, A., C. Nemetz, Reduction of primer-dimer formation during generation of expression fragments by PCR, in Cell-free Protein Expression, ed. J. R. Swartz, 2003, Springer, New York. 51 Boyer, M. E., J. R. Swartz, Simultaneous expression and maturation of the ironsulfur protein ferredoxin in a cell-free system. Biotechnol. Bioeng., 2006, 94(1), 128–138. 52 Sawasaki, T. et al., A bilayer cell-free protein synthesis system for high-throughput screening of gene products. FEBS Lett., 2002, 514(1), 102–105.

References mization and in vitro metabolic engineer53 Woodrow, K. A., I. O. Airen, J. R. Swartz, ing. Science, 2004, 304(5669), 428–431. Reliable generation of linear DNA tem55 Service, R. F., Proteomics – Protein chips plates for expressing functional genomic map yeast-kinase network. Science, 2005, libraries. J. Proteome Res., 2006, 5(12), 307(5717), 1854–1855. 3288–3300. 54 Jung, G. Y., G. Stephanopoulos, A functional protein chip for pathway opti-

67

69

4

Cell-free Technology for Rapid Production of Patient-specific Fusion Protein Vaccines Aaron R. Goerke, Junhao Yang, Gregory Kanter, Ronald Levy, and James R. Swartz

4.1 Introduction 4.1.1 Lymphoma and Fusion Protein Vaccine Treatments

Non-Hodgkin’s lymphoma (NHL) is a heterogeneous group of lymphoproliferative malignancies with differing patterns of behavior and responses to treatment [1]. The disease usually originates in the lymphoid tissues and can spread to other organs. The incidence of NHL is increasing [2], and, using standard treatment, the five-year survival rate is only 50–60%. Although 30% of patients with NHL can be completely cured, relapse is common within 2 years following aggressive radiation, chemotherapy, and immunotherapy treatment. The limited success of radiation and chemotherapy treatments has suggested that every malignancy is unique and requires individualized biological treatment for increased efficacy. One individualized strategy for lymphoma immunotherapy is to use tumor idiotypic antigens as vaccines to elicit an anti-idiotype immune response. Immunization of animals with the Id (idiotype) protein expressed by the B cell lymphoma (a monoclonal antibody) can induce specific protection against subsequent tumor challenge [3]. Several generations of Id vaccines have been designed and investigated with a carcinogen-induced 38C13 murine B cell tumor as a model system [4]. Initially, the syngeneic Id showed effective tumor immunity when it was conjugated to a strong carrier such as keyhole limpet hemocyanin (KLH) and mixed with an adjuvant [5]. Later, it was demonstrated that a soluble “built in” adjuvant could be fused to the full antibody to substitute for a carrier protein and an adjuvant [6, 7]. Fusion proteins composed of the immunogen specific Id (in this case, the variable region of the antibody) and cytokines, such as granulocyte-macrophage colony stimulating factor (GM-CSF), interleukin-2, interleukin-4 and interferon-γ , produced protective anti-tumor immune responses in mice. A third generation of vaccines consisted of naked DNA encoding the Id-GM-CSF fusion structure and these also protected mice from tumor challenge [8]. Furthermore, single-chain variable fragment (scFv) fusion proteins containing the variable region of the heavy chain Cell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

70

4 Cell-free Technology for Rapid Production

(VH ) and light chain (VL ) can be linked by a flexible peptide spacer (Gly4 Ser)3 to murine GM-CSF (GM) or an immunoenhancing peptide derived from interleukin1β. These fusion proteins served as effective vaccines to elicit protection against a murine B cell lymphoma [9]. 4.1.2 Comparing Cell-free and In Vivo Production Systems

Most malignancies of B lymphocytes represent the transformation and expansion of a single B cell, which expresses an immunoglobulin (idiotype) on its surface containing unique variable region sequences. Therefore, one important feature of B cell lymphoma immunotherapy is the extreme variability in the vaccine antigen. It is this variability that requires patient-specific vaccination. The current commercial design of idiotype vaccines is to use a whole immunoglobulin derived from each B cell tumor and to conjugate this to KLH, a highly immunogenic molecule. Several different in vivo production methods have been used to produce the immunoglobulin (Ig) portion of the vaccine [10, 11]. Two on-going clinical trials use a full-length Ig fused to KLH, which is co-administered with GM-CSF. The Id-specific vaccine proteins are either produced by mammalian cells (Genitope Incorporated) or by insect cells (Favrille Incorporated), and the associated time and labor costs impose significant barriers for the production of patient-specific vaccines. Using an E. coli-based cell-free protein expression system, we can overcome many limitations of traditional in vivo production systems. Advantages of a cellfree system include but are not limited to the rapid production of a single protein product, high-throughput production, and control of the reaction environment to direct metabolism and improve protein folding. While plasmid-encoded expression of the tumor-specific Ig variable region genes is required for our present research, it is possible to express patient-specific vaccines from linear DNA templates which are amplified by PCR [12]. This provides additional time savings compared with common in vivo recombinant DNA production methods. Because the cell-free reaction mixture is not enclosed by a cell membrane, the purification is simplified relative to other E. coli processes. The product can be purified, characterized, and ready to use as a vaccine in a matter of days. Therefore, the cell-free technology can provide a much more rapid and cost-effective method for producing many proteins in parallel for patient-specific treatments.

4.2 Developing the Fusion Protein Construct and the Cell-free Production Process

4.2 Developing the Fusion Protein Construct and the Cell-free Production Process 4.2.1 Fusion-protein Production in the Cell-free System

Since the Id structure alone is normally very weak as an immunogen, prospective B cell lymphoma vaccine proteins usually contain an immunoenhancer. We use murine GM-CSF as the immunoenhancer in this work because it is a very effective fusion partner for a B cell lymphoma vaccine [6]. Five recombinant vaccine forms were evaluated. They include scFv (VH -VL ), VH -VL -GM, VL -VH -GM, GM-VH -VL and GM-VL -VH as listed in Fig. 4.1a. Here, GM represents murine GM-CSF; VH and VL represent the variable regions of the mouse B lymphoma 38C13 Id heavy chain and light chain, respectively. All these constructs contain a 5′ cat DNA fragment that encodes the amino-terminal five amino acids of the E. coli chloramphenicol acetyl transferase (cat) gene. We find that this modification often increases the protein yield (data not shown). The construction of the expression plasmids for these five recombinant proteins has been reported by Yang et al. [13]. All these constructs were cloned into the expression plasmid pK7 under the control of the standard T7 promoter. The recombinant proteins were expressed in the combined transcription and translation cell-free system as described by Jewett and Swartz [14]. Normally, the reaction mixture contains the following components, 20 mm magnesium glutamate, 10 mm ammonium glutamate, 170 mm potassium glutamate, 1.2 mm ATP, 0.86 mm each of GTP, UTP and CTP, 34 µg mL–1 folinic acid, 170.6 µg mL–1 E. coli tRNAs, 2 mm each of the twenty amino acids, 2 mm additional cysteine, 30 mm pyruvate, 0.33 mm NAD, 0.27 mm CoA, 2.7 mm oxalic acid, 1 mm putrescine, 1.5 mm spermidine, 10 mm Tris, 10.5 mm L-[U-14 C]leucine, 0.1 mg mL–1 T7 RNA polymerase, 0.0133 mg mL–1 plasmid DNA and 24% NMR2 S30 E. coli cell extract. The single-chain variable fragment and the four GM-CSF/scFv fusion proteins, encoded by pK7.catscFv, pK7.catscFvGM, pK7.catVL -VH -GM, pK7.catGM-VL -VH and pK7.catGM-VH -VL , were expressed in the cell-free system. These combined transcription/translation cell-free reactions were normally conducted at a total volume of 30 µL in an Eppendorf tube at 30 ◦ C in a temperature-controlled incubator. After 5 hours, 5 µL of the reaction mixture was applied to a 10% Bis-Tris SDS-PAGE gel and analyzed by autoradiography. As shown in Fig. 4.1b, cell-free expression was limited to the expected products as indicated by the appropriate molecular weight of the expression products (28.5 and 42.8 kDa). 4.2.2 Oxidized Reaction Conditions and DsbC Increase Soluble Protein Yield

Disulfide bonds play a very important role in the structure and function of GM-CSF (two disulfide bonds) and scFv (two disulfide bonds). We observed previously that proper folding of a recombinant protein with disulfide bonds requires a relatively

71

72

4 Cell-free Technology for Rapid Production

Fig. 4.1 Fusion protein constructs and cell-free protein synthesis results. (a) Diagrams of scFv and GM-CSF/scFv fusion proteins expressed in the cell-free system. (b) Autoradiography analysis of the cell-free expression. Five microliters of the total reaction product was separated by a 10% Bis-Tris SDS-PAGE gel. 1. scFv (VH -VL ); 2. VH -VL -GM; 3. VL -VH -GM; 4. GM-VL -VH ; 5. GM-VH -VL .

oxidizing redox potential and a disulfide isomerase [15–17]. However, the S30 cell extract prepared from E. coli is normally in a reducing state relative to disulfide bonds. To minimize protein misfolding, we modified the sulfhydral redox potential of the cell-free reaction mixture by two measures. First, we used iodoacetamide (IAM) to pretreat the S30 cell extract to inactivate the free sulfhydryl groups required by the cytoplasmic oxidoreductases. In this way, the sulfhydral redox potential of the cell-free reaction mixture can be stabilized without a significant loss of protein synthesis activity [16]. Second, we used a glutathione buffer composed of 4 mm oxidized glutathione (GSSG) and 1 mm reduced glutathione (GSH) to adjust the redox potential. Combining these two measures, we maintain a controllable and stabilized oxidizing environment. Although an oxidizing environment can support efficient disulfide bond formation, it can not guarantee correct disulfide linkages. When expressed in E. coli, recombinant proteins with disulfide bonds are normally secreted into the periplasm for disulfide bond formation. To catalyze disulfide exchange, E. coli uses DsbC as a disulfide isomerase [18]. DsbC is maintained in its active reduced state by another membrane protein, DsbD [19]. In our previous report about cell-free expression of murine GM-CSF, DsbC was very important for

4.2 Developing the Fusion Protein Construct and the Cell-free Production Process Table 4.1 Total and soluble yields of the scFv and murine GM-VL -VH

protein in the cell-free expression system under different reaction conditions. Expr. Conditions scFv (µg mL–1 ) number Glutathione buffer (mM) IAMa) DsbCb) Total Soluble

GM-VL -VH (µg mL–1 ) Total Soluble

1 2 3 4 5 6

326 301 238 322 344 359

– – 4 GSSG/1 GSH 4 GSSH/1 GSH 5 GSH 4 GSSG/1 GSH

– – – + + +

– + + + – –

a)

Final IAM concentration was 0.5 mm.

b)

Final DsbC concentration was 115 µg mL–1 .

277 250 221 240 241 259

37 37 92 109 46 69

20 20 75 139 25 48

the production of active GM-CSF [17]. Therefore, in the expression of vaccine fusion proteins, DsbC was also added to catalyze rearrangement of incorrectly formed disulfide bonds. The concentration of IAM for cell extract pretreatment, the GSH and GSSG concentrations, and the DsbC concentration were optimized for production of soluble protein. The total protein yield was calculated by the amount of L-[U-14 C]leucine incorporated into the final product, and the soluble protein yield was determined after the reaction mixture was centrifuged at 15 000g for 15 min. Table 4.1 shows the effects of IAM pretreatment, the glutathione buffer, and DsbC on the expression of scFv and GM-VL -VH . Clearly, each of the modifying factors contributes to the improved soluble protein yields. Interestingly, though, there was no increase in soluble yield with only the addition of DsbC, indicating that DsbC could not promote disulfide formation by air oxidation in the unmodified cell-free system. The soluble yield of scFv and GM-VL -VH were only 37 and 20 µg mL–1 under the unmodified reaction conditions. They were increased to 109 and 139 µg mL–1 under the modified conditions with 4 mm GSSG/1 mm GSH, 0.5 mm IAM cell extract pretreatment (30 min incubation at room temperature) and 115 µg mL–1 DsbC. The expression of the other fusion proteins was also investigated. The same optimized conditions produced maximal soluble protein yields. Under the conditions of Experiment number 4 in Table 4.1, the soluble yields of GM-VH -VL , VH -VL -GM, and VL -VH -GM reached 109, 76 and 59 µg mL–1 , respectively. 4.2.3 GM-CSF is more Active at the N-terminus of the Fusion Protein Vaccine

The biological activity of the murine GM-CSF portion of the cell-free expressed fusion proteins was assayed using a murine GM-CSF-dependent cell line, NFS-60 [20]. The soluble cell-free expressed fusion proteins and the standard murine GMCSF (E. coli derived murine GM-CSF from R&D Systems) were serially diluted in

73

74

4 Cell-free Technology for Rapid Production Table 4.2 Cell proliferation activities of the cell-free expressed fusion vaccine constructs.

Construct

ED50 (pM)a)

(R+D) standard GM-CSF VH -VL -GM VL -VH -GM GM-VH -VL GM-VL -VH

21.0 ± 0.8 41.6 ± 3.0 181.8 ± 13.0 11.9 ± 0.6 12.1 ± 3.1

a)

ED50 is the concentration of murine GM-CSF or the fusion protein that gave one-half the maximum incorporation of [3 H]thymidine into NFS-60 cells.

triplicate. Equal volumes of RPMI media (Invitrogen, NY) with 10% FCS and the log phase NSF-60 cell culture were added to the serially diluted cell-free products and the murine GM-CSF standard to test their ability to stimulate cell proliferation. NFS-60 cells were inoculated at a concentration of 5000 cells in flat-bottom 96-well tissue culture plates (Falcon Microtest 96). After incubation at 37 ◦ C and 5% CO2 for 16–20 hours, 50 µL of [3 H]thymidine (Amersham) was added to each well at a final concentration of 6.7 µCi mL–1 , and proliferation was monitored by the incorporation of [3 H]thymidine. Following 8–10 hours of incubation at 37 ◦ C and 5% CO2 the cells were harvested onto glass fiber filter mats and washed. The [3 H]thymidine incorporation was measured using a Wallach 1450 Microbeta scintillation counter (Perkin Elmer Life Sciences). Both standard murine GM-CSF and cell-free synthesized GM-CSF fusion proteins stimulated NFS-60 proliferation (Fig. 4.2) while an irrelevant product produced in the cell-free system did not (data not shown). The ED50 , the concentration of the protein that gave one-half the maximum incorporation of [3 H]thymidine into NFS-60 cells, was also assessed for the standard murine GM-CSF and for each fusion construct (Table 4.2). The results suggest that the two fusion constructs with GM-CSF at the amino-terminus were most active. The specific activities of GM-VL -VH and GM-VH -VL were approximately twice as high as the murine GM-CSF standard. This suggests the cell-free system produces active disulfide bonded fusion proteins with GM-CSF activity comparable to that produced by an in vivo system. In contrast, although VH -VL -GM (HLG) showed higher GM-CSF specific activity than VL -VH -GM (LHG), both were significantly less active than the standard. Figure 4.2 indicates that the two cell-free expressed fusion proteins with carboxyterminal GM-CSF had much lower cell proliferation activity than the GM-CSF standard. In contrast, the fusion proteins with amino-terminal GM-CSF exhibited high biological activity. Two possible reasons were suggested to explain this phenomenon. First, the amino-terminal GM-CSF might fold more effectively than the carboxy-terminal GM-CSF. It was evident from our previous work that murine GMCSF itself could fold correctly with high efficiency during cell-free expression [17]. Therefore, the possibility of poor folding for carboxy-terminal GM-CSF could be due to interference from the scFv. Assuming a co-translational mechanism for protein folding, the amino-terminal part of a growing peptide begins to fold as soon

4.2 Developing the Fusion Protein Construct and the Cell-free Production Process

Fig. 4.2 Stimulation of NSF-60 cell proliferation by GM-CSF/scFv

fusion proteins. The soluble cell-free expressed fusion proteins and the standard murine GM-CSF were tested for their ability to stimulate proliferation of an NSF-60 cell line. Cell-free reaction products were added according to the radioactivity recovered after TCA precipitation. R+D, murine GM-CSF standard from R&D systems; GLH, GM-VL -VH ; GHL, GM-VH -VL ; LHG, VL -VH -GM; HLG; and VH -VL -GM.

as it has been synthesized and prior to the completion of the entire polypeptide chain. For the N-terminal GM-CSF fusions, the GM-CSF has the opportunity to fold as the scFv domain is still emerging from the ribosome, thereby decreasing the interference of the carboxy-terminal scFv on GM-CSF folding. The second possible reason might be improved accessibility for GM-CSF receptor binding when the GM-CSF was moved to the amino-termini of the fusion proteins. According to a model based on the three-dimensional structure of GM-CSF, the activity related residues form a continuous surface that encircles the molecule at one end that is very close to the amino-terminus [21]. Thus the receptor binding might be hindered if scFv is connected to the amino-terminus of GM-CSF even if the GM-CSF is properly folded. Moving the scFv fragment to the carboxy-terminus of GM-CSF might avoid this steric hindrance. 4.2.4 New Linker Improves Fusion Protein Stability

As we attempted to purify the GM-VL -VH fusion protein, it quickly degraded (Fig. 4.3b). This instability might be due to the poor folding of the scFv domain. Because the folding of the N-terminal murine GM-CSF is acceptable, the assumed poor folding of the C-terminal scFv might be due to interference by the N-terminal murine GM-CSF. Therefore we designed a new linker that contains the structure

75

76

4 Cell-free Technology for Rapid Production

Fig. 4.3 Autoradiography of the purified vaccine constructs with and without the Im9 linker. (a) Schematic of fusion vaccines: 1. GM-VL -VH ; 2. GM-Im9-VL -VH . (b) Autoradiography analysis of purification fractions 1. Loaded sample of GM-VL -VH ; 2. elution fraction of GM-VL -VH ; 3. loaded sample of GM-Im9-VL -VH ; 4. elution fraction of GM-Im9-VL -VH .

of a whole protein, E. coli Im9, to separate the two domains of the B cell lymphoma vaccine protein. It is assumed that this larger linker will decrease the interaction of the two protein domains during folding. The linkers in fusion protein structures are normally short peptides [22]. However, in the present study, we use a bacterial immunity protein (Im9) that has 85 amino acids as a linker [23]. It contains four α-helices in its structure and folds very quickly. We reasoned also that the rapid folding of Im9 would reduce the chance of the linker interfering with the folding of the two protein domains it connects. We further reasoned that the spatial separation provided by Im9 might improve the folding of the C-terminal scFv.

4.2 Developing the Fusion Protein Construct and the Cell-free Production Process

To produce the new vaccine, the Im9 linker gene is inserted into the GM-VL -VH sequence to replace the sequence encoding the GGGGS linker. The new chimeric gene yields a new fusion protein, murine GM-Im9-VL -VH (Fig. 4.3a). The DNA sequence encoding the Im9 protein and the construction of the new fusion protein is described in Ref. [24]. The fusion proteins encoded by pK7.catGM-VL -VH and pK7.catGM-Im9-VL -VH were expressed in 6 well tissue-culture plates (Falcon) when they were produced at the 1 mL scale. The cell-free reaction is carried out at 30 ◦ C for 4 hours. The soluble yield of murine GM-Im9-VL -VH under the modified reaction conditions (Experiment 4 in Table 4.1) reached 146 µg mL–1 as compared with 139 µg mL–1 for GM-VL -VH . The cell-free expressed murine GM-VL -VH and GM-Im9-VL -VH were purified with a 1 mL HisTrap chelating column (Amersham Biosciences) as described in our previous report [17]. The purification samples were analyzed by SDS-PAGE and autoradiography as described previously [13]. Comparing the old construct with the GGGGS linker, the new fusion protein with the Im9 linker shows less degradation during purification (Fig. 4.3a and b). 4.2.5 Expression and Purification Scale-up for Vaccine Protein Production

A simple and inexpensive scale-up technology for cell-free protein expression has been developed in our laboratory [25]. To produce fusion protein vaccines for tumor challenge studies, the in vitro reactions were scaled up to 30–40 mL. A 5 mL reaction mixture was placed as a large drop in a standard sterile Fisherbrand Petri dish (100 × 15 mm) and incubated at 30 ◦ C for 4 hours. Normally, the protein yield of a big-drop scale-up expression is higher than the small scale expression in a test tube. This is due to the increase in surface to volume ratio, enabling good gas transfer and efficient energy regeneration and also providing more hydrophobic surface area. After 4 hours, the soluble protein was separated by centrifugation at 15 000g for 15 min. The soluble fraction was loaded on a 5 mL Ni-NTA column (Qiagen), which was equilibrated with 10 mm imidazole, 50 mm phosphate buffer (pH 8.0), and 300 mm NaCl. The column was then washed with 30 mL of 25 mm imidazole in the same buffer and eluted with 250 mm imidazole. The purified products were then concentrated with Amicon Ultra-15 Centrifugal Filter Units (5000 MWCO) and dialyzed with 7000 MWCO Slide-A-Lyzer (PIERCE) units against phosphatebuffered saline (PBS). Lastly, the vaccine was formulated to a final concentration of 0.1% Tween-20 (Sigma) and sterilized by filtration through a 0.2 mm filter (Nalgene). The purified vaccines, shown in Fig. 4.4, were stored at 4 ◦ C before vaccination.

77

78

4 Cell-free Technology for Rapid Production

Fig. 4.4 Purified protein products from cell-free reactions. After 4 h of cell-free expression, products were purified from the cell-free reaction supernatant using a nickel affinity column. Shown is a Coomassie staining of the partially purified products separated by a 10% Bis-Tris SDS-PAGE gel. 1. Murine GM-Im9-VL VH ; 2. VL VH -IL1β9aa; 3. scFv (VH VL ) (Id-KLH and DsbC data not shown).

4.3 Fusion Proteins Raise Protective Antibodies 4.3.1 Design of Vaccine Constructs and Mouse Studies

We next determined the ability of the cell-free produced vaccines to protect mice from a tumor challenge. Two experimental vaccines were tested. The first was the murine GM-Im9-VL -VH vaccine described previously. The second was a vaccine containing the same 38C13 single chain antibody fragment fused to a nine amino acid peptide from interleukin-1β (IL1β9aa, [9]). It was interesting to test such constructs because, if effective, they would lead to much simpler protein products. Our experience is that smaller and less complicated proteins tend to have higher levels of expression in the cell-free system. Evaluating IL1β9aa-scFv vaccines would also test the robustness of our cell-free system. As a positive control we used the in vivo produced, whole IgG antibody derived from the 38C13 B cell tumor. Similar to the Genitope vaccine candidate, the antibody was conjugated with KLH. The scFv by itself was included to asses its stability and to investigate the importance of

4.3 Fusion Proteins Raise Protective Antibodies

the attached immunoenhancer. The negative control was cell-free produced E. coli disulfide isomerase (DsbC). Mice were subjected to three rounds of vaccination (bi-weekly) and then challenged with 400 38C13 tumor cells (intraperitoneally) two weeks after the third vaccination. The mice were vaccinated three times intraperitoneally with 50 µg of idiotype Ig conjugated to KLH, 15 µg of the appropriate scFv, or the molar equivalent of scFv fusion proteins or cell-free produced DsbC. Each protein was diluted into 200 to 500 µL of PBS. Before the tumor challenge, 38C13 cells were thawed, washed three times in RPMI/10% FCS/Pen-Strep/L-Glutamine and grown in culture for 3 days. The cells were then washed three times with PBS and diluted to the correct concentration prior to injection. 4.3.2 Fusion Protein Vaccination Protects against Aggressive Tumors

Figure 4.5a and b summarizes the mouse survival data [24]. In two separate experiments C3H/Hen female mice age 6–8 weeks were obtained from Harlan Sprague– Dawley (San Diego, CA) and were housed in the Laboratory Animal Facility at Stanford University Medical Center. The carcinogen-induced IgM/κ 38C13 murine B cell lymphoma cell line used has been described by Maloney et al. [26]. All of the cell-free vaccines tested provided protection as compared with mice immunized with the negative control protein (DsbC). It was concluded that the murine GMIm9-VL VH fusion vaccine works as well as mammalian cell produced full-length tumor IgG fused to KLH and better than the VL -VH -IL1β9aa vaccine (previously reported to be effective by Hakim et al.) and the scFv (VH VL ) [9]. Moreover, the murine GM-Im9-VL VH fusion protected 90% of the mice vaccinated. These results also clearly show, for the first time, that a cell-free system can be used to produce tumor specific anti-cancer vaccines that successfully provide protection against cancer in vivo. 4.3.3 Antibody Generation is enhanced by Fusion Partners

The data shown in Fig. 4.5b suggest that specific elements of the vaccine construct contribute to vaccine efficacy. Specifically we began to evaluate the relative importance of the IL1β9aa, murine GM-CSF and/or Im9 sequences in protecting the micel during a tumor challenge study. In these experiments the scFv-fusion vaccine containing the nine amino acid sequence of 1L1β resulted in a 10% increase in survival rate over VH -VL alone. This is not statistically significant. It is not clear whether the added stability or the functionality of the murine GMIm9 fusion, on the N-terminus, is the cause of the 20–30% increase in mouse survival rate, as compared with VL VH (Fig. 4.5b). Im9 may be beneficial as a carrier protein to improve the folding of the scFv or might directly stimulate the immune system. Im9 is an “immunity” protein produced by E. coli bacteria to protect against colicin molecules they use to gain a competitive advantage over other bacte-

79

80

4 Cell-free Technology for Rapid Production

Fig. 4.5 Cell-free-produced vaccine proteins protect mice from tumor challenge better than VL -VH -IL1β9aa and comparably to the Id-KLH vaccine. (a) First tumor challenge experiment (n = 10 mice per group). (b) Second tumor challenge experiment (n = 10 mice per group).

rial strains [27]. Or rather, the GM-CSF may be specifically enhancing the immune response by stimulating dendritic cells. Additional work is underway to evaluate vaccine domains that contribute to efficacy and to identify a vaccine that is clinically and economically advantageous.

4.4 Conclusions and Projections

The cell-free system offers speed and cost advantages over in vivo systems for the production of NHL patient specific vaccines. First, the cell-free system successfully produces complex synthetic mammalian proteins when the reaction environment is modified to promote oxidative folding. Second, we have successfully scaled-up vaccine synthesis to rapidly and efficiently produce active vaccines in amounts required for the mouse tumor protection studies. Together, these achievements allow complex fusion protein vaccines to be produced in approximately one-tenth of the time and with considerable cost savings relative to the use of mammalian cells. These advances expand the utility of the cell-free system and demonstrate exciting new possibilities for the field of patient specific medicine. The advantages of cell-free system protein expression, such as short expression time, easily multiplexed production, and simpler downstream processing, are ideally suited for protein-based personalized medicine applications. In the near future we will investigate cell-free production of vaccines with human variable regions and will develop rapid assays for product characterization, both of which are required for the development of an efficient patient-specific production platform. Our goal is to make it economically feasible to provide every NHL patient with a personalized therapeutic vaccine.

References

Acknowledgments

A. R. Goerke is supported by a Doctoral Fellowship from Merck & Company, Inc. Special thanks to the Merck Biologics and Engineering Department for additional funding. We also gratefully acknowledge funding from the Stanford University Bio-X IAP Program.

References 1 Armitage, J. O., Treatment of nonHodgkin’s lymphoma. N. Engl. J. Med. 1993, 328, 1023–1030. 2 www.lymphoma.org. 3 Campbell, M. J., Carroll, W., Kon, S., Thielemans, K., Rothbard, J. B., Levy, S., Levy, R. Idiotype vaccination against murine B cell lymphoma. Humoral and cellular responses elicited by tumorderived immunoglobulin M and its molecular subunits. J. Immunol. 1987, 139, 2825–2833. 4 Timmerman, J. M., Levy, R. The history of the development of vaccines for the treatment of lymphoma. Clin. Lymphoma 2000, 1, 129–139; discussion 140. 5 Kaminski, M. S., Kitamura, K., Maloney, D. G., Levy, R. Idiotype vaccination against murine B cell lymphoma. Inhibition of tumor immunity by free idiotype protein. J. Immunol. 1987, 138, 1289–1296. 6 Tao, M. H., Levy, R. Idiotype/granulocytemacrophage colony-stimulating factor fusion protein as a vaccine for B-cell lymphoma. Nature 1993, 362, 755–758. 7 Chen, T. T., Tao, M. H., Levy, R. Idiotypecytokine fusion proteins as cancer vaccines. Relative efficacy of IL-2, IL-4, and granulocyte-macrophage colonystimulating factor. J. Immunol. 1994, 153, 4775–4787. 8 Syrengelas, A. D., Levy, R. DNA vaccination against the idiotype of a murine B cell lymphoma: mechanism of tumor protection. J. Immunol. 1999, 162, 4790–4795. 9 Hakim, I., Levy, S., Levy, R. A nine-amino acid peptide from IL-1beta augments antitumor immune responses induced by protein and DNA vaccines. J. Immunol. 1996, 157, 5503–5511. 10 McCormick, A. A., Kumagai, M. H., Hanley, K., Turpen, T. H., Hakim, I., Grill,

11

12

13

14

15

16

17

L. K., Tuse, D., Levy, S., Levy, R. Rapid production of specific vaccines for lymphoma by expression of the tumor-derived single-chain Fv epitopes in tobacco plants. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 703– 708. McCormick, A. A., Reinl, S. J., Cameron, T. I., Vojdani, F., Fronefield, M., Levy, R., Tuse, D. Individualized human scFv vaccines produced in plants: humoral anti-idiotype responses in vaccinated mice confirm relevance to the tumor Ig. J. Immunol. Methods 2003, 278, 95–104. Michel-Reydellet, N., Woodrow, K., Swartz, J. Increasing PCR fragment stability and protein yields in a cell-free system with genetically modified Escherichia coli extracts. J. Mol. Microbiol. Biotechnol. 2005, 9, 26–34. Yang, J., Kanter, G., Voloshin, A., Michel-Reydellet, N., Velkeen, H., Levy, R., Swartz, J. R. Rapid expression of vaccine proteins for B-cell lymphoma in a cell-free system. Biotechnol. Bioeng. 2005, 89, 503–511. Jewett, M. C., Swartz, J. R. Mimicking the Escherichia coli cytoplasmic environment activates long-lived and efficient cellfree protein synthesis. Biotechnol. Bioeng. 2004, 86, 19–26. Kim, D. M., Swartz, J. R. Efficient production of a bioactive, multiple disulfidebonded protein using modified extracts of Escherichia coli. Biotechnol. Bioeng. 2004, 85, 122–129. Yin, G., Swartz, J. R. Enhancing multiple disulfide bonded protein folding in a cellfree system. Biotechnol. Bioeng. 2004, 86, 188–195. Yang, J., Kanter, G., Voloshin, A., Levy, R., Swartz, J. R. Expression of active murine granulocyte-macrophage colony-

81

82

4 Cell-free Technology for Rapid Production

18

19

20

21

22

stimulating factor in an Escherichia coli cell-free system. Biotechnol. Prog. 2004, 20, 1689–1696. Zapun, A., Missiakas, D., Raina, S., Creighton, T. E. Structural and functional characterization of DsbC, a protein involved in disulfide bond formation in Escherichia coli. Biochemistry 1995, 34, 5075–5089. Missiakas, D., Schwanger, F., Raina, S. Identification and characterization of a new disulfide isomerase-like protein (DsbD) in Escherichia coli. EMBO J. 1995, 14, 3415–3424. Miyajima, A., Otsu, K., Schreurs, J., Bond, M. W., Abrams, J. S., Arai, K. Expression of murine and human granulocyte-macrophage colonystimulating factors in S. cerevisiae: mutagenesis of the potential glycosylation sites. EMBO J. 1986, 5, 1193–1197. Walter, M. R., Cook, W. J., Ealick, S. E., Nagabhushan, T. L., Trotta, P. P., Bugg, C. E. Three-dimensional structure of recombinant human granulocytemacrophage colony-stimulating factor. J. Mol. Biol. 1992, 224, 1075–1085. George, R. A., Heringa, J. An analysis of protein domain linkers: their classification

23

24

25

26

27

and role in protein folding. Protein Eng. 2002, 15, 871–879. Ferguson, N., Capaldi, A. P., James, R., Kleanthous, C., Radford, S. E. Rapid folding with and without populated intermediates in the homologous four-helix proteins Im7 and Im9. J. Mol. Biol. 1999, 286, 1597–1608. Kanter, G., Yang, J., Voloshin, A., Swartz, J. R., Levy, R. Cell-free production of scFv fusion proteins: An effective and efficient approach for custom lymphoma vaccines. Blood 2007, 109, 3393–3399. Voloshin, A. M., Swartz, J. R. Efficient and scalable method for scaling up cell free protein synthesis in batch mode. Biotechnol. Bioeng. 2005, 91, 516–521. Maloney, D. G., Kaminski, M. S., Burowski, D., Haimovich, J., Levy, R. Monoclonal anti-idiotype antibodies against the murine B cell lymphoma 38C13: characterization and use as probes for the biology of the tumor in vivo and in vitro. Hybridoma 1985, 4, 191–209. James, R., Kleanthous, C., Moore, G. R. The biology of E colicins: paradigms and paradoxes. Microbiology 1996, 142(Pt 7), 1569–1580.

83

5

Bacterial Cell-free System for Highly Efficient Protein Synthesis Takanori Kigawa, Takayoshi Matsuda, Takashi Yabuki, and Shigeyuki Yokoyama

5.1 Overview

We have been developing and using an E. coli cell extract-based coupled transcription–translation cell-free system to establish it as a protein expression method. Development includes many issues, such as cell extract preparation, template construction, reaction conditions, reaction format, and automation. This chapter describes our developments in the bacterial cell-free protein synthesis. These developments have improved the efficiency, productivity, and throughput of our cellfree system, enabling us to use it as one of the standard expression methods in our group. Our system has the most successful applications, especially for protein production for structure determination, among existing cell-free protein synthesis systems.

5.2 Introduction

In recent years, cell-free protein synthesis has widely become regarded as one of the useful protein expression methods. The problem of low productivity associated with cell-free synthesis has been solved by modification of the reaction conditions [1–5] and the development of continuous methods [6–10]. In addition to the conventional E. coli, rabbit, and wheat germ systems, new cell-free systems based on eukaryotic cell extracts have been developed [11–14]. As well as these developments, there is a growing need for protein expression methods optimized to a high-throughput format to prepare an extremely large number of protein samples efficiently and rapidly. One area of strong demand is in structural genomics and proteomics research, which aims to determine representative protein structures and/or functions [see International Structural Genomics Organization (ISGO), http://www.isgo.org] [15].

Cell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

84

5 Bacterial Cell-free System for Highly Efficient Protein Synthesis

We present here an E. coli cell extract-based coupled transcription–translation system, which we have developed to be efficient and productive, and routinely use for many kinds of applications.

5.3 Coupled Transcription–Translation System based on E. coli Extract

Our bacterial cell-free system is fundamentally based on the Zubay system [16], which uses the cell extract prepared from E. coli, where transcription from DNA to mRNA and translation from mRNA to protein are carried out in a coupled manner. The step to prepare mRNA from DNA is not always necessary, as a DNA molecule can be used as a template for coupled transcription–translation in the protein synthesis system. Because of this simplicity, the system may be applied even to largescale production. The E. coli coupled transcription–translation system had not been highly reproducible, partly because the endogenous RNA polymerase in the extract was unstable. We improved the reproducibility by using exogenous T7 RNA polymerase for transcription [17]. Other highly productive RNA polymerases such as SP6 phage polymerase may also be used for this purpose.

5.4 DNA Template Construction

Since T7 RNA polymerase is used for transcription in our system, as described above, the optimal DNA construct is almost compatible with the recombinant T7 expression system [18], also known as the pET system (Novagen, USA), which is one of the most popular protein expression methods using alive E. coli cells. Therefore, protein production can be tried by both in vivo and in vitro expression methods with the same DNA construct without re-cloning. We have developed a method for PCR to produce the expression construct, which can be used as the template suitable for the cell-free system (T. Yabuki, unpublished); we named the method “two-step PCR” (Fig. 5.1). First, the open reading frame (ORF) or a domain fragment of ORF is amplified by PCR using gene-specific primers and/or universal primer. Second, the first PCR product, a T7 promoter fragment with the tag-coding sequence, a T7 terminator fragment, and the universal primer (these are non-gene-specific and thus in common with all the genes) are subjected to overlapping PCR, and the construct that expresses a fusion protein under the control of the T7 promoter is produced. This method has an advantage that different tags can be attached simply by changing the common tag-coding fragment at the second PCR step with the same specific primer pair for each target. This method is so efficient and robust that we routinely use it in our highthroughput protein expression pipeline, as described in Chapter 6.

5.5 Preparation of Cell Extract from E. coli

Fig. 5.1 The “two-step PCR” method. First, the open reading frame

(ORF) or a domain fragment of ORF is amplified by PCR using gene-specific primers and/or universal primer. Second, the first PCR product, a T7 promoter fragment with the tag-coding sequence, a T7 terminator fragment, and the universal primer are subjected to overlapping PCR, and the construct that expresses a fusion protein under the control of the T7 promoter is produced.

5.5 Preparation of Cell Extract from E. coli

Cell extract preparation is one of the key steps to obtaining successful and reproducible results using the cell-free protein synthesis. We have made numerous efforts to develop the protocol for the extract preparation from E. coli cells. Our protocol previously described in detail [19] is fundamentally based on that of Pratt [20]; however, our modifications dramatically improved the reproducibility of the extract preparation and thus the productivity of the cell-free synthesis. One of the important modifications is the E. coli strain used for cell extract preparation. Previously, E. coli strains that lack the major RNase, such as the MRE600 and A19 strains, were used for cell extract preparation [16, 20]. Instead of these strains, we usually use BL21-derivative strains containing extra copies of the genes for minor tRNAs, such as BL21 codon-plus strains (Stratagene, USA), Rosetta strains (Novagen, USA), or originally developed strains [21]. The use of the extract prepared from the strain

85

86

5 Bacterial Cell-free System for Highly Efficient Protein Synthesis

Fig. 5.2 Batch-mode cell-free reaction. The reaction solution (30 µL) consisted of 58 mm HEPES-KOH buffer (pH 7.5) containing 1.8 mm DTT, 1.2 mm ATP, 0.8 mm each of CTP, GTP, UTP, 80 mm creatine phosphate, 0.25 mg mL–1 creatine kinase, 4.0% PEG 8000, 0.64 mm 3′ ,5′ -cyclic AMP, 68 mm l-(–)-5-formyl-5,6,7,8-tetrahydrofolic acid, 175 µg mL–1 E. coli total tRNA, 210 mm potassium glutamate, 27.5 mm ammonium

acetate, 10.7 mm magnesium acetate, 1.5 mm each of the 20 amino acids, 0.133 mg mL–1 T7 RNA polymerase, 7.2 µL S30 extract, and 4 µg mL–1 plasmid pK7-CAT [1] (•) or 4 µg mL–1 linear DNA fragment amplified by PCR from pK7-CAT plasmid (). The reaction mixture was incubated at 37 ◦ C for 1 hour. CAT productivity was calculated from CAT activity of the reaction as previously described [19].

overproducing minor tRNAs circumvents the low productivity due to rare codons [21]. We have already proposed that a BL21-derivative strain lacking RNase could be one of the best strains for the extract source [19], and mRNA stability was increased by using the extract prepared from the BL21-Star (DE3) strain (Invitrogen, USA) which is deficient in the major RNase (RNase E) [22]. The cell disruption method is another important modification. A FRENCH press (Thermo Electron, USA) is a popular device to disrupt the E. coli cells for the extract preparation; however, in our experience it is difficult to obtain reproducible cell disruption results on a laboratory-scale preparation by using a FRENCH press. Instead, we use disruption with small glass beads with a MSK cell homogenizer (B. Braun, Germany) and the Multi-beads Shocker (Yasui Kikai, Japan), which is usually used to disrupt cells with rigid cell walls, such as small algae and yeast. During the last five years, over 18 L of cell extract, which corresponds to more than 60 L of the cell-free reaction mixture, have been prepared by only one or two staff members, and almost all of the extract has reproducible and sufficient activity for cell-free synthesis.

5.6 Batch-mode Cell-free Reaction

5.6 Batch-mode Cell-free Reaction

Conventional batch-mode cell-free reaction is quite convenient for quick protein production because it can be completed in 1–2 hours, and the productivity reaches 600–800 µg of protein per 1 mL reaction mixture in the best case, and, in addition, PCR-amplified linear DNA fragment is utilized as a template for protein synthesis without any cloning procedures (Fig. 5.2). We have established a high-throughput automated protein-expression method by combining the batch-mode cell-free protein synthesis with PCR, which is performed on multi-well plates (Fig. 5.3) and is

Fig. 5.3 SDS-PAGE analysis of MBP-fused

Arabidopsis proteins. Template DNAs were constructed by the two-step PCR method as MBP-fused form from 96 RAFL Arabidopsis thaliana cDNA clones [25]. Protein products encoded on cDNAs were produced by batch-mode cell-free reaction as described in Fig. 5.2. For ten clones, template DNAs were not properly generated mainly due to the inconsistency of the sequence information.

For 74 out of 86 clones, excluding the previous 10 clones, the product could be identified on the gel. Two lanes are paired for each cDNA (Left: total fraction, Right: soluble fraction). A portion (3 µL) of the reaction (30 µL) was analyzed on 12.5% Perfect NT gel (DRC, Japan) and then stained with Coomassie Brilliant Blue. Black dot indicates the position of the product.

87

88

5 Bacterial Cell-free System for Highly Efficient Protein Synthesis

thus adapted for robotics [23, 24]. This method is widely applicable not only to the screening of protein constructs suitable for the structural analysis but also to the functional analysis of proteins in a genome scale.

5.7 Dialysis-mode Cell-free Reaction

The continuous system [6] is a major technological breakthrough in cell-free synthesis and is quite suitable for protein production. In the past, we focused on a continuous flow cell-free (CFCF) system [27]. For the last decade, however, we have been developing and using the continuous exchange cell-free (CECF) system, i.e., the dialysis-mode cell-free system, mainly for structural biology and structural

Fig. 5.4 Small-scale dialysis-mode cell-free system. (A) Schematic diagram of small-scale dialysis-mode cell-free system. (B) A photograph of the system.

5.7 Dialysis-mode Cell-free Reaction

proteomics. Several formats of the dialysis-mode cell-free system, including smallscale and large-scale formats, have been developed. The small-scale dialysis system [28] is composed of a Slide-A-Lyzer MINI Dialysis Unit with a molecular weight cut off (MWCO) of 10 kDa (PIERCE, USA) and Cryogenic Vial (NALGEN, USA) as a reservoir and, in a typical case, 30 µL of the internal solution is dialyzed against 300 µL of the external solution (Fig. 5.4). The external solution is gently mixed by a small stirring bar during incubation, and the reaction is usually carried out at 30 ◦ C for 6–16 hours. This system is mainly used as a protein production method in microgram quantities for biochemical and biological studies. A large-scale system for protein production in milligram quantities is much simpler (Fig. 5.5). The internal solution in a dialysis tube (Spectra/Por 7, MWCO: 15 kDa, Spectrum, USA) is dialyzed against the external solution with gentle shaking, and the reaction is performed at 30 ◦ C for several hours to overnight.

Fig. 5.5 Large-scale dialysis-mode cell-free system. (A) Schematic

diagram of large-scale dialysis-mode cell-free system. (B) A photograph of such a system (3 mL internal/30 mL external solutions). The container lid is closed during incubation to avoid evaporation of the external solution.

89

90

5 Bacterial Cell-free System for Highly Efficient Protein Synthesis

Fig. 5.6 CAT protein production by the dialysis-mode cell-free system with the plasmid template and the PCR-amplified linear template. The internal solution consisted of 58 mm HEPES-KOH buffer (pH 7.5) containing 1.8 mm DTT, 1.2 mm ATP, 0.8 mm each of CTP, GTP, UTP, 80 mm creatine phosphate, 0.25 mg mL–1 creatine kinase, 4.0% PEG 8000, 0.64 mm 3′ ,5′ -cyclic AMP, 68 mm l-(–)-5-formyl-5,6,7,8-tetrahydrofolic acid, 175 µg mL–1 E. coli total tRNA, 210 mm potassium glutamate, 27.5 mm ammonium acetate, 10.7 mm magnesium acetate, 1.5 mm each of the 20 amino acids, 0.05% sodium

azide, 30% S30 extract, 0.066 mg mL–1 T7 RNA polymerase, and either 1 µg mL–1 plasmid DNA (•) or 4 µg mL–1 PCR-amplified DNA (). The external solution, the volume of which is 10× larger than that of the internal solution, contained the components of the internal solution except for the creatine kinase, template DNA, T7 RNA polymerase, and S30 extract. The reaction was performed at 30 ◦ C for 10 hours. (a) Translation kinetics; (b) SDS-PAGE analysis. Total (T) and soluble (S) fractions corresponding to 0.6 µL of the internal solution were loaded. No template DNA reaction is marked with a dash (–).

5.8 Template DNA

Fig. 5.7 SDS-PAGE analysis of the product of

small-scale cell-free protein synthesis. The reaction conditions were the same as those in Fig. 5.6. Samples (0.3 µL each) of the total fraction (T) or the soluble fraction (S) were analyzed. 1. The SH3 domain of human Rho guanine exchange factor 16 (PDB ID: 1X6B); 2. the DNA-binding domain of human transcriptional adaptor 2-like protein (PDB ID: 1X41); 3. the eRF1 domain of human pelota gene product (PDB ID: 1X52); 4. a potential DNA-binding domain of human endothelial

differentiation-related factor 1 (PDB ID: 1X57); 5. the NUDIX domain of human A/G-specific adenine DNA glycosylase (PDB ID: 1X51); 6. the sterile alpha motif (SAM)/Pointed domain of human Friend leukemia integration 1 transcription factor (PDB ID: 1X66); 7. the actin depolymerization factor/cofilin-like domain of human Drebrin-like protein (PDB ID: 1X67); 8. the PH domain of human protein kinase D2 (PDB ID: 2COA); and 9. the HMG box-like domain of human homolog of Maelstrom (PDB ID: 2CTO).

Productivity reaches over 7-mg protein per mL of reaction mixture, after 10 hours of incubation, for CAT protein (Fig. 5.6). Figure 5.7 shows the production of several human proteins whose solution structures have been deposited to the PDB; in most cases, produced proteins give a thick main band on the SDS-PGAE gel. The reaction scale is from 3 mL internal/30 mL external format to 9 mL internal/90 mL external format with almost the same productivity per reaction mixture volume if the proportion of the surface area of the dialysis tube to the reaction volume is kept at optimum (Section 5.10). Thanks to the high productivity of our dialysis system and the property of the cell extract we prepared, PCR-amplified linear DNA fragments can be used as a template even in the dialysis system though the productivity is approximately 70% of that with the plasmid template (Fig. 5.6). As described in Chapter 6, we have prepared more than 21 000 protein samples using the dialysis system with the PCRamplified linear DNA template, and then successfully measured NMR spectra.

5.8 Template DNA

As described previously [19], the quality of the template DNA is important for successful cell-free protein synthesis (Fig. 5.8). Usually, plasmid DNA purified by commercially available kits (Qiagen, Promega etc.) has acceptable quality with lit-

91

92

5 Bacterial Cell-free System for Highly Efficient Protein Synthesis

Fig. 5.8 Effects of the template DNA quality and quantity on productivity. The CAT protein was produced by the small-scale dialysis-mode cell-free system (30 µL internal/300 µL external solutions) with pK7-CAT plasmid purified using a Wizard Plus Midiprep kit (Promega, USA) according to the vendor protocol (◦) or that with an additional pre-wash treatment by 40% 1-propanol/4.2 m guanidine-HCl (•).

tle RNase contamination if the plasmid is washed with the pre-wash buffer (40% 1-propanol/4.2 m guanidine-HCl) just after it is loaded on the column. Figure 5.8 also indicates that the productivity is strongly influenced by the template DNA concentration, suggesting that the optimum concentration should be examined for each protein to obtain the best results.

5.9 Reaction Temperature

We usually perform the bacterial cell-free synthesis reaction at 37 ◦ C for the batchmode reaction and at 30 ◦ C for the dialysis-mode reaction. Although the productivity is reduced to some extent, the bacterial cell-free reaction can be carried out at lower temperatures. Figure 5.9 shows the effect of the incubation temperature on the productivity; about a half of the productivity is achieved even at 24 ◦ C, suggesting that some of the protein that is less stable at higher temperature is expected to be produced properly at the lower temperatures.

5.10 Surface Area of the Dialysis Membrane

Fig. 5.9 Bacterial cell-free system at various temperatures. The CAT

protein was produced by the small-scale dialysis-mode cell-free system (30 µL internal/300 µL external solutions) at 24 (), 27 (), and 30 (•) ◦ C.

5.10 Surface Area of the Dialysis Membrane

As mentioned above, the proportion of the surface area of the dialysis tube to the reaction volume is important for the success of the dialysis reaction. Figure 5.10 shows how the protein synthesis is affected by the surface area of dialysis membrane, indicating that 1.0 mm2 of membrane surface area per µL of reaction is the optimal proportion for obtaining the maximum protein synthesis rate.

5.11 Stable-isotope Labeling for NMR Spectroscopy

We have reported that the cell-free system is extremely useful for three kinds of stable-isotope labeling of proteins: amino-acid selective [17, 2, 9], uniform [9], and site-directed [9, 26, 35, 31] labeling. By simply replacing the amino acid(s) of interest in the cell-free reaction mixture with labeled one(s), amino acid-selective or uniform labeling can be achieved. We have recently developed a new system that uses D-glutamate in place of L-glutamate for efficient stable-isotope labeling [29]. The stereo-array isotope labeling (SAIL) method [32], which is now expected to

93

94

5 Bacterial Cell-free System for Highly Efficient Protein Synthesis

Fig. 5.10 Effect of the surface area of the dialysis membrane on the cell-free protein synthesis. The CAT protein was produced by the small-scale dialysis-mode cell-free system at 30 ◦ C using the same unit described above with various reaction volumes.

Internal/external solutions (µL/µL) are 20/200 (•), 30/300 (), 50/500 (), 70/700 (◦), and 90/900 (open triangle), which corresponds to ratios of membrane surface area per reaction mixture (mm2 µL–1 ) of 1.41, 0.94, 0.57, 0.40, and 0.31, respectively.

expand the molecular size limit of NMR spectroscopy, is one application of “uniform” stable-isotope labeling using the cell-free system. Instead of expensive purified labeled amino acids, a relatively inexpensive stable-isotope labeled algal amino acid mixture can be used [9]. In this case, labeled cysteine, tryptophan, asparagine, and glutamine are still required for uniform labeling, because most of the commercially available algal amino acid mixture lacks these four amino acids. Some vendors are now selling an amino acid mixture that consists of algal amino acid mixture supplemented with cysteine, tryptophan, asparagine, and glutamine. We have prepared over 1500 uniformly [13 C, 15 N]-labeled samples, indicating that our system is readily applicable to stable-isotope labeling of proteins.

5.12 Selenomethionine Incorporation for X-Ray Crystallography

We have also reported that the cell-free system is useful for preparing selenomethionine substituted proteins for X-ray crystallography [35], and application to the structure determination (e.g., [33, 34]). Selenomethionine can be incorporated into protein by simply replacing methionine with selenomethionine; in this way over 95% of the methionine residues were substituted with selenomethionine [35].

5.13 Automation

Therefore, the cell-free synthesis system could become a powerful protein expression method for high-throughput structure determinations by X-ray crystallography.

5.13 Automation

We have originally developed fully automated cell-free protein synthesis systems in various formats. One such system can perform the two-step PCR reaction for template generation, the batch mode cell-free protein synthesis reaction (30 µL scale), and, optionally, the fluorescence measurement of GFP-fused product for 768 samples at the same time within 9 hours. Another system that integrates the dialysis mode cell-free system (1 mL of internal solution) and the affinity purification can prepare 96 kinds of partially purified proteins in milligram quantities within a half day. These automated systems are now fully operated in our group and are a strong driving force for our structural genomics and proteomics project. The application of these systems to our high-throughput protein expression pipeline are described in the next chapter.

5.14 Conclusion

Cell-free protein synthesis is now an extremely powerful protein expression method, especially for high-throughput applications. As described, our E. coli cellfree system is quite simple and productive due to our developments. The system is now one of the routinely used expression methods in our group and is used for many of applications, including high-throughput protein expression pipeline for our mammalian structural genomics and proteomics project in the RIKEN Structural Genomics/Proteomics Initiative (RSGI; http://www.rsgi.riken.jp) [23]. So far, by using the large-scale system, we have prepared about 2500 samples for structure determination by NMR spectroscopy or X-ray crystallography. Therefore, among the cell-free protein synthesis systems, our system has by far the largest number of successful applications, especially to protein production for structure determination.

References 1 Kim, D.-M., Kigawa, T., Choi, C.-Y., Yokoyama, S. (1996) A highly efficient cell-free protein synthesis system from Escherichia coli. Eur. J. Biochem. 239: 881– 886.

2 Yabuki, T., Kigawa, T., Dohmae, N., Takio, K., Terada, T., Ito, Y., Laue, E. D., Cooper, J. A., Kainosho, M., Yokoyama, S. (1998) Dual amino acid-selective and site-directed stable-isotope labeling of the human c-

95

96

5 Bacterial Cell-free System for Highly Efficient Protein Synthesis

3

4

5

6

7 8

9

10

11

12

13

Ha-Ras protein by cell-free synthesis. J. Biomol. NMR 11: 295–306. Kim, D. M., Swartz, J. R. (1999) Prolonging cell-free protein synthesis with a novel ATP regeneration system. Biotechnol. Bioeng. 66: 180–188. Madin, K., Sawasaki, T., Ogasawara, T., Endo, Y. (2000) A highly efficient and robust cell-free protein synthesis system prepared from wheat embryos: plants apparently contain a suicide system directed at ribosomes. Proc. Natl. Acad. Sci. U.S.A. 97: 559–564. Kim, T. W., Kim, D. M., Choi, C. Y. (2006) Rapid production of milligram quantities of proteins in a batch cell-free protein synthesis system. J. Biotechnol. 24: 373–380. Spirin, A. S., Baranov, V. I., Ryabova, L. A., Ovodov, S. Y., Alakhov, Y. B. (1988) A continuous cell-free translation system capable of producing polypeptides in high yield. Science 242: 1162–1164. Davis, J., Thompson, D., Beckler, G. S. (1996) Promega Notes Mag. 54: 14–18. Kim, D. M., Choi, C. Y. (1996) A semicontinuous prokaryotic coupled transcription/translation system using a dialysis membrane. Biotechnol. Prog. 12: 645–649. Kigawa, T., Yabuki, T., Yoshida, Y., Tsutsui, M., Ito, Y., Shibata, T., Yokoyama, S. (1999) Cell-free production and stableisotope labeling of milligram quantities of proteins. FEBS Lett. 442: 15–19. Sawasaki, T., Hasegawa, Y., Tsuchimochi, M., Kamura, N., Ogasawara, T., Kuroita, T., Endo, Y. (2002) A bilayer cell-free protein synthesis system for high-throughput screening of gene products. FEBS Lett. 514: 102–105. Tarui, H., Imanishi, S., Hara, T. (2000) A novel cell-free translation/glycosylation system prepared from insect cells. J. Biosci. Bioeng. 90: 508–514. Katzen, F., Kudlicki, W. (2006) Efficient generation of insect-based cell-free translation extracts active in glycosylation and signal sequence processing. J. Biotechnol. 125: 194–197. Mikami, S., Masutani, M., Sonenberg, N., Yokoyama, S., Imataka, H. (2006) An efficient mammalian cell-free translation system supplemented with translation factors. Protein Expr. Purif. 46: 348–357.

14 Wakiyama, M., Kaitsu, Y., Yokoyama, S. (2006) Cell-free translation system from Drosophila S2 cells that recapitulates RNAi. Biochem. Biophys. Res. Commun. 343: 1067–1071. 15 Stevens, R. C., Yokoyama, S., Wilson, I. A. (2001) Global efforts in structural genomics. Science 294: 89–92. 16 Zubay, G. (1973) In vitro synthesis of protein in microbial systems. Annu. Rev. Genet. 7: 267–287. 17 Kigawa, T., Muto, Y., Yokoyama, S. (1995) Cell-free synthesis and amino acidselective stable isotope labeling of proteins for NMR analysis. J. Biomol. NMR 6: 129–134. 18 Studier, F. W., Moffatt, B. A. (1986) Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J. Mol. Biol. 189: 113–130. 19 Kigawa, T., Yabuki, T., Matsuda, N., Matsuda, T., Nakajima, R., Tanaka, A., Yokoyama, S. (2004) Preparation of Escherichia coli cell extract for highly productive cell-free protein expression. J. Struct. Funct. Genomics 5: 63–68. 20 Pratt, J. M. (1984) Coupled transcriptiontranslation in prokaryotic cell-free system. In: Hames, B. D., Higgins, S. J. (eds) Coupled Transcription-Translation in Prokaryotic Cell-Free System, pp. 179–209, Oxford: IRL Press. 21 Chumpolkulwong, N., Sakamoto, K., Hayashi, A., Iraha, F., Shinya, N., Matsuda, N., Kiga, D., Urushibata, A., Shirouzu, M., Oki, K., Kigawa, T., Yokoyama, S. (2006) Translation of ‘rare’ codons in a cell-free protein synthesis system from Escherichia coli. J. Struct. Funct. Genomics. 22 Ahn, J. H., Chu, H. S., Kim, T. W., Oh, I. S., Choi, C. Y., Hahn, G. H., Park, C. G., Kim, D. M. (2005) Cell-free synthesis of recombinant proteins from PCR-amplified genes at a comparable productivity to that of plasmid-based reactions. Biochem. Biophys. Res. Commun. 338: 1346–1352. 23 Yokoyama, S., Hirota, H., Kigawa, T., Yabuki, T., Shirouzu, M., Terada, T., Ito, Y., Matsuo, Y., Kuroda, Y., Nishimura, Y., Kyogoku, Y., Miki, K., Masui, R., Kuramitsu, S. (2000) Structural genomics projects in Japan. Nat. Struct. Biol. 7(Suppl): 943–945.

References 24 Yokoyama, S. (2003) Protein expression systems for structural genomics and proteomics. Curr. Opin. Chem. Biol. 7: 39–43. 25 Seki, M., Narusaka, M., Kamiya, A., Ishida, J., Satou, M., Sakurai, T., Nakajima, M., Enju, A., Akiyama, K., Oono, Y., Muramatsu, M., Hayashizaki, Y., Kawai, J., Carninci, P., Itoh, M., Ishii, Y., Arakawa, T., Shibata, K., Shinagawa, A., Shinozaki, K. (2002) Functional annotation of a full-length Arabidopsis cDNA collection. Science 296: 141–145. 26 Hirao, I., Ohtsuki, T., Fujiwara, T., Mitsui, T., Yokogawa, T., Okuni, T., Nakayama, H., Takio, K., Yabuki, T., Kigawa, T., Kodama, K., Nishikawa, K., Yokoyama, S. (2002) An unnatural base pair for incorporating amino acid analogs into proteins. Nat. Biotechnol. 20: 177–182. 27 Kigawa, T., Yokoyama, S. (1991) A continuous cell-free protein synthesis system for coupled transcription-translation. J. Biochem. (Tokyo) 110: 166–168. 28 Matsuda, T., Kigawa, T., Koshiba, S., Inoue, M., Aoki, M., Yamasaki, K., Seki, M., Shinozaki, K., Yokoyama, S. (2006) Cell-free synthesis of zinc-binding protein. J. Struct. Funct. Genomics 7: 93–100. 29 Matsuda, T., Koshiba, S., Tochio, N., Seki, E., Iwasaki, N., Yabuki, T., Inoue, M., Yokoyama, S., Kigawa, T. (2007) Improving cell-free protein synthesis for stable-isotope labeling. J. Biomol. NMR 37: 225–229. 30 Kigawa, T., Yamaguchi-Nunokawa, E., Kodama, K., Matsuda, T., Yabuki, T., Matsuda, N., Ishitani, R., Nureki, O., Yokoyama, S. (2002) Selenomethionine incorporation into a protein by cell-free synthesis. J. Struct. Funct. Genomics 2: 29–35.

31 Kodama, K., Fukuzawa, S., Nakayama, H., Kigawa, T., Sakamoto, K., Yabuki, T., Matsuda, N., Shirouzu, M., Takio, K., Tachibana, K., Yokoyama, S. (2006) Regioselective carbon-carbon bond formation in proteins with palladium catalysis; new protein chemistry by organometallic chemistry. Chembiochem 7: 134–139. 32 Kainosho, M., Torizawa, T., Iwashita, Y., Terauchi, T., Mei Ono, A., Guntert, P. (2006) Optimal isotope labelling for NMR protein structure determinations. Nature 440: 52–57. 33 Wada, T., Shirouzu, M., Terada, T., Ishizuka, Y., Matsuda, T., Kigawa, T., Kuramitsu, S., Park, S. Y., Tame, J. R., Yokoyama, S. (2003) Structure of a conserved CoA-binding protein synthesized by a cell-free system. Acta Crystallogr. D Biol. Crystallogr. 59: 1213–1218. 34 Arai, R., Kukimoto-Niino, M., UdaTochio, H., Morita, S., Uchikubo-Kamo, T., Akasaka, R., Etou, Y., Hayashizaki, Y., Kigawa, T., Terada, T., Shirouzu, M., Yokoyama, S. (2005) Crystal structure of an enhancer of rudimentary homolog (ERH) at 2.1 Å resolution. Protein Sci. 4: 888–893. 35 Kiga, D., Sakamoto, K., Kodama, K., Kigawa, T., Matsuda, T., Yabuki, T., Shirouzu, M., Harada, Y., Nakayama, H., Takio, K., Hasegawa, Y., Endo, Y., Hirao, I., Yokoyama, S. (2002) An engineered Escherichia coli tyrosyl-tRNA synthetase for site-specific incorporation of an unnatural amino acid into proteins in eukaryotic translation and its application in a wheat germ cell-free system. Proc. Natl. Acad. Sci. U.S.A. 99: 9715–9720.

97

99

6

The Use of the Escherichia coli Cell-free Protein Synthesis for Structural Biology and Structural Proteomics Takanori Kigawa, Makoto Inoue, Masaaki Aoki, Takayoshi Matsuda, Takashi Yabuki, Eiko Seki, Takushi Harada, Satoru Watanabe, and Shigeyuki Yokoyama

6.1 Overview

The RIKEN Structural Genomics/Proteomics Initiative (RSGI) aims to study both structures and molecular functions of protein families. Within this framework, the RIKEN Genomic Sciences Center is mainly targeting mouse, human, and Arabidopsis thaliana, from the viewpoint of domain families involved in particular biological systems such as cell signaling, nucleic acid binding, and human diseases. Domain(s) belonging to selected families are further screened in terms of suitability for structure determination. The screening features the E. coli cell-free protein synthesis system, which is more suitable for efficient expression of a large number of constructs than in vivo protein expression systems. In the initial stage of screening the domain fragments are synthesized by small-scale cell-free reactions. Several different expression constructs are tested for the productivity and the solubility of the product, and good constructs are selected. In the next stage, milligram quantity of 15 N-labeled samples are produced in parallel using a dialysis cell-free system, and are then partially purified by affinity chromatography. This procedure is fully automated and 96 samples in sub-milligram to milligram quantity can be prepared within 12 hours. Throughout the screening stages, the protein samples were prepared directly from PCR-amplified linear DNA fragments without any cloning procedures. In the third stage, 1 H-15 N HSQC spectra are measured to investigate whether selected domains are properly folded and, thus, suitable for structure determination. The winners in the third screening stage are subjected to larger-scale cell-free production of the uniformly 13 C/15 N-labeled samples for NMR spectroscopy, and are then subject to structure determination. Those samples judged to be more suitable for X-ray crystallography are prepared in the selenomethionine-substituted form for structure determination by MAD phasing. We have determined about 1300 structures by NMR spectroscopy according to this workflow, indicating that our cell-free protein synthesis is quite suitable for largescale protein preparation for structural biology and structural proteomics.

Cell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

100

6 The Use of the Escherichia coli Cell-free Protein Synthesis

6.2 Introduction

Structural genomics and proteomics (SG) are genome/proteome-scale projects of structural biology [1]. The International Structural Genomics Organization (ISGO, http://www.isgo.org) was organized by over 25 projects from countries such as Canada, China, Japan, Korea, France, Germany, the UK, and the USA. Vast numbers of proteins from many kind of organisms have been selected as targets for structure determination ([2], see also Target DB on the PDB site, http://targetdb.pdb.org). Protein production is one of the most important steps in structural biology. This process should be scaled up for parallel production of hundreds or thousands of targets for an SG project. Prokaryotic expression with an E. coli host is most frequently used in SG projects, although eukaryotic hosts, such as the yeasts Pichia pastoris and S. cerevisiae, are also used. High-throughput methods for cloning and expression have been developed to achieve a high success rate in protein production [3–6]. Cell-free protein synthesis can be used for large-scale protein production [7]. The productivity of the E. coli coupled transcription–translation system has been improved [8] to be more than 7 mg per mL of reaction mixture (see Chapter 5). This method has been used successfully for stable-isotope labeling of proteins for NMR spectroscopy [9, 10, 8] and selenomethionine incorporation into protein for X-ray crystallography [11]. Furthermore, the cell-free system could produce proteins directly from PCR-amplified linear DNA fragments, without any cloning procedures. All of the reactions can be carried out in a multi-well plate format, which could be easily adapted to automated procedures. Thus, it has been widely noted as a promising and suitable protein expression method for SG projects [12–15]. The RIKEN Structural Genomics/Proteomics Initiative (RSGI), in which the authors are involved, is the largest SG project in Japan and aims to study both structures and molecular functions of protein families. In this chapter we describe the use of the E. coli cell-free protein synthesis for high-throughput expression pipeline (Fig. 6.1) in the RSGI. The system is now routinely used for the expression screening of mouse, human, and Arabidopsis proteins and protein domains, as well as large-scale production such as stable isotope labeling for NMR structure determination and selenomethionine incorporation for X-ray crystallography.

6.3 High-throughput Expression by PCR-based Small-scale Cell-free Protein Synthesis for Initial Screening

As we describe in Chapter 5, we have developed a method for PCR to produce the expression construct that can be used as the template for protein production (“twostep PCR”). First, the open reading frame (ORF) or a domain fragment of ORF is amplified by PCR using gene-specific primers and/or universal primer. Second,

6.3 High-throughput Expression by PCR-based Small-scale Cell-free Protein Synthesis

Fig. 6.1 A high-throughput expression pipeline based on cell-free protein synthesis.

the first PCR product, a T7 promoter fragment with the tag-coding sequence, a T7 terminator fragment, and the universal primer are subjected to overlapping PCR to obtain the construct that expresses a fusion protein under the control of the T7 promoter. With the second PCR product as the template, coupled transcription–translation reactions for cell-free protein synthesis in batch mode is carried out. The N- or C-terminal tagged forms (mostly the N-terminal His-tagged form) of proteins and

101

102

6 The Use of the Escherichia coli Cell-free Protein Synthesis

protein domains encoded by the mouse, human, and Arabidopsis cDNAs are produced by this method, and well-produced and soluble proteins are selected as candidates for structural analysis. The advantages of this method are: (a) Simple protocol: All reactions, including PCR and cell-free protein synthesis, can be carried out in multi-well plate format, and are thus suited to robotics. (b) Rapid expression: All reactions from PCR to protein synthesis are completed in one day. (c) Productivity: 10–800 µg of proteins can be produced per mL of cell-free protein synthesis reaction. (d) No necessity of cloning: Template DNA construction is completed by PCR without time-consuming ligation or transformation procedure. (e) Variety in tags: Just by changing the tag-coding fragment at the second PCR step, different tags can be attached. This method is widely applicable not only to the screening of protein constructs suitable for structural analysis, but also to the functional analysis of proteins on a genome scale. We usually generate several expression constructs with moderately different terminals for one target protein/protein domain (multi-construct strategy). Typically, a combination of two to four variations at intervals of several to ten residues for both N- and C-termini, which makes 4–16 variations in the terminal residues, are constructed and examined with respect to the productivity and solubility. The C-terminal GFP-fusion construct [16] is used to investigate both the productivity and solubility of the products on the basis of the fluorescence as a measure of the amount of soluble protein. We are therefore able to simply and efficiently find the best constructs, which exhibit the strongest fluorescence. Usually, a maximum of four best constructs are selected for the next stage, even if all of the tested constructs exhibit strong fluorescence. With the SH2 domain from human tyrosineprotein kinase BTK, for example, we have generated 16 different constructs, with residues 250, 260, 270, and 280 for the N-terminus each with residues 361, 371, 381, and 391 for the C-terminus (Fig. 6.2). Interestingly, soluble products with adequate fluorescence were obtained for the constructs with residues 250, 260, and 270 at the N-terminus and residues 381 and 391 at the C-terminus, whereas residues 280 and 361 are the N- and C-terminal domain boundaries, respectively, as predicted by the Pfam database [17]. Thus, the two-step PCR reaction, the batch mode cell-free protein synthesis reaction (20–30 µL), and the fluorescence measurement of GFP-fused products compose the first stage of the initial screening. Experiments at this first stage are now fully automated with the integrated machine “The CellFree eXpress” (Fig. 6.3A), which we originally developed (Aoki, unpublished), and are completed in 9 hours for 768 (8 × 96 wells) samples. At the second stage of the initial screening, the selected constructs are further subjected to a small-scale dialysis-mode cell-free protein synthesis reaction (30 µL internal/300 µL external), which is also carried out in the 96-well format. The solubility and productivity are assessed by SDS-PAGE analysis with the native, nonGFP-fusion construct. Linear DNA fragments generated by the two-step PCR are still used as the template at this stage, so that the construct conversion from the GFP-fusion into the native form is quite easy and simple. By the end of March 2006 we had completed the first stage of the initial screening for 288 445 constructs of 31 126 proteins/domains, mostly from higher eukaryotes,

6.4 Fully Automated Protein Production using Middle-scale Cell-free Protein Synthesis

Fig. 6.2 Fluorescence of the GFP-fusion SH2 domain constructs.

Fig. 6.3 Fully automated cell-free protein synthesis systems. (A) “The

Cell-Free eXpress”, which can perform the two-step PCR reaction, the batch mode cell-free protein synthesis reaction, and the fluorescence measurement of GFP-fused products of 768 samples in parallel. (B) A system assembled based on TECAN Freedom EVO 200, which can perform middle-scale dialysis-mode cell-free synthesis, and successive affinity purification for 96 samples in parallel.

mouse, human, and Arabidopsis thaliana, and then selected 39 866 constructs of 11 875 proteins/domains for the next, second stage. At the second stage, 23 229 constructs of 7609 proteins/domains have passed to the next step.

6.4 Fully Automated Protein Production using Middle-scale Cell-free Protein Synthesis

For the protein selected at the previous stage, 15 N-labeled samples are produced, using middle-scale dialysis-mode cell-free synthesis (1 mL internal/10 mL exter-

103

104

6 The Use of the Escherichia coli Cell-free Protein Synthesis

Fig. 6.4 Representative 1 H-15 N HSQC spectra in the HSQC

screening. Spectra of samples categorized into “A”, “B”, “C”, and “X” are presented (see text for the definition of A–C and X).

nal), and are then partially purified by affinity chromatography (Fig. 6.4). This process is also fully automated with the system that we assembled using TECAN Freedom EVO 200 (Fig. 6.3B, Aoki, unpublished), and up to 96 samples can be prepared in parallel within a half day. The final productivity, which varies from 0.2 to 3 mg depending on the target, is sufficient, in most cases, for the NMR experiment at the later stage, even when PCR-amplified liner DNA fragments are used as the templates.

6.5 NMR Screening

As an NMR spectrum gives the most direct and reliable information about the property of the sample in respect of NMR, we measured the 1 H-15 N HSQC spectra of 15 N-labeled proteins prepared at the previous stage, and judge whether the sample is suitable for structure determination. The spectra are collected on an AVANCE 600 spectrometer equipped with a triple-resonance CryoProbe and an

6.6 Large-scale Protein Production for Structure Determination Table 6.1 Statistics of NMR screening as of the end of March, 2007

(see text for the definition of ranks A–C and X). Rank

Construct

Protein/protein domain

A B C X Total

8289 6112 5515 1871 21 787

2310 2212 2000 1015 7537

automatic sample changer (Bruker Biospin), where data collection is performed automatically. We also use analytical gel filtration chromatography to investigate the apparent molecular mass of the sample, which suggests that the sample behaves as a monomer, oligomer, or their equilibrium. On the bases of the separation and line shape of the HSQC cross peaks, the samples are classified into four categories: “A” (suitable for structure determination), “B” (folded, but exhibits severe signal overlap and/or line broadening), “C” (not folded), and “X” (weak signal) (Fig. 6.4). The structure of most “A” samples is expected to be completed within a few months. For better behavior of proteins in the “B” and “C” categories, we usually try several approaches, such as fine-tuning the domain boundary and optimizing the solvent conditions. As of the end of March, 2007, 21 787 constructs of 7537 proteins/domains mainly from mouse, human, and Arabidopsis thaliana were subjected to “NMR screening” (Table 6.1), and 2310 protein/protein domains of them were judged as suitable samples for NMR analysis (Rank A). Interestingly, 2000 proteins/domains are in the “not folded” state – probably because they need some additional partner such as protein, nucleic acid or small ligand for their proper folding, and then co-expression with the putative partner would solve the folding problem. Only weak signals could be observed for 1871 constructs, although most (1330 samples) were in sufficient amount (above 0.5 mg) for NMR measurement. They might be in some aggregated but soluble state.

6.6 Large-scale Protein Production for Structure Determination

The constructs selected at the previous stage are, finally, cloned into the plasmid vector and then their nucleotide sequences are examined for confirmation. Then, the uniformly [13 C/15 N]-labeled samples for NMR spectroscopy are prepared using the large-scale (in typical cases, 9 mL internal/90 mL external) dialysis-mode cell-free reaction for 6–12 h [18]. The product is purified by His-tag affinity chromatography (HisTrap, GE HealthCare, Sweden, or TALON, Clonetech, USA), incubated with protease (in almost all cases, TEV protease is used) to cleave off the His-tag, and are then treated again with the His-tag affinity column to remove the His-tag peptide. The resultant is further subjected to ion-exchange or gel-filtration

105

106

6 The Use of the Escherichia coli Cell-free Protein Synthesis

Fig. 6.5 Purification of the zf-TRAF domain from human TNF receptor-associated factor 4 protein. The supernatant fraction of the cell-free reaction (lane 1), the flow-through fraction of the first HisTrap chromatography (lane 2), the elution fraction of the first

HisTrap chromatography (lane 3), the fraction after TEV protease cleavage (lane 4), the flow-through fraction of the second HisTrap chromatography (lane 5), and the elution fraction of the second HisTrap chromatography (lane 6) are shown.

chromatography, if the sample purity is insufficient. Figure 6.5 shows purification of the zf-TRAF domain from human TNF receptor-associated factor 4 whose solution structure we have determined and deposited to the PDB (PDB ID: 2YUC). By March 2007 we had prepared 2165 samples. Some of the samples at the previous HSQC stage are judged to be more suitable for X-ray crystallography. Consequently, selenomethionine-substituted protein samples for multi-wavelength anomalous diffraction (MAD) phasing [19] are prepared using the cell-free synthesis in which methionine in the reaction mixture is simply replaced by selenomethionine [11]. Selenomethionine substitution is almost complete in most cases and is much higher than that by the in vivo expression method. To accelerate large-scale NMR structural biology, the RIKEN GSC Large-scale NMR Facility, which houses 40 high-field NMR spectrometers (including three 900 MHz systems), was constructed, and a software package, KUJIRA, for the systematic and interactive NMR data analysis [20] and the program CYANA for automated structure calculation [21] were developed and used. Using our high-throughput expression pipeline based on the bacterial cell-free protein synthesis in combination with the facility and the software, we have successfully determined about 1300 structures of proteins/domains from higher eukaryotes, mouse, human, and Arabidopsis thaliana, by NMR spectroscopy, and

6.7 Discussion

Fig. 6.6 Some structures we have determined by NMR spectroscopy.

most of their coordinates have already been deposited to PDB (for details see http://www.rsgi.riken.go.jp/Target/index.html). Figure 6.6 shows some of the structures.

6.7 Discussion

We have established a high-throughput expression pipeline based on the bacterial cell-free protein synthesis. The pipeline is successfully operated, at least for NMR structural genomics and proteomics of protein domains from higher eukaryotes, as described in this chapter. Improved reaction conditions are expected to increase the success rate of the pipeline. The pipeline would be applicable to X-ray crystallography, though some tuning and/or modification would be required. Other types of the cell-free systems, including those from human and insect cells and wheat germ, could be used if the productivity and success rate are matched with the demand of the project. Other large-scale structural genomics projects, such as the Center for Eukaryotic Structural Genomics (CESG, http://www.uwstructuralgenomics.org/) and the NorthEast Structural Genomics Consortium (NESG, http://www.nesg.org/), have announced the use of wheat germ cell-free protein synthesis in their expression pipeline. At the CESG, they now produce 16 samples per week for NMR screening [22] and, thus, on the assumption of similar statistics to those of our NMR screening, they may expect at least 3–4 new samples suitable for NMR analysis per week.

107

108

6 The Use of the Escherichia coli Cell-free Protein Synthesis

References 1 Mittl, P. R., Grutter, M. G. (2001). Structural genomics: opportunities and challenges. Curr. Opin. Chem. Biol. 5: 402– 408. 2 Chen, L., Oughtred, R., Berman, H. M., Westbrook, J. (2004). TargetDB: a target registration database for structural genomics projects. Bioinformatics 20: 2860– 2862. 3 Christendat, D., Yee, A., Dharamsi, A., Kluger, Y., Gerstein, M., Arrowsmith, C. H., Edwards, A. M. (2000). Structural proteomics: prospects for high throughput sample preparation. Prog. Biophys. Mol. Biol. 73: 339–345. 4 Knaust, R. K., Nordlund, P. (2001). Screening for soluble expression of recombinant proteins in a 96-well format. Anal. Biochem. 297: 79–85. 5 Dieckman, L., Gu, M., Stols, L., Donnelly, M. I., Collart, F. R. (2002). High throughput methods for gene cloning and expression. Protein Expr. Purif. 25: 1–7. 6 Gilbert, M., Albala, J. S. (2002). Accelerating code to function: sizing up the protein production line. Curr. Opin. Chem. Biol. 6: 102–105. 7 Spirin, A. S., Baranov, V. I., Ryabova, L. A., Ovodov, S. Y., Alakhov, Y. B. (1988). A continuous cell-free translation system capable of producing polypeptides in high yield. Science 242: 1162–1164. 8 Kigawa, T., Yabuki, T., Yoshida, Y., Tsutsui, M., Ito, Y., Shibata, T., Yokoyama, S. (1999). Cell-free production and stableisotope labeling of milligram quantities of proteins. FEBS Lett. 442: 15–19. 9 Kigawa, T., Muto, Y., Yokoyama, S. (1995). Cell-free synthesis and amino acidselective stable isotope labeling of proteins for NMR analysis. J. Biomol. NMR 6: 129–134. 10 Yabuki, T., Kigawa, T., Dohmae, N., Takio, K., Terada, T., Ito, Y., Laue, E. D., Cooper, J. A., Kainosho, M., Yokoyama, S. (1998). Dual amino acid-selective and site-directed stable-isotope labeling of the human cHa-Ras protein by cell-free synthesis. J. Biomol. NMR 11: 295–306. 11 Kigawa, T., Yamaguchi-Nunokawa, E., Kodama, K., Matsuda, T., Yabuki, T., Matsuda, N., Ishitani, R., Nureki, O.,

12

13

14

15

16

17

18

19

Yokoyama, S. (2002). Selenomethionine incorporation into a protein by cell-free synthesis. J. Struct. Funct. Genomics 2: 29–35. Yokoyama, S., Hirota, H., Kigawa, T., Yabuki, T., Shirouzu, M., Terada, T., Ito, Y., Matsuo, Y., Kuroda, Y., Nishimura, Y., Kyogoku, Y., Miki, K., Masui, R., Kuramitsu, S. (2000). Structural genomics projects in Japan. Nat. Struct. Biol. 7(Suppl.): 943–945. 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. Busso, D., Kim, R., Kim, S. H. (2003). Expression of soluble recombinant proteins in a cell-free system using a 96-well format. J. Biochem. Biophys. Methods 55: 233– 240. Vinarov, D. A., Lytle, B. L., Peterson, F. C., Tyler, E. M., Volkman, B. F., Markley, J. L. (2004). Cell-free protein production and labeling protocol for NMR-based structural proteomics. Nat Methods 1: 149–153. Waldo, G. S., Standish, B. M., Berendzen, J., Terwilliger, T. C. (1999). Rapid protein-folding assay using green fluorescent protein. Nat. Biotechnol. 17: 691–695. Bateman, A., Coin, L., Durbin, R., Finn, R. D., Hollich, V., Griffiths-Jones, S., Khanna, A., Marshall, M., Moxon, S., Sonnhammer, E. L., Studholme, D. J., Yeats, C., Eddy, S. R. (2004). The Pfam protein families database. Nucleic Acids Res. 32: D138–141. Matsuda, T., Koshiba, S., Tochio, N., Seki, E., Iwasaki, N., Yabuki, T., Inoue, M., Yokoyama, S., Kigawa, T. (2007) Improving cell-free protein synthesis for stable-isotope labeling. J. Biomol. NMR 37: 225–229. Hendrickson, W. A., Horton, J. R., LeMaster, D. M. (1990). Selenomethionyl proteins produced for analysis by multiwavelength anomalous diffraction (MAD): a vehicle for direct determination of threedimensional structure. EMBO J. 9: 1665– 1672.

References 20 Kabayashi, N., Iwahara, J., Koshiba, S., Tomizawa, T., Tochio, N., Güntert, P., Kigawa, T., Yokoyama, S. (in press, 2007). KWIRA, a package of integrated modules for systematic and interactive analysis of NMR data directed to high-throughput NMR structure studies. J. Biomol. NMR. 21 Herrmann, T., Guntert, P., Wuthrich, K. (2002). Protein NMR structure deter-

mination with automated NOE assignment using the new software CANDID and the torsion angle dynamics algorithm DYANA. J. Mol. Biol. 319: 209–227. 22 Vinarov, D. A., Markley, J. L. (2005). High-throughput automated platform for nuclear magnetic resonance-based structural proteomics. Expert Rev. Proteomics 2: 49–55.

109

111

7

The Wheat Germ Cell-free Protein Synthesis System Tatsuya Sawasaki and Yaeta Endo

7.1 Overview

Among cell-free protein synthesis systems, the wheat germ-based translation system is of special interest for its eukaryotic nature: it has significant advantages for the high-throughput production of eukaryotic multi-domain proteins in the folded state. Here we describe how this highly efficient cell-free expression system is built and review its application to today’s functional and structural biology.

7.2 Development of a Highly Efficient Eukaryotic Cell-free Protein Synthesis System

Currently, three strategies are being used for protein production: chemical synthesis, in vivo expression, and cell-free protein synthesis. The first two methods have severe limitations. Chemical synthesis is not practical for peptides longer than 20 residues [1], and in vivo expression cannot produce those proteins that interfere with host cell physiology [2–4]. Cell-free translation systems, in contrast, can synthesize proteins with high accuracy at a speed approaching in vivo rates [5, 6], and they can express proteins that seriously interfere with cell physiology. However, cell-free systems in general were at one time relatively inefficient because of their instability [7] until Spirin et al. introduced the first continuous cell-free translation system, which in fact works much more efficiently than conventional batch systems [8]. To adapt the cell-free system for industrial applications, Swartz et al. have been developing ways to remove several limitations of the E. coli cell-free system through their focus on basic biochemical reactions. The recent fruitful results of their work have been reviewed [9]. We took a different approach. We found that plants contain endogenous inhibitors of translation [10, 11] and demonstrated that elimination of these inhibitors leads to an extraordinarily stable and efficient translation system [12]. We, however, recognized that certain critical improvements must be made to adapt this new cell-free system to the needs of high-throughput applications in modern proCell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

112

7 The Wheat Germ Cell-free Protein Synthesis System

teomics. To improve its performance, we scrutinized the following critical design issues [13]: (a) optimization of the 5′ and 3′ UTRs of mRNA; (b) elimination of the 5′ -7 mGpppG (cap) and poly(A)-tail; (c) design of PCR primers so that we may generate transcription templates directly from Escherichia coli cells carrying cDNAs, bypassing the time-consuming cloning steps; (d) construction of an expression vector specialized for the massive production of proteins; (e) development of a transcription and translation reaction in one tube to eliminate the mRNA purification step; (f) invention of the bilayer reaction, which enables us to perform continuous exchange mode translation for high-volume production without using a membrane; and (g) robotic automation of all transcription and translation steps. The resulting system exhibited attractive features essential for the systematic study of protein structures and functions and for a series of applications delineating the scope of modern proteomics. 7.2.1 Preparation of a Highly Active and Robust Extract from Wheat Embryos

One of the most convenient and promising eukaryotic cell-free translation systems is based on wheat embryos in which all the necessary components of translation are stored under a dried state. Since the first report by Johnston et al. [14], the solvent flotation enrichment of viable, intact embryos from wheat seeds has commonly been used for the preparation of wheat embryos [15]. We found, however, that conventional wheat germ extracts contained the RNA N-glycosidase tritin and other inhibitors of translation such as thionin, ribonucleases, deoxyribonucleases, and proteases. We hypothesized that those inhibitors, originating from endosperm, caused the instability of then existing wheat germ cell-free systems. We then measured the modification of ribosomes by tritin as a sensitive marker of the contaminants [16]. When germs are isolated from dry wheat seeds by conventional procedures [14], their microscopic examination reveals that they contain some white material and several white and brownish granules (Fig. 7.1A). Our analysis of ribosomal RNAs from protein synthesis reaction performed on the extract of those embryos confirmed the depurination of ribosomes (Fig. 7.1B). After 4 h of incubation, 24% of the ribosome population had been depurinated, as judged by the aniline-dependent formation of a specific RNA fragment (arrow, corresponding to bands on lanes 6 and 9 are of the fragment marker). At the start of incubation, however, 7% of the population had already been depurinated. When RNA was extracted directly from embryos by guanidine isothiocyanate-phenol, little formation of the aniline-induced fragment was observed (lanes 7 and 8). Thus, we concluded that depurination had occurred during the extract preparation and continued during protein synthesis reaction. The observed extent of depurination meant considerable damage to protein synthesis, since inactivation of any one of the actively translating ribosomes on an mRNA results in the blockage of the respective polyribosome and the cessation of translation [16]. Attempts to neutralize harmful enzymes with synthetic RNA aptamers, which can tightly bind to ribosome-inactivating protein (RIP) [17] and

7.2 Development of a Highly Efficient Eukaryotic Cell-free Protein Synthesis System

Fig. 7.1 Removal of tritin from embryos [10].

One extract was prepared from unwashed embryos and the other from washed embryos (A), and both were subjected to the depurination assay (B). Translation mixtures prepared from the unwashed embryo extract were incubated for 0, 1, 2, 3, and 4 h (lanes 1–5, respectively), and those from the washed embryo extract for 0, 2, and 4 h (lanes 10–12, respectively). Isolated RNA was treated with

acid/aniline, and then separated on 4.5% polyacrylamide gel. Additionally, RNA was directly extracted from embryos with guanidine isothiocyanate-phenol and analyzed before (lane 7) and after (lane 8) treatment with acid/aniline. The fragment marker (lanes 6 and 9) was incubated in the presence of gypsophilin, a highly active RIP from Gypsophila elegance; the arrow indicates the aniline-induced fragment.

other ribonucleases, were unsuccessful. Fortunately, extensive washing of the embryos proved highly effective in removing those contaminants originating from endosperm (right-hand panel, Fig. 7.1A). Subsequent depurination assay detected no aniline-induced cleavage (lanes 10–12 in Fig. 7.1B) [12], which signified minimal, if any, depurination during extract preparation and incubation. Thus, simple washing of embryos could apparently eliminate not only tritin but also other endogenous translation inhibitors such as thionin, ribonucleases, deoxyribonucleases, and proteases. The cell-free system prepared from washed embryos had much higher translational activity and stability than the conventional system (compare Fig. 7.2A and B). When dihydrofolate reductase (DHFR) was synthesized with 5′ -capped mRNA containing 549 nts of 3′ UTR and a poly(A) tail, the new system containing 24% extract supported nearly linear kinetics for over 4 h, while the conventional system ceased to function after 1.5 h. When the washed extract content was increased from 24 to 48%, the initial rate of amino acid incorporation doubled, but the reaction stopped

113

114

7 The Wheat Germ Cell-free Protein Synthesis System

Fig. 7.2 Protein syntheses in the extract solutions prepared from unwashed and washed embryos [10]. The batch system contained either 12 µL (24%) or 24 µL (48%) of washed embryo extract (A) and unwashed embryo extract (B). Protein synthesis was quantified by measuring hot trichloroacetic acid insoluble radioactivity. Arrows indicate the addition of substrates.

after 1 h. This halt, however, was caused by a shortage of substrates rather than the irreversible inactivation of ribosomes or factors necessary for translation; as shown by an arrow in Fig. 7.2A, once amino acids, ATP, and GTP were replenished, translation resumed at a rate similar to the initial reaction rate. In contrast, when the extract content in the conventional system was increased from 24 to 48%, protein synthesis decreased. Furthermore, once stopped, protein synthesis in 24% extract could not be resumed by replenishment of substrates. This indicated irreversible damage by contaminants from endosperm (Fig. 7.2B). The high protein synthesis activity of washed embryos could also be demonstrated by sucrose density gradient analysis: significant formation of polysomes was observed after 1 h of incubation, and, after 2 h, the formation shifted towards heavier polysomes with a concomitant decrease in the amount of 80S monosomes [12]. When cell-free translation was performed in dialysis mode (a version of Spirin’s continuous-exchange cellfree, or CECF, protein synthesis system) [18], the reaction proceeded for over 60 h, yielding 0.8–4.0 mg of functionally active proteins per mL of reaction mixture [12]. Thus, the extensive washing of wheat embryos to eliminate endosperm contaminants gave birth to an extract with high stability and activity. The high efficiency

7.2 Development of a Highly Efficient Eukaryotic Cell-free Protein Synthesis System

of our system can be attributed to at least two factors: high initiation, elongation, and termination rates (efficient usage and recycling of all the translation factors, ribosomes, mRNA, tRNA, etc.), on the one hand, and low endogenous ribonuclease activity (retention of heavy polysomes for a prolonged time), on the other hand. 7.2.1.1 Protocol for the Preparation of Wheat Germ Extract [12] Isolation of Wheat Embryos

1. Grind the wheat seeds in a mill (Roter Speed Mill, model pulverisette 14, Fritsh, Germany). 2. Collect embryos by sieving through a 710–850-µm mesh. 3. Select the intact embryos by solvent flotation using a solvents cyclohexane and carbon tetrachloride in the ratio of 1: 2.5 (vol/vol). 4. Dry overnight in a fume hood. 5. Wash three times with ten volumes of sterile water under vigorous stirring. 6. Sonicate for 3 min in a 0.5% Nonidet P-40 solution by using a Bronson model 2210 sonicator (Yamato, Japan). Preparation of the Extract

1. Grind 5 g of isolated wheat embryo to a fine powder in liquid nitrogen. 2. Add 5 mL of extraction buffer and vortex the mixture briefly. 3. Centrifuge the embryo homogenate (30 000g; 30 min) and retain the supernatant. 4. Gel-filtrate the supernatant by using a Sephadex G-25 (fine) column, equilibrated with two volumes of Extraction buffer (40 mm HEPES-KOH, pH 7.6, 100 mm potassium acetate, 5 mm magnesium acetate, 2 mm calcium chloride, 4 mm DTT and 0.3 mm of each of the 20 amino acids). 5. Collect the void fraction. 6. Centrifuge the fraction (30 000g; 10 min) and retain the final supernatant. 7. Adjust to 200 A260 −units mL−1 with Extraction buffer. 8. Divide into small aliquots and store at –80 ◦ C until use. 7.2.2 mRNA 5′ and 3′ UTRs which Enhance Translation

The 5′ and 3′ UTRs of eukaryotic mRNA play a crucial role in the regulation of gene expression: they control mRNA’s translational efficiency, stability, and localization. The cap and poly(A), found on almost all eukaryotic mRNAs, stimulate translation initiation and stabilize mRNA by their synergistic action. For in vitro translation, however, the use of cap and poly(A) is problematic. The cap molecules and capped mRNA fragments are accumulated during translation due to unavoidable digestion of mRNAs by residual ribonuclease activities, bind to the initiation factor eIF4E and thus strongly competes with the capped mRNA for the initiation factor, inhibiting translation initiation. This leads to a narrow range of optimum concentration of capped mRNA and requires the tedious work of accurately de-

115

116

7 The Wheat Germ Cell-free Protein Synthesis System

termining the amount of mRNA for efficient translation prior to the start of each reaction. This problem is especially serious when protein synthesis concerns (a) the high-throughput materialization of genetic information from large numbers of genes and (b) production of large amounts of protein in CECF for a long period of time, in which mRNA needs to be replenished to prolong the reaction. In addition, the long plasmid-encoded poly(dT/dA) is usually unstable during replication in host cells, and the cap-analogue for mRNA synthesis is costly. To solve those problems, we endeavored to find such 5′ and 3′ UTRs as would enhance mRNA translation without cap and poly(A) in a pSP65 vector. Initial candidates for our study were found among those genomes of plant RNA viruses that contained a positive-sense RNA but had neither cap nor poly(A) [19]. We first focused on the 5′ UTR. After screening 5′ enhancer elements consisting of less than 100 nts, we chose the 71-nt UTR omega sequence (71) of TMV. It appeared the most promising because of its well-known function as a general enhancer of translation both in vivo and in vitro [19–21]. A template coding for luciferase with the 71 5′ UTR was synthesized by in vitro transcription with a 549-nt 3′ UTR, with the 100-nt poly(A) left intact. Activity assays in the batch system (Fig. 7.3A) demonstrated that this mRNA had (a) higher activity than the mRNA without cap and poly(A) and (b) 34% higher activity than the capped and polyadenylated mRNA. These observations are consistent with an earlier report [18]. We tried to improve the template activity further by changing the 5′ -end of 71 and found that GAA (GAA-71) was the most effective, producing a dramatic translational enhancement (nearly 77% of the control level), although we currently have no explanation for the enhancing mechanism. A commercially available SP6 polymerase transcription kit can be used to create 5′ mRNA sequences starting with GAA. Next, we studied the influence of 3′ UTR on mRNA activity. By systematically examining the activity of luciferase mRNAs that contained 3′ UTRs of different lengths and sequences and 5′ UTR consisting of GAA-71 we ascertained that, for efficient translation in the wheat cell-free system, 3′ UTR’s length was more important than its sequence [13]. This result is also consistent with the presence of an exosome, which degrades mRNA from the 3′ end as reported by others [22, 23]. Figure 7.3B shows the optimal concentration for each mRNA. The template that contained GAA-71 as 5′ enhancer and a 3′ UTR made of 1626 nts exhibited the highest template activity, which approached that of the original mRNA with a 5′ cap, a 3′ UTR made of 549 nts, and a poly(A) made of 100 nts. This designed template has worked well in a wide range of mRNA concentrations, allowing us to forego optimization of mRNA concentration. This is a highly desirable feature, especially when aiming for high-throughput protein synthesis. Although the underlying mechanism governing translation enhancement is still unknown, our experiments have demonstrated that both 5′ and 3′ UTRs are important for retaining high translational activity. In eukaryotes ranging from protozoan to human cells, polysomes take circular and helical configurations to form circular polysomes, while those in prokaryotic cells take linear configurations during translation. It is believed that circular polysomes are formed as the result of an interaction between the cap and the

7.2 Development of a Highly Efficient Eukaryotic Cell-free Protein Synthesis System

Fig. 7.3 Effect of UTRs on the template

activity and optimum concentration of mRNA [11]. Protein synthesis activity in the presence of [14 C]leucine (0.05 µCi/25 µL reaction volume) was measured for each of the following luciferase mRNAs: with 5′ cap and a 549-nt 3′ UTR (3′ -549) plus 100-nt pA at the 3′ terminal (cap/549pA); with 5′ cap and 3′ -549 but lacking pA (cap/549); with 5′ GUA-71 and 3′ -549 plus pA (GUA/549pA); with 5′ GAA- and 3′ -18 but without pA (GAA/18); with 5′ GAA- and 3′ -549 plus pA (GAA/549pA); with 5′ GAA- and 3′ -549 but

without pA (GAA/549); with 5′ GAA- and 3′ -1178 (GAA/1178); with 5′ GAA- and 3′ -1626 (GAA/1626); and with 3′ 1626-nt sequence from the negative strand of the plasmid DNA (GAA/N-1626). Reactions were carried out (A) in 25 µL containing either 2.5 pmol of mRNA (cap/549pA, cap/549) or 5 pmol of mRNA (others) and (B) in 25 µL containing 2.5 pmol of mRNA (cap/549pA) or 5 pmol of mRNA (others). After incubation at 26 ◦ C for the indicated periods of time in (A), or 4 h in (B), hot-TCA insoluble radioactivity was measured.

poly(A), mediated by several protein factors such as initiation factors and poly(A)binding protein (PABP) [24–26]. The circular configuration leads to a high reinitiation reaction rate, with ribosomes on the mRNA protecting the latter from ribonuclease attack. Electron microscopic observation of the reaction mixture programmed with the 5′ -GAA--luciferase-3′ -1626 mRNA revealed the formation of circular polysomes, which were very similar to, though somewhat smaller than, those obtained with the capped mRNA containing a 549-nt 3′ UTR and a poly(A) (compare B + E with H + I in Fig. 7.4). Furthermore, the capped but poly(A)-lacking mRNA gave practically the same circular-type polysomes (Fig. 7.4F and G). In contradiction to general understanding, these biochemical and morphological examinations indicated that neither 5′ cap nor 3′ -pA is essential for the architecture of circular polysomes [24–26]. Our findings here also seem consistent with early electron-microscopic observations of purified polysomes as a circular configuration [27]. In any case, they explain the high efficiency of the wheat cell-free translation reaction at the morphological level.

117

118

7 The Wheat Germ Cell-free Protein Synthesis System

Fig. 7.4 EM images of polysomes that formed during wheat cell-free translation: Protein synthesis was performed at 26 ◦ C for 1 h with luciferase mRNA having 5′ cap and pA (cap549pA). Specimens for EM were prepared by surface-spreading and shadowing. (A) Blank without mRNA; (B) protein synthesis with capped and polyadenylated luciferase

mRNA; (C)–(E) magnified images of polysomes C, D, and E as shown by arrows in (B). Effect of 5′ and 3′ enhancer sequences without cap and pA tail (F–I). Translation reactions were programmed with the cap549 mRNA for (F) and the 1626 mRNA for (H) and incubated for 1 h. Bars equal 100 nm.

As part of our effort to put the result of our study into practice, we selected from a random library 5′ UTR sequences that enhance downstream translation (enhancers E01 and E02) [28]. They were shorter than the 5′ UTR mentioned above. The selection was made to take advantage of the wheat cell-free translation system’s characteristic that the sequence upstream of the coding region affects the efficiency of initiation and sufficiently promotes the formation of polysomes in vitro. Although we could not find any 37-nt sequence that had a comparable activity to that of the omega, it was named E02 with that of GAA-, we succeeded in obtaining full activity nearly equal to that of GAA- by combining two short enhancers with different sequences chosen from the selected sequence pool. This is analogous to the reported finding of an in vivo experiment that two copies of the 22-nt element in the  sequence gave full activity, while one copy did not [20]. We did not find any similarity between the selected sequences and  or other

7.2 Development of a Highly Efficient Eukaryotic Cell-free Protein Synthesis System

known 5′ UTR sequences [28]. The selected sequences are the first non-natural sequences that stimulate downstream translation in plants. We have been using these enhancers for the high-throughput protein production to be described below. 7.2.3 Split-primer PCR for Genome-wide Generation of DNAs for Transcription

Cell-free protein synthesis includes in vitro transcription to prepare mRNAs under the control of phage RNA polymerase promoter, and mRNAs are generally prepared from gene-carrying linearized plasmids. Of the steps involved in this procedure, cloning cDNA into an expression vector is one of the most laborious and time consuming. A preferred alternative is direct synthesis from PCR-generated linear cDNAs [29, 30]. In this method, the start and stop signals for transcription and translation are incorporated into primers for PCR to generate autonomous coding fragments. The current PCR-aided cell-free systems have several remaining difficulties: (a) low productivity, which necessitates protein detection by immunological methods, (b) inability to synthesize proteins of high molecular weights, and (c) the short halflife of the linear DNA strand, which prevents productivity improvement by means of CFCF. These limitations mainly stem from two problems: (a) nonspecific annealing such as primer-dimer effect [31] that prevents PCR from generating full length DNAs and (b) rapid degradation of linear DNAs and mRNAs by contaminating nucleases, especially in the E. coli translation system. The second problem is not an issue in our wheat germ cell-free system, which, unlike the other cellfree systems as reported in the literature, has no detectable nuclease activity. Thus, to establish a general and practical method for the in vitro expression of proteins, we focused our efforts on developing ways to minimize the primer–dimer effect during PCR [13]. Figure 7.5 shows the results of protein synthesis in the wheat germ cell-free system in which mRNAs from four different genes were selected as models and transcribed according to a conventional PCR strategy [32]. With a unique primer for each gene (see legend), we used four specific primers (Fig. 7.5A-a): primer-1 to cover the entire promoter sequence of phage SP6 polymerase; primer-2 to introduce both the SP6 promoter and GAA-71 as translation enhancer; primer-3 to connect the sequence of primer-2 to the ORF by virtue of containing 19 nts that are complementary to the 5′ sequence in each ORF; and primer-4 to create the 3′ UTR. Each PCR product was used for transcription reactions, and synthesized RNAs were separated on agarose gel (Fig. 7.5A-e). Although PCR-generated DNA fragments appeared as a single major band apparently at the correct size in each lane (Fig. 7.5A-d), most transcripts were shorter than 500 nts and migrated to the bottom of the gel (arrow, Fig. 7.5A-e). As confirmed by DNase-l treatment (results not shown), the upper bands of each lane were mostly DNA templates. After ethanol precipitation and subsequent washing of the precipitate, cell-free protein synthesis was carried out in a batch mode reaction. The translation products were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and ana-

119

120

7 The Wheat Germ Cell-free Protein Synthesis System

Fig. 7.5 PCR-based expression of cDNA information using (A) conventional method and (B) a new strategy with split-primers: (B-a) design of the split-primer for introducing the required UTRs into cDNA sequences; (B-b) and (B-c) expected PCR-generated DNAs and mRNA, respectively; (B-d)

PCR-generated DNA; (B-e) transcripts resulted from transcription reaction performed in a volume of 100 µL containing a 10-µL aliquot from a PCR sample; (B-f) autoradiography of translation products from batch mode translation (50 µL, 4 h) using all the transcripts as shown in (B-e).

lyzed by autoradiography (Fig. 7.5A-f). The analysis revealed that, for each mRNA, the main products were mostly short peptides of less than 10 kDa (marked with arrowhead) rather than full-length proteins. Expected products appeared only as faint bands even after a long exposure (asterisks). These results can be explained as follows: as the full-length DNA is amplified under the current strategy [Fig. 7.5A-

7.2 Development of a Highly Efficient Eukaryotic Cell-free Protein Synthesis System

b-(1)], short DNA fragments containing promoter sequences are also produced in large numbers through nonspecific amplification [31]. Possible artifacts include those DNAs that contain the promoter, 5′ enhancer, and possibly short parts of the ORF [Fig. 7.5A-b-(2)], and those which contain the N-terminal portion but lack the 5′ enhancer [Fig. 7.5A-b-(3)]. The RNA molecules can be produced as a single band on the gel by in vitro transcription reactions. The problem noted here can be partially overcome by tuning PCR conditions, such as adjusting the molar ratio of the primer to the cDNA and the annealing temperature [31]. But, as this strategy was deemed impractical for high-throughput synthesis, we explored a different approach to minimizing nonspecific products. The new strategy for PCR amplification was to adopt a split-primer (Fig. 7.5B) [13]. The idea was to separate the single primer for the promoter sequence into two parts (Fig. 7.5B-a). We expected that only those DNA fragments (Fig. 7.5B-b) that have the full sequence, including the complete promoter, translation enhancer, ORF, and 3′ UTR, could serve as transcription templates to produce full-length mRNAs (Fig. 7.5B-c) without any further purification of the complete construct. The test results confirmed our expectation: When unpurified PCR products were used as transcription templates, RNA products migrated and formed single bands at their expected sizes (Fig. 7.5B-e), forming patterns similar to DNA fragment patterns on the agarose gel (Fig. 7.5B-d). Figure 7.5B-f demonstrates the efficient production of four proteins, with each one present (besides fewer small peptides) as a single major band in the autoradiogram [13]. The four proteins were produced from the same four genes as were used in our first strategy. Notably, the reaction conditions for PCR and transcription were the same for all four genes and were not fine-tuned purposefully to create an optimal environment for each gene. We also confirmed that, once the reaction volume was appropriately adjusted at the PCR step, it was not necessary to measure or adjust the mRNA concentration prior to translation. This split-primer method is generally applicable to the design of DNA templates to produce proteins with desired tags for such purposes as purification and immobilization, and for marking the reporter protein. 7.2.3.1 Protocol for “Split-primer” PCR [13] Design of Primers As shown in Fig. 7.5B the split-primers PCR technique involves four major steps:

1. Designing the target specific primer (primer-3) in such a way that its 3′ end has target-specific sequence (5′ terminal of open reading frame) and the 5′ end has a portion of the 5′ terminal site of omega sequence. 2. Primer-2 has the full-length omega sequence and a portion of the SP6 promoter sequence at its 5′ end. 3. Primer-1 contains a portion of the 5′ sequence of the SP6 promoter (underlined 14-mer) and 9-mer out of the 14-mer at the 3′ site is designed so as to form a base pairing with 9-mer at 5′ site in the underlined sequence (another half of the promoter) of Primer-2. 4. Primer-4 is specific for sequence at the 3′ UTR of the vector.

121

122

7 The Wheat Germ Cell-free Protein Synthesis System

Therefore, for each clone of cDNA, primer-3 is the only specific primer covering the 5′ end portion of open reading frame and the remaining primers are common for all cDNAs. Based on this strategy, the following four primers have been designed: Primer-1:5′ -GCGTAGCATTTAGGTGACACT (the underlined sequence is the 5′ half of the promoter). Primer-2: 5′ -GGTGACACTATAGAAGTATTTTTACAACAATTACCAACAACAA CAACAAACAACAACAACATTACATTTTACATTCTACAACTACCACCCACCACC ACCAATG (underlined sequence is the 3′ -half of the promoter and the sequence in italic denotes the annealing region of primer 1 or primer 3). Primer-3: 5′ -CCACCCACCACCACCAatgnnnnnnnnnnnnnnnn (the 5′ -coding region of a target gene is in lowercase) and Primer-4: 5′ -AGCGTCAGACCCCGTAGAAA; this is designed following a reported PCR method [13], which covers pUC ori. We describe here the use of omega sequence, but enhancers E01 and E02, which we selected, can be equally utilized. First PCR 1. Prepare 60 µL of a PCR reaction [200 µm each of dNTP, 1.5 U of ExTaq DNA polymerase (Takara Bio Inc., Otsu, Japan), 10 nm of primer-3 and primer-4, 3 ng of plasmid or 3 µL of E. coli (over-night culture) as template, and the buffer supplied by the manufacturer]. 2. Set the PCR thermo-cycler (model MP, Takara Bio Inc., Otsu, Japan) for 4 min denaturation at 94 ◦ C followed by 30 cycles of amplification: 98 ◦ C for 10 s, 55 ◦ C for 30 s, and 72 ◦ C for 5 min, depending on the length of gene (1 kb min−1 ). Second PCR To equip the first PCR amplified target gene with omega and SP6 promoter sequence, the 30 µL second PCR was carried out by mixing 3 µL of the first PCR product (without any purification), with final concentrations of each 100 nm of primer-1 and primer-4, and 1 nm primer-2 with the same cycling conditions as in first PCR. 7.2.4 Bilayer Translation Reaction Method

The CFCF described by Spirin [8] is well suited for well-ordered protein production using the efficient wheat germ cell-free system. It may, however, not be suitable for the massive screening of gene products from comprehensive cDNA libraries; equipped with a semi-permeable membrane in the reaction chamber, the CFCF apparatus is mechanically complicated to manage. Biochemical analyses such as the functional analysis of enzymes require only tens of micrograms. Yet, regular batch mode reactions cannot synthesize a sufficient amount of protein for such analyses, because they can work only for an hour or so. Therefore, we needed a reaction system for analyzing gene products from cDNA libraries – a system simpler and more cost-effective than the CFCF and yet capable of synthesizing larger amounts of proteins than batch mode reactions. Subsequently, we developed the bilayer-

7.2 Development of a Highly Efficient Eukaryotic Cell-free Protein Synthesis System

Fig. 7.6 Novel bilayer cell-free protein synthesis. (A) Schematic

illustration of the method and (B) GFP syntheses in the wheat germ cell-free system, with incubation by the regular batch-mode reaction (◦) and the bilayer method (). Protein synthesis was quantified as described in Fig. 7.2; inset: autoradiograms, with GFP marked by the arrowhead.

translation method (Fig. 7.6). The method does not require any membrane [33], because the translation mixture is simply overlaid with the substrate mixture. The new system worked for more than 10× longer than a batch mode reaction, yielding sub-milligram of proteins. This longevity can be explained as the method enables the continuous supply of substrates, and removal (dilution) of small byproducts, through a phase between the translation mixture and substrate mixture by diffusion. It performed well to synthesize sufficient amounts of proteins for the functional analysis and solubility determination of gene products from cDNA libraries. 7.2.5 Transcription and Translation in One Tube

For our study, we used purified mRNA as the translation template, which is a common practice in eukaryotic cell-free protein synthesis. The purification step, however, was cumbersome and time consuming. To examine the possibility of skipping this step, we tested the use of the whole mixture of transcription containing mRNA along with NTPs, a high concentration of magnesium ion, spermidine, DTT, and pyrophosphate/phosphate in the bilayer reaction system. Surprisingly, direct addition of the whole mixture produced template activity almost identical to that of

123

124

7 The Wheat Germ Cell-free Protein Synthesis System

the purified mRNA reported (data not shown). Under the conditions employed, the concentration of magnesium ion was 12 mm – far more than the well-known optimum concentration of around 3 mm [7, 34]. We followed up this unusual observation with the titration of magnesium ion to find that the system in fact worked well over a wide range of magnesium ion concentration (5–18 mm) in the presence of NTPs and 1.3 mm spermidine. This apparently wide range of may have been due to high concentrations of NTPs, because NTPs are known to reduce the free magnesium ion concentration. In fact, experiments similar to the above, except for the use of purified mRNA, have confirmed that typical values of optimum magnesium concentration lie in a narrow range [7, 12]. The new system that performs transcription and translation in one tube without mRNA purification also worked well in the batch reaction [35]. 7.2.5.1 Protocol for One-tube Protein Synthesis Reaction 1. Prepare 250 µL of transcription reaction [80 mm HEPES-KOH, pH 7.8, 16 mm magnesium acetate, 2 mm spermidine and 10 mm DTT, 2.5 mm each of NTPs, 250 U of SP6 RNA polymerase (80 U µL−1 stock solution, Promega, Madison, WI), 250 U of RNasin (80 U µL−1 stock solution, Promega, Madison, WI) and 25 µg of high quality circular plasmid, or 2.5-µg PCR product as for the transcription template]. 2. Incubate the reaction mixture at 37 ◦ C for 3 h. 3. Add 250 µL of wheat embryo extract (200 A260 −units per mL of the stock solution), and 2 µL of 20 mg mL−1 creatine kinase, and then mix well (reaction mixture). 4. Prepare 5.5 mL of Translational substrate buffer (TSB) [30 mm HEPES-KOH, pH 7.8, 100 mm potassium acetate, 2.7 mm magnesium acetate, 0.4 mm spermidine, 2.5 mm DTT, 0.3 mm amino acid mix, 1.2 mm ATP, 0.25 mm GTP and 16 mm creatine phosphate] in a six-well plate (Whatman Inc., Clifton, NJ). 5. Carefully release 500 µL of the translation mixture at the bottom of titer plate well by inserting the tip to the bottom of the well. Owing to the higher density of the step 3 mix compared with the step 4 mix, one can see clearly two layers. 6. Place a sealing film and then coverlet on the plate to avoid evaporation. 7. Keep the plate in an incubator at 17 ◦ C (15–26 ◦ C) for 16–20 h without shaking. Recommendations for the Incubation Temperature with Regard to the Quality of Products The translation reaction in our system works under wide temperature range: at 15–28 ◦ C practically the same amount of protein is obtained, and even at 10 ◦ C the productivity decreases only by 30–40% of that of 17 ◦ C. Generally, however, the lower the temperature the higher the quality of protein produced. The effect is remarkable, especially for the synthesis of proteases. It is very difficult to synthesize protease in intact form in the absence of specific inhibitors at higher temperature due to their self-digestion (attacks on each other) during incubation. But translation and purification at 0–4 ◦ C decreases the enzyme activity, and one can obtain intact and highly active proteases, although the yield is decreased to 10–15% of that at the optimum temperature, 17 ◦ C.

7.2 Development of a Highly Efficient Eukaryotic Cell-free Protein Synthesis System

7.2.6 Reaction Methods for Large-scale Protein Production

As described earlier, we devised 5′ and 3′ UTRs and incorporated them into a plasmid to construct an efficient expression vector specialized for the wheat cell-free system. The plasmid, named pEU (plasmid of Ehime University), originated from pSP65. Designed to synthesize mRNAs, it contains GAA-71 at the 5′ end and a long UTR (1626 nts) at the 3′ end [13]. It has a multiple cloning site (MCS) between the two UTRs (Fig. 7.7A). Incubation after simply adding the mRNA transcribed from this circular pEU (without linearization) into the dialysis bag of the CECF system resulted in the production of a large amount of protein (Fig. 7.7B, arrow). Densitometric quantification of the Coomassie brilliant blue (CBB)-stained band indicated that the reaction continued for up to 14 days, yielding 9.7 mg of GFP in an active form in a reaction volume of 1 mL containing 40 A260 units of the extract. Although it was necessary to replenish the bag with the mRNA every 48 h, the amount of protein produced exceeded that of the endogenous protein in the reaction mixture (7.5 mg mL−1 ). This indicated that a two-fold purification would already be sufficient to yield the product in a pure form. This CECF-based reaction method, however, had a weak point: it had a long incubation time and thus it was not very practical, particularly from the standpoint of general stability of protein. We, therefore, tried to develop another reaction method to cut the production time, guided by the experimental evidence that, in batch mode reactions, halted translation in a translation mixture could be restarted with the same initial reaction velocity as a new batch when the mixture was subjected to ultrafiltration and the ingredients were replaced with fresh substrates (results not shown). We took advantage of this characteristic of the cleansed wheat extract system and devised a

Fig. 7.7 A cell-free expression vector and its

performance [11]. (A) Schematic illustration of pEU, (B) SDS-PAGE analysis of GFP produced in 14 days of reaction. The mRNA produced by transcription of circular pEU was used for the translation reaction in the dialysis

membrane system and was replenished every 48 h. A 0.1-µL aliquot of the reaction mixture was run on the gel and the resultant protein bands were stained with CBB (Coomassie brilliant blue). The arrow shows GFP and “st” designates an authentic GFP band (0.5 µg).

125

126

7 The Wheat Germ Cell-free Protein Synthesis System

Fig. 7.8 Discontinuous batch translation system. Incubate the reaction mixture in batch mode until translation cease takes place; this usually takes 30 min when the extract concentration is at around 120A260 . The mixture is then subjected to ultrafiltration and

the ingredients are replaced with fresh substrates using VIVAFLOW 50 (molecular weight cutoff 10 000, purchased from VIVASCIENCE Hannover, Germany) (A, B). After repeating this process five times, 150 mg of GFP was produced in 50 mL in 5 h (C).

discontinuous batch translation method, whose principle is depicted in Fig. 7.8A and the actual apparatus is shown in Fig. 7.8B. With this reaction method, one can increase the extract concentration in the reaction mixture while keeping the translation efficiency fairly constant. Adapting this reaction method with concentrated extract (120 A260 -units mL−1 ), the system worked; 150 mg of GFP in a reaction volume of 50 mL was produced in 5 h (Fig. 7.8C).

7.3 Completion of Protocols for the Wheat Cell-free System

Combining all the elemental technologies described above, we could establish two protocols for synthesizing proteins in test tubes for practical use (Fig. 7.9). One protocol is for parallel production and consists of the following steps: (a) in silico selection of suitable genes from a database, (b) construction of templates for transcription by the split-PCR, (c) transcription, and (d) the bilayer translation in which the solution resulting from transcription is directly used as the mRNA source. To produce proteins with tags for the reporter and for such purposes as purification

7.3 Completion of Protocols for the Wheat Cell-free System

Fig. 7.9 Completion of two protocols for protein synthesis based on

the wheat germ cell-free system.

and mutant protein preparation, DNA constructs can be generated by the PCR as well. The other protocol is for massive preparation (bold arrow) for structural biology and consists of: (a) selection of suitable gene products from the parallel production described above and subsequent functional screening, (b) cloning of the genes into pEU and transcription of mRNA, (c) fine tuning of translational conditions such as the concentrations of pertinent ions, and (d) protein production that incorporates either bilayer, dialysis, or discontinuous batch reactions. Proteins with a desired purification tag can easily be affinity purified, and the tag portion can be eliminated if so desired. 7.3.1 Performance of the Wheat Cell-free System

Cell-free systems lend themselves to high-throughput formats suitable for the parallel production of numerous proteins. There have been reports published on cellfree systems made from Escherichia coli that demonstrate the possibility of producing large amounts of proteins for structural analyses [36, 37]. In fact, E. coli systems of various designs have been employed [30, 38]. E. coli systems, however, have two serious limitations, one intrinsic and the other technically unavoidable. The intrinsic limitation concerns the folding of multi-domain proteins, which are found more often in eukaryotes than in prokaryotes. Prokaryotic cytosol has the fundamental ability to support effective co-translational protein folding. Multi-domain proteins in eukaryotic translation systems, however, have a much better chance to fold into their correct conformations than in prokaryotic systems [39, 40]. The second limitation is that, in general, PCR-generated DNA fragments are not efficiently transcribed and translated in E. coli systems, even though several PCR-based methods have been devised to alleviate this limitation [41, 42]. This is because the mRNAs and DNA-degrading enzymes originating from cells contaminate the systems and decrease the stability of the templates, thus reducing the yield. In contrast, mRNAs added to the eukaryotic wheat system are stable for a long time.

127

128

7 The Wheat Germ Cell-free Protein Synthesis System

To evaluate the performance of the cell-free method in high-throughput screening, we carried out parallel protein syntheses from 27 genes originating with E. coli cells carrying the cDNAs [13]. In 50 out of the total 54 cases (authentic and fusion proteins with GST), a clearly visible CBB-stained protein band was obtained (Fig. 7.10), and the yield after a 36-h incubation was 0.1–2.3 mg per mL of the reaction volume in the dialysis cup mode reaction (CECF); the was yield estimated by densitometric scanning of the bands using BSA as the standard (Table 7.1). Some gene products were recovered in the supernatant (S) and some others in the precipitate phase (P) after centrifugation at 30 000g for 15 min. Notably, we could not detect any dependency of the productivity and solubility of the proteins on the gene sources. Furthermore, the system had little preference in codon usage, which is a prerequisite for genome-wide protein expression. In fact, we could express malaria proteins from 76% AT-rich cDNAs of Plasmodium falciparum [43]. The most important requirement for an expression system, however, is that it produces quality products. To verify the quality of proteins expressed in the wheat germ cell-free system and its general applicability, we tested a subset of product proteins for their functional activity. PHOT1 gene from Arabidopsis thaliana codes for a protein of 120 kDa, which noncovalently binds to flavin mononucleotide (FMN) and presumably acts as a chromophore in light-dependent autophosphorylation [44]. As shown in Fig. 7.11(A), the protein that was responsible for the blue light responsive activity could be produced only when the reaction was carried out in the presence of FMN [45]. The results demonstrated that a prosthetic group is required during translation and folding into the holoenzyme, which supports the notion that co-translational folding takes place on eukaryotic ribosomes. Figure 7.11(B) shows another example in which the polyhedrins of baculoviruses synthesized in the cell-free system were active enough to be assembled in vitro into polyhedra but are smaller than those of natural baculoviruses [46]. We next examined the foldedness of proteins produced in the cell-free system [13]. For our NMR spectroscopic analysis, we selected flowering locus T protein (FT protein, 19.8 kDa). Originally named from phenotype on genetic works of late flowering mutants [47], it has not yet been biochemically characterized. The synthesized FT protein was recovered mostly in soluble form, but it was not obvious whether it had folded correctly. We therefore measured by NMR the heteronuclear single-quantum coherence (HSQC) of 15 N-labeled FT protein containing four 15 Nlabeled amino acids, Gly, Ala, Leu, and Gln. Dispersion of resonances measured in the 2D 15 N-1 H correlation spectrum exhibited a reasonable number of signals that demonstrated a folded protein in solution (Fig. 7.12). Although the system has not yet been optimized for producing membrane proteins in full size in soluble, folded state, there is the possibility of combining detergents or liposomes [37, 48].

7.3 Completion of Protocols for the Wheat Cell-free System

Fig. 7.10 High-throughput production of proteins from cDNA libraries

for screening [11]. Authentic (A) and GST-fused (G) proteins were produced from 54 different cDNAs by semi-automated PCR/transcription and translation. The reaction mixtures were separated by SDS-PAGE and stained with CBB (Coomassie brilliant blue). T and S mark, respectively, the total reaction mixture and supernatant fraction after centrifugation at 30 000g for 15 min.

129

130

7 The Wheat Germ Cell-free Protein Synthesis System Table 7.1 An example of proteins synthesized in the PCR-directed wheat germ cell-free system.

Proteins encoded by cDNAs

Annotation clone no. name

Arabidopsis X56062 (From GenBank and MIPS) chlorophyll a/b-binding protein Agamous-like gene 9 (AGL9) AF015552 Flowering locus T (FT) AF152096 HY5 AB005456 Flowering locus F (FLF) AF116527 Hypothetical protein At1g69630a) From a commercial cDNA libraryb) Putative heat shock protein 40 AL021749 Heat shock protein 70-3 Y17053 Putative s-Adenosylmethionine AY037214 synthetase NADPH thioredoxin reductase Z23108 Putative ACC oxidase AF370155 Putative fructokinase AF387001 Rubisco activase X14212 Glutaredoxin At4g15660a) Chlorophyllase 2 AF134302 Human (From GenBank) M22349 Neuron-specific gamma-2 enolased) zeta-Crystallin/quinone L13278 reductase X11-like protein AB014719 Importin alpha 1 NM_002266 M17851 Glyceraldehyde-3-phosphate dehydrogenase Enolase 3d) NM_001976 APBA3 d),e) NM_004886 JAK binding proteind) NM_003745 From a commercial cDNA libraryb) Phosphoglycerate kinase 1 XM_010102 β-Actin X00351 Hypothetical protein FLJ10652 XM_006938 Hypothetical protein FLJ10559 XM_001479 a) b) c) d) e)

M.W.

Authentic Total Sup (mg) (%)

Fusion Total (mg)

Sup (%)

25 995 0.2

30

0.5

90

At01

29 065 19 808 18 462 21 864 11 311

0.7 0.3 0.4 0.2 0.1

30 100 90 100 40

0.8 0.8 1.5 0.4 0.1

90 100 90 100 100

At02 At03 At04 At05 At06

38 189 1.8 71 144 0.9 42 793 1.5

30 100 100

1.0 –c) 0.6

80 –c) 100

At07 At08 At09

40 635 36 677 35 276 51 981 11 311 34 902

10 10 10 20 –c) 10

0.5 1.2 0.6 0.6 0.4 0.4

20 100 100 80 80 70

At10 At11 At12 At13 At14 At15

47 266 1.0

100

0.5

100

Hs01

35 205 2.3

80

1.3

100

Hs02

82 480 0.5 57 859 –c) 36 051 0.4

100 –c) 70

0.2 80 Hs03 0.2 30 Hs04 0.9 100 Hs05

46 956 1.7 61 451 0.9 23 550 0.4

80 100 30

0.9 0.2 0.2

100 100 100

Hs06 Hs07 Hs08

43 965 41 735 41 539 35 237

100 100 –c) 50

0.7 0.3 0.2 1.0

100 100 10 70

Hs09 Hs10 Hs11 Hs12

0.1 1.0 1.0 0.4 –c) 0.2

1.0 1.3 –c) 0.7

MIPS Arabidopsis thaliana database MAtDB (http://mips.gsf.de/proj/plant/jsf/athal/index.jsp). Lambda ZAP II Library, products of Stratagene cat #937010 and #936204 respectively. Below the detectable level. These genes were cloned from tissue (heart, brain, kidney, liver, placenta) cDNAs (BioChain Institute, Inc., cat #0516001). Amyloid beta (A4) precursor protein-binding, family A, member 3.

7.3 Completion of Protocols for the Wheat Cell-free System

Fig. 7.11 Quality of eukaryotic gene products expressed in the wheat

cell-free system. (A) Blue-light-induced autophosphorylation activity of the PHOT1 gene product. (B) Polyhedron formation in vitro. After the translation of polyhedrin gene from baculoviruses in the cell-free system, the translation mixture was kept for 48 h at 4 ◦ C. Microscopic observation revealed the formation of small polyhedra.

Fig. 7.12 HSQC spectrum of the hypothetical

protein of flowering locus T protein produced in the cell-free system [11]. The FT protein was synthesized in the same way as in Fig. 7.10 except that Ala, Leu, Gly, and Gln in both translation and substrate mixtures were replaced with their 15 N-labeled forms (ISOTEC INC.). After incubation for 48 h, the reaction mixture (1 mL) was dialyzed against

10 mm phosphate buffer (pH 6.5) overnight, and then centrifuged at 30 000g for 10 min. The supernatant containing 30 µm of the protein was directly subjected to NMR spectroscopy. The spectrum was recorded on a Bruker DMX-500 spectrometer at 25 ◦ C, and 2048 scans were averaged for the final 1 H–15 N HSQC spectrum.

131

132

7 The Wheat Germ Cell-free Protein Synthesis System

7.3.2 Robotic Automation of the Cell-free Protein Synthesis

One of the ultimate goals of our technology has been the automated execution of the protocols for the parallel materialization of genetic information for screening, on the one hand, and large-scale production of quality proteins for structural biology, on the other hand. That goal has been achieved by the advent of robotized protein synthesizers (http://www.cfsciences.com). One of them, named GenDecoder, incorporates both the transcription and the bilayer translation. It can handle four 96-well multi-titer plates (384 samples) in a fully automated manner in a 24-hour cycle (Fig. 7.13A). A more recent version of GenDecoder, which we named Super GenDecoder, can handle up to 6000 samples in a 60-hour cycle (not shown). They are both based on a 96-well format and may therefore not be suitable to produce proteins on a massive scale. In view of this, we developed two more robots, named Protemist 211 and DT (Fig. 7.13B and C). Protemist 211 can synthesize six proteins in parallel in a large scale (0.5–3.0 mg). Protemist DT is equipped with an affinitypurification unit and can produce 70–80% pure products overnight (Fig. 7.13D).

7.4 Application to High-throughput Biochemical Annotation of Genetic Information

With the sequencing of genomes, attention has turned to the structures and functions of proteins. Although cell-based expression systems have been widely used, they have certain limitations in terms of the quality, quantity, or high-throughput production of proteins. Many of these limitations can be circumvented by the use of the wheat cell-free translation system. Besides being eukaryotic, this system has other advantages over others in that its raw material (wheat seeds) is readily available at a low cost and that, wheat being a foodstuff, bio-pollution and ethical issues concerning the system would be minimal. 7.4.1 Genome-wide Functional Analysis

To ascertain the performance of our approach to the high-throughput synthesis of gene products and biochemical characterization, we performed cell-free protein synthesis starting with E. coli cells carrying cDNAs from certain organisms. Transcription and translation were performed automatically by the protein synthesizer GenDecoder equipped with four standard 96-well microtiter plates for both transcription and translation. We chose 439 genes encoding protein kinases (PKs) out of 1064 annotated PKs of Arabidopsis thaliana (http://plantsp.sdsc.edu/) [45], 292 out of 540 mouse PKs (http://kinase.com/mouse/), and 214 out of 518 human PKs (http://198.202.68.14/human/kinome/). Each kinase gene was fused with the gene coding GST, and the DNA templates were generated by the split-primer PCR. After protein synthesis, each sample was spun down at 30 000g for 15 min (some

7.4 Application to High-throughput Biochemical Annotation of Genetic Information

Fig. 7.13 GenDecoder and Protemist series

specialized for functional and structural analyses of genetic information: (A) GenDecoder equipped with three robotic arms for pipetting and plate transfer, one incubator for transcription and translation, four plates of maximum capacity, and a centrifuge for mRNA recovery after ethanol precipitation. (B) Protemist 211 for massive protein production using the discontinuous method. (C) Protemist DT for massive protein production using a 6-well plate. (D) The result of robotic protein synthesis using Protemist

DT in (C). The translation mixture was separated on SDS-polyacrylamide gel and stained with CBB (Coomassie brilliant blue). Proteins fused with GST were synthesized. Total reaction mixture (3 µL) or affinity purified samples (corresponding to 60 µL of the translation mixture) were analyzed. Proteins synthesized in a fusion form were (1) Akt1, (2) DHFR, (3) GFP, (4) GUS, (5) Junk2, (6) p38, (M) molecular weight markers (ranging from 14 400 to 94 000, Amersham Biosciences), (T) total protein, and (E) eluate fraction of 10 mm glutathione.

insoluble gene products were lost in the precipitate at this step). The recovered supernatant was mixed with glutathione-coated beads to isolate the product fused with GST, and the partially affinity-purified protein was treated with phosphatase. Then, as expedient substrates, histone H1 and myelin basic protein (MBP) (commonly used substrates for serine and threonine kinases) were added to each sample. After incubation in the presence of [γ -32 P]ATP and magnesium, the aliquot of each sample was separated with SDS-PAGE and analyzed by autoradiography;

133

134

7 The Wheat Germ Cell-free Protein Synthesis System

Fig. 7.14 Biochemical screening of products from human cDNAs encoding protein kinases: methods for protein synthesis and analysis are described in the text. The eleven products indicated by a small filled circle were separated by SDS-PAGE and (A) stained with CBB (Coomassie brilliant blue) and (B) subjected to autoradiography. Arrowheads indicate expedient substrates. M denotes molecular weight markers (ranging for 14 400 to 94 000).

calcium-dependent protein kinases can be determined through similar kination reactions in the presence of calcium instead of magnesium. In the autoradiograms, 207 out of the 439 PKs of Arabidopsis thaliana, 80 out of the 292 mouse PKs, and 86 out of the 214 human PKs tested displayed autophosphorylation activity (data not shown). This indicated that at least the kinase domain of each of those proteins that were produced in the wheat cell-free system was folded into its active form. Figure 7.14 shows examples of the analytical results. Each of the 11 human kinases shown in Figure 7.14A was synthesized and affinity purified by the protein synthesizer Protemist DT, and its kinase activity was assayed in a similar way. The analysis demonstrated that each enzyme had its own unique substrate specificity – thus the products are in enzymatically active forms (Fig. 7.14B). Among the genes that could be successfully expressed in our wheat cell-free system were those that could hardly be expressed in E. coli cells, and if expressed at all then only as low quality products. These difficult proteins were well produced in our new system in a folded and, thus, active state (Table 7.2). 7.4.2 Preparation of Protein for NMR Spectroscopy

One of the bottlenecks in high-throughput structural determination of proteins is the step to produce proteins in folded state. The other limiting step is amino acid-specific, selective labeling for assigning the signals collected in the 1 H-15 N HSQC spectra in NMR spectroscopy, or selenomethionine labeling for MAD phasing in X-ray crystallography. Cell-free systems have the general advantage over cell

7.4 Application to High-throughput Biochemical Annotation of Genetic Information Table 7.2 Successfully expressed “difficult proteins” in the wheat cell-free system.

Species

Activity

Gene name

Ref.

Cypoviruse Bacillus Thermococcus Aquifex Arthrobactor Arabidopsis Rice Rice Rice Mouse Human Human

Crystalline particle formation Endonuclease (Bam HI) DNA synthesis Methyltransferase Sarcosine oxidation Protein phosphorylation by light Anthranilate synthase Shikimate kinase EPSP synthase Immunoglobulin induction DNA binding DNA binding

VP3 Restriction enzyme KOD DNA polymerase tRNA (Gm18) methyltransferase Sarcosine oxidase PHOT1 Anthranilate synthase alpha subunit Shikimate kinase EPSP synthase Osteopontin c-fos c-jun

46 33 33 49 33 45 50 51 51 52 53 53

expression systems in that cell-free products for NMR measurements do not need extensive purification, because none of the endogenous proteins are labeled during the translation. Cell-free systems derived from E. coli, however, contain high levels of amino-acid-metabolizing enzymes, since the extracts are prepared from cells in the exponential growth phase. As a result, many types of inhibitors of amino acid metabolism, or in certain cases suitable mutant strains, have to be added to avoid scrambling of the label among amino acids. In the wheat-based system, in contrast, most of those enzymatic activities were expected to be very low, since the embryos are in hibernation. Nevertheless, two transaminases and one synthetase were active during the translation reaction. Our subsequent search for their inhibitors was successful in that the interconversions of amino-15 N between Ala and Glu (alanine transaminase) and between Glu and Asp (aspartate transaminase) were completely inhibited by adding β-chloro-l-alanine and l-methionine sulfoximimate, and that the other leaking pathway between Glu and Asp (glutamine synthetase) was suppressed by aminooxyacetate [54]. The improved methodology is expected to pave the way for a high-throughput, labeled protein synthesis system suited for estimating the foldedness of proteins [55] and intermolecular interaction between proteins of interest and ligands, besides assigning signals in the NMR field without protein purification. However, from the view point of economy, only 15–25% of amino acid additions can be incorporated into proteins even in our efficient cell-free system. This is an economic issue to be solved. One solution would be to recover unincorporated amino acids after translation using a conventional amino acid isolation method, which would benefit structural biology. The fruitful results of the application of the wheat cell-free protein synthesis system to NMR structure determination are presented at http://www.uwstructuralgenomics.org/ [56, 57], although X-ray analysis results have yet to be published.

135

136

7 The Wheat Germ Cell-free Protein Synthesis System

7.5 Conclusion

The improved wheat germ cell-free translation system is especially powerful in the high-throughput production of eukaryotic multi-domain proteins in the folded state. We have succeeded in the robotic automation of the protocols, as embodied in a series of machines that can perform all the steps preceding protein purification. They hold the promise of increasing the throughput and decreasing the cost of protein production. Protein microarrays, which have enormous potential in biosciences and medical fields, are one of the promising applications of the cell-free system. Recent progress of elemental technologies in this field is remarkable [58, 59], but the goal appears to be quite far from realization. One of the most difficult tasks is to establish a technique for storage, i.e., a technique to keep each immobilized protein intact in such a tiny space, which would be totally different from that used for DNA chips. In this light, we propose a possible use of the cell-free system. A prospective scheme would be to take advantage of the stability and capability of the system and store all the ingredients involved in transcription and translation in lyophilized form together with the DNA template. Adding water prior to the preparation of chips would then start the production of proteins fused with immobilization tags. The accumulated information on the structures and functions of gene products should revolutionize our understanding of biology and fundamentally alter practice in medicine, as well as influencing other industries.

References 1 Blaschke, U. K., Silberstein, J., Muir,T. timized in vitro system. Arch. Biochem. W., Protein engineering by expressed proBiophys. 1996, 328, 9–16. tein ligation. Methods Enzymol. 2000, 328, 7 Roberts, B. E., Paterson, B. M., Efficient 478–496. translation of tobacco mosaic virus RNA 2 Henrich, B., Lubitz, W., Plapp, R., Lysis and rabbit globin 9S RNA in a cell-free of Escherichia coli by induction of cloned system from commercial wheat germ. phi X174 genes. Mol. Gen. Genet. 1982, Proc. Natl. Acad. Sci. U.S.A. 1973, 70, 185, 493–497. 2330–2334. 3 Goff, S. A., Goldberg, A. L., An increased 8 Spirin, A. S., Baranov, V. I., Ryabova, content of protease La, the lon gene prodL. A., Ovodov, S. Y., Alakhov, Y. B., A uct, increases protein degradation and continuous cell-free translation system blocks growth in Escherichia coli. J. Biol. capable of producing polypeptides in high Chem. 1987, 262, 4508–4515. yield. Science 1988, 242, 1162–1164. 9 Swartz, J., Developing cell-free biology for 4 Chrunyk, B. A., Evans, J., Lillquist, J., industrial applications. J. Ind. Microbiol. Young, P., Wetzel, R., Inclusion body Biotechnol. 2006, 33, 476–485. formation and protein stability in se10 Barbieri, L., Battelli, M. G., Stirpe, quence variants of interleukin-1 beta. J. F., Ribosome-inactivating proteins from Biol. Chem. 1993, 268, 18053–18061. plants. Biochim. Biophys. Acta 1993, 1154, 5 Kurland, C. G., Translational accuracy in 237–282. vitro. Cell 1982, 28, 201–202. 11 Ogasawara, T., Sawasaki, T., Morishita, 6 Pavlov, M. Y., Ehrenberg, M., Rate of R., Ozawa, A., Madin, K., Endo, Y., A new translation of natural mRNAs in an opclass of enzyme acting on damaged ri-

References

12

13

14

15

16

17

18

19

20

21

22

23

bosomes: ribosomal RNA apurinic site specific lyase found in wheat germ. EMBO J. 1999, 18, 6522–6531. Madin, K., Sawasaki, T., Ogasawara, T., Endo, Y., A highly efficient and robust cell-free protein synthesis system prepared from wheat embryos: plants apparently contain a suicide system directed at ribosomes. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 559–564. Sawasaki, T., Ogasawara, T., Morishita, R., Endo, Y., A cell-free protein synthesis system for high-throughput proteomics. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 14652–14657. Johnston, F. B., Stern, H., Mass isolation of viable wheat embryos. Nature 1957, 179, 160–161. Erickson, A. H., Blobel, G., Cell-free translation of messenger RNA in a wheat germ system. Methods Enzymol. 1983, 96, 38–50. Wool, I. G., Gluck, A., Endo, Y., Ribotoxin recognition of ribosomal RNA and a proposal for the mechanism of translocation. Trends Biochem. Sci. 1992, 17, 266–269. Hirao, I., Madin, K., Endo, Y., Yokoyama, S., Ellington, A. D., RNA aptamers that bind to and inhibit the ribosomeinactivating protein, pepocin. J. Biol. Chem. 2000, 275, 4943–4948. Kim, D. M., Choi, C. Y., A semicontinuous prokaryotic coupled transcription/translation system using a dialysis membrane. Biotechnol. Prog. 1996, 12, 645–649. Zaccomer, B., Haenni, A. L., Macaya, G., The remarkable variety of plant RNA virus genomes. J. Gen. Virol. 1995, 76(Pt 2), 231–247. Gallie, D. R., Walbot, V., Identification of the motifs within the tobacco mosaic virus 5′ -leader responsible for enhancing translation. Nucleic Acids Res. 1992, 20, 4631–4638. Gallie, D. R., Translational control of cellular and viral mRNAs. Plant Mol. Biol. 1996, 32, 145–158. Beelman, C. A., Parker, R., Degradation of mRNA in eukaryotes. Cell 1995, 81, 179–183. Anderson, J. S., Parker, R. P., The 3′ to 5′ degradation of yeast mRNAs is a gen-

24

25

26

27

28

29

30

31

32

33

34

eral mechanism for mRNA turnover that requires the SKI2 DEVH box protein and 3′ to 5′ exonucleases of the exosome complex. EMBO J. 1998, 17, 1497–1506. Le, H., Tanguay, R. L., Balasta, M. L., Wei, C. C., Browning, K. S., Metz, A. M., Goss, D. J., Gallie, D. R., Translation initiation factors eIF-iso4G and eIF-4B interact with the poly(A)-binding protein and increase its RNA binding activity. J. Biol. Chem. 1997, 272, 16247–16255. Tarun, S. Z., Wells, Jr., S. E., Deardorff, J. A., Sachs, A. B., Translation initiation factor eIF4G mediates in vitro poly(A) taildependent translation. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 9046–9051. Craig, A. W., Haghighat, A., Yu, A. T., Sonenberg, N., Interaction of polyadenylate-binding protein with the eIF4G homologue PAIP enhances translation. Nature 1998, 392, 520–523. Warner, J. R., Knopf, P. M., Rich, A., A multiple ribosomal structure in protein synthesis. Proc. Natl. Acad. Sci. U.S.A. 1963, 49, 122–129. Sawasaki, T., Morishita, R., Gouda, M. D., Endo, Y., Methods for highthroughput materialization of genetic information based on wheat germ cellfree expression system. Methods Mol. Biol. 2007, 375, 95–106. Gurevich, V. V., Use of bacteriophage RNA polymerase in RNA synthesis. Methods Enzymol 1996, 275, 382–397. Hanes, J., Pluckthun, A., In vitro selection and evolution of functional proteins by using ribosome display. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 4937–4942. Erlich, H. A., Gelfand, D., Sninsky, J. J., Recent advances in the polymerase chain reaction. Science 1991, 252, 1643–1651. Sambrook, J., Russell, D. W., Molecular Cloning: A Laboratory Manual, (3rd edition), Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 2001. Sawasaki, T., Hasegawa, Y., Tsuchimochi, M., Kamura, N., Ogasawara, T., Kuroita, T., Endo, Y., A bilayer cell-free protein synthesis system for high-throughput screening of gene products. FEBS Lett. 2002, 514, 102–105. Craig, D., Howell, M. T., Gibbs, C. L., Hunt, T., Jackson, R. J., Plasmid cDNA-

137

138

7 The Wheat Germ Cell-free Protein Synthesis System

35

36

37

38

39

40

41

42

43

44

directed protein synthesis in a coupled eukaryotic in vitro transcription-translation system. Nucleic Acids Res. 1992, 20, 4987– 4995. Sawasaki, T., Morishita, R., Gouda, M. D., Endo, Y., Methods Mol. Biol., in the press. Kigawa, T., Yabuki, T., Yoshida, Y., Tsutsui, M., Ito, Y., Shibata, T., Yokoyama, S., Cell-free production and stable-isotope labeling of milligram quantities of proteins. FEBS Lett. 1999, 442, 15–19. Klammt, C., Lohr, F., Schafer, B., Haase, W., Dotsch, V., Ruterjans, H., Glaubitz, C., Bernhard, F., High level cell-free expression and specific labeling of integral membrane proteins. Eur. J. Biochem. 2004, 271, 568–580. Noren, C. J., Anthony-Cahill, S. J., Griffith, M. C., Schultz, P. G., A general method for site-specific incorporation of unnatural amino acids into proteins. Science 1989, 244, 182–188. Netzer, W. J., Hartl, F. U., Recombination of protein domains facilitated by cotranslational folding in eukaryotes. Nature 1997, 388, 343–349. Hartl, F. U., Hayer-Hartl, M., Molecular chaperones in the cytosol: from nascent chain to folded protein. Science 2002, 295, 1852–1858. Lesley, S. A., Brow, M. A., Burgess, R. R., Use of in vitro protein synthesis from polymerase chain reaction-generated templates to study interaction of Escherichia coli transcription factors with core RNA polymerase and for epitope mapping of monoclonal antibodies. J. Biol. Chem. 1991, 266, 2632–2638. Martemyanov, K. A., Spirin, A. S., Gudkov, A. T., Direct expression of PCR products in a cell-free transcription/translation system: synthesis of antibacterial peptide cecropin. FEBS Lett. 1997, 414, 268–270. Tsuboi, T., [Malaria transmission-blocking vaccine development]. Seikagaku 2005, 77, 646–649. Sakai, T., Kagawa, T., Kasahara, M., Swartz, T. E., Christie, J. M., Briggs, W. R., Wada, M., Okada, K., Arabidopsis nph1 and npl1: blue light receptors that mediate both phototropism and chloroplast relocation. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 6969–6974.

45 Sawasaki, T., Hasegawa, Y., Morishita, R., Seki, M., Shinozaki, K., Endo, Y., Genome-scale, biochemical annotation method based on the wheat germ cell-free protein synthesis system. Phytochemistry 2004, 65, 1549–1555. 46 Ikeda, K., Nakazawa, H., Shimo-Oka, A., Ishio, K., Miyata, S., Hosokawa, Y., Matsumura, S., Masuhara, H., Belloncik, S., Alain, R., et al., Immobilization of diverse foreign proteins in viral polyhedra and potential application for protein microarrays. Proteomics 2006, 6, 54–66. 47 Kobayashi, Y., Kaya, H., Goto, K., Iwabuchi, M., Araki, T., A pair of related genes with antagonistic roles in mediating flowering signals. Science 1999, 286, 1960–1962. 48 Pornillos, O., Chen, Y. J., Chen, A. P., Chang, G., X-ray structure of the EmrE multidrug transporter in complex with a substrate. Science 2005, 310, 1950–1953. 49 Hori, H., Kubota, S., Watanabe, K., Kim, J. M., Ogasawara, T., Sawasaki, T., Endo, Y., Aquifex aeolicus tRNA (Gm18) methyltransferase has unique substrate specificity. TRNA recognition mechanism of the enzyme. J. Biol. Chem. 2003, 278, 25081–25090. 50 Kanno, T., Kasai, K., Ikejiri-Kanno, Y., Wakasa, K., Tozawa, Y., In vitro reconstitution of rice anthranilate synthase: distinct functional properties of the alpha subunits OASA1 and OASA2. Plant Mol. Biol. 2004, 54, 11–22. 51 Kasai, K., Kanno, T., Akita, M., IkejiriKanno, Y., Wakasa, K., Tozawa, Y., Identification of three shikimate kinase genes in rice: characterization of their differential expression during panicle development and of the enzymatic activities of the encoded proteins. Planta 2005, 222, 438–447. 52 Miyazaki, T., Ono, M., Qu, W. M., Zhang, M. C., Mori, S., Nakatsuru, S., Nakamura, Y., Sawasaki, T., Endo, Y., Nose, M., Implication of allelic polymorphism of osteopontin in the development of lupus nephritis in MRL/lpr mice. Eur. J. Immunol. 2005, 35, 1510–1520. 53 Miyamoto-Sato, E., Takashima, H., Fuse, S., Sue, K., Ishizaka, M., Tateyama, S., Horisawa, K., Sawasaki, T., Endo, Y., Yanagawa, H., Highly stable and efficient mRNA templates for mRNA-protein fu-

References sions and C-terminally labeled proteins. Nucleic Acids Res. 2003, 31, e78. 54 Morita, E. H., Shimizu, M., Ogasawara, T., Endo, Y., Tanaka, R., Kohno, T., A novel way of amino acid-specific assignment in (1)H-(15)N HSQC spectra with a wheat germ cell-free protein synthesis system. J. Biomol. NMR 2004, 30, 37–45. 55 Morita, E. H., Sawasaki, T., Tanaka, R., Endo, Y., Kohno, T., A wheat germ cell-free system is a novel way to screen protein folding and function. Protein Sci. 2003, 12, 1216–1221. 56 Vinarov, D. A., Lytle, B. L., Peterson, F. C., Tyler, E. M., Volkman, B. F., Markley, J. L., Cell-free protein produc-

tion and labeling protocol for NMR-based structural proteomics. Nat. Methods 2004, 1, 149–153. 57 Vinarov, D. A., Markley, J. L., Highthroughput automated platform for nuclear magnetic resonance-based structural proteomics. Expert Rev. Proteom. 2005, 2, 49–55. 58 Zhu, H., Snyder, M., Protein chip technology. Curr. Opin. Chem. Biol. 2003, 7, 55–63. 59 Merkel, J. S., Michaud, G. A., Salcius, M., Schweitzer, B., Predki, P. F., Functional protein microarrays: just how functional are they? Curr. Opin. Biotechnol. 2005, 16, 447–452.

139

141

8

Cell-free Expression of Integral Membrane Proteins for Structural Studies Christian Klammt, Daniel Schwarz, Ines Lehner, Solmaz Sobhanifar, Frank Löhr, Johan Zeelen, Clemens Glaubitz, Volker Dötsch, and Frank Bernhard

8.1 Overview

Cellular expression systems are often very inefficient for the high-level production of membrane protein. Toxic effects, instability or formation of inclusion bodies are frequently observed effects that prevent the synthesis of sufficient amounts of functional protein. The development of preparative scale cell-free expression systems has provided new, alternative tools with several attractive benefits for the production of membrane proteins. Unique and fascinating properties are the possibilities to synthesize recombinant membrane proteins directly into detergent micelles or into liposomes of defined composition. Considerable success has already been made with the expression of structurally diverse membrane proteins in cell-free systems and this chapter summarizes recent approaches. We discuss distinct applications with a special focus on the cell-free production of functionally folded membrane proteins.

8.2 Introduction

Integral membrane proteins (IMPs) are embedded into cellular membranes by multiple hydrophobic transmembrane segments (TMSs) and they control numerous essential functions like transport activities, energy generation, signal perception and communication of the cell with its environment. The topology of IMPs is generally dominated by α-helical structures, while typical β-barrel arrangements are prevalent in IMPs inserted into the outer membrane of Gram-negative bacteria. An average whole cell proteome is supposed to consist to 20–40% of IMPs. Many pharmaceutical studies currently focus on IMPs as they play key roles in various global human diseases. IMPs provide thus an estimated 60% of all modern drug targets. A basic prerequisite for a directed drug design as well as for the understanding of biological functions is knowledge of the three-dimensional structure of Cell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

142

8 Cell-free Expression of Integral Membrane Proteins for Structural Studies

a protein. Relatively high amounts of protein, in the range of several 100 mg, are often needed for structural approaches by either X-ray crystallography or nuclear magnetic resonance (NMR) spectroscopy. With few exceptions, where the natural abundance of a protein offers the opportunity of high yield preparations, heterologous cellular expression systems of prokaryotic or eukaryotic origin have to be used to produce sufficient amounts of protein. Unfortunately, the hydrophobic nature of IMPs and their distinct localization in cellular membranes often causes tremendous problems upon their synthesis in conventional expression systems. Blocking of cellular protein targeting systems, complex formation with cellular membrane proteins or disintegration of membranes are frequent effects that result in toxicity to living cells, leading to low expression rates. Cell-free (CF) expression systems offer, in principle, the possibility to eliminate toxic effects of recombinant proteins as no living and metabolic active environments are necessary. Only care has to be taken that the essential translation and transcription processes will not be affected by the overproduced protein. CF systems are therefore predestined for the synthesis of even strong toxins that cannot obtained with conventional in vivo expression systems [1]. Moreover, numerous additional benefits are intrinsic properties of CF expression systems. Expression conditions like pH, redox potential or buffer systems could easily be modified. A highly valuable characteristic is the open nature of the reaction. This allows the addition of nearly any compound at any time-point directly into the reaction. In contrast to cellular expression systems, there is no selection of additives by specific transport mechanisms or risk of metabolic conversion or even breakdown of the added compound. The proteolytic degradation of synthesized proteins could be inhibited by the addition of protease inhibitors and their folding into functional conformations could be facilitated by supplemented chaperones, cofactors or other helper proteins. Recombinant proteins could further be stabilized by providing ligands, substrates or inhibitors. A unique and fascinating option is the generation of artificial hydrophobic environments in CF expression reactions for the production of soluble IMPs. Different designs of preparative scale CF expression systems with protein yields are possible, starting from a few 100 µg up to several mg per mL of reaction volume. Batch systems with only one compartment can be carried out in various common labware containers such as simple plastic reaction tubes. More advanced systems like continuous exchange CF (CECF) reactions are composed of two compartments, holding a reaction mixture (RM) and a feeding mixture (FM) with a volume ratio of RM : FM usually between 1:10 and 1:20 [2–4]. The two compartments are separated by a semipermeable membrane that ensures efficient exchange of precursors from the FM into the RM. Inhibitory breakdown products are continuously removed from the translation machinery by diffusion into the FM. CECF systems give greater yields than batch systems, and considerable efforts have recently been made to improve the productivity and longevity of batch CF reactions [5–10]. Several groups have published reliable and detailed protocols for the set-up of CF reactions [11–13]. In addition to the individual CF designs, commercial systems are also available that can result in the high-level expression of IMPs [14–

8.3 Specific Characteristics for the Cell-free Expression of Membrane Proteins

16]. A key compound of CF expression systems is a cellular extract that can be derived from either bacterial origin, mostly Escherichia coli cells, or from eukaryotic origin, preferentially selected wheat germs. High quality bacterial extracts can be obtained from various common laboratory strains like BL21 derivatives [12] or RNAse deficient strains such as A19 or D10 [17, 18]. The preparation of bacterial extracts is a reliable and routine technique [11, 18–21] and an important step is the removal of the cellular mRNA that virtually eliminates any background expression in subsequent CF expression reactions. Nucleotides and amino acids can be added to release residual mRNA from the ribosomes in a “run-off” translation procedure [19, 20]. The non-protected mRNA will then rapidly become degraded by endogenous RNAses. We prefer a modification of this step by adjusting the extract to 400 mm NaCl followed by incubation at 42 ◦ C for 45 min [11]. This treatment causes the efficient dissociation of mRNA and ribosomes while no addition of precursors is necessary. So far only E. coli extracts have been used for the CF production of IMPs. This might be mostly attributed to the more difficult preparation procedure of wheat germ extracts [22]. However, CF expression systems based on eukaryotic cell extracts might be considered in the future as they could provide several advantages, especially for the production of eukaryotic IMPs, such as improved folding pathways or the possibility of posttranslational modifications. CF reactions with E. coli extracts are operated as coupled transcription–translation systems with added circular plasmid DNA or linear DNA fragments as a template [3]. Even several different templates could be transcribed simultaneously in a CF reaction, thus enabling the coexpression of different proteins, e.g., various subunits of heterooligomeric IMP complexes. The ratio of the recombinant proteins could furthermore easily be manipulated by varying the amounts of the supplied templates. An efficient transcription is provided by placing the target genes under control of a T7 promoter and by addition of T7 RNA polymerase into the reaction. This chapter summarizes recent advancements in the CF production of IMPs on a preparative scale; results with non-preparative scale CF systems are not covered. CF expression systems generally provide an attractive alternative technique for protein targets that cannot efficiently be produced in conventional in vivo expression systems. The technique does not require special equipment and can be operated in most standard biochemical laboratories. CF expression of IMPs is still an emerging application but the achievements already obtained are highly promising and make a rapidly increasing request for this technology very likely.

8.3 Specific Characteristics for the Cell-free Expression of Membrane Proteins

Different modes are feasible for the expression of IMPs in CF systems (Fig. 8.1). Most cellular membranes have been removed during extract preparation and only spurious amounts of lipids might remain [14]. Standard CF reactions therefore result, consequently, in the production of IMPs as precipitates as no hydrophobic compartments are present in the RM. The CF reaction protocol in that mode is

143

144

8 Cell-free Expression of Integral Membrane Proteins for Structural Studies

Fig. 8.1 Cell-free expression modes. The CF expression of IMPs can be run in three different modes: IMPs can be expressed as precipitate (Mode A), solubilized in an appropriate detergent and further reconstituted into proteoliposomes. Alternatively, the IMPs can be soluble expressed in the presence of detergents

(Mode B), purified and reconstituted into proteoliposomes, or expressed in presence of preformed liposomes (Mode C) and isolated from the RM by density gradient centrifugation. CD, circular dichroism; EM, electron microscopy; NMR, nuclear magnetic resonance.

virtually identical to the production of globular proteins [11, 14]. Critical components are the ions Mg2+ and K+ because of their narrow optimal concentration ranges, usually 12–17 mm and 270–330 mm, respectively. Optimization screens have, therefore, often to be employed before high levels of protein synthesis can be obtained. Supplementation of CF reactions with detergents or lipids generates

8.3 Specific Characteristics for the Cell-free Expression of Membrane Proteins

Fig. 8.2 Overproduction of integral

membrane proteins by cell-free expression. SDS-polyacrylamide gel analysis of IMPs from different families that have been CF produced either as precipitate or in soluble form in the presence of detergents. Samples (1 µL) of the soluble or of the resuspended insoluble part of the RM were analyzed on 10% tricine SDS gels (lanes 1–3, 10), 12% SDS gels (lanes 4, 13–16) or 16.5% SDS gels (lanes 5–9, 11, 12). The overproduced IMPs are indicated by arrows and proposed secondary structures are

illustrated on top; detergents are given in parenthesis in the following, if applicable. Lane 1, EmrE; lane 2, SugE; lane 3, Tbsmr; lane 4, Hsmr (DPC) purified by Ni-NTA and additional arrows indicate oligomers; lane 5, RM control; lane 6, YfiK (Brij-58); lane 7, YfiK (Brij-98); lane 8, YfiK; lane 9, TehA, lane 10, TehA; lane 11, SecE; lane 12, Tsx; lane 13, porcine V2R; lane 14, human V2R; lane 15, rat CRF; lane 16, human ETB. P, precipitate; S, soluble expressed; and p, purified.

preformed micelles or liposomes in the RM. This option is an attractive feature specific to CF expression and not possible with any other expression system. It enables the direct synthesis of IMPs into micelles composed of the desired detergents or into liposomes of defined compositions. Critical steps in common IMP preparation protocols like the disintegration of cellular membranes or repeated transfers of IMPs into micelles of different compositions can thus be reduced or even completely avoided. In the best case, the IMPs can become inserted in the desired environment straight after translation and the resulting proteomicelles or proteoliposomes could be directly used for further analysis. A high level production of IMPs is, therefore, generally possible with three different CF expression modes: (A) the production of IMP precipitates without any hydrophobic additives [11, 23], (B) the synthesis of IMPs into micelles in presence of detergents [13, 14, 23, 24], and (C) the synthesis of IMPs into liposomes in the presence of lipids [16] (Fig. 8.2). 8.3.1 Cell-free Expression of Membrane Proteins in the Presence of Detergents or Lipids

Detergents have to be added at concentrations above their specific critical micellar concentrations (c.m.c.s) to become effective for the solubilization of IMPs. However, detergents also interfere non-specifically with any hydrophobic protein regions and they may thus affect the productivity of the CF system. Several deter-

145

146

8 Cell-free Expression of Integral Membrane Proteins for Structural Studies

gents such as sodium deoxycholate, sodium cholate and N-laurylsarcosine as well as dodecyl-phosphocholine (DPC) and n-octyl-β-glucopyranoside (β-OG) severely inhibit CF systems already at low concentrations and they are not suitable for the soluble expression of IMPs [13, 14, 24]. However, most commonly employed relatively mild detergents appear to be tolerated by CF systems at concentrations that exceed several times the proposed specific c.m.c.s [13, 14, 24] (Table 8.1). Table 8.1 Efficiency of selected detergents for the cell-free production of membrane proteins.

8.3 Specific Characteristics for the Cell-free Expression of Membrane Proteins Table 8.1 (continued)

Evaluation of the most effective detergent for the CF production of a specific IMP should be one of the primary subjects of initial optimization screens. Most IMPs are probably likely to become soluble when expressed in various different detergents. The bi-chain-phosphocholines 1,2-dioctanoyl-sn-glycero-3phosphocholine (diC8 PC), 1,2-diheptanoyl-sn-glycero-3-phosphocholine (DHPC), the alkyl-glucosides n-dodecyl-β-d-maltoside (DDM) and n-decyl-β-d-maltoside (DM), the alkyl-ether polyoxyethylene-sorbitan-monolaurate 20 (Tween 20) as well as the poly(ethylene-glycol) derivatives polyethylene-glycol derivatives 4-(1,1,3,3tetramethyl-butyl)phenyl-ether (TX-100) and polyethylene-glycol 400 dodecyl-ether (Thesit) are some detergents that resulted in the production of preparative amounts of protein [13, 14, 23, 24] (Table 8.1). In contrast, all these detergents have not been very effective for the CF expression of other IMPs like the porcine vasopressin type 2 receptor (V2R), a member of the family of G-protein coupled receptors (GPCRs).

147

148

8 Cell-free Expression of Integral Membrane Proteins for Structural Studies

Various polyoxyethylene-alkyl-ethers and the steroid derivative digitonin are exceptionally suitable for the CF expression of structurally different IMPs [13, 14, 24]. EmrE, Tsx and even the V2R protein and other members of the GPCR family could be produced in amounts of up to several mg of soluble protein in one mL of RM. Some structural details of these detergents, such as the length of polyoxyethylene chains, are important for the efficiency of IMP solubilization. Long-chain derivatives like polyoxyethylene-(23)-lauryl-ether (Brij-35), polyoxyethylene-(20)-cetyl-ether (Brij-58), polyoxyethylene-(20)-stearyl-ether (Brij78) and polyoxyethylene-(20)-oleyl-ether (Brij-98) resulted in a high level soluble expression of the different IMPs. In the presence of shorter chain length derivatives, like polyoxyethylene-(10)-cetyl-ether (Brij-56) or polyoxyethylene-(10)-oleylether (Brij-97), IMPs have still been synthesized in high amounts but they remained almost quantitatively as precipitate. The supplied final detergent concentration deserves special consideration. Most detergents become inhibitory above certain concentrations and some guidelines for the supply of detergents into CF reactions have to be considered [13, 14, 24]. The particular micellar concentration (Cmic ) of a detergent in the RM should be at least equal to the estimated molar concentration of the synthesized IMP at the end of a CF reaction. An excess of synthesized IMP might form undesired precipitates or even heterogeneous micelles that could prevent further structural approaches. The Cmic is difficult to calculate as it is a result of the specific c.m.c. as well as of the aggregation number, the proposed number of detergent molecules per micelle. These parameters, however, are highly variable as they depend on several environmental factors like pH and temperature as well as on the topology of the solubilized IMP [25]. A good compromise C mic has thus to be found to ensure optimal solubilization of the synthesized amount of protein while avoiding inhibitory effects of too high detergent concentrations. Detergents with a low aggregation number resulting in a high C mic at still relatively low molar concentrations appear, generally, to be more suitable for the efficient solubilization of CF produced IMPs. However, specific characteristics of detergents further contribute to their behavior in CF reactions. DDM and digitonin already became inhibitory above 15 × c.m.c. (1.8 mm) and 4.5 × c.m.c. (3.3 mm), respectively, while, e.g., Brij-58 and Brij-78 are completely tolerated at even 170 × c.m.c. (12.8 mm) and 280 × c.m.c. (12.9 mm), respectively. The amount of soluble produced IMP in a CF reaction as a function of detergent concentration follows plateau-like kinetics and the yield remains constant above a certain concentration level. However, not all of the synthesized protein might become soluble and it is common for some residual precipitated IMP to be present at the end of the CF reaction. For the CF expression of soluble Tsx protein, optimal detergent concentrations have been determined for TX-100 (7 × c.m.c.), Brij-58 (47 × c.m.c.), Brij-78 (76 × c.m.c.), digitonin (2.2 × c.m.c.) and DDM (15 × c.m.c.) [24]. Increased yields of functionally folded mechanosensitive channel MscL were obtained in CF reactions with mixed micelles of TX-100 and E. coli lipids or in the presence of liposomes prepared from E. coli lipid mixtures [14]. In addition, selenomethionine labeled EmrE has been synthesized in the presence

8.3 Specific Characteristics for the Cell-free Expression of Membrane Proteins

of 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) liposomes; the quality of the isolated protein was sufficient for structural analysis by X-ray crystallography [16]. However, no further characterization of the proteoliposomes, e.g., by freeze– fracture analysis, has been done. Subsequent purification steps involved the solubilization of the IMPs with detergents, and the efficiencies of IMP reconstitution into the provided liposomes during the CF reaction are still unclear. Nevertheless, these first data provide promising results and the expression of IMPs into artificial liposomes might become an important technique for membrane protein analysis. 8.3.2 Detergents for the Efficient Resolubilization of Cell-free Produced Membrane Proteins

For the CF production of IMP precipitates, standard reaction protocols could instantly be used and no time-consuming evaluation of detergents is needed. High yields of synthesized IMP are ensured as no inhibitory effects of supplied detergents are present. In addition, the IMPs could be obtained in highly purified form in only a few steps [11]. This expression mode mostly resembles conventional in vivo approaches of the production of IMPs in the form of inclusion bodies. However, structural differences might exist between cellular inclusion bodies and CF produced precipitates. IMP precipitates obtained by CF reactions usually solubilize rapidly upon addition of relatively mild detergents and they do not require intensive denaturation and refolding steps as is known from the solubilization of inclusion bodies [14, 23, 24]. Denaturizing agents like guanidinium·HCl or excessive amounts of urea could thus be avoided, and gentle mixing of the IMP precipitate with a suitable detergent at a final concentration of 1–5% at room temperature is often sufficient for a quantitative solubilization. In particular, 1-myristoyl-2hydroxy-sn-glycero-3-[phospho-rac-(1-glycerol)] (LMPG) and related derivatives like 1-palmitoyl-2-hydroxy-sn-glycero-3-[phospho-rac-(1-glycerol)] (LPPG) exhibit outstanding properties in the efficient solubilization of CF-produced precipitates of structurally different IMPs [24, 26]. Other mild detergents like DDM, DPC, TX-100, DHPC or 3-[(3-cholamidopropyl)dimethylammonio]-1-propansulfonate (CHAPS) are further effective in the solubilization of specific IMPs like EmrE, Tsx or the mechanosensitive channel McsL, but they failed to solubilize the GPCR protein V2R [14, 24] (Table 8.1). Always most important is the preparation of functionally folded proteins – the solubilization of precipitated IMPs might generally be more critical than the production as soluble proteins in the presence of detergents. Some proteins like the transporter EmrE or the channel MscL can be produced in a functional form in both ways. Solubilization out of CF produced precipitates as well as direct soluble expression in the presence of detergents resulted in active proteins [14, 23, 24]. However, some IMPs like the nucleoside transporter Tsx could only produced in fully functional form with the soluble mode of CF expression [24].

149

150

8 Cell-free Expression of Integral Membrane Proteins for Structural Studies

8.4 Case Studies for the High Level Cell-free Expression of Membrane Proteins

The application of CF expression to the production of membrane proteins is an emerging technique and we describe here recent approaches of the preparative scale CF production of structurally different IMPs. The available data are still confined to relatively few proteins, but the considerable structural diversity of the synthesized IMPs and recent success in the expression of GPCR proteins already indicate an enormous potential of the CF technique for the high-level production of difficult IMPs (Figs. 8.2 and 8.3 below). The functionality of the synthesized IMPs, and also the corresponding yields obtained with in vivo expression systems, are discussed if data are available. The yields of recombinant protein obtained from CECF systems can easily be manipulated by varying the RM/FM ratio or by refreshing the complete FM during the reaction. The specific reaction conditions should, therefore, be considered when the efficiencies of different CF reactions are compared. It is even more difficult to compare yields of CF reactions with those of in vivo reactions. If cell volumes should be compared, one could consider that 1 L of an outgrown E. coli culture in standard Luria broth (10 g tryptone, 10 g yeast extract, 5 g NaCl) contains a cell volume equivalent to approx. 5–10 mL of RM. However, this calculation considers neither the preparation of S30 extracts, enzymes or reaction mixtures nor fermentation and purification procedures. In general, CF expression is a relatively expensive technique and it might, therefore, be the first choice primarily for specific labeling purposes or if a target protein cannot become expressed at all or only at low amounts in other systems. 8.4.1 α-Helical Transporters

Transporters control the conditions between the inside and outside of their hostmembrane, such as the uptake of substrates, the efflux of toxic substances or the regulation of ion concentrations. The 110 amino acid small multidrug transporter EmrE has been the subject of several approaches of high-level production by CF expression. Conventional in vivo expression of EmrE routinely yields approx. 1 mg of protein per liter of E. coli culture [27]. In CF systems based on E. coli S30 extracts, 2–3 mg per mL of RM in preparative scale set-ups and 1–2 µg of protein per 20 µL of RM in analytical set-ups have been reported [23]. Supplied detergents did not alter the production rate of EmrE and directed most of the protein into the soluble fraction. Pull down experiments as well as crosslinking studies with CF-produced EmrE revealed a homodimer as the functional conformation, and activity assays of the transporter in DDM micelles showed specific substrate binding of [3 H]tetraphenylphosphonium (TPP+ ). The determined K d of 2.3 nm matches very nicely with the K d of 2.8 nm obtained from in vivo produced EmrE [28]. The methyl viologen uptake of reconstituted EmrE was measured to verify the specific substrate transport across a membrane. A H+ gradient dependent accumulation of the substrate against its concentration gradient that could be determined to a

8.4 Case Studies for the High Level Cell-free Expression of Membrane Proteins

rate of 7.5 nmol min–1 per µg of CF produced EmrE that is very similar to that isolated from E. coli cells at 5.5 nmol min–1 µg–1 [23]. CF-produced EmrE precipitate could be solubilized in DDM and reconstituted into liposomes based on E. coli lipids. These proteoliposomes also showed specific transport of the substrate ethidium bromide [11]. Recorded NMR spectra indicated in addition an identical conformation of CF-produced EmrE to that obtained by in vivo expression. In further CF expression studies, the four homologous bacterial multidrug transporters TBsmr, BPsmr, Psmr and Hsmr could be produced [23] (Fig. 8.2). All four proteins can also be produced in an in vivo E. coli system but at least the Psmr protein apparently showed higher expression levels in the CF system. High-level CF expression of the mechanosensitive channel MscL from E. coli, a homopentamer of 14 kDa subunits that each consist of two TMSs, resulted in functional protein [14]. Conductance measurements on reconstituted MscL protein in giant liposomes were performed by patch clamp assays. Soluble CF-produced MscL in the presence of TX-100 showed comparable activity (8.3 ± 1.8 open channels per patch) to that in an E. coli cells overproduced channel (9.3 ±5.5 open channels per patch). The addition of an amino-terminal poly(His)6 -tag lowered the numbers of channels per patch of MscL protein isolated from both expression systems. The full activity of MscL could be restored by removing the tag upon cleavage with the factor Xa protease. Crosslinking experiments indicated the formation of the functional pentameric state of MscL already in detergent micelles. The in vitro expression levels could be increased to a final yield of approx. 3.6 mg MscL protein per mL RM by adding 18 µg mL–1 E. coli lipids. The bacterial α-helical tellurite transporter TehA (36 kDa, 10 TMSs) and the cysteine transporter Yfik (22 kDa, 6 TMSs) were not produced in detectable amounts in E. coli cells. The yields in CECF systems for both protein and of the C-terminal truncated derivative TehA (24 kDa, 7 TMSs) reached levels up to 2.7 mg mL–1 RM (Fig. 8.2). No functional analysis of the two proteins has been made but structural evaluation by NMR spectroscopy revealed strong evidence of folded proteins [11, 26]. Expression in an E. coli based system as well as the CF production of the 6.1 kDa α apoprotein of the light harvesting complex (LH1) from Rhodospirillum rubrum yielded 1.2 mg L–1 cell-culture and 0.7 mg mL–1 RM, respectively [15]. The CF obtained α-LH1 precipitate could be functionally refolded in a buffer system containing 0.5–2.0% TX-100. Recorded CD spectra were essentially identical to that of native α-LH1 protein and the in vitro formation of a structural complex with the β-subunit of the LH1 complex was demonstrated by its specific spectral absorbance pattern. CF expression of homologues of eukaryotic glutamate transporters from the four bacterial species E. coli, Aeropyrum pernix, Pyrococcus furiosus and P. horikoshii have been compared (Fig. 8.3). The approx. 45 kDa proteins have eight proposed TMSs and form putative trimers. All four proteins have been produced as precipitate and could subsequently be solubilized in the detergents zwittergent 3–12 or zwittergent 3–14. The highest expression level was obtained for the E. coli homologue (ca. 2 mg mL–1 RM) followed with approx. 30% of that yield from the homologue of P.

151

152

8 Cell-free Expression of Integral Membrane Proteins for Structural Studies

Fig. 8.3 Cell-free expression of bacterial glutamate transporter. CF expression of different eukaryotic glutamate transporter homologues as precipitate. After CF reaction, the IMP pellets were suspended in a volume equal to the RM and 5 µL samples were analyzed on a 12% SDS gel. Sizes of marker proteins (kDa) are indicated. Arrows indicate the synthesized glutamate transporter homologues.

furiosus. The two other proteins have been produced only at a significantly lower level in the CF system. The data demonstrate that the expression efficiencies of even closely related IMPs can be very different. Optimized codon usage of the target genes, supplementation of rare codon tRNAs or adjusting the amino acid pool according to the composition of the specific target proteins might be considered as further approaches to improve the expression yields. 8.4.2 G-Protein Coupled Receptors

Eukaryotic GPCRs represent the largest single class of receptors and they are predicted to consist of seven TMSs. Ligand binding induces conformational changes leading to the activation of cytoplasmatic G-proteins. GPCRs and GPCR-dependent signaling pathways of higher organisms have become prominent drug targets and there is a rapidly increasing demand for the identification of new GPCR ligand analogues. The structural characterization of these proteins is an essential prerequisite for directed drug screening approaches. However, GPCRs are generally of low natural abundance and efficient overexpression systems are therefore indispensable. Escherichia coli strains show in most cases only very low GPCR expression levels, with often less than 200 receptors per cell [29, 30]. A rare exception with up to 3500 copies per E. coli cell (resulting in 0.66 mg L–1 culture) of functional muscarinic m1 receptor has been reported [31]. However, the overproduced GPCRs generally accumulate in E. coli as inclusion bodies. The possibility of posttranslational modifications make yeast strains interesting for the overproduction of the eukaryotic GPCRs, and improved expression conditions resulted in the detection of several functionally synthesized GPCRs [32]. Reported expression levels were generally approx 1–2 pmol mg–1 membrane protein with some exceptions of 10 pmol mg–1 membrane protein in case of the serotonin 5-HT5A , β2 - and α2 adrenergic and endothelin B receptors [33]. The closest alternative to their native environment is the overexpression of GPCRs in mammalian or insect cells, espe-

8.4 Case Studies for the High Level Cell-free Expression of Membrane Proteins

cially when post-translational modifications are a prerequisite for functional studies. The expression levels usually do not exceed 5–10 pmol mg–1 membrane protein in adherent cell-lines and they are often lower in suspension cultures [34]. The human β2 adrenergic receptor (β2AR), the human M2 muscarinic acetylcholine receptor (M2) and the rat neurotensin receptor (NTR) could be synthesized in batch CF systems in amounts of 150–200 µg synthesized protein per mL RM [13]. The proteins had to be produced as translational fusions C-terminal to thioredoxin and almost no protein production was detected with the native non-fused coding regions. The authors speculated that stabilization of the GPCRs by the Nterminal thioredoxin or improved translation efficiencies might account for that observation. The final size of the recombinant IMPs was between 53 and 109 kDa and even the larger constructs were synthesized without significant loss of efficiency. In CECF systems, the final yield was improved to approx. 1 mg GPCR per mL RM within 8 hours of reaction. The CF-produced thioredoxin fused β2AR protein was reconstituted into preformed phospholipid vesicles by dialysis, and binding studies were performed with the radioactive labeled substrate, [3 H]dihydroalprenolol [13]. The calculated K d of the CF expressed β2AR was determined to 5.5 nm, which is comparable to the K d of 3.9 nm of β2AR isolated from Sf9 insect cells. Competition studies with non-labeled dihydroalprenolol resulted in EC50 s of 2.2 ×10−8 m with CF expressed β2AR and of 1.0 ×10−8 m with Sf9-expressed β2AR in one-site competition analyses. Up to 3 mg per mL RM can be obtained for the porcine vasopressin receptor type 2 (V2R) containing only a small amino-terminal T7 tag fusion in an individual CECF set-up [24]. CF expression without the 14 amino acid T7 tag fusion resulted only in spurious expression, which is in agreement with the above-mentioned observations with the CF expression of the GPCRs β2AR, M2 and NTR. The same observations were made for three further GPCR proteins, the human endothelin B receptor (ETB), the rat corticotropin releasing factor receptor (CRF) and the human V2R (Fig. 8.2). 8.4.3 β-Barrel Proteins

The outer membrane of Gram-negative bacteria forms a protective permeability barrier around the cells and serves as a molecular filter for hydrophilic substances. Outer membrane proteins (OMPs) often show the general architecture of membrane-spanning β-barrels, which are self-closed β-sheets. This topology contrasts the typical structure of inner membrane proteins with single or multiple membrane-spanning α-helices. Based on their transport mechanisms, OMPs can be divided into the classes of general porins, substrate-specific transporters and active transporters. The substrate-specific transporters contain low-affinity substratebinding sites that are saturable and allow efficient diffusion of substrates at very shallow concentration gradients. The E. coli nucleoside transporter Tsx belongs to this class [35, 36]. Tsx is, so far, the only CF-produced OMP and example of the high level production and functional reconstitution of this group of transport pro-

153

154

8 Cell-free Expression of Integral Membrane Proteins for Structural Studies

teins [24]. Tsx was expressed in standard CF set-ups at levels up to 4 mg per mL of RM either as precipitate or as soluble proteins in the presence of various detergents like Brij-35, Brij-78, TX-100 or others (Fig. 8.2). Conductance measurements of CF-produced Tsx in planar membranes indicated a clear dependence of activity on the mode of expression. Only soluble expressed Tsx in the presence of distinct detergents showed pore-forming activity. The high-level production of IMPs does not, therefore, always result in functionally folded protein. Solubilized Tsx precipitate did not show any activity and the efficient folding of soluble produced protein into its active conformation depended also on the presence of specific detergents like TX-100.

8.5 Structural Characterization of Cell-free Produced Membrane Proteins

IMPs are grossly underrepresented in the protein structural database. The rare and often inhomogeneous IMP samples obtained from eukaryotic in vivo expression systems with often incomplete post-translational modifications make structural determinations very difficult. CF expression could now provide access to otherwise difficult to obtain IMPs. The availability of high amounts of overproduced IMPs, the option of IMP expression directly into a detergent micelle with the desired properties, and many other possibilities to stabilize or to modify the synthesized IMPs, are valuable properties of the CF technique. The versatility, speediness and advantages of CF systems in the production of labeled IMPs promise, therefore, considerable progress for high-resolution structural analysis by NMR spectroscopy as well as by X-ray crystallography. 8.5.1 Crystallization of Cell-free Produced Membrane Proteins

The crystallization of membrane proteins is a complicated process due to nonpredictable influences of additives like detergents or lipids and it is mostly limited by the lack of sufficient amounts of protein. Over 25 000 protein structures have been solved by X-ray crystallography but only some 60 of them represent members of the family of IMPs. Recently, the first X-ray structure taking advantage of a CFproduced IMP was presented [16]. SeMet-labeled samples of the small multidrug transporter EmrE were produced in a commercially available CF batch system (Invitrogen, Carlsbad, USA) based on E. coli lysates. The RM contained 4.5 mm SeMet and was supplemented with DMPC (2 mg mL–1 ) and with substrates such as TPP+ . EmrE was produced as soluble protein in the presence of the preformed DMPC liposomes at an average of 0.2 mg per mL of RM. EmrE crystals used for structure determination and SeMet-EmrE-TPP+ crystals were grown using similar conditions. Analysis revealed an antiparallel EmrE homodimer with one bound TPP+ , which is consistent with other biochemical studies. The two subunits adopt slightly different tertiary folds, which seem to play an important role in substrate transport and

8.5 Structural Characterization of Cell-free Produced Membrane Proteins

H+ antiport. This first report demonstrates that the quality of CF-produced IMP samples is sufficient for crystallization and the technique might therefore attract increasing attention for sample preparation for X-ray analysis studies. 8.5.2 Cell-free Expression as a Tool for High-resolution NMR Spectroscopy

This section describes strategies in the sample preparation of IMPs by CF expression in combination with high-resolution NMR spectroscopy. Sample preparation has considerable influence on spectrum quality. An indispensable prerequisite for the NMR spectroscopy of larger proteins is the preparation of samples that have been labeled individually or with combinations of the stable isotopes 2 H, 13 C and 15 N. The full labeling of a protein requires the growth of the expression strain in defined medium supplemented with precursors that are labeled with the desired isotope. For amino acid specific labeling approaches, usually E. coli strains have to be employed containing auxotrophic mutations that correspond to the provided labeled amino acid. Conventional expression systems can cause considerable problems upon protein labeling, such as incomplete label incorporation, label scrambling, metabolic conversion of the labeled precursors or reduced yields if compared with the production of the unlabeled recombinant protein. Especially, combinatorial approaches in which several different amino acid types have to be labeled can become extremely difficult and inefficient. CF expression provides, generally, an interesting alternative tool for the fast and efficient production of labeled protein samples. Complete label incorporation is ensured as no unlabeled amino acids are present in the reaction. Moreover, no auxotrophic mutations are needed as any amino acid can be simply replaced in the reaction mixtures by a labeled derivative. This option is very attractive for combinatorial labeling approaches and protein derivatives with any label combinations can be generated with the same efficiencies as the unlabeled proteins as the overall reaction conditions do not have to be modified. Sample preparation with in vivo expression systems can, furthermore, take several days to weeks [37, 38] while CF reactions are completed within 12 hours. A further aspect is the preparation of highly deuterated protein samples to attenuate transverse relaxation rates and to improve sensitivity and resolution in NMR spectra. Deuterated protein samples are usually generated by growing the expression strains in almost pure D2 O supplemented with the required nutrients. However, state of the art NMR methodologies for protein backbone assignment rely on the back-exchange of 2 H to 1 H or at least on a high occupancy of protons at the amide positions of deuterated proteins during their isolation in aqueous solutions [39]. An incomplete back-exchange may lead to significant loss of signals and poor back-exchange may be aggravated for hydrophobic IMPs and, in particular, for residues in the hydrophobic TMS regions [40]. Therefore, unfolding and refolding of isolated proteins that have been produced in conventional expression systems by employing denaturants such as 6 m urea or guanidinium HCl are often necessary [41]. With CF expression, deuterated amino acids can be provided and all labile deuterated positions will already efficiently back-exchange in the aqueous re-

155

156

8 Cell-free Expression of Integral Membrane Proteins for Structural Studies

action mixtures before the amino acids become inserted into the growing peptide chain. A complete back-exchange of deuterated IMPs and other proteins at their amide positions is, therefore, generally ensured and no critical denaturation steps are necessary. To study protein structure and dynamics by NMR spectroscopy, the backbone of a protein must first be assigned. Given the limitations involved in the analysis of IMPs, certain strategies must be implemented for an efficient designation. A significant challenge in assignment inherent in NMR studies remains in the size limitation for proteins. Membrane proteins tend to be large with multiple domains, and the necessity for subsequent solubilization in detergent may further significantly increase the final size of the protein/micelle composite, leading to considerable line broadening due to restricted molecular tumbling. Fortunately, this problem can be alleviated to a certain extent by using suitable detergents that obviously allow relatively free movements of inserted protein. The detergent class of lyso-phosphoglycerols, including LMPG and LPPG, is especially suitable for this purpose [26, 42, 43]. Three E. coli IMPs, the cysteine exporter Yfik (21.2 kDa), the C-terminally truncated tellurite transporter TehA (24 kDa), and SecE (13.6 kDa), a component of the secretion machinery, have given promising NMR heteronuclear single quantum correlation (HSQC) spectra when dissolved in LMPG (Fig. 8.4A– C). In all cases, over 90% of the expected signals are detectable and backbone assignments as a prerequisite for their structural analysis appear, therefore, to be feasible. A limitation with most IMPs is their often high content of α-helical secondary structures, giving rise to poor chemical shift dispersion and leading to considerable spectral overlap compared with other structural elements such as β-sheets. However, spectral overlap especially, in the range 7.5–8.5 ppm, can be diminished significantly by the analysis of amino acid type specifically labeled samples (Fig. 8.4D–I). Limited spectral dispersion has further been addressed by the use of certain combinatorial labeling strategies combined with sample analysis in highfield spectrometers [26, 44, 45]. An important advancement is the development of the Transverse Relaxation Optimized Spectroscopy (TROSY) experiment, which reduces transverse relaxation and allows one to obtain high-resolution spectra for molecules with higher molecular weights [46]. One of the primary steps in assignment strategies is the measurement of 3D or 4D triple resonance (13 C/15 N-labeled) spectra and, often, protein backbones can be assigned by measuring the TROSYtype experiments, such as HNCA, HN(CO)CA, HNCACB, HN(CO)CACB, HNCO and HN(CA)CO, which allow the sequential identification of backbone residues. The 24 kDa truncated version of the putative bacterial tellurite transporter TehA with seven TMSs was the first IMP whose backbone assignment was approached by taking advantage of CF expression strategies [26]. Standard non-selective triple resonance experiments resulted only in approx. 55% assignment and high degrees of signal overlap in the α-helical IMP required further strategies for the full backbone assignment. The analysis of ten further samples having each a different amino acid type specifically labeled resulted only in approx. 10% additional assignments. This low rate of success was accredited to redundancy, such that in many cases an identified amino acid type was associated with more than one N- or C-terminal

8.5 Structural Characterization of Cell-free Produced Membrane Proteins

Fig. 8.4 Liquid-state NMR of cell-free

produced transporters. 1 H-15 N TROSY-HSQC spectra of CF expressed 15 N labeled IMPs. The proteins were expressed as precipitates and solubilized in LMPG. (A) U-15 N labeled SecE; (B) U-15 N labeled YfiK; (C) U-15 N labeled TehA; (D) 15 N-threonine labeled TehA; (E) 15 N-tryptophan labeled TehA; (F) 15 N-isoleucine labeled TehA; (G) 15 N-methionine labeled TehA; (H) 15 N-phenylalanine labeled TehA; and (I) 15 N-arginine labeled TehA. The protein precipitates were dissolved in 25 mm

potassium phosphate buffer (pH 6.0) containing 5% LMPG (A, B, D–I) or dissolved in LMPG, purified by Ni-NTA chromatography and equilibrated in 20 mm MES/Bis-Tris (pH 6.0) with 2% LMPG (C). Spectra were taken with protein concentrations of 1 mm for SecE, 1 mm for YfiK, 0.1 mm for TehA and 0.9 mm for selective TehA samples (D–I) at 40 ◦ C on a (A) 600, (B) 700, (C) 900, or (D–I) 800 MHz NMR spectrometer – all equipped with cryogenic 1 H [13 C/15 N] triple-resonance probes.

connectivity, leading to ambiguous sequence assignments. Of much higher success was the application of a combinatorial selective labeling method that relies on the 13 C labeling of certain amino acid types concomitant with the simultaneous 15 N labeling of others [44]. In this approach, three samples were produced having each two to three different 15 N labeled amino acids types together with two 13 C labeled types (Fig. 8.5). The rationale behind this strategy is that by using this combinatorial base together with 2D versions of the HNCO experiment and 15 N/1 H TROSY spectra, 15 N-labeled amino acids types that are preceded by

157

158

8 Cell-free Expression of Integral Membrane Proteins for Structural Studies

Fig. 8.5 Combinatorial labeling approach. Example of a combinatorial labeling scheme with three differentially 15 N and 13 C labeled samples of the tellurite transporter TehA. The labeling scheme is shown in the table on top. In the [15 N, 1 H]-TROSY spectra, backbone amide protons of the 15 N-labeled amino acids are visible. The crossed lines indicate the peak to be identified. The indicated peak at 117.5/7.78 is identified as phenylalanine as it is present in samples 1 and 2, but not in 3. The corresponding HNCO 13 C-labeled

spectra show the amide crosspeaks after carbonyl transfer (large arrow) and indicate that the preceding residue of this phenylalanine must be a glycine, as no crosspeaks are visible in samples 2 and 3 without 13 C-glycine. The analyzed phenylalanine residue can now be localized in a Gly-Phe pair in the primary sequence, which in this example was identified as Phe97 of TehA. All spectra were recorded on a Bruker Avance 600 MHz spectrometer.

types can readily be selected (Fig. 8.5). If the identified combination of two amino acids in the protein has a singular occurrence, the specific site can be unambiguously assigned and may act as an anchor point for further sequential assignments. It is predicted that 40–50% of all amino acids within a protein are part

8.5 Structural Characterization of Cell-free Produced Membrane Proteins

of a unique amino acid pair, and algorithm programs that assist with the selection of amino acid types to be used for such combinatorial specific labeling schemes have been made available (http://www.Biophyschem.uni-frankfurt.de/AKDoetsch/ projects/download/combilabel.m). In combination with the information obtained from the non-selective and from the amino acid specific labeled samples, the combinatorial approach afforded a total of 85% assignments, which is sufficient for structural analysis. However, data re-analysis revealed that the combinatorial approach alone, along with the nonselective labeling experiments, would have been enough to yield these 85% assignment of TehA. Expensive and time consuming single selective labeling experiments might therefore not be necessary for future assignments of IMPs. Accordingly, five combinatorial 15 N labeled samples of the C-terminal 16 kDa domain of the τ subunit of the E. coli DNA polymerase III holoenzyme yielded the same information as derived from 19 individual amino acid type selectively labeled samples [43]. Given the considerable degree of signal overlap in IMPs, the use of CF expression systems for the generation of combinatorial and non-specifically labeled samples in combination with 15 N-HSQC-TROSY spectra and their analysis by the differential HNCO strategy might significantly assist backbone assignments in the near future. 8.5.3 Applications of Cell-free Expression for Solid-state NMR

Solid-state NMR (ssNMR), like the more frequently used solution-state NMR, offers the possibility to study the atomic structure, exchange processes, spin diffusion and molecular dynamics of proteins. Additional information is provided on torsion angles, atomic orientation and very precise internuclear distances up to 15 Å by using the available chemical shift anisotropy, quadrupole coupling and strong dipolar couplings. The ssNMR technique is very versatile as no inherent size limit on molecules for investigation exists. However, ssNMR is hampered by fast relaxation, insensitivity (>1 µmol protein are required for 2D spectroscopy) and broader line-width than observed with solution-state NMR. Well-resolved spectra containing only selected interactions can nevertheless be obtained using magic angle sample spinning (MAS) and recoupling techniques. These remove anisotropic interactions by fast sample rotation (in practice 5–20 kHz) about the magic angle (54.7 ◦ with respect to the magnetic field) and reintroduce selectively desired interactions using appropriate pulse sequences [47]. ssNMR enables the study of proteins in a multitude of different states, including frozen solution [48], protein aggregates [49], 2D crystals [50], 3D crystals [51] and in proteoliposomes [52]. Proteoliposomes are clearly the samples of choice for IMPs as they are closest to the proteins native environment. Lipid reconstituted samples also offer the option of investigating the dynamics of an IMP upon ligand binding [53] or in response to changes in the lipid environment [54]. One of the key problems in the study of uniformly labeled large IMPs is spectral overlap. However, ssNMR is not only focused on structural biology but offers unique possibilities for

159

160

8 Cell-free Expression of Integral Membrane Proteins for Structural Studies

Fig. 8.6 Solid-state NMR spectroscopy of a cell-free produced multidrug transporter. (I) 15 N-cross polarization (CP)-MAS spectra of 15 N-Leu-EmrE directly after CF synthesis (A), after solubilization in DDM (B) and of 15 N-Phe-Hsmr after CF expression, solubilization in DDM and reconstitution into E. coli lipid liposomes (C). Experiments were performed with a 5 kHz sample spinning rate at 253 K using a Bruker 7-mm MAS probe at 40.5 MHz for EmrE, and with a 8 kHz sample spinning rate at 230 K using a Bruker 4-mm MAS probe at 60.84 Hz for Hsmr. The CP contact time was 1.5 ms and 50 000 scans were accumulated; for Hsmr the CP contact

time was 0.75 ms and 30 000 scans were accumulated. (II) 13 C-CP-MAS spectra of 13 C-Glu-EmrE in DOPC. (A) After reconstitution into DOPC, most resonances are obscured by the overall naturally abundant 13 C signals of lipid and protein. (B) Applying double-quantum filtering techniques, the natural abundance background can be suppressed, allowing selective observation of the labeled sites. (Reproduced with permission from M. Lorch, I. Lehner, A., Siarheyeva, D. Basting, N. Pfleger, T. Manolikas, C. Glaubitz, Biochem. Soc. Trans., 2005, 33, 873–877.)

functional biophysical studies and it has been widely used to study membranebound peptides. Such studies require efficient residue- and site-selective labeling of, for example, active sites in IMPs. For these labeling schemes, the CF expression system offers the, so far, largest possible flexibility. Figure 8.6 shows preliminary ssNMR spectra of CF-expressed and amino acid selectively labeled SMR proteins. CF generated 15 N amino acid specifically labeled samples of EmrE and its homologue Hsmr have been prepared as precipitate, frozen DDM micelles and as reconstituted protein in E. coli lipid mixture and analyzed by ssNMR (Fig. 8.6I). The best resolution was obtained of the reconstituted Hsmr protein and some signals of individual amino acid residues are already visible. Among SMR proteins the residue Glu14 is highly conserved and suggested to be involved in substrate binding [28]. A 13 C glutamate selective labeling of EmrE was performed by CF expression to monitor the active site (Fig. 8.6II). The protein was reconstituted in 1,2-dioleoyl-snglycero-3-phosphocholine (DOPC) liposomes and signal overlap due to the natural abundance of 13 C signals from the lipid and from side chains of the protein could be reduced by applying specific double-quantum filtering techniques. Only signals

Abbreviations

of the labeled Glu residues remain visible and possible applications of this approach are the observation of the glutamate protonation state and its modulation by substrate binding as well as precise distance measurements between the active site glutamate and bound substrates.

Abbreviations

Brij-35, polyoxyethylene-(23)-lauryl-ether Brij-56, polyoxyethylene-(10)-cetyl-ether Brij-58, polyoxyethylene-(20)-cetyl-ether Brij-78, polyoxyethylene-(20)-stearyl-ether Brij-97, polyoxyethylene-(10)-oleyl-ether Brij-98, polyoxyethylene-(20)-oleyl-ether β-OG, n-octyl-β-glucopyranoside CD, circular dichroism CECF, continuous exchange cell-free CF, cell-free CHAPS, 3-[(3-cholamidopropyl)dimethylammonio]-1-propansulfonate c.m.c., critical micellar concentration C mic , micellar concentration CP, cross polarization DHPC, 1,2-diheptanoyl-sn-glycero-3-phosphocholine diC8 PC, 1,2-dioctanoyl-sn-glycero-3-phosphocholine DM, n-decyl-β-d-maltoside DDM, n-dodecyl-β-d-maltoside DMPC, 1,2-dimyristoyl-sn-glycero-3-phosphocholine DOPC, 1,2-dioleoyl-sn-glycero-3-phosphocholine DPC, dodecyl-phosphocholine FM, feeding mixture GPCR, G-protein coupled receptor HSQC, heteronuclear single quantum correlation IMP, integral membrane protein LMPG, 1-myristoyl-2-hydroxy-sn-glycero-3-[phospho-rac-(1-glycerol)] LPPG, 1-palmitoyl-2-hydroxy-sn-glycero-3-[phospho-rac-(1-glycerol)] MAS, magic angle sample spinning NMR, nuclear magnetic resonance OMP, outer membrane protein RM, reaction mixture SDS, sodium dodecyl sulfate ssNMR, solid-state NMR Thesit, polyethylene-glycol 400 dodecyl-ether TMS, transmembrane segment TPP+ , tetraphenylphosphonium TROSY, transverse relaxation optimized spectroscopy

161

162

8 Cell-free Expression of Integral Membrane Proteins for Structural Studies

TX-100, 4-(1,1,3,3-tetramethyl-butyl)phenyl-polyethylene glycol (Triton X-100) Tween 20, polyoxyethylene-sorbitan-monolaurate 20.

References 1 Martemyanov, K. A., Shirokov, V. A., Kurnasov, O. V., Gudkov, A. T., Spirin, A. S. (2001). Cell-free production of biologically active polypeptides: Application to the synthesis of antibacterial peptide cecropin. Protein Expr. Purif. 21, 456–461. 2 Alakov, Y. B., Baranov, V. I., Ovodov, S. J., Ryabova, L. A., Spirin, A. S., Morozov, I. J. (1995). Method of preparing polypeptides in cell-free translation system in U.S. Pat. 5,478,730. 3 Kim, D. M., Choi, C. Y. (1996). A semicontinuous prokaryotic coupled transcription/translation system using a dialysis membrane. Biotechnol. Prog. 12, 645–649. 4 Kigawa, T., Yabuki, T., Yoshida, Y., Tsutsui, M., Ito, Y., Shibata, T., Yokoyama, S. (1999). Cell-free production and stableisotope labeling of milligram quantities of proteins. FEBS Lett. 442, 15–19. 5 Kim, D. M., Swartz, J. R. (1999). Prolonging cell-free protein synthesis with a novel ATP regeneration system. Biotechnol. Bioeng. 66, 180–188. 6 Kim, D. M., Swartz, J. R. (2000). Oxalate improves protein synthesis by enhancing ATP supply in a cell-free system derived from Escherichia coli. Biotechnol. Lett. 22, 1537–1542. 7 Kim, D. M., Swartz, J. R. (2000). Prolonging cell-free protein synthesis by selective reagent additions. Biotechnol. Prog. 16, 385–390. 8 Jewett, M. C., Swartz, J. R. (2004). Mimicking the Escherichia coli cytoplasmic environment activates long-lived and efficient cell-free protein synthesis. Biotechnol. Bioeng. 86, 19–26. 9 Jewett, M. C., Swartz, J. R. (2004). Substrate replenishment extends protein synthesis with an in vitro translation system designed to mimic the cytoplasm. Biotechnol. Bioeng. 87, 465–472. 10 Michel-Reydellet, N., Calhoun, K., Swartz, J. (2004). Amino acid stabilization for cell-free protein synthesis by modi-

11

12

13

14

15

16

17

18

fication of the Escherichia coli genome. Metab. Eng. 6, 197–203. Klammt, C., Lohr, F., Schafer, B., Haase, W., Dotsch, V., Ruterjans, H., Glaubitz, C., Bernhard, F. (2004). High level cell-free expression and specific labeling of integral membrane proteins. Eur. J. Biochem. 271, 568–580. Torizawa, T., Shimizu, M., Taoka, M., Miyano, H., Kainosho, M. (2004). Efficient production of isotopically labeled proteins by cell-free synthesis: a practical protocol. J. Biomol. NMR 30, 311–325. 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. Berrier, C., Park, K. H., Abes, S., Bibonne, A., Betton, J. M., 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. Shimada, Y., Wang, Z. Y., Mochizuki, Y., Kobayashi, M., Nozawa, T. (2004). Functional expression and characterization of a bacterial light-harvesting membrane protein in Escherichia coli and cell-free synthesis systems. Biosci. Biotechnol. Biochem. 68, 1942–1948. Pornillos, O., Chen, Y. J., Chen, A. P., Chang, G. (2005). X-ray structure of the EmrE multidrug transporter in complex with a substrate. Science 310, 1950–1953. Kigawa, T., Yabuki, T., Matsuda, N., Matsuda, T., Nakajima, R., Tanaka, A., Yokoyama, S. (2004). Preparation of Escherichia coli cell extract for highly productive cell-free protein expression. J. Struct. Funct. Genomics 5, 63–68. Ozawa, K., Jergic, S., Crowther, J. A., Thompson, P. R., Wijffels, G., Otting, G., Dixon, N. A. (2005). Cell-free protein synthesis in an autoinduction system for

References

19

20

21

22

23

24

25

26

27

28

NMR studies of protein-protein interactions. J. Biomol. NMR 32, 235–241. Zubay, G. (1973). In vitro synthesis of protein in microbial systems. Annu. Rev. Genet. 7, 267–287. Liu, D. V., Zawada, J. F., Swartz, J. R. (2005). Streamlining Escherichia coli S30 extract preparation for economical cellfree protein synthesis. Biotechnol. Prog. 21, 460–465. Yamane, T., Ikeda, Y., Nagasaka, T., Nakano, H. (2005). Enhanced cell-free protein synthesis using a S30 extract from Escherichia coli grown rapidly at 42 degrees C in an amino acid enriched medium. Biotechnol. Prog. 21, 608–613. Endo, Y., Sawasaki, T. (2003). Highthroughput, genome-scale protein production method based on the wheat germ cell-free expression system. Biotechnol. Adv. 21, 695–713. 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. Klammt, C., Schwarz, D., Fendler, K., Haase, W., Dotsch, V., Bernhard, F. (2005). Evaluation of detergents for the soluble expression of alpha-helical and beta-barrel-type integral membrane proteins by a preparative scale individual cell-free expression system. FEBS J. 272, 6024–6038. Garavito, R. M., Ferguson-Miller, S. (2001). Detergents as tools in membrane biochemistry. J. Biol. Chem. 276, 32403– 32406. Trbovic, N., Klammt, C., Koglin, A., Lohr, F., Bernhard, F., Dotsch, V. (2005). Efficient strategy for the rapid backbone assignment of membrane proteins. J. Am. Chem. Soc. 127, 13504–13505. Winstone, T. L., Duncalf, K. A., Turner, R. J. (2002). Optimization of expression and the purification by organic extraction of the integral membrane protein EmrE. Protein Expr. Purif. 26, 111–121. Gutman, N., Steiner-Mordoch, S., Schuldiner, S. (2003). An amino acid cluster around the essential Glu-14 is part of the substrate- and proton-binding domain of EmrE, a multidrug transporter

29

30

31

32

33

34

35

36

37

38

39

from Escherichia coli. J. Biol. Chem. 278, 16082–16087. Massotte, D. (2003). G protein-coupled receptor overexpression with the baculovirus-insect cell system: a tool for structural and functional studies. Biochim. Biophys. Acta 1610, 77–89. Sarramegna, V., Talmont, F., Demange P., Milon, A. (2003). Heterologous expression of G-protein-coupled receptors: comparison of expression systems fron the standpoint of large-scale production and purification. Cell Mol. Life Sci. 60, 1529–1546. Hulme, E. C., Curtis, C. A. (1998). Purification of recombinant M1 muscarinic acetylcholine receptor. Biochem. Soc. Trans. 26, S361. André, N., Cherouati, N., Prual, C., Steffan, T., Zeder-Lutz, G., Magnin, T., Pattus, F., Michel, H., Wagner, R., Reinhart, C. (2006). Enhancing functional production of G protein-coupled receptors in Pichia pastoris to levels required for structural studies via a single expression screen. Protein Sci. 15, 1115–1126. Reilander, H., Weiss, H. M. (1998). Production of G-protein-coupled receptors in yeast. Curr. Opin. Biotechnol. 9, 510–517. Stanasila, L., Pattus, F., Massotte, D. (1998). Heterologous expression of Gprotein-coupled receptors: human opioid receptors under scrutiny. Biochimie 80, 563–571. Bannwarth, M., Schulz, G. E. (2003). The expression of outer membrane proteins for crystallization. Biochim. Biophys. Acta 1610, 37–45. Ye, J., van den Berg, B. (2004). Crystal structure of the bacterial nucleoside transporter Tsx. EMBO J. 23, 3187–3195. Kigawa, T., Muto, Y., Yokoyama, S. (1995). Cell-free synthesis and amino acidselective stable isotope labeling of proteins for NMR analysis. J. Biomol. NMR 6, 129–134. Ozawa, K., Dixon, N. E., Otting, G. (2005). Cell-free synthesis of 15N-labeled proteins for NMR studies. IUBMB Life 57, 615–622. Tugarinov, V., Muhandiram, R., Ayed, A., Kay, L. E. (2002). Four-dimensional NMR spectroscopy of a 723-residue protein: chemical shift assignments and sec-

163

164

8 Cell-free Expression of Integral Membrane Proteins for Structural Studies

40

41

42

43

44

45

46

47

ondary structure of malate synthase. J. Am. Chem. Soc. 124, 10025–10035. Chill, J. H., Louis, J. M., Miller, C., Bax, A. (2006). NMR study of the tetrameric KcsA potassium channel in detergent micelles. Protein Sci. 15, 684–698. Tugarinov, V., Hwang, P. M., Kay, L. E. (2004). Nuclear magnetic resonance spectroscopy of high-molecular-weight proteins. Annu. Rev. Biochem. 73, 107–146. Krueger-Koplin, R. D., Sorgen, P. L., Krueger-Koplin, S. T., Rivera-Torres, I. O., Cahill, S. M., Hicks, D. B., Grinius, L., Krulwich, T. A., Girvin, M. E. (2004). An evaluation of detergents for NMR structural studies of membrane proteins. J. Biomol. NMR 28, 43–57. Roosild, T. P., Greenwald, J., Vega, M., Castronovo, S., Riek, R., Choe, S. (2005). NMR structure of Mistic, a membraneintegrating protein for membrane protein expression. Science 307, 1317–1321. Parker, MJ., Aulton-Jones, M., Hounslow, A. M., Craven, C. J. (2004). A combinatorial selective labeling method for the assignment of backbone amide NMR resonances. J. Am. Chem. Soc. 126, 5020–5021. Wu, P. S., Ozawa, K., Jergic, S., Su, X. C., Dixon, N. E., Otting, G. (2006). Aminoacid type identification in 15N-HSQC spectra by combinatorial selective 15Nlabelling. J. Biomol. NMR 34, 13–21. Pervushin, K., Riek, R., Wider, G., Wuthrich, K. (1997). Attenuated T2 relaxation by mutual cancellation of dipole-dipole coupling and chemical shift anisotropy indicates an avenue to NMR structures of very large biological macromolecules in solution. Proc. Natl. Acad. Sci. U.S.A. 94, 12366–12371. Laws, D. D., Bitter, H. M., Jerschow, A. (2002). Solid-state NMR spectroscopic methods in chemistry. Angew. Chem. Int. Ed. 41, 3096–3129.

48 Luca, S., White, J. F., Sohal, A. K., Filippov, D. V., van Boom, J. H., Grisshammer, R., Baldus, M. (2003). The conformation of neurotensin bound to its G protein-coupled receptor. Proc. Natl. Acad. Sci. U.S.A. 100, 10706–10711. 49 Ritter, C., Maddelein, M. L., Siemer, A. B., Luhrs, T., Ernst, M., Meier, B. H., Saupe, S. J., Riek, R. (2005). Correlation of structural elements and infectivity of the HET-s prion. Nature 435, 844–848. 50 Hiller, M., Krabben, L., Vinothkumar, K. R., Castellani, F., van Rossum, B. J., Kuhlbrandt, W., Oschkinat, H. (2005). Solid-state magic-angle spinning NMR of outer-membrane protein G from Escherichia coli. Chembiochem 6, 1679–1684. 51 Lorch, M., Fahem, S., Kaiser, C., Weber, I., Mason, A. J., Bowie, J. U., Glaubitz, C. (2005). How to prepare membrane proteins for solid-state NMR: A case study on the alpha-helical integral membrane protein diacylglycerol kinase from E. coli. Chembiochem 6, 1693–1700. 52 Mason, A. J., Siarheyeva, A., Haase, W., Lorch, M., van Veen, H. W., Glaubitz, C. (2004). Amino acid type selctive isotope labelling of the multidrug ABC transporter LmrA for solid-state NMR studies. FEBS Lett. 568, 117–121. 53 Patching, S. G., Brough, A. R., Herbert, R. B., Rajakarier, J. A., Henderson, P. J., Middleton, D. A. (2004). Substrate affinities for membrane transport proteins determined by 13C cross-polarization magic-angle spinning nuclear magnetic resonance spectroscopy. J. Am. Chem. Soc. 126, 3072–3080. 54 Yamaguchi, S., Tuzi, S., Bowie, J. U., Saito, H. (2004). Secondary structure and backbone dynamics of Escherichia coli diacylglycerol kinase, as revealed by sitedirected solid-state 13C NMR. Biochim. Biophys. Acta 1698, 97–105.

165

9

Cell-free Production of Membrane Proteins in the Presence of Detergents Jean-Michel Betton and Marika Miot

9.1 Introduction

Integral membrane proteins, including enzymes, receptors and transporters, play essential and functionally diverse roles in living cells. Analysis of the complete genomic sequences for several organisms indicate that 25–30% of the genes may encode membrane-spanning α helix proteins [1]. However, these predominant membrane proteins account for less than 1% of the three-dimensional structures in the Protein Data Bank [2]. The major difficulty impeding structural analysis of membrane proteins is the lack of generally applicable systems for their overproduction in quantities sufficient for crystallization studies. Indeed, overexpression of genes encoding membrane proteins in Escherichia coli, the most widely used host for high-level production of recombinant proteins, results in the formation of inclusion bodies and/or toxic effects leading to cell death [3]. The conventional explanation for this toxicity suggests that E. coli, like most cellular hosts, lacks the capacity to integrate an extra protein load into its cytoplasmic membrane. The biogenesis of α-helical membrane proteins requires precise coordination between a membrane-embedded translocon, a protein-conducting channel, and a translating ribosome to correctly insert transmembrane (TM) helices into the lipid bilayer [4]. Once inserted, these TM helices interact within the membrane to form tertiary and/or quaternary structure. Based on the observation that hydrophobic TM helices are generally independently stable in the lipid bilayer, the folding of αhelical membrane proteins has been postulated to follow a two-state model, which recently has been modified to include a third state to allow for the incorporation of non TM helix elements [5]. However, this simple model may not fully consider that other cellular factors like the translocon could regulate the recognition, orientation and insertion of TM helices and facilitate protein folding at the membrane level [6]. Although still far from clear, the mechanistic analysis of membrane protein biogenesis is most advanced for the translocation process of TM helices. Indeed, structural and functional studies of the heterotrimeric Sec (SecY in prokaryotes or Sec61 in eukaryotes) translocon suggest that a TM helix leaves the protein-conducting channel by a lateral exit gate when another TM helix enters the translocon [7, 8]. In vivo, Cell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

166

9 Cell-free Production of Membrane Proteins in the Presence of Detergents

these coordinated inter- and intraprotein interactions determine both the topology and folding of membrane proteins. Thus, high cellular levels of recombinant membrane proteins could quickly saturate the translocation machinery and competitively exclude other essential host membrane proteins. Recent progress in cell-free expression systems has been made to improve the protein yield up to several milligrams [9]. However, the main property of these in vitro systems is their open design, which permits a direct manipulation of the biosynthesis reactions, notably by the addition of exogenous molecules [10]. Indeed, recent reports indicate that soluble membrane proteins can be efficiently and functionally produced in cell-free systems supplemented with detergents [11–14]. These promising studies imply that, in absence of translocons, some integral membrane proteins may spontaneously insert, during or after translation, into detergent micelles and correctly fold or assemble into a functional structure. This chapter describes such an attractive strategy and evaluates the production of two structurally different membrane histidine protein kinases.

9.2 Histidine Protein Kinases

Histidine protein kinases (HPKs) are the receptor in protein phosphotransfer by the two-component systems, and play a central role in signaling pathways required for environmental adaptation of prokaryotes and lower eukaryotes [15]. The large majority of HPKs are homodimeric membrane proteins in which each subunit contains a short N-terminal cytoplasmic fragment followed by a transmembrane α helix and an extracellular or periplasmic sensing domain connected via a second membrane-spanning α helix to a C-terminal cytoplasmic kinase domain. In some cases, the N-terminal region of HPKs contains several TM helices that would be expected to form a distinct hydrophobic domain embedded within the membrane bilayer. These latter HPKs appear to be the prokaryotic correlates of the seven-transmembrane G-protein-coupled receptors that mediate various different sensory responses in eukaryotic cells. Whatever the membrane topology of their sensory input domains, all HPKs share a common output mechanism consisting of the ATP-dependent phosphorylation of a specific histidine residue and subsequent transfer of the phosphoryl group to an aspartate residue in the receptor domain of a cognate response regulator (RR) protein, usually a transcription factor. The response is proportional to the degree of RR phosphorylation, which depends not only on the efficiency of the autokinase and transfer reactions but also, in many cases, on an intrinsic phosphatase activity of the HPK. Thus, the phosphoaccepting histidine residue of HPKs participates directly in autokinase, phosphotransferase and phosphatase activity. Because of their hydrophobic TM helices, biochemical studies of membrane HPKs have frequently been restricted to their soluble cytoplasmic domains, and no structural or functional information is available for signal perception and transmission of full-length HPKs [16]. Therefore, two different full-length HPKs, the CpxA and DesK proteins, from E. coli and Bacillus sub-

9.2 Histidine Protein Kinases

Fig. 9.1 TMHMM analysis of CpxA and DesK membrane proteins.

The 457 and 370 amino acid residues of CpxA (A) and DesK (B), respectively, are indicated on the x-axis. The dashed vertical lines indicate predicted transmembrane regions. Designations of the different TM helices are shown.

tilis, respectively, were chosen to evaluate their cell-free production in the presence of detergents. The CpxA protein is activated by extracytoplasmic stress, including the presence of envelope misfolded proteins, and regulates via CpxR, its cognate response regulator, the expression of several genes coding mainly for periplasmic folding factors [17]. The DesK protein is activated by a decrease in membrane lipid fluidity, and phosphorylates DesR, a transcription factor that regulates expression of the des gene coding for a 5 acyl lipid desaturase [18]. The predicted membrane topologies of CpxA and DesK, using the hidden Markov model topology program TMHMM [1], are shown in Fig. 9.1A and B, respectively. While CpxA displays two predicted TM helices, consistent with the canonical topology of HPKs, at least, five TM helices were predicted for DesK. Therefore, although both HPKs contain a similar cytoplasmic histidine kinase domain, their transmembrane and extracytoplasmic sensor domains are structurally very different.

167

168

9 Cell-free Production of Membrane Proteins in the Presence of Detergents

9.3 Materials and Methods 9.3.1 Plasmids

The cpxA gene from E. coli and desK from Bacillus subtilis were cloned under the control of T7 promoter into pIVEX plasmids. Forward and reverse PCR primers were designed to introduce NdeI and XmaI sites upstream and downstream of both genes, respectively. The two PCR products were subcloned into two different pIVEX vectors (pIVEX2.4d and pIVEX2.3d, Roche) which allow addition of a polyhistidine (His6 ) tag either at the N- or the C-terminus of the corresponding proteins. Plasmid pIV3-MalE, carrying the wild-type malE gene without signal sequence, has been described previously [19]. Plasmid pIVEX-GFP, encoding the green fluorescent protein (GFP) cycle 3 variant, was provided by C. Nemetz from Roche. 9.3.2 Cell-free Protein Production

Batch and continuous exchange cell-free (CECF) reactions were performed in the RTS100 and RTS500 E. coli lysates, respectively, as described in the Instruction Manual from Roche. For analytical experiments in RTS100, the reaction was initiated by adding 0.5 µg of pIVEX plasmid DNA, and protein production after 2 h at 30 ◦ C was detected by immunoblotting with anti-His6 monoclonal antibody conjugated with peroxidase (Roche) using diaminobenzidine and H2 O2 as substrates. Protein solubility was determined by collecting a total protein sample from the cell-free reaction mixture and a soluble protein sample from the supernatant after centrifugation (20 000g for 10 min). For preparative experiments in the RTS500, the reaction was initiated by adding 15 µg of pIVEX plasmid DNA and 1% Brij35, and protein production after 20 h at 30 ◦ C was analyzed by SDS-PAGE followed by Coomassie blue staining. 9.3.3 Protein Purification

Membrane proteins were purified by immobilized-metal affinity chromatography (IMAC) at 4 ◦ C. The CECF reaction mixtures (1 mL) were diluted with the same volume of phosphate buffer (50 mm Na2 HPO4 , 0.3 m NaCl at pH 8), containing 0.5% Brij35 and 20 mm imidazole, and incubated for 30 min at 4 ◦ C. After centrifugation for 15 min at 20 000g, supernatants were loaded onto Ni-NTA columns (1 mL) equilibrated in the same buffer. Following extensive washes, proteins were eluted with the phosphate buffer containing 0.5% Brij35 and 250 mm imidazole. The MalE protein was purified by affinity chromatography on an amylose resin (Biolabs), as previously described [20]. After purification, all protein samples were

9.4 Results and Discussion

analyzed by SDS-PAGE with Coomassie blue staining, and protein concentrations were determined from absorbance at 280 nm with the following extinction coefficients (m–1 cm–1 ): 65 430 for CpxA, 49 390 for DesK, and 68 750 for MalE. 9.3.4 Structural and Functional Protein Characterizations

Electrospray ionization mass spectrometry was performed with an API 365 triplequadrupole mass spectrometer (ABI-MDS-SCIEX) on purified MalE proteins (0.1 mg mL–1 ) dissolved in water–methanol–formic acid (50:50:5, v/v/v). Differential scanning calorimetry (DSC) endotherms were measured with an VP-DSC calorimeter (MicroCal). Purified MalE proteins, extensively dialyzed against 50 mm glycine buffer at pH 8, were loaded (0.5 mg) into the calorimetric cell and heated at a constant scan rate of 1 ◦ C min–1 . Data analysis was performed with the software provided by the manufacturer. Dynamic light scattering (DLS) and far ultraviolet circular dichroism (far UV-CD) analysis were performed at 25 ◦ C on a DynaPro 800 (Protein Solutions) and an Aviv 215 spectropolarimeter (AvivBiomedical), respectively, with protein solutions at 0.5 mg mL–1 in 10 mm sodium phosphate, 0.5% Brij35, pH 7.5 buffer. Quantitative analysis of far-UV-CD spectra was carried out with the CONTIN/LL method included in the CDPro package [21]. Autokinase activity of purified CpxA and DesK proteins, at a final concentration of 5 µm, were assayed in Tris buffer (50 µm Tris-HCl, pH 8, 0.1 m NaCl, 1 mm MgCl2 , 0.5% Brij35 at 30 ◦ C). Reactions were initiated by adding ATP solution (50 µm ATP and 5 µCi of [γ -33 P]ATP), and aliquots were withdrawn at various time points. Then, the samples were separated by SDS-PAGE and dried gels analyzed by autoradiography.

9.4 Results and Discussion 9.4.1 Analytical Cell-free Production of His6 -tagged Proteins

To facilitate the purification of both membrane proteins, a His6 tag was introduced by subcloning the corresponding PCR product into the pIVEX plasmids. Since tag fusion can alter expression level and/or protein activity, two expression plasmids were constructed with each gene, producing proteins with a His6 tag either at their N-termini (pIV4-CpxA and pIV4-DesK) or at their C-termini (pIV3-CpxA and pIV3DesK). Furthermore, previous studies have reported that the presence of mRNA structures, such as hairpin loops in the translation initiation region, could inhibit the binding of ribosomes, thus limiting the yield of cell-free protein production [22]. However, such problems can be simply circumvented by testing several different constructions, to introduce a variation in the translation initiation region [19]. The pIVEX plasmids greatly facilitated this strategy, since multiple cloning sites

169

170

9 Cell-free Production of Membrane Proteins in the Presence of Detergents

Fig. 9.2 Analysis of cell-free production of CpxA and DesK using batch technology. After synthesis, whole protein samples were fractionated by 12% SDS-PAGE, and the gel was stained with Coomassie blue (A) or transferred and subjected to immunoblotting with anti-His6 antibodies coupled to

peroxidase (B). Lanes M, molecular weight marker; lanes 1, empty pIVEX2.4d; lanes 2, pIV3-DesK; lanes 3, pIV4-DesK; lanes 4, pIV3-CpxA; lanes 5, pIV4-CpxA. Arrows indicate the location of the Desk and CpxA proteins.

were designed to allow tag fusion, either at the N- or C-terminal of a target protein, by conserving the same restriction site and reading frame compatibility. The study began by testing protein yields in batch technology without detergents. When analyzed by SDS-PAGE followed by Coomassie blue staining, the steadystate production of the four His6 -tagged proteins is detectable at a similar level (Fig. 9.2A). While a single band at the expected size is resolved to both CpxA proteins, only aggregated species were visible at the top of the gel for DesK proteins. Similar behavior was observed with DesK produced in E. coli (D. Albanesi, personal communication). However, the immunoblot with anti-His6 antibodies revealed only proteins produced from pIV4-CpxA and pIV3-DesK plasmids (Fig. 9.2B). A single band with an approximate molecular mass of 50 kDa was observed for Ntagged CpxA, whereas multiple immunoreactive bands were resolved to C-tagged DesK, which is consistent with monomeric, dimeric and oligomeric species. These observations suggest that antibody recognition of the His6 -tag at the C-terminus of CpxA or at the N-terminus of DesK can be strongly hampered by protein conformation. Although other factors like accessibility or protein degradation may be invoked, we have frequently observed from a larger set of expression constructs that poor immunoblotting detection of His6 -tagged proteins is associated with their weak binding on IMAC column under native conditions. The functional properties of affinity tag fusions should, therefore, be assessed at the beginning of each new cell-free protein production development. Consequently, only pIV4-CpxA and pIV3-DesK plasmids were used in the following experiments.

9.4 Results and Discussion

9.4.2 Detergents Compatible with Cell-free Synthesis

Detergent molecules self-associate to form micelles at a given concentration [23], the critical micelle concentration (c.m.c.). Ideally, complete and stable solubilization of membrane proteins occurs when the detergent concentration exceeds the c.m.c. [24]. Since protein yields with CECF technology could reach milligram amounts (see below), binding of membrane proteins to the detergent micelles is no longer negligible. The ratio of protein to detergent becomes an important experimental variable in cell-free production of membrane proteins, and it might therefore be necessary to supply the detergent at a concentration well above the c.m.c. [11]. However, as detergents associate with the hydrophobic surfaces of proteins, they also behave as disaggregating agents, and could denature proteins when used at high concentration. To evaluate the productivity of the RTS (rapid translation system) system under these conditions, synthesis of GFP was performed with the batch technology in the presence of various detergents commonly used in membrane protein biochemistry. These detergents were initially mixed to native lysates at three-times their c.m.c. before starting the reactions. As shown in the Fig. 9.3, the presence of all selected nonionic detergents with a low c.m.c., Tween 20 (T20), Brij 35 (Bj), n-dodecyl β-d-maltoside (Dm) and Triton X-100 (Tx), did not modify the steady-state production of GFP. In contrast, the presence of the nonionic detergent n-octyl β-d-glucoside (Og), with a high c.m.c., completely inhibited GFP

Fig. 9.3 Detergents compatible with cell-free GFP production. GFP

was synthesized in the absence of detergent (lane 1) or in the presence of 0.12 mm Tween 20 (lane 2), 0.28 mm Brij35 (lane 3), 0.6 mm n-dodecyl β-d-maltoside (lane 4), 0.9 mm Triton X-100 (lane 5), 72 mm n-octyl β-d-glucoside (lane 6), 12 mm deoxycholate (lane 7), and 21 mm CHAPS (lane 8). Whole protein samples were fractionated by 12% SDS-PAGE, and the gel was analyzed by immunoblotting as described for Fig. 9.2.

171

172

9 Cell-free Production of Membrane Proteins in the Presence of Detergents

Fig. 9.4 Production of soluble CpxA in nonionic detergent micelles. CpxA was produced from pIV4-CpxA in the presence of the following detergents at 3×c.m.c.: Tween 20 (lanes 1), n-dodecyl β-d-maltoside (lanes 2), Brij35 (lanes 3), Triton X-100 (lanes 4), and in the absence of any detergent (lanes 5). After synthesis at 30 ◦ C for 2 h, whole and soluble protein samples were analyzed by immunoblotting as described for Fig. 9.2.

production. A similar result was obtained in the presence of the ionic detergent deoxycholate (Dx). The inhibitory effect of the zwitterionic detergent CHAPS was much less than that of Og or Dx, but the decreased yield of GFP observed with this detergent precludes its future use for membrane proteins. Although performed on a limited range of detergents, these data suggest that only nonionic detergents with low c.m.c. can be used in cell-free expression systems without decreasing protein yield. Next, the solubility of CpxA produced in the presence of detergents compatible with cell-free expression was assessed, as described in Section 9.3.2, and the results are shown in Fig. 9.4. When CpxA was synthesized in the absence of detergent, the protein sedimented in the pellet (or insoluble fraction), as expected for a membrane protein, whereas when synthesized in the presence of nonionic detergent micelles it remained in the supernatant (or soluble fraction). In this specific case, Brij35 seems to be the most effective detergent for production of CpxA-detergent micelles. Interestingly, successful cell-free production of other soluble membrane proteins in the presence of this detergent has also been reported recently [13, 14]. A further advantage of using Brij35 micelles is that they do not absorb UV light, thus facilitating the structural characterization of solubilized membrane proteins by spectroscopic methods.

9.4 Results and Discussion

9.4.3 Fidelity of In Vitro Biosynthesis Reactions in the Presence of Brij35

The above results indicate that the overall performance of cell-free protein production is not modified by the presence of detergents at high concentrations. However, the high fidelity of both the T7 RNA polymerase and the translation machinery could be affected under these conditions, leading to amino acid substitutions in the synthesized protein. To investigate this possibility, maltose-binding protein (MalE) was produced in a CECF reaction containing 1% Brij35, and its structural integrity, assessed by electrospray ionization-mass spectrometry (ESI-MS) and differential scanning calorimetry (DSC), was compared with MalE produced in the absence of detergent. MalE is ideally suited for such physical techniques because it is easy to purify by one chromatography step. In both cases, the protein yields obtained were very similar, but an additional protein band was detected when Brij35 was added to CECF reactions, independently of cell-free biosynthesis reactions (Fig. 9.5A). This unidentified band presumably results from protein degradation after dissociation of complexes induced by the presence of detergent. However, no significant differences were measured in the average molecular weight of MalE synthesized either in the absence (40 831.4 ± 1.5 Da) or in the presence of Brij35 (40 831 ± 2 Da). Both experimental values agree with the expected molecular weight of MalE from its primary structure. Next, the thermal stability of MalE was analyzed by DSC. Both proteins exhibit a single peak with a Tm of 62.29 ± 0.04 ◦ C (Fig. 9.5B), indicating that they have an identical DSC endotherm. Within experimental errors,

Fig. 9.5 Quality control of CECF protein

production in the presence of Brij35. (A) Production and purification of MalE protein. Control reactions without template plasmids (lane 1), reaction mixtures with pIV3-MalE (lane 2) after 20 h in the absence of detergent (lanes headed –) or in the presence of 1%

Brij35 (lanes headed +), and purified protein (lanes 3) were analyzed by SDS-PAGE followed by Coomassie blue staining. Arrows indicate the location of an unidentified band (see text). (B) DSC endotherm of purified MalE produced in the absence and in the presence (lower curve at 70 ◦ C) of 1% Brij35.

173

174

9 Cell-free Production of Membrane Proteins in the Presence of Detergents

no detectable misincorporation of amino acids into MalE is observed under these experimental conditions. 9.4.4 High-level Production of Functional HPKs in CECF Technology

Finally, CECF reactions were performed with the two selected plasmids to produce CpxA and DesK in the presence of 1% Brij35 (a final concentration that corresponds to 90 × c.m.c.). Under these conditions, both proteins were correctly produced and solubilized in detergent micelles (Fig. 9.6A and B). Although DesK did not enter the gel, as previously observed with batch technology, the protein was found exclusively in the supernatant after centrifugation of native cell-free lysates. Next, these soluble proteins were purified by IMAC in the presence of 0.5% Brij35. After pooling the protein-containing fractions, the yield of purified protein was about 2 mg for CpxA and 1.5 mg for DesK, as determined by UV spectrophotometry. The presence of polymeric species did not significantly affect the binding of His6 -tagged DesK to the Ni-NTA column.

Fig. 9.6 CECF production and purification of CpxA from pIV4-CpxA (A) and of DesK from pIV3-DESK (B), in the presence of 1% Brij35, were analyzed by SDS-PAGE followed by Coomassie blue staining. Lanes 1, control reactions without plasmids; lanes 2, soluble reaction mixtures at 30 ◦ C for 20 h; lanes 3, unbound proteins from IMAC; lanes 4–11, elution fractions from IMAC; lanes M, molecular weight markers.

9.4 Results and Discussion

Fig. 9.7 Normalized far-UV CD spectra of native (•) and denatured (9

m urea) (◦) CpxA. The solid line corresponds to the spectra reconstructed from the best fits with CONTIN/LL analysis.

Dynamic light scattering analysis of both purified HPKs showed major assemblies with a hydrodynamic radius for DesK (19.2 ± 2.3 nm), slightly larger than for CpxA (16.2 ± 1.5 nm). However, high polydispersity was found for DesK, indicating a heterogeneous distribution of particle size for these protein–detergent micelles. For this reason the following structural analysis was only performed on CpxA. Figure 9.7 shows typical far-UV CD spectra of native and denatured CpxA, reflecting the variation in secondary structure content. Denatured CpxA exhibited a typical monotonic curve, whereas the spectra of native CpxA showed a predominance of α-helices with strong negative peaks at 208 and 222 nm. The reconstructed spectra, deduced from CONTIN/LL analysis, fit the experimental data well, and indicated α-helix and β-strand contents of 41 ± 2% and 9 ± 2%, respectively. Furthermore, secondary structure prediction from the amino acid sequence of CpxA, using the PSI-Pred algorithm [25], gave results in close agreement with the experimental data. Indeed, the secondary structure content of CpxA predicted by PSI-Pred was 42% for α-helices and 10% for β-strands. Thus, the protein solubilized by Brij35 micelles appears to form properly folded secondary structures. The first step in the two-component signal transduction paradigm is the autophosphorylation of a conserved histidine residue in HPKs [15]. Therefore, a functional structure of both purified proteins can be assessed by their ATP-dependent autokinase activity. As shown in Fig. 9.8A, both detergent-solubilized HPKs can undergo autophosphorylation in the presence of [γ -33 P]ATP. Since dimerization is necessary for the function of HPKs [26], it can be assumed that a significative fraction of CpxA and DesK is correctly assembled as homodimers in the detergent

175

176

9 Cell-free Production of Membrane Proteins in the Presence of Detergents

Fig. 9.8 Autokinase activity of purified CpxA and DesK proteins. (A) Purified DesK and CpxA were incubated in the presence of [γ -33 P]ATP at 25 ◦ C, and aliquots were removed at indicated times. The reactions were stopped by the addition of SDS gel loading buffer, and samples were fractionated by SDS-PAGE. After electrophoresis, the gel was dried and exposed for autoradiography.

(B) CECF reaction mixtures (lane 1, pIVEX2.4d; lane 2, pIV3-CpxA; lane 3, pIV3-CpxAHN) and purified proteins (lane 4, wild-type CpxA; lane 5, variant CpxA-H248N) were analyzed by SDS-PAGE followed by Coomassie blue staining. (C) Autokinase activity of wild-type CpxA (lane 1) and variant CpxA-H248N (lane 2) after 20 min incubation with [γ -33 P]ATP, as described in (A).

micelles. Based on sequence comparison with EnvZ, one of the best characterized HPK, the conserved amino acid His-248 of CpxA would be the phosphoacceptor residue. In agreement with this prediction, a CpxA-H248N variant, where the histidine 248 was substituted by an asparagine residue (Fig. 9.8B), was unable to undergo autophosphorylation, confirming that phosphorylation takes place at this specific residue (Fig. 9.8C). For DesK, a previous study performed on its soluble cytoplasmic kinase domain identified histidine 188 as the phosphoacceptor residue [18]. The activity of purified DesK was rapidly lost after two days. In contrast, CpxA, which is less sensitive to storage at 4 ◦ C, exhibited no decrease of autokinase activity even after one week. A further possibility to improve protein stability would be to produce DesK in the presence of mixed lipid-detergent micelles, since it is known that complete membrane protein delipidation might lead to protein inactivation [27]. However, it appears that no such requirements apply for the functionally active CpxA, produced and purified in the absence of lipids. The total amount of CpxA purified from one CECF reaction makes it possible to carry on preliminary crystallization trials.

9.5 Conclusions

9.5 Conclusions

We have demonstrated that cell-free expression in the presence of Brij35 micelles is well adapted for the rapid obtention of purified membrane HPKs. The protein yields of functional proteins generated by this in vitro technology are suitable for crystallization studies and structure determination. Since similar results have been obtained for a significant number of membrane proteins, cell-free protein production in the presence of detergents should play an important role in structural as well as in biochemical studies of these notoriously difficult but essential proteins.

Acknowledgments

We thank D. Albanesi and S. Hunke for providing expression plasmids, A. Haouz for DLS determination, F. Shaeffer for DSC experiments, A. Chaffotte for CD analysis, and F. Saul for carefully reading the manuscript. This research was supported by funds from the Institut Pasteur and the Centre National de la Recherche Scientifique (CNRS).

References 1 Krogh, A., Larsson, B., von Heijne, G., 7 Van den Berg, B., Clemons, W. M., Jr., Sonnhammer, E. L., Predicting transneina Collinson, I., Modis, Y., Hartmann, E., Harrison, S. C., Rapoport, membrane protein topology with a hidden T. A., X-ray structure of a proteinMarkov model: application to complete genomes. J. Mol. Biol., 2001, 305, 567–580. conducting channel. Nature, 2004, 427, 36–44. 2 White, S. H., The progress of membrane 8 Sadlish, H., Pitonzo, D., Johnson, A. E., protein structure determination. Protein Skach, W. R., Sequential triage of transSci., 2004, 13, 1948–1949. membrane segments by Sec61alpha dur3 Tate, C. G., Overexpression of maming biogenesis of a native multispanning malian integral membrane proteins for structural studies. FEBS Lett., 2001, 504, membrane protein. Nat. Struct. Mol. Biol., 94–98. 2005, 12, 870–878. 9 Spirin, A. S., High-throughput cell-free 4 White, S. H., von Heijne, G., The masystems for synthesis of functionally acchinery of membrane protein assembly. tive proteins. Trends Biotechnol., 2004, 22, Curr. Opin. Struct. Biol., 2004, 14, 397–404. 538–545. 5 Engelman, D. M., Chen, Y., Chin, C. N., 10 Betton, J. M., Rapid translation system Curran, A. R., Dixon, A. M., Dupuy, A. D., Lee, A. S., Lehnert, U., Matthews, (RTS): a promising alternative for recombinant protein production. Curr. Protein E. E., Reshetnyak, Y. K., Senes, A., Popot, Peptide Sci., 2003, 4, 73–80. J. L., Membrane protein folding: beyond 11 Berrier, C., Park, K. H., Abes, S., Bithe two stage model. FEBS Lett., 2003, bonne, A., Betton, J. M., Ghazi, A., Cell555, 122–125. free synthesis of a functional ion channel 6 Ott, C. M., Lingappa, V. R., Integral memin the absence of a membrane and in the brane protein biosynthesis: why topology presence of detergent. Biochemistry, 2004, is hard to predict. J. Cell Sci., 2002, 115, 43, 12585–12591. 2003–2009.

177

178

9 Cell-free Production of Membrane Proteins in the Presence of Detergents 12 Elbaz, Y., Steiner-Mordoch, S., Danieli, T., Schuldiner, S., In vitro synthesis of fully functional EmrE, a multidrug transporter, and study of its oligomeric state. Proc. Natl. Acad. Sci. U.S.A., 2004, 101, 1519–1524. 13 Klammt, C., Schwarz, D., Fendler, K., Haase, W., Dotsch, V., Bernhard, F., Evaluation of detergents for the soluble expression of alpha-helical and betabarrel-type integral membrane proteins by a preparative scale individual cell-free expression system. FEBS J., 2005, 272, 6024–6038. 14 Ishihara, G., Goto, M., Saeki, M., Ito, K., Hori, T., Kigawa, T., Shirouzu, M., Yokoyama, S., Expression of G protein coupled receptors in a cell-free translational system using detergents and thioredoxin-fusion vectors. Protein Expr. Purif., 2005, 41, 27–37. 15 Hoch, J. A., Two-component and phosphorelay signal transduction. Curr. Opin. Microbiol., 2000, 3, 165–170. 16 Marina, A., Waldburger, C. D., Hendrickson, W. A., Structure of the entire cytoplasmic portion of a sensor histidinekinase protein. EMBO J., 2005, 24, 4247– 4259. 17 Raivio, T. L., Silhavy, T. J., Periplasmic stress and ECF sigma factors. Annu. Rev. Microbiol., 2001, 55, 591–624. 18 Albanesi, D., Mansilla, M. C., de Mendoza, D., The membrane fluidity sensor DesK of Bacillus subtilis controls the signal decay of its cognate response regulator. J. Bacteriol., 2004, 186, 2655–2663. 19 Roge, J., Betton, J. M., Use of pIVEX plasmids for protein overproduction in Escherichia coli. Microb. Cell Fact, 2005, 4, 18.

20 Betton, J. M., Hofnung, M., In vivo assembly of active maltose binding protein from independently exported protein fragments. EMBO J., 1994, 13, 1226–1234. 21 Sreerama, N., Woody, R. W., Estimation of protein secondary structure from circular dichroism spectra: comparison of CONTIN, SELCON, and CDSSTR methods with an expanded reference set. Anal. Biochem., 2000, 287, 252–260. 22 Voges, D., Watzele, M., Nemetz, C., Wizemann, S., Buchberger, B., Analyzing and enhancing mRNA translational efficiency in an Escherichia coli in vitro expression system. Biochem. Biophys. Res. Commun., 2004, 318, 601–614. 23 Garavito, R. M., Ferguson-Miller, S., Detergents as tools in membrane biochemistry. J. Biol. Chem., 2001, 276, 32403–32406. 24 le Maire, M., Champeil, P., Moller, J. V., Interaction of membrane proteins and lipids with solubilizing detergents. Biochim. Biophys. Acta, 2000, 1508, 86– 111. 25 McGuffin, L. J., Bryson, K., Jones, D. T., The PSIPRED protein structure prediction server. Bioinformatics, 2000, 16, 404–405. 26 Yang, Y., Inouye, M., Intermolecular complementation between two defective mutant signal-transducing receptors of Escherichia coli. Proc. Natl. Acad. Sci. U.S.A., 1991, 88, 11057–11061. 27 Zhang, H., Kurisu, G., Smith, J. L., Cramer, W. A., A defined proteindetergent-lipid complex for crystallization of integral membrane proteins: The cytochrome b6f complex of oxygenic photosynthesis. Proc. Natl. Acad. Sci. U.S.A., 2003, 100, 5160–5163.

179

10

Novel Techniques using PCR and Cell-free Protein Synthesis Systems for Combinatorial Bioengineering Hideo Nakano and Tsuneo Yamane

10.1 Introduction

Cell-free protein synthesis systems, which are sometimes called in vitro expression or rapid translation systems, have recently attracted considerable attention as an alternative method of synthesizing proteins in the post-genome era, though the first such system was developed over 50 years ago. Previously, the productivity of the system was so low, due to its short reaction time, that it had been used only for limited biochemical analysis in conjunction with radioisotope labeling. That situation, however, was drastically changed by Spirin et al.’s report that cell-free protein synthesis could be prolonged for over 20 hours by employing a continuous flow system [1]. Following this report, cell-free protein synthesis systems have been improved to produce a large amount of proteins, resulting in high concentration (>1 mg mL–1 ) in the case of some intracellular proteins [2]. However, some proteins cannot be obtained in an active conformation by using conventional systems, owing to the lack of correct folding or appropriate post-translational modifications. In particular, proteins that are secreted across membranes in original host cells have seldom been produced in their active form, since these proteins often require several post-translational modifications, such as disulfide bridge formation. Adjustment of reaction conditions and the addition of some chaperones, though, have enabled the correct folding of such proteins. First, we briefly describe some examples of the expression of proteins with disulfide bonds that cannot be formed under conventional conditions. A protein that requires a special chaperone-like protein for correct folding will be also described. Second, a novel method for protein library construction using single-molecule PCR and cell-free protein synthesis is described. Some examples, using the system for the purpose of directionally evolving proteins, are also introduced. Finally, we describe a high-throughput monoclonal antibody production technique using single-cell RT-PCR and a cell-free protein synthesis system.

Cell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

180

10 Novel Techniques using PCR and Cell-free Protein Synthesis Systems

10.2 Improvements in the Escherichia coli Cell-free Protein Synthesis Systems

One of the advantages of cell-free protein synthesis systems is that they readily allow adjustment of various conditions, such as temperature, redox conditions, and chaperones for translation and folding of individual proteins. Originally, cell-free protein synthesis reactions were performed under rather reducing conditions, because the cytoplasm has a reducing environment in normal cells. However, the E. coli cell-free protein synthesis system was shown to be relatively stable under oxidative conditions, since an active single-chain Fv (scFv) containing a disulfide bond was synthesized successfully in oxidative glutathione buffer [3]. In addition, some industrially important enzymes, such as Burkholderia cepacia lipase with a single disulfide bond and Streptomyces antibioticus phospholipase D with four disulfide bonds, were also synthesized in the active form by removing dithiothreitol both from the reaction mixture and the cell-free extract [4, 5]. For B. cepacia lipase, the coexpression of its specific chaperone or the addition of the purified one was also required for the active production of the lipase. A Fab fragment of a catalytic antibody was also successfully produced in an enzymatically active form with an intermolecular disulfide bond, suggesting that the system could produce multimer proteins as well as monomer ones [6]. Phanerochaete chrysosporium manganese peroxidase, which includes five disulfide bonds, Mn2+ , Ca2+ and heme in its catalytic center, cannot be produced in an active form in recombinant systems, possibly due to the complicated structure of the enzyme; even the cell-free system could not produce the enzyme as an active form under conventional conditions. However, lowering the reaction temperature from 37 to 25 ◦ C increased the solubility of the protein significantly, and the addition of heme and protein disulfide isomerase greatly enhanced the production of the active form of the enzyme [7]. Table 10.1 shows typical reaction conditions of the cell-free protein synthesis for the manganese peroxidase. Recently, we have found that cultivation of E. coli cells at 42 ◦ C in a rich medium from which S30 cell-free extract is prepared enhances soluble fraction of some proteins in the cell-free protein synthesis system at 37 ◦ C, possibly because the extract contains a greater amount of heat-shock proteins and/or folding-assisting proteins [8]. These examples show the high adjustability and flexibility of the cell-free system for various types of proteins, and suggest that the system can be a promising expression platform for engineering and evolving proteins.

10.3 High-throughput Construction of a Protein Library by SIMPLEX 10.3.1 Development of SIMPLEX

Because the cell-free protein synthesis system can directly utilize PCR products as a template for transcription and translation, and because it is compatible with a

10.3 High-throughput Construction of a Protein Library by SIMPLEX Table 10.1 Typical reaction conditions of cell-free protein synthesis for

manganese peroxidase (the reaction mixture is incubated at 25 ◦ C for 180 min). Component

Concentration

Component

Concentration

Tris acetate, pH 7.4 ATP Each of GTP, CTP and UTP Creatine phosphate Each of 20 kinds of amino acids Poly(ethylene glycol) 6000 Folinic acid E. coli tRNAs Ammonium acetate Mg(OAc)2

56.4 mm 1.2 mm 1 mm 40 mm 0.7 mm 4.1% (w/w) 35 µg mL–1 0.2 mg mL–1 36 mm 10 mm

KOAc Creatine kinase Rifampicin T7 RNA polymerase S30 extract Hemin Fungal PDI GSH GSSG

100 mm 0.15 mg mL–1 10 µg mL–1 7.7 µg mL–1 28.3% (v/v) 10 µm 0.5 µm 1.0 mm 0.1 mm

multi-well plate format, it can be considered a quite appropriate system for highthroughput generation and screening of proteins. Accordingly, our group developed a novel protein library construction system using both amplification of single DNA molecules and the cell-free protein synthesis system, and we named this system SIMPLEX: single-molecule PCR-linked in vitro expression. Briefly, template DNA molecules, which include a T7 promoter and a terminator sequence necessary for cell-free protein synthesis (the T7 terminator sequence can greatly enhance the productivity of proteins), are diluted typically to one molecule per well, amplified by PCR, and transcribed and translated by the cell-free system (Fig. 10.1). The reduction of the accumulation of primer dimers is the most important point for the successful amplification of a single DNA molecule that is large enough to encode fully functional proteins (typically more than 1 kbp). To amplify a single DNA molecule specifically, we first employed the nested PCR method using PCR with nested primers after primary amplification [9]. However, the two-step PCR protocol was not suitable for high-throughput PCR to obtain a larger protein library. Therefore, we established a one-step method of amplifying a single DNA molecule by using single-primer PCR and hot-startable DNA polymerase [10]. The use of single-primer PCR against a homo-tailed template drastically reduced the accumulation of primer dimers, because the primer dimers made of a single primer have a panhandle structure, which effectively represses the entry of a free primer to the dimer, thereby preventing further accumulation of the byproduct. Use of hot-startable DNA polymerase also helped to reduce the formation of the primer dimers. Eventually, even a single DNA molecule was able to be amplified, and the resulting DNA library was transformed into a protein library ready for functional screening by the cell-free protein synthesis system.

181

182

10 Novel Techniques using PCR and Cell-free Protein Synthesis Systems

Fig. 10.1 Schematic drawing of a novel protein library construction system named SIMPLEX.

10.3.2 Quality of the SIMPLEX-based Protein Library

The quality of the protein library is as important as the number of molecules that can be handled at one time. If the amount of protein produced varies from clone to clone, for example, it would be very difficult to identify actual candidates from the sea of noise. We therefore examined the uniformity of the amount of proteins obtained by SIMPLEX. The amounts of PCR product amplified from a single green fluorescent protein (GFP) gene were found to be almost equal, and the amounts of GFP proteins produced in each well were also highly uniform. The relative standard deviation was about 8%, which is almost the same as that of the liquid volume of micropipettes. Since the relative standard deviation index was more than 25% using a conventional E. coli in vivo expression system, SIMPLEX can provide a much more highly uniform protein library than the conventional cell-based system [11]. 10.3.3 Expansion of the SIMPLEX-based Library

Although the size of a protein library generated by SIMPLEX would be ultimately unlimited, the production of very large libraries is currently not feasible due to its

10.3 High-throughput Construction of a Protein Library by SIMPLEX

Fig. 10.2 Schematic representation of

expanded SIMPLEX-based library construction and screening. 1: DNA templates in the gene pool are diluted to a specified number of molecules per well and amplified by multiplex-PCR, yielding a primary PCR library. 2: The PCR library is converted into the primary protein library by means of cell-free protein synthesis. 3: The protein library is screened for desired properties. 4: The genes

of positive wells are collected. 5: The pooled DNA is diluted to one molecule per well, and amplified by single-molecule PCR (SM-PCR), yielding the secondary PCR library. 6: The PCR library is converted into a secondary protein library by means of cell-free protein synthesis. 7: The protein library is screened to isolate single genes encoding a protein with the desired properties.

cost. A miniaturized PCR system that could empower the SIMPLEX is not commercially available at present. Therefore, the practically achievable library size is limited by the capacity of existing thermocyclers and the requirement of costly DNA polymerase. To increase the achievable size of the SIMPLEX-based library, a PCR of multiple molecules of templates (in this case, five molecules) was tried for an initial construction of the library. Each of the five molecules in the tube were amplified equally even after extensive amplification [12]. This important result means that the searchable number of variations can be easily multiplied by increasing the average

183

184

10 Novel Techniques using PCR and Cell-free Protein Synthesis Systems

number of molecules in a PCR well using a double-screening method (Fig. 10.2). For example, amplification of a five-molecule PCR instead of a single-molecule PCR in a 384-well plate can actually yield approximately 1900 variations of proteins at one time. The SIMPLEX system has the following advantages over cell-based systems: (a) the library size is theoretically unlimited; (b) the total time necessary to obtain a protein library from each DNA molecule is only about 5 hours; (c) the system can be easily automated; (d) cytotoxic proteins can be expressed; (e) various assay methods can be applied because the library is constructed on a multi-well plate; and (f) since the system requires no biological cloning or cultivation of recombinant cells, and is performed exclusively on a multi-plate format, its throughput is considerably higher than that of the conventional library method using agar plates. 10.3.4 Application of SIMPLEX for Combinatorial Engineering of Proteins

To demonstrate the utility of the system we have tried to engineer proteins having particular properties using the SIMPLEX technology with focused combinatorial mutation. First, the enantioselectivity of B. cepacia lipase was targeted [13]. The wildtype lipase has high enantiopreference toward the (S)-form of p-nitrophenyl 3phenylbutyrate. We subsequently tried to create a mutant lipase with reversed enantioselectivity. To introduce mutations effectively, a semi-rational and semirandom approach was employed. Four amino acid residues (L17, F119, L167, and L266) in the hydrophobic substrate-binding pocket of the lipase were selected based on a reaction intermediate model of the lipase and the substrate (Fig. 10.3). Since these residues were located in the hydrophobic region, only seven hydrophobic amino acids (G, A, V, L, I, M, F) were used as substitutes for the four selected residues, to design a mutation library. Then the combinatorial mutation library was constructed by SIMPLEX and screened for (R) and (S)-configurations of

Fig. 10.3 Model of the reaction intermediate of KWI-56 lipase and the (S)-configuration of 3-phenylbutyrate (3PB).

10.3 High-throughput Construction of a Protein Library by SIMPLEX Table 10.2 Summary of purified lipase variants with reversed

enantioselectivity (Ref. [12] and unpublished results). Variant no. WT 1 2 3 5 a) b)

c)

Mutation sitea)

L17F,F119L,L167G, L266V,(T251A) L17F,F119L,L167A,L266V,(D21N) L17F,F119V,L167G,L266V,(A24V, I139V) F119L, L167G, L266V,(Y1757)

Specific Activity (U mg–1 )b)

Enantioselectivityc)

60.2 90.6 16.6 40.0 24.1

ES = 33 ER = 38 ER = 33 ER = 84 ER = 45

Parentheses indicate mutations outside of the initial design. Specific activities of the wild type and variants were measured using (S)- and(R)- forms, respectively, of p-nitrophenyl 3-phenylbutyrate as substrate. Enantioselectivity was calculated from the degree of conversion and the corresponding enantiomeric excess [22], which were determined by chiral HPLC using (RS)-ethyl 3-phenylbutyrate as substrate.

p-nitrophenyl 3-phenylbutyrate. Some combinations of amino acid substitutions in the four positions of the lipase were found to be effective for changing the enantiopreference from the (S)-form substrate to the (R)-form without loss of specific activity (Table 10.2). This result strongly suggests that the semi-rational and semirandom combinatorial design of a mutant library followed by a high-throughput screening using the SIMPLEX technology should be a powerful tool to engineer the enantioselectivity of enzymes. Second, the hydrogen peroxide tolerance of P. chrysosporium manganese peroxidase was improved using the SIMPLEX system [7]. In this case, a mutant manganese peroxidase library containing three randomized amino acid residues located in the H2 O2 -binding pocket of the enzyme was designed and constructed on a 384-well plate using SIMPLEX. The following screening for improved H2 O2 stability gave four positive mutants, one of which showed a nine times higher stability against H2 O2 than the wild type. Third, the SIMPLEX was applied for improving the affinity of a single-chain Fv. A pool of anti-human serum albumin scFv mutant genes was created by in vitro saturation mutagenesis of a parent’s CDR-H3 region, followed by the library construction with SIMPLEX. Thereafter, all expressed scFv mutants were screened by competition ELISA. Two mutants showing improved affinities were selected and kinetically characterized, and showed significantly improved Kd s [14]. As described above, different properties of proteins have been improved in a high-throughput manner by the SIMPLEX technology. Nevertheless, the limited capacity of thermocyclers obviously restricts the usability of the system. To address this issue, a miniaturized PCR device that can increase the size and reduce the cost of the library would be desirable [15].

185

186

10 Novel Techniques using PCR and Cell-free Protein Synthesis Systems

10.4 Development and Application of SICREX

Monoclonal antibodies (Mabs) have been approved for many practical applications in research and human healthcare since they were first successfully generated by hybridoma technology. The technical difficulties inherent in generating stable hybridomas led researchers to attempt to develop genetic techniques that do not require B-cell transformation and tissue culture. To date, various approaches for the generation of Mabs have been established, including phage display [16] and ribosome display [17]. However, these methods cannot provide original pairings of the light-chain and heavy-chain genes of an antibody. In contrast, the single-cell reverse transcription polymerase chain reaction (single-cell RT-PCR) technique has been developed and used for the amplification of antibody genes from single B cells [18], enabling the regeneration of original in vivo pairings of the light- and heavy-chain genes of an antibody. However, the amplified antibody genes have to be cloned into vectors for their expression [19], which requires laborious and time-consuming effort, and limits the feasibility of the system. In contrast, an Escherichia coli cell-free protein synthesis system without a reducing reagent can produce an active Fab fragment from the light-chain (Lc) and Fd portion of heavy-chain (Hc) genes without gene cloning. Therefore, we have exploited the advantages of single-cell RT-PCR and cell-free protein synthesis systems, and have developed a novel approach by combining these two techniques in a method termed SICREX (single cell RT-PCR-linked in vitro expression) [20]. Figure 10.4 illustrates the entire in vitro method for the efficient generation and screening of Mabs. Briefly, single B cells are isolated from the spleen of a mouse

Fig. 10.4 Illustration of hybridoma technology and a novel monoclonal antibody screening method named SICREX.

10.4 Development and Application of SICREX

Fig. 10.5 Schematic illustration for the generation of a Fab antibody by

single-cell RT-PCR and cell-free protein synthesis. T7P, rbs and T7T represent aT7 promoter, a ribosome binding site and a T7 terminator, respectively.

immunized with antigen or human peripheral blood. The genes encoding the Lc and Hc of the isolated cells are separately amplified by single-cell RT-PCR. This is actually the most critical step in the whole procedure, because the specific amplification of even such a small number of mRNAs with degenerated primers is challenging. To overcome this problem, two-step PCR was utilized (Fig. 10.5). The first set of primers has a tag sequence at their 5′ end, enabling the following single-

187

188

10 Novel Techniques using PCR and Cell-free Protein Synthesis Systems

primer PCR that can increase the specificity of the amplified product greatly. The two-step PCR method, however, is laborious and sometimes causes contamination by already amplified fragments, because the tubes are sometimes opened at an early stage of PCR. To address this problem, the reaction was performed simultaneously with a low concentration of the tagged primers, with the result that a simpler and stable amplification was possible [21]. Then a T7 promoter, a ribosome-binding site and a T7 terminator were combined by the overlapping PCR technique. A pair of the thus-manipulated DNA fragments of Lc and Hc genes from the same cell were placed into the same reaction tube, and directly transcribed and translated in vitro, yielding a Fab fragment, followed by screening by an enzyme-linked immunosorbent assay (ELISA). Using this SICREX approach, the high-throughput generation of antigen-specific Mabs becomes possible from various cells and tissue sources.

10.5 Conclusion

Cell-free protein synthesis systems can now produce various types of proteins that meet special requirements, such as disulfide bond formation between intra- and intermolecules, hetero-dimerization, specific chaperones, and incorporation of heme or metal ions. The usefulness of the system has been expanded greatly. Since the cell-free system can use PCR products directly as template for transcription and translation and is compatible with a multiplate format, direct combination with PCR techniques leads to useful applications. One such application is SIMPLEX, a novel protein library construction method that uses single-molecule PCR and cellfree protein synthesis, and that has already been applied to various types of protein evolutionary engineering. Another application is a new technique called SICREX, which uses single-cell RT-PCR and cell-free protein synthesis to obtain monoclonal antibodies; this method enables the quick acquisition of monoclonal antibodies as Fab fragments from single B cells of various sources, including human sources.

References 1 Spirin, A. S., Baranov, V. I., Ryabova, L. A., Ovodov, S. Y., Alakhov, Y. B., Science 1988, 242 1162–1164. 2 Kigawa, T., Yabuki, T., Yoshida, Y., Tsutsui, M., Ito, Y., Shibata, T., Yokoyama, S., FEBS Lett. 1999, 442 15–19. 3 Ryabova, L. A., Desplancq, D., Spirin, A. S., Plückthun, A., Nat. Biotechnol. 1997, 15, 79–84. 4 Yang, J., Kobayashi, K., Iwasaki, Y., Nakano, H., Yamane, T., J. Bacteriol. 2000, 182, 295–302.

5 Iwasaki, Y., Nishiyama, T., Kawarasaki, Y., Nakano, H., Yamane, T., J. Biosci. Bioeng. 2000, 89, 506–508. 6 Jiang, X., Ookubo, Y., Fujii, I., Nakano, H., Yamane, T., FEBS Lett. 2002, 514, 290– 294. 7 Miyazaki-Imamura, C., Oohira, K., Kitagawa, R., Nakano, H., Yamane, T., Takahashi, H., Protein Eng. 2003, 16, 423–428. 8 Yamane, T., Ikeda, Y., Nagasaka, T., Nakano, H., Biotechnol. Prog. 2005, 21, 608–613.

References 9 Ohuchi, S., Nakano, H., Yamane, T., Nucleic Acids Res. 1998, 26, 4339–4346. 10 Nakano, H., Kobayashi, K., Ohuchi, S., Sekiguchi, S., Yamane, T., J. Biosci. Bioeng. 2000, 90, 456–458. 11 Rungpragayphan, S., Kawarasaki, Y., Imaeda, T., Kohda, K., Nakano, H., Yamane, T., J. Mol. Biol. 2002, 318, 395–405. 12 Rungpragayphan, S., Nakano, H., Yamane, T., FEBS Lett. 2003, 540, 147–150. 13 Koga, Y., Kato, K., Nakano, H., Yamane, T., J. Mol. Biol. 2003, 331, 585–592. 14 Rungpragayphan, S., Haba, M., Nakano, H., Yamane, T., J. Mol. Catal. B. Enzymatic 2004, 28, 223–228. 15 Nagai, H., Murakami, Y., Morita, Y., Yokoyama, K., Tamiya, E., Anal. Chem. 2001, 73, 1043–1047.

16 McCafferty, J., Griffiths, A. D., Winter, G., Chiswell, D. J., Nature 1990, 348, 552–554. 17 Hanes, J., Plückthun, A., Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 4937–4942. 18 Wang, X., Stollar, B. D., J. Immunol. Methods 2000, 244, 217–225. 19 Babcook, J. S., Leslie, K. B., Olsen, O. A., Salmon, R. A., Schrader, J. W., Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 7843– 7848. 20 Jiang, X., Suzuki, H., Hanai, Y., Wada, F., Hitomi, K., Yamane, T., Nakano, H., Biotechnol. Prog. 2006, 22, 979–988. 21 Ali, M., Hitomi, K., Nakano, H., J. Biosci. Bioeng. 2006, 101, 284–286. 22 Chen, C.-S., Fujimoto, Y., Girdaukas, G., Sih, C. J., J. Am. Chem. Soc. 1982, 104, 7294–7299.

189

191

11

Gene Cloning and Expression in Molecular Colonies Alexander B. Chetverin, Timur R. Samatov, and Helena V. Chetverina

11.1 A Gap in Cell-free Biotechnology

Practical cell-free biotechnology began with the chemical synthesis of nucleic acids [1] and with the use of restriction endonucleases [2] and ligases [3] to manipulate DNA sequences. The invention of these tools gave an enormous impulse to all biotechnology areas. Since then, further progress in biotechnology as a whole has largely been determined by the ever-increasing number of operations performed outside the cell. The invention of PCR [4], and other DNA and RNA amplification reactions, and the development of cell-free protein synthesis systems discussed in other chapters of this book have had the greatest impact. In several aspects, in vitro methods are advantageous over the traditional approaches that utilize living cells and organisms. The transfer of biotechnology to the cell-free level provides for a tighter control of all processes, for a greater variation of physicochemical parameters, for the selection from a larger number of genetic variants, and for further multiplication of structural diversity of nucleic acids and proteins by incorporation of unnatural nucleotides, amino acids and their chemical analogs. In addition, procedures that do not involve living cells are more amenable to automation and are less susceptible to natural selection. Currently, the principal components of cell-free biotechnology aimed at obtaining proteins with desired properties are as follows (Fig. 11.1). DNA or RNA preparation isolated from cells is used to create a genetic library, such as a cDNA library, which serves as starting material for the PCR amplification of individual genes. Numerous variants of the natural genes are produced by means of introducing mutations, sequence randomization and recombinations in vitro. The resulting diverse sequences are sorted using oligonucleotide arrays, microbeads, water-in-oil emulsions, as well as ribosome display techniques, to obtain genetic pools that are enriched in the sequences of interest. Those sequences are used as templates for the synthesis of desired proteins, and this can now be done in cell-free systems, including large-scale reactors capable of producing milligrams of proteins. Hence, the cell-free format has been developed for almost every step on the way from a natural gene to an engineered protein. However, until recently there was an imporCell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

192

11 Gene Cloning and Expression in Molecular Colonies

Fig. 11.1 Components of modern cell-free biotechnology. All the indicated operations are carried out in vitro, with the exception of isolation, expression and testing of individual clones, which are still performed using living cells.

tant exception: isolation of individual molecular clones from DNA or RNA pools obtained by in vitro manipulations, as well as selection of clones according to properties of the encoded proteins, was exclusively performed using living cells, thereby restricting the power of in vitro methods. The present chapter discusses a recently developed methodology that provides for gene cloning, expressing, and screening entirely in vitro, thus filling the last major gap in cell-free biotechnology. The relevant procedures are described in detail at the end of the chapter.

11.2 Molecular Colony Technique

The new methodology is based on the molecular colony technique (MCT), the idea of which is to amplify nucleic acids in a gel layer rather than in solution. This generates a 2D pattern of spherical molecular colonies, each consisting of many copies (a clone) of a single starting RNA or DNA molecule [5–7]. To produce a sufficient number of copies for a colony to be detectable, the enzyme system present in the gel must be able to exponentially amplify nucleic acids. We conceived the idea of MCT [8] while attempting to choose between two controversial hypotheses explaining the mysterious “spontaneous” RNA synthesis by Qβ replicase (RNA-directed RNA polymerase of phage Qβ) in the absence of any added template: (1) contamination of the reaction medium with replicase templates [9] and (2) de novo template-free synthesis of RNA in the test tube via random nucleotide polymerization, with the fortuitous formation of replicating molecules [10, 11]. By carrying out RNA amplification in an agarose layer containing Qβ replicase

11.3 Gene Cloning in Molecular Colonies

we demonstrated the generation of individual RNA colonies that could be visualized by staining with ethidium bromide and counted. This enabled us to discover the source of replicating molecules and to conclude that the so-called spontaneous synthesis is caused by airborne templates [8, 12]. Later, we used a similar procedure to study chemical reactions between single RNA molecules [13]. It was tempting to apply the Qβ replicase version of MCT for in vitro cloning of mRNAs, using natural Qβ replicase templates as replicable RNA vectors [14]. Unfortunately, such recombinant RNAs proved incapable of exponential amplification [15]. Even worse, these RNAs undergo spontaneous deletions of the mRNA inserts, resulting in exponentially amplifying shorter RNAs that rapidly outgrow the original recombinant templates [16]. Therefore, we switched to other enzymatic reactions that provide for the exponential amplification of nucleic acids, each of which can be utilized by MCT [5–7], including PCR and isothermal amplification reactions, such as 3SR (self-sustained sequence replication [17]), NASBA (nucleic acid sequence-based amplification [18]), SDA (strand displacement amplification [19]), and even RCA (rolling circle amplification [20]) and LAMP (loop-mediated DNA amplification [21]). The most promising results were obtained with the PCR version of MCT (PCRMCT). As far as PCR involves repeated sample heating, temperature resisting media, such as a polyacrylamide gel, are used instead of agarose. Originally, we prepared amplification gels by polymerizing acrylamide in the presence of all the PCR reagents, including DNA polymerase [6, 7]. Our protocol was followed by Church et al. [22], who slightly modified it by immobilizing one of the PCR primers on the gel matrix using the acrydite chemistry [23] and termed the method “polony technology”. Later we found that more consistent results can be obtained by first preparing an “empty” gel, washing it to remove any soluble substances, drying, and soaking the dehydrated gel in a complete PCR cocktail [24]. Compared with the original protocol, this ensured better preservation of the DNA polymerase activity. Also, this permits many gels to be prepared for future use, eliminating the need for repeating this cumbersome procedure in every MCT experiment. PCR-MCT has been employed in several applications, such as single nucleotide polymorphism genotyping and gene expression analysis [25–31], massively parallel sequencing of DNA fragments [32], studies on alternative pre-mRNA splicing [33], and extremely sensitive and reliable quantitative diagnostics [24, 34]. Recently, we demonstrated the suitability of PCR-MCT for the in vitro cloning and screening of entire genes [35], as discussed below.

11.3 Gene Cloning in Molecular Colonies

In principle, molecular clones can be obtained in vitro by diluting a nucleic acid sample to less than one molecule per compartment, and then amplifying the compartment contents. If a compartment received just one template molecule, the amplified material will represent an individual clone. This strategy is employed in

193

194

11 Gene Cloning and Expression in Molecular Colonies

“digital” PCR [36], and also in sorting on oligonucleotide arrays [37], on microbeads [38], or on microbeads contained in water-in-oil emulsion vesicles [39, 40]. The difference between these approaches and MCT is the same as the difference between cloning bacterial cells by exhaustive dilution of a liquid culture and by growing bacterial colonies on a nutrient agar. In contrast to the colony techniques, whether bacterial or molecular, which are inherently clonal methods, the dilution procedures do not automatically generate molecular clones, and the clonal purity of the resulting preparations needs to be verified by direct methods. Therefore, the use of terms “sorting” and “enrichment”, rather than “cloning”, is more justified in these cases. For the same reasons, we can conclude that, at present, MCT is the only method that can provide for true gene cloning in vitro. Unlike diagnostic applications wherein molecular colonies consist of relatively short DNA fragments, cloning of entire genes requires that much larger DNA sequences are amplified. It is also important that sufficient number of gene copies is accumulated in each colony for the clones to be detectable by their expression products. These goals can be achieved by optimizing the gel porosity and the composition of the PCR cocktail. Optimizing the gel porosity is a delicate balancing task. Molecular colonies form because the gel matrix prevents convection and limits the motion of reaction products. This, in turn, may slow down the gene amplification and expression reactions by obstructing the mobility of participating reagents and catalysts, especially of such giant biomolecules as ribosomes. Both polymerization monomer (acrylamide) and crosslinking reagent (N,N′ -methylene-bisacrylamide) concentrations are important. In our experiments, 5% acrylamide and 1% of crosslinks produced the best results [35]. In such a gel, 1.6-kbp-long DNA fragments produces colonies that contain up to 108 copies. Moreover, as discussed below, transcription and translation reactions can also be performed in this gel. Regarding the composition of PCR cocktail, we excluded detergents, salts (such as potassium chloride and ammonium sulfate), and SH-containing reagents, such as 2-mercaptoethanol or dithiothreitol, which were found to be inhibitory. Small amounts of a hyperthermophilic Pwo DNA polymerase (from Pyrococcus woesei [41]), in addition to Taq DNA polymerase (from Thermus aquaticus), enhance the amplification of long PCR products (Fig. 11.2). Reportedly, Pwo DNA polymerase also improves the fidelity of PCR due to its proof-reading 3′ → 5′ exonuclease activity [42]. At the same time, it is not necessary to perform “hot start” using special forms of DNA polymerases that become active upon heating, because the formation of primer-dimers does not significantly interfere with the amplification of molecular colonies. Compared with solution PCR, a higher concentration of a carrier protein (bovine serum albumin), at least 1 mg mL–1 , is required to prevent the enzyme adsorption on glass surfaces. Unlike Church and colleagues, we do not immobilize PCR primers on the gel matrix, because this does not aid DNA amplification and may interfere with the subsequent gene expression. Also, if the colony DNA is not covalently immobilized on the gel matrix it can be easily picked for further amplification.

11.3 Gene Cloning in Molecular Colonies

Fig. 11.2 Effect of Pwo DNA polymerase on the growth of colonies of long DNA molecules. A 1570-bp-long DNA fragment containing a sequence coding for green fluorescent protein (GFP) was PCR-amplified as described in Section 11.8.2 in a polyacrylamide gel that received approximately 30 template molecules, in the presence or in the absence of 0.02 ng µL–1 of Pwo DNA polymerase. DNA colonies were visualized by hybridization with a radioactively labeled probe (Section 11.8.3).

Fig. 11.3 A procedure for in vitro gene cloning using the molecular

colony technique, consisting of isolation of the total RNA preparation from a biological source, synthesis of cDNAs for all poly(A)-tailed mRNAs, and growing individual colonies as described in Section 11.8.2. In this example, clones of luciferase cDNA were obtained from the genetic material derived from a single cell [35].

With the above gel and PCR cocktail, we were able to clone a 1700 nucleotidelong luciferase cDNA sequence from dried lanterns of firefly Luciola mingrelica [43], which were stored in a freezer for over 10 years (Fig. 11.3). To this end, a cDNA library was synthesized using the total RNA isolated by a guanidine thiocyanatephenol method [44], the RNase H– Moloney murine leukemia virus reverse transcriptase, and oligo(dT), which served as a primer for reverse transcription of all the poly(A)-containing mRNAs. A portion of the cDNA preparation corresponding to 10 pg of the input RNA, which approximates the RNA content of a single animal cell [45], was then seeded on a PCR gel containing luciferase gene-specific oligonucleotide primers. Several tens of specific DNA colonies had grown (Fig. 11.3), suggesting that the genetic material of a single cell is sufficient for MCT cloning without preliminary amplification. The material picked from a colony can be further amplified by the solution PCR to produce the full-sized luciferase cDNA. Such a result could not be achieved with in vivo cloning techniques, which recover sequences at 0.01–0.0001% efficiency [46] due to the need to insert a sequence into a cloning vector and in transformation of cells with the ligation product – each of these steps is very inefficient.

195

196

11 Gene Cloning and Expression in Molecular Colonies

11.4 Gene Expression in Molecular Colonies: Transcription

At present, efficient cell-free transcription systems employing thermostable DNAdependent RNA polymerases are not available. Therefore, DNA amplification and transcription steps have to be separated, and this can be done by various means. For example, after completion of PCR, DNA colonies can be partially transferred to another gel, either through a direct contact with the amplification gel or by means of a blotting membrane. However, since the efficiency of such a transfer is typically within 10% [35], the preferred method is to perform transcription in situ, i.e., in the same gel in which PCR was carried out. A simple way to do this is to dry the PCR gel and then to load it with a transcription cocktail. Gel can be dried either under vacuum (Section 11.8.4) or by using a procedure employed for in situ translation (Section 11.8.5). Certain concern may arise as to whether DNA colonies will smear upon re-swelling the gel as during this process the gel contacts a free liquid. We observed no indication of smearing. Moreover, if the amplification gel was dried after blotting, then re-swollen with a PCR cocktail and again subjected to several PCR cycles, the amount of DNA in a colony increases, but the colony pattern remained essentially unchanged, suggesting that DNA molecules did not migrate from their home colonies. A likely explanation of this fact is that when a gel soaks a limited amount of the cocktail there occurs a unidirectional current of the liquid inside the gel, which does not allow DNA molecules to leave. With the use of phage T7 RNA polymerase, 10–100 copies of a 1700 nucleotidelong luciferase mRNA were synthesized per DNA molecule in each colony [35], which is comparable with the yield of transcription in solution. Cloning and transcription in molecular colonies could be used for direct screening of RNA species, such as ribozymes and aptamers. Compared with the currently employed RNA selection strategies [46], this would eliminate the need for cloning the RNA-encoding plasmids in bacterial cells and would allow the clones arranged in a 2D pattern to be tested in situ, thus greatly increasing the performance of the screening procedure.

11.5 Gene Expression in Molecular Colonies: Translation

In contrast to transcription, in situ translation cannot be achieved by merely drying the PCR gel and re-swelling it with a translation reaction mixture. This is because the components of the PCR cocktail, primarily the PCR buffer, are potent inhibitors of translation. In particular, the pH requirements of PCR and translation were found to be incompatible, with the optimal pH being 8.6 and 7.6, respectively [35]. Therefore, to perform in situ translation, the components of the PCR cocktail must be removed from the gel, which can be done by the following procedure. We found that solutions useful for precipitating nucleic acids and washing or extracting the resulting pellets are generally applicable to nucleic acids embedded in a gel. Earlier, we developed an efficient procedure for removing the substances

11.5 Gene Expression in Molecular Colonies: Translation

inhibitory to PCR from nucleic acid preparations isolated from whole blood. The procedure consists of precipitation of nucleic acids with ethanol, and then extraction of the resulting pellet with a saline alcohol solution whose composition was found empirically [24]. The latter solution contains a lower concentration of ethanol than the precipitating solution to increase the solubility of non-nucleic acid components. The salts help to keep nucleic acids precipitated and also help to remove proteins by weakening nucleic acid–protein interactions [34]. Also extracted are short nucleic acid fragments like PCR primers and, of course, nucleotides. The same solution can be used for washing the gel containing DNA colonies. Since DNA is already entrapped in the gel, the preliminary precipitation is not needed. The gel is directly extracted with saline alcohol, followed by washing with 70% ethanol to remove salts. The recipe of the saline alcohol given in Section 11.8.5 is not the only possible one. Both the organic solvent and salt moieties can be varied. Optimal concentrations of organic solvent(s) and salt(s) in the solution should be determined in preliminary experiments. Rather than by extracting a gel, it is convenient to test solutions by extracting nucleic acid pellets and analyzing the residue by electrophoresis. Washing removes from the gel all the PCR components capable of inhibiting translation, and fixes DNA molecules within their home colonies. The gel can then be dried and used for in situ translation. Translation can be performed in two steps, by first carrying out transcription (Section 11.4) and then translation itself. However, a combined transcription–translation is easier to perform, and it also results in a higher yield of the synthesized protein. To date, the most satisfactory results were obtained with a combined system consisting of phage T7 RNA polymerase and wheat germ lysate. The transcription and translation moieties of this system have compatible requirements, and the wheat germ lysate is known to have a very low content of nucleases. As always with cell-free translation, it is important to determine the optimal Mg2+ concentration in preliminary test-tube experiments, preferably for every new batch of lysate. Compared with a recipe reported elsewhere [47], we use higher concentrations of rNTPs, which results in a 3.5-fold increase of the protein yield due to boosting the transcription. Using this system, we were able to demonstrate synthesis of the green fluorescent protein (GFP) from jellyfish Aequorea victoria in molecular colonies. The colonies were generated by PCR amplification of a DNA sequence containing the cDNA for the redshifted variant of GFP [48] flanked by the 5′ -terminal untranslated region of obelin mRNA [49] and the 3′ -terminal untranslated region of the satellite tobacco necrosis virus genome [50]. These translational enhancers are very effective in this type of cell-free system [47]. The experiments were carried out as follows. After in-gel amplification of the GFP-encoding sequences, the PCR gel was washed as discussed above and soaked in the combined transcription–translation cocktail. Subsequent incubation at 25 ◦ C resulted in the appearance of fluorescent spots, whose number approximated to the number of DNA templates introduced into the gel. Moreover, the fluorescent pattern coincided with the pattern of DNA colonies revealed by hybridization with a radioactively labeled GFP mRNA, indicating that

197

198

11 Gene Cloning and Expression in Molecular Colonies

fluorescence resulted from expression of the GFP gene (see Ref. [35] and Fig. 11.4B below). These experiments have demonstrated that translation in molecular colonies results in the synthesis and folding of a functionally active protein. Additional experiments have shown that translation occurs in the gel almost as efficiently as in a liquid medium. Comparison with GFP standards suggests that, on average, 109 GFP molecules are synthesized in each expressed colony, i.e., about ten protein molecules per each DNA template [35]. This corresponds to a surface density of 40 pg of protein per square millimeter, which means that molecular colonies produce sufficient amounts of protein to be screened in situ either immunochemically or according to a function of the protein, such as binding of a ligand or catalyzing a biochemical reaction.

11.6 Gene Expression in Molecular Colonies: The Role of Thiol Compounds

The results of GFP synthesis in molecular colonies reported in Ref. [35] were obtained with the use of wheat germ lysate from Roche Diagnostics. However, when we replaced it with a home-made lysate prepared by a modification [47] of the procedure of Erickson and Blobel [51] we could not observe fluorescing colonies. The most apparent explanation was that our lysate had a too low activity, but this proved incorrect. In control experiments, in which protein synthesis was monitored by the incorporation of [14 C]leucine in test tubes, our lysate produced at least as much GFP protein as did the Roche’s lysate. One possibility was that GFP was synthesized in colonies in sufficient amounts but did not fluoresce because it failed to mature. Maturation of a newly synthesized GFP involves formation of a chromophore that requires oxidation of an intermediate by the dissolved oxygen [52]. It was, therefore, natural to suspect that the reaction conditions provided by our lysate were too reducing and hence unfavorable for the chromophore formation. We were unaware of the composition of the Roche’s lysate, but knew that the only reducing component of our lysate was a 4 mm dithiothreitol (DTT), which increased the final DTT concentration in the reaction cocktail to 2.8 mm, i.e., 1.2 mm higher than is provided by the reaction buffer. If this suppressed the formation of GFP chromophore, then decreasing the DTT concentration would help. However, it did not. On the contrary, our reaction conditions turned out to be insufficiently reducing, and that a higher concentration of DTT (or another thiol-containing reagent) is needed. Figure 11.4A shows the results of an experiment illustrating this point. DNA colonies were mimicked by spotting dried gels with miniature aliquots of a serially diluted plasmid carrying the GFP cDNA. The arrangement of spots is shown on the accompanying scheme. The gels were then soaked in reaction cocktails containing varying concentrations of several thiols, and transcription–translation was performed as above, with the washing step omitted. In addition to what is indicated, each gel contained 1.2 mm DTT originating from lysate. A gel soaked in

11.6 Gene Expression in Molecular Colonies: The Role of Thiol Compounds

Fig. 11.4 Gene expression in molecular

colonies. (A) Effect of thiol compounds on the synthesis of active (fluorescing) GFP in molecular colony mimics prepared by spotting a polyacrylamide gel with the indicated amounts of DNA template. Combined transcription–translation was carried out for 2 h as described in Section 11.8.5, using a wheat germ lysate obtained from Roche Diagnostics or a home-made lysate prepared according to a published protocol [47, 51]. (B) Images of GFP-synthesizing molecular colonies obtained by monitoring the GFP fluorescence (top row) or by hybridization of a blotting membrane with a GFP gene-specific radioactive probe (bottom row). The colonies were obtained by PCR amplification of approximately 30 template molecules per gel,

followed by a 1-hour expression using the home-made wheat germ lysate described in (A). Fluorescent images were obtained by scanning the gels with a ScanArray™ Express microarray reader (Perkin Elmer), using a blue laser (488 nm, which is identical to the absorption maximum of this GFP mutant) for excitation and a 508-nm filter for registering the GFP emission. All images were obtained at identical settings of the reader, so that signal intensities can be directly compared. Indicated are final concentrations of thiols introduced into the gels with the reaction buffer. DTT = dithiothreitol, MES = 2-mercaptoethane-sulfonic acid. (These experiments were carried out with the technical assistance of Z. V. Valina.)

199

200

11 Gene Cloning and Expression in Molecular Colonies

a cocktail based on Roche’s lysate served as a control. As shown, while with the Roche’s lysate the GFP fluorescence was detectable at each of the DNA spots under the standard conditions (1.6 mm DTT), no fluorescence was observed with our lysate when a spot contained less than 3 × 108 DNA molecules. Increasing the DTT concentration in the buffer to 3.8 mm (5 mm total) allowed the GTP fluorescence to be detected with our lysate in each spot, down to 3 × 107 DNA molecules. A similar effect was produced by a 20-mm 2-mercaptoethane-sulfonic acid (MES). The brightest fluorescence was observed when the reaction cocktail contained cysteamine (10 mm). With the increased concentration of thiols, our lysate was able to produce detectable amounts of fluorescing GFP in molecular colonies, too, and, again, the 10-mm cysteamine had the highest effect (Fig. 11.4B). These observations indicate that gene expression in molecular colonies requires a higher concentration of thiol compounds than does the standard solution format. The most plausible explanation is that molecular colonies are expressed in a thin gel layer, in which case the surface-to-volume ratio of the reaction medium is higher than, for example, in a test tube. Hence, the reaction medium is more susceptible to oxidation by atmospheric oxygen, which may affect either the components of the cell-free system or the nascent protein (e.g., by inducing the formation of incorrect S–S bond during its folding) or both.

11.7 Conclusions

The molecular colony technology described in this chapter allows the entire process of cloning, expression and screening of genes to be performed in vitro, without any use of living cells. This is the first technology that provides for true molecular cloning, and it possesses several advantages over cell-based cloning techniques. It requires neither cloning vectors nor cell transformation. As a result, up to 100% of a genetic library can be cloned and tested compared with 0.0001–0.01% when cloning is performed in living cells. Cloned genes need not be isolated, because molecular colonies comprise a genetically pure DNA. The overall process time is dramatically reduced: it takes hours instead of days. Since living cells are not involved, genes can be amplified and expressed in the absence of natural selection and in the presence of unnatural nucleotides and amino acids. Many genes and their expression products can be tested directly in molecular colonies, because there are no cell walls or membranes. Yet, since the genes and their expression products are entrapped in a thin layer of gel matrix, they can be interrogated in situ under conditions different from the transcription–translation by merely soaking into the gel any desired analytes. As far as each gene is physically linked to its expression product by being located in the same molecular colony, such a molecular display can serve as an alternative of the phage display and other display techniques. Unlike in vivo and most in vitro display methods, a protein or peptide is linked to its gene without fusing it to a tag sequence or to another protein; therefore, its native fold and properties are not disturbed.

11.8 Molecular Colony Protocols

Further developments of this technology will likely include replacement of PCR with isothermal methods, such as 3SR [17], NASBA [18], SDA [19], RCA [20, 53] or LAMP [21] amplification. Particularly promising is the use of expression systems entirely composed of purified components, such as the PURESYSTEM (see Ref. [54] and Chapter 2 by Ueda in this volume). The absence of endogenous nucleases and phosphatases increases the performance of such systems, whereas the absence of any cellular components that are not directly involved in transcription or translation provides for the in situ assay of expression products for any enzymatic, ligand-binding, immunologic, or another function that is normally present in a cell lysate.

11.8 Molecular Colony Protocols 11.8.1 Amplification Gels

Polyacrylamide was prepared in gels in shallow wells (14 mm in diameter, 0.4 mm deep) drilled in 1-mm thick glass microscopic slides at the Institute’s workshop. Alternatively, one can use Frame-Seal™ incubation chambers (MJ Research, now part of BioRad) adhered to a glass slide [22] or another suitable device. We do not recommend using Teflon™-coated slides because Teflon suppresses acrylamide polymerization. It is important that gels are covalently attached to the glass surface. To this end, prior to casting gels, the slides are defatted, such as by overnight soaking in 0.5 m NaOH followed by rinsing with water and drying, and then the inner surface of the wells is treated with PlusOne™ Bind-Silane A174 (Amersham Biosciences, now part of GE Healthcare) by applying to each well 80 µL of a diluted reagent (0.4% of the silane in an aqueous solution of 2% acetic acid and 80% ethanol). After passive evaporation of the solvent, the slides are rinsed with water, and each one is washed in 20 mL of ethanol by a 15-min shaking, rinsed again, and then dried. Into each well is poured ≈70 µL of a degassed solution containing 5% acrylamide, 0.05% N,N′ -methylene-bisacrylamide and 0.04% ammonium persulfate, mixed immediately before use with TEMED (N,N,N′ ,N′ -tetramethylethylenediamine) to the final concentration of 0.5%. During filling, the wells are covered by sliding-over with another, well-free, microscopic slide pre-treated with PlusOne™ Repel-Silane ES (Amersham Biosciences) according to the manufacturer’s instructions. After incubation at room temperature for 40 min and then at 4 ◦ C for 1 h, the upper slides are removed, and the lower (gel-containing) slides are each boiled 3 × 20 min in 40 mL of deionized water and dried overnight at room temperature. The dry gels retain their properties for at least 1.5 months upon storage in a closed container at room temperature.

201

202

11 Gene Cloning and Expression in Molecular Colonies

11.8.2 Growing DNA Colonies

Before experiments, the gels are reconstituted to their original volume by soaking in a complete PCR cocktail. To this end, 2–3 µL of mineral oil are applied as a ring around each well, the well is half-covered with a coverslip, and PCR cocktail (65–70 µL, to completely fill the well without allowing air bubbles) is poured under the coverslip, with simultaneous sliding the slip over the well until the well is fully covered. The covered well is then sealed with either adhesive PCR Foil (such as sold by Eppendorf) or Autoclave Tape (3M). The gels are incubated for 1.5 h at 4 ◦ C or otherwise (e.g., during 30 min at 60 ◦ C) prior to PCR so that all the liquid is absorbed. The recommended composition of the PCR cocktail is 50 mm Tris-HCl (pH 8.6 at 25 ◦ C), 1 µg µL–1 bovine serum albumin (fraction V, Amersham Biosciences or Roche Diagnostics), 2.5 mm MgCl2 , 0.2 mm each of dNTPs, 3.6 ng µL–1 of Taq DNA polymerase, 0.02 ng µL–1 of Pwo DNA polymerase, 0.2 µm each of PCR primers (whose Tm s are preferably 10 ◦ C above the annealing temperature, as calculated online at the Primer 3 web site: http://frodo.wi.mit.edu/cgibin/primer3/primer3_www.cgi) and a nucleic acid sample containing 10–100 template molecules. We use home-made DNA polymerases modified by His6 -tag at the N terminus and isolated as described in Refs. [24, 41], but commercially available enzymes can be used as well, taking into account that 1 unit corresponds to ≈8 ng of Taq DNA polymerase [55] and ≈32 ng of Pwo DNA polymerase [41], respectively. It is advisable to experimentally determine the optimal concentrations of DNA polymerases from a given vendor. PCR is carried out in a temperature cycler (we use UNO-Thermoblock™ from Biometra) equipped with a flatbed “in situ” module. Glass slides carrying the amplification gels are placed directly on the surface of the heating block pre-heated to 94 ◦ C, without any liquid in between, and are subjected to 40 PCR cycles, each consisting of melting at 94 ◦ C for 20 s, annealing at 55 ◦ C for 20 s, and extension at 72 ◦ C for 1–2 min, depending on the length of amplified sequences, followed by incubation at 72 ◦ C for 5 min. 11.8.3 Detection of Molecular Colonies

We detect DNA or RNA (following in situ transcription) colonies by hybridization on blotting membranes with radioactively labeled probes as described below. Membrane hybridization with fluorescent oligonucleotides can be used as well (Kravchenko, Chetverina and Chetverin, in preparation). Coverslips are carefully removed, and the gels are blotted with 14-mm diameter nylon membrane discs. Membranes modified with positively charged groups are preferred, such as Hybond™ N+ (Amersham Biosciences) and BioTrans(+) (ICN, now MP Biomedicals). Two layers of Whatman 3 filter paper, a rubber spacer, and a load (≈500 g gel–1 ) are mounted atop the membranes. After a 30-min blotting,

11.8 Molecular Colony Protocols

the membranes are washed in 70% ethanol (20 mL per membrane; if a membrane adheres to the gel during blotting, it can be easily detached upon wetting with 70% ethanol) and dried. The membranes are then incubated for 2 min on a filter paper moistened with 1.5 m NaCl and 0.5 m NaOH (at 80 ◦ C, to denature DNA) or with 6% formaldehyde, 50% formamide (freshly deionized), 50 mm Na-phosphate, pH 7.0, and 1 mm EDTA (at 65 ◦ C, to denature RNA), followed by UV illumination at 160 mJ cm–2 in Stratalinker™ (Stratagene) to immobilize nucleic acids. After washing in 1× SSPE (10 mm Na-phosphate, pH 7.7, 180 mm NaCl, 1 mm EDTA) containing 0.5% SDS (5 mL of the solution per membrane), the membranes are incubated for 30 min at 60 ◦ C in 4× SSPE containing 50% formamide and 1% SDS (1 mL of the solution per membrane) and then hybridized overnight at 55 ◦ C in a sealed plastic bag between filter paper sheets (1.5 × 1.5 cm) moistened in the same solution containing a 32 P-labeled probe (106 cpm mL–1 ; 80 µL per sheet). As a probe, we use either RNA synthesized in the presence of a 32 P-labeled NTP by transcribing an appropriate plasmid with T7 RNA polymerase or an oligodeoxynucleotide 5′ -phosphorylated using [γ -32 P]ATP and polynucleotide kinase [13]. After hybridization, the membranes are washed in four changes of 1×SSPE containing 0.5% SDS (15 min at room temperature, 5 mL of the solution per membrane) and exposed to either a radiography film (such as Hyperfilm MP from Amersham Biosciences) or a storage phosphor screen, such as screen MS, which is then scanned with a Cyclone™ imager (Packard Instrument, now a division of Perkin Elmer). 11.8.4 Transcription in Molecular Colonies

After drying in vacuo (e.g., in a freeze-drier), a PCR gel is overlaid under a coverslip with 65–70 µL of a reaction mixture containing 100 mm Tris-HCl (pH 8.0 at 25 ◦ C), 20 mm MgCl2 , 1 mm spermidine, 0.2 mm EDTA, 40 mm dithiothreitol, 4 mm each of rNTPs, and 50 ng µL–1 T7 RNA polymerase, and incubated at 4 ◦ C for 1 h, to allow all the liquid to be absorbed. Transcription is carried out by incubating the gel at 37 ◦ C for 2 h in a humidity chamber, such as in a plastic bag containing a piece of wet filter paper. 11.8.5 Protein Synthesis in Molecular Colonies

After PCR, gels are twice extracted by shaking for 10 min in saline alcohol prepared by mixing 45 volumes of 96% ethanol with 55 volumes of 200 mm Na-citrate, 300 mm NaCl, and 0.4 mm EDTA (to avoid precipitation of the salts, the saline alcohol should be kept above 20 ◦ C), followed by washing three times for 10 min in 70% ethanol. Gels are washed as follows. A slide containing three gels is placed, upside-down, 5 mm above the bottom of a glass Petri dish of 95 mm internal diameter. A washing solution is then poured into the dish in an amount sufficient to cover the slide (40–45 mL). Washing is performed without agitation, taking advantage of natural convection, so that the denser solution diffusing out from the

203

204

11 Gene Cloning and Expression in Molecular Colonies

gel fells down on the bottom of the dish without being mixed with the body of the washing solution, so that the gels mostly contact fresh portions of the solution. After drying in vacuo, each gel is overlaid under a coverslip with 65 µL of the following reaction mixture prepared on ice and cleared in the Eppendorf MiniSpin™ centrifuge for 15 min at maximum speed immediately before soaking the gel: 30% (v/v) wheat germ lysate (Roche Diagnostics), 25 mm HEPES, pH 7.6, 2.5 mm MgCl2 , 20 mm K-acetate, 1.6 mm dithiothreitol, 4% glycerol, 2 mm ATP, 1 mm GTP, 1 mm CTP, 1 mm UTP, 0.25 mm spermidine, 8 mm creatine phosphate, 60 µg mL–1 of creatine phosphokinase, 0.1 mm each of amino acids, 50 µg mL–1 yeast tRNA, and 35 ng µL–1 T7 RNA polymerase. Lysates from other sources can also be used, but in this case a different concentration of DTT or of another thiol-containing reagent may be required (Section 11.6 and Fig. 11.4A). The Mg2+ concentration may also need to be adjusted for every new batch of lysate (Section 11.5). Subsequently, the gel-carrying slide is immediately (without incubation at 4 ◦ C) placed into a humidity chamber and incubated there at 25 ◦ C for up to 2 h, after which protein synthesis levels off.

Acknowledgements

We thank Z. V. Valina for technical assistance. This work was supported in part by the program “Molecular and Cell Biology” of the Russian Academy of Sciences and an International Research Scholar’s Award from the Howard Hughes Medical Institute to A.B.C.

References 1 Khorana, H. G., in Nobel Lectures, Physiology or Medicine 1963–1970, Elsevier, Amsterdam, 1972. 2 Kelly, T. J., Jr., Smith, H. O., J. Mol. Biol. 1970, 51, 393–409. 3 Weiss, B., Richardson, C. C., Proc. Natl. Acad. Sci. U.S.A. 1967, 57, 1021–1028. 4 Saiki, R. K., Gelfand, D. H., Stoffel„ Scharf, S. J., Higuchi, R., Horn, G. T., Mullis, K. B., Erlich, H. A., Science 1988, 239, 487–491. 5 Chetverina, H. V., Chetverin, A. B., Nucleic Acids Res. 1993, 21, 2349–2353. 6 Chetverin, A. B.,Chetverina, H. V., Russian Federation Pat. 2,048,522, 1995. 7 Chetverin, A. B., Chetverina, H. V., U.S. Pat. 5,616,478, 1997. 8 Chetverin, A. B., Chetverina, H. V., Munishkin, A. V., J. Mol. Biol. 1991, 222, 3–9.

9 Hill, D., Blumenthal, T., Nature 1983, 301, 350–352. 10 Sumper, M., Luce, R., Proc. Natl. Acad. Sci. U.S.A. 1975, 72, 162–166. 11 Biebricher, C. K., Eigen, M., Luce, R., Nature 1986, 321, 89–91. 12 Chetverin, A. B., FEBS Lett. 2004, 567, 35–41. 13 Chetverin, A. B., Chetverina, H. V., Demidenko, A. A., Ugarov, V. I., Cell 1997, 88, 503–513. 14 Chetverin, A. B., Spirin, A. S., Prog. Nucleic Acid Res. Mol. Biol. 1995, 51, 225–270. 15 Morozov, I. Yu., Ugarov, V. I., Chetverin, A. B., Spirin, A. S., Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 9325– 9329. 16 Chetverina, H. V., Demidenko, A. A., Ugarov, V. I., Chetverin, A. B., FEBS Lett. 1999, 450, 89–94.

References 17 Guatelli, J. C., Whitfield, K. M., Kwoh, D. Y., Barringer, K. J., Richman, D. D., Gingeras, T. R., Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 1874–1878. 18 Compton, J., Nature 1991, 350, 91–92. 19 Walker, G. T., Little, M. C., Nadeau, J. G., Shank, D. D., Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 392–396. 20 Fire, A., Xu, S.-Q., Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 4641–4645. 21 Notomi, T., Okayama, H., Masubuchi, H., Yonekawa, T., Watanabe, K., Amino, N., Hase, T., Loop-mediated isothermal amplification of DNA. Nucleic Acids Res. 2000, 28, e63. 22 Mitra, R. D., Church, G. M., Nucleic Acids Res. 1999, 27, e34. 23 Boles, T. C., Kron, S. J., Adams, C. P., U.S. Pat. 5,932,711, 1999. 24 Chetverina, H. V., Samatov, T. R., Ugarov, V. I., Chetverin, A. B., BioTechniques 2002, 33, 150–156. 25 Mitra, R. D., Butty, V. L., Shendure, J., Williams, B. R., Housman, D. E., Church, G. M., Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 5926–5931. 26 Merritt, J., DiTonno, J. R., Mitra, R. D., Church, G. M., Edwards, J. S., Nucleic Acids Res. 2003, 31, e84. 27 Butz, J., Wickstrom, E., Edwards, J., BMC Biotechnol. 2003, 3, 11. 28 Mikkilineni, V., Mitra, R. D., Merritt, J., DiTonno, J. R., Church, G. M., Ogunnaike, B., Edwards, J. S., Biotechnol. Bioeng. 2004, 86, 117–124. 29 Butz, J. A., Roberts, K. G., Edwards, J. S., Biotechnol. Prog. 2004, 20, 1836–1839. 30 Butz, J. A., Yan, H., Mikkilineni, V., Edwards, J. S., BMC Genetics 2004, 5, 3. 31 Merritt, J., Ogunnaike, B. A., Edwards, J. S., Biotechnol Bioeng. 2005, 92, 519–531. 32 Mitra, R. D., Shendure, J., Olejnik, J., Olejnik, E. K., Church, G. M., Anal. Biochem., 2003, 320, 55–65. 33 Zhu, J., Shendure, J., Mitra, R. D., Church, G. M., Science 2003, 301, 836– 838. 34 Chetverina, H. V., Falaleeva, M. V., Chetverin, A. B., Anal. Biochem., 2004, 334, 376–381. 35 Samatov, T. R., Chetverina, H. V., Chetverin, A. B., Nucleic Acids Res. 2005, 33, e145.

36 Vogelstein, B., Kinzler, K. W., Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 9236–9241. 37 Chetverin, A. B., Kramer, F. R., Bio/Technology 1994, 12, 1093–1100. 38 Brenner, S., Williams, S. R., Vermaas, E. H., Storck, T., Moon, K., McCollum, C., Mao, J.-I., Luo, S., Kirchner, J. J., Eletr, S., DuBridge, R. B., Burcham, T., Albrecht, G., Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 1665–1670. 39 Sepp, A., Tawfik, D. S., Griffiths, A. D., FEBS Lett. 2002, 532, 455–458. 40 Dressman, D., Yan, H., Traverso, G., Kinzler, K. W., Vogelstein, B., Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 8817–8822. 41 Dabrowski, S., Kur, J., Protein Expr. Purif. 1998, 14, 131–138. 42 Sambrook, J., Russell, D. W., Molecular Cloning, 3rd edn., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 2001. 43 Devine, J. H., Kutuzova, G. D., Green, V. A., Ugarova, N. N., Baldwin, T. O., Biochim. Biophys. Acta 1993, 1173, 121– 132. 44 Chomczynski, P., Sacchi, N., Analyt. Biochem. 1987, 162, 156–159. 45 Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K., Watson, J. D., Molecular Biology of the Cell, 3rd edn., Garland, New York, 1994. 46 Roberts, R. W., Ja, W. W., Curr. Opin. Struct. Biol. 1999, 9, 521–529. 47 Shaloiko, L. A., Granovsky, I. E., Ivashina, T. V., Ksenzenko, V. N., Shirokov, V. A., Spirin, A. S., Biotechnol. Bioeng. 2004, 88, 730–739. 48 Cormack, B. P., Valdivia, R. H., Falkow, S., Gene 1996, 173, 33–38. 49 Illarionov, B. A., Bondar, V. S., Illarionova, V. A., Vysotski, E. S., Gene 1995, 153, 273–274. 50 Timmer, R. T., Benkowski, L. A., Schodin, D., Lax, S. R., Metz, A. M., Ravel, J. M., Browning, K. S., J. Biol. Chem. 1993, 268, 9504–9510. 51 Erickson, A. H., Blobel, G., Methods Enzymol. 1983, 96, 38–50. 52 Cody, C. W., Prasher, D. C., Westler, W. M., Prendergast, F. G., Ward, W. W., Biochemistry 1993, 32, 1212–1218. 53 Lizardi, P. M., Huang, X., Zhu, Z., BrayWard, P., Thomas, D. C., Ward, D. C., Nat. Genet. 1998, 19, 225–232.

205

206

11 Gene Cloning and Expression in Molecular Colonies 54 Shimizu, Y., Inoue, A., Tomari, Y., Suzuki, T., Yokogawa, T., Nishikawa, K., Ueda, T., Nat. Biotechnol. 2001, 19, 751– 755.

55 Roche Applied Science, Taq DNA Polymerase pack insert 2004 (www.roche-applied-science.com/packinsert/03734927001a.pdf).

207

12

Large-Scale Batch Reactions for Cell-free Protein Synthesis Alexei M. Voloshin and James R. Swartz

12.1 Introduction 12.1.1 Cell-free Protein Synthesis

Cell-free protein synthesis systems can harness and focus coupled processes of transcription and translation to produce one desired target protein without the need for living cellular organisms. This is accomplished by extracting the necessary catalytic machinery (for energy generation, translation, and folding) from living cells in the form of crude extract. This extract is then incubated with substrates and other components (salts, cofactors, energy source, RNA polymerase, DNA template encoding the target protein, and amino acids) to synthesize and fold the protein of interest (Fig. 12.1). The lack of a cell wall and freedom from the need to satisfy other biochemical needs, typically associated with growth and maintenance of a living cell, gives cell-free protein synthesis systems several key advantages over conventional rDNA protein expression technology. More energy and other resources can be focused on processes directed toward synthesis and folding of the target protein. Conditions such as pH and redox potential can be easily adjusted and non-native substrates and catalysts can be readily and precisely added to facilitate protein expression and folding. These features make the cell-free systems particularly well-suited for the study of complex cellular processes with much greater ease than can be done in living cells. One can incrementally strip away the complexity of the processes associated with a living cell and still leave the biochemical system of interest intact. Because the cell-free synthesis reactions can be done in a variety of formats, this type of technology is also well suited for multiplexed protein production that can be used for various of applications such as high throughput protein library screening [1–5] or rapid expression of custom proteins for medical applications [6]. Expression of toxic protein products is yet another application for which cell-free systems are particularly well suited since there is no need to maintain the viability of the host cell [7]. Similarly, incorporation of unnatural amino acids into the target protein’s polypepCell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

208

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

Fig. 12.1 Schematic diagram of combined transcription and translation in a cell-free system.

tide chain is potentially easier in cell-free systems and is of critical value not only for the study of protein chemistry [8–11], but also to other research fields such as fluorescence spectroscopy at the single molecule level, where the unambiguous location and multiplicity of fluorescent tags on a polypeptide chain are of paramount importance [12]. 12.1.2 Comparing Cell-free Reaction Configurations; Advantages of Batch Mode

Nevertheless, several serious road blocks remain on the way to cell-free systems becoming a widely adopted technology and an attractive alternative to conventional in vivo protein synthesis. One of these main barriers is the inability to easily and efficiently increase the volume of protein synthesis reactions. The reaction scales are limited to, in most cases, tens of microliters producing tens of micrograms of target protein per batch. In some cases, milliliter scales are realized, but the low efficiency, high cost, and reactor complexity of such reactions make them unsuitable for wide adoption and certainly unusable for industrial applications [9, 13–16]. Several commercial kits for in vitro expression are available from such companies as Roche Diagnostics (Basel, Switzerland), Invitrogen (Carlsbad, CA), and Active Motiff (Carlsbad, CA); but all fail to deliver the simplicity, robustness, and cost effectiveness that is needed for in vitro synthesis technology to be widely adopted as a large scale alternative to conventional recombinant protein expression. Clearly, if cell-free technology is to deliver on its promise of producing protein products quickly and easily with all of the potential advantages, an effective scale-up methodology must be developed. Several reactor types have been developed in an effort to increase the scale of the cell-free reactions. The main types are batch, continuous, semi-continuous (continuous exchange), and hollow-fiber reactors (Fig. 12.2).

12.1 Introduction

Fig. 12.2 Four major types of bioreactors for carrying out cell-free

protein synthesis. The feeding solution contains all of the low molecular weight reagents and includes NTPs, energy source, amino acids, cofactors, salts, and buffers.

Continuous mode cell-free reactors (CFCF) (Fig. 12.2b) were pioneered by Spirin and co-workers in 1988 [17]. The main purpose of this configuration was to solve the problem of reaction longevity. Spirin and co-workers demonstrated that the CFCF design could maintain a constant rate of protein expression for tens of hours. The cell-free reaction chamber is continuously supplied with a substrate feeding buffer containing ATP, GTP, and amino acids, while reaction byproducts and synthesized polypeptide are removed across an ultrafiltration membrane. Spirin’s experiments showed that cell-free protein synthesis reactions were limited by substrate supply and/or byproduct accumulation, and that macromolecular assembly of catalytic machinery prevented its loss by flow-though. However, such reactor designs use substrates too inefficiently to be economical for large scale production. Several laboratories have improved upon Spirin’s work [18–21], but none of them was able to significantly increase the substrate utilization efficiency of the CFCF approach.

209

210

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

The hollow fiber membrane reactor (HFMR) (Fig. 12.2d) is a direct extension of CFCF systems [22, 23]. This reactor design significantly increases the filtration surface area, which enhances substrate replenishment and waste removal rates relative to earlier CFCF formats. Similar to other CFCF reactors, the HFMR format has seen little use due to inefficient substrate utilization, difficulty in scaling up cell-free systems requiring rapid gas exchange, and the complicated reactor configuration, which is difficult to scale past several milliliters in reaction volume. Realizing that the reactor design complexity are the weak points of the CFCF format, Spirin proposed the semi-continuous (continuous exchange) flow system (CECF) [24] (Fig. 12.2c). This scheme supplies substrates and removes byproducts by diffusion rather than flow. In such a reactor, a large volume of support reagents is separated from the reaction mixture by a dialysis membrane. Kim and Choi [15] demonstrated the effectiveness of this rector design in practice, producing 1.2 mg protein mL−1 reaction. More recently, Kigawa et al. [9] reported the expression of milligram quantities (6 mg mL−1 ) of chloramphenicol acetyl transferase (CAT) with a continuous exchange system. While the CECF system was an improvement over the CFCF counterpart, it was still far too inefficient in substrate utilization for industrial deployment and did not allow the scale-up of more economical cell-free systems that require oxygen for oxidative phosphorylation. Batch mode reactors (Fig. 12.2a) offer the simplest and most attractive operational mode. The reaction components are mixed in a reactor (be it a test-tube or an industrial size stirred tank) and are incubated for a set period of time. This mode offers simple reactor design and little operator involvement during the reaction. It is also the preferred mode of recombinant protein production, so it will see the least resistance from the pharmaceutical industry for cell-free protein synthesis adoption. Until recently, the batch mode scale-up was severely limited by a decrease of protein synthesis yields at scales larger than 15 µL. In this chapter we discuss recent developments in the area of batch cell-free protein synthesis system scale-up and demonstrate that this mode of operation can be successfully scaled-up to 1-L reaction volumes and beyond while enhancing the productivity of the system compared to the small-scale test-tube reactions. Even more importantly, the batch mode scale-up is capable of scaling-up much more efficient and economical cellfree systems that utilize oxidative phosphorylation to supply the energy for protein synthesis.

12.2 Challenges for Extending Batch Duration and Productivity 12.2.1 Providing Energy

Escherichria coli cell-free systems that do not require oxygen for oxidative phosphorylation are typically driven by compounds that contain high energy phosphate bonds, such as creatine phosphate, phosphoenolpyruvate, or acetyl phos-

12.2 Challenges for Extending Batch Duration and Productivity

phate (Fig. 12.3). These compounds can directly donate the phosphate bond to ADP through a reaction catalyzed by an appropriate kinase enzyme. These systems have a relatively short synthesis time since the energy provided by the substrate depletes quickly, usually in < 1 h. Additionally, these systems are plagued by the accumulation of free phosphate which eventually inhibits protein synthesis. To prolong ATP regeneration within the cell-free system (and thereby increase protein production yields), alternative techniques activate more of the available catalytic inventory in the S30 extract. In this way, ATP can be produced more efficiently from multi-step reactions, rather than relying simply on one-step phosphorylation. Kim and Swartz demonstrated two such systems in 1999 and 2001 [25, 26] that exploited networks surrounding pyruvate. Both are able to recycle phosphate and avoid the inhibitory byproduct formation common to conventional high-energy phosphate donors (e.g. phosphoenyl pyruvate (PEP)) (Fig. 12.4). One scheme uses pyruvate oxidase from either Lactobacillus or Pediococcus, in the presence of molecular oxygen, to catalyze the conversion of pyruvate to acetyl phosphate (Fig. 12.4a). The acetyl phosphate then drives the formation of ATP via the endogenous acetate kinase present in the cell extract. An alternative method involves activating the formation of acetyl phosphate from pyruvate by the addition of nicotinamide adenine dinucleotide (NAD) and coenzyme-A (CoA), required cofactors for endogenous pyruvate dehydrogenase (Fig. 12.4b). This forms the basis for the PANOx system reported in 2001 [26]. In addition to activating networks surrounding the central branch-point of carbon metabolism, ATP has also been regenerated by activating the Embden-Meyerhof pathway. Conversion of glucose-6-phosphate (G6P) to acetate and lactate via pyruvate was also reported to fuel protein biosynthesis by regenerating ATP [26]. Prolonging energy generation by using the entire glycolytic pathway results in higher protein production than the conventional PEP system (228 µg protein mL−1 from 33mm G6P after a 2 h incubation). In another example, biochemical reactions from 3-phosphoglycerate to pyruvate were exploited to regenerate ATP [27]. By increasing the duration of protein synthesis to 2 h, approximately twice the amount of protein was produced.

Fig. 12.3 Energy generation by direct phosphate transfer reactions.

211

212

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

Fig. 12.4 Secondary energy regeneration systems exploiting anaerobic metabolism. (a) Pyruvate oxidase system. (b) PANOx system. Abbreviations: Ack, acetate kinase; ATP, adenosine triphosphate; CoA, Coenzyme A; Ldh, lactate dehydrogenase; NAD, nicotinamide adenine dinucleotide; Pdh, pyruvate dehydrogenase; Pox, pyruvate oxidase.

Fig. 12.5 Aerobic energy generation in cell-free systems using central metabolism and oxidative phosphorylation. Abbreviations: ETC, electron transport chain; PMF, proton-motive force.

12.2 Challenges for Extending Batch Duration and Productivity

Fig. 12.6 Typical ATP concentration profile in a cell-free protein

synthesis reaction utilizing the aerobic energy generation system.

While the above schemes are effective at delivering the energy to drive the protein synthesis in CFPS systems, all of these energy sources are prohibitively expensive for large-scale reactions and some accumulate harmful byproducts such as phosphate that limit the protein synthesis potential. Recently, researchers in the Swartz laboratory have engineered the CFPS system to utilize oxidative phosphorylation to drive the synthesis of ATP (Fig. 12.5) [28]. The electron transport chain cascade resides in the inner membrane vesicles formed during the cell-extract preparation process. The metabolic pathways have been activated by mimicking the intracellular chemical environment more closely in the CFPS system. This opens the door for the use of several inexpensive energy substrates such as pyruvate, glutamate, and glucose [29]. Additionally, the new scheme can provide a stable energy source for protein synthesis for longer than 6 h (Fig. 12.6). In at least one sense, however, the newly developed systems have increased in complexity as the oxidative phosphorylation uses oxygen as the final electron acceptor. Thus, efficient delivery of oxygen to the system becomes a critical parameter in the reactor design. Nonetheless, the inexpensive nature of the energy substrates makes these new systems much more desirable for use in large-scale cell-free reactions. One of the most interesting of these, the glutamate-phosphate Cytomim system, generates reducing equivalents as glutamate is processed by the TCA cycle. The ATP supply for the resulting oxidative phosphorylation is sufficient to allow the use of nucleotide monophosphates and free phosphate as the NTP source. 12.2.2 Stabilizing the Substrates

Substrate stability (apart from the energy source) in CFPS systems has been substantially improved over the past several years. Much effort has been focused on the DNA and mRNA templates and on amino acid stabilization.

213

214

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

To convert stored genetic information into protein, cell-free systems can use either mRNA (translation only) or DNA (combined transcription–translation) as templates. For the translation only systems, the presence of ribonucleases in the S30 extract requires that steps be taken to protect the message from degradation. Chemical modification of the nucleic acid backbone [30–33], as well as incorporation of secondary structural elements at the 3′ and 5′ ends of the message [34], have been successfully demonstrated. Alternatively, one can also remove the enzymes responsible for mRNA degradation from the extract by genetic modification of the extract E. coli strain. It has been shown [35] that using the extract prepared an from E. coli strain lacking the genes that encode for the C-terminal portion of ribonuclease E and polynucleotide phosphorylase (A19 rne pnp) increases mRNA concentration in the cell-free reaction relative to extract from the wild type A19 strain. To regenerate mRNA within the system, two approaches have been employed. The first inserts the message into a template recognized by QB replicase. This RNAdependent RNA polymerase continuously produces mRNA in the reaction [36–38]. The second, which is the more widely used approach, combines transcription and translation by producing mRNA from a DNA template. Either endogenous RNA polymerase is used, or exogenous phage RNA polymerase, typically T7 RNAP, is added to the reaction mixture or is overexpressed within the extract source cells [39–42]. Linear templates, such as PCR products, can also function as templates in CFPS systems. This type of template is particularly well suited for high-throughput expression of protein libraries in a multiplexed format (96 well) and for ultra-rapid development of protein production processes such as those required for deploying a vaccine or antidote for an epidemic outbreak. Expression using linear templates bypasses the lengthy process required for cloning genes into plasmid vectors, transforming cells, growing these cells and then isolating the plasmid. Until recently, the poor stability of linear templates in the presence of S30 extract severely limited their use. Protein yields in CFPS systems using linear templates were less than 20 µg mL−1 [18, 43–46]. To increase PCR product stability several nuclease activities can be eliminated from the source strain used in extract preparation, or these can be inhibited. One approach demonstrated by Sitaraman et al. stabilizes linear DNA templates by the addition of the bacteriophage λ GAM protein (an inhibitor of RecBCD endo and exo-nuclease activity) [47]. As a result, in vitro protein synthesis was improved to a level almost comparable with expression from plasmid DNA. Michel-Reydellet and colleagues demonstrated another successful stabilization strategy by modifying the extract E. coli strain [35]. By removing the gene for endonuclease I (encoded by endA) and replacing the recBCD operon (part of the E. coli recombination system) with the λ phage Red recombination system (encoded by exo and bet) in the bacteriophage, an extract was generated that stabilizes PCR products and produces equivalent protein yields from either plasmid or linear DNA templates. Amino acid instability is also a limitation for many cell-free systems. Conclusive evidence for this first came in 2000, when Kim and Swartz demonstrated that arginine, cysteine, and tryptophan are depleted over the course of a reaction [48]. Replenishment of these substrates increased protein yields, presumably because

12.2 Challenges for Extending Batch Duration and Productivity

polypeptide formation was less restricted by the availability of these amino acids. Another approach, used in the PANOx system, is simply to increase the initial amino acid concentrations [26]. While substrate replenishment is a satisfactory solution, this approach increases costs and operational complexity. A more efficient solution is to engineer the system such that the reactions that deplete particular amino acids can be controlled. For example, protein yields can be increased by using inhibitory agents or by genetically modifying the cell extract source strain to remove the enzyme acting upon a specific amino acid. Kim and Choi hypothesized that cysteine depletion was a result of the glutamate-cysteine ligase reaction, part of the glutathione formation pathway [15]. Rather than periodically feeding cysteine to the reaction, addition of an inhibitor to the ligase reaction, l-buthionine, was shown to increase protein synthesis yields by 23%. Deletion of genes encoding nonessential enzymes that lead to substrate depletion provides another solution. Inactivating the gshA gene to avoid the presence of glutamate-cysteine ligase in the cell extract has stabilized cysteine concentrations [49]. Furthermore, deleting the speA and tnaA genes to avoid arginine decarboxylase and tryptophanase expression stabilized arginine and tryptophan concentrations, respectively, in the cell-free system [50]. It was originally thought that the degradation of only three amino acids was present in the cell-free system. Subsequent reports, however, have determined that serine is also exhausted. This effect was discovered after the conventional growth media was altered. Mutant strains have been constructed to stabilize serine concentrations within the in vitro system by deleting the sdaA and sdaB genes, which encode for two deaminase enzymes [49]. In addition to obvious concerns about amino acid depletion, the production of amino acids within the cell-free system can also be detrimental if resources are directed away from protein synthesis. To this end, Kim and Swartz reported that alanine, asparagine, and aspartic acid were produced in cell-free protein synthesis reactions [48]. Formation of aspartic acid and asparagine suggested the possibility of phosphoenolpyruvate synthetase (Pps) activity. In energy regeneration systems utilizing PEP as the secondary energy source, the pyruvate kinase reaction catalyzes the formation of pyruvate. The accumulated pyruvate can be used as a substrate in numerous metabolic reactions, including the Pps reaction which forms PEP at the expense of two high-energy phosphate bonds. The interconversion between PEP and pyruvate potentially sets up a futile cycle, independent of aspartic acid and asparagine formation. However, PEP can also be converted to oxaloacetate, the precursor to aspartic acid and asparagine. The continued increase of aspartic acid and asparagine concentration after apparent PEP depletion suggested that PEP was regenerated by pyruvate via Pps. The Pps reaction reduced the amount of energy available for protein synthesis since the conversion of ATP into AMP is necessary to drive the reaction. Adding oxalic acid, a known inhibitor of Pps, was reported to enhance synthesis yields and reduce the formation of aspartate from the cellfree system [50]. This is the first component of the PANOx system [26]. Alanine production is likely a result of glutamate pyruvate transaminase.

215

216

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

12.3 Scale-up of Reactions not Requiring Oxygen in Batch Mode 12.3.1 Test-tube Scale-up Results are Disappointing

During the scale up of cell-free protein synthesis reactions in batch mode in the PANOx [26] system that utilizes phosphoenol pyruvate (PEP) for energy production (as discussed in Section 12.2), the specific yield of the target protein decreases dramatically with increasing reaction volume (Fig. 12.7). In our case, the reactions were conducted in 1.5 mL polypropylene test tubes. The specific protein yields decreased as the reaction volume exceeded 30 µL. This effect appears to be general and protein target independent. We show this effect on the cell-free expression of two very different proteins. One is the E. coli chloramphenicol acetyl transferase (CAT), and the other is a synthetic multi-domain mammalian cancer vaccine construct (GLH), which is a fusion between murine granulocyte macrophage colony stimulating factor and a murine lymphoma tumor B-cell receptor scFv [6]. The expression of both targets suffers from yield reduction as the reaction scale increases past 15 µL. This suggests that there exists a fundamental limitation to system performance as the reaction volume increases. We observed that at 15 µL scale the reaction takes the shape of a drop on the bottom of a test tube. As the reaction volume increases, the surface-to-volume ratio decreases due to the geometrical constraints of the test tube. We reasoned that this had two possible consequences for the reaction. The first effect is a reduction in total gas transfer rate. For cell-free systems requiring oxygen for energy generation, this surface area reduction could limit the energy available for protein synthesis. However, systems that do not require gas transfer, such as the PANOx system, do not require a large gas–liquid surface area for the transfer of gases yet these systems also exhibit decreases in specific productivity with increased reaction scale. We hypothesized that the hydrophobic surface (both plastic and gas–liquid interface) in contact with the cell-free reaction might be important in promoting efficient protein expression. Thus, a high surface area to volume ratio would be beneficial for such systems. 12.3.2 Thin-film Format Conserves Performance

To solve the batch mode scale-up problem, we decided to conduct our cell-free reactions in a thin-film (TF) format (Fig. 12.8). Such a format directly addresses the problem of surface-to-volume ratio, and, potentially, can scale to any volume. In our case, the reactions were conducted on a hydrophobic plastic surface (polystyrene) in a sealed chamber to limit evaporation. CAT and GLH synthesis reactions conducted in the thin film format did not suffer yield reduction as the scale increased (Fig. 12.9). Total, soluble, and active yields of CAT and the GLH B-cell cancer vaccine candidate were preserved. In these experiments, the activity of the GLH

12.3 Scale-up of Reactions not Requiring Oxygen in Batch Mode

Fig. 12.7 Yields of (a) E. coli CAT and (b) the cancer vaccine candidate GLH both decrease during conventional test tube scale-up of PANOx cell-free reactions.

vaccine candidate was judged by the surface-immobilized anti-mGM-CSF antibody binding recognition of the mGM-CSF domain (BD OptEIA Mouse mGMCSF ELISA Kit, BD Biosciences, San Diego, CA). This method showed only partial mGM-CSF activity compared to a in vivo produced mGM-CSF without the scFv domain. This is probably because the mGM-CSF binding epitope availability is affected by the scFv domain. A more biologically relevant cell proliferation assay con-

217

218

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

Fig. 12.8 Simple cell-free batch reaction geometries.

firmed that the mGM-CSF domain of the GLH fusion was fully active when used to stimulate cell growth [6]. We then measured protein synthesis yields as a function of time at large and small scales. Figure 12.10 summarizes product accumulation kinetics in the PANOx system. Using the TF scale-up format, protein production kinetics at the 500 µL scale mimicked those of 15 µL reactions. Conventional test tube batch scale-up yields are dramatically lower and synthesis appeared to stop sooner. 12.3.3 Investigating Fundamental Influences

The hydrophobic plastic surface of the reaction chambers (polypropylene test tube and polystyrene surface), and hydrophobic gas–liquid interfaces enhance protein production and folding. It was previously found [51–53] that misfolded polypeptide chains, lipids, and small molecule hydrophobic byproducts can bind and inhibit cellular synthesis and folding machinery. Hydrophobic surfaces may remove these species from the reaction by competitive binding. If this is the case, any hydrophobic surface capable of adsorption will work. Figure 12.11 shows results of an experiment where the gas–liquid interface is replaced by a hydrophobic plastic surface. The film thickness was maintained at approximately 1.5 mm, and the PANOx system was used since it does not require oxygen transfer. Results indicate that replacement of a gas–liquid hydrophobic interface with a solid hydrophobic surface preserves CAT product yields as reaction scale increases.

12.4 Scale-up of Reactions Requiring Oxygen 12.4.1 Test-tube Scale up is Disastrous

To provide a cheaper, more robust, and long-lasting energy source to fuel the protein synthesis reactions, the oxidative phosphorylation pathway and part of central metabolism were activated by Jewett and Swartz [28]. This energy production

12.4 Scale-up of Reactions Requiring Oxygen

Fig. 12.9 The thin film reaction geometry conserves protein yield with

the PANOx system as the cell-free reaction scale increases.

scheme relies heavily on the availability of dissolved oxygen in the cell-free reaction mixture to act as the final electron acceptor (Section 12.2.1). Thus, efficient oxygen delivery to the cell-free reaction becomes a vital element that needs to be considered when scaling up such cell-free systems. Test-tube scale up of the Cytomim system, driven by pyruvate and glutamate as the primary energy sources and utilizing oxidative phosphorylation for direct

219

220

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

Fig. 12.10 CAT accumulation in the PANOx system. Reactions were

performed in either 1.5 mL polypropylene test-tubes (T-T) or thin films (T-F).

Fig. 12.11 PANOx reactions investigating the effect of a hydrophobic

gas–liquid interface versus a hydrophobic plastic surface at the reaction boundaries. The air–liquid interface of the thin film format is replaced by a hydrophobic plastic (polystyrene)–liquid interface. Thin film reactions were run side by side with 15 µL test-tube reactions.

12.4 Scale-up of Reactions Requiring Oxygen

Fig. 12.12 CAT (top) and GLH vaccine candidate (bottom) synthesis

reactions in the Cytomim system conducted in a test tube at increasing reaction volumes. The protein yield drops off sharply as the reaction volume increases.

ATP synthesis, fails to maintain the protein synthesis yields observed at 15 µL level (Fig. 12.12) due to limiting gas–liquid interfacial area and, thus, inadequate oxygen diffusion rate into the cell-free reaction. This phenomenon is target protein independent since the primary energy supply of the system is affected.

221

222

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

Fig. 12.13 Surface-to-volume ratio for different reaction geometries.

The thin film format conserves the interfacial area to volume ratio as the reaction volume increases.

12.4.2 Thin-film Format Conserves Performance

The thin film format (Fig. 12.8), developed by Voloshin and Swartz [54], solves the problem of simple batch scale-up of cell-free systems requiring oxygen, including the Cytomim system. The key to the thin film geometry is that, unlike the test tube, the thin film provides increased oxygen transfer areas at larger reaction volumes (Fig. 12.13). In this way, the CFPS reaction can be scaled to larger volumes with less severe limitations in oxygen delivery. Figure 12.14 summarizes the scale-up reaction data for CAT and GLH in the Cytomim system. In both cases the specific yields are fully conserved as the reaction scale increases past the 15 µL scale. The time course data (Fig. 12.15) shows that the protein synthesis kinetics in the thin film closely mimic the 15 µL scale test tube reaction. The thin film scale-up approach, coupled with the efficiency, robustness, and lower cost of the Cytomim system significantly increases the attractiveness and practicality of the cell-free protein technology for research and commercial applications. The CFPS can now be used to synthesize quantities of material needed for in vivo studies as well as for X-ray crystallography and NMR structural investigations without requiring special equipment or expensive energy substrates at moderate reaction scales. 12.4.3 Stirred Tank Aerated Reactor Format Requires Antifoaming Agents

While the thin format is successful at scaling up various CFPS systems, it suffers from several inherent limitations. Evaporation may become problematic at very

12.4 Scale-up of Reactions Requiring Oxygen

Fig. 12.14 The thin film reaction geometry conserves protein yield in

the Cytomim system as the cell-free reaction scale increases.

large (100 L+) reaction volumes. Addition of reagents and sampling during the reaction is also problematic because of the lack of lateral mixing in a thin film. Gas transfer rates, although enhanced by the large surface-to-volume ratio of the thin film geometry, may still be limited by surface area. This is particularly important for the new systems that utilize oxidative phosphorylation [28]. Commercial adoption of the thin film geometry for pharmaceutical protein production is also likely to see resistance by industry since the idea of using such a geometry for recombinant protein production is new and untested. If the CFPS technology is to be adapted as a general industrial and laboratory technology platform, a scale-up method must be developed that can avoid all of the limitations of the thin film approach and, at the same time, be efficient and universal.

223

224

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

Fig. 12.15 Cytomim system protein synthesis performance in thin film

reactions (T-F) closely mimics that of small scale test tube reactions (T-T). The 500 µL test tube reactions are constrained due to oxygen transfer limitation.

Fig. 12.16 Conceptual diagram of the stirred tank reactors used for cell-free protein synthesis.

Colleagues in the Swartz laboratory have developed an approach for batch scaleup of cell-free protein systems that gives significantly more control over gas exchange rate, addition of reagents, and sampling during the reaction. This method involves performing cell-free protein synthesis in a stirred tank reactor (Fig. 12.16). The research has focused on scaling up the more efficient cell-free systems, such as the Cytomim system, that utilize central metabolism and oxidative phosphorylation for energy generation. The stirred reactor approach is rapidly scalable to volumes of 1 L and beyond, and the reactor can be easily modified to run in batch, continuous, or semi-continuous modes. In addition, development of stirred tank reactor scale-up technology for cell-free systems allows us to conduct in vitro protein synthesis reactions in conventional fermenters, a format that is trusted by the recombinant protein expression community.

12.4 Scale-up of Reactions Requiring Oxygen

Fig. 12.17 Effect of addition of antifoam on synthesis of CAT in the

cell-free system. Reactions were performed in a 15 µL batch mode at 37 ◦ C for 6 hours in a test tube.

Performing cell-free reactions in a stirred tank reactor using the oxidative phosphorylation driven CFPS system requires oxygen delivery. In a stirred tank reactor, this is done by sparging oxygen-containing gas through the reaction mixture. The sparge rate, bubble size, and stir rate are adjusted to vary the amount of gas–liquid interfacial area and, thus, the oxygen transfer rate. The main limitation of this approach, when applied to the cell-free system, is the foaming produced by the cell extract components. The reaction foams out immediately even at low (< 1 volume of gas per volume of reaction mixture per minute (VVM)) gas flow rates. In conventional bacterial fermentations, antifoam agents have been successfully used to reduce foaming. Several classes of antifoam agents are available on the market. Although the exact mechanism of antifoam action is not known [55], all of the antifoams contain several active hydrophobic species that help to break up the bubbles. Hydrophobicity is tuned empirically [56]. While it was suspected that the addition of hydrophobic species to cell-free reactions might be detrimental to the systems’ performance, the recent experiments with protein synthesis in thin films [54] strongly suggested that hydrophobic surfaces may, in fact, be beneficial for the cell-free system. The effects of several antifoam agents on the performance of the cell-free system (synthesis of CAT) were studied at the 15 µL scale in test tubes (Fig. 12.17). Our antifoam choices spanned a large spectrum of products, ranging from glycol based agents (O-30, Sigma-Aldrich, St. Louis, MO), to poly-propylene based Triton X-705 (Sigma-Aldrich), to the ethylene-propylene oxide block-copolymer Pluronic R 31R1 (BASF, Florham Park, NJ), and to the heterogeneous silicone-based SE-15 (Sigma-Aldrich). None of the antifoam agents exerted

225

226

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

any deleterious effects on the cell-free system up to 0.01% v/v concentration. Antifoam concentrations in a typical bacterial fermentation range between 0.001% and 0.01% v/v. In cell-free systems, concentrations of 0.001% are sufficient to suppress foaming even at gas flow rates above 1 VVM. In fact, several antifoam agents enhanced the CAT synthesis yields, which further suggests that addition of hydrophobic molecular surfaces is beneficial for cell-free protein synthesis. Clearly, antifoams can be used in cell-free systems. 12.4.4 Enhanced O2 Transfer Increases ATP Concentrations

Cytomim glutamate-phosphate systems utilize oxygen as the final electron acceptor in the energy production cascade, as was demonstrated by Jewett and Swartz [28]. Measurements of oxygen concentration in such a thin-film reaction (Fig. 12.18) reveal that oxygen concentration drops to zero almost immediately after the reaction begins. The affinity constant of the electron transport chain for oxygen is reported to be in the sub-micromolar range [57, 58], which would explain why the reaction continues to synthesize protein even at very low oxygen concentrations. Although the sensor that was used (MI-730, Microeletrodes Inc., Bedford, NH) is not able to accurately measure oxygen concentrations in the sub-micromolar range, it is believed that the actual oxygen concentration in these reactions hovers near the affinity constant of the high affinity E. coli terminal oxidase cytochrome d. This was further confirmed by the fact that replacing the headspace gas with nitrogen instead of air resulted in essentially no protein synthesis. These results strongly suggested hat the supply of oxygen to the test-tube and thin-film CFPS reactions limits protein synthesis in the Cytomim system. To alleviate the oxygen limitation, the cell-free synthesis reaction was conducted in a 2 mL stirred tank reactor with oxygen feed. Pluronic 31R1 antifoam agent was used at a concentration of 0.01% v/v to prevent foaming. We compared the stirred tank reactor performance with that of a 2 mL thin film and to a 15-µL testtube reaction. As shown in Figure 12.18, the dissolved oxygen (DO) concentration in the stirred reactor was maintained far above the affinity of the electron transport chain. Steady-state ATP levels were increased by the faster oxygen delivery (Fig. 12.19). The ATP concentrations remained significantly higher in the stirred tank reactor than in the thin-film and 15-µL test tube reactors. Not surprisingly, the stirred tank reactor was also considerably more productive than the other two formats (Fig. 12.20). The initial rate of protein synthesis in the stirred tank was significantly higher than in the oxygen-limited formats. This indicated that low availability of oxygen in the cell-free mixture was, in fact, a serious limitation in thin film and test tube formats that adversely affected protein production. Relative CAT activity remained similar in all three formats at approximately 85% of the soluble yield after the 6-hour reaction.

12.4 Scale-up of Reactions Requiring Oxygen

Fig. 12.18 Dissolved oxygen concentration profile in the

glutamate-phosphate Cytomim system during CAT synthesis. Measurements were conducted using Clark Cell oxygen microprobes (Microelectrodes, Bedford, NH) during CAT synthesis reactions (37 ◦ C, 6 hours). Test tube experiments were conducted at the 15 µL scale, while thin film and stirred tank reactors were at the 2 mL scale.

Fig. 12.19 ATP concentration in a Cytomim CFPS system with different

reactor geometries. The stirred tank reactor approach is able to maintain considerably higher ATP concentrations throughout the duration of the synthesis compared with the thin film and test tube formats.

227

228

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

Fig. 12.20 CAT synthesis time course for a 2 mL stirred tank reactor

(S-T), 2 mL thin film (T-F), and 15 µL test-tube (T-T).

12.4.5 Protein Production in 1-liter Batch Reactions

To further evaluate the stirred tank scale-up approach we chose a complex, mammalian, pharmaceutically relevant protein, human insulin-like growth factor-1 (IGF-1) and conducted a 1-L cell-free reaction. Our experiment utilized all of the recent advances in the field developed to make the cell-free reactions economical and, potentially, commercially viable. We used high cell density techniques to produce the extract source cells [59] and the streamlined extract preparation procedure developed by Zawada and Liu [60] to rapidly attain the volume of extract required for the 1-L cell-free reaction. The extract is made from the E. coli strain, KC6, developed by Calhoun et al. [49] that includes stabilization of all amino acids. We utilized a cell-free protein synthesis system, based on the highly efficient Cytomim system developed by Jewett and Swartz [28], that employs oxidative phosphorylation for energy generation. For scale-up, we utilized the stirred reactor approach because it removes oxygen transfer limitation for energy generation and is the most scalable method for large reaction volumes. Additionally, the stirred tank reactor technology is widely used and accepted for recombinant protein expression in vivo. We chose IGF-1 as our expression target because it is a complex mammalian protein of high therapeutic value and because bacterial expression systems are currently unable to express IGF-1 in soluble form [61]. Similar to the chloramphenicol acetyl transferase (CAT) scale-up work discussed in Section 12.4.3, cell-free synthesis reactions of IGF-1 suffer from oxygen limitation when performed in a test-tube or in a thin film. The stirred tank approach (at 2 mL scale) solves the oxygen delivery problem and provides much more robust ATP generation. As a result, IGF-1 yields increase substantially relative to the other

12.4 Scale-up of Reactions Requiring Oxygen

Fig. 12.21 IGF-1 product accumulation in the Cytomim cell-free

protein synthesis system. The stirred tank reactor format offers significant increases in IGF-1 yields.

formats (Fig. 12.21). Clearly, the stirred reactor outperformed the test tube and the thin-film reactors and is the best reactor type for large volume reactions. To demonstrate scalability of the stirred-tank approach, we conducted IGF-1 synthesis reactions at the 2 mL, 50 mL, and the 1 L scales. For the 50 mL reaction volume, a custom stirred tank reactor was constructed with a marine type impeller. For the 1 L reactions, we used a standard BioFlo 3000 cell culture reactor from New Brunswick Scientific (Edison, NJ) to show that the cell-free reaction can be performed in a conventional cell culture reactor.

229

230

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

Fig. 12.22 IGF-1 product accumulation profile in stirred tank reactors

at different reaction volumes.

The dissolved oxygen (DO) was kept above 20% air saturation (at 30 ◦ C reaction temperature) by adjusting the agitation rate. The gas flow rate (60% O2 , 40% N2 ) was controlled at 3 VVM. To control foaming, 31R1 Pluronic-R surfactant (BASF, Florham Park, NJ) at 1/12000 v/v concentration was used. No significant foaming was observed and previous studies (Section 12.4.2) show that this surfactant does not adversely affect the cell-free system performance at the concentration used. Figure 12.22 summarizes the data for IGF-1 synthesis in stirred reactors at three different reaction volumes. It is clear from the results that the stirred tank reactor demonstrated excellent scalability from the 2 mL to the 1 L scale (Fig. 12.22). At the 1 L reaction scale, the

12.5 Conclusions and Projections

cell-free reaction produced 435 ± 43 mg total and 386 ± 3 mg of soluble IFG-1. Results were very similar for the 50 and 2 mL stirred tank reactions. The fraction of soluble protein (89% of total synthesized protein) is substantially higher than that achieved in the bacterial fermentation [61]. This fact makes the refolding process unnecessary, which greatly simplifies the IGF-1 downstream processing. The excellent scalability exhibited by the stirred tank reactor approach has a very important advantage as it allows one to tune the reaction conditions for optimal product expression at very small scales of several milliliters and potentially even smaller. In fact such small reactor systems can be used in a highly parallel fashion to simultaneously screen a large number of reaction conditions side by side. Once the optimal conditions are found, the system can be immediately scaled up to a much larger size for actual production without any change in the performance of the system. Rapid expression time, easy adjustment of reaction conditions, and rapidly scalable reactor technology now make cell-free protein synthesis a powerful and convenient tool for commercial applications as well as for the laboratory.

12.5 Conclusions and Projections

Efficient and simple batch scale-up methodology removes a serious barrier to the adoption of cell-free protein synthesis technology as an attractive alternative to in vivo protein expression. Both of the approaches discussed in this chapter are capable of scaling up CFPS to large reaction volumes while conserving system productivity. More importantly these scale-up methods are readily applied to the new, more efficient cell-free systems driven by oxidative phosphorylation. Consequently, multiple applications of the technology, and the full advantages that the proponents of cell-free protein synthesis always talk about, can now be realized. We discuss here two of the more obvious applications. 12.5.1 Personalized Medicine

One of the main advantages of cell-free protein synthesis technology is the extremely short expression cycle. This, combined with the ability to perform synthesis reactions in a highly parallel fashion, makes this technology particularly attractive for production of personalized therapeutics. Yet the deployment of cell-free technology to solve such problems has been limited due to the inability to rapidly scale-up the synthesis reactions to produce the quantity of protein needed for immunization (typically on the order of several milligrams of purified material per patient). The individuality of these vaccine targets also requires that the scale-up method is inexpensive and can be done in a multiplexed format. We believe that the thin film approach discussed in this chapter is ideally suited for this application. Yang et al. [6] and Kanter et al. [62] have successfully employed the thin film approach to demonstrate the potential of cell-free technology for treating time-critical conditions such

231

232

12 Large-scale Batch Reactions for Cell-free Protein Synthesis

as B-cell lymphoma. Murine B-cell lymphoma vaccines were rapidly synthesized in cell-free system for in vivo studies. The resulting vaccine proteins successfully protected mice from tumor challenge. These studies demonstrate that the speed, efficiency, and flexibility of the cell-free technology can be used for synthesis of very specific and effective medicines for cancer treatment. Even more important is the realization that the CFPS platform can produce very effective therapeutics against unexpected and fast changing targets. Such applications include responding to influenza viruses as well as providing fast response against biological warfare. 12.5.2 Large-scale Pharmaceutical Production

The results from the 1-L IGF-1 experiments demonstrate that the cell-free system can be engineered to scale efficiently to large volumes in conventional fermenters. This fact allows us, for the first time, to seriously consider the application of CFPS technology for large scale pharmaceutical production. In the case of IGF-1, the flexibility of the cell-free system can be used to fold the mammalian protein in the reaction and produce mostly soluble material, in contrast to the in vivo E. coli production. Even though the efficiency and the cost-effectiveness of the cell-free system have been substantially improved over the past few years, work is continuing to further simplify the system to make it more commercially attractive even at very large (1000+ L) scales.

References 1 Angenendt, P., Nyarsik, L., Szaflarski, W., Glokler, J., Nierhaus, K. H., Lehrach, H., Cahill, D. J., Lueking, A. Cell-free protein expression and functional assay in nanowell chip format. Anal Chem. 2004, 76, 1844–1849. 2 Rungpragayphan, S., Nakano, H., Yamane, T. PCR-linked in vitro expression: A novel system for high-throughput construction and screening of protein libraries. FEBS Lett. 2003, 540, 147–150. 3 Miyazaki-Imamura, C., Oohira, K., Kitagawa, R., Nakano, H., Yamane, T., Takahashi, H. Improvement of H2 O2 stability of manganese peroxidase by combinatorial mutagenesis and high-throughput screening using in vitro expression with protein disulfide isomerase. Protein Eng. 2003, 16, 423–428. 4 Takahashi, F., Ebihara, T., Mie, M., Yanagida, Y., Endo, Y., Kobatake, E., Aizawa, M. Ribosome display for selection of active dihydrofolate reductase mu-

5

6

7

8

tants using immobilized methotrexate on agarose beads. FEBS Lett. 2002, 514, 106–110. Hanes, J., Pluckthun, A. In vitro selection and evolution of functional proteins by using ribosome display. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 4937–4942. Yang, J., Kanter, G., Voloshin, A., Michel-Reydellet, N., Velkeen, H., Levy, R., Swartz, J. R. Rapid expression of vaccine proteins for B-cell lymphoma in a cell-free system. Biotechnol. Bioeng. 2005, 89, 503–511. Martemyanov, K. A, Shirokov, V. A., Kurnasov, O. V., Gudkov, A. T., Spirin, A. S. Cell-free production of biologically active polypeptides: Application to the synthesis of antibacterial peptide cecropin. Prot. Expr. Purif. 2001, 21, 456–461. Kigawa, T., Yamaguchi-Nunokawa, E., Kodama, K., Matsuda, T., Yabuki, T., Matsuda, N., Ishitani, Nureki, O., Yokoyama, S. Selenomethionine incorporation into

References

9

10

11

12

13

14

15

16

17

18

a protein by cell-free synthesis. J. Struct. Funct. Genom. 2002, 2, 29–35. Kigawa, T., Yabuki, T., Yoshida, Y., Tsutsui, M., Ito, Y., Shibata, T., Yokoyama, S. Cell-free production and stable-isotope labeling of milligram quantities of proteins. FEBS. Lett. 1999, 442, 15–19. Noren, C. J., Anthony-Cahill, S. J., Griffith, M. C., Schultz, P. G. A general method of site-specific incorporation of unnatural amino acids into proteins. Science 1989, 94, 182–188. Hirao, I., Ohtsuki, T., Fujiwara, T., Mitsui, T., Yokogawa, T., Okuni T, Nakayama, H., Takio, K., Yabuki, T., Kigawa, T., Kodama, K., Yokogawa, T., Nishikawa, K., Yokoyama, S. An unnatural base pairing for incorporating amino acid analogs into proteins. Nat. Biotechnol. 2002, 20, 177– 182. Perroud, T. D., Bokoch, M. P., Zare, R. N. Cytochrome c conformations resolved by the photon counting histogram: Watching the alkaline transition with single-molecule sensitivity. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 17537–17882. Jewett, M. C., Swartz, J. R. Rapid expression and purification of 100 nmol quantities of active protein using cell-free protein synthesis. Biotechnol. Prog. 2004, 20, 102–109. Lamla, T., Stiege, W., Erdmann, V. A. An improved protein bioreactor. Mol. Cell. Proteomics 2002, 1, 466–471. Kim, D. M., Choi, C. Y. A semicontinuous prokaryotic coupled transcription/translation system using a dialysis membrane. Biotechnol. Prog. 1996, 12, 645–649. Chekulayeva, M. N., Kurnasov, O. V., Shirokov, V. A., Spirin, A. S. Continuousexchange cell-free protein-synthesizing system: Synthesis of HIV-1 Antigen Nef. Biochem. Biophys. Res. Commun. 2001, 280, 914–917. Spirin, A. S., Baranov, V. I., Ryabova, L. A., Ovodov, S. Y., Alakhov, Y. B. A continuous cell-free translation system capable of producing polypeptides in high yield. Science 1988, 242, 1162–1164. Kudlicki, W., Kramer, G., Hardesty, B. High efficiency cell-free synthesis of proteins: Refinement of the coupled

19

20

21

22

23

24

25

26

27

28

29

transcription/translation system. Anal biochem. 1992, 206, 389–393. Alexandrov, A., Kolosova, I. M., Kolosov, M. I. mRNA stabilization in continuous flow translation system. Biochem. Mol. Biol. Int. 1996, 38, 1111–1116. Kolosov, M. I., Kolosova, I. M., Alakhov, V. Y., Ovodov, S. Y., Alakhov, Y. B. Preparative in vitro synthesis of bioactive human interleukin-2 in a continuous flow translation system. Biotechnol. and Appl. Biochem. 1992, 16, 125–133. Kigawa, T., Yokoyama, S. A continuous cell-free protein synthesis system for coupled transcription-translation. J. Biochem. 1991, 110, 166–168. Nakano, H., Shinbata, T., Okumura, R., Sekiguchi, S., Fujishiro, M., Yamane, T. Efficient coupled transcription/translation from PCR template by a hollow fiber membrane bioreactor. Biotechnol. Bioeng. 1999, 64, 194–199. Yamamoto, Y. I., Nagahori, H., Yao, S., Zhang, S. T., Suzuki, E. Hollow fiber reactor for continuous flow cell-free protein production. J. Chem. Eng. Jpn. 1996, 29, 1047–1050. Alakov, Y. B., Baranov, V. I., Ovodov, S. Y., Ryabova, L. A., Spirin, A. S., Morozov, I. Y. Method of preparing polypeptides in a cell-free translation system. 1995, US Pat. 5,478,730. Kim, D. M., Swartz, J. R. Prolonging cellfree protein synthesis with a novel ATP regeneration system. Biotechnol. Bioeng. 1999, 66, 180–188. Kim, D. M., Swartz, J. R. Regeneration of ATP from glycolytic intermediates for cellfree protein synthesis. Biotechnol. Bioeng. 2001, 74, 309–316. Sitaraman, K., Esposito, D., Klarmann, G., Le Grice, S. F., Hartley, J. L., Chatterjee, D. K. A novel cell-free protein synthesis system. J. Biotechnol. 2004, 10, 257– 263. Jewett, M. C., Swartz, J. R. Mimicking the Escherichia coli cytoplasmic environment activates long lived and efficient cellfree protein synthesis. Biotechnol. Bioeng. 2004, 86, 19–26. Calhoun, K. A., Swartz, J. R. Energizing cell-free protein synthesis with glucose metabolism. Biotechnol. Bioeng. 2005, 90, 606–613.

233

234

12 Large-scale Batch Reactions for Cell-free Protein Synthesis 30 Tohda, H., Chikazumi, N., Ueda, T., Nishikawa, K., Wantanabe, K. Efficient expression of E. coli dihydrofolate reductase gene by an in vitro translation system using phosphorothioate mRNA. J. Biotechnol. 1994, 34, 61–69. 31 Ueda, T., Tohda, H., Chikazumi, N., Eckstein, F., Wantanabe, K. 1991. Phosphorothioate-containing RNAs show mRNA activity in the prokaryotic translation system in vitro. Nucleic Acids Res. 1991, 19, 547–552. 32 Ovodov, S. Y., Alakhov, Y. B. mRNA acetylated at 2′ -OH-groups of ribose residues is functionally active in cell-free translation system from wheat embryos. FEBS Lett. 1990, 270, 111–114. 33 Bald, R., Brumm, K., Buchholz, B., Furste, J. P., Hartmann, R. K., Jaschke, A., Kretschmer-Kazemi Far, R., Lorenz, S., et al. New possibilities in RNA research through RNA engineering. In: Structural Tools for the analysis of Protein-Nucleic Acid Complexes, Lilley, D., Heumann, H., Suck, D., eds. Basel, Switzerland: Birkhauser Verlag 1992, 449–466. 34 Yoshizawa, S., Ueda, T., Ishido, Y., Miura, K., Watanabe, K., Hirao, I. Nuclease resistance of an extraordinary thermostable mini-hairpin DNA fragment, d(GCGAAGC) and its application to in vitro protein synthesis. Nucleic Acids Res. 1994, 22, 2217–2221. 35 Michel-Reydellet, N., Woodrow, K., Swartz, J. R. Increasing PCR fragments stability and protein yields in a cell-free system with genetically modified Escherichia coli extracts. J. Mol. Microbiol. Biotechnol. 2005, 9, 26–34. 36 Morozov, I. Y., Ugarov, V. I., Chetverin, A. B., Spirin, A. S. Synergism in replication and translation of messenger RNA in a cell-free system. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 9325–9329. 37 Chetverin, A. B., Spirin, A. S. Replicable RNA vectors: Prospects for cell-free gene amplification and cloning. Prog. Nucleic. Acid. Res. Mol. Biol. 1995, 51, 225–270. 38 Katanaev, V. L., Kurnasov, O. V., Spirin, A. S. Viral QB RNA as a high expression vector for mRNA translation in a cell-free system. FEBS Lett. 1995, 359, 89–92. 39 Kitaoka, Y., Nishimura, N., Niwano, M. Cooperativity of stabilized mRNA and en-

40

41

42

43

44

45

46

47

48

49

hanced translation activity in the cell-free system. J. Biotechnol. 1996, 48, 1–8. Nishimura, N., Kitaoka, Y., Mimura, A., Takahara, Y. Continuous protein synthesis system with Escherichia coli S30 extract containing endogenous T7 RNA Polymerase. Biotechnol. Lett. 1993, 15, 785–790. Nevin, D. E., Pratt, J. M. A coupled in vitro transcription-translation system for the exclusive synthesis of polypeptides from the T7 promotor. FEBS Lett. 1991, 291, 259–263. Chen, H. Z., Zubay, G. Prokaryotic coupled transcription-translation. Methods Enzymol. 1983, 101, 674–690. Ohuchi, S., Nakano, H., Yamane, T. In vitro method for the generation of protein libraries using PCR amplification of a single DNA molecule and coupled transcription/translation. Nucleic Acids Res. 1998, 26, 4339–4346. Burks, E. A., Chen, G., Georgiou, G., Iverson, B. L. In vitro scanning saturation mutagenesis of an antibody binding pocket. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 412–417. Lesley, S. A., Brow, M. A. D., Burgess, R. R. Use of in vitro synthesis from polymerase chain reaction-generated templates to study interaction of Escherichia coli transcription factors with core RNA polymerase and for epitope mapping of monoclonal antibodies. J. Biol. Chem. 1991, 266, 2632–2638. Pratt, J., Boulnois, G., Darby, V., Orr, E., Wahle, E., Holland, I. Identification of gene products programmed by restriction endonucleases DNA fragments using and E. coli in vitro system. Nucleic Acids Res. 1981, 9, 4459–4474. Sitaraman, K., Esposito, D., Klarmann, G., Le Grice, S. F., Hartley, J. L., Chatterjee, D. K. A novel cell-free protein synthesis system. J. Biotechnol. 2004, 10, 257– 263. Kim, D. M., Swartz, J. R. Prolonging cell free protein synthesis by selective reagent additions. Biotechnol. Prog. 2000, 16, 385– 390. Calhoun, K. A. & Swartz, J. R. Total amino acid stabilization during cell-free protein synthesis reactions. J. Biotechnol. 2006, 123, 193–203.

References 50 Kim, D. M., Swartz, J. R. Oxalate improves protein synthesis by enhancing ATP supply in cell-free system derived from Escherichia coli. Biotechnol. Lett. 2000, 22, 1537–1542. 51 Bukau, B., Horwich, A. L. The Hsp70 and Hsp60 Chaperone Machines. Cell 1998, 92, 351–366. 52 Diamant, S., Ben-Zvi, A. P., Bukau, B., Goloubinoff, P. Size-dependent disaggregation of stable protein aggregates by the DnaK chaperone machinery. J. Biol. Chem. 2000, 275, 21107–21113. 53 Matts, R. L., Hurst, R., Zuoyu, X. Denatured proteins inhibit translation in hemin-supplemented rabbit reticulocyte lysate by inducing the activation of the heme-regulated eif-2α kinase. Biochemistry 1993, 32, 7323–7328. 54 Voloshin, A., Swartz, J. R. Efficient and scalable method for scaling up cell-free protein synthesis in batch mode. Biotechnol. Bioeng. 2005, 91, 516–21. 55 Denkov, N. D., Cooper, P., Martin, J. Mechanisms of action of mixed solidliquid antifoams. 1. Dynamics of foam film rupture. Langmuir. 1999, 15, 8514– 8529. 56 Marinova, K. G., Denkov, N. D., Branlard, P., Giraud, Y., Deruelle, M. Optimal hydrophobicity of silica in mixed oil-silica antifoams. Langmuir. 2002, 18, 3399–3403.

57 D’Mello, R., Hill, S., Poole, R. K. The oxygen affinity of cytochrome bo’ in Escherichia coli determined by the deoxygenation of oxyleghemoglobin and oxymyoglobin: Km values for oxygen are in the submicromolar range. J. Bacteriol. 1995, 177, 867–870. 58 D’Mello, R., Hill, S., Poole, R. K. The cytochrome bd quinol oxidase in Escherichia coli has an extremely high oxygen affinity and two oxygen-binding haems: Implications for regulation of activity in vivo by oxygen inhibition. Microbiology. 1996, 142, 755–763. 59 Zawada, J. F., Swartz, J. R. Maintaining rapid growth in high-density Escherichia coli fermentations. Biotechnol. Bioeng. 2004, 89, 407–415. 60 Liu, D. V., Zawada, J. F., Swartz, J. R. Streamlining Escherichia coli S30 extract preparation for economical cell-free protein synthesis. Biotechnol. Prog. 2005, 21, 460–465. 61 Chang, J. Y., Swartz, J. R. Protein Folding: In vivo and In vitro (ed. Cleland, J. L.). American Chemical Society, Washington, D.C. 1993, Ch. 14, 178–188. 62 Kanter, G., Yang, G., Voloshin, A., Swartz, J. R., Levy R. Cell-free production of scFv fusion proteins: An effective and efficient approach for custom lymphoma vaccines. 2007, 109, 3393–3399.

235

237

Index a acetyl phosphate 18 ff., 211 adenosine triphosphate (ATP) 1, 7, 15, 17, 18ff., 28, 48, 90ff., 169, 175, 181f., 211ff., 215, 233ff. alfalfa mosaic virus (AMV) 10, 27, 32 amino acids 1ff., 35, 43, 55, 62, 76, 86, 90, 94, 114f., 135ff., 143, 155ff., 181, 184, 191, 200, 204, 209, 215ff. aminoacyl-tRNA synthetases (ARS or ARSases) 4f., 35, 43, 44, 53 arginine 6, 12f., 157, 214f. aspartic acid (aspartate) 6, 13, 176, 215 b B cell lymphoma immunotherapy 70ff., 81, 232 barley yellow dwarf virus (BYDV) 10 batch cell-free reaction format 3, 5, 17f., 19f., 29, 42, 56ff., 86ff., 92, 95, 101ff., 210, 212ff., 214, 220 batch mode reactor, see batch cell-free reaction format bilayer translation reaction method, see continuous-exchange cell-free (CECF) systems brome mosaic virus (BMV) 10 c cap structure (m7 Gppp) 9, 10 cell disruption 86 cell extracts 1, 5, 8, 9, 11ff., 16, 19, 23, 55, 57, 61, 83, 143 – bacterial S30 22, 150 – bacterial S100 2, 168 – concentration 22f. – E. coli extract (ECE) 1f., 5, 11ff., 44, 72f., 85ff., 143 – genetic modifications 12f.

– preparation 5f., 11, 14ff., 21, 54, 58, 83, 85f. – rabbit reticulocyte lysate (RRL) 1, 6, 8ff., 41 – wheat germ extract (WGE) 2, 6, 8ff., 15, 41, 87, 109, 111ff., 119, 123, 127ff., 199 cell-free expression systems, see cell-free protein synthesis systems cell-free gene expression, see cell-free protein synthesis systems cell-free production process, see cell-free protein synthesis systems cell-free protein production, see cell-free protein synthesis systems cell-free protein synthesis (CFPS) systems 1ff., 11, 54ff., 74ff., 82, 87ff., 94f., 99, 101ff., 115ff., 123, 127f., 135, 140, 175ff., 180ff., 202ff., 217, 220f., 226ff. – advantages 58f., 69, 79, 100, 108, 129, 139, 15, 176, 182, 226f. – applications 1, 52, 55ff., 79, 82f., 94, 108f., 133, 137, 157, 189, 202f., 217, 226ff. – automation 82, 94, 109, 129, 133, 186 – classification 1ff. – dialysis-mode 15, 87ff., 100ff., 111 – disadvantages 8, 55 – eukaryotic types 5ff. – functional genomic analysis 51ff., 60ff. – historical landmarks 1ff. – inhibition 8 – large-scale 103f., 122, 129, 202ff. – limitations 54, 69, 108, 124, 217f. – middle-scale 101f. – prokaryotic types 2ff., 16ff., 51, 95, 113, 124ff., 138, 162

Cell-free Protein Synthesis. Edited by Alexander Spirin and James Swartz Copyright © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-31649-6

238

Index – protein yields 53ff., 63ff., 72, 138, 166ff., 209f. – reaction scale 90, 203, 211, 217ff., 225 – reaction temperature 91, 176 – reactor types 203f. – robotic automation 109, 129f., 133 – small-scale 87ff., 90ff., 97ff., 205 – substrates stability 208ff. cell growth 13 chaperones 19, 21, 24ff., 38, 42ff., 48ff., 55, 138, 175f., 184 chloramphenicol acetyl transferase (CAT) 13ff., 18, 56ff., 70ff., 76, 85, 89ff., 93, 205, 209ff., 226ff., 235 coenzyme A (CoA) 6, 7, 70, 90, 96, 206f. combinatorial bioengineering 175ff. combined transcription-translation 3, 8f., 11, 192, 209 continuous-exchange cell-free (CECF) systems 4, 8ff., 16f., 87, 111, 113, 122, 125, 138ff., 146ff., 157, 164, 167, 169ff., 172 – bilayer translation reaction format 119f., 123f., 129 – dialysis bag format 4ff., 15ff., 58, 87ff., 100ff., 118, 122ff., 149, 216 – flat dialysis membrane format 4, 92, 205 – hollow fiber format 4, 203, 205, 228 continuous-flow cell-free (CFCF) systems 4ff., 19f., 87, 116, 119ff., 204ff. co-translational glycosylation 5ff., 38 co-translational protein folding 5, 13, 21ff., 38ff., 52ff., 61, 69, 124, 161 coupled replication-translation 8 coupled transcription-translation 2ff., 8, 19f., 23, 41, 44, 83ff.,196ff., 209 creatine phosphate (CP) 15ff., 20, 41, 43, 85, 121, 177, 199, 205 cysteine 6, 12f., 70, 93, 147, 152, 209ff. cytidine triphosphate (CTP) 18, 43, 48, 70, 85, 89, 177, 199 cytokines 68 “cytomim” system 20, 208, 214ff., 221ff. d detergents 5, 24ff., 48, 125, 141ff., 158ff., 173, 189 dialysis cell-free system, see continuousexchange cell-free (CECF) systems differential scanning calorimetry (DSC) 165, 169 dihydrofolate reductase (DHFR) 16, 24f., 63,110

disulfide-bonded proteins 5, 21, 24 disulfide bonds 5, 12, 19, 24, 40, 70ff., 175f. disulfide isomerase, see protein disulfide isomerase dithiothreitol (DTT) 19, 43, 48, 85, 89, 112, 120ff., 193f., 199 DNA template 2f., 8, 12f., 19, 54–58, 61f., 69, 83, 116, 129, 133, 179, 192ff., 202, 209f. – circular 6, 12 – linear 6, 8, 12ff., 41, 54–59, 61–67, 69, 85, 90, 97f., 100, 116ff., 139, 209f. – stability 54–63, 65 e electrospray ionization-mass spectrometry (ESI-MS) 169 elongation factors (EF’s) 36ff., 40, 49 endogenous DNA 2f. endonucleases 13, 55f., 139, 196 energy generation, see energy supply systems energy substrates 10f., 208, 217 energy supply systems 6, 18ff., 53, 216 enhancing sequences 7f. Escherichia coli extract (ECE) 5ff., 13ff., 50, 80ff., 96ff., 109, 124, 148, 159, 176, 182 – cell growth 13f., 213 – genetics 5, 11ff., 55 – preparation 5, 13f., 84f. eukaryotic extracts 3ff., 8f., 21 – advantages 8, 108 – disadvantages 8 – rabbit reticulocyte lysate (RRL) 1, 6ff., 8ff., 41 – wheat germ extract (WGE) 2, 4, 6ff., 15, 30f., 41, 112, 139 eukaryotic initiation factors (eIF’s), see initiation factors exogenous messages 1f. exonucleases 13, 55, 192 expression templates (ET) 55ff., 61, 63 expression vectors 7ff., 40, 54, 61, 116, 122 f Fab fragments 24, 184 firefly luciferase 21–23, 113f., 190ff. flat membrane reactors 4 flavin adenine dinucleotide (FAD) 23, 58 folding ligands 23, 132, 138 folinic acid (folate) 19, 22, 70, 177 functional genomic analysis 51ff., 82, 94, 96f., 132 functional proteomics 51 fusion protein production 40, 68, 74f., 78f.

Index fusion protein vaccines 68, 76ff., 79 – patient-specific vaccination 69ff., 79 – protection against aggressive tumors 78 g genetic constructs 7ff. – eukaryotic 8ff. – prokaryotic 7ff. genetic modifications (of cell extracts), see cell extracts genomic target – isolation and expression 66 globin 1f., 6, 9 globin mRNA 2 glucose 12f., 18, 20ff., 55, 206f. glucose-6-phosphate 20, 206 glutamic acid (glutamate) 6, 17ff., 55, 156f., 208f., 210f., 214, 221f. glutathione 12f., 71, 130, 176, 210 glutathione reductase 12f., 65 glyceraldehydes 3-P dehydrogenase 12, 127 glycosylase 38, 90 glycosylation 5ff. – co-translational 5f. – post-translational 5 G-proteins 148 G-protein coupled receptors (GPCR) 143f., 148f., 157, 162 granulocyte macrophage colony stimulating factor (GM-CSF) 14, 24, 68ff., 78f., 212ff. green fluorescent protein (GFP) 24, 94, 100ff., 120ff., 130, 164f., 168, 178, 190ff. guanosine triphosphate (GTP) 1, 17, 18, 22, 36ff., 48, 85 h His6 -tagged proteins 165ff. histidine protein kinase (HPK) 162f., 170ff. hollow fiber membrane reactor (HFMR), see continuous-exchange cell-free (CECF) systems, hollow fiber format human proinsulin 22, 23f. human telomerase 24 hybridoma technology 182 i immunoglobulins 69, 80, 132 in vitro expression system, see cell-free protein synthesis (CFPS) systems in vitro protein folding, see protein folding in vitro translation system, see cell-free protein synthesis (CFPS) systems inclusion bodies 55, 137, 145, 148, 161

initiation codon 7 initiation factors 36, 40, 53, 114 – eukaryotic (eIF’s) 8–10, 112 – prokaryotic (IF’s) 28ff., 42, 51 initiation of translation, see translation initiation inner membrane proteins 149ff. insulin-like growth factor-1 (IGF-1) 223ff. integral membrane protein (IMP) 29, 46, 137ff., 142ff., 173f. interaction proteomics 51 interleukins 68f., 77 internal ribosome entry sites (IRESs) 9ff. iodoacetamide (IAM) 12ff., 71ff. k keyhole limpet hemocyanin (KLH) 68ff. l β-lactamase 53, 62ff. large-scale batch reaction, see cell-free protein synthesis (CFPS) systems, batch cellfree reaction format large-scale protein production, see cell-free protein synthesis (CFPS) systems lipids 139ff., 144, 147ff., 172, 213 liposomes 125, 137, 140ff., 144ff., 155f. lipoxygenase (LOX) 7 luciferase, see firefly luciferase luciferin 22f. lymphoma immunotherapy 68f. leader sequence (of mRNA) 9f. (see also 5’-UTR) – capped 6, 9ff., 110, 112ff. – obelin 11, 192 – poly(A) 2, 9ff., 109ff. – uncapped 6, 9f. m magnesium cations (Mg2+ ) 16f., 120ff. manganese peroxidase 24, 176f., 181 membrane proteins 25, 39, 46ff., 52, 125, 137ff., 148ff., 161ff., 173 – crystallization 150f., 159, 172f. – high level cell-free expression 146 – membrane integration 46f. – resolubilization 145ff. – specific characteristics for the cellfree expression 139ff. – structural characterization 148, 150ff., 168 membrane targeting system 39, 44, 46ff.

239

240

Index messenger RNA (mRNA) 1ff., 15, 19, 22, 24, 34ff., 43, 52–57, 62f., 83ff., 109–118, 124ff., 147, 175, 188–192, 208 – capped 6, 9f., 110, 112ff. – capping 9f. – deadenylylation 9 – decapping 9 – degradation 2, 5, 8, 52, 116 – endogenous 1ff., 83, 139 – eukaryotic 2, 112 – exogenous 1f. – folding 2 – life-time 7f. – polyadenylylation 9f. – polycistronic 7 – secondary structure 7 – stabilization 13 – synthesis 8, 19, 110 – uncapped 6, 9f. methionine 6, 12f., 19, 42, 48, 93, 104 micelles 137, 141ff., 144, 146f., 147, 156, 167, 168ff. microsomes 6 misfolded proteins 21, 163, 213 mitochondrial rhodanese 24 molecular colonies 186ff. – detection 197 – gene cloning 186ff. – gene expression: role of thiol compounds 193ff. – gene expression: transcription 191ff., 202 – gene expression: translation 191ff., 199 – protocols 196ff. molecular colony technique (MCT) 187ff. MS2 phage RNA 7 multiple cloning site (MCS) 122 murine urokinase 24 n nicotinamide adenine dinucleotide (NAD) 6, 19 non-Hodgkin’s lymphoma (NHL) 68, 79 nuclear magnetic resonance (NMR) spectroscopy 92ff., 103f., 138 – high-resolution NMR spectroscopy 150f. – NMR screening 102f. – preparation of protein 131f. – solid-state NMR (ssNMR) 155ff. nucleases 2, 8, 13, 41f., 55, 62f., 116, 192, 196, 209

nucleic acid programmable protein arrays (NAPPA) 51 nucleolytic attack, see nucleases nucleolytic degradation, see nucleases nucleoside triphosphates (NTP) 5, 8, 19ff., 41, 119ff., 197, 204, 208 – concentration 8, 119 – hydrolysis 5, 20, 36, 41 – regeneration 19ff. nucleotide kinases 16 o omega () leader 10 open reading frame (ORF) 35f., 40, 58f., 63f., 83f., 98, 116f., 118 outer membrane protein (Omp or OMP) 13, 137, 157, 160f. oxalic acid (oxalate) 5, 17, 19, 70, 210 oxygen 14, 193, 195, 205f., 211ff., 216ff., 220ff. p PANOx energy system 14, 20, 57, 206f., 210f. pH 18ff., 48, 76, 85, 98f., 112, 121, 128, 138 phenyl hydrazine 6 phosphatases 14f., 20, 130, 162, 196 phosphate 3, 14ff., 20, 41, 43, 48, 76, 85, 98, 120, 153, 164ff., 198ff., 205ff., 210 phosphoenyl pyruvate (PEP) 12, 14, 17, 19f., 206, 210f. phosphorylation 5, 20, 162, 172, 206ff., 213ff. picornaviruses 9 plasmid 3, 5, 8, 13f., 41ff., 48, 53ff., 70, 89, 195, 113, 119ff., 164, 193, 198, 210 – circular 8, 139 – linearized 8, 12f., 116 poly(A) leader 10 poly(A) tail, poly(A) 3’-sequence 6, 8, 10, 110 polyethyleneglycol (PEG) 15, 18, 141 polymerase chain reaction (PCR) 3, 13, 40, 53ff., 83f., 98f., 175ff., 197 – conventional PCR 116ff. – PCR-molecular colony technique (PCR-MCT) 188 – single-cell reverse transcription polymerase chain reaction (RT-PCR) 175, 182 – single-molecule PCR 175, 177ff. – split-primer PCR 116ff., 129, 177 polyribonucleotides 2 polysomes 14f., 111f., 113f.

Index post-translational modifications 38f., 45, 149, 175 (see also glycosylation) pox viruses 10 promoter 3, 41, 116ff. prosthetic groups 23, 63 proteases 12, 24, 53, 109f., 121 protein disulfide isomerase (PDI) 21, 24, 30f., 40, 66 protein folding 5, 13, 21ff., 38f., 42f., 44ff., 50f., 69, 124 – cell extract concentration 23 – co-translational 21ff., 29, 38, 124 – effects of chaperones and foldases 24f. – effects of detergents 24f. – effects of folding ligands 23 – post-translational 38, 175 – temperature effects 21 protein kinases 38, 51, 162f. protein library 176ff. protein maturation 38ff., 49 protein synthesizer 129f. proteoliposome 140f., 155 proteomics 97 PURE system 11, 41ff. putrescine 19, 70 pyruvate 18f., 55, 205ff., 210f., 214 r rabbit reticulocyte lysate (RRL), see cell extracts Rapid Translation System (RTS) 175 recombinant proteins 70f., 138f, 161f., 223 recombinant RNA 188 redox potential 12, 21, 24, 71 release factor (RF) 35ff. (see also termination factors) repressors 7 response regulator protein 162 rhodanese, see mitochondrial rhodanese ribonuclease (RNase) 2, 6ff., 84f., 139 – micrococcal 2 – RNase E 13 ribosomes 1ff., 14, 21, 30ff., 41f., 109 ribosome binding site (RBS) 7f. ribosome recycling factor (RRF) 36 ribosome-inactivating protein (RIP) 15, 109f. (see also tritin) riboswithes 7 rifampicin 3 RNA polymerases 2f., 13, 19, 83, 191f., 198f. – animal virus-associated 3, 9 – endogenous E. coli 3

– SP6 bacteriophage 3, 83, 116 – T7 bacteriophage 3, 7, 19 s satellite tobacco necrosis virus (STNV) 10 secretory protein 40, 47f. selenomethionine incorporation 93, 98 self-inhibition phenomenon 8, 11 semi-continuous flow system (SCCF) 205 sequential cell-free protein synthesis 52ff. sequential expression analysis 54ff. serine 5, 12, 130, 210 Shine-Dalgarno (SD) sequence 7f., 36, 41 single cell RT-PCR-linked in vitro expression (SICREX) 182 – application 182ff. – development 182ff. single-chain antibodies 24 single-molecule PCR-linked in vitro expression (SIMPLEX) 176ff. – application 180ff. – expansion of SIMPLEX-based library 178ff. – quality of SIMPLEX-based protein library 178 spermidine 19, 70, 120f. split-primer PCR, see polymerase chain reaction (PCR) SP6 promoter 116, 118f. stereo-array isotope labeling (SAIL) 92 stirred tank reactor 219ff. – conceptual diagram 219 structural genomics (SG) 52, 82, 194 structural proteomics 97ff., 106 sucrose gradient analysis 14, 111 S30 extract, see cell extract S100 extract, see cell extract t temperature effects 21 termination factors 36 thin film format 211ff., 217, 226 thionins 6, 108f. thioredoxin reductase 12, 23, 65, 149 tissue plasminogen activator (TPA) 20 tobacco mosaic virus (TMV) 10f., 113 transcription factors 163 transcription-translation systems, see combined transcription-translation, coupled transcription-translation transfer RNA (tRNA) 19, 35f., 40f., 84f. translation 1ff., 56f., 120 – cap-dependent 9f., 109 – cap-independent 9f.

241

242

Index – general scheme 32, 133 translation factors 19, 23, 35, 41 translation-enhancing domains (TED) 10 translation initiation 7f., 49, 54, 112f., 165 transmembrane proteins 47, 161 transmembrane segments (TMSs) 137, 147f. transposon mutagenesis 61 trigger factor 23, 38ff. tritin 6, 15, 109 tryptophan 6, 12, 16, 93, 209f. turnip yellow mosaic virus (TYMV) 10 T7 promoter 19, 41, 44, 83, 99, 177 T7 terminator 13, 41, 83,99, 177

unnatural amino acids 5, 202 untranslated regions (of eukaryotic mRNA) (UTRs) – 3’-UTR 9ff., 113ff., 123, 192 – 5’-UTR 9ff., 33, 119ff., 129, 192 uridine triphosphate (UTP) 18 urokinase, see murine urokinase

u ultrafiltration membrane

x x-ray crystallography

3f., 42, 122, 204

v vaccine

68ff., 75ff., 209ff., 226f.

w wheat embryos 6, 109, 111f. wheat germ extract (WGE), see cell extracts

93, 97f., 150