Green Fluorescent Protein: Properties, Applications and Protocols

GREEN FLUORESCENT PROTEIN

METHODS OF BIOCHEMICAL ANALYSIS

Volume 47

GREEN FLUORESCENT PROTEIN Properties, Applications, and Protocols SECOND EDITION

Edited by

Martin Chalfie Steven R. Kain

A JOHN WILEY & SONS, INC., PUBLICATION

Copyright © 2006 by John Wiley & Sons, Inc. All rights reserved Published by John Wiley & Sons, Inc., Hoboken, New Jersey Published simultaneously in Canada No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com. Library of Congress Cataloging-in-Publication Data: Green fluorescent protein : properties, applications, and protocols / edited by Martin Chalfie and Steven R. Kain.—2nd ed. p. ; cm.—(Methods of biochemical analysis) Includes bibliographical references and index. ISBN-13 978-0-471-73682-0 (pbk.) ISBN-10 0-471-73682-1 (pbk.) 1. Green fluorescent protein—Laboratory manuals. [DNLM: 1. Green Fluorescent Proteins—analysis—Laboratory Manuals. 2. Green Fluorescent Proteins—biosynthesis—Laboratory Manuals. 3. Green Fluorescent Proteins—diagnostic use—Laboratory Manuals. 4. Luminescent Agents—analysis—Laboratory Manuals. QU 25 G795 2005] I. Chalfie, Martin. II. Kain, Steven. III. Series. QP552.G73G467 572¢.6—dc22

2005 2004029639

Printed in the United States of America 10 9 8 7 6 5 4 3 2 1

CONTENTS

First Edition Preface

vii

Preface

xi

Contributors

xiii

1.

DISCOVERY OF GREEN FLUORESCENT PROTEIN Osamu Shimomura

2.

PHOTONS FOR REPORTING MOLECULAR EVENTS: GREEN FLUORESCENT PROTEIN AND FOUR LUCIFERASE SYSTEMS J. Woodland Hastings and James G. Morin

15

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN William W. Ward

39

THE THREE-DIMENSIONAL STRUCTURE OF GREEN FLUORESCENT PROTEIN AND ITS IMPLICATIONS FOR FUNCTION AND DESIGN George N. Phillips, Jr.

67

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN David A. Zacharias and Roger Y. Tsien

83

3. 4. 5. 6.

DISCOVERY AND PROPERTIES OF GFP-LIKE PROTEINS FROM NONBIOLUMINESCENT ANTHOZOA Konstantin A. Lukyanov, Dmitry M. Chudakov, Arkady F. Fradkov, Yulii A. Labas, Mikhail V. Matz, and Sergey Lukyanov

1

121

7.

EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS Mikhail V. Matz, Yulii A. Labas, and Juan Ugalde

139

8.

THE USES OF GREEN FLUORESCENT PROTEIN IN PROKARYOTES Raphael H. Valdivia, Brendan P. Cormack, and Stanley Falkow

163

9.

THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS Amy L. Hitchcock, Jason A. Kahana, and Pamela A. Silver

179 v

vi

CONTENTS

10.

USES OF GFP IN CAENORHABDITIS ELEGANS Oliver Hobert and Paula Loria

203

11.

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA Tulle Hazelrigg and Jennifer H. Mansfield

227

12.

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS Jim Haseloff and Kriby R. Siemering

259

13.

USES OF GFP IN TRANSGENIC VERTEBRATES Sean Magason, Adam Amsterdam, Nancy Hopkins, and Shuo Lin

285

14.

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS Theresa H. Ward and Jennifer Lippincott-Schwartz

15.

PRACTICAL CONSIDERATIONS FOR USE OF REEF CORAL FLUORESCENT PROTEINS IN MAMMALIAN CELLS: APPLICATIONS IN FLUORESCENCE MICROSCOPY AND FLOW CYTOMETRY Yu Fang, Olivier Déry, Michael Haugwitz, Pierre Turpin, and Steven R. Kain

16.

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP Nicola Bevan and Stephen Rees

17.

REASSEMBLED GFP: DETECTING PROTEIN–PROTEIN INTERACTIONS AND PROTEIN EXPRESSION PATTERNS Thomas J. Magliery and Lynne Regan

305

339

361

391

Methods and Protocols Steven R. Kain

407

Index

423

FIRST EDITION PREFACE Now it is such a bizarrely improbable coincidence that anything so mind-bogglingly useful could have evolved purely by chance that some thinkers have chosen to see it as a final and clinching proof of the nonexistence of God. Douglas Adams, Hitchhikers Guide to the Galaxy

In 1955, Davenport and Nicol reported that the light-producing cells of the jellyfish Aequorea victoria fluoresced green when animals were irradiated with long-wave ultraviolet. Five years later, Shimomura et al. (1962) described a protein extract from this jellyfish that could produce this fluorescence. Independently, Morin and Hastings (1971) found the same protein a few years later. This protein, now called the Green Fluorescent Protein (GFP), was studied for many years in virtual obscurity. However, with the cloning and expression of A. victoria GFP (Prasher et al., 1992; Chalfie et al., 1994), interest in this protein has grown enormously. To steal a phrase from a recent movie, GFP has gone from “zero to hero.” As of January, 1998 at least 500 scientific publications have been published with the term “GFP” in their titles or abstracts. In the last 3 years hundreds of people have used GFP to mark proteins, cells, and organisms in a wide range of prokaryotic and eukaryotic species. They have used GFP to investigate fundamental questions in cell biology, developmental biology, neurobiology, and ecology. The interest in GFP goes beyond its utility as a biological marker. The protein is intrinsically intriguing, and investigators have sought to understand its structure, fluorescent properties, and biochemistry. This increased interest in GFP, serves as an important reminder of the usefulness of studying the biology of organisms that are not among the chosen “model” systems. The usefulness of GFP as a biological marker derives from the finding that the protein’s fluorescence requires no other cofactor: The fluorophore forms from the cyclization of the peptide backbone. This feature makes the molecule a virtually unobtrusive indicator of protein position in cells. Indeed, use of GFP as a tag suggests that the protein does not alter the normal function or localization of the fusion partner. Because permeabilization for substrate entry and fixation are not needed to localize GFP, proteins, organelles, and cells marked with this protein can be examined in living tissue. This ability to examine processes in living cells has permitted biologists to study the dynamics of cellular and developmental processes in intact tissues. In addition to the broad impact of GFP technology on basic research, several companies have also incorporated this important reporter into more applied efforts such as high throughput drug screening, evaluation of viral vectors for human gene therapy, biological pest control, and monitoring genetically altered microbes in the environment. Most notable on this list are applications for GFP in drug discovery, here the potential for real time kinetics, ease of use, and cost savings provided by this reporter are leading to the replacement of other markers such as firefly luciferase and b-galactosidase. As the development of GFP technology continues to expand, the instrument companies are introducing new and better instruments for detecting GFP fluorescent. Finally, two U.S. patents have issued (as of July, 1997) on GFP and its variants, with many more certain to appear in the next few years. vii

viii

FIRST EDITION PREFACE

As editors we find ourselves in the exciting, yet frustrating, position of producing a book that, in some aspects, will be out of date as it is published. The excitement comes from seeing the wealth of information being discovered about GFP and the many uses that people are finding for this molecule. The frustration results from the same source: New applications and information about GFP are published weekly, and no book on this subject can remain current. For example, as we write this preface, two papers have appeared on single molecule fluorescence of GFP (Dickson et al., 1997; Pierce et al., 1997), three on modifying GFP to measure calcium levels (Miyawaki et al., 1997; Persechini et al., 1997; Romoser et al., 1997), two on conditions that make GFP fluoresce red (Elowitz et al., 1997; Sawin and Nurse, 1997), and one on converting GFP to a voltage indicator (Siegel and Isacoff, 1997). We feel, however, that the contents of this volume serve as an important foundation for strategies that utilize GFP, and should guide the reader in using the marker in his or her system. We are in a period of rapid development of GFP as a tool for the biological sciences as people adapt the molecule for use in different organisms, generate variants with altered properties, and discover new ways that the protein can be used. Despite the intrinsic incompleteness of this enterprise, we have asked our colleagues to summarize the state of GFP research and they have done an admirable job. We are grateful that so many of the initial investigators that pioneered the study and use of GFP consented to write chapters for this volume. We have organized this book into four sections. We start with two introductory chapters by Osamu Shimomura on the discovery of GFP and by Woody Hastings and James Morin on bioluminescence and biofluorescence in nature. The second section describes the biochemistry and molecular biology of GFP. Bill Ward has written a very useful description of the biochemistry of GFP, pointing out both the gaps in our knowledge and the importance of physical chemical criteria for evaluating new variants of GFP. George Phillips then discusses the structure of GFP and implications of this structure for its function as a fluorescent molecule. In the last chapter in this section, Roger Tsien and Douglas Prasher describe many of these variants, their uses, and how they were derived. The third section documents various biological applications of GFP. The people we asked to contribute these chapters are the major developers of GFP in the various organisms described. As described above these chapter are incomplete in that new information and application are developing at a very rapid rate. Nonetheless, each of these chapters provides insights into how GFP is being applied to particular species. We urge readers not to look only at the organism they love best, since approaches used for one organism may prove important when applied to others. For example, the use of species-specific codon usage, presumably by allowing greater production of protein, has been very important for GFP expression in mammalian cells. Also Andy Fire has found that GFP (and bgalactosidase) expression is elevated in the nematode Caenorhabditis elegans when artificial introns are interspersed in the cDNA sequence. Both of these observations may be important for those considering optimizing GFP expression in their organisms. Finally, we asked the contributors in the third section to provide protocols on the purification of GFP and its application in various organisms and Bill Ward to contribute information on purifying GFP. Sharyn Endow and David Piston have admirably taken on the formidable task of collecting, editing, and adding to this material for the fourth section of this book. In particular, they have provided outstanding protocols for visualizing and recording GFP fluorescence. We are just beginning to learn about and use GFP, and, as always, many questions remain. Much still needs to be learned about the chemistry of fluorophore formation and

ix

FIRST EDITION PREFACE

the role of the protein structure in this formation. Additional variants are needed. In particular, variants with spectra that do not significantly overlap with those of existing variants would be very useful. Such variants could be used in multiple labeling experiments, but they may have an even greater potential. Specifically, the use of fluorescence resonance energy transfer between two fluorescent proteins would enable the generation of a system analogous to the yeast two-hybrid system (Fields and Song, 1989) to look at protein: protein interactions. The advantage of such a system is that it would not require transcription as a readout of the interaction, and could therefore be used anywhere in the cell (e.g., cytosol, plasma membrane, mitochondria). Morever, suitably marked molecules would allow the testing of protein interactions in situ in a variety of organisms. Finally, as we learn more about the properties of these protein, we need to take advantage of this information to optimize GFP fluorescence intensity, excitation and emission spectra, and protein and message stability for different uses. In the next few years, we will undoubtedly see many more uses for this protein. The future does look bright for GFP. The editors of a book have, perhaps, the easiest jobs; everyone contributes to an effort that they get the credit for. As this was the first volume that either of us had edited, we are particularly grateful for all the help that we have been given. Foremost we wish to thank the contributors who graciously consented to write chapters and then put up with our requests for rewrites and for “just a little more information” with great good humor. We are indebted to David Ades and Kaaren Janssen for starting us on this endeavor. We will get even. Finally, we are most obligated to Colette Bean, our editor at John Wiley, for showing us the ropes, keeping us on schedule, and getting us over the anxieties of producing this volume. Martin Chalfie Steven Kain

REFERENCES Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263:802–805. Davenport, D., and Nicol, J. A. C. Luminescence in Hydromedusae. Proc. R. Soc. London Ser. B 144:399–411. Dickson, R. M., Cubitt, A. B., Tsien, R. Y., and Moener, W. E. (1997). On/off blinking and switching behavior of single molecules of green fluorescent protein. Nature (London) 388:355–358. Elowitz, M. B., Surette, M. G., Wolf, P. E., Stock, J., and Leibler, S. (1997). Photoactivation turns green fluorescent protein red. Curr. Biol. 7:809–812. Fields, S., and Song, O. K. (1989). A novel genetic system to detect protein-protein interactions. Nature (London) 340:245–246. Miyawaki, A., Llopis, J., Heim, R., McCaffery, J. M., Adams, J. A., Ikura, M., and Tsien, R. Y. (1997). Fluorescent indicators for Ca2+ based on green fluorescent proteins and calmodulin. Nature 388:882–887. Morin, J. G., and Hastings, J. W. (1997). Biochemistry of the bioluminescence of colonial hydroids and other coelenterates. J. Cell. Physiol. 77:305–312. Persechini, A., Lynch, J. A., and Romoser, V. A. (1997). Novel fluorescent indicator proteins for monitoring free intracellular Ca2+. Cell Calcium 22:209–216. Pierce, D. W., Hom-Booher, N., and Vale, R. D. (1997). Imaging individual green fluorescent proteins. Nature (London) 388:338.

x

FIRST EDITION PREFACE

Prasher, D. C., Eckenrode, V. K., Ward, W. W., Prendergast, F. G., and Cormier, M. J. (1992). Primary structure of the Aequorea victoria green-fluorescent protein. Gene 111:229–233. Romoser, V. A., Hinkle, P. M., and Persechini, A. (1997). Detection in living cells of Ca2+dependent changes in the fluorescence emission of an indicator composed of two green fluorescent protein variants linked by a calmodulin-binding sequence. A new class of fluorescent indicators J. Biol. Chem. 272:13270–13274. Sawin, K. E., and Nurse, P. (1997). Photoactivation of green fluorescent protein. Curr. Biol. 7:R606–R607. Shimomura, O., Johnson, F. H., and Saiga, Y. (1962). Extraction, purification, and properties of aequorin, a bioluminescent protein from the luminous hydromedusan, Aequorea. J. Cell. Comp. Physiol. 59:223–239.

PREFACE In the preface to the first edition of this book seven years ago, we predicted that we were just beginning to see the usefulness of GFP. Although we expected many new uses for this molecule, we are amazed at the extent to which GFP, it derivatives, and similar fluorescent proteins have been used in biology today. A very approximate estimate of the general usefulness of these proteins can be seen in the number of journal articles that have citations to them. In 2004 roughly 50%, 35%, 60%, and 20% of the articles in Cell, Development, Journal of Cell Biology, and Neuron, respectively, mentioned or used these proteins (values were obtained by searching journal web site for articles with the words GFP, CFP, YFP or dsRed and then estimating the total number of articles published in the year). The fluorescent proteins are not only a general tool in basic biological research; they have also been used extensively by industry. As one unusual example, a group of entrepreneurs in the San Francisco Bay Area are pursuing a venture called Canary, Inc. which uses GFP as the basis for detecting landmines and other unexploded remnants of war. Another indicator of the growing use and importance of GFP is that this year saw the publication of the first book about GFP for the general public (Zimmer, 2005). The last seven years have also seen the introduction of many new fluorescent proteins and derivatives. Perhaps the most striking change in the field has been the discovery of the coral fluorescent proteins (Matz et al., 1999). These proteins not only provide a wealth of new colors, but also demonstrate that these types of proteins exist in a wide range of organisms. People continue to modify the fluorescent proteins and discover interesting new properties and uses. As the first edition was coming to press, we noted the GFP-based calcium sensors had just been developed. Now many more derivatives have been produced. One intriguing discovery was made by Ghosh et al., 2000. They split GFP into two separate polypeptioles. Coexpression of these proteins alone did not yield any fluorescence. Remarkably however, fluorescence could be reconstituted when covalently linked interacting protein domains brought the two parts of GFP together. This discovery has already led to an alternative to fluorescence resonance energy transfer (FRET) as a way of examining protein-protein interactions and a combinatorial method of labeling cells. These and other advances rendered the first edition of this book considerably out-ofdate. In the hope of bringing these more recent discoveries to the attention of a general audience, we were persuaded to edit a second edition of this book. As we noted in the preface to the first edition, trying to evaluate the field of fluorescent proteins is very difficult, because it is a moving target. The field is changing all the time. This dynamic feature of the field reflects its strength, but also means that reviews will always be incomplete. For that reason, we are delighted that so many of our previous authors were gracious enough to consider updating and rewriting their contributions. In addition, we are delighted to include as new authors, researchers who have done so much to move the field in new directions. As for the first editions, our editors at Wiley have been particularly supportive and diligent. These people include Luna Han, who first convinced us that this enterprise was xi

xii

PREFACE

worthwhile, and Darla Henderson and Danielle Lacourciere, who saw this second edition to completion. We are grateful for their help. Finally, we once again look forward to being astonished by even newer uses for this remarkable collection of proteins. Judging by the past, the future continues to look bright. Martin Chalfie Steven Kain

REFERENCES Matz, M. V., Fradkov, A. F., Labas, Y. A., Savitsky, A. P., Zaraisky, A. G., Markelov, M. L., and Lukyanov, S. A. (1999) Fluorescent proteins from nonbioluminescent Anthozoa species. Nat. Biotechnol. 17:969–973. Zimmer, M. (2005) Glowing Genes: A Revolution in Biotechnology, Prometheus Press, 250 pp.

CONTRIBUTORS Adam Amsterdam Center for Cancer Research, Massachusetts Institute of Technology, Cambridge, MA 02139 Nicola Bevan Screening Development and Compound Profiling, GlaxoSmithKline, Stevenage, Herts, SG1 2NY, United Kingdom Dmitry M. Chudakov Shemyakin and Ovchinnikov Institute of Bioorganic Chemistry RAS, Moscow 117997, Russia Brendan P. Cormack Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford, CA 94305 Olivier Déry BD Biosciences Clontech, Palo Alto, CA 94303 Stanley Falkow Rocky Mountain Laboratories, National Institute of Allergy and Infections Diseases, Hamilton, MT; and Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford, CA 94305 Yu Fang BD Biosciences Clontech, Palo Alto, CA 94303 Arkady F. Fradkov Shemyakin and Ovchinnikov Institute of Bioorganic Chemistry RAS, Moscow 117997, Russia Jim Haseloff Division of Cell Biology, MRC Laboratory of Molecular Biology, CB2 2QH Cambridge, United Kingdom. Present address: Department of Plant Sciences, University of Cambridge, Cambridge CB2 3EA. United Kingdom J. Woodland Hastings Department of Molecular and Cellular Biology, Harvard University, Cambridge, MA 02138 Michael Haugwitz BD Biosciences Clontech, Palo Alto, CA 94303 Tulle Hazelrigg Department of Biological Sciences, Columbia University, New York, NY 10027 Amy L. Hitchcock Department of Molecular and Celluar Biology, Harvard University, Cambridge, MA Oliver Hobert Department of Biochemistry and Molecular Biophysics, Columbia University, College of Physicians and Surgeons, New York, NY 10032 Nancy Hopkins Center for Cancer Research, Massachusetts Institute of Technology, Cambridge, MA 02139 Jason A. Kahana Department of Alzheimer’s, Research, Merck Research Laboratories, West Point, PA xiii

xiv

CONTRIBUTORS

Steven R. Kain Agilent Technologies, Inc. 3500 Deer Creek Road Palo Alto, CA 94304 Yulii A. Labas Institute of Biochemistry RAS, 117071 Moscow, Russia Shuo Lin Department of Molecular, Cell, and Developmental Biology, UCLA, Los Angeles, CA 90095 Jennifer Lippincott-Schwartz Department of Cell Biology and Metabolism, NICHD, NIH, Bethesda, MD 20892 Paula Loria Department of Biochemistry and Molecular Biophysics, Columbia University, College of Physics and Surgeons, New York, NY 10032 Konstantin A. Lukyanov Shemyakin and Ovchinnikov Institute of Bioorganic Chemistry RAS, Moscow 117997, Russia Sergey Lukyanov Shemyakin and Ovchinnikov Institute of Bioorganic Chemistry RAS, Moscow, 117997, Russia Thomas J. Magliery Department of Molecular Biophysics & Biochemistry, Yale University, New Haven, CT 06520. Present address: Department of Chemistry and Department of Biochemistry, The Ohio State University, Columbus, OH Jennifer H. Mansfield Department of Genetics, Harvard Medical School, Boston, MA 02115. Present address: Department of Biological Sciences, Columbia University, New York, NY 10027 Mikhail V. Matz Whitney Laboratory, University of Florida, St. Augustine, FL 32080 Sean Megason Beckman Institute of Biological Imaging, California Institute of Technology, Pasadena, CA 91125 James G. Morin Section of Ecological Systematics, Cornell University, Ithaca, NY 14850 George N. Phillips Jr. Department of Biochemistry, University of Wisconsin—Madison, Madison, Wi 53706 Stephen Rees Screening and Compound Profiling GlaxoSmithKline, Stevenage, Herts, SG1 2NY, United Kingdom Lynne Regan Department of Molecular Biophysics & Biochemistry and Department of Chemistry, Yale University, New Haven, CT 06520 Osamu Shimomura The Photoprotein Laboratory, Falmouth, MA 02540 Kirby Siemering Division of Cell Biology, MRC Laboratory of Molecular Biology, Cambridge, United Kingdom, CB2 2QH Pamela Silver Department of Systems Biology, Harvard Medical School, Boston, MA; Department of Cancer Biology, Dana-Farber Cancer Institute, Boston, MA 02115 Roger Y. Tsien Department of Pharmacology, University of California, San Diego, La Jolla, CA 92093 Pierre Turpin BD Biosciences Clontech, Palo Alto, CA 94303

CONTRIBUTORS

Juan Ugalde Laboratory of Bioinformatics and Gene Expression, University of Chile, Santiago, Chile Raphael Valdivia Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford, CA. 94305 Present address: Department of Molecular Genetics and Microbiology, Duke University, Durham , NC 27710 Theresa H. Ward Immunology Unit, Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London WC1E 7HT, United Kingdom William W. Ward Department of Biochemistry and Microbiology, Rutgers University, Cook College, New Brunswick, NJ 08901 David A. Zacharias The Whitney Laboratory for Marine Bioscience, University of Florida, Department of Neuroscience, St. Augustine, FL

xv

Figure 1.1. Mid-summer specimens of Aequorea aequorea photographed in natural environment (top) and in seawater supplemented with KCl in darkroom (bottom), both at the University of Washington’s Friday Harbor Laboratories.

Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

Figure 2.7. Streaks of luminescent bacteria photographed by their own light, showing two strains of Photobacterium fischeri, one of which emits yellow light by virtue of having YFP (yellow fluorescent protein). The other lacks YFP, emitting only blue light.

Figure 2.10. Transgenic tobacco plant carrying the firefly luciferase gene photographed by its own light. The continuous luminescence occurs following the uptake of luciferin by the roots. [From Ow et al. (1986).]

Figure 2.11. Bacterial colonies carrying four different beetle luciferase genes cloned from the ventral organ, distinguished by their different luminescence colors: green, yellow-green, yellow and orange (Wood et al., 1989).

3

2 5 1 6 4

9

11

8 7

10

10 A Figure 4.1. End-on (top left) and side (top right) views of the cylindrical b-can structure of GFP. Eleven strands of b-sheet form an antiparallel barrel with short helices forming lids on each end. The fluorophore is inside the can, as a part of a distorted a-helix, which runs along the axis of the cylinder. The GFP usually forms dimers in the crystal, aligned largely along the sides of the cylinders. Drawing by Ribbons (Carson, 1987), coordinates (Protein Data Bank entry 1GFL). [From Yang et al. (1996.) Reprinted with permission from Nature Biotechnology.]

Phe165

Phe165 Val150

Gln183 Arg96 Gln69 Thr62

Ile167 His148 Thr203 Tyr145

Gln94 Gly67

Tyr66

Ser205

Gln183 Arg96 Gln69 Thr62

Ile167 His148 Thr203

Val61 Glu222 Ser65

Val150

Tyr145 Val68

Gln94 Gly67

Tyr66 Val61 Glu222 Ser65

Val68

Ser205

Figure 4.3. Stereoview of the fluorophore and its environment. His148, Gln94, Arg96, and Glu222 can be seen on opposite ends of the fluorophore and probably stabilize anionic resonant forms. Water molecules, charged, polar, and nonpolar side chains all contact the fluorophore in various ways.

Figure 5.3. Visual appearance of E. coli expressing four differently colored mutants of GFP. Clockwise from upper right: Blue mutant P4-3 (= Y66H, Y145F) (Heim and Tsien, 1996); cyan mutant W7 (Y66W, N146I, M153T, V163A, N212K) (Heim and Tsien, 1996); green mutant S65T (Heim et al., 1995); yellow mutant 10C (= S65G, V68L, S72A, T203Y) (Ormo et al., 1996). In each of these lists of mutants, the mutation most responsible for the special alterations is underlined, while the other substitutions improve folding or brightness. The bacteria were streaked onto nitrocellulose, illuminated with a Spectraline B-100 mercury lamp (Spectronics Corp., Westbury, NY) emitting mainly at 365 nm, and photographed with Ektachrome 400 slide film through a low-fluorescence 400 nm and a 455-nm colored glass long-pass filter in series. The relative brightness of the bacteria in this image is not a good guide to the true brightness of the GFP mutants. Expression levels are not normalized, and the 365 nm excites the blue and cyan mutants much more efficiently than the green and yellow mutants, but the blue emission is significantly filtered by the 455-nm filter required to block violet haze.

Figure 5.4. The crystal structure of dimeric GFP (1GFL) (Yang et al., 1996a). The residues A206 (red), L221 (orange), and F223 (lavender) are shown as ball-and-stick representations. Replacing any of these residues with the positively charged residues lysine or arginine effectively monomerizes the protein.

Figure 5.5. GFP biosensors. (A) GFP can be engineered to be directly sensitive to a small molecule of interest. (B) Insertion of a conformationally dynamic domain into GFP can result in a chimera in which the fluorescence properties of GFP are modulated by a change in conformation of the domain.

Figure 5.5. (continued) (C) Similarly, proteins or peptides with dyanmic, associative properties can be fused to the N and C termini of circularly permuted GFPs, thereby reporting on the changes in the association in response to a stimulus.

Figure 5.6. A backbone representation of the three-dimensional structure of GFP (1EMG) (Elsliger et al., 1998). The residues where circular permutations are permitted while retaining fluorescence are color highlighted. E142, hot pink; Y143, gray; Y145, dark blue; H148, fuchsia; D155, yellow; H169, red; E172, light blue; D173, orange; A227, Cyan; I229, light purple. These residues represent sites where the main chain can be interrupted. In most cases, resumption of GFP sequence can occur one to four residues following the initial interruption.

Figure 7.2. (This also appears in color insert.) Copepoda species that yielded fluorescent GFPlike proteins (from Shagin et al., 2004). Images were taken by fluorescent microscope using combined illumination with white light and standard FITC filter set. (a) Pontellina plumata. Inset magnifies the head. (b) Labidocera aestiva. (c) Compare Pontella meadi.

Figure 8.1. Fluorescence images of sporulating B. subtilis cells expressing transcriptional and translational GFP fusions. Two sporangia are shown per panel. (A) Forespore-specific expression of a sF-dependent SspE2G-GFP fusion. (B) Mother cell-specific expression of a sE-dependent cotEgfp fusion. (C) Localization of a SpoIVFB-GFP translational fusion (note localized fluorescence seen as a shell at one end of each sporangium). Courtesy of O. Resnekov and C. Webb, Harvard University.

Figure 8.4. Laser scanning confocal images of R. meloliti infection threads in plant root hairs. The R. meloliti bearing a plasmid with a trp-gfp fusion was used to infect alfalfa plants. Infection threads can be seen within individual root hair as they extend toward the main root body (stained red with propidium iodide).

Figure 8.5. Visualization of S. typhimurium intracellular-specific gene expression by fluorescence microscopy. S. typhimurium bearing a pagA::gfp fusion shows gene induction inside an infected mammalian cell but not in the extracellular medium. The corresponding DIC images show the relative topology of bacteria with respect to the infected cell.



Figure 10.1. Examples of subcellular structures visualized with GFP. (A) Presynaptic specializa-

tions: Transgenic animals expressing a synaptobrevin-GFP fusion construct reveal localization of GFP to synaptic sites in all neurons. Here, punctate fluorescence can be seen in the SAB motor neurons. (Reprinted from Nonet, M., Visualization of synaptic specializations in live C. elegans with synaptic vesicle protein-GFP fusions, J. Neurosc. Methods, 89:33–40. Copyright © 1999, with permission from Elsevier.) (B) Splicing speckles: Live transgenic animals expressing rescuing unc75::GFP show GFP localization in subnuclear puncta predicted to be splicing speckles (Loria et al., 2003). Here, multiple puncta can be seen in the nucleus of a ventral cord motorneuron (arrows). The corresponding Nomarski-DIC image is on the right. (C) Transcription factor target sites: elt1::GFP binds to its own promoter and leads to discrete fluorescent foci in nuclei. The embryonic gut nuclei of eight-cell embryos homozygous for a transgenic array containing fully rescuing elt2::GFP are shown. Many of the nuclei show two striking and intense foci of fluorescence (arrows). The image represents a stack of serially collected 400-nm optical sections projected without further manipulation. [Courtesy of Fukushige et al. (1999).] (D) Dense bodies: Transgenic animals expressing a rescuing unc-97::GFP fusion construct show localization to discrete spots and lines that correspond to dense bodies (DB) and M lines (M) of the body wall muscle. The expression of the unc-97::GFP reporter gene can be monitored in live or fixed animals. Some subcellular structures appear more crisp in formaldehyde-fixed animals (shown here), although they are also distinctly visible in live animals. (Reproduced from The Journal of Cell Biology, 1999, 144:53 by copyright permission of the Rockefeller University Press.) (E) Nuclear spindles: Multiphoton image series of GFP::b-tubulin in a live wild-type embryo from metaphase through telophase reveals centrosome dynamics. Elapsed time from the first frame is shown in minutes:seconds. (Reprinted from Molecular Biology of the Cell, 2001, 12:1751–1764) with permission by the American Society for Cell Biology.)

A

C

E

B

D

A

B

C

Figure 10.2. Examples of axon anatomy visualized with GFP. (A,B) Growth cone: Confocal micrograph of live L2 larvae at 17 hours post-hatching showing GFP driven under the unc-47 promoter that expresses in the ventral nerve cord, the DD and VD cell bodies (open arrows), the DD commissures (arrowheads), and the VD growth cones (solid arrows). (B) High magnification confocal micrograph of VD growth cones shown above (arrows). Filopodia extend from the round central mass of the growth cone on left. The middle growth cone is anvil-shaped. The right growth cone is extending a single finger toward the dorsal nerve cord. Existing embryonic DD commissures are marked with arrowheads. [Reprinted from Knobel et al. (1999), Development 126:4489–4498 with permission from The Company of Biologists Ltd.] (C) Axon co-labeling: Double labeling of axons in the ventral nerve cord (schematic). 3D image stacks of the ventral cord of double-labeled animals were recorded with a confocal microscope and subjected to a deconvolution algorithm to improve spatial resolution. Image shows an interneuron labeled with CFP (glr-1::GFP) and motorneuron axons labeled with GFP (unc-4::GFP). The image on the right is a cross-section through the ventral nerve cord at the position marked by the arrowhead in the left image. The orientation of images is depicted in the schematic. (Reprinted by permission of Wiley-Liss, Inc., a subsidiary of John Wiley and Sons, Inc. from Hutter, H., New ways to look at axons in Caenorhabditis elegans, Microscopy Research and Technique, copyright © 2000.)

D

E * * * *

*

Figure 10.2. (continued) (D) PVT axon morphology: The pioneer neuron PVT has a large cell body (large arrow), situated in the pre-anal ganglion that sends out an anteriorly-directed process in the ventral nerve cord (small arrows). Original EM reconstructions suggested that the axon of PVT terminated in the posterior body. A zig-2::GFP reporter shows strong expression in PVT. This analysis shows that the PVT axon in fact extends the entire length of the nerve cord and terminates within the nerve ring (arrowhead). Asterisk denotes gut autofluorescence. (E) PVD axon morphology: Previously, the processes of the PVD interneuron were not completely reconstructed by EM. Analysis of a GFP reporter for PVD shows that the axon displays an elaborate branching pattern not previously appreciated (arrow indicates position of cell body, and asterisk denotes gut autofluorescence).

A

B wild type

mec-3

Figure 10.4. Examples of the use of GFP as a tool to identify cells. (A) Identifying cells for electrophysiological recording: The gcy-5::GFP reporter was used to identify the chemosensory neuron ASER for in situ patch-clamp recording. (Top) Nomarski-DIC micrograph showing exposed neuronal cell bodies and a recording pipette sealed to ASER. Scale bar is 10 mm, anterior is left. (Bottom) Fluorescent micrograph of the same field above showing the GFP label in ASER, which allowed the unambiguous identification of the neuron. [Courtesy of Goodman et al. (1998).] (B) Analyzing of mutant cells in vitro: GFP-positive touch neurons from wild type (left) and mec-3 mutant animals (right) were enriched by fluorescence-activated cell sorting, cultured in vitro, and used to isolate RNA for DNA microarray analysis. This technique allowed the identification of mec-3-dependent genes and demonstrated that, unlike whole-worm RNA analysis, genes expressed in only a few cells can be identified systematically. [Reprinted from Zhang et al. (2002) by copyright permission of the Nature Publishing Group.]

Figure 11.4. GFP-tagged Gag proteins of two non-LTR retrotransposons, HeT-A and TART, shown in interphase Drosophila tissue culture cells. When transfected singly, HeT-A GFP-Gag is targeted to telomeres, but TART GFP-Gag is not. A. HeT-A GFP-Gag. B. TART GFP-Gag. Left panels: GFPGag; middle panels—anti-HOAP, which labels the telomeres; right panels—superimposed images, with DAPI-stained chromosomes. When co-transfected, Het-A Gag recruits TART-Gag to telomeres (not shown here). (From Rashkova et al., 2002. Courtesy of Mary-Lou Pardue and Svetlana Rashkova.)

Figure 11.7. GFP-a-Tubulin (Tub) in dividing germ cells in the ovary. During oogenesis, a germ stem cell gives rise to a daughter cystoblast, which subsequently undergoes 4 incomplete mitotic divisions to produce a cyst of 16 interconnected germ cells, one of which is destined to become an oocyte. The germ cells of developing cysts are connected by the spectrin-rich fusome, which plays an important role in orienting these cells and in oocyte specification. GFP-Tub was expressed in the germ cells by the Gal4-UAS system. In this mitotic cyst producing 8 germ cells, one end of each spindle is associated with the fusome (red). Later, after 16 cells are formed, the fusome is required to polarize the interphase microtubule network, an event that accompanies oocyte specification. (From Grieder et al., 2000. Courtesy of Allan Spradling.)

Figure 11.11. GFP-labeled nos mRNA in living (A, D, E) and fixed (B, C, F) Drosophila egg chambers and embryos. (A) GFP-labeled nos RNA is visible in the oocyte of early egg chambers. Excess MCP-GFP fusion protein that is not bound to nos RNA enters the nurse cell nuclei. In these egg chambers, MCP-GFP fusion protein alone is also expressed in the follicle cells. (B) During midoogenesis, GFP-labeled nos RNA is transiently localized to the anterior margin of the oocyte. Nurse cell and follicle nuclei appear yellow/orange due to the overlap of Hoescht (red) and unbound MCP-GFP (green). (C) Z-series projection of the posterior of a stage 13 oocyte showing particles of GFP-labeled nos RNA at the cortex. (D) GFP-labeled nos RNA is also detected in particles at the posterior cortex of the early embryo. (E) GFP-labeled nos RNA in pole cells during gastrulation. (F) GFPlabeled nos RNA overlaps Vasa protein (detected by anti-Vasa immunostaining in red) in newly formed pole cells. (From Forrest and Gavis, 2003. Figure and legend courtesy of Elizabeth Gavis.)

Figure 11.14. A GFP reporter to study innate immunity. The drosomycin (dros) promoter was used to drive expression of GFP in transgenic flies. A. dros-GFP is induced strongly in flies that are immunized (challenged by microbial infection), but only at low levels in unchallenged flies (compare the top and bottom flies). B. dros-GFP is expressed in the fat body of immunized larvae (top) but not in control larvae (bottom). C. Dissected fat body of an immunized adult. Higher magnification image of the immunized larva shown in B. (From Ferrandon et al., 1998. Courtesy of Dominique Ferrandon.)



Figure 11.13. GFP as a reporter for transcriptional silencing. Reporter constructs were designed to measure the effects of heterochromatic gene silencing on two adjacent genes, mini-white and UASGFP. With this system, de-repressive effects of Gal4-induction of UASGFP in the context of flanking heterochromatin could be determined. Shown are eyes of flies carrying two different reporter insertions, x21 (A) and x18.4.1 (B), and two different Gal4 drivers (A5CGAL in A, and GMRGAL in B). The left panels are light microscopy images of the eyes, and the middle and right panels are fluorescent images showing the red pigments (middle) or both the red pigments and GFP (right). In some cells, uncoupling of UASGFP and mini-white expression occurs: the white lines indicate areas where GFP is expressed, but mini-white is silenced. (From Ahmad and Henikoff, 2001. Coutesy of Steve Henikoff and Kami Ahmad.)

Figure 11.15. Two-color GFP shows that RNAi is cell-autonomous in Drosophila embryos. In the control panel (left), all segments of the embryo express a GFP fusion protein from a transgene driven by the poly-ubiquitin promoter. The posterior domain of each segment (two segments are shown) also expresses a GFP fusion protein from a UAS-regulated transgene in combination with en-GAL4. The overlap of GFP and CFP appears yellow. In the right panel, both transgenes are expressed in addition to an en-Gal4-induced ds RNA that targets the RNA encoding GFP. Expression of the GFP fusion is specifically repressed in the en domain, indicating that RNAi is cell autonomous and cannot spread to the anterior compartments. (From van Roessel et al., 2003.

A

B

C

D

E

F

Figure 11.16. GFP balancer chromosomes. Balancer chromosomes were constructed that carry UAS-GFP and Kruppel (Kr)-GAL4. Flies bearing these balancers express GFP in the Kr pattern. In embryos, zygotic expression of GFP commences during germ band extension and can be detected at all subsequent stages. A. Stages 4–5, cellularization. The yellow signal is yolk. B. Stages 8, early germ band extension. C. Stages 9–12, germ band extension. GFP is first detected. D. Stages 13–14, germ band retraction. E. Stages 16. F. Stages 17, late embryo. a = amnioseroasa, bo = bolwig’s organ, cd = central domain, s = spiracles, y = yolk. (From Casso et al., 2000. Courtesy of Tom Kornberg and Dave Casso.)

Figure 13.1. Stable transgenic zebrafish expressing GFP in specific tissues. (A, B) GATA-1 GFP expression in hematopoietic cells (Long et al., 1997). (C, D) GATA-2 BAC GFP expression in neuronal cells (Shuo Lin, unpublished). (E, F) Rag-1 GFP expression in olfactory sensory neurons (Jessen et al., 1999). (G, H) Rag-1 BAC GFP expression in thymus (Jessen et al., 1999). (I) Insulin GFP expression in pancreatic beta cells (Huang et al., 2001). (J) POMC GFP expression in pituitary cells (Liu et al., 2003). (K) FLK GFP expression in vascular cells (Cross et al., 2003).

Figure 13.2. GFP expression in transgenic mouse and chick. (A, B) GFP expression in neural tube and neural crest following electroporation of chick with a GFP encoding plasmid (green) and anti-HNK1 immunostaining (red) to mark neural crest. (A) Lateral view of whole mount (Maria Elena de Bellard and Marianne Bronner-Fraser, unpublished). (B) Cross section through neural tube with DAPI staining (blue) to mark nuclei (Ed Coles and Marianne Bronner-Fraser, unpublished observations). (C) Yolk sac of an E9.5 transgenic mouse showing e–globin GFP expression in red blood cells (Dyer et al., 2001; Elizabeth Jones, unpublished observations). (D) Section through cerebellum of Calbindin BAC GFP transgenic mouse showing expression in Purkinje cells (Xiangdong William Yang and Nat Heintz, unpublished observations).

A

ER

ssKDEL-GFP

ER exit sites

Sec13-GFP

Golgi

GalT-GFP

PM + Golgi

GPI-GFP

B

Figure 14.1. Examples of GFP chimeras and their subcellular localization. (A) Seady-state distribution of several proteins. (B) Confocal images of a cell expressing the secretory cargo protein VSVG-GFP imaged by time lapse as the protein leaves the Golgi apparatus. Eight images at 10-s intervals were overlaid. (Boxed areas) The route of a single post-Golgi carrier to the cell periphery. [Courtesy of Hirschberg et al. (1998).] A

B

Control

Staurosporin

Figure 15.4. Dual-color analysis for monitoring Bid activation with DsRed2. HeLa cells were transiently cotransfected with plasmids encoding the fusion protein Bid-DsRed2 and a mitochondriatargeted ZsGreen1 (ZsGreen1-Mito). (A) Before induction of apoptosis, Bid-DsRed2 is localized in the cytosol and ZsGreen1-Mito labels the mitochondria. (B) After induction of apoptosis with 1 mM staurosporine for 3 h, the relocalization of Bid-DsRed2 to mitochondria as revealed by the colocalization with the mitochondria marker ZsGreen1-Mito. Images were taken with a 100¥ objective using Chroma filter sets hq460/40x, 490dclp, and hq515/30m for ZsGreen1 and using hq545/50x, 580dcxr, and hq630/60m for DsRed2.

Figure 15.5. Detection of three fluorescent proteins by fluorescent microscopy. HeLa cells were separately transfected with plasmids pAmCyan1-N1, pZsYellow1-N1, and pHcRed1-N1, mixed, and observed by microscopy using Chroma Technology Corp. filter sets d440/40x, 470dcxr, and d500/40m for AmCyan1, using hq500/40, 530dclp, and hq550/40m for ZsYellow1, and using hq575/50x, 610dclp, and hq640/50m for HcRed1.

GABA R2 cyan

GABA RI yellow

Overlay

FRET

Figure 16.3. Confocal visualization of GABA-B receptor heterodimerization. HEK 293T cells were transfected with the fusion proteins, GABA-BR2/cyanRCFP and GABA-BR1/yellowRCFP. From left to right in the figure, the images show cellular expression of GABA-BR2/cyanRCFP (excitation at 433 nM, emission at 475 nM), GABA-BR1/yellowRCFP (excitation at 488 nM, emission at 525 nM; the overlay of the first two images demonstrate that both proteins are expressed at the same site), and the FRET signal (excitation at 433 nM; emission at 525 nM). The FRET event demonstrates that the proteins are in close proximity.

Figure 17.1. Schematic of GFP dissection. (A) The original system used by Ghosh et al. (2000) split GFP at 157–158. The reassembled GFP, fused to antiparallel leucine zipper peptides (blue), is depicted with the N- and C-terminal fragments are colored green and red, respectively. (B) The dissection points discussed in the text are highlighted. Those in bold have been the most generally successful. Created with PyMOL (http://www.pymol.org) from PDB entries 1EMA and 1SER.

Figure 17.2. Multicolor reassembly of fluorescent proteins. Reassembly of CFP(155–238) with (A) YFP(1–172), (B) GFP(1–172), (C) BFP(1–172), and (D) CFP(1–172) results in yellow, green, blue, and cyan cells. [Adapted from Hu and Kerppola (2003) with permission.]

1 DISCOVERY OF GREEN FLUORESCENT PROTEIN Osamu Shimomura The Photoprotein Laboratory, Falmouth, MA

1.1 DISCOVERY OF GFP It was early July in 1961. Dr. Frank Johnson and I were studying the bioluminescence of the jellyfish Aequorea aequorea (see Section 1.4 concerning the species name) at the Friday Harbor Laboratories of the University of Washington, located on a small island near Victoria, British Columbia, Canada. Since early morning of that day, we were trying to develop a practical method to isolate the light-emitting matter of the jellyfish, a substance later named “aequorin” (cf. Shimomura et al., 1962; Shimomura, 1995a), of which we had found the basic principle of solubilization and extraction the day before. In the course of our experiments, however, I became deeply annoyed and also puzzled when I realized that the light emitted from the extract was clearly blue, contrary to our expectation of green light identical to the luminescence of live specimens. A mature specimen of A. aequorea looks like a transparent, hemispherical umbrella, with its mouth at the underside of the body (Fig. 1.1, top). Average mature specimens measure 7–10 cm in diameter. Due to the high transparency of the body, the jellyfish can function as a magnifier lens when the mouth is fully open. The light organs, consisting of about 200 tiny granules, are distributed evenly along the edge of the umbrella, making a full circle. Soaking a specimen of the jellyfish in a dilute potassium chloride (KCl) solution in a darkroom causes the light organs to luminesce, exhibiting a ring of bright green light in the darkness (Fig. 1.1, bottom). If a specimen is soaked in distilled water, a green ring is first observed, which gradually changes into blue with the progress of the cytolysis of cells. Under an ultraviolet light, a specimen of fresh jellyfish exhibits a ring of brilliant green fluorescence, similar to the luminescence caused by KCl. Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

1

2

DISCOVERY OF GREEN FLUORESCENT PROTEIN

Figure 1.1. Mid-summer specimens of Aequorea aequorea photographed in natural environment (top) and in seawater supplemented with KCl in darkroom (bottom), both at the University of Washington’s Friday Harbor Laboratories. See color insert.

The margin of the umbrella containing the light organs can be cut off with a pair of scissors, yielding a 2- to 3-mm-wide strip called the “ring.” When the rings obtained from 20–30 jellyfish were squeezed through a rayon gauze, a dimly luminescent, turbid liquid called the “squeezate” is obtained. The granules of light organs in the squeezate can be collected by filtration or centrifugation. When the granules are mixed with dilute neutral

DISCOVERY OF GFP

buffer solutions, they are cytolyzed and emit light. When mixed with a pH 4.0 buffer, however, the granules are cytolyzed without light emission, preserving the light-emitting activity in the solution. After the removal of cell debris by centrifugation, the pH 4.0 cellfree solution can be luminesced by the addition of a neutral buffer solution containing Ca2+. These are the outline of the procedure we were doing on that day in July 1961, and I saw that the luminescence of the neutralized solution was blue, contrary to our expectation. I doubled, then tripled, the number of the jellyfish used in each experiment in order to make the final luminescence stronger and clearer, but these efforts only helped to confirm my observation. My question concerning the seeming discrepancy remained in my consciousness, until we found an explanation more than 10 years later. After returning to Princeton University with the jellyfish extracts, we purified the light-emitting substance. The substance obtained was a protein capable of emitting light in the presence of Ca2+; the protein was named aequorin. During the purification of aequorin, we noticed the existence of a green fluorescent protein in the jellyfish extract. Upon column chromatography, a green fluorescent band moved closely together with the band of aequorin on a Sephadex G-100 column and moved ahead of the aequorin band on a DEAE-cellulose column. Although the presence of a green fluorescent substance in the light organs was previously known (Davenport and Nicol, 1955), it was the first time that the substance was isolated and recognized to be a protein. Our observation was mentioned in our first full article on the purification and characterization of aequorin (Shimomura et al., 1962), in a footnote, as follows: A protein giving solutions that look slightly greenish in sunlight though only yellowish under tungsten lights, and exhibiting a very bright, greenish fluorescence in the ultraviolet of a Mineralite, has also been isolated from the squeezates. No indications of a luminescent reaction of this substance could be detected.

The first measurements of the luminescence spectrum of aequorin and the fluorescence spectrum of the green protein were reported quickly (Johnson et al., 1962). The luminescence spectrum of aequorin was broad, with a peak at 460 nm. The fluorescence spectrum of the green protein was sharp, with a peak at 508 nm. Apparently, the light organs of the jellyfish contain these two proteins, aequorin and the green protein, of which the former emits blue light in the presence of Ca2+ and the latter emits green fluorescence when excited. The green protein was later called green fluorescent protein, GFP (Hastings and Morin, 1969). One average-sized specimen contains 20–30 mg of aequorin (Shimomura and Johnson, 1979), and each of its about 200 light organs contains approximately 0.1 mg of aequorin and 0.02 mg of GFP (Morise et al., 1974; Cutler, 1995). How can a protein, aequorin, luminesce just by the addition of Ca2+, even in the absence of oxygen? Why is the luminescence of a live jellyfish green, while aequorin emits blue light? Regarding the first question, it seems clear that the luminescence is produced by an intramolecular chemical reaction of aequorin triggered by Ca2+. Thus, we would need to understand the mechanism of this intramolecular reaction, which was a formidable task to accomplish at the time. To answer the second question, it would be necessary to consider two possibilities: (1) a filtering effect by the green protein or something else that shifts the emission maximum of aequorin luminescence to longer wavelength and (2) an energy transfer from aequorin molecules to the green protein by a certain mechanism. Considering that the fluorescent protein was created by nature presumably under some selective pressure, the possibility of an energy transfer would be more likely. In those days, however, we were not concerned with the details of the energy-transfer mechanism;

3

4

DISCOVERY OF GREEN FLUORESCENT PROTEIN

we merely assumed that the green protein absorbed the blue light of aequorin, and then reemitted the absorbed energy as green light (i.e., an energy transfer by the trivial mechanism). We deferred the studies of these subjects for the next 5 years, because of various reasons. Aequorin is an unusual protein that contains an energy-producing source for light emission inside the molecule, resembling a luciferin in this respect. However, it seemed inappropriate to designate aequorin a luciferin because of its heat-labile and nondiffusible nature. In 1965, we discovered the second example of a bioluminescent protein that contains the energy source of luminescence in the molecule from the marine tubeworm Chaetopterus, and we proposed to use the general term “photoprotein” to refer to this type of protein (Shimomura and Johnson, 1966). Thus, a photoprotein is a naturally occurring bioluminescent protein that is capable of emitting light in proportion to the amount of the protein (Shimomura, 1984). The term is now widely used, and many different kinds of photoprotein are presently known—for example, Ca2+-sensitive photoproteins from coelenterates (aequorin, obelin, mnemiopsin) and protozoa (thalassicolin); superoxide-activated photoproteins from scaleworm (polynoidin) and the clam Pholas (pholasin); and an ATP-activated photoprotein from a Sequoia millipede Luminodesmus (Shimomura, 1984).

1.2 ISOLATION AND PROPERTIES OF THE GREEN FLUORESCENT PROTEIN Ridgway and Ashley (1967) reported the first successful application of aequorin bioluminescence. They microinjected aequorin into barnacle muscle single fibers, and they monitored the concentration changes of Ca2+ that occur during muscle contraction. The study clearly demonstrated the usefulness and importance of aequorin in the studies of intracellular calcium, causing a rush of requests for this photoprotein. For the efficient and productive use of aequorin, detailed knowledge on the properties of aequorin and the mechanism of light emission became necessary. Thus, we decided to try to solve the chemical mechanism of aequorin luminescence, an intramolecular reaction. It seemed to be an exceedingly difficult, almost unachievable undertaking at the time. After several years of strenuous efforts, however, we had the luck to be able to uncover a large part of the intramolecular chemistry involved in the Ca2+ triggered luminescence of aequorin, including the chemical structure of the functional moiety “coelenterazine” in the protein and also the means to regenerate spent aequorin into the original, active aequorin (Shimomura and Johnson, 1969, 1972, 1973, 1975). During the same period, green fluorescent proteins similar to Aequorea GFP were found in a number of other bioluminescent coelenterates (Hastings and Morin, 1969; Morin and Hastings, 1971a,b; Wampler et al., 1971, 1973; Cormier et al., 1973, 1974; Morin, 1974); those green fluorescent proteins apparently function as the light emitter of in vivo bioluminescence, as in the case of Aequorea. Green fluorescent protein was not found in the jellyfish of Scyphozoa (such as Pelagia and Periphylla) and Ctenophora (such as Mnemiopsis and Beroë). The following genera of bioluminescent coelenterates contain GFP: Class Hydrozoa The jellyfish Aequorea The jellyfish Mitrocoma (synonym Halistaura) The hydroid Obelia The jellyfish Phialidium (hydroid Clytia)

ISOLATION AND PROPERTIES OF THE GREEN FLUORESCENT PROTEIN

Class Anthozoa Acanthoptilum The sea cactus Cavernularia The sea pansy Renilla The sea pen Ptilosarcus and Pennatula Stylatula Concerning the mechanism of energy transfer from the excited state of photoprotein molecule to GFP molecule, Morin and Hastings (1971b) suggested for the first time that the mechanism of coelenterate bioluminescence possibly involves the Förster-type radiationless energy transfer. To clarify the mechanism of energy transfer involved in the emission of green light from the jellyfish Aequorea, we isolated and purified the green fluorescent protein from the jellyfish, and then we studied its properties in detail (Morise et al., 1974). The purified Aequorea GFP was easily crystallized by decreasing the ionic strength of the solvent (Fig. 1.2). We investigated the energy transfer from aequorin molecule to GFP molecule during the Ca2+-triggered luminescence reaction of aequorin, under two sets of conditions: one with high concentrations of GFP (1.7–5.5 mg/ml) and the other with relatively low concentrations of GFP (0.15–1.1 mg/ml). In the presence of the high concentrations of GFP, apparently an energy transfer by the trivial (radiative) mechanism takes place, at least to some extent. Namely, the light emitted from aequorin (emission lmax 460 nm) is absorbed by GFP (lmax 400 and 480 nm),

Figure 1.2. The fluorescence photomicrograph of the crystals of Aequorea GFP formed in a low ionic strength aqueous solution, by dialysis against pure water. The fluorescence of GFP crystal is strongly anisotropic (Inoué et al., 2002). The view field shown is about 0.5 mm wide. Photograph by Dr. Shinya Inoué.

5

6

DISCOVERY OF GREEN FLUORESCENT PROTEIN

followed by reemission of the absorbed energy from GFP as fluorescence (emission lmax 509 nm). In this mechanism, the extent of energy transfer and the spectral shape of emitted light are dependent on the GFP concentration. It is clear, however, that GFP cannot absorb all the light emitted from aequorin, because the luminescence emission of aequorin extends to about 600 nm on the red side of wavelength whereas GFP can absorb light up to only about 510 nm. Therefore, a complete energy transfer by the trivial mechanism is clearly impossible. In any event, a very high concentration of GFP (with a very high level of absorbance) is required to obtain a significant extent of energy transfer by the trivial mechanism. Under such a condition, the self-absorption of GFP would strongly affect the spectral shape of the fluorescence emitted from GFP, in two ways: (1) a very steep decrease in the light intensities below 510 nm and (2) a red shift of the fluorescence peak position. The actual luminescence spectrum should be the sum of the aequorin luminescence unabsorbed by GFP and the GFP fluorescence distorted by self-absorption; it would be unthinkable that such a spectrum coincides with the true, undistorted spectrum of GFP fluorescence or with the luminescence from the live Aequorea. When aequorin was luminesced with Ca2+ in a low ionic strength buffer (10 mM sodium phosphate) in the presence of relatively low concentrations of GFP (about 0.15 mg/ml), the emission spectrum of aequorin was little affected by GFP. However, when a small amount of fine particles of diethylaminoethyl (DEAE) cellulose or DEAE Sephadex (anion exchangers) was mixed in advance to the same solution, the Ca2+triggered luminescence of the clouded mixture became spectrally identical with the in vivo bioluminescence of Aequorea, indicating the occurrence of an efficient energy transfer from the aequorin light emitter to GFP. It should be pointed out that the amounts of aequorin and GFP, as well as the volume used, were kept equal in the aforementioned experiments (i.e., the overall concentrations and the absorbance values were unchanged); the only difference was the DEAE material added in the latter experiment. The interpretation of the above-mentioned finding is as follows. Under the conditions used, the DEAE cellulose particles had co-adsorbed GFP and aequorin by anion exchange mechanism, greatly increasing the local concentrations of the two proteins around the particles. The co-adsorption perhaps made the distance between the GFP molecules and the aequorin molecules sufficiently short (roughly 30 Å) to make the Förster-type (radiationless) energy transfer workable. Thus, the result observed was the green light that spectrally matches with the in vivo luminescence and the fluorescence emission of GFP. Because the radiationless process is not significantly influenced by the concentration of GFP and does not require a very high concentration of GFP, the energy transfer can take place without being significantly affected by the absorbance of GFP. On the basis of the above experiments and discussion, the energy transfer involved in the emission of green light from live Aequorea is considered to be mostly, if not entirely, a radiationless process. The quantum yield of the Ca2+-triggered aequorin luminescence is approximately 0.16 at 23–24°C (Shimomura and Johnson, 1970; Shimomura, 1986), and that of aequorin coadsorbed with GFP is the same as that of aequorin alone (Morise et al., 1974). In a live specimen of Aequorea, each light organ (0.4 ¥ 0.2 ¥ 0.1 mm) is packed with photogenic cells (average size 10 mm), and each photogenic cell is again densely packed with fine particles (diameter 0.5 mm), according to Davenport and Nicol (1955). It is believed that these particles contain high concentrations of aequorin and GFP. In the particles, aequorin molecules and GFP molecules must be very closely and tightly arranged, if they are not directly bound to each other, to allow an efficient energy transfer by a radiationless process. In fact, the concentration of aequorin and GFP in the photogenic cells are previously estimated to be 5% each, or 10% altogether, of the weight of the cells

ISOLATION AND PROPERTIES OF THE GREEN FLUORESCENT PROTEIN

(Morise et al., 1974). In a more recent estimate, the concentration of GFP was estimated at 2.5% (Cutler, 1995). Another kind of green fluorescent protein, the GFP of the sea pansy Renilla, was purified and physicochemically characterized (Ward and Cormier, 1979). There are substantial differences between the bioluminescence systems of Aequorea and Renilla, though in both systems the light energy is provided by the oxidation of coelenterazine. The in vivo bioluminescence reaction of Renilla requires coelenterazine (the luciferin), Renilla luciferase, Renilla GFP, and molecular oxygen, whereas that of Aequorea requires only aequorin, Ca2+, and Aequorea GFP. The fluorescence emission peak of Renilla GFP (509 nm) is identical to that of Aequorea GFP, but its absorption spectrum (lmax 498 nm) is markedly different from that of Aequorea GFP (lmax 400 nm and 480 nm). Addition of coelenterazine to a solution containing Renilla luciferase results in the emission of blue light. However, when Renilla GFP has been added to the luciferase solution before the addition of coelenterazine, green luminescence is emitted with a threefold increase in the quantum yield, clearly indicating the occurrence of radiationless energy transfer (Ward and Cormier, 1979). Thus, in the case of Renilla, there must be a sufficiently strong binding affinity between the molecules of luciferase and GFP, to make the distance between the chromophores sufficiently short for the energy transfer by radiationless process. It appears that the affinity between Renilla luciferase and Renilla GFP is much greater than that between aequorin and Aequorea GFP. The fluorescence quantum yields of Aequorea GFP and Renilla GFP are nearly equal in a range of 0.7–0.8 (Morise et al., 1974; Kurian et al., 1994; Chapter 4, this volume). However, Renilla GFP significantly increases the quantum yield of bioluminescence, but Aequorea GFP does not, as noted earlier. The difference must come mainly from the difference in the fluorescence quantum yields of coelenterazine light-emitters in the two systems, on the basis of the following discussion. The quantum yield of bioluminescence, Qbl, can be expressed, in a practical way, as the product of the yield of the excited state generated, E, and the fluorescence quantum yield of the light emitter, Qf. Thus, Qbl = EQf, where the values of Qbl, E, and Qf cannot exceed 1. In the Ca2+-triggered light emission of aequorin, the photoprotein is decomposed into apoaequorin, coelenteramide, and carbon dioxide, wherein apoaequorin, coelenteramide, and calcium ions form a complex called the blue fluorescent protein “BFP” (Shimomura and Johnson, 1970). The aequorin luminescence is emitted from BFP or, more precisely, from the amide anion of coelenteramide in excited state (i.e., coelenterazine light emitter) bound to apoaequorin in BFP (Hori et al., 1973; Shimomura, 1995b). The quantum yield Qbl of aequorin luminescence is 0.16 as already noted. The fluorescence quantum yield Qf of BFP measured a few seconds after the light emission was 0.12, but this value is considered to be inaccurate on the basis that BFP is a dissociable equilibrium complex and also that the conformation of apoaequorin changes after the light-emitting reaction (Morise et al., 1974). At present, there seems to be no way to measure an accurate value of Qf at the moment of light emission in the case of aequorin luminescence. However, because the Qbl of aequorin luminescence and the Qbl of the luminescence of aequorin coadsorbed with GFP are equal (0.16), the Qf of BFP at the moment of light emission should be equal to that of GFP (0.7–0.8) when the energy transfer from BFP to GFP is 100%. In the luciferase-catalyzed Renilla bioluminescence, quantum yield Qbl is increased three times by the addition of Renilla GFP, and the fluorescence quantum yield Qf of Renilla GFP is 0.7–0.8 as already noted. The bioluminescence quantum yield Qbl in the absence of GFP was reported at 0.055 (Matthews et al., 1977) and 0.1 (Inouye and

7

8

DISCOVERY OF GREEN FLUORESCENT PROTEIN

Shimomura, 1997). Assuming the energy transfer at 100%, the fluorescence quantum yield Qf of the coelenterazine light-emitter of the Renilla system is calculated to be approximately 0.15–0.27, which is significantly lower than the Qf value for the coelenterazine light emitter in BFP (0.7–0.8). Regarding the nature of the chromophore, it is believed that the GFPs of Aequorea, Renilla, and many other coelenterates contain an identical chromophore (Ward and Cormier, 1978; Ward et al., 1980); the only exception presently known is the GFP of the jellyfish Phialidium that shows a blue-shifted fluorescence emission peak at 497 nm (Levine and Ward, 1982).

1.3

DISCOVERY OF THE STRUCTURE OF GFP CHROMOPHORE

In 1979, I was interested in the chemical structure of the chromophore of Aequorea GFP, which had never been studied before. From a papain digest of heat-denatured GFP, I isolated a small peptide containing the chromophore. I synthesized a model compound of the chromophore. Based on the resemblance between this model compound and the chromophore of the peptide, I was able to deduce the structure of the GFP chromophore to be the structure D in Fig. 1.3 (Shimomura, 1979). It might look as though I were very lucky in my guesswork, because the data obtained from the peptide were clearly insufficient to elucidate the structure of the chromophore. In fact, several people questioned me as to how I could guess the imidazolone structure. The truth is that I was certainly lucky, but not only in my guesswork. CH3 CH3

Rb Ra

Figure 1.3. (A) A tentative structure of Cypridina luciferin proposed in 1959. (B) One of the model compounds synthesized to test the feasibility of the structure A. (C) A model compound of GFP chromophore synthesized. (D) The chromophore of GFP proposed in 1979. Both Ra and Rb are peptide residues.

A NOTE ON THE SPECIES NAME OF THE JELLYFISH FROM WHICH AEQUORIN AND GFP WERE ISOLATED

In the period of the late 1950s, I was studying the structure of the luciferin of the ostracod Cypridina at the laboratory of Professor Y. Hirata, Nagoya University. The techniques for structure determination available at the time were not as sophisticated as at present. The modern techniques that would produce clear-cut information, such as nuclear magnetic resonance (NMR), high-resolution mass spectroscopy, and high-performance liquid chromatography (HPLC), were not available. In an early stage of our study on Cypridina luciferin, we arrived at a tentative structure that contained an imidazolone ring, A (Hirata et al., 1959). To test the absorption spectrum of this tentative structure, we synthesized various imidazolone compounds that contained one double bond conjugated with the imidazolone ring (Shimomura and Eguchi, 1960), although the results eventually showed that structure A was incorrect. Compound B was one of the imidazolones synthesized at that time. When I obtained the chromophore-bearing peptide from Aequorea GFP in 1979, I immediately noticed a close resemblance in spectroscopic and other properties between the chromophore-bearing peptide obtained from GFP and the imidazolone compound B, which was synthesized some 20 years before. A small difference found in the wavelength of the absorption peak was thought to be the effect of a phenolic OH, based on the evidence that acid hydrolysis of the peptide yielded p-hydroxybenzaldehyde. I synthesized a new model compound C. The spectroscopic properties of compound C were in satisfactory agreement with those of the peptide. Thus, structure D was proposed as the chromophore of GFP (Shimomura, 1979). The chromophore structure was confirmed later to be correct, although the side chains were different (Cody et al., 1993). I learned in 1979 that W. W. Ward of Rutgers University, the pioneer of the isolation of the photosensitive ctenophore photoproteins (Ward and Seliger, 1974a,b), had been working on Aequorea GFP in addition to Renilla GFP. I thought my role was over and decided to discontinue my work on GFP. Since then, the work on Aequorea GFP by Ward and others has steadily progressed, finally developing into the successful cloning of GFP (Prasher et al., 1992), a memorable event that established the basis of using GFP. The cloning was soon followed by the expression of GFP in living organisms by Chalfie et al. (1994) that triggered the explosive popularity of GFP and made the foundation of the present volume.

1.4 A NOTE ON THE SPECIES NAME OF THE JELLYFISH FROM WHICH AEQUORIN AND GFP WERE ISOLATED This brief discussion concerning the names of Aequorea species is included here in consideration of the problems and confusions induced by the recent common use of the species name Aequorea victoria in place of Aequorea aequorea (and Aequorea forskalea). The species names A. aequorea (Forskal, 1775) and A. forskalea (Peron and Lesueur, 1809) are synonymous, and both names have been commonly used; the decision of priority between them appears to be a matter of opinion. The species A. aequorea is highly variable in both form and color (Mayer, 1910) and is distributed very widely—Mediterranean; Atlantic coasts, from Norway to South Africa and Cape Cod to Florida; northeastern Pacific; east coast of Australia; and Iranian Gulf (Kramp, 1968). According to Mayer (1910), A. victoria (Murbach and Shearer, 1902) from the northeastern Pacific is probably a variety of A. aequorea. He stated “I cannot distinguish this medusa from Aequorea forskalea of the Atlantic and Mediterranean. Were it described from the Atlantic I would not hesitate to designate it A. forskalea.” Mayer’s opinion has been overwhelmingly accepted until recently (Russell, 1953; Kramp, 1965, 1968); thus the jellyfish we collected

9

10

DISCOVERY OF GREEN FLUORESCENT PROTEIN

in the Friday Harbor area have been called A. aequorea. The situation changed, however, after Arai and Brinckmann-Voss (1980) reported their conclusion to separate A. victoria from the species A. aequorea, based on their study of about 40 specimens collected from more than 10 different areas around Vancouver Island (Friday Harbor included). Their reasons were that A. victoria has much more regularly serrated mouth lobes and a much larger (thus, taller), almost hemispherical lens in the stomach region, when compared with A. aequorea from the Mediterranean. It is not clear in the Arai and Brinckmann-Voss article why the conclusion to separate A. victoria from A. aequorea was made on the basis of the comparison between the former (from British Columbia and Puget Sound) and the latter from only the Mediterranean; their use of only a few specimens per study area, collected probably on a single occasion, brings about another problem. It has been well documented that a wide intraspecific variation of A. aequorea by geography exists (Mayer, 1910; Russell, 1953; Kramp, 1959, 1965). The Mediterranean form of A. aequorea is only one of many variations of this species. Therefore, the difference between A. victoria and A. aequorea cannot be fully determined by the comparison between the former and the latter from only the Mediterranean; to determine the difference, A. victoria should be compared with various other forms of A. aequorea. It seems particularly intriguing to compare A. victoria with the varieties of A. aequorea from the northern and western Atlantic. In our record, we collected a large number of A. aequorea at Woods Hole, MA, in the summer of 1987, when there was a strong easterly wind; those medusae appeared to be indistinguishable from the average specimens of Aequorea obtained at Friday Harbor in both form and the composition of aequorin isoforms. Most of the several million specimens of Aequorea used for biochemical research had been collected around Friday Harbor, where the specimens were extremely abundant at least until 1988 (since then, they virtually disappeared from the area for unknown reason). If all those medusae were a single species of A. victoria, as implied by Arai and Brinckmann-Voss (1980), it seems that A. victoria must have a very wide variation, like A. aequorea. We have collected over 1 million specimens of Aequorea in the vicinity of Friday Harbor in 17 summers between 1961 and 1988. All the specimens were mature (>7 cm in diameter), and they were collected, handled, and excised individually. More than several times during our operation, we observed pronounced changes in the form of the jellyfish collected. The jellyfish can drift far and widely by current, tide, and wind in groups, and the changes that we observed usually lasted for only a few days but occasionally continued for several weeks. The bell height of the medusae were sometimes markedly higher than usual relative to the diameter (thus taller than hemispherical), and sometimes much flatter and saucer-like. In one of these occasions, we thought that the jellyfish we had collected were a wrong species because they were too flat; we suspended our operation until we had an assurance by a jellyfish expert in the lab that they were indeed a variety of A. aequorea (as known then). If all those medusae at Friday Harbor are the variations of A. victoria, the situation would be very confusing. Because both species have very wide variations without any clear difference between them, distinction of the two species would be extremely difficult. A detailed discussion on the species name of Aequorea is given by Claudia E. Mills (2003). Despite the high intraspecific variability that causes confusions, the species names A. aequorea and A. forskalea have been accepted and used by the majority of researchers for the period of at least 60 years until 1980. To avoid further confusion, and with reference to discussion in other chapters of this volume, the name A. victoria should be considered as a synonym, not as the name of a separate species, until the difference between A.

REFERENCES

victoria and A. aequorea (or A. forskalea) is firmly established on genetic basis. Until that time, the species name A. aequorea (or A. forskalea) should have priority.

ACKNOWLEDGMENTS Our work on the jellyfish Aequorea was initiated by the late Professor Frank H. Johnson, whose contribution to the project was enormous and immeasurable. I sincerely thank all the people who contributed directly or indirectly to this project. The work was made possible by the excellent facilities of the Friday Harbor Laboratories, University of Washington, and research grants from the National Science Foundation and National Institutes of Health.

REFERENCES Arai, M. N., and Brinckman-Voss, A. (1980). Hydromedusae of British Columbia and Puget Sound. Can. Bull. Fish Aquat. Sci. Bull. 204:1–181. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green-fluorescent protein as a marker for gene expression. Science 263:802–805. Cody, C. W., Prasher, D. C., Westler, W. M., Prendergast, F. G., and Ward, W. W. (1993). Chemical structure of the hexapeptide chromophore of the Aequorea green-fluorescent protein. Biochemistry 32:1212–1218. Cormier, M. J., Hori, K., Karkhanis, Y. D., Anderson, J. M., Wampler, J. E., Morin, J. G., and Hastings, J. W. (1973). Evidence for similar biochemical requirements for bioluminescence among the coelenterates. J. Cell. Physiol. 81:291–298. Cormier, M. J., Hori, K., and Anderson, J. M. (1974). Bioluminescence in coelenterates. Biochim. Biophys. Acta 346:137–164. Cutler, M. W. (1995). Characterization and energy transfer mechanism of green fluorescent protein from Aequorea victoria. Ph.D. dissertation, Rutgers University, New Brunswick, NJ. Davenport, D., and Nicol, J. A. C. (1955). Luminescence of hydromedusae. Proc. R. Soc. London Ser. B 144:399–411. Forskal, P. (1775). Descriptiones animalium avium, amphibiorum, piscium, insectorum, vermium: Quae in itinere orientali observavit Petrus Forskal. Post mortem auctoris edidit Carsten Niebuhr. 164 pages. Ex Officina Moller Hauniae (Copenhagen). Hastings, J. W., and Morin, J. G. (1969). Comparative biochemistry of calcium-activated photoproteins from the ctenophore, Mnemiopsis and the coelenterates. Aequorea, Obelia, Pelagia and Renilla. Biol. Bull. 137:402. Hirata, Y., Shimomura, O., and Eguchi, S. (1959). The structure of Cypridina luciferin. Tetrahedron Lett. 5:4–9. Hori, K., Wampler, J. E., and Cormier, M. J. (1973). Chemiluminescence of Renilla (sea pansy) luciferin and its analogues. Chem. Commun. 492–493. Inouye, S., and Shimomura, O. (1997). The use of Renilla luciferase, Oplophorus luciferase, and apoaequorin as bioluminescent reporter protein in the presence of coelenterazine analogues as substrate. Biochem. Biophys. Res. Commun. 233:349–353. Inoué, S., Shimomura, O., Goda, M., Shribak, M., and Tran, P. T. (2002). Fluorescence polarization of green fluorescence protein. Proc. Natl. Acad. Sci. USA 99:4272–4277. Johnson, F. H., Shimomura, O., Saiga, Y., Gershman, L. C., Reynolds, G. T., and Waters, J. R. (1962). Quantum efficiency of Cypridina luminescence, with a note on that of Aequorea. J. Cell. Comp. Physiol. 60:85–104.

11

12

DISCOVERY OF GREEN FLUORESCENT PROTEIN

Kramp, P. L. (1959). The hydromedusae of the Atlantic ocean and adjacent waters. Dana Report No. 46. Carlsberg Foundation, Copenhagen, Denmark. Kramp, P. L. (1965). The hydromedusae of the Pacific and Indian Oceans. Dana Report No. 63. Carlsberg Foundation, Copenhagen, Denmark. Kramp, P. L. (1968). The hydromedusae of the Pacific and Indian oceans, sections II and III. Dana Report No. 72. Carlsberg Foundation, Copenhagen, Denmark. Levine, L. D., and Ward, W. W. (1982). Isolation and characterization of a photoprotein, “phialidin,” and a spectrally unique green-fluorescent protein from the bioluminescent jellyfish Phialidium gregarium. Comp. Biochem. Physiol. 72B:77–85. Matthews, J. C., Hori, K., and Cormier, M. J. (1977). Purification and properties of Renilla reniformis luciferase. Biochemistry 16:85–91. Mayer, A. G. (1910). Medusae of the world, Vol. II (Hydromedusae). Carnegie Institute Washington Publication, Washington, D.C., pp. 231–498. Mills, C. E. (2003). http://faculty.washington.edu/cemills/Aequorea.html. Morin, J. G. (1974). Coelenterate bioluminescence. In Coelenterate Biology. Reviews and Perspectives, Muscatine, L., and Lenhoff, H. M., Eds., Academic, New York, pp. 397–438. Morin, J. G., and Hastings, J. W. (1971a). Biochemistry of the bioluminescence of colonial hydroids and other coelenterates. J. Cell. Physiol. 77:305–311. Morin, J. G., and Hastings, J. W. (1971b). Energy transfer in a bioluminescent system. J. Cell. Physiol. 77:313–318. Morise, H., Shimomura, O., Johnson, F. H., and Winant, J. (1974). Intermolecular energy transfer in the bioluminescent system of Aequorea. Biochemistry 13:2656–2662. Murbach, L., and Shearer, C. (1902). Preliminary report on a collection of medusae from the coast of British Columbia and Alaska. Ann. Mag. Nat. Hist. Ser. 7 9:71–73. Peron, F., and Lesueur, C. A. (1809). Tableau des caracteres generiques et specifiques de toutes les especes de Meduses connues jusqu’a ce jour. Ann. Mus. Hist. Nat. Paris 14:325–366. Prasher, D. C., Eckenrode, V. K., Ward, W. W., Prendergast, F. G., and Cormier, M. J. (1992). Primary structure of the Aequorea victoria green fluorescent protein. Gene 111:229–233. Ridgway, E. B., and Ashley, C. C. (1967). Calcium transients in single muscle fibers. Biochem. Biophys. Res. Commun. 29:229–234. Russell, F. S. (1953). The Medusae of the British Isles, Vol. I: Anthomedusae, Leptomedusae, Limnomedusae, Trachymedusae and Narcomedusae, Cambridge University Press, London, 530 pages. Shimomura, O. (1979). Structure of the chromophore of Aequorea green fluorescent protein. FEBS Lett. 104:220–222. Shimomura, O. (1984). Bioluminescence in the sea: photoprotein systems. Symp. Soc. Exp. Biol. 39:351–372. Shimomura, O. (1995a). A short story of aequorin. Biol. Bull. 189:1–5. Shimomura, O. (1995b). Cause of spectral variation in the luminescence of semisynthetic aequorins. Biochem. J. 306:537–543. Shimomura, O., and Eguchi, S. (1960). Studies on 5-imidazolone. I–II. Nippon Kagaku Zasshi 81:1434–1439. Shimomura, O., and Johnson, F. H. (1966). Partial purification and properties of the Chaetopterus luminescence system. In Bioluminescence in Progress, Johnson, F. H., and Haneda, Y., Eds., Princeton University Press, Princeton, NJ, pp. 495–521. Shimomura, O., and Johnson, F. H. (1969). Properties of the bioluminescent protein aequorin. Biochemistry 8:3991–3997. Shimomura, O., and Johnson, F. H. (1970). Calcium binding, quantum yield, and emitting molecule in aequorin bioluminescence. Nature 227:1356–1357.

REFERENCES

Shimomura, O., and Johnson, F. H. (1972). Structure of the light-emitting moiety of aequorin. Biochemistry 11:1602–1608. Shimomura, O., and Johnson, F. H. (1973). Chemical nature of the light emitter in bioluminescence of aequorin. Tetrahedron Lett. 31:2963–2966. Shimomura, O., and Johnson, F. H. (1975). Regeneration of the photoprotein aequorin. Nature 256:236–238. Shimomura, O., and Johnson, F. H. (1979). Comparison of the amounts of key components in the bioluminescence systems of coelenterates. Comp. Biochem. Physiol. 64B:105–107. Shimomura, O., Johnson, F. H., and Saiga, Y. (1962). Extraction, purification and properties of aequorin, a bioluminescent protein from the luminous hydromedusan, Aequorea. J. Cell. Comp. Physiol. 59:223–239. Wampler, J. E., Hori, K., Lee, J., and Cormier, M. J. (1971). Structured bioluminescence. Two emitters during both the in vitro and the in vivo bioluminescence of the sea pansy, Renilla. Biochemistry 10:2903–2909. Wampler, J. E., Karkhanis, Y. D., Morin, J. G., and Cormier, M. J. (1973). Similarities in the bioluminescence from the Pennatulacea. Biochim. Biophys. Acta 314:104–109. Ward, W. W., and Cormier, M. J. (1978). Energy transfer via protein–protein interaction in Renilla bioluminescence. Photochem. Photobiol. 27:389–396. Ward, W. W., and Cormier, M. J. (1979). An energy transfer protein in coelenterate bioluminescence. Characterization of the Renilla green fluorescent protein. J. Biol. Chem. 254:781–788. Ward, W. W., and Seliger, H. H. (1974a). Extraction and purification of calcium-activated photoproteins from ctenophores. Biochemistry 13:1491–1499. Ward, W. W., and Seliger, H. H. (1974b). Properties of mnemiopsin and berovin, calcium-activated photoproteins. Biochemistry 13:1500–1510. Ward, W. W., Cody, C. W., Hart, R. C., and Cormier, M. J. (1980). Spectrophotometric identity of the energy transfer chromophores in Renilla and Aequorea green fluorescent proteins. Photochem. Photobiol. 31:611–615.

13

2 PHOTONS FOR REPORTING MOLECULAR EVENTS: GREEN FLUORESCENT PROTEIN AND FOUR LUCIFERASE SYSTEMS J. Woodland Hastings Department of Molecular and Cellular Biology, Harvard University, Cambridge, MA

James G. Morin Section of Ecological Systematics, Cornell University, Ithaca, NY

2.1 INTRODUCTION During the course of evolution, bioluminescence has repeatedly appeared where it serves biological functions important to the organism. Functions may differ among organisms and a given organism may utilize luminescence in more than one way (Morin, 1983; Hastings, 1983; Hastings and Morin, 1991). The different specific recognized functions may be classed under three major rubrics: defensive (to help deter predators), offensive (to aid in predation), and communication (e.g., for courtship or mating). Within each category a number of different specific strategies are recognized; for example, luminescence may be used defensively as a decoy to divert, as a flash to frighten, or as ventral luminescence to camouflage the silhouette. In terms of the total number of different species, the emission of bioluminescence is rather rare, but it occurs in many phylogenetically different groups (Table 2.1; Harvey, 1952; Herring, 1978). In those groups that do emit light, the biochemical and physiological mechanisms responsible for it are often very different, as are its several functional roles. Indeed, luciferase (the enzyme) and luciferin (the substrate) are generic terms, and quite different in the different groups. Thus, the organism from which they are obtained must be specified.

Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

15

Photobacterium Vibrio Xenorhabdus Panus, Armillaria Pleurotus Gonyaulax Pyrocystis Noctiluca

Bacteria

Vargula; Cypridina

Crustacea Ostracod

Meganyctiphanes Gaussia

Aldehyde Clam luciferin, structure?, Cu2+ Bacterial symbionts (490)

Latia Pholas Heteroteuthis

Shrimp Copepods and others

N-isovalyeryl-3 amino propanal Unknown (465) Unknown (510) Unknown (530)

Diplocardia Chaetopterus Odontosyllis Acholoë

Linear tetrapyrrole (470) Unknown

Imadazopyrazine nucleus (465)

Ca2+, coelenterazine/aequorin Imidazo pyrazine nucleus (460–510), GFP as accessory emitter Ca2+, coelenterazine (460)

Aequorea Obelia Renilla Mnemiopsis; Beroë

Reduced flavin and long chain aldehyde (475–540) YFP and LUMP as accessory emitters Unknown (535) Linear tetrapyrrole Cell organelles (scintillons) (470)

Luciferins and Other Factors (Emission Max. (nm))

Cnidaria Jellyfish Hydroid Sea Pansy Ctenophores Annelids Earthworms Chaetopterid worm Syllid fireworm Scale worm Molluscs Limpet Clam Squid

Dinoflagellates

Mushrooms

Representative Genera

Type of Organism

TABLE 2.1. Representatives of the Major Bioluminescent Organisms

Squirts enzyme and substrate Diversion, decoy, courtship Photophores; camouflage Deter predators

Exuded luminescence in all three. Photophores and symbiotic bacteria in some squid. Functions: diversion, decoy, camouflage, probably others

Cellular exudates or intracellular flashes sometimes very bright. To divert or deter; courtship

Bright flashes; frighten or deter

Bright flash or train of flashes To frighten or deter

pH change causes short (0.1 s) bright flashes To frighten or deter

Steady bright glow after autoinduction of luciferase Symbiosis Steady dim glow; to attract insects

Displays Features and Functions

16 PHOTONS FOR REPORTING MOLECULAR EVENTS

Symbiotic luminous bacteria (~490) Self-luminous, Vargula-type luciferin, Nutritionally obtained (485) Self-luminous, biochemistry unknown Self-luminous, biochemistry unknown Self-luminous, biochemistry unknown

Cyclothone

Neoscopelus Tarletonbeania

Midwater fishes

Symbiotic luminous bacteria (~490) Symbiotic luminous bacteria (~490)

Leiognathus Photoblepharon

Cryptopsaras Porichthys

Unknown

Cell organelles evolved from bacteria (480–500)

Pyrosoma

Isistius

Trains of rapid flashes; frighten, divert

Biochemistry unknown, Ca2+

Ophiopsila

Camouflage, courtship, deterrence, capture prey Many photophores, ventral and lateral Photophores: lateral, on tongue Sexual dimorphism; males have dorsal (police car) photophores

Ventral luminescence; camouflage, courtship

Ventral luminescence and flashes; camouflage, attract and capture prey, courtship, deter predators, communication

Ventral glow; camouflage

Brilliant trains of flashes; function unknown Stimulated by light and other factors

Lure to attract prey

Flashes, specific kinetic patterns Deter predators; courtship, mating

Benzothiazole, ATP, Mg2+ Similar chemistry in all coleoptera (most 550–580) ATP in Arachnocampa (460–480)

Photinus, Photuris Pyrophorus Phengodes, Phrixothrix Arachnocampa, Orfelia

Angler fish Midshipman

Fishes Cartilaginous fishes Bony fishes Ponyfish Flashlight fish

Insects Coleopterids (beetles) firefly click beetles railroad worm Diptera (flies) Echinoderms Brittle stars Chordates Tunicates

INTRODUCTION

17

18

PHOTONS FOR REPORTING MOLECULAR EVENTS

Bioluminescence is thus not an evolutionarily conserved function; in the different groups of organisms the genes and proteins involved are mostly unrelated, and evidently originated and evolved independently. How many times this may have occurred is difficult to say, but it has been estimated that present day luminous organisms come from as many as 30 different evolutionarily distinct origins (Hastings, 1983; Hastings and Morin, 1991). Cnidarian luminescence, with green fluorescent protein (GFP) present as an accessory emitter in some but not all species, is thus only one of a wide array of luminescent systems and biochemistries. The genes and proteins of several of the other systems (notably bacterial and firefly) have been used for many different analytical and reporter purposes (DeLuca, 1978; Hastings et al., 1997). Their diversity allows for many different possible applications in which photons are the reporter, but none of the other proteins or systems possesses the unique features of GFP. The biochemistries of only four of the different luminous systems are known in detail (Table 2.1), namely, bacteria, dinoflagellates, cnidarians, and fireflies (Hastings and Morin, 1991; Wilsman and Hastings, 1998). Although some information is known for another half-dozen or so, we confine our review to the four best described three dimensional crystal structures of all four have been reported. While these systems differ in the structures of the luciferins (Fig. 2.1) and luciferases, all systems have some features in common at the chemical level. All known luciferases are oxygenases that utilize molecular oxygen to

Figure 2.1. Structures of four different luciferins, oxygen-containing intermediates, and postulated emitters (see Table 2.1).

CNIDARIANS, CTENOPHORES, AND GFP

oxidize the associated luciferin, giving an intermediate enzyme-bound peroxide, whose breakdown then results in the production of an intermediate or product directly in its excited singlet state. In most systems, emission occurs from the luciferase-bound substrate-derived excited molecule (see Fig. 2.1), but an accessory secondary emitter occurs in certain cnidarians and some bacteria. With GFP in cnidarians, the mechanism responsible was postulated and later confirmed to involved Förster-type energy transfer (Morin and Hastings, 1971b; Morise et al., 1974; Ward et al., 1980). GFP is unusual in that its chromophore is a part of the (modified) primary structure of the protein, thus not subject to dissociation (Cody et al., 1993; Heim et al., 1994). Its use as a transgene reporter, pioneered by Chalfie et al. (1994), relies on this feature and its fluorescence alone.

2.2 CNIDARIANS, CTENOPHORES, AND GFP Luminescence is common and widely distributed in these groups (Morin, 1974; Cormier, 1981; Herring, 1978). In the ctenophores (comb jellies), they comprise over one-half of all genera, whereas in the cnidarians it is about 6%. These organisms are mostly sessile, sedentary, or planktonic, and upon stimulation they emit light as flashes. Hydroids such as Obelia occur as plant-like growths, typically adhering to rocks and kelp below low-tide level in many of the world’s oceans. Upon stimulation, a conducted scintillating emission emanates as a wave along the colony from individual photocytes (cells specialized for light emission); repetitive waves may occur from a single stimulus. Aequorea, a hydromedusan that is very abundant in the San Juan Islands region of the northwest United States, has been extensively used for biochemical studies (Shimomura and Johnson, 1975; Cormier et al., 1989). The biochemistry of the sea pansy, Renilla, which occurs near shore on sandy bottoms, has also been elucidated (Cormier, 1981). Early observations, attributable to what we now know as GFP, were reported by several investigators, including an emission spectrum of the bioluminescence of the sea pen Pennatula phosphorea showing a narrow bandwidth emission in the green (Nicol, 1958), which is now known to be characteristic of GFP. Shimomura et al. (1962) later noted that the bioluminescence of Aequorea extracts was blue while that of the intact organism was green, attributing this to a protein that fluoresced green in extracts (Johnson et al., 1962). More suggestive of a relationship between luminescence and fluorescence was the observation of green fluorescence in cells located in the vicinity of bioluminescence activity by Titschack (1964) in the pennatulacean Veretillum cynomorium. In none of these studies, however, was the relationship of the green fluorescence to the bioluminescence clearly established. We discovered GFP quite independently while we were examining both the mechanisms controlling luminescent flashes and the biochemical underpinnings of the luminescence in the colonial hydroid Obelia geniculata. Our biochemical studies quickly expanded to studying calcium activated photoproteins in a variety of coelenterates, including several species of hydrozoans (Obelia, Aequorea, Clytia), pennatulaceans (the sea pens Renilla and Ptilosarcus), the scyphozoan jellyfish Pelagia, and the ctenophore Mnemiopsis (Morin et al., 1968; Hastings and Morin 1969a,b). Measurements of emission spectra of reactions in extracts gave wide bandwidth curves peaking in the blue, but somewhat different in different species; the blue emissions of the luciferase systems of Aequorea, Obelia, and Renilla exhibit maxima at 460, 472, and 486 nm, respectively (Fig. 2.2). In vivo luminescent spectra from all three, however, were narrow, peaking in the green at

19

20

PHOTONS FOR REPORTING MOLECULAR EVENTS

Figure 2.2. Emission spectra for bioluminescence in vivo for Obelia, Aequorea, and Renilla compared with spectra from emission of in vitro reactions isolated from these same organisms. [From Morin and Hastings (1971b).]

Figure 2.3. Luminescent flash (L) and luminescent potential (E) bursts recorded concurrently from an O. geniculata photocyte and hydranth, respectively. Responses are to a train of stimuli applied about once every 2 s, indicated by solid triangles beneath the lower trace (horizontal bar = 1 s); lower (E) ordinate, vertical bar = 1 mV; upper (L) ordinate, light intensity in relative units. [From Morin and Cooke (1971a).]

about 508 nm, matching fluorescence emission spectrum of GFP (Morin and Hastings, 1971a). So while the underlying biochemistry may differ somewhat, emissions are all in the green peaking at about 508 nm so long as the luciferases interact appropriately with GFP in order to transfer excitation energy. GFP is thus an accessory emitter protein of the cnidarian luminescent system, deriving its excitation by nonradiative energy transfer in association with the luciferase reaction, which in the absence of GFP emits blue light (Morin and Hastings, 1971b). Many of the properties of GFP, such as its thermal stability and remarkable resistance to proteolysis, derive from its barrel structure, dubbed a b-can, with 11 b strands arranged protectively around a central chromophore (Ormö et al., 1996; Yange et al., 1996). Solvent access to the inside cavity in these “lanterns” is blocked on top and bottom by short segments of a-helices, although some water molecules are immobilized inside; there would clearly be no room for an enzyme to catalyze chromophore formation. The biochemical studies provided critical information for our physiological and morphological studies on the colonial hydroid O. geniculata, where action potentials propagate through electrically excitable epithelial cells. These action potentials spread incrementally through a colony and repetitively elicit flashes from individual photocytes (Fig. 2.3), which are located along the length of the stems and pedicels but not in the polyps themselves (see color Fig. 2.4; Morin et al., 1968; Morin and Reynolds, 1969;

21

CNIDARIANS, CTENOPHORES, AND GFP

(a)

(b)

Figure 2.4. Fluorescence micrograph of a living colony of Obelia species showing photocytes visualized by GFP. Height of field shown is about 3 mm in (a) and 1.5 mm in (b). (a) Dispersed photocytes (bright green spots) in an upright of O. geniculata (two polyps also shown at lower right and upper left). (b) Concentrated photocytes at the tip of a pedicel below the base of a hydranth of O. bidentata (= bicuspidata).

Morin and Cooke, 1971a,b). Indeed, it was the green fluorescence of the protein that allowed us, in conjunction with image intensification, to identify the photocytes by the colocalization of the fluorescence and bioluminescence of the photocytes and to establish irrefutably the association of GFP with the luminescent system (Fig. 2.5; Morin and Reynolds, 1969, 1970, 1974). In different species, GFP is confined to discrete photocytes (~10–20 mm in diameter), which are either dispersed [Fig. 2.4a] or clumped [Fig. 2.4B] in specific locations within the gastrodermis of the colonies (Morin and Reynolds, 1969, 1970, 1974; Morin, 1974). In measuring in vivo bioluminescence from single Obelia photocytes, GFP allowed us to identify their location prior to stimulation, so as to record photometrically via a fine-tipped (0.5 mm) light guide (Fig. 2.3). Trains of action potentials, termed luminescent potentials, initiated by single stimuli, propagate via electrically excitable epithelial cells (rather than neurons). These action potentials spread incrementally through a colony and repetitively excite the photocytes, and they also couple to other neuroid conducting systems such as those governing polyp contraction. Based on the number of photocytes, as determined with the aid of GFP, we were able to calculate from photometric measurements that each cell could emit about 1–2 ¥ 108 quanta/cell (Shorey and Morin, 1974). Studies of the GFP in the hydroid Obelia were also instrumental in providing the first demonstration that gap junctions can pass chemical signals in excitable tissues (Dunlap

22

PHOTONS FOR REPORTING MOLECULAR EVENTS

Figure 2.5. Bioluminescence (A), fluorescence (B), and rear illumination (C) pictures of an Obelia geniculata upright. Note that the luminescent and fluorescent spots (six in each) directly superimpose. The scale bar indicates 200 mm. [From Morin and Reynolds (1974).]

et al., 1987; Brehm et al., 1989). They showed that calcium actually enters neighboring nonluminescent, but electrically excitable, epithelial cells via voltage-dependent calcium channels and then into the photocytes, which are nonexcitable, via secondary calcium diffusion through gap junctions. This calcium then triggers the luminescence from the calcium activated photoprotein (obelin) with subsequent energy transfer to and emission from the GFP. Finally, by using GFP fluorescence as a reporter for the spatial distribution of luminescent cells in pennatulaceans (sea pens; see color Fig. 2.6) and photometry to measure the temporal aspects of the light emission, we have been able to infer that luminescence in sea pens and probably all cnidarians functions as an aposematic signal to deter damage to the colonies by potential aggressors or predators such as fishes and crustaceans (Morin, 1976, 1983). This inference has been experimentally verified for both pennatulaceans and brittle stars by Grober (1988a,b). At the biochemical level, the luciferin (coelenterazine) is the same in different cnidarian luminescent systems. Coelenterazine possesses an imidazopyrazine skeleton (Fig. 2.1) and is notable for its widespread phylogenetic distribution (Thomson et al., 1997), but whether the reason is nutritional or genetic (hence, possible evolutionary relatedness) has not yet been elucidated. But there are differences between the cnidarian anthozoan and hydrozoan systems with regard to the site of calcium action. In the anthozoan, coelenterazine is sequestered by a Ca2+-sensitive binding protein, and Ca2+ causes its release, thus triggering the in vivo flash. The Renilla luciferase reaction (EC 1.13.12.5) does not itself require calcium (Lorenz et al., 1991). In the hydrozoan Aequorea, calcium reacts instead at the luciferase stage, namely with aequorin, a luciferase-bound hydroperoxy coelenterazine intermediate, poised for the completion of the reaction. Aequorin was isolated by Shimomura et al. (1962) from the jellyfish Aequorea [in the presence of ethylenediaminetetraacetate (EDTA) to chelate calcium] and shown to emit light simply upon the addition of Ca2+, which is presumably the trigger in vivo (Hastings and Morin 1971; Blinks et al., 1982; Cormier et al., 1989). It was postulated (Hastings and Gibson, 1963) that in vivo luciferin coelenterazine reacts with oxygen, catalyzed by its luciferase (EC1.13.12.5), to form the hydroperoxide in a calcium-free compartment (the photocyte), where it is stored. Excitation allows Ca2+ to enter and bind to the protein (which possesses homology with calmodulin; Lorenz et al., 1991), changing its conformation so that the reaction continues, but without the need for free oxygen at this stage. It had been reported in the early literature (Harvey, 1952) that coe-

23

CNIDARIANS, CTENOPHORES, AND GFP

(a)

T T

T

M (b)

Figure 2.6. Fluorescence micrographs of photocytes visualized by GFP in living pennatulacean (sea pen) colonies. Width of field shown is about 0.8 mm in (a) and 1.6 mm in (b–d). (a) Photocytes in a cluster of five siphonozooids (water pumping polyps) of Renilla kollikeri. (b) Photocytes clustered in the lateral-axial region of the tentacles and oral disk of an autozooid (feeding polyp) of Renilla kollikeri (mouth [M] and base of three of eight tentacles [T] shown).

lenterates could inexplicably emit bioluminescence without oxygen. The explanation is now evident. Crystal structure determinations of aequorin (Head et al., 2000) and obelin (Liu et al., 2000) have confirmed them to be luciferase-bound peroxy coelenterazine intermediates; the two are very similar and represent an entirely new luciferase fold. As mentioned above, both possess homology with calmodulin and other calcium binding E-F hand proteins. The

24

PHOTONS FOR REPORTING MOLECULAR EVENTS

(c)

T T

T M

T T

T

T

(d)

Figure 2.6. (continued) (c) Photocytes clustered in only the two outer (of the eight) chambers within the calyx of the column (and not the tentacles) of an autozooid of Acanthoptilum gracile. (d) Photocytes clustered laterally along the length of each of the eight tentacles (T) of an autozooid of Ptilosarcus guerneyi (M = mouth).

monomer is considered to be the active species and is compact, having a 25Å radius. Each monomer is predominantly helical, being composed of 4 E-F hand motifs, of which three are able to bind calcium. The structure has been depicted as two cups joined “rim to rim” in which the bottom cup is composed of the first four helices and the top cup by the last four helices.

BACTERIA

Aequorin luminescence has been widely used for the detection and measurement of calcium, most especially in living cells, into which aequorin can be microinjected (Blinks et al., 1982). The first such experiment was reported by Ridgeway and Ashly (1968), in which they detected a calcium transient accompanying the contraction of single muscle fibers. Since then there have been many analogous applications (Blinks et al., 1982), making aequorin an important tool in analytical biochemistry, physiology, and developmental biology. Apoaequorin, which functions as the luciferase in this system, has been cloned and expressed in other cell types (Inouye et al., 1986, 1989; Tanahashi et al., 1990) where, in the presence of exogenously added coelenterazine, it serves to monitor intracellular calcium levels. For example, expressed as a transgene in Dictyostelium, it was used to monitor intracellular calcium changes in response to cyclic adenosinemonophosphate (cAMP) stimulation (Saran et al., 1994). In tobacco and Arabidopsis plants, the expressed transgene revealed circadian oscillations in free cytosolic calcium (Johnson et al., 1995); when targeted to the chloroplast, circadian chloroplast rhythms were likewise observed.

2.3 BACTERIA Luminous bacteria (see color Fig. 2.7) occur ubiquitously in sea water samples. A primary habitat where most species abound is in association with another (higher) organism, dead (saprophytes) or alive (parasites or symbionts), where growth and propagation occur. Specific associations involve specialized light organs (e.g., in fish and squid; Ruby, 1996) in which a pure culture of luminous bacteria is maintained at a high density and at high light intensity (Nealson and Hastings, 1991). Parasitic and commensal relationships are also known. Terrestrial luminous bacteria are rare, the best described being those harbored by nematodes that are parasitic on insects such as caterpillars.

Figure 2.7. Streaks of luminescent bacteria photographed by their own light, showing two strains of Photobacterium fischeri, one of which emits yellow light by virtue of having YFP (yellow fluorescent protein). The other lacks YFP, emitting only blue light. See color insert.

25

26

PHOTONS FOR REPORTING MOLECULAR EVENTS

Luminous bacteria emit light continuously, peaking at about 490 nm if no accessory protein is present. When strongly expressed, a single bacterium may emit 104–105 photons s-1. The luciferase is a flavin mixed-function monooxygenase (EC 1.14.14.3), and its presence is diagnostic for a bacterial symbiotic involvement in the luminescence of a higher organism. The pathway constitutes a shunt of cellular electron transport at the level of flavin; reduced flavin mononucleotide (FMN) (Fig. 2.1) reacts with oxygen in the presence of bacterial luciferase to produce an intermediate peroxy flavin, which then reacts with a long-chain aldehyde (tetradecanal) to form the acid and the luciferase-bound hydroxy flavin in its excited state (Hastings et al., 1985; Baldwin and Zeigler, 1992). Although there are two substrates in this case, the flavin can claim the name luciferin on etymological grounds, since it forms (bears) the emitter. The bioluminescence quantum yield has been estimated to be about 30%, the same as the fluorescence quantum yield of FMN. Curiously, no other flavin monoxygenases have been found to emit light, even at very low quantum yields, and no genes with significant sequence similarities have been recorded in any of the databases. Bacterial luciferases are heterodimeric (alpha–beta) proteins (~80 kDa) in all species; they are relatively simple, having no metals, disulfide bonds, prosthetic groups, or nonamino acid residues. The crystal structure of V. harveyi luciferase has been determined at both 2.4Å and 1.5Å resolution (Fisher et al., 1995, 1996), but so far only in the absence of substrates. Each subunit adopts the shape of an a/b barrel, and the two barrels are roughly superimposable. The b strands in the core of the superimposed structures overlap very well, while the more peripheral helical and coil elements show more structural divergence. For both subunits the classic eight-stranded barrel topology is interrupted in several locations. For example, helix a4a is located between b4 and a4, and a long coil is located between helix a4 and b5. The most significant disruption in the classic (b/a)8 topology involves a long insertion between b7 and a7. In the b subunit, this insertion comprises about 50 residues and contains a7a, a7b, and b7a, while in the a subunit it contains an additional 29 residues. It is significantly disordered in both reported crystal structures and is located in a protease labile region that is thought to be involved in interactions with the reduced flavin substrate. The heterodimer interface contains both hydrogen bonds and hydrophobic interactions and is dominated by a four-helix bundle motif involving a2 and a3 from both subunits related by a pseudo two-fold rotation axis. An interesting feature of the reaction is its inherent slowness: At 20°C the time required for a single catalytic cycle is about 20 s. The luciferase peroxy flavin itself has a long lifetime; at low temperatures (0 to -20°C) it has been isolated, purified, and characterized (Hastings et al., 1973). It can be further stabilized by long-chain alcohols and amines, which bind at the aldehyde site. However, its crystal structure has not been determined. Two major operons contain genes for the luciferase and other proteins associated with the luminescent system, including enzymes that serve to maintain the supply of myristic aldehyde (Fig. 2.8a; Meighen, 1991). There are also genes that specifically control the development and expression of luminescence. This fascinating mechanism is called “autoinduction” (Nealson et al., 1970), in which the transcription of the luciferase and aldehyde synthesis genes of the lux operon is regulated by genes of the operon itself. A substance produced by the cells called autoinducer (a homoserine lactone; Eberhard et al., 1981; Fig. 2.8b) is a product of the lux I gene. The ecological implications are evident: In planktonic bacteria, a habitat where luminescence has no apparent value, autoinducer cannot accumulate, and no luciferase synthesis occurs (Nealson and Hastings, 1991). However, in the confines of a light organ, high autoinducer levels are reached and

BACTERIA

Figure 2.8. Organization of the lux genes (a) and the homoserine structure of autoinducer (b) in Vibrio fischeri. The operon on the right, transcribed from the 5¢ to the 3¢ end, carries genes for synthesis of autoinducer (lux I), for luciferase a and b peptides (lux A and B), and for aldehyde production (lux C, D, and E). Lux R, transcribed from the operon on the left, codes for a receptor molecule that binds autoinducer, controlling the transcription of the right operon. Other genes, lux F (N), G, and H (right), are associated with the operon but with still uncertain functions. The genes coding for accessory fluorescent proteins Lump and YFP are located to the left.

the luciferase genes are transcribed. Interestingly, it has recently been discovered that an autoinduction-type mechanism, now dubbed quorum sensing (Fuqua et al., 1994), similarly controls the expression of other specific genes in several different groups of bacteria. Bacterial lux genes have been used as reporters in numerous instances (Chatterjee and Meighen, 1995), such as for visualizing gene expression in Streptomyces (Schauer et al., 1988) and following the circadian regulation of transcription in cyanobacteria (Kondo et al., 1994), to name only two. In these cases it is necessary to supply exogenous aldehyde (as a vapor), but the reduced flavin substrate need not be added, since it is generally present in all cells. Bacterial luciferase is also useful in many analytical applications, where flavin or aldehyde, or any enzyme linked to nicotinamide adenine dinucleotide (NAD) or NAD phosphate (NADP), can be assayed (Hastings et al., 1997). Several species and strains of luminous bacteria also contain accessory proteins that, like GFP, serve as secondary emitters. These include both blue- and red-shifted emissions. As with GFP, light is emitted from the luciferase reaction alone (in the absence of a second emitter protein), with the emission peaking at about 490 nm. The blue-shifted emission, due to a lumazine protein (LUMP), peaks at about 475–480 nm; the dissociable chromophore is identified as 6,7-dimethyl-8-ribitylumazine (Small et al., 1980; Petushkov et al., 1995a). A yellow emission peaking at 540 nm in a strain of V. fischeri is due to an analogous yellow fluorescent protein (YFP; recently shown to have homologies with LUMP) in which the chromophore is flavin mononucleotide (FMN) or riboflavin (Hastings et al., 1985; Macheroux et al., 1987; Karatani and Hastings, 1993; Petushkov et al., 1995b). In the YFP system, evidence has been obtained that energy transfer alone cannot account for the yellow emission (Eckstein et al., 1990). In that case a direct population of

27

28

PHOTONS FOR REPORTING MOLECULAR EVENTS

the excited state of the accessory emitter may occur, without the intermediacy of the luciferase-bound excited state.

2.4

DINOFLAGELLATES

Dinoflagellates occur ubiquitously in the oceans as planktonic forms, and they contribute substantially to the bioluminescence commonly seen at night (especially in summer) when the water is disturbed. They occur primarily in surface waters, and many species are photosynthetic. In the phosphorescent bays (e.g., in Puerto Rico and Jamaica), high densities of a single species (Pyrodinium bahamense) usually occur. The so-called red tides are blooms of dinoflagellates, and some of these are bioluminescent. About 6% of all dinoflagellate genera contain luminous species, but since there are no luminous dinoflagellates among the fresh water species, the proportion of luminous forms in the ocean is higher. As a group, dinoflagellates are important as symbionts, notably for contributing photosynthesis and carbon fixation in animals, especially corals. But unlike bacteria, no luminous dinoflagellates are known from symbiotic niches. Bioluminescent flashing is postulated to help reduce predation either by directly diverting predators or by revealing the location of the predators to their predators (Buskey et al., 1983; Hastings and Morin, 1991; Mensinger and Case, 1992; Fleisher and Case, 1995). Luminescence in dinoflagellates is emitted from many small (~0.5 mm) cortical locations. The structures have been identified as novel organelles, termed the scintillons (flashing units). They occur as outpocketings of the cytoplasm into the cell vacuole, like balloons, with their necks remaining connected (Fig. 2.9). Scintillons contain only two major proteins, dinoflagellate luciferase (LCF) and luciferin binding protein (LBP) (Desjardins and Morse, 1993); the latter sequesters luciferin and prevents it from reacting

Figure 2.9. A cartoon depicting scintillons of dinoflagellate, the organelles responsible for flashing light emission. They are formed as outpocketings of the cytoplasm projecting into the acidic vacuole.

DINOFLAGELLATES

with luciferase. Ultrastructurally, these proteins can be identified by immunolabeling (Nicolas et al., 1987, 1991; Fritz et al., 1990) and visualized with image intensification by their bioluminescent flashing following stimulation (Johnson et al., 1985), as well as by the fluorescence of luciferin, the emission spectrum of which is the same as the bioluminescence. Dinoflagellate luciferin is a novel tetrapyrrole related to chlorophyll (Fig. 2.1). Activity (quantum yield, 0.2) can be obtained in extracts made at pH 8 simply by shifting the pH from 8 to 6; it occurs in both soluble and particulate (scintillon) fractions (Fogel and Hastings, 1971, 1972). The existence of activity in both fractions is explained by the rupture of some scintillons during extraction, while others seal off at the neck and form closed vesicles. With the scintillon fraction, the in vitro activity occurs as a flash (~100 ms), very similar to that of the living cell, and the kinetics are independent of the dilution of the suspension. For the soluble fraction, the kinetics depend on dilution, as in enzyme reactions. A distinctive feature of the reaction is that the binding of luciferin to LBP is pHdependent, being bound at pH 8 and free at pH 6. Thus, the flashing of dinoflagellates in vivo is postulated to result from a transient pH change in the scintillons, triggered by an action potential in the vacuolar membrane which, while sweeping over the scintillon, opens ion channels that allow protons from the acidic vacuole to enter (Fig. 2.9). The genes for the two dinoflagellate luminescence proteins have been cloned and sequenced (Lee et al., 1993; Bae and Hastings, 1994; Li et al., 1997); there are no introns in either gene. Both proteins are synthesized and destroyed each day, mediated translationally for LBP by proteins that bind to its mRNA 3¢ untranslated region (Johnson et al., 1984; Morse et al., 1989; Mittag et al., 1994). Both of the cloned genes produce active proteins; when expressed in vitro, LBP exhibits a pH-dependent binding of luciferin while LCF catalyzes the oxidation of luciferin to give light. LCF has an interesting and unusual feature (Li et al., 1997). The approximately 140kDa protein has three tandem repeat domains (~377, 377, and 375 aa long, with no spacer sequences between). Recombinant proteins expressed from the three individual domains of the messenger ribonucleic acid (mRNA) are all separately active as luciferases. This means that in the scintillon, three different sites in the molecule could be concurrently contributing to the activity of the luciferase: a three-ring circus with the same act in all three rings. Although the three dimensional structure of full-length dinoflagellate luciferase has not been determined, crystal structures for two of the three contiguous homologous domains have been solved (Lui et al., 2003; Schultz et al., 2005). The two are very similar, consistent with their high protein sequence identity (Li et al., 1997). The overall size of a domain is about 45 by 50 by 50Å, and each comprises three subdomains, an N-terminal region of 75 residues, a C-terminal 10 stranded b-barrel, and a highly conserved central region forming a barrel that is the most likely location for binding the luciferin and the active site. However, the tetrapyrrole luciferin would be unable to fit within the barrel in its geometry at pH 8. There is strong evidence that the barrel shape is altered at low pH because of changes in the interaction between the N-terminal subdomain and the helixturn-helix region that resides above the barrel. Genes for LBP and LCF have no homologies or similarities with other luciferases or other sequences in any of the data bases. This distinctiveness is consistent with the hypothesis that luciferases have arisen independently in evolution. However, the 5¢ ends of both genes are about 50% homologous over a 90-nt region; for the luciferase, this constitutes the entire remainder of its sequence outside the three repeat regions. Both proteins bind luciferin, but since this region is not needed for luciferase activity, it must have some other

29

30

PHOTONS FOR REPORTING MOLECULAR EVENTS

function. It might be a sequence for targeting the proteins to the vacuolar membrane in the formation of scintillons.

2.5

FIREFLIES AND OTHER INSECTS

Out of a total of approximately 75,000 insect genera, there are only about 100 classed as luminous. But where seen, their luminescence is impressive, most notably in the many species of beetles: the fireflies and their relatives. Fireflies themselves possess ventral light organs on posterior segments; the South American railroad worm, Phrixothrix, has paired green lights on the abdominal segments and red head lights; and the click and fire beetles, Pyrophorini, have both “running lights” (dorsal) and “landing lights” (ventral). From ceiling perches, the dipteran cave glow worms (true flies, not beetles; they occur in New Zealand and Australia) use their light to attract flying prey, which are then entrapped. In fireflies, communication in courtship is the major function of luminescence in fireflies; one sex emits a flash as a signal, to which the other responds, usually in a speciesspecific pattern (Lloyd, 1977, 1980; Case, 1984). The time delay between the two may be a signaling feature; for example, it is precisely 2 s in some North American species. But the flashing pattern (e.g., trains distinctive in duration and/or intensity) is also important in some cases, as is the kinetic character of the individual flash (duration; onset and decay kinetics). In some species, flickering occurs within the flashes, sometimes at very high frequencies (~40 Hz). Fireflies in Southeast Asia are particularly noteworthy for their synchronous flashing; congregations of many thousands form in single trees, where the males produce an all-night-long courtship display of synchronous flashing (Buck and Buck, 1976). The adult firefly light organ comprises a series of photocytes arranged in rosettes, positioned radially around a central well, through which run nerves and trachea, the latter carrying oxygen to the cells (Ghiradella, 1977). Within the photocytes, organelles containing luciferase have been identified with peroxisomes on the basis of immunochemical labeling (Hanna et al., 1976). This identification is supported by the presence of a Cterminal peroxisomal signal sequence in luciferase (Conti et al., 1996). Although flashing is initiated by a nerve impulse that travels to the light organ, the nerve terminals in the light organ are not on photocytes but on tracheolar cells, which regulate the supply of oxygen (Case and Strause, 1978), suggesting that these cells control the flash. In support of this theory, there is a strong positive relationship between the flashing ability and the extent of the tracheal supply system in different species. On the other hand, rapid kinetics, complex waveforms, multiple flashes, and high-frequency flickering all seem unlikely to be regulated by a gas in solution. However, although oxygen might diffuse slowly, it reacts very rapidly in this system; the half rise-time of luminescence with the anaerobic enzyme intermediate (luciferase–luciferyl adenylate) is less than 10 ms (Hastings et al., 1953). Also, possibilities alternate to oxygen seem unlikely. The flash is not directly triggered by an action potential, and none of the ions typically gated by membrane potential changes (Na+, K+, and Ca2+) appear to be candidates for controlling firefly luminescence chemistry. The firefly system was the first in which the biochemistry was extensively studied. It had been known since before 1900 that cell-free extracts could continue to emit light for several minutes or hours, and that after the complete decay of the light, emission could be restored by adding a second extract, prepared by using boiling water to extract the cells (cooled before adding). The enzyme luciferase was assumed to be in the first (cold water)

FIREFLIES AND OTHER INSECTS

extract (with all the luciferin substrate being used up during the emission), whereas the enzyme would be denatured by the hot-water extraction, leaving some substrate intact. This test was referred to as the luciferin–luciferase reaction, and it was already in the first part of this century that luciferins and luciferases from the different major groups would not cross react, indicative of their independent evolutionary origins (Harvey, 1952). McElroy (1947) discovered that the addition of adenosine triphosphate (ATP) to an “exhausted” cold-water extract resulted in bioluminescence. This showed that luciferin had not actually been used up in the cold-water extract. But ATP could not be the emitter, since it does not have the appropriate fluorescence. It was thus discovered that firefly luciferin, which is a unique benzothiazole (Fig. 2.1), was still present in large amounts in the “exhausted” cold-water extract, and that it was ATP that was used up, but available in the hot-water extract. With the elucidation of the luciferin structure, ATP was shown to be required to form the luciferyl adenylate intermediate, which with the adenylate as the leaving group then reacts with oxygen to form a cyclic luciferyl peroxy species (Fig. 2.1). This breaks down to yield CO2 and an excited state of the carbonyl product (McElroy and DeLuca, 1978; Wood, 1995). A remarkably high quantum yield of 0.88 was reported (Seliger and McElroy, 1960). In reactions in which luminescence has decreased to a low level (this may continue for days), it was found that emission is greatly increased by coenzyme A (CoA), but the reason for this was obscure. The recent discovery that long-chain acyl-CoA synthetase (EC 6.2.1.3) has homologies with firefly luciferase (EC 1.13.12.7) both explains this observation and indicates the evolutionary origin of the gene (Wood, 1995). Firefly luciferase has been cloned and expressed in other organisms, including Escherichia coli and tobacco, and its crystal structure has recently been determined (Conti et al., 1996; Franks et al., 1998). It comprises a large N-terminal domain and a smaller Cterminal domain linked to the former by a flexible, four-residue coil. The N-terminal domain contains three distinct regions, two b-sheets and one b-barrel. The b-sheets are flanked on each side by helices and are related by a dyad axis of pseudo-symmetry creating a region of ababa topology within this domain. These two b-sheets flank the third structural element within the N-terminal domain, a distorted antiparallel b-barrel made of 8 b-strands. The N-terminus serves as a “cap” to this large domain and contains a small a+b motif. There is a cleft exposed to water between the N- and C-terminal distinct domains. The residues that are most conserved among all beetle luciferases and the other ATPactivating enzymes are located on the surfaces facing this cleft and on the coil connecting the domains. As a result, early analysis suggested that the active site would be located in this region. However, the cleft is too wide to allow both surfaces to interact with luciferin simultaneously, suggesting that a conformation change occurs on binding ATP or adenylate. A second hypothesis, based on structures of firefly luciferase with bromoform bound, places the active site deeper inside the enzyme somewhat away from the cleft. Unfortunately, firely luciferase has not yet been solved with luciferin bound, leaving somewhat open the questions about active site location as well as concerning domain closure during the bioluminescence reaction. To visualize expression, luciferin must be added exogenously; tobacco “lights up” when the roots are dipped in luciferin (Ow et al., 1986; see color Fig. 2.10). Luciferase catalyzes both the luciferin activation with ATP and the subsequent steps leading to the excited product. There are some beetles in which the light from different organs is a different color, and there is additional color variation between individuals of the same species.

31

32

PHOTONS FOR REPORTING MOLECULAR EVENTS

Figure 2.10. Transgenic tobacco plant carrying the firefly luciferase gene photographed by its own light. The continuous luminescence occurs following the uptake of luciferin by the roots. [From Ow et al. (1986).] See color insert.

Figure 2.11. Bacterial colonies carrying four different beetle luciferase genes cloned from the ventral organ, distinguished by their different luminescence colors: green, yellow-green, yellow and orange (Wood et al., 1989). See color insert.

In Pyrophorus plagiophthalamus, the same ATP-dependent luciferase reaction with the same luciferin occurs in the different organs, but no accessory emitter proteins have been implicated in any of these cases. Instead, differences in the luciferases appear to responsible. Different (but closely homologous) genes from a single organism have been cloned in E. coli and shown to fall into four color classes (see color Fig. 2.11; Wood et al., 1989;

REFERENCES

Wood, 1995). The chemical basis for the color differences remains to be elucidated (McCapra, 1997). Firefly luciferase has been extensively used in analytical applications for the measurement of ATP (Brolin and Wettermark, 1992; Hastings et al., 1997). The cloned gene has also been used as a reporter gene in a number of studies, most recently to monitor circadian regulation of transcription of genes in higher plants (Millar et al., 1995). The use of the genes that result in different colors of luminescence has also been explored. More recently, the simultaneous use of two different luminous systems, for example, firefly and Renilla, for assays of two different substances, has been reported (Sherf et al., 1997).

2.6 CONCLUSION Bioluminescence occurs in many different species in phylogenetically diverse groups. Among the different groups, the type and method of display of the light, its color, and its function may be very different. In two groups (and only two), bacteria and cnidaria, some of the luminous species possess accessory proteins carrying chromophores, which may serve as secondary emitters and shift the spectrum of the light. The diversity of luminous organisms is indicative of what has been firmly established over the past several decades: Bioluminescent systems in different major groups are not evolutionarily conserved, so that the genes coding for the proteins (e.g., luciferases and accessory proteins) are not homologous. The consequent biochemical diversity offers a marvelous menu for many different specific analytical and reporter applications (Hastings et al., 1997), featuring noninvasive reporting by light emission, as exemplified by GFP.

REFERENCES Bae, Y.-M., and Hastings, J. W. (1994). Cloning, sequencing and expression of dinoflagellate luciferase DNA from a marine alga, Gonyaulax polyedra. Biochem. Biophys. Acta 121:449–456. Baldwin, T., and Ziegler, M. (1992). The biochemistry and molecular biology of bacterial bioluminescence. In Chemistry and Biochemistry of Flavoenzymes, Vol. III, Müller, F. Ed., CRC Press, Boca Raton, FL, pp. 467–530. Blinks, J. R., Wier, W. G., Hess, P., and Prendergast, F. G. (1982). Measurement of C concentrations in living cells. Prog. Biophys. Mol. Biol. 40:1–114. Brehm, P., Lechleiter, J., Smith, S., and Dunlap, K. (1989). Intercellular signaling as visualized by endogenous calcium-dependent bioluminescence. Neuron 3:191–198. Brolin, S., and Wettermark, G. (1992). Bioluminescence Analysis, Vol. 71, VCH Publishers, New York, pp. 176–182. Buck, J. B., and Buck, E. (1976). Synchronous fireflies. Sci. Am. 234:74–85. Buskey, E. J., Mills, L., and Swift, E. (1983). The effects of dinoflagellate bioluminescence on the swimming behavior of a marine copepod. Limnol. Oceanogr. 28:575–579. Case, J. (1984). Firefly behavior and vision. In Insect Communication, Lewis, T., Ed. Harcourt, Brace, Jovanovich, London, pp. 195–222. Case, J. F., and Strause, L. G. (1978). Neurally controlled luminescent systems. In Bioluminescence in Action, Herring, P. J., Ed., Academic Press, New York, pp. 331–366. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263:802–805.

33

34

PHOTONS FOR REPORTING MOLECULAR EVENTS

Chatterjee, J., and Meighen, E. A. (1995). Biotechnological applications of bacterial bioluminescence (lux) genes. Photochem. Photobiol. 62:651–650. Cody, C. W., Prasher, D. C., Westler, W. M., Prendergast, F. G., and Ward, W. W. (1993). Chemical structure of the hexapeptide chromophore of the Aequorea green-fluorescent protein. Biochemistry 32:1212–1218. Conti, E., Franks, N. T., and Brick, P. (1996) Crystal structure of firefly luciferase throws light on a super family of adenylate-forming enzymes. Structure 4:287–298. Cormier, M. J. (1981). Renilla and Aequorea bioluminescence. In Bioluminescence and Chemiluminescence, DeLuca, M., and McElroy, W. D., Eds., Academic, New York, pp. 225–233. Cormier, M. J., Prasher, D. C., Longiaru, M., and McCann, R. O. (1989). The enzymology and molecular biology of the Ca++-activated photoprotein, aequorin. Photochem. Photobiol. 49:509–512. DeLuca, M., Ed. (1978). Bioluminescence and chemiluminescence. Methods Enzymol. 57:1–653. Desjardins, M., and Morse, D. (1993). The polypeptide components of scintillons, the bioluminescence organelles of the dinoflagellate. Gonyaulax polyedra. Biochem. Cell Biol. 71:176–182. Dunlap, K., Takeda, K., and Brehm, P. (1987). Activation of a calcium-dependent photoprotein by chemical signalling through gap junctions. Nature (London) 325:60–62. Eberhard, A., Burlingame, A. L., Eberhard, C., Kenyon, G. L., Nealson, K. H., and Oppenheimer, N. H. (1981). Structural identification of autoinducer of Photobacterium fischeri Luciferase. Biochemistry 20:2444–2449. Eckstein, J., Cho, K. W., Colepicolo, P., Ghisla, S., Hastings, J. W., and Wilson, T. (1990). A timedependent bacterial bioluminescence emission spectrum in an in vitro single turnover system: Energy transfer alone cannot account for the yellow emission of Vibrio fischeri Y-1. Proc. Natl. Acad. Sci. 87: 1466–1470. Fisher, A. J., Thompson, T. B., Thoden, J. B., Baldwin, T. O., and Rayment, I. (1996). The 1.5-Å resolution crystal structure of bacterial luciferase in low salt conditions. J. Biol. Chem. 271: 21956–21968. Fisher, J., Rauschel, F. M., Baldwin, T. O., and Rayment, I. (1995). The three-dimensional structure of bacterial luciferase from Vibrio harveyi at 2.4 Å resolution. Biochemistry 34:6581–6586. Fleisher, K. J., and Case, J. F. (1995). Cephalopod predation facilitated by dinoflagellate luminescence. Biol. Bull. 189:263–271. Fogel, M., and Hastings, J. W. (1971). A substrate binding protein in the Gonyaulax bioluminescence reaction. Arch. Biochem. Biophys. 142:310–321. Fogel, M., and Hastings, J. W. (1972). Bioluminescence: Mechanism and mode of control of scintillon activity. Proc. Nat. Acad. Sci. 69:690–693. Franks, N., Jenkins, A., Conti, E., Lieb, W., and Brick, P. (1998). Structural basis for the inhibition of firefly luciferase by a general anesthetic. Biophys. J. 75: 2205–2211. Fritz, L., Morse, D., and Hastings, J. W. (1990). The circadian bioluminescence rhythm of Gonyaulax is related to daily variations in the number of light emitting organelles. J. Cell Sci. 95:321–328. Fuqua, W. C., Winans, S. C., and Greenberg, E. P. (1994). Quorum sensing in bacteria: The LuxRLufamily of cell density-responsive transcriptional regulators. J. Bacteriol. 176:269–275. Ghiradella, H. (1977). Fine structure of the tracheoles of the lantern of a photurid firefly. J. Morphol. 153:187–204. Grober, M. S. (1988a). Brittle-star bioluminescence functions as an aposematic signal to deter crustacean predators. Anim. Behav. 36:493–501. Grober, M. S. (1988b). Responses of tropical reef fauna to brittle-star luminescence. J. Exp. Mar. Biol. Ecol. 115:157–168. Hanna, C. H., Hopkins, T. A., and Buck, J. (1976). Peroxisomes of the firefly lantern. J. Ultrastruct. Res. 57:150–162.

REFERENCES

Harvey, E. N. (1952). Bioluminescence. Academic, New York. Hastings, J. W. (1983). Biological diversity, chemical mechanisms and evolutionary origins of bioluminescent systems. J. Molec. Evol. 19:309–321. Hastings, J. W., and Gibson, Q. H. (1963). Intermediates in the bioluminescent oxidation of reduced flavin mononucleotide: J. Biol. Chem. 238:2537–2554. Hastings, J. W., and Morin, J. G. (1969a). Calcium-triggered light emission in Renilla. A unitary biochemical scheme for coelenterate bioluminescence. Biochem. Biophys. Res. Commun. 37:493–498. Hastings, J. W., and Morin, J. G. (1969b). Comparative biochemistry of calcium activated photoproteins from the ctenophore, Mnemiopsis and the coelenterates Aequorea, Obelia, Pelagia and Renilla. Biol. Bull. 137:402. Hastings, J. W., and Morin, J. G. (1971). Kinetics of calcium-triggered luminescence of aequorin and other photoproteins. In Contractility of Muscle Cells and Related Processes, Podolsky, R. J., Ed., Prentice-Hall, Engelwood Cliffs, NJ, pp. 99–104. Hastings, J. W., and Morin, J. G. (1991). Bioluminescence. In Neural and Integrative Animal Physiology, Prosser, C. L., Ed., Wiley-Interscience, New York, pp. 131–170. Hastings, J. W., McElroy, W. D., and Coulombre, J. (1953). The effect of oxygen upon the immobilization reaction in firefly luminescence. J. Cell Comp. Physiol. 42:137–150. Hastings, J. W., Potrikus, C. J., Gupta, S., Kurfürst, M., and Makemson, J. C. (1985). Biochemistry and physiology of bioluminescent bacteria. Adv. Microbial. Physiol. 26:235–291. Hastings, J. W., Kricka, L., and Stanley, P., Eds. (1997). Bioluminescence and Chemiluminescence: Molecular Reporting by Photons. Wiley, Chichester. Head, J. F., Inouye, S., Teranishi, K., and Shimomura, O. (2000). The crystal structure of the phosphoprotein aequorin at 2.3Å resolution. Nature 405: 372–376. Heim, R., Prasher, D. C., and Tsien, R. Y. (1994). Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proc. Natl. Acad. Sci. USA 9:12501–12504. Herring, P. J., Ed. (1978). Bioluminescence in Action. Academic, New York, 570 pp. Inouye, S., Sakaki, Y., Goto, T., and Tsuji, F. I. (1986). Expression of apoaequorin complementary DNA in Escherichia coli. Biochemistry 25:8425–8429. Inouye, S., Aoyama, S., Miyata, T., Tsuji, F., and Sakaki, Y. (1989). Overexpression and purification of the recombinant Ca2+-binding protein, Apoaequorin, J. Biochem. 105:473–477. Johnson, C. H., Roeber, J., and Hastings, J. W. (1984). Circadian changes in enzyme concentration account for rhythm of enzyme activity in Gonyaulax. Science 223:1428–1430. Johnson, C. H., Inoué, S., Flint, A., and Hastings, J. W. (1985). Compartmentation of algal bioluminescence: Autofluorescence of bioluminescent particles in the dinoflagellate Gonyaulax as studied with the image-intensified video microscopy and flow cytometry. J. Cell Biol. 100:1435–1446. Johnson, C. H., Knight, M. R., Kondo, T., Masson, P., Sedbrook, J., Haley, A., and Trewavas, A. (1995). Circadian oscillations of cytosolic and chloroplastic free calcium in plants. Science 269:1863–1865. Johnson, F. H., Shimomura, O., Saiga, Y., Gershman, G., Reynolds, G. T., and Waters, J. R. (1962). Quantum efficiency of Cypridina luminescence with a note on that of Aequorea. J. Cell. Comp. Physiol. 60:85–103. Karatani, H., and Hastings, J. W. (1993). Two active forms of the accessary yellow fluorescence protein of the luminous bacterium Vibrio fischeri strain Y1. J. Photochem. Photobiol. B. 18:227–232. Kondo, T., Tsinoremas, N., Golden, S., Johnson, C., Kutsuna, S., and Ishiura, M. (1994). Circadian clock mutants of cyanobacteria. Science 266:1233–1236.

35

36

PHOTONS FOR REPORTING MOLECULAR EVENTS

Lee, D.-H., Mittag, M., Sczekan, S., Morse, D., and Hastings, J. W. (1993). Molecular cloning and genomic organization of a gene for luciferin-binding protein from the dinoflagellate Gonyaulax polyedra. J. Biol. Chem. 268:8842–8850. Li, L., Hong, R., and Hastings, J. W. (1997). Three functional luciferase domains in a single polypeptide chain. Proc. Natl. Acad. Sci. USA 94:8954–8958. Liu, L., Im, H., Cegielski, M., LeMagueres, P., Schultz, L. W., Krause, K. L., and Hastings, J. W. (2003). Characterization and crystallization of active domains of a novel luciferase from a marine dinoflagellate. Acta Crystallogr. D 59: 761–764. Liu, Z.-J., Vysotski, E. S., Chen, C.-J., Rose, J. P., Lee, J., and Wang, B.-C. (2000). Structure of the Ca2+-regulated photoprotein obelin at 1.7Å resolution determined directly from its sulfur substructure. Protein Sci. 9: 1–9. Lloyd, J. E. (1977). Bioluminescence and communication. In How Animals Communicate, Sebeok, T. A., Ed., Indiana University Press, Bloomington, IN, pp. 164–183. Lloyd, J. E. (1980). Male Photuris fireflies mimic sexual signals of their females’ prey. Science 210:669–671. Lorenz, W. W., McCann, R. O., Longiaru, M., and Cormier, M. J. (1991). Isolation and expression of cDNA encoding Renilla reniformis luciferase. Proc. Natll. Acad. Sci. USA 88:4438–4442. Macheroux, P., Steinerstauch, P., Ghisla, S., Buntic, R., Colepiciolo, P., and Hastings, J. W. (1987). Purification of the yellow fluorescent protein from V. fischeri and identity of the flavin chromophore. Biochem. Biophys. Res. Commun. 146:101–106. McCapra, F. (1997). Mechanisms in chemiluminescence and bioluminescence—unfinished business. In Hastings, J. W., Kricka, L. J., and Stanley, P. E., Ed., Bioluminescence and Chemiluminiscence, Wiley, Chichester, pp. 7–15. McElroy, W. D. (1947). The energy source for bioluminescence in an isolated system. Proc. Natl. Acad. Sci. USA 33:342–346. McElroy, W. D., and DeLuca, M. (1978). Chemistry of firefly luminescence. In Herring, P. J., Ed., Bioluminescence in Action, Academic, New York, pp. 109–128. Meighen, E. A. (1991). Molecular biology bacterial bioluminescence. Microbiol. Rev. 55:123–142. Mensinger, A. F., and Case, J. F. (1992). Dinoflagellate luminescence increases susceptibility of zooplankton to teleost predation. Mar. Biol. 112:207–210. Millar, A. J., Carré, I. A., Strayer, C. A., Chua, N.-H., and Kay, S. A. (1995). Circadian clock mutants in Arabidopsis identified by luciferase imaging. Science 267:1161–1166. Mittag, M., Lee, D.-H., and Hastings, J. W. (1994). Circadian expression of the luciferin-binding protein correlates with the binding of a protein to its 3¢ untranslated region. Proc. Natl. Acad. Sci. USA 91:5257–5261. Morin, J. G. (1974). Coelenterate bioluminescence. In Coelenterate Biology: Reviews and New Perspectives, Muscatine, L., and Lenhoff, H., Eds., Academic, New York, pp. 397–438. Morin, J. G. (1976). Probable functions of bioluminescence in the Pennatulacea (Cnidaria, Anthozoa). In Coelenterate Ecology and Behavior, Mackie, G. O., Ed., Plenum, New York, pp. 629–638. Morin, J. G. (1983). Coastal bioluminescence: Patterns and functions. Bull. Mar. Sci. 33:787–817. Morin, J. G., and Cooke, I. M. (1971a). Behavioural physiology of the colonial hydroid Obelia. II. Stimulus-initiated electrical activity and bioluminescence. J. Exp. Biol. 54:707–721. Morin, J. G., and Cooke, I. M. (1971b). Behavioural physiology of the colonial hydroid Obelia. III. Characteristics of the bioluminescent system. J. Exp. Biol. 54:723–735. Morin, J. G., and Hastings, J. W. (1971a). Biochemistry of the bioluminescence of colonial hydroids and other coelenterates. J. Cell Physiol. 77:305–311. Morin, J. G., and Hastings, J. W. (1971b). Energy transfer in a bioluminescent system. J. Cell Physiol. 77:313–318.

REFERENCES

Morin, J. G., and Reynolds, G. T. (1969). Fluorescence and time distribution of photon emission of bioluminescent photocytes in Obelia geniculata. Biol. Bull. 137:410. Morin, J. G., and Reynolds, G. T. (1970). Luminescence and related fluorescence in coelenterates. Biol. Bull. 139:430–431. Morin, J. G., and Reynolds, G. T. (1974). The cellular origin of bioluminescence in the colonial hydroid Obelia. Biol. Bull. 147:397–410. Morin, J. G., Reynolds, G. T., and Hastings, J. W. (1968). Excitatory physiology and localization of bioluminescence in Obelia. Biol. Bull. 135:429–430. Morise, H., Shimomura, O., Johnson, F. H., and Winant, J. (1974). Intermolecular energy transfer in the bioluminescent system. Biochemistry 13:2656–2662. Morse, D., Milos, P. M., Roux, E., and Hastings, J. W. (1989). Circadian regulation of the synthesis of substrate binding protein in the Gonyaulax bioluminescent system involves translational control. Proc. Natl. Acad. Sci. USA 86:172–176. Nealson, K., and Hastings, J. W. (1991). The luminous bacteria. In The Prokaryotes, 2nd ed., Vol. I, Part 2, Chapter 25, Balows, A., Trüper, H. G., Dworkin, M., Harder, W., and Schleifer, K. H., Eds., Springer-Verlag, New York, pp. 625–639. Nealson, K., Platt, T., and Hastings, J. W. (1970). The cellular control of the synthesis and activity of the bacterial luminescent system. J. Bacteriol. 104:313–322. Nicol, J. A. C. (1958). Observations on the luminescence of Pennatula phosphorea, with a note on the luminescence of Virgularia mirabilis. J. Mar. Biol. Assoc. UK 37:551–563. Nicolas, M.-T., Nicolas, G., Johnson, C. H., Bassot, J-M., and Hastings, J. W. (1987). Characterization of the bioluminescent organelles in Gonyaulax polyedra (dinoflagellates) after fast-freeze fixation and antiluciferase immunogold staining. J. Cell Biol. 105:723–735. Nicolas, M.-T., Morse, D., Bassot, J.-M., and Hastings, J. W. (1991). Colocalization of luciferin binding protein and luciferase to the scintillons of Gonyaulax polyedra revealed by immunolabeling after fast-freeze fixation. Protoplasma 160:159–166. Ormö, M., Cubitt, A. B., Kallio, K., Gross, L. A., Tsien, R. Y., and Remington, S. J. (1996). Crystal structure of the Aequorea victoria green fluorescent protein. Science 273:1392–1395. Ow, D. W., Wood, K. V., DeLuca, M., de Wet, J. R., Helinski, D. R., and Howell, S. H. (1986). Transient and stable expression of the firefly luciferase gene in plant cells and transgenic plants. Science 234:856–859. Petushkov, V. N., Gibson, B. G., and Lee, J. (1995a). Properties of recombinant fluorescent proteins from Photobacterium leiognathi and their interaction with luciferase intermediates. Biochemistry 34:3300–3309. Petushkov, V. N., Gibson, B. G., and Lee, J. (1995b). The yellow bioluminescence bacterium, Vibrio fischeri Y1, contains a bioluminescence active riboflavin protein in addition to the yellow fluorescence FMN protein. Biochem. Biophys. Res. Commun. 211:774–779. Ridgeway, E. B., and Ashley, C. C. (1968). Simultaneous recording of membrane potential, calcium transient and tension in single muscle fibres. Nature (London) 219:1168–1169. Ruby, E. G. (1996). Lessons from a cooperative, bacterial-animal association: the Vibrio fischeriEuprymna scolopes light organ symbiosis. Annu. Rev. Microbiol. 50:591–624. Saran, S., Nakao, H., Tasaki, M., Iida, H., Tsuji, F. I., Nanjundiah, V., and Takeuchi, I. (1994). Intracellular free calcium level and its response to cAMP stimulation in developing Dictyostelium cells transformed with jellyfish apoaequorin cDNA. FEBS Lett. 337:43–47. Schauer, A., Ranes, M., Santamaria, R., Guijarro, J., Lawlor, E., Mendez, C., Chater, K., and Losick, R. (1988). Gene expression in the filamentous bacterium Streptomyces coelicolor. Science 240:768–772. Schultz, L. W., Liu, L., Cegielski, M., and Hastings, J. W. (2005). Crystal structure of a pHregulated luciferase catalyzing the bioluminescent oxidation of an open tetrapyrrole. Proc. Natl. Acad. Sci. USA 102:1378–1383.

37

38

PHOTONS FOR REPORTING MOLECULAR EVENTS

Seliger, H. H., and McElroy, W. D. (1960). Spectral emission and quantum yield of firefly bioluminescence. Arch. Biochem. Biophys. 88:136. Sherf, B., Navarro, S., Hannah, R., and Wood, K. V. (1997). Co-reporter technology integrating firefly and Renilla luciferase assays. In Hastings, J. W., Kricka, L. J., and Stanley, P. E., Eds., Co-reporter Technology Integrating Firefly and Renilla Luciferase Assays, Wiley, Chichester, pp. 228–231. Shimomura, O., and Johnson, F. H. (1975). Regeneration of the photoprotein aequorin. Nature (London) 256:236–238. Shimomura, O., Johnson, F. H., and Saiga, Y. (1962). Extraction, purification and properties of aequorin, a bioluminescent protein from the luminous hydromedusan, Aequorea J. Cell. Comp. Physiol. 59:223–240. Shorey, J., and Morin, J. G. (1974). Quantification of light produced from hydrozoan photocytes. Biol. Bull. 147:499. Small, E. D., Koka, P., and Lee, J. (1980). Lumazine protein from the bioluminescent bacterium Photobacterium phosphoreum. J. Biol. Chem. 255:8804–8810. Tanahashi, H., Ito, T., Inouye, S., Tsuji, F. I., and Sakaki, Y. (1990). Photoprotein aequorin: use as a reporter enzyme in studying gene expression in mammalian cells. Gene 96:249–255. Thomson, C. M., Herring, P. J., and Campbell, A. K. (1997). The widespread occurrence and tissue distribution of the imidazolopyrazine luciferins. J. Biolum. Chemilum. 12:87–91. Titschack, H. (1964). Untersuchungen Über das Leuchten der Seefeder Veretillum cynomorium (Pallas). Vie Milieu 15:547–563. Ward, W. W., Cody, C. W., Hart, R. C., and Cormier, M. J. (1980). Spectrophotometric identity of the energy transfer chromophores in Renilla and Aequorea green-fluorescent proteins. Photochem. Photobiol. 31:611–615. Wilson, T., and Hastings, J. W. (1998). Bioluminescence. Annu. Rev. Cell Dev. Biol. 14:197–230. Wood, K. V. (1995). The chemical mechanism and evolutionary development of beetle bioluminescence. Photochem. Photobiol. 62:662–673. Wood, K. V., Lam, Y. A., Seliger, H. H., and McElroy, W. D. (1989). Complementary DNAs encoding click beetle luciferases can elicit bioluminescence of different colors. Science 244:700–702. Yang, F., Moss, L. G., and Phillips Jr., G. N. (1996). The molecular structure of green fluorescent protein. Nat. Biotechn. 14:1246–1251.

3 BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN William W. Ward Department of Biochemistry and Microbiology, Rutgers University, Cook College, New Brunswick, NJ

3.1 INTRODUCTION The recent popularity of green fluorescent protein (GFP) as a research tool in cellular and developmental biology (Chalfie, 1995; Hassler, 1995; Kain et al., 1995; Prasher, 1995; Stearns, 1995) requires that we look very carefully at the chemical and physical properties of the GFP molecule and its chromophore. Unfortunately, the chemical and physical characterizations of native and recombinant forms of GFP and numerous mutants of the original Aequorea victoria derived clone (Chalfie et al., 1994) have not, and cannot, keep pace with the proliferation of GFP mutants and the accelerating pace in GFP applications. After 30 years of research on the prototype native GFP molecules from the jellyfish, Aequorea victoria (Morin and Hastings, 1971; Morise et al., 1974; Prendergast and Mann, 1978), and the sea pansy, Renilla reniformis (Wampler et al, 1971, 1973; Morin, 1974; Cormier et al., 1974; Ward, 1979; Ward and Cormier, 1979), these proteins are still incompletely characterized; much less is known about the chemical and physical properties of the available mutants of GFP. In this chapter, the known chemical and physical properties of GFP are summarized to provide a sound basis for the qualitative and quantitative interpretations of data generated in its applications. Nonetheless, because so much remains unknown about the biochemical properties of these molecules, users of GFP should cautiously interpret their GFP derived data.

Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

39

40

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

3.2

BIOLOGICAL FUNCTION OF GREEN FLUORESCENT PROTEIN

Biologically, GFP acts to shift the color of bioluminescence from blue to green in luminous coelenterates (jellyfish, hydroids, sea pansies, and sea pens) and to increase the quantum yield of light emission (Ward, 1979). All coelenterates utilize the same luciferin (coelenterate-type luciferin or coelenterazine) in their bioluminescence reactions (Cormier et al., 1973; Hori and Cormier, 1973; Hori et al., 1973; Wampler et al., 1973; Ward and Cormier, 1975; Inoue et al., 1977a,b; Shimomura and Johnson, 1979; Shimomura et al., 1980), producing a protein-bound oxyluciferin (Hori et al., 1973, 1975, 1977) that emits blue light in the absence of GFP. In the presence of GFP, however, the emitted light is green and identical in spectral properties to the fluorescence emission spectrum of GFP (Ward, 1979) when excited directly by exogenous radiation. Such spectral shifts are known to occur in spectroscopy by one of two general mechanisms: (1) radiative (trivial) energy transfer in which the donor molecule emits light that is subsequently absorbed and reemitted by the acceptor and (2) radiationless (often called Förster type) energy transfer in which excitation energy is transferred, without photon emission, from the donor molecule to the acceptor (Ward, 1979). Efficient trivial transfer from blue-emitting oxyluciferin to GFP requires a relatively high concentration of GFP and a sufficiently long pathlength. In a 1-cm pathlength fluorometric cuvette, for example, 90% of the incident blue light could be absorbed by wild-type GFP at a concentration of 5–10 mg ml-1 (absorbance = 1.0 at 480 nm). But, with a very short pathlength as would be seen in animal cells (10 mm diameter), the GFP concentration would need to be 1000¥ as great for trivial transfer to operate efficiently. Clearly, intracellular protein concentrations on the order of 10,000 mg ml-1 cannot be achieved. Even if all oxyluciferin emission were absorbed by GFP, the quantum yield in a trivial transfer system can be no greater than the product of the quantum yields of donor and acceptor. In the coelenterate systems, the bioluminescence quantum yield for oxyluciferin is about 0.10 (Hori et al., 1973; Hart et al., 1979) and the fluorescence quantum yield for GFP is about 0.80 (Ward and Cormier, 1979). Thus the maximum overall quantum yield, by a trivial mechanism, would be 0.10 ¥ 0.80 = 0.08, a slight decrease from oxyluciferin emission alone. But, a radiationless system traps excitation energy directly by resonance transfer, so long as donor and acceptor molecules are relatively close to each other (<100 Å). No light is actually emitted by the donor. Close proximity of donor and acceptor can be achieved by high concentrations of donor and acceptor or by chemical interactions that favor heterodimer formation at much lower concentrations. In the Renilla reniformis system, luciferase-bound oxyluciferin and GFP are brought into intimate contact via protein–protein interaction facilitating radiationless energy transfer at submicromolar protein concentrations (<0.1 mg ml-1 GFP). Such radiationless energy transfer, unlike trivial transfer, can result in a potential enhancement in overall quantum yield. Thus, in the coelenterates, if the yield of excited-state oxyluciferin is 100%, transfer of excitation energy to GFP could result in overall quantum yield of 0.80- to 10-fold greater than the maximum achievable by trivial transfer. In practice, an enhancement in radiative quantum yield of 3¥ has been measured in vitro in the R. reniformis system using native coelenterate-type luciferin (Ward and Cormier, 1976, 1978a, 1979). With certain low quantum yield luciferin analogues (Hart et al., 1979) an enhancement of 200¥ has been measured. Efficient energy transfer of this sort occurs in the Renilla system at low concentrations of GFP (10-6 M) and low concentrations of salt (<0.1 M NaCl) implicating electrostatic protein–protein association between dimeric Renilla GFP and Renilla luciferase in the mechanism (Ward and Cormier, 1978a,b). Radiationless energy transfer between protein-bound oxyluciferin and GFP has been demonstrated in this system by five

PHYSICAL CHARACTERISTICS OF GFP

independent physical means including intraphyletic cross-reactions (Ward and Cormier, 1978a). However, at similarly low-protein concentration levels, the formation of a complex of Aequorea GFP and aequorin (the blue light emitting photoprotein of Aequorea) in aqueous solution has not been demonstrated. This lack of association led to our premature conclusion that energy transfer in Aequorea occurs by a trivial mechanism (Ward and Cormier, 1978a), while earlier Morise et al. (1974), who coimmobilized aequorin and Aequorea GFP to an ion exchange gel, had reached the opposite conclusion. Years later, we demonstrated that radiationless energy transfer between aequorin and Aequorea GFP does occur in solution, but only at very high protein concentrations (Cutler, 1995; Cutler and Ward, 1993, 1997). To eliminate trivial transfer at high-protein concentrations (10– 20 mg ml-1), experiments were performed in microcapillary tubes (200-mm diameter). Our calculations indicate that GFP concentration within Aequorea photocytes (cells where bioluminescence originates) is approximately 25 mg ml-1 (Cutler, 1995). Aequorin concentration appears to be several times higher. The conditions in these microcapillary tubes simulate intracellular protein concentrations and diameter of the photogenic mass such that aequorin and Aequorea GFP form a complex that emits green light upon Ca2+ addition. In such tubes, the pathlength of emitted light is too short for significant trivial transfer from aequorin to GFP. Present evidence suggests that the functional unit in Aequorea photocytes is a heterotetramer (GFP2 · Aequorin2) (Cutler and Ward, 1997). Very high protein concentration (>10 mg ml-1), comparable to that found within the jellyfish photocytes, is required for formation of the heterotetramer (Cutler and Ward, 1997).

3.3 NATURAL SOURCES OF GFP Green fluorescent proteins are found in a large number of bioluminescent coelenterates within the classes hydrozoa and anthozoa (Morin and Hastings, 1971; Cormier et al., 1974; Morin, 1974; Ward, 1979; Levine and Ward, 1982; Prasher, 1995). It is not clear, however, that they occur at all within the scyphozoans and there are no reports of GFPs in the closely related phylum ctenophora. All known naturally occurring GFPs, and most of the characterized mutants of wild-type recombinant Aequorea GFP (Heim et al., 1994, 1995; Heim and Tsien, 1996; Ehrig et al., 1995; Cubitt et al., 1995; Delagrave et al., 1995; Cormack et al., 1996; Crameri et al., 1996), emit light with wavelength maxima in the 490- to 520-nm range; most are centered at 508–509 nm. However, the range of excitation (absorption) maxima of purified and partially characterized native GFPs is much greater (395–498 nm). Curiously, the only naturally occurring GFP molecule with an excitation maximum in the ultraviolet (UV) region (Morise et al., 1974) is the one found in the hydrozoan jellyfish A. victoria (lmax = 395 nm), and this is the only GFP for which the gene has been cloned (Prasher et al., 1992; Chalfie et al., 1994). The second most blue-shifted excitation maximum of a natural GFP is seen in the GFP of Halistaura (Mitrocoma) cellularia (lmax = 465 nm). Excitation and emission spectra of four naturally occurring GFPs are shown in Fig. 3.1.

3.4 PHYSICAL CHARACTERISTICS OF GFP Most of our knowledge of the physical characteristics of GFP comes from work on R. reniformis GFP and A. victoria GFP (Tables 3.1–3.4). Only minimal information is avail-

41

42

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

Figure 3.1. Excitation (upper) and emission (lower) spectra of the green fluorescent proteins from Aequorea (A), Halistaura (Mitrocoma) (H), Phialidium (P), and Renilla (R). Excitation maxima are at 395, 465, 485, and 498 nm, respectively. Emission maxima are at 509, 497, 498, and 508 nm, respectively. Data were collected with a fully corrected Spex fluorometer at room temperature in 10 mM Tris-ethylenediamenetetracetic acid (EDTA) buffer at pH 8.0. Data were collected by Dr. Richard Ludescher in the Rutgers University Department of Food Science.

able about the 20 other naturally occurring GFPs that have been studied—usually with nothing more than the wavelength of peak emission being reported (Cormier et al., 1974; Ward, 1981; Levine and Ward, 1982; Prasher, 1995). All known GFPs are acidic, compact, globular molecules with monomer molecular weights (Fig. 3.2) of about 27 kDa (see Chapter 5 by Phillips for details of X-ray structure). With the exception of Aequorea GFP and recombinant forms thereof, which are monomers in dilute aqueous solution (Prendergast and Mann, 1978; W. Ward and T. Spires, unpublished), all other GFPs are stable, nondissociable dimers (2 ¥ 27 kDa) that remain dimeric (Fig. 3.3) unless denatured

43

PHYSICAL CHARACTERISTICS OF GFP

TABLE 3.1. Aequorea and Renilla GFP: Comparative Physical Data Aequorea Monomer molecular weighta

Renilla 27 kDab

27 kDa 26.9 kDac 4.6–5.1d 508e–509 nm 0.72–0.78e 0.80

Isoelectric point(s) (pI) Fluorescence emission maximum Fluorescence quantum yield Molar extinction coefficient (monomer) el1M (liter mol-1 cm-1) l = 498 nm l = 475 nm l = 397 nm l = 280 nm Absorption ratio (highest purity achieved) 498 nm/280 nm 397 nm/380 nm

5.34 ± 0.07b 509 nmb 0.80b

133,000b 53,000b <1,000b 22,000b

3,000 14,000 27,600 22,000

5.6b–6.0 1.25

a At moderate protein concentration of Aequorea GFP (<0.5 mg mL-1) the monomeric form predominates. At higher protein concentrations of Aequorea GFP (>2.0 mg mL-1) the dimeric form predominates. Renilla GFP is dimeric (2 ¥ 27 kDa) at all concentrations unless denatured. b From Ward and Cormier (1979). c From Prasher et al. (1992). Based upon sequence of cDNA. d From Cutler (1995). Nine isoforms have been characterized. e From Morise et al. (1974).

TABLE 3.2. The Amino Acid Compositions of Renillas and Aequorea GFP Amino Acids Lysine Histidine Arginine Half-cystine Methionine Aspartic acid Asparagine Glutamic acid Glutamine Threonine Serine Proline Glycine Alanine Valine Isoleucine Tyrosine Phenylalanine Tryptophan Amino sugars a

Renilla GFP Nearest Integer per 27,000 Daa

Aequorea GFP from cDNA Sequence gfp 10b

19 8 7 2c 9

20 10 6 2 6 18 13 16 8 15 10 10 22 8 17 12 11 13 1 0

} 20 } 27 17 15 11 22 14 18 14 11 13 0d 0e

From Ward and Cormier (1979). Each value represents the average from hydrolyses of 24, 48, and 72 h unless otherwise indicated. b From Prasher et al. (1992). c Determined as cysteic acid following performic acid oxidation. d Determined by hydrolysis in the presence of thioglycolate. e Determined by hydrolysis with p-toluenesulfonic acid.

Ser

Met

Glu

Ile

Thr

His

Gly

Met

Asp

Glu

Leu

Tyr

Lys

Residue 108

Residue 141

Residue 172

Residue 219

Residue 230

Residue 231

Residue 232

Residue 233

Residue 234

Residue 235

Residue 236

Residue 237

Residue 238

Lys

Tyr

Leu

Glu

Asp

Met

Gly

His

Thr

Val

Glu

Leu

Thr

Phe

Gly

His

Thr

Val

Glu

Leu

Thr

Phe

Gln

c

b

a

25,953

25,943

His

Thr

Val

Glu

Leu

Thr

Phe

Gln

M2

25,816

25,806

Thr

Val

Glu

Leu

Thr

Phe

Gln

M3

L2

25,830

25,824

Thr

Val

Lys

Met

Ser

Tyr

Gln

25,841

25,838

Thr

Ile

Lys

Met

Ser

Tyr

Gln

N-Acetylated serine

L1

Thr

Ile

Glu

Met

Ser

Tyr

Gln

L3

25,841

25,839

Shaded amino acids represent those that have been determined directly by cDNA sequencing of C-terminal protein sequencing. From Cutler (1995). An additional “c” isoform has been found (data not shown).

26,009

Tyr

Residue 100

Gln

Experimental MW (Da)

Gln

Residue 80

Met

M1

26,000

Met

N-terminal amino acid

gfp10 cDNA

Calculated MW (Da)

gfp2 gene

Structure Element

TABLE 3.3. Structural Differences Among Native and Recombinant GFP Isoformsa,b

26,279

26,279

Asp

Met

Gly

His

Thr

Ile

Glu

Met

Ser

Tyr

Gln

C1

26,393

26,391

Glu

Asp

Met

Gly

His

Thr

Ile

Glu

Leu

Ser

Tyr

Gln

C2c

26,845

26,836

Lys

Tyr

Leu

Glu

Asp

Met

Gly

His

Thr

Val

Glu

Leu

Thr

Phe

Arg

Ala

rGFP

44 BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

45

PHYSICAL CHARACTERISTICS OF GFP

TABLE 3.4. Molar Extinction Coefficients of Native and Variant Green Fluorescent Proteins Molar Extinction Coefficienta (liter mol-1/cm-1) per Monomeric Unit GFP Source 1. Renilla GFPb Dimer 2. Aequorea GFP Monomerc Monomerd Dimer 3. wt Recombinant Monomere Monomerd Dimer 4. “Cycle 3” variant Monomer Monomerd 5. S-65-T variant Monomer/dimerf Monomer/dimerd 6. Mut 1 variant Monomer/dimerf Monomer/dimerd

397 nm

475 nm

489 nm

498 nm 133,000

27,600 25,000 30,000

14,000 11,000 3,000

30,000 28,000 34,000

12,000 11,000 2,500

30,000 30,300

8,000 8,500 58,000 56,000 57,000 55,000

Based on E44 = 44,000 liter mol-1 cm-1 for the denatured protein in 0.1 M NaOH, unless otherwise indicated. From Ward and Cormier (1979); Renilla GFP has never been shown to exist as a fluorescent monomer. c From Ward (1981); lmax for native Aequorea GFP at 395 and 470 nm. d Based on measured extinction coefficients at 292 nm for Trp (3590 liter mol-1 cm-1) and Tyr (2340 liter mol-1 cm-1) in 0.1 NaOH. Each variant contains 1 Trp and 11 Tyr, \ e292 = 29,300. e Average of seven determinations. f Average of five determinations. a b

(Fig. 3.4) (Ward and Cormier, 1979; W. Ward, A. S. Sawyer, and T. Spires, unpublished results). Aequorea and Renilla GFPs and many of the mutants of recombinant wild-type GFP are isolated, by ion exchange chromatography or isoelectric focusing, as a family of closely related isoforms (Cutler, 1995) with isoelectric points between 4.6 and 5.4 (for purification details see Protocol 1 by González and Ward in Protocol I, Section I.E). In the jellyfish or the sea pansy, a contributor to micro-heterogeneity is the presence of multiple GFP genes or alleles having several point mutations (Prasher et al., 1992). These mutations produce as many as five internal amino acid substitutions, some of which generate differences in the isoelectric points of resulting isoforms. Recombinant GFPs, products of single gene transfections, are not expected to have internal amino acid substitutions and should purify as a single isoform by ion exchange chromatography or isoelectric focusing. But this is not usually the case. Full-length wild-type recombinant GFP is a protein containing 238 amino acids, the C-terminal segment of which has the sequence His-GlyMet-Asp-Glu-Tyr-Lys (Prasher et al., 1992). Unlike the core of the protein, which is highly resistant to proteolysis (Roth, 1985), this C-terminal “tail” is quite susceptible to attack by carboxypeptidases and by nonspecific proteases such as proteinase K and pronase

46

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

1

2

3

4

1

2

3

4

5

5

6

6

7

7

8

8

9

9

Figure 3.2. Sodium dodecyl sulfate (SDS) gel electrophoresis of recombinant GFP samples. Samples were heated 10 min in 70 mM SDS solution containing 0.05 M dithiothreitol in bis–Tris buffer prior to application to a 1 mm, 4–20% Tris–Glycine polyacrylamide gel (Novex, San Diego, CA) with MOPS/SDS running buffer. Standards include molecular-weight ladders (lanes 1 and 8) containing standards of 200, 120, 110, 100, 90, 80, 70, 60, 50, 40, 30, 20, and 10 kDa (Gibco, Gaithersburg, MD) and low-molecular-weight standards (lanes 2 and 9). The low molecular weight standards include ovalbumin, carbonic anhydrase, ß-lactoglobulin, lysozyme, bovine trypsin, and insulin a and b chains having molecular weights of 43, 29, 18.4, 14.3, 6.2, and 3 kDa, respectively. Sample lanes are as follows: Upper gel: lane 3-Mut 1, lane 4-S65T, lane 5-“cycle 3,” lane 6-Y66H, lane 7-wild type recombinant. Lower gel: lane 3-wild-type recombinant, lane 4native Renilla mulleri, lane 5-native Renilla reniformis, lane 6-native Halistaura (Mitrocoma) cellularia, lane 7-native Phialidium gregarum. All bands of native and recombinant GFP migrate as monomeric proteins of molecular weight 25–26 kDa in this system.

PHYSICAL CHARACTERISTICS OF GFP

Figure 3.3. Semilog plot of the molecular weights of Aequorea GFP and Phialidium GFP vs elution time from a Phenomenex SEC S-2000 size exclusion high-performance liquid chromatography (HPLC) column. The six reference standards are Blue Dextran (BD), ferritin (FERR), bovine serum albumin (BSA), ovalbumin (OVAL), carbonic anhydrase (CA), and myoglobin (MYO) having molecular weights of 2000, 440, 67, 43, 29, and 17 kDa, respectively. Based on these standards, Aequorea GFP elutes at an apparent molecular weight of 28.5 kDa and Phialidium GFP elutes at an apparent molecular weight of 52 KDa. These apparent molecular weights are independent of GFP protein concentration for all detectable sample dilutions below 1 mg ml-1. Sample concentrations as low as 1 mg/ml-1, which can be detected with an on-line fluorimetric monitor, show no shifts in apparent molecular weight for any of the GFPs tested. The column was equilibrated with and eluted with an aqueous pH 6.5 buffer consisting of 50 mM sodium phosphate, 100 mM sodium chloride, and 0.02% w/v sodium azide. Flow was maintained at 0.5 ml/min and a sample volume of 80 ml was injected at time zero. Blue dextran and ferritin mark the column void volume (7.8 ml), which elutes at 15.6 min. All recombinant forms of GFP tested under these conditions (wt, S65T, “cycle 3,” Mut 1, and Y66H) elute as monomers (26–30 kDa) while, with the exception of Aequorea GFP, all native forms of GFP [R. reniformis, R. mulleri, R. kollikeri, Phialidium gregarum, and Halistaura (Mitrocoma) cellularia] elute as dimers (50–54 kDa).

47

48

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

Figure 3.4. GFP fluorescence versus pH. Solid circles for Renilla GFP, open circles for Aequorea GFP. Samples incubated 5 ± 0.5 min at 22 ± 2°C before fluorescence was measured on a Turner 110 fluorimeter with blue lamp, Ditric FITC excitation filter, and Corning 3–70 emission filter. Buffers were: 0.05 M each of glycine · HCl, sodium phosphate, sodium citrate (pH 2–11), 0.025 M Na2HPO4 (pH 11–11.9), 0.05 M KCl · NaOH buffer (pH 12–13), and 0.1–1.0 M NaOH (pH 13–14). [From Ward (1981).]

(Cutler, 1995). Because this tail contains two basic amino acid residues (His and Lys) and two acidic amino acid residues (Asp and Glu), partial proteolytic cleavage can generate a rich array of isoforms including many that are separable by ion exchange chromatography, isoelectric focusing, and native gel electrophoresis. We have identified and sequenced nine isoforms of native Aequorea GFP having five internal substitutions and five different C-terminal truncations (Table 3.3). Fortunately, we see a much lower degree of microheterogeneity in recombinant forms of GFP, especially if we take precautions to reduce C-terminal proteolysis (by working at 0–4°C and by adding the serine protease inhibitor phenylmethylsulfonyl fluoride (PMSF).

3.4.1

GFP Stability

One of the great advantages of GFP as a reporter of gene expression is its high level of stability. This stability appears to be the consequence of its unique three-dimensional structure (Ormö et al., 1996; Yang et al., 1996a, 1997; Wu et al., 1997). Eleven beta strands surround and protect the chromophore that is positioned near the geometric center of a “beta-can” (see Chapter 4 by Phillips). Short loop regions and distorted alpha helices cap both ends of the “can.” The protection is so complete that classical fluorescence quenching agents such as acrylamide, halides, and molecular oxygen have almost no effect on GFP fluorescence (Rao et al., 1980). The purified wild-type recombinant GFP derived from gfp 10 (Chalfie et al., 1994) shows no significant differences in its stability properties when compared with natural GFP purified from A. victoria. In most respects, purified R. reni-

PHYSICAL CHARACTERISTICS OF GFP

formis GFP is even more stable than A. victoria GFP and its chromophore is less responsive to external perturbations (Ward et al., 1982). Renilla GFP, for example, has a broader pH stability profile than Aequorea GFP (Fig. 3.4) and much greater stability in protein denaturing solutions such as 8 M urea, 1% SDS, and 6 M guanidine hydrochloride (Ward and Cormier, 1979; Ward et al., 1982; W. Ward, A. S. Sawyer, and T. Spires, unpublished results).

3.4.2

Denaturation and Renaturation

The Tm (temperature at which one-half of the endogenous fluorescence is lost) for Aequorea GFP is 76°C (Bokman and Ward, 1981; Ward, 1981). For Renilla GFP the value is 70°C (Ward, 1981) and for Phialidium GFP the Tm is 69°C (Levine and Ward, 1982). In both Aequorea and Renilla GFP, the far-UV circular dichroism (CD) signal (at 205, 207.5, and 252.2 nm) decays in parallel with the fluorescence as the temperature is raised [such CD changes are indicative of the loss of secondary and tertiary structure] that fluorescence depends on the intact secondary and tertiary structure of the protein (Bokman and Ward, 1981). Although heat-denatured GFP does not renature effectively, fully denatured GFP will recover most or all of its original fluorescence following other conditions of denaturation as described below. The fluorescence of Renilla GFP is completely unperturbed over a wide pH range from about 5.5–12.6 (Ward, 1981) (Fig. 3.4). Between pH 4.5 and 5.5, the fluorescence is metastable and fluorescence bleaching increases over time, especially at elevated temperature. Renilla GFP is stable at its isoelectric point (pH 5.34) for several days at 4°C but is much less stable at room temperature. Renilla GFP will tolerate a pH of 12.6 for an hour or more, but pH values above 12.6 cause almost instantaneous (but reversible) loss of fluorescence as the protein denatures. Native Aequorea GFP, the wild-type recombinant protein (Chalfie et al., 1994), and the “cycle 3” variant of GFP (Crameri et al., 1996) show a similar pattern of fluorescence as Renilla GFP at the acidic end of the pH spectrum, but they are less stable than Renilla GFP at their isoelectric points (pH 4.7–5.1). These proteins will bleach almost completely during a 3-h long isoelectric focusing experiment at 15°C. At 10°C or below, they are stable for at least 3 h. These proteins are dramatically different from Renilla GFP under alkaline conditions (Ward et al., 1982). Beginning at about pH 10.0 and continuing to pH 12.2, native Aequorea GFP and the related recombinant forms (wild type and “cycle 3”) undergo large shifts in their absorption (excitation) spectra (Fig. 3.4) (Ward et al., 1982; González et al., 1997). The familiar peak at 395 nm (Morise et al., 1974) drops in intensity and the shoulder at 475 nm increases three-fold (Fig. 3.5) (Ward et al., 1982). If, at pH 12, fluorescence is excited in the 475-nm region (e.g. FITC optics or 488-nm argon laser line), the intensity of fluorescence will appear to triple, making quantitation of fluorescence in the alkaline range more difficult. However, when, at pH 12, Aequorea GFP is excited in the UV (e.g., 365 nm), fluorescence will appear to drop by two-fold as compared with pH 8. The pH dependent spectral shifts in GFP absorption/excitation spectra in the extreme alkaline range may be the result of ionization of tyrosine at position 66 in the chromophore and/or deprotonation of arginine at position 96 that stabilizes the enol form of the chromophore imidazolone (Fig. 3.6). Both Renilla and Aequorea GFPs will recover fluorescence after acid or base denaturation. With native Aequorea GFP, we have demonstrated up to 80% recovery of fluorescence following denaturation in acid (pH 1), base (pH 13), 6 M guanidine · HCl, and

49

50

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

Figure 3.5. Aequorea GFP absorption spectra at various pH values. Samples incubated 30 min at 22 ± 2°C at pH: 5.46 (a), 8.08 (b), 10.22 (c), 11.07 (d), 11.55 (e), 13.0 (f) and 1.0 (g). For curves a–e the buffer contained 0.01 M each of sodium citrate, sodium phosphate and glycine. Sample f was in 0.1 M NaOH, and sample g was in 0.1 M HCl. Note isosbestic points at 422 and 405 nm. [From Ward (1981).]

8 M urea (Ward and Bokman, 1982); however, reproducibility has always been a problem and yields of renatured GFP vary greatly. Renaturation occurs rapidly (half-time in tens of seconds) upon adjustment to pH 8 (from acid or base) or following 12-fold dilution into aqueous buffer (from guanidine or urea). The percentage recovery is greater if the experiments are performed at 0°C. Denatured protein slowly oxidizes in air (half-time <3 days). After 7 days in air at 0–4°C, no recovery of fluorescence is seen. However, full fluorescence is restored by the addition of 2-mercaptoethanol (1 mM) (Surpin and Ward, 1989). The wild-type recombinant form of Aequorea GFP behaves in a very similar manner with respect to denaturation and renaturation. We have found, however, that the S65T variant recovers 100% from base denaturation (A. S. Sawyer, D. Gonzalez, and W. Ward, unpublished results). We have been unable to renature R. reniformis GFP from 6 M guanidine · HCl despite many attempts, but this protein will renature from acid or base as described above (Sawyer and Ward, unpublished). Better fluorescence recovery (up to 80%) is seen in renaturation from base than from acid (20–40%).

3.4.3

Effect of Proteases

When used at moderate concentration (0.1 mg ml-1), none of the common proteases has any effect on GFP fluorescence, even after 24 h incubation under optimum conditions for the protease. This protease resistance is true for Renilla GFP (subtilisin) and Aequorea GFP (trypsin, chymotrypsin, thermolysin, elastase, ficin, proteinase K, chymopapain, papain, subtilisin, pancreatin, bromelain, and pronase). At higher protease concentrations (43 mg ml-1), following 60-h incubation with each of these 12 proteases, individually, only bromelain and pronase appear to have affected Aequorea GFP fluorescence significantly

51

PHYSICAL CHARACTERISTICS OF GFP

Gln94 O

(a)

NH2

NH Arg96

H2N NH2+

O

N

HN

N

HO

N

O N

His148

Phe

NH2

NH +H

–

OH

HO Glu222

(b)

Gly

2N

NH2

O

O

H

N

NH

N

N

O

Gly N

His148

HO

Phe

OH O

Figure 3.6. Tautomeric forms of the GFP chromophore showing crucial interactions with four amino acid side chains. Form (a) is the protonated phenolic form of the chromophore with a keto oxygen on the imidazole ring. This form is thought to be responsible for the absorption band centered near 395 nm. Form (b) is the quinone–enol form of the chromophore with a full negative charge on the imidazolone oxygen. This form is thought to be responsible for the absorption band centered near 475 nm (wild type), 489 nm (S-65T, Mut 1), or 498 nm (Renilla GFP). [Diagram from Yang, et al. (1997) and Youvan and Michel-Beyerle (1996).]

(Roth and Ward, 1983; Roth, 1985). Protease concentrations as high as 1 mg ml-1 (trypsin, chymotrypsin, papain, subtilisin, thermolysin, and pancreatin) have shown no effect on Aequorea GFP fluorescence (Bokman and Ward, 1981). In fact, GFP is so stable that it tolerates massive contamination with bacteria or fungi when, inadvertently, it is stored for months at room temperature in buffer or in buffered sucrose solutions without antimicrobial agents (W. Ward, unpublished results). Isolation of the chromophoric hexapeptide from Aequorea GFP by papain digestion (Cody et al., 1993) required prior heat denaturation of the protein, as papain has no effect on undenatured GFP.

3.4.4

Effect of Organic Solvents

A number of years ago, following purification of native Aequorea GFP by hydrophobic interaction chromatography on a phenyl sepharose column, in which we used 50% ethylene glycol as the eluting solvent, we measured a GFP recovery of 150% (Robart and Ward, 1990). The “150% recovery” was traced to a direct organic solvent effect on the fluorescence excitation spectrum of GFP (but not on the fluorescence emission spectrum). Since then, we have surveyed 25 different water-miscible organic solvents at 10% increments (from 10–90% v/v in 10 mM Tris–EDTA buffer at pH 8.0). Every organic solvent that we have surveyed, even glycerol, perturbs the absorbance/excitation spectrum of native

52

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

and recombinant wild-type Aequorea GFP (González et al., 1997; A. S. Sawyer, E. Castriciones, and W. Ward, unpublished results). In each case, the solvent induces a spectral shift, qualitatively and quantitatively similar to the shift produced in the GFP absorption spectrum when pH is changed from 8.0–12.0 (Ward et al., 1982). The 395-nm absorption peak decreases in intensity while the 475-nm peak increases severalfold, suggesting a transition in the electronic form of the chromophore from that represented in Fig 3.6a to that represented in Fig. 3.6b. The solvent-induced apparent increase in fluorescence, observed when exciting GFP with optics that favor the longer wavelength excitation band (Ditric FITC broad-band interference filter, Ditric Optics) has, at times, exceeded fourfold. The most effective solvent in promoting this shift is acetonitrile, which induces large spectral shifts at concentrations as low as 5% v/v. Similar organic solvent experiments with native R. reniformis GFP and with Aequorea GFP variants S65T and Mut 1 (EGFP) show no changes at all in excitation or emission spectra, suggesting that in Renilla GFP, S65T, and Mut 1 the chromophore may always be in the electronic form represented in Fig. 3.6b. In most solvents, at very high organic solvent concentration (>60% v/v) both Aequorea and Renilla GFPs lose fluorescence intensity without undergoing further spectra shifts, consistent with solvent-induced denaturation.

3.4.5

Effect of Detergents and Chaotropes

Surprisingly, none of the GFP types we have studied shows any significant change in the intensity or peak position of excitation or emission when observed within 60 min of treatment with a wide array of anionic, cationic, zwitterionic, and nonionic detergents (0.01, 0.10, 1.0% w/v) or chaotropic agents at room temperature (González et al., 1997; A. S. Sawyer, E. Castriciones, and W. Ward, unpublished results). Examples among the detergents include: SDS, CHAPS, CTAB, Tween 80, and Triton X-100. Among chaotropic agents, those that show little or no effect on GFP fluorescence include: 8 M urea, 4 M guanidine · HCl, 1 M guanidine · SCN, and 4 M KI. Over longer periods of time or at slightly higher temperature (40°C) both Renilla GFP and Aequorea GFP will lose fluorescence in detergents or chaotropes as the proteins slowly denature, but always Aequorea GFP fades more rapidly, suggesting, again, a more rigid tertiary conformation in the case of Renilla GFP.

3.4.6

Effect of Fixatives and Preservatives

To the cellular and developmental biologists and microscopists who work with GFP, the most useful stability property of GFP may be its tolerance to fixatives such as formaldehyde and glutaraldehyde (Chalfie et al., 1994). Retention of GFP fluorescence in these traditional fixatives provides an opportunity to view and localize GFP in preserved tissue. We have, for example, been able to store solutions of GFP for several weeks in 3% buffered formaldehyde solutions (10% formalin) without appreciable loss of fluorescence. For longterm storage of preserved tissues, formaldehyde may be preferable to glutaraldehyde, as the latter tends to yellow with age and to develop its own bluish fluorescence. Thus opaque internal organs genetically transformed with GFP can be fixed with 10% formalin and prepared by frozen sectioning for microscopic localization of GFP. Water soluble embedding, with substances such as carbowax, followed by conventional microtome sectioning could possibly be used instead of cryosectioning. Because GFP loses all fluorescence in absolute ethanol, it is very unlikely that GFP will tolerate complete dehydration, as required for paraffin or plastic imbedding.

SPECTROSCOPY AND CHROMOPHORE STRUCTURE

3.5 SPECTROSCOPY AND CHROMOPHORE STRUCTURE 3.5.1

Chromophore Structure

Shimomura (1979) was first to propose a structure for the Aequorea GFP chromophore, which he released from the protein by papain digestion. He also noted a similarity between this chromophore and coelenterate-type luciferin (coelenterazine). Later we showed (Cody et al., 1993) that the papain limit digest chromopeptide of Aequorea GFP is a cyclized hexapeptide derived from an internal portion of the GFP primary sequence (Phe64-SerTyr-Gly-Val-Gln69). Mass spectroscopy of the isolated hexapeptide indicates the loss of 20 mass units, consistent with dehydration (-18) and dehydrogenation (-2) reactions. Peptide sequencing, two-dimensional nuclear magnetic resonance (2D NMR), and amino acid analysis following further proteolytic degradation of the hexapeptide with carboxypeptidase and pronase show the locations of posttranslational changes. The data demonstrate that cyclization involves condensation between the carboxyl carbon of serine and the amino nitrogen of glycine and the dehydrogenation of the tyrosine methylene bridge. The structure of this chromophore, shown in Figure 3.7a (Cody et al., 1993), has been confirmed by X-ray crystallography (see Chapter 4 by Phillips). The hexapeptide chromophore structure in Renilla reniformis GFP appears to be identical to that of Aequorea GFP except that Val68 and Gln69 are replaced by Asp and Arg, respectively (SanPietro et al., 1993). Several chromophore variants of GFP that produce a fluorescent product have been isolated. Serine at position 65 has been replaced by Thr, Ala, Cys, Leu, and Gly, tyrosine at position 66 has been replaced by Phe, Trp, and His (Cubitt et al., 1995; Delagrave et al., 1995). To date, no functional GFP has been reported in which Gly67 is replaced by any other amino acid, suggesting either that changes at this position do not affect

Figure 3.7. (a) Structure of the chromophore of the GFP. (b) Structure of coelenterate-type luciferin (coelenterazine).

53

54

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

fluorescent properties of GFP or that there is an essential role for Gly67 in chromophore formation (Delagrave et al., 1995). The chemical mechanism for chromophore formation is not completely understood; however, it is clear that molecular oxygen is required (Davis et al., 1994; Heim et al., 1994). Furthermore, the ability to express GFP in a cell-free translation system (Kolb et al., 1996) reinforces the belief that no other cofactors or enzymes are required for cyclization of the apoprotein to make a fluorescent chromophore. Heim et al. (1994) and Cubitt et al. (1995) proposed a plausible mechanism for GFP chromophore formation in which cyclization to form the imidazolone ring precedes tyrosine side chain oxidation by molecular oxygen, the latter reaction occurring with a time constant of about 4 h. Others have created or selected for GFP variants in E. coli with postinduction “greening” rate constants of less than 2 h. The “cycle 3” variant, for example (Crameri et al., 1996), reaches 50% of maximum fluorescence in E. coli in 95 min postinduction.

3.5.2

The Question of Brightness

Most of the applications of GFP in cellular and developmental biology published in the years 1994–1996 have utilized the original wild-type clone of GFP that is based on the gfp 10 gene (Prasher et al., 1992; Chalfie et al., 1994). While some of the fluorescence images derived from the GFP product of this gene are indeed spectacular, such results are seen when the wild-type gene is controlled by an exceptionally strong promoter. Systems with weak promoters may give disappointing results, so there has been a demand for a “brighter” form of GFP. To the GFP applications expert, “brighter” may mean higher signal and improved contrast in fluorescence microscopy, confocal microscopy, or fluorescence activated cell sorting (FACS). Here, critical concerns include the intensity and spectral distribution of the exciting lamp, the selection of excitation and emission filters, and the relative photostability of the GFP in question. To the molecular biologist, a “brighter” GFP may be one that is more efficiently transcribed and translated because of optimum plasmid copy number more efficient protein folding, elimination of cryptic introns, improved codon usage, and absence of GFP-containing inclusion bodies. To the biochemist and biophysicist, “brighter” means higher fluorescence quantum yield, higher molar extinction coefficient, and/or greater absorption cross section. The quantitation of any one of these parameters associated with “brightness,” from photostability, to plasmid copy number, to molar extinction coefficient, is a nontrivial process. As a biochemist, I will focus on the “brightness” factors derived from quantum yield, molar extinction coefficient, and absorption cross-section.

3.5.3

GFP Quantum Yield

Now, there seems to be general agreement that the fluorescence quantum yield for R. reniformis GFP (Ward and Cormier, 1979; Ward, 1979) and A. victoria GFP (Morise et al., 1974; Kurian et al., 1994) is 0.8. Earlier references, citing fluorescence quantum yields for Renilla GFP of 0.3 (Cormier et al., 1974; Wampler et al., 1971; Ward and Cormier, 1976), are in error. Quantum yield measurements require the collection of fluorescence emission spectra from the unknown and from an accepted reference standard that closely approximates the unknown in spectral emission. Fluorescein satisfies these requirements for all forms of GFP except Y66H and related blue emission mutants. Both fluorophores are adjusted to the same low-absorbance value spectrophotometrically (generally 0.0100 absorbance

SPECTROSCOPY AND CHROMOPHORE STRUCTURE

units) at a wavelength common to both excitation spectra (470 nm can be used for all “nonblue” GFPs). In the case of the native GFP and the wild-type and cycle 3 recombinant GFPs, measurement of a concentrated protein stock solution followed by dilution to 0.0100 absorbance units may introduce a large error. In the 470-nm region, all three of these forms of GFP deviate greatly (by a factor of 4 or 5) from Beer’s law upon dilution (Fig. 3.5), especially the “cycle 3” variant (Morise et al., 1974; Ward et al., 1982; W. Ward, D. González, and A. S. Sawyer, unpublished results). Fluorescence emission spectra are then recorded and the integrated spectra are compared mathematically to derive a quantum yield for the unknown. If the unknown and reference standard differ significantly in the shapes of their spectral emissions, then corrections for spectral sensitivity of the emission monochromator and photomultiplier tube must be employed. Further questions arise when dealing with unknowns having bimodal excitation spectra (native Aequorea GFP, wild-type recombinant, and the “cycle 3” variant), the peaks of which vary in relative intensity as a function of protein concentrations, pH, temperature, and ionic strength (Ward et al., 1982), solvent composition of the external environment (Robart and Ward, 1990; González et al., 1997), and exposure to intense light (Cubitt et al., 1995; Patterson et al., 1997). Estimates of quantum yields, ranging from 0.21–0.77, for seven different GFP variants expressed in E. coli, including wild-type (0.77), have been reported (Heim and Tsien, 1996). However, raw data, calculations, and methodology are not reported in this publication, so the potential for independent interpretation is limited. Patterson et al. (1997) report quantum yields for wild-type, S65T, aGFP, EGFP, and EBFP of 0.79, 0.64, 0.79, 0.60, and 0.17, respectively. Methodology is clearly described. In general, it would appear that enhanced “brightness” of GFP mutants cannot be attributed to improved fluorescence quantum yields. A value of 0.8 is very close to the theoretical limit of 1.0. Furthermore, all reported fluorescence quantum yields of so-called “brighter” mutants are, in fact, lower than values reported for the native GFPs of Renilla and Aequorea. If “brighter” variants exist, it must be for other reasons.

3.5.4

Molar Extinction Coefficient

Accurate measurement of the molar extinction coefficient of GFP requires the following: (a) a very highly purified and correctly folded protein (e.g., >95%), with greater than 95% conversion of the chromogenic tripeptide (-Ser65 Tyr Gly-, or equivalent) into the mature fluorescent chromophore; (b) a reliable measurement of total protein concentration that is insensitive to changes in amino acid composition, (c) a measurement of absorbance at the wavelength of interest; and (d) a precise evaluation of the absorbance at that wavelength as a function of external conditions (pH, ionic strength, temperature, buffer composition, organic solvent strength, presence of buffer additives, etc.) as well as a function of protein concentration (e.g., is Beer’s law obeyed?). If it is assumed that an improperly folded protein (or one lacking the mature chromophore) can be separated from a properly folded, mature GFP by a combination of traditional methods of high-resolution protein purification (gel filtration, ion exchange, and hydrophobic interaction chromatography plus native gel electrophoresis and isoelectric focusing), then it should be possible to demonstrate empirically that a preparation of GFP is pure and properly formed. Furthermore, a GFP molecule with an improperly formed chromophore, one that has failed to cyclize, should be readily apparent by high-resolution mass spectrometry and, due to its different pI from that of mature GFP, should be separable from it by ion exchange chromatography or isoelectric focusing. In our many years

55

56

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

of purifying and characterizing native and recombinant GFPs by traditional multiphase chromatographic and electrophoretic means, we have never seen any evidence of improperly folded or improperly cyclized GFP in our final product. If, however, one were to purify a recombinant GFP molecule containing a polyhistidine N- or C-terminal tail by immobilized metal affinity chromatography (IMAC) on a nickel column, for example, at the exclusion of other high-resolution methods (Inouye and Tsuji, 1994), it would be impossible to separate improperly folded or improperly cyclized GFP from properly folded GFP. Furthermore, IMAC purification of GFP having an N-terminal polyhistidine tag is likely to produce a mixed population of incomplete GFP molecules including dead-end products of translation, many of which would not be expected to fluoresce. Spectroscopic measurements to determine molar extinction coefficients of such a product could, in fact, be meaningless. The results of other attempts to physically characterize IMAC purified GFP could be equally meaningless. Each of the four required conditions listed above (a–d) appears to have been met prior to the establishment of the molar extinction coefficient for R. reniformis GFP at its absorption wavelength maximum of 498 nm (Ward and Cormier, 1979; Ward et al., 1980; Ward, 1981). The value for the molar extinction coefficient of Renilla GFP is 133,000 liter mol-1 cm-1 for the monomer and 266,000 liter mol-1 cm-1 for the dimer (Tables 3.1 and 3.4). We also established the molar extinction coefficient for fully denatured R. reniformis GFP under strongly alkaline conditions (0.1 M NaOH). Under these conditions (pH 13) Renilla GFP loses all of its fluorescence in 1–3 min at room temperature and the absorption spectrum shifts to a single broad band centered at 447 nm (Fig. 3.5, curve f). The molar extinction coefficient for this “denatured chromophore” is 44,000 liter mol-1 cm-1 (Ward et al., 1980; Ward, 1981). Then, because the absorption spectrum of denatured A. victoria GFP at pH 13 (0.1 M NaOH) is identical to that of denatured R. reniformis GFP at pH 13 (Ward et al., 1980), we were able to back calculate the extinction coefficient of native Aequorea GFP at 395 nm (in 1 mM sodium phosphate buffer at pH 7.2) and arrive at a value of 27,600 liter mol-1 cm-1 (Ward and Bokman, 1982), a value that varies substantially with pH, temperature, ionic strength, and total GFP concentration (Ward, 1981). This comparative method, based upon the common absorption characteristics of denatured GFP, is suitable for calculating extinction coefficients (traceable back to the original measurements with Renilla GFP) for all GFPs, natural or recombinant, so long as the chromophore contains a dehydrotyrosine residue conjugated to the imidazolone group. Thus, this method works for the S65T mutant and for Mut 1 (Cormack et al., 1996; Yang, et al., 1996), but not for Y66H, Y66W, or Y66F. Table 3.4 summarizes the molar extinction coefficients we have determined. Values based on measurements of the “base-denatured chromophore” are in excellent agreement with those determined independently by measurements of known tyrosine and tryptophan content in the respective GFP variants. They also agree closely with those reported recently by Patterson et al. (1997). The spectrum of denatured GFP at pH 13 is stable for 30–60 min at room temperature, but at higher temperature 40°C or after longer incubation time, the peak at 447 nm degrades irreversibly to a new spectral form peaking near 335 nm (Ward, 1981). When using this method for determining relative extinction coefficients, it is essential to follow the absorption spectrum of the denatured GFP over time to insure total denaturation without chromophore degradation. Determination of the molar extinction coefficients of Y66H and related Tyr66 variants requires alternate methods for measuring protein concentration. Summation of tryptophan and tyrosine absorbencies (Table 3.4, footnote d) or use of the BCA protein assay method (Pierce Chem. Co.) are suitable alternatives.

ABSORPTION/EXCITATION CROSS-SECTION

3.6 ABSORPTION/EXCITATION CROSS-SECTION The true fluorescence efficiency of GFP is the product of its absorption cross section and its quantum yield. The absorption/excitation cross-section of a molecule refers to the number of photons absorbed divided by the number of photons incident (Jagger, 1977). Thus, the more closely the exciting light (spectrum of the exciting light X spectral transmission of the exciting filter) matches the absorption spectrum of GFP, the higher the fluorescence efficiency. For a GFP molecule with a broad absorption spectrum but relatively low light extinction across the spectrum (e.g., wild-type recombinant GFP derived from Aequorea gfp10), it may be necessary to employ broad band exciting light (e.g., 350– 490 nm) of moderately high intensity to generate sufficient fluorescence. GFP molecules with higher extinction coefficients (e.g., S65T, Mut 1, and other so-called “bright” mutants) may give satisfactory results with exciting lights that are narrower in spectral output (e.g., argon lasers with line spectrum emission at 488 nm). Fluorescence intensity will be proportional to the product of the GFP quantum yield and absorption cross section, as described above, so long as the light intensity is below saturation levels (light levels that are so intense as to create a photostationary state where all molecules of GFP are always excited). When light intensity is high enough to maintain a photostationary state, as may occur in some cases in fluorescence microscopy, chromophores with very high extinction coefficients offer no particular advantage over those with lower extinction coefficients; for example, every chromophore is excited at all times. Under such conditions, all that matters is the number of chromophores in the illuminated field and the fluorescence quantum yield and relative photostability of the fluorophore.

3.6.1

Monomer/Dimer Equilibrium

With one exception, all GFP molecules that we have studied are stable nondissociable dimers in dilute aqueous solution (Fig. 3.3). Relative to six other standard proteins of known molecular weight, they migrate as dimers (2 ¥ 27 kDa) on a Phenomenex SEC S-2000 size exclusion HPLC column in an aqueous pH 6.5 buffer solution comprised of 50 mM sodium phosphate and 100 mM NaCl. The chromatographic positions, thus the apparent molecular weights, are unaffected by protein dilution, even at levels below 1 mg GFP/ml. The GFPs surveyed include those from R. reniformis (Ward and Cormier, 1979) R. mulleri, R. kollikeri, Phialidium gregarium, and Halistaura (Mitrocoma) cellularia (Fig. 3.3) (W. Ward, unpublished results). The only exception, so far, is the GFP from A. victoria. At moderate protein concentrations (<1 mg/ml), Aequorea GFP migrates as a 26– 29 kDa monomer under the chromatographic conditions described above (Fig. 3.3). So do all of the recombinant forms of Aequorea GFP. In this regard, we have studied wild-type GFP (Chalfie et al., 1994), the DNA shuffled cycle 3 mutant (Crameri et al., 1996), S65T (Cubitt et al., 1995), Y66H (Cubitt et al., 1995), and Mut 1 (Cormack et al., 1996; Yang et al., 1996b). All GFP variants, whether native or recombinant, have monomer molecular weights of 27 kDa as determined by SDS gel electrophoresis (Fig. 3.2). Although they are monomers in dilute solution, native A. victoria GFP and all recombinant forms of GFP derived therefrom, including wild-type, S65T, Mut 1, cycle 3 variant, and Y66H, will form dimers at high-protein concentration. The concentration required to demonstrate dimer formation exceeds the loading capacity of the Phenomenex SEC S2000 column described above. But dimer can be demonstrated clearly on a Bio-Rad P-

57

58

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

100 BioGel column (fine particle size, 1 ¥ 47 cm). This column is routinely run in our lab in 10 mM Tris–EDTA at pH 8.0 with 1 M (NH4)2SO4 and 0.02% NaN3. The high concentration of ammonium sulfate favors dimer formation by promoting hydrophobic interactions between GFP monomers. Gravity-generated flow rates of 2 ml/h are maintained and 200–500-ml samples of GFP (up to 100 mg ml-1) are loaded. This protein concentrationdependent shift in apparent molecular weight of wild-type recombinant GFP is shown in Fig. 3.8. As we have seen strong evidence in our laboratory (Cutler, 1995; Cutler and Ward, 1997) for hydrophobically driven dimer formation in native Aequorea GFP, in wild-type GFP, and in recombinant variants of GFP, we would expect to see a crystallographic unit cell that is also dimeric. Thus, when GFP crystals are grown in high-salt media, the unit cell is, in fact, a dimer (Perozzo et al., 1988; M. A. Perozzo, personal communication). But when the medium contains 22–26% polyethylene glycol, a solvent that does not favor hydrophobic interaction, monomers form the unit cell of the crystal (Ormö et al., 1996). If a nonpolar solvent (58% 2-methyl-2,4,-pentanediol) is used to promote crystallization of GFP, the unit cell is a dimer, but in this case most of the dimer interface contacts are polar (Yang et al., 1996a, 1997). The dimer interface appears, by computer-generated molecular modeling of the Yang et al. (1996a) coordinates, to contain both hydrophobic

Figure 3.8. Shift in apparent molecular weight of wild-type recombinant Aequorea GFP as a function of GFP concentration. The GFP samples are chromatographed at 24 ± 2°C on a BioRad P-100 BioGel column (225-ml total volume) equilibrated in 10 mM Tris–EDTA buffer (pH 8.0) containing 1 M (NH4)2SO4 and 0.02% NaN3. The concentrations of GFP samples loaded onto the column are 12 mg/ml (A), 4.2 mg/ml (B), 0.84 mg/mL (C), and 0.17 mg/mL (D). The elution behavior is consistent with there being nearly all dimer in sample (A), nearly all monomer in sample (D), and intermediate forms (in equilibrium) in samples (B and C).

A MOLECULAR MECHANISM FOR SUPPRESSION OF CHROMOPHORE ABSORPTION

and hydrophilic contacts. A slight rotation around the long axis of the “beta can” can cause a switch from predominantly hydrophilic contacts at the dimer interface to predominantly hydrophobic contacts (M. A. Perozzo, personal communication). If such a rotation occurs as a function of crystallization solvent, it has little effect on the overall crystal structure, but it might be expected to affect the absorption spectrum of the crystalline protein. If wild-type recombinant GFP is being used as a reporter of gene expression in living cells, it is likely to be of sufficient intracellular concentration (>5 mg ml-1) as to be nearly all in the dimer form. In the photocytes of the jellyfish A. victoria, for example, GFP intracellular concentration (assuming uniform distribution of the protein throughout the cytoplasm) has been estimated to be 25 mg ml-1 (Cutler and Ward, 1997). Dimerization would not be of concern to those using GFP as a reporter if it were not for the fact that wild-type recombinant GFP and related mutants like the cycle 3 variant undergo large changes in absorption (excitation) spectra upon dimerization. The extinction coefficient of the peak at 395 nm increases about 15%, but the extinction coefficient of the shoulder near 475 nm (where most microscopists prefer to excite GFP) drops fourfold to fivefold (Fig. 3.5) from about 12,000 liter mol-1 cm-1 to less than 3000 liter mol-1 cm-1 (Ward et al., 1982). Thus, in the GFP concentration range from 0.2 to 10 mg ml-1, the higher the concentration of intracellular GFP, the more poorly each chromophore absorbs and is excited by blue light. The Ser65 mutants of GFP (e.g., S65T and Mut 1) do not show a concentrationdependent suppression of the blue absorption peak near 489 nm. They retain their relatively high extinction coefficients of 55,000–58,000 liter mol-1 cm-1 at 489 nm (Table 3.4), even when they dimerize at high protein concentration. Retention of a concentrationindependent molar extinction coefficient is a distinct advantage in quantitating intracellular fluorescence signals. However, the Ser65 mutants are more sensitive to pH differences than wild-type GFP and show a steep pH-dependent loss of fluorescence below pH 7.0 (González et al., 1997; Patterson et al., 1997; A. S. Sawyer, D. González, and W. Ward, unpublished results). This loss of fluorescence intensity below pH 7.0 for S65T and Mut 1 may present a problem with quantitation and calibration of fluorescence signals within cells if the cytoplasmic or organellar pH is lower than 7.0 and/or variable.

3.7 A MOLECULAR MECHANISM FOR SUPPRESSION OF CHROMOPHORE ABSORPTION For more than two decades, we have been aware that the 475-nm shoulder in the absorption spectrum of native A. victoria GFP is strongly suppressed at high-protein concentrations (Morise et al., 1974). As the protein concentration is raised from 0.112 to 18.6 mg ml-1, the 475-nm absorption shoulder is suppressed fourfold (Fig. 3.5) while the main peak at 395 nm increases by 15% (Ward et al., 1982). At even higher protein concentrations, and particularly in the presence of antichaotropic salts (e.g., ammonium sulfate), the shoulder suppression reaches fivefold (Cutler, 1995; Cutler and Ward, 1997) such that the molar extinction coefficient at 475 nm drops below 3000 liter mol-1 cm-1. The same behavior is seen with the wild-type recombinant GFP and the cycle 3 mutant (González et al., 1997; Patterson et al., 1997; W. Ward, unpublished results). It is also clear from calibrated gel filtration chromatography (Cutler, 1995; Cutler and Ward, 1997) that native Aequorea GFP has a strong tendency to form dimers at high-protein concentration and that ammonium sulfate (1 M) promotes dimerization at substantially lower GFP concentrations. High concentrations of antichaotropic salts, such as ammonium sulfate, are known to favor hydrophobic interactions, so we have proposed (Cutler, 1995) that GFP monomers inter-

59

60

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

act at hydrophobic interfaces when they form dimers in aqueous solution, especially in the presence of antichaotropic salts. In organic solvents, such as the 58% 2-methyl-2,4pentane-diol used to induce crystal formation with wild-type recombinant GFP (Yang et al., 1996a, 1997), hydrophobic interactions between monomers would be inhibited by the organic solvent. In fact, Yang et al. (1996a) observe more hydrophilic contacts than hydrophobic contacts between GFP monomers in their crystals. This result is not at all surprising considering the crystallization solvent. However, just beyond these hydrophilic interfaces, on the surface of the GFP cylinder, there exists a region of strong surface hydrophobicity dominated by residues such as Leu221, Phe223, Pro211, and Tyr39 (Perozzo, personal communication). Amino acid residues 221 and 223 project into the hydrophobic interface while the bridge amino acid residue, Glu222, projects toward the interior of the beta can. Glu222 is hydrogen bonded to the hydroxyl residue of Ser65 in wild-type recombinant GFP (Yang et al., 1996a). Thus an unbroken chain exists, connecting surface hydrophobic residues to the chromophore via Glu222. This chain appears to “regulate” the distance between the keto oxygen of the chromophore imidazolone and the electron withdrawing groups Arg96 and Glu94 that are positioned above the imidazolone ring of the chromophore. A retracted chain (resulting from hydrophobic contacts between monomers at residues 221 and 223) “pulls” on the chromophore, separating the keto oxygen from the electron-withdrawing residues Arg96 and Glu94. An expanded chain (resulting from the relaxation of hydrophobic contacts at the dimer interface or direct elongation of the “chain”) allows the chromophore to slip closer to Arg96 and Glu94, shortening the distance between the keto oxygen and residues 96 and 94. As the wild-type GFP dimer forms in aqueous solution (GFP dimerization is strongly favored in high ionic strength medium such as the cytoplasm of prokaryotic and eukaryotic cells where recombinant GFPs are being expressed], corresponding hydrophobic interfaces collapse upon each other. In this process, leucine221 and Phe223 are drawn into the interface and farther from the surface of the protein. This interfacial collapse exerts a “pull” on the chromophore, via Glu222, decreasing the electron-withdrawing influence of Arg96 and Glu94. Removed from the influence of these residues, the chromophore assumes the tautomeric form shown in Fig. 3.6a. This tautomeric form absorbs light at a wavelength maximum of 395 nm. The alternate tautomeric form, shown in Fig. 3.6b is suppressed and so its corresponding spectral form (lmax = 475 nm) is also suppressed. In the case of S65T and Mut 1, the orientation of the 2° hydroxyl group (serine to threonine substitution) at position 65 may, in effect, lengthen the chain connecting the hydrophobic interface with the chromophore. This lengthening allows the chromophore to remain close to residues 96 and 94 at all times, favoring the quinone–enol form of the chromophore shown in Fig. 3.6b. As a consequence, S65T and Mut 1 display red-shifted absorption spectra (lmax = 489 nm), which do not shift further upon protein dimerization. In the case of Renilla GFP, which is always dimeric, the interfacial contact between dimers may position the chromophore very close to electron-withdrawing groups (amino acid residues in Renilla GFP analogous to Arg96 and Glu94 in Aequorea GFP) at all times. Thus, the absorption peak is shifted to 498 nm and the protein fails to respond spectrally to a wide array of external perturbants. Water-miscible organic solvents are capable of inducing large absorption/excitation spectral shifts (toward the red) in Aequorea GFP (Robart and Ward, 1990) and related recombinant forms (wild-type recombinant and the cycle 3 variant) (González et al., 1997). Such spectral shifts are observed by absorption spectroscopy at moderate protein concentration (0.1–1.0 mg ml-1) and by fluorescence spectroscopy at much lower protein con-

FUTURE PROSPECTS

centration (0.1–1.0 mg ml-1), where Aequorea GFP is entirely monomeric, suggesting that these organic solvents must be affecting the monomer directly. No solvent-dependent shift in monomer–dimer equilibrium is involved. If the organic solvent interacts with surface residues Leu221 and Phe223, as might be expected, it may force these residues deeper into the protein interior. A retraction of these residues, communicated to the chromophore via Glu222, could allow the quinone–enol form of the chromophore to slip into the electron-withdrawing pocket formed by Arg96 and Glu94. Consistent with this model is the fact that all GFP forms with bimodal absorption/excitation spectra in the UV and visible regions of the spectrum (native, wild-type, and cycle 3 variants) undergo these large solvent-induced spectral shifts. Those that are already red-shifted (e.g., Renilla GFP, S65T, and Mut 1) show no further spectral shifts upon solvent addition.

3.8 FUTURE PROSPECTS It is a daunting process to predict future directions of research, especially when one is so close to the field. No one, for example, within the bioluminescence community (even as late as 1990) predicted that GFP chromophore formation would be autocatalytic or that cDNA from A. victoria could ever be expressed in heterologous systems as a functional fluorescent protein. No one envisioned GFPs applications in developmental biology, in protein trafficking, in gene reporting, or in high throughput screening in cell-based assays. The GFP was just that other interesting protein in Aequorea that a handful of “aequorin researchers” stockpiled in their freezers and occasionally thawed out to study its magnificent physical properties. When I started working in Milton Cormier’s lab in 1973, GFP was almost unknown, except within a small circle of bioluminescence researchers. None of us ever expected to see GFP on the covers of Science, Nature Biotechnology, Bioluminescence and Chemiluminescence, Biotechniques, the CLONTECH catalog, or BioRad’s “Explorer” series of GFP-based instructional kits for high schools. So, having established my record as a prognosticator, here are my predictions. I believe that the next research frontier for GFP will be in the area of cell-based diagnostics and high throughput homogeneous screening of drugs, organic chemicals, toxicants, food additives, herbicides, pesticides, mutagens, carcinogens, and teratogens. Customdesigned cell lines in which GFP expression is controlled by specific promoters will be used to screen these sorts of compounds minutes after their robotic application to multiwell plates. The carefully designed specificity of each promoter-GFP response will establish directly and immediately the biochemical mode of action of the substance being tested. The information generated in such biochemically and pharmacologically specific screening assays will be of such high quality as to eliminate the need for many costly levels of secondary testing. The search for Aequorea GFP variants with higher extinction coefficients, more rapid “greening” rates, and red-shifted excitation and emission spectra will continue to be fruitful, but the focus of the search will shift to other species of coelenterates that already display spectral biodiversity nearly as great as the range of Aequorea GFP variants. Incremental improvements in GFP will further stimulate development of instrumentation that can better handle the relatively high noise level of fluorometric screening methods. I see GFP as one of several coelenterate bioluminescence reporters that will be used together, to an increasing extent, in cell-based diagnostics. Renilla luciferase and aequorin are already cloned and available for nonhomogeneous assays requiring luciferin (coelen-

61

62

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

terazine) addition. The need to add coelenterazine as an additional reagent, however, places the coelenterate system into the same category as other luciferin–luciferase diagnostic assays—only the luciferase has been cloned. But, there exists the hope that coelenterazine may soon be cloned as well, enabling Renilla luciferase, aequorin, and coelenterazine to be used together in a variety of homogeneous diagnostic assays. Development of rapid, sensitive, and homogeneous screening methods based on the coelenterate bioluminescence system could dominate this field over the next decade. In the field of education, GFP will soon become the tool of choice to illustrate all aspects of biotechnology in the classroom laboratory from cell transformation to immobilized metal ion affinity chromatography. GFP based educational modules will be created for all educational levels, primary through postgraduate, “bringing to light” such diverse fields as general biology, environmental testing, microbiology, molecular biology, toxicology, biochemistry, and molecular medicine. The best prediction for the future of GFP was written 2 years ago by Susan Hassler whose editorial in Bio/Technology (Hassler, 1995) says it all, “Green Fluorescent Protein: The Next Generation.”

ACKNOWLEDGMENTS This work was sponsored, in part, by a grant from the National Science Foundation– Advanced Technological Education (DUE-9602356) and research contracts from BioRad Corporation and Clontech Laboratories. The manuscript was critically reviewed by Daniel González. Figures and tables were prepared by Peter Anderson and Dr. Frank Petersen. Much of the previously unpublished work has been contributed by Dr. Mark Cutler, Amy Roth, Daniel González, Anita Sawyer, Tom Spires, and Elmer Castriciones.

REFERENCES Bokman, S. H., and Ward, W. W. (1981). Renaturation of Aequorea green-fluorescent protein. Biochem. Biophys. Res. Commun. 101:1372–1380. Chalfie, M. (1995). Green fluorescent protein. Photochem. Photobiol. 62:651–656. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green-fluorescent protein as a marker for gene expression. Science 263:802–805. Cody, C. W., Prasher, D.C., Westler, W. M., Prendergast, F. G., and Ward, W. W. (1993). Chemical structure of the hexapeptide chromophore of the Aequorea green-fluorescent protein. Biochemistry 32:1212–1218. Cormack, B., Valdivia, R., and Falkow, S. (1996). FACS-optimized mutants of the green fluorescent protein (GFP). Gene 173:33–38. Cormier, M. J., Hori, K., Karkhanis, Y. D., Anderson, J. M., Wampler, J. E., Morin, J. G., and Hastings, J. W. (1973). Evidence for similar biochemical requirements for bioluminescence among the coelenterates. J. Cell. Physiol. 81:291–297. Cormier, M. J., Hori, K., and Anderson, J. M. (1974). Bioluminescence in coelenterates. Biochim. Biophys. Acta. 346:137–164. Crameri, A., Whitehorn, E. A., Tate, E., and Stemmer, W. P. C. (1996). Improved green fluorescent protein by molecular evolution using DNA shuffling. Nat. Biotechnol. 14:315–319.

REFERENCES

Cubitt, A. B., Heim, R., Adams, S. R., Boyd, A. E., Gross, L. A., and Tsien, R. Y. (1995). Understanding, improving and using green fluorescent proteins. TIBS 20:448–455. Cutler, M. W. (1995). Characterization and energy transfer mechanism of the green-fluorescent protein from Aequorea. victoria. Ph.D. Thesis, Rutgers University, New Brunswick, NJ. Cutler, M. W., and Ward, W. W. (1993). Protein–protein interactions in Aequorea bioluminescence. Bioluminescence Symposium, Maui, HI, Nov. 5–10. Cutler, M. W., and Ward, W. W. (1997). Spectral analysis and proposed model for GFP dimerization. In Bioluminescence and Chemiluminescence; Molecular Reporting with Photons, Hastings, J. W., Kricka, L. J., and Stanley, P. E., Eds., Wiley, New York, pp. 403–406. Davis, D. F., Ward, W. W., and Cutler, M. W. (1994). Post-translational chromophore formation in recombinant GFP from E. coli requires oxygen. In Bioluminescence and Chemiluminescence, Fundamentals and Applied Aspects, Campbell, A. K., Kricka, L. J., and Stanley, P. E., Eds., Wiley, New York, pp. 596–599. Delagrave, S., Hawtin, R. E., Silva, C. M., Yang, M. M., and Youvan, D. C. (1995). Red-shifted excitation mutants of the green fluorescent protein. Bio/Technol. 13:151–154. Ehrig, T., O’Kane, D. J., and Prendergast, F. G. (1995). Green-fluorescent protein mutants with altered fluorescence excitation spectra. FEBS. Lett. 367:163–166. González, D., Sawyer, A., and Ward, W. W. (1997). Spectral perturbations of mutants of recombinant Aequorea victoria green-fluorescent protein (GFP). Photochem. Photobiol. 65:21S. Hart, R. C., Matthews, J. C., Hori, K., and Cormier, M. J. (1979). Renilla reniformis bioluminescence: Luciferase-catalyzed production of nonradiating excited states from luciferin analogues and elucidation of the excited state species involved in energy transfer to Renilla green fluorescent protein. Biochemistry 18:2204–2210. Hassler, S. (1995). Green fluorescent protein: The next generation. Bio/Technol. 13:103. Heim, R., Cubitt, A. B., and Tsien, R. Y. (1995). Improved green fluorescence. Nature. (London) 373:663–664. Heim, R., Prasher, D. C., and Tsien, R. Y. (1994). Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proc. Natl. Acad. Sci. USA 91:12501–12504. Heim, R., and Tsien, R. Y. (1996). Engineering green fluorescent protein for improved brightness, longer wavelengths, and fluorescence resonance energy transfer. Curr. Biol. 6:178–182. Hori, K., and Cormier, M. J. (1973). Structure and chemical synthesis of a biologically active form of Renilla (sea pansy) luciferin. Proc. Natl. Acad. Sci. USA 70:120–123. Hori, K., Wampler, J. E., Matthews, J. C., and Cormier, M. J. (1973). Identification of the product excited states during the chemiluminescent and bioluminescent oxidation of Renilla (sea pansy) luciferin and certain of its analogs. Biochemistry 12:4463–4469. Hori, K., Anderson, J. M., Ward, W. W., and Cormier, M. J. (1975). Renilla luciferin as the substrate for calcium induced photoprotein bioluminescence. Assignment of luciferin tautomers in aequorin and mnemiopsin. Biochemistry 14:2371–2376. Hori, K., Charbonneau, H., Hart, R. C., and Cormier, M. J. (1977). Structure of native Renilla reniformis luciferin. Proc. Natl. Acad. Sci. USA 74:4285–4287. Inoue, S., Kakoi, H., Murata, M., Toto, T., and Shimomura, O. (1977a). Complete structure of Renilla luciferin and luciferyl sulfate. Tetrahedron Lett. 31:2685–2689. Inoue, S., Okada, K., Kakoi, H., and Goto. T. (1977b). Fish bioluminescence I. Isolation of a luminescent substance from a myctophina fish, Neoscopelus microchir, and identification of it as Oplophorus luciferin. Chem. Lett. 1977:257–258. Inouye, S., and Tsuji, F. I. (1994). Aequorea green fluorescent protein: expression of the gene and fluorescence characteristics of the recombinant protein. FEBS Lett. 341:277–280. Jagger, J. (1977). Phototechnology and biological experimentation. In The Science of Photobiology, Smith, K. C., Ed., Plenum, New York, pp. 1–26.

63

64

BIOCHEMICAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN

Kain, S. R., Adams, M., Kondepudi, A., Yang, T.-T. Ward, W. W., and Kitts, P. (1995). Green fluorescent protein as a reporter of gene expression and protein localization. BioTechniques 19:650–655. Kolb, V. A., Makeyev, E. V., Ward, W. W., and Spirin, A. S. (1996). Synthesis and maturation of green fluorescent protein in a cell-free translation system. Biotech. Lett. 18:1447–1452. Kurian, E., Fisher, P. J., Ward, W. W., and Prendergast, F. G. (1994). Characterization of secondary and tertiary structure of the green fluorescent protein from A. victoria. J. Biolum. Chemilum. 9:333. Levine, L. D., and Ward, W. W. (1982). Isolation and characterization of a photoprotein, “phialidin”, and a spectrally unique green-fluorescent protein from the bioluminescent jellyfish Phialidium. gregarium. Comp. Biochem. Physiol. 72B:77–85. Morin, J. G. (1974). Coelenterate bioluminescence. In Coelenterate Biology: Reviews and New Perspectives, Muscatine, L., and Lenhoff, H., Eds., Academic, New York, pp. 397–438. Morin, J. G., and Hastings., J. W. (1971). Energy transfer in a bioluminescent system. J. Cell. Physiol. 77:313–318. Morise, H., Shimomura, O., Johnson, F. H., and Winant, J. (1974). Intermolecular energy transfer in the bioluminescent system of Aequorea. Biochemistry 13:2656–2662. Ormö, M., Cubitt, A. B., Kallio, K., Gross, L. A., Tsien, R. Y., and Remington, S. J. (1996). Crystal structure of the Aequorea victoria green fluorescent protein. Science 273:1392–1395. Patterson, G. H., Knobel, S. M., Sharif, W. D., Kain, S. R., and Piston, D. W. (1997). Use of the green fluorescent protein and its mutants in quantitative fluorescence microscopy. Biophys. J. 73:2782–2790. Perozzo, M. A., Ward, K. B., Thompson, R. B., and Ward, W. W. (1988). X-ray diffraction and timeresolved fluorescence analyses of Aequorea green fluorescent protein crystals. J. Biol. Chem. 263:7713–7716. Prasher, D. C. (1995). Using GFP to see the light. Trends Genet. 11:320–323. Prasher, D. C., Eckenrode, V. K., Ward, W. W., Prendergast, F. G., and Cormier, M. J. (1992). Primary structure of the Aequorea victoria green-fluorescent protein. Gene 111:229–233. Prendergast, F. G., and Mann, K. G. (1978). Chemical and physical properties of aequorin and the green-fluorescent protein isolated from Aequorea. forskalea. Biochemistry 17:3448–3453. Rao, B., Kemple, M., and Prendergast, F. (1980). Proton nuclear magnetic resonance and fluorescence spectroscopic studies of segmental mobility in aequorin and a green fluorescent protein from Aequorea. forskalea. Biophys. J. 32:630–632. Robart, F. D., and Ward, W. W. (1990). Solvent perturbations of Aequorea green-fluorescent protein. Photochem. Photobiol. 51:92s. Roth, A. (1985). Purification and protease susceptibility of the green-fluorescent protein of Aequorea aequorea with a note on Halistaura. M.S. Thesis. Rutgers University, New Brunswick, NJ. Roth, A. F., and Ward, W. W. (1983). Conformational stability after protease treatment in Aequorea GFP. Photochem. Photobiol. 37S:S71. SanPietro, R. M., Prendergast, F. G., and Ward, W. W. (1993). Sequence of the chromogenic hexapeptide of Renilla green-fluorescent protein. Photochem. Photobiol. 57:63s. Shimomura, O. (1979). Structure of the chromophore of Aequorea green fluorescent protein. FEBS Lett. 104:220–222. Shimomura, O., and Johnson, F. H. (1979). Comparison of the amounts of key components in the bioluminescence systems of various coelenterates. Comp. Biochem. Physiol. 64B:105–107. Shimomura, O., Inoue, S., Johnson, F. H., and Haneda, Y. (1980). Widespread occurrence of coelenterazine in marine bioluminescence. Comp. Biochem. Physiol. 65B:435–437. Stearns, T. (1995). Green fluorescent protein. The green revolution. Curr. Biol. 5:262–264.

REFERENCES

Surpin, M. A., and Ward, W. W. (1989). Reversible denaturation of Aequorea green-fluorescent protein–thiol requirement. Photochem. Photobiol. 49:62S. Wampler, J. E., Hori, K., Lee, J., and Cormier, M. J. (1971). Structured bioluminescence. Two emitters during both the in vitro and the in vivo bioluminescence of Renilla. Biochemistry 10: 2903–2910. Wampler, J. E., Karkhanis, Y. D., Morin, J. G., and Cormier, M. J. (1973). Similarities in the bioluminescence from the Pennantulacea. Biochim. Biophys. Acta 314:104–109. Ward, W. W. (1979). Energy transfer processes in bioluminescence. In Photochemical and Photobiological Reviews, Vol. 4, Smith, K., Ed., Plenum, New York, pp. 1–57. Ward, W. W. (1981). Properties of the coelenterate green-fluorescent proteins. In Bioluminescence and Chemiluminescence: Basic Chemistry and Analytical Applications, DeLuca, M., and McElroy, D, W., Eds., Academic, New York, pp. 235–242. Ward, W. W., and Bokman, S. H. (1982). Reversible denaturation of Aequorea green-fluorescent protein: Physical separation and characterization of the renatured protein. Biochemistry 21: 4535–4540. Ward, W. W., Cody, C., Hart, R. C., and Cormier, M. J. (1980). Spectrophotometric identity of the energy transfer chromophores in Renilla and Aequorea green-fluorescent proteins. Photochem. Photobiol. 31:611–615. Ward, W. W., and Cormier, M. J. (1975). Extraction of Renilla-type luciferin from the calciumactivated photoproteins aequorin, mnemiopsin and berovin. Proc. Natl. Acad. Sci. 72:2530–2534. Ward, W. W., and Cormier, M. J. (1976). In vitro energy transfer in Renilla bioluminescence. Michael Kasha Symposium—Electronic Processes and Energy Transfer in Organic, Inorganic and Biological Systems (special edition). J. Phys. Chem. 80:2289–2291. Ward, W. W., and Cormier, M. J. (1978a). Energy transfer via protein–protein interaction in Renilla bioluminescence. Photochem. Photobiol. 27:389–396. Ward, W. W., and Cormier, M. J. (1978b). Protein–protein interactions as measured by bioluminescence energy transfer. Methods Enzymol. 62:257–267. Ward, W. W., and Cormier, M. J. (1979). An energy transfer protein in coelenterate bioluminescence: Characterization of the Renilla green-fluorescent protein (GFP). J. Biol. Chem. 254:781–788. Ward, W. W., Prentice, H. J., Roth, A. F., Cody, C. W., and Reeves, S. C. (1982). Spectral perturbations of the Aequorea green-fluorescent protein. Photochem. Photobiol. 35:803–808. Wu, C.-K., Liu, Z.-J., Rose, J. P., Inouye, S., Tsuji, F., Tsien, R. Y., Remington, S. J., and Wang, B.-C. (1997). The three-dimensional structure of green fluorescent protein resembles a lantern. In Bioluminescence and Chemiluminescence, Molecular Reporting With Photons, Hastings, J. W., Kricka, L. J., and Stanley, P. E. Eds., Wiley, New York, pp. 399–402. Yang, F., Moss, L. G., and Phillips, G. N., Jr. (1996a). The molecular structure of green fluorescent protein. Nature Biotech. 14:1246–1251. Yang, F., Moss, L. G., and Phillips, G. N., Jr. (1997). The three-dimensional structure of green fluorescent protein. In Bioluminescence and Chemiluminescence, Molecular Reporting With Photons, Hastings, J. W., Kricka, L. J., and Stanley, P. E., Eds., Wiley, New York, pp. 375–382. Yang, T. T., Cheng, L., and Kain, S. R. (1996b). Optimized codon usage and chromophore mutations provide enhanced sensitivity with the green fluorescent protein. Nucleic Acids Res. 24(22):4592–4593. Youvan, D. C., and Michel-Beyerle, M. E. (1996). Structure and fluorescence mechanism of GFP. Nat. Biotechnol. 14:1219–1220.

65

4 THE THREE-DIMENSIONAL STRUCTURE OF GREEN FLUORESCENT PROTEIN AND ITS IMPLICATIONS FOR FUNCTION AND DESIGN George N. Phillips, Jr. Department of Biochemistry, University of Wisconsin—Madison, Madison, WI

4.1 INTRODUCTION Green fluorescent protein (GFP) is a naturally fluorescent protein first isolated from the jellyfish Aequorea victoria (Shimomura et al., 1962). It converts the blue chemiluminescence of other proteins, aequorin or luciferase, into green fluorescent light (Morin and Hastings, 1971; Ward, 1979), presumably to reduce scattering and, hence, improve penetration of the light over longer distances. Since its initial discovery in jellyfish, this protein fold has been found to occur broadly in sea creatures, forming colored, if not fluorescent, pigments (Matz et al., 2002), and the same protein fold is also found in mammals in the form of an extracellular basement membrane component called nidogen (Hopf et al., 2001). The molecular cloning of GFP cDNA from the Pacific jellyfish, A. victoria (Prasher et al., 1992), and the demonstration by Chalfie et al. (1994) that this GFP can be functionally expressed in bacteria and nematodes have opened exciting new avenues of investigation in cell, developmental, and molecular biology, as pointed out in prior reviews (Chalfie, 1995; Heim and Tsien, 1996) and by the many techniques and applications described in this volume. As a consequence of this interest, knowledge of the threedimensional (3D) structure of GFP and its relatives has become highly desirable for understanding the origins and mechanisms of the diversity of chromophore behaviors and for engineering modified GFP’s for various purposes. Quite a number of researchers are currently carrying out structural studies on GFP, its relatives, and the mutations of these relatives. Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

67

68

THE THREE-DIMENSIONAL STRUCTURE OF GFP AND ITS IMPLICATIONS FOR FUNCTION AND DESIGN

Green fluorescent protein from the jellyfish, A. Victoria, has several isotypes (Cutler, 1995); the cDNA of GFP that was first cloned and expressed, called TU#58 (Chalfie et al., 1994), is comprised of 238 amino acids. As first expressed in recombinant form, the N-terminal methionine was replaced by methionine-alanine, and there was also an inadvertent substitution of arginine for glutamine at position 80, probably due to a polymerase chain reaction (PCR) error in the original cloning. For the purposes of this chapter, I refer to this form as “wild type,” since the key amino acids are identical to the one of the native isotypes, and its spectral properties are like those from the native protein.

4.2 SPECTRAL AND PHYSICAL PROPERTIES OF GREEN FLUORESCENT PROTEIN The wild-type fluorescence absorbance/excitation peak is at 395 nm with a minor peak at 475 nm [molar absorbances of roughly 30,000 and 7000 M-1 cm-1, respectively (Ward, 1979; Ward and Cormier, 1979; Kahana and Silver, 1996)]. The normal emission peak is at 508 nm. Interestingly, continued excitation at wavelengths that excite the major 395-nm peak led to a decrease over time of the 395-nm excitation peak and a reciprocal increase in the 475-nm excitation band (Chalfie et al, 1994; Cubitt et al., 1995). This interconversion effect is especially evident with irradiation of GFP by ultraviolet (UV) light. Femtosecond time-resolved spectroscopic studies have revealed that the two states corresponding to two major absorption bands can interconvert quickly in the excited state (Chattoraj et al., 1996; Lossau et al., 1996). Both Chattoraj et al. (1996) and Lossau et al. (1996) showed that excited-state deprotonation transfer is a key process in GFP photochemistry [see also Youvan and Michel-Beyerle (1996)]. Both groups demonstrated a slowing of kinetic features upon deuteration in the picosecond decay of the excited protonated species and the concomitant rise of the green fluorescence of the deprotonated fluorophore. While deuterium effects are consistent with proton movement, they are not in themselves proof. In the initial work, both groups simply incubated the protein in deuterated buffer; therefore, neither the extent nor the position of deuteration can be determined. However, Lossau et al. (1996) and Brejc et al. (1997) further substantiated a deprotonation mechanism through the comparative spectroscopic study of wild-type GFP containing the Phe64–Leu mutation plus Tyr66–His (blue emitter) or Ser65–Thr (red-shifted excitation), or GFP with only the Ser65–Thr mutant. The Tyr66–His mutant has a fluorophore that does not deprotonate in the excited state, and the Ser65–Thr mutant has fluorophores that are mostly deprotonated in the electronic ground state, respectively. For a more complete description, see Remington (2000). The use of semiempirical quantum mechanical modeling has also been used to study the relative stability; proton affinities, geometrical structures, and absorption spectra of wild-type and several mutant GFPs suggest an alternative site for deprotonation [Voityuk et al. (1997); see also review by Zimmer (2002)]. Agreement between the computed and observed absorption spectra of the fluorophore in aqueous solution at acidic and basic pH suggests that the structure of important states of the fluorophore in the native and mutant proteins can be identified: The absorption maximum of GFP at 477 nm is assigned to a zwitterion where the phenolic oxygen of Tyr66 is deprotonated and the nitrogen of the heterocyclic ring is protonated. Furthermore, the high-energy absorption peak at 397 nm is assigned to the excitation of the protonated form of the nitrogen in the heterocycle. This conclusion is further corroborated by the agreement between the calculated absorption energies at 355, 433, and 387 nm in the Tyr66–Phe, Tyr66–Trp, and Tyr66–His mutants as

THE b-CAN STRUCTURE OF GFP

compared to the experimental values of 360, 436, and 382 nm, respectively. In summary, these computations suggest that the protonated nitrogen constitutes a crucial factor in the function of GFP. Physical and chemical studies of purified GFP also identified several important characteristics that relate directly to its structure. GFP is very resistant to denaturation, requiring treatment with 6 M guanidine hydrochloride at 90°C or pH of less than 4.0 or greater than 12.0. Partial to near total renaturation occurs within minutes following reversal of denaturing conditions by dialysis or neutralization (Ward and Bokman, 1982). Over a nondenaturing range of pH, increasing pH leads to a reduction in fluorescence by 395-nm excitation and an increased absorption at 475 nm (Ward et al., 1982). Because GFP in crystallurm exhibits nearly identical fluorescence spectra and excited-state lifetimes to that for GFP in aqueous solution (Perozzo et al., 1988) and fluorescence is not an inherent property of the isolated fluorophore, the elucidation of its 3D structure of GFP (Ormo et al., 1996; Yang et al., 1996) helped provide an explanation for the generation of fluorescence in the mature protein, as well for as the mechanism of autocatalytic fluorophore formation. Furthermore, the discovery or development of fluorescent proteins with varied emission and excitation or other characteristics based on the predicted changes in the structure has dramatically expanded biological applications of GFP and its variants.

4.3 THE b-CAN STRUCTURE OF GFP The structure of the wild-type protein was solved by Yang et al. (1996), and that of the Ser65Thr mutant was solved by Ormo et al. (1996). The density maps of both determinations of GFP were very clear, revealing quite regular b-barrels with 11 strands on the outside of cylinders (Fig. 4.1). These cylinders have a diameter of about 30 Å and a length of about 40 Å. Inspection of the density within the cylinders revealed the fluorescent center of the molecule, a modified tyrosine side chain and cyclized protein backbone as a part of an irregular a-helical segment. Small sections of a helices and loops also form caps on the ends of the cylinders. This motif, with a single a helix inside a very uniform cylinder of b-sheet structure, represents a new protein class, which we have named the b-can (Yang et al., 1996). The regularity of the b-can of GFP is quite remarkable. The 11 strands of the sheet form an almost seamless symmetrical structure, the only irregularities being between two of the strands. In fact, the structure is so regular that water molecules on the outside of the can also form “stripes” around the surface of the cylinder. A surprising number of water molecules are also found inside the can. The tightly constructed b-barrel would appear to serve the role of protecting the fluorophore well, providing overall stability and resistance to unfolding by heat and denaturants. The previously known proteins that most closely resemble the b-can fold of GFP are porin, which has not 11 but 16 antiparallel strands and has no “lids” at the ends of the barrel (Kreusch et al., 1994), and strepavidin, which is a smaller, eight-stranded antiparallel b-barrel (Fig. 4.2) (Hendrickson et al., 1989). Unlike streptavidin, both GFP and porin have water molecules inside the barrel, as well as small segments of polypeptide chain inside. In the case of porin, whose function is to allow passage of small molecules through its center, its design needs to be open, whereas GFP’s function is better served with a closed structure that can restrict access to the fluorophore and perhaps also contain damaging free radical photoproducts. Because of the smaller number of strands in streptavidin and hence smaller inside diameter, its center consists simply of side chains originating from the staves of the barrel.

69

70

THE THREE-DIMENSIONAL STRUCTURE OF GFP AND ITS IMPLICATIONS FOR FUNCTION AND DESIGN

3

2 5 1 6 4

9

11

8 7

10

10 A Figure 4.1. End-on (top left) and side (top right) views of the cylindrical b-can structure of GFP. Eleven strands of b-sheet form an antiparallel barrel with short helices forming lids on each end. The fluorophore is inside the can, as a part of a distorted a-helix, which runs along the axis of the cylinder. The GFP usually forms dimers in the crystal, aligned largely along the sides of the cylinders. Drawing by Ribbons (Carson, 1987), coordinates (Protein Data Bank entry 1GFL). [From Yang et al. (1996.) Reprinted with permission from Nature Biotechnology.] See color insert.

THE FLUOROPHORE AND ITS ENVIRONMENT

Figure 4.2. Tertiary structures of porin (top left), GFP (bottom), and strepavidin (top right) showing three sizes of antiparallel b-barrel proteins. Porin is open on both ends with a water channel through the middle, GFP has water and protein on the inside, but is sealed on both ends, and streptavidin is too small to have anything in the core except side chains from the strands of the barrel.

4.4 THE FLUOROPHORE AND ITS ENVIRONMENT Analysis of a hexapeptide derived by proteolysis of purified GFP led to the prediction that the fluorophore originates from an internal Ser–Tyr–Gly sequence, which is posttranslationally modified to a 4-(p-hydroxybenzylidene)-imidazolidin-5-one structure (Cody et al.,

71

72

THE THREE-DIMENSIONAL STRUCTURE OF GFP AND ITS IMPLICATIONS FOR FUNCTION AND DESIGN

1993). Studies of recombinant GFP expression in Escherichia coli led to a proposed sequential mechanism initiated by a rapid cyclization between Ser64 and Gly67 to form a imidazolin-5-one intermediate followed by a much slower (hours) rate-limiting oxygenation of the Tyr side chain by O2 (Heim et al., 1994). Extensive combinatorial mutagenesis (Delagrave et al., 1995) and other mutational studies (Heim et al., 1995; Cormack et al., 1996) suggested that the Gly67 is required in the functional fluorophore. While no known cofactors or enzymatic components are required for this apparently autocatalytic process, it is rather thermosensitive with the yield of fluorescently active to total GFP protein decreasing at temperatures greater than 30°C (Lim et al., 1995). However, once the functional form of GFP is produced, it is quite thermostable. The critical fluorophore-forming sequence, Ser–Tyr–Gly, occurs many times in proteins. How is it that cyclization occurs in GFP, but not in these other proteins? Two factors appear to be required for fluorophore formation: (a) close proximity of the backbone atoms of amino acids 65 and 67 and (b) ACID–base chemistry to catalyze the cyclization. Close proximity is achieved by the removal of steric hindrance by a side chain at position 67, where a glycine is present. As described above, despite many mutations at position 67, no functionally fluorescent GFP’s have been found with anything but glycine at position 67, including naturally found GFP analogs (Wall et al., 2000). Among the 3D structures in the protein data bank with a Ser–Tyr–Gly sequence, 10 of the 206 found have the required proximity of backbone atoms (Zimmer and Branchini, 1997), so steric factors seem to be necessary but not sufficient for cyclization. Arginine at position 96 is close by and could act as a base, withdrawing electrons by hydrogen bonding with carbonyl oxygen of Ser65 and activating the carbonyl carbon for nucleophilic attack by the amide nitrogen of Gly67. Aspects of this scheme have been supported by ab initio calculations and by database searches of similar compounds and protein sequences (Branchini et al., 1997). Model compounds identical to the hydroxyphenyl imidizolidinone core of the fluorophore have been synthesized and shown not to be significantly fluorescent in solution (Niwa et al., 1997). The protein and its strategically placed acids and bases at the edges of the fluorophore are implicated in providing key resonance stabilization. The remarkable cylindrical fold of the protein seems ideally suited for the function of the protein— it provides a scaffold that surrounds the fluorophore by 360°, keeping it planar and providing a wide range of possible protein side-chain interactions. Together with the short a-helices and loops on the ends, the barrel structure forms a single compact domain and does not have obvious clefts for easy access of diffusible ligands to the fluorophore. The fluorophore would appear to be protected from collisional quenching by oxygen (K < 0.004 M-1s-1) (Rao et al., 1980), and hence reduction of the quantum yield. For comparison, the bimolecular quenching rate for free tryptophan by oxygen is about 1010 M-1s-1 and for tryptophan within small proteins is on the order of 109 M-1s-1 (Lakowicz and Weber, 1973). GFP from Renilla reniformis (a sea pansy) may have even more specialized structures for maintaining the rigidity of the fluorophore, because its absorption and emission spectra are essentially mirror images (Ward, 1979). This relationship is a hallmark of highly immobilized fluorescent molecules, since it implies that the environment of the fluorophore is exactly the same during the absorbing and emitting states. Perhaps more seriously, photochemical damage by the formation of singlet oxygen from the collision of oxygen with the fluorophore in the excited state is reduced by the restricted access of the structure. Such collisions could result in reactive oxygen species. Thus, the can structure may serve to (a) protect the fluorophore from inactivation and (b) contain any reactive products that form inadvertently. Recent evidence suggests that

STRUCTURES OF AEQUOREA GREEN FLUORESCENT PROTEIN MUTANTS AND OTHER RELATED PROTEINS

Figure 4.3. Stereoview of the fluorophore and its environment. His148, Gln94, Arg96, and Glu222 can be seen on opposite ends of the fluorophore and probably stabilize anionic resonant forms. Water molecules, charged, polar, and nonpolar side chains all contact the fluorophore in various ways. See color insert.

toxicity from the radical species or the resultant hydrogen peroxide can indeed cause cause damage to cells (Liu et al., 1999). The fluorophore is located on the central helix within a couple of angstroms of the geometric center of the cylinder. The pocket containing the fluorophore has a surprising number of charged residues in the immediate environment (Fig. 4.3). The environment around the fluorophore includes both apolar and polar amino acid side chains and immobilized water molecules. Both Phe64 and Phe46 are near the fluorophore and separate the single tryptophan, Trp63, from direct contact with fluorophore (closest distance of 13 Å). Most of the other polar residues in the pocket form an extensive hydrogen-bonding network on the side of Tyr66 that requires abstraction of protons in the oxidation process. It is tempting to speculate that these residues help abstract the protons. Atoms in the side chains of Thr203, Glu222, and Ile167 are in van der Waals contact with Tyr66 so their mutation would have direct steric effects on the fluorophore and would also change its electrostatic environment if the charge were changed, as suggested previously (Ehrig et al., 1995). Remington (2000) and Palm et al. (1997) have further discussed the structural basis of spectral variations. Additional testing may show that mutation of other residues near the fluorophore also affect the absorption and/or emission spectra.

4.5 STRUCTURES OF AEQUOREA GREEN FLUORESCENT PROTEIN MUTANTS AND OTHER RELATED PROTEINS The location of certain amino acid side chains in the vicinity of the fluorophore also begins to explain the fluorescence and the behavior of certain mutants of the protein. At least two resonant forms of the fluorophore can be drawn: one with a partial negative charge on the benzyl oxygen of Tyr66 and one with the charge on the carbonyl oxygen of the imidizolidone ring (Fig. 4.4). Interestingly, basic residues appear to form hydrogen bonds with

73

74

THE THREE-DIMENSIONAL STRUCTURE OF GFP AND ITS IMPLICATIONS FOR FUNCTION AND DESIGN

Figure 4.4. Schematic diagram of the resonant forms of the fluorophore with nearby basic amino acids, His148, Gln94, and Arg96 and the acid Glu222. The bases appear to stabilize anionic oxygen atoms at the opposite ends of the fluorophore, and the acid forms a hydrogen bond with the hydroxyl of Ser65.

each of these oxygen atoms: (a) His48 with Tyr66 and (b) Arg96 with the imidizolidone. These bases presumably act to stabilize and possibly further delocalize the charge on the fluorophore. Kinetic studies have suggested that proton transfers may play an important role in the excitation-fluorescence process (Chattoraj et al., 1996; Youvan and MichelBeyerle, 1996). Based on this structure, the carbonyl oxygen of the imidizolidone, arising originally from the backbone of Tyr66, is a likely candidate for the acceptor, with Arg96 as the donor. Another possibility is the Tyr66 distal oxygen–His148 donor–acceptor pair or the Glu222 interactions. Perturbation of the interaction by mutation to His at position 66 will result in only the high-energy absorption peak and a blue-shifted emission band, whereas disruption of the Ser65 hydroxyl–Glu222 interactions will result in a red-shifted absorption maximum and an unchanged emission spectrum. The Tyr66 to Phe mutation has been reported to have dramatically reduced fluorescence (Heim et al., 1994), presumably due to poorer charge delocalization. The structure of the GFP mutant, Ser65–Thr (Ormo et al., 1996), shows only minor differences in the b-can structure when compared to the wild-type protein. The different absorption properties are consistent with the general trend that disruption of the interaction between the Ser65 hydroxyl and the Glu222 carboxylate eliminate the 395-nm absorption peak. Inspection of the structure of the Ser65–Thr GFP confirms that this interaction has been changed in this mutant as well, via a rotation of the hydroxyl of Thr65 by about 120° about the Ca–Cb bond relative to the wild-type protein (Fig. 4.5). Glutamic acid at

STRUCTURES OF AEQUOREA GREEN FLUORESCENT PROTEIN MUTANTS AND OTHER RELATED PROTEINS

2.8 2.6

2.8 2.6

Figure 4.5. Comparison of the wild-type and Ser65 Thr structures in the vicinity of the fluorophore. Consistent with the pattern that disruption of the Glu222–Ser65 interactions lead to elimination or severe reduction of the absorption at 395 nm, there are different hydrogenbonding arrangements in the Ser65 Thr mutant. [Drawing by Ribbons, coordinates from entry 1GFL and 1EMA from the protein data bank.]

position 222 has corresponding adjustments. The net result is a longer hydrogen bond in the Ser65–Thr structure. Ormo et al. (1996) postulated that the pattern of hydrogen bonding is different in the Thr65 mutation and that the net result is that an anionic Glu222 cannot coexist with an anionic fluorophore without serine at position 65. The availability of E. coli clones expressing GFP has led to extensive mutational analysis of GFP, with many crystal structures appearing that test ideas about the designed rearrangements of side chains resulting from mutagenesis [see Kimata et al. (1999)]. Screens of random and directed point mutations for changes in fluorescent behavior have uncovered a number of informative amino acid substitutions. Mutation of Ser to Thr, Ala, Cys, or Leu causes a loss of the 395-nm excitation peak with a major increase in blue excitation (Delagrave et al., 1995; Heim et al., 1995). When combined with Ser65 mutants, mutations at other sites near the fluorophore such as Val68–Leu and Ser72–Ala can further enhance the intensity of green fluorescence produced by excitation at 488 nm (Delagrave et al., 1995; Cormack et al., 1996). However, amino acid substitutions significantly outside this region also affect the protein’s spectral character. For example, Ser202–Phe and Thr203–Ile both cause the loss of excitation in the 475-nm region with preservation of 395-nm excitation (Heim et al., 1994; Ehrig et al., 1995). The residue Ile167–Thr change results in reversed ratio of 395- to 475-nm sensitivity (Cubitt et al., 1995), while Glu222–Gly is associated with the elimination of only the 395-nm excitation (Ehrig et al., 1995). The pH dependence of the excitation bands at 395 and 475 nm (Ward et al., 1982) is almost certainly due to His148, whose Nd atom is 3.3 Å from the Tyr66 hydroxyl oxygen atom of the fluorophore, although NMR pKa measurements or mutagenesis studies would be needed for confirmation. The structures of other proteins closely related to GFP have been determined to reveal variations on a theme (Remington, 2000) with those from corals being particularly diverse (Verkhvsha and Lukyanov, 2004). The range of colors now span the entire visible spectrum from purple/blue to green to yellow to the far red, with crystal structures revealing secrets of their coloring (Petersen et al., 2003).

75

76

THE THREE-DIMENSIONAL STRUCTURE OF GFP AND ITS IMPLICATIONS FOR FUNCTION AND DESIGN

4.6 STRUCTURE-BASED ENGINEERING OF GREEN FLUORESCENT PROTEIN Mutations in regions of the sequence adjacent to the fluorophore—that is, in the range of positions 65–67—have been systematically explored (Delagrave et al., 1995); some have significant wavelength shifts and most suffer a loss of fluorescence intensity. For example, mutation of the central Tyr to Phe or His shifts the excitation bands, but there is an overall loss of intensity. Secondary mutations to compensate for the deleterious intensity effects may also now be possible. The Ser65–Thr mutant is particularly interesting because of its reported increase in fluorescence intensity (Heim et al., 1994, 1995). These authors suggested that this effect is through improved conversion of the tyrosine to dehydrotyrosine. This is not likely, however, because we see an essentially fully cyclized structure in the wild-type protein (Yang et al., 1996). The report of improvements in the “brightness” of cells produced by DNA shuffling of GFP constructs, comprising mutations Phe99–Ser, Met153–Thr, and Val1–Ala as numbered in the TU#58 system (Crameri et al., 1996), is impossible to explain from the structures, because the quantum yields are close to unity to begin with. Positions 153 and 163 are on the surface of the protein and may exert their effects through improved solubility and/or reduced aggregation. The Phe–Ser mutation at first glance would appear to destabilize the core of the protein, and at present we have no good ideas how it would improve the fluorescence properties of GFP.

4.7

CONTROL OF ACTIVATION OF THE FLUOROPHORE

The mechanism of activation of the fluorophore from ordinary protein structure is consistent with a nonenzymatic cyclization mechanism like that of Asn–Gly deamidation (Wright, 1991) followed by oxidation of the tyrosine to dehydrotyrosine, as previously suggested. The role of molecular oxygen in this mechanism and in GFP fluorescence is paradoxical, however. Molecular oxygen is proposed to be needed for oxidation of tyrosine to form an extended aromatic system, but oxygen must also be excluded from regular interactions with the fluorophore or else collisional quenching of the fluorescence or damaging photochemistry will occur. The low bimolecular quenching rate suggests that the protein’s design sacrifices efficient fluorophore formation for stability and higher quantum yields once fully formed. Once formed, the dehydrotyrosine can be chemically reduced by sodium dithionite to produce a mixture of d- and l-amino acid geometries at the carbon alpha of tyrosine 66 (Yang, F. personnel communication). Because most enzymatic reactions are stereospecific, this result implies that the catalysis of the original dehydrogenation may, in fact, be nonenzymatic or may involve amino acid side chains that are (a) only transiently available in the oxidation and (b) not present (or needed) in the reduction. The mechanisms for formation of the chromophore in other colored proteins is a topic of current interest.

4.8

GFP TRUNCATION AND FUSION CONSTRUCTS

Truncation of more than seven amino acids from the C-terminus or more than the Nterminal Met led to total loss of fluorescence (Dopf and Horiagan, 1996). These N- and C-termini truncation studies and the fluorescent fusion products are now understandable, given the structure of the protein. Since the C-terminus loops back outside the cylinder

CONTROL OF DIMERIZATION OF GREEN FLUORESCENT PROTEIN

and the last seven or so amino acids are disordered, it should not be critical to have them present and further addition would seem to be easily tolerated. These residues do not form a stave of the barrel. The role of the N-terminus is a little less clear, because the first strand in the barrel does not begin until amino acid 10 or 11. Thus barrel formation does not require the N-terminal region. The N-terminal segment, is however, an integral part of the “cap” on one end of the protein, and it may be essential in folding events or in protecting or forming the fluorophore. Again, extensions at the N-terminus would not disrupt the motif structure of the protein. In fact, the GFP crystal structure solved by Wu et al. (1997) has a 37-residue His tag on the N-terminus, of which only two amino acids are ordered and hence visible in the electron density map. The basic assembly of GFP seems quite robust. Cyclic permutants of the GFP sequence are possible, and inserts or whole domains of other proteins can be inserted into the GFP sequence, forming appendages that do not destroy fluorescence and yet allow sensing of other activities by monitoring changes in the GFP fluorescence [Baird et al., 1999; Akemann et al., 2001]. Thus the b-can scaffold of GFP is becoming a framework for dramatically reengineered protein structures.

4.9 CONTROL OF DIMERIZATION OF GREEN FLUORESCENT PROTEIN Green fluorescent protein can form homodimers in solution and in crystals. The equilibrium dissociation constant, Kd, is approximately 100 mM as measured by analytical ultracentrifugation (F. Yang and G. N. Phillips, unpublished results). This weak association is consistent with (a) the observation that most, but not all, crystal forms exist as dimers and (b) correlation microscopy of purified GFP, which revealed a diffusion coefficient consistent with monomers at low-protein concentrations (Terry et al., 1996). The fluorescence of GFP also changes on dimerization (Ward et al., 1982). The DsRed homolog is a tetramer (Wall et al., 2000), but maintains the usual dimer interface as a part of its 222 symmetry. In both the wild-type and native crystal structures, the crystallographic contacts are all rather tenuous, consisting of a few amino acid side chains for each. In contrast, the dimer symmetry is maintained by extensive contacts and is consistent in many structures (Yang et al., 1996; Wu et al., 1997; M. Perozzo, W. Ward, and K. Ward, personnel communication) and thus is likely to be the source of the dimerization seen in solution studies. The dimer contacts are fairly tight and consist of a core of hydrophobic side chains from each of the two monomers and a wealth of hydrophilic contacts. The smaller hydrophobic patch could conceivably be involved in physiological interactions with aequorin, because there would be a natural advantage to close proximity for efficient energy transfer. There are fluorescence changes on dimerization (Cutler, 1995), and subtle rearrangements of the fluorophore may also occur upon dimerixation (Wu et al., 1997). Control of the dimerization will be important for fluorescence resonance energy transfer (FRET) studies of protein–protein interactions using GFP (Heim and Tsien, 1996; Mitra et al., 1996), because one would not want to induce association and hence resonance energy transfer between the differently colored GFP proteins by mechanisms other than those of the target protein interactions. Dimerization of GFP and the concornmitent spectral changes have also been exploited in the construction of novel reporters of calcium and calmodulin activities (Romoser et al., 1997; Persechini et al., 1997). A calmodulin target sequence was inserted as a linker between a red-shifted and a blue-shifted GFP sequence to yield a dimer GFP that exhibits FRET changes when calcium binds to calmodulin

77

78

THE THREE-DIMENSIONAL STRUCTURE OF GFP AND ITS IMPLICATIONS FOR FUNCTION AND DESIGN

(Romoser et al., 1997). These authors went on to fuse calmodulin to the molecule to yield a complete calcium indicator that is tunable by changing the particular sequences of the calmodulin target (Persechini et al., 1997). Targetability of similar GFP-based Ca2+ indicators for the cytosol, nucleus, and endoplasmic reticulum has also been achieved by adding appropriate localization signals (Miyawaki et al., 1997).

4.10

IMPLICATIONS OF THE STRUCTURE ON FOLDING MECHANISMS

The pattern of connections of the antiparallel b strands (Figs. 4.6 and 4.7) suggests that the initial stages of folding the b-can involve the coalescence of two- or three-stranded segments of contiguous polypeptide chain, followed by assembly of the segments into a complete can around the central helix. Many variants of GFP fold with different rates or extents relative to wild-type protein, producing better markers in that less protein winds up in inclusion bodies. One change, Val163–Arg, increases the temperature tolerance for functional GFP expression (Kahana and Silver, 1996). This effect may also be the basis of several known improvements in total fluorescence by GFP (Cormack et al., 1996), and further studies in this area are certainly warranted. Folding kinetics are usually complicated, and it is probably difficult to predict results of mutants in advance, but GFP could be used as a convenient reporter of folding in biophysical studies of b-barrel formation that might lead to better understanding of folding phenomena.

Figure 4.6. Stereoview of the polypeptide trace of GFP, showing the N- and C-termini and the central helix. Note that not much of either of the termini can be cleaved without losing a stave in the barrel structure. [Figure produced with RasMol (Sayle and Milner-White, 1995.]

REFERENCES

Figure 4.7. Schematic diagram of secondary structure components and their connections. The protein appears to be built of sets of two or three adjacent b strands that come together to form the can structure.

4.11

CONCLUSIONS AND FUTURE PROSPECTS

New insights into the mechanism of the photochemistry of GFP and its relatives should be of great interest to cell biologists, especially if such new mechanisms suggest improved procedures for imaging. In the complex problems of multicolor analyses and quantitation of FRET, careful attention to the biophysical details will be essential. Furthermore, future studies may include imaging devices that have the capability of acquiring full excitation, emission, and absorption spectra for every pixel in a scene. Clearly for biological purposes, it may also sometimes be desirable to sacrifice long-term stability for “real-time” color development. In fact, fast photodestruction could be desirable to follow increases and then decreases in reporter gene applications. The discoveries of blinking (Dickson et al., 1997) and light-induced “kindling” species (Chudakov et al., 2003) also extend the range of potential uses. Another potential use of GFP is in electron microscopy. If the photochemistry can be controlled, the free radical reactions could in principle be used to produce deposits of metal stains for high-resolution localization experiments. The 3D structures of GFP and its relatives have provided a physicochemical basis of many observed features of the proteins, including their stability, protection of their fluorophores or chromophores, behavior of mutants, dependence of the spectra on pH, and oligomerization properties. The structures will also allow directed mutation studies to complement random mutagenesis and also improve combinatorial approaches.

REFERENCES Akemann, W., Raj, C. D., and Kno, T. (2001). Functional characterization of permuted enhanded green fluorescent protein comprising varying linker peptides. Photochem. Photobiol. 42:356–363. Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (1999). Circular permutation and receptor insertion within green fluorescent proteins. Proc. Natl. Acad. Sci. USA 96:11241–11246. Branchini, B. R., Lusins, J. O., and Zimmer, M. (1997). A molecular mechanics and database analysis of the structural preorganization and activation of the chromophore-containing hexapeptide fragment in green fluorescent protein. J. Biomol. Struct. Funct. 14:1–8.

79

80

THE THREE-DIMENSIONAL STRUCTURE OF GFP AND ITS IMPLICATIONS FOR FUNCTION AND DESIGN

Brejc, K., Sixma, T. K., Kitts, P. A., Kain, S. R., Tsien, R. Y., Ormo, M., and Remington, S. J. (1997). Structural basis for dual excitation and photoisomerization of the Aequorea victoria green fluorescent protein. Proc. Natl. Acad. Sci. USA 94:2306–2311. Carson, M. (1987). Ribbon models of macromolecules. J. Mol. Graphics 5:103–106. Chalfie, M. (1995). Green fluorescent protein. Photochem. Photobiol. 62:651–656. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263:802–805. Chattoraj, M., King, B. A., Bublitz, G. U., and Boxer, S. G. (1996). Ultra-fast excited state dynamics in green fluorescent protein: Multiple states and proton transfer. Proc. Natl. Acad. Sci. USA 93:8362–8367. Chudakov, D. M., Belousov, V. V., Zaraisky, A. G., Novoselov, V. V., Staroverov, D. B., Zorov, D. B., Lukyanov, S., and Lukyanov, K. A. (2003). Kindling fluorescent proteins for precise in vivo labeling. Nat. Biotechnol. 21:191–194. Cody, C. W., Prasher, D. C., Westler, W. M., Prendergast, F. G., and Ward, W. W. (1993). Chemical structure of the hexapeptide chromophore of the Aequorea green-fluorescent protein. Biochemistry 32:1212–1218. Cormack, B. P., Valdivia, R. H., and Falkow, S. (1996). FACS-optimized mutants of the green fluorescent protein (GFP). Gene 173:33–38. Crameri, A. E., Whitehom, E. A., Tate, E., and Stemmer, W. P. C. (1996). Improved green fluorescent protein by molecular evolution using DNA shuffling. Nat. Biotechnol. 14:315– 319. Cubitt, A. B., Heim, R., Adams, S. R., Boyd, A. E., Gross, L. A., and Tsien, R. Y. (1995). Understanding, improving and using green fluorescent proteins. TIBS 20:448–455. Cutler, M. W. (1995). Characterization and energy transfer mechanism of the green-fluorescent protein. Rutgers, the State University of New Jersey, New Brunswick, NJ. Delagrave, S., Hawtin, R. E., Silva, C. M., Yang, M. M., and Youvan, D. C. (1995). Red-shifted excitation mutants of the green fluorescent protein. Biotechnology 13:151–154. Dickson, R. M., Cubitt, A. B., Tsien, R. Y., and Moerner, W. E. (1997). On/off blinking and switching behaviour of single molecules of green fluorescent protein. Nature 388:355–358. Dopf, J., and Horiagan, TM. (1996). Deletion mapping of the Aequorea victoria green fluorescent protein. Gene 173:39–44. Ehrig, T., O’Kane, D. J., and Prendergast, F. G. (1995). Green-fluorescent protein mutants with altered fluorescence excitation spectra. FEBS Lett. 367:163–166. Heim, R., and Tsien, R. Y. (1996). Engineering green fluorescent protein for improved brightness, longer wavelengths and fluorescence resonance energy transfer. Curr. Biol. 6:178–182. Heim, R., Prasher, D. C., and Tsien, R. Y. (1994). Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proc. Natl. Acad. Sci. USA 91:12501–12504. Heim, R., Cubitt, A. B., and Tsien, R. Y. (1995). Improved green fluorescence. Nature (London) 373:663–664. Hendrickson, W. A., Pahler, A., Smith, J. L., Satow, R., Meritt, E. A., and Phizackerley, R. D. (1989). Crystal structure of core strepavadin determined from multiwavelength anomalous diffraction of synchrotron radiation. Proc. Natl. Acad. Sci. USA 86:2190–2194. Hopf, M., Goehring, W., Ries, A., Templ, R., and Hohenester, E. (2001). Crystal structure and mutational analysis of a perlecan-binding fragment of nidogen-1. Nat. Struct. Biol. 8:634–640. Kahana, J., and Silver, P. A. (1996). Use of the A. victorea green fluorescent protein to study protein dynamics in vivo. Curr. Protocols Mol. Biol. 9.7.22–9.7.28. Kreusch, A., Newbueser, A., Schlitz, E., Weckessser, I., and Schultz, G. E. (1994). The structure of the membrane channel porin from Rhodopseudomonas blastica at 2.0 Angstrom resolution. Protein Sci. 3:58.

REFERENCES

Lakowicz, J., and Weber, G. (1973). Quenching of fluorescence by oxygen. A probe for structural fluctuation in macromolecules. Biochemistry 12:4161–4170. Lim, C. R., Kimata, K., Oka, M., Nomaguchi, K., and Kohno, K. (1995). Thermosensitivity of a green fluorescent protein utilized to reveal novel nuclear-like compartments. J. Biochem. (Tokyo) 118:13–17. Liu, H.-S., Jan, M.-S., Chou, C.-K., Chen, P.-H., and Ke, N.-J. (1999). Is green fluorescent protein toxic to living cells? Biochem. Biophys. Res. Commum. 260:712–717. Lossau, H., Kumer, A., Heinecke, R., Poellinger-Dammer, F., Kompa, C., Bieser, G., Jonsson, T., Silva, C. M., Yang, M. M., Youvan, D. C., and Michel-Beyerle, M. E. (1996). Time-resolved spectroscopy of wild type and mutant green fluorescent proteins reveals excited state deprotonation consistent with fluorophore–protein interactions. Chem. Phys. 213:1–16. Matz, M. V., Lukyanov, K. A., and Lukyanov, S. A. (2002). Family of the green fluorescent protein: Journey to the end of the rainbow. Bioessays 24:953–959. Mitra, R. D., Silva, C. M., and Youvan, D. C. (1996). Fluorescence resonance energy transfer between blue-emitting and red-shifted excitation derivatives of the green fluorescent protein. Gene 173:13–17. Miyawaki, A., Llopis, J., Heim, R., McCaffery, J. M., Adams, J. A., Ikung, M., and Tsien. R. Y. (1997). Fluorescent indicators for the Ca2+ based on green fluorescent proteins and calmodulin. Nature 388:822–887. Morin, J. G., and Hastings, J. W. (1971). Energy transfer in a bioluminescent system. J. Cell Physiol. 77:313–318. Niwa, H., Matsuno, T., Kojima, T., Kubota, M., Hirano, T., Ohashi, M., Inouye, T., Ohmiya, Y., and Tsuji, F. I. (1977). Aequorea green fluorescent protein: Structural elucidation of the chormophore. In Bioluminescence and Chemoluminescence, Hastings, J. W., Kricka, L. J., and Stanley, P. E., Eds., Wiley, Chichester, pp. 2395–2398. Ormo, M., Cubitt, A., Kallio, K., Gross, L., Tsien, R. Y., and Remington, S. J. (1996). Crystal structure of the Aequorea victoria green fluorescent protein. Science 273:1392–1395. Palm, G. J., Zdanov, A., Gaitanaris, G. A., Stauber, R., Pavlakis, G. N., and Wlodawer, A. (1997). The structural basis for spectral variations in green fluorescent protein. Nat. Struct. Biol. 4:361–365. Palm, G. J., and Wlodawer, A. (1999). Spectral variants of green fluorescent protein. Methods Enzymol. 302:378–394. Perozzo, M. A., Ward, K. B., Thompson, R. B., and Ward, W. W. (1988). X-ray diffraction and timeresolved fluorescence analyses of Aequorea green fluorescent protein crystals. J. Biol. Chem. 263:7713–7716. Persechini, A., Lynch, J. A., and Romoser, V. A. (1997). Novel fluorescent indicator proteins for monitoring free intracellular Ca2+. Cell Calcium 22:209–216. Peterson, J., Wilmann, P. G., Beddoe, T., Oakley, A. J., Devenish, R. J., Prescott, M., and Rossjohn, J. (2003). The 2.0 Å crystal structure of eqFP611, a far red fluorescent protein from the sea anemone, Entacmaea quadricolor. J. Biol. Chem. 278:44626–44631. Prasher, D. C., Eckenrode, V. K., Ward, W. W., Prendergast, F. G., and Cormier, M. J. (1992). Primary structure of the Aequorea victoria green-fluorescent protein. Gene 11 1:229–233. Rao, B. D. N., Kemple, M. D., and Prendergast, F. G. (1980). Proton nuclear magnetic resonance and fluorescence spectroscopic studies of segmental mobility in aequorin and a green fluorescent protein from Aequorea forskalea. Biophys. J. 32:630–632. Remington, S. J. (2000). Structural basis for understanding spectral variations in green fluorescent protein. Methods Enzymol. 305:196–211. Romoser, V. A., Hinkle, P. M., and Persechini, A. (1997). Detection on living cells of Ca2+ dependent changes in the fluorescence emission of an indicator composed of two green fluorescent protein variants linked by a calmodulin binding sequence. J. Biol. Chem. 272:13270–13274.

81

82

THE THREE-DIMENSIONAL STRUCTURE OF GFP AND ITS IMPLICATIONS FOR FUNCTION AND DESIGN

Sayle, R., and Milner-White, E. (1995). RasMol: Biomolecular graphics for all. TIBS 20:374–375. Shimomura, O., Johnson, F. H., and Saiga, Y. (1962). Extraction, purification, and properties of aequorin, a bioluminescent protein from the luminous hydromedusan, Aquorea. J. Cell Comp. Physiol. 59:223–239. Terry, B. R., Matthews, E. K., and Haseloff, J. (1996). Molecular characterisation of recombinant green fluorescent protein by fluorescence correlation microscopy. Biochem. Biophys. Res. Commun. 2l7:21–27. Verhusha, V. V., and Lukyanov, K. A. (2004). The molecular properties and applications of Anthozoa fluorescent proteins and chromoproteins. Nat. Biotechnol. 22:289–296. Voityuk, A. A., Michel-Beyerle, M. E., and Rosch, N. (1997). Protonation effects on the chromophore of green fluorescent protein—Quantum chemical study of the absorption spectrum. Chemi. Phys. Lett. 272:162–167. Wall, M. A., Socolich, M., and Rnaganathan, R. (2000). The structural basis for red fluorescence in the tetrameric GFP homolog DsRed. Nat. Struct. Biol. 7:1133–1138. Ward, W. W. (1979). Energy transfer processes in bioluminescence. In Photochemical and Photobiological Reviews, Vol. 4, Smith, K., Ed., Plenum, New York, pp. 1–57. Ward, W. W., and Bokman, S. H. (1982). Reversible denaturation of Aequorea green-fluorescent protein: Physical separation and characterization of the renatured protein. Biochemistry 21:4535–4540. Ward, W. W., and Cornier, M. J. (1979). An energy transfer protein in coelenterate bioluminescence: Characterization of the Renilla green-fluorescent protein. J. Biol Chem. 254:781–788. Ward, W. W., Prentice, H., Roth, A., Cody, C., and Reeves, S. (1982). Spectral perturbations of the Aequorea green fluorescent protein. Photochem. Photobiol. 35:803–808. Wright, H. T. (1991). Nonenzymatic deamidation of asparaginyl and glutaminyl residues in proteins. Crit. Rev. Biochem. Mol. Biol. 26:1–52. Wu, C.-K., Liu, Z.-J., Rose, J. P., Inouye, S., Tsuji, F., Tsien, R. Y., Remington, S. J., and Wang, B.-C. (1997). The three-dimensional structure of green fluorescent protein resembles a lantern. In Bioluminescence and Chemoluminescence, Hastings, J. W., Kricka, L. J., and Stanley, P. E., Eds., Wiley, Chichester, pp. 399–402. Yang, F., Moss, L. G., and Phillips, G. N., Jr. (1996). The molecular structure of green fluorescent protein. Nat. Biotechnol. 14:1246–1251. Youvan, D., and Michel-Beyerle, M. E. (1996). Structure and fluorescence mechanism of GFP. Nat. Biotechnol. 14:1219–1220. Zimmer, M. (2002). Green fluorescent protein (GFP): Applications, structure and related photophysical behavior. Chem. Rev. 102:759–781. Zimmer, M., and Branchini, B. (1997). A computational and database analysis of the structural preorganization and activation involved in chromophore formation if green fluorescent protein. In Bioluminescence and Chemoluminescence, Hastings, J. W., Kricka, L. J., and Stanley, P. E., Eds., Wiley, Chichester, pp. 407–410.

5 MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN David A. Zacharias The Whitney Laboratory for Marine Bioscience, University of Florida Department of Neuroscience, St. Auqustine, FL

Roger Y. Tsien Department of Pharmacology, University of California, San Diego, La Jolla, CA

5.1 INTRODUCTION Likely not since the discovery by McElroy (McElroy, 1947) of the involvement of ATP in the reaction catalyzed by firefly luciferase has there been as much interest in bioluminescence as currently exists in the scientific community. The interest was reignited, in large part, by the work of Chalfie et al. (1994), first showing the usefulness of the Aequorea victoria GFP. While Aequorea was first shown to fluoresce when irradiated with ultraviolet light in 1955 (Davenport and Nichol, 1955), it was not for another 40 years that it was shown for the first time that there exists a genetically encoded reporter molecule that is detectable in the absence of an enzymatic substrate or cofactor in a variety of cell types. The fluorescent properties of GFP make it especially useful in living cells and tissue. Recently, interest has surged again because there has been an exciting expansion in the discovery and characterization of homologous fluorescent proteins from distantly related sea creatures (discussed extensively in other chapters in this book). The protein sequence, derived from the cDNA nucleotide sequence (Prasher et al., 1992), contains 238 amino acid residues and enabled determination of the chromophore structure (Shimomura, 1979; Cody et al., 1993). The p-hydroxybenzylideneimidazolinone chromophore is formed by the autocatalytic cyclization of Ser65, Tyr66, and Gly67 and dehydrogenation of the tyrosine. A mechanism for formation of the chromophore, which Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

83

84

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

is strongly fluorescent only in the intact protein (Ward and Bokman, 1982), has been proposed (Heim et al., 1994; Cubitt et al., 1995) and can be generally described as a threestep process in which cyclization is initiated by nucleophilic attack of the nitrogen atom of Glycine at position 67 on the carbonyl carbon of serine at position 65. This reaction creates the five-membered imidazolone ring. The carbonyl oxygen of S65 is then dehydrated, and finally the Ca–Cb bond of Y66 is oxidized to conjugate the ring systems. While these steps must occur to arrive at the mature chromophore, the energetics and physical nature of the chromophore and its immediate environment that are necessary to drive the reaction have not been completely elucidated. Changing the Arg at position 96 to Ala (R96A) slows the cyclization reaction from minutes to months, allowing the isolation and crystallization of GFP in a state just prior to cyclization of the chromophore, providing a fixed image of a physical state of the protein that would be difficult to capture otherwise (Barondeau et al., 2003). This experimental result led the authors to propose a “conjugation-trapping mechanism” in which the thermodynamically unfavorable cyclization reaction of the chromophore is coupled to an electronic conjugation trapping reaction to yield the mature chromophore (Barondeau et al., 2003). Nucleotide sequences derived from Aequorea indicate that at least five variants of GFP exist (Table 5.1). Four of the DNA sequences are encoded by cDNAs, while the fifth is encoded in the exons of a gfp gene. The variants differ generally by conservative amino acid replacements, suggesting they may have nearly identical physical properties. One of the genomic clones with GFP sequence contains three exons that can be matched to the TABLE 5.1. Heterogeneous Amino Acid Residues Derived from the gfp Nucleotide Sequences Locus1

Residue Position

AEVGFPA2

AEVGFP3

AVGFP14

AVGFP25

AEVGFPB6

14 25 30 45 84 100 108 141 154 157 172 209 212 213 219 226 228

Ile His Ser Lys Phe Phe Thr Leu Ala Gln Glu Lys Asn Glu Val Ala Gly

Ile Gln Ser Lys Phe Tyr Thr Met Gly Pro Lys Lys Asn Glu Ile Ala Gly

Val Gln Ser Asn Phe Tyr Thr Met Gly Pro Lys Lys Asn Glu Ile Ala Arg

Ile Gln Arg Lys Leu Tyr Thr Met Gly Pro Lys Gln His Gly Val Ser Gly

Ile His Ser Lys Phe Phe Ser Met Ala Gln Glu Lys Asn Glu Ile Ala Gly

1 As assigned by GenBank. 2 Derived from the gfp10 cDNA reported by Prasher et al. (1992), Accession No. M62653. 3 Derived from the gfp cDNA reported by Inouye & Tsuiji (1994), Accession No. L29345. 4 Derived from a cDNA submitted to GenBank by Watkins & Campbell, Accession No. X83959. 5 Derived from a cDNA submitted to GenBank by Watkins & Campbell, Accession No. X83960. 6 Derived from the exons of the gene encoded by gfp2 (Prasher et al., 1992), Accession No. M62654. Source: With permission from the Annual Review of Biochemistry, volume 67 © 1998 by Annual Reviews www.annualreviews.org.

MUTATIONAL STRATEGIES

cDNA, while a fourth exon must exist to account for the 5¢ end of the cDNA (Prasher et al., 1992). The tripeptide encoding the chromophore is located near the 3¢ end of exon II (Prasher et al., 1992). A similarly large number of sequence variants exists in corals, and systematic analysis of the sequences of these variants from an evolutionary perspective is shedding considerable light on the genetic basis behind the diversity in coloration (Kelmanson and Matz, 2003; Chapter 3, this volume). A flood of activity describing the use and optimization of the wild-type GFP and its modification followed a report (Chalfie et al., 1994) where it was shown that the chromophore forms when GFP is expressed from its cDNA in a prokaryote (E. coli) or a eukaryote (C. elegans); for reviews see Prasher (1995), Simon (1996), Tsien (1998), Zacharias et al. (2000), Matz et al. (2002), Zhang et al. (2002), Zimmer (2002), Miyawaki (2003), and other chapters in this volume. The earliest modifications of wild-type GFP were directed at improving the ability to express the protein in mammalian cell systems at 37°C. Other improvements that followed in rapid succession included simplification of the excitation spectrum, increasing the fluorescence intensity of the protein by increasing the extinction coefficient as well as increasing the speed and efficiency of folding, making different colors or spectral mutants, reducing sensitivity to pH and halides, and most recently reducing the tendency of the protein to dimerize. In this chapter we will describe, in general terms, the methods and strategies used to generate existing mutants of A. victoria GFP. We then present a classification of A. victoria GFPs based on the chemical structure and photophysical behavior of their chromophores and then catalog and describe many of the useful and interesting mutants that have been created and characterized to date.

5.2 MUTATIONAL STRATEGIES Many strategies have been used to generate mutations in GFP; the degree of sophistication and power has risen considerably since the earliest methods were developed, partly because GFP itself is such a convenient reporter and testbed for the development of improved mutagenic strategies. We will present the strategies roughly in order of increasing degree of complexity: (1) random mutagenesis by chemical mutagens, (2) error-prone polymerase chain reaction (PCR) over the entire or a portion of the coding sequence, (3) deliberate site-directed mutations to test specific hypotheses (eventually aided by the crystal structure), (4) randomization (possibly with codon biases) of a predetermined, limited stretch of residues, (5) DNA “shuffling,” (6) heteroduplex recombination, and (7) directed evolution to change or incorporate into fluorescent proteins (FPs) certain properties more rapidly than could be accomplished by rational design methodology. 1. Random Mutagenesis by Chemical Mutagens. The cDNA vector encoding GFP can be treated with mutagens such as hydroxylamine or nitrous acid before transfection into a host (Sikorski and Boeke, 1991). As far as mutagenic strategies for GFP are concerned, this method has been largely supplanted by protocols that rely solely on molecular biology. 2. Error-Prone PCR. cDNA can be amplified by PCR under conditions where the polymerase fidelity is compromised—for example, by replacing Mg2+ with Mn2+ in the reaction mixture and/or restricting the concentration of one nucleoside triphosphate at a time (Muhlrad et al., 1992). Error-prone PCR has the advantages of confining the mutations only to a desired region of the gene between the two

85

86

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

primers and in producing a greater variety of mutations at the nucleic acid level. Nevertheless, most mutations are typically changes of just one base in a relevant codon, so that amino acids for which codons differ in two or three positions from the starting codon are much less likely to be accessed. The importance of Y66, I167, T203, and E222 (Heim et al., 1994; Ehrig et al., 1995) were discovered this way, as well as some of the folding mutations such as Y145F, M153T, and V163A (Heim and Tsien, 1996), C70V (Zapata-Hommer and Griesbeck, 2003), and F46L (Nagai et al., 2002). This method is presumably random at the nucleic acid level and is a strategy that is still used commonly as a first step in some of the methods described below. 3. Deliberate Site-Directed Mutations. This method is generally used to mutagenize GFP to test specific hypotheses about certain amino acid replacements. This was the approach used to make monomeric versions of GFP (Zacharias et al., 2001). The coordinates from a crystal structure of dimeric GFP (Yang et al., 1996a) were used to predict which residues were responsible for the weak tendency to dimerize (Table 5.5; Fig. 5.4). The exchange of the aliphatic residues for positively charged residues was effective in each instance of reducing or eliminating the ability to dimerize. An earlier example of systematic single amino acid replacements was exchanging Thr203 for aromatic residues such as Tyr (Delagrave et al., 1995; Ormo et al., 1996). See Table 5.4 for a description of the resulting changes in the spectral characteristics imparted on the protein by these substitutions. Another example is testing the effect of adding one specific mutation known to impart a desirable property, such as efficient folding, to another mutant version of GFP to see if the benefits would be additive (Nagai et al., 2002). 4. Randomization of a Predetermined Stretch of Amino Acids. Delagrave et al. (1995) randomized positions 64–65 and 67–69 with the random codon NNK, where N is any base and K is G or T. They found (F64M, S65G, Q69L) as their favored mutant, code-named “RSGFP4” (Delagrave et al., 1995). Later, Cormack et al. (1996) randomized positions 55 through 74 with a 10% probability of substituting NNK for each wild-type codon. This produced three different mutants: (F64L, S65T), (S65A, V68L, S72A), and (S65G, S72A). All of these have excitation spectra shifted to 470–490 nm because of their replacement of Ser65, plus one or more mutations that improve folding efficiency. Subsequent crystal structures have provided some guidance by defining proximity to the chromophore in three-dimensional space rather than as a distance defined by the sequence of the amino acids. This knowledge has allowed the rational design of mutants with desired characteristics such as the introduction of simple metal-binding sites engineered into the staves of the beta barrel [e.g., Richmond et al. (2000)] (see Fig. 5.5A). 5. DNA Shuffling. This method (Crameri et al., 1996), reviewed in Giver and Arnold (1998) and in Minshull and Stemmer (1999), is one way to recombine an existing set of mutations spread throughout the gene of interest or to combine different mutations from each of two or more copies of the same gene. For example, it would be an ideal method to mix the individual (or sets of) beneficial mutations, such as those that improve folding, and produce composite proteins containing multiple mutations that might synergize with each other. In brief, a collection of cDNAs, each of which contains only one or a few mutations, is subjected to limited digestion with DNAse. The fragments are re-annealed in a PCR-like reaction with

SCREENING METHODS

nucleoside triphosphates but without primers. This process allows fragments with different mutations to re-sort with each other, repairs the breaks, and introduces additional point mutations. The resulting mixture is finally amplified by conventional PCR with primers, spliced into an expression vector, and screened. Obviously this cycle can be repeated as many times as desired. Using this technique, Crameri et al. (1996) [reviewed in Minshull and Stemmer (1999)] achieved a significant increase in brightness over wild-type GFP. Inspection of the mutations they generated (F99S, M153T, V163A) found that two of the three are ones found by other methods of random mutagenesis. So the individual mutants produced by DNA shuffling are not necessarily unique, but the potential for synergistically recombining them is a main attraction. 6. Heteroduplex Recombination. This is a labor-intensive twist on the polymerasebased DNA shuffling method. The general strategy consists of creating libraries of chimeric DNA sequences derived from homologous, nonidentical parental sequences (Volkov et al., 1999). cDNA heteroduplexes formed by denaturation and annealing are transformed into bacteria where the endogenous DNA repair system fix regions of nonidentity within the heteroduplexes, thereby creating a library of sequences comprised of elements of all parents. Among the potential advantages of this method is the possibility of simultaneously combining desired features from many individual, homologous cDNAs. 7. Directed Evolution. This methodology requires fairly loose definition. Realistically, the process can comprise one, several, or theoretically all of the mutagenic processes listed above. Generally, it is the repetitive mutagenesis of a cDNA encoding a protein that is subsequently expressed and exposed to some selective pressure or development criterion in an effort to capture a protein that has altered characteristics that are desirable. The combination of mutagenesis on many levels with selective pressure can yield remarkable changes in protein activity by incorporation of mutations that might never have been made rationally. Some examples where directed evolution has proven useful include the following: (a) the creation of allosteric GFP biosensors for beta-lactamase activity (Doi and Yanagawa, 1999), (b) to monitor folding efficiencies of proteins to which GFPs are fused (Waldo et al., 1999), and (c) perhaps the most dramatic example, the generation of a monomeric version the tetrameric red fluorescent protein [Campbell et al. (2002); see also Gibbs et al. (2001)] and subsequently a rainbow of color variatious from the monomer RFP (Shaner et al., 2004). Using fluorescent proteins as a reporter of the success of mutagenic strategies has probably decreased the development time and increased the efficiency of protocols designed to alter protein activity by mutagenesis.

5.3 SCREENING METHODS 1. Visual screening requires the least expensive equipment. The minimum is a source of excitation light, either 365 nm from a “black-light” illuminator or ~480 nm from a xenon lamp and interference filter or monochromator that can illuminate Petri dishes. Observation is through a long-pass filter. In the case of UV excitation, an external UV-blocking filter is advisable for health reasons. For blue excitation, one can tape pieces of yellow or orange gelatin filters (e.g., Kodak Wratten filters) over

87

88

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

disposable lab safety glasses. Alternatively, one can scan the dish with a fluorescence microscope with a low-power objective, typically 4¥, using the appropriate filter cube to select excitation and emission wavelengths. The main advantage of visual screening is low capital cost; however, it is both laborious and insensitive to small changes in wavelength below the capacity of human color vision to recognize easily. 2. Video imaging with a frame grabber and computerized false-color display of spectral features was dubbed digital imaging spectroscopy by Youvan et al. (1995). This method is similar to the excitation or emission ratioing now common in imaging physiological indicators inside cultured mammalian cells, except that the optics illuminate and observe a macroscopic rather than microscopic field of view. Digital imaging is more objective and sensitive than human vision to small changes in spectra, but the combined optical and computer setup is relatively expensive. Custom-designed instruments are becoming commonplace and being used successfully in labs (Delagrave et al., 1995; Sawano and Miyawaki, 2000; Griesbeck et al., 2001; Nagai et al., 2002; Neal Woodbury, personal communication) interested in developing fluorescent proteins for uses in cell biology. At least one platform is commercially available from Genetix, www.genetix.com. 3. Fluorescence-activated cell sorting is one of the most obvious high-throughput screening methods to search for useful mutants, especially when the sorting criterion is an unusual ratio of emissions at two wavelengths and no information on the spatial distribution of the protein is necessary. The earliest example of its use was to screen for random mutants that increased fluorescence intensity (Cormack et al., 1996). Two recent examples demonstrate the power of this screening technique, in combination with other innovative, directed-evolution techniques, to generate a broad range of spectral mutants of monomeric red fluorescent protein (Shaner et al., 2004; Wang et al., 2004). 4. High-throughput microscopy (HTM) is a methodology that is finding a home in large-scale drug discovery programs and, because of the increased abundance of commercially available, lower-priced and more powerful detection platforms, is becoming feasible for the academic researcher as well. Such systems provide throughput similar to that of FACS, but have the ability to quantify cellular morphology including the spatial distribution and concentration of fluorescently labeled molecules within the cells. While no group has as yet used the technology to discover useful mutants of GFP, such a task is certainly well within the capability of the technology. In many cases, GFP and GFP-fusion proteins fold with better efficiency when expressed in mammalian cells versus bacterial expression systems. The trapping of nascent proteins in inclusion bodies in bacterial expression systems could foreseeably prevent discovery of useful, but hard-to-fold, mutants of GFP. HTM would allow screening of very large mutagenic libraries of GFP expressed in mammalian cells. The ability to easily configure the excitation and emission characteristics in some HTM systems such as the EIDAQ100 (www.Q3DM.com) would allow users to screen for and select mutants in the exact format that the mutants would be imaged in later experiments. While existing protocols have not been adapted for the isolation of mutant cDNAs when using HTM to identify interesting mutants expressed by vertebrate cells in culture, there appear to be no technological barriers to doing so.

CLASSIFICATION OF SPECTRAL MUTANTS BY CHROMOPHORE TYPE

5.4 CLASSIFICATION OF SPECTRAL MUTANTS BY CHROMOPHORE TYPE Classification of spectral mutants of Aequorea GFP by chromophore type has provided a logical and useful guide to users of the GFPs since its introduction in a comprehensive form in 1998 (Tsien, 1998). Here we update the compendium to include unique mutants described since 1998. All known GFP variants may be divided into seven classes based on the distinctive component of their chromophores (Table 5.2): class 1, or wild type, is a mixture of neutral phenol and anionic phenolate; class 2 is a phenolate anion; class 3 is a neutral phenol; class 4 is a phenolate anion with stacked p-electron system; class 5 is an indole; class 6 is an imidazole; and class 7 is a phenyl. Each class has distinct wavelength distributions for their excitation and emission spectra (Table 5.1). The first four classes are derived from polypeptides with Tyr at Position 66, while classes 5–7 result from Trp, His, and Phe at position 66. The chromophore structures for each class are shown in Fig. 5.2 together with typical respective fluorescence spectra.

Class 1: Wild-Type Mixture of Neutral Phenol and Anionic Phenolate [Wild-Type Green Fluorescent Proteins (wtGFPs)] Wild-type GFP has the most complex spectra of all the Aequorea GFPs. The major excitation peak is at 395 nm and is about three times larger than the secondary peak at 475 nm. In normal solution, excitation of the major peak gives rise to an emission maximum at 508 nm whereas exciting the secondary, minor peak results in an emission maximum at 503 nm (Heim et al., 1994). The fact that the emission maxima are dependent on the excitation maxima indicates that there are at least two chemically distinct populations of chromophore that do not equilibrate within the lifetime of the excited state. Though the nature of the transition from one state to the other is not completely understood, the simplest explanation is that the minor peak at 475 nm arises from a deprotonated, anionic chromophore while the major peak at 395 nm represents a protonated or neutral chromophore (Heim et al., 1994; Cubitt et al., 1995; Niwa et al., 1996; Bell et al., 2000). A third, transitional, intermediate state has been identified using time-resolved fluorescence spectroscopy and hole burning (Chattoraj et al., 1996; Creemers et al., 1999; Zimmer, 2002), but exactly what this state represents and what the exact mechanism by which this occurs is still under investigation.

Class 2: Phenolate Anion in Chromophore [Green Fluorescent Proteins (GFPs)] GFPs of this class are perhaps the most commonly used due to their relative brightness, simple excitation, and emission spectra and the fact that the spectra match closely those of fluorescein. The prototypical GFP of this class incorporates the mutation S65T. This substitution results in a dramatic increase in the amplitude (five- to sixfold) and a red shift (from 470–475 nm to 489–490 nm) of the anionic peak while also dramatically suppressing the neutral phenol peak at 395 nm (Delagrave et al., 1995; Heim et al., 1995; Cheng et al., 1996). Formation of the mature chromophore in S65T (Heim et al., 1995) was about four times faster than in the wild type, and folding was fairly efficient at low temperatures as in the wild type at temperatures lower than 37°C. For this reason, a lot of effort was

89

399 (20) 399 (29) 399 (44)

488 (42) 487 (57.5)

488 (55–57)

489 (52–58)

395–397 (25–30) 470–475 (9.5–14) 397 (30) 475 (6.5–8.5)

lexc (e)b

Class 4, phenolate anion with stacked p-electron system (yellow fluorescent proteins-YFPs) -S65G, S72A, T203F — 512 (65.5) -S65G, S72A, T203H — 508 (48.5) -S65G, S72A, T203Y Topaz 514 (94.5) -S65G, V68L, S72A, T203Y 10C 514 (83.4) -10C + Q69K 10CQ69K 516 (62) -10C + F64L, M153T, V163A Venus 515 (92.2) S175G -10C + V68L, Q69M Citrine 516 (77)

H9 H9-40/Sapphire T-Sapphire

— Emerald

-F64L, S65T, V163A -S65T, S72A, N149K, M153T, I167T

Class 3, neutral phenol -S202F, T203I -T203I, S72A, Y145F -Q69M, C70V, V163A, S175G

EGFP

-F64L, S65T

Class 2, phenolate anion (green fluorescent proteins-GFPs) -S65T

Cycle 3

Wild Type

Class 1, wild type -None or Q80R

-F99S, M153T, V163A

Common Name

Mutationa

TABLE 5.2. Spectral Characteristics of the Major Classes of GFP Variants

(Cubitt et al., 1999) (Cubitt et al., 1999) (Cubitt et al., 1999) (Cubitt et al., 1999) (Cubitt et al., 1999) (Nagai et al., 2002; Rekas et al., 2002) (Griesbeck et al., 2001)

529 (0.76)

(Cubitt et al., 1999) (Zapata-Hommer & Griesbeck, 2003) (Zapata-Hommer & Griesbeck, 2003)

(Patterson et al., 1997; Ward, 1997; Cubitt et al., 1999) (Patterson et al., 1997; Ward, 1997; Cubitt et al., 1999) (Cubitt et al., 1999) (Cubitt et al., 1999)

(Patterson et al., 1997; Ward, 1997)

(Patterson et al., 1997; Ward, 1997)

Referenced

522 (0.70) 518 (0.78) 527 (0.60) 527 (0.61) 529 (0.71) 528 (0.57)

511 (0.60) 511 (0.64) 511 (0.60)

511 (0.58) 509 (0.68)

507–509 (0.60)

509–511 (0.64)

506 (0.79)

504 (0.79)

lem (QY)c

90 MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

442

(Cubitt et al., 1995)

(Cubitt et al., 1999) (Cubitt et al., 1999) (Patterson et al., 1997; Cubitt et al., 1999)

(Cubitt et al., 1999)

(Cubitt et al., 1999)

(Heim et al., 1994) (Cubitt et al., 1999)

Source: Ann. Rev. Biochem. a Substitutions from the primary sequence of GFP are given as the single-letter code for the amino acid being replaced, its numerical position in the sequence, and the singleletter code for the replacement. Note that many valuable mutants have been left out of this table for reasons of brevity and because quantitative spectral and brightness data were not available; therefore omission does not imply denigration. Phenotypically neutral substitutions such as Q80R, H231L have been omitted. b lexc is the peak of the excitation spectrum in units of nanometers. E in parentheses is the absorbance extinction coefficient in units of 103 M-1 cm-1. Estimates of extinction coefficients have tended to increase as expression and purification are optimized; obsolete older values have been omitted. Two numbers separated by a dash indicate a range of estimates from different authors working under slightly different conditions. Two numbers on separate lines indicate two distinct peaks in the excitation spectrum and have been designated as “shoulders” as they are also the minor of the two peaks. c lem is the peak of the emission spectrum in units of nanometers. QY in parentheses is the fluorescence quantum yield, which is dimensionless. The best figure of merit for the overall brightness of properly matured GFPs is the product of e and QY. See footnote to b for explanation of values. d References only for the quantitative spectral and brightness data. References to the origin and use of the mutants have been omitted for lack of space.

360

Class 7, phenyl in chromophore Y66F

485 476 (0.42) 505 (shoulder) 476 (0.40) 505 (shoulder) 495 (0.39)

448 (0.24) 446 (0.30) 440–447 (0.17–0.26)

436 434 (23.9) 452 (shoulder) 434 (23.9) 452 (shoulder) 435 (21.2)

Class 6, imidazole in chromophore (blue fluorescent proteins- BFPs) -Y66H BFP 384 (21) -Y66H, Y145F P4-3 382 (22.3) -F64L, Y66H, Y145F EBFP 380–383 (26.3–31)

Class 5, indole chromophore (cyan fluorescent proteins-CFPs) -Y66W — -Y66W, N146I, M153T, W7 V163A -F64L, S65T, Y66W W1B/ECFP N146I, M153T, V163A -S65A, Y66W, S72A W1C N146I, M153T, V163A

CLASSIFICATION OF SPECTRAL MUTANTS BY CHROMOPHORE TYPE

91

92

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

put into finding mutants that incorporated the benefits of S65T as well as having increased brightness at 37°C. Folding mutations that have worked well in combination with S65T include F64L (Cormack et al., 1996) and V163A (Kahana and Silver, 1996), V68L (Cormack et al., 1996), and I167T. Other mutations that improve folding used in combination with S65T have also proved useful. The photophysical mechanism by which replacement of Ser65 promotes chromophore ionization is reviewed in Tsien (1998) and Zimmer (2002).

Class 3: Neutral Phenol in Chromophore (GFP400 or Sapphire) In the class 2 mutants, the protonated state of the chromophore is favored and the anionic state is suppressed. Conversely, suppression of this protonation can be accomplished by mutating Tyr203 to Ile, a change that basically eliminates the excitation peak at 475 nm, leaving only the lower wavelength peak at 399 nm, with the emission still occurring at 511 nm. This mutant produces the greatest separation of the excitation and emission maxima for any of the Aequorea GFPs. This mutant called, sapphire [Turbo sapphire, GFP400, or H9-40 when combined with other mutations that improve the folding (ZapataHommer and Griesbeck, 2003)], has the greatest Stokes shift (greater than 100 nm) of any GFP mutant. This protein has proved to be the only published, successful FRET donor when dsRED was used as a FRET acceptor (Mizuno et al., 2001). The complexity of the excitation spectra of dsRED and the fact that it has been, until recently (Campbell et al., 2002), an obligate tetramer composed of individual subunits with spectra very similar to GFP has meant that only a donor such as sapphire, with such a large Stoke’s shift and blue-shifted excitation maximum, would do the job without unintentional cross-excitation of the immature, spectrally GFP-like, monomeric members of the dsRED tetramer. A full discussion of how the ionization state is influenced by the interior shape of the barrel of GFP and the sizes and shapes of the side chains of residue 203 can be found in Tsien (1998).

Class 4: Phenolate Anion with Stacked p-Electron System [Yellow Fluorescent Proteins (YFPs)] This class of chromophore, the furthest red-shifted of all the Aequorea mutants, results from stacking an aromatic ring next to the phenolate anion of the chromophore. Residue 203 is properly positioned to provide an aromatic side chain. To promote ionization of the chromophore, position 65 must be Gly or Thr instead of Ser. All four aromatic residues at position 203 (His, Trp, Phe, and Tyr—in order of least to greatest red-shifting potential) increase the excitation and emission spectra up to 20 nm (Ormo et al., 1996). See Table 5.3. All of them cause a red shift in both the exciation and emission spectra by adding polarizability around the chromophore and by extending the p–p interaction, thereby reducing the excited-state energy level. These replacements were rationally designed, using information from the crystal structures of S65T (Ormo et al., 1996; Yang et al., 1996b). Given that all three of the nucleotides of the codon for Thr (ACA) must be replaced in order to swap-in any one of the aromatic residues, it is unlikely that these substitutions would have been discovered by random mutagenesis. The crystal structure of the T203Y mutant (Wachter et al., 1998) confirms that the aromatic side chain of Tyr stacks next to the chromophore. Mutations at residue Q69 like Q69K (Cubitt et al., 1999) and Q69M (Heikal et al., 2000; Griesbeck et al., 2001) cause an additional 1- to 2-nm red shift, making these the “reddest” mutants of Aequorea GFP yet discovered.

93

CLASSIFICATION OF SPECTRAL MUTANTS BY CHROMOPHORE TYPE

TABLE 5.3. Effects of Different Aromatic Amino Acids at Position 203 of GFPa Mutations

lexc

e

lem

Comments and References

T203H T203H, S65G, S72A T203F, S65G, S72A T203Y, S65T T203Y, F64L, S65G, S72A T203Y, S65G, V68L, S72A

512 508 512 513 513 513

19.4

14.5 30.8 36.5

524 518 522 525 525 527

T203W, S65G, S72A

502

33

512

(Ormo et al., 1996) Cubitt (unpublished) (Dickson et al., 1997) (Ormo et al., 1996) (Ormo et al., 1996) (Ormo et al., 1996) quantum yield 0.64 (Ormo et al., 1996)

Abbreviations: lexc, excitation maximum in nanometers; e, extinction coefficient in M-1 cm-1 at_lexc; lem, emission maximum in nanometers.

a

The mutation T203H appears to be the crucial alteration within photoactivatable GFPs (Patterson and Lippincott-Schwartz, 2002), discussed later under photoisomerization.

Class 5: Indole in Chromophore Derived from Y66W [Cyan Fluorescent Proteins (CFPs)] An indole is formed in the chromophore when Tyr66 is replaced with a Trp (Y66W). The excitation and emission wavelengths are intermediate between the anionic phenolate and the neutral phenol-type chromphores, giving the protein a cyan appearance (see Table 5.2). The indole chromophore is bulky and requires a large number of additional mutations to restore the protein to reasonable brightness (Heim and Tsien, 1996). Many of these additional, beneficial mutations are the same as those required to restore brightness to the spectral mutants of other classes and are likely functioning in the same capacity. CFPs usually have two distinct peaks in their excitation spectra, two peaks in their emission spectra, and two lifetimes in their excited state decay. This microheterogeneity has been attributed to two distinct conformations visible in the crystal structure, associated with differing solvent exposures of Y145 and H148 ((Hyun Bae et al., 2003)). Therefore these two residues have been varied. The best improvement is the mutant S72A/Y145A/H148D, which has been dubbed “Cerulean” (Rizzo et al., 2004). Cerulean is reported to be 2.5-fold brighter than ECFP, due to improvements of about 1.5-fold in extinction coefficient and 1.7-fold in quantum yield. Cerulean is slightly more photostable than ECFP and shows single exponential kinetics for the decay of its excited state. Curiously, Cerulean retains the double-humped excitation and emission spectra of ECFP, suggesting that the molecular basis for this spectral shape remains somewhat mysterious.

Class 6: Imidazole in Chromophore Derived from Y66H [Blue Fluorescent Proteins (BFPs)] Exchanging Tyr66 for His puts an imidazole in the chromophore (Heim et al., 1994) and results in blue-shifting the excitation and emission wavelengths yet a bit more than replacement by the indole in class 5. As such, this class has been dubbed blue florescent proteins (BFP). Crystal structures (Palm et al., 1997; Wachter et al., 1997; Palm and Wlodawer, 1999) have been solved. As has been the case in the other classes, the brightness of these proteins is improved by additional mutations such as F64L (Patterson et al., 1997) and

94

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

Y145F (Heim and Tsien, 1996; Yang et al., 1998). Optimizing the codon usage to improve expression in mammalian cells was also beneficial (Yang et al., 1996b). However, even with these improvements, BFP remains among the least-used of the spectral mutants because it is easily photobleached and is dim [due to a very low fluorescence quantum yield (Rizzuto et al., 1996; Palm and Wlodawer, 1999; Kummer et al., 2002)], and the excitation wavelengths necessary to cause fluorescence are relatively more phototoxic and induce more tissue or cellular autofluorescence.

Class 7: Phenyl in Chromophore Derived from Y66F Replacement of Tyr66 with Phe (Y66F) results in the lowest known excitation (360 nm) and emission (442 nm) maxima (Cubitt et al., 1995). While this substitution illustrates nicely that any aromatic residue at position 66 can form a chromophore, it has been of little practical use. The described classification above (and Table 5.2) provides information concerning the fundamental chemical and photophysical nature of each fluorophore type. Ultimately it is direct manipulation, of these fluorophores or their environment, that is the goal of most mutagenic experiments. The following sections include descriptions of how certain amino acid changes in A. victoria GFP affect fluorophore behavior and ultimately how, why, and when it glows.

5.5 MUTANTS THAT COVALENTLY ALTER THE FLUOROPHORE: Y66FHW The most fundamental mutations that do not destroy fluorescence entirely are those that alter the covalent core structure of the chromophore. Such mutations are inevitably at position 66, because the crucial p-electron conjugated framework is derived from Tyr66 plus the invariant carbonyl carbon and amino nitrogen of residues 65 and 67, respectively. All the aromatic amino acids allowable by the genetic code—namely, phenylalanine, histidine, and trytophan—have been tested and found to generate fluorescent proteins of very different wavelengths (Heim et al. 1994; Heim and Tsien, 1996). The structures of the chromophores expected to be formed from these amino acids and the observed excitation and emission wavelengths are compared with wild-type GFP in Fig. 5.2. The rank order of shortest to longest excitation wavelengths is as expected from the electronic properties of the side chains. The benzene ring of phenylalanine has no electron donor group to conjugate to the electron-withdrawing carbonyl at the other end of the chromophore, so Phe gives the shortest wavelengths. The imidazole of His has a moderately electron-donating HN< group, but this is weakened by the electron-withdrawing =N— on the same ring. Next after imidazole is the nonionized phenol of tyrosine, which is responsible for the 395-nm excitation maximum of wild-type GFP. Yet more electron-rich is the >NH in the indole of tryptophan, which also has the largest conjugated system. The strongest electron-donating group is the phenolate anion of tyrosine, which is responsible for the minor excitation maximum of wild-type GFP at 470 nm. The chromophore with an ionized phenolate has about twice the extinction coefficient as that with a neutral phenol (Chattoraj et al., 1996). Therefore wild-type GFP, whose 395-nm peak is typically about three times as high as its 470-nm peak, is approximately a 6 : 1 mixture of neutral and anionic chromophores. This ratio has recently been confirmed by X-ray crystallography of wild-type GFP, in which two internal isomers corresponding to neutral and anionic chromophores

MUTANTS THAT COVALENTLY ALTER THE FLUOROPHORE: Y66FHW

in an 85 : 15 ratio can be resolved (Brejc et al., 1997). Five unnatural tyrosine-analog substitutions (Wang et al., 2003) have also been substituted at position 66 on the background of the Cycle 3 GFP variant (Crameri et al., 1996) as a showcase for incorporation of unnatural amino acids. Similarly, the spectral properties of these mutant GFPs, including the absorbance and fluorescence maxima and quantum yields, correlate with the structural and electron-donating ability of the substituents on the amino acids (Wang et al., 2003). Similar to most classes of GFP chromophore variants, practical applications of position-66 mutants have required additional amino acid substitutions to increase folding efficiency and fluorescence intensity (Heim and Tsien, 1996; Mitra et al., 1996; Siemering et al., 1996). The visual appearance of bacteria expressing first-generation improved versions of Y66H (blue) and Y66W (cyan) are shown in Fig. 5.3; normalized excitation and emission spectra are in Fig. 5.1. Based on crystallographic information obtained from the mutant Y66H (Palm et al., 1997; Wachter et al., 1997; Palm and Wlodawer, 1999), it was suggested that an increased flexibility around the chromophore led to a reduced quantum yield rather than poor folding efficiency as is often the case among the various classes. However, picosecond time-resolved fluorescence measurements from Y66F and

Figure 5.1. Fluorescence excitation and emission spectra for typical members of the major classes of GFP mutants together with the chromophore structures believed to be responsible for the spectra. The spectra have been normalized to a maximum amplitude of 1. When only one structure is represented in the figure, both excitation and emission spectra arise from the same state of chromophore protonation. The actual GFP proteins depicted are (a) wild type, (b) emerald, (c) H9-40, (d) topaz, (e) W1B, and (f) P4-3. The detailed substitutions within each of these variants are listed in Table 5.1.

95

96

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

Figure 5.2. Structures and fluorescence wavelengths of fluorophores resulting from different aromatic amino acids at position 66 of GFP. Each structure is labeled by the amino acid occupying position 66 before chromophore formation. Excitation (exc) and emission (em) peak wavelengths are given in nanometers. The native amino acid, Tyr, gives two peaks depending on whether the phenolic hydroxyl is neutral or ionized. For the latter, two of the possible resonance structures are drawn. Spectral data from Heim and Tsien (1996) and Cubitt et al. (1995).

Y66H indicate that while the free volume of the chromophore, cannot be discounted, the primary channel for loss of excitation energy is dependent more on hydrogen bond interactions between the chromophore and its immediate protein environment (Kummer et al., 2002). Heterocycle formation begins even with a nonaromatic residue such as Gly at 66 (Barondeau et al., 2003), although no useful fluorescence properties have been reported for such mutations.

5.6 MUTANTS WITH AN ALTERED RATIO BETWEEN THE TWO WT EXCITATION PEAKS: I167T, T203I, T203C, E222G, S65GACTVL, Q69L; CIRCULAR PERMUTATIONS, DURABLE PHOTOISOMERIZATION The next most fundamental mutations are those that alter the ionization of Tyr66 and change the ratio between the excitation peaks due to the neutral (~395 nm) and anionic (~470 nm) species. Mutation of Thr203 to Ile (T203I) suppresses the anionic peak (Heim et al., 1994; Ehrig et al., 1995; Zapata-Hommer and Griesbeck, 2003). This result is readily understandable from crystal structures because the hydroxyl of Thr203 donates a hydrogen bond to the chromophore phenolate (Ormo et al., 1996; Yang et al., 1996a; van Thor et al., 2002). In the neutral form of the chromophore, the hydroxyl of Thr203 rotates away from the protonated phenol (Brejc et al., 1997; van Thor et al., 2002). Ile cannot form a hydrogen bond, so the anionic phenolate would be destabilized relative to the neutral phenol. However, once the neutral chromophore absorbs light and reaches the excited state, the phenol ionizes (Chattoraj et al., 1996; Brejc et al., 1997; van Thor et al., 2002) despite

MUTANTS WITH AN ALTERED RATIO BETWEEN THE TWO WT EXCITATION PEAKS

the relatively poor solvation, so the subsequent emission still peaks at 511 nm. The most recent mutant published in this class is called Turbo or “T-Sapphire” (Zapata-Hommer and Griesbeck, 2003). The mutations Q69M/C70V/V163A/S175G were added on the background of H9-40 (see Table 5.2), resulting in a protein that remained insensitive to most cellular pHs (pKa 4.9), became fluorescent much faster, and recovered from denaturation more quickly and to a greater extent. Of these, V163A and S175G are common folding mutations and Q69M was known to suppress pH and halide sensitivity of YFPs (Griesbeck et al., 2001), C70V was a serendipitous, PCR-induced mutation that rescued the otherwise nonfluorescent H9-40 Q69M/V163A/S175G. Replacement of T203 with Cys (on the background S65T/H148G-“deGFP1” or S65T/C48S/H148C-“deGFP4”) results in ratiometric sensors in which the absorbance/excitation and the emission maxima vary in a pH-dependent way between the protonated (~400 nm) and the anionic (~508 nm) state of the chromophore (Hanson et al., 2002; McAnaney et al., 2002). Interestingly, these deGFPs appear to support excited-state proton transfer through a mechanism novel for a GFP that includes Ser147 and two water molecules. It permits rapid proton transfer between the chromophore hydroxyl and the bulk solvent at high pH. This network rearranges, and Ser147 is removed from contact with the chromophore, thereby eliminating the proton relay at low pH (Hanson et al., 2002). Many researchers have been interested in shifting the excitation to longer rather than shorter wavelengths, because ultraviolet excitation is potentially injurious to cells, excites more cellular autofluorescence, and generally requires more expensive optics and detection instrumentation. Shifts toward longer wavelengths require favoring the anionic peak. The first such mutants were I167T and I167V, which gave anionic excitation peaks about twice the amplitude of the neutral peaks (Heim et al., 1994). Given the ratio of extinction coefficients, these mutants probably contain about a 1 : 1 molar ratio of the two species. The crystal structure of GFP (Ormo et al., 1996; Yang et al., 1996a) allowed one to rationalize why Thr at 167 gives a moderately, but not overwhelmingly, higher ratio of anion to neutral than does wild-type Ile167. Thr should point its hydroxyl toward the phenolate without coming close enough to form a direct hydrogen bond. The dipole moment of the hydroxyl should therefore favor the phenolate. This rationalization, however, does not explain why Val at 167, which lacks the hydroxyl, produces much the same effect as Thr. Perhaps the smaller steric bulk of either Val or Thr relative to Ile is more important. An effective way of suppressing the neutral species was found to be mutation of Ser65 to any of a variety of small uncharged amino acids including Gly, Ala, Cys, Thr, or Leu (Heim et al., 1995). Such mutations cause essentially complete ionization of the chromophore even in the ground state and simplify the excitation spectrum to a single peak between 470 and 490 nm, with an amplitude about six times higher than that of the 475nm subsidiary peak of the wild-type protein. Thus at such blue excitation wavelengths, the mutants are about six times brighter per molecule than wild type. The improved brightness of such Ser65 mutants when excited in this blue region is understandable, because the wild-type protein was only about one-sixth ionized. The exact positions of the excitation and emission maxima depend on which amino acid replaces Ser65, as shown in Table 5.4. Ala gives the shortest wavelengths, Thr or Gly the longest. Very polar or bulky residues such as Arg, Asn, Asp, Phe, or Trp at position 65 do not seem to be tolerated (Heim et al., 1995). The mutant with Thr at 65, or S65T, was chosen by Heim et al. (1995) for further analysis and use because it combined the longest peak wavelengths (489 nm excitation, 511 nm emission) with the most conservative substitution. The visual appearance of bacteria expressing S65T is included in Fig. 5.3, and its excitation and emission spectra are

97

98

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

TABLE 5.4. Effects of Different Amino Acids at Position 65 of GFPa Amino Acid

lexc

e

lem

Q

Ser (wt)

395 475

21,000 7,150

508 503

0.77

Ser (wt)

404 471

504 482

Ala Ala

471 481

503 507

Cys Leu Thr Thr Thr

479 484 489 488

507

Gly

490

505

Gly

501

511

Arg, Asn, Asp Phe, Trp

47,400 39,200 52,900

507 510 511

Comments and References At room temperature (Heim et al., 1995); the 395- and 475-nm peaks probably represent excitation of the neutral and anionic fluorophore respectively At 77K in 1 : 1 glycerol-water to slow down proton transfers (Chattoraj et al., 1996) (Heim et al., 1995) Measured in combination with V68L, S72A (Cormack et al., 1996) (Heim et al., 1995) R. Heim (unpublished data) (Heim et al., 1995) R. Heim (unpublished data) Measured in combination with F64L (Cormack et al., 1996) Measured in combination with F64M, Q69L (Delagrave et al., 1995) Measured in combination with S72A (Cormack et al., 1996) Weak or no fluorescence (Heim et al., 1995)

a Abbreviations: lexc, excitation maximum in nanometers; e, extinction coefficient in M-1 cm-1 at lexc; lem, emission maximum in nanometers; Q, fluorescence quantum yield. Discrepancies of 2–3 nm between different laboratories in estimating lexc and lem are probably not significant.

shown in Fig. 5.1. S65T also proved to have at least two other advantages over wild type: (1) a fourfold increase in the rate of the final oxygen-dependent step in the fluorophore formation and (2) a considerable increase in photostability (Heim et al., 1995). The increased photostability of S65T is manifest in two ways. Wild-type GFP, when irradiated at wavelengths short enough to excite the neutral species, tends to photoisomerize toward the anionic species (Cubitt et al., 1995; Chattoraj et al., 1996). Therefore, excitation of the 395-nm peak rapidly diminishes its amplitude and boosts that of the 470-nm peak (Chalfie et al., 1994). This conversion involves transfer of a proton from the neutral chromophore to Glu222 (Brejc et al., 1997; van Thor et al., 2002) and can partially reverse upon standing in the dark over the time scale of hours to days (Chattoraj et al., 1996). Since the fluorophore of S65T is already entirely anionic, there is no way for photoisomerization to push the ionization any further. In addition to the wild-type photoisomerization, both wild-type and mutant GFPs irreversibly photobleach. The reciprocal interaction between Glu222 and the fluorophore predicts that if Glu222 were replaced by a nonionizable group, the fluorophore would be permanently anionic. Indeed, Ehrig et al. (1995) had found E222G by random mutagenesis and visual screening and had shown that its only excitation peak is at 481 nm, consistent with full ionization of the fluorophore. Delagrave et al. (1995) combinatorially mutagenized positions 64–69 and screened colonies by imaging spectroscopy. They found six mutants, which they described as “redshifted” because their excitation spectra peaked near 490 nm, while the position of the emission maxima were largely unchanged. These mutations are now interpretable as

MUTANTS WITH AN ALTERED RATIO BETWEEN THE TWO WT EXCITATION PEAKS

Figure 5.3. Visual appearance of E. coli expressing four differently colored mutants of GFP. Clockwise from upper right: Blue mutant P4-3 (= Y66H, Y145F) (Heim and Tsien, 1996); cyan mutant W7 (Y66W, N146I, M153T, V163A, N212K) (Heim and Tsien, 1996); green mutant S65T (Heim et al., 1995); yellow mutant 10C (= S65G, V68L, S72A, T203Y) (Ormo et al., 1996). In each of these lists of mutants, the mutation most responsible for the special alterations is underlined, while the other substitutions improve folding or brightness. The bacteria were streaked onto nitrocellulose, illuminated with a Spectraline B-100 mercury lamp (Spectronics Corp., Westbury, NY) emitting mainly at 365 nm, and photographed with Ektachrome 400 slide film through a low-fluorescence 400 nm and a 455-nm colored glass long-pass filter in series. The relative brightness of the bacteria in this image is not a good guide to the true brightness of the GFP mutants. Expression levels are not normalized, and the 365 nm excites the blue and cyan mutants much more efficiently than the green and yellow mutants, but the blue emission is significantly filtered by the 455-nm filter required to block violet haze. See color insert.

emphasizing the anionic fluorophore at the expense of the neutral form. Many biologists have misunderstood the term “red-shifted” to imply that the fluorescence is actually red, but in fact the emission spectra are not significantly changed from the normal green emission of wild-type protein, peaking at 505–510 nm. Five out of six of these mutants had Gly, Ala, Cys, or Leu at position 65, which are undoubtedly the key substitutions responsible for the shift in excitation spectra. Photoisomerization of GFP was discovered just after the cDNA for GFP was cloned (Chalfie et al., 1994; Cubitt et al., 1995). Exposure to light at ~400 nm causes photoisomerization in wild-type GFP that involves a shift in the chromophore population from the protonated form to the unprotonated. van Thor et al. (2002) have provided evidence for a two-step mechanism to explain this conversion. The first step is decarboxylation of Glu222 followed by structural rearrangements of Thr203 and His148. The most dramatic example

99

100

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

of photoconversion of Aequorea GFP being useful in cell biology was described by Patterson and Lippincott-Schwartz (2002). The mutation T203H (on a wild-type GFP background) reduces the absorbance/excitation peak at 490 nm to an even greater extent that in the T203I mutant. Illumination of this mutant called photoactivatable GFP (PAGFP) with laser light at 413 nm caused a dramatic shift in the excitation spectrum from the peak at 400 nm to the peak at 490 nm. The optical contrast between the photoactivated and nonphotoactivated PA-GFP was up to 100-fold when excited with laser light at 488 nm (60-fold in cells) and was stable at 37°C for at least 1 week. This mutant is potentially very valuable for tracking the location and fate of a certain subpopulation of activated proteins (Patterson and Lippincott-Schwartz, 2002). An even more dramatic example of photoactivation occurs in the coral protein Kaede (Ando et al., 2002), in which UV illumination changes the emission from green to red due to a photochemically driven betaelimination between the GFP-like chromophore and the imidazole within an immediately adjacent histidine (Mizuno et al., 2001). Circular permutants (cp) of EYFP and EFYP with calmodulin inserted at position 145 (Camgaroo) have two absorbance peaks (396–404 and 490–496 nm) corresponding to the two major excitation peaks of wild-type GFP (Baird et al., 1999). In both cases, and unlike GFPs, excitation of the most blue peak (~400 nm) does not result in fluorescent emission. Insertion of calmodulin at position 145 of YFP modulates, in a Ca2+-dependent manner, the relative magnitudes of these two absorption maxima; Ca2+ increases the protonated (~400 nm) peak while dramatically diminishing the nonprotonated (~496 nm) peak, while increasing Ca2+ at the same pH reverses the relative absorption profile and increases the fluorescence seven- to eightfold when excited at ~496 nm. The conformational change in calmodulin presumably torques the staves of the beta barrel in such a way that in the Ca2+bound state, the chromophore is more susceptible to quenching by acid (Fig. 5.2). These will be discussed in further detail later. The only other mutant with an increased ratio of anionic to neutral excitation peaks that cannot be explained by the above substitutions of S65 and E222 is “RSGFP1” of Delagrave et al. (1995), which is F64G, V68L, Q69L. Mutations of F64 and V68 are known to affect folding efficiency (see below) but not wavelengths, so Q69L is most likely responsible for the increased ionization of the fluorophore. The crystal structure (Ormo et al., 1996) shows that Q69 anchors a cluster of water molecules that also participate in solvating the Glu222 carboxylate. Replacement by Leu would disrupt this hydrogen-bonding network and destabilize the carboxylate anion more than the neutral carboxyl of Glu222, thereby indirectly promoting fluorophore ionization.

5.7

MUTANTS THAT MORE SUBTLY MODIFY THE ENERGY LEVELS

A few mutants shift both the excitation and emission wavelength distributions. Obviously the major effort has been directed toward obtaining mutants with longer rather than shorter wavelengths. It is this class of mutants that can most accurately be termed “red-shifted.” One of the best understood mutants is T203Y, which was designed rationally when the crystal structure of GFP (Ormo et al., 1996) revealed the close proximity of Thr203 to the fluorophore. Replacement of that aliphatic residue by aromatic residues next to the fluorophore was intended to increase the local polarizability of the chromophore. Starting from S65 mutants that were already fully ionized, T203Y increased the excitation and emission maxima by 24 and 16 nm, respectively, to 513 and 527 nm, respectively (Ormo et al., 1996). These wavelengths, which are the longest so far published for an Aequorea GFP

MUTANTS THAT IMPROVE FOLDING AT 37°C

mutant, permit detection through at least some standard filter sets for rhodamines—for example 510- to 560-nm excitation, 565-nm dichroic, 572- to 647-nm emission. Although pure 527 nm is itself still green, the long tail of emission at yet longer wavelengths makes the emission yellowish to the eye and distinguishable from S65T in side-by-side comparisons. Figure 5.3 shows such a comparison captured on color film. The hue difference is somewhat more impressive in real life than in this picture. T203F and T203H were almost as effective (Table 5.3), whereas T203W produced less of a shift, perhaps because its steric bulk was excessive or because dipole moments of the new substituents are playing contributory roles. Another mutant of even smaller and less easily explained effect is M153A, which increases the excitation and emission wavelengths of S65T by 15 and 3 nm, respectively (Heim and Tsien, 1996).

5.8 MUTANTS THAT IMPROVE FOLDING AT 37°C A large class of mutations improves the percentage yield of GFP molecules that fold correctly and become fluorescent without seeming to affect the spectral properties of those properly matured proteins. The yields of fluorescent protein expressed from the original jellyfish gene fall steeply as the temperature increases above about 15–20°C, which is not surprising given the low temperature of Puget Sound, where Avictoria victoria lives. Also, high-level expression in E. coli tends to give extensive deposition of nonfluorescent protein in inclusion bodies, in which the chromophore has not even formed (Siemering et al., 1996). Both of these problems can be greatly ameliorated by suitable amino acid substitutions, which have been found mostly by random mutagenesis and visual or flowcytometric screening for brighter bacteria at temperatures up to 37°C. Many have been arrived at independently by several groups (Cormack et al., 1996; Crameri et al., 1996; Heim and Tsien, 1996; Kahana and Silver, 1996; Siemering et al., 1996; Kimata et al., 1997; Nagai et al., 2001), which may imply that existing mutational strategies are approaching saturation. In our opinion, all new GFP constructs should incorporate some subset of mutations that improve maturation because they do no harm and can often produce large increases in brightness. The only exception would be experiments in which one desires to shut off the formation of newly fluorescent GFP molecules by raising the temperature (Kaether and Gerdes, 1995; Lim et al., 1995). Note that even wild-type GFP is fairly heat-stable once properly folded and matured (Lim et al., 1995), becoming denatured only above 65°C (Ward and Bokman, 1982); only during the folding process is it highly temperaturesensitive (Siemering et al., 1996). Although folding mutations are quite valuable, one should not expect indefinite further improvements in GFP brightness. The ultimate brightness of any fluorophore is limited at the molecular level by the product of extinction coefficient and fluorescence quantum efficiency. Folding mutations do not significantly improve the brightness of GFP molecules that are correctly folded (Siemering et al., 1996), and unfortunately few laboratories characterizing GFP mutants have reported extinction coefficients or quantum yields along with their studies. The fact that the brightness of bacteria expressing very high levels of GFP can be raised by a factor of 20 to 50 (Cormack et al., 1996; Crameri et al., 1996; Siemering et al., 1996) is more indicative of the wretched folding efficiency of the wildtype protein at high temperature and concentrations; at lower temperatures or expression levels—that is, when one wants to be able to detect as few GFP molecules as possible— the improvement due to folding mutations is not as impressive. Furthermore, the improve-

101

102

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

ment factors due to individual mutations do not multiply: Two mutations that separately improve folding efficiency by factors of x and y generally give much less than an improvement of xy when combined (Zapata-Hommer and Griesbeck, 2003). Folding efficiency can asymptotically approach but can never exceed 100%. The existing crystal structures do not clearly explain how most of the folding mutations exert their favorable effects. Many of the mutations may only make a difference while the protein is still relatively disordered and in an unfavorable environment, whereas the crystals are typically grown from mature protein produced under conditions chosen to maximize successful expression. In a few cases, such as F99S and M153T, for which the side chains face outward, the mutations reduce patches of surface hydrophobicity, possibly inhibiting aggregation and enhancing brightness. However, these mutations did not alter the overall speed of fluorescence development at 37°C compared to wild type (Crameri et al., 1996). Mutations V163A and S175G together actually slow the final aerobic development of fluorescence (Siemering et al., 1996), even though they greatly improve the yield of properly matured protein. The folding and unfolding reactions of Cycle3 (F99S, M153T, V163A) and wild-type GFP were compared using fluorescence and circular dichroism spectroscopy and measurement of hydrogen exchange reactions using Fourier transform infrared spectroscopy (Fukuda et al., 2000). The results illustrate that while folding and unfolding were relatively slow processes, they were basically the same for both proteins and that the decreased exterior surface hydrophobicity of Cycle3 contributed more to the general increase in brightness of Cycle3 than did the rates of folding. Similarly, Miyawaki and colleagues (Nagai et al., 2002) identified a novel mutation (F46L) that decreased the time required to form mature, fluorescent YFP. When they looked closely at the rates of the folding and the oxidation of the chromophore, the mutation F46L slowed the rate of folding slightly, compared to YFP without F46L, but increased the rate of oxidation of the chromophore so that the resulting maturation time was reduced (Nagai et al., 2002). Griesbeck et al. (2001) reported a series of mutations Q69M/C70V/V163A/ S175G that improved the folding of GFP400 or Sapphire. The fact that Q69M has some beneficial effects on YFP had been reported (Griesbeck et al., 2001) previously, but C70V, which was the fortunate result of a PCR error, turned out to be crucial because Sapphire with Q69M/V163A/S175G was nonfluorescent. F46L, F99S, and M153T were found not to improve the overall efficiency of folding or brightness of Sapphire. An understanding at the molecular level of how other folding mutations work, especially those affecting buried side chains, may require detailed comparison of the resulting final crystal structures with the baseline structures already on hand, as well as continued investigations of the folding intermediates and dynamics.

5.9

MUTATIONS THAT MODULATE AGGREGATION

Currently, Aequorea GFP is the only fluorescent protein known that is not an obligate homo-oligomer in its natural state; the GFP from Renilla is an obligate dimer (Ward and Cormier, 1979; Ward, 1998) and RFP (dsRED from coral) is an obligate tetramer (Baird et al., 2000) (as are all other characterized fluorescent proteins form coral species see other chapters in this volume). Quite some time before a rigorous determination of the homoaffinity was made, it was known that even Aequorea GFP dimerized to some degree in solution (Yang et al., 1996b). GFP crystallizes as either a dimer (Yang et al., 1996a; Palm et al., 1997; Battistutta et al., 2000) or a monomer (Ormo et al., 1996; Brejc et al., 1997; Wachter et al., 1997; Wachter et al., 1998; Wachter et al., 2000). In the dimeric

103

MUTATIONS THAT MODULATE AGGREGATION

crystal structure, the unit cell consists of two monomers associated in a slightly-twisted, head-to-tail fashion via many hydrophilic contacts as well as several hydrophobic contacts. The very different solvent conditions used by each group are sufficient to explain the differing results (Ward, 1998). Residing within a large hydrophobic patch, residues A206, L221 and F223 (Fig. 5.4) are sufficient to cause formation of the dimer at relatively low concentrations in solution and in living cells. However, changing these residues singly or in combination to positively charged residues such as A206K, L221K and F223R (Fig. 5.4; Table 5.5), effectively eliminated the interaction of the monomers (Zacharias et al., 2001); the resulting monomeric GFPs have been termed mGFPs. To determine the strength

Figure 5.4. The crystal structure of dimeric GFP (1GFL) (Yang et al., 1996a). The residues A206 (red), L221 (orange), and F223 (lavender) are shown as ball-and-stick representations. Replacing any of these residues with the positively charged residues lysine or arginine effectively monomerizes the protein. See color insert.

TABLE 5.5. Mutations that Eliminate Dimerization of GFPs Mutation wild type L221K F223R L221K, F223R A206K

Quantum Yield (Ward, 1998)

Extinction Coefficient1

Kd2

0.67 0.67 0.53 0.68 0.62

67 64 65 59 79

0.11 9.7 4.8 2.4 743

Fluorescence and dissociation of wild-type YFP versus monomeric YFP. 1 Extinction coefficient ¥ 1000 in M-1 cm-1 at # lexc. 2 The dissociation constant Kd (mM) as measured by sedimentation equilibrium analytical ultracentrifugation (Zacharias et al., 2002). 3 Due to the extreme monomeric nature of this protein it was difficult to determine an accurate dissociation constant for a hypothetical dimer.

104

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

of the interaction in solution, Zacharias et al. (2001) used sedimentation equilibrium analytical ultracentrifugation to characterize the affinity of GFPs with the wild-type interface as well as the mGFPs. In significant contrast to X-ray crystallography the experimental conditions used in the analytical ultracentrifugation experiments approximates cellular physiological conditions and were able to provide definitive information (McRorie and Voelker, 1993) about the affinity of the complex. The dimer dissociation constants for wtGFP and several of the mutants are compiled in Table 5.5. Other mutations are also thought to affect the state of GFP aggregation. F99S and M153T first described in relation to aggregation by Stemmer and colleagues (Crameri et al., 1996) reduce obvious patches of surface hydrophobicity and could inhibit aggregation, but no dissociation constant has yet been determined for Cycle3. Indeed, the triple mutant (F99S, M153T, V163A) (Crameri et al., 1996) has a diffusion coefficient inside mammalian cells one order of magnitude higher than that of wild-type GFP, implying a corresponding reduction in binding to other macromolecules (Yokoe and Meyer, 1996). Because V163 points into the interior of the protein (Ormo et al., 1996) and because F99S and M153 face outwards, the latter two are most likely the culprits in wild-type GFP. However, the triple mutation did not alter the overall speed of fluorescence development at 37°C compared to wild-type GFP (Crameri et al., 1996). Mutations V163A and S175G together actually slow the final aerobic development of fluorescence (Siemering et al., 1996), even though they greatly improve the yield of properly matured protein. The crystal structure of cyclized Cycle3 GFP was determined (Hofmann et al., 2002). The authors found a crystallographic dimer interface different than those previously reported and concluded that various polar and nonpolar patches on the surface of GFP could serve to dimerize the proteins in ways previously undescribed. However, it is unlikely that the dimer interface described in this work exists under physiological conditions with great affinity in non-Cycle3 versions of GFP because it is clear from sedimentation equilibrium analytical ultracentrifugation experiments on GFP containing A206K (and other mutations) (Zacharias et al., 2001) that there is virtually no remaining affinity when the more commonly observed dimer interface is altered. Further biophysical characterization of Cycle3 and mutants derived from it is clearly warranted. The issue of GFP oligomerization is significant for several reasons. Most of the potential for trouble arises when GFP or its spectral mutants are fused to other proteins to track protein localization or expression or to measure interactions by fluorescence resonance energy transfer (FRET) (most often CFP and YFP). If GFP dimerizes in the context of being part of a fusion protein, it could also foreseeably dimerize the protein to which it is fused. The situation could become even “stickier” if the host protein is itself an oligomer. When measuring the interactions of molecules by FRET, the fluorophores used to report the interactions must not themselves in any way influence—or worse yet, create—the interactions being measured. Obviously, if the fluorophores have affinity for each other, then doubt is cast on the accuracy of any measurement made to the presumed interaction of the host proteins. The problems associated with GFP dimerization are most troublesome when measuring intermolecular FRET in a two-dimensional space such as a membrane (Fung and Stryer, 1978; Wolber and Hudson, 1979; Dewey and Hammes, 1980; Snyder and Freire, 1982; Dewey and Datta, 1989; Yguerabide, 1994; Zimet et al., 1995). In this situation, we found that wtGFPs were very likely to dimerize even when expressed at very low surface densities (Zacharias et al., 2001). Since the monomerizing mutations alter nothing but the homoaffinity of GFP, we recommend including them (preferably A206K) in all GFP expression constructs where dimerization is not desirable.

TRUNCATIONS OF GFP

Another interesting phenomenon associated with dimerization of type I (wild-type) GFPs is that when they are dimerized, there is a suppression of the ~500-nm peak in the absorbtion/excitation spectrum and a concomitant but relatively smaller increase in the ~400-nm peak (Morise et al., 1974). In GFPs, a dimer-induced shift of the excitation spectrum ~100 nm toward the blue (neutral form of the chromophore) will result in off-peak excitation (unless one changes to appropriate filters) and much lower fluorescence output. In YPF, excitation near the 400-nm peak yields a nonfluorescent species. Additionally, for FRET-related experiments a shift in the absorbance of YFP (acceptor) toward the ~400-nm peak will result in decreased spectral overlap with the emission spectrum of a CFP (donor) and therefore reduce the efficiency of FRET. This particular phenomenon was exploited to make non-FRET-based sensors for molecular proximity (De Angelis et al., 1998). It has been a point of curiosity as to why GFP and FPs in general should oligomerize. It has been proposed, in the case of nonbioluminescent corals (and likely the anemones; also anthozoans), that FPs are acting as a protective barrier between the harmful UVA irradiation from the sun and the resident symbiotic organisms that produce energy by photosynthesis (Salih et al.,2000). Corals living closer to the surface of the water are exposed to greater amounts of UVA irradiation from the sun, and it is in these Salih and co-workers found that these corals had the greatest abundance and diversity of FPs. They went on to show that the FP complexes dissipate excess energy at wavelengths of low photosynthetic activity and also serve to reflect visible and infrared light. The complexes are able to do this is by capturing light at the bluer, more damaging end of the spectrum and to shuttle it by FRET to red-shifted (FRET acceptors), oligomeric, partner proteins, finally “spilling” the energy in a range that is not compatible with photosynthesis. In this view, oligomerization between FPs with different chromophores (Gross et al., 2000; Salih et al., 2000; Cotlet et al., 2001; Dove, 2001; Garcia-Parajo et al., 2001) serves to maximize FRET (Zacharias, 2002). However, this type of explanation is less appropriate for bioluminescent sea pansies and jellyfish. In their native species, these FPs serve as acceptors for luciferases, and there are no known spectral differences among FPs expressed in the same organism, so oligomerization does not promote intersubunit FRET. A speculative alternative comes from the observations that resistance to photobleaching of DsRed mutants decreased as the stoichiometry was reduced from tetrameric to dimeric to monomeric (Campbell et al., 2002), and that the most monomeric FP (Aequorea) comes from 49°N latitude, obligate dimeric FPs (Renilla) from 31–33°N latitude, and obligate tetramers from corals in tropical waters. Perhaps oligomerization protects the FPs themselves from photobleaching by decreasing surface-to-volume ratio, hindering access of oxygen to the chromophores. The higher the intensity of solar irradiation, the more selection pressure for photostability. Water temperature also is inversely correlated with latitude, and thermostability and photostability requirements run in parallel. Further characterization of more related FPs from geographically disparate sources should help to clarify the reasons for these interesting distinctions (Zacharias, 2002).

5.10

TRUNCATIONS OF GFP

A frequently asked question about GFP is whether it can be significantly reduced in size. In the two most systematic investigations of this possibility, Dopf and Horiagon (1996) and Kim and Kaang (1998) produced a family of genetic truncations from the N- and C-

105

106

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

termini. The N-terminal methionine could be replaced by a polyhistidine tag (Dopf and Horiagon, 1996), but deletion of residues 2–8 prevented fluorescence or chromophore development. The C-terminus was slightly more tolerant, in that 6 (Dopf and Horiagon, 1996–9) (Kim and Kaang, 1998) but not 13 (Dopf and Horiagon, 1996; Kim and Kaang, 1998) residues could be eliminated. These narrow limits are in good agreement with the crystal structure (Ormo et al., 1996), in which Met1 and residues 230–238 are too disordered to be located accurately. There are no major regions of internal sequence that appear dispensable. The core domain of GFP, residues 2–229, seems to be a monolithic entity required to shield the fluorophore from solvent. A few local loops might conceivably be shortened somewhat, but the slight net reduction in size would hardly be worth the effort. Fortunately, it seems possible to target full-length GFP to essentially any compartment in the cell, so its size has not yet severely restricted its versatility. A report (Wiedenmann et al., 2000) of a coral chromoprotein with only 148 amino acids proved to be an artifact of incorrect DNA sequencing combined with cleavage of the peptide backbone during chromophore formation (Martynov et al., 2001; Wiedenmann et al., 2002).

5.11

GFP BIOSENSORS

The crystal structure of GFP gives us a picture of a protein that appears to be rather monolithic, seemingly without much tolerance for gross rearrangements, insertions, or deletions. However, during some semirandom mutagenesis experiments, Baird et al. (1999) made the serendipitous discovery that a six-residue peptide (FKTRHN) had been inserted into CFP at position 145 without destroying its fluorescence (e.g., Fig. 5.5B). This suggested that the two halves of CFP had the capability to fold autonomously and that circular permutations generated by fusing the original N- and C-termini and breaking the main chain backbone at other locations might be tolerated (Fig. 5.5C). A systematic procedure (Graf and Schachman, 1996) to find other potential break points uncovered 10 sites within CFP that could serve as new N- and C-termini while retaining fluorescent proteins (Fig. 5.6). Another report (Topell et al., 1999) that followed closely on the heels of the report by Baird and colleagues found 20 possible breakpoints, but among them only five retained significant fluorescence. When new termini are generated at position 145, a minimum of 3 residues must form the linker between the wild-type termini to reestablish fluorescence, and linkers of four to seven residues resulted in restoration of fluorescence to the same level as EGFP (Akemann et al., 2001). Aside from their sheer novelty, such circular permutations and insertions have proved to be useful in generating single-fluorophore, GFP-based sensors for Ca2+ (Nagai et al., 2001) and for changing the orientation of a fluorophores to try to improve the a FRET signal (Baird et al., 1999). GFPs have also been completely circularized by split intein technology (Hofmann et al., 2002). Some of the same breakpoints identified in circularly permuted GFP can also be locations to serve as acceptor sites into which other proteins can be fused (Fig. 5.5B). The first such insertion of a whole, functional protein was Xenopus calmodulin into position 145 of YFP (Baird et al., 1999), resulting in a calcium sensor named Camgaroo which has proven useful in measuring calcium inside of mitochondria and the mushroom bodies in the brain of Drosophila (Yu et al., 2003). Other groups combined insertions with circular permutations of EGFP (Nakai et al., 2001) or EYFP (V68L/Q69K) (Nagai et al., 2001) and created a multimember suite of calcium sensors each with a different fluorometric readout for calcium. One sensor for Zn2+, with an apparent Kd of 400 mM made by inserting the zinc finger motif from zif268 into YFP at position 145, gave a maximum response

GFP BIOSENSORS

Figure 5.5. GFP biosensors. (A) GFP can be engineered to be directly sensitive to a small molecule of interest. (B) Insertion of a conformationally dynamic domain into GFP can result in a chimera in which the fluorescence properties of GFP are modulated by a change in conformation of the domain. (C) Similarly, proteins or peptides with dyanmic, associative properties can be fused to the N and C termini of circularly permuted GFPs, thereby reporting on the changes in the association in response to a stimulus. See color insert.

of about 1.7-fold change in fluorescence intensity (Baird et al., 1999); another group engineered Zn2+-binding sites directly into the chromophore of BFP (Richmond et al., 2000). The resulting protein had about a twofold change in fluorescence intensity with a Kd for Zn2+ of 50 mM. The slow on-rate for Zn2+ for the second probe (T1/2 of >4 hours) limits its usefulness in cell biological experiments. Similarly, attempts have been made (Richmond

107

108

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

Figure 5.6. A backbone representation of the three-dimensional structure of GFP (1EMG) (Elsliger et al., 1998). The residues where circular permutations are permitted while retaining fluorescence are color highlighted. E142, hot pink; Y143, gray; Y145, dark blue; H148, fuchsia; D155, yellow; H169, red; E172, light blue; D173, orange; A227, Cyan; I229, light purple. These residues represent sites where the main chain can be interrupted. In most cases, resumption of GFP sequence can occur one to four residues following the initial interruption. See color insert.

et al., 2000) to rationally design sites, on the external surface of the barrel (Fig. 5.5A) of YFP (10C) that could bind to metals. Mutants on the 10C background (S147H/Q204H and S202D or F223E) were generated that successfully quenched YFP fluorescence when exposed to various divalent cations, but nothing was reported of the selectivity concerning the ability of these mutants to discriminate among the cations. It would be ideal if robust, selective sensors of such simplicity could be generated for cellular analytes. Similar to the case where Zn2+ sites were engineered directly into the chromophore, individual GFPs with only small modifications can be made into sensors for pH and halides (Fig. 5.5A).

5.12

pH

High pH, 11–12, causes a relative redistribution of the two absorbance/excitation maxima of wtGPF toward the longer wavelength (470 nm) peak while low pH (pKa ~ 5.5) causes a quenching of fluorescence (Bokman and Ward, 1981; Ward et al., 1982). A fluorophore with sensitivity to variations in pH is often viewed as a lemon, but in the case of GFPs, the lemon has been made into some very informative lemonade. While it is good to have a fluorescent probe that emits stably over a broad range of pHs, it was also apparent that the pH-dependent change in fluorescence behavior could be exploited to measure pH in

109

HALIDES

TABLE 5.6. pka Values of Gfps Used as ph Indicators FP

pKa

Mutations

Reference

GFP YFP GFP GFP GFP GFP CFP

4.8 5.7 5.9 6.0 6.1 6.15 6.4

(Kneen et al., 1998) (Griesbeck et al., 2001) (Kneen et al., 1998) (Kneen et al., 1998) (Kneen et al., 1998) (Llopis et al., 1998) (Llopis et al., 1998)

YFP GFP GFP GFP

7.1 7.3 8.0 8.0

T203I Q69M S65 F64L, S65T Y66H F64L, S65T, H231L K26R, F64L, S65T, Y66W, N146I, M153T, V163A, N164H, H231L S65G, S72A, T203Y, H231L S65T/C48S/H148C/T203C S65T, H148G, T203C H148G

(Llopis et al., 1998) (Hanson et al., 2002) (Hanson et al., 2002) (Maysuyama et al., 2000)

FP indicates the spectral mutant upon which the indicated mutations were incorporated. pKa is the pH value at which 50% of the molecules are fluorescent.

subcellular domains (Llopis et al., 1998). pKa values range for GFPs vary widely covering the range of most cellular pHs (Table 5.6). pHluorins comprise a set of pH-sensitive GFPs used to monitor exocytosis (Miesenbock et al., 1998; Sankaranarayanan et al., 2000). These sensors are targeted to the luminal side of secretory vesicles where the pH is below the pKa of pHluorin (which causes them to be nonfluorescent). Upon fusion, exocytosis, and exposure of the membrane-associated pHluorin to the extracellular milieu held at a desired physiological pH, the pHluorin becomes fluorescent (ecliptic pHluorin) or shifts its excitation maximum from 395 nm to 475 nm (ratiometric pHluorin). The broad range of pKa values for the fluorophores is generated by the diversity mutations in and around the fluorophore. Many of the physical reasons for the various pH-sensitive behaviors are summarized in a theoretical study (Scharnagl et al., 1999) that incorporated a broad range of existing physical data from many experimental sources. In YFP (class 4 chromophores) the mutation Q69M (named “Citrine”) retains virtually identical excitation and emission spectra but lowers the pKa of the chromophore to 5.7, renders it insensitive to chloride, increases the photostability over previous versions of YFP by about twofold, and improves expression at 37°C in cells (Griesbeck et al., 2001). In the crystal structure of Citrine, the Met at position 69 is well-ordered, tightly packed into the cavity, and unlikely to be able to undergo the same sort of conformational change that is seen with the apo- and iodidebound forms of EYFP.

5.13

HALIDES

Wachter and Remington (1999) first reported halide sensitivity of a YFP (S65G/V68L/ S72A/T302Y and H148Q or H148G). Shortly thereafter, they and Verkman’s group made more detailed biophysical characterization of the nature of the Cl- sensitivity of YFP (T203Y/S65G/V68L/S72A/H148Q) and made the first steps toward developing it as the first genetically encoded Cl- sensor (Jayaraman et al., 2000). Almost simultaneously, Kuner and Augustine (2000) discovered the Cl- sensitivity of YFP while using CFP (K26R/F64L/S65T/Y66W/N146I/M153T/V163A/N164H/H231L) and YFP (S65G/S72A/ K79R/T203Y/H231L) to study protein–protein interactions by FRET. They fused CFP and

110

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

YFP with a short intervening peptide linker (ENLYFQG) to make a protein that, in the resting state, had a high degree of FRET, and they called it “Clomeleon.” When Cl- bound to YFP, quenching its fluorophore and rendering it a nonfunctional FRET acceptor, CFP subsequently became dequenched or brighter. This ratiometric sensor for Cl- was able to measure small changes in Cl- at physiological concentrations. The crystal structure of YFP H148Q has been solved (Wachter et al., 2000), and the selecitive halide-binding site was found in a small, amphiphilic, buried cavity adjacent to R96. The halide ion was found to be hydrogen-bonded to the phenol group of tyrosine at position 203, illustrating why this mutation is critical to the formation of a halide-sensitive GFP (Wachter et al., 2000). Conversely, YFPs have been rendered virtually insensitive to the effects of halide binding by the mutation Q69M (Griesbeck et al., 2001) or F64L/M153T/V163A/S175G (“Venus”) (Nagai et al., 2002). In the first case, the bulkier side chain of Met fills the cavity into which chloride ion could reside with Q69, and it is conformationally stable. Even if a conformational change in the thioether side chain of Met were permitted in this space, it is unlikely to contribute to halide binding because it is incapable of hydrogen bonding in the same manner as the carboxamide nitrogen of a Gln side chain (Griesbeck et al., 2001). In the second case, F64L induced large conformational changes in the molecule, leading to the removal of halide sensitivity by preventing ion access to the binding site (Rekas et al., 2002). The “Venus” variant of GFP is also very insensitive to changes in pH.

5.14

INSERTION OF GFP INTO OTHER PROTEINS

Insertion of GFPs into other proteins is another important twist on this theme. An early example was the insertion of GFP [GFPDC; Chalfie et al. (1994)] into a nonconducting mutant of the Shaker K+ channel (Siegel and Isacoff, 1997) and subsequent, improved generations thereof (Guerrero et al., 2002). This first version of the fusion was able to monitor changes in membrane voltage with a maximal fractional fluorescence change of 5.1%. Similarly, wtGFP was inserted into an intracellular loop of a reversibly nonconducting form of the rat mu I skeletal muscle voltage-gated sodium channel (Ataka and Pieribone, 2002). The resulting protein called SPARC (sodium channel protein-based activity reporting construct) can report depolarizing pulses as short as 2 ms and does not inactivate during prolonged depolarizations, but the size of its optical response is very small.

5.15

TANDEM CONCATENATIONS OF TWO GFPS

The availability of GFP mutants of different colors, UV-excited blue emitters and blueexcited green emitters, enables fluorescence resonance energy transfer (FRET) from one to the other. FRET is strongly dependent on the angular orientation and distance of the fluorophores from one another, falling off steeply as the distance exceeds the Förster distance R0 at which FRET is 50% efficient (Tsien et al., 1993; Lakowicz, 1999). For the blue emitter P4-3, containing the point mutations Y66H and Y145F, donating energy either to S65T or S65C, R0 is calculated to be about 40 Å (Heim and Tsien, 1996), assuming that the mutual orientation of the fluorophores is random or freely tumbling. The larger the R0, the better; this is because GFP is a cylindrical structure of about 12-Å radius and 42-Å length (Ormo et al., 1996), and so much of R0 is used up simply within the two GFPs. A systematic study determined the Förster distances between all homo and hetero

SILENT AND LOSS-OF-FUNCTION MUTATIONS

pairings of BFP, CFP, GFP, YFP, and DsRed [Patterson et al., 2000; see also Wu (1994)] The maximum R0 measured for any pair was 56.4 Å between EGFP and EYFP. The Förster distance between CFP and YFP is 49.2 Å. This aspect, combined with greater distance between the peaks of excitation and a favorable overlap integral (J), makes CFP and YFP the pair used most commonly in FRET studies. Sensors that track the activity of proteases were one of the earliest applications for which FRET between concatenated GFPs was exploited (Xu et al., 1998 Heim, 1999; Harpur et al., 2001; Luo et al., 2001; Tawa et al., 2001) and reviewed (Jones et al., 2000). Sensors for caspase 3 have been used successfully as reporters in high-throughput drug discovery programs (Tawa et al., 2001). The first examples of these constructs included the concatenation of the genes encoding S65C or S65T (GFPs) and P4-3 (BFP) with an intervening 25-residue linker connecting the two GFP-derived domains (Heim and Tsien, 1996). Likewise, BFP5 and RSGFP4 have been fused with a 20-residue linker sensitive to factor Xa (Mitra et al., 1996). In either case, before protease cleavage, UV excitation gives rise to some blue emission but also substantial green emission due to FRET from the blue- to the green-emitting domain. After protease cleavage to separate the two domains, FRET is abolished, the blue emission is increased, and the green emission is nearly abolished. For the S65C:P4-3 construct, the ratio between blue and green emission intensities increased by a factor of 4.6 upon cleavage (Heim and Tsien, 1996), while the RSGFP4::BFP5 fusion showed about a 1.9-fold increase in emission ratio (Mitra et al., 1996). In the case of the S65C/P4-3 construct, the large change in ratio between blue and green emissions was shown to result from separation of the two fluorophores rather than from an effect on either one separately, because control experiments with the two separate proteins showed no spectral sensitivity to protease under matching conditions. Another concatenated, intramolecular FRET-based reporter that has been used broadly is the calcium sensor CaMeleon (Miyawaki et al., 1997). In this sensor, CFP and YFP flank calmodulin and a Ca2+-calmodulin-binding peptide, M13 from myosin light-chain kinase. Wheng CaMeleon encounters a change in Ca2+, M13 and calmodulin respond either by associating in increased [Ca2+] or dissociating in decreased [Ca2+]. The conformational change that occurs causes a change in the FRET efficiency (which is low at lower [Ca2+] and high at higher [Ca2+]) largely via a change in kappa squared or the intermolecular angle of fluorophore orientation (Atsushi Miyawaki, personal communication). This probe has been especially useful in reporting calcium changes in subcellular domains like the nucleus and endoplasmic reticulum (Arnaudeau et al., 2002; Demaurex and Frieden, 2003; Malli et al., 2003; Palme et al., 2004) and caveolae (Isshiki et al., 2002), as well as in Drosophila (Reiff et al., 2002; Liu et al., 2003) and C. elegans (Kerr et al., 2000). The most recent versions of CaMeleon incorporate YFPs (Citrine) that are insensitive to pH and Halides (Griesbeck et al., 2001). Most recently, conformationally sensitive concatenations of GFP have been used to track the activity of kinases (Ting et al., 2001; Zhang et al., 2001b; Violin et al., 2003), elegantly showing links between activity of these kinases and other signal transduction pathways.

5.16

SILENT AND LOSS-OF-FUNCTION MUTATIONS

During any random or semirandom mutagenesis screen, the great majority of colonies typically are either indistinguishable from the starting phenotype or of significantly reduced brightness. In principle, these mutants could be sequenced to provide a list of neutral or deleterious substitutions, but such a list would be laborious to collect and of negligible

111

112

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

interest to those wishing to improve GFP or obtain novel properties. Perhaps this list deserves compilation, because it might increase understanding of how GFP folds and builds its chromophore. A few mutations are known to be neutral, such as the ubiquitous Q80R, which may have arisen from a PCR error in the initially distributed cDNA clone (Chalfie et al., 1994). The substitutions within the natural isoforms are presumably all permissive for fluorescence, although other properties may well be altered. As mentioned earlier, mutations of S65 to bulky or highly polar residues, mutations of Y66 to any nonaromatic amino acid, and mutations of G67 to anything else are probably not tolerated.

5.17 NUCLEIC ACID CHANGES THAT DO NOT CHANGE THE PREDICTED AMINO ACID SEQUENCE—THAT IS, OPTIMIZATION OF CODON USAGE AND ELIMINATION OF CRYPTIC SPLICE SITES GFP expression levels can often be increased by redesigning the nucleic acid sequence in ways that should have no significant effect on the final protein sequence. For example, the codon usage in the jellyfish gene is not optimal for mammalian cells, so the gene has been resynthesized with mammalian-preferred codons (Crameri et al., 1996; Levy et al., 1996; Zolotukhin et al., 1996). Translation in eukaryotes can be optimized by inclusion of an optimal translation-initiation sequence (Kozak, 1989). This redesign sometimes involves inserting a new codon that begins with G immediately after the start (AUG) codon. This introduces an extra amino acid such as Ala or Val, which in some articles adds one to the numbering of all amino acids from 2 upwards (Crameri et al., 1996), whereas we prefer to call it 1a to preserve wild-type numbering. Fortunately, the N-terminus is tolerant of such additions. For ease of comparison of mutants, this chapter numbers residues according to their position in the original gfp gene. In plant cells, mRNA derived from the original gfp gene undergoes undesired splicing, which can be eliminated by codon changes (Haseloff and Amos, 1995; see also Chapter 12). GFP cDNA coding sequences have also been altered to reflect the codon bias, and thereby increase the level of expression, of a wide variety of organisms such as Chlamydomonas (Franklin et al., 2002), yeast (GeramiNejad et al., 2001), paramecium (Hauser et al., 2000), and sugar beets (Zhang et al., 2001a).

5.18

ODDS AND ENDS

Perhaps one of the most interesting and persistent questions concerning the existence of GFPs is that of their functional role in the animal: Why should they glow? What advantage is afforded to the creatures who harbor such a protein? The recent discoveries of GFPlike proteins from nonbioluminescent Anthozoan organisms indicates that the proteins primary function cannot be linked exclusively to bioluminescence. Similarly, discoveries of chromoproteins in these same animals indicate that the proteins function may not even necessarily be tied to fluorescence. Konstantin Lukyanov’s group (Gurskaya et al., 2003) has cloned a colorless, nonfluorescent GFP (acGFPL) from Aequorea coerulescens. They showed convincingly that the protein was not an artifact of cloning and that fluorescence could be imparted by a reintroducing the invariant G222 which existed naturally in acGFPL as E222. In the living organism, this protein cannot serve as an acceptor for the bioluminescence energy of aequorin, suggesting that this protein may have some completely unique role in the jellyfish. When one considers that the major absorbtion of wild-

REFERENCES

type Aequorea victoria GFP is at 396 nm with only a minor peak at 475 nm (the primary emission band of aequorin), it seems plausible that the primary role of normally fluorescing A. victoria GFP is not necessarily as an acceptor for aequorin. The absorbance maximum of wild-type A. victoria GFP can be shifted to the 475-nm peak by a singlepoint mutation at position 65, a seemingly easy evolutionary maneuver that would optimize the transfer of energy between the two proteins. Perhaps future screens will produce the optimal GFP from Aqueorea. It seems that the structure of GFP has persisted throughout the evolutionary tree and has possibly ended up as an invariant component called Nidogen of basement membranes in organisms up to humans (Hopf et al., 2001). Despite having only a 10% sequence identity with Aequorea GFP, all of the general structural components similar to GFP exist with remarkable similarity. The Nidogen residues that are equivalent to the chromophoreforming SYT residues in GFP are IGG. Several other residues within GFP that are known to be critical for forming fluorescence are also represented in Nidogen by residues that would eliminate the opportunity to form a functional fluorophore. These final two examples suggest that the striking beta-barrel structure of GFP has served several important functions in the many organisms known to express such a structure. We may simply count ourselves fortunate that at some point in the past the evolution chanced upon a version that glows.

REFERENCES Akemann, W., Raj, C. D., and Knopfel, T. (2001). Functional characterization of permuted enhanced green fluorescent proteins comprising varying linker peptides. Photochem. Photobiol. 74: 356–363. Ando, R., Hama, H., Yamamoto-Hino, M., Mizuno, H., and Miyawaki, A. (2002). An optical marker based on the UV-induced green-to-red photoconversion of a fluorescent protein. Proc. Natl. Acad. Sci. USA 99:12651–12656. Arnaudeau, S., Frieden, M., Nakamura, K., Castelbou, C., Michalak, M., and Demaurex, N. (2002). Calreticulin differentially modulates calcium uptake and release in the endoplasmic reticulum and mitochondria. J. Biol. Chem. 277:46696–46705. Ataka, K., and Pieribone, V. A. (2002). A genetically targetable fluorescent probe of channel gating with rapid kinetics. Biophys. J. 82:509–516. Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (1999). Circular permutation and receptor insertion within green fluorescent proteins. Proc. Natl. Acad. Sci. USA 96:11241–11246. Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2000). Biochemistry, mutagenesis, and oligomerization of DsRed, a red fluorescent protein from coral. Proc. Natl. Acad. Sci. USA 97: 11984–11989. Barondeau, D. P., Putnam, C. D., Kassmann, C. J., Tainer, J. A., and Getzoff, E. D. (2003). Mechanism and energetics of green fluorescent protein chromophore synthesis revealed by trapped intermediate structures. Proc. Natl. Acad. Sci. USA. 100:12111–12116. Battistutta, R., Negro, A., and Zanotti, G. (2000). Crystal structure and refolding properties of the mutant F99S/M153T/V163A of the green fluorescent protein. Proteins 41:429–437. Bell, A. F., He, X., Wachter, R. M., and Tonge, P. J. (2000). Probing the ground state structure of the green fluorescent protein chromophore using Raman spectroscopy. Biochemistry 39: 4423–4431. Bokman, S. H., and Ward, W. W. (1981). Renaturation of Aequorea green-fluorescent protein. Biochem. Biophys. Res. Commun. 101:1372–1380.

113

114

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

Brejc, K., Sixma, T. K., Kitts, P. A., Kain, S. R., Tsien, R. Y., Ormo, M., and Remington, S. J. (1997). Structural basis for dual excitation and photoisomerization of the Aequorea victoria green fluorescent protein. Proc. Natl. Acad. Sci. USA 94:2306–2311. Campbell, R. E., Tour, O., Palmer, A. E., Steinbach, P. A., Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2002). A monomeric red fluorescent protein. Proc. Natl. Acad. Sci. USA 99:7877–7882. Campbell, R. E., Tour, O., Palmer, A. E., Steinbach, P. A., Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2002). A monomeric red fluorescent protein. Proc. Natl. Acad. Sci. USA 99:7877–7882. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263:802–805. Chattoraj, M., King, B. A., Bublitz, G. U., and Boxer, S. G. (1996). Ultra-fast excited state dynamics in green fluorescent protein: Multiple states and proton transfer. Proc. Natl. Acad. Sci. USA 93:8362–8367. Cheng, L., Fu, J., Tsukamoto, A., and Hawley, R. G. (1996). Use of green fluorescent protein variants to monitor gene transfer and expression in mammalian cells. Nat. Biotechnol. 14:606–609. Cody, C. W., Prasher, D. C., Westler, W. M., Prendergast, F. G., and Ward, W. W. (1993). Chemical structure of the hexapeptide chromophore of the Aequorea green-fluorescent protein. Biochemistry 32:1212–1218. Cormack, B. P., Valdivia, R. H., and Falkow, S. (1996). FACS-optimized mutants of the green fluorescent protein (GFP). Gene 173:33–38. Cotlet, M., Hofkens, J., Habuchi, S., Dirix, G., Van Guyse, M., Michiels, J., Vanderleyden, J., and De Schryver, F. C. (2001). Identification of different emitting species in the red fluorescent protein DsRed by means of ensemble and single-molecule spectroscopy. Proc. Natl. Acad. Sci. USA 27:27. Crameri, A., Whitehorn, E. A., Tate, E., and Stemmer, W. P. (1996). Improved green fluorescent protein by molecular evolution using DNA shuffling. Nat. Biotechnol. 14:315–319. Creemers, T. M., Lock, A. J., Subramaniam, V., Jovin, T. M., and Volker, S. (1999). Three photoconvertible forms of green fluorescent protein identified by spectral hole-burning. Nat. Struct. Biol. 6:557–560. Cubitt, A. B., Heim, R., Adams, S. R., Boyd, A. E., Gross, L. A., and Tsien, R. Y. (1995). Understanding, improving and using green fluorescent proteins. Trends Biochem. Sci. 20:448–455. Cubitt, A. B., Woollenweber, L. A., and Heim, R. (1999). Understanding structure-function relationships in the Aequorea victoria green fluorescent protein. Methods Cell Biol. 58:19–30. Davenport, D., and Nichol, J. A. C. (1955). Proc. R. Soc. London, Ser. B 144:399–411. De Angelis, D. A., Miesenbock, G., Zemelman, B. V., and Rothman, J. E. (1998). PRIM: proximity imaging of green fluorescent protein-tagged polypeptides. Proc. Natl. Acad. Sci. USA 95:12312–12316. Delagrave, S., Hawtin, R. E., Silva, C. M., Yang, M. M., and Youvan, D. C. (1995). Red-shifted excitation mutants of the green fluorescent protein. Biotechnology (NY) 13:151–154. Demaurex, N., and Frieden, M. (2003). Measurements of the free luminal ER Ca(2+) concentration with targeted “cameleon” fluorescent proteins. Cell Calcium 34:109–119. Dewey, T. G., and Hammes, G. G. (1980). Calculation on fluorescence resonance energy transfer on surfaces. Biophys. J. 32:1023–1035. Dewey, T. G., and Datta, M. M. (1989). Determination of the fractal dimension of membrane protein aggregates using fluorescence energy transfer. Biophys. J. 56:415–420. Doi, N., and Yanagawa, H. (1999). Design of generic biosensors based on green fluorescent proteins with allosteric sites by directed evolution. FEBS Lett. 453:305–307. Dopf, J., and Horiagon, T. M. (1996). Deletion mapping of the Aequorea victoria green fluorescent protein. Gene 173:39–44. Dove, S. G., Hoegh-Guldberg, O., and Ranganathan, S. (2001). Major colour patterns of reefbuilding corals are due to a family of GFP-like proteins. Coral Reefs 19:197–204.

REFERENCES

Ehrig, T., O’Kane, D. J., and Prendergast, F. G. (1995). Green-fluorescent protein mutants with altered fluorescence excitation spectra. FEBS Lett. 367:163–166. Fukuda, H., Arai, M., and Kuwajima, K. (2000). Folding of green fluorescent protein and the cycle3 mutant. Biochemistry 39:12025–12032. Fung, B. K., and Stryer, L. (1978). Surface density determination in membranes by fluorescence energy transfer. Biochemistry 17:5241–5248. Garcia-Parajo, M. F., Koopman, M., van Dijk, E. M., Subramaniam, V., and van Hulst, N. F. (2001). The nature of fluorescence emission in the red fluorescent protein DsRed, revealed by singlemolecule detection. Proc. Natl. Acad. Sci. USA 27:27. Gerami-Nejad, M., Berman, J., and Gale, C. A. (2001). Cassettes for PCR-mediated construction of green, yellow, and cyan fluorescent protein fusions in Candida albicans. Yeast 18:859–864. Gibbs, M. D., Nevalainen, K. M., and Bergquist, P. L. (2001). Degenerate oligonucleotide gene shuffling (DOGS): A method for enhancing the frequency of recombination with family shuffling. Gene 271:13–20. Giver, L., and Arnold, F. H. (1998). Combinatorial protein design by in vitro recombination. Curr. Opin. Chem. Biol. 2:335–338. Graf, R., and Schachman, H. K. (1996). Random circular permutation of genes and expressed polypeptide chains: Application of the method to the catalytic chains of aspartate transcarbamoylase. Proc. Natl. Acad. Sci. USA 93:11591–11596. Griesbeck, O., Baird, G. S., Campbell, R. E., Zacharias, D. A., and Tsien, R. Y. (2001). Reducing the environmental sensitivity of yellow fluorescent protein. Mechanism and applications. J. Biol. Chem. 276:29188–29194. Gross, L. A., Baird, G. S., Hoffman, R. C., Baldridge, K. K., and Tsien, R. Y. (2000). The structure of the chromophore within DsRed, a red fluorescent protein from coral. Proc. Natl. Acad. Sci. USA 97:11990–11995. Guerrero, G., Siegel, M. S., Roska, B., Loots, E., and Isacoff, E. Y. (2002). Tuning Flash: Redesign of the dynamics, voltage range, and color of the genetically encoded optical sensor of membrane potential. Biophys. J. 83:3607–3618. Gurskaya, N. G., Fradkov, A. F., Pounkova, N. I., Staroverov, D. B., Bulina, M. E., Yanushevich, Y. G., Labas, Y. A., Lukyanov, S., and Lukyanov, K. A. (2003). A colourless green fluorescent protein homologue from the non-fluorescent hydromedusa Aequorea coerulescens and its fluorescent mutants. Biochem. J. 373:403–408. Hanson, G. T., McAnaney, T. B., Park, E. S., Rendell, M. E., Yarbrough, D. K., Chu, S., Xi, L., Boxer, S. G., Montrose, M. H., and Remington, S. J. (2002). Green fluorescent protein variants as ratiometric dual emission pH sensors. 1. Structural characterization and preliminary application. Biochemistry 41:15477–15488. Harpur, A. G., Wouters, F. S., and Bastiaens, P. I. (2001). Imaging FRET between spectrally similar GFP molecules in single cells. Nat. Biotechnol. 19:167–169. Haseloff, J., and Amos, B. (1995). GFP in plants. Trends Genet. 11:328–329. Hauser, K., Haynes, W. J., Kung, C., Plattner, H., and Kissmehl, R. (2000). Expression of the green fluorescent protein in Paramecium tetraurelia. Eur. J. Cell. Biol. 79:144–149. Heikal, A. A., Hess, S. T., Baird, G. S., Tsien, R. Y., and Webb, W. W. (2000). Molecular spectroscopy and dynamics of intrinsically fluorescent proteins: Coral red (dsRed) and yellow (Citrine). Proc. Natl. Acad. Sci. USA 97:11996–12001. Heim, R. (1999). Green fluorescent protein forms for energy transfer. Methods Enzymol. 302:408–423. Heim, R., and Tsien, R. Y. (1996). Engineering green fluorescent protein for improved brightness, longer wavelengths and fluorescence resonance energy transfer. Curr. Biol. 6:178–182. Heim, R., Prasher, D. C., and Tsien, R. Y. (1994). Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proc. Natl. Acad. Sci. USA 91:12501–12504.

115

116

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

Heim, R., Cubitt, A. B., and Tsien, R. Y. (1995). Improved green fluorescence. Nature 373:663–664. Hofmann, A., Iwai, H., Hess, S., Pluckthun, A., and Wlodawer, A. (2002). Structure of cyclized green fluorescent protein. Acta Crystallogr. D Biol. Crystallogr. 58:1400–1406. Hopf, M., Gohring, W., Ries, A., Timpl, R., and Hohenester, E. (2001). Crystal structure and mutational analysis of a perlecan-binding fragment of nidogen-1. Nat. Struct. Biol. 8:634–640. Hyun Bae, J., Rubini, M., Jung, G., Wiegand, G., Seifert, M. H., Azim, M. K., Kim, J. S., Zumbusch, A., Holak, T. A., Moroder, L., Huber, R., and Budisa, N. (2003). Expansion of the genetic code enables design of a novel “gold” class of green fluorescent proteins. J. Mol. Biol. 328:1071–1081. Isshiki, M., Ying, Y. S., Fujita, T., and Anderson, R. G. (2002). A molecular sensor detects signal transduction from caveolae in living cells. J. Biol. Chem. 277:43389–43398. Jayaraman, S., Haggie, P., Wachter, R. M., Remington, S. J., and Verkman, A. S. (2000). Mechanism and cellular applications of a green fluorescent protein-based halide sensor. J. Biol. Chem. 275:6047–6050. Jones, J., Heim, R., Hare, E., Stack, J., and Pollok, B. A. (2000). Development and application of a GFP-FRET intracellular caspase assay for drug screening. J. Biomol. Screen 5:307–318. Kaether, C., and Gerdes, H. H. (1995). Visualization of protein transport along the secretory pathway using green fluorescent protein. FEBS Lett. 369:267–271. Kahana, J. A., and Silver, P. A. (1996). Use of the A. Victoria green fluorescent protein to study protein dynamics in vivo. In Current Protocols in Molecular Biology, Aosobel, F. M., Brent, R., Kingston, R. E., Moore, D. E., Seidman, J. G., Smith, J. A., and Struhl, K., Eds., (New York, NY, John Wiley & Sons, Inc.), pp. 9.7.22–29.27.27. Kelmanson, I. V., and Matz, M. V. (2003). Molecular basis and evolutionary origins of color diversity in great star coral Montastraea cavernosa (Scleractinia: Faviida). Mol. Biol. Evol. 20:1125–1133. Kerr, R., Lev-Ram, V., Baird, G., Vincent, P., Tsien, R. Y., and Schafer, W. R. (2000). Optical imaging of calcium transients in neurons and pharyngeal muscle of C. elegans. Neuron 26:583–594. Kim, H. K., and Kaang, B. K. (1998). Truncated green fluorescent protein mutants and their expression in Aplysia neurons. Brain. Res. Bull. 47:35–41. Kimata, Y., Iwaki, M., Lim, C. R., and Kohno, K. (1997). A novel mutation which enhances the fluorescence of green fluorescent protein at high temperatures. Biochem. Biophys. Res. Commun. 232:69–73. Kozak, M. (1989). The scanning model for translation: An update. J. Cell Biol. 108:229–241. Kummer, A. D., Wiehler, J., Schuttrigkeit, T. A., Berger, B. W., Steipe, B., and Michel-Beyerle, M. E. (2002). Picosecond time-resolved fluorescence from blue-emitting chromophore variants Y66F and Y66H of the green fluorescent protein. Chembiochem 3:659–663. Kuner, T.. and Augustine G. J. (2000). A genetically encoded ratiometric indicator for chloride: capturing chloride transients in cultured hippocampal neurons. Neuron 27:447–459. Lakowicz, J. (1999). Principles of Fluorescence Spectroscopy, 2nd ed., Plenum, New York. Levy, J. P., Muldoon, R. R., Zolotukhin, S., and Link, C. J., Jr. (1996). Retroviral transfer and expression of a humanized, red-shifted green fluorescent protein gene into human tumor cells. Nat. Biotechnol. 14:610–614. Lim, C. R., Kimata, Y., Oka, M., Nomaguchi, K., and Kohno, K. (1995). Thermosensitivity of green fluorescent protein fluorescence utilized to reveal novel nuclear-like compartments in a mutant nucleoporin NSP1. J. Biochem. (Tokyo) 118:13–17. Liu, L., Yermolaieva, O., Johnson, W. A., Abboud, F. M., and Welsh, M. J. (2003). Identification and function of thermosensory neurons in Drosophila larvae. Nat. Neurosci. 6:267–273. Llopis, J., McCaffery, J. M., Miyawaki, A., Farquhar, M. G., and Tsien, R. Y. (1998). Measurement of cytosolic, mitochondrial, and Golgi pH in single living cells with green fluorescent proteins. Proc. Natl. Acad. Sci. USA 95:6803–6808.

REFERENCES

Luo, K. Q., Yu, V. C., Pu, Y., and Chang, D. C. (2001). Application of the fluorescence resonance energy transfer method for studying the dynamics of caspase-3 activation during UV-induced apoptosis in living HeLa cells. Biochem. Biophys. Res. Commun. 283:1054–1060. Malli, R., Frieden, M., Osibow, K., and Graier, W. F. (2003). Mitochondria efficiently buffer subplasmalemmal Ca2+ elevation during agonist stimulation. J. Biol. Chem. 278:10807– 10815. Martynov, V. I., Savitsky, A. P., Martynova, N. Y., Savitsky, P. A., Lukyanov, K. A., and Lukyanov, S. A. (2001). Alternative cyclization in GFP-like proteins family. The formation and structure of the chromophore of a purple chromoprotein from Anemonia sulcata. J. Biol. Chem. 276:21012–21016. Matz, M. V., Lukyanov, K. A., and Lukyanov, S. A. (2002). Family of the green fluorescent protein: Journey to the end of the rainbow. Bioessays 24:953–959. McAnaney, T. B., Park, E. S., Hanson, G. T., Remington, S. J., and Boxer, S. G. (2002). Green fluorescent protein variants as ratiometric dual emission pH sensors. 2. Excited-state dynamics. Biochemistry 41:15489–15494. McElroy, W. D. (1947). The energy source for biolominescence in an isolated system, Proc. Natl. Acad. Sci. USA 33:342–345. McRorie, D., and Voelker, P. (1993). Self-Associating Systems in the Analytical Ultracentrifuge, Beckman Instruments, Fullerton, CA. Miesenbock, G., De Angelis, D. A., and Rothman, J. E. (1998). Visualizing secretion and synaptic transmission with pH-sensitive green fluorescent proteins. Nature 394:192–195. Minshull, J., and Stemmer, W. P. (1999). Protein evolution by molecular breeding. Curr. Opin. Chem. Biol. 3:284–290. Mitra, R. D., Silva, C. M., and Youvan, D. C. (1996). Fluorescence resonance energy transfer between blue-emitting and red-shifted excitation derivatives of the green fluorescent protein. Gene 173:13–17. Miyawaki, A., Llopis, J., Heim, R., McCaffery, J. M., Adams, J. A., Ikura, M., and Tsien, R. Y. (1997). Fluorescent indicators for Ca2+ based on green fluorescent proteins and calmodulin. Nature 388:882–887. Miyawaki, A. (2003). Visualization of the spatial and temporal dynamics of intracellular signaling. Dev. Cell. 4:295–305. Mizuno, H., Sawano, A., Eli, P., Hama, H., and Miyawaki, A. (2001). Red fluorescent protein from Discosoma as a fusion tag and a partner for fluorescence resonance energy transfer. Biochemistry 40:2502–2510. Morise, H., Shimomura, O., Johnson, F. H., and Winant, J. (1974). Intermolecular energy transfer in the bioluminescent system of Aequorea. Biochemistry 13:2656–2662. Muhlrad, D., Hunter, R., and Parker, R. (1992). A rapid method for localized mutagenesis of yeast genes. Yeast 8:79–82. Nagai, T., Sawano, A., Park, E. S., and Miyawaki, A. (2001). Circularly permuted green fluorescent proteins engineered to sense Ca2+. Proc. Natl. Acad. Sci. USA 98:3197–3202. Nagai, T., Ibata, K., Park, E. S., Kubota, M., Mikoshiba, K., and Miyawaki, A. (2002). A variant of yellow fluorescent protein with fast and efficient maturation for cell-biological applications. Nat. Biotechnol. 20:87–90. Nakai, J., Ohkura, M., and Imoto, K. (2001). A high signal-to-noise Ca(2+) probe composed of a single green fluorescent protein. Nat. Biotechnol. 19:137–141. Niwa, H., Inouye, S., Hirano, T., Matsuno, T., Kojima, S., Kubota, M., Ohashi, M., and Tsuji, F. I. (1996). Chemical nature of the light emitter of the Aequorea green fluorescent protein. Proc. Natl. Acad. Sci. USA 93:13617–13622. Ormo, M., Cubitt, A. B., Kallio, K., Gross, L. A., Tsien, R. Y., and Remington, S. J. (1996). Crystal structure of the Aequorea victoria green fluorescent protein. Science 273:1392–1395.

117

118

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

Palm, G., and Wlodawer, A. (1999). Spectral variants of green fluorescent protein. Methods Enzymol. 302:378–394. Palm, G. J., Zdanov, A., Gaitanaris, G. A., Stauber, R., Pavlakis, G. N., and Wlodawer, A. (1997). The structural basis for spectral variations in green fluorescent protein. Nat. Struct. Biol. 4:361–365. Palmer, A. E., Jin, C., Reed, J. C., and Tsien, R. Y. (2004). Bcl-2-mediated alterations in endoplasmic reticulum Ca2+ analyzed with an improved genetically encoded fluorescent sensor. Proc. Natl. Acad. Sci. USA 101:17404–17409. Patterson, G. H., and Lippincott-Schwartz, J. (2002). A photoactivatable GFP for selective photolabeling of proteins and cells. Science 297:1873–1877. Patterson, G. H., Knobel, S. M., Sharif, W. D., Kain, S. R., and Piston, D. W. (1997). Use of the green fluorescent protein and its mutants in quantitative fluorescence microscopy. Biophys. J. 73:2782–2790. Patterson, G. H., Piston, D. W., and Barisas, B. G. (2000). Forster distances between green fluorescent protein pairs. Anal. Biochem. 284:438–440. Prasher, D. (1995). Using GFP to see the light. Trends Genet. 11:320–232. Prasher, D. C., Eckenrode, V. K., Ward, W. W., Prendergast, F. G., and Cormier, M. J. (1992). Primary structure of the Aequorea victoria green-fluorescent protein. Gene 111:229–233. Reiff, D. F., Thiel, P. R., and Schuster, C. M. (2002). Differential regulation of active zone density during long-term strengthening of Drosophila neuromuscular junctions. J. Neurosci. 22:9399–9409. Rekas, A., Alattia, J. R., Nagai, T., Miyawaki, A., and Ikura, M. (2002). Crystal structure of Venus, a yellow fluorescent protein with improved maturation and reduced environmental sensitivity. J. Biol. Chem. 4:4. Richmond, T. A., Takahashi, T. T., Shimkhada, R., and Bernsdorf, J. (2000). Engineered metal binding sites on green fluorescence protein. Biochem. Biophys. Res. Commun. 268:462–465. Rizzo, M. A., Springer, G. H., Granada, B., and Piston, D. W. (2004). An improved cyan fluorescent protein variant useful for FRET. Nat. Biotechnol. 22:445–449. Rizzuto, R., Brini, M., De Giorgi, F., Rossi, R., Heim, R., Tsien, R. Y., and Pozzan, T. (1996). Double labelling of subcellular structures with organelle-targeted GFP mutants in vivo. Curr. Biol. 6:183–188. Salih, A., Larkum, A., Cox, G., Kuhl, M., and Hoegh-Guldberg, O. (2000). Fluorescent pigments in corals are photoprotective. Nature 408:850–853. Sankaranarayanan, S., De Angelis, D., Rothman, J. E., and Ryan, T. A. (2000). The use of pHluorins for optical measurements of presynaptic activity. Biophys. J. 79:2199–2208. Sawano, A., and Miyawaki, A. (2000). Directed evolution of green fluorescent protein by a new versatile PCR strategy for site-directed and semi-random mutagenesis. Nucleic Acids Res. 28:E78. Scharnagl, C., Raupp-Kossmann, R., and Fischer, S. F. (1999). Molecular basis for pH sensitivity and proton transfer in green fluorescent protein: Protonation and conformational substates from electrostatic calculations. Biophys. J. 77:1839–1857. Shaner, N. C., Campbell, R. E., Steinbach, P. A., Giepmans, B. N., Palmer, A. E., and Tsien, R. Y. (2004). Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nat. Biotechnol. 22:1567–1572. Shimomura, O. (1979). Structure of the chromophore of Aequorea green fluorescent protein. FEBS Lett. 104:220–222. Siegel, M. S., and Isacoff, E. Y. (1997). A genetically encoded optical probe of membrane voltage. Neuron 19:735–741. Siemering, K. R., Golbik, R., Sever, R., and Haseloff, J. (1996). Mutations that suppress the thermosensitivity of green fluorescent protein. Curr. Biol. 6:1653–1663.

REFERENCES

Sikorski, R. S., and Boeke, J. D. (1991). In vitro mutagenesis and plasmid shuffling: from cloned gene to mutant yeast. Methods Enzymol. 194:302–318. Simon, S. M. (1996). Cellular probes on the move. Nat. Biotechnol. 14:1221. Snyder, B., and Freire, E. (1982). Fluorescence energy transfer in two dimensions. A numeric solution for random and nonrandom distributions. Biophys. J. 40:137–148. Tawa, P., Tam, J., Cassady, R., Nicholson, D. W., and Xanthoudakis, S. (2001). Quantitative analysis of fluorescent caspase substrate cleavage in intact cells and identification of novel inhibitors of apoptosis. Cell Death Differ. 8:30–37. Ting, A. Y., Kain, K. H., Klemke, R. L., and Tsien, R. Y. (2001). Genetically encoded fluorescent reporters of protein tyrosine kinase activities in living cells. Proc. Natl. Acad. Sci. USA 98:15003–15008. Topell, S., Hennecke, J., and Glockshuber, R. (1999). Circularly permuted variants of the green fluorescent protein. FEBS Lett. 457:283–289. Tsien, R. Y. (1998). The green fluorescent protein. Annu. Rev. Biochem. 67:509–544. Tsien, R. Y., Bacskai, B., and Adams, S. R. (1993). FRET for studying intracellular signalling. Trends Cell Biol. 3:242–245. van Thor, J. J., Gensch, T., Hellingwerf, K. J., and Johnson, L. N. (2002). Phototransformation of green fluorescent protein with UV and visible light leads to decarboxylation of glutamate 222. Nat. Struct. Biol. 9:37–41. Violin, J. D., Zhang, J., Tsien, R. Y., and Newton, A. C. (2003). A genetically encoded fluorescent reporter reveals oscillatory phosphorylation by protein kinase C. J. Cell Biol. 161:899–909. Volkov, A. A., Shao, Z., and Arnold, F. H. (1999). Recombination and chimeragenesis by an in vitro heteroduplex formation and in vivo repair. Nucleic Acids Res. 27:e18 i–vi. Wachter, R. M., King, B. A., Heim, R., Kallio, K., Tsien, R. Y., Boxer, S. G., and Remington, S. J. (1997). Crystal structure and photodynamic behavior of the blue emission variant Y66H/Y145F of green fluorescent protein. Biochemistry 36:9759–9765. Wachter, R. M., Elsliger, M. A., Kallio, K., Hanson, G. T., and Remington, S. J. (1998). Structural basis of spectral shifts in the yellow-emission variants of green fluorescent protein. Structure 6:1267–1277. Wachter, R. M., and Remington, S. J. (1999). Sensitivity of the yellow variant of green fluorescent protein to halides and nitrate. Curr. Biol. 9:R628–R629. Wachter, R. M., Yarbrough, D., Kallio, K., and Remington, S. J. (2000). Crystallographic and energetic analysis of binding of selected anions to the yellow variants of green fluorescent protein. J. Mol. Biol. 301:157–171. Waldo, G. S., Standish, B. M., Berendzen, J., and Terwilliger, T. C. (1999). Rapid protein-folding assay using green fluorescent protein. Nat. Biotechnol. 17:691–695. Wang, L., Jackson, W. C., Steinbach, P. A., and Tsien, R. Y. (2004) Evolution of new nonantibody proteins via iterative somatic hypermutation. Proc. Natl. Acad. Sci. USA 101:16745–16749. Wang, L., Xie, J., Deniz, A. A., and Schultz, P. G. (2003). Unnatural amino acid mutagenesis of green fluorescent protein. J. Org. Chem. 68:174–176. Ward, W. W. (1998). In GFP Green Fluorescent Protein Properties, Applications and Protocols, Chalfie, M., and Kain, S, Ed. Wiley-Liss, New York, pp. 45–75. Ward, W. W., and Bokman, S. H. (1982). Reversible denaturation of Aequorea green-fluorescent protein: Physical separation and characterization of the renatured protein. Biochemistry 21:4535–4540. Ward, W. W., and Cormier, M. J. (1979). An energy transfer protein in coelenterate bioluminescence. Characterization of the Renilla green-fluorescent protein. J. Biol. Chem. 254:781–788. Ward, W., Prentice, H., Roth, A., Cody, C., and Reeves, S. (1982). Spectral perturbations of the Aequorea green fluorescent protein. Photochem. Photobiol. 35:803–808.

119

120

MOLECULAR BIOLOGY AND MUTATION OF GREEN FLUORESCENT PROTEIN

Wiedenmann, J., Elke, C., Spindler, K. D., and Funke, W. (2000). Cracks in the beta-can: Fluorescent proteins from Anemonia sulcata (Anthozoa, Actinaria). Proc. Natl. Acad. Sci. USA 97:14091–14096. Wiedenmann, J., Schenk, A., Rocker, C., Girod, A., Spindler, K. D., and Nienhaus, G. U. (2002). A far-red fluorescent protein with fast maturation and reduced oligomerization tendency from Entacmaea quadricolor (Anthozoa, Actinaria). Proc. Natl. Acad. Sci. USA 99:11646–11651. Wolber, P. K., and Hudson, B. S. (1979). An analytic solution to the Forster energy transfer problem in two dimensions. Biophys. J. 28:197–210. Wu, P., and Brand, L. (1994). Resonance energy transfer: Methods and applications. Anal. Biochem. 218:1–13. Xu, X., Gerard, A. L., Huang, B. C., Anderson, D. C., Payan, D. G., and Luo, Y. (1998). Detection of programmed cell death using fluorescence energy transfer. Nucleic Acids Res. 26:2034–2035. Yang, F., Moss, L. G., and Phillips, G. N., Jr. (1996a). The molecular structure of green fluorescent protein. Nat. Biotechnol. 14:1246–1251. Yang, T. T., Kain, S. R., Kitts, P., Kondepudi, A., Yang, M. M., and Youvan, D. C. (1996b). Dual color microscopic imagery of cells expressing the green fluorescent protein and a red-shifted variant. Gene. 173:19–23. Yang, T. T., Sinai, P., Green, G., Kitts, P. A., Chen, Y. T., Lybarger, L., Chervenak, R., Patterson, G. H., Piston, D. W., and Kain, S. R. (1998). Improved fluorescence and dual color detection with enhanced blue and green variants of the green fluorescent protein. J. Biol. Chem. 273:8212–8216. Yguerabide, J. (1994). Theory for establishing proximity relations in biological membranes by excitation energy transfer measurements. Biophys. J. 66:683–693. Yokoe, H., and Meyer, T. (1996). Spatial dynamics of GFP-tagged proteins investigated by local fluorescence enhancement. Nat. Biotechnol. 14:1252–1256. Youvan, D. C., Goldman, E., Delagrave, S., and Yang, M. M. (1995). Digital imaging spectroscopy for massively parallel screening of mutants. Methods Enzymol. 246:732–748. Yu, D., Baird, G. S., Tsien, R. Y., and Davis, R. L. (2003). Detection of calcium transients in Drosophila mushroom body neurons with camgaroo reporters. J. Neurosci. 23:64–72. Zacharias, D. A., Baird, G. S., and Tsien, R. Y. (2000). Recent advances in technology for measuring and manipulating cell signals. Curr. Opin. Neurobiol. 10:416–421. Zacharias, D. A., Violin, J. D., Newton, A. C., and Tsien, R. Y. (2001). Partitioning of monomeric lipid modified GFPs inot lipid rafts of living cells. Science 296:913–916. Zacharias, D. A. (2002). Sticky caveats in an otherwise glowing report: Oligomerizing fluorescent proteins and their use in cell biology. Sci. STKE 2002:PE23. Zapata-Hommer, O., and Griesbeck, O. (2003). Efficiently folding and circularly permuted variants of the Sapphire mutant of GFP. BMC Biotechnol. 3:5. Zhang, C. L., Chen, D. F., McCormac, A. C., Scott, N. W., Elliott, M. C., and Slater, A. (2001a). Use of the GFP reporter as a vital marker for Agrobacterium-mediated transformation of sugar beet (Beta vulgaris L.). Mol. Biotechnol. 17:109–117. Zhang, J., Ma, Y., Taylor, S. S., and Tsien, R. Y. (2001b). Genetically encoded reporters of protein kinase A activity reveal impact of substrate tethering. Proc. Natl. Acad. Sci. USA 98:14997–15002. Zhang, J., Campbell, R. E., Ting, A. Y., and Tsien, R. Y. (2002). Creating new fluorescent probes for cell biology. Nat. Rev. Mol. Cell Biol. 3:906–918. Zimet, D. B., Thevenin, B. J., Verkman, A. S., Shohet, S. B., and Abney, J. R. (1995). Calculation of resonance energy transfer in crowded biological membranes. Biophys. J. 68:1592–1603. Zimmer, M. (2002). Green fluorescent protein (GFP): Applications, structure, and related photophysical behavior. Chem. Rev. 102:759–781. Zolotukhin, S., Potter, M., Hauswirth, W. W., Guy, J., and Muzyczka, N. (1996). A “humanized” green fluorescent protein cDNA adapted for high-level expression in mammalian cells. J. Virol. 70:4646–4654.

6 DISCOVERY AND PROPERTIES OF GFP-LIKE PROTEINS FROM NONBIOLUMINESCENT ANTHOZOA Konstantin A. Lukyanov, Dmitry M. Chudakov, and Arkady F. Fradkov Shemyakin and Ovchinnikov Institute of Bioorganic Chemistry RAS, Moscow, Russia

Yulii A. Labas Institute of Biochemistry RAS, Moscow, Russia

Mikhail V. Matz Whitney Laboratory, University of Florida, St. Augustine, FL

Sergey Lukyanov Shemyakin and Ovchinnikov Institute of Bioorganic Chemistry RAS, Moscow, Russia

6.1 INTRODUCTION Green fluorescent protein (GFP) was discovered in hydroid medusa Aequorea victoria (synonyms A. forskalea, A. aequorea) more than 40 years ago (Johnson et al., 1962; Chapter 1 of this volume). After that, GFPs were found in several bioluminescent Hydrozoa and Anthozoa species (Chalfie, 1995). In all these examples, GFPs played role of secondary emitter within bioluminescent systems. The association of GFPs with bioluminescence was possibly the main reason why researchers did not search for GFP-like proteins in nonbioluminescent corals for a long time. We were lucky to clone genes for GFP-like proteins from nonbioluminescent Anthozoa for the first time (Matz et al., 1999). Fortune really smiled on us: Several months after the work was started, we cloned the first GFP from the sea anemone Anemonia majana. In comparison to many other marine coelenterates, working on reef Anthozoa is particuGreen Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

121

122

DISCOVERY AND PROPERTIES OF GFP-LIKE PROTEINS FROM NONBIOLUMINESCENT ANTHOZOA

larly convenient because one can readily buy the specimens for several dollars in aquarium shops throughout the world. We did not organize expeditions on a ship or bathyscaphe, but found brightly fluorescent and colored sea anemones and corallimorphs (mushroom anemones) in Moscow instead. Shortly after the publication of our paper, other groups independently reported finding of GFP-like proteins in Anthozoa species (Wiedenmann et al., 2000; Dove et al., 2001). At present, it is widely accepted that vivid fluorescent and nonfluorescent coloration of coral polyps is mainly determined by numerous GFP homologs (Matz et al., 1999, 2002; Fradkov et al., 2000; Lukyanov et al., 2000; Wiedenmann et al., 2000, 2002; Salih et al., 2000; Gurskaya et al., 2001a; Dove et al., 2001; Labas et al., 2002; Ando et al., 2002).

6.2

COLOR DIVERSITY WITHIN ANTHOZOA GFP HOMOLOGS

The most interesting feature of the coral GFP-like proteins is their color variety as determined by the proteins emission properties. Four main color groups have been recognized thus far (Labas et al., 2002): (a) green, yellow, and red fluorescent proteins (FPs) and (b) nonfluorescent chromoproteins (CPs) of different hues, from orange to blue (Fig. 6.1, left column). The green FPs are divided into three subgroups. The first subgroup contains cyan FPs, which are relatively blue-shifted and have broad spectra (Fig. 6.1A). These proteins possess excitation maxima at 440–460 nm and emission maxima at ~480–490 nm, along with spectral widths (width of the peak at half of maximal intensity) of emission curves of about 50 nm. The second subgroup includes the majority of the green FPs, proteins that have excitation–emission maxima at 490–510 and 500–520 nm, respectively, and a narrow fluorescence peak (spectral width about 25 nm; Fig. 6.1B). A characteristic feature of the third subgroup is dual-peaks excitation spectrum usually having a major peak at around 400 nm and a minor peak at 470–490 nm (Fig. 6.1C). There was found only one yellow FP, zoanYFP (zFP538) (Matz et al., 1999). It has a narrow emission spectrum with a peak at 538 nm and has an excitation spectrum with a major peak at 528 nm and a minor peak at 494 nm (Fig. 6.1D). Red FPs possess emission maxima greater than 570 nm. Often these proteins go through green-emitting stage during their posttranslational maturation. RFPs can be subdivided into two subgroups. The first subgroup is represented by drFP583 (Matz et al., 1999) (commercial name DsRed, available from Clontech Laboratories), the most well-studied RFP to date. These RFPs are characterized by a broad emission spectrum (spectral width about 50–60 nm) with a peak between 570 and 610 nm (Fig. 6.1E). The main characteristic feature of the second RFP subgroup is need of UV or violet light irradiation for red fluorophore formation, as it was first shown by Ando et al. (2002) and later confirmed for other similar proteins (Wiedenmann et al., 2004; M. Matz and K. Lukyanov, unpublished). In the dark these proteins mature only to the green FP form. Short-wavelength light irradiation causes their fast transformation into red FP form. The resulting red emission spectra are rather narrow (spectral width about 25 nm) and have a pronounced shoulder at about 630 nm (Fig. 6.1F). The last color class of GFP-like proteins unites chromoproteins (CPs) that effectively absorb but emit little or no light (Lukyanov et al., 2000; Wiedenmann et al., 2000; Gurskaya et al., 2001a; Dove et al., 2001). Known CPs possessed single absorption maxima at 560–590 nm (Fig. 6.1G). Curiously, in this region of spectra, a subtle shift of the absorption maximum may lead to a significant change in the perceived CP color, so

123

COLOR DIVERSITY WITHIN ANTHOZOA GFP HOMOLOGS

W ild types A

4 45

M utants dstrG FP

4 83

H

4 60

4 93

6 10 5 42

dstrG FP N 68S I112S G reen

+ Red

B

zoanG FP

I

zoanG FP A 63G N 65K N 68D

5 25 5 37 4 94

G reen

4 94 5 06

C

hcriG FP

4 99

4 05

zoanY FP D 68N

J

G reen + Yellow

4 81

D

zoanY FP

Yellow

5 24 5 37

K

zoanY FP K 65M

4 94 5 06

4 94

E

Yellow

G reen

5 58

5 83

D sR ed

L

D sR ed Y 121H

4 80 5 00

Red

G reen

F

5 08 5 20

5 80

m cavR FP

M

hcriC P

N

D sR ed S 148C I165N K 167M Chrom o S 203A

5 61

5 72

G

5 90

Chrom o

5 78

6 40

Far-red hcriC P C 148 S 400

500 600 Wavelength, nm

700

400

500 600 Wavelength, nm

700

Figure 6.1. Spectral properties of wild-type and mutant GFP-like proteins of the main color groups. Spectra for wild-type proteins are shown on the left, and spectra for corresponding mutants with altered color are shown on the right. Dashed lines represent excitation spectra for FPs or absorption spectra for CPs. Solid lines correspond to the emission spectra.

that to a human observer these proteins may appear as soft hues of purple, crimson, lilac, and even blue. In some CPs, extremely weak (quantum yield <0.001) red and far-red fluorescence (590–640 nm) can be detected. Two main reasons for the color differences can be suggested. First, it could arise from distinct noncovalent interactions of the chromophore with its microenvironment. Second, chemically distinct chromophores can determinate bathochromic spectral shifts. The first

124

DISCOVERY AND PROPERTIES OF GFP-LIKE PROTEINS FROM NONBIOLUMINESCENT ANTHOZOA

way was well documented for number of Aequorea GFP mutants (Remington, 2000). In Anthozoa proteins, an unexpected diversity of chromophore structures was found. Red FP DsRed was shown to derive its spectral quality from an additional autocatalytic dehydrogenation of the aC–N bond of Gln65, which extends the GFP-like chromophore by two strongly electron-withdrawing double bonds (Gross et al., 2000; Wall et al., 2000; Yarbrough et al., 2001). The enlarged system of conjugated double bonds leads to significantly red-shifted spectra. The first crystal structure for non-fluorescent CP, Rtms5, from Montipora efflorescens demonstrated that the DsRed-like chromophore here exists in unusual trans-conformation and is nonplanar (Prescott et al., 2003). Another conformational state of a DsRed-like chromophore was found in far-red fluorescent protein eqFP611 from Entacmaea quadricolor (Petersen et al., 2003). The chromophore of this protein is in the trans-conformation, but in contrast to Rtms5, it is coplanar. Thus, a coplanar chromophore probably ensures high fluorescence quantum yield, while a trans non-planar conformation is characteristic for non-fluorescent chromoproteins. Generally, the DsRed-like structure appeared to be the most widespread chromophores among red-shifted fluorescent and chromoproteins. Other modifications, however, have been reported. In the photoconvertible Kaede (Ando et al., 2002) a red-emitting chromophore is formed upon UV irradiation. The process results in the cleavage of the protein backbone between the aC and amide N of His65 and the formation of a double bond between the aC and bC of His65 (Mizuno et al., 2003). Crystallographic studies showed that the yellow fluorescent protein zoanYFP (zFP538) from Zoanthus contains a three-ring chromophore (Remington et al., 2005). The third ring was proposed to result from a transimination reaction in which a transiently appearing DsRed-like acylimine is attacked by the terminal amino group of Lys65, cleaving the polypeptide backbone. The chromophore within a photoactivatable mutant of the asulCP (asFP595) chromoprotein was found to be in trans non-planar conformation (Quillin et al., 2005). Moreover, a protein backbone break between Cys64 and Met65 results in a novel chemical structure consisting of C=O group at aC of Met65 in conjugation with GFP-like chromophore (Quillin et al., 2005; Yampolsky et al., 2005). Thus, together with the ordinary GFP-like chromophore, at least six chromophore structures are embodied in coral’s fluorescent and chromoproteins, which are distributed between the above-mentioned color classes in such a way that each color class can be characterized by a specific type (or types) of chromophore. It would be very desirable from both scientific and practical points of view to achieve understanding the correlation between protein structure and spectral properties. In spite of the fact that the number of GFP-like proteins that have been cloned to date is quite significant (about 100), it is still difficult to pinpoint amino acid positions that are responsible for a particular type of spectra. Some preliminary conclusions, however, can be made on the basis of structural and mutagenesis data, as will be discussed below.

6.3

SEQUENCE COMPARISON

The amino acid sequence identity of Anthozoa fluorescent proteins (FPs) with Aequorea victoria GFP is rather low, about 25%. Nevertheless, all key secondary structure elements of GFP including 11 b-strands and turn motifs between them are clearly detectable in their sequences (Fig. 6.2). Amino acid alignment demonstrates conservation of many positions (especially glycines, aspartates, and prolines) apparently important for organization of secondary and tertiary protein structure. Thus, the overall b-can fold structure of GFP (Ormo et al., 1996; Yang et al., 1996) looks well-conserved in the Anthozoa FPs. This conclu-

125

SEQUENCE COMPARISON

10 GFP dstrGFP hcriGFP zoanGFP zoanYFP mcavRFP DsRed asulCP

20

30

40

50

MSKGEELFTGVVPILVELDGDVNGHKFSVSGEGEGDATYGKLTLKFICTTG-KLPVPWPT MSWSKSVIKEEMLIDLHLEGTFNGHYFEIKGKGKGKPNEGTNTVTLEVTKGGPLPFGWHI ---MCSYIKETMQSKVYMEGKVNDHNFKCTAEGKGEPYKGSQSLTITVTEGGPLPFAFDI MAQSKHGLTKEMTMKYRMEGCVDGHKFVITGEGIGYPFKGKQAINLCVVEGGPLPFAEDI MAHSKHGLKEEMTMKYHMEGCVNGHKFVITGEGIGYPFKGKQTINLCVIEGGPLPFSEDI ----MSVIKSVMKIKLRMEGSVNGHNFVIVGEGEGKPYEGTQSMDLTVKEGAPLPFAYDI MRSSKNVIKEFMRFKVRMEGTVNGHEFEIEGEGEGRPYEGHNTVKLKVTKGGPLPFAWDI ---MASFLKKTMPFKTTIEGTVNGHYFKCTGKGEGNPFEGTQEMKIEVIEGGPLPFAFHI

60 GFP dstrGFP hcriGFP zoanGFP zoanYFP mcavRFP DsRed asulCP

70

80

90

100

110

LVTTFSYGVQCFSRYPDHMKQHDFFKSAMPEGYVQERTIFFKDDGNYKTRAEVKFE--GDT LCPQFQYGNKAFVHHPDDIP--DYLKLSFPEGYTWERSMHFEDGGLCCITNDISLT--GNC LSHAFRYGNKVFAKYPKDHP--DFFKQSLPEGFTWERVSNYEDGGVLTVKQETSLE--GDC LSAAFNYGNRVFTEYPQDIV--DYFKNSCPAGYTWDRSFLFEDGAVCICNADITVSVEENC LSAGFKYGDRIFTEYPQDIV--DYFKNSCPAGYTWGRSFLFEDGAVCICNVDITVSVKENC MTTVFHYGNRVFAKYPKHIP--DYFKQVFPEGYSWERSMNFEDGGICTARNEITME--GDC LSPQFQYGSKVYVKHPADIP--DYKKLSFPEGFKWERVMNFEDGGVVTVTQDSSLQ--DGC LSTSCMYGSKTFIKYVSGIP--DYFKQSFPEGFTWERTTTYEDGGFLTAHQDTSLD--GDC

120 GFP dstrGFP hcriGFP zoanGFP zoanYFP mcavRFP DsRed asulCP

130

140

150

160

170

LVNRIELKGIDFKEDGNILGHKLEYNYNSHNVYIMADKQKNGIKVNFKIRHNIEDGSVQL FNYDIKFTGLNFPPNGPVVQKKTTG-WEPSTERLYPR--DGVLIGDIHHALTVEGGGHYV IICKIKAHGTNFPADGPVMQKRTNG-WEPSTETVIPR-GGGILMRDVPALKLLGNKGHLL MYHESKFYGVNFPADGPVMKKMTDN-WEPSCEKIIPVPKQGILKGDVSMYLLLKDGGRLR IYHKSIFNGMNFPADGPVMKKMTTN-WEASCEKIMPVPKQGILKGDVSMYLLLKDGGRYR FFNKVRFDGVNFPPNGPVMQKKTLK-WEPSTEKMYVR--DGVLTGDINMALLLEGGGHYR FIYKVKFIGVNFPSDGPVMQKKTMG-WEASTERLYPR--DGVLKGEIHKALKLKDGGHYL LVYKVKILGNNFPADGPVMQNKAGR-WEPATEIVYEV--DGVLRGQSLMALKCPGGRHLT

180 GFP dstrGFP hcriGFP zoanGFP zoanYFP mcavRFP DsRed asulCP

190

200

210

220

230

ADHYQQNTPIGDG-PVLLPDNHYLSTQSALSKDPNEKRDHMVLLEFVTAAGITHGMDELYK CDIKTVYRAKK---PVKMPGYHYVDTKLVIRSNDKEFM-KVEEHEIAVARHHPLQSQ CVMETTYKSKK---KVNLPKPHFHHLRMEKDSVSDDEK-TIEQHENVRASYFNDSGK CQFDTVYKAKSV--PRKMPDWHFIQHKLTREDRSDAKNQKWHLTEHAIASGSALP CQFDTVYKAKSV--PSKMPEWHFIQHKLLREDRSDAKNQKWQLTEHAIAFPSALA CDFRTTYRAKKK--GVKLPDYHFVDHSIEILRHDKEYT-EVKLYEHAEAHSGLPRGQRKA VEFKSIYMAKK---PVQLPGYYYVDSKLDITSHNEDYT-IVEQYERTEGRHHLFL CHLHTTYRSKKPASALKMPGFHFEDHRIEIMEEVEKGK-CYKQYEAAVGRYCDAAPSKLGHN

Figure 6.2. Polypeptide sequence alignment of the GFP-like protein family. Representatives from the different color groups are shown. Numbering is based on Aequorea GFP. Introduced gaps are represented by dashes. Invariant Tyr66, Gly67, Arg96, and Glu222 residues are underlined. Residues whose side chains form the interior of the b-barrel are shaded. Amino acid positions crucial for color transitions are labeled in white on black.

sion was proved when the crystal structures of some coral GFP homologs were elucidated (Wall et al., 2000; Yarbrough et al., 2001; Petersen et al., 2003; Prescott et al., 2003; Quillin et al., 2005). In GFP the fluorophore is formed inside the protein globule by modification of amino acids at positions 65–67 (Ser–Tyr–Gly) (Shimomura 1979; Cody et al., 1993; Ormo et al., 1996; Yang et al., 1996). In Anthozoa FPs, chromophore-forming Tyr66 and Gly67 are absolutely invariant (to simplify comparison of different proteins we will use numbering in accordance to GFP; see Fig. 6.2). Also, Arg96, which probably participates in chromophore cyclization, occurs in all known GFP-like proteins. Strong conservation of Glu222 suggests great importance of this position. A noticeable feature of Anthozoa GFP-like proteins is presence of 3–4 Trp residues, some of which are positionally well-

126

DISCOVERY AND PROPERTIES OF GFP-LIKE PROTEINS FROM NONBIOLUMINESCENT ANTHOZOA

conserved, that adjoin the fluorophore or chromophore (e.g., at positions 94 and 145) and that cause these proteins to have a pronounced absorption/excitation peak at 280 nm.

6.4

OLIGOMERIC STATE OF ANTHOZOA GFP-LIKE PROTEINS

DsRed is an obligate tetramer even at nanomolar concentrations (Baird et al., 2000; Heikal et al., 2000; Wall et al., 2000; Yarbrough et al., 2001; Vrzheshch et al., 2000; Mizuno et al., 2001; Wiehler et al., 2001; Luonis et al., 2001; Cotlet et al., 2001; GarciaParajo et al., 2001). Several other coral FPs are also oligomeric (Verkhusha et al., 2001; Gurskaya et al., 2001a; Yanuchevich et al., 2002; Petersen et al., 2003; Prescott et al., 2003; Quillin et al., 2005). Probably, all GFP-like proteins from nonbioluminescent Anthozoa are tetrameric, and this state is important for their functioning in coral polyps. Crystal structure of DsRed confirmed the tetramerization of the protein (Wall et al., 2000; Yarbrough et al., 2001). Each DsRed monomer contacts the two adjacent protein molecules by two chemically distinct interfaces: one hydrophobic, the other hydrophilic (Table 6.1). The hydrophobic interface includes a central cluster of closely packed hydrophobic residues surrounded by polar side chains. The hydrophilic interface contains many salt bridges and hydrogen bonds between polar residues and buried water molecules and also includes an unusual “clasp” formed by several C-end residues of each monomer. Comparison of putative interface-forming residues in different FPs demonstrates high conservation of tetrameric structure among Anthozoa GFP-like proteins (Table 6.1). At the same time, some specific features can be found in almost each protein. For instance, rfloRFP probably does not contain salt bridges around the hydrophobic interface because it carries Val19, Pro26, and His124 rather than the interacting Glu19, Lys124, and Glu26 in DsRed. Also, rfloRFP lacks Arg157 that would interact with Glu101 in a hydrophilic interface. Also, all three closely homologous FPs from Zoanthus—zoanGFP, zoanYFP, and zoan2RFP—possess clearly distinct hydrophobic interfaces (differences at positions 109, 124, 126, 128) but very similar hydrophilic interfaces (the only pronounced difference is the absence of the salt bridges between positions 101 and 157 for zoanRFP). A stacking interaction between tyrosines 126th of the adjacent monomers could be suggested for zoanGFP and zoanRFP but not for zoanYFP. zoanYFP and zoanRFP apparently contain no salt bridges between positions 19 and 124. Probably, each protein aims to form homorather than heterotetramer. In addition, fluorescence microscopy of live corals demonstrates that FPs of different colors are usually predominantly expressed in different cells of the organism. We thus believe that the presence of FP heterotetramers in nature suggested by Wall et al. (2000) is unlikely.

6.5

COLOR TRANSITIONS BY MUTAGENESIS

Site-directed and random mutagenesis have been used to identify residues needed for FP color and function. The first color transitions were obtained for the nonfluorescent chromoprotein asFP595 (asulCP) from sea anemone Anemonia sulcata by Lukyanov et al. (2000). It was found that a single amino acid substitution A148S produced a red fluorescent protein. Also, dual-color (green and red) fluorescent mutants were generated. Since then, many mutants with altered color have been described. A number of investigators have generated DsRed mutants with dual-color or green fluorescence (Fig. 6.1L) (Baird et al., 2000; Terskikh et al., 2000, 2002; Wiehler et al., 2001; Mizuno et al., 2001).

127

COLOR TRANSITIONS BY MUTAGENESIS

TABLE 6.1. DsRed Teramer Interface-Forming Residues and Corresponding Residues in Some Other GFP-Like Proteinsa

Hydrophilic interface

Hydrophobic interface

Amino Acid Positions and Their Interactions

a

DsRed (mRFP1)

amaj GFP

zoan GFP

zoan YFP

zoan RFP

rflo RFP

asul CP

hcri CP

T

T



97

V

T

S

S

S

T



105

V

V

V

V

V

V

F

I

 

107

T

T

I

I

I

T

T

T



109

T

S

N

N

S

T

H

H

 

126

I (R)

H

Y

N

Y

M

L

L



128

V (T)

V

V

M

V

A

N

T



184

I (T)

S

V

V

I

I

T

S



19

E

D

E

E

E

V

E

E



124

K

T

K

I

T

H

K

K



26

E

Y

K

K

K

P

Y

Y



101

E

E

E

E

E

E

E

E



157

R (E)

C

K

K

S

C

V

E

 

166

H (K)

T

S

S

S

A

L

V



180

E

Q

Q

Q

Q

D

H

H



153

Y

T

I

M

I

F

Y

Y



176

H

N

R

R

R

H

H

R



178

L (D)

R

R

R

R

R

T

I



151

R

K

K

K

K

I

I

V



168

A (R)

F

Y

Y

Y

S

A

A



146

E

D

E

E

E

E

E

E



198

Y (A)

N

W

W

W

Y

F

F



200

Y (K)

V

F

F

F

F

F

F



147

A

P

P

A

P

P

P

P



149

T

F

C

C

C

T

T

T



229

H (S)

T

S

S

S

S

C

S

 

230

L (T)

S

A

A

A

P

D

D



231

F (G)

V

L

L

L

L

A

L



232

L (A)

F

P

A

P

Q

A

P



223

R

H

H

H

H

G

A

A



225

E

V

I

I

I

V

V

V



204

K

R

K

K

K

C

R

R

Positions connected by arrows are in direct contact in the adjacent monomers. In DsRed column, mutations for monomeric mRFP1 variant is shown in parentheses.

128

DISCOVERY AND PROPERTIES OF GFP-LIKE PROTEINS FROM NONBIOLUMINESCENT ANTHOZOA

Green

68+112 37, 42, 65, 68, 70, 84, 94, 121, 201, 203, 224

Red

63 + 65, 65 + 68 68

148 + 165 148, + 167 165 + 203

Yellow

Chromo

Figure 6.3. Main color transitions achieved by mutagenesis within the four main groups of the GFP-like proteins. Numbers near arrows represent key amino acid residues substituted in mutants with changed color.

We demonstrated interconversion of green and yellow FPs (Fig. 6.1I–K), appearance of additional red peak in cyan FP (Fig. 6.1H), and transformation of cyan and green FPs to forms with intermediate peak fluorescence (Gurskaya et al., 2001b). Both far-red fluorescent mutants of nonfluorescent CPs (Fig. 6.1N) (Gurskaya et al., 2001a) and a true nonfluorescent chromoprotein on the basis of DsRed (Fig. 6.1M) (Bulina et al., 2002) have been made. The main color transitions achieved by mutagenesis and corresponding crucial amino acid positions are shown in Fig. 6.1 (right row) and Fig. 6.3. Using the available 3D structures for GFP and DsRed, one can see that the majority of positions responsible for color change are placed in the nearest proximity to the chromophore. In particular, positions 65, 68, 148, 165, 167, and 203 appeared to be the main color determinants. Noteworthily, some of these positions (e.g., 65, 167, 203) were found to be very important for spectral properties of Aequorea GFP (Remington, 2000). Substitutions at these points resulted in drastic changes in a ratio between protonated and deprotonated forms of the chromophore (Heim et al., 1994, 1995; Ehrig et al., 1995; Patterson and LippincottSchwartz, 2002) or in significant red shift of excitation-emission spectra (Ormo et al., 1996). Special attention has focused on creation of red-shifted FPs to provide an additional color for multicolor labeling and to provide better imagining reagents for expression in heterologous systems such as animal tissues. Animal tissues are much more translucent to far-red light than to shorter-wavelength light. Thus the longer the light wavelength, the higher the sensitivity and a greater efficiency of fluorescence detection in whole-body imaging. The known naturally occurring red FPs have fluorescence maxima no more than 611 nm (Wiedenmann et al., 2002). A further red shift was achieved in mutants of naturally nonfluorescent chromoproteins. HcRed, a far-red fluorescent mutant of the chromoprotein hcriCP from Heteractis crispa, has excitation-emission spectra peaks at 590 and 645 nm (Gurskaya et al., 2001a). At least two types of mutants with altered color properties look very promising. First, a DsRed mutant, E5, that changes color from green to red with time represents a novel tool (“fluorescent timer”) for visualizing the up- and down-regulation of target promoters (Terskikh et al., 2000). Second, conversion of CPs into FPs opened up a novel source of far-red FPs such as HcRed (Gurskaya et al., 2001a). Possibly, even more redshifted proteins can be created if chromoproteins with red-shifted absorption spectra are found.

FP MODIFICATION BY MUTAGENESIS

6.6 FP MODIFICATION BY MUTAGENESIS Great efforts have been devoted to overcoming some properties undesirable for the use of FP as fluorescent labels. Slow folding rate was an especially serious problem for DsRed, the most popular red FP (Baird et al., 2000). Using random mutagenesis, Terskikh et al. (2002) generated a faster maturing mutant, E57. In addition, Bevis and Glick (2002) described very fast DsRed variants with maturation half-times of 0.7–1.3 h. Many FPs form very high molecular weight aggregates in vivo and in vitro (Mizuno et al., 2001; Jakobs et al., 2000; Yanushevich et al., 2002). FP aggregation may impede targeting to cell compartments and interfere with investigation of protein interactions, and it may result in considerable cellular toxicity. Notably, the aggregates retain bright fluorescence, meaning that they contain properly folded protein molecules. One possible cause of such aggregation may be “sticky” hydrophobic patches on the molecular surface (Eaton and Hofrichter, 1990); alternatively, aggregation may result from electrostatic interactions between positively and negatively charged surfaces (Serina et al., 1996; Himanen et al., 1997). The first explanation seems unlikely for DsRed, since the outer surface of the DsRed tetramer does not contain pronounced hydrophobic areas (Wall et al., 2000; Yarbrough et al., 2001). At the same time, a calculation of the DsRed electrostatic potential shows that the protein surface is mostly negatively charged, except for a positively charged short N-terminal region for each monomer. Possibly, each DsRed tetramer can form up to four salt bridges with adjacent tetramers. Once created, this net-like “polymeric” structure should be very stable, due to four valencies for electrostatic interactions. In support of this hypothesis, (Yanushevich et al., 2002) created low-aggregating mutants of several FPs by site-directed substitutions of basic residues near the N-termini for negatively charged or neutral residues. Also, low-aggregating DsRed variants with practically the same mutations were found independently by random mutagenesis (Bevis and Glick, 2002). Still, the most important drawback of Anthozoa FPs as fluorescent labels is their oligomeric nature. Oligomerization does not preclude the use of FPs as reporters of gene expression, but is a grave problem in applications where FP is fused to a target protein, especially if this protein is also oligomeric. Although some successful DsRed protein fusions have been reported (Mas et al., 2000; Ayoob et al., 2001; Charpilienne et al., 2001), often DsRed chimeras formed intracellular aggregates (Mizuno et al., 2001; Lauf et al., 2001; Qin et al., 2001; Campbell et al., 2002). Crystallographic studies provided a basis for mutagenesis of tetramer interfaceforming residues to generate nonoligomerizing FPs. The first successful modification was made on the far-red protein HcRed (Gurskaya et al., 2001a). The L126H substitution within a hydrophobic interface transformed this protein into a dimeric form. A monomeric HcRed has not yet been generated. For DsRed, mutation at the same position (I126R) also resulted in a dimeric protein that was almost non-fluorescent (Campbell et al., 2002). Additional mutagenesis resulted in a bright dimeric DsRed. More importantly, a true monomeric DsRed was generated by a multiple steps of random and site-directed mutagenesis (totally 33 substitutions compared to DsRed). This protein named mRFP1 possessed advanced characteristics: redshifted spectra (excitation-emission at 584 and 607 nm, respectively), absence of residual green fluorescence, and very fast maturation (Campbell et al., 2002). Very recently, different monomeric color variants well-suited for protein labeling have been created on the base of mRFP1 (Shaner et al., 2004; Wang et al., 2004) as well as from-wild type tetrameric fluorescent proteins (Karasawa et al., 2004).

129

130

DISCOVERY AND PROPERTIES OF GFP-LIKE PROTEINS FROM NONBIOLUMINESCENT ANTHOZOA

Additional approaches have been used to overcome oligomerization of Anthozoa RFP fusion proteins or its effects. Covalently head-to-tail linking of two copies of dimeric mutants of DsRed (Campbell et al., 2002) and HcRed (Fradkov et al., 2002) to form tandem dimers produces nonoligomerizing, monomer-like tags. Fusion of tandem HcRed1 with b-actin and fibrillarin demonstrated its superiority in in vivo labeling of fine cytoskeletal structures and tiny nucleoli. The resultant labeling patterns were indistinguishable from those produced by analogous EGFP-fusion constructs. Dimeric DsRed mutant linked in tandem and fused with connexin43 were also properly trafficked to the membrane and successfully formed functional connexon channels, but were unable to assemble a large gap junction. Simultaneous coexpression of Anthozoa FP-tagged proteins with an excess of either fusion partner alone (Lauf et al., 2001) or nonfluorescent mutant of respective free FP (Gavin et al., 2002; Bulina et al., 2003) sometimes rescues the targeting and function of fusion constructs. In the last case, the resulting FP heterotetramers contain only a single target polypeptide and, therefore, can be considered pseudomonomeric. Besides practical applications, heterooligomers provides a useful model for the biophysical characterization of FPs. Tetrameric nature of FPs complicates data interpretation and hides true behavior of monomers because of their very close interactions within tetramers (Garcia-Parajo et al., 2001; Cotlet et al., 2001). We believe that heterotetramers consisting of one fluorescent monomer and three transparent mutant monomers should provide unique information about properties of tetrameric FPs in the pseudomonomeric state in single-molecule and time-resolved studies.

6.7

PHOTOACTIVATABLE PROBES

GFP photobleaching is a widely used approach for tracking intracellular protein movement (White and Stelzer, 1999; Reits and Neefjes, 2001; Lippincott-Schwartz et al., 2001). Although intentional photobleaching is a powerful technique, it does not allow direct tracking of an object’s path or velocity within a living cell. Direct tracking becomes possible only with the introduction of a photoactivatable fluorescent marker. Considerable progress in developing GFP-like proteins, including Anthozoa FPs, that can be activated by light has been achieved in the last few years (Table 6.2) (Marchant et al., 2001; Patterson and Lippincott-Schwartz, 2002; Ando et al., 2002; Chudakov et al., 2003a; Chudakov, 2004; Wiedenmann et al., 2004; Ando et al., 2004).

6.7.1

DsRed “Greening”

Marchant et al. (2001) demonstrated that multiphoton infrared excitation rapidly changes the fluorescence of DsRed from red to green when viewed by conventional epifluorescence. This technique, named DsRed greening, can optically mark individual cells, cellular compartments, and proteins. DsRed greening is based on the selective bleaching of red-emitting monomers within the DsRed tetramers. Single-molecule spectroscopy has shown that most DsRed tetramers contain both green and red fluorescent monomers in various proportions (Garcia-Parajo et al., 2001; Cotlet et al., 2001). In native tetramers, most of the green emission passes to the neighboring red fluorescent monomers through FRET. As a result, the red fluorescence dominates the DsRed emission spectrum. When the red monomers are bleached, the green fluorophores lose their FRET acceptors, and green fluorescence increases up to 2- to

UV-Violet

UV-Violet UV-Violet

Green

UV-Violet

PA-GFP

PS-CFP

Kaede mEosFP

KFP1

Dronpa

Activating Light

3-photon IR UV-Violet

DsRed

Protein

Blue

Blue

No No

No

No

No

Quenching Light

Reversible and Irreversible Reversible

Irreversible Irreversible

Irreversible

Irreversible

Irreversible

Reversibility of Photoactivation

TABLE 6.2. Main Properties of Photoactivatable GFP-Like Proteins

None to Green

None to Red

Green to Red Green to Red

Cyan to Green

None to Green

Red to Green

Fluorescence Changes

No data

30–70

2000 No data

1500

100

15

Contrast

Monomer

Tetramer

Tetramer Monomer

Monomer

Monomer

Tetramer

Oligomeric State

Ando et al., 2004

Patterson and LippincottSchwartz, 2002 Chudakov et al., 2004 Ando et al., 2002 Wiedenmann et al., 2004 Chudakov et al., 2003a

Marchant et al., 2001

Reference

PHOTOACTIVATABLE PROBES

131

132

DISCOVERY AND PROPERTIES OF GFP-LIKE PROTEINS FROM NONBIOLUMINESCENT ANTHOZOA

2.5-fold. Because during the “greening” procedure the red fluorescence decreases 5- to 7-fold, the resulting contrast (change in red-to-green fluorescence intensity ratio) reaches 10–15 times. Disadvantages of this photolabeling technique are that tetrameric DsRed must be used and two fluorescent colors are required. This latter requirement impedes the concurrent use of several fluorescent labels.

6.7.2

UV-Induced Green-to-Red Photoconversion

Ando et al. (2002) characterized a fluorescent protein named “Kaede” (“maple leaf ” in Japanese) from the stony coral Trachyphyllia geoffroyi. When exposed to ultraviolet or violet (350–400 nm) light, Kaede is transformed from a green fluorescent to a bright red fluorescent protein. This activation of Kaede results in greater than a 2000-fold increase in the ratio of red to green fluorescence. Kaede represents an excellent photoactivatable fluorescent marker for organelle and cells photolabeling. However, Keade is unsuitable for fusion protein studies because of its tetrameric structure. Importantly, the monomeric Kaede-like protein, mEosFP, has been recently developed (Wiedenmann et al., 2004). mEosFP is the first monomeric red fluorescent photoactivatable protein reported, making it valuable supplement to the photoactivatable proteins pallet. Chromophore formation in mEosFP, however, occurs at temperatures below 30°C, substantially limiting applications in mammalian cells.

6.7.3

Kindling Fluorescent Proteins

The chromoprotein asulCP (asFP595, asCP) from the sea anemone Anemonia sulcata possesses a unique type of photoconversion (Lukyanov et al., 2000). Initially nonfluorescent, asulCP becomes red fluorescent (kindles), with excitation-emission maxima at 575 and 595 nm, respectively, in response to intense green light irradiation. The protein then relaxes back to its initial nonfluorescent state, or it can be quenched instantly by blue light irradiation. Both kindling in green and quenching in blue light are reversible processes for the wild-type protein. Mutagenesis studies showed that positions 148, 165, and 203 largely determine the kindling effect (Chudakov et al., 2003b). Based on the asulCP mutants properties, we proposed a model whereby kindling is related to trans–cis isomerization of the excited chromophore, from “chromo” to fluorescent state. Before kindling, the asulCP’s chromophore is stabilized in the trans-configuration, instead of known cis-configuration in GFP and DsRed. The kindling effect is caused by the excited chromophore isomerizing to the fluorescent cis-configuration, close to that of GFP and DsRed. Following this logic, the kindling properties were transferred to the two other coral chromoproteins, hcriCP and cgigCP by site-directed mutagenesis at positions 148, 165, and 203 (Chudakov et al., 2003b). Extending this model, we also proposed that cis- and trans-configuration of the chromophore are common for the GFP-like fluorescent proteins and chromoproteins, respectively. The last supposition was recently supported by the first crystal structure for the chromoprotein (Prescott et al., 2003), which showed a trans-configuration of the chromophore. We developed a group of photoactivatable fluorescent proteins: kindling fluorescent proteins (KFPs) based on asulCP, cgigCP, and hcriCP (Chudakov et al., 2003a, 2003b). Initially nonfluorescent, these proteins become red fluorescent for several minutes (“kindle” reversibly) in response to irradiation with intense light of the definite wavelength. The capability of reversible fluorescence kindling is the unique property of KFPs,

133

PHOTOACTIVATABLE PROBES

Fluorescence

which can be very useful in a number of applications. Routinely, reversible kindling allows us to visualize a whole KFP fluorescent pattern and, after that, perform precise irreversible photolabeling. Also, it can be used to monitor fast intracellular movements during a much shorter period than KFP relaxation half-life time. More intense or prolonged irradiation causes irreversible kindling of most KFPs (Fig. 6.4). Kindled KFPs are fluorescent in the red part of spectrum (excitation at about 580 nm, emission at 600–630 nm for different KFPs). Contrast between the kindled and initial KFP reaches 30- to 70-fold for the best variants. The main disadvantage of KFPs is the tetrameric nature of these proteins, which hampers their use in protein fusions. However, KFPs can be successfully used for photolabeling and tracking of cell organelles and living cells (Chudakov et al., 2003a) (Fig. 6.5). Red light penetrates living tissues better, and it is preferable as a fluorescent signal

0

10

20 30 Time (min)

40

Figure 6.4. KFP1 kindling. Reversible (gray line) and irreversible (black line) kindling of KFP1. Fluorescence emission at 600 nm was measured in 560-nm excitation light. Time zero is set at the commencement of irradiation with kindling light (532-nm laser line). Laser irradiation was stopped after 2 minutes for reversible and after 20 minutes for irreversible kindling.

Figure 6.5. Photoactivation and tracking of KFP1 within mammalian cell. Images were taken using Zeiss LSM 510 confocal microscope 24 h after transient cell transfection with a plasmid encoding KFP1 under CMV promoter. (A) Almost no fluorescence could be observed before kindling upon irradiation by 1% power of 1-mW 543-nm laser. (B) The upper part of the cell was shortly irradiated by 100% power 543-nm laser, causing kindling of the KFP1. (C–E) Starting immediately after kindling, diffusion of the activated protein was observed using 1% power laser as an excitation light source.

134

DISCOVERY AND PROPERTIES OF GFP-LIKE PROTEINS FROM NONBIOLUMINESCENT ANTHOZOA

in cell migration studies, such as investigations into morphogenesis, metastasis, and inflammation.

6.7.4

Dronpa

Recently, a green fluorescent protein form Pectiniidae sp. and its monomeric mutant named Dronpa capable of reversible photoconversion have been characterized (Ando et al., 2004). Initially, Dronpa fluoresces green. Intense irradiation with the blue light leads to protein quenching to the non-fluorescent form with absorption peak at 390 nm. Dronpa can be reversibly activated back to the green fluorescent form using irradiation at about 400 nm. Thus, Dronpa requires a “pre-quenching” step before photoactivation. Remarkably, numerous quenching-activation cycles can occur within a single cell.

6.8

CONCLUSIONS AND PERSPECTIVES

The discovery of GFP-like proteins in Anthozoa species has led to a significant expansion of our knowledge on this amazing protein family. The most interesting feature of the novel proteins is a very broad diversity of their spectral properties. For fundamental protein science, this fact promotes investigation of structural basis for the color diversity. For practical needs, FPs of different color are very important for multicolor labeling and FRETbased applications. During the past four years a considerable progress was achieved in both understanding structure of Anthozoa proteins and their improvement for use in in vivo labeling. The area of investigation and engineering of GFP homologs is far from completion. Massive crystallographic, biochemical, and biophysics studies should be performed to provide an adequate understanding of structure–function relations in this protein family and to enable us to create FP variants with desired properties. The evolutionary diversity of GFP-like proteins in animals should be investigated to understand origin and ancestral biological function of these proteins and to provide novel sources of FPs such as the recently discovered copepods with GFPs (Shagin et al., 2004). For biotechnological applications, further development of the following probes appeared to be of the high demand. First, truly far-red FPs should be either found among natural FPs or created on the base of chromoproteins with absorption in far-red region. This direction is especially important for imaging on the level of whole animals since far-red light penetrates through the tissues deeply. Second, a new generation of sensitive and highly contrast fluorescent molecular sensors should be engineered for real-time monitoring changes in intracellular concentration of ions, radicals, secondary messengers, and other key events. Also, several monomeric photoactivatable tags of distinct colors and activating light sensitivity should be developed to track simultaneously different target proteins. Research into fluorescent probes and into novel microscopy and tomography techniques should continue in the next few years to produce novel breakthroughs in the noninvasive imaging of biomolecular function in living systems ranging from single cells to whole animals.

REFERENCES Ando, R., Hama, H., Yamamoto-Hino, M., Mizuno, H., and Miyawaki, A. (2002). An optical marker based on the UV-induced green-to-red photoconversion of a fluorescent protein. Proc. Natl. Acad. Sci. USA 99:12651–12656.

REFERENCES

Ando, R., Mizuno, H., and Miyawaki, A. (2004). Regulated fast nucleocytoplasmic shuttling observed by reversible protein highlighting. Science 306:1370–1373. Ayoob, J. C., Shaner, N. C., Sanger, J. W., and Sanger, J. M. (2001). Expression of green or red fluorescent protein (GFP or DsRed) linked proteins in nonmuscle and muscle cells. Mol. Biotechnol. 17:65–71. Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2000). Biochemistry, mutagenesis, and oligomerization of DsRed, a red fluorescent protein from coral. Proc. Natl. Acad. Sci. USA 97:11984–11989. Bevis, B. J., and Glick, B. S. (2002). Rapidly maturing variants of the Discosoma red fluorescent protein (DsRed). Nat. Biotechnol. 20:83–87. Bulina, M. E., Chudakov, D. M., Mudrik, N. N., and Lukyanov, K. A. (2002). Interconversion of Anthozoa GFP-like fluorescent and non-fluorescent proteins by mutagenesis. BMC Biochem. 3:7. Bulina, M. E., Verkhusha, V. V., Staroverov, D. B., Chudakov, D. M., and Lukyanov, K. A. (2003). Heterooligomeric tagging diminishes non-specific aggregation of target proteins fused with Anthozoa fluorescent proteins. Biochem. J. 371:109–114. Campbell, R. E., Tour, O., Palmer, A. E., Steinbach, P. A., Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2002). A monomeric red fluorescent protein. Proc. Natl. Acad. Sci. USA 99:7877–7882. Chalfie, M. (1995). Green fluorescent protein. Photochem. Photobiol. 62:651–656. Charpilienne, A., Nejmeddine, M., Berois, M., Parez, N., Neumann, E., Hewat, E., Trugnan, G., and Cohen, J. (2001). Individual rotavirus-like particles containing 120 molecules of fluorescent protein are visible in living cells. J. Biol. Chem. 276:29361–29367. Chudakov, D. M., Belousov, V. V., Zaraisky, A. G., Novoselov, V. V., Staroverov, D. B., Zorov, D. B., Lukyanov, S., and Lukyanov, K. A. (2003a). Kindling fluorescent proteins for precise in vivo photolabeling. Nat. Biotechnol. 21:191–194. Chudakov, D. M., Feofanov, A. V., Mudrik, N. N., Lukyanov, S., and Lukyanov, K. A. (2003b). Chromophore environment provides clue to “kindling fluorescent protein” riddle. J. Biol. Chem. 278:7215–7219. Chudakov, D. M. Verkhusha, V. V., Staroverov, D. B., Souslova, E. A., Lukyanov, S., and Lukyanov, K. A. (2004). Photoswitchable cyan fluorescent protein for protein tracking. Nat. Biotechnol. 22:1435–1439. Cody, C. W., Prasher, D. C., Westler, W. M., Prendergast, F. G., and Ward, W. W. (1993). Chemical structure of the hexapeptide chromophore of the Aequorea green fluorescent protein. Biochemistry 32:1212–1218. Cotlet, M., Hofkens, J., Habuchi, S., Dirix, G., van Guyse, M., Michiels, J., Vanderleyden, J., and de Schryver, F. C. (2001). Identification of different emitting species in the red fluorescent protein DsRed by means of ensemble and single-molecule spectroscopy. Proc. Natl. Acad. Sci. USA 98:14398–14403. Dove, S. G., Hoegh-Guldberg, O., and Ranganathan, S. (2001). Major colour patterns of reefbuilding corals are due to a family of GFP-like proteins. Coral Reefs 19:197–204. Eaton, W. A., and Hofrichter, J. (1990). Sickle cell hemoglobin polymerization. Adv. Protein Chem. 40:63–279. Ehrig, T., O’Kane, D. J., and Prendergast, F. G. (1995). Green-fluorescent protein mutants with altered fluorescence excitation spectra. FEBS Lett. 367:163–166. Fradkov, A. F., Chen, Y., Ding, L., Barsova, E. V., Matz, M. V., and Lukyanov, S. A. (2000). Novel fluorescent protein from Discosoma coral and its mutants possesses a unique far-red fluorescence. FEBS Lett. 479:127–130. Fradkov, A. F., Verkhusha, V. V., Staroverov, D. B., Bulina, M. E., Yanushevich, Y. G., Martynov, V. I., Lukyanov, S., and Lukyanov, K. A. (2002). Far-red fluorescent tag for protein labelling. Biochem. J. 368:17–21.

135

136

DISCOVERY AND PROPERTIES OF GFP-LIKE PROTEINS FROM NONBIOLUMINESCENT ANTHOZOA

Garcia-Parajo, M. F., Koopman, M., van Dijk, E. M., Subramaniam, V., and van Hulst, N. F. (2001). The nature of fluorescence emission in the red fluorescent protein DsRed, revealed by singlemolecule detection. Proc. Natl. Acad. Sci. USA 98:14392–14397. Gavin, P., Devenish, R. J., and Prescott, M. (2002). An approach for reducing unwanted oligomerisation of DsRed fusion proteins. Biochem. Biophys. Res. Commun. 298:707–713. Gross, L. A., Baird, G. S., Hoffman, R. C., Baldridge, K. K., and Tsien, R. Y. (2000). The structure of the chromophore within DsRed, a red fluorescent protein from coral. Proc. Natl. Acad. Sci. USA 97:11990–11995. Gurskaya, N. G., Fradkov, A. F., Terskikh, A., Matz, M. V., Labas, Y. A., Martynov, V. I., Yanushevich, Y. G., Lukyanov, K. A., and Lukyanov, S. A. (2001a). GFP-like chromoproteins as a source of far-red fluorescent proteins. FEBS Lett. 507:16–20. Gurskaya, N. G., Savitsky, A. P., Yanushevich, Y. G., Lukyanov, S. A., and Lukyanov, K. A. (2001b). Color transitions in coral’s fluorescent proteins by site-directed mutagenesis. BMC Biochem. 2:6. Heikal, A. A., Hess, S. T., Baird, G. S., Tsien, R. Y., and Webb, W. W. (2000). Molecular spectroscopy and dynamics of intrinsically fluorescent proteins: Coral red (dsRed) and yellow (Citrine). Proc. Natl. Acad. Sci. USA 97:11996–12001. Heim, R., Prasher, D. C., and Tsien, R. Y. (1994). Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proc. Natl. Acad. Sci. USA 91:12501–12504. Heim, R., Cubitt, A. B., and Tsien, R. Y. (1995). Improved green fluorescence. Nature 373:663–664. Himanen, J. P., Popowicz, A. M., and Manning, J. M. (1997). Recombinant sickle hemoglobin containing a lysine substitution at Asp-85(alpha): Expression in yeast, functional properties, and participation in gel formation. Blood 89:4196–4203. Jakobs, S., Subramaniam, V., Schonle, A., Jovin, T. M., and Hell, S. W. (2000). EFGP and DsRed expressing cultures of Escherichia coli imaged by confocal, two-photon and fluorescence lifetime microscopy. FEBS Lett. 479:131–135. Johnson, F. H., Shimomura, O., Saiga, Y., Gershman, L. C., Reynolds, G. T., and Waters, J. R. (1962). Quantum efficiency of Cypridina luminescence, with a note on that of Aequorea. J. Cell Comp. Physiol. 60:85–104. Karasawa, S., Araki, T., Nagai, T., Mizuno, H., and Miyawaki, A. (2004). Cyan-emitting and orangeemitting fluorescent proteins as a donor/acceptor pair for fluorescence resonance energy transfer. Biochem. J. 381:307–312. Labas, Y. A., Gurskaya, N. G., Yanushevich, Y. G., Fradkov, A. F., Lukyanov, K. A., Lukyanov, S. A., and Matz, M. V. (2002). Diversity and evolution of the green fluorescent protein family. Proc. Natl. Acad. Sci. USA 99:4256–4261. Lauf, U., Lopez, P., and Falk, M. M. (2001). Expression of fluorescently tagged connexins: A novel approach to rescue function of oligomeric DsRed-tagged proteins. FEBS Lett. 498:11–15. Lippincott-Schwartz, J., Snapp, E., and Kenworthy, A. (2001). Studying protein dynamics in living cells. Nat. Rev. Mol. Cell Biol. 2:444–456. Lounis, B., Deich, J., Rosell, F. I., Boxer, S. G., and Moerner, W. E. (2001). Photophysics of DsRed, a red fluorescent protein, from the ensemble to the single-molecule level. J. Phys. Chem. B 105:5048–5054. Lukyanov, K. A., Fradkov, A. F., Gurskaya, N. G., Matz, M. V., Labas, Y. A., Savitsky, A. P., Markelov, M. L., Zaraisky, A. G., Zhao, X., Fang, Y., Tan, W., and Lukyanov, S. A. (2000). Natural animal coloration can be determined by a nonfluorescent green fluorescent protein homolog. J. Biol. Chem. 275:25879–25882. Marchant, J. S., Stutzmann, G. E., Leissring, M. A., LaFerla, F. M., and Parker, I. (2001). Multiphoton-evoked color change of DsRed as an optical highlighter for cellular and subcellular labeling. Nat. Biotechnol. 19:645–649. Mas, P., Devlin, P. F., Panda, S., and Kay, S. A. (2000). Functional interaction of phytochrome B and cryptochrome 2. Nature 408:207–211.

REFERENCESS

Matz, M. V., Fradkov, A. F., Labas, Y. A., Savitsky, A. P., Zaraisky, A. G., Markelov, M. L., and Lukyanov, S. A. (1999). Fluorescent proteins from nonbioluminescent Anthozoa species. Nat. Biotechnol. 17:969–973. Matz, M. V., Lukyanov, K. A., and Lukyanov, S. A. (2002). Family of the green fluorescent protein: Journey to the end of the rainbow. Bioessays 24:953–959. Mizuno, H., Mal., T. K., Tong, K. I., Ando, R., Furuta, T., Ikura, M., and Miyawaki, A. (2003). Photo-induced peptide cleavage in the green-to-red conversion of a fluorescent protein. Mol. Cell 12:1051–1058. Mizuno, H., Sawano, A., Eli, P., Hama, H., and Miyawaki, A. (2001). Red fluorescent protein from Discosoma as a fusion tag and a partner for fluorescence resonance energy transfer. Biochemistry 40:2502–2510. Ormo, M., Cubitt, A. B., Kallio, K., Gross, L. A., Tsien, R. Y., and Remington, S. J. (1996). Crystal structure of the Aequorea victoria green fluorescent protein. Science 273:1392– 1395. Patterson, G. H., and Lippincott-Schwartz, J. (2002). A photoactivatable GFP for selective photolabeling of proteins and cells. Science 297:1873–1877. Petersen, J., Wilmann, P. G., Beddoe, T., Oakley, A. J., Devenish, R. J., Prescott, M., and Rossjohn, J. (2003). The 2.0-A crystal structure of eqFP611, a far red fluorescent protein from the sea anemone Entacmaea quadricolor. J. Biol. Chem. 278:44626–44631. Prescott, M., Ling, M., Beddoe, T., Oakley, A. J., Dove, S., Hoegh-Guldberg, O., Devenish, R. J., and Rossjohn, J. (2003). The 2.2A crystal structure of a pocilloporin pigment reveals a nonplanar chromophore conformation. Structure (Camb.) 11:275–284. Qin, H., Shao, Q., Belliveau, D. J., and Laird, D. W. (2001). Aggregated DsRed-tagged Cx43 and overexpressed Cx43 are targeted to lysosomes in human breast cancer cells. Cell Commun. Adhes. 8:433–439. Quillin, M. L., Anstrom, D. M., Shu, X., O’Leary, S., Kallio, K., Chudakov, D. M., and Remington, S. J. (2005). The kindling fluorescent protein from Anemonia sulcata: Dark state structure at 1.38-Å resolution. Biochemistry, 44:5774–5787. Reits, E. A., and Neefjes, J. J. (2001). From fixed to FRAP: Measuring protein mobility and activity in living cells. Nat. Cell Biol. 3:145–147. Remington, S. J. (2000). Structural basis for understanding spectral variations in green fluorescent protein. Methods Enzymol. 305:196–211. Remington, S. J., Wachter, R. M., Yarbrough, D. K., Branchaud, B., Anderson, D. C., Kallio, K., and Lukyanov, K. A. (2005). zFP538, a yellow-fluorescent protein from Zoanthus, contains a novel three-ring chromophore. Biochemistry 44:202–212. Salih, A., Larkum, A., Cox, G., Kuhl, M., and Hoegh-Guldberg, O. (2000). Fluorescent pigments in corals are photoprotective. Nature 408:850–853. Serina, L., Bucurenci, N., Gilles, A. M., Surewicz, W. K., Fabian, H., Mantsch, H. H., Takahashi, M., Petrescu, I., Batelier, G., and Barzu, O. (1996). Structural properties of UMP-kinase from Escherichia coli: Modulation of protein solubility by pH and UTP. Biochemistry 35:7003–7011. Shagin, D. A., Barsova, E. V., Yanushevich, Y. G., Fradkov, A. F., Lukyanov, K. A., Labas, Y. A., Semenova, T. N., Ugalde, J. A., Meyers, A., Nunez, J. M., Widder, E. A., Lukyanov, S. A., and Matz, M. V. (2004). GFP-like proteins as ubiquitous metazoan superfamily: evolution of functional features and structural complexity. Mol. Biol. Evol. 21: 841–850. Shaner, N. C., Campbell, R. E., Steinbach, P. A., Giepmans, B. N., Palmer, A. E., and Tsien, R. Y. (2004). Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nat. Biotechnol. 22:1567–1572. Shimomura, O. (1979). Structure of the chromophore of Aequorea green fluorescent protein. FEBS Lett. 104:220–222.

137

138

DISCOVERY AND PROPERTIES OF GFP-LIKE PROTEINS FROM NONBIOLUMINESCENT ANTHOZOA

Terskikh, A., Fradkov, A., Ermakova, G., Zaraisky, A., Tan, P., Kajava, A. V., Zhao, X., Lukyanov, S., Matz, M., Kim, S., Weissman, I., and Siebert, P. (2000). “Fluorescent timer”: Protein that changes color with time. Science 290:1585–1588. Terskikh, A. V., Fradkov, A. F., Zaraisky, A. G., Kajava, A. V., and Angres, B. (2002). Analysis of DsRed mutants. Space around the fluorophore accelerates fluorescence development. J. Biol. Chem. 277:7633–7636. Verkhusha, V. V., Akovbian, N. A., Efremenko, E. N., Varfolomeyev, S. D., and Vrzheshch, P. V. (2001). Kinetic analysis of maturation and denaturation of DsRed, a coral-derived red fluorescent protein. Biochemistry (Moscow) 66:1342–1351. Vrzheshch, P. V., Akovbian, N. A., Varfolomeyev, S. D., and Verkhusha, V. V. (2000). Denaturation and partial renaturation of a tightly tetramerized DsRed protein under midly acidic conditions. FEBS Lett. 487:203–208. Wall, M. A., Socolich, M., and Ranganathan, R. (2000). The structural basis for red fluorescence in the tetrameric GFP homolog DsRed. Nat. Struct. Biol. 7:1133–1138. Wang, L., Jackson, W. C., Steinbach, P. A., and Tsien, R. Y. (2004). Evolution of new nonantibody proteins via iterative somatic hypermutation. Proc. Natl. Acad. Sci. USA 101:16745–16749. White, J., and Stelzer, E. (1999). Photobleaching GFP reveals protein dynamics inside live cells. Trends Cell Biol. 9:61–65. Wiedenmann, J., Elke, C., Spindler, K. D., and Funke, W. (2000). Cracks in the beta-can: Fluorescent proteins from Anemonia sulcata (Anthozoa, Actinaria). Proc. Natl. Acad. Sci. USA 97:14091–14096. Wiedenmann, J., Ivanchenko, S., Oswald, F., Schmitt, F., Rocker, C., Salih, A., Spindler, K. D., and Nienhaus, G. U. (2004). EosFP, a fluorescent marker protein with UV-inducible green-to-red fluorescence conversion. Proc. Natl. Acad. Sci. USA 101:15905–15910. Wiedenmann, J., Schenk, A., Rocker, C., Girod, A., Spindler, K. D., and Nienhaus, G. U. (2002). A far-red fluorescent protein with fast maturation and reduced oligomerization tendency from Entacmaea quadricolor (Anthozoa, Actinaria). Proc. Natl. Acad. Sci. USA 99:11646–11651. Wiehler, J., von Hummel, J., and Steipe, B. (2001). Mutants of Discosoma red fluorescent protein with a GFP-like chromophore. FEBS Lett. 487:384–389. Yampolsky, I. V., Remington, S. J., Martynov, V. I., Potapov, V. K., Lukyanov, S., and Lukyanov, K. A. (2005). Synthesis and properties of the chromophore of asFP595 chromoprotein from Anemonia sulcata. Biochemistry, 44:5788–5793. Yang, F., Moss, L. G., and Phillips, G. N., Jr. (1996). The molecular structure of green fluorescent protein. Nat. Biotechnol. 14:1246–1251. Yanushevich, Y. G., Staroverov, D. B., Savitsky, A. P., Fradkov, A. F., Gurskaya, N. G., Bulina, M. E., Lukyanov, K. A., and Lukyanov, S. A. (2002). A strategy for the generation of nonaggregating mutants of Anthozoa fluorescent proteins. FEBS Lett. 511:11–14. Yarbrough, D., Wachter, R. M., Kallio, K., Matz, M. V., and Remington, S. J. (2001). Refined crystal structure of DsRed, a red fluorescent protein from coral, at 2.0-A resolution. Proc. Natl. Acad. Sci. USA 98:462–467.

7 EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS Mikhail V. Matz Whitney Laboratory, University of Florida, St. Augustine, FL

Yulii A. Labas Institute of Biochemistry RAS, Moscow, Russia

Juan Ugalde Laboratory of Bioinformatics and Gene Expression, University of Chile, Santiago, Chile

7.1 INTRODUCTION The current dataset of sequenced and spectroscopically characterized GFP-like proteins from representatives of the phylum Cnidaria includes about 140 members that can be classified into four color types: green, yellow, and orange-red fluorescent proteins and purple-blue nonfluorescent chromoproteins. These color types apparently possess specific chromophore structures (Labas et al., 2002; see also Chapter 7 in this volume); it must be noted, however, that there may be more than one chromophore structure per color type. To generate these diversity of chromophores, the proteins need to perform additional and/or different autocatalytic reactions—in other words, act as different enzymes (Cody et al., 1993; Gross et al., 2000; Prescott et al., 2003). In spite of this, three-dimensional structures of proteins belonging to different color types are virtually identical to each other (Ormo et al., 1996; Prescott et al., 2003; Wall et al., 2000; Yarbrough et al., 2001). In addition to cnidarian proteins, a few more GFP homologs are currently known. These are the six recently described fluorescent proteins of the green color type from planktonic copepods (phylum Arthropoda, class Crustacea; Shagin et al., 2004), and Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

139

140

EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS

neither fluorescent nor colored G2F domains of some extracellular matrix proteins, such as nidogens (also called entactins). G2F domains are found in all bilaterian genomes sequenced thus far and represent the most remote members of the protein family, with less than 10% protein sequence identity to Cnidarian fluorescent proteins. Even though this difference is large enough to preclude homology detection between G2F domains and Cnidarian GFP-like proteins by sequence-based search tools such as BLAST, the proteins have strikingly similar three-dimensional structure (Hopf et al., 2001), which is an unequivocal evidence of homology. In contrast to functions related to transformation of light as in Cnidarian GFP-like proteins (discussed below), G2F domains serve as proteinbinding modules. Taken together, the family of GFP-like proteins represents a remarkable case of great functional diversity produced on the basis the protein structure that changes very little (Labas et al., 2002). In this chapter, we review this diversity and speculate about its origins. We also devote specific attention to the evolution of colors in reef-dwelling Anthozoa.

7.2 7.2.1

FUNCTIONS OF GFP-LIKE PROTEINS Bioluminescent Organisms

Green fluorescent protein (GFP) was discovered in hydroid medusa Aequorea victoria (synonyms A. forskalea, A. aequorea) more than 40 years ago (Johnson et al., 1962; Shimomura et al., 1962; see also Chapter 1 of this volume). GFPs were also detected in several other bioluminescent Hydrozoa and Anthozoa species, such as other jellyfish, sea pansies, sea pens, and hydroid polyps (Johnson et al., 1962; Morin and Hastings, 1971; Szent-Gyorgyi et al., 2001; Ward and Cormier, 1979). In all these examples, GFPs play the role of secondary emitter: They accept the energy from primary light-producing protein and reemit it, thereby making the system glow green instead of blue (Morin and Hastings, 1971; Morise et al., 1974b; Ward and Cormier, 1976). This energy transfer is radiationless (Morin and Hastings, 1971; Ward and Cormier, 1979) and therefore requires binding of GFP to primary emitter protein (in Aequorea it is a Ca2+-activated photoprotein called aequorin; in Renilla it is a luciferase–oxyluciferin excited-state complex). One Renilla luciferase molecule complexes with one GFP homodimer, which involves electrostatic interactions (Ward and Cormier, 1979). Presumably, in Aequorea system, aequorin and GFP form an electrostatically bound heterotetramer (Cutler and Ward, 1997). Renilla GFP itself exists as a tight dimer (Ward and Cormier, 1979), while Aequorea GFP dimerizes only at high protein concentrations (Cutler and Ward, 1997), with the dimerization constant being 0.11 mM (Zacharias et al., 2002). Thus far, all the known natural GFP-like proteins from bioluminescent organisms are green. Their biological significance is not yet clear, especially taking into account that it is not yet clear why their host animals glow at all (Buck, 1978). Still, it is sometimes said that their action in bioluminescence—production of green light instead of blue—might help to achieve better penetration of emission light through oceanic water (Miyawaki, 2002). Although water in open ocean is in fact most transparent to blue light, green indeed may become the least attenuated when there is noticeable amount of chlorophyll in the water—a common situation near shore or during seasonal phytoplankton blooms in temperate waters (Jerlov, 1976). From this standpoint, it seems logical that shallow-water benthic organisms such as sea pens and all of the hydroid polyps, as well as pelagic organisms from temperate regions such as jellyfish Aequorea, show green color of biolumines-

FUNCTIONS OF GFP-LIKE PROTEINS

cence rather than characteristic blue of most other luminous creatures of the ocean (Herring, 1978). However, it is not immediately clear why these organisms would need remote detection in the first place. Indeed, their luminescence is classified as deterring— that is, aimed to scare off a potential predator or any other organism that might inadvertently cause contact damage (e.g., Hastings, 1995; Labas, 1977; Mackie, 1995a; Morin, 1976). Most importantly, it is triggered by mechanical contact, meaning that the subject is in the immediate vicinity, so there seems to be no need for the increased light transmission through the water unless with a purpose to warn other conspecific individuals. Although such intraspecific signaling was described in planktonic tunicates (Mackie, 1995b) and some ctenophores (Labas, 1980), it is unknown in GFP-containing Cnidaria. The “transmission-increase” hypothesis can be rescued by assuming that bioluminescence of the green-emitting animals actually serves as the “burglar alarm” (Burkenroad, 1943)— that is, is designed to reveal the location of primary predator (or disturber) to secondary predators (Fleisher and Case, 1995). A more plausible explanation, however, would be that green light appears brighter to a potential attacker due to adaptations of the attacker’s visual pigments. In Aequorea bioluminescent system, GFP binding condenses all the emission energy into a narrow green peak at 508 nm, rather than having it spread over a wide spectral region between violet and yellow. This actually makes the peak intensity of luminescence about twice higher in presence of GFP than in its absence (Gorokhovatsky et al., 2003). Such a “monochromized” emission can be expected to elicit higher response in subjects if it matches their peak of visual sensitivity. Indeed, many coastal fishes and crustaceans that are active during the day have maximum of their vision sensitivity around 500 nm (Marshall et al., 2003; Partridge, 1990; Partridge and Cummings, 1999). An alternative explanation of GFPs presence in bioluminescent animals may be that their function is not directly related to the change in emission color, but rather helps to improve the efficiency of bioluminescence. Thus, the luminescence quantum yield can be increased due to the possibility of energy transfer from the primary light-producing enzyme with inefficient chromophore to an alternative efficient chromophore, before the energy is dissipated into heat. Binding of light-producing protein to more than one secondary emitter molecule would further increase this chance. This solution appears especially likely as a rescue scenario, useful in the cases when the originally efficient bioluminescent system became suboptimal due to accumulation of random mutations. In favor of the possibility of such a scheme is the fact that luminescence quantum yield of Renilla luciferase is approximately threefold lower in the absence of GFP than in complex with GFP (Ward and Cormier, 1979). In Aequorea system, however, the quantum yield does not change as a result of complexing of aequorin with GFP (Gorokhovatsky et al., 2003; Morise et al., 1974a). Was GFP an original team member within emerging bioluminescent system, or was it recruited later? The recruitment variant certainly appears preferable, since it breaks the process of evolution of a complex system into two steps (Lenski et al., 2003). Then, where was GFP and what was it doing before getting a new job in bioluminescence? This was exactly the line of reasoning that in 1999 lead out team to the study of fluorescent, but nonbioluminescent, Anthozoa (Matz et al., 1999).

7.2.2

Nonbioluminescent Organisms

Hydroid Medusae. Recently, several GFP-like proteins were cloned from hydromedusae other than Aequorea (Shagin et al., 2004). Unfortunately, with the exception of one

141

142

EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS

medusa identified as Phiallidium sp. (suborder Leptomedusae, same as Aequorea), these specimens could not be identified beyond the suborder Anthomeduzae. None of these specimens exhibited bioluminescence in vivo, although Phialidium is known to have green bioluminescence, and several luminous species are known within suborder Anthomedusae (Herring, 1978). Notably, the protein from Phialidium was not green, but yellowfluorescent, and a protein from one of the anthomedusae was clearly belonging to nonfluorescent (chromoprotein) color type. All these proteins, together with GFPs from Aequorea species, were contained within a single clade in the phylogenetic tree; the chromoprotein was basal and all three leptomedusae sequences were within one of the subclades (Shagin et al., 2004). This arrangement indicates that color diversity in Hydroida evolved independently from the color diversification events seen elsewhere in the phylogenetic tree (see below). In addition, the basal position of chromoprotein suggests that the Hydroida chromoprotein lineage separated from the fluorescent proteins before separation of suborders Anthomedusae and Leptomedusae. Functions of these proteins in Hydroida, especially of the nonfluorescent one, are mysterious; it is hoped that the upcoming studies will soon clarify this issue. Genes Coding for GFP-like Proteins in Corals. It is currently believed that the major color determinants of the reef-building corals are GFP-like proteins, which are responsible for the majority of colors superimposed upon the overall brownish hue provided by endosymbiotic zooxanthellae (Dove et al., 2001; Lukyanov et al., 2000; Matz et al., 1999). A unique feature of a GFP-like protein as a pigment is that the structure and molecular environment of its mature chromophore, and therefore the resulting color, are determined by a single molecular sequence, namely, by the sequence of the GFP-like protein itself. This fact provides a unique opportunity to apply the comprehensive suite of methods for molecular sequence analysis to directly address questions related to color evolution (Matz et al., 2002). Two of the basic questions regarding the evolution of anthozoan colors that can be addressed even without knowledge of their exact function are, How many genes for GFPlike proteins are there in a single coral species? and How are intraspecific color variations produced? Many corals exhibit these variations, ranging from fluorescent blue-green to fluorescent red and nonfluorescent purple or blue (Mazel, 1995, 1997; Veron, 2000). Two alternative mechanisms of generating diversity of coloration are possible: polymorphism and phenotypic plasticity (polyphenism). Polymorphism is the situation when the color appearance of an organism is determined by combination of alleles at the color-coding loci. This explanation of color diversity in corals seemed particularly attractive considering the relative ease of certain types of color conversion by random mutations (Gurskaya et al., 2001b; Lukyanov et al., 2000; see also Chapter 7 in this volume). In contrast, phenotypic plasticity, or polyphenism, would mean the existence of the same collection of genes coding for GFP-like proteins in all color morphs; the differences in color appearance are due to the changes in relative levels of expression of these genes. In this case, the color can be a much more flexible character during the organism’s lifetime than it is allowed under the polymorphism model. In our recent study (Kelmanson and Matz, 2003) we undertook exhaustive cloning of cDNAs for GFP-like proteins from three colonies of the great star coral Montastraea cavernosa (Scleractinia, Faviida), which represented distinct color morphs: red, green, and mixed green-red. In complete agreement with a phenotypic plasticity scenario, these specimens were found to express the same collection of GFP-like proteins, produced by at least four, and possibly up to seven, different genetic loci. These genes code for three basic

143

FUNCTIONS OF GFP-LIKE PROTEINS

Figure 7.1. Fluorescent properties of GFP-like proteins found M. cavernosa. Horizontal axis shows wavelength in nanometers; vertical axis shows fluorescence intensity; solid lines show excitation, dotted lines show emission. Positions of maxima are indicated for each curve.

colors—cyan, green, and red (Fig. 7.1)—and are expressed differently relative to one another in different morphs; specifically, we found that intensity of a spectroscopic signature of a particular protein is proportional to the abundance of corresponding mRNA. Phylogenetic analysis of the new sequences indicated that gene lineages leading to these three colors diverged at about the time of separation of coral families (see below for more detail on phylogeny), with the split between cyan proteins and the rest of colors being the most ancient. This latter finding came quite unexpected, since cyan GFP-like proteins were previously considered merely a variation within a green type on the basis of their chromophore structure (Labas et al., 2002). Apparently, at least in studies devoted to color evolution and ecology, cyan should be regarded as a separate type of GFP-like proteins, most probably having specific function in corals. Photoprotection Hypothesis. Despite their prominence in nonbioluminescent corals, the function of GFP-like proteins there remains controversial. The early idea that

144

EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS

GFP-like proteins may serve as ultraviolet-blocking agents (Catala, 1959) has been substantially modified in the course of recent research; at present, photoprotection in the more general sense of the word is being discussed as the most likely functional assignment (Kawaguti, 1944). In the key work on this subject, Salih et al. (2000) demonstrated a positive correlation between content of fluorescent proteins and efficiency of photosystem II of the zooxanthellae under excess light stress. On the basis of supporting confocal microscopy data, Salih et al. (2000) suggested that fluorescent pigments might be achieving this effect by means of light-screening, energy dissipation via fluorescence, and scattering of excess visible and ultraviolet light. These function fits best the spectral properties of green fluorescent proteins: In addition to efficient scattering, they would convert the light from the waveband very near the maximum of photosynthesis action spectrum into the waveband where photosynthetic pigments absorb 1.5–2 times less (Kinzie and Hunter, 1987; Kinzie et al., 1984). Notably, Salih et al. (2000) also found that fluorescent morphs are more resistant to mass bleaching of corals during periods of heat stress, presumably because GFP-like proteins provide protection from excess light that may lead to oxidative stress, thereby controlling at least one of the stress components. Although the photoprotection function of fluorescent proteins already made its way into the BBC documentary “Blue Planet,” the issue is still far from being settled. Thus, Mazel et al. (2003) recently argued that in two Caribbean corals of the genus Montastraea, the green fluorescent protein absorption, emission, and reflection must have negligible impact on the level of solar radiation reaching the zooxanthellae, so that the photoprotection mechanisms intrinsic to the photosynthesis systems of zooxanthellae would be much more efficient (Gorbunov et al., 2001). Furthermore, it is clear that the photoprotection function of greens cannot be attributed to fluorescent proteins of other color types, since their absorption/emission bands do not match the peaks of photosynthesis action spectrum. For non-green fluorescent proteins, Salih et al. (2000) suggested that they could serve as enhancers of the photoprotection function of greens by means of fluorescent coupling: The energy transfer from green to yellow-red pigments would result in further conversion of the incoming light into wavelengths barely affecting photosynthesis. The physiological relevance of this mechanism remains to be investigated. Finally, the function of nonfluorescent chromoproteins remains a total mystery. These proteins, which absorb very intensely but do not fluoresce at all, are widespread in sea anemones and Indo-Pacific corals and are most commonly localized to terminal parts of the organism, such as branch or tentacle tips. They were first described in stony coral Pocillopora damicornis under the name “pocilloporins,” and they were found to be neither photoprotectants, nor UV screens, nor photosynthetic accessory pigments (Dove et al., 1995). The only physiological difference that was found between pigmented and nonpigmented morphs of P. damicornis was slightly decreased calcification rate in the pigmented morph (Takabayashi and HoeghGuldberg, 1995). Possible Functions Other than Photoprotection. What else can be the function of GFP-like proteins in nonbioluminescent Anthozoa, except photoprotection? There ought to be something, since GFP-like proteins were cloned from several anthozoans that do not contain zooxanthellae, such as Alcyonarians Dendronephtya and Clavularia (Labas et al., 2002; Matz et al., 1999). Moreover, what is the function of color diversity, which remains largely unexplained even if the photoprotection function is accepted? Addressing this problem is an exciting challenge, since it is possible that during such study a key to the general evolutionary roots of the color diversity of the coral reefs will emerge.

FUNCTIONS OF GFP-LIKE PROTEINS

First of all, it can be imagined that green fluorescent proteins are indeed for photoprotection, while other colors are just “play of nature”—that is, randomly mutated variants of greens that have not been eliminated by natural selection. This explanation of natural color diversity may be valid for some organisms (Fox, 1974), but for corals it seems extremely unlikely. Random mutation and drift would work as the only mechanism generating the color diversity if probabilities of forward and backward conversions between different colors would be approximately equal. As Labas et al. (2002) recently argued, this is not the case with GFP-like proteins, where green type represents a simplest organization and other colors (yellow, red, and nonfluorescent) are more complicated systems since they require at least one additional autocatalytic reaction to synthesize their chromophores. This brings about a very unequal probability of color conversion in different directions, as it is seen from mutagenesis experiments: There are many ways to produce a green protein out of any other color by just a single mutation (Baird et al., 2000; Lukyanov et al., 2000; Terskikh et al., 2000; Wiehler et al., 2001; see also Chapter 6 in this volume), but not the other way around (Gurskaya et al., 2001b). So, it can be expected in the absence of selection for non-green colors all such proteins would eventually be driven into green state by random mutations and drift (Labas et al., 2002). To prove that there is indeed a color-directed natural selection, an in-depth study of distribution of selection pressures among different sites in GFP-like proteins is currently underway in Matz’s lab. Clarification of this issue would also deal with the annoying possibility that coral coloration is just a by-product with no adaptive value of its own, as suggested by Wicksten (1989) for some bright colors in sessile invertebrates. Salih et al. (2000), in addition to their photoprotection hypothesis, also discussed the possibility that in low-light conditions fluorescent proteins may serve the opposite function—namely, act as a light-collecting device—when the fluorescent layer is positioned below the zooxanthellae instead of above them. This photosynthesis-aid function is indeed established for blue-fluorescent pigment of deep-water coral Leptoseris fragilis from Red Sea (Schlichter and Fricke, 1990; Schlichter et al., 1994); however, in this case the pigment is unlikely to be a GFP-like protein since it is extractable by chlorophorm. Observations of distribution of fluorescent granules with respect to zooxanthellae within shade- and light-adapter corals seemed to support the photosynthesis-aid idea (Salih et al., 2000). However, a dedicated study demonstrated that, although the energy of the absorbed light is efficiently transferred between different types of fluorescent proteins within the coral, it never ends up at the photosynthetic pigments of zooxanthellae (Gilmore et al., 2003), which constitutes a strong argument against photosynthesis-aid function. Photoreception is one of the areas where participation of GFP-like proteins can be imagined. Gorbunov and Falkowski (2002) recently undertook a very inventive study of low-light photoreception in stony corals, including great star coral Montastraea cavernosa, for which GFP-like proteins are most extensively characterized (Kelmanson and Matz, 2003). They found that the action spectrum for tentacle retraction in M. cavernosa had a maximum around 480 nm and half bandwidth of about 105 nm, which does not match either absorption or emission maxima of any of the known GFP-like proteins from this coral (Fig. 7.1), but rather resembles an invertebrate rhodopsin. Moreover, the photoreceptors appeared to be localized exclusively in the coral’s tentacles, whereas GFP-like proteins in this species are found in high concentration not only there, but also in oral disk and coenosteum (polyp sides and base). In some hydroid medusae, the position of the photosensitivity maximum (480–550 nm) is closer to what can be expected from a GFP-like protein; however, the overall shape of the photosensitivity spectrum is still very unlike GFP characteristics (Arkett, 1985; Weber, 1982). Therefore, it must be concluded that

145

146

EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS

thus far the experimental data do not support involvement of GFP-like proteins in photoreception. Some researchers, including one of the founders of GFP field William W. Ward, suppose that the color diversity in corals may be aimed to an observer. To put it simply, corals may have different colors to appear colorful—for example, to attract fishes that would weed out the fouling algae (Ward, 2002). To our knowledge, no study has been devoted thus far to determine whether fishes or other reef organisms can use the color of a coral as a recruitment cue. Interestingly, some territorial herbivorous fishes such as damselfishes are able to condition their habitat to fit their needs, thereby acting as important determinants of the community structure (e.g., Ceccarelli et al., 2001; Hata and Nishihira, 2002). If coral colors affect the distribution of such organisms on the reef, the color diversity of corals may indirectly lead to increase of the complexity of the habitat on the scale of the whole reef, bringing about more ecological niches and increase in species richness. In this regard, it is important to ask whether the GFP-like proteins possess necessary spectral characteristics to be well seen underwater. A recent study by Mazel and Fuchs (2003) addressed this issue, demonstrating that orange-red (emission max about 580 nm) and green (emission max about 515 nm) fluorescent proteins are quite suitable to produce strong visual effects; this is because red ones stand out of the blue-green background, and green ones emit at high efficiency and at the wavelengths where human eye is very sensitive (this is also true for the eyes of most reef fishes (Partridge, 1990). In turn, the cyan proteins are expected to produce less of a visual effect due to minimal spectral shift in comparison to downwelling light and relatively unsaturated color. The study of Mazel and Fuchs does not include non-fluorescent chromoproteins. Since these proteins have their visual effect through absorption, their colors will be seen only when they are irradiated by relatively full-spectrum light (Lukyanov et al., 2000; Gurskaya et al., 2001a). This light is available only in shallow water, where the corresponding wavelengths are not yet depleted by the water column (Jerlov, 1976). So, one would predict that if nonfluorescent coloration is mainly for the sake of colorfulness, the nonfluorescent colors would be most abundant in shallow waters. This expectation is not fulfilled in the case of Pocillopora damicornis, pink morphs of which tended to be more abundant in deeper water than on reef top (Takabayashi and Hoegh-Guldberg, 1995). Still, on the scale of the whole reef rather than for a single species, non-fluorescently colored corals indeed tend to be more abundant in shallow habitat such as reef flat (Salih, 2005). Finally, there is a possibility that coloration has no adaptive value in adult corals, but is important at a larval, recruit, or juvenile stage. It is tempting to speculate that GFP-like proteins may be involved in negotiations between zooxanthellae and coral while establishing the symbiotic relationships. In most corals, this happens during immediate postsettlement period, which is also known as the most critical stage in a lifetime of a coral (e.g., Van Woesik, 2000). Note on Oligomerization. There is one peculiar feature that apparently unites all GFP-like proteins from nonbioluminescent Anthozoa irrespective of their color: They all form oligomeric complexes (Baird et al., 2000; Gurskaya et al., 2001a; Wall et al., 2000; Yanushevich et al., 2002; Yarbrough et al., 2001; see also Chapter 7 in this volume). The minimal complex usually includes four monomers; only the red fluorescent protein from sea anemone Entacmaea quadricolor was demonstrated to form less then four-member oligomers (Wiedenmann et al., 2002). In the most well-studied red-emitting anthozoan protein DsRed, which is tetrameric, this oligomerization is a prerequisite for proper chro-

FUNCTIONS OF GFP-LIKE PROTEINS

mophore maturation, so that straightforward disruption of the interface between monomers by mutagenesis leads to loss of fluorescence (Campbell et al., 2002; Sacchetti et al., 2002; see also Chapter 6 in this volume). From the point of view of functionality, it is interesting to note that in red proteins that have a green-emitting intermediate maturation stage, oligomerization leads to pure red fluorescence output despite significant proportion of the protein pool remains green (Garcia-Parajo et al., 2001). This happens due to efficient resonant energy transfer between monomers within the complex (Heikal et al., 2000), so that the energy absorbed by still-green monomers gets channeled down into red fluorescence (Marchant et al., 2001). The same mechanism was suggested for another type of redemitting GFP-like proteins from stony corals on the basis of the shape of excitation and emission spectra (Labas et al., 2002). It therefore may be possible that oligomerization of Anthozoa fluorescent proteins evolved specifically for the purpose of rescuing the inefficiently maturing red-shifted fluorescence. This idea is similar to the one described earlier concerning possible efficiency-improving function of GFP-like proteins in bioluminescent systems. However, in conjunction with the fact that all known GFP-like proteins from subclass Zoantharia (hexacorals), including green ones, are oligomeric, this would imply that they all descended from a red-shifted ancestor. Although the possibility that some natural green proteins may have descended from a red ancestor has been suggested (Labas et al., 2002), it seems rather unlikely that the whole tree of Zoantharian GFP-like proteins had a red root. Nevetheless, it would be very interesting to address this issue using phylogenetic approaches that allow reconstruction of ancestral protein sequences, coupled with synthesis and study of predicted ancestral genes (Ugalde et al., 2004).

7.2.3

Copepoda GFP-like Proteins

Recently, six green GFP-like proteins were cloned from four Copepoda species of family Pontellidae that display bright green fluorescence (Shagin et al., 2004; Fig. 7.2), putting the end to the idea that fluorescent GFP-like proteins represent a Cnidaria-specific innovation. Spectral properties of these new proteins in native and denatured states are very similar to GFP from Aequorea victoria in analogous conditions, suggesting that they possess identical chromophore (strictly speaking, identical to GFP chromophore in deprotonated state). They also appear to be monomeric according to gel-filtration test, again, similarly to GFP from A. victoria. Extending the analogy with GFP, it would be tempting to associate their function with bioluminescence, since order Copepoda harbors many luminescent species, if not for the fact that no such species have been found in the Pontellidae family (Herring, 1988). It can be imagined that the species containing GFPlike proteins are in fact luminescent, but simply escaped their characterization as such, maybe due to seasonal and/or regional variations in bioluminescence capability. Recognizing this possibility, Shagin et al. (2004) nevertheless suggested an attractive alternative: Copepods might use pronounced differences in fluorescence localization between the species (Fig. 7.2) for visual recognition of conspecific individuals. Obviously, function of GFP-like proteins in Copepoda must await future studies.

7.2.4

G2F Domains

G2F is a name of profile in SMART protein domain database (Schultz et al., 2000), which stands for “Globular-2 fragment” and refers to a part of human multidomain protein of basal membranes called nidogen or entactin (Hohenester and Engel, 2002). It consists of a canonical EGF (epidermal growth factor-like) domain linked to a domain that is identi-

147

148

EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS

Figure 7.2. (This also appears in color insert.) Copepoda species that yielded fluorescent GFPlike proteins (from Shagin et al., 2004). Images were taken by fluorescent microscope using combined illumination with white light and standard FITC filter set. (a) Pontellina plumata. Inset magnifies the head. (b) Labidocera aestiva. (c) Compare Pontella meadi. See color insert.

cal in its 3D fold to the beta-can of cnidarian GFP-like proteins (Hopf et al., 2001). Apart from nidogens, this GFP-like domain is found in a different modular context in human extracellular matrix proteins related to fibulins (Yanagisawa et al., 2002), in C. elegans hemicentin (Vogel and Hedgecock, 2001), and in a protein from tunicate Halocynthia (GenBank accession number BAA03127). Nidogen-like protein, containing G2F domain,

PHYLOGENY

is also found in Drosophila (GenBank accession number NM__136731). The Tyr-Gly pair that is converted into chromophore in fluorescent GFP-like proteins is not conserved in G2Fs, and the presumably essential catalytic residues inside the barrel are also missing (see Chapters 5, 6, and 7 in this volume). Indeed, none of the G2Fs is fluorescent. Instead, they serve as protein-binding modules, which is particularly well studied for G2F domain of human nidogen: It binds to the third immunoglobulin-like (IG) domain of another extracellular matrix protein called perlecan (Kvansakul et al., 2001). Nidogens are not structural proteins like laminins or collagens, but rather contribute to the correct organization of the extracellular matrix during development; the protein-binding function of G2F domain is essential for this (e.g., Tunggal et al., 2003; Willem et al., 2002). The region of G2F domain that forms the binding surface (corresponding to beta-strands 1, 2, 3, and 11 in GFP) is unusually well conserved among all G2Fs, suggesting that their function is similar (Hopf et al., 2001). Notably, in the tetrameric structure of red fluorescent protein DsRed from corallimorph (Wall et al., 2000; Yarbrough et al., 2001), the corresponding region does not participate in forming the interface between individual monomers, but is exposed on the surface of the tetramer. Although it does not show any increased conservation in fluorescent proteins, one may reason that if coral GFP-like proteins ever bind to anything, this same region ought to be involved simply because there is almost no choice— not much more of a free protein surface remains available. Unfortunately, at least to our knowledge, it is not known for sure which regions of the molecule are responsible for binding of GFPs to light-producing proteins in bioluminescent systems. It can be expected, however, that these will not be the sites involved in GFP dimerization, located on the strands 7, 10, and 11 (Phillips, 1998)—which only slightly overlaps with the binding patch of G2Fs. It is therefore tempting to speculate that protein binding by beta-strands 1, 2, 3, and 11 of the beta-can was the most ancient function of GFP-like proteins, preceding the fluorescent capability.

7.3 PHYLOGENY 7.3.1

Deep-Level Relationships

The evolutionary route of the GFP/G2F protein family could be imagined as two separate lineages that diverged by descent: One, in Cnidaria, retained fluorescent properties, while another, in Bilateria, lost fluorescence and became part of an extracellular matrix protein. It is also possible, however, that these two lineages originated as a result of ancient gene duplication preceding separation of Cnidaria and Bilateria, so that both G2Fs and fluorescent GFPs could be found within a single genome, in Cnidaria as well as in Bilateria. The position of Copepoda fluorescent proteins within the tree helps to choose between these two possibilities. Under the first scenario (“G2Fs for Bilateria, GFPs for Cnidaria”), Copepoda proteins can be either (a) G2Fs that evolved fluorescence independently of GFPs or (b) products of horizontal gene transfer from one of the Cnidarian lineages. Under the second scenario (“ancient duplication”), Copepoda should be seen as an outgroup with respect to Cnidaria, but still much more related to them than to bilaterian G2Fs. Results of phylogenetic analysis (Fig. 7.3A) confirm the second variant (Shagin et al., 2004). First, there is a very strong statistical support for monophyly of the clade uniting deuterostome and protostome G2Fs, which indicates that Copepoda GFPs did not come from that lineage, but were already separate at the time of common ancestor of deuterostomes and protostomes. Second, Copepoda GFPs did not come from any of the known Cnidaria

149

150

EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS

Figure 7.3. (A) Phylogenetic tree of GFP-like proteins and G2F domains. The topology and branch lengths follow the maximum-likelihood tree constructed by PAUP* program (model GTR+G+I). The clade of Anthozoa proteins is represented schematically. The support values above the branches are tree-puzzle indexes obtained by analysis of protein (normal) and cDNA (bold) sequences, as well as by bayesian support from cDNA analysis using MrBayes (underlined). Below the branches are support values obtained by nonparametric bootstrap under the following criteria: maximum likelihood (normal), minimum evolution with LogDet distances (bold), and maximum parsimony (underlined). (B) Details of phylogeny of Anthozoa GFP-like proteins. Only nodes showing maximum likelihood bootstrap support exceeding 50% are shown; support values from minimum evolution and parsimony analysis are usually higher. The four ancient clades that existed before separation of subclasses Alcyonaria and Zoantharia (Labas et al., 2002) are marked in italic. Scale bar: distance of average 0.1 replacements per nucleotide.

PHYLOGENY

lineages either, ruling out the possibility of horizontal gene transfer from there: The current data prohibit such a transfer after separation of Hydrozoa and Anthozoa. Third, Copepoda sequences are significantly more similar to cnidarian GFPs than to G2Fs: The average maximum-likelihood distance between Copepoda and Cnidaria cDNA sequences is 2.2 replacements/nucleotide (2.1–2.2 with 95% confidence), which is about four times less than the average distance between Copepoda GFPs and bilaterian G2Fs (7.8–9.6 replacements/nucleotide with 95% confidence). Accepting the “ancient duplication” scenario implies that the common ancestor of Cnidaria and Bilateria had both G2F and GFP proteins. Moreover, by that time, GFP lineage may have already evolved fluorescence: Copepoda GFPs apparently follow exactly the same structural solution to fluorescence as Cnidaria proteins (Shagin et al., 2004), which is most parsimoniously explained by appearance of the solution only once before Cnidaria and Bilateria separation. In turn, this means that descendants of the fluorescent GFP lineage are likely to pop up in virtually any animal. So, when somebody reports a fluorescent or colored GFP from a flatworm or a monkey, we will be impressed, but not surprised.

7.3.2

Relationships Between Anthozoa Proteins

The phylogenetic tree for Anthozoa GFP-like proteins (Shagin et al., 2004) is shown in Fig 7.3B. It is a magnified and rescaled clade of the complete tree that appears in Fig 7.3A. The nomenclature of the proteins follows the principle proposed by Labas et al. (2002) with one alteration: Cyan proteins (see example in Fig. 7.1), originally regarded as a subtype of greens on the basis of chromophore similarity, are now denoted “CyFP” in recognition of their possible functional specialization (Kelmanson and Matz, 2003). This tree is slightly different from the one in our first report (Labas et al., 2002) since it (i) includes more sequences that recently became available, (ii) excludes some sequences that possibly experienced gene conversion (Kelmanson and Matz, 2003), and (iii) was constructed as a part of the bigger tree that also included Hydrozoa and Copepoda proteins. It is important to note that out of five measures of clade robustness, only four are high enough to support monophyly of Anthozoa (puzzling support values for protein and DNA alignments, bootstrap values for maximum likelihood and maximum parsimony). It is therefore still probable that Anthozoa proteins are paraphyletic with respect to Hydrozoa; that is, Anthozoa may include Hydrozoa proteins as descendants of one of their lineage. Besides excluding sequences that are likely to be a result of gene conversion, we also left only a few sequences to represent some large groups of highly similar proteins, to facilitate computation. Thus, within clade D, only five proteins from great star coral Montastraea cavernosa (sequence names beginning with “mcav”) were kept out of 18 cloned from different color morphs (Kelmanson and Matz, 2003); and the huge group of 60 very similar sequences of chromoproteins from stony corals and corallimorphs that was recently published within a patent by Karan et al. (2002) was represented by a single chromoprotein aaspCP from Acropora aspera (clade B).

7.3.3

Color Diversity

The most important result of phylogenetic analysis concerning color evolution is that the color diversity originated independently within different lineages (Fig. 7.3B, Fig. 7.4). A clade denoted D in Fig. 7.3B shows as many as three apparently independent cases of

151

152

EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS

Figure 7.4. (This also appears in color insert.) (A) Phylogeny of GFP-like proteins (black and dark gray lines) superimposed upon the tree representing standard taxonomy of the organisms from which they were cloned (gray). The extant proteins are represented by rectangles marked according to their color (see legend). Question marks indicate that the extent of distributions of fluorescent GFP-like proteins in Bilateria, as well as possible presence of G2Fs in Cnidaria, is not yet studied. Note multiple diversification events at the base of Anthozoa clade and several independent cases of color separation. The common ancestor of all GFP-like proteins (black dot at the base of the tree) existed before separation of Cnidaria and Bilateria and presumably was neither fluorescent nor colored, but able to bind other proteins. (B) Alternative scenario for Anthozoa clade, involving horizontal gene transfer from Zoantharia to Alcyonaria.

cyan/green and red color diversification: one within a group of Scleractinia proteins, another in Ricordea proteins, and yet another within Alcyonarian branch. We must point out at this moment that fluorescent protein of alcyonarian Dendronephtya, originally described as green (Labas et al., 2002), turned out to be red fluorescent, so it is denoted dendRFP within the tree. The confusion was due to the fact that, similarly to other red proteins of apparently same chromophore organization, Dendronephtya protein required long-wave UV irradiation for the conversion from the green-emitting form into the redemitting form (K. Lukyanov, personal communication). This peculiar feature was first

PHYLOGENY

described for tgeoRFP, or Kaede (Ando et al., 2002), and later confirmed for similar proteins mcavRFP and rfloRFP (M. Matz, unpublished observations). It is tempting to speculate that independent origination of color diversity of reef Anthozoa was due to some environmental and/or ecological factors that were common to all organisms of the reef. Therefore, there is an exciting possibility that identifying such factors would provide a key to the biological significance of coral reef colorfulness in general. The first question to ask is, When did it happen? Many biologists not specializing in zoology tend to assume that reef-building corals have been around for many hundreds of millions of years, since they come from the phylum that branches off at the basis of the animal tree of life. In fact, the first primitive scleractinian corals appeared in the fossil record only in Triassic (about 240 million years ago [Ma], which was the beginning of the dinosaur age), with no known common ancestor, and very rapidly diversified into up to nine suborders (Stanley, 2003; Veron, 1995). Close attention has been given to the origination of color diversity within suborder Faviina (part of clade D), which covers the three-color palette of the great star coral Montastraea cavernosa (Kelmanson and Matz, 2003). On the basis of topology of the phylogenetic tree, it was inferred that gene duplication events that led to the split between the three main colors—cyan, green, and red— happened before the Montastraea genus became separate and, in fact, at about the time of separation of the families Faviidae, Mussidae, and Trachyphyllidae. The upper time boundary for this color split event is provided by the fact that, according to the fossil record, Mussidae and Faviidae separated in the middle to upper Jurassic (about 180–160 Ma; Veron, 1995, 2000). Before that, corals already underwent one radiation in Triassic followed by extinction at the Triassic–Jurassic boundary (Stanley, 2003; Veron, 1995). It seems plausible that color diversity appeared after this first extinction event (206–180 Ma), along with the origination of many modern coral families during the early to middle Jurassic recovery period. However, it is also possible that color diversity originated already during first diversification of corals in Triassic, so the lower time boundary should be pushed back to the first coral appearance, 240 Ma. The boundary is unlikely to go further back in time, since it would imply independent Paleozoic origin of scleractinian body plan within “robust” corals represented in clade D. It would be very interesting to know when the color diversity was produced in other coral lineages—in particular, in suborders that are recorded since Triassic, such as Archaeocoeniina or Fungiina (Veron, 2000), and whether this event can really be linked to development of modern coral reefs as ecosystems. Evolution of red-emitting proteins is especially interesting to look at, since they represent a higher-complexity level of organization in comparison to cyans and greens (Gross et al., 2000; Labas et al., 2002; Shagin et al., 2004; Ugalde et al., 2004) and are currently known from all the four ancient clades, providing a unique example of convergent molecular evolution of a complex feature (red fluorescence). Different structural solutions apparently have been found in different clades. Thus, in clade B, the red-emitters DsRed and dis2RFP possess wide and skewed emission spectra and no fluorescent green intermediate in the course of maturation; zoan2RFP from clade C and equaRFP from clade A show emission spectra similar to DsRed, but proceed through brightly fluorescent green stage during maturation (“fluorescent timer” phenotype); finally, red-emitters from clade D (tgeoRFP, mcavRFP1, rfloRFP, and dendRFP) have a peculiar narrow emission spectrum, exhibit “timer” phenotype, and require long-wave UV-A light to complete maturation (Ando et al., 2002; Labas et al., 2002; Wiedenmann et al., 2002; see also Chapter 7 in this volume for examples of the spectra). Thus far, there are not too many documented cases of convergent evolution at molecular level [see Zakon (2002) for a recent review], and, to

153

154

EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS

our knowledge, none of these cases can be claimed as convergent evolution of complexity. Uniqueness of the situation and extraordinary tractability of the primary function (fluorescence) in GFP-like proteins makes evolution of red fluorescence very promising for in-depth molecular phylogenetic analysis combined with site-specific mutagenesis as a validation tool (Ugalde et al., 2004).

7.3.4

Ancient Diversity of GFP-like Proteins

The grouping of proteins in the Anthozoa tree does not follow color types, chromophore structures, or taxonomic position of the host organisms. The four ancient gene lineages of Zoantharia proteins identified previously [clades A, B, C, and D of Labas et al. (2002)] were supported upon analysis of a more complete dataset that included Hydrozoa and Copepoda sequences, along with Alcyonaria proteins in one of the clades (D). The absence of monophyletic lineages for Alcyonaria and Zoantharia proteins is strongly supported by likelihood-based tests of the tree topology (AU test: P = 0.001, KH test: P = 0.002, SH test: P = 0.002; Kishino and Hasegawa, 1989; Shimodaira, 2002; Shimodaira and Hasegawa, 1999). No previous molecular study suggested that either Alcyonaria or Zoantharia as organisms are polyphyletic (Berntson et al., 2001; Chen et al., 1995; France et al., 1996; Won et al., 2001), so we originally preferred to explain our observation of polyphyly by differential gene loss after a series of ancient duplications and/or differential imprinting of ancient polymorphisms (Fig. 7.4A). This implies that several GFP gene lineages—at least the four corresponding to A, B, C and D clades in the tree—must have existed before separation of the subclasses Alcyonaria and Zoantharia (Labas et al., 2002), as shown in Fig. 7.4A. This scenario, however, creates a dating problem, provided that the divergence between Mussidae and Faviidae proteins (branching off of the scubGFP1 in Fig. 7.3B) most probably happened no earlier than at the Permian–Triassic boundary 250 Ma (see above). Even if this most conservative estimation is used as a calibration point, the Alcyonaria/Zoantharia divergence within clade D would be dated to about 380 Ma (M. Matz, unpublished observations), which is still much later than suggested by fossil record: Definite Alcyonarians are known from lower Ordovician, 480 Ma (Lindstrom, 1978), and may have been already around in pre-Cambrian, 600 Ma (Clarkson, 1998). This discrepancy in dating can be explained in two ways (unless polyphyly of Alcyonaria is allowed): first, by much slower evolution rates of GFP-like proteins in Palaeozoic; or second, by assuming that Alcyonarian proteins of clade D were acquired from Zoantharia relatively recently by means of horizontal gene transfer (Fig. 7.4B). Neither of these variants seems favorable at the present state of knowledge, so the issue must await further studies. GFP-like proteins from two Zoantharia orders—Corallimorpharia (mushroom anemones) and Scleractinia (stony corals)—were also polyphyletic (P < 0.0001 in AU, KH, and SH tests). In contrast to Alcyonaria/Zoantharia story, in this case it is likely that the GFP-like tree topology reflects the true organismal phylogeny. Thus, currently known Scleractinia GFP-like proteins are split into two groups that correspond to the clades of “robust” and “complex” corals, which were first identified on the basis of mitochondrial 16s ribosomal gene analysis (Romano and Palumbi, 1996, 1997). Although in several molecular studies the monophyly of Scleractinia was supported (Berntson et al., 2001; Chen et al., 1995, 2002; France et al., 1996; Won et al., 2001), an extended analysis of 28s rDNA and mitochondrial 16s rDNA from 20 of 24 coral families and all seven suborders challenged the monophyly of Scleractinia, suggesting that aragonite skeleton may

CONCLUSIONS

have evolved independently several times (Romano and Cairns, 2000). Furthermore, Stanley and Fautin (2001) proposed that skeleton in corals is an ephemeral character and that transitions between skeletonized and soft-bodied forms happened many times in both directions in response to changes in seawater chemistry [see Stanley (2003) for a more extended discussion on this subject]. Since soft-bodied order Corallimorpharia is very similar to stony corals and often appears paraphyletic with respect to them (Berntson et al., 2001), it is the most likely group to provide soft-bodied forms for these transitions— and maybe receive back the secondarily soft-bodied ones. The phylogeny of GFP-like proteins may be reflecting this very evolutionary scenario.

7.4 CONCLUSIONS GFP-like proteins perform a variety of functions, although their exact biological roles in many cases are not fully understood. There are two major gene lineages of GFP-like proteins in animals, existing since the common ancestor of Cnidaria and Bilateria (Fig. 7.4). One of these lineages corresponds to fluorescent proteins, a great variety of which are known from Cnidaria and which were recently reported in Arthropoda. Within bioluminescent systems, GFP-like proteins act as secondary emitters, making the luminescence spectrum a sharp green peak instead of wide blue one. This may be an adjustment for better fit to visual systems of potential subjects in green coastal waters. Apart from the change of emission color, in some cases the biological significance of GFPs may consist in improvement of bioluminescence efficiency. Fluorescent proteins in nonbioluminescent Anthozoa may be providing photoprotection for endosymbiotic algae; however, this function remains controversial: Intrinsic photoprotection mechanisms of zooxanthellae seem to be more powerful than the expected effects of fluorescent proteins; moreover, under photoprotection hypothesis the diversity of colors remains poorly explained. Among possible alternatives, photosynthesis aid and photoreception functions were rendered unlikely by the recent data, while the idea that corals display their colors to be observed by other members of reef community may deserve better attention. In modern corals and their relatives, color diversity evolved independently several times. One of these cases—origination of cyan, green, and red colors in suborder Faviina—happened no later than mid-Jurassic (about 180–160 million years ago), and it may have happened even earlier in Triassic right after appearance of first stony corals and coral reefs 240 million years ago. Independent evolution of red fluorescence in four lineages provides a unique example of convergent evolution of complexity at the molecular level. The function of recently discovered green fluorescent proteins from apparently nonbioluminescent Copepoda is unknown. Finding of GFP-like fluorescent proteins in Bilateria representatives suggests that these proteins can be present in virtually any animal. G2F domains of some extracellular matrix proteins represent the other major gene lineage of GFP family. G2F domains are neither colored nor fluorescent and serve as protein-binding modules, participating in control of extracellular matrix formation during development. Curiously, it appears that if there is one functional feature that unites all GFP-like proteins, it is not fluorescence, but instead the ability to bind other proteins. It is possible that protein binding by beta-sheets 1, 2, 3, and 11 was part of the function of the common ancestor of all GFP-like proteins, preceding the fluorescence capability.

155

156

EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS

ACKNOWLEDGMENTS We wish to thank Dr. J. E. N. (“Charlie”) Veron (AIMS, Australia) for the discussion on coral systematics and palaeontology, and we wish to thank Dr. Nick V. Grishin (University of Texas) for providing access to computing resources. This work was supported by NIH grant RO1 GM066243-1 and a US Department of Defense (SERDP program grant) to M.V.M. and Russian Foundation for Basic Research grant 02-04-49717 to Y.A.L. J.U.’s participation was supported by Grass Foundation, Latin American exchange program.

REFERENCES Ando, R., Hama, H., Yamamoto-Hino, M., Mizuno, H., and Miyawaki, A. (2002). An optical marker based on the UV-induced green-to-red photoconversion of a fluorescent protein. Proc. Natl. Acad. Sci. USA 99:12651–12656. Arkett, S. A. (1985). The shadow response of a hydromedusan (Polyorchis penicillatus): Behavioral mechanisms controlling diel and ontogenic vertical migration. Biol. Bull. 169:297–312. Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2000). Biochemistry, mutagenesis, and oligomerization of DsRed, a red fluorescent protein from coral. Proc. Natl. Acad. Sci. USA 97: 11984–11989. Berntson, E. A., Bayer, F. M., McArthur, A. G., and France, S. C. (2001). Phylogenetic relationships within the Octocorallia (Cnidaria: Anthozoa) based on nuclear 18S rRNA sequences. Marine Biol. 138:235–246. Buck, J. B. (1978). Functions and evolution of bioluminescence. In Bioluminescence in Action, Herring, P. J., Ed., Academic Press, London. Burkenroad, M. D. (1943). A possible function of bioluminescence. J. Mar. Res. 5:161–164. Campbell, R. E., Tour, O., Palmer, A. E., Steinbach, P. A., Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2002). A monomeric red fluorescent protein. Proc. Natl. Acad. Sci. USA 99:7877–7882. Catala, R. (1959). Fluorescence effect from corals irradiated with ultra-violet rays. Nature 183:949. Ceccarelli, D. M., Jones, G. P., and McCook, L. J. (2001). Territorial damselfishes as determinants of the structure of benthic communities on coral reefs. Oceanogr. Mar. Biol. 39: 355–389. Chen, C. A., Odorico, D. M., Tenlohuis, M., Veron, J. E. N., and Miller, D. J. (1995). Systematic relationships within the Anthozoa (Cnidaria, Anthozoa) using the 5¢-end of the 28s Rdna. Mol. Phylogenet. Evol. 4:175–183. Chen, C. A., Wallace, C. C., and Wolstenholme, J. (2002). Analysis of the mitochondrial 12S rRNA gene supports a two-clade hypothesis of the evolutionary history of scleractinian corals. Mol. Phylogenet. Evol. 23:137–149. Clarkson, E. N. K. (1998). Invertebrate Palaeontology and Evolution, 4th Ed., Blackwell Science, Oxford. Cody, C. W., Prasher, D. C., Westler, W. M., Prendergast, F. G., and Ward, W. W. (1993). Chemical structure of the hexapeptide chromophore of the Aequorea green-fluorescent protein. Biochemistry 32:1212–1218. Cutler, M. W., and Ward, W. W. (1997). Spectral analysis and proposed model for GFP dimerization. In Bioluminescence and Chemiluminescence, Hastings, J. W., Kricka, L, J., and Stanley, P. E., Eds., Wiley, Chichester, pp. 403–406. Dove, S. G., Hoegh-Guldberg, O., and Ranganathan, S. (2001). Major colour patterns of reefbuilding corals are due to a family of GFP-like proteins. Coral Reefs 19:197–204. Dove, S. G., Takabayashi, M., and HoeghGuldberg, O. (1995). Isolation and partial characterization of the pink and blue pigments of pocilloporid and acroporid corals. Biol. Bull. 189:288–297.

REFERENCES

Fleisher, K. J., and Case, J. F. (1995). Cephalopod predation facilitated by dinoflagellate luminescence. Biol. Bull. 189:263–271. Fox, D. L. (1974). Biochromes. Occurrence, distribution and comparative biochemistry of prominent natural pigments in the marine world. In The Significance of Zoochromes, Needham, A. E. Ed., Springer-Verlag, New York. France, S. C., Rosel, P. E., Agenbroad, J. E., Mullineaux, L. S., and Kocher, T. D. (1996). DNA sequence variation of mitochondrial large-subunit rRNA provides support for a two-subclass organization of the Anthozoa (Cnidaria). Molecular Mar. Biol. and Biotechnol. 5:15–28. Garcia-Parajo, M. F., Koopman, M., van Dijk, E., Subramaniam, V., and van Hulst, N. F. (2001). The nature of fluorescence emission in the red fluorescent protein DsRed, revealed by singlemolecule detection. Proc. Natl. Acad. Sci. USA 98:14392–14397. Gilmore, A. M., Larkum, A., Salih, A., Itoh, S., Shibata, Y., Bena, C., Yamasaki, H., Papina, M., and Van Woesik, R. (2003). Simultaneous time resolution of the emission spectra of fluorescent proteins and zooxanthellar chlorophyll in reef-building corals. Photochem. Photobiol. 77:515–523. Gorbunov, M. Y., and Falkowski, P. G. (2002). Photoreceptors in the cnidarian hosts allow symbiotic corals to sense blue moonlight. Limnol. Oceanogr. 47:309–315. Gorbunov, M. Y., Kolber, Z. S., Lesser, M. P., and Falkowski, P. G. (2001). Photosynthesis and photoprotection in symbiotic corals. Limnol. Oceanogr. 46:75–85. Gorokhovatsky, A. Y., Rudenko, N. V., Marchenkov, V. V., Skosyrev, V. S., Arzhanov, M. A., Burkhardt, N., Zakharov, M. V., Semisotnov, G. V., Vinokurov, L. M., and Alakhov, Y. B. (2003). Homogeneous assay for biotin based on Aequorea victoria bioluminescence resonance energy transfer system. Anal. Biochem. 313:68–75. Gross, L. A., Baird, G. S., Hoffman, R. C., Baldridge, K. K., and Tsien, R. Y. (2000). The structure of the chromophore within DsRed, a red fluorescent protein from coral. Proc. Natl. Acad. Sci. USA 97:11990–11995. Gurskaya, N. G., Fradkov, A. F., Terskikh, A., Matz, M. V., Labas, Y. A., Martynov, V. I., Yanushevich, Y. G., Lukyanov, K. A., and Lukyanov, S. A. (2001a). GFP-like chromoproteins as a source of far-red fluorescent proteins. FEBS Lett. 507:16–20. Gurskaya, N. G., Savitsky, A. P., Yanushevich, Y. G., Lukyanov, S. A., and Lukyanov, K. A. (2001b). Color transitions in coral’s fluorescent proteins by site-directed mutagenesis. BMC Biochem. 2:6. Hastings, J. W. (1995). Bioluminescence. In Cell Physiology Source Book, Sperelakis, N., Ed., Academic Press, San Diego, CA. Hata, H., and Nishihira, M. (2002). Territorial damselfish enhances multi-species coexistence of foraminifera mediated by biotic habitat structuring. Journal of Exp. Mar. Biol. Ecol. 270:215– 240. Heikal, A. A., Hess, S. T., Baird, G. S., Tsien, R. Y., and Webb, W. W. (2000). Molecular spectroscopy and dynamics of intrinsically fluorescent proteins: Coral red (dsRed) and yellow (Citrine). Proc. Natl. Acad. Sci. USA 97:11996–12001. Herring, P. J. (1978). Bioluminescence of invertebrates other than insects. In Bioluminescence in Action, Herring, P. J., Ed., Academic Press, London. Herring, P. J. (1988). Copepod luminescence. Hydrobiologia 167/168:183–195. Hohenester, E., and Engel, J. (2002). Domain structure and organisation in extracellular matrix proteins. Matrix Biol. 21:115–128. Hopf, M., Gohring, W., Ries, A., Timpl, R., and Hohenester, E. (2001). Crystal structure and mutational analysis of a perlecan-binding fragment of nidogen—1. Nat. Struct. Biol. 8:634–640. Jerlov, N. G. (1976). Marine Optics. Elsevier, Amsterdam. Johnson, F. H., Gershman, L. C., Waters, J. R., Reynolds, G. T., Saiga, Y., and Shimomura, O. (1962). Quantum efficiency of Cypridina luminescence, with a note on that of Aequorea. J. Cell. Comp. Physiol. 60:85–104.

157

158

EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS

Karan, M., Brugliera, F., Mason, J., Jones, E. L., Dove, S. G., Hoegh-Guldberg, O., and Prescott, M. (2002). Cell visual characteristic-modifying sequences. In Patent: WO 02070703-A 179, Nufarm Australia Limited, The University of Queensland. Kawaguti, S. (1944). On the phsiology of reef corals. VI. Study of the pigments. Palao Trop. Biol. Stn. Stud. 2:617–674. Kelmanson, I., and Matz, M. (2003). Molecular basis and evolutionary origins of color diversity in great star coral Montastraea cavernosa (Scleractinia: Faviida). Mol. Biol. Evol. 20:1125– 1133. Kinzie, R. A., and Hunter, T. (1987). Effect of light quality on photosynthesis of the reef coral Montipora-Verrucosa. Mar. Biol. 94:95–109. Kinzi, R. A., Jokiel, P. L., and York, R. (1984). Effects of light of altered spectral composition on coral zooxanthellae associations and on zooxanthellae in vitro. Mar. Biol. 78:239–248. Kishino, H., and Hasegawa, M. (1989). Evaluation of the maximum-likelihood estimate of the evolutionary tree topologies from DNA-sequence data, and the branching order in Hominoidea. J. Mol. Evol. 29:170–179. Kvansakul, M., Hopf, M., Ries, A., Timpl, R., and Hohenester, E. (2001). Structural basis for the high-affinity interaction of nidogen-1 with immunoglobulin-like domain 3 of perlecan. EMBO J. 20:5342–5346. Labas, Y. A. (1977). Triggering and regulatory mechanisms of ciliary beating in Ctenophores. 1. Coordination of ciliary beating with intracellular luminescence and muscle contraction. Tsitologya 19:514–521. Labas, Y. A. (1980). Luminescent signalling in the Ctenophores. In Theoretical and Practical Importance of Coelenterates, Naumov, D. V., and Stepanjanc, S. D., Eds., Zoological Institute Academy Science USSR, Leningrad, pp. 41–49. Labas, Y. A., Gurskay, N. G., Yanushevic, Y. G., Fradko, A. F., Lukyano, K. A., Lukyanov, S. A., and Matz, M. V. (2002). Diversity and evolution of the green fluorescent protein family. Proc. Natl. Acad. Sci. USA 99:4256–4261. Lensk, R. E., Ofria, C., Pennock, R. T., and Adami, C. (2003). The evolutionary origin of complex features. Nature 423:139–144. Lindstrom, M. (1978). An octocoral from the Lower Ordovican of Sweden. Geol. Palaeontol. 12:41–52. Lukyano, K. A., Fradko, A. F., Gurskay, N. G., Mat, M. V., Laba, Y. A., Savitsk, A. P., Markelov, M. L., Zaraisk, A. G., Zhao, X., Fang, Y., Tan, W., and Lukyanov, S. A. (2000). Natural animal coloration can be determined by a nonfluorescent green fluorescent protein homolog. J. Biol. Chem. 275:25879–25882. Mackie, G. O. (1995a). Defensive strategies in planktonic coelenterates. Mar. Fresh. Behav. Physiol. 26:119–129. Mackie, G. O. (1995b). Unconventional signalling in Tunicates. Mar. Fresh. Behav. Physiol. 26:197–205. Marchant, J. S., Stutzman, G. E., Leissrin, M. A., LaFerla, F. M., and Parker, I. (2001). Multiphoton-evoked color change of DsRed as an optical highlighter for cellular and subcellular labeling. Nature Biotechnol. 19:645–649. Marshall, N. J., Cronin, T. W., and Frank, T. M. (2003). Visual adaptations in crustaceans: Chromatic, developmental, and temporal aspects. In Sensory Processing in Aquatic Environments, Collin, S. P., and Marshall, N. J., Eds., Springer, New York, pp. 343–372. Matz, M. V., Fradko, A. F., Laba, Y. A., Savitsk, A. P., Zaraisk, A. G., Markelov, M. L., and Lukyanov, S. A. (1999). Fluorescent proteins from nonbioluminescent Anthozoa species. Nat. Biotechnol. 17:969–973. Matz, M. V., Lukyanov, K. A., and Lukyanov, S. A. (2002). Family of the green fluorescent prote: Journey to the end of the rainbow. Bioessays 24:953–959.

REFERENCES

Mazel, C. H. (1995). Spectral measurements of fluorescence emission in Caribbean Cnidarians. Mar. Ecol. Prog. Ser. 120:185–191. Mazel, C. H. (1997). Coral fluorescence characteristics: Excitation–emission spectra, fluorescence efficiencies, and contribution to apparent reflectance. Ocean Opt. XIII, SPIE 2963:240–245. Mazel, C. H., and Fuchs, E. (2003). Contribution of fluorescence to the spectral signature and perceived color of corals. Limnol. Oceanogr. 48:390–401. Mazel, C. H., Lesse, M. P., Gorbuno, M. Y., Barr, T. M., Farrel, J. H., Wyman, K. D., and Falkowski, P. G. (2003). Green-fluorescent proteins in Caribbean corals. Limnol. Oceanogr. 48:402–411. Miyawaki, A. (2002). Green fluorescent protein-like proteins in reef anthozoa animals. Cell Struct. Funct. 27:343–347. Morin, J. G. (1976). Probable functions of bioluminescence in Pennatulacea (Cnidaria, Anthozoa). In Coelenterate Ecology and Behavior, Mackie, G. O., Ed., Plenum, New York, pp. 629– 638. Morin, J. G., and Hastings, J. W. (1971). Energy transfer in a bioluminescent system. J. Cell. Physiol. 77:313–318. Morise, H., Shimomur, O., Johnson, F. H., and Winant, J. (1974a). Intermolecular energy-transfer in bioluminescent system of Aequorea. Biochemistry 13:2656–2662. Morise, H., Shimomura, O., Johnson, F. H., and Winant, J. (1974b). Intermolecular energy-transfer in bioluminescent system of Aequorea. Biochemistry 13:2656–2662. Ormo, M., Cubit, A. B., Kallio, K., Gros, L. A., Tsien, R. Y., and Remington, S. J. (1996). Crystal structure of the Aequorea victoria green fluorescent protein. Science 273:1392–1395. Partridge, J. C. (1990). The colour sensitivity and vision of fishes. In Light and Life in the Sea, Herrin, P. J., Campbel, A. K., Whitfield, M., and Maddock, L., Eds., Cambridge University Press, Cambridge, pp. 167–184. Partridge, J. C., and Cummings, M. E. (1999). Adaptation of visual pigments to the aquatic environment. In Adaptive Mechanisms in the Ecology of Vision, Arche, S. N., Djamgoz, M. B. A., Loew, E. R., Partridge, J. C., and Vallerga, S., Eds., Kluwer Academic Publishers, Dordrecht, pp. 251–283. Phillips, G. N., Jr. (1998). The three-dimensional structure of the green fluorescent protein and its implications for function and design. In Green Fluorescent Protein: Properties, Applications and Protocols, Chalfie, M., and Kain, S. R., Eds., Wiley-Liss, New York, pp. 77–96. Prescott, M., Ling, M., Beddoe, T., Oakley, A. J., Dove, S., Hoegh-Guldberg, O., Devenish, R. J., and Rossjohn, J. (2003). The 2.2 A crystal structure of a pocilloporin pigment reveals a nonplanar chromophore conformation. Structure 11:275–284. Romano, S. L., and Cairns, S. D. (2000). Molecular phylogenetic hypotheses for the evolution of scleractinian corals. Bull. Mar. Sci. 67:1043–1068. Romano, S. L., and Palumbi, S. R. (1996). Evolution of scleractinian corals inferred from molecular systematics. Science 271:640–642. Romano, S. L., and Palumbi, S. R. (1997). Molecular evolution of a portion of the mitochondrial 16S ribosomal gene region in scleractinian corals. J. Mol. Evol. 45:397–411. Sacchetti, A., Subramaniam, V., Jovin, T. M., and Alberti, S. (2002). Oligomerization of DsRed is required for the generation of a functional red fluorescent chromophore. FEBS Lett. 525:13–19. Salih, A. (2005). An exploration of light regulating pigments of reef building corals from macro to micro and nano-scales. In Impact Science—Science Foundation for Physics, Nicholls, J. A., and Collins, R. E., Eds., School of Physics, University of Sydney, Sydney. Salih, A., Larkum, A., Cox, G., Kuhl, M., and Hoegh-Guldberg, O. (2000). Fluorescent pigments in corals are photoprotective. Nature 408:850–853. Schlichter, D., and Fricke, H. W. (1990). Coral host improves photosynthesis of endosymbiotic algae. Naturwissenschaften 77:447–450.

159

160

EVOLUTION OF FUNCTION AND COLOR IN GFP-LIKE PROTEINS

Schlichter, D., Meier, U., and Fricke, H. W. (1994). Improvement of photosynthesis in zooxanthellate corals by autofluorescent chromatophores. Oecologia 99:124–131. Schultz, J., Cople, R. R., Doerks, T., Ponting, C. P., and Bork, P. (2000). SMART: A web-based tool for the study of genetically mobile domains. Nucleic Acids Res. 28:231–234. Shagin, D. A., Barsova, E. V., Yanushevich, Y. G., Fradkov, A. F., Lukyanov, K. A., Labas, Y. A., Semenova, T. N., Ugalde, J. A., Meyers, A., Nunez, J. M., Widder, E. A., Lukyanov, S. A., Matz, M. V. (2004). GFP-like proteins as ubiquitous metazoan superfamily: evolution of functional features and structural complexity. Mol. Biol. Evol. 21:841–850. Shimodaira, H. (2002). An approximately unbiased test of phylogenetic tree selection. Syst. Biol. 51:492–508. Shimodaira, H., and Hasegawa, M. (1999). Multiple comparisons of log-likelihoods with applications to phylogenetic inference. Mol. Biol. Evol. 16:1114–1116. Shimomura, O., Johnson, F. H., and Saiga, Y. (1962). Extraction, purification and properties of Aequorin, a bioluminescent protein from luminous hydromedusan, Aequorea. J. Cell. Comp. Physiol. 59:223–239. Stanley, G. D. (2003). The evolution of modern corals and their early history. Earth Sci. Rev. 60:195–225. Stanley, G. D., and Fautin, D. G. (2001). Paleontology and evolution—The origins of modern corals. Science 291:1913–1914. Szent-Gyorgy, C. S., Bryan, B. J., and Szczepaniak, W. (2001). Patent US 6232107-B. Takabayashi, M., and Hoegh-Guldberg, O. (1995). Ecological and physiological differences between two color morphs of the coral Pocillopora damicornis. Mar. Biol. 123:705–714. Terskikh, A., Fradkov, A., Ermakova, G., Zaraisky, A., Tan, P., Kajav, A. V., Zhao, X., Lukyanov, S., Matz, M., Kim, S., Weissman, I., and Siebert, P. (2000). “Fluorescent timer”: Protein that changes color with time. Science 290:1585–1588. Tunggal, J., Wartenberg, M., Paulsson, M., and Smyth, N. (2003). Expression of the nidogen-binding site of the laminin gamma 1 chain disturbs basement membrane formation and maintenance in F9 embryoid bodies. J. Cell Sci. 116:803–812. Ugalde, J. A., Chang, B. S. W., and Matz, M. (2004). Evolution of coral pigments recreated. Science 305:1433. Van Woesik, R. (2000). Modelling processes that generate and maintain coral community diversity. Biodivers. Conserv. 9:1219–1233. Veron, J. E. N. (1995). Corals in Space and Time, UNSW Press, Sydney. Veron, J. E. N. (2000). Corals of the World, Australian Institute of Marine Science, Townsville MC. Vogel, B. E., and Hedgecock, E. M. (2001). Hemicentin, a conserved extracellular member of the immunoglobulin superfamily, organizes epithelial and other cell attachments into oriented lineshaped junctions. Development 128:883–894. Wall, M. A., Socolich, M., and Ranganathan, R. (2000). The structural basis for red fluorescence in the tetrameric GFP homolog DsRed. Nat. Struct. Biol. 7:1133–1138. Ward, W. W. (2002). Fluorescent proteins: Who’s got ‘em and why? In Bioluminescence and Chemiluminescence, Stanley, P. E., and Kricka, L. J., Eds., World Scientific, Cambridge, UK, pp. 123– 126. Ward, W. W., and Cormier, M. J. (1976). In vitro energy-transfer in Renilla bioluminescence. J. Phys. Chem. 80:2289–2291. Ward, W. W., and Cormier, M. J. (1979). An energy transfer protein in coelenterate bioluminescence. Characterization of the Renilla green-fluorescent protein. J. Biol. Chem. 254:781–788. Weber, C. (1982). Electrical activities of a type of electroretinogram recorded from the ocellus of a jellyfish, Polyorchis penicillatus (Hydromedusae). J. Exp. Zool. 223:231–243. Wicksten, M. K. (1989). Why are there bright colors in sessile marine invertebrates? Bull. Mar. Sci. 45:519–530.

REFERENCES

Wiedenmann, J., Schenk, A., Rocker, C., Girod, A., Spindler, K. D., and Nienhaus, G. U. (2002). A far-red fluorescent protein with fast maturation and reduced oligomerization tendency from Entacmaea quadricolor (Anthozoa, Actinaria). Proc. Natl. Acad. Sci. USA 99:11646–11651. Wiehler, J., von Hummel, J., and Steipe, B. (2001). Mutants of Discosoma red fluorescent protein with a GFP-like chromophore. FEBS Lett. 487:384–389. Willem, M., Miosge, N., Halfter, W., Smyth, N., Jannetti, I., Burghart, E., Timpl, R., and Mayer, U. (2002). Specific ablation of the nidogen-binding site in the laminin gamma 1 chain interferes with kidney and lung development. Development 129:2711–2722. Won, J. H., Rho, B. J., and Song, J. I. (2001). A phylogenetic study of the Anthozoa (phylum Cnidaria) based on morphological and molecular characters. Coral Reefs 20:39–50. Yanagisawa, H., Davi, E. C., Starche, B. C., Ouchi, T., Yanagisawa, M., Richardson, J. A., and Olson, E. N. (2002). Fibulin-5 is an elastin-binding protein essential for elastic fibre development in vivo. Nature 415:168–171. Yanushevich, Y. G., Starovero, D. B., Savitsk, A. P., Fradko, A. F., Gurskay, N. G., Bulin, M. E., Lukyanov, K. A., and Lukyanov, S. A. (2002). A strategy for the generation of nonaggregating mutants of Anthozoa fluorescent proteins. FEBS Lett. 511:11–14. Yarbrough, D., Wachte, R. M., Kallio, K., Matz, M. V., and Remington, S. J. (2001). Refined crystal structure of DsRed, a red fluorescent protein from coral, at 2.0-A resolution. Proc. Natl. Acad. Sci. USA 98:462–467. Zacharias, D. A., Violi, J. D., Newton, A. C., and Tsien, R. Y. (2002). Partitioning of lipid-modified monomeric GFPs into membrane microdomains of live cells. Science 296:913–916. Zakon, H. H. (2002). Convergent evolution on the molecular level. Brain Behav. Evol. 59:250–261.

161

8 THE USES OF GREEN FLUORESCENT PROTEIN IN PROKARYOTES Raphael H. Valdivia* and Brendan P. Cormack Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford, CA

Stanley Falkow Rocky Mountain Laboratories, National Institute of Allergy and Infectious Diseases, Hamilton, MT; and Department of Microbiology and Immunology, Stanford University School of Medicine, Stanford, CA

8.1 INTRODUCTION The green fluorescent protein (GFP) of Aequorea victoria is a unique tool that permits the monitoring of gene expression and protein localization in living cells. Green fluorescent protein is stable, does not require cofactors for activity (Chalfie et al., 1994; Inouye and Tsuji, 1994a), and can be functionally expressed in different bacterial species. Because of GFPs unique properties, it can be used as a reporter of gene expression, dynamic processes during bacterial development, and the behavior of single bacteria in complex environments. Several properties of wild-type GFP, however, limit its applications in prokaryotes: (1) GFP tends to precipitate in the cytoplasm as insoluble, nonfluorescent inclusion bodies; (2) the posttranslational formation of a functional chromophore occurs approximately 2 h after synthesis of GFP (Heim et al., 1995); and (3) the magnitude of the fluorescence signal obtained is low compared to that of other reporter proteins [e.g., b-galactosidase (LacZ), chloramphenicol acetyl transferase (CAT), luciferase (Lux)]. Despite these concerns, GFP has already had a significant impact in the fields of bacterial development, pathogenesis, * Present address: Department of Genetics and Microbiology, Duke University, Durham, NC. Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

163

164

THE USES OF GREEN FLUORESCENT PROTEIN IN PROKARYOTES

TABLE 8.1. Examples of Bacterial Species in Which GFP Has Been Successfully Expressed Bacterial Species

References

Bacillus subtilis

Arigoni et al. (1995), Barak et al. (1996), Lewis and Errington (1996), Resnekov et al. (1996), Sharpe and Errington (1996), Webb et al. (1995) Lee and Falkow (unpublished results) Skerker and Shapiro (unpublished results) Chalfie et al. (1994), Cormack et al. (1996), Delagrave et al. (1995), Leff and Leff (1996), Heim et al. (1994), Ma et al. (1996) Covacci (1996) Martin and Tompkins (unpublished results) Dhandayuthapani et al. (1995), Kremer et al. (1995), Via et al. (1996) Valdivia et al. (1996) Dhandayuthapani et al. (1995), Kremer et al. (1995), Valdivia et al. (1996) Licking and Kaiser (unpublished results) Burlage et al. (1996), Christensen et al. (1996) Tombolini et al. (1997) Gage et al. (1996) Kain et al. (1995), Valdivia and Falkow (1996), Valdivia et al. (1996) Valdivia et al. (1996) Hinnebusch et al. (1996)

Bartonella henselae Caulobacter crescentus Escherichia coli Helicobacter pylori Legionella pneumophila Mycobacterium bovis BCG Mycobacterium marinum Mycobacterium smegmatis Myxococcus xanthus Pseudomonas putida Pseudomonas fluorescens Rhizobium meloliti Salmonella typhimurium Yersinia pseudotuberculosis Yersinia pestis

and ecology. Moreover, the advent of new generations of GFP variants with increased sensitivity and solubility should widen the spectrum of uses to which GFP is applied. Green fluorescent protein cDNA has been expressed in a variety of both grampositive and gram-negative bacteria. The bacterial DNA GC content does not seem to pose a barrier to expression, since gfp is expressed in GC rich organisms such as Mycobacteria sp. (Dhandayuthapani et al., 1995; Kremer et al., 1995; Valdivia et al., 1996) and in AT-rich bacteria such as Helicobacter pylori (Covacci, personal communication). Codon optimization has not been required to achieve detectable levels of the fluorescent protein. However, in several eukaryotic systems (see Chapter 12) gfp codon optimization has resulted in increased fluorescence, and it would not be surprising to find that speciestailored optimizations will lead to enhanced fluorescence in prokaryotic systems as well. Table 8.1 shows a summary of the different bacterial species in which GFP has been successfully expressed. While GFP does not require any additional cofactors for fluorescence, the amount of fluorescence is not always proportional to the total pool of GFP. The assembly of the GFP chromophore requires molecular oxygen; therefore GFP fluorescence often decreases under anaerobic or strong reducing environments (Heim et al., 1994; Inouye and Tsuji, 1994b). Temperature can also affect GFP fluorescence because the tendency of GFP to precipitate into nonfluorescent inclusion bodies increases with temperature (Ogawa et al., 1995; Cormack et al., 1996). The fluorescence signal from GFP can be enhanced by growing bacteria aerobically at low temperatures (25–30°C) (Heim et al., 1994; Webb et al., 1995). This requirement is less important when using some of the enhanced GFP mutants discussed in this chapter.

BACTERIAL DEVELOPMENT AND CELL BIOLOGY

Unlike conventional bacterial gene reporters such as lacZ, lux, or cat, GFP is not an enzyme and thus there is no signal amplification derived from multiple substrate cleavage by one molecule of reporter protein. The fluorescence signal in a particular cell, therefore, depends largely on two broad parameters: the rate of synthesis of functional protein and, because GFP is very stable, the rate of dilution as the cell divides. Most reports of GFP expression in bacteria have been performed with gfp on multicopy copy plasmids (Chalfie et al., 1994; Dhandayuthapani et al., 1995; Gage et al., 1996; Hinnebusch et al., 1996; Ma et al., 1996; Valdivia et al., 1996). But single-copy fusions with strong promoters in Pseudomonas sp. have also yielded fluorescent bacteria (Burlage et al., 1996; Christensen et al., 1996; Tombolini et al., 1997). In B. subtilis, single copy gene constructs driving GFP–protein fusions have been routinely imaged, even though the promoters are only of moderate strength (Arigoni et al., 1995; Webb et al., 1995; Lewis and Errington, 1996; Resnekov et al., 1996; Sharpe and Errington, 1996). The rate of dilution is probably lower here than in other systems because B. subtilis ceases to divide after sporulation and translational activity from the GFP protein fusion is essentially cumulative. In contrast, the pool of posttranslationally modified GFP is rapidly diluted in a fast-growing organism. For example, Tombolini et al. (1997) followed the fluorescence of GFP-tagged P. fluorescens by flow cytometry during different stages of growth. Individual bacteria were virtually nonfluorescent during the exponential phase of growth and did not achieve maximal fluorescence until stationary phase. Several recently isolated GFP mutants provide greater sensitivity, faster kinetics of formation, and greater protein solubility (Delagrave et al., 1995; Heim et al., 1995; Crameri et al., 1996; Cormack et al., 1996). These variants improve sensitivity in most, if not all, systems, and some should be sensitive enough as to allow for the imaging of gene fusions driven from moderate strength single-copy constructs, especially with advanced fluorescence imaging systems (CCDcamera, laser scanning confocal microscopes). Indeed, in Salmonella typhimurium we have routinely imaged single copy gfp fusions with the aid of enhanced GFPs (unpublished observations). A review of these mutants is covered in detail elsewhere. The use of these GFP variants has been crucial to some of the applications described below. Where appropriate, we will point out which GFP variant was used in each particular example described.

8.2 BACTERIAL DEVELOPMENT AND CELL BIOLOGY Green fluorescent protein has been successfully used as a tool to study cell differentiation in two models of bacterial development, the cyanobacterium Anabaena and B. subtilis, where cell specific gene expression has been observed with the use of gfp transcriptional and translational fusions. In addition, GFP has been used to visualize bacterial subcellular structures such as the cytoskeletal division apparatus during E. coli division and to observe localization of proteins important in B. subtilis spore formation. The following examples illustrate how GFP is uniquely suited for the study of bacterial development.

8.2.1

Spore Formation in B. subtilis

Under certain environmental conditions, B. subtilis undergoes asymmetrical cell division leading to the formation of a forespore and a mother cell [reviewed in Errington (1993) and in Stragier and Losick (1996)]. A cascade of sigma factors controls the transcription of forespore and mother cell specific genes that lead to different developmental outcomes

165

166

THE USES OF GREEN FLUORESCENT PROTEIN IN PROKARYOTES

Figure 8.1. Fluorescence images of sporulating B. subtilis cells expressing transcriptional and translational GFP fusions. Two sporangia are shown per panel. (A) Forespore-specific expression of a sF-dependent SspE2G-GFP fusion. (B) Mother cell-specific expression of a sE-dependent cotEgfp fusion. (C) Localization of a SpoIVFB-GFP translational fusion (note localized fluorescence seen as a shell at one end of each sporangium). Courtesy of O. Resnekov and C. Webb, Harvard University. See color insert.

(Losick and Stragier, 1992). Understanding cell-specific gene expression during sporulation has been key to the dissection of this developmental program. Cell-specific gene expression has been analyzed either by immunofluorescence or with the use of fluorogenic enzyme substrates (Lewis et al., 1994; Harry et al., 1995; Pogliano et al., 1995). GFP is a powerful addition to these methods. Transcriptional and translational GFP fusions can be used to study cell-specific gene expression and the localization of proteins important in bacterial development, obviating the need to raise antibodies to specific cellular components. Most importantly, GFP allows analyses to be performed in living cells. The potential of GFP for the study of B. subtilis development has been demonstrated in two separate investigations. Webb et al. (1995) visualized cell-specific gene expression in the forespore and the mother cell at different stages of sporulation: Fusions of GFP to either sspE-2G or csfB resulted in forespore-specific fluorescence. By contrast, translational GerE-GFP or transcriptional cotE-gfp fusions showed mother-cell-specific fluorescence (see color Fig. 8.1). A GFP fusion to the spore coat protein CotE was localized to the region surrounding the forespore and appeared as uneven green halos on mature spores. Lewis and Errington (1996), using the S65T variant of GFP (Heim et al., 1995), showed that a DacF-GFP fusion localizes to the forespore. By contrast, a SpoIVA-GFP fusion localized to the mother cell only if a wild-type copy of spoIVA was coexpressed. Full-length SpoIVA fused to GFP did not retain its native function, and the strain was rendered sporulation deficient. This result points out that caution must be exercised when interpreting results from GFP fusions. Optimally, to avoid artifactual results, one would like to verify that any translational GFP fusion can functionally replace the native protein. Another note of caution is indicated by the paradoxical difference in fluorescence intensity between CsfB-GFP and GerE-GFP translation fusions. Analogous LacZ fusions to these two proteins show much higher expression of GerE than CsfB, while the reverse seems to be true for the GFP fusions (Webb et al., 1995). Not withstanding these notes of caution, these experiments show that GFP fusions (transcriptional and translational) behave largely as predicted by other methods. In other work, GFP has played a role in understanding the molecular functions of these proteins. For example, Arigoni et al. (1995) and Barak et al. (1996) independently showed that SpoIIE, a key protein in triggering the developmental fate of progeny after cell division, localizes to a sharp zone close to the pole of sporangia prior to septum formation. Resnekov et al. (1996) using an enhanced GFP mutant (F64S) showed that SpoIVFB, which is proposed to be a proteolytic activator of a mother-cell-specific sigma factor (sk), localizes initially to the sporulation septum and subsequently to the forespore. The use of SpoIVFB-GFP fusions was central to this work since raising and purifying ade-

167

BACTERIAL DEVELOPMENT AND CELL BIOLOGY

quate anti-SpoIVB antibodies for immunofluorescece proved to be exceedingly difficult. The localization of SpoIVFB to the mother cell membrane surrounding the sporangium lends support to a proposed model in which a sporangia protein (SpoIVB) couples development in the forespore to mother cell transcription by activating the proteolytic conversion of pro-sk to sk, thereby activating sk-directed gene expression. Sharpe and Herrington (1996) used GFP (S65T) to analyze DNA transfer. One of the major events during B. subtilis sporulation is the translocation of one of the daughter chromosomes into the forespore. Bacteria deficient in SpoIIIE are unable to translocate the entire chromosome and are arrested after 30% transfer (Wu et al., 1995). The same region of the chromosome is always transferred, suggesting an interaction between a centrosomelike segment of the B. subtilis chromosome and one or more bacterial targeting proteins (Wu and Errington, 1994). Wu and Errington used a forespore-specific dacF::gfp fusion to show that Soj and Spo0J, homologs of proteins needed for efficient plasmid partitioning in other bacterial systems, are important in the specificity of chromosomal translocation. The dacF mutant is located in a region of the chromosome that is not translocated to the forespore in an spoIIIE mutant. Forespore fluorescence from dacF-gfp fusions could not be detected in spoIIIE mutants but could be detected in double mutants of spoIIIE and a deletion of soj-spo0J.

8.2.2

Heterocyst Formation in Anabaena

The cyanobacterium Anabaena grows as long filaments of photosynthetic cells that, under conditions of fixed-nitrogen starvation, undergo genetic changes leading to the development of specialized nitrogen-fixing cells known as heterocysts (reviewed in Haselkorn, 1992). W. J. Buikema and R. Haselkorn (personal communication) showed that fusions of gfp and the developmentally regulated gene hetR were expressed only in the cells destined to become heterocysts (see color Fig. 8.2). GFP fluorescence was observed only during

Figure 8.2. Heterocyst-specific gene expression in the cyanobacterium Anabaena. Fluorescence from nitrogen-starved Anabena cells bearing a plasmid with a hetR-gfp fusion. Green fluorescence is observed preferentially in heterocyst where hetR is exclusively expressed. Courtesy of W. Buikema, University of Chicago.

168

THE USES OF GREEN FLUORESCENT PROTEIN IN PROKARYOTES

Figure 8.3. Localization of FtsZ-GFP during septum formation in live E. coli.

early stages of heterocyst formation. This downregulation of fluorescence is likely not the result of a change in hetR transcription, but rather the result of the anaerobic environment present within the nitrogen-fixing heterocyst, and the resulting failure to form a functional GFP chromophore. Indeed, in a strain deficient for the deposition of glycolipids important for maintaining the anaerobic environment in heterocysts (Hg1K) (Black et al., 1995), the hetR::gfp fusion was fluorescent throughout heterocyst development.

8.2.3

Cell Division in E. coli

During bacterial division, a complex network of proteins localize to the midpoint between daughter chromosomes in the diving cell. Ma et al. (1996) used GFP fusions to localize two of these proteins, FtsZ and FtsA, in living bacteria. To visualize the formation of the division ring in actively dividing cells, these authors utilized a fast chromophore assembly GFP mutant (S65A/V68L/S782A) (Cormack et al., 1996). Both FtsA-GFP an FtsZGFP fusions localize to the division site and support the concept of a multiprotein septator complex. A polymeric form of FtsZ-GFP could be seen as a ring structure in actively dividing cells (Fig. 8.3) and as localized aggregates and spiral tubules when the fusion protein was overexpressed. Deletion analysis of the FtsZ-GFP fusions defined the NH2 terminus of FtsZ as necessary for polymerization. The behavior of the FtsZ-GFP fusion was consistent with the noted similarities between FtsZ and tubulin (Erickson, 1995). For example, FtzZ-GFP polymerizes to form long tubules in E. coli cells, the C-terminus domain of FtsZ is not essential for this polymerization, and furthermore, C-terminal truncations produce tubules that are unusually large and stable, suggesting a lack of dynamic instability as seen in C-terminal truncations of tubulin. FtsA-GFP appeared to localize to the membrane and to ring structures in the division plane at both early and late stages of septum formation. FtsA-GFP also colocalizes with FtsZ into tubular structures when FtsZ is over-produced, suggesting that FtsA directly interacts with FtsZ. The use of these fusions will allow for the simple identification of mutants that alter the pattern of FtsZ and FtsA localization. More importantly, prokaryotic cell division components can now be followed dynamically during cell division.

8.3

BACTERIA IN COMPLEX ENVIRONMENTS

Green fluorescent protein expression is uniquely suited to track live bacteria within complex environments and can be conveniently assayed within fractionated samples. Bacterial localization, association, and multiplication, as monitored by fluorescence, can be

BACTERIA IN COMPLEX ENVIRONMENTS

followed temporally and spatially. These features have tremendous implications for the study of bacterial behavior in natural habitats such as soil and biofilms and in the study of host colonization by bacterial pathogens and symbionts. Thus, GFP should facilitate not only the localization of individual bacteria in complex microbial environments, but also the analysis of gene regulation in those environments.

8.3.1

Bacterial–Host Interactions

8.3.1.1 Bacteria in an Animal Host. Green fluorescent protein can be used to study the interactions between pathogenic bacteria and their mammalian hosts (Dhandayuthapani et al., 1995; Kremer et al., 1995; Valdivia et al., 1996). GFP has been expressed in S. typhimurium, Y. pseudotuberculosis, and Mycobacteria sp. with no adverse effect on the ability of these pathogenic organisms to interact with their hosts or cause disease (Valdivia et al., 1996). Bacteria–host interactions can be followed microscopically by epifluorescence, and can be quantitated with a spectrofluorimeter or a flow cytometer. This quantitation is very sensitive; for Y. pseudotuberculosis, the bacterial load present within single infected mammalian cells can be determined by flow cytometric analysis (Valdivia et al., 1996). GFP simplifies the analysis of the cell biology of bacterial infections since endogenous labeling of the microorganism guarantees a constant level of fluorescence signal and obviates the need to raise antibodies against the pathogen for immunolabeling. Furthermore, since GFP is intracellular, there is a reduced potential of interfering with surface contact between the bacteria and host cells. The GFP-labeled organisms have been used to visualize host cell actin rearrangements incuded by S. typhimurium invasion (Kain et al., 1995), visualize the interaction of Y. pseudotuberculosis with cytoskeletal components in murine macrophages (Hromockyj, Amieva, and Falkow, unpublished results), and purify M. bovis BCG containing vesicles after homogenization of infected macrophages (Dhandayuthapani et al., 1995). GFP-tagged bacteria can also be visualized in the tissues of experimentally infected animals. For example, M. marinum expressing gfp has been imaged in cryosections of chronically infected frog spleens up to 5 weeks postinfection (Valdivia et al., 1996) and from lungs sections of M. bovis BCG-infected mice (Kremer et al., 1995). In addition, Hinnebusch et al. (1996) used GFP tagged plague bacillus, Y. pestis, to visualize the colonization and blockage of the flea midgut. The GFP-tagged Y. pestis bearing a deletion in the hemin storage locus (hms) were unable to block the foregut of a colonized flea. This blockage is important in the transmission of plague since it profoundly affects the feeding behavior of the insect. A flea colonized with Hms+ Y. pestis is unable to feed and thus aggressively attempts to take a blood meal. Eventually, a successful feeding occurs in which the bacterial mass that blocked the foregut is regurgitated into the host’s bloodstream. In addition to histology, flow cytometry can also identify specific classes of infected cells within an organ. For example, we have sorted infected cells from the spleens of mice infected with GFP labeled S. typhimurium (R. H. Valdivia, D., Monack, and S. Falkow, unpublished results). This technique is particularly helpful in the study of bacterial pathogenesis, because it can potentially identify subsets of cells specifically targeted by intracellular pathogens during acute and chronic infections. Furthermore, gfp gene fusions allow one to measure bacterial gene expression in infected animal tissues. 8.3.1.2 Bacteria in a Plant Host. GFP labeled bacteria can also be used to analyze bacterial–plant interactions. The clearest illustration of this is the interaction

169

170

THE USES OF GREEN FLUORESCENT PROTEIN IN PROKARYOTES

Figure 8.4. Laser scanning confocal images of R. meloliti infection threads in plant root hairs. The R. meloliti bearing a plasmid with a trp-gfp fusion was used to infect alfalfa plants. Infection threads can be seen within individual root hair as they extend toward the main root body (stained red with propidium iodide). See color insert.

between Rhizobium sp. and their plant host. The symbiotic relationship between a plant host and Rhizobium begins with the infection of a root hair and lead to the formation of a plant nodule that is colonized by nitrogen-fixing Rhizobium [reviewed in Fisher and Long (1992)]. Gage et al. (1996) recently examined R. meloliti growth and behavior during the early stages of nodule formation using bacteria expressing gfp (S65T) (see color Fig. 8.4). From time-lapse observations of infected root hairs, the investigators determined that bacterial growth occurs only from bacteria in the tip region of the infection thread. Interestingly, bacterial fluorescence was found at all stages of colonization, including the nodules. This result is surprising since the nodule environment is thought to be oxygenfree in order to permit efficient nitrogen fixation. This finding suggests that the S65T GFP mutants might be less dependent than wild-type GFP on free-oxygen for chromophore assembly. Tombolini et al. (1997) also examined the adherence of GFP tagged P. fluorescens to the roots of Japanese lotus plants. In this particular work, P. fluorescens was chromosomally tagged with gfp (I167T) fused to a strong constitutive promoter that was present in a Tn5 delivery vector.

8.3.2

Bacterial Ecology and Behavior

GFP labeled bacteria have been used in the fields of bioremediation (the use of biological agents to remove toxic contaminants from soil and water, for example) and bacterial

GREEN FLUORESCENT PROTEIN AS A GENETIC TOOL

ecology. One obvious application is the tracking of bacteria in soil, biofilms, and complex microbial communities. Another is the monitoring of gene transfer in bacterial populations. Christensen et al. (1996) monitored conjugation-mediated DNA transfer in situ by following the expression of gfp in recipient cells of P. putida. Donor cells had a (toluene degradation) TOL plasmid containing gfp under the control of the Pø10 promoter of bacteriophage T7. Recipient cells constitutively expressed the T7 RNA polymerase. Transfer of the TOL plasmid to recipient cells was monitored by following the T7-driven synthesis of GFP. In addition, donor cells expressed luciferase, facilitating identification of donor and recipient cells. By monitoring the DNA flux within a mixed community over time, the authors provided new insights into the dynamics of horizontal gene transfer. Specifically, they found that conjugative transfer of DNA occurs very rapidly upon initial bacterial contact and is limited under poor growth conditions. While these conclusions took into consideration the lag in posttranslational chromophore oxidation, the analysis might be further refined by the use of fast-chromophore assembly GFP mutants, such as those described previously (Heim et al., 1994; Cormack et al., 1996). GFP also makes a suitable marker to follow genetically engineered microorganisms as they move through porous materials or in aquatic environments (Leff and Leff, 1996). Burlage et al. (1996) showed that either E. coli expressing gfp from a plasmid or P. putida expressing gfp from a Tn5gfp chromosomal insertion can be tracked by fluorimetry as they elute from sand columns. Tombolini et al. (1997) showed that GFP tagged P. fluorescens can be visualized in soil samples. Bacterial fluorescence was easily detected, even after prolonged carbon starvation conditions, suggesting that GFP tags will be useful in monitoring bacteria growing under energy limiting conditions (i.e., in soil and water samples).

8.4 GREEN FLUORESCENT PROTEIN AS A GENETIC TOOL GFP synthesis is easily assayed as green fluorescence. This fluorescence can be visualized directly on culture plates upon illumination with either blue- or long-wave ultraviolet (UV) light (Chalfie et al., 1994). For some applications, a qualitative comparison of fluorescence intensity between two bacterial colonies bearing different gfp fusion is sufficient to determine gross differences in levels of gene expression. However, spectrofluorimetry provides a more accurate, quantitative measurement of GFP fluorescence. Spectrofluorimetric measurements from gfp-expressing bacteria are simple, and, because it does not require cell lysis or the addition of exogenous substrates, can be monitored in the same sample over time. Kremer et al. (1995) used the spectrofluorimetric measurement of GFP fluorescence to assess bacterial sensitivity to antibiotics. While GFP is not a vital marker, the levels of GFP synthesis roughly correlate with the levels of overall protein synthesis. Therefore, the slow decline in bacterial fluorescence during drug exposure is an indirect measure of the antibiotic’s adverse effects on bacterial metabolism. Fluorimetry has also been used to compare the relative strength of different mycobacterial promoters in M. smegmatis (Dhandayuthapani et al., 1995). Dhandayuthapani et al. examined expression of the ahpC gene, which encodes alkyl hydroperoxide reductase whose levels have been linked to resistance to the antimycobacterial drug isoniazid. The investigators compared the levels of fluorescence expressed by ahpC-gfp fusions to demonstrate that the ahpC promoter region from Mycobacterium tuberculosis had substantially lower transcriptional activity than its Mycobacterium leprae counterpart, potentially explaining why M. tuberculosis, unlike other mycobacteria, is exquisitely sensitive to isoniazid.

171

172

THE USES OF GREEN FLUORESCENT PROTEIN IN PROKARYOTES

8.4.1

Flow Cytometry and Bacterial Genetics

The use of fluorescence as a reporter of gene expression permits the use of fluorescencebased technologies that not only quantitify fluorescence but also physically separate microorganisms on the basis of their relative fluorescence intensities (reviewed in Parks et al., 1989; Shapiro, 1995). A fluorescence-activated flow cytometer reads the fluorescence intensity of every particle that passes through the laser sensing area. This useful feature allows one to examine a large population of cells and determine the levels of gene expression for each bacterium in the sample. Escherichia coli, S. typhimurium, Y. pseudotuberculosis, P. fluorescens, and Mycobacteria sp. have been successfully analyzed by flow cytometry (Dhandayuthapani et al., 1995; Kremer et al., 1995; Cormack et al., 1996; Tombolini et al., 1996; Valdivia et al., 1996). Unlike spectrofluorimetry, which gives an average fluorescence for a sample, flow cytometry provides a multiparameter record of all cells sampled (Parks et al., 1989). This information can be displayed as histograms or contour plots that provide the frequency of bacteria with particular light scatter and fluorescence characteristics. For example, Dhandayuthapani et al. (1995) have used flow cytometry to characterize four different mycobacterial promoters fused to gfp. The per bacterium fluorescence intensity mimicked that found by spectrofluorimetry. Furthermore, Dhandayuthapani et al. (1995) could separate a mixed population of mycobacteria bearing a transcriptionally weak (mtrA::gfp) and strong (hsp60::gfp) gene fusions using a fluorescence activated cell sorter (FACS). In another example, demonstrating the power of this approach, Cormack et al. (1996) used FACS to isolate highly fluorescent GFP mutants from a library of over 4 ¥ 106 bacteria carrying different GFP chromophore mutations. These genetic approaches to isolating productive gfp fusions have also been used to identify genes induced under complex or poorly defined conditions. In conventional bacterial genetics inducible genes are often isolated by screening the expression of fusions with a measurable reporter gene. For example, bacteria bearing random lacZ gene fusions can be scored on nutrient agar plates, in the presence of an inducing stimulus, for the synthesis of b-galactosidase. Positive fusions are then assayed for the synthesis of bgalactosidase in the absence of the inducer and thus inducible (and repressible) gene fusions can be identified. However, if the induction conditions are hard to replicate in solid media or are detrimental to bacterial growth, such genes are often difficult to isolate. Some of these problems can be overcome by using FACS to isolate bacteria bearing gfp gene fusions. The transient expression of gfp is a phenotype that is easily scored by the cell sorter. The physical separation of individual bacteria on the basis of fluorescence is analogous to the manual screening of colonies on agar plates, but the processivity of the FACS machine (2–3 thousand bacteria per second; Parks et al., 1989), makes this screening process similar in power to genetic selection. Furthermore, FACS can discriminate among different fluorescence intensities. Theoretically, individual bacteria bearing inducible gene fusions with any absolute fluorescence level can be specifically isolated (Parks et al., 1989). We will use two examples to illustrate how these flow cytometric based gene selection and enrichment strategies, termed differential fluorescence induction (DFI) (Valdivia and Falkow, 1996), have been successfully applied to the isolation of inducible genes. 8.4.1.1 Isolation of Acid-inducible Genes in S. typhimurium. The response of bacteria to acidic conditions has been difficult to study because bacterial growth is hindered at low pH. We developed a DFI enrichment cycle to identify bacterial genes that

GREEN FLUORESCENT PROTEIN AS A GENETIC TOOL

are induced by transient exposure to highly acidic conditions (pH 4.5) (Valdivia and Falkow, 1996). A library of S. typhimurium bearing random DNA fragments fused to gfp (S65G/S72A) (Cormack et al., 1996) in a plasmid vector was exposed to media at pH 4.5 for 2 h and all fluorescent bacteria in the pool were sorted. The sorted sample was expanded and exposed to media at neutral pH. Since acid-inducible fusions will not be expressed at neutral pH, bacteria bearing nonfluorescent fusions at pH 7 were collected. This nonfluorescent bacterial population was exposed to acidic pH again, and all fluorescent organisms were sorted. The final collected pool was highly enriched (~30–50%) for bacteria bearing gene fusions whose activity was upregulated under acidic conditions. The DNA sequence analysis of eight of these acid-inducible gene fusions revealed high homology to promoter regions from genes involved in stress response and multidrug resistance; and genes with previously reported pH regulated activity (Valdivia and Falkow, 1996). FACS can also be used to identify loci that regulate the activity of a gene of interest. For example, we have used an acid-inducible aas::gfp fusion (Valdivia and Falkow, 1996) to isolate miniTn5 insertions that abolish this fusion’s induction at acidic pH (R. H. Valdivia, M. Rathman, and S. Falkow, unpublished data). After a generalized insertional mutagenesis, over 107 S. typhimurium bearing an aas::gfp fusion were exposed to pH 4.5 and sampled by FACS. Nonfluorescent organisms were present at a frequency of 0.01%. These bacteria were collected and the miniTn5 insertions present within this sorted population were transduced into a nonmutagenized S. typhimurium background. The inability to induce aas::gfp under acidic conditions cotransduced with the insertion elements. Several insertions mapped to the ompR/envZ locus. OmR/EnvZ is a two component system known to regulate the expression of several genes in response to changes in osmolarity (reviewed in Pratt and Silhavy, 1995). The finding that this locus is also involved in the acid response is not surprising since low pH regulates the expression of at least one porin gene, under the control of OmpR (Foster et al., 1994). 8.4.1.2 Isolation of Macrophage-Inducible Genes in S. typhimurium. S. typhimurium is a facultative intracellular pathogen that modifies the macrophage’s phagocytic vacuole to permit its own survival (reviewed in Garcia del Portillo and Finlay, 1995) Several bacterial genes are preferentially expressed in the intracellular environment (Alpuche-Aranda et al., 1992; Garcia del Portillo et al., 1992; Fierer et al., 1993), but the mechanisms that allow S. typhimurium to survive in this adverse environment are poorly understood. To explore the genetic basis of intracellular survival, we have applied DFI selections to identify genes from pathogenic bacteria that are induced within the host cell (R. H. Valdivia and S. Falkow, unpublished results). We infected macrophages with S. typhimurium bearing random gfp gene fusions (see above) and sampled them by FACS. Macrophages containing GFP-fluorescent bacteria were collected, lysed, and the released bacteria grown in the absence of host cells. Bacteria that did not fluoresce under these conditions were collected by FACS and used for a second round of macrophage infection. Bacteria recovered from fluorescent macrophages contained gfp fusions that were upregulated in the host cell’s intracellular environment. Bacteria bearing individual gene fusions were tested by fluorescence microscopy to demonstrate upregulation of gfp expression in the intracellular environment. In a nonsaturating screen, we identified 18 macrophageinducible loci, including two previously identified acid-inducible genes (Valdivia et al., 1996) (see color Fig. 8.5). Some of these genes are known to be upregulated intracellularly or to be important for in vivo survival. The large majority of the identified genes, however, have not been previously described. The role of these genes in Salmonella pathogenesis is currently under investigation.

173

174

THE USES OF GREEN FLUORESCENT PROTEIN IN PROKARYOTES

Figure 8.5. Visualization of S. typhimurium intracellular-specific gene expression by fluorescence microscopy. S. typhimurium bearing a pagA::gfp fusion shows gene induction inside an infected mammalian cell but not in the extracellular medium. The corresponding DIC images show the relative topology of bacteria with respect to the infected cell. See color insert.

GFP is particularly well suited for the in vivo imaging of bacterial gene expression within host cells. Via et al. (1996) showed both microscopically and by flow cytometry that an mtrA::gfp is upregulated in M. bovis BCG during its residence in macrophages. A direct quantitation of gfp expression by intracellular organisms was determined by the flow cytometric analysis of bacteria released after mechanical disruption of infected cells. The MtrA protein is homologous to response regulators of many bacterial two-component systems. The finding that the transcription of mtrA is upregulated intracellularly makes this gene a putative candidate as a mycobacterial virulence factor. We used similar flow cytometric quantitation to examine the kinetics of S. typhimurium gene induction with phagocytic cells, and showed that there are at least two classes of macrophage inducible genes: the first class induced within 1 h of cell entry, and the second induced only after 4 h (Valdivia and Falkow 1996), suggesting that the bacterial response to macrophage internalization is transcriptionally complex. This result emphasizes the advances that GFP has brought to the analysis of gene regulation, and hints at the promise it holds.

8.5

PERSPECTIVES

Microbes exist in nature in complex interactive communities. Appreciation of this fact has increased the interest in following particular bacterial cells and their gene expression in mixed populations. GFP is probably the single-best experimental tool available for this purpose, and it enters the repertoire of research tools just as microbiologists begin to shift from the study of microorganisms grown in the laboratory to their study in the “wild.” With the use of GFP and other available reporting molecules, like LacZ and Lux, we will undoubtedly see an increased focus on microbial interactions in the “real world.” In the field of host–parasite relationships, particularly among the medically important microbes, the study of the initial interaction of the microbe with the innate elements of the immune system is now seen as key to understanding the pathogenesis of infection. GFP and similar reporters will play an important role in revealing the details of the interaction between host and microbe that occurs at the earliest times after exposure to infectious agents. Such

REFERENCES

studies will undoubtedly aid in the design of novel antiinfective agents and in the development of vaccines. GFP also adds to the family of new experimental approaches (Mahan et al., 1993) that identify genes expressed only in the natural microbial habitat. Recently, a negative selection method using signature-tagged transposition (Hensel et al., 1995) has proved successful for identifying genes essential in one environment (e.g., pathogen genes essential inside the animal host), but dispensible under another set of growth conditions. Detection of GFP expression provides another powerful approach to the identification of genetic sequences expressed under unique and complex environmental conditions. Since GFP is extremely stable, it has the advantage of permitting even transient gene expression to be detected by flow cytometry. In our hands, the combined use of signature-tagged transposition and the GFP-based DFI provides complementary information about genetic sequences essential for S. typhimurium growth within the murine host. We believe that this and similar GFP-based methods that exploit the features of contemporary cell sorting technology have the potential to become highly useful tools for the identification of novel genetic sequences important for growth under conditions too complex to be duplicated easily in the laboratory. One can expect the complete nucleotide sequence of the most medically and commercially important microorganisms to be widely available in the next decade. These sequences will be extensively analyzed by computer algorithms to identify homologous sequences with known biochemical motifs. The fact remains that understanding the roles of these genetic sequences in the biology of the microbe will necessarily include understanding where, when, and how particular genes are expressed. The use of GFP can provide insight that sequences cannot into the coordinate transcription and assembly of products underlying such complex activities as cell division or the contact-dependent translocation of proteins into mammalian cells. The use of GFP and its derivatives will permit the exploration of many functional facets of the biology of microbes.

ACKNOWLEDGMENTS We wish to thank W. Buikema, W. Margolin, D. Gage, F. Bruijn, O. Resnekov, and C. Webb for contributing unpublished data, figures, and helpful discussions.

REFERENCES Alpuche-Aranda, C. M., Swanson, J. A., Loomis, W. P., and Miller, S. I. (1992). Salmonella typimurium activates virulence gene transcription within acidified macrophage phagosomes. Proc. Natl. Acad. Sci. USA 89:10079–10083. Arigoni, F., Pogliano, K., Webb, C. D., Stragier, P., and Losick, R. (1995). Localization of protein implicated in establishment of cell type to sites of asymmetric division. Science 270:637–640. Barak, I., Behary, J., Olmedo, G., Guzman, P., Brown, D. P., Castro, E., Walker, D., Westpheling, J., and Youngman, P. (1996). Structure and function of the Bacillus SpoIIE protein and its localization in sites of sporulation septum assembly. Mol. Microbiol. 19:1047–1060. Black, K., Buikema, W. J., and Haselkorn, R. (1995). The hglK gene is required for the localization of heterocyst-specific glycolipids in the cyanobacterium Anabaena strain PCC 7120. J. Bacteriol. 172:6440–6448. Burlage, R. S., Yang, Z. K., and Mehlhorn, T. (1996). A transposon for green fluorescent protein transcriptional fusions: application for bacterial transport experiments. Gene 173:53–58.

175

176

THE USES OF GREEN FLUORESCENT PROTEIN IN PROKARYOTES

Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green Fluorescent Protein as marker of gene expression. Science 263:802–805. Christensen, B. B., Sternberg, C., and Molin, S. (1996). Bacterial plasmid conjugation on semisolid surfaces monitored with the green fluorescent protein (GFP) from Aequorea victoria as a marker. Gene 173:59–65. Cormack, B. P., Valdivia, R. H., and Falkow, S. (1996). FACS-optimized mutants of the green fluorescent protein (GFP). Gene 173:33–38. Crameri, A., Whitehorn, E. A., Tate, E., and Stemmer, P. C. (1996). Improved green fluorescent protein by molecular evolution using DNA shuffling. Nat. Biotech. 14:315–319. Delagrave, S., Hawtin, R. E., Silva, C. M., Yang, M. M., and Youvan, D. C. (1995). Red-shifted excitation mutants of the green fluorescent protein. Bio/technology 13:151–155. Dhandayuthapani, S., Via, L. E., Thomas, C. A., Horowitz, P. M., Deretic, D., and Deretic, V. (1995). Green fluorescent protein as a marker for gene expression and cell biology of mycobacterial interactions with macrophages. Mol. Microbiol. 17:901–912. Driks, A., Roels, S., Beall, B., Jr., Moran, C. P., and Losick, R. (1994). Subcellular localization of proteins involved in the assembly of the spore coat of Bacillus subtilis. Genes Dev. 8:234–244. Errington, J. (1993). Bacillus subtilis sporulation: Regulation of gene expression and control of morphogenesis. Microbiol. Rev. 57:1–33. Erickson, H. P. (1995). FtsZ, a prokaryotic homolog of tubulin? Cell 80:367–370. Fierer, J., Eckmann, L., Fang, F., Pfeifer, C., Finlay, B. B., and Guiney, D. (1993). Expression of the Salmonella virulence plasmid gene spvB in cultured macrophages and nonphagocytic cells. Infect. Immunol. 61:5231–5236. Fisher, R. F., and Long, S. R. (1992). Rhizobium-plant signal exchange. Nature (London) 356:655–660. Foster, J. W., Park, Y. K., Bang, I. S., Karem, K., Betts, H., Hall, H. K., and Shaw, E. (1994). Regulatory circuits involved with pH-regulated gene expression in Salmonella typhimurium. Microbiology 140:341–352. Gage, D. J., Bobo, T., and Long, S. R. (1996). Use of green fluorescent protein to visualize the early events of symbiosis between Rhizobium meloliti and alfalfa, Medicago sativa. J. Bacteriol 178:7159–7166. Garcia del Portillo, F., and Finlay, B. B. (1995). The varied lifestyles of intracellular pathogens within eukaryotic vacuolar compartments. Trends Microbiol. 3:373–380. Garcia del Portillo, E., Foster, J. W., Maguire, M. E., and Finlay, B. B. (1992). Characterization of the micro-environment of Salmonella typhimurium-containing vacuoles within MDCK epithelial cells. Mol. Microbiol. 6:3289–3297. Harry, E. J., Pogliano, K., and Losick, R. (1995). Use of immunofluorescence to visualize cellspecific gene expression during sporulation in Bacillus subtilis. J. Bacteriol. 177:3386–3393. Haselkorn, R. (1992). Developmentally regulated gene rearrangements in prokaryotes. Ann. Rev. Genet. 26:113–130. Heim, R., Prasher, D. C., and Tsien, R. Y. (1994). Wavelength mutations and post-translational autoxidation of green fluorescent protein. Proc. Natl. Acad. Sci. USA 91:12501–12504. Heim, R., Cubitt, A. B., and Tsien, R. Y. (1995). Improved green fluorescence. Nature (London) 373:663–664. Hensel, M., Shea, J., Gleeson, C., Jones, M., Dalton, E., and Holden, D. (1995). Simultaneous identification of bacterial virulence genes by negative selection. Science 269:400–403. Hinnebusch, B. J., Perry, R. D., and Schwan, T. G. (1996). Role of Yersinia pestis hemin storage (hms) locus in the transmission of plague by fleas. Science 273:367–370. Inouye, S., and Tsuji, F. I. (1994a). Aequorea green fluorescent protein: Expression of the gene and fluorescent characteristics of the recombinant protein. FEBS Lett. 341:277–280.

REFERENCES

Inouye, S., and Tsuji, F. I. (1994b). Evidence for redox forms of the Aequorea green fluorescent protein. FEBS Lett. 351:211–214. Kain, S. R., Adams, M., Kondepudi, A., Yang, T.-T., Ward, W. W., and Kitts, P. (1995). Green fluorescent protein as a reporter of gene expression and protein localization. BioTech. 19:640–655. Kremer, L., Baulard, A., Estaquier, J., Poulain-Godefroy, O., and Locht, C. (1995). Green fluorescent protein as a new expression marker in mycobacteria. Mol. Microbiol. 17:913–922. Leff, L. G., and Leff, A. A. (1996). The use of green fluorescent protein to monitor survival of genetically engineered bacteria in aquatic environments. Appl. Environ. Microbiol. 62:3486– 3488. Lewis, P. J., and Errington, J. (1996). Use of green fluorescent protein for detection of cell-specific gene expression and subcellular protein localization during sporulation in Bacillus subtilis. Microbiology 142:733–740. Lewis, P. J., Nwoguh, C. E., Barer, M. R., Harwood, C. J., and Errington, J. (1994). Use of digitized video microscopy with a fluorogenic enzyme substrate to demonstrate cell- and compartment-specific gene expression in Salmonella enteriditis and Bacillus subtilis. Mol. Microbiol. 13:655–662. Losick, R., and Stragier, P. (1992). Crisscross regulation of cell-type specific gene expression during development in B. subtilis. Nature (London) 355:601–604. Ma, X., Ehrhardt, D. W., and Margolin, W. (1996). Co-localization of cell division proteins FtsZ and FtsA to cytoskeletal structures in living Escherichia coli. cells using green fluorescent protein. Proc. Natl. Acad. Sci. USA 93:12998–13003. Mahan, M. J., Slauch, J. M., and Mekalanos, J. J. (1993). Selection of bacterial virulence genes that are specifically induced in host tissues. Science 259:686–688. Ogawa, H., Inouye, S., Tsuji, F. I., Yasuda, K., and Umesono, K. (1995). Localization, trafficking and temperature-dependence of the Aequorea green fluorescent protein in cultured vertebrate cells. Proc. Natl. Acad. Sci. USA 92:11899–11903. Parks, D. R., Herzenberg, L. A., and Herzenberg, L. A. (1989). Flow cytometry and fluorescenceactivated cell sorting. In Fundamental Immunology, Paul, W. E., Ed., Raven, New York, pp. 781–802. Prasher, D. C., Eckenrode, V. K., Ward, W. W., Prendergast, F. G., and Cormier, M. J. (1992). Primary structure of the Aequorea victoria green fluorescent protein. Gene 111:229–233. Pratt, L., and Silhavy, T. J. (1995). Porin regulon of Escherichia coli. In Two-Component Signal Transduction. Hoch, J. A., and Silhavy, T. J., Eds., The American Society of Microbiology, Washington, DC, pp. 105–127. Pogliano, K., Harry, E. J., and Losick, R. (1995). Visualizing the subcellular localization of sporulation proteins in Bacillus subtilis using immunofluorescence microscopy. Mol. Microbiol. 18:459–470. Resnekov, O., Alpera, S., and Losick, R. (1996). Subcellular localization of proteins governing the proteolytic activation of a developmental transcription factor in Bacillus subtilis. Genes Cells 1:529–542. Shapiro, H. M. (1995). Practical Flow Cytometry, 3rd ed. Wiley-Liss, New York. Sharpe, M. E., and Errington, J. (1996). The Bacillus subtilis soj-spoOJ locus is required for a centromere-like function involved in prespore chromosome partitioning. Mol. Microbiol. 21:501–509. Stragier, P., and Losick, R. (1996) Molecular genetics of sporulation in Bacillus subtilis. Ann. Rev. Genet. 30:297–241. Tombolini, R., Unge, A. Davey, M. E., de Bruijn, F. J., and Jansson, J. K. (1997). Flow cytometric and microscopic analysis of GFP-tagged Psedumonas fluorescens bacteria. FEMS Microbiol Ecol. 22:17–28.

177

178

THE USES OF GREEN FLUORESCENT PROTEIN IN PROKARYOTES

Valdivia, R. H., and Falkow, S. (1996). Bacterial genetics by flow cytometry: Rapid isolation of Salmonella typhimurium acid-inducible promoters by differential fluorescence induction. Mol. Microbiol. 22:367–378. Valdivia, R. H., Hromockyj, A. E., Monack, D., Ramakrishnan, L., and Falkow, S. (1996). Applications for the green fluorescent protein (GFP) in the study of host–pathogen interactions. Gene 173:47–52. Via, L. E., Curcic, R., Mudd, M. H., Dhandayuthapani, S., Ulmer, R. J., and Deretic, V. (1996). Elements of signal transduction in Mycobacterium tuberculosis. In vitro phosphorylation and in vivo expression of the response regulator MtrA. J. Bacteriol. 178:3314–3321. Webb, C. D., Decatur, A., Teleman, A., and Losick, R. (1995). Use of green fluorescent protein for visualization of cell-specific gene expression and subcellular protein localization during sporulation in Bacillus subtilis. J. Bacteriol. 177:5906–5911. Wu, L. J., and Errington, J. (1994). Bacillus subtilis SpoIIE protein required for DNA segregation during asymmetric cell division. Science 264:572–575. Wu, L. J., Lewis, P. J., Allmansberger, R., Hauser, P. M., and Errington, J. (1995). A conjugationlike mechanism for prespore chromosome partitioning during sporulation in Bacillus subtilis. Genes Dev. 9:1316–1326.1.

9 THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS Amy L. Hitchcock Department of Molecular and Cellular Biology, Harvard University, Cambridge, MA

Jason A. Kahana Department of Alzheimer’s Research, Merck Research Laboratories, West Point, PA

Pamela A. Silver Department of Systems Biology, Harvard Medical School, Boston, MA; Department of Cancer Biology, Dana-Farber Cancer Institute, Boston, MA

9.1 INTRODUCTION The budding yeast Saccharomyces cerevisiae and the fission yeast Schizosaccharomyces pombe serve as excellent model systems for studying a variety of cell-biological processes. Basic processes such as cell cycle progression, intracellular macromolecular transport, gene expression, and metabolism have been conserved throughout eukaryotic evolution, making these yeasts a convenient means to study these phenomena. Both budding and fission yeast are easy to culture and grow extremely quickly. Yeast can be manipulated genetically, and the genomes of both S. cerevisiae and S. pombe have been fully sequenced. Finally, the availability of a variety of molecular genetic tools, including auxotrophic and dominant selectable markers, promoters, extrachromosomal plasmids, genomic deletion and integration techniques, and conditional-lethal alleles greatly facilitates studies of gene expression, localization, and function. These powerful tools have made S. cerevisiae in particular a paradigm of genomics and proteomics research, including DNA chip technology for transcriptional profiling and genome-wide localization, as well as large-scale protein localization, protein–protein interaction, and biochemical genomic analyses Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

179

180

THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS

[reviewed in Bader et al. (2003)]. Such genomic- and proteomic-based studies in turn have rapidly accelerated the pace of hypothesis-driven cell biological discoveries in this model organism, as well as in higher eukaryotes. The pathogenic yeast Candida albicans has also become the focus of intense research in recent years due to its ability to cause superficial as well as systemic infection in individuals with impaired immunity. Furthermore, C. albicans displays interesting biology distinct from S. cerevisiae and S. pombe, including its ability to assume diverse cell morphologies such as hyphae. With the recent completion of a draft of the C. albicans genome, along with the development of molecular genetic techniques, researchers are equipped to investigate the molecular mechanisms of virulence and to develop antifungal therapies for this important human pathogen. However, yeasts have several shortcomings as model organisms. In particular, the analysis of subcellular structures using light microscopic techniques presents an experimental challenge due to the small size of the cells, between 5 and 10 mm in diameter. The thick yeast cell wall prevents the microinjection of fluorescent markers. Furthermore, in contrast to the relatively flat mammalian cells cultured on plastic, the significant thickness (~60% of diameter) of yeast cells prevents the visualization of larger organelles (such as the nucleus) by standard light microscopy. Nonetheless, microscopy techniques mostly based on fluorescence have become standard practice in yeast laboratories. A number of fluorescent vital dyes that label specific cellular compartments, nuclear and mitochondrial DNA, and the cell wall are available for use in yeast [a compilation is provided in Kohlwein (2000)]. Protein localization studies, which represent an important application in yeast cell microscopy, have also traditionally required cell fixation/permeabilization prior to detection by fluorescently labeled probes such as antibodies or small molecule ligands. In each of these cases, the extensive preparation and manipulation of the cells prior to visualization increases the potential for artifacts; as a result, the researcher must perform careful controls to properly interpret the observed fluorescence of a protein or cellular compartment of interest. Thus, a convenient and noninvasive method for studying cytology would make yeast a more “complete” system. The green fluorescent protein (GFP), an intrinsically fluorescent protein cloned from the Aequorea victoria jellyfish, has revolutionized microscopic studies of yeast cytology. Notably, this genetically encoded fluorophore can serve as a molecular tag for proteins, organelles, and other structures in living cells, without the need for fixation, antibodies, or exogenously provided cofactors (Tsien, 1998). Furthermore, the development of spectral variants of GFP, including the cyan and yellow fluorescent proteins (CFP and YFP respectively), and the discovery of a red fluorescent protein (RFP, also called DsRed) from the coral Discosoma, have paved the way for in vivo colocalization and fluorescence resonance energy transfer (FRET) studies in yeast (Campbell et al., 2002; Matz et al., 1999; Tsien, 1998). In this chapter, we will address some technical aspects of expressing GFP fusion genes in yeast and will review the tremendous possibilities for the dynamic, in vivo applications of GFP (and other related fluorescent proteins) in yeast extending far beyond studies of protein localization.

9.2

EXPRESSION OF GFP IN YEASTS

The budding and fission yeasts are particularly amenable to studies utilizing GFP. A wide variety of gene expression systems enable the researcher to express GFP and GFP fusion genes at levels that best fit the application at hand. Most pertinent for in vivo protein local-

EXPRESSION OF GFP IN YEASTS

ization studies, GFP gene fusion expression can be easily modulated to approximate the physiological expression levels of the unmodified gene of interest. Several options are available for experiments requiring higher levels of expression or inducible/repressible expression, including well-characterized regulated yeast promoters and high-copy plasmids. Finally, the simple genetics of S. cerevisiae and S. pombe greatly facilitate assessment of the functionality of GFP fusion proteins. GFP technology has also been successfully applied to cell biological studies of the medically relevant yeast C. albicans, despite significant molecular genetic challenges. In this section we will describe several yeast-based systems for expressing recombinant gene fusions. While we will focus largely on the expression of GFP gene fusions, it is important to note that these systems are also relevant for the expression of gene fusions to the GFP variants CFP and YFP, as well as DsRed. Some considerations regarding the expression of these spectrally diverse fluorescent proteins in yeast will be discussed in Section 9.2.5.

9.2.1

S. cerevisiae Gene Expression Systems

Powerful molecular genetic resources and techniques that have been developed for budding yeast allow tremendous control and ease in recombinant gene expression. In most cases, researchers need to express a GFP gene fusion at levels that are similar to those of the chromosomally encoded, untagged gene of interest. To this end, GFP fusion genes can be expressed extrachromosomally on autonomously replicating, centromeric (ARS/CEN) plasmids that mimic the behavior of single chromosomes in yeast (often called “singlecopy” plasmids). Because plasmid loss can occur at a measurable rate, transformed yeast should be maintained in selective media. With few exceptions, an S. cerevisiae gene, along with a sufficient promoter (5¢) and terminator (3¢) sequence—usually ~300–600 base pairs (bp) upstream and downstream, respectively—can be expressed from an ARS/CEN plasmid at levels comparable to those from the endogenous chromosomal locus. Our laboratory has had tremendous success with a second approach for expressing GFP fusion genes at physiological levels: integration of in-frame GFP sequence into the chromosomal open reading frame (ORF) of the gene of interest. Several PCR-based approaches that target GFP integrations in budding yeast have been recently described (Knop et al., 1999; Longtine et al., 1998; Prein et al., 2000). The Yeast Resource Center (YRC) at the University of Washington (http://depts.washington.edu/~yeastrc/) generously provides compatible template plasmids that additionally allow for the targeted integration of CFP, YFP, and DsRed by these approaches. By expressing the fusion gene from the endogenous chromosomal locus, it is subject to any cis-acting regulatory controls (such as chromatin structure) that might normally modulate the gene’s expression. A second advantage of integrated gene fusions is that they are stably inherited by daughter cells (even in the absence of selection), resulting in very consistent expression from cell to cell in a clonally derived population. Third, this approach eliminates the potential problem of simultaneously expressing a tagged and untagged form of a protein in a yeast cell; in a haploid yeast strain, the only version of the protein expressed is the GFP fusion. Finally, assuming an easily monitored phenotype for loss of function, this benefit also can allow for the rapid assessment of fusion protein functionality (see Section 9.2.4). Higher level or regulated expression of recombinant genes is also desirable in some experimental settings, and is easily accomplished in S. cerevisiae. First, several strong constitutive and inducible promoters have been characterized for their ability to drive expression of genes in budding yeast. A commonly used strong promoter is derived from the ADH1 promoter (pADH1), which normally drives constitutive expression of the alcohol

181

182

THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS

dehydrogenase gene in glucose-containing medium (Ruohonen et al., 1995). For mediumstrength constitutive gene expression, our laboratory and others have had success with the NOP1 promoter [e.g., Bailer et al. (1998)]. The most commonly utilized inducible promoter, pGAL1-10, is activated in the presence of galactose and strongly repressed in glucose (Yocum and Johnston, 1984). The MET3 promoter (pMET3) is active in the absence of methionine, whereas addition of this amino acid results in the strong repression of pMET3-driven expression (Cherest et al., 1987). Finally, pCUP1 drives moderatelevel expression in normal media and is strongly induced in the presence of soluble Cu2+ ion (Karin et al., 1984). In each of these examples, a given gene fusion can be expressed at relatively high levels (up to ~0.5–1% total cellular protein) under conditions that promote otherwise normal cell metabolism and growth. Importantly, none of these small molecule inducers/repressors cause fluorescence that would interfere with GFP visualization. Thus, for a gene whose level of expression under its endogenous promoters is too low to permit visualization using GFP fluorescence, a strong constitutive or inducible promoter may be a good solution. Furthermore, these promoters are often the most convenient system for expressing GFP fusions of non-yeast genes (or any ORF without a yeast promoter). In addition to the use of strong constitutive or inducible promoters, 2m circle-derived (“high-copy”) plasmids offer a second option for the overexpression of genes under the control of their own promoter in S. cerevisiae. Typically, high-copy plasmids contain 2m circle sequences that serve as an origin of DNA replication and direct plasmid amplification to ~20-40 copies/cell. Therefore, the “dosage” of a 2m plasmid-encoded gene fusion can be increased while still allowing for regulation by the gene’s endogenous 5¢ and 3¢ sequences. In this way, the researcher may be able to avoid possible toxic effects of gene overexpression, for example, during inappropriate periods of the cell cycle or at the levels of induced promoters. As for ARS/CEN plasmids, maintenance of selective conditions is recommended to prevent 2m plasmid loss. A number of specific expression systems for S. cerevisiae are described in more detail in the Methods section of this chapter (Section 9.4.1).

9.2.2

S. pombe Gene Expression Systems

In contrast to S. cerevisiae, single-copy plasmids are not available for gene expression in S. pombe because fission yeast centromeres are too large to be encompassed on a reasonably sized plasmid. Plasmids containing the fission yeast ars1+ sequence (an origin of replication) transform S. pombe with high efficiency and replicate to high copy number (~15-80) per cell. However, these plasmids are mitotically unstable and are prone to producing polymers with various numbers of repeat units. Inclusion of an stb (stability) element renders ars1+-containing plasmids more stable such that they are roughly 10% as stable as S. cerevisiae ARS/CEN plasmids. Due to their high copy number, it is uncertain but possible that expression of a GFP gene fusion (as regulated by its endogenous 5¢ and 3¢ sequences) from an ars1+ stb plasmid will produce expression levels similar to that of a chromosomally encoded gene. The reader is referred to a study by Craven et al. (1998) that describes a series of plasmids that allow expression of recombinant gene fusions, including to GFP, in S. pombe. A good solution to the caveats of plasmid-based gene expression in fission yeast is to integrate GFP coding sequence into the chromosomally encoded gene of interest by homologous recombination. Expression of such integrated fusion genes provides the same benefits described above for S. cerevisiae (see Section 9.2.1), and a PCR-based approach for

EXPRESSION OF GFP IN YEASTS

targeted integration of the GFP sequence in S. pombe has been described in a study by Bahler et al. (1998). The Yeast Resource Center (YRC) at the University of Washington (http://depts.washington.edu/~yeastrc/) generously provides compatible template plasmids that additionally allow for the targeted integration of YFP, CFP, and DsRed in fission yeast by this approach. A number of promoters have been characterized in S. pombe for their ability to direct high level or inducible gene expression, including the adh1+ promoter (constitutive high expression), fbp1+ (carbon source responsive), and a tetracycline-repressible version of the cauliflower mosaic virus (CaMV) 35S promoter (Faryar and Gatz, 1992; Hoffman and Winston, 1989). The nmt1+ promoter, which is the most commonly used promoter in S. pombe gene expression studies, drives high-level gene expression and can be repressed (although not necessarily fully) in the presence of thiamine (Maundrell, 1990). Additionally, activity of the nmt1+ promoter can be modulated to drive intermediate levels of expression by varying the levels of thiamine in the growth media and/or through the use of attenuated promoter mutants (Basi et al., 1993; Maundrell, 1993). Importantly, thiamine does not autofluoresce, and therefore does not interfere with the visualization of GFP in fission yeast.

9.2.3

C. albicans Gene Expression Systems

The development of molecular genetic techniques in C. albicans has lagged behind that for its distant relatives S. cerevisiae and S. pombe. First, C. albicans is a presumed obligate diploid and does not have a complete sexual cycle, thus preventing basic genetic analysis of and approaches to its biological processes. Second, C. albicans displays alternate codon usage, translating the CUG codon as a serine rather than leucine. For this reason, many heterologous markers and protein tags do not function in C. albicans unless the CUG codons are first modified. Nonetheless, many molecular genetic techniques are now available in this pathogenic yeast, including those required for the expression of GFP and GFP fusion genes [reviewed in Berman and Sudbery (2002) and in De Backer et al. (2000)]. Importantly, codon-optimized versions of GFP have been generated that successfully direct expression of this fluorescent protein in C. albicans. The alteration of the single CUG codon (L201) in wild-type GFP to the preferred leucine codon UUG is necessary for GFP fluorescence in this pathogenic yeast (Cormack et al., 1997; Morschhauser et al., 1998). Additional optimization of GFP based on C. albicans codon usage and characterized chromophore mutations appears to further enhance expression and function of GFP in C. albicans (Cormack et al., 1997). Based on these studies, the yEGFP version of GFP developed by Cormack et al. (1997), which is fully codon-optimized for expression in this pathogenic yeast, may be the best option for in vivo fluorescence studies. Additionally, the CFP and YFP spectral variants of yEGFP have recently been generated, paving the way for colocalization and FRET studies in C. albicans (Gerami-Nejad et al., 2001). GFP fusion genes can be expressed in C. albicans from extrachromosomal ARScontaining plasmids; it should be noted, however, that these plasmids are extraordinarily unstable in C. albicans, leading to wide variation in copy number from cell to cell, genomic integration, and plasmid loss in the absence of selection. Therefore, such plasmids are unlikely to express encoded gene fusions at consistent levels cell to cell; also, the ability to approximate physiological levels of expression, even using the gene’s endogenous promoter, is questionable. Nonetheless, the reader is referred to a recent study introducing a panel of ARS-containing vectors designed for gene expression, including as GFP fusions, in C. albicans (Park and Choi, 2002). Conveniently, these plasmids are of reasonable size

183

184

THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS

(5–7.5 kb), thus allowing for the cloning of large inserts, contain selectable markers for both C. albicans and E. coli, and, finally, in some cases contain the inducible MET3 promoter (see below). An attractive alternative to plasmid-based gene expression is to integrate GFP fusion genes into the C. albicans genome by targeted homologous recombination. Such stable integrants, as for S. cerevisiae and S. pombe, typically display more uniform fusion protein expression that more closely mimics the endogenous expression levels of the untagged gene. A PCR-based method for the directed integration of GFP, CFP, and YFP at the 3¢ end of chromosomal ORFs in this pathogenic yeast has recently been described (GeramiNejad et al., 2001). Finally, high-level or inducible gene expression of gene fusions can be controlled by one of several C. albicans promoters that direct high level (pADH1, pACT1) or inducible/repressible (pMAL2, pMET3, pPCK1) gene expression (Berman and Sudbery, 2002; De Backer et al., 2000).

9.2.4

Function and Proper Localization

With a molecular weight of 29 kilodaltons (kDa), GFP represents a significant structural addition to most tagged proteins. In the absence of any other means to confirm protein localization (e.g., biochemical fractionation with the untagged or alternatively tagged protein), one should determine whether the GFP fusion protein localizes properly. The most direct means of assaying proper localization is to determine if the GFP fusion protein is functional, using the most stringent tests possible. Ideally, the fusion gene will complement all known phenotypes of a null or loss-of-function mutant when expressed at physiological levels. Additionally, one should determine the growth rates of cells expressing only the GFP–tagged form of the protein, as well as confirm any known biochemical activities or protein–protein interactions. Yeast genetics is particularly helpful in testing fusion protein functionality: Knockout and conditional-lethal (e.g., temperature-sensitive) alleles are available for most yeast genes, although not every gene has a detectable loss-offunction phenotype. If the GFP fusion is expressed from a plasmid, transformation allows it to be directly tested for the ability to rescue a conditional or nonlethal phenotype of a yeast gene mutation or knockout. To test for the ability of a plasmid-encoded GFP fusion gene to fully rescue an essential gene deletion, a “plasmid-swap” experiment can sometimes be performed [depending on the markers and strains available; see Rothstein (1991)]. If GFP is integrated at a chromosomal gene locus, functionality tests are simplified since the only form of the gene expressed (in a haploid strain) is the GFP fusion. If a gene displays a conditional or nonlethal loss-of-function phenotype, the haploid strain carrying the GFP fusion can be directly tested for the mutant phenotype. For genes that are essential, the viability of the integrated haploid can attest to the (at least partial) functionality of the gene fusion. A caveat to interpretations of integrated GFP fusion gene functionality is the possibility that extragenic mutations that suppress the lethal phenotype of a nonfunctional gene fusion might have been acquired during strain construction. To avoid this possibility, integrations should be performed in a diploid strain in which the second allele of the gene can complement, if necessary, any nonfunctional integration. Subsequent functionality tests can then be performed with haploid GFP fusion integrants isolated by sporulation of the heterozygous diploid. It is important to note that full functionality of a GFP fusion gene does not necessarily ensure that the major fluorescent signal observed reflects the functional localization of the translated protein. For example, a relatively small fraction of the fusion protein

EXPRESSION OF GFP IN YEASTS

population may localize properly and provide function, while the majority mislocalizes due to the addition of the GFP tag. It is also possible, but in our experience uncommon, that the expression of a particular GFP fusion protein may be toxic to yeast cells, even in the presence of the endogenous, untagged protein product. We have found that this most often occurs during fusion protein overexpression, and it may be alleviated by expressing the fusion protein at levels that more closely mimic those of the endogenous protein.

9.2.5 Expression of GFP Spectral Variants and Red Fluorescent Proteins in Yeast In vivo, multispectral imaging has become feasible in yeast and other cell types through the discovery and development of fluorescent proteins with absorbance and emission spectra distinct from GFP. These fluorescent proteins include mutant versions of GFP with blue, cyan, and yellow fluorescence (BFP, CFP, and YFP respectively), as well as a red fluorescent protein (RFP, also called DsRed) from the sea coral of the genus Discosoma [reviewed in Lippincott-Schwartz and Patterson (2003)]. While CFP and YFP have proven to be reliable in vivo fluorescent markers in S. cerevisiae, S. pombe, and C. albicans, the third GFP variant BFP displays (at best) weak fluorescence in yeast, and also rapidly photobleaches, and the short UV wavelengths required for excitation can be damaging to living cells (Hailey et al., 2002; Tatchell and Robinson, 2002). As a result, the CFP/YFP pair is preferable for dual labeling in yeast cells (see Section 9.3.2.1). Furthermore, CFP can serve as a FRET donor to YFP, permitting in vivo interaction studies with these two fluorophores (see Section 9.3.3). Resources and references for the expression of CFP and YFP fusion genes in yeasts have been included in previous sections (see Sections 9.2.1–9.2.3). Fluorescent proteins with emission wavelengths in the red region of the spectrum have more recently been cloned from marine organisms including the Discosoma coral (Matz et al., 1999). The longer wavelength of light emitted by these red-shifted GFP homologs minimizes light scattering and autofluorescence by living cells and, importantly, has opened the door for in vivo triple-labeling experiments (with CFP and YFP) and the potential for new FRET partners. Unfortunately, the Discosoma red fluorescent protein, now termed DsRed, has two major drawbacks as an in vivo fluorescence marker. First, DsRed requires >48 hours to reach >90% of maximal fluorescence, a length of time that is prohibitive for use in a fast-growing organism such as yeast (Baird et al., 2000). The second drawback of DsRed is that it exists in solution and in vivo as a tetramer (Baird et al., 2000). While this oligomerization does not preclude the use of DsRed as an organelle or expression marker in yeast, it does complicate interpretation of the localization and function of DsRed fusion proteins. Directed optimization of DsRed, however, appears to have successfully addressed these shortcomings. Bevis and Glick (2002) reported on the isolation of a DsRed variants that mature 10–15 times faster than DsRed, albeit with lower fluorescence intensity. Encouragingly, these authors found that a mitochondrially targeted DsRed variant, in contrast to the parental DsRed, gave consistently strong fluorescence in S. cerevisiae cells from growing cultures. As mentioned earlier, the Yeast Resource Center has a template plasmid available for the targeted integration of one of these DsRed variants (termed DsRed.T1.N1) into the S. cerevisiae and S. pombe genomes (see Sections 9.2.1 and 9.2.2); however, researchers should be aware that this DsRed variant is still prone to tetramerization. To further improve DsRed as an in vivo fluorescent marker, Campbell et al. (2002) subjected one of these rapidly maturing DsRed variants to stepwise evolution, leading ultimately to the isolation of both a monomeric (mRFP) and tandem dimer (effectively

185

186

THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS

monomeric, termed tdimer2) DsRed. Several studies have confirmed that both mRFP and tdimer2 can be expressed successfully in S. cerevisiae [see, for example, Huh et al. (2003), Malinska et al. (2003), and Sheth and Parker (2003)]. Thus, the mRFP and tdimer2 DsRed variants appear to be the best currently available option for expression of red fluorescence in yeast. Resources necessary for the expression of these monomeric DsRed variants in yeast should be available, with the proper permissions, from the aforementioned authors.

9.3

UTILIZATION OF GFP AND GFP FUSION PROTEINS IN YEAST

GFP has revolutionized microscopic studies of dynamic, in vivo protein localization in the yeast cell. Recent years have witnessed a second revolution, in which GFP technology has been successfully applied to studies of organelle function and inheritance, in vivo DNA, RNA, and lipid localization, and the mapping of in vivo protein–protein interactions (see Fig. 9.1). Finally, GFP has become a permanent addition to the yeast geneticists’ tool kit,

Figure 9.1. In vivo macromolecular localization via GFP. (A) Proteins can be localized in living cells by expressing them as GFP fusions. (B) DNA can be visualized by expressing GFP fused to a DNA-binding protein (DBP) or to a DNA binding domain (DBD) that recognizes specific DNA sequences or regions. (C) RNA molecules engineered to contain hairpin tertiary structures can be localized in vivo by expressing GFP fusions to RNA binding domains (RBDs) that recognize such hairpins. (D) Phosphoinositides have been visualized in living cells by expressing GFP fusions to lipid-binding domains (LBDs) such as the plextrin homology (PH) and FYVE domains. Details regarding each of these approaches to localizing proteins, nucleic acids, and lipids are given in Section 9.3.2.

UTILIZATION OF GFP AND GFP FUSION PROTEINS IN YEAST

serving as a reporter for gene expression studies and in large-scale forward-genetic visual screens. In this section, we review several studies that highlight the diverse, creative, and powerful uses of GFP and GFP fusions in yeast.

9.3.1

Organelle Structure, Function, and Inheritance

GFP fusion proteins have served as valuable markers for intracellular compartments in studies of yeast organelle structure, function, and inheritance. Additionally, these reporters can be used to confirm colocalization of a protein or other macromolecule of interest to a specific subcellular compartment. GFP fusion proteins that have been utilized for studies of nearly every organelle and compartment in yeast have been compiled with corresponding references in two recent reviews (Kohlwein, 2000; Tatchell and Robinson, 2002). Typically, organelles are visualized by fusing GFP to either (1) a protein that is well characterized to localize exclusively to the organelle/compartment of interest or (2) the minimal amino acid sequence required for protein targeting to the organelle/compartment of interest. As an example of the first approach, Prinz et al. (2000) studied the dynamic structure of the cortical ER in S. cerevisiae by monitoring a GFP fusion to the wellcharacterized ER protein Sec63p (see Fig. 9.2). The second approach is exemplified by the recent description of a mitochondrially localized GFP construct (mtGFP) generated by fusing the mitochondrial presequence (targeting signal) at the N-terminus of GFP

Figure 9.2. Expression of Sec63-GFP highlights the ER of living yeast cells. The continuous perinuclear ER is best seen in images acquired while focusing on the center of the cells (top panels), while the cortical ER network is best seen with focal planes close to the periphery of the cell (bottom panels). Bars, 5 mm. [Reproduced from Prinz et al. (2000) with permission by The Rockefeller University Press.]

187

188

THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS

(Westermann and Neupert, 2000). This mtGFP reporter has been used, for example, to characterize factors required for normal mitochondrial morphology in the budding yeast cell (Messerschmitt et al., 2003). Similarly, the inheritance of peroxisomes has been monitored by in vivo time-lapse microscopy of budding yeast expressing a GFP fusion to the type I peroxisomal targeting sequence (GFP-PTS1; Hoepfner et al., 2001). Dual labeling of peroxisomes and the actin cytoskeleton, in part through the use of the GFP variants CFP and YFP, along with the use of yeast mutants, further allowed these authors to demonstrate that peroxisome movement into the daughter cell is dependent upon the action of the Myo2p myosin motor protein along actin filaments.

9.3.2

Localization Studies

9.3.2.1 Protein Localization. GFP fusions have been utilized to determine the in vivo steady-state localization of many proteins, thereby guiding hypotheses of protein function. In fact, a large-scale study cataloging the in vivo steady-state localizations of >4000 S. cerevisiae proteins (representing 75% of the yeast proteome) by GFP tagging and microscopy has recently been published (Huh et al., 2003). The true worth of GFP, however, has been revealed in studies of dynamic protein movement and trafficking within the yeast cell. For example, several groups have utilized GFP fusion to better understand the signals, mechanisms, and regulation of nucleocytoplasmic transport. Görner et al. (1998) found that GFP fusions to the stress-response transcription factors Msn2p and Msn4p are normally localized to the cytoplasm, but accumulate rapidly in the nucleus when stress is applied. By localizing Msn2p-GFP in a panel of yeast strains deleted or mutated for nuclear transport receptors (also termed importins/exportins or karyopherins), we found that the exportin Msn5p was responsible for Msn2p-GFP nuclear export [our unpublished results; see also Chi et al. (2001) and Gorner et al. (2002)]. Figure 9.3 depicts the constitutive nuclear localization of Msn2p-GFP in Dmsn5 cells, even in the absence of stress. Jacquet et al. (2003) tracked the stress-activated nuclear translocation of Msn2pGFP in real time by high-resolution time-lapse video microscopy. Strikingly, these authors found that an intermediate stress response (triggered by the microscope excitation light) caused the entire population of Msn2p-GFP to shuttle repeatedly into and out of the nucleus with a periodicity of a few minutes. This dynamic behavior of Msn2p and its implications for the autoregulation of Msn2p localization would not have been appreciated without the capability afforded by GFP to follow protein localization in living cells in real time. Further examples of the utility of GFP in studies of nucleocytoplasmic transport are plentiful. Convenient GFP reporter proteins have been generated to monitor the rates and requirements for nuclear localization signal (NLS)- and nuclear export signal (NES)dependent nuclear transport in yeast (Roberts and Goldfarb, 1998; Shulga et al., 1996; Stade et al., 1997; Taura et al., 1998). Fusions of GFP to components of the ribosome have led to the development of in vivo assays to identify the nuclear transport pathway(s) of this large riboprotein complex (Hurt et al., 1999; Stage-Zimmermann et al., 2000). Finally, GFP fusion proteins have been effectively used to study the localization and movements of nuclear transport factors and nuclear pore complexes themselves, shedding further light on the striking dynamics of the nucleocytoplasmic transport machinery (Belgareh and Doye, 1997; Bucci and Wente, 1997; Seedorf et al., 1999; Stade et al., 1997). Studies of the dynamic behavior, function, and regulation of the mitotic spindle in yeast have also benefited greatly by the use of GFP fusion proteins. Using a fusion of the

UTILIZATION OF GFP AND GFP FUSION PROTEINS IN YEAST

Figure 9.3. Stress- and Msn5-regulated nuclear transport of Msn2-GFP in S. cerevisiae. Budding yeast cells expressing a GFP fusion to the Msn2 transcription factor were visualized by fluorescence and Nomarski differential interference contrast (DIC) microscopy in the absence and presence of stress (0.4 M NaCl). Msn2-GFP can be seen to localize throughout the cell in the absence of stress and to relocalize tightly to the nucleus in the presence of stress (left panels). The constitutive localization of Msn2-GFP in Dmsn5 cells (right panels) indicates that the Msn5 nuclear transport receptor is required for nuclear export of this transcription factor.

budding yeast spindle pole body (SPB; the yeast equivalent of the centrosome) antigen Nuf2p to GFP and time-lapse fluorescence microscopy in living cells, we determined the rate and polarity of mitotic spindle growth in vivo [see Fig. 9.4 and Kahana et al. (1995)]. Similarly, Carminati and Stearns (1997) expressed GFP fusions to the microtubule component a-tubulin (Tub1p) to localize astral and mitotic spindle microtubules in vivo. From their time-lapse observations of GFP-labeled microtubules through the cell cycle, the authors were able to document the role of astral microtubules in mitotic spindle positioning. More recently, Maddox et al. (2000) studied the polarity and dynamics of the minus ends of budding yeast microtubules at the SPB through the in vivo monitoring of a GFPtagged a-tubulin subunit by fluorescent speckle microscopy (FSM) and fluorescence redistribution after photobleaching (FRAP). Based on their findings, the authors were able to conclude that the dynamic assembly and disassembly of astral and spindle microtubules most likely occurs at their plus, and not minus, ends. Finally, GFP-tagged tubulin expression has also allowed assessment of microtubule dynamics in the mitotic spindle of S. pombe (Mallavarapu et al., 1999) and observation of microtubules in budding and hyphal forms of C. albicans (Bachewich et al., 2003; Hazan et al., 2002). An exciting recent development to GFP-based studies in yeast is the use of the YFP and CFP spectral variants, as well as the related red fluorescent protein DsRed, to make real-time observations of the dynamic colocalization of proteins in living cells. The power in this approach is exemplified in an elegant study by Browning et al. (2003) in which the authors investigated the in vivo movement of cargo proteins along microtubules in fission yeast. In this study, a fusion of the Tea2p kinesin protein to YFP was observed by timelapse fluorescence microscopy in S. pombe cells that coexpressed a tubulin-CFP fusion. In real time, Tea2p-YFP could be seen loading onto microtubules in the middle of the cell near the nucleus and traveling at the tips of polymerizing microtubules toward the end of

189

190

THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS

Figure 9.4. Time-lapse fluorescence microscopy montage of budding yeast cells expressing the centrosomal antigen Nuf2-GFP. In frames in which the centrosomes (white dots) are parfocal, the distance between them is measured in CCD pixels (14.7 pixels = 1 mm). A time stamp is included at the bottom of each frame.

the cell. By introducing ATPase mutations into the Tea2p fusion protein, the movement of this kinesin to the cell ends was shown to require motor activity. By monitoring the movement of wild-type Tea2p-YFP in various mutant backgrounds, the authors were further able to identify proteins that regulate Tea2p cell-end transport both at early (microtubule loading) and late (cell-end anchoring) steps. 9.3.2.2 DNA Localization. GFP-based methods for monitoring in vivo DNA localization have revolutionized studies of chromosomal movement, localization, interactions, and architecture in the context of nuclear processes including mitosis, meiosis, DNA transcription, silencing, repair, and recombination. Generally, yeast DNA can be visualized in two ways with GFP-based systems. First, endogenous proteins that interact with DNA (such as histones, centromere components, or silencing factors) can be tagged with GFP and utilized as an indirect monitor of the dynamics of the DNA region(s) they bind.

UTILIZATION OF GFP AND GFP FUSION PROTEINS IN YEAST

As an example of this approach, Pidoux et al. (2000) utilized a GFP fusion to the Swi6p heterochromatin-binding protein as a centromeric and telomeric marker during real-time analysis of chromosome segregation in S. pombe; by monitoring GFP-Swi6p in chromosome segregation mutants, the authors were able to observe the wide range of behavior displayed by lagging chromosomes. GFP fusions to histone proteins (H2B and H4) have also been utilized to monitor nuclear migration and movement during cell division in wildtype and mutant budding yeast (Hoepfner et al., 2000; Thrower et al., 2003). As a second GFP-based approach to localizing DNA, specific chromosomal sites can be “tagged” with direct repeats of bacterial operator sequences (typically derived from the lac or tet operons); these tags can then be visualized using GFP fusions to the appropriate repressor (DNA-binding) protein [reviewed in Belmont (2001)]. An excellent example of this second approach can be found in a study of the dynamic localization of chromosomal regions during interphase in budding yeast. Heun et al. (2001) tagged early and late replication origins, as well as centromeric and telomeric loci, with tandem lac operator sequences; these four chromosomal regions were then monitored by expression of LacIGFP and time-lapse microscopy. Coexpression of a GFP-tagged nuclear pore protein allowed for the measurement of DNA movements relative to either the nuclear periphery or the calculated center of the nucleus. These experiments revealed that interphase chromatin is highly dynamic (in the case of replication origins, moving distances ≥0.5 mm within seconds) and that different chromosomal domains display varying degrees of constraint on their movement. This GFP-based chromosomal tagging technique has also been successfully applied to S. pombe (Nabeshima et al., 1998). Despite the obvious power of this approach, a valid concern is that the integration of bacterial operator sequences (often up to 10 kb) might significantly interfere with the normal, physiological behavior of a given chromosomal region. Gasser (2002) has suggested that such inserts do not substantially alter local chromatin structure, but such a finding would almost certainly depend on the specific genomic region. Therefore, inclusion of functional tests with the altered genomic region would significantly increase the credibility of any such studies. 9.3.2.3 RNA Localization. Studies of RNA localization and movement in yeast, especially for single transcripts, have been difficult due to the relative insensitivity of fluorescence in situ hybridization (FISH) techniques and the inability to perform injections of fluorescently labeled RNA. Furthermore, the extensive manipulations required for such techniques often call into question the physiological relevance of resulting observations. Within the past five years, a number of research groups have developed systems that allow for the GFP-based in vivo localization of RNA [reviewed in Brodsky and Silver (2002)]. Two components are required for GFP-based RNA imaging: an RNA-binding protein (RBP) fusion to GFP and an RNA (typically the RNA of interest) engineered to contain the binding sites (typically a hairpin structure) for the RBP. Two hairpin–RBP interactions have been used for GFP-based RNA imaging: one derived from the bacteriophage MS2 capsid protein and the other from human splicing protein U1A (each along with its cognate RNA hairpin binding sites). In yeast coexpressing these two components, the hairpincontaining RNA is bound by the RBP-GFP fusion; in this way, GFP fluorescence serves as an indirect indicator of RNA localization. Functional tests with the altered RNA of interest can be performed to attempt to allay concerns that the introduced hairpin structures (often in tandem repeats) interfere with physiological localization and/or function of the RNA. These GFP-based RNA localization techniques, combined with yeast genetics, have contributed significantly to the identification of factors required for mRNA nuclear export

191

192

THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS

(Brodsky and Silver, 2000) and bud-tip localization of the ASH1 mRNA (Beach and Bloom, 2001; Beach et al., 1999; Bertrand et al., 1998). Additionally, these techniques have allowed the relatively rapid screening and identification of additional nuclear encoded mRNAs that localize to the mitochondria or bud tip in yeast (Corral-Debrinski et al., 2000; Marc et al., 2002; Shepard et al., 2003). Finally, a recent and particularly elegant study employed a combination of GFP-based protein- and RNA-localization techniques to investigate the significance of cytoplasmic foci formed by mRNA decay factors in budding yeast (Sheth and Parker, 2003). By colocalizing mRNA degradation intermediates in vivo to these foci, Sheth and Parker identified these foci as actual sites of mRNA decay, rather than as storage or assembly sites for decay factors. As such, this finding adds tremendously to our growing appreciation for the importance of subcytoplasmic “compartments” in mRNA sequestration and regulation. 9.3.2.4 Lipid Localization. The development of GFP-based reporters for the in vivo detection of specific phosphorylated pools of the eukaryotic membrane component phosphatidylinositol (PtdIns) has greatly aided studies of the compartmentalized regulatory roles of these signaling lipids in both mammalian and yeast cells. Such GFP-based lipid reporters, sometimes termed FLAREs (for fluorescent lipid-associated reporters), take advantage of protein domains that have been identified in yeast and mammalian proteins to bind with high affinity and specificity to the inositide head group of particular phosphoinositides (see Balla et al., 2000; Balla and Varnai, 2002). By expressing a GFP fusion to any one of these phosphoinositide-binding modules, the lipid population corresponding to that module’s specificity can in principle be detectable by fluorescence microscopy of single living yeast cells. Stefan et al. (2002) visualized cellular pools of PtdIns(4,5)P2 at the budding yeast plasma membrane by expressing GFP fusions to a PtdIns(4,5)P2-binding plextrin homology (PH) domain derived from mammalian phospholipase C d1. Their finding that this lipid pool mislocalized in yeast mutated for a class of phosphoinositide 5-phospatases, concurrent with defects in cell morphology and membrane trafficking, underscores the importance of proper PtdIns(4,5)P2 compartmentalization to these biological processes. The same authors, using a GFP fusion to the PH domain derived from the mammalian FAPP1 protein, localized PtdIns(4)P pools to intracellular Golgi compartments, consistent with studies associating this phosphoinositide with transport of proteins from the Golgi (Stefan et al., 2002). In an earlier study from the same group, PtdIns(3)P was localized in vivo to endocytic compartments by a GFP-FYVE fusion protein, in agreement with its function in protein sorting in the late secretory pathway of yeast (Burd and Emr, 1998). Finally, GFP-FYVE has additionally been expressed in fission yeast to investigate the role of PtdIns(3)P in formation of the forespore membrane during sporulation (Onishi et al., 2003). It is worth noting that as protein domains with novel lipid-binding specificities continue to be identified, the repertoire of lipids (phosphoinositides and others) that can be detected and studied in living yeast cells by GFP will almost certainly expand.

9.3.3

Fluorescence Resonance Energy Transfer

The GFP derivatives CFP and YFP can be used in fluorescence resonance energy transfer (FRET) studies to detect in vivo interactions between two tagged proteins. During FRET, the excited donor fluorophore (CFP) directly transfers energy to the acceptor fluorophore (YFP); FRET can be detected, therefore, by measuring emission from YFP in the presence of wavelengths excitatory to CFP. Because FRET only occurs if the donor and accep-

UTILIZATION OF GFP AND GFP FUSION PROTEINS IN YEAST

tor moieties are very close in space (maximum separation of ~25–35 Å), the presence of a FRET indicates a high probability that the two tagged proteins directly interact. Yeastbased in vivo FRET studies especially benefit from the ability to express CFP- and YFPfusion proteins at controlled and consistent physiological levels from integrated genomic fusions. This technical advantage greatly increases the likelihood that the fusion protein will maintain functional localizations and interactions and that consistent FRET values will be obtained from cell to cell. Readers are directed to a recent method-based review by Hailey et al. (2002) of FRET applications in yeast. Several groups have applied the FRET assay to studies of yeast protein–protein interactions. In one study from our laboratory, potential pairwise interactions between YFPtagged nuclear pore proteins (termed nucleoporins) and CFP-tagged nuclear transport receptors were tested by microscopic FRET analysis (Damelin and Silver, 2000). Coimmune precipitation was able to confirm at least one novel receptor–nucleoporin interaction detected by FRET, bolstering the credibility of results obtained by this method. In all, this work revealed that distinct nuclear import and export receptors traverse the nuclear pore complex by overlapping, but not identical, pathways. A refined molecular model of protein-protein interactions within the nuclear pore complex was generated in a second study by testing for FRET interactions between nucleoporins themselves (Damelin and Silver, 2002). Finally, FRET has been utilized by Blumer and colleagues to demonstrate and map sequences required for the in vivo oligomerization of the G-protein coupled mating factor receptor Ste2p at the yeast plasma membrane (Overton and Blumer, 2000; Overton et al., 2003).

9.3.4

Gene Expression and Genetic Studies with GFP

GFP has become one of the most widely used reporters of gene expression, largely because the addition of exogenous substrates and/or cell disruption is not required for fluorescence, along with the ease of detection by fluorescence microscopy, spectrophotometry, and flow cytometry. A study by Li et al. (2000), for instance, monitored the in vivo kinetics of induction from the yeast GAL1 promoter using GFP as a reporter. The stability of GFP can be advantageous in studies of promoter activation, where low expression levels and sensitivity are issues. If, however, decreases and/or dynamic changes in gene expression must be monitored, the stability of the GFP protein can mask downregulation of transcription. To address this concern, a destabilized form of GFP has been generated for yeast by the addition of the ubiquitin/proteasome-targeting PEST motif of the constitutively unstable yeast Cln2p protein (Mateus and Avery, 2000). Notably, GFP-Cln2pPEST displayed a halflife of ~30 minutes as opposed to the ~7-hour half-life of unaltered GFP. Using this destabilized GFP reporter, the authors were able to monitor dynamic gene expression from the Cu2+-regulated CUP1 promoter as well as the cell-cycle regulated CLN2 promoter. Along similar lines, silencing- and promoter-responsive GFP expression constructs have been developed to specifically monitor processes such as mating type switching and the response to DNA damage in yeast (Laney and Hochstrasser, 2003; Walmsley et al., 1997). Finally, GFP has found important use as a reporter for several large-scale forward genetic screens in yeast. Because GFP can be rapidly observed by microscopy, it has become feasible to screen yeast temperature-sensitive or knockout libraries for mutants that improperly localize a given GFP fusion protein. For example, Ryan and Wente (2002) screened a library of temperature-sensitive mutants for mislocalized GFP-labeled nuclear pore complexes (NPCs) in an effort to identify yeast mutants affecting NPC assembly.

193

194

THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS

Interestingly, their identification of yeast mutated for regulators of the small GTPase Ran in this screen suggests for the first time a potential role for the Ran GTPase cycle in NPC assembly (Ryan et al., 2003). If the specific parameters permit, one can pre-enrich for mutants of interest by fluorescence activated cell sorting (FACS) of GFP-expressing yeast. Hammell et al. (2002) successfully identified yeast mutants affected for mRNA nuclear export by first enriching by FACS for mutants that failed to express a GFP fusion to the heat-shock response protein Ssa4p after shift to 42°C. More labor-intensive secondary screening (in situ hybridization to detect the SSA4 mRNA) then identified the subset of mutants that failed to express Ssa4p-GFP due to blocked mRNA export. The results of this genetic screen nicely reinforce the interdependence of mRNA 3¢ processing and nuclear export.

9.4

METHODS

The following two sections outline several methods for GFP fusion protein expression in S. cerevisiae and the detection of such proteins by fluorescence microscopy. For more detail on methods regarding S. pombe and C. albicans expression systems, as well as yeast DNA-, RNA-, and lipid-localization, FRET microscopy techniques and quantification, the reader is referred to reviews cited in the relevant sections above.

9.4.1

S. cerevisiae GFP Expression Methods

9.4.1.1 Extrachromosomal (Plasmid) GFP Constructs. We have designed a series of yeast plasmid vectors for the construction of GFP fusion proteins in budding yeast. The pCGF (C-terminal GFP fusion) series of high-copy vectors are designed for pGAL1-10 inducible expression of yeast gene fusions to the 3¢ end of GFP [orienting GFP at the N-terminus of the protein of interest; see Kahana and Silver (1996)]. Once an ORF is cloned into a pCGF vector and transformed into S. cerevisiae, high-level fusion gene expression can be induced by addition of galactose to the growth medium. Notably, the GAL1-10 inducible promoter can be extremely useful for expression of non-yeast cDNAs or yeast genes normally expressed at levels too low to be seen with GFP. While this system may be useful for some proteins, the atemporal and typically highlevel expression may cause several problems. First, such expression may, depending on the gene fused, have toxic side effects. Second, overexpression of a GFP fusion gene may cause it to mislocalize. For instance, when a Nuf2p-GFP fusion is expressed from a pCGF vector in the presence of 2% galactose, fluorescence is observed throughout the cell as opposed to its normal exclusive localization to the spindle pole body (SPB). Some of these problems can be ameliorated by repressing expression of the GFP fusion gene with glucose after sufficient induction in galactose. We have observed that expression of Nuf2p-GFP from a pCGF vector using a one-hour galactose “pulse” followed by a six-hour glucose “chase” leads to accurate localization of the fusion protein [at ~1000–5000 molecules Nuf2p-GFP/SPB; see Kahana et al. (1995)]. Presumably this effect is attributable to dilution of protein level after turnover and cell division during the glucose chase. Moreover, mixtures of galactose and glucose can give low- and intermediate-level expression, thus modulating protein levels even further. A second potential solution to problems with GFP fusion gene misexpression and protein mislocalization can be to direct fusion gene expression by the gene’s endogenous promoter. To do this in a plasmid-based system, we developed the centromeric (ARS/CEN)

METHODS

pNGF (N-terminal GFP fusion) vector that contains GFP followed by the NUF2 3¢ terminator sequence (Kahana and Silver, 1996). Once a gene and its upstream (5¢) promoter sequence (~500 bp is usually sufficient) are ligated in frame into pNGF, the resulting construct should express the fusion (with GFP oriented at the C-terminus of the gene of interest) at a level similar to that of the endogenous gene. Because the fusion is not overexpressed, toxic side effects are not likely, and functionality can be readily tested (see Section 9.2.4). 9.4.1.2 Integrating GFP Constructs. While centromeric plasmids such as pNGF are often able to approximate physiological expression levels of GFP fusion proteins in S. cerevisiae, experiments utilizing such constructs can still be hindered by the need to maintain plasmid selection, the variable expression of the fusion protein from cellto-cell, and/or complications/competition effects due to coexpression of the endogenous (untagged) chromosomal version of the gene (unless the experiment is performed in a knockout strain). To circumvent these remaining inconveniences, our laboratory has most often sought to express GFP fusion genes that are stably integrated into the yeast genome, most often at the 3¢ (translated C-terminus) of the chromosomal ORF for the gene of interest. As one approach to do this, we generated a plasmid based on pNGF that lacks ARS/CEN sequences—and as such can only be stably transmitted in budding yeast if integrated into the genome. By cloning a short region of the 3¢ end of the gene of interest (~200–400 bp) upstream of and in-frame with GFP in this “integrating” pNGF plasmid, and subsequently linearizing the resulting plasmid within the gene-specific sequence, integration can be targeted specifically at the 3¢ end of the ORF of interest by homologous recombination [e.g., Seedorf et al. (1999)]. As a second approach to generate integrated C-terminal fusions of GFP to genes of interest, we have also had significant success with the PCR-based method introduced by Knop et al. (1999). By this approach, a module containing GFP and a dominant selectable marker is amplified by PCR with primers that contain ~45 bp of flanking homology to the desired site of integration.

9.4.2

Assessing Expression

Several approaches can assess GFP fusion expression. The first, and most straightforward, is to look for fluorescence by microscopy. While this method will show that the GFP moiety is being accurately expressed, it does not prove that the entire fusion protein is being made. For instance, if an ORF is ligated out-of-frame into a pCGF vector, cells will fluoresce in the presence of galactose due to expression of unfused GFP. Conversely, if a protein is rapidly degraded within the cell, the GFP may not mature into its fluorescent form in time to be detected. Hence, the most accurate method of assessing expression is the immunoblot, which allows detection of immature as well as mature GFP. If antibodies against the protein being fused are available, they should recognize the fusion protein (which should run ~28 kDa larger than the unfused protein). Furthermore, such antibodies can be used to assess the relative levels of the fused to the unfused protein by the relative intensities of the signals on the blots. If antibodies to the untagged protein are not available, anti-GFP monoclonal and polyclonal antibodies are commercially available from a variety of commercial sources. To eliminate the possibility of interpreting a nonspecific anti-GFP reactive band as evidence for GFP fusion protein expression, a negative control experiment (e.g., yeast lysate without GFP) should be performed. Once fusion protein expression has been confirmed, it is additionally important to assess, if at all possible, its functionality. If a GFP fusion to a protein of interest retains

195

196

THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS

functionality, the researcher can be more confident that this reporter will accurately reflect the localizations and interactions of the endogenous (untagged) protein. A more detailed discussion of assessing GFP fusion protein functionality is provided in Section 9.2.4.

9.4.3

Microscopy

Because yeast are among the smallest eukaryotes (~5–10 mm in diameter), a high level of magnification must be used in fluorescence microscopy. For visual observations, we typically use a 60¥ or 100¥ oil immersion objective lens and a 10¥ eyepiece lens. For maximal brightness and resolution, we always use objectives with a numerical aperture of 1.4 (the highest commercially available). For recording images with a digital camera, we typically set the magnification in such a manner as to project 1 mm of the specimen onto 8-15 pixels of the detector. Using a CCD camera with 6.8 ¥ 6.8-mm2 pixels, we need only a total magnification of 60¥ to achieve about 9 pixels/mm sample (or ~0.1 mm resolution). To achieve this, we use a 60¥ 1.4 N.A. oil immersion objective lens without an intermediate eyepiece or projection lens. Both yeast and many types of yeast media exhibit yellow autofluorescence when excited with ultraviolet or blue light. Thus, a fluorescence filter set that maximizes GFP detection while minimizing autofluorescence must be used. We have found that the use of a standard “barrier pass” FITC filter set (excitation 460–500 nm, Dichroic 505 nm; Barrier 510-560 nm; Chroma Technology No. #41001 or equivalent) with the S65T isolate of GFP (excitation 488 nm, emission 520 nm) gives the highest signal/noise ratio for detection. Furthermore, the use of low-fluorescence media is often advantageous. Rich media such as YPD (yeast extract, peptone, and dextrose) tends to have high-background fluorescence. Less rich media such as “synthetic complete” generally fluoresce much less brightly (Adams et al., 1997). Furthermore, media that lacks tryptophan tends to have the lowest levels of autofluorescence. Budding yeast that have mutations in the ADE1 or ADE2 genes tend to accumulate a metabolic intermediate that interferes with the observation of GFP. Under normal room lighting, colonies of ade1 or ade2 cells appear pink; when observed by epifluorescence in FITC filter sets, individual cells from these colonies appear bright yellow or green. While it has been reported that the addition of supplemental adenine to yeast media diminishes the pinkness of colonies, we have observed that this method does not completely ablate the autofluorescence observed by fluorescence microscopy. Thus, use of ade1 and ade2 strains should be avoided with GFP. GFP fusion proteins are often not affected by the presence of formaldehyde or other chemicals typically used in immunofluorescence protocols. Thus, a GFP fusion may be able to be colocalized with a protein being detected by immunofluorescence or with DNA stains such as 4¢,6-diamidino-2-phenylindole (DAPI).

9.5

CONCLUSIONS

To summarize, GFP is an extremely useful tool for studying dynamic protein, nucleic acid, and lipid localization in yeast. Furthermore, studies of organelle movement and function, protein–protein interaction, gene expression, and genetics have benefited greatly from this tool. Conversely, the use of yeast as a model system offers many advantages for expression of GFP. As a result, many of the newest uses for GFP have been developed in this

REFERENCES

model eukaryote. Undoubtedly the coming years will see even more diverse, creative, and powerful uses for this “genetic fluorophore” in yeast.

REFERENCES Adams, A., Gottschling, D. E., Kaiser, C. A., and Stearns, T. (1997). Methods in Yeast Genetics, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Bachewich, C., Thomas, D. Y., and Whiteway, M. (2003). Depletion of a polo-like kinase in Candida albicans activates cyclase-dependent hyphal-like growth. Mol. Biol. Cell 14:2163–2180. Bader, G. D., Heilbut, A., Andrews, B., Tyers, M., Hughes, T., and Boone, C. (2003). Functional genomics and proteomics: Charting a multidimensional map of the yeast cell. Trends Cell Biol. 13:344–356. Bahler, J., Wu, J. Q., Longtine, M. S., Shah, N. G., McKenzie, A., 3rd, Steever, A. B., Wach, A., Philippsen, P., and Pringle, J. R. (1998). Heterologous modules for efficient and versatile PCRbased gene targeting in Schizosaccharomyces pombe. Yeast 14:943–951. Bailer, S. M., Siniossoglou, S., Podtelejnikov, A., Hellwig, A., Mann, M., and Hurt, E. (1998). Nup116p and nup100p are interchangeable through a conserved motif which constitutes a docking site for the mRNA transport factor gle2p. EMBO J. 17:1107–1119. Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2000). Biochemistry, mutagenesis, and oligomerization of DsRed, a red fluorescent protein from coral. Proc. Natl. Acad. Sci. USA 97:11984–11989. Balla, T., Bondeva, T., and Varnai, P. (2000). How accurately can we image inositol lipids in living cells? Trends Pharmacol. Sci. 21:238–241. Balla, T., and Varnai, P. (2002). Visualizing cellular phosphoinositide pools with GFP-fused proteinmodules. Sci. STKE 2002:PL3. Basi, G., Schmid, E., and Maundrell, K. (1993). TATA box mutations in the Schizosaccharomyces pombe nmt1 promoter affect transcription efficiency but not the transcription start point or thiamine repressibility. Gene 123:131–136. Beach, D. L., and Bloom, K. (2001). ASH1 mRNA localization in three acts. Mol. Biol. Cell 12:2567–2577. Beach, D. L., Salmon, E. D., and Bloom, K. (1999). Localization and anchoring of mRNA in budding yeast. Curr. Biol. 9:569–578. Belgareh, N., and Doye, V. (1997). Dynamics of nuclear pore distribution in nucleoporin mutant yeast cells. J. Cell Biol. 136:747–759. Belmont, A. S. (2001). Visualizing chromosome dynamics with GFP. Trends Cell Biol. 11:250–257. Berman, J., and Sudbery, P. E. (2002). Candida Albicans: A molecular revolution built on lessons from budding yeast. Nat. Rev. Genet. 3:918–930. Bertrand, E., Chartrand, P., Schaefer, M., Shenoy, S. M., Singer, R. H., and Long, R. M. (1998). Localization of ASH1 mRNA particles in living yeast. Mol. Cell 2:437–445. Bevis, B. J., and Glick, B. S. (2002). Rapidly maturing variants of the Discosoma red fluorescent protein (DsRed). Nat. Biotechnol. 20:83–87. Brodsky, A. S., and Silver, P. A. (2000). Pre-mRNA processing factors are required for nuclear export. RNA 6:1737–1749. Brodsky, A. S., and Silver, P. A. (2002). Identifying proteins that affect mRNA localization in living cells. Methods 26:151–155. Browning, H., Hackney, D. D., and Nurse, P. (2003). Targeted movement of cell end factors in fission yeast. Nat. Cell Biol. 5:812–818. Bucci, M., and Wente, S. R. (1997). In vivo dynamics of nuclear pore complexes in yeast. J. Cell Biol. 136:1185–1199.

197

198

THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS

Burd, C. G., and Emr, S. D. (1998). Phosphatidylinositol(3)-phosphate signaling mediated by specific binding to RING FYVE domains. Mol. Cell 2:157–162. Campbell, R. E., Tour, O., Palmer, A. E., Steinbach, P. A., Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2002). A monomeric red fluorescent protein. Proc. Natl. Acad. Sci. USA 99:7877–7882. Carminati, J. L., and Stearns, T. (1997). Microtubules orient the mitotic spindle in yeast through dynein-dependent interactions with the cell cortex. J. Cell Biol. 138:629–641. Cherest, H., Kerjan, P., and Surdin-Kerjan, Y. (1987). The Saccharomyces cerevisiae MET3 gene: Nucleotide sequence and relationship of the 5¢ non-coding region to that of MET25. Mol. Gen. Genet. 210:307–313. Chi, Y., Huddleston, M. J., Zhang, X., Young, R. A., Annan, R. S., Carr, S. A., and Deshaies, R. J. (2001). Negative regulation of Gcn4 and Msn2 transcription factors by Srb10 cyclin-dependent kinase. Genes Dev. 15:1078–1092. Cormack, B. P., Bertram, G., Egerton, M., Gow, N. A., Falkow, S., and Brown, A. J. (1997). Yeastenhanced green fluorescent protein (yEGFP)a reporter of gene expression in Candida albicans. Microbiology 143 (Pt 2):303–311. Corral-Debrinski, M., Blugeon, C., and Jacq, C. (2000). In yeast, the 3¢ untranslated region or the presequence of ATM1 is required for the exclusive localization of its mRNA to the vicinity of mitochondria. Mol. Cell Biol. 20:7881–7892. Craven, R. A., Griffiths, D. J., Sheldrick, K. S., Randall, R. E., Hagan, I. M., and Carr, A. M. (1998). Vectors for the expression of tagged proteins in Schizosaccharomyces pombe. Gene 221:59– 68. Damelin, M., and Silver, P. A. (2000). Mapping interactions between nuclear transport factors in living cells reveals pathways through the nuclear pore complex. Mol. Cell 5:133–140. Damelin, M., and Silver, P. A. (2002). In situ analysis of spatial relationships between proteins of the nuclear pore complex. Biophys. J. 83:3626–3636. De Backer, M. D., Magee, P. T., and Pla, J. (2000). Recent developments in molecular genetics of Candida albicans. Annu. Rev. Microbiol. 54:463–498. Faryar, K., and Gatz, C. (1992). Construction of a tetracycline-inducible promoter in Schizosaccharomyces pombe. Curr. Genet. 21:345–349. Gasser, S. M. (2002). Visualizing chromatin dynamics in interphase nuclei. Science 296:1412–1416. Gerami-Nejad, M., Berman, J., and Gale, C. A. (2001). Cassettes for PCR-mediated construction of green, yellow, and cyan fluorescent protein fusions in Candida albicans. Yeast 18:859– 864. Gorner, W., Durchschlag, E., Martinez-Pastor, M. T., Estruch, F., Ammerer, G., Hamilton, B., Ruis, H., and Schuller, C. (1998). Nuclear localization of the C2H2 zinc finger protein Msn2p is regulated by stress and protein kinase A activity. Genes Dev. 12:586–597. Gorner, W., Durchschlag, E., Wolf, J., Brown, E. L., Ammerer, G., Ruis, H., and Schuller, C. (2002). Acute glucose starvation activates the nuclear localization signal of a stress-specific yeast transcription factor. EMBO J. 21:135–144. Hailey, D. W., Davis, T. N., and Muller, E. G. (2002). Fluorescence resonance energy transfer using color variants of green fluorescent protein. Methods. Enzymol. 351:34–49. Hammell, C. M., Gross, S., Zenklusen, D., Heath, C. V., Stutz, F., Moore, C., and Cole, C. N. (2002). Coupling of termination, 3¢ processing, and mRNA export. Mol. Cell Biol. 22:6441–6457. Hazan, I., Sepulveda-Becerra, M., and Liu, H. (2002). Hyphal elongation is regulated independently of cell cycle in Candida albicans. Mol. Biol. Cell 13:134–145. Heun, P., Laroche, T., Shimada, K., Furrer, P., and Gasser, S. M. (2001). Chromosome dynamics in the yeast interphase nucleus. Science 294:2181–2186. Hoepfner, D., Brachat, A., and Philippsen, P. (2000). Time-lapse video microscopy analysis reveals astral microtubule detachment in the yeast spindle pole mutant cnm67. Mol. Biol. Cell 11:1197–1211.

REFERENCES

Hoepfner, D., van den Berg, M., Philippsen, P., Tabak, H. F., and Hettema, E. H. (2001). A role for Vps1p, actin, and the Myo2p motor in peroxisome abundance and inheritance in Saccharomyces cerevisiae. J. Cell Biol. 155:979–990. Hoffman, C. S., and Winston, F. (1989). A transcriptionally regulated expression vector for the fission yeast Schizosaccharomyces pombe. Gene 84:473–479. Huh, W. K., Falvo, J. V., Gerke, L. C., Carroll, A. S., Howson, R. W., Weissman, J. S., and O’Shea, E. K. (2003). Global analysis of protein localization in budding yeast. Nature 425:686–691. Hurt, E., Hannus, S., Schmelzl, B., Lau, D., Tollervey, D., and Simos, G. (1999). A novel in vivo assay reveals inhibition of ribosomal nuclear export in ran-cycle and nucleoporin mutants. J. Cell Biol. 144:389–401. Jacquet, M., Renault, G., Lallet, S., De Mey, J., and Goldbeter, A. (2003). Oscillatory nucleocytoplasmic shuttling of the general stress response transcriptional activators Msn2 and Msn4 in Saccharomyces cerevisiae. J. Cell Biol. 161:497–505. Kahana, J. A., Schnapp, B. J., and Silver, P. A. (1995). Kinetics of spindle pole body separation in budding yeast. Proc. Natl. Acad. Sci. USA 92:9707–9711. Kahana, J. A., and Silver, P. A. (1996). Use of the A. Victoria green fluorescent protein to study protein dynamics in vivo. In Current Protocols in Molecular Biology, Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. E., Seidman, J. G., Smith, J. A., and Struhl, K., Eds., (New York, NY, John Wiley & Sons, Inc.), pp. 9.7.22–29.27.27. Karin, M., Najarian, R., Haslinger, A., Valenzuela, P., Welch, J., and Fogel, S. (1984). Primary structure and transcription of an amplified genetic locus: The CUP1 locus of yeast. Proc Natl Acad Sci USA 81:337–341. Knop, M., Siegers, K., Pereira, G., Zachariae, W., Winsor, B., Nasmyth, K., and Schiebel, E. (1999). Epitope tagging of yeast genes using a PCR-based strategy: More tags and improved practical routines. Yeast 15:963–972. Kohlwein, S. D. (2000). The beauty of the yeast: Live cell microscopy at the limits of optical resolution. Microsc. Res. Tech. 51:511–529. Laney, J. D., and Hochstrasser, M. (2003). Ubiquitin-dependent degradation of the yeast Mat(alpha)2 repressor enables a switch in developmental state. Genes Dev 17:2259–2270. Li, J., Wang, S., VanDusen, W. J., Schultz, L. D., George, H. A., Herber, W. K., Chae, H. J., Bentley, W. E., and Rao, G. (2000). Green fluorescent protein in Saccharomyces cerevisiae: real-time studies of the GAL1 promoter. Biotechnol. Bioeng. 70:187–196. Lippincott-Schwartz, J., and Patterson, G. H. (2003). Development and use of fluorescent protein markers in living cells. Science 300:87–91. Longtine, M. S., McKenzie, A., 3rd, Demarini, D. J., Shah, N. G., Wach, A., Brachat, A., Philippsen, P., and Pringle, J. R. (1998). Additional modules for versatile and economical PCR-based gene deletion and modification in Saccharomyces cerevisiae. Yeast 14:953–961. Maddox, P. S., Bloom, K. S., and Salmon, E. D. (2000). The polarity and dynamics of microtubule assembly in the budding yeast Saccharomyces cerevisiae. Nat. Cell Biol. 2:36–41. Malinska, K., Malinsky, J., Opekarova, M., and Tanner, W. (2003). Visualization of protein compartmentation within the plasma membrane of living yeast cells. Mol. Biol. Cell 14:4427–4436. Mallavarapu, A., Sawin, K., and Mitchison, T. (1999). A switch in microtubule dynamics at the onset of anaphase B in the mitotic spindle of Schizosaccharomyces pombe. Curr. Biol. 9:1423–1426. Marc, P., Margeot, A., Devaux, F., Blugeon, C., Corral-Debrinski, M., and Jacq, C. (2002). Genomewide analysis of mRNAs targeted to yeast mitochondria. EMBO Rep. 3:159–164. Mateus, C., and Avery, S. V. (2000). Destabilized green fluorescent protein for monitoring dynamic changes in yeast gene expression with flow cytometry. Yeast 16:1313–1323. Matz, M. V., Fradkov, A. F., Labas, Y. A., Savitsky, A. P., Zaraisky, A. G., Markelov, M. L., and Lukyanov, S. A. (1999). Fluorescent proteins from nonbioluminescent Anthozoa species. Nat. Biotechnol. 17:969–973.

199

200

THE USES OF GREEN FLUORESCENT PROTEIN IN YEASTS

Maundrell, K. (1990). nmt1 of fission yeast. A highly transcribed gene completely repressed by thiamine. J. Biol. Chem. 265:10857–10864. Maundrell, K. (1993). Thiamine-repressible expression vectors pREP and pRIP for fission yeast. Gene 123:127–130. Messerschmitt, M., Jakobs, S., Vogel, F., Fritz, S., Dimmer, K. S., Neupert, W., and Westermann, B. (2003). The inner membrane protein Mdm33 controls mitochondrial morphology in yeast. J. Cell Biol. 160:553–564. Morschhauser, J., Michel, S., and Hacker, J. (1998). Expression of a chromosomally integrated, single-copy GFP gene in Candida albicans, and its use as a reporter of gene regulation. Mol. Gen. Genet. 257:412–420. Nabeshima, K., Nakagawa, T., Straight, A. F., Murray, A., Chikashige, Y., Yamashita, Y. M., Hiraoka, Y., and Yanagida, M. (1998). Dynamics of centromeres during metaphase-anaphase transition in fission yeast: Dis1 is implicated in force balance in metaphase bipolar spindle. Mol. Biol. Cell 9:3211–3225. Onishi, M., Koga, T., Morita, R., Nakamura, Y., Nakamura, T., Shimoda, C., Takegawa, K., Hirata, A., and Fukui, Y. (2003). Role of phosphatidylinositol 3–phosphate in formation of forespore membrane in Schizosaccharomyces pombe. Yeast 20:193–206. Overton, M. C., and Blumer, K. J. (2000). G-protein-coupled receptors function as oligomers in vivo. Curr. Biol. 10:341–344. Overton, M. C., Chinault, S. L., and Blumer, K. J. (2003). Oligomerization, biogenesis, and signaling is promoted by a glycophorin A-like dimerization motif in transmembrane domain 1 of a yeast G protein-coupled receptor. J. Biol. Chem. 278:49369–49377. Park, N. H., and Choi, W. (2002). Vectors designed for efficient molecular manipulation in Candida albicans. Yeast 19:1057–1066. Pidoux, A. L., Uzawa, S., Perry, P. E., Cande, W. Z., and Allshire, R. C. (2000). Live analysis of lagging chromosomes during anaphase and their effect on spindle elongation rate in fission yeast. J. Cell Sci. 113 (Pt 23):4177–4191. Prein, B., Natter, K., and Kohlwein, S. D. (2000). A novel strategy for constructing N-terminal chromosomal fusions to green fluorescent protein in the yeast Saccharomyces cerevisiae. FEBS Lett. 485:29–34. Prinz, W. A., Grzyb, L., Veenhuis, M., Kahana, J. A., Silver, P. A., and Rapoport, T. A. (2000). Mutants affecting the structure of the cortical endoplasmic reticulum in Saccharomyces cerevisiae. J. Cell Biol. 150:461–474. Roberts, P. M., and Goldfarb, D. S. (1998). In vivo nuclear transport kinetics in Saccharomyces cerevisiae. Methods Cell Biol. 53:545–557. Rothstein, R. (1991). Targeting, disruption, replacement, and allele rescue: Integrative DNA transformation in yeast. Methods Enzymol. 194:281–301. Ruohonen, L., Aalto, M. K., and Keranen, S. (1995). Modifications to the ADH1 promoter of Saccharomyces cerevisiae for efficient production of heterologous proteins. J. Biotechnol. 39:193–203. Ryan, K. J., McCaffery, J. M., and Wente, S. R. (2003). The Ran GTPase cycle is required for yeast nuclear pore complex assembly. J. Cell Biol. 160:1041–1053. Ryan, K. J., and Wente, S. R. (2002). Isolation and characterization of new Saccharomyces cerevisiae mutants perturbed in nuclear pore complex assembly. BMC Genet 3:17. Seedorf, M., Damelin, M., Kahana, J., Taura, T., and Silver, P. A. (1999). Interactions between a nuclear transporter and a subset of nuclear pore complex proteins depend on Ran GTPase. Mol. Cell Biol. 19:1547–1557. Shepard, K. A., Gerber, A. P., Jambhekar, A., Takizawa, P. A., Brown, P. O., Herschlag, D., DeRisi, J. L., and Vale, R. D. (2003). Widespread cytoplasmic mRNA transport in yeast: Identification

REFERENCES

of 22 bud-localized transcripts using DNA microarray analysis. Proc. Natl. Acad. Sci. USA 100:11429–11434. Sheth, U., and Parker, R. (2003). Decapping and decay of messenger RNA occur in cytoplasmic processing bodies. Science 300:805–808. Shulga, N., Roberts, P., Gu, Z., Spitz, L., Tabb, M. M., Nomura, M., and Goldfarb, D. S. (1996). In vivo nuclear transport kinetics in Saccharomyces cerevisiae: A role for heat shock protein 70 during targeting and translocation. J. Cell Biol. 135:329–339. Stade, K., Ford, C. S., Guthrie, C., and Weis, K. (1997). Exportin 1 (Crm1p) is an essential nuclear export factor. Cell 90:1041–1050. Stage-Zimmermann, T., Schmidt, U., and Silver, P. A. (2000). Factors affecting nuclear export of the 60S ribosomal subunit in vivo. Mol. Biol. Cell 11:3777–3789. Stefan, C. J., Audhya, A., and Emr, S. D. (2002). The yeast synaptojanin-like proteins control the cellular distribution of phosphatidylinositol (4,5)-bisphosphate. Mol. Biol. Cell 13:542–557. Tatchell, K., and Robinson, L. C. (2002). Use of green fluorescent protein in living yeast cells. Methods Enzymol. 351:661–683. Taura, T., Krebber, H., and Silver, P. A. (1998). A member of the Ran-binding protein family, Yrb2p, is involved in nuclear protein export. Proc. Natl. Acad. Sci. USA 95:7427–7432. Thrower, D. A., Stemple, J., Yeh, E., and Bloom, K. (2003). Nuclear oscillations and nuclear filament formation accompany single-strand annealing repair of a dicentric chromosome in Saccharomyces cerevisiae. J. Cell Sci. 116:561–569. Tsien, R. Y. (1998). The green fluorescent protein. Annu. Rev. Biochem. 67:509–544. Walmsley, R. M., Billinton, N., and Heyer, W. D. (1997). Green fluorescent protein as a reporter for the DNA damage-induced gene RAD54 in Saccharomyces cerevisiae. Yeast 13:1535–1545. Westermann, B., and Neupert, W. (2000). Mitochondria-targeted green fluorescent proteins: Convenient tools for the study of organelle biogenesis in Saccharomyces cerevisiae. Yeast 16:1421–1427. Yocum, R. R., and Johnston, M. (1984). Molecular cloning of the GAL80 gene from Saccharomyces cerevisiae and characterization of a gal80 deletion. Gene 32:75–82.

201

10 USES OF GFP IN CAENORHABDITIS ELEGANS Oliver Hobert and Paula Loria Department of Biochemistry and Molecular Biophysics, Center for Neurobiology and Behavior, Columbia University, College of Physicians and Surgeons, New York, NY

10.1

INTRODUCTION

Green fluorescent protein has become one of the standard tools of research in C. elegans. Its use in C. elegans appears even more widespread than in other metazoan model organisms, including Drosophila, not just because GFP was first introduced into C. elegans (Chalfie et al., 1994), but because several characteristics of C. elegans specifically favor the use of GFP. A key advantage of C. elegans for the use of GFP is the animal’s transparency and thin diameter (<100 mm), which allow visualization of GFP in living animals under a nonconfocal, regular compound microscope or even a regular dissecting scope equipped with a fluorescent light source. Another implicit advantage of C. elegans is the ease and speed with which transgenic animals expressing a GFP reporter can be constructed. A strong disadvantage of C. elegans that also favors the use of GFP is the technical difficulty in using RNA in situ analysis for gene expression analysis. While being the standard tool in virtually all other model organisms studied to date, RNA in situ analysis lacks the cellular resolution to visualize gene expression profiles in individual C. elegans cell types (for example, individual neurons) past the very early embryonic stage. As such, GFP reporter transgenes have become the major tool for expression pattern analysis in C. elegans. The use of GFP is of course not limited to expression pattern analysis. In this chapter we will follow-up on a remark made by Andy Fire and colleagues in the previous edition of this book chapter; they noted that “the field is just beginning to realize the breadth of Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

203

204

USES OF GFP IN CAENORHABDITIS ELEGANS

questions to which GFP technology can be applied” (Fire et al., 1998). Now, seven years later, we indeed have a broader feel for the myriad of applications in which GFP technology has proven to be useful. We will provide here an overview of the uses that have emerged over the past few years and will expand on some of the issues that Fire and colleagues mentioned in the previous edition of this chapter. We will specifically emphasize, though by no means entirely focus on, the uses of gfp in the nervous system, a bias that arises not only from the fact that one-third of the worm’s cells are neurons but also from our own personal research focus. Before discussing the uses of GFP, it is perhaps worth emphasizing the advantages of GFP over other reporter gene tags, the most commonly used before and after the advent of GFP being bGAL (O’Kane and Gehring, 1987). The specific advantage of GFP over bGAL is the ability to visualize reporter gene expression in live animals rather than in fixed preparations. This not only provides the trivial advantage of speed (no fixation required) but, more importantly, allows the visualization of “real time” processes, such as movement of proteins, organelles, or cells. The same animals can thus be watched over the course of several days if necessary. Also, the fact that animals do not have to be sacrificed for visualization of reporter gene expression is a significant added advantage if one utilizes genetic approaches requiring the isolation of individual animals with a mutant phenotype from a large population of mutagenized animals (in the course of mapping and/or genetic mutant screening). Moreover, fixation methods required for staining for bGal activity also tend to disrupt fine aspects of anatomy. It should nevertheless be kept in mind that many of the examples discussed below would work similarly well with bGal (e.g., determination of expression patterns).

10.2 10.2.1

USES OF GFP Use of GFP for Expression Pattern Analysis

The most basic use of GFP lies in analyzing gene expression patterns. In this approach, presumptive regulatory regions of the gene of interest (usually including 5¢ upstream regulatory regions with or without the inclusion of exons and introns of the gene of study) are fused to the coding region of GFP. In comparison to antibody staining, the obvious advantages of GFP reporters are the speed, low costs, and ease with which these tools can be constructed. These factors become critical if one aims to study whole gene families. For example, genome sequence-based expression pattern analyses of predicted odorant receptors (Troemel et al., 1995), nuclear hormone receptors (Miyabayashi et al., 1999), neuropeptide encoding genes (Nathoo et al., 2001), ion channels (Salkoff et al., 2001), Gproteins (Jansen et al., 1999), and immunoglobulin genes (Aurelio et al., 2002) were all conducted using GFP reporters. The results of these studies also point to a key issue in applying GFP reporter gene technology: One has to take the data for what they are—that is, not as a proof for the endogenous gene expression profile but merely as a hypothesisgenerating result. For example, the expression pattern of scores of orphan transmembrane receptors in chemosensory neurons immediately suggested a role for these gene as chemosensory receptors (Sengupta et al., 1996; Troemel et al., 1995), a notion later confirmed experimentally (Zhang et al., 1997). Or, for example, the unusual temporal aspects of expression of a family of Ig-domain proteins, suggested by GFP reporter analyses, revealed the unanticipated existence of postembryonic maintenance mechanisms in the nervous system, a notion confirmed by cell ablation and genetic analysis (Aurelio et al.,

USES OF GFP

2002). Similarly, on a more routine basis, if one positionally clones a gene based on a mutant phenotype, GFP expression analysis can help suggest where the gene of interest may act. Further studies, such as mosaic analysis or tissue-specific rescue experiments, are necessary to verify the GFP expression data; such verifications are advisable even if the expression pattern is determined using antibody staining. An additional advantage of GFP over antibody staining often emerges if a protein is highly localized within a cell. For example, if a protein is exclusively localized to axons within a fascicle, it is difficult, if not impossible, to identify the cell from which the axon emanates. In these cases, a transcriptional GFP reporter gene fusion which contains no sequences that may target the protein to subcellular sites and which therefore fills and labels the whole cell often allows the easy identification of the nature of the expressing cell. Second, if a protein is expressed only at very low levels and/or is highly unstable, GFP reporters are more likely to reveal sites of expression given the multicopy nature of transgenic arrays and the stability of GFP. In one particularly noteworthy example, Baum and co-workers observed that the expression of the ina-1 gene, which the authors had demonstrated to act in a specific neuronal type based on mosaic analysis, was not revealed by antibody staining (which readily detected protein expression in other, unrelated cell types) but was revealed through a GFP reporter (Baum and Garriga, 1997). Reporter gene analysis also has serious potential drawbacks. While RNA in situ analysis (which due to its lack of cellular resolution in postembryonic animals is rarely used in worms) or antibody staining measure the actual presence of the gene product, reporter gene constructs only provide an inference of where a gene is expressed because additional regulatory information may be missing from the reporter gene construct. In the majority of cases examined so far, most transcriptional regulatory information is contained within the first several kilobases upstream of the start of transcription. However, regulatory information often is localized within introns, particularly when the first intron is disproportionately large. Transcriptional regulatory control can also be exerted through elements located 3¢ of the coding sequence (e.g., Conradt and Horvitz, 1999). An example for the potentially incomplete nature of reporter constructs is provided by the initial characterization of the homeobox gene vab-7 which focused on its role in the patterning of posterior muscle and epidermal cells, which expressed a transcriptional vab-7 reporter gene fusion (Ahringer, 1996). Subsequently, antibody staining revealed expression of vab-7 in the nervous system which then led to the characterization of vab-7 function in nervous system development (Esmaeili et al., 2002). In addition, the absence of coding regions of the gene in a transcriptional, “promoter-only” reporter construct or the addition of GFP coding sequences to an inherently unstable protein in a translational reporter fusion may obscure posttranscriptional or posttranslational levels of gene expression control. Many of the above-mentioned drawbacks are minimized if one studies a gene for which a mutant phenotype is available. In such a case, a “rescuing” GFP reporter—that is, a reporter construct that contains all the coding region of the gene of interest and that is capable of rescuing the mutant phenotype—provides confidence that the sites of GFP fusion protein expression reflect the authentic sites of gene function.

10.2.2

Use of GFP to “Decode” Regulatory Sequences

The previous section focused on the use of reporter genes to study the expression of individual genes or gene families. Yet, there is a more fundamental issue attached to the study of gene expression patterns, which can be illustrated in the following way. A genome can

205

206

USES OF GFP IN CAENORHABDITIS ELEGANS

be considered to contain at least two distinct kinds of “coding information.” The code embedded in exons and translated into proteins and a cis-regulatory code, composed of transcription factor binding sites that determine spatial and temporal properties of gene expression (Davidson et al., 2002). Unlike amino acid codes, the nature of the cisregulatory code is a highly degenerate and context-dependent one. While computational methods have improved to identify and predict aspects of cis-regulatory logic, these approaches at best provide a mere suggestion that necessitates experimental confirmation. In the majority of studies to date, cis-regulatory information has been decoded through “promoter-bashing” methodology, which involves fusing fragments of genomic DNA surrounding a gene locus of interest to GFP to examine the expression pattern and then further subdividing the genomic regions into smaller elements until minimal, cis-regulatory elements required for cell specific expression are defined (e.g., Cui and Han, 2003; Eastman et al., 1999; Gupta and Sternberg, 2002; Harfe and Fire, 1998; Kirouac and Sternberg, 2003; Wenick and Hobert, 2004). To date, these approaches have largely been utilized in the context of studying a single gene. It is clear that in the future these approaches will be more focused on defining cis-regulatory elements for individual cell types rather than for individual genes. For example, microarray analysis identified scores of genes expressed in the pharynx, the feeding organ of C. elegans. Sequence comparison subsequently identified a conserved element in the cis-regulatory region of these genes, which, through gfp reporter analysis, was then shown to constitute a cis-regulatory “code” for pharynx expression (Gaudet and Mango, 2002). Cis-regulatory codes identified in this way can then be used to search the genome for more genes expressed in a specific cell type (Wenick and Hobert, 2004).

10.2.3

Use of GFP for Determining Protein Localization

If fused to the coding region of a gene and expressed under the control of either the gene’s own promoter or a heterologous promoter, the resulting GFP fusion protein can give valuable insights into the subcellular localization of a gene product. Particularly striking examples of localizing proteins to subcellular compartments include the GFP-based visualization of proteins at presynaptic specializations (Fig. 10.1A) (Nonet, 1999). The successful localization of a protein to a specific subcellular compartment provides immediate suggestions about its function that could not have been easily inferred by the primary sequence of the protein. For example, the visualization of a GFP tagged RNA binding protein, UNC-75, at dynamic subnuclear speckles suggested a role for this protein in premRNA splicing (Fig. 10.1B) (Loria et al., 2003). Deletion of amino acids required for correct subnuclear localization of GFP tagged UNC-75 also abolished the ability of UNC75::GFP to rescue the locomotory defects of unc-75 mutant animals, demonstrating the functional significance of subnuclear localization. In a more unusual twist of visualizing protein localization through reporter gene technology, a GATA-type transcription factor was found to localize on a target gene (its own promoter) in the nucleus of transgenic worms (Fig. 10.1C) (Fukushige et al., 1999). Localization of extracellular proteins is also possible (Vogel and Hedgecock, 2001). A potential caveat that one needs to bear in mind is that GFP tagged proteins expressed from multicopy arrays may lead to an artificial overexpression of the protein that can consequently suffuse the whole cell and thus obscure normal localization. The generation of multiple independent transgenic lines generated with different concentrations of injected DNA may circumvent this problem. A failure to see a specific subcellular localization may also be caused by the mere addition of the GFP moiety that may disrupt a structural determinant normally required for localization.

207

USES OF GFP

An advantage of using GFP reporter fusions for localization studies is their amenability for structure function analysis. Either by defining minimal domains sufficient for localization or by introducing mutations into “full-length” reporter constructs, domains of the protein required to localize to specific subcellular compartments can easily be deduced. For example, deletion and mutational analysis of the synaptic vesicle protein synaptogyrin (SNG-1) has defined a 38-amino-acid sequence within the C terminus of SNG-1 and a single arginine in the cytoplasmic loop between transmembrane domain 2 and 3 that are required for SNG-1 localization. These studies suggest that these two domains may represent components of signals that target synaptogyrin for endocytosis from the plasma membrane and direct synaptogyrin to synaptic vesicles, respectively (Zhao and Nonet, 2001). In another example, the previously enigmatic structural determinants of membrane

A

C

B

D

Figure 10.1. Examples of subcellular structures visualized with GFP. (A) Presynaptic specializations: Transgenic animals expressing a synaptobrevin-GFP fusion construct reveal localization of GFP to synaptic sites in all neurons. Here, punctate fluorescence can be seen in the SAB motor neurons. (Reprinted from Nonet, M., Visualization of synaptic specializations in live C. elegans with synaptic vesicle protein-GFP fusions, J. Neurosc. Methods, 89:33–40. Copyright © 1999, with permission from Elsevier.) (B) Splicing speckles: Live transgenic animals expressing rescuing unc75::GFP show GFP localization in subnuclear puncta predicted to be splicing speckles (Loria et al., 2003). Here, multiple puncta can be seen in the nucleus of a ventral cord motorneuron (arrows). The corresponding Nomarski-DIC image is on the right. (C) Transcription factor target sites: elt1::GFP binds to its own promoter and leads to discrete fluorescent foci in nuclei. The embryonic gut nuclei of eight-cell embryos homozygous for a transgenic array containing fully rescuing elt2::GFP are shown. Many of the nuclei show two striking and intense foci of fluorescence (arrows). The image represents a stack of serially collected 400-nm optical sections projected without further manipulation. [Courtesy of Fukushige et al. (1999).] (D) Dense bodies: Transgenic animals expressing a rescuing unc-97::GFP fusion construct show localization to discrete spots and lines that correspond to dense bodies (DB) and M lines (M) of the body wall muscle. The expression of the unc-97::GFP reporter gene can be monitored in live or fixed animals. Some subcellular structures appear more crisp in formaldehyde-fixed animals (shown here), although they are also distinctly visible in live animals. (Reproduced from The Journal of Cell Biology, 1999, 144:53 by copyright permission of the Rockefeller University Press.) See color insert.

208

USES OF GFP IN CAENORHABDITIS ELEGANS

E

Figure 10.1. (continued) (E) Nuclear spindles: Multiphoton image series of GFP::b-tubulin in a live wild-type embryo from metaphase through telophase reveals centrosome dynamics. Elapsed time from the first frame is shown in minutes:seconds. (Reprinted from Molecular Biology of the Cell, 2001, 12:1751–1764) with permission by the American Society for Cell Biology.) See color insert.

localization of a putative chloride channel could be deduced from the expression and localization of GFP reporter constructs (Berry et al., 2003).

10.2.4

Use of GFP to Visualize Cellular Anatomy

Tissue, organ, and cell anatomy have traditionally been analyzed using either Nomarskibased light microscopy techniques, which are aided by the transparency of worms, or electron microscopy (EM). GFP has provided new opportunities to visualize aspects of anatomy that either cannot be visualized by Nomarski optics or are too tedious to analyze by EM (Fig. 10.2). Examples include gonadal sheath cells whose complex and elaborate structures were delineated with a GFP marker (Hall et al., 1999), growth cone migration and remodeling (Fig. 10.2A,B) (Knobel et al., 1999), filopodia migration of epithelial cells (Raich et al., 1999), distal tip cell morphology (Fitzgerald and Greenwald, 1995), the visualization of age-dependent changes in cell morphology (Herndon et al., 2002), or the visualization of individual steps of pharynx development (Portereiko and Mango, 2001). The visualization of the dynamics of some of these events through the use of live imaging has been particularly advantageous [e.g., growth cone migration (Fig. 10.2A,B; Knobel et al., 1999) or epithelial cell fusion (Mohler et al., 1998; Raich et al., 1999)]. These visualizations make it possible through the use of classic forward genetics to now identify genes that are involved in these processes.

209

USES OF GFP

A

B

C

Figure 10.2. Examples of axon anatomy visualized with GFP. (A,B) Growth cone: Confocal micrograph of live L2 larvae at 17 hours post-hatching showing GFP driven under the unc-47 promoter that expresses in the ventral nerve cord, the DD and VD cell bodies (open arrows), the DD commissures (arrowheads), and the VD growth cones (solid arrows). (B) High magnification confocal micrograph of VD growth cones shown above (arrows). Filopodia extend from the round central mass of the growth cone on left. The middle growth cone is anvil-shaped. The right growth cone is extending a single finger toward the dorsal nerve cord. Existing embryonic DD commissures are marked with arrowheads. [Reprinted from Knobel et al. (1999), Development 126:4489–4498 with permission from The Company of Biologists Ltd.] (C) Axon co-labeling: Double labeling of axons in the ventral nerve cord (schematic). 3D image stacks of the ventral cord of double-labeled animals were recorded with a confocal microscope and subjected to a deconvolution algorithm to improve spatial resolution. Image shows an interneuron labeled with CFP (glr-1::GFP) and motorneuron axons labeled with GFP (unc-4::GFP). The image on the right is a cross-section through the ventral nerve cord at the position marked by the arrowhead in the left image. The orientation of images is depicted in the schematic. (Reprinted by permission of Wiley-Liss, Inc., a subsidiary of John Wiley and Sons, Inc. from Hutter, H., New ways to look at axons in Caenorhabditis elegans, Microscopy Research and Technique, copyright © 2000.) See color insert.

210

USES OF GFP IN CAENORHABDITIS ELEGANS

D

E * * * *

*

Figure 10.2. (continued) (D) PVT axon morphology: The pioneer neuron PVT has a large cell body (large arrow), situated in the pre-anal ganglion that sends out an anteriorly-directed process in the ventral nerve cord (small arrows). Original EM reconstructions suggested that the axon of PVT terminated in the posterior body. A zig-2::GFP reporter shows strong expression in PVT. This analysis shows that the PVT axon in fact extends the entire length of the nerve cord and terminates within the nerve ring (arrowhead). Asterisk denotes gut autofluorescence. (E) PVD axon morphology: Previously, the processes of the PVD interneuron were not completely reconstructed by EM. Analysis of a GFP reporter for PVD shows that the axon displays an elaborate branching pattern not previously appreciated (arrow indicates position of cell body, and asterisk denotes gut autofluorescence). See color insert.

Other subcellular structures that could previously only be observed through antibody staining can also now be easily assessed in live, nonfixed animals [for example, dense bodies of muscle cells (Fig. 10.1D; Hobert et al., 1999) or the mitotic spindle (Fig. 10.1E; Strome et al., 2001)]. Double labeling with GFP and the GFP variant CFP (cyan fluorescent protein), in combination with the use of deconvolution approaches, has also served well to visualize axon anatomy to a high resolution (Fig. 10.2C; Hutter, 2000). GFP markers have revealed aspects of neuroanatomy not precisely defined by EM (Fig. 10.2C,D). An anatomical database of C. elegans that is currently being created makes extensive use of a plethora of GFP markers (http://www.wormatlas.org). Lastly, intracellular organelles, such as mitochondria or the endoplasmic reticulum, are now also easily amenable for visualization using specific GFP markers (Labrousse et al., 1999; Rolls et al., 2002). One of the key advantages of using GFP in a genetically amenable organism such as C. elegans is that the powerful visualization of individual structures can be combined with genetic approaches to identify mutants in which these anatomical features are disrupted. Initial screens for axon pathfinding or cell migration mutants in C. elegans relied on the rather tedious need to either antibody-stain or bGal-stain populations of mutated animals, to fill animals with fluorescent dye, or to visualize cells through Nomarski optics (e.g., Colavita and Culotti, 1998; Forrester and Garriga, 1997; Hedgecock et al., 1985). With the advent of GFP technology, these types of screens have now become much easier (e.g., Zallen et al., 1999; Kim and Wadsworth, 2000) and have been extended to even visualize rather subtle aspects of axon outgrowth, such as axon branching (Fig. 10.3A; Bülow

211

USES OF GFP

A B Figure 10.3. Examples of the use of GFP to screen for anatomical mutants. (A) Axon morphology: Expression of the C. elegans homolog of the Kallmann syndrome gene kal-1 in the interneuron AIY causes a highly penetrant axon-branching phenotype (Bülow et al., 2002). The AIY interneuron can be visualized in wild type (left panel) and the overexpressing transgenic animals (middle panel) using a cell-specific GFP marker (ttx-3::GFP). Arrows indicate the cell bodies and arrowheads mark ectopic branches in AIY. This GFP phenotype was used as the basis for a suppressor screen that was successful in isolating suppressors of the AIY branching phenotype (right panel) (Bülow et al., 2002). (B) Presynaptic sites: Transgenic worms expressing the GFP-tagged synaptic vesicle marker synaptobrevin (SNB-1::GFP) were used in forward genetic screens to identify mutants that disrupt the structure of presynaptic termini. Shown on the left is a wild-type transgenic animal expressing SNB::GFP in the GABAergic neurons of the ventral cord. GFP localizes to discrete puncta along the ventral nerve cord (arrows), which may represent individual presynaptic termini. In rpm-1 mutants isolated in the screen (right panel), GFP puncta are irregular in size, are larger than those in wild type (arrow), and are spaced unevenly, leaving gaps without GFP (arrowheads). [Courtesy of Zhen et al. (2000).]

et al., 2002). Even more subtle anatomical features than axon outgrowth and cell migration have been examined with a marker, snb-1::gfp, which visualizes presynaptic specializations (Fig. 10.1A; Nonet, 1999) as well as a marker, glr-1::gfp, that detects postsynaptic neurotransmitter receptor clustering (Rongo et al., 1998). Candidate mutant testing and forward genetic screens using these markers have uncovered molecular mechanisms that underlie the construction of pre- and postsynaptic specializations (e.g., Rongo and Kaplan, 1999; Rongo et al., 1998; Schaefer et al., 2000; Zhen et al., 2000; Zhen and Jin, 1999) (Fig. 10.3B). Neuronal degeneration induced by expression of Huntingtin-type polyQ proteins has also been visualized by GFP and used for genetic screens aimed at suppressing this phenotype (Faber et al., 1999, 2002). It is important to consider that reporter gene arrays can have detrimental effects on the anatomy of a cell. We and others have observed several instances of this phenomenon in the nervous system where a subset of several independent transgenic lines, each containing the same reporter gene, shows defects in neuroanatomy (Toms et al., 2001; O. Hobert, unpublished observations). Furthermore, even if GFP-labeled cellular anatomy appears normal in all available transgenic lines in a wildtype background, some transgenic lines may be “sensitized”—that is, more prone to show anatomical defects in specific mutant backgrounds (Forrester and Garriga, 1997). Since DNA injected into worms is assembled into extrachromosomal arrays containing multiple copies of the injected DNA and since the copy number of the injected DNA on this array may vary from line to line, these effects are likely due to excessive levels of injected DNA. Having apparently too

212

USES OF GFP IN CAENORHABDITIS ELEGANS

much DNA on the array can cause transcription factor titration effects. That is, the presence of multiple transcription factor binding sites in the multicopy reporter gene array can sequester a relevant transcription factor, preventing it from acting on its normal target genes and hence causing the anatomical defects. This scenario has been explicitly observed in one example in the nervous system (Toms et al., 2001).

10.2.5

Use of GFP as a Cell Fate Marker

Individual cell types in any multicellular organism are defined by the expression of certain “gene batteries”—that is, the combinatorial expression of a defined set of genes (Davidson et al., 2002). GFP-based cell fate markers have become invaluable tools to assess the development of individual cell types in specific mutant backgrounds. Such cell fate markers allow one to precisely determine at what step a specific gene may act to define the development of a given cell type. To give one example, with the use of a battery of six different GFP-based markers that monitor the fate and development of a single interneuron class, disruption of the ttx-3 transcription factor was shown to affect all subtype specific properties of this neuron class, but not its pan-neuronal properties, defined by 3 pan-neuronally expressed gfp reporter gene constructs. A transcription factor located more downstream of ttx-3, the ceh-23 homeobox gene, was found to affect only a subset of differentiation characteristics of this neuron, namely the maintenance of expression of a single orphan transmembrane receptor (Altun-Gultekin et al., 2001). These sorts of approaches have been widely used in other cell types as well (e.g., Tsalik et al., 2003; Eastman et al., 1999; Esmaeili et al., 2002; Hallam et al., 2000; Sarafi-Reinach et al., 2001; Sarafi-Reinach and Sengupta, 2000; Satterlee et al., 2001; Winnier et al., 1999).

10.2.6

GFP to Measure Neuronal Function and Plasticity

GFP serves to visualize not only the anatomical features of a cell but also their functional properties. The small neurons of C. elegans and its hydrostatic cuticle have made electrophysiological studies of neuron function extremely difficult. More recently, electrophysiological and optical indicators based on GFP have been utilized successfully in C. elegans (Kerr et al., 2000; Khatchatouriants et al., 2000; Samuel et al., 2003). The calcium indicator protein cameleon (Miyawaki et al., 1997) has been used to image calcium transients in both neurons and pharyngeal muscle (Kerr et al., 2000). The technique (described in detail in Chapter 6 of this volume) relies on calmodulin-dependent FRET between the GFP variants CFP and YFP, expressed under cell-specific promoters. An increase in cellular calcium leads to an increase in the observable YFP/CFP fluorescence intensity ratio. A related approach employing pH-sensitive “synaptopHluorins” (Miesenbock et al., 1998) was also extended to C. elegans (Samuel et al., 2003). This reporter technique relies on the lumen of a synaptic vesicle being more acidic than the extracellular environment. Hence, GFP attached to the luminal surface of a synaptic vesicle through synaptobrevin will encounter an increase in pH upon synaptic vesicle fusion and exocytosis. By using a pH-sensitive variant of GFP (superecliptic pHluorin)(Miesenbock et al., 1998; Sankaranarayanan et al., 2000) fused to C. elegans synaptobrevin, the synaptic activity of a set of thermosensory neurons could be inferred under different experimental conditions and in different mutant backgrounds (Samuel et al., 2003). Lastly, neuronal activity can also be visualized using standard, transcriptional GFP reporters. For example, a variety of environmental, sensory inputs affect the transcription of a set of putative chemosensory receptor genes (Nolan et al., 2002; Peckol et al., 2001; Troemel et al., 1999).

USES OF GFP

The major appeals of all these techniques are twofold. First, they allow one to correlate patterns of behavior with patterns of neuronal activity in single cells and/or groups of cells, thus providing a basis for understanding animal behavior on a cellular level. These patterns of behavior may be complex and entail processes such as the experiencedependent modification of specific behavioral outputs. Second, these approaches allow one to assess the effect of manipulating gene function on the activity of a neuronal circuit, thus paving the way to understanding behavior on a molecular level.

10.2.7

GFP to Visualize Other Cellular and Physiological Processes

Besides visualizing physiological processes in the nervous system, one can use GFP to visualize a variety of other dynamic physiological and cellular processes. These include monomeric axonal transport (Orozco et al., 1999; Zhou et al., 2001), cell cycle (Hong et al., 1998), oxidant stress (Link and Johnson, 2002), and the unfolded protein response (Urano et al., 2002). In addition, uptake of secreted GFP fusion proteins has been used to study endocytosis and to uncover genes needed for its regulation (Fares and Greenwald, 2001; Grant and Hirsh, 1999). The study of chromosome dynamics in living animals has also largely benefited from the use of GFP. For example, the dynamics of chromosome segregation can be visualized with GFP-tagged histone and has identified proteins required for this process (Kaitna et al., 2000). GFP tagging of b- and g-tubulin furthermore enabled the visualization of mitotic spindle dynamics (Fig. 10.1E) (Strome et al., 2001). The use of GFP has gone beyond the mere visualization of the dynamics of specific cellular processes, such as chromosome segregation, but has now also included the visualization of the dynamics of GFP-tagged proteins themselves. FRAP ( fluorescence recovery after photobleaching) has been used to correlate protein recruitment to the centrosome during different stages of the cell cycle (Leidel and Gonczy, 2003), to examine the dynamics of clathrin-mediated endocytosis (Greener et al., 2001), to explore synaptic transmission (Samuel et al., 2003), and to monitor the trafficking of proteins through the ER (Rolls et al., 2002).

10.2.8

Use of GFP for Identifying and Isolating Cells

Laser Ablation. Traditionally, laser ablation of specific cells required their identification by Nomarski optics, an often difficult task, particularly for densely clustered neurons in the head ganglia (Bargmann and Avery, 1995). Ablation of GFP-expressing cells is much simpler because (a) GFP expression more reliably identifies a cell, whose position can be somewhat variable, and (b) loss of GFP confirms that a cell has been successfully killed. This GFP-based approach has, for example, been used to elucidate the function of several interneuron classes in the head ganglia of C. elegans (Tsalik and Hobert, 2003). Electrophysiological Recording. Relative cell positions in C. elegans are disrupted with the manipulations (cuticle incisions or filet preparations) needed to expose cells for electrophysiological recordings. GFP-labeled cells, however, can still be identified in these preparations (Fig. 10.4A), thus enabling the recording from chemosensory neurons (Goodman et al., 1998) and interneurons (Brockie et al., 2001). Determination of Expression Pattern by Co-labeling. The identity of the cells that express a given GFP reporter gene is traditionally inferred from its characteristic position and shape. This can be a difficult feat for a nontrained eye, specifically within the

213

214

USES OF GFP IN CAENORHABDITIS ELEGANS

A

B wild type

mec-3

Figure 10.4. Examples of the use of GFP as a tool to identify cells. (A) Identifying cells for electrophysiological recording: The gcy-5::GFP reporter was used to identify the chemosensory neuron ASER for in situ patch-clamp recording. (Top) Nomarski-DIC micrograph showing exposed neuronal cell bodies and a recording pipette sealed to ASER. Scale bar is 10 mm, anterior is left. (Bottom) Fluorescent micrograph of the same field above showing the GFP label in ASER, which allowed the unambiguous identification of the neuron. [Courtesy of Goodman et al. (1998).] (B) Analyzing of mutant cells in vitro: GFP-positive touch neurons from wild type (left) and mec-3 mutant animals (right) were enriched by fluorescence-activated cell sorting, cultured in vitro, and used to isolate RNA for DNA microarray analysis. This technique allowed the identification of mec-3-dependent genes and demonstrated that, unlike whole-worm RNA analysis, genes expressed in only a few cells can be identified systematically. [Reprinted from Zhang et al. (2002) by copyright permission of the Nature Publishing Group.] See color insert.

main head ganglia, which contain many densely packed neurons whose cell positions can sometimes be variable. This difficulty can be overcome by using animals in which defined cell types are labeled with a different fluorescent marker, such as RFP. We have, for example, used this approach extensively in a search for genes expressed in the AIY interneuron class. We have first generated transgenic animals that express RFP under control of an AIY-specific cis-regulatory element. We then examined the expression of a variety of genes, identified through genomic approaches, for their expression in AIY by tagging their promoter region with GFP and injecting the constructs into the RFP reporter-

USES OF GFP

containing background. Co-labeling of GFP and RFP allowed us to very quickly and reliably identify genes that are expressed in AIY (Wenick and Hobert, 2004). Culturing and Sorting Cells. After disruption of animals, GFP-labeled muscle or neurons were isolated and grown in culture (Christensen et al., 2002; Zhang et al., 2002) (Fig. 10.4B). By using cell-type specific GFP markers, an individual neuron type can thus be identified when grown in culture. This approach has already been used to analyze the effect of gene disruption on in vitro properties of the neuron; by isolating touch sensory neurons from mec-3 mutant animals, the effect of this gene on axon outgrowth could be assessed (Zhang et al., 2002). GFP-labeled cells were also successfully sorted using conventional fluorescence-activated cell sorting (FACS) techniques (Christensen et al., 2002; Zhang et al., 2002). A fruitful application of the approach has been the isolation of mRNAs from GFP labeled and sorted touch neurons, followed by an examination of gene expression profiles through microarray analysis (Zhang et al., 2002). GFP as a Protein Tag. GFP-tagged proteins can serve multiple purposes at the same time. They not only can serve as tools to visualize protein localization but also can be used as affinity tags for purification. For example, the GFP-tagged glutamate receptor GLR-1 was shown to localize to postsynaptic specializations and was then immunopurified with anti-GFP antibodies to detect ubiquitination of the GLR-1 protein (Burbea et al., 2002). The mutation of a set of lysine residues in the context of the GLR-1::GFP reporter then allowed the determination that these residues are required for appropriate localization of the reporter in vivo and for correct ubiquitination, as assessed by an analysis of GLR1::GFP anti-GFP immunoprecipitates. Anti-GFP immunoprecipitation of GFP tagged proteins have also been used to reveal the in vivo existence of complexes of dsRNA-binding proteins required for RNA interference (Tabara et al., 2002).

10.2.9

Other Uses of GFP as an Experimental Tool

Injection Marker. As evident from the examples mentioned above, several applications of GFP are completely independent of the nature of the sequences to which GFP has been fused and simply use the GFP fusion construct as a tool. Another very common application is the use of a GFP reporter as an injection marker. Two types of injection markers have been traditionally used. One is a marker DNA that codes for a mutated collagen, rol-6, which confers a dominant “roller” phenotype on injected worms (Mello et al., 1991). The problem with this approach is that the roller phenotype can interfere with a mutant phenotype under study and/or it leads to morphological distortions of the worm that hamper an anatomic analysis of the transgenic worms. The alternative injection marker approach has been to use DNA coding for specific wildtype genes (e.g., lin-15, unc-4, dpy20, or pha-1 gene) that were injected into the respective mutant background (Jin, 1999). Transgenic animals could hence be scored through the rescue of the respective mutant phenotype. The problem with this approach has been manifold: (a) The phenotype of the injection marker or the mutant background that is rescued through the injection marker can interfere with the studied mutant phenotype. (b) In order to compare isogenic backgrounds, the rescued mutant phenotype must be included when the transgenic array is transferred from one genetic background to another genetic background. (c) If the experiment consists of an attempt to rescue a mutant phenotype with a specific piece of DNA, one has to first construct a double mutant of the mutant gene of interest and the mutant used as transformation marker.

215

216

USES OF GFP IN CAENORHABDITIS ELEGANS

Many of these problems have been largely overcome by the use of GFP injection markers, which provide a noninterfering, easy-to-score, dominant phenotype—that is, the fluorescent labeling of a specific group of cells. We routinely use unc-122::gfp, which labels coelomocytes (Loria et al., 2004), elt-2::gfp, which labels the intestine (Hawkins and McGhee, 1995), ttx-3::gfp, which labels a single interneuron class (Hobert et al., 1997), and ceh-22::gfp, which labels the pharynx (Okkema and Fire, 1994). Each of these reporters gives a strong enough fluorescent signal so that animals can be picked under a standard dissecting microscope equipped with a fluorescent light source. More importantly, each of these injection markers labels distinct regions of the animal. This is of critical importance if the animals to be injected shall be scored for a gfp phenotype. For example, if one wants to score neuroanatomy in the ventral nerve cord with a gfp marker, one wants to avoid GFP injection markers that interfere with the visualization of the cord. The pharynx-specific ceh-22::gfp is a good choice in this case. In contrast, if one wants to score anatomy in the nerve ring, one wants to avoid an injection marker such as ceh22::gfp or ttx-3::gfp, but wants to use, for example, unc-122::gfp which labels cells located in the mid-body region. GFP injection markers are of no use, however, if one undertakes GFP-based expression pattern analysis since the fluorescence of the GFP injection marker may interfere with the analysis of the GFP-tagged experimental DNA. Mosaic Analysis. Mosaic analysis is used to determine the site of action of a gene. In C. elegans, mosaicism is produced by the somatic loss of a free duplication or an unstable extrachromosomal array carrying rescuing DNA for the mutant phenotype being studied and marker DNA that rescues or produces a score-able phenotype (Herman, 1995). Recent studies have made use of a ubiquitously expressed GFP marker, sur-5, which, due to the ease of assessing array loss, allows one to screen through a much greater number of animals in a search for the appropriate mosaic animal (Yochem et al., 1998). In addition, single-cell-specific GFP or RFP markers allow one to easily identify very specific array losses. Using this approach, it was for example possible to assess whether a gene acts in the left versus right bilateral analog of a sensory neuron pair to affect signaling between the two cells (Sagasti et al., 2001).

10.2.10

Use of GFP Variants

The use of modified versions of GFP has extended the utility in C. elegans by allowing the simultaneous observation of multiple fluorescent proteins (Miller et al., 1999; Hutter, 2000). Red-, cyan-, and yellow-shifted variants of GFP (DsRed, CFP, YFP), described in other chapters of this volume, are now routinely used in C. elegans and are available in the vectors made available from the Fire lab (http://www.ciwemb.edu/pages/firelab.html). As described earlier, one application of double labeling is to facilitate cell identification. Specific known cells can be labeled with one fluorescent reporter (reference cells) to assess the expression of a gene of interest with a second fluorescent marker. Moreover, CFP and YFP variants were used to label different groups of neurons and their axons in the same animal (Fig. 10.2C; Hutter, 2000). These double-labeling experiments allow the determination of spatial relationships of axons within axon bundles, an achievement previously only possible through tedious EM reconstructions. Even analyses of axons within the nerve ring are approachable with these techniques. With these duallabeling techniques at hand, it is now possible to analyze the genetic and cellular requirements for correct placement of axons into defined fascicles. A similar strategy of dual

TECHNICAL ASPECTS

labeling has been used to visualize sets of vulval cell types during development (Inoue et al., 2002). Dual labeling of proteins with GFP variants has also facilitated analysis of protein subcellular localization. For example, the mechanosensory proteins MEC-6 and MEC-4 were shown to colocalize in a punctate pattern in touch cells using YFP- and CFP-tagged proteins (Chelur et al., 2002). Labeling of presynaptic and postsynaptic proteins with different GFP variants also allows the visualization of the correct alignment of synaptic specializations. This approach has been used successfully to label glutamatergic synapses in a subset of ventral cord interneurons (Burbea et al., 2002). Dual labeling and colocalization studies with GFP variants have been extended to the analysis of subcellular compartments such as the endoplasmic reticulum (Rolls et al., 2002). It should be noted that while the spectral characteristics of DsRed make it an attractive fluorescent marker, even improved versions of DsRed, such as DsRed2 (Clontech, Inc.), show a rather long maturation time relative to the fast development of C. elegans. For example, the promoter of the gcy-7 gene is turned on in mid-embryogenesis as assessed with GFP, yet the corresponding dsRed2 construct only begins to yield red fluorescence in mid-larval stages (unpublished observations). To its advantage, however, DsRed2 does not show the toxicity associated with DsRed1.

10.3

TECHNICAL ASPECTS

Creation of GFP Reporter Gene Constructs. Andy Fire (Carnegie Institution) has constructed a widely used vector kit that contains a variety of different expression vectors (http://www.ciweb.edu/pages/firelab.html). Key structural elements of these vectors are a multiple cloning site, the GFP coding sequence which contains amino acid changes to enhance the spectral properties of GFP (see other chapters), and a heterologous, canonical 3¢UTR, that of the unc-54 gene (Fire et al., 1990). In addition, artificial introns have been engineered in the GFP coding sequences, which was found to have dramatic effects on the expression levels of GFP. Traditionally, a fragment of DNA— containing either the putative 5¢ regulatory region of a gene under study (= “transcriptional fusion”) or the whole genomic locus including all coding sequence (= “translational fusion”)—is subcloned into any of the standard GFP expression vectors. Several recent alternatives to the subcloning approach aided the production of GFP reporters (Fig. 10.5): 1. PCR Fusion (Hobert, 2002). Here, a genomic locus (e.g., containing the upstream regulatory region of a gene) is amplified by PCR. The 3¢ primer is engineered to contain an overlap to sequences from the GFP expression vector. In parallel, an amplicon of GFP is produced. These two amplicons are then joined in a fusionPCR reaction that relies on the short overlap between the two PCR fragments, thus allowing mutual priming (Fig. 10.5A). The key advantage of this approach is speed; dozens of fusion PCR products can be constructed within a single day and injected into worms without any further purification required. Other miscellanous subcloning problems are avoided. This technique is routinely used in our lab and in the worm community. In several instances tested, we have not observed differences in expression patterns of subcloned versus PCR-fused reporters.

217

218

USES OF GFP IN CAENORHABDITIS ELEGANS

A

C

A PCR

GFP

unc-54 3’UTR D

B PCR

A* unc-54 3’UTR fusion PCR

GFP

D*

Inject PCR mix

B

GFP URA3

PCR

URA3

GFP

Homologous recombination in yeast (selection for URA) URA3 Counterselection against URA

Recover DNA from yeast, inject

C

PCR fusion product (see panel A ) Homologous recombination in worms Inject PCR mix

Figure 10.5. Techniques for generating GFP fusion constructs. (A) PCR fusion approach (Hobert, 2002): Primer A* and D* are directly nested to primer A and D, respectively. The overhang in primer B is about 20 nucleotides. (B) Recombination cloning (Hawkins et al., 2003): The plasmid from which the PCR product derives contains two direct repeats of GFP separated by a selected and counter-selectable marker, URA3. The overhang in the primers required for in vivo recombination need to be at least 50 nt long and are homologous to the sites of desired gene insertion. (C) In vivo recombination (Tsalik et al., 2003): Variation of A and B. To apply this to generating reporter gene constructs, a fusion PCR product is first generated with the same technique as described in panel A. This fusion construct can be small because its sole purpose is to create a region of >500 bp that overlaps with the recombination substrate. This recombination substrate is a second PCR fragment amplified from worm genomic DNA. Both PCR products are co-injected. As a control that the resultant GFP expression is indeed a result of recombination, the second PCR fragment is also injected alone. The only limiting factor for the technique is the size of the genomic PCR product. In that regard, it is virtually impossible to beat the yeast homologous recombination technique, in which one can recombine GFP, at least in theory, into several hundred kilobase YAC clones. One pitfall to the in vivo recombination technique is that one has no control over whether the in vivo recombination has worked successfully, except in those cases where a selectable scheme is employed (i.e., rescue a mutant phenotype with a GFP-tagged locus that has been generated through in vivo recombination). Alternatively, one can visualize successful recombination if one knows that only the recombined DNA can produce a protein that is correctly localized to specific subcellular sites.

TECHNICAL ASPECTS

2. Recombination Cloning (Hawkins et al., 2003). While being more timeconsuming than the PCR fusion method, this technique—which relies on an elegant yeast recombination scheme (Fig. 10.5B)—has the invaluable advantage of being able to (a) handle virtually unlimited sizes of genomic DNA and (b) allow one to drop GFP anywhere within the genomic locus (the latter could in theory also be achieved in a triple-PCR fusion approach). 3. In Vivo Recombination Approach (Tsalik et al., 2003; Fig. 10.5C). This approach relies on the observation that overlapping pieces of DNA injected into C. elegans will undergo efficient homologous recombination, provided that the homologous region is >500 bp (Maryon et al., 1998; Mello et al., 1991). The approach uses a technical combination of the two above approaches and combines several of their strengths—that is, speed and fewer restrictions in terms of size (see legend to Fig. 10.5C). Creation of Transgenic Lines that Express GFP Fusion Proteins. DNA injected into the gonad of a worm will form extrachromosomal arrays that contain multiple copies of the injected DNA (Jin, 1999; Mello et al., 1991). The multicopy nature of the arrays has advantages in the form of potentially offering a higher level of sensitivity, but also disadvantages such as potential overexpression artifacts, gene silencing, or promoter-titration artifacts. Since extrachromosomal arrays created by DNA injection are not stably transmitted through mitosis (used as an advantage in mosaic analysis mentioned above), one often induces these arrays to become integrated into chromosomes through the use of highenergy irradiation. A problem of this integration approach is that its high mutagenicity rate necessitates thorough backcrossing following irradiation. Due to the repetitive structure of arrays and problems of gene silencing associated with it, mosaicism may even be observed with chromosomally integrated arrays (Hsieh et al., 1999). New developments in the field of array formation have alleviated some of these problems. The generation of “complex arrays” through the injection of the desired reporter gene, together with a complex mixture of heterologous DNA, allows expression of a transgene in the germline, which could not be observed with regular arrays (Kelly et al., 1997). In addition, reporters expressed from “complex arrays” appear less mosaic and more stably expressed. Moreover, DNA delivery to C. elegans via microparticle bombardment can also yield expression in the germline, likely due to spontaneous chromosomal integration of low-copy number arrays (Praitis et al., 2001). The low-copy number of the reporter gene delivered by particle bombardment has the added advantage of yielding expression levels of the reporter construct that are more likely to reflect endogenous gene expression levels, thus minimizing the potential problems arising from reporter gene overexpression mentioned in previous chapters. Visualization of GFP. The level of expression of GFP can in some cases be so low that GFP fluoresence cannot be observed. In some documented cases, the presence of GFP was nevertheless revealed through the use of antibody staining against GFP, thus illustrating the higher sensitivity of antibody-staining procedures (Levitan and Greenwald, 1998). Common Artifacts. Notable fluorescent artifacts are the autofluorescence of particles in the gut and the nucleoli of hypodermal cells. Another artifact commonly observed

219

220

USES OF GFP IN CAENORHABDITIS ELEGANS

is posterior gut fluorescence. Unlike the autofluorescence of the gut and hypodermal nucleoli, the source of this fluorescence clearly is GFP. The GFP signal may be a consequence of the canonical 3¢UTR derived from the unc-54 gene, present in the commonly used GFP vectors. It is curious to note that this posterior gut expression becomes more prevalent when less regulatory information is provided in the sequences fused 5¢ to the GFP coding sequences. Co-injected DNA (e.g., a gfp and rfp reporter) as well as extrachromosomal arrays can also show poorly understood interactions among themselves. For example, we observe that they can sometimes silence each other’s expression. In these cases, it is advisable to use independently created and chromosomally integrated arrays and to cross strains that carry these integrated arrays. Lastly, another inherent problem of GFP is the need for chromophore maturation, which makes its difficult to temporally correlate the onset of GFP fluorescence to the onset of gene expression as assessed by GFP reporters. Induction of GFP expression under a strong, heat-inducible promoter yields visible fluorescence in <30 minutes after heat shock; however, more weakly expressed promoters may cause a longer delay since enough GFP protein may have to accumulate to observe a sufficient fluorescent signal. In cases where this is an important issue, antibody staining against the GFP protein is a viable option.

10.4

CONCLUSION

GFP has now permeated so many different branches of research in C. elegans that it is hard to imagine life without it. Many of the applications described above will be even more widely used in the future. It is to be expected that GFP will be developed as a marker to visualize many more than the physiological processes mentioned above. For example, GFP reporters that monitor aging or the metabolic state of an animal are likely to be discovered, and the sensitivities of existing tools (e.g., those that measure synaptic activity) are likely to be enhanced. GFP will also be useful to systematically decode the cisregulatory information provided in nonprotein/RNA coding regions of the genome. In a more applied sense, it can also be envisioned to use transgenic gfp worms as biosensors (Candido and Jones, 1996) or tools for drug screening and discovery. There are few limits to the imagination of what GFP can be used for.

ACKNOWLEDGMENTS Special thanks to Andy Fire for his generous supply of vectors to the community over the years. Thanks to Frank Slack, Yishi Jin, Piali Sengupta, Susan Strome, Jim Powers, and Martin Chalfie for inputs that helped writing this chapter.

REFERENCES Ahringer, J. (1996). Posterior patterning by the Caenorhabditis elegans even-skipped homolog vab-7. Genes Dev. 10:1120–1130.

REFERENCES

Altun-Gultekin, Z., Andachi, Y., Tsalik, E. L., Pilgrim, D., Kohara, Y., and Hobert, O. (2001). A regulatory cascade of three homeobox genes, ceh-10, ttx-3 and ceh-23, controls cell fate specification of a defined interneuron class in C. elegans. Development 128:1951–1969. Aurelio, O., Hall, D. H., and Hobert, O. (2002). Immunoglobulin-domain proteins required for maintenance of ventral nerve cord organization. Science 295:686–690. Bargmann, C. I., and Avery, L. (1995). Laser Killing of Cells in Caenorhabditis elegans, Vol. 48, Epstein, H. F., and Shakes, D., Academic Press, San Diego, CA, pp. 225–250. Baum, P. D., and Garriga, G. (1997). Neuronal migrations and axon fasciculation are disrupted in ina-1 integrin mutants. Neuron 19:51–62. Berry, K. L., Bülow, H. E., Hall, D. H., and Hobert, O. (2003). A C. elegans CLIC-like protein required for intracellular tube formation and maintenance. Science 302:2134–2137. Brockie, P. J., Mellem, J. E., Hills, T., Madsen, D. M., and Maricq, A. V. (2001). The C. elegans glutamate receptor subunit NMR-1 is required for slow NMDA-activated currents that regulate reversal frequency during locomotion. Neuron 31:617–630. Bülow, H. E., Berry, K. L., Topper, L. H., Peles, E., and Hobert, O. (2002). Heparan sulfate proteoglycan-dependent induction of axon branching and axon misrouting by the Kallmann syndrome gene kal-1. Proc. Natl. Acad. Sci. USA 99:6346–6351. Burbea, M., Dreier, L., Dittman, J. S., Grunwald, M. E., and Kaplan, J. M. (2002). Ubiquitin and AP180 regulate the abundance of GLR-1 glutamate receptors at postsynaptic elements in C. elegans. Neuron 35:107–120. Candido, E. P., and Jones, D. (1996). Transgenic Caenorhabditis elegans strains as biosensors. Trends Biotechnol. 14:125–129. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263:802–805. Chelur, D. S., Ernstrom, G. G., Goodman, M. B., Yao, C. A., Chen, L., O’Hagan, R., and Chalfie, M. (2002). The mechanosensory protein MEC-6 is a subunit of the C. elegans touch-cell degenerin channel. Nature 420:669–673. Christensen, M., Estevez, A., Yin, X., Fox, R., Morrison, R., McDonnell, M., Gleason, C., Miller, D. M., 3rd, and Strange, K. (2002). A primary culture system for functional analysis of C. elegans neurons and muscle cells. Neuron 33:503–514. Colavita, A., and Culotti, J. G. (1998). Suppressors of ectopic UNC-5 growth cone steering identify eight genes involved in axon guidance in Caenorhabditis elegans. Dev. Biol. 194:72–85. Conradt, B., and Horvitz, H. R. (1999). The TRA-1A sex determination protein of C. elegans regulates sexually dimorphic cell deaths by repressing the egl-1 cell death activator gene. Cell 98:317–327. Cui, M., and Han, M. (2003). Cis regulatory requirements for vulval cell-specific expression of the Caenorhabditis elegans fibroblast growth factor gene egl-17. Dev. Biol. 257:104–116. Davidson, E. H., Rast, J. P., Oliveri, P., Ransick, A., Calestani, C., Yuh, C. H., Minokawa, T., Amore, G., Hinman, V., Arenas-Mena, C., et al. (2002). A genomic regulatory network for development. Science 295:1669–1678. Eastman, C., Horvitz, H. R., and Jin, Y. (1999). Coordinated transcriptional regulation of the unc25 glutamic acid decarboxylase and the unc-47 GABA vesicular transporter by the Caenorhabditis elegans UNC-30 homeodomain protein. J. Neurosci. 19:6225–6234. Esmaeili, B., Ross, J. M., Neades, C., Miller, D. M., 3rd, and Ahringer, J. (2002). The C. elegans even-skipped homologue, vab-7, specifies DB motoneurone identity and axon trajectory. Development 129:853–862. Faber, P. W., Alter, J. R., MacDonald, M. E., and Hart, A. C. (1999). Polyglutamine-mediated dysfunction and apoptotic death of a Caenorhabditis elegans sensory neuron. Proc. Natl. Acad. Sci. USA 96:179–184.

221

222

USES OF GFP IN CAENORHABDITIS ELEGANS

Faber, P. W., Voisine, C., King, D. C., Bates, E. A., and Hart, A. C. (2002). Glutamine/proline-rich PQE-1 proteins protect Caenorhabditis elegans neurons from huntingtin polyglutamine neurotoxicity. Proc. Natl. Acad. Sci. USA 99:17131–17136. Fares, H., and Greenwald, I. (2001). Genetic analysis of endocytosis in Caenorhabditis elegans: Coelomocyte uptake defective mutants. Genetics 159:133–145. Fire, A., Harrison, S. W., and Dixon, D. (1990). A modular set of lacZ fusion vectors for studying gene expression in Caenorhabditis elegans. Gene 93:189–198. Fire, A., Kelly, W. G., Hsu, M., Xu, S., Ahnn, J., Harfe, B. D., Kostas, S. A., and Hsieh, J. (1998). The uses of green fluorescent protein in Caenorhabditis elegans. In Green Fluorescent Protein. Properties, Applications and Protocols, Chalfie, M., and Kain, S., Eds., Wiley-Liss, New York, pp. 153–168. Fitzgerald, K., and Greenwald, I. (1995). Interchangeability of Caenorhabditis elegans DSL proteins and intrinsic signalling activity of their extracellular domains in vivo. Development 121: 4275–4282. Forrester, W. C., and Garriga, G. (1997). Genes necessary for C. elegans cell and growth cone migrations. Development 124:1831–1843. Fukushige, T., Hendzel, M. J., Bazett-Jones, D. P., and McGhee, J. D. (1999). Direct visualization of the elt-2 gut-specific GATA factor binding to a target promoter inside the living Caenorhabditis elegans embryo. Proc. Natl. Acad. Sci. USA 96:11883–11888. Gaudet, J., and Mango, S. E. (2002). Regulation of organogenesis by the Caenorhabditis elegans FoxA protein PHA-4. Science 295:821–825. Goodman, M. B., Hall, D. H., Avery, L., and Lockery, S. R. (1998). Active currents regulate sensitivity and dynamic range in C. elegans neurons. Neuron 20:763–772. Grant, B., and Hirsh, D. (1999). Receptor-mediated endocytosis in the Caenorhabditis elegans oocyte. Mol. Biol. Cell 10:4311–4326. Greener, T., Grant, B., Zhang, Y., Wu, X., Greene, L. E., Hirsh, D., and Eisenberg, E. (2001). Caenorhabditis elegans auxilin: a J-domain protein essential for clathrin-mediated endocytosis in vivo. Nat. Cell Biol. 3:215–219. Gupta, B. P., and Sternberg, P. W. (2002). Tissue-specific regulation of the LIM homeobox gene lin-11 during development of the Caenorhabditis elegans egg-laying system. Dev. Biol. 247: 102–115. Hall, D. H., Winfrey, V. P., Blaeuer, G., Hoffman, L. H., Furuta, T., Rose, K. L., Hobert, O., and Greenstein, D. (1999). Ultrastructural features of the adult hermaphrodite gonad of Caenorhabditis elegans: Relations between the germ line and soma. Dev. Biol. 212:101–123. Hallam, S., Singer, E., Waring, D., and Jin, Y. (2000). The C. elegans NeuroD homolog cnd-1 functions in multiple aspects of motor neuron fate specification. Development 127:4239–4252. Harfe, B. D., and Fire, A. (1998). Muscle and nerve-specific regulation of a novel NK-2 class homeodomain factor in Caenorhabditis elegans. Development 125:421–429. Hawkins, M. G., and McGhee, J. D. (1995). elt-2, a second GATA factor from the nematode Caenorhabditis elegans. J. Biol. Chem. 270:14666–14671. Hawkins, N. C., Garriga, G., and Beh, C. T. (2003). Creating precise GFP fusions in plasmids using yeast homologous recombination. Biotechniques 34:74–78, 80. Hedgecock, E. M., Culotti, J. G., Thomson, J. N., and Perkins, L. A. (1985). Axonal guidance mutants of Caenorhabditis elegans identified by filling sensory neurons with fluorescein dyes. Dev. Biol. 111:158–170. Herman, R. K. (1995). Mosaic analysis. In Caenorhabditis elegans—Modern Biological Analysis of an Organism, Vol. 48, Epstein, H. F., and Shakes, D., Academic Press, San Diego, CA, pp. 123–147. Herndon, L. A., Schmeissner, P. J., Dudaronek, J. M., Brown, P. A., Listner, K. M., Sakano, Y., Paupard, M. C., Hall, D. H., and Driscoll, M. (2002). Stochastic and genetic factors influence tissue-specific decline in ageing C. elegans. Nature 419:808–814.

REFERENCES

Hobert, O. (2002). PCR fusion-based approach to create reporter gene constructs for expression analysis in transgenic C. elegans. Biotechniques 32:728–730. Hobert, O., Mori, I., Yamashita, Y., Honda, H., Ohshima, Y., Liu, Y., and Ruvkun, G. (1997). Regulation of interneuron function in the C. elegans thermoregulatory pathway by the ttx-3 LIM homeobox gene. Neuron 19:345–357. Hobert, O., Moerman, D. G., Clark, K. A., Beckerle, M. C., and Ruvkun, G. (1999). A conserved LIM protein that affects muscular adherens junction integrity and mechanosensory function in Caenorhabditis elegans. J. Cell Biol. 144:45–57. Hong, Y., Roy, R., and Ambros, V. (1998). Developmental regulation of a cyclin-dependent kinase inhibitor controls postembryonic cell cycle progression in Caenorhabditis elegans. Development 125:3585–3597. Hsieh, J., Liu, J., Kostas, S. A., Chang, C., Sternberg, P. W., and Fire, A. (1999). The RING finger/Bbox factor TAM-1 and a retinoblastoma-like protein LIN-35 modulate context-dependent gene silencing in Caenorhabditis elegans. Genes Dev. 13:2958–2970. Hutter, H. (2000). New ways to look at axons in caenorhabditis elegans. Microsc. Res. Tech. 48:47–54. Inoue, T., Sherwood, D. R., Aspock, G., Butler, J. A., Gupta, B. P., Kirouac, M., Wang, M., Lee, P. Y., Kramer, J. M., Hope, I., et al. (2002). Gene expression markers for Caenorhabditis elegans vulval cells. Gene Exp. Patterns 2:235–241. Jansen, G., Thijssen, K. L., Werner, P., van der Horst, M., Hazendonk, E., and Plasterk, R. H. (1999). The complete family of genes encoding G proteins of Caenorhabditis elegans. Nat. Genet. 21:414–419. Jin, Y. (1999). Transformation. In C. elegans—A Practical Approach, Hope, I., Ed., Oxford University Press, Oxford, pp. 69–96. Kaitna, S., Mendoza, M., Jantsch-Plunger, V., and Glotzer, M. (2000). Incenp and an aurora-like kinase form a complex essential for chromosome segregation and efficient completion of cytokinesis. Curr. Biol. 10:1172–1181. Kelly, W. G., Xu, S., Montgomery, M. K., and Fire, A. (1997). Distinct requirements for somatic and germline expression of a generally expressed Caernorhabditis elegans gene. Genetics 146:227–238. Kerr, R., Lev-Ram, V., Baird, G., Vincent, P., Tsien, R. Y., and Schafer, W. R. (2000). Optical imaging of calcium transients in neurons and pharyngeal muscle of C. elegans. Neuron 26:583–594. Khatchatouriants, A., Lewis, A., Rothman, Z., Loew, L., and Treinin, M. (2000). GFP is a selective non-linear optical sensor of electrophysiological processes in Caenorhabditis elegans. Biophys. J. 79:2345–2352. Kim, S., and Wadsworth, W. G. (2000). Positioning of longitudinal nerves in C. elegans by nidogen. Science 288:150–154. Kirouac, M., and Sternberg, P. W. (2003). cis-Regulatory control of three cell fate-specific genes in vulval organogenesis of Caenorhabditis elegans and C. briggsae. Dev. Biol. 257:85–103. Knobel, K. M., Jorgensen, E. M., and Bastiani, M. J. (1999). Growth cones stall and collapse during axon outgrowth in Caenorhabditis elegans. Development 126:4489–4498. Labrousse, A. M., Zappaterra, M. D., Rube, D. A., and van der Bliek, A. M. (1999). C. elegans dynamin-related protein DRP-1 controls severing of the mitochondrial outer membrane. Mol. Cell 4:815–826. Leidel, S., and Gonczy, P. (2003). SAS-4 is essential for centrosome duplication in C. elegans and is recruited to daughter centrioles once per cell cycle. Dev. Cell 4:431–439. Levitan, D. and Greenwald, I. (1998). LIN-12 protein expression and localization during vulval development in C. elegans. Development 125:3101–3109. Link, C. D. and Johnson, C. J. (2002). Reporter transgenes for study of oxidant stress in Caenorhabditis elegans. Methods Enzymol. 353:497–505.

223

224

USES OF GFP IN CAENORHABDITIS ELEGANS

Loria, P. M., Duke, A., Rand, J. B., and Hobert, O. (2003). Two neuronal, nuclear-localized RNAbinding proteins involved in synaptic transmision. Curr. Biol. 13:1317–1323. Loria, P. M., Hodgkin, J., and Hobert, O. (2004). A conserved postsynaptic transmembrane protein affecting neuromuscular signaling in C. elegans. J. Neurosci. 24:2191–2201. Maryon, E. B., Saari, B., and Anderson, P. (1998). Muscle-specific functions of ryanodine receptor channels in Caenorhabditis elegans. J. Cell Sci. 111(Pt 19):2885–2895. Mello, C. C., Kramer, J. M., Stinchcomb, D., and Ambros, V. (1991). Efficient gene transfer in C. elegans: Extrachromosomal maintenance and integration of transforming sequences. EMBO J. 10:3959–3970. Miesenbock, G., De Angelis, D. A., and Rothman, J. E. (1998). Visualizing secretion and synaptic transmission with pH-sensitive green fluorescent proteins. Nature 394:192–195. Miller, D. M., 3rd, Desai, N. S., Hardin, D. C., Piston, D. W., Patterson, G. H., Fleenor, J., Xu, S., and Fire, A. (1999). Two-color GFP expression system for C. elegans. Biotechniques 26:914–918, 920–921. Miyabayashi, T., Palfreyman, M. T., Sluder, A. E., Slack, F., and Sengupta, P. (1999). Expression and function of members of a divergent nuclear receptor family in Caenorhabditis elegans. Dev. Biol. 215:314–331. Miyawaki, A., Llopis, J., Heim, R., McCaffery, J. M., Adams, J. A., Ikura, M., and Tsien, R. Y. (1997). Fluorescent indicators for Ca2+ based on green fluorescent proteins and calmodulin. Nature 388:882–887. Mohler, W. A., Simske, J. S., Williams-Masson, E. M., Hardin, J. D., and White, J. G. (1998). Dynamics and ultrastructure of developmental cell fusions in the Caenorhabditis elegans hypodermis. Curr. Biol. 8:1087–1090. Nathoo, A. N., Moeller, R. A., Westlund, B. A., and Hart, A. C. (2001). Identification of neuropeptide-like protein gene families in Caenorhabditiselegans and other species. Proc. Natl. Acad. Sci. USA 98:14000–14005. Nolan, K. M., Sarafi-Reinach, T. R., Horne, J. G., Saffer, A. M., and Sengupta, P. (2002). The DAF7 TGF-beta signaling pathway regulates chemosensory receptor gene expression in C. elegans. Genes Dev. 16:3061–3073. Nonet, M. L. (1999). Visualization of synaptic specializations in live C. elegans with synaptic vesicle protein-GFP fusions. J. Neurosci. Methods 89:33–40. O’Kane, C. J., and Gehring, W. J. (1987). Detection in situ of genomic regulatory elements in Drosophila. Proc. Natl. Acad. Sci. USA 84:9123–9127. Okkema, P. G., and Fire, A. (1994). The Caenorhabditis elegans NK-2 class homeoprotein CEH-22 is involved in combinatorial activation of gene expression in pharyngeal muscle. Development 120:2175–2186. Orozco, J. T., Wedaman, K. P., Signor, D., Brown, H., Rose, L., and Scholey, J. M. (1999). Movement of motor and cargo along cilia. Nature 398:674. Peckol, E. L., Troemel, E. R., and Bargmann, C. I. (2001). Sensory experience and sensory activity regulate chemosensory receptor gene expression in Caenorhabditis elegans. Proc. Natl. Acad. Sci. USA 98:11032–11038. Portereiko, M. F., and Mango, S. E. (2001). Early morphogenesis of the Caenorhabditis elegans pharynx. Dev. Biol. 233:482–494. Praitis, V., Casey, E., Collar, D., and Austin, J. (2001). Creation of low-copy integrated transgenic lines in Caenorhabditis elegans. Genetics 157:1217–1226. Raich, W. B., Agbunag, C., and Hardin, J. (1999). Rapid epithelial-sheet sealing in the Caenorhabditis elegans embryo requires cadherin-dependent filopodial priming. Curr. Biol. 9:1139– 1146. Rolls, M. M., Hall, D. H., Victor, M., Stelzer, E. H., and Rapoport, T. A. (2002). Targeting of rough endoplasmic reticulum membrane proteins and ribosomes in invertebrate neurons. Mol. Biol. Cell 13:1778–1791.

REFERENCES

Rongo, C., and Kaplan, J. M. (1999). CaMKII regulates the density of central glutamatergic synapses in vivo. Nature 402:195–199. Rongo, C., Whitfield, C. W., Rodal, A., Kim, S. K., and Kaplan, J. M. (1998). LIN-10 is a shared component of the polarized protein localization pathways in neurons and epithelia. Cell 94:751–759. Sagasti, A., Hisamoto, N., Hyodo, J., Tanaka-Hino, M., Matsumoto, K., and Bargmann, C. I. (2001). The CaMKII UNC-43 activates the MAPKKK NSY-1 to execute a lateral signaling decision required for asymmetric olfactory neuron fates. Cell 105:221–232. Salkoff, L., Butler, A., Fawcett, G., Kunkel, M., McArdle, C., Paz-y-Mino, G., Nonet, M., Walton, N., Wang, Z. W., Yuan, A., et al. (2001). Evolution tunes the excitability of individual neurons. Neuroscience 103:853–859. Samuel, A. D., Silva, R. A., and Murthy, V. N. (2003). Synaptic activity of the AFD neuron in Caenorhabditis elegans correlates with thermotactic memory. J. Neurosci. 23:373–376. Sankaranarayanan, S., De Angelis, D., Rothman, J. E., and Ryan, T. A. (2000). The use of pHluorins for optical measurements of presynaptic activity. Biophys. J. 79:2199–2208. Sarafi-Reinach, T. R., and Sengupta, P. (2000). The forkhead domain gene unc-130 generates chemosensory neuron diversity in C. elegans. Genes Dev. 14:2472–2485. Sarafi-Reinach, T. R., Melkman, T., Hobert, O., and Sengupta, P. (2001). The lin-11 LIM homeobox gene specifies olfactory and chemosensory neuron fates in C. elegans. Development 128:3269– 3281. Satterlee, J. S., Sasakura, H., Kuhara, A., Berkeley, M., Mori, I., and Sengupta, P. (2001). Specification of thermosensory neuron fate in C. elegans Requires ttx-1, a homolog of otd/Otx. Neuron 31:943–956. Schaefer, A. M., Hadwiger, G. D., and Nonet, M. L. (2000). rpm-1, a conserved neuronal gene that regulates targeting and synaptogenesis in C. elegans. Neuron 26:345–356. Sengupta, P., Chou, J. H., and Bargmann, C. I. (1996). odr-10 encodes a seven transmembrane domain olfactory receptor required for responses to the odorant diacetyl. Cell 84:899–909. Strome, S., Powers, J., Dunn, M., Reese, K., Malone, C. J., White, J., Seydoux, G., and Saxton, W. (2001). Spindle dynamics and the role of gamma-tubulin in early Caenorhabditis elegans embryos. Mol. Biol. Cell 12:1751–1764. Tabara, H., Yigit, E., Siomi, H., and Mello, C. C. (2002). The dsRNA binding protein RDE-4 interacts with RDE-1, DCR-1, and a DExH-box helicase to direct RNAi in C. elegans. Cell 109:861–871. Toms, N., Cooper, J., Patchen, B., and Aamodt, E. (2001). High copy arrays containing a sequence upstream of mec-3 alter cell migration and axonal morphology in C. elegans. BMC Dev. Biol. 1:2. Troemel, E. R., Chou, J. H., Dwyer, N. D., Colbert, H. A., and Bargmann, C. I. (1995). Divergent seven transmembrane receptors are candidate chemosensory receptors in C. elegans. Cell 83:207–218. Troemel, E. R., Sagasti, A., and Bargmann, C. I. (1999). Lateral signaling mediated by axon contact and calcium entry regulates asymmetric odorant receptor expression in C. elegans. Cell 99:387–398. Tsalik, E. L., and Hobert, O. (2003). Functional mapping of neurons that control locomotory behavior in Caenorhabditis elegans. J. Neurobiol. 56:178–197. Tsalik, E. L., Niacaris, T., Wenick, A. S., Pau, K., Avery, L., and Hobert, O. (2003). LIM homeobox gene-dependent expression of biogenic a mine receptors in restricted regions of the C. elegans nervous system. Dev. Biol. 263:81–102. Urano, F., Calfon, M., Yoneda, T., Yun, C., Kiraly, M., Clark, S. G., and Ron, D. (2002). A survival pathway for Caenorhabditis elegans with a blocked unfolded protein response. J. Cell Biol. 158:639–646.

225

226

USES OF GFP IN CAENORHABDITIS ELEGANS

Vogel, B. E., and Hedgecock, E. M. (2001). Hemicentin, a conserved extracellular member of the immunoglobulin superfamily, organizes epithelial and other cell attachments into oriented lineshaped junctions. Development 128:883–994. Wenick, A. S., and Hobert, O. (2004). Genome cis-regulatory architecture and trans-acting regulators of a single interneuron-specific gene battery in C. elegans. Dev. Cell 6:757–770. Winnier, A. R., Meir, J. Y., Ross, J. M., Tavernarakis, N., Driscoll, M., Ishihara, T., Katsura, I., and Miller, D. M., 3rd. (1999). UNC-4/UNC-37-dependent repression of motor neuron-specific genes controls synaptic choice in Caenorhabditis elegans. Genes Dev. 13:2774–2786. Yochem, J., Gu, T., and Han, M. (1998). A new marker for mosaic analysis in Caenorhabditis elegans indicates a fusion between hyp6 and hyp7, two major components of the hypodermis. Genetics 149:1323–1334. Zallen, J. A., Kirch, S. A., and Bargmann, C. I. (1999). Genes required for axon pathfinding and extension in the C. elegans nerve ring. Development 126:3679–3692. Zhang, Y., Chou, J. H., Bradley, J., Bargmann, C. I., and Zinn, K. (1997). The Caenorhabditis elegans seven-transmembrane protein ODR-10 functions as an odorant receptor in mammalian cells. Proc. Natl. Acad. Sci. USA 94:12162–12167. Zhang, Y., Ma, C., Delohery, T., Nasipak, B., Foat, B. C., Bounoutas, A., Bussemaker, H. J., Kim, S. K., and Chalfie, M. (2002). Identification of genes expressed in C. elegans touch receptor neurons. Nature 418:331–335. Zhao, H., and Nonet, M. L. (2001). A conserved mechanism of synaptogyrin localization. Mol. Biol. Cell 12:2275–2289. Zhen, M., Huang, X., Bamber, B., and Jin, Y. (2000). Regulation of presynaptic terminal organization by C. elegans RPM-1, a putative guanine nucleotide exchanger with a RING-H2 finger domain. Neuron 26:331–343. Zhen, M., and Jin, Y. (1999). The liprin protein SYD-2 regulates the differentiation of presynaptic termini in C. elegans. Nature 401:371–375. Zhou, H. M., Brust-Mascher, I., and Scholey, J. M. (2001). Direct visualization of the movement of the monomeric axonal transport motor UNC-104 along neuronal processes in living Caenorhabditis elegans. J. Neurosci. 21:3749–3755.

11 GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA Tulle Hazelrigg and Jennifer H. Mansfield* Department of Biological Sciences, Columbia University, New York, NY

11.1

INTRODUCTION

GFP from the jellyfish Aequorea victoria was first used as a reporter for gene expression in bacteria and C. elegans by Chalfie et al. (1994). This initial article established a fundamental property of GFP that was both surprising and exciting: It did not require factors unique to the jellyfish for its fluorescence, allowing its detection in living, unperturbed cells of transgenic organisms. The ability to express GFP in living cells of diverse species has revolutionized biological imaging. The first GFP fusion protein reported was with the Drosophila protein Exuperantia (Exu), a protein required for subcellular RNA localization in oocytes (Wang and Hazelrigg, 1994). This work established additional important properties of GFP: that both GFP and its fusion partner can retain their biological activity and localize like the fusion partner. Since then, GFP fusions have become a standard technique for studying protein trafficking and localization in many species and types of cells. In combination with the other well-developed genetic tools of Drosophila, GFP has been used as a reporter for gene expression, to study protein trafficking and localization in living cells, to mark cell structures and organelles, to mark clones in mosaic animals, and as a tool in genetic screens. Fundamental biological processes have been studied using GFP, including the morphogenesis of tissues, the control of the cell cycle, mitosis and meiosis, gametogenesis, innate immunity, neurogenesis, morphogen gradients, subcellular RNA localization, and the regulation of transcription and translation. Genetic screens * Present address: Department of Genetics, Harvard Medical School, Boston, MA. Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

227

228

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

using GFP have identified new mutations, genes expressed in selected tissues or patterns, and proteins localized to particular cellular domains. GFP has also facilitated the isolation of specific genotypes or types of cells from populations for applications such as microarrays. In the following we review GFP use in Drosophila and provide examples that illustrate its applications.

11.2

STRATEGIES FOR EXPRESSING GFP IN DROSOPHILA

The versatile P-element transformation system of Drosophila allows for the ready production of transgenic flies expressing GFP. In many cases a gene’s natural promoter has been used to drive either GFP expression alone or a GFP fusion protein. Inducible promoters have also been used, including heat shock promoters and promoters regulated by the Gal4-responsive upstream activator sequence (UAS), allowing for both temporal and tissue-specific control of expression. The power of the Gal4 system lies in its versatility: A single reporter construct or GFP fusion protein may be expressed in various spatial and temporal patterns, depending on the Gal4-expressing driver with which it is combined (Brand and Perrimon, 1993; reviewed in Duffy, 2002). The wide variety of Gal4 expression patterns generated by the Drosophila research community makes it possible to express GFP in nearly every tissue type, or in specific cells within a tissue, without having to isolate tissue-specific promoters. The utility of this approach for driving GFP expression was first demonstrated by Yeh et al. (1995), who showed that GFP could be expressed efficiently when driven by Gal4 expressed in a variety of tissues.

11.2.1

GFP-Based Transformation Vectors

GFP vectors for P-element-mediated transformation include vectors in which GFP acts as the screenable marker for selecting transformants, vectors in which GFP functions as a gene expression reporter for inserted enhancer and promoter elements, and vectors for expressing GFP fusion proteins. Horn et al. (2000) reported that GFP provides a more sensitive marker for transformation than the traditional mini-white gene. The authors performed a P-element insertion screen using a vector that contains both mini-white and GFP driven from three copies of an artificial promoter (3xP3-EGFP) expressed in the eye. In every case where white expression was detected in eyes and ommatidia, GFP could also be detected. An additional 20% of insertions were recovered by GFP fluorescence that lacked detectable white expression. Since the 3xP3 promoter is also expressed in embryonic, larval, and pupal tissues, transformants can also be identified at these earlier stages. In addition to the P-element vector, 3xP3-EGFP was also used as a marker for successful transformation with three additional transposable elements: the Tc1/mariner element Mos1, the TTAA element PiggyBac, and the hAT element Hermes. In addition to these GFP vectors, 3xP3-ECFP and 3xP3-EYFP vectors have been constructed with these transposable elements, creating a set of versatile vectors for Drosophila transformation and possible applications with nonmodel insects and other species (Horn and Wimmer, 2000). Barolo et al. (2000) made several reporter vectors for enhancer/promoter analysis with GFP or lacZ reporters. These vectors, known as the Pelican vectors, carry the mini-white gene for screening and have copies of the 0.4-kb gypsy insulator flanking GFP (or lacZ) to reduce position effects of flanking DNA on gene expression. These vectors have exten-

STRATEGIES FOR EXPRESSING GFP IN DROSOPHILA

sive multiple cloning sites (MCS), with 14–19 restriction sites, located upstream of the reporter. There are five separate Pelican GFP vectors that carry either (a) an enhanced GFP variant or (b) the same GFP with a nuclear localization signal (nls). These vectors are also distinguished by different minimal promoters and whether they contain UAS elements. Parker et al. (2001) constructed a vector to express fusion proteins with GFP at the N-terminus. In this vector, called pP {UAS-EGFP}, the GFP fusion is Gal4 inducible. A MCS, located at the 3¢ end of GFP, allows cloning of cDNA or genomic fragments of a gene of interest in frame with GFP. Nagel et al. (2002) developed several useful vectors that utilize GFP. pGreeni, pBluei, and pYelli are small (4.2 kb total) vectors with the screenable markers GFP, CFP, or YFP, driven by the 3xP3 regulatory element that allows detection in embryos, larvae, pupae, or adults. In addition to their increased sensitivity for screening, along with the ability to screen at earlier stages than mini-white, the use of GFP color variants allows transposition of multiple P-elements to be tracked simultaneously. pFlipG is another small vector, with FLP recombinase target (FRT) sites flanking both GFP and a polylinker cloning site. Since GFP and a second gene cloned into the polylinker are both excised simultaneously by the FLP recombinase, GFP can be used to identify FLP/FRT cell clones and also to screen for targeted gene knockouts (Rong and Golic, 2000). A final vector, pudsGFP, was designed for simultaneous Gal4 activation of an inserted gene of interest and GFP, both driven by separate UAS elements. In combination with another vector, pHIBS, pudsGFP can be used for constructing dsRNA vectors for Gal4-driven RNA interference (RNAi). Briefly, a sequence of interest is cloned into the pHIBS MCS, downstream of the 72-nucleotide Hairless intron I. By digesting the resulting clone with two different sets of enzymes, the sequence of interest may be cloned in both sense and antisense orientations into the pUdsGFP vector. In the final construct, inverted DNA segments are cloned so that they are interrupted by an intron, to produce a loopless hairpin dsRNA. Since both dsRNA and GFP are inducible by Gad4, cells that express dsRNA are marked with GFP.

11.2.2

Transient Expression

Bossing et al. (2002) combined the inducible Gal4/UAS system with DNA injections to show that protein localization in embryos could be assayed without generating stable lines of transgenic animals. The authors first showed that a UAS-GFP reporter could be injected into embryos and was expressed only in cells expressing Gal4. They then demonstrated that injection of plasmids containing UAS-regulated GFP fusion genes, including UASCFP-actin, UAS-GFP-tubulin, and UAS-engrailed-GFP, could be used to quickly assess subcellular localization of GFP fusion proteins.

11.2.3

Detection of GFP: Sensitivity and Timing Issues

Fusions of GFP to b-galactosidase have been used to compare the sensitivity of each reporter (Timmons et al., 1997). In this study, Gal4-inducible gene constructs expressing GFP fused to either the N- or C-terminus of b-galactosidase were crossed to a large set of Gal4 drivers to test expression in different tissue types at various stages of development. Detection of the two reporters was generally quite similar, but there were some differences: In the nervous system, GFP was a better reporter than b-galactosidase, whereas in ovaries and testes b-galactosidase was more reliable. GFP fluorescence was less frequently detected in eye imaginal discs than in the wing, haltere, or leg discs. In general, b-

229

230

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

galactosidase staining offered more sensitivity than GFP fluorescence, probably because of enzymatic amplification. Several mutant versions of GFP have been isolated that affect the excitation or emission properties of the protein (see Chapter 6 in this volume). As in other organisms, GFPS65T produces a brighter signal than wild-type GFP in Drosophila (Endow and Komma, 1996, 1997; Timmons et al., 1997). The variant EGFP has been used in Drosophila as a reporter of gene expression and to mark a variety of subcellular structures [see, for example, Morin et al. (2001), Halfon et al. (2002), Zhang et al. (2002), and Cox and Spradling (2003)]. The GFP color variants EYFP and ECFP have also been used successfully in Drosophila (Halfon et al., 2002; Rashkova et al., 2002; Van Roessel et al., 2002). [Throughout the remainder of this review, we generally use the terms GFP, YFP, and CFP to refer to these variants, without specifying the particular variant used in each case (e.g., GFPS65T versus EGFP); the reader is referred to the original articles for this information.] The concentration of GFP within a cell is a critical factor for its detection. As might be expected, increasing GFP copy number within a cell yields better signal detection. When GFP is expressed under control of Gal4, increasing the copy number of Gal4 in the genome can also enhance expression (Cox and Spradling, 2003; Noguchi and Miller, 2003). In addition, lower levels of GFP can be detected if it is localized to certain subcellular regions—for example, to the nucleus—since a high local concentration produces signal bright enough for detection. Most cellular autofluorescence occurs within the excitation and emission wavelengths of GFP and can obscure GFP signals (Aubin, 1979; Niswender et al., 1995). For instance, fluorescence from yolk in Drosophila oocytes and young embryos can make GFP detection difficult. This problem can be addressed by selecting appropriate excitation and/or emission filters, or by using variants of GFP that excite or emit at different wavelengths than those producing the problematic autofluorescence (e.g., Halfon et al., 2002). Targeting GFP to the nucleus, when appropriate, may also help since there is generally little autofluorescence in nuclei (Aubin, 1979). Several studies have addressed the timing with which GFP fluorescence appears. The acquisition of fluorescence, which requires chromophore maturation, varies with the type of GFP variant used and may also vary with cell type and cellular compartment, so conclusions obtained with any particular study are not completely generalizable. Davis et al. (1995) examined the acquisition of GFP fluorescence in Drosophila embryos expressing a nuclear-localized GFP (nlsGFP) driven from the polyubiquitin promoter. When maternally provided, GFP fluorescence could be detected at low levels quite early in embryogenesis. When the protein was expressed zygotically but not maternally, fluorescence appeared approximately 3–5 h after the protein was detectable by Western blot at comparable levels to maternally provided protein. Brand (1995) estimated a similar time lag of 3 hours for GFP signal detection. GFP fluorescence was estimated to develop more rapidly in conditions where the timing of transcription or translation was tightly regulated (Hazelrigg et al., 1998). In these experiments, two gfp-bicoid (bcd) constructs were designed to allow control over either the induction of transcription or the initiation of translation. When transcription was regulated by the hsp70 promoter, GFP-Bcd fluorescence in the nuclei of larval salivary glands was observed within 40 minutes following a 10-minute heat shock induction (Fig. 11.1). When bcd’s promoter drove expression of a g fp-bcd mRNA containing regulatory elements to direct initiation of translation after eggs are laid, GFP signal was first detected in syncytial stage embryos, within 1–2 h after the initiation of translation.

STRATEGIES FOR EXPRESSING GFP IN DROSOPHILA

Figure 11.1. Time requirements for GFP maturation in vivo. The time required for GFP signal acquisition was determined after induction of transcription in Drosophila salivary glands, using a hsp70-gfp-bcd transgene (From Hazelrigg et al., 1998). This construct contains the original gfp cDNA. a–d. Larvae were heat-shocked for 10 minutes and the salivary glands were examined for GFP fluorescence either immediately (a), or after 20 (b), 40 (c) or 60 (d) min. e. GFP-Bcd binds to many bands in the salivary gland polytene chromosomes. (From Hazelrigg et al., 1998.)

231

232

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

11.3 USING GFP FUSION PROTEINS TO STUDY CELL BIOLOGY AND DEVELOPMENT GFP has been used to tag many Drosophila proteins, providing a wealth of information about their localization and the cell structures with which they are associated. The single most important advantage of GFP-tagged proteins, as compared to immunohistochemistry, is that GFP fusion proteins can be detected in living cells. GFP fusions provide a dynamic view of protein localization, and the structures these proteins associate with, as these change with development or in response to experimental manipulation. GFP also circumvents the problem of background signal due to nonspecific antibody binding, and in some cases makes it possible, by crossing the transgene into a background that lacks an expressed endogenous gene, to tag every target protein molecule in the cell. Additionally, in cases where structures are labile to fixation, imaging with GFP fusion proteins in live cells may be the only way to visualize these structures.

11.3.1

Original Studies

The first protein to be tagged with GFP in any organism was Exuperantia (Exu), a Drosophila protein required for the localization of bicoid (bcd) mRNA in the oocyte (Wang and Hazelrigg, 1994; Fig. 11.2). This early study established several fundamental properties of GFP fusion proteins. First, GFP fusion proteins generally localize like their fusion partners, and the fusion proteins are usually functional. GFP-Exu, with GFP fused to either the N- or C-terminus of Exu, was able to rescue the function of an exu null allele, pro-

Figure 11.2. GFP-Exu in Drosophila egg chambers. Exuperantia (Exu) is required for RNA localization in the oocyte. GFP-Exu, the first reported GFP fusion protein, showed that GFP-tagged proteins are able to function like their fusion partners and are powerful tools for in vivo analysis of protein transport and localization. A. In early stages of oogenesis, GFP-Exu accumulates in the oocyte (arrows) of each egg chamber, and in cytoplasmic particles in the nurse cells. B. In older egg chambers GFP-Exu particles are prominent in all of the germ cells, and accumulate transiently at the anterior and posterior poles of the oocyte. These particles contain RNAs destined for the oocyte poles (From Wang and Hazelrigg, 1994).

USING GFP FUSION PROTEINS TO STUDY CELL BIOLOGY AND DEVELOPMENT

viding strong evidence that it was localized like endogenous Exu protein. Second, GFP fusion proteins provide unparalleled information about protein localization dynamics in living cells. GFP-Exu fluorescence revealed cytoplasmic particles that function as microtubule-associated RNA transport particles (Wang and Hazelrigg, 1994; Theurkauf and Hazelrigg, 1998; Cha et al., 2001). Analysis of these particles in live egg chambers, by time-lapse confocal analysis, established basic information such as (a) their rates and directionality of movement in the cytoplasm and (b) the microtubule-dependence of these movements. Finally, the GFP tag on Exu demonstrated the usefulness of GFP as a molecular handle for biochemical analysis, with the subsequent purification of an Exu RNP complex from ovary extracts (Wilhelm et al., 2000).

11.3.2

Morphogenesis of Tissues

Labeling cells with GFP has dramatically augmented the study of morphogenesis. In an early study, a fusion of GFP to the indirect flight-muscle-specific actin promoter, Act88F (Barthmaier and Fyrberg, 1995), was used to mark indirect flight muscle precursor cells and to follow muscle development. The GFP-marked muscle fibers could be visualized through the cuticle, in live pupae and in adults. Ward et al. (2003) used GFP to examine several processes during Drosophila metamorphosis, including imaginal disc eversion and the destruction of both the salivary glands and larval muscles. The movement and fusion of cells to form the tracheal system in embryos has been studied with a nuclear-targeted GFP-b-galactosidase fusion protein, GFPN-lacZ (Shiga et al., 1996). The processes examined in these studies have traditionally been difficult to study, depending on examination of histological preparations of tissues, which do not reveal the dynamics of morphogenetic processes. A fusion of GFP to the C-terminal region of Moesin (GFP-Moe), which targets GFP to the actin cytoskeleton (Edwards et al., 1997), has proven to be a valuable tool for studying morphogenetic changes in a wide variety of cell types. GFP-Moe has been used to study dorsal closure, development of the nervous system, formation and histolysis of muscles, cell movements and morphogenesis in oogenesis, and the function of the actin scaffold in bristle development (Edwards et al., 1997; Kiehart et al., 2000; Guild et al., 2002; Dutta et al., 2002; Hutson et al., 2003; Dorman et al., 2004). GFP fused directly to actin has also been used for studying the formation of actin bundles in migrating border cells, dorsal closure, and glial cell migrations (Verkhusha et al., 1999; Jacinto et al., 2000, 2002; Sepp and Auld, 2003). Marking cell membranes is another useful technique for studying morphogenetic movements and changes in cell shape. A fusion of GFP to the myristylation domain of GAP43 (Ritzenthaler et al., 2000), which targets the protein to cell membranes, has been used to observe the cell rearrangements that occur during embryogenesis to produce the stigmatophore, a structure that forms part of the posterior spiracles (Brown and CastelliGair Hombria, 2000). Similarly, a GFP-a-catenin fusion targets GFP to cell membranes, and it was used to study the cell movements that accompany wound healing and dorsal closure in Drosophila embryos (Wood et al., 2002). Edwards et al. (1997) were essentially able to pulse label a population of ovarian follicle cells using the actin-associated GFP-Moe fusion protein described above. By using the heat-inducible hsp70 promoter to drive expression in ovaries, the authors took advantage of the fact that the polar follicle cells (PFCs), specialized pairs of follicle cells at the anterior and posterior ends of developing egg chambers, cease proliferation early in egg chamber development. Following heat shock, these cells were strongly fluorescent, since

233

234

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

the neighboring cells went on to divide and dilute the GFP signal, leaving the undivided PFCs with high levels of GFP. The authors used this technique to show that the two anterior PFCs invariably become central cells in the group of follicle cells known as border cells and also form a specialized PFC projection that functions as a scaffold during formation of the micropyle, the site of sperm entry into the oocyte.

11.3.3

Organelle Function and Transport

GFP has been used to examine organelle organization and movements within Drosophila cells. Cox and Spradling (2003) expressed Gal4-inducable GFP fused to the human COX VIII mitochondrial targeting signal (Mito-GFP). In ovarian germ cells, Mito-GFP was localized to mitochondria and allowed the authors to track individual mitochondrial movements. The morphology of the endoplasmic reticulum during oogenesis and embryogenesis has been studied using a fusion of GFP to protein disulfide isomerase, PDI-GFP (Bobinnec et al., 2003). Observations of the changes in ER structure and localization during mitosis in early embryos led the authors to propose a novel role for ER in regulating mitosis (Fig. 11.3).

11.3.4

Chromosome Structure

The role of retrotransposons in forming telomeres has been studied using GFP-tagged proteins. In Drosophila, telomeres consist of arrays of two non-LTR retrotransposons, Het-A and TART. Analysis of tissue culture cells singly or co-transfected with GFP tagged Gag proteins from each retrotransposon revealed that Het-A Gag localizes to chromosome ends

Figure 11.3. GFP-tagged Protein Disulfide Isomerase (PDI) reveals dynamic rearrangements of the ER during mitosis. This time-lapse series shows a single cleavage cycle in a living syncytial stage embryo. Images were collected every 30 sec, and span a total time of 8 min, 30 sec. At prophase, ER material begins to accumulate at the spindle poles. By metaphase the ER is also highly concentrated around the edges of the spindle. Late in telophase, a distinct accumulation of ER material occurs at the midzone of the spindle. (From Bobinnec et al., 2003. Courtesy of Yves Bobinnec.)

USING GFP FUSION PROTEINS TO STUDY CELL BIOLOGY AND DEVELOPMENT

Figure 11.4. GFP-tagged Gag proteins of two non-LTR retrotransposons, HeT-A and TART, shown in interphase Drosophila tissue culture cells. When transfected singly, HeT-A GFP-Gag is targeted to telomeres, but TART GFP-Gag is not. A. HeT-A GFP-Gag. B. TART GFP-Gag. Left panels: GFPGag; middle panels—anti-HOAP, which labels the telomeres; right panels—superimposed images, with DAPI-stained chromosomes. When co-transfected, Het-A Gag recruits TART-Gag to telomeres (not shown here). (From Rashkova et al., 2002. Courtesy of Mary-Lou Pardue and Svetlana Rashkova.) See color insert.

independently and also recruits TART Gag to the ends of chromosomes (Rashhkova et al., 2002; Fig. 11.4). Furthermore, the regions of each protein required for both heterodimer formation and telomere localization were determined using a series of deletion constructs fused to GFP, CFP, and YFP (Rashkova et al., 2003). This analysis suggested an interesting symbiotic relationship between these two retrotransposons. Since Tart, but not Het A, encodes a reverse transcriptase required for its chromosomal insertion, it appears that each retrotransposon plays sequential roles in integration at telomeres. GFP fusions to evolutionarily diverged proteins from different species can be a powerful method to define localization domains. CID is a specialized H3-like histone that localizes to centromeres. Using GFP-tagged CIDs from different species expressed in D. melanogaster tissue culture cells, Vermaak et al. (2002) found that D. bipectinata CID did not associate with D. melanogaster centromeres. This allowed a domain swapping experiment with D. melanogaster and D. bipectinata CIDs to identify a domain that targets CID to centromeres.

11.3.5

Meiosis and Mitosis

Multiple spindle proteins, and in some cases mutant versions of these proteins, have been tagged with GFP to visualize mitosis and meiosis. These include a-tubulin, the meiotic centromere protein Mei-S322, the spindle-associated unconventional kinesin Ncd, cyclin B, the APC/C regulators Fizzy/Cdc20 and Fzy-related/Cdh1, the centrosome proteins Centrosomin, D-TACC, and the XMAP215/TOG family member Mini Spindles, as well as others (Kerrebrock et al., 1995; Endow and Komma, 1996; Huang and Raff, 1999; Grieder et al., 2000; Lee et al., 2001; Megraw et al., 2002; Raff et al., 2002). An early example, GFP-tagged Ncd (Fig. 11.5 and Fig. 11.6), demonstrates the sorts of information that can be obtained with these fusions (Endow and Komma, 1996, 1997). These experiments revealed details about the localization of Ncd on spindles that had not been detected by immunologic detection of the protein. By injecting embryos containing Ncd-GFP with rhodamine-labeled histones or rhodamine-labeled tubulin, Endow and Komma determined the distribution of Ncd relative to chromosomes and microtubules and observed spindle dynamics over time. GFP-tagged proteins that are associated with chromosomes throughout the cell cycle provide markers for chromosome behavior during mitosis and meiosis. Thus fusion of GFP to the histone variant His2Avd (GFP-His2AvD) has proven to be a very useful marker for chromosomes during both interphase and mitosis (Clarkson and Saint, 1999). Sites within

235

236

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

Figure 11.5. Ncd-GFP on metaphase spindles in a living Drosophila embryo. Ncd, a minus-end directed Kinesin, is present on spindle microtubules, and at the centrosomes. Ncd-GFP also labels fibers that extend from pole to pole of each metaphase spindle. These fibers were not detected by anti-Tubulin labeling. (From Endow and Komma, 1996. Courtesy of Sharyn Endow.)

individual chromosomes have been marked by a clever method that uses elements of the E. coli lac operon (Vazquez et al., 2001, 2002). In this system, a GFP-Lac repressor fusion is expressed in flies carrying chromosomal insertions of a 10-kb array of 256 copies of the lac operator site to which the Lac repressor binds. These marked chromosomes have been used to (a) study the nuclear localization of chromosomes during interphase in cultured spermatocytes and (b) follow chromosome pairing during male meiosis (Vazquez et al., 2001, 2002). Savoian and Rieder (2002) developed a method for short-term culture of larval neuroblasts to study mitosis, using cells derived from larvae expressing GFP-tagged proteins. These included GFP-Fzy to label kinetochores, GFP-Fzr to label centrosomes, His2ADVGFP to label chromosomes, and GFP-a-tubulin to label microtubules. The neuroblast division is asymmetric, giving rise to another neuroblast and a smaller ganglion mother cell. The GFP-tagged proteins provided beautiful and informative images of spindle formation and chromosome behavior in single cells as they progressed through the cell cycle. Asymmetric cytokinesis was shown to occur as the result of repositioning of the spindle at anaphase. Grieder et al. (2000) created a Gal4-inducible transgene for expressing a N-terminal fusion of GFP to tubulin (a-tubulin 84B). GFP-tubulin is an excellent marker for both cytoplasmic and spindle microtubules. This construct was made in a UAS P-element optimized for germline expression, allowing the authors to examine spindle and cytoplasmic microtubules in forming egg chambers (Fig. 11.7). These studies revealed a central role

USING GFP FUSION PROTEINS TO STUDY CELL BIOLOGY AND DEVELOPMENT

Figure 11.6. Time lapse analysis of Ncd-GFP localization during cleavage cycle 9 in a live embryo. Images, collected at 30 second intervals, are shown for: interphase (frames 1 and 2), nuclear envelope breakdown (frame 3), metaphase (frames 4–7), metaphase/early anaphase (frame 8), midanaphase (frame 9), late anaphase (frame 10), and telophase (frames 11 and 12). (From Endow and Komma, 1996. Courtesy of Sharyn Endow.)

237

238

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

Figure 11.7. GFP-a-Tubulin (Tub) in dividing germ cells in the ovary. During oogenesis, a germ stem cell gives rise to a daughter cystoblast, which subsequently undergoes 4 incomplete mitotic divisions to produce a cyst of 16 interconnected germ cells, one of which is destined to become an oocyte. The germ cells of developing cysts are connected by the spectrin-rich fusome, which plays an important role in orienting these cells and in oocyte specification. GFP-Tub was expressed in the germ cells by the Gal4-UAS system. In this mitotic cyst producing 8 germ cells, one end of each spindle is associated with the fusome (red). Later, after 16 cells are formed, the fusome is required to polarize the interphase microtubule network, an event that accompanies oocyte specification. (From Grieder et al., 2000. Courtesy of Allan Spradling.) See color insert.

played by the fusome, a structure that connects newly formed germ cells in a cyst, in orienting microtubules during germ cell divisions and in the establishment of egg chamber polarity.

11.3.6

Neuronal Function and Connectivity

In the nervous system, GFP has been used to visualize axonal projections and the synaptic connections made by neurons. Specific types of neurons have been labeled by either (a) driving GFP expression directly from the promoter of a gene expressed in that neuron or (b) the Gal4 system. In two early studies, GFP was expressed in the photoreceptor cells using a promoter responsive to the glass gene product (Plautz et al., 1996; Potter et al., 1996). In live adult flies, GFP could be detected in the eyes and ocelli. Axons extending from ocelli were filled with GFP, and the fluorescent signal could be detected through the adult cuticle. Recently, cholinergic neurons were identified and studied by inducing UASGFP expression with the promoter of the choline acetyltransferase (Cha) gene driving Gal4 expression (Salvaterra and Kitamoto, 2001; Fig. 11.8). When expressed in neurons in this manner, GFP fills the cell body and processes of the neuron, allowing visualization of the entire cell. Mushroom bodies (MB) play important roles in olfactory processing, learning, and memory. GAL-4-inducible GFP-synaptobrevin (Syb), which labels synaptic vesicles, has been used to visualize the synapses made by MB neurons (Ito et al., 1998). Using a Gal4 driver expressed specifically in the MB, the presynaptic sides of MB neurons were labeled with GFP-Syb, allowing mapping of the connections of these neurons to the brain.

USING GFP FUSION PROTEINS TO STUDY CELL BIOLOGY AND DEVELOPMENT

Figure 11.8. GFP-labeled cholinergic neurons. The reporter of the cholinergic locus (cha) was used to drive Gal4 and UAS-GFP expression in cholinergic neurons. GFP fills the entire neuron, allowing the cell bodies and their processes to be mapped. A. Wing margin with presumed chemosensory neurons. The arrows point to neurons adjacent to an internal wing vein. B. Presumed chemosensory neurons in the labial palps. C. Large neurons in the femur of the leg. D. Clusters of neurons in the distal tarsal segments (*) of the leg. E. Neurons in the second (*) and third segments of the antenna. The arrowhead indicates the base of the arista. (From Salvaterra and Kitamoto, 2001. Courtesy of Paul Salvaterra.)

GFP has also made it possible to (a) identify subsets of neurons expressing specific receptors and (b) map their projection patterns. Bhalerao et al. (2003) expressed UASGFP with Gal4 driven by the promoters for genes expressing five different odorant receptors. This allowed them to identify the olfactory neurons that express each receptor and to map their projections in the antennal lobes of the brain. Similarly, Hiroi et al. (2002) expressed GFP under the control of the promoters of six different gustatory receptors (Fig. 11.9). This allowed them to (a) identify these neurons in live adults and (b) make electrophysiological recordings from these cells to measure their responses to different sugars.

11.3.7

Morphogen Gradients

GFP can also reveal the localization of extracellular proteins, such as secreted molecules. This property has provided insights into how morphogen gradients are formed in populations of cells in developing tissues. The Decapentaplegic (Dpp) gradient has been examined in imaginal discs with biologically active GFP-Dpp fusions (Entchev et al., 2000; Teleman and Cohen, 2000; Gibson et al., 2002). GFP-Dpp forms an unstable extracellu-

239

240

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

Figure 11.9. GFP-labeled chemosensory neurons in the labellum. Promoters of 6 gustatory receptor (GR) genes were used to drive GFP expression by the Gal4-UAS system. A–F. GFPexpressing neurons driven by each Gr-Gal4. G–I. Nomarski images superimposed with fluorescent images, for 3 of the Gr-GFPs. The arrowheads point to sensilla associated with the GFP-expressing neurons. (From Hiroi et al., 2002. Courtesty of Teiichi Tanimura.)

lar gradient that could be detected up to 80 mm away from expressing cells in wing discs, a distance of multiple cell diameters. Analysis of GFP-Dpp in mutant backgrounds showed that the Dpp receptor, Dynamin, and Coronin are required for gradient formation (Entchev et al., 2000; Bharathi et al., 2004), suggesting that endocytosis is a critical step in gradient formation. Similarly, Wingless (Wg) gradient formation has been directly visualized in embryos with a GFP-Wg fusion (Pfeiffer et al., 2002). GFP-Wg revealed that much of Wg is retained on the surface of secreting cells, dependent on heparin sulfate proteoglycans, and that some GFP-Wg is recycled via endocytosis by expressing cells.

USING GFP FUSION PROTEINS TO STUDY CELL BIOLOGY AND DEVELOPMENT

11.3.8

Transcription Factors

GFP fusions have allowed analysis of the nuclear localization patterns of transcription factors in space and time, sometimes with unexpected results. For instance, the Polycomb (Pc) protein, a chromatin protein involved in gene silencing, was tagged with GFP, and its nuclear distribution was studied in different types of cells (Dietzel et al., 1999). PcGFP was found to localize to chromosomes in spermatocytes, as expected, but also to the nucleolus. In the germinal vesicle, Pc-GFP was associated with the DNA and also with the endobody. Jil-1 is a nuclear kinase that phosphorylates histone H3. A Jil-1-GFP fusion allowed live imaging of polytene chromosomes in the salivary glands, where Jil-1 was found to be increased on the single male X chromosome, consistent with it playing a role in dosage compensation (Jin et al., 1999; Fig. 11.10).

11.3.9

Tagged RNAs

Forrest and Gavis (2003) adapted a system developed in yeast for tagging mRNA molecules with GFP (Bertrand et al., 1998). To learn about the mechanism of nanos (nos) mRNA localization during oogenesis, the authors generated a transgene for expressing the nos mRNA with an insertion of six tandem copies of a stem loop that binds the bacterophage MS2 coat protein (MCP). A second transgene was made that expresses a fusion of MCP to GFP from the inducible hsp83 promoter. When expressed together, these transgenes allowed visualization of nos mRNA localization by GFP fluorescence (Fig. 11.11).

Figure 11.10. GFP-tagged Jil-1, a nuclear histone-H3 kinase. In salivary gland nuclei, Jil-1-GFP accumulates most strongly on the X chromosome in male cells, shown here, but not in female cells, suggesting that Jil-1 plays a role in dosage compensation. (From Jin et al., 1999. Courtesy of Kristen Johansen.)

241

242

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

Figure 11.11. GFP-labeled nos mRNA in living (A, D, E) and fixed (B, C, F) Drosophila egg chambers and embryos. (A) GFP-labeled nos RNA is visible in the oocyte of early egg chambers. Excess MCP-GFP fusion protein that is not bound to nos RNA enters the nurse cell nuclei. In these egg chambers, MCP-GFP fusion protein alone is also expressed in the follicle cells. (B) During midoogenesis, GFP-labeled nos RNA is transiently localized to the anterior margin of the oocyte. Nurse cell and follicle nuclei appear yellow/orange due to the overlap of Hoescht (red) and unbound MCP-GFP (green). (C) Z-series projection of the posterior of a stage 13 oocyte showing particles of GFP-labeled nos RNA at the cortex. (D) GFP-labeled nos RNA is also detected in particles at the posterior cortex of the early embryo. (E) GFP-labeled nos RNA in pole cells during gastrulation. (F) GFP-labeled nos RNA overlaps Vasa protein (detected by anti-Vasa immunostaining in red) in newly formed pole cells. (From Forrest and Gavis, 2003. Figure and legend courtesy of Elizabeth Gavis.) See color insert.

This elegant method allows RNA trafficking to be studied in vivo. Prior to this report, injection of fluorescently labeled RNA was used for this purpose. However, proper localization of RNA may depend on its nuclear history or a cytoplasmic transport pathway difficult to recapitulate by injection, and both cells and the RNA to be injected may be damaged during injection.

11.3.10

FRAP Analysis with GFP Fusion Proteins

Recent studies have demonstrated the power of FRAP (fluorescence recovery after photobleaching) to study protein turnover and changes in the architecture of structures labeled with GFP-tagged proteins. In these studies, strong laser light is used to photobleach GFP at a specific site within a cell. Due to the relatively long time required for a given molecule of GFP to reacquire fluorescence, rapid recovery indicates exchange with an

USING GFP FUSION PROTEINS TO STUDY CELL BIOLOGY AND DEVELOPMENT

Figure 11.12. FRAP analysis with GFP fusion proteins. Shown are living embryos expressing GFPTub (top panels), GFP-Fzr (middle panels), and DTACC-GFP (bottom panels). A region of the embryo was bleached with intense laser light at interphase of nuclear cycle 10–12. The recovery time after bleaching was determined, allowing an assessment of the relative rates of turnover of these proteins at the spindle. The arrows point to the centrosomes of adjacent spindles, one of which lies in the bleached section. Time points, from left to right, are 0:0 min, 0:30 min, 1:20 min and 3:00 min. (From Raff et al., 2002. Courtesy of Jordan Raff.)

unbleached pool of GFP. Spindle structures are readily amenable to FRAP analysis. The association of cell cycle and other proteins with microtubules and centrosomes has been examined by FRAP using GFP-tubulin, DTACC-GFP, and GFP-tagged forms of the APC/C regulators Fizzy/Cdc20 and Fzy-related/Cdh1 (Raff et al., 2002; Fig. 11.12) FRAP analysis of GFP-labeled Aurora-A kinase (AurA) was also used to study its association with centrosomes (Berdnik and Knoblich, 2002). Actin cycling in the progressing cones of individuating spermatids was studied by FRAP using a GFP-actin fusion protein (Noguchi and Miller, 2003). GFP-actin was photobleached in a line across the cone of an individuating spermatid, and fluorescence recovery within the photobleached area was measured over time to reveal the rates of actin turnover at various locations within the cone. The study of synaptic vesicles has also been enhanced by FRAP. In one study, synaptic vesicles were visualized using a synaptotagmin (syt)-GFP fusion protein, and FRAP analysis was used to determine the in vivo dynamics of vesicles at individual synaptic boutons (Zhang et al., 2002).

243

244

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

11.4

USING GFP REPORTERS TO STUDY GENE REGULATION

GFP has been used to define DNA enhancer elements that regulate transcription, as well as regulatory elements within RNAs that regulate translation or RNA stability. In this application, GFP is advantageous over traditional reporters because no processing is required to visualize the signal and because the dynamic expression pattern of reporters can more easily be monitored. Work by Sano et al. (2002) illustrates the use of GFP reporters for defining tissue-specific regulatory elements associated with promoters. They used a GFP reporter gene to define a 40-bp cis-acting DNA regulatory element that is both necessary and sufficient for germline expression of the vasa gene. Ahmed and Henikoff (2001) used a GFP reporter to study transcriptional silencing imposed by heterochromatin in position effect variegation (PEV). In their study, a Pelement that carried adjacent mini-white (w+) and UAS-GFP genes was inserted in heterochromatin, where PEV resulted in mosaic expression of both w+ and GFP in the eye. Activation of the UAS-GFP gene at different times in development, using various Gal4 drivers, revealed that early promoter activation protected the UAS-GFP gene from subsequent gene silencing by PEV. In the adult eye, the ability to detect both GFP and w+ allowed them to determine if Gal4-activation of UAS-GFP also affected expression of the adjacent w+ gene. Although some uncoupling of w+ and GFP expression was observed, their evidence suggested that Gal4 bound at the UAS-GFP promoter also protected the nearby white gene from silencing by heterochromatin (Fig. 11.13).

Figure 11.13. GFP as a reporter for transcriptional silencing. Reporter constructs were designed to measure the effects of heterochromatic gene silencing on two adjacent genes, mini-white and UASGFP. With this system, de-repressive effects of Gal4-induction of UASGFP in the context of flanking heterochromatin could be determined. Shown are eyes of flies carrying two different reporter insertions, x21 (A) and x18.4.1 (B), and two different Gal4 drivers (A5CGAL in A, and GMRGAL in B). The left panels are brightfield images of the eyes, and the middle and right panels are fluorescent images showing the red pigments (middle) or both the red pigments and GFP (right). In some cells, uncoupling of UASGFP and mini-white expression occurs: the white lines indicate areas where GFP is expressed, but mini-white is silenced. (From Ahmad and Henikoff, 2001. Coutesy of Steve Henikoff and Kami Ahmad.) See color insert.

USING GFP REPORTERS TO STUDY GENE REGULATION

Protection from microorganisms is governed by Drosophila’s innate immunity system, which utilizes a set of genes encoding small anti-microbial peptides. GFP has been used as a reporter to study the expression profiles of these genes and how they are induced by microbial challenge. These studies have shown that flies have two independently regulated innate immunity systems, namely, the systemic and the local epithelial responses. Tzou et al. (2000) examined GFP reporters driven by five antimicrobial gene promoters, and they found gene-specific patterns of expression in different epithelial tissues. Levashina et al. (1998) used GFP as a reporter to define a 1.5-kb infection-inducible promoter fragment of the metchnikowin (metch) gene, active in the fat body and circulating hemocytes of larvae. Using a drosomycin (dros)-GFP reporter, Ferrandon et al. (1998; Fig. 11.14) found that in addition to fat body expression induced in response to microbial challenge, dros is also expressed in a number of other tissues in larvae and adults, including salivary glands, trachea, labial glands in the proboscis, specific sets of epithelial cells, and sperm storage organs in the female reproductive tract. GFP has been used as a reporter for the expression of the micro-RNA (miRNA) bantam (Brennecke et al., 2003). miRNAs are small noncoding RNAs that bind to complementary sequences in target RNAs and downregulate expression. Brennecke et al.

Figure 11.14. A GFP reporter to study innate immunity. The drosomycin (dros) promoter was used to drive expression of GFP in transgenic flies. A. dros-GFP is induced strongly in flies that are immunized (challenged by microbial infection), but only at low levels in unchallenged flies (compare the top and bottom flies). B. dros-GFP is expressed in the fat body of immunized larvae (top) but not in control larvae (bottom). C. Dissected fat body of an immunized adult. Higher magnification image of the immunized larva shown in B. (From Ferrandon et al., 1998. Courtesy of Dominique Ferrandon.) See color insert.

245

246

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

generated a sensor construct with GFP expressed from a ubiquitously expressed tubulin promoter, with sequences complementary to bantam in its 3¢ UTR. These sequences conferred bantam regulation on the GFP mRNA, allowing bantam’s dynamic expression pattern during development to be visualized by the loss of GFP signal. One advantage of using GFP for this application is that, unlike antibodies or enzymatic reporters in which an amplification step is involved in detection, the relative levels of GFP in different regions of a tissue are more accurate readouts of quantitative differences between cells. They identified the pro-apototic gene hid as a target of bantam by showing that its 3¢ UTR, which contains bantam complementary sequences, confers bantam regulation on a GFP-hid 3¢ UTR reporter. Van Roessel et al. (2002) used the absence of one fluorescent signal in cells expressing two different color fluorescent proteins to show that RNAi is cell-autonomous in at least some cells in Drosophila embryos (Figure 11.15). They expressed GFP in all cells of the embryo using a polyubiquitin promoter. The embryos also carried a Gal4-regulated transgene expressing an inverted repeat (hairpin) sequence for GFP RNA. Expression of this double-stranded RNA by engrailed-Gal4 induction led to degradation of the GFP

Figure 11.15. Two-color GFP shows that RNAi is cell-autonomous in Drosophila embryos. In a, all segments of the embryo express a GFP fusion protein (green) from a transgene driven by the poly-ubiquitin promoter. The posterior domain of each segment (two segments are shown) also expresses a CFP fusion protein (red) from a UAS-regulated transgene in combination with enGAL4. The overlap of GFP and CFP appears yellow. In b, both transgenes are expressed in addition to an en-Gal4-induced ds RNA that targets the RNA encoding GFP. Expression of the GFP fusion is specifically repressed in the en domain, indicating that RNAi is cell autonomous and cannot spread to the anterior compartments. Panels c and d show the GFP channel only for the embryos shown in a and b, respectively. (From van Roessel et al., 2003. Courtesy of Peter van Roessel and Andrea Brand.) See color insert.

GFP AS A GENETIC TOOL

message in engrailed stripes and therefore loss of the GFP signal in these cells. A UASCFP transgene (the divergent CFP sequence is not targeted by the dsRNA) was simultaneously expressed in these embryos, allowing CFP expression to mark cells in which the dsRNA construct was coexpressed. RNAi was shown to be cell-autonomous, since GFP expression was lost only in cells expressing CFP.

11.5

GFP AS A GENETIC TOOL

Many standard tools for Drosophila genetics have been updated and improved by the addition of GFP, including GFP balancers and chromosomes designed for clonal analysis with GFP markers. GFP has also proven invaluable as a screening tool in a variety of contexts.

11.5.1

Balancers

GFP balancers are particularly useful for the study of embryos, where traditional markers are scarce. These balancers have been used to purify large populations of homozygous mutant embryos in a fluorescence-activated embryo sorter on a scale large enough to extract mRNA for transcriptional profiling experiments using cDNA microarrays (Furlong et al., 2001a,b). They have also allowed smaller-scale hand sorting of homozygous embryos, facilitating both (a) phenotypic analysis of embryonic lethal mutations and (b) molecular analysis. Casso et al. (2000) modified the FM7c, CyO, and TM3 balancer chromosomes by the addition of P-elements carrying UAS-GFP and Kruppel-Gal4. GFP expression in embryos bearing any of these balancers commences during gastrulation and persists, in the endogenous Kruppel pattern, throughout development and into adulthood (Fig. 11.16). Live animals homozygous for a mutation of interest can be identified (by their lack of GFP signal) and sorted from their siblings as early as stage 9 of embryogenesis. A separate set of CyO and TM3 balancers was constructed that allows either GFP or YFP visualization beginning at embryonic stage 8 and continuing though subsequent stages of development (Halfon et al., 2002). These balancers utilize a twist-Gal4 driver to express bicistronic GFP or YFP transcripts, which are brighter than a single copy of GFP.

11.5.2

Clonal Analysis

GFP has proven a valuable marker for clonal analysis. Because GFP’s signal is dosesensitive, GFP allows the detection of mutant clones from surrounding heterozygous cells as well as from wild-type cells in twin spots. For instance, Brown and Castelli-Gair Hombria (2000) generated clones of cells homozygous for a mutation in the essential transcription factor grain (grn) in leg imaginal discs of heterozygous animals, using the Flp/FRT recombination system [reviewed in Theodosiou and Xu (1998)]. These experiments used an FRT-bearing chromosome arm and a distal marker, ubi-GFPnls, which expresses GFP with a nuclear localization signal, from the ubiquitin promoter (Neufeld et al., 1998). grn mutant clones were recognized by their lack of GFP signal, since ubiGFPnls was carried on the homologous chromosome. Since GFP is brighter when present in two copies than in one, grn+/grn+ twin spots could also be identified by their stronger GFP signal. By examining twin spots in larval imaginal discs by GFP, and also examining clones in adult legs identified with cuticular markers, the authors were able to show that grn-associated defects in tissue shape occur during leg eversion, not earlier.

247

248

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

A

B

C

D

E

F

Figure 11.16. GFP balancer chromosomes. Balancer chromosomes were constructed that carry UAS-GFP and Kruppel (Kr)-GAL4. Flies bearing these balancers express GFP in the Kr pattern. In embryos, zygotic expression of GFP commences during germ band extension and can be detected at all subsequent stages. A. Stages 4–5, cellularization. The yellow signal is yolk. B. Stages 8, early germ band extension. C. Stages 9–12, germ band extension. GFP is first detected. D. Stages 13–14, germ band retraction. E. Stages 16. F. Stages 17, late embryo. a = amnioserosa, bo = Bolwig’s organ, cd = central domain, s = spiracles, y = yolk. (From Casso et al., 2000. Courtesy of Tom Kornberg and Dave Casso.) See color insert.

Datar et al. (2000) used ubi-GFPnls to mark clones of cells overexpressing the cell cycle genes CycD and Cdk4. Clones were generated using the flip-out Gal4 system, in which heat-shock-induced FLP produces clones of cells that express Gal4, which in turn induces the expression of UAS-regulated targets. Following clonal induction, the tissue of interest was dissociated and FACS-sorted, using the GFP signal to separate wild-type from overexpressing cells. Cell size and DNA content were simultaneously measured during the FACS sort, allowing the authors to directly compare cell cycle progression in wildtype and experimental cells within the same tissue. A very useful system for clonal analysis called MARCM (mosaic analysis with a repressible cell marker) was developed to positively mark cells with GFP (Lee and Luo, 1999, 2001). In this system, the clone expresses GFP in a background of heterozygous cells that do not (Fig. 11.17). This allows a marked cell to stand out with clarity in a situation where neighboring cells might otherwise obscure identification. When neurons are marked in this manner, the entire cell body and cell processes are labeled, allowing the cell projections to be readily mapped. The MARCM system is based on the fact that Gal80 represses the activity of Gal4. Clones are induced, using the FLP/FRT system, in heterozygous cells of the genotype FRT, UAS-GFP/FRT, tubulin-Gal80. A Gal4 transgene

GFP AS A GENETIC TOOL

Figure 11.17. Clonal analysis with GFP using the MARCM method. Cells expressing GFP are produced with the hsFLP-FRT method. UAS-GFP expression occurs, after mitotic recombination, as a result of losing a repressor of UAS-promoters, Gal80, that is expressed from a tub-Gal80 transgene on the homologous chromosome arm. A. This diagram, which represents neuroblast divisions, illustrates how neuronal clone size differs depending on the stage of mitotic recombination. B–D. Examples of neuronal clones in third instar larvae marked with GFP. (From Lee and Luo, 1999. Courtesy of Liqun Luo.)

can be present on the same chromosome arm as UAS-GFP, or elsewhere in the genome. Only cells that are homozygous for the FRT, UAS-GFP chromosome arm will express GFP, due to absence of tubulin-Gal80 in these cells. When a mutation is present on the FRT, UAS-GFP chromosome arm, only homozygous mutant cells are labeled with GFP. Also, ectopic expression and gain-of-function studies can be done with MARCM, when a gene to be clonally expressed is present on the UAS-GFP chromosome arm. The beauty of this system is apparent not only in the CNS, but also in other tissues (e.g., Brumby et al., 2003).

249

250

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

11.5.3 Screens The uses for GFP as a screening tool continue to grow. In some cases, embryos or larvae can be screened and selected a generation earlier than possible with screens that rely on observations of fixed tissues. Several groups have used GFP transgenes to highlight cells or tissues of interest in mutant backgrounds or to report changes in gene expression in mutants. Barthmair and Fyrberg (1995), for example, used Act88F-GFP, expressed in the developing indirect flight muscles, to screen through a collection of mutants for those altering indirect flight muscle development. Merabet et al. (2002) constructed a GFP reporter regulated by Hoxresponsive elements from the wingless and empty spiracles promoters and crossed the transgene to a collection of chromosomal deficiencies. They then assayed for changes in the GFP fluorescence pattern to identify deficiencies uncovering Hox modifier genes. In other cases, localization of a GFP fusion protein has been used to screen for mutants. Martin et al. (2003) used such a screen to identify mutations in essential genes that alter anterior–posterior patterning of the oocyte. Mutations were identified by alterations in GFP-Staufen localization in female germline clones, produced by the FRT-ovoD system. Classical enhancer trap screens can also be conducted more efficiently using GFP. Gerlitz et al. (2002) used a UAS-GFP reporter to screen through a collection of insertions of a Gal4-bearing P-element for those expressed in imaginal discs. This approach, which allows detection through the cuticle in living animals, is amenable to relatively highthroughput screening, and the authors report the identification of 2000 P-element insertions that drive GFP expression in imaginal discs (Gerlitz et al., 2002). Scully et al. (2002) developed an enhancer trap element with a reporter optimized for detecting both temporal and spatial expression patterns. This element, Pluc+, expresses a fusion of GFP (with five copies of the Myc epitope appended at its C-terminus) to luciferase: GFP-5XMyc-Luc. The luciferase moiety is an accurate reporter for temporal profiling of gene expression. GFP, on the other hand, along with the Myc epitope, provides a tag for precise spatial analysis of gene expression. In an initial screen, they demonstrated the value of this enhancer trap vector for identifying genes expressed in temporal patterns. By assaying bioluminescence activity in embryos over time, they identified an insertion in the gene regular (rgr), which is expressed in a circadian pattern. Spatial information about gene expression was provided by the GFP and Myc moieties, which showed that rgr is expressed in a subset of neurons in the brain. A screen for proteins that localize in the developing egg chamber was conducted by constructing an ovarian cDNA library in a P-element vector designed for germline expression of the cDNAs as fusions to GFP (Nakamura et al., 2001). The library was transformed into embryos by standard P-element transformation techniques, and adult females were dissected to observe the localization pattern of GFP in ovaries. Using this method, the authors identified a DEAD box RNA-helicase, Me31B, that is localized to cytoplasmic RNPs and is required for translational repression of localized mRNAs in the oocyte. GFP has been developed very successfully for use in protein trap screens. Morin et al. (2001) constructed a vector called PTT that carries GFP flanked by splice acceptor and donor sites. When this element is inserted in an intron, the GFP sequence is treated as an exon, generating GFP fusions. Because intron/exon junctions can occur in any frame, three separate PTT elements were constructed, in each possible reading frame relative to

GFP AS A GENETIC TOOL

the splice sites. GFP fusion proteins created in this manner are expressed from a gene’s natural promoter at its normal location in the genome, along with the regulatory elements that normally control expression of the gene in cis. In addition to identifying new genes for proteins expressed in particular patterns, this approach has produced (a) GFP tagged proteins for new and previously identified genes and (b) several markers for cell types and subcellular regions. This protein trap strategy has been adopted as a large-scale gene targeting screen, called “Flytrap,” conducted in several laboratories (Kelso et al., 2004). Information about Flytrap insertions can be obtained at the Flytrap website (http://flytrap.med.yale.edu). Additional protein trap lines are described at the Protein Trap database (http//biodev.obs-vlfr.fr/gavdos/protrap.htm). More recently, a smaller protein trap element, the Wee-P, was developed to minimize disruption of gene function associated with its insertion, to tag proteins at their N-termini or internally, and to transpose at a high frequency because of its small size (Clyne et al., 2003). Nonlethal protein trap insertions in essential genes are desirable, because they can be studied in homozygotes, allowing for higher levels of expression than heterozygous insertions. The Wee-P element was created in a two-step process. First, transformants were identified by the presence of a mini-white marker in a P-element also carrying GFP flanked by splice acceptor and donor sites. Then, because the white gene was flanked with FRT elements, a smaller element was created in vivo by excising white with FLP. The resulting element, Wee-P, is a streamlined 1.9-kb element that jumps with high efficiency. GFP’s AUG was maintained in this construct, and a Drosophila translation initiation consensus sequence was added. Since P-elements have a tendency to insert in 5¢ UTRs, this allows a protein to be tagged with GFP forming the N-terminus. However, GFP can also be expressed as an internal exon. As in Morin et al. (2003), Wee-P elements were created in 3 different reading frames relative to the splice junctions. Fig. 11.18 shows representative protein trap lines generated by Wee-P insertions. Further information about Wee-P technology and available insertion lines is available at Graeme Davis’s website: http://www.ucsf.edu/davislab/projects/wee-p_update.htm.

11.6

CONCLUDING REMARKS

There has been a virtual explosion of GFP used in Drosophila research in the past several years. The GFP fusion proteins produced in labs around the world now constitute a Drosophila cell biology kit for visualizing many different cell structures, including spindles and cytoplasmic microtubules, the actin cytoskeleton, nuclei, mitochondria, endoplasmic reticulum, the golgi, plasma membranes, chromosomes, and RNAs. This kit will continue to grow, and with the ongoing protein trap screens the very real possibility exists of someday tagging every Drosophila protein with GFP. As these fluorescent proteins become more widespread, cell and developmental biologists are studying biological processes as they happen in living cells. We are bound to learn new and unexpected things in the future as our view of the cell becomes ever more a view of a living, changing entity, in contact with other living cells and responding to its environment.

251

252

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

Figure 11.18. Proteins marked in vivo by the Wee-P GFP protein trap method. The Wee-P element was created in a two-step insertion and deletion process. The final element is a small, 1.9 Kb streamlined P-element that allows GFP to be inserted as an internal domain in a trapped protein, utilizing the splice acceptor and donor sites that flank GFP, or as an N-terminal fusion. Examples shown are from third-instar larval tissues. A. Wee-P114:GFP:Cam at a neuromuscular junction. B. Wee-P114:GFP:Cam in a peripheral sensory neuron. C. Wee-P:GFP:Cib in the optic lobes and some cells of the ventral ganglion. Wee-P:GFP:Cib in the cytoplasm of cells of the chordotonal organs (D) and the tracheal system (E). F. Wee-P:GFP:Sec61 in the cytoplasm of epithelial cells. Wee-P:GFP:Kis is nuclear, as shown in imaginal discs at low magnification (G), and appears punctate in the nuclei at higher magnification (H). (Clyne et al., 2003. Courtesy of Graeme Davis.)

REFERENCES Ahmad, K., and Henikoff, S. (2001). Modulation of a transcription factor counteracts heterochromatic gene silencing in Drosophila. Cell 104:839–847. Aubin, J. E. (1979). Autofluorescence of viable cultured mammalian cells. J. Histochem. Cytochem. 27:36043.

REFERENCES

Barolo, S., Carver, L. A., and Posakony, J. W. (2000). GFP and beta-galactosidase transformation vectors for promoter/enhancer analysis in Drosophila. Biotechniques 29:726–732. Barthmaier, P., and Fyrberg, E. (1995). Monitoring development and pathology of Drosophila indirect flight muscles using green fluorescent protein. Dev. Biol. 169:770–774. Berdnik, D., and Knoblich, J. A. (2002). Drosophila Aurora-A is required for centrosome maturation and actin-dependent asymmetric protein localization during mitosis. Curr. Biol. 12:640–647. Bertrand, E., Chartrand, P., Schaefer, M., Shenoy, S. M., Singer, R. H., and Long, R. M. (1998). Localization of ASH1 mRNA particles in living yeast. Mol. Cell 2:437–445. Bhalero, S., Sen, A., Stocker, R., and Rodrigues, V. (2003). Olfactory neurons expressing identified receptor genes project to subsets of glomeruli within the antennal lobe of Drosophila melanogaster. J. Neurobiol. 54:577–592. Bharathi, V., Pallavi, S. K., Bajpai, R., Emerald, B. S., and Shashidhara, L. S. (2004). Genetic characterization of the Drosophila homologue of coronin. J. Cell Sci. 117:1911–1922. Bobinnec, Y., Marcaillou, C., Morin, X., and Debec, A. (2003). Dynamics of the endoplasmic reticulum during early development of Drosophila melanogaster. Cell Motil. Cytoskeleton 54:217–225. Bossing, T., Barros, C. S., and Brand, A. H. (2002). Rapid tissue-specific expression assay in living embryos. Genesis 34:123–126. Brand, A. H., and Perrimon, N. (1993). Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118:401–415. Brand, A. H. (1995). GFP in Drosophila. Trends Genet. 11:324–325. Brennecke, J., Hipfner, D. R., Stark, A., Russell, R. B., and Cohen, S. M. (2003). bantam encodes a developmentally regulated microRNA that controls cell proliferation and regulates the proapoptotic gene hid in Drosophila. Cell 113:25–36. Brown, S., and Castelli-Gair Hombria, J. (2000). Drosophila grain encodes a GATA transcription factor required for cell rearrangement during morphogenesis. Development 127:4867–4876. Brumby, A. M., and Richardson, H. E. (2003). scribble mutants cooperate with oncogenic Ras or Notch to cause neoplastic overgrowth in Drosophila. EMBO J. 22:5769–5779. Casso, D., Ramirez-Weber, F., and Kornberg, T. B. (2000). GFP-tagged balancer chromosomes for Drosophila melanogaster. Mech. Dey. 91:451–454. Cha, B. J., Koppetsch, B. S., and Theurkauf, W. E. (2001). In vivo analysis of Drosophila bicoid mRNA localization reveals a novel microtubule-dependent axis specification pathway. Cell 106:35–46. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263:802–805. Clarkson, M., and Saint, R. (1999). A His2AvDGFP fusion gene complements a lethal His2AvD mutant allele and provides an in vivo marker for Drosophila chromosome behavior. DNA Cell Biol. 18:457–462. Clyne, P. J., Brotman, J. S., Sweeney, S. T., and Davis, G. (2003). Green fluorescent protein tagging Drosophila proteins at their native genomic loci with small P elements. Genetics 165:1433–1441. Cox, R. T., and Spradling, A. C. (2003). A Balbiani body and the fusome mediate mitochondrial inheritance during Drosophila oogenesis. Development 130:1579–1590. Datar, S. A., Jacobs, H. W., de la Cruz, A. F., Lehner, C. F., and Edgar, B. A. (2000). The Drosophila cyclin D-Cdk4 complex promotes cellular growth. EMBO J. 19:4543–4554. Davis, I., Girdham, C. H., and O’Farrell, P. H. (1995). A nuclear GFP that marks nuclei in living Drosophila embryos; maternal supply overcomes a delay in the appearance of zygotic fluorescence. Dev. Biol. 170:726–729.

253

254

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

Dietzel, S., Niemann, H., Bruckner, B., Maurange, C., and Paro, R. (1999). The nuclear distribution of Polycomb during Drosophila melanogaster development shown with a GFP fusion protein. Chromosoma 108:83–94. Dorman, J. B., James, K. E., Fraser, S. E., Kiehart, D. P., and Berg, C. A. (2004). Bullwinkle is required for epithelial morphogenesis during Drosophila oogenesis. Dev. Biol. 267:320–341. Duffy, J. B. (2002). GAL4 system in Drosophila: A fly geneticist’s Swiss army knife. Genesis 34(1–2):1–15. Dutta, D., Bloor, J. W., Ruiz-Gomez, M., Vijay Raghavan, K., and Kiehart, D. P. (2002). Real-time imaging of morphogenetic movements in Drosophila using Gal4-UAS-driven expression of GFP fused to the actin-binding domain of moesin. Genesis 34:146–151. Edwards, K. A., Demsky, M., Montague, R. A., Weymouth, N., and Kiehart, D. P. (1997). GFP-Moesin illuminates actin cytoskeleton dynamics in living tissue and demonstrates cell shape changes during morphogenesis in Drosophila. Dev. Biol. 191:103–117. Endow, S., and Komma, D. (1996). Centrosome and spindle function of the Drosophila Ncd microtubule motor visualized in live embryos using Ncd-GFP fusion proteins. J. Cell Sci. 109:2429–2442. Endow, S., and Komma, D. (1997). Spindle dynamics during meiosis in Drosophila oocytes. J. Cell Biol. 137:1321–1336. Entchev, E. V., Schwabedissen, A., and Gonzalez-Gaitan, M. (2000). Gradient formation of the TGF-beta homolog Dpp. Cell 103:981–991. Ferrandon, D., Jung, A. C., Criqui, M., Lemaitre, B., Uttenweiler-Joseph, S., Michaut, L., Reichhart, J., and Hoffmann, J. A. (1998). A drosomycin-GFP reporter transgene reveals a local immune response in Drosophila that is not dependent on the Toll pathway. EMBO J. 17:1217–1227. Forrest, K. M., and Gavis, E. R. (2003). Live imaging of endogenous RNA reveals a diffusion and entrapment mechanism for nanos mRNA localization in Drosophila. Curr. Biol. 13:1159–1168. Furlong, E. E., Profitt, D., and Scott, M. P. (2001a). Automated sorting of live transgenic embryos. Nat. Biotechnol. 19:153–156. Furlong, E. E., Andersen, E. C., Null, B., White, K. P., and Scott, M. P. (2001b). Patterns of gene expression during Drosophila mesoderm development. Science 293:1629–1633. Gerlitz, O., Nellen, D., Ottiger, M., and Basler, K. (2002). A screen for genes expressed in Drosophila imaginal discs. Int. J. Dev. Biol. 46:173–176. Gibson, M. C., Lehman, D. A., and Schubiger, G. (2002). Lumenal transmission of Decapentaplegic in Drosophila imaginal discs. Dev. Cell 3:451–460. Grieder, N. C., de Cuevas, M., and Spradling, A. C. (2000). The fusome organizes the microtubule network during oocyte differentiation in Drosophila. Development 127:4253–4264. Guild, G. M., Connelly, P. S., Vranich, K. A., Shaw, M. K., and Tilney, L. G. (2002). Actin filament turnover removes bundles from Drosophila bristle cells. J. Cell Sci. 115:641–653. Halfon, M. S., Gisselbrecht, S., Lu, J., Estrada, B., Keshishian, H., and Michelson, A. M. (2002). New fluorescent protein reporters for use with the Drosophila Gal4 expression system and for vital detection of balancer chromosomes. Genesis 34:135–138. Hazelrigg, T., Liu, N., Hong, Y., and Wang, S. (1998). GFP expression in Drosophila tissues: Time requirements for formation of a fluorescent product. Dev. Biol. 199:245–249. Hiroi, M., Marion-Poll, F., and Tanimura, T. (2002). Differentiated response to sugars among labellar chemosensilla in Drosophila. Zoolog. Sci. 19:1009–1018. Horn, C., and Wimmer, E. A. (2000). A versatile vector set for animal transgenics. Dev. Genes Evol. 201:630–637. Horn, C., Jaunich, B., and Wimmer, E. A. (2000). Highly sensitive, fluorescent transformation marker for Drosophila transgenesis. Dev. Genes Evol. 210:623–629.

REFERENCES

Huang, J., and Raff, J. W. (1999). The disappearance of cyclin B at the end of mitosis is regulated spatially in Drosophila cells. EMBO J. 18:2184–2195. Hutson, M. S., Tokutake, Y., Chang, M. S., Bloor, J. W., Venakides, S., Kiehart, D. P., and Edwards, G. S. (2003). Forces for morphogenesis investigated with laser microsurgery and quantitative modeling. Science 300:145–149. Ito, K., Suzuki, K., Estes, P., Ramaswami, M., Yamamoto, D., and Strausfeld, N. J. (1998). The organization of extrinsic neurons and their implications in the functional roles of the mushroom bodies in Drosophila melanogaster Meigen. Learn Mem. 5:52–77. Jacinto, A., Wood, W., Balayo, T., Turmaine, M., Martinez-Arias, A., and Martin, P. (2000). Dynamic actin-based epithelial adhesion and cell matching during Drosophila dorsal closure. Curr. Biol. 10:1420–1426. Jacinto, A., Wood, W., Woolner, S., Hiley, C., Turner, L., Wilson, C., Martinez-Arias, A., and Martin, P. (2002). Dynamic analysis of actin cable function during Drosophila dorsal closure. Curr Biol. 12:1245–1250. Jin, Y., Wang, Y., Walker, D. L., Dong, H., Conley, C., Johansen, J., and Johansen, K. M. (1999). JIL-1: A novel chromosomal tandem kinase implicated in transcriptional regulation in Drosophila. Mol. Cell 4:129–135. Kelso, R. J., Buszczak, M., Quinones, A. T., Castiblanco, C., Mazzalupo, S., and Cooley, L. (2004). Flytrap, a database documenting a GFP protein-trap insertion screen in Drosophila melanogaster. Nucleic Acids Res. 32:D418–D420. Kerrebrock, A. W., Moore, D. P., Wu, J. S., and Orr-Weaver, T. L. (1995). Mei-S332, a Drosophila protein required for sister-chromatid cohesion, can localize to meiotic centromere regions. Cell 83:247–256. Kiehart, D. P., Galbraith, C. G., Edwards, K. A., Rickoll, W. L., and Montague, R. A. (2000). Multiple forces contribute to cell sheet morphogenesis for dorsal closure in Drosophila. J. Cell Biol. 149:471–490. Lee, T., and Luo, L. (1999). Mosaic analysis with a repressible cell marker for studies of gene function in neuronal morpholgenesis. Neuron 22:451–461. Lee, T., and Luo, L. (2001). Mosaic analysis with a repressible cell marker (MARCM) for Drosophila neural development. Trends Neurosci. 24:251–254. Lee, M. J., Gergely, F., Jeffers, K., Peak-Chew, S. Y., and Raff, J. W. (2001). Msps/XMAP215 interacts with the centrosomal protein D-TACC to regulate microtubule behaviour. Nat. Cell Biol. 3:643–649. Levashina, E. A., Ohresser, S., Lemaitre, B., and Imler, J. L. (1998). Two distinct pathways can control expression of the gene encoding the Drosophila antimicrobial peptide metchnikowin. J. Mol. Biol. 278:515–527. Martin, S. G., Leclerc, V., Smith-Litiere, K., and St. Johnston, D. (2003). The identification of novel genes required for Drosophila anteroposterior axis formation in a germline clone screen using GFP-Staufen. Development 130:4201–4215. Megraw, T. L., Kilaru, S., Turner, F. R., and Kaufman, T. C. (2002). The centrosome is a dynamic structure that ejects PCM flares. J. Cell Sci. 115:4707–4718. Merabet, S., Catala, F., Pradel, J., and Graba, Y. (2002). A green fluorescent protein reporter genetic screen that identifies modifiers of Hox gene function in the Drosophila embryo. Genetics 162:189–202. Morin, X., Daneman, R., Zavortink, M., and Chia, W. (2001). A protein trap strategy to detect GFP-tagged proteins expressed from their endogenous loci in Drosophila. PNAS 98:15050– 15055. Nagel, A. C., Maier, D., and Preiss, A. (2002). Green fluorescent protein as a convenient and versatile marker for studies on functional genomics in Drosophila. Dev. Genes Evol. 212:93–98.

255

256

GREEN FLUORESCENT PROTEIN APPLICATIONS IN DROSOPHILA

Nakamura, A., Amikura, R., Hanyu, K., and Kobayashi, S. (2001). Me31B silences translation of oocyte-localizing RNAs through the formation of cytoplasmic RNP complex during Drosophila oogenesis. Development 128:3233–3242. Neufeld, T. P., de la Cruz, A. F., Johnston, L. A., and Edgar, B. A. (1998). Coordination of growth and cell division in the Drosophila wing. Cell 93:1183–1193. Niswender, K. D., Blackman, S. M., Rohde, L., Magnuson, M. A., and Piston, D. W. (1995). Quantitative imaging of green fluorescent protein in cultured cells: Comparison of microscopic techniques, use in fusion proteins and detection limits. J. Microsc. 180:109–116. Noguchi, T., and Miller, K. G. (2003). A role for actin dynamics in individualization during spermatogenesis in Drosophila melanogaster. Development 130:1805–1816. Parker, L., Gross, S., and Alphey, L. (2001). Vectors for the expression of tagged proteins in Drosophila. Biotechniques 31:1280–1286. Pfeiffer, S., Ricardo, S., Manneville, J. B., Alexandre, C., and Vincent, J. P. (2002). Producing cells retain and recycle Wingless in Drosophila embryos. Curr. Biol. 12:957–962. Plautz, J. D., Day, R. N., Dailey, G. M., Welsh, S. B., Hall, J. C., Halpain, S., and Kay, S. A. (1996). Green fluorescent protein and its derivatives as versatile markers for gene expression in living Drosophila melanogaster, plant and mammalian cells. Gene 173:83–87. Potter, S. M., Wang, C.-W., Garrity, P. A., and Fraser, S. E. (1996). Intravital imaging of green fluorescent protein using two-photon laser-scanning microscopy. Gene 173:25–31. Raff, J. W., Jeffers, K., and Huang, J. Y. (2002). The roles of Fzy/Cdc20 and Fzr/Cdh1 in regulating the destruction of cyclin B in space and time. J. Cell Biol. 157:1139–1149. Rashkova, S., Karam, S. E., Kellum, R., and Pardue, M. L. (2002). Gag proteins of the two Drosophila telomeric retrotransposons are targeted to chromosome ends. J. Cell Biol. 159:397–402. Rashkova, S., Athanasiadis, A., and Pardue, M. L. (2003). Intracellular targeting of Gag proteins of the Drosophila telomeric retrotransposons. J. Virol. 77:6376–6384. Ritzenthaler, S., Suzuki, E., and Chiba, A. (2000). Postsynaptic filopodia in muscle cells interact with innervating motoneuron axons. Nat. Neurosci. 3:1012–1017. Rong, Y. S., and Golic, K. G. (2000). Gene targeting by homologous recombination in Drosophila. Science 288:2013–2018. Salvaterra, P. M., and Kitamoto, T. (2001). Drosophila cholinergic neurons and processes visualized with Gal4/UAS-GFP. Gene. Expr. Patterns 1:73–82. Sano, H., Nakamura, A., and Kobayahsi, S. (2002). Identification of a transcriptional regulatory region for germline-specific expression of vasa gene in Drosophila melanogaster. Mech. Dev. 112:129–139. Savoian, M. S., and Rieder, C. L. (2002). Mitosis in primary cultures of Drosophila melanogaster larval neuroblasts. J. Cell Sci. 115:3061–3072. Scully, A. L., Zelhof, A. C., and Kay, S. A. (2002). A P element with a novel fusion of reporters identifies regular, a C2H2 zinc-finger gene downstream of the circadian clock. Mol. Cell Neurosci. 19:501–514. Sepp, K. J., and Auld, V. J. (2003). RhoA and Rac1 GTPases mediate the dynamic rearrangement of actin in peripheral glia. Development 130:1825–1835. Shiga, Y., Tanaka-Matakatsu, M., and Hayashi, S. (1996). A nuclear GFP/b-galactosidase fusion protein as a marker in living Drosophila. Dev. Growth Differ. 38:99–106. Teleman, A. A., and Cohen, S. M. (2000). Dpp gradient formation in the Drosophila wing imaginal disc. Cell 103:971–980. Theodosiou, N. A., and Xu, T. (1998). Use of FLP/FRT system to study Drosophila development. Methods 14:355–365.

REFERENCES

Theurkauf, W. E., and Hazelrigg, T. I. (1998). In vivo analyses of cytoplasmic transport and cytoskeletal organization during Drosophila oogenesis: Characterization of a multistep anterior localization pathway. Development 125:3655–3666. Timmons, L., Becker, J., Barthmaier, P., Fyrberg, C., Shearn, A., and Fyrberg, E. (1997). Green fluorescent protein/b-galactosidase double reporters for visualizing Drosophila gene expression patterns. Dev. Genet. 20:338–347. Tzou, P., Ohresser, S., Ferrandon, D., Capovilla, M., Reichhart, J. M., Lemaitre, B., Hoffmann, J. A., and Imler, J. L. (2000). Tissue-specific inducible expression of antimicrobial peptide genes in Drosophila surface epithelia. Immunity 13:737–748. Van Roessel, P., Hayward, N. M., Barros, C. S., and Brand, A. H. (2002). Two-color GFP imaging demonstrates cell-autonomy of GAL4-driven RNA interference in Drosophila. Genesis 34:170–173. Vazquez, J., Belmont, A. S., and Sedat, J. W. (2001). Multiple regimes of constrained chromosome motion are regulated in the interphase Drosophila nucleus. Curr. Biol. 11:1227–1239. Vazquez. J., Belmont, A. S., and Sedat, J. W. (2002). The dynamics of homologous chromosome pairing during male Drosophila meiosis. Curr. Biol. 12:1473–1483. Verkhusha, V. V., Tsukita, S., and Oda, H. (1999). Actin dynamics in lamellipodia of migrating border cells in the Drosophila ovary revealed by a GFP-actin fusion protein. FEBS Lett. 445:395–401. Vermaak, D., Hayden, H. S., and Henikoff, S. (2002). Centromere targeting element within the histone fold domain of Cid. Mol. Cell Biol. 22:7553–7561. Wang, S., and Hazelrigg, T. (1994). Implications for bcd mRNA localization from spatial distribution of exu protein in Drosophila oogenesis. Nature (London) 369:400–403. Ward, R. E., Reid, P., Bashirullah, A., D’Avino, P. P., and Thummel, C. S. (2003). GFP in living animals reveals dynamic developmental responses to ecdysone during Drosophila metamorphosis. Dev. Boil. 256:389–402. Wilhelm, J. E., Mansfield, J., Hom-Booher, N., Wang, S., Turck, C. W., Hazelrigg, T., and Vale, R. D. (2000). Isolation of a ribonucleoprotein complex involved in mRNA localization in Drosophila oocytes. J. Cell Biol. 148:427–440. Wood, W., Jacinto, A., Grose, R., Woolner, S., Gale, J., Wilson, C., and Martin, P. (2002). Wound healing recapitulates morphogenesis in Drosophila embryos. Nat. Cell Biol. 4:907–912. Yeh, E., Gustafson, K., and Boulianne, G. L. (1995). Green fluorescent protein as a vital marker and reporter of gene expression in Drosophila. Proc. Natl. Acad. Sci. 92:7036–7040. Zhang, Y. Q., Rodesch, C. K., and Broadie, K. (2002). Living synaptic vesicle marker: Synaptotagmin-GFP. Genesis 34:142–145.

257

12 THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS Jim Haseloff* and Kirby R. Siemering Division of Cell Biology, MRC Laboratory of Molecular Biology, Hills Road, Cambridge, United Kingdom

12.1

INTRODUCTION

Marker genes have proved extremely useful for reporting gene expression in transformed plants. The b-glucuronidase gusA (GUS) gene (Jefferson et al., 1987) has been used extensively. Transformed cells or patterns of gene expression within plants can be identified histochemically, but this is generally a destructive test and is not suitable for assaying primary transformants, or for following the time course of gene expression in living plants, or as a means of rapidly screening segregating populations of seedlings. The green fluorescent protein (GFP) from Aequorea victoria shares none of these problems, because its intrinsic fluorescence can be seen in living cells. In addition, there has been intense interest in its use as a marker for transgenic plants. Unmodified gfp has been successfully expressed at high levels in tobacco plants using the cytoplasmic ribonucleic acid (RNA) viruses potato virus X (Baulcombe et al., 1995) and tobacco mosaic virus (Heinlein et al., 1995). In these experiments, the gene was directly expressed as a viral mRNA in infected cells, and spectacularly high levels of GFP fluorescence were seen. In contrast to the efficient RNA virus-mediated expression of GFP, variable results have been obtained with transformed cells and plants. Although green fluorescence has been seen in gfp transformed protoplasts of citrus (Niedz et al., 1995) and maize (Hu and * Present address: Department of Plant Sciences, University of Cambridge, Cambridge, United Kingdom. Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

259

260

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

Cheng, 1995; Sheen et al., 1995), we and others have seen no fluorescence in Arabidopsis and other transformed plant species (Haseloff and Amos, 1995; Haseloff et al., 1997). Hu and Cheng (1995) reported that no signal was seen in gfp transformed Arabidopsis thaliana protoplasts. Reichel et al. (1996) also failed to detect fluorescence in gfp transformed Arabidopsis, tobacco, or barley protoplasts. Sheen et al. (1995) also saw no expression of a CAB2-driven gfp gene in transgenic Arabidopsis plants and Pang et al. (1996) saw little or no expression in gfp transformed wheat, corn, tobacco, and Arabidopsis plants. There appears to be a need for substantial improvement of the wild-type gfp gene for use in plants. In this chapter, we describe some of the pitfalls affecting gfp expression and detection in plants, and describe modified forms of the gene and new techniques which have helped to overcome these problems.

12.2

CRYPTIC SPLICING OF gfp mRNA IN ARABIDOPSIS

Useful expression of the gfp cDNA (cyclic deoxyribonucleic acid) in plants requires that (a) the GFP apoprotein be produced in suitable amounts within plant cells, and (b) the nonfluorescent apoprotein undergoes efficient posttranslational modification to produce the mature GFP. The high levels of GFP fluorescence seen in plants infected with suitable RNA virus vectors (Baulcombe et al., 1995; Heinlein et al., 1995) demonstrate that the protein can undergo efficient posttranslational modification in plants. However, the expression of integrated copies of the gene has proved problematic. We have used Agrobacterium-mediated transformation to produce transgenic Arabidopsis plantlets containing a cauliflower mosaic virus 35S promoter-driven gfp cDNA (Haseloff et al., 1997). However, at no stage during the transformation procedure did we detect GFP related fluorescence, using an ultraviolet (UV) lamp illumination and epifluorescence microscopy. Therefore we used polymerase chain reaction (PCR)-based methods to verify the correct insertion of the gene and to check mRNA transcription and processing in these transformed plantlets. Samples of DNA and mRNA were separately extracted, and gfp sequences were amplified via PCR from the separate extracts and analysed. While the expected full-length gfp product was obtained after amplification of the integrated gene, RT-PCR of gfp mRNA sequences gave rise to a truncated product. The shortened RT-PCR product was cloned and sequenced, and a deletion of 84 nucleotides was found between residues 380–463 of the GFP coding sequence (Fig. 12.1). The missing sequence bears close similarity to known plant introns and it is likely that expression of gfp in Arabidopsis is curtailed by aberrant mRNA splicing, with an 84 nucleotide sequence being recognized as a cryptic intron. This explanation would also account for the efficient expression of gfp from RNA virus vectors that replicate in the cytoplasm, and thus evade splicing. The nucleotide sequences bordering the deletion are shown in Fig. 12.1 and demonstrate similarity to known plant introns. Matches were found for sequences that are conserved at the 5¢ and 3¢ splice sites of plant introns [reviewed in Luehrsen et al. (1994)] and for conserved branch point nucleotides in plant introns (Liu and Filipowicz, 1996; Simpson et al., 1996). The excised gfp sequence contains a high AU content (68%) that has also been shown to be important for recognition of plant introns (Hanley and Schuler, 1988; Wiebauer et al., 1988; Goodall and Filipowicz, 1989; Goodall and Filipowicz, 1991). It is likely that this 84 nucleotide region of the jellyfish gfp cDNA sequence is efficiently recognized as an intron when transcribed in Arabidopsis, resulting in an in frame deletion and the production of a defective protein product, which is predicted to be 28 aa shorter. Subsequently, an artificial neural network program has been

CRYPTIC SPLICING OF gfp mRNA IN ARABIDOPSIS

Figure 12.1. (opposite) Cryptic splicing of gfp transcripts in transgenic Arabidopsis thaliana. (A) Restriction endonuclease digestion of PCR fragments derived from gfp DNA and mRNA sequences. Sequences corresponding to the integrated gfp gene and to mRNA transcripts were isolated and separately amplified using PCR techniques, and incubated with various restriction endonucleases. The radiolabeled fragments were fractionated by electrophoresis in a 5% polyacrylamide gel, and are shown labeled with the source of the amplified sequences (DNA or mRNA) and the name of the restriction endonuclease used for digestion, or not (uncut). The mRNA derived sequences appeared to lack sites for Dra I and Acc I, and to contain a corresponding deleted region of 80–90 nucleotides. Restriction endonuclease fragments that are smaller than those expected of the gene sequence have been indicated with a white asterisk. (B) Sequence analysis of cloned gfp mRNAs. Autoradiograph and sequence of the amplified gfp mRNA sequence. Nucleotides 380–463 are absent from the transcribed sequence, and the site of this 84 nucleotide deletion is arrowed. (C) Schematic diagram of the gfp gene sequence, which shows the positions of restriction endonuclease sites used for the analysis of PCR amplified mRNA transcripts, and the location of the cryptic intron, shown with dark shading. Sequences that are similar to those normally found at splice sites and branch points of plant introns are shown below. Splice sites are arrowed, and the putative lariant branch point is shown in reverse type (Haseloff et al., 1997).

261

262

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

used to correctly predict the presence of the Arabidopsis cryptic intron in the gfp coding sequence (Hebsgaard et al., 1996), and similar cryptic splicing has been seen now in other plant species (Schuler, personal communication). It should be noted that the borders of the cryptic intron do not coincide with any of the natural spliced junctions found after processing of the gfp mRNA in Aequorea victoria (Prasher et al., 1992). No full-length gfp mRNA is detectable by RT-PCR, and so misprocessing must be close to complete in transformed Arabidopsis plantlets. It has been claimed that gfp fluorescence has been detected in bombarded Arabidopsis tissues (Sheen et al., 1995). However, in these experiments, leaf tissue was treated with methanol prior to microscopic examination, where methanol causes rapid and irreversible bleaching of GFP (Ward et al., 1980). Local wounding due to particle bombardment can cause punctate patterns of bright autofluorescence, and this type of experiment needs to be interpreted with care. The same authors saw no expression of a CAB2-driven gfp gene in transgenic Arabidopsis plants (Sheen et al., 1995). It is likely that elimination of cryptic splicing is essential for proper expression of the gfp gene in Arabidopsis.

12.3

REMOVAL OF THE CRYPTIC INTRON

It has proved necessary to destroy this cryptic intron to ensure proper expression in plants. We have altered the codon usage for GFP, deliberately mutating recognition sequences at the putative 5¢ splice site and branch point and decreasing the AU content of the intron. All of the sequence modifications affected only codon usage, and this modified gene, mgfp4, encodes a protein product that is identical to that of the jellyfish (Fig. 12.2). When the mgfp4 sequence was inserted behind the 35S promoter and introduced into Arabidopsis using the root transformation technique (Valvekens et al., 1988), bright green fluorescent plant cells were detected within 2–3 days of cocultivation. As cell proliferation continued, the brightest clumps of callus and developing shoot tissue were so intensely fluorescent that they were clearly visible by eye, using a 100-W long wavelength handheld UV lamp (UV Products, B100AP). We have also adapted an inverted fluorescence microscope (Leitz DM-IL) to allow more sensitive, higher magnification observation of cells in sterile culture during transformation and regeneration. The microscope was fitted with a filter set (Leitz-D excitation BP355–425, dichroic 455, emission LP460) suitable for the main 395-nm excitation and 509-nm emission peaks of GFP, and we used a 7-mm threaded extension tube with a 4¥ objective (EF 4/0.12) to give a greater working distance above the microscope stage. This allows the convenient direct observation of transformed tissues and plantlets within sealed inverted Petri dishes. The ease with which fluorescent proteins can be monitored in living tissues allows new approaches for improving transformation and regeneration of intractable or slowgrowing plant species. During our own regeneration experiments, we observed a wide range of GFP fluorescence intensities in 35S-mgfp4 transformed plantlets, which we expect arose from position-dependent modulation of gene expression in different transformants. It proved difficult to regenerate fertile plants from the brightest transformants, with cells remaining as a highly fluorescent callus or mass of shoots after several months of culture. It is possible that high levels of GFP expression were mildly toxic or interfered with differentiation. This toxicity is of special concern with a fluorescent molecule such as GFP, which would be expected to generate free radicals upon excitation, and which undergoes oxidative modification and could possess catalytic properties. The conditions that we have used for plant regeneration should provide a stringent test for any deleterious effect due

EXPRESSION OF GFP IN OTHER PLANTS

Figure 12.2. (opposite) Sequence comparison of gfp and the modified mgfp5-ER gene. The sequence of gfp is as described for the gfp10 cDNA (Prasher et al., 1992), except that codon 80 contains a change from CAG to CGG resulting in replacement of a glutamine with arginine, as noted by Chalfie et al. (1994). Both the gfp and mgfp5-ER gene cassettes are flanked by restriction endonuclease sites for BamHI and SacI, a ribosome binding site (RBS) for bacterial expression and the sequence AACA upstream of the start codon for improved plant translation. The cryptic plant intron present in gfp (Haseloff et al., 1997) is shown underlined with the 5¢ and 3¢ splice sites arrowed. Nucleotide sequence alterations present in mgfp5-ER are shown outlined in gray. Most alterations are silent and all amino acid substitutions are shown in reverse type below the nucleotide sequence. The mgfp5-ER gene cassette contains additional sequences shown in bold face type, which comprise a 5¢ terminal signal peptide and 3¢ HDEL sequence. An EcoRI site was used to link the signal peptide and coding sequences.

to GFP. The 35S promoter was used to drive expression of the protein at high levels throughout the plant, including meristematic cells, and regeneration took place under continual illumination, allowing the possibility for GFP mediated phototoxicity. Despite poor regeneration of the brightest transformants, we managed to obtain over 50 separate transgenic Arabidopsis lines, most of which contained levels of GFP that were easily detectable by microscopy.

12.4

EXPRESSION OF GFP IN OTHER PLANTS

Some expression of wild-type gfp has been seen in plant protoplasts of tobacco (Reichel et al., 1996), Citrus sinensis (Niedz et al., 1995) and maize (Hu and Cheng, 1995; Sheen et al., 1995; Pang et al., 1996), and so aberrant splicing of gfp mRNA may not be as efficient in other plant species, as in transgenic Arabidopsis. However, the mgfp4 gene has

263

264

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

proved useful for expression studies in other plants, which share features involved in intron recognition (Luehrsen et al., 1994). Experiments with tobacco and barley protoplasts (Reichel et al., 1996) demonstrated that mgfp4 derived sequences are expressed at much higher levels than the wild-type gene in these species. The mgfp4 is also expressed efficiently in soybean cells (Plautz et al., 1996). There have been reports of improved GFP expression in mammalian cells after alteration of gene codon usage (Haas et al., 1996; Zolotukhin et al., 1996). Increased levels of expression have been attributed to improved rates of translation due to optimized codon usage. However, this “humanization” of gfp also leads to alteration of the cryptic intron sequence, and expression of sGFP (Haas et al., 1996) been shown to result in 20-fold increased fluorescence in maize protoplasts (Chiu et al., 1996). The increased levels of expression may be due, at least in part, to an effect on RNA processing. Other workers have also found it necessary to deliberately alter the codon usage of gfp for efficient expression in plants. Transient expression of the synthetic pgfp gene gave rise to about 20-fold more fluorescence than wild-type gfp in maize and tobacco protoplasts (Pang et al., 1996). It is possible that altered mRNA sequences affect posttranscriptional processing in animal cells as well. However, introns found in animals, including A. victoria (Prasher et al., 1992), share a conserved polypyrimidine tract adjacent to the 3¢ splice site, reviewed in (Green, 1992), and introns in yeast cells possess a requirement for additional conserved sequences (UACUAAC) located at the branch point (Langford et al., 1984). The lack of these additional features may help to minimize recognition of the cryptic intron and aberrant processing of gfp mRNA in fungal and animal cells.

12.5

LOCALIZATION OF GFP IN PLANT CELLS

In transgenic Arabidopsis cells, GFP is found throughout the cytoplasm, but appears to accumulate within the nucleoplasm (Haseloff and Amos, 1995; Chiu et al., 1996; Grebenok et al., 1997; Haseloff et al., 1997; Kohler et al., 1997). It is excluded from vacuoles, organelles, and other bodies in the cytoplasm and is excluded from the nucleolus (Fig. 12.3). A similar subcellular distribution of GFP was seen in all Arabidopsis cell types examined in our experiments, and red autofluorescent chloroplasts provide an effective counter-fluor for GFP in the upper parts of the plant. Cytoplasmic streaming and the movement of organelles could be observed in these living cells. In addition to cell ultrastructure, the architecture of the intact tissue was also clearly discernible, and the arrangement of different cell types could be seen in longitudinal optical sections of root tips and cotyledons. For example, cells within the epidermis of the cotyledon contain few mature chloroplasts and could be distinguished from layers of neighboring mesophyll cells, and flies of developing cells around the primary root meristem are clearly evident (Fig. 12.3). While the mgfp4 gene was proving useful as a marker in transgenic Arabidopsis, it was also clear from the initial studies that it could bear improvement. While we were able to generate 35S-mgfp4 transformed cells and calli that were intensely fluorescent, and easily detectable by eye under long wavelength UV illumination, it proved difficult to regenerate fertile plants from the brightest transformants. It is possible that very high levels of GFP expression are mildly toxic or interfere with regeneration, perhaps due to the fluorescent or catalytic properties of the protein. In jellyfish photocytes, where high levels of GFP are well tolerated, the protein is found sequestered in cytoplasmic granules (Davenport and Nichol, 1955). In contrast, the mature protein is found throughout the cytoplasm and accumulates within the nucleoplasm of transformed Arabidopsis cells. If GFP

SUBCELLULAR TARGETING OF GFP

Figure 12.3. Confocal images of 35S-mgfp4 transformed Arabidopsis plants. The 35S-mgfp4 transformed seedlings were grown in sterile agar culture and mounted intact in water for confocal microscopy. Images were collected using a BioRad MRC-600 instrument equipped with Nikon Optiphot microscope and Nikon planapo 60¥ water immersion lens. The GFP and chlorophyll were excited using the 488- and 568-nm lines, respectively, of a 25-nm krypton–argon ion laser. The green and red emissions were collected in separate channels and combined using Adobe Photoshop. (A) The shoot apical meristem is shown. Individual vacuolate cells that each contain a layer of green fluorescent cytoplasm containing red autofluorescent chloroplasts, can be distinguished. (B) A emerging leaf primordia, positioned at the shoot apex between two cotyledons. (C) Mesophyll cells within a cotyledon show large numbers of mature chloroplasts. (D) An optical section of a single hypocotyl cortex cell, showing mature chloroplasts. (E) Cells from within the root meristem. The GFP accumulates within nuclei, but is excluded from nucleoli, and is found throughout the cytoplasm where various endomembrane compartments are shown in negative relief. (F) Median longitudinal optical section of a root tip.

is a source of fluorescence-related free radicals, for example, it might be advisable to target the protein to a more localized compartment within the plant cell.

12.6

SUBCELLULAR TARGETING OF GFP

We have fused several targeting peptides to GFP, and directed the protein to different subcellular compartments. The targeted forms of the mgfp4 gene were initially tested by expression in Saccharomyces cerevisiase. The modified genes were introduced into yeast cells on a multicopy vector and expressed fluorescent protein was visualized using confocal microscopy (Fig. 12.4). Unmodified protein is normally found throughout the cytoplasm and nucleoplasm of yeast cells. The addition of a peptide containing the SV40 T-antigen NLS (amino acids APKKKRKVEDPR) to the N- or C-terminus of the protein does little to alter its distribution (not shown). However, if the NLS-GFP protein is fused to a larger protein, such as that encoding b-galactosidase, the fusion protein is exclusively found in the nucleoplasm (Fig. 12.4d). We have also fused to GFP a mitochondrial targeting sequence from the yeast cytochrome c oxidase IV protein (amino acids MLSLRQSIRFFKPATRTLCSSR). This sequence confers mitochondrial localization to GFP in both yeast (Fig. 12.4b) and Arabidopsis cells. Kohler et al. (1997) fused a similar

265

266

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

Figure 12.4. Localization of GFP in yeast. Various peptide targeting sequences were fused to GFP in order to direct the protein to different subcellular compartments. The modified proteins were expressed in yeast cells and visualized by confocal microscopy. (a) Unmodified GFP is found throughout the cytoplasm and nucleoplasm. (b) Fusion of N-terminal sequences from yeast cytochrome oxidase B subunit IV results in mitochondrial localization of GFP. (c) N-terminal fusion of the signal sequence from yeast carboxypeptidase Y and C-terminal fusion of the amino acids HDEL results in retention of GFP within the endoplasmic reticulum. (d) N-terminal fusion of a nuclear localization sequence (NLS) from the SV40 T-antigen ensurers nuclear import of GFP, however, only larger forms of GFP are efficiently retained within the nucleus. In this case, NLSGFP has been fused in Escherichia coli b-galactosidase. This results in exclusive localization of the fusion protein within nucleis. The outlines of the yeast cells, obtained using phase contrast optics, are superimposed in this image.

localization sequence to mgfp4 and demonstrated the utility of the encoded fluorescent protein as a precise marker for mitochondria in Arabidopsis. In addition, we fused the yeast carboxypeptidase Y (amino acids MKAFTSLLCGLGLSTTLAKA) and Arabidopsis basic chitinase (amino acids MKTNLFLFLIFSLLLSLSSA) signal sequences to GFP, and have successfully targeted the protein to the secretory pathway (Figs. 12.4c and 12.5). It would be highly advantageous to produce relatively high levels of fluorescence for routine screening of GFP expression in transgenic plants and we tested the targeted forms of GFP in Arabidopsis. The modified genes were placed behind the 35S promoter, introduced into Arabidopsis by Agrobacterium-mediated root transformation (Valvekens et al., 1988), and we tested for localization of the protein and fluorescence intensity in regenerated plants. The one variant that showed a substantial improvement over unmodified GFP was one that was targeted to the endoplasmic reticulum (ER) (Haseloff et al., 1997). This targeted form of GFP contains an N-terminal signal peptide derived from an Arabidopsis vacuolar basic chitinase and the C-terminal amino acid sequence HDEL (Fig. 12.2), to ensure entry into the secretory pathway and retention of the protein within the lumen of the ER. By using this modified gene (mgfp4-ER), it has been possible to regenerate intensely fluorescent and fertile plantlets consistently. Fluorescence within these plants

SUBCELLULAR TARGETING OF GFP

Figure 12.5. Images of 35S-mgfp4-ER transformed seedlings. The ER-localized form of GFP was visualized in transgenic Arabidopsis seedlings using the procedures described for Fig. 12.2. (A) An optical section of the apical meristem showing the junction of the hypocotyl and cotyledons, and (B) an emerging first leaf. (C) Cells within the mesophyll of a cotyledon, packed with mature red autofluorescent chloroplasts. (D) A view of epidermal cells within the hypocotyl, showing the reticulate distribution of GFP within the endomembrane system and the appearance of green fluorescence within maturing plastids. Mature chloroplasts are brightly red autofluorescent in these cells. (E) Cells within the root meristem clearly display the characteristic perinuclear distribution expected for the ER-localized GFP. This perinuclear distribution is also seen in the shoot (panel A). (F) Median longitudinal optical section of a root tip.

could be readily observed by eye using a long wavelength UV lamp. The mgfp4-ER expressing plants were examined by confocal microscopy, and fluorescent protein was found mainly within the endomembrane system. The protein is excluded from the nucleus, shows a perinuclear distribution, and is found associated with the ER that forms a characteristic reticulate network in highly vacuolate cells. In highly cytoplasmic meristematic cells, the nuclei and orientation of cell divisions can be clearly distinguished. Localization of the modified protein to cytoplasmic organelles was also evident, to what appear to be large leucoplasts or proplastids. For example, an optical section of a hypocotyl epidermal cell is shown in Figure 12.5 and this includes a thin portion of cytosol that is pressed between the cell wall and vacuole. Such hypocotyl cells in mgfp4-ER transformed seedlings appear to contain a spectrum of developing plastids that range from the brightly green fluorescent to those that take on a yellow, orange, or red appearance in dual channel confocal micrographs (Fig. 12.6). We presume that this is due to increasing chlorophyll synthesis, and that the green fluorescent plastids may be the maturing precursors of chloroplasts in these cells. These green fluorescent plastids are also found within the chloroplastfree epidermal cells of leaves and cotyledons, but are not found within the underlying mesophyll cells that are packed with mature chloroplasts. It seems likely that these organelles are proplastids and are capable of developing into chloroplasts, but we cannot exclude the possibility that they are some specialized form of leucoplast. The accumulation of mgfp4-ER protein within leucoplasts or developing proplastids, in addition to its entry into the secretory pathway and retention in the endoplasmic reticulum, may indicate misrecognition of the N-terminal signal peptide. Proplastid accumulation of GFP is not seen in the 35S-mgfp4 transformed plants. If the mgfp4-ER encoded

267

268

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

Figure 12.6. Subcellular distribution of mgfp4-ER. The distribution of GFP and GFP-ER is shown in cells of the hypocotyl epidermis of transformed Arabidopsis. The cytosol forms a thin layer at the periphery of these highly vacuolate cells. The cytoplasmic form (mgfp) is excluded from endomembrane components and plastids within the cytosol, and forms a negative stain for these components. A single chlorplast, with its red fluorescent chlorophyll contents, can be seen in this image. Several nonfluorescent cigar-shaped bodies that appear to be some kind of plastids are also evident. In contrast, mgfp-ER is found within the endoplasmic reticulum and unexpectedly within the plastid-like organelles. The distribution of these labeled plastids is mainly limited to epidermal cells of the shoot, and varying degrees of chlorophyll fluorescence can be detected within the organelles, indicating that they may be developing pro-plastids.

signal peptide is inefficiently recognized prior to docking and cotranslational transport of the protein into the lumen of the ER, a proportion of GFP bearing fused terminal sequences may be produced in the cytoplasm. If so, it is possible that the neglected signal peptide may act as a transit sequence for plastid entry. Alternatively, there may be some direct exchange between developing plastids and the endomembrane system. We see no free cytoplasmic fluorescence, and the protein is sorted very efficiently to the ER or to plastids.

IMPROVED MATURATION OF GFP

Figure 12.7. Transgenic 35S-mgfp4-ER Arabidopsis seedlings. Both 5-day-old wild-type (left) and 35S-mgfp4-ER transgenic (right) seedlings were mounted in water on a Leitz DM-IL inverted fluorescence microscope and illuminated with long-wavelength UV light (Leitz-D filter set, excitation BP355–425, dichroic 455, emission LP460). Seedlings were visualized using a 4¥ objective (EF 4/0.12) and a Sony DXC-930P videocamera with F100-MPU framestore. A montage of the entire seedlings was assembled from collected videoimages using Adobe Photoshop.

It is unclear whether the beneficial effects of targeting GFP to the ER are due to increased levels or safer accumulation of mature GFP within cells. For example, if accumulation of fluorescent protein leads to the generation of free radicals in illuminated cells, it is conceivable that removing GFP from the nucleus could protect cells from DNA damage due to such short-range highly reactive species. However, it is also possible that the fusion of peptide targeting sequences may improve the properties of the protein itself, or that the localization of GFP to the lumen of the ER may improve its maturation and accumulation. The maturation of the GFP aproprotein is sensitive to temperature, and the apoprotein readily misfolds under certain conditions (Siemering et al., 1996). The lumen of the ER is known to contain components, such as chaperones and peptidyl prolyl isomerases that aid protein folding (Fischer, 1994), and secretion and retention of GFP within the ER may allow improved formation and accumulation of the mature fluorescent protein. These improvements have allowed us to routinely generate transgenic Arabidopsis plants that contain high levels of GFP fluorescence (Fig. 12.7).

12.7

IMPROVED MATURATION OF GFP

The green fluorescent protein is normally produced within photocytes of the jellyfish A. victoria, and must undergo a series of posttranslational maturation steps to produce the fluorescent form of the protein. Expression of GFP in a number of heterologous systems has been described as poor or variable. For example, strong promoters and decreased incubation temperatures have been required for efficient expression of gfp in mammalian cells (Kaether and Gerdes, 1995; Ogawa et al., 1995; Pines, 1995). Other researchers found that development of fluorescence is favored by a lower incubation temperature during expression of GFP in bacteria (Webb et al., 1995) and yeast (Lim et al., 1995). These

269

270

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

observations suggested that expression of GFP in heterologous cells may be far from optimal. We have clearly demonstrated that maturation of the wild-type GFP is temperature sensitive, due to a defect in the folding of the GFP apoprotein. We have produced mutant forms of GFP that have improved folding and spectral properties (Siemering et al., 1996; Haseloff et al., 1997). These new GFPs are cured of the cryptic intron and are expressed brightly in plant cells. The mgfp4 gene was subjected to random mutagenesis, expressed in E. coli at 37°C and colonies were screened for increased fluorescence. Brighter mutants were isolated, and mapped by recombination with the wild-type mgfp4 gene. Sequencing of the brightest mutant (GFPA) revealed two amino acid differences, V163A and S175G. The mutant GFP produced up to 35-fold increased fluorescence in bacterial cells, while the difference in protein levels was not nearly enough to account for this. The result suggested that a large proportion of wild-type GFP that is expressed in cells at 37°C is nonfluorescent. Experiments with a GFP–nucleoplasmin fusion protein have indicated that maturation of GFP to the fluorescent form may be sensitive to temperature during expression in the yeast. S. cerevisiae (Lim et al., 1995). To test whether the same could be true of expression in E. coli and whether the substitutions present in GFPA enhance maturation by suppressing any such sensitivity, we examined expression of GFP and GFPA over a range of different temperatures. Strains expressing GFP or GFPA were grown overnight at temperatures ranging between 25°C and 42°C. For each culture, the fluorescence values were measured and normalized against the amount of recombinant protein present in the cells to give a measure of the proportion of intracellular GFP that is fluorescent at different temperatures. The proportion of GFP that is fluorescent steadily decreases with increasing incubation temperatures (Fig. 12.8), indicating that either mature GFP or the maturation pathway leading to its formation is temperature sensitive. Mature GFP is a highly stable protein whose fluorescence in vitro in unaffected by temperatures up to 42°C (Bokman and Ward, 1981), and we confirmed that the fluorescence of the mature protein is unaltered in bacterial cells at 42°C. Therefore, higher incubation temperatures must interfere with the posttranslational maturation of GFP, rather than causing inactivation of the mature protein. We confirmed that expression of GFP is also temperature sensitive in yeast and demonstrated that this is suppressed by the substitutions present in GFPA. These results indicate that the thermosensitivity of GFP maturation may be a common phenomenon that can be suppressed by the amino acid substitutions present in GFPA (Siemering et al., 1996). The posttranslational maturation of GFP presumably involves initial folding of the apoprotein into an active conformation, to allow the cyclization and oxidation reactions that form the chromophore (Cody et al., 1993; Heim et al., 1994; Cubitt et al., 1995). The mature protein must then be correctly folded to maintain its fluorescent properties, to protect the chromophore from solvent effects (Ward et al., 1980). In principle, any of these processes could be sensitive to temperature and thus be responsible for the observed thermosensitivity of GFP during maturation. Since the oxidation reaction involved in GFP chromophore formation appears to require molecular oxygen (Heim et al., 1994), oxidation rates can be measured after growth under anaerobic conditions by measuring the development of fluorescence after admission of air. We measured the rates of oxidation of GFP and GFPA expressed in anaerobically grown yeast at both 25°C and 37°C. The time constant measured for the oxidation of GFP at 37°C (5.9 ± 0.1 min) was found to be approximately threefold faster than that measured at 25°C, indicating that the posttranslational oxidation of the GFP chromophore is not the step responsible for the temperature sensitivity of maturation. In confirmation of this

271

IMPROVED MATURATION OF GFP

A

2500

mGFP-4 2000 fluorescence (normalised)

mGFP-A

1500

1000

500

0 25

30

37

42

temperature (∞C)

Figure 12.8. Improved mutants of GFP. (A) Improved thermotolerance of GFP. Bacterial cells expressing GFP and GFPA (V163A, S175G) were grown at different temperatures. The GFP fluorescence values were measured and normalized with respect to the amount of intracellular recombinant protein for cultures grown at 25°C, 30°C, 37°C, and 42°C (Siermering et al., 1996). (B) Excitation and emission spectra of GFP, GFPA (V163A, S175G), and GFP5 (V163A, 1167T, S175G). Protein concentrations were 23.5 mg ml-1 in PBS (pH 7.4). All spectra have been normalized to a maximum value of 1.0 (Siemering et al., 1996).

272

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

conclusion, the time constants derived for GFPA were somewhat slower than those measured for GFP (Siemering et al., 1996). In contrast, we found that the folding of GFP is clearly temperature-sensitive and the substitutions present in GFPA enhance proper folding at increased temperatures. We examined the solubilities of the two proteins during expression in E. coli at 25°C and 37°C. Fluorescence was found almost exclusively in the soluble fraction. At 25°C, both GFP and GFPA were found predominantly in the soluble fraction, indicating that proper folding of both proteins is relatively efficient at this temperature. At 37°C, however, the majority of GFP was found as nonfluorescent protein in the insoluble fraction, whereas most of GFPA was still present in the soluble fraction. To obtain information on which species in the maturation pathway of GFP misfolds at higher temperatures, we examined the absorption spectrum of denatured protein isolated from inclusion bodies. If GFP undergoes cyclization of the chromphore prior to aggregation, protein from inclusion bodies should show an absorption in the near UV/blue region that is characteristic of the GFP chromophore in either the mature or chemically reduced state (Ward et al., 1980; Inouye and Tsuji, 1994). On the other hand, if unmodified GFP (apo-GFP) is the aggregating species, no such absorption should be observable in this region. GFP was purified from the inclusion bodies of bacterial cells grown at 37°C and, as a positive control, from the soluble fraction of cells grown at 25°C. Protein derived from cells grown at 25°C showed a characteristic absorption peak similar to that of acid-denatured GFP (Ward et al., 1980). By contrast, protein purified from inclusion bodies of cells grown at 37°C showed no such absorption, indicating that the aggregating species has not formed a chromophore. Taken together, the results indicate that the temperature sensitivity of GFP maturation is due primarily to the failure of the unmodified apoprotein to fold into its catalytically active conformation at higher temperatures. Furthermore, the amino acid substitutions present in GFPA suppress this defect by enhancing proper folding at elevated temperatures.

12.8

MODIFICATION OF FLUORESCENCE SPECTRA

The fluorescence excitation spectrum of GFP and GFPA exhibits peaks at wavelengths of 395 and 475 nm, with the 395-nm peak predominating. This property is useful for simple detection of the protein using a long-wavelength UV source. Ultraviolet illumination is not efficiently detected by the human eye and a suitable long-wavelength UV lamp can be used to excite GFP for simple observation of transformed plant material without obscuring the green emission. However, efficient blue light excitation (~470 nm) is essential for use with imaging devices such as confocal microscopes or cell sorters, which are equipped with argon laser sources. Recently, it has been demonstrated that the relative amplitudes of the excitation peaks of GFP can be altered by means of mutagenesis (Heim et al., 1994; Delagrave et al., 1995; Ehrig et al., 1995; Heim et al., 1996). These mutations appear to affect the microenvironment of the chromophore so as to influence the equilibrium between two spectroscopic states of the chromophore (Heim et al., 1994; Ehrig et al., 1995). One of these mutations, I167T, has been shown to increase the amplitude of the 475-nm excitation peak relative to that of the 400-nm peak (Heim et al., 1994). We recombined the I167T substitution with the substitutions present in GFPA to increase the amplitude of the 475-nm peak relative to the 395-nm excitation peak, to produce a variant (GFP5), which has two excitation

MODIFICATION OF FLUORESCENCE SPECTRA

peaks (maxima at 395 and 473 nm) of almost exactly equal amplitude and an emission spectrum (lmax = 507 nm) largely unchanged from that of wild type. The GFP5 variant retains a thermotolerant phenotype, and bacterial cells grown at 37°C fluoresce 39- and 111-fold more intensely than cells expressing GFP, when excited at 395 and 473 nm, respectively. The broad excitation spectrum of GFP5 allows both efficient UV and blue light excitation of the protein. For example, the expression of gfp5 gene fusions can be rapidly scored after transformation of microbial colonies or plant tissues by simple inspection with a UV lamp. The same material is well-suited for laser scanning confocal microscopy. In addition, plants are highly autofluorescent, and the use of a dual-wavelength excitation mutant like GFP5 also enables faint signals to be easily distinguished from autofluorescence during microscopy, by alternating the excitation sources. For example chloroplasts are intensely fluorescent but are less efficiently excited by UV light. We routinely use longwavelength UV excitation for visual and microscopic screening of transformed tissues. Autofluorescence can also be an advantage. For example, UV light excites a faint blue fluorescence in Arabidopsis cell walls, and this “counterstain” allows roots growing in agar culture to be easily located and scored for GFP fluorescence. Recently, screened we have several thousand Arabidopsis transformants for root specific “enhancer-trap” expression patterns, and this feature was very useful (J. Haseloff and S. Hodge, unpublished results). In contrast, widely used GFP variants that contain the S65T mutation (Heim et al., 1995; Cormack et al., 1996) provide optimized properties for blue light excitation, but are not useful for detection by long-wavelength UV light. It is possible to manipulate the fluorescence spectra of GFPA by introducing additional substitutions into the protein without deleteriously affecting its improved folding characteristics. The Y66H substitution dramatically blue shifts both the excitation and emission spectra of GFP to give a “blue fluorescent protein” (Heim et al., 1994). The GFPA containing the Y66H substitution was found to have identical fluorescence spectra to those of the corresponding GFP (Y66H) protein (excitation maximum = 384 nm, emission maximum = 448 nm), and it gave rise to 29-fold more fluorescence when expressed at 37°C and threefold more fluorescence when expressed at 25°C. In addition, a number of workers obtained GFP variants that show brighter fluorescence in heterologous cell types, and it is likely that the improved properties of these proteins is due largely to improved folding. For example, the V163A mutation present in GFPA has also been generated independently by at four different groups (Crameri et al., 1996; David and Vrestra, 1996; Heim et al., 1996; Kohler et al., 1997) and this residue may play a pivotal role in folding of the protein. Cormack et al. (1996) introduced random amino acid substitutions throughout the 20 residues flanking the chromophore of GFP. They used fluorescence activated cell sorting to select variants that fluoresced 20- to 35-fold more intensely than wild type, and noted that the mutant proteins had improved solubility during expression in bacteria. The mutant proteins persumably have improved folding properties. One of these variants [GFPmut1 (Cormack et al., 1996)] contains two amino acid differences, F64L and S65T, located within the central a helix of the protein, adjacent to the chromophore. The V163A and S175G mutations that we have isolated are positioned on the outer surface of the protein (Ormö et al., 1996; Yang et al., 1996) and recombination of these two sets of mutations appears to result in markedly improved fluorescence in bacterial, plant and animal cells (Zoenicka-Goetz et al., 1996, 1997; J. Haseloff and K. R. Siemering, unpublished results). It is possible that the mutations affect separate steps of the folding or maturation process, and that their benefit is additive.

273

274

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

12.9

MODIFIED GFP GENES FOR PLANT EXPRESSION

Expression of the wild-type gfp gene has given poor results in a number of plant systems, and we have found it necessary to alter the gene for our experiments with transgenic Arabidopsis plants. As outlined above, we have (a) altered codon usage to remove a cryptic plant intron, (b) added peptide sequences to allow targeting of the protein to the lumen of the endoplasmic reticulum, and mutated the protein to (c) improve folding of the apo-GFP during posttranslational maturation (V163A, S175G), and (d) provide equalised UV and blue light excitation (I167T). These alterations have all been incorporated into a single highly active form of the gene (mgfp5-ER), which we now routinely use for monitoring gene expression and marking cells in live transgenic plants (Siemering et al., 1996; Haseloff et al., 1997). Removal of the cryptic intron appears to be essential for gfp expression in Arabidopsis, and other workers have observed improved expression in plants using gfp genes containing “humanized” or synthetic codon usage (Chiu et al., 1996; Pang et al., 1996) Altered codon usage of the gfp gene appears to be a crucial requirement for efficient expression in plant cells. Improvements in the folding, spectral properties, and subcellular localization of the protein provide secondary improvements that allow the accumulation of high levels of fluorescent protein in plant cells.

12.10

IMAGING GFP IN PLANT CELLS

GFP can be visualized directly in living plant tissue, unlike commonly used markers such as b-glucuronidase, which requires a prolonged and lethal histochemical staining procedure (Jefferson et al., 1987). The GFP is therefore finding application in three broad areas: (1) for the dynamic visualization of labeled protein within the cells, and at a larger scale, (2) for the selective labeling and monitoring of whole plant cells within growing plant tissue, and (3) for the identification of individual transgenic plants expressing GFP. For example, different peptide domains can be fused to GFP to allow the decoration of particular structures within cells and/or to observe the subcellular distribution of the fusion protein. In addition, use of an active GFP marker gene allows transgenic cells to be scored by simple observation during a plant transformation experiment, throughout regeneration to the adult plant and its progeny. The use of tissue specific promoters to drive expression of GFP also allows the selective labeling of particular cell types within intact transformed plants. In these cases, it is beneficial to express GFP at high levels within the marked cells to aid detection, and to minimize any deleterious effects of GFP expression. We have found the optimized mgfp5-ER gene very useful for this type of experiment. The dynamic properties of labeled cells or subcellular features can be resolved at high resolution in whole plant tissues using fluorescence microscopy techniques, however, the use of intact tissue imposes some additional constraints on the imaging process. Intact plant tissue proves a difficult subject for fluorescence microscopy as it consists of deep layers of highly refractile cell walls and aqueous cytosol and contains various autofluorescent and light scattering components. There are two approaches to the difficulties imposed by these conditions: to fix and to clear the tissue with a high refractive index mounting medium, or to directly image living tissue using suitably corrected microscope optics. In our experience, it has proved difficult to effectively clear Arabidopsis wholemounts without causing artifacts or losing GFP fluorescence, and there are considerable advantages to working with living tissues. Thus we have mainly pursued the second

IMAGING GFP IN PLANT CELLS

approach. The natural autofluorescence and depth of intact plant tissue means that out of focus blur often obscures high-magnification views obtained with a conventional epifluorescence microscope. However, the technique of laser scanning confocal microscopy can be used to optically section GFP expressing plant tissues. Confocal imaging allows precise visualization of fluorescent signals within a narrow plane of focus, with exclusion of outof-focus blur, and the technique permits the reconstruction of three dimensional (3D) structures from serial optical sections. Arabidopsis seedlings can simply be mounted in water for microscopy, and examined using a long-working distance water immersion objective to minimize the effects of spherical aberration when focusing deep into an aqueous sample (Haseloff et al., 1995). Young seedlings (3–7 days old) can be grown on agar culture media, and then placed in a drop of water (100–200 ml) on a glass slide. A glass coverslip is lowered gently to flatten and cover the seedling. Even with the use of a specialized water immersion objective such as the Nikon 60¥ planapochromat, N.A. 1.2 (working distance 220 mm), image quality degrades rapidly for optical sections deeper than 50–80 mM within the tissue. Ideally, the tissue of interest should be positioned immediately below the cover slip and depression slides should be avoided unless this is ensured. Despite these limitations, the small size of Arabidopsis seedlings allows very useful imaging and, for example, median longitudinal optical sections can be easily obtained from intact root tips (e.g., Fig. 12.9).

Figure 12.9. Dual-channel imaging of GFP/chlorophyll/propidium iodide. A BioRad MRC-600 confocal microscope equipped with an argon–krypton mixed-gas laser and K1/K2 filter blocks was used for green and ref fluorescence imaging of GFP expressing Arabidopsis seedlings. (a) Separate images of a 35S-mgfp4 transformed cotyledon mesophyll cell showing the green fluorescence channel with the GFP signal distributed throughout the nucleoplasm and cytoplasm, the red fluorescent signal of the chloroplasts, and the combined dual channel image. There is little spillover between the two channels. (b) Separate green, red, and combined fluorescence images are also shown for a propidium iodide stained root tip of a GFP expressing enhancer trap line J0571 (Haseloff and Hodge, unpublished results). The mgfp5-ER gene is expressed strongly in the root cortex and endodermis of this line. Propidium iodide provides a distinct counterfluor that outlines cells in the living root tip.

275

276

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

The blue 488-nm wavelength line of the commonly used argon ion or krypton–argon lasers is ideal for exciting GFP, and this can be used in combination with a fluorescein/rhodamine or Texas Red® filter set for dual channel imaging of GFP and chlorophyll for photosynthetic tissues. For nonphotosynthetic tissues, a fluorescent counterstain that can be distinguished from GFP is often very useful. For example, Arabidopsis seedlings can be placed in a solution of 10 mg ml-1 propidium iodide for 5–20 min, before being directly mounted in water for confocal microscopic examination of roots. Propidium iodide is red fluorescent and highly charged and does not enter living cells. It stains the walls of living cells within the root tip and fills dead cells (van der Berg et al., 1995) (Fig. 12.10). In a similar way the red fluorescent dye nile red can be used to stain neutral lipids, and rhodamine 6B can be used to stain the casparian strip and lignified cells within living roots (J. Haseloff, unpublished results).

12.11

VISUALIZING SUBCELLULAR DYNAMICS

The expression of GFP within an organism produces an intrinsic fluorescence that colors normal cellular processes, and high-resolution optical techniques can be used noninvasively to monitor the dynamic activities of these living cells. For time-lapse studies, it is very important that GFP fluorescence be bright, to minimize levels of illumination that can cause phototoxicity and photobleaching during observation. The modified mgfp5-ER gene that is described above has proved very useful for generating highly fluorescent transgenic Arabidopsis plants that are suitable for intensive time-lapse studies. During Texas Red® is a registered trademark of Molecular Probes, Inc., Eugene, OR.



Figure 12.10. (opposite) Time-lapse confocal microscopy of subcellular processes. A transgenic

Arabidopsis seedling expressing the mgfp5-ER gene was mounted in water and a small segment of the hypocotyl epidermis was observed using a BioRad MRC-600 laser scanning confocal microscope. The laser light was attenuated by 99% using a neutral density filter, and the confocal aperture was stopped down. Two channel, single scan images were collected at the rate of 1/2 s for 20 min, and transferred to an Apple Macintosh computer. The large data file was then converted to full-color numbered PICT files using the program PicMerge, and finally converted to a Quicktime movie for analysis and videorate playback. A section corresponding to 4.5 min of the original observation was chosen and representative frames are presented here. Each frame is marked with the time (minute: second) that had elapsed from the first chosen frame. Two schematic diagrams are shown. A key for the identities of cellular structures and organelles is shown in diagram A. Nuclei (N), chloroplasts (C), endoplasmic reticulum (ER), proplastids or leucoplasts (P), and the position of the cell wall (CW) are shown (scale bar = 10 m). In the second diagram (B), the positions of proplastids throughout the 4.5-min period of the experiment is shown. The plastids were located in each frame, and their cumulative positions within the cell were plotted frame by frame, indicated by black dots. The position of one plastid is also plotted, with a series of red lines representing the successive orientations of the long axis of the plastid. This particular organelle is indicated by a white cross on the timed confocal images, from its appearance in the field of view at 0:20. The images 0:00–0:25 each contain an arrow that indicates a ring-like feature within the ER that provides a morphological landmark. Image 3:40 contains an arrow that indicates the formation of a transient filamentous structure that appears associated with rapid vesicular and plastid traffic.

VISUALIZING SUBCELLULAR DYNAMICS

277

278

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

confocal microscopy experiments, we have routinely observed high rates of cytoplasmic streaming within living specimens, and we have used short-term time-lapse observations to gain a better understanding of the relative movements of cellular components. Arabidopsis seedlings that expressed the mgfp5-ER gene at high levels were simply mounted in water for confocal microscopy, which allowed observation for up to 2 h. Hypocotyl epidermal cells form ideal specimens for viewing the various components of the cytoplasm. The cells are large, highly vacuolate and surface borne. An extremely thin layer of cytoplasm is squeezed between the wall and the vacuole of these cells. The thinness of the layer greatly limits the movement of cytoplasmic components to within a single plane of focus of the microscope, and objects can be rapidly tracked across a portion of the cell without the need for refocusing. A seedling can easily be mounted so that the hypocotyl is pressing closely against the microscope coverslip, and the layer of cytoplasm beneath the outer wall of an epidermal cell will be only a few microns from the surface, allowing high optical resolution. To follow rapid movements in cell it is necessary to use a correspondingly fast sampling rate. We have collected time-lapse confocal images at up to two frames per second, which requires almost continual laser scanning with a BioRad MRC-600 microscope. Living specimens have been examined for up to an hour without appreciable phototoxic or bleaching effects, but this is only possible with bright samples, which allow attenuation of the exciting laser light. A short segment of a time-lapse experiment is shown in Fig. 12.11. A section of hypocotyl epidermis was monitored at a rate of 0.5 frames per second for about 20 min and representative confocal images are shown for a 4.5-min period. Cellular components are clearly recognizable in the optical sections, and their identity is indicated in a schematic diagram (Fig. 12.11a). The cells contain green fluorescent proplastids and highly reticulate endomembranes. The nuclei are outlined due to the peripheral distribution of the ER, and the reticulate surfaces of a nucleus can be seen in the cell that is central to the field of view. A cross section of a nucleus can also be seen in the adjacent lower cell. Chloroplasts are red autofluorescent, and characteristically small and spheroid in these hypocotyl epidermal cells. When a time course of images is played at videorate, proplastids and what appears to be vesicular material move vigorously and erratically through the cells. The distribution of all proplastids was plotted frame by frame though this experiment and the path of one example is shown in Fig. 12.11b. These plastids more with uneven velocities, up to 20 m/s, along irregular paths that may correspond to underlying cytoskeletal elements such as actin. In contrast, the endoplasmic reticulum, which is presumably associated with cortical microtubules, undergoes relatively slower rearrangement. A relatively stable feature of the ER is indicated with an arrow in Fig. 12.11, panels 0:00–0:25, while nearby proplastids undergo substantial movement. Chloroplasts and nuclei moved only slowly during the 20-min time course of the experiment. These cells contain an ER retained form of GFP, and we expect the protein to be cycled in vesicles between the lumen of the ER and the cis golgi. A rapid and irregular movement of small vesicle-like particles is seen throughout cells during the time course. Although these small movements are difficult to see in still images, we also see the transient formation of extended filamentous structures (Fig. 12.11, panel 3:40), which are comprised of a larger amount of this fluorescent material, and are associated with rapid movement of both vesicularlike material and proplastids. The location of the cis golgi in these micrographs is unclear, although small regions of punctate fluorescence can be seen associated with endomembranes.

Figure 12.11. (opposite) Time-lapse confocal microscopy of root development. Seeds of the Arabidopsis enhancer trap line J0571 (J. HAseloff and S. Hodge, unpublished results) were germinated and grown in agar medium on a coverglass. After 10 days of growth, an emerging lateral root was visualized by confocal time-lapse microscopy. The root tip was imaged through the coverglass of the tissue culture vessel. A median longitudinal optical section was collected every 2 min over a 6-h period. (a) Representative frames from a 320-min period are shown, labeled with the time of collection (minute: second). Cell divisions and growth of the labeled cortex and endodermis cell layers are evident. Individual cells in the process of mitotic division are arrowed. One endodetermis cell is marked with an asterisk, and its behavior is shown in more detail. (b) Frames collected at shorter intervals are shown for the marked cell. The times of image collection are indicated (minute: second). The ER-localization of the GFP marker allows clear visualization of nuclear division and phragmoplast formation in these cells.

280

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

12.12

MARKING CELLS WITHIN THE PLANT

The expression of GFP can be limited to particular cell types within a plant, to provide a means for visualizing the behavior of individual cells within the living organism. We have developed a scheme for targeted gene expression in plants, which is based on a method widely used in Drosophila (Brand and Perrimon, 1993). We have randomly inserted a gene for a foreign transcription activator (a derivative of the yeast GAL4 gene) into the Arabidopsis genome using Agrobacterium tumefaciens mediated transformation. We have thus generated a library of Arabidopsis lines, which each express the GAL4 derivative in a particular pattern, dependent on adjacent genomic DNA sequences. The inserted DNA has also been engineered to contain a GAL4-responsive mgfp5-ER gene, and so interesting patterns of GAL4 gene expression are immediately and directly visible, with each GAL4 expressing cell marked by green fluorescence within the endoplasmic reticulum. Importantly, GAL4 expression within these lines allows precise targeted ectopic gene expression. A chosen target gene can be cloned under the control of a GAL4 responsive promoter, separately transformed into Arabidopsis, and maintained silently in the absence of GAL4. Genetic crossing of this single line with any of the library of GAL4 containing lines allows a specific activation of the target gene in particular tissue and cell types, and the phenotypic consequences of misexpression, including those lethal to the organism, can be conveniently studied (J. Haseloff, S. Hodge, and H. M. Goodman, unpublished results). We have used in vivo detection of GFP to develop a new and efficient enhancer-trap screening procedure. As our particular interest is in the cells of the Arabidopsis root tip, we have modified the plant transformation protocol to include an auxin induction of roots from regenerating shootlets. More than 7500 transformants were then generated, planted in grid patterns in sterile culture dishes and directly screened by fluorescence microscopy for GAL4-driven GFP expression within roots. Several hundred lines with interesting patterns of root expression were chosen, documented, transferred to soil and grown to seed, to both amplify and self-hybridize the lines. Consequently, we have a collection of 250 Arabidopsis lines with distinct and stable patterns of GAL4 and GFP expression in the root. These GAL4-GFP lines provide a valuable set of markers, where particular cell types are tagged and can be visualised with unprecedented ease and clarity in living plants. The GAL4-GFP enhancer-trap screen was designed to yield markers for the Arabidopsis root meristem, which is our choice of a model system. The simple and wellcharacterized architecture of the root (Dolan et al., 1993) enables simple analysis of GAL4mediated perturbations of cell fate within the meristem. The Arabidopsis root meristem consists of a plate of quiescent cells surrounded by initials that divide to produce distal root cap cells, and also lay down continuous cell files proximally. Behind the tip, the newly formed cells of the root undergo differentiation and expansion to build a conserved arrangement of cell types within the mature root. We have generated GAL4-GFP lines that precisely mark particular cell types within the meristem, and one example is shown in Figure 12.11. Arabidopsis line J0571 exhibits GAL4 directed expression of GFP within the cortex and endodermis of the root, including the initials shared by these two cell files. Five-day-old seedlings can be briefly counterstained with 10 mg ml-1 propidium iodide and mounted in water for confocal microscopy. Optical sectioning allows very simple and precise imaging of the GFP labeled cells within the root meristem (Fig. 12.11a). The behavior of these cells within the developing root meristem can be observed using timelapse techniques. The GFP expressing seedlings can be planted in sterile agar media and grown in coverslip-based vessels. The roots grow down through the media and then along the surface of the coverslip. The roots are then ideally positioned for microscopic imaging

CONCLUSIONS

through the base of the vessel. A series of images are shown in Fig. 12.11b that illustrate 2 h during the growth of a root tip of Arabidopsis line J0571. Confocal optical sections were collected at 2-min intervals. The cortical and endodermal cell files and their initials are clearly seen due to the expression of GAL4 driven mgfp5-ER in this line. The localization of GFP to the endoplasmic reticulum, and its consequent perinuclear distribution, ensures that the cell nuclei are clearly evident in these meristematic cells. In addition, the processes of cell division can be seen within the living plant. The breakdown of the nuclear membrane, segregation of chromosomes and formation of the daughter nuclei and cell wall plate are reflected in changes of the distribution of the ER localized GFP. Also, the cell nuclei appears to possess a larger volume prior to cell division, consistent with an extra, newly replicated DNA complement. This characteristic may be useful for scoring DNA replication within living cells. Such GFP expressing lines allow the simple, noninvasive observation of events within living plants at an unprecedented level of detail. The GFP can now be used as a cellular marker to illuminate the defective behavior of mutant plants, or the perturbations induced by reverse genetic techniques.

12.13

CONCLUSIONS

In order to overcome problems with the expression of GFP in plant cells, and with the safe accumulation and detection of GFP in whole Arabidopsis plants, we have engineered improvements to the gfp gene. The modified gene contains (a) altered codon usage to remove a cryptic plant intron (b) added peptide sequences to allow targeting of the protein to the lumen of the endoplasmic reticulum, and (c) mutations that improve folding of the apoprotein during posttranslational maturation (V163A, S175G) and provide equalized UV and blue light excitation (I167T). This highly modified variant (mgfp5-ER) is proving useful as a safe and bright marker in transgenic plants. We expect that the mgfp5-ER gene and its derivatives will also be useful in work with transgenic fungi and animals, where at least some similar problems may be encountered. A major use for GFP will be as a replacement for the b-glucuronidase gene, which is widely used as a reporter for promoter and gene fusions in transformed plants. The GUS gene product can be localized or quantified using histochemical techniques, but these are generally destructive tests (Jefferson et al., 1987). In contrast, GFP can be directly seen in living tissues. For example, high levels of fluorescence intensity are obtained in GFP transformed bacterial and yeast colonies, allowing simple screening for GFP expression with the use of a hand-held UV lamp. Such an assay for gene expression in living plants will be a very useful tool for plant transformation and breeding experiments. Many transformation techniques give rise to regenerating tissues that are variable or chimeric, and require testing of the progeny of the primary transformants. Potentially, GFP expressing tissues could be monitored using in vivo fluorescence, avoiding any need for destructive testing, and the appropriate transformants could be rescued and directly grown to seed. Similarly, in vivo fluorescence will be an easily scored marker for field testing in plant breeding, allowing transgenes linked to the GFP gene to be easily followed, and provide a potential alternative to antibiotic resistance markers. Unlike enzyme markers, GFP can be visualized at high resolution in living cells using confocal microscopy. The images are not prone to fixation or staining artifacts, and can be of exceptional clarity. Moreover, the activities of living cells, such as cytoplasmic streaming, are clearly evident during microscopy. Ordinarily, movement within a sample

281

282

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

is a nuisance, placing constraints on the use of sometimes lengthy techniques for noise reduction during confocal microscopy, such as frame averaging. However, we have shown that it is also possible to monitor dynamic events by time-lapse confocal microscopy, and this combination of a vital fluorescent reporter with high-resolution optical techniques shows much promise for use in cell biological and physiological experiments. Genetic systems such as that of Arabidopsis provide a large resource of potentially informative mutants, and there has been much recent improvement in techniques for determining the molecular basis of a particular phenotype. The use of fluorescent proteins will provide further tools for examining the biology of mutant cells. The ability to simply and precisely monitor both particular cells and subcellular structures that have been highlighted with a fluorescent signal will improve both the screening for particular abnormal phenotypes and the characterization of dynamic process.

REFERENCES Baulcombe, D. C., Chapman, S., and Cruz, S. S. (1995). Jellyfish green fluorescent protein as a reporter for virus infections. Plant J. 7:1045–1053. Bokman, S. H., and Ward, W. W. (1981). Renaturation of Aequorea green fluorescent protein. Biochem. Biophys. Res. Commun. 101:1372–1380. Brand, A. H., and Perrimon, N. (1993). Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118:401–415. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263:802–805. Chiu, W. L., Niwa, Y., Zeng, W., Hirano, T., Kobayashi, H., and Sheen, J. (1996). Engineered GFP as a vital reporter in plants. Curr. Biol. 6:325–330. Cody, C. W., Prasher, D. C., Westler, W. M., Prendergast, F. G., and Ward, W. W. (1993). Chemical-structure of the hexapeptide chromophore of the Aequorea green-fluorescent protein. Biochemistry 32:1212–1218. Cormack, B. P., Valdivia, R. H., and Falkow, S. (1996). FACS-optimized mutants of the green fluorescent protein (GFP). Gene 173:33–38. Crameri, A., Whitehorn, E. A., Tate, E., and Stemmer, W. P. C. (1996). Improved green fluorescent protein by molecular evolution using dna shuffling. Nat. Biotechnol. 14:315–319. Cubitt, A. B., Heim, R., Adams, S. R., Boyd, A. E., Gross, L. A., and Tsien, R. Y. (1995). Understanding, improving and using green fluorescent proteins. Trends Biochem. Sci. 20:229–233. Davenport, D., and Nichol, J. A. C. (1955). Luminescence in Hydromedusae. Proc. R. Soc., Ser. B. 144:399– 411. David, S., and Viestra, R. (1996). Soluble derivatives of green fluorescent protein (GFP) for use in Arabidopsis thaliana. Weeds World 3:43–48. Delagrave, S., Hawtin, R. E., Silva, C. M., Yang, M. M., and Youvan, D. C. (1995). Red-shifted excitation mutants of the green fluorescent protein. Bio-Technology 13:151–154. Dolan, L., Janmaat, K., Willemsen, V., Linstead, P., Poethig, S., Roberts, K., and Scheres, B. (1993). Cellular organisation of the Arabidopsis thaliana root. Development 119:71–84. Ehrig, T., Okane, D. J., and Prendergast, F. G. (1995). Green fluorescent protein mutants with altered fluorescence excitation spectra. FEBS Lett. 367:163–166. Fischer, G. (1994). Peptidyl-prolyl isomerases and their effectors. Angew. Chem. Int. Ed. Engl. 33:1415–1436. Goodall, G. J., and Filipowicz, W. (1989). The AU-rich sequences present in the introns of plant nuclear pre-mRNAs are required for splicing. Cell 58:473–483.

REFERENCES

Goodall, G. J., and Filipowicz, W. (1991). Different effects of intron nucleotide composition and secondary structure on pre-mRNA splicing in monocot and dicot plants. EMBO J. 10:2635–2644. Grebenok, R. J., Pierson, E., Lambert, G. M., Gong, F. C., Afonso, C. L., Haldeman-Cahil, R., Carrington, J. C., and Galbraith, D. W. (1997). Green fluorescent protein fusions for efficient characterization of nuclear targeting. Plant J. 11:573–586. Green, M. R. (1991). Biochemical mechanisms of constitutive and regulated pre-mRNA splicing. Annu. Rev. Cell Biol. 7:559–599. Haas, J., Park, E. C., and Seed, B. (1996). Codon usage limitation in the expression of hiv-1 envelope glycoprotein. Curr. Biol. 6:315–324. Hanley, B. A., and Schuler, M. A. (1988). Plant intron sequences: evidence for distinct groups of introns. Nucleic Acids Res. 16:7159–7176. Haseloff, J., and Amos, B. (1995). GFP in plants. Trends Gen. 11:328–329. Haseloff, J., Siemering, K. R., Prasher, D. C., and Hodge, S. (1997). Removal of a cryptic intron and subcellular localisation of green fluorescent protein are required to mark transgenic Arabidopsis plants brightly. Proc. Natl. Acad. Sci. USA 94:2122–2127. Hebsgaard, S. M., Korning, P. G., Tolstrup, N., Engelbrecht, J., Rouze, P., and Brunak, S. (1996). Splice site prediction in Arabidopsis thaliana pre-mRNA by combining local and global sequence information. Nucleic Acids Res. 24:3439–3452. Heim, R., Prasher, D. C., and Tsien, R. Y. (1994). Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proc. Natl. Acad. Sci. USA 91:12501–12504. Heim, R., and Tsien, R. Y. (1996). Engineering green fluorescent protein for improved brightness, longer wavelengths, and fluorescence resonance energy transfer. Curr. Biol. 6:178–182. Heinlein, M., Epel, B. L., Padgett, H. S., and Beachy, R. N. (1995). Interaction of tobamovirus movement proteins with the plant cytoskeleton. Science 270:1983–1985. Hu, W., and Cheng, C. L. (1995). Expression of Aequorea green fluorescent protein in plant cells. FEBS Lett. 369:331–334. Inouye, S., and F. I. Tsuji (1994). Evidence for redox forms of the Aequorea green fluorescent protein. FEBS Lett. 351:211–214. Jefferson, R. A., Kavangh, T. A., and Bean, M. W. (1987). GUS fusions: b-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J. 6:3901–3907. Kaether, C., and Gerdes, H. H. (1995). Visualization of protein-transport along the secretory pathway using green fluorescent protein. FEBS Lett. 369:267–271. Kohler, R. H., Zipfel, W. R., Webb, W. W., and Hanson, M. R. (1997). The green fluorescent protein as a marker to visualize plant mitochondria in vivo. Plant J. 11:613–621. Langford, C. J., Klinz, F.-J., Donath, C., and Gallwitz, D. (1984). Point mutations identify the conserved, intron-contained TACTAAC box as an essential splicing signal in yeast. Cell 36:645–653. Lim, C. R., Kimata, Y., Oka, M., Nomaguchi, K., and Kohno, K. (1995). Thermosensitivity of green fluorescent protein fluorescence utilized to reveal novel nuclear-like compartments in a mutant nucleoporin nsp1. J. Biochem. 118:13–17. Liu, H. X., and Filipowicz, W. (1996). Mapping of branchpoint nucleotides in mutant pre-mRNAs expressed in plant cells. Plant J. 9:381–389. Luehrsen, K. R., Taha, S., and Walbot, V. (1994). Nuclear pre-mRNA processing in higher plants. Prog. Nucleic Acid Res. Mol. Biol. 47:149–193. Niedz, R. P., Sussman, M. R., and Satterlee, J. S. (1995). Green fluorescent protein—an in vivo reporter of plant gene-expression. Plant Cell Rep. 14:403–406. Ogawa, H., Inouye, S., Tsuji, F. I., Yasuda, K., and Umesono, K. (1995). Localization, trafficking, and temperature-dependence of the Aequorea green fluorescent protein in cultured vertebrate cells. Proc. Natl. Acad. Sci. USA 92:11899–11903.

283

284

THE USES OF GREEN FLUORESCENT PROTEIN IN PLANTS

Ormö, M., Cubitt, A. B., Kallio, K., Gross, L. A., Tsien, R. Y., and Remington, S. J. (1996). Crystal structure of the Aequorea victoria green fluorescent protein. Science 273:1392–1395. Pang, S. Z., DeBoer, D. L., Wan, Y., Ye, G., Layton, J. G., Neher, M. K., Armstrong, C. L., Fry, J. E., Hinchee, M. A. W., and Fromm, M. E. (1996). An improved green fluorescent protein gene as a vital marker in plants. Plant Physiol. 112:893–900. Pines, J. (1995). GFP in mammalian cells. FEBS Lett. 11:326–327. Plautz, J. D., Day, R. N., Dailey, G. M., Welsh, S. B., Hall, J. C., Halpain, S., and Kay, S. A. (1996). Green fluorescent protein and its derivatives as versatile markers for gene expression in living Drosophila melanogaster, plant and mammalian cells. Gene 173:83–87. Prasher, D. C., Eckenrode, V. K., Ward, W. W., Prendergast, F. G., and Cormier, M. J. (1992). Primary structure of the Aequorea Victoria green-fluorescent protein. Gene 111:229–233. Reichel, C., Mathur, J., Eckes, P., Langenkemper, K., Reiss, B., Koncz, C., Schell, J., and Maas, C. (1996). Enhanced green fluorescence by the expression of an Aequorea victoria green fluorescent protein mutant in mono- and dicotyledonous plant cells. Proc. Natl. Acad. Sci. USA 93:5888–5893. Sheen, J., Hwang, S. B., Niwa, Y., Kobayashi, H., and Galbraith, D. W. (1995). Green-fluorescent protein as a new vital marker in plant-cells. Plant J. 8:777–784. Siemering, K. R., Golbik, R., Sever, R., and Haseloff, J. (1996). Mutations that suppress the thermosensitivity of green fluorescent protein. Curr. Biol. 6:1653–1663. Simpson, C. G., Clark, G., Davidson, D., Smith, P., and Brown, J. W. S. (1996). Mutation of putative branchpoint consensus sequences in plant introns reduces splicing efficiency. Plant J. 9:359–380. Valvekens, D., Van Montagu, M., and Van Lijsebettens, M. (1988). Agrobacterium tumefaciensmediated transformation of Arabidopsis thaliana root explants by using kanamycin selection. Proc. Natl. Acad. Sci. USA 85:5536–5540. van der Berg, C., Willemsen, V., Hage, W., Wiesbeek, P., and Scheres, B. (1995). Cell fate in the Arabidopsis root meristem determined by directional signalling. Nature (London) 378:62–65. Ward, W. W., Cody, C. W., Hart, R. C., and Cormier, M. J. (1980). Spectrophotomeric identity of the energy transfer chromophores in Renilla and Aequorea green fluorescent proteins. Photochem. Photobiol. 31:611–615. Webb, C. D., Decatur, A., Teleman, A., and Losick, R. (1995). Use of green fluorescent protein for visualization of cell-specific gene expression and subcellular protein localization during sporulation in Bacillus subtilis. J. Bacteriol. 177:5906–5911. Wiebauer, K., Herrero, J.-J., and Filipowicz, W. (1988). Nuclear pre-mRNA processing in plants: distinct modes of 3¢-splice-site selection in plants and animals. Mol. Cell. Biol. 8:2042–2051. Yang, F., Moss, L. G., and Phillips, G. N. J. (1996). The molecular structure of green fluorescent protein. Nature Biotech. 14:1246–1251. Zoenicka-Goeta, M., Pines, J., Ryan, K., Siemering, K. R., Haseloff, J., Evans, M. J., and Gurdon, J. B. (1996). An indelible lineage marker for Xenopus using a mutated green fluorescent protein. Development 122:3719–3724. Zoenicka-Goetz, M., Pines, J., Siemering, K. R., Haseloff, J., and Evans, M. J. (1997). Following cell fate in the living mouse embryo. Development 124:1133–1137. Zolotukhin, S., Potter, M., Hauswirth, W., Guy, J., and Muzyczka, N. (1996). A “humanized” green fluorescent protein cDNA adapted for high level expression in mammalian cells. J. Virol. 70:4646– 4654.

13 USES OF GFP IN TRANSGENIC VERTEBRATES Sean Megason Beckman Institute of Biological Imaging, California Institute of Technology, Pasadena, CA

Adam Amsterdam and Nancy Hopkins Center for Cancer Research, Massachusetts Institute of Technology, Cambridge, MA

Shuo Lin Department of Molecular, Cell, and Developmental Biology, UCLA, Los Angeles, CA

13.1

INTRODUCTION

The ability to express exogenous DNA in vertebrate animals has been invaluable to a wide range of biological studies. Our understanding of processes as diverse as gene expression, cell lineage relationships, and gene function have all benefited from the use of “transgenic” animals. Traditionally, analysis of reporter genes such as lacZ in transgenic animals usually involved killing the animal, precluding many types of experiments. The discovery that green fluorescent protein (GFP) could be used as a reporter for gene expression in living animals (Chalfie et al., 1994) has revolutionized the use of transgenics for studying many aspects of biology from development to neuroscience. GFP has a number of advantages over traditional transgenic markers. Most importantly, GFP can be imaged noninvasively in live animals. GFP also provides superior spatial and temporal resolution. Finally the use of GFP fusion proteins and variants of GFP enable a tremendous variety of experiments to probe biological function. A great deal of progress has been made in recent years

Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

285

286

USES OF GFP IN TRANSGENIC VERTEBRATES

in the use of GFP, in techniques for generating transgenic animals, and in techniques for imaging animals. These advances have combined to lead to a dramatic increase in the power of using GFP in transgenic vertebrates since the first edition of this book. This chapter will try to summarize many of these advances and indicate how these techniques might be useful for research involving vertebrate model animals.

13.2

TYPES OF TRANSGENICS VERTEBRATES

The word “transgenic” is used to refer to an organism whose cells contain exogenous DNA. There are two predominant distinctions: whether the DNA is integrated into the host’s chromosome or not (stable versus transient transgenics) and whether or not all of the cells of the organism contain the DNA (nonmosaic versus mosaic). Often the term “transgenic” is only used for the case where DNA has integrated into the host genome and is present in every cell in the animal, a situation best confirmed by passage through the germline. However, it is equally appropriate to think of animals in which only some cells contain the foreign DNA, which may or may not be stably maintained, as transgenic. For many types of experiments, such transgenics are the only feasible option, and for others they are actually more appropriate than transgenics in which every cell harbors the integrated transgene. We will therefore discuss the creation and use of both transient and stable transgenics in this review. Transgenics can be defined methodologically into two major classes: those organisms into which the DNA was introduced (G0, the founder generation) and those (F1, F2, etc., subsequent generations) who have inherited stably integrated copies of it through the germline. The former can be divided into three phenomenological classes: transient mosaic, stable mosaic, and stable nonmosaic. The last class is often treated as functionally equivalent to a germline transgenic, since both involve integration of the transgene in all of their cells (see Table 13.1).

13.3 ISSUES AFFECTING THE SUCCESSFUL USE OF GFP IN TRANSGENIC VERTEBRATES There are a number of factors affecting the successful use of GFP in transgenic vertebrates, including: the expression levels of GFP; the tissue and organism of interest; the GFP variant being used; and the method of visualization. Each of these factors must be considered when designing an experiment involving GFP transgenic vertebrates to increase its chances of success. TABLE 13.1. Methods for Making Different Types of Transgenics Type of Transgenic

Methods

G0 transient G0 stable mosaic G0 stable nonmosaic

DNA microinjection; episomal virus; electroporation Integrating virus; cell transplantation Microinjection into egg pronucleus (mouse); sperm nuclear transplantation (frog) Inheritance of integrated transgene through germline of G0 transgenic

Germline stable nonmosaic

ISSUES AFFECTING THE SUCCESSFUL USE OF GFP IN TRANSGENIC VERTEBRATES

13.3.1

Expression Levels

Expressing GFP at high enough levels for visual detection was initially a concern since GFP reporters do not employ enzymatic amplification as do other transgenic reporters such as lacZ and alkaline phosphatase. The first reports on the use of GFP in vertebrates relied on the use of strong regulatory elements such as the ef1a enhancer in fish (Amsterdam et al., 1995) and the CMV-IE enhancer in mammals and frog (Ikawa et al., 1995; Kroll and Amaya, 1996). An initial concern was that GFP might not be detectable when expressed at lower levels such as those produced by tissue-specific enhancers. Fortunately, however, a number of reports over the past 5 years have demonstrated that GFP can be readily detected in transgenics using tissue-specific enhancers of developmentally important genes such as transcription factors under the right circumstances (Long et al., 1997; Plautz et al., 1997; Pownall et al., 1998). Moderate to low levels of GFP can be detected if the specimen has low levels of autofluorescence, the signal is not too deep in a tissue, and the tissue is transparent. In zebrafish, a number of transgenics have been made that express GFP under the control of enhancers from developmentally important genes (Moss et al., 1996; Higashijima et al., 1997; Meng et al., 1997; Long et al., 1997; Jessen et al., 1999). In many cases, transgenic fish reveal the expression of the gene of interest better than in situ hybridization. A transgenic zebrafish expressing GFP under the regulatory control of the Rag1 gene revealed expression in olfactory neurons that had been previously missed by in situ (Jessen et al., 1999). Since GFP diffuses to fill the entire cell in which it is expressed, GFP transgenics allow single positive cells to be identified with very high resolution. Cellular processes such as axons can be traced over long distances in GFP transgenics (Feng et al., 2000). Since the tissue must be permeablized for chromogenic detection of lacZ or alkaline phosphatase, the signal from these reporters can diffuse into adjacent cells, resulting in a lower resolution signal relative to GFP transgenics. Confocal microscopy of GFP transgenics allows positive cells deep within a tissue to be identified that might be obscured by more superficial positive cells using chromogenic detection. Zebrafish transgenic lines often contain multiple copies of the transgene, leading to brighter signals. GFP knock-in experiments in mice using gene-targeting in ES cells provide the ultimate acid test for the ability to detect GFP in transgenic embryos since these embryos contain only a single copy of GFP. Several such GFP knock-in lines have been made. GFP can be readily detected using confocal and even fluorescent dissecting microscopes from single-copy knock-in insertions of GFP into the loci of the transcription factors Hoxa1, Hoxc13, and CBFb (Godwin et al., 1998; Kundu et al., 2002b). In the case of Hoxc13, an equivalent lacZ knock-in was also available (Goodwin and Capecchi, 1998), allowing for a direct comparison of the sensitivity of these two markers. Comparing mice embryos containing the single-copy lacZ or GFP alleles showed that the sensitivity of these two markers was comparable (Godwin et al., 1998). Although achieving sufficient expression levels for detection should still be considered as a possible limitation when designing experiments using GFP transgenics, the data so far are promising in this regard.

13.3.2

Tissue and Organism

Another issue that must be considered when designing an experiment using GFP transgenics is the tissue and organism that are going to be studied. The tissue and organism dictate many factors that affect the quality of the GFP signal that can be detected. These

287

288

USES OF GFP IN TRANSGENIC VERTEBRATES

factors include: the depth of the signal; the amount of light scattering and absorption of the tissue; and the amount of background caused by autofluorescence. Typically, GFP signals that are more superficial are easier to detect than deeper signals since a deeper signal must pass though more tissue and will be absorbed more. The distance that a GFP signal can penetrate is very dependent on the tissue. Tissue that is uniform and transparent will allow a GFP signal to be transmitted with less degradation than tissues that are opaque. In practice, younger embryos are typically easier to image since they are smaller, more transparent, and have less autofluorescence. The species of the organism and its stage of development have a large impact on how well GFP can be imaged in a transgenic. Zebrafish embryos can be imaged nicely over a wide range of development from fertilization until the larval stage since they are small and optically transparent. One of the principal advantages of using GFP in transgenics is that GFP allows imaging of developmental processes to be done in live embryos over time. Since zebrafish embryos develop freely outside their mother, vital imaging is much easier in zebrafish than in mice or chick. Zebrafish embryos begin developing melanocytes at 24 h of development, which can interfere with imaging, but melanin synthesis can be easily blocked using phenylthiourea (PTU). Iridophores begin to form during the second day of zebrafish development. These highly reflective cells can pose a problem in imaging certain tissues such as the eye. In other organisms, imaging GFP transgenics is only possible at certain stages of development. Mouse embryos can be imaged well up until about embryonic day 12.5 and chick embryos until about Stage 24. After these stages, the embryos become too large and opaque for optimal imaging of GFP. Performing time-lapse imaging in mouse and chick requires more sophisticated culture conditions than in zebrafish or Xenopus. Xenopus embryos are opaque due to the presence of yolk throughout the embryo from the late stages of oogenesis until larval stages. GFP can be used in Xenopus if it is expressed at high expression levels and in superficial tissues, but Xenopus is not as well-suited for imaging GFP as is zebrafish or early stage mouse and chick embryos. The tissue being studied also affects the quality of imaging from a GFP transgenic because tissues vary in their degree of transparency and autofluorescence. Tissues that are not transparent cause a loss of GFP signal, while tissues that are autofluorescent cause an increase in background. Extracellular matrix, fat deposits, and pigment can decrease the transparency of a tissue. Embryos can be cleared by soaking them in agents that match the refractive index of the tissue such as glycerol or benzyl benzoate. Embryos must be fixed in paraformaldehyde prior to clearing to prevent GFP from leeching out so clearing is not compatible with time-lapse imaging. Autofluorescence in vertebrate embryos is a problem in the gut, yolk, and red blood cells. The spectrum of autofluorescence is typically different from that of the fluorescent protein being used. Thus, utilizing proper filter sets can reduce the effects of autofluorscence.

13.3.3

GFP Variants

There are now a number of fluorescent proteins in addition to the originally discovered GFP (termed wild-type GFP or wtGFP). GFP variants differ in their excitation spectra, emission spectra, brightness, toxicity, codon-optimization, photostability, protein stability, maturation time, and suitability as a fusion partner. Currently, the best fluorescent proteins for making transgenics are enhanced green fluorescent protein (EGFP), the yellow fluorescent proteins citrine (Greisbeck et al., 2001) and venus (Nagai et al., 2002), and the DsRed derivative mCherry (Shaner et al., 2004). These proteins are bright, photostable,

ISSUES AFFECTING THE SUCCESSFUL USE OF GFP IN TRANSGENIC VERTEBRATES

monomeric, nontoxic, mature quickly, and are codon-optimized for expression in vertebrates. The excitation and emission spectra of these proteins are also compatible with the most commonly used laser lines and filter sets for microscopy. Other variants may be useful in special cases. Destabilized GFP variants with decreased half-lives such as d2EGFP (Clontech) may be useful for monitoring rapid changes in gene expression levels during development that might be obscured by the long half-life EGFP. The DsRed variant called Timer (Clontech), which changes colors from green to red over time, may also be useful for following rapid changes in gene expression levels in a transgenics. GFP is a small protein that will freely diffuse to fill the entire volume of cells in which it is expressed. GFP will diffuse to fill the nucleus, cytoplasm, and even cellular projections such as axons, allowing the morphology of cells to be visualized with high resolution. Sometimes, however, it is advantageous to only label certain parts of GFP-positive cells in a transgenic. Nuclear-localized variants of GFP such as the histone H2B-EGFP fusion (Kanda et al., 1998) or the fusion of EGFP to the SV40 nuclear localization signal (Clontech) are useful when it is necessary to clearly identify single cells such as for cell counting. Other uses for GFP fusion proteins are discussed below.

13.3.4

Visualization Methods

GFP transgenics will most often be visualized using fluorescence microscopy to detect the fluorescence of GFP. GFP transgenics can be visualized on any type of fluorescence microscope including dissecting (stereo) microscopes, compound microscopes, and laserscanning microscopes (confocal or multiphoton). 13.3.4.1 Visualizing GFP Transgenics with a Fluorescent Dissecting Microscope. Many lines of GFP transgenic vertebrates can be visualized using a fluorescent dissecting microscope. Fluorescent dissecting microscopes are generally not as sensitive as laser-scanning microscopes, but they allow the embryos to be easily manipulated while they are visualized. GFP-positive embryos can be identified and sorted for further experiments using a fluorescent dissecting microscope. This technique is particularly useful for establishing GFP transgenic lines in embryos that can be vitally imaged such as zebrafish, Medaka, and Xenopus. Fluorescent dissecting microscopes can also be used to dissect GFP-positive tissue from the embryo for further molecular analysis. Several manufacturers produce dissecting microscopes equipped for fluorescence. The Leica MZFLIII uses epi-illumination provided by a mercury arc lamp for excitation and contains a rotating filter wheel with filter sets for use with different fluorophores including EGFP and mRFP1. 13.3.4.2 Visualizing GFP Transgenics with Confocal or Multiphoton Microscopy. GFP transgenic vertebrates can also be visualized using confocal or multiphoton microscopy. These forms of microscopy eliminate out-of-focus light, allowing thin optical sections to be captured deep within a specimen. This capability is especially useful for imaging GFP transgenics as whole animals since the thickness of these specimens generates a large amount of haze from out-of-focus light using traditional fluorescence microscopy. Confocal and multiphoton microscopy can reveal the detailed expression pattern of GFP in deep tissues even if more superficial tissues are also GFP-positive. Modern confocal and multiphoton microscopes are advantageous for imaging live animals because they have computer controlled shutters that limit the exposure of the specimen to light, and they have very sensitive photodetectors that allow lower light levels to be used

289

290

USES OF GFP IN TRANSGENIC VERTEBRATES

for imaging. These advances allow time-lapse imaging of living embryos to be performed over long periods of time without phototoxicity to the embryo or bleaching of GFP (Megason and Fraser, 2003). 13.3.4.3 Other Methods for Visualizing GFP Transgenics. There are several other methods available for visualizing GFP transgenics in addition to the ones discussed above. GFP transgenic embryos can be sectioned rather than imaged as whole-mounts. Paraffin embedding destroys the fluorescence of GFP, but vibratome sectioning and frozen sectioning work well with GFP. Embryos should be fixed in paraformaldehyde prior to sectioning. Paraformaldehyde slightly decreases the fluorescence of GFP but preserves the morphology of the tissue well. Immunostaining can be performed on frozen sections of GFP transgenic embryos to allow GFP and an antibody marker to be visualized simultaneously (Megason and McMahon, 2002). If methods must be used that destroy the fluorescence of GFP, then antibody staining or in situ hybridization can be used to detect GFP expression. A final and rather different method for visualizing GFP is “GFP Goggles” (bls-ltd.com). This device contains a blue light source mounted like a miner’s helmet, and the goggles contain green filters. These GFP goggles allow fast, macroscopic viewing of GFP-positive animals. They can be worn in an animal facility and used to visually genotype lines of GFP transgenic animals with expression in the eyes or skin.

13.4 METHODS FOR CREATING TRANSGENICS IN DIFFERENT VERTEBRATE SPECIES 13.4.1

Fish

13.4.1.1 Plasmid Injection. Transient and stable GFP transgenics can be made in both medaka and zebrafish by injecting plasmid DNA into the embryo at early cleavage stages (Ozato et al., 1986; Stuart et al., 1988). In zebrafish, injected plasmid DNA is converted into a high-molecular-weight form and amplified during the cleavage stage. It is subsequently degraded during gastrulation and retained in only a mosaic of cells (Stuart et al., 1988). The mosaic distribution of injected DNA in injected embryos can be a problem for using transient transgenic fish. Zebrafish embryos injected with plasmids containing GFP under the control of ubiquitous promoters such as the Xenopus EF1a enhancer show GFP expression in only a minority of cells (Amsterdam et al., 1995). Plasmid injection has not been used often for functionally characterizing genes by overexpression analysis because of the mosaic expression. RNA injection is more frequently used for functional molecular analysis. Transient transgenesis using plasmid-based GFP reporters can, however, be used for characterizing regulatory elements. By superimposing the expression patterns observed in a number of embryos injected with plasmid-based GFP reporters, the regulatory elements controlling the expression of GATA-2 and GATA-1 were determined (Meng et al., 1997; Long et al., 1997). The mosaicism of GFP in transient transgenics can be useful for following individual cells in time-laspse imaging (Koster and Fraser, 2001) or for identifying the projections of individual neurons (Downes et al., 2002). Stable GFP transgenics can also be generated by injection of linearized plasmid DNA. Injected plasmid DNA is only inherited by a small fraction of cells, resulting in a mosaicly transgenic germline (Stuart et al., 1988, 1990; Culp et al., 1991). Typically only 5–10% of injected fish will carry the transgene in their germline, and these will pass on the trans-

METHODS FOR CREATING TRANSGENICS IN DIFFERENT VERTEBRATE SPECIES

gene to only 5–10% of their progeny. In subsequent generations, the transgene will be nonmosaic and passed on at normal Mendelian frequencies. Stable transgenics in zebrafish that faithfully expressed GFP in desired tissues were first created using the zebrafish GATA-1 promoter (Long et al., 1997). Since then, a great number of stable transgenic GFP lines have been created using tissue-specific enhancers in zebrafish (Fig. 13.1).

Figure 13.1. Stable transgenic zebrafish expressing GFP in specific tissues. (A, B) GATA-1 GFP expression in hematopoietic cells (Long et al., 1997). (C, D) GATA-2 BAC GFP expression in neuronal cells (Shuo Lin, unpublished). (E, F) Rag-1 GFP expression in olfactory sensory neurons (Jessen et al., 1999). (G, H) Rag-1 BAC GFP expression in thymus (Jessen et al., 1999). (I) Insulin GFP expression in pancreatic beta cells (Huang et al., 2001). (J) POMC GFP expression in pituitary cells (Liu et al., 2003). (K) FLK GFP expression in vascular cells (Cross et al., 2003). See color insert.

291

292

USES OF GFP IN TRANSGENIC VERTEBRATES

13.4.1.2 BAC Injection. In vertebrates, elements regulating the expression of a gene can be located hundreds of kilobases away from the promoter. It is difficult to identify such elements using plasmid-based GFP reporters. The use of larger insert vectors such as YACs (yeast artificial chromosomes), BACs (bacterial artificial chromosomes), and PACs (P1-derived artificial chromosomes) for generating transgenics largely overcomes these problems (Strauss et al., 1993; Yang et al., 1997). BACs and PACs are advantageous over YACs because they grow in bacteria rather than in yeast and they are more stable. These constructs are too large to be manipulated using traditional cloning techniques such as restriction digestion and ligation, but they can be very precisely manipulated using homologous recombination (Yang et al., 1997). Jessen et al. (1998) used chi-stimulated homologous recombination in bacteria to target GFP to a BAC containing the GATA-2 gene. Injection of the GFP modified BAC into zebrafish led to proper expression of GFP in the GATA-2 domain, whereas plasmid-based constructs did not. Similar results were obtained for the rag1 locus (Jessen et al., 1999). 13.4.1.3 Meganuclease. The meganuclease I-SceI has been shown to dramatically increase the efficiency of GFP transgenesis in the Medaka fish (Thermes et al., 2002). I-SceI meganuclease is a homing endonuclease from yeast with an 18-bp recognition sequence (Colleaux et al., 1988). Although homing endonucleases tolerate some degeneracy in their recognition sequences, they are still expected to cleave rarely within a vertebrate genome. Thermes and colleagues flanked a GFP transgene of interest with I-SceI sites and co-injected this DNA with I-SceI enzyme into Medaka embryos (Thermes et al., 2002). Uniform, promoter-dependent expression of GFP was increased in the injected embryos, and the transgenesis frequency was raised to 30.5%. Amazingly, many of the injected animals passed on the transgene to almost 50% of their offspring. This technique has the potential to dramatically increase the efficiency of transgenesis in fish. 13.4.1.4 Transposons. Transposons are small mobile DNA elements that can catalyze their own excision and integration into DNA. The components necessary for transposition can be separated into trans components (e.g. transposase) and cis components (e.g. terminal repeats). Transgenesis can be performed using transposons by designing plasmids that contain a transgene flanked by a transposon’s cis components and coinjecting the plasmid with transposase into an embryo. Two transposon systems have shown promise in recent years in fish: Sleeping Beauty (Ivics et al., 1997; Davidson et al.) and Tol2 (Koga et al., 1996; Kawakami et al., 2004). With both of these systems, transgenes can be inserted into vectors that contain the transposon’s required cis elements and coinjected with RNA encoding transposase into fertilized eggs. For Sleeping Beauty, coinjection of a transposon based EGFP expression plasmied with transposase generated expressing germ-line transgenics at a rate of 31% compared to 5% without transposase (Davidson et al., 2003). An analogous experiment with Tol2, generated transgenics founders at a rate of 50%, and many of these fish had multiple transposon insertions (Kawakami et al., 2004). In addition to generating transgenics more efficiently than standard DNA injection, transposon based transgenesis also produces single-copy insertions rather than the head-to-tail concatmers found with standard DNA injection. Transgene concatamers can provide higher levels of expression than single copy insertions, but single copy insertions are more ideal for some applications such as gene trapping and the generation of reporters and other substrates for site-directed recombinases such as Cre and

METHODS FOR CREATING TRANSGENICS IN DIFFERENT VERTEBRATE SPECIES

FLP. Transposons likely have limits on their insert size. For Sleeping Beauty, a decreaes in transgenic frequency from 31% to 10% was seen by increasing the size of the transposon from 2 kb to 5 kb (Davison et al., 2003). Transposon based transgensis can be performed using tissue specific enhancers as along as potential issues with insert size are kept in mind. Sleeping Beauty was used to generated zebrafish expressing GFP in the lens of the eye using a 490 bp Xenopus g-crystallin promoter (Davidson et al., 2003). Likewise, Tol2 was used to generate transgenic zebrafish that express GFP in the anterior neural plate and eye anlage using a 1.6 kb enhancer from the gene six3.2 (Kawakami et al., 2004). 13.4.1.5 Somatic Nuclear Transfer. The cloning of Dolly the sheep using somatic nuclear transfer from a cultured cell (Campbell et al., 1996) generated interest that animal cloning could be used as a method for generating transgenic organisms by using donor nuclei from cells that were genetically modified in culture. Animal cloning has now been successful performed on the animal model systems of mouse (Wakayama et al., 1998) and zebrafish (Lee et al., 2002). This technique has also been used to successfully generate GFP transgenic livestock in goats (Reggio et al., 2001), pigs (Lai et al., 2002), and cows (Bordignon et al., 2003). The use of somatic nuclear transfer for generating transgenic organisms for research is especially promising in zebrafish. Unlike in mice, embryonic stem cells have not been developed for zebrafish that would allow very precise genetic modification in culture followed by recreating an organism with the desired modification. Using somatic nuclear transfer of cultured cells, it may be possible to perform GFP knock-ins and other types of very precise genetic modification in zebrafish (Lee et al., 2002).

13.4.2

Xenopus

13.4.2.1 Sperm Nuclear Transplantation. As in fish, injection of plasmid DNA into Xenopus embryos results in a mosaic distribution of DNA and only a small fraction of cells that express the injected DNA (Etkin and Pearman, 1987). Since Xenopus require a long time to mature, it is not practical to raise a large number of plasmid-injected embryos to maturity and screen for those that transmit stable, nonmosaic transgene insertions as is done in fish. Sperm nuclear transplantation overcomes the problems of mosaic expression resulting from plasmid injection (Kroll and Amaya, 1996). In this method, purified DNA containing the transgene of interest is mixed with decondensed sperm nuclei. The restriction enzyme used to isolate the transgene can also be added to the sperm/ transgene mixture to help the transgene integrate in a process termed restriction-enzymemediated integration (REMI). Addition of the restriction enzyme can improve the efficiency of transgenesis but is not essential. The sperm nuclei are then injected into oocytes to initiate development. This technique results in stable, nonmosaic expression of the transgene in the injected embryos and can be used to generate a large number of transgenic embryos per day. Sperm nuclear transplantation has been used to perform a gene trap screen using GFP to identify several lines of Xenopus laevis that express GFP in a tissuerestricted manner (Bronchain et al., 1999). The sperm nuclear transplantation technique has also been extended to Xenopus tropicalis to generate GFP transgenic with tissuespecific expression patterns (Offield et al., 2000; Chae et al., 2002; Hirsch et al., 2002). Xenopus tropicalis is a relative of Xenopus laevis that is more useful for genetics because

293

294

USES OF GFP IN TRANSGENIC VERTEBRATES

it has a shorter generation time, and its genome is diploid rather than pseudo-tetraploid like Xenopus laevis.

13.4.3

Chick

13.4.3.1 Virus. Infection with retroviruses has been the traditional method for ectopic gene expression in chick (Logan and Tabin, 1998). Retrovirus infection results in stable, mosaic transgene expression. Unfortunately, retroviruses are limited in the size of the transgene that they can carry. Rous sarcoma virus (RSV) is the retrovirus that is most commonly used for overexpressing genes in chick. RCAS, the replication competent form of RSV that is standardly used in chick, can accept inserts up to 2.5 kb. The viral titer drops significantly with inserts larger than 3.0 kb, and the inserts tend to be truncated during viral production. A replication-incompetent form of RSV can accept larger inserts up to 4 kb (Boerkoel et al., 1993). Since GFP is only ~800 bp, it is possible to make fusions of GFP to a protein used for overexpression in RCAS as long as the protein’s coding region is less than 1.7 kb. An internal ribosome entry sequence (IRES) can be used to coexpress a gene of interest with GFP in chicks (Dai et al., 2001). Enhancer analysis is not routinely performed using retroviruses because of the limit on the size of the insert and the difficulty of making retroviruses containing inserts with their own promoters or splice sites, although it is possible (Petropoulos et al., 1992). Retroviruses that express GFP have been successfully used to label neurons for time-lapse, lineage analysis in chick (Okada et al., 1999). 13.4.3.2 Electroporation. A more promising method of generating GFP transgenic chicks is using in ovo electroporation. In this technique, plasmid DNA is injected into the tissue of interest, and a series of brief, square pulses of voltage are applied across the tissue using microelectrodes (Muramatsu et al., 1997; Itasaki et al., 1999). A range of different tissues can be targeted by adjusting the method of DNA injection and the position of the electrodes, including the neural tube, endoderm, somites, eye, surface ectoderm, and limb (Fig. 13.2A,B). Electroporation results in mosaic, transient expression. The degree of mosaicism depends on the tissue, but some tissues such as the neural tube can be electroporated very efficiently such that >95% cells are GFP-positive (Megason and McMahon, 2002). Expression typically fades in 3–4 days. GFP can be coexpressed from a ubiquitous enhancer/promoter such as CMV/b-actin along with a gene of interest utilizing an IRES (internal ribosome entry sequence) to mark which cells were transfected since the expression is mosaic (Megason and McMahon, 2002). Tissue-specific enhancers can also be used in electroporation (Itasaki et al., 1999). Electroporation has been most widely used in chick, but it is also applicable to other species including mice (Itasaki et al., 1999), ascidians (Corbo et al., 1997), zebrafish (Swartz et al., 2001; Tawk et al., 2002), and Xenopus (Eide et al., 2000). The quickness, ease, and flexibility of electroporation make it a very promising technique for transient transgenesis.

13.4.4

Mice

13.4.4.1 DNA Injection. Pronuclear injection has been a valuable technique for generating transgenic mice for over 20 years (Gordon and Ruddle, 1981). DNA injected into the pronuclei of mouse zygotes integrates to generate stable, nonmosaic transgenics at fairly high efficiency (10–50%). A number of lines of mice have been generated that ubiquitously express GFP (Takada et al., 1997; Okabe et al., 1997; Chiocchetti et al., 1997).

METHODS FOR CREATING TRANSGENICS IN DIFFERENT VERTEBRATE SPECIES

Figure 13.2. GFP expression in transgenic mouse and chick. (A, B) GFP expression in neural tube and neural crest following electroporation of chick with a GFP encoding plasmid (green) and anti-HNK1 immunostaining (red) to mark neural crest. (A) Lateral view of whole mount (Maria Elena de Bellard and Marianne Bronner-Fraser, unpublished). (B) Cross section through neural tube with DAPI staining (blue) to mark nuclei (Ed Coles and Marianne Bronner-Fraser, unpublished observations). (C) Yolk sac of an E9.5 transgenic mouse showing e–globin GFP expression in red blood cells (Dyer et al., 2001; Elizabeth Jones, unpublished observations). (D) Section through cerebellum of Calbindin BAC GFP transgenic mouse showing expression in Purkinje cells (Xiangdong William Yang and Nat Heintz, unpublished observations). See color insert.

Over the last several years, over 100 lines have also been generated that express GFP in a tissue-restricted manner (Fig. 13.2C). As in fish, using BACs to make GFP transgenics is also advantageous in mice to increase the likelihood of the transgene containing all of the elements required for proper expression of GFP (Yang et al., 1997) (Fig. 13.2D). A large-scale effort directed by Nathaniel Heintz, Mary-Beth Hatten, and Alexandra Joyner is currently underway to target GFP to a large number of BACs, generate transgenics for the BACs, and analyze their expression patterns (Gong et al., 2003). The project aims to analyze 1000 genes per year. This project has the potential to greatly benefit our knowledge of gene expression and provide a valuable collection of modified BACs. 13.4.4.2 Knock-in. Transgenic mice can also be generated using gene targeting in embryonic stem cells to “knock-in” GFP into a locus of interest. Knock-ins generally require more work to create than do transgenics created through pronuclear injection, but knock-ins allow for more precise control of the transgene. In knock-ins, GFP is inserted into the endogenous loci, ensuring that all of the regulatory elements required for proper expression of the gene of interest are present. Sequences inserted into a locus via target-

295

296

USES OF GFP IN TRANSGENIC VERTEBRATES

ing such as selection cassettes can interfere with the regulation of the locus, so care must be taken that interfering sequences are removed in the ES cells prior to generating the mice. Selection cassettes used in targeting can be removed using the cre-loxP system or through retargeting while using negative selection. GFP knock-ins in mice can be designed in several ways: GFP can be expressed in place of the endogenous gene such that the endogenous gene is knocked out (Godwin et al., 1998); GFP can be expressed as a protein fusion with the endogenous gene (Kundu et al., 2002a,b); or GFP can be expressed from an internal ribosome entry sequence (IRES) such that both GFP and the endogenous gene are expressed.

13.5 13.5.1

USES OF GFP TRANSGENICS IN VERTEBRATES GFP as a Coexpression Marker

GFP transgenics can be used for a variety purposes in vertebrates. Perhaps the most straightforward use of GFP is as a coexpression marker. In this scenario a transgenic is created that expresses both GFP and a gene of interest. The individual animals and cells expressing the gene of interest can then be identified simply and prescisely by visualizing GFP. Many forms of transgenesis such as electroporation result in mosaic expression. Coelectroporation a GFP expression plasmid along with an expression plasmid of interest has been used to identify which regions of the embryo were successfully transfected (Araki and Nakamura, 1999). For single-cell resolution of which cells were transfected by in ovo electroporation, a construct containing a gene of interest followed by an IRES-GFP cassette can be used (Megason and McMahon, 2002). GFP can also be used for identifying which animals are transgenic when generating transgenic vertebrates. In transgenic Xenopus embryos made using sperm nuclear transfer using both a GFP expression construct and another construct of interest, 94% of the transgenic embryos contained both transgenes cointegrated (Hartley et al., 2001). Double promoter plasmids containing a gene and promoter of interest along with GFP under the control of the crystalline promoter have been used in Xenopus to identify transgenics by looking for GFP in their eyes (FU et al., 2002). Cointegration of a detectable marker with the transgene of interest can also be used in mice to identify transgenics (Overbeek et al., 1991). GFP can also be used for genotyping and sex-typing: By generating mice embryos using a father containing a GFP insertion on his X-chromosome, the sex of embryos was determined noninvasively at embryonic day 2.75 by GFP visualization long before overt sexual differentiation occurs at E12.5 (Hadjantonakis et al., 1998).

13.5.2

GFP as a Marker for Cell Types

GFP can also be used as a marker for a cell type of interest by using an enhancer that drives expression of GFP in that cell type in transgenics. Currently, immunhistochemistry with cell-type-specific antibodies and in situ hybridization with cell-type specific probes are the standard methods used for marker analysis. These techniques are advantageous in that they allow a number of different markers to be assayed, but they can only be used on fixed specimens. Using GFP transgenics as markers allows cell types of interest to be studied in living specimens. GFP expressed under control of the Oct4 enhancer was used to mark primordial germ cells in the mouse (Anderson et al., 2000). Time-lapse imaging revealed that PGCs originate from the posterior primitive streak and begin migrating toward the future site of the allantois. Using GFP as a marker for cell types also allows

USES OF GFP TRANSGENICS IN VERTEBRATES

cellular morphology to be visualized at much higher resolution than by immunhistochemistry or in situ hybrization. GFP transgenics are particularly useful for marking neural cell types. Feng and colleagues used the Thy1 promoter to generate transgenics using four different spectral variants of GFP (Feng et al., 2000). They generated 25 different lines that each marked different populations of neurons presumably due to transgene integration effects. GFP diffused within expressing neurons to beautifully mark the entire cell from dentrititic spines to the nerve terminal of axons several centimeters long. GFP transgenics can also be used to mark cell types of interest to be purified by fluorescentactivated cell sorting (FACS). Transgenic mice expressing GFP under the control of the L7 promoter were used to purify live Purkinje cells (Tomomura et al., 2001). A knock-in of GFP into the Hoxa13 locus in mice was used to purify limb mesenchymal cells that were then used to show Hoxa13 null cells are defective in forming chondrogenic condensations in vitro (Stadler et al., 2001).

13.5.3

GFP as a Marker for Gene Expression Patterns

GFP transgenics can also be used to study gene expression patterns by placing GFP under the control of the regulatory elements of a gene of interest. This technique is similar to using GFP to mark cell types, but the focus is on the gene being marked rather than the cell type being marked. Knock-ins and BAC transgenics are the preferred method for marking a gene expression pattern with GFP because these methods are more likely to result in transgenics that faithfully recapitulate the expression pattern of the gene of interest compared to plasmid-based transgenics. GFP transgenics are advantageous over in situ hybrization for marking gene expression in several ways. Since GFP can reveal gene expression patterns in live animals, they could make it easier to study rapid and dynamic changes in gene expression during development, such as the oscillation of gene expression in the presegmental plate during somitogenesis, or in response to exogenous factors. Once a GFP transgenic line is established, it allows gene expression to be assayed much more easily than in situ hybridization. GFP transgenics also provide better spatial and temporal resolution of gene expression than does in situ hybridization because GFP transgenics can be imaged at cellular resolution continuously over development using time-lapse, confocal microscopy. However, GFP transgenics have some potential problems for marking gene expression patterns relative to in situ hybridization. One must ensure that GFP from a transgenic is expressed in the same pattern as the gene being marked, usually by comparison with in situ hybridization at individual time points. GFP takes 1–2 h to become fluorescent after its transcription is initiated. Because of the long half-life of normal GFP protein, GFP fluorescence can remain long after its transcription has ended, although destabilized variants of GFP may reduce this problem (Li et al., 1998). These effects can cause shifts in the timing of GFP fluorescence relative to its transcription. As described above, a large-scale effort is currently underway to analyze the gene expression patterns of thousands of genes in mice using BAC GFP transgenics. Similar efforts may also be performed in the coming years in zebrafish and Xenopus. Gene trapping using GFP is also being used to mark expression patterns from a number of genes in Xenopus (Bronchain et al., 1999) and zebrafish (Kawakami et al., 2004; Parinor et al., 2004; Balciunas et al., 2004).

13.5.4

GFP Transgenics for Enhancer Analysis

GFP transgenics can be used for mapping the regulatory elements that control the expression of a gene. Enhancer analysis is usually begun by isolating a large DNA fragment from

297

298

USES OF GFP IN TRANSGENIC VERTEBRATES

the locus being studied that is capable of directing expression of GFP in the proper pattern in transgenics. Successively smaller fragments are then used to map the location and tissue-specificity of different positive and negative regulatory elements in the enhancer. GFP transgenics were used to map the regulatory elements of the GATA1, GATA2, and rag1 loci in zebrafish (Long et al., 1997; Meng et al., 1997; Jessen et al., 1999). Data obtained from a number of transient transgenics can be pooled to simplify this analysis in zebrafish (Long et al., 1997). GFP transgenics were used in Xenopus to identify separate regulatory elements that direct endodermal and mesodermal expression of the transcription factor HNF3a (Ryffel and Lingott, 2000). Electroporation in chick can also be used for enhancer analysis (Itasaki et al., 1999). Electroporation in chick can be done using mouse enhancers and is much easier, quicker, and cheaper than creating transgenics in mice (Timmer et al., 2001). Electroporation in chick can be used for initial dissection of an enhancer followed by confirmation using mouse transgenics.

13.5.5

GFP Fusions for Examining Protein Function in Vivo

Although not yet fully realized, a potentially exciting use of GFP in transgenic vertebrates is the use of GFP protein fusions for studying the in vivo function of proteins. In this method the coding sequence of GFP is fused to the coding sequence of another protein to generate a chimeric protein containing both GFP and the protein of interest. If designed correctly, GFP fusions are often functional because of the compact, monomeric nature of GFP. Long et al. (2000) used transgenic zebrafish expressing a death receptor/GFP fusion protein to demonstrate the critical role of this protein in negative regulation of erythropoiesis. One use of GFP fusions is to examine the subcellular distribution of a protein in vivo. Many proteins change their subcellular distribution depending on their functional state. GFP fusions were used to show that the Wnt signaling component Axin redistributes from the cytoplasm to the membrane in response to Wnt signaling (Cliffe et al., 2003). GFP fusions were also used to show that the hedgehog signaling component Smoothened redistributes from internal compartments to the cell surface upon hedgehog signaling (Zhu et al., 2003). Functional GFP fusions can even be formed for secreted signaling molecules. A fusion of decapentaplegic to GFP was used to visualize the fromation of a morphogen gradient in the fly wing disc (Teleman and Cohen, 2000). GFP fusions are particularly useful in neurosecience. Protein fusions to ECFP and EYFP have been used in conjunction with fluorescent resonant energy transfer (FRET) to detect activation of G-proteincoupled potassium channels (Riven et al., 2003). pH-, Ca2+-, and voltage-sensitive variants of GFP generated through protein fusions can be used for monitoring neural activity noninvasively (Miesenbock et al., 1998; Miyawaki et al., 1999; Sakai et al., 2001). A novel GFP fusion protein can be used for mapping neural connectivity: A fusion of GFP to a nontoxic fragment of tetanus toxin is transferred across synapses in a retrograde direction in transgenic mice, allowing for mapping of neural circuits using transgenics (Maskos et al., 2002).

13.6

CONCLUSION

It has been less than a decade since the first use of GFP as a marker (Chalfie et al., 1994), yet this little protein has already revolutionized many areas of biology. Parallel advances in imaging and genetic manipulation over the past decade have further benefited the use of GFP. Because of the extra time and expense involved in using vertebrates compared to

REFERENCES

using invertebrates or in vitro approaches, many techniques for using GFP have only recently been applied to vertebrates. Already though, the use of GFP in vertebrates has contributed immensely to investigations into both their embryonic development and function as adults. The coming decade will undoubtedly see a dramatic broadening in the use of currently available GFP techniques in vertebrates in addition to the development of novel GFP techniques for studying biology.

REFERENCES Anderson, R., Copeland, T. K., Scholer, H., Heasman, J., and Wylie, C. (2000). The onset of germ cell migration in the mouse embryo. Mech. Dev. 91:61–68. Amsterdam, A., Lin, S., and Hopkins, N. (1995). The Aequorea victoria green fluorescent protein can be used as a reporter in live zebrafish embryos. Dev. Biol. 171:123–129. Araki, I., and Nakamura, H. (1999). Engrailed defines the position of dorsal di-mesencephalic boundary by repressing diencephalic fate. Development 126:5127–5135. Balciunas, D., Davidson, A. E., Sivasubbu, S., Hermanson, S. B., Welle, Z., and Ekker, S. C. (2004) Enhancer trapping in zebrafish using the Sleeping Beauty transposon. BMC Genomics 5:62–76. Boerkoel, C. F., Federspiel, M. J., Salter, D. W., Payne, W., Crittenden, L. B., Kung, H. J., and Hughes, S. H. (1993). A new defective retroviral vector system based on the Bryan strain of Rous sarcoma virus. Virology 195:669–679. Bordignon, V., Keyston, R., Lazaris, A., Bilodeau, A. S., Pontes, J. H., Arnold, D., Fecteau, G., Keefer, C., and Smith, L. C. (2003). Transgene expression of green fluorescent protein and germ line transmission in cloned calves derived from in vitro-transfected somatic cells. Biol. Reprod. 68:2013–2023. Bronchain, O. J., Hartley, K. O., and Amaya, E. (1999). A gene trap approach in Xenopus. Curr. Biol. 9:1195–1198. Campbell, K. H., McWhir, J., Ritchie, W. A., and Wilmut, I. (1996). Sheep cloned by nuclear transfer from a cultured cell line. Nature 380:64–66. Campbell, R. E., Tour, O., Palmer, A. E., Steinbach, P. A., Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2002). A monomeric red fluorescent protein. Proc. Natl. Acad. Sci. USA 99: 7877–7882. Chae, J., Zimmerman, L. B., and Grainger, R. M. (2002). Inducible control of tissue-specific transgene expression in Xenopus tropicalis transgenic lines. Mech. Dev. 117:235–241. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263:802–805. Chiocchetti, A., Tologano, E., Hirsch, E., Silengo, L., and Altruda, F. (1997). Green fluorescent protein as a reporter of gene expression in transgenic mice. Biochim. Biophys. Acta 1352:193–202. Cliffe, A., Hamada, F., and Bienz, M. (2003). A role of dishevelled in relocating axin to the plasma membrane during wingless signaling. Curr. Biol. 13:960–966. Colleaux, L., D’Auriol, L., Galibert, F., and Dujon, B. (1988). Recognition and cleavage site of the intron-encoded omega transposase. Proc. Natl. Acad. Sci. USA 85:6022–6026. Corbo, J. C., Levine, M., and Zeller, R. W. (1997). Characterization of a notochord-specific enhancer from the Brachyury promoter region of the ascidian, Ciona intestinalis. Development 124:589–602. Culp, P., Nusslein-Volhard, C., and Hopkins, N. (1991). High-frequency germ-line transmission of plasmid DNA sequences injected into fertilized zebrafish eggs. Proc. Natl. Acad. Sci. USA 88:7953–7957.

299

300

USES OF GFP IN TRANSGENIC VERTEBRATES

Cross, L. M., Cook, M. A., Lin, S., Chen, J. N., and Rubinstein, A. L. (2003). Rapid analysis of angiogenesis drugs in a live fluorescent zebrafish assay. Arterioscler. Thromb. Vasc. Biol. 23(5):911–912. Dai, C., Celestino, J. C., Okada, Y., Louis, D. N., Fuller, G. N., and Holland, E. C. (2001). PDGF autocrine stimulation dedifferentiates cultured astrocytes and induces oligodendrogliomas and oligoastrocytomas from neural progenitors and astrocytes in vivo. Genes Dev. 15:1913–1925. Downes, G. B., Waterbury, J. A., and Granato, M. (2002). Rapid in vivo labeling of identified zebrafish neurons. Genesis 34:196–202. Dyer, M. A., Farrington, S. M., Mohn, D., Munday, J. R., and Baron, M. H. (2001). Indian hedgehog activates hematopoiesis and vasculogenesis and can respecify prospective neurectodermal cell fate in the mouse embryo. Development 128:1717–1730. Eide, F. F., Eisenberg, S. R., and Sanders, T. A. (2000). Electroporation-mediated gene transfer in free-swimming embryonic Xenopus laevis. FEBS Lett. 486:29–32. Etkin, L. D., and Pearman, B. (1987). Distribution, expression and germ line transmission of exogenous DNA sequences following microinjection into Xenopus laevis eggs. Development 99:15–23. Feng, G., Mellor, R. H., Bernstein, M., Keller-Peck, C., Nguyen, Q. T., Wallace, M., Nerbonne, J. M., Lichtman, J. W., and Sanes, J. R. (2000). Imaging neuronal subsets in transgenic mice expressing multiple spectral variants of GFP. Neuron 28:41–51. Fu, L., Buchholz, D., and Shi, Y. B. (2002). Novel double promoter approach for identification of transgenic animals: A tool for in vivo analysis of gene function and development of gene-based therapies. Mol. Reprod. Dev. 62:470–476. Godwin, A. R., and Capecchi, M. R. (1998). Hoxc13 mutant mice lack external hair. Genes Dev. 12:11–20. Godwin, A. R., Stadler, H. S., Nakamura, K., and Capecchi, M. R. (1998). Detection of targeted GFP-Hox gene fusions during mouse embryogenesis. Proc. Natl. Acad. Sci. USA 95:13042–13047. Gong, S., Zheng, C., Doughty, M. L., Losos, K., Didkovsky, N., Schambra, U. B., Nowak, N. J., Joyner, A., Leblanc, G., Hatten, M. E., and Heintz, N. (2003). A gene expression atlas of the central nervous system based on bacterial artificial chromosomes. Nature 425: 917–925. Gordon, J. W., and Ruddle, F. H. (1981). Integration and stable germ line transmission of genes injected into mouse pronuclei. Science 214:1244–1246. Griesbeck, O., Baird, G. S., Campbell, R. E., Zacharias, D. A., and Tsien, R. Y. (2001). Reducing the environmental sensitivity of yellow fluorescent protein. Mechanisms and applications. J. Biol. Chem. 276:29188–29194. Hadjantonakis, A. K., Gertsenstein, M., Ikawa, M., Okabe, M., and Nagy, A. (1998). Non-invasive sexing of pre-implantation stage mammalian embryos. Nat. Genet. 19:220–222. Hartley, K. O., Hardcastle, Z., Friday, R. V., Amaya, E., and Papalopulu, N. (2001). Transgenic Xenopus embryos reveal that anterior neural development requires continued suppression of BMP signaling after gastrulation. Dev. Biol. 238:168–184. Higashijima, S., Okamoto, H., Ueno, N., Hotta, Y., and Eguchi, G. (1997). High-frequency generation of transgenic zebrafish which reliably express GFP in whole muscles or the whole body by using promoters of zebrafish origin. Dev. Biol. 192:289–299. Hirsch, N., Zimmerman, L. B., Gray, J., Chae, J., Curran, K. L., Fisher, M., Ogino, H., and Grainger, R. M. (2002). Xenopus tropicalis transgenic lines and their use in the study of embryonic induction. Dev. Dyn. 225:522–535. Ikawa, M., Kominami, K., Yoshimura, Y., Tanaka, K., Nishimune, Y., and Okabe, M. (1995). A rapid and non-invasive selection of transgenic embryos before implantation using green fluorescent protein (GFP). FEBS Lett. 375:125–128.

REFERENCES

Itasaki, N., Bel-Vialar, S., and Krumlauf, R. (1999). ‘Shocking’ developments in chick embryology: Electroporation and in ovo gene expression. Nat. Cell Biol. 1:E203–E207. Ivics, Z., Hackett, P. B., Plasterk, R. H., and Izsvak Z. (1997). Molecular reconstruction of Sleeping Beauty, a Tc1-like transposon from fish, and its transposition in human cells. Cell 91:501–510. Jessen, J. R., Meng, A., McFarlane, R. J., Paw, B. H., Zon, L. I., Smith, G. R., and Lin, S. (1998). Modification of bacterial artificial chromosomes through chi-stimulated homologous recombination and its application in zebrafish transgenesis. Proc. Natl. Acad. Sci. USA 95:5121–5126. Jessen, J. R., Willett, C. E., and Lin, S. (1999). Artificial chromosome transgenesis reveals longdistance negative regulation of rag1 in zebrafish, Nat. Genet. 23:15–16. Kanda, T., Sullivan, K. F., and Wahl, G. M. (1998). Histone-GFP fusion protein enables sensitive analysis of chromosome dynamics in living mammalian cells. Curr. Biol. 8:377–385. Kawakami, K., Takeda, H., Kawakami, N., Kobayashi, M., Matsuda, N., and Mishina, M. (2004). A transposon-mediated gene trap approach identifies developmentally regulated genes in zebrafish. Dev. Cell 7:133–144. Koga, A., Suzuki, M., Inagaki, H., Bessho, Y., and Hori, H. (1996). Transposable element in fish. Nature 383:30. Koster, R. W., and Fraser, S. E. (2001). Direct imaging of in vivo neuronal migration in the developing cerebellum. Curr. Biol. 11:1858–1863. Kroll, K. L., and Amaya, E. (1996). Transgenic Xenopus embryos from sperm nuclear transplantations reveal FGF signaling requirements during gastrulation. Development 122:3173–3183. Kundu, M., Javed, A., Jeon, J. P., Horner, A., Shum, L., Eckhaus, M., Muenke, M., Lian, J. B., Yang, Y., Nuckolls, G. H., Stein, G. S., and Liu, P. P. (2002a). Cbfbeta interacts with Runx2 and has a critical role in bone development. Nat. Genet. 32:639–644. Kundu, M., Chen, A., Anderson, S., Kirby, M., Xu, L., Castilla, L. H., Bodine, D., and Liu, P. P. (2002b). Role of Cbfb in hematopoiesis and perturbations resulting from expression of the leukemogenic fusion gene Cbfb-MYH11. Blood 100:2449–2456. Lai, L., Park, K. W., Cheong, H. T., Kuhholzer, B., Samuel, M., Bonk, A., Im, G. S., Rieke, A., Day, B. N., Murphy, C. N., Carter, D. B., and Prather, R. S. (2002). Transgenic pig expressing the enhanced green fluorescent protein produced by nuclear transfer using colchicine-treated fibroblasts as donor cells. Mol. Reprod. Dev. 62:300–306. Lee, K. Y., Huang, H., Ju, B., Yang, Z., and Lin, S. (2002). Cloned zebrafish by nuclear transfer from long-term-cultured cells. Nat. Biotechnol. 20:795–799. Li, X., Zhao, X., Fang, Y., Jiang, X., Duong, T., Fan, C., Huang, C. C., and Kain, S. R. (1998). Generation of destabilized green fluorescent protein as a transcription reporter. J. Biol. Chem. 273:34970–34975. Liu, N. A., Huang, H., Yang, Z., Herzog, W., Hammerschmidt, M., Lin, S., and Melmed, S. (2003). Pituitary corticotroph ontogeny and regulation in transgenic zebrafish. Mol. Endocrinol. 17:959–966. Logan, M., and Tabin, C. (1998). Targeted gene misexpression in chick limb buds using avian replication-competent retroviruses. Methods 14:407–420. Long, Q., Meng, A., Wang, H., Jessen, J. R., Farrell, M. J., and Lin, S. (1997). GATA-1 expression pattern can be recapitulated in living transgenic zebrafish using GFP reporter gene. Development 124:4105–4111. Long, Q., Huang, H. G., Shafizadeh, E., Liu, N., and Lin, S. (2000). Stimulation of erythropoiesis by inhibiting a new hematopoietic death receptor in transgenic zebrafish. Nature Cell Biology 2(8):549–552. Maskos, U., Kissa, K., St. Cloment, C., and Brulet, P. (2002). Retrograde trans-synaptic transfer of green fluorescent protein allows the genetic mapping of neuronal circuits in transgenic mice. Proc. Natl. Acad. Sci. USA 99:10120–10125.

301

302

USES OF GFP IN TRANSGENIC VERTEBRATES

Megason, S. G., and Fraser, S. E. (2003). Digitizing life at the level of the cell: High-performance laser-scanning microscopy and image analysis for in toto imaging of development. Mech. Dev. 120:1407–1420. Megason, D. G., and McMahon, A. P. (2002). A mitogen gradient of dorsal midline Wnts organizes growth in the developing CNS. Development 129:2087–2098. Meng, A., Tang, H., Ong, B. A., Farrell, M. J., and Lin, S. (1997). Promoter analysis in living zebrafish embryos identifies a cis-acting motif required for neuronal expression of GATA-2. Proc. Natl. Acad. Sci. USA 94:6267–6272. Miesenbock, G., De Angelis, D. A., and Rothman, J. E. (1998). Visualizing secretion and synaptic transmission with pH-sensitive green fluorescent proteins. Nature 394:192–195. Miyawaki, A., Griesbeck, O., Heim, R., and Tsien, R. Y. (1999). Dynamic and quantitative Ca2+ measurements using improved cameleons. Proc. Natl. Acad. Sci. USA 96:2135–2140. Moss, J. B., Price, A. L., Raz, E., Driever, W., and Rosenthal, N. (1996). Green fluorescent protein marks skeletal muscle in murine cell lines and zebrafish. Gene 173:89–98. Muramatsu, T., Mizutani, Y., Ohmori, Y., and Okumura, J.-I. (1997). Comparison of three non-viral transfection methods for foreign gene expression in early chicken embryos in ovo. Biochem. Biophys. Res. Commun. 230:376–380. Nagai, T., Ibata, K., Park, E. S., Kubata, M., Mikoshiba, K., and Miyawaki, A. (2002). A variant of yellow fluorescent protein with fast and efficient maturation for cell-biological applications. Nat. Biotechnol. 20:87–90. Offield, M. F., Hirsch, N., and Grainger, R. M. (2000). The development of Xenopus tropicalis transgenic lines and their use in studying lens developmental timing in living embryos. Development 127:1789–1797. Okabe, M., Ikawa, M., Kominami, K., Nakanishi, T., and Nishimune, Y. (1997). “Green mice” as a source of ubiquitous green cells. FEBS Lett. 407:313–319. Okada, A., Lansford, R., Weimann, J. M., Fraser, S. E., and McConnell, S. K. (1999). Imaging cells in the developing nervous system with retrovirus expressing modified green fluorescent protein. Exp Neurol. 156:394–406. Overbeek, P. A., Aguilar-Cordova, E., Hanten, G., Schaffner, D. L., Patel, P., Lebovitz, R. M., and Lieberman, M. W. (1991). Coinjection strategy for visual identification of transgenic mice. Transgenic Res. 1:31–37. Ozato, K., Kondoh, H., Inohara, H., Iwamatsu, T., Wakamatsu, Y., and Okada, T. S. (1986). Production of transgenic fish: Introduction and expression of chicken delta-crystallin gene in medaka embryos. Cell Differ. 19:237–244. Parinov, S., Kondrichin, I., Korzh, V., and Emelyanov, A. (2004). Tol2 transposon-mediated enhancer trap to identify developmentally regulated zebrafish genes in vivo. Dev. Dyn. 231:449–459. Petropoulos, C. J., Payne, W., Salter, D. W., and Hughes, S. H. (1992). Appropriate in vivo expression of a muscle-specific promoter by using avian retroviral vectors for gene transfer. J. Virol. 66:3391–3397. Plautz, J. D., Kaneko, M., Hall, J. C., and Kay, S. A. (1997). Independent photoreceptive circadian clocks throughout Drosophila. Science 278:1632–1635. Pownall, M. E., Isaacs, H. V., and Slack, J. M. (1998). Two phases of Hox gene regulation during early Xenopus development. Curr Biol. 8:673–676. Reggio, B. C., James, A. N., Green, H. L., Gavin, W. G., Behboodi, E., Echelard, Y., and Godke, R. A. (2001). Cloned transgenic offspring resulting from somatic cell nuclear transfer in the goat: Oocytes derived from both follicle-stimulating hormone-stimulated and nonstimulated abattoirderived ovaries. Biol. Reprod. 65:1528–1533. Riven, I., Kalmanzon, E., Segev, L., and Reuveny, E. (2003). Conformational rearrangements associated with the gating of the G protein-coupled potassium channel revealed by FRET microscopy. Neuron 38:225–235.

REFERENCES

Ryffel, G. U., and Lingott, A. (2000). Distinct promoter elements mediate endodermal and mesodermal expression of the HNF1alpha promoter in transgenic Xenopus. Mech Dev. 90:65–75. Sakai, R., Repunte-Canonigo, V., Raj, C. D., and Knopfel, T. (2001). Design and characterization of a DNA-encoded, voltage-sensitive fluorescent protein. Eur. J. Neurosci. 13:2314–2318. Shaner, N. C., Campbell, R. E., Steinbach, P. A., Giepmans, B. N., Palmer, A. E., and Tsien, R. Y. (2004). Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nat. Biotechnol. 22:1567–1572. Stadler, H. S., Higgins, K. M., and Capecchi, M. R. (2001). Loss of Eph-receptor expression correlates with loss of cell adhesion and chondrogenic capacity in Hoxa13 mutant limbs. Development 128:4177–4188. Strauss, W. M., Dausman, J., Beard, C., Johnson, C., Lawrence, J. B., and Jaenisch, R. (1993). Germ line transmission of a yeast artificial chromosome spanning the murine alpha 1(I) collagen locus. Science 259:1904–1907. Stuart, G. W., McMurray, J. V., and Westerfield, M. (1988). Replication, integration and stable germline transmission of foreign sequences injected into early zebrafish embryos. Development 103:403–412. Stuart, G. W., Vielkind, J. R., McMurray, J. V., and Westerfield, M. (1990). Stable lines of transgenic zebrafish exhibit reproducible patterns of transgene expression. Development 109:577–584. Swartz, M., Eberhart, J., Mastick, G. S., and Krull, C. E. (2001). Sparking new frontiers: Using in vivo electroporation for genetic manipulations. Dev. Biol. 233:13–21. Takada, T., Iida, K., Awaji, T., Itoh, K., Takahashi, R., Shibui, A., Yoshida, K., Sugano, S., and Tsujimoto, G. (1997). Selective production of transgenic mice using green fluorescent protein as a marker. Nat. Biotechnol. 15:458–461. Tawk, M., Tuil, D., Torrente, Y., Vriz, S., and Paulin, D. (2002). High-efficiency gene transfer into adult fish: A new tool to study fin regeneration. Genesis 32:27–31. Teleman, A. A., and Cohen, S. M. (2000). Dpp gradient formation in the Drosophila wing imaginal disc. Cell 103:971–980. Thermes, V., Grabher, C., Ristoratore, F., Bourrat, F., Choulika, A., Wittbrodt, J., and Joly, J. S. (2002). I-SceI meganuclease mediates highly efficient transgenesis in fish. Mech. Dev. 118:91–98. Timmer, J., Johnson, J., and Niswander, L. (2001). The use of in ovo electroporation for the rapid analysis of neural-specific murine enhancers. Genesis. 29:123–132. Tomomura, M., Rice, D. S., Morgan, J. I., and Yuzaki, M. (2001) Purification of Purkinje cells by fluorescence-activated cell sorting from transgenic mice that express green fluorescent protein. Eur. J. Neurosci. 14:57–63. Wakayama, T., Perry, A. C., Zuccotti, M., Johnson, K. R., and Yanagimachi, R. (1998). Full-term development of mice from enucleated oocytes injected with cumulus cell nuclei. Nature. 394369–394374. Yang, X. W., Model, P., and Heintz, N. (1997). Homologous recombination based modification in Escherichia coli and germline transmission in transgenic mice of a bacterial artificial chromosome. Nat. Biotechnol. 15:859–865. Zhu, A. J., Zheng, L., Suyama, K., and Scott, M. P. (2003). Altered localization of Drosophila Smoothened protein activates Hedgehog signal transduction. Genes Dev. 17:1240–1252.

303

14 THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS Theresa H. Ward Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London, United Kingdom

Jennifer Lippincott-Schwartz Department of Cell Biology and Metabolism, NICHD, NIH, Bethesda, MD

14.1

INTRODUCTION

The use of green fluorescent protein (GFP) chimeras in the study of cell behavior and dynamics is now ubiquitous in all fields of mammalian cell biology. This extensive use is due to (a) the development of new GFP variants and optimized cell expression strategies that produce bright, stable fluorescent signals and (b) advances in fluorescent imaging methods and microscopy systems that make it simple to analyze protein geography, movement, and chemistry in living cells. Here, we discuss several GFP-based techniques including time-lapse imaging, photobleaching, photoactivation, and fluorescence resonance energy transfer (FRET) that have allowed protein dynamics, function, and expression to be analyzed in living mammalian cells. We further describe how these techniques have led to the identification of new pathways and mechanisms essential for mammalian cell homeostasis, which traditional biochemical approaches have been unable to address.

14.2

GFP AND GFP VARIANTS

Initial breakthrough discoveries demonstrated that the gene for GFP from the jellyfish Aequorea victoria contained all of the information necessary for proper synthesis of a Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

305

306

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

fluorescent protein in non-jellyfish species without the need for ancillary jellyfish proteins (Chalfie et al., 1994), that GFP could be utilized in many cell systems (Heim et al., 1994; Inouye and Tsuji, 1994), and that GFP could be attached to a protein of interest to yield a fluorescent chimera (Wang and Hazelrigg, 1994; Kaether and Gerdes, 1995; Ogawa et al., 1995; Olson et al., 1995; Cole et al., 1996; Rizzuto et al., 1996) while retaining the tagged protein’s function [e.g., NMDA (Marshall et al., 1995)]. However, widespread use of GFP as a fluorescent protein tag in mammalian cells only occurred once GFP variants with improved folding, spectral, and expression properties were generated (Table 14.1; see Fig. 14.1 for example fluorescence microscopy images demonstrating organelle localization). Constructed by random and site-directed mutagenesis, many of these variants include the amino acid substitution Ser65 to Thr65 (S65T), which converts the major and minor absorbance peaks of wild-type GFP (wtGFP) to a single absorbance peak at ~489 nm and results in accelerated fluorophore formation (Heim et al., 1995; Heim and Tsien, 1996). The variants also have the codon usage in wtGFP converted to forms more

TABLE 14.1. Spectral Characteristics of the Major Fluorescent Proteinsa Fluorescent Protein

Amino Acid Substitution

BFP

F64L, Y66H, Y145F, V163A F64L, S65T, Y66W, N146I, M153T, V163A F64L, S65T, Y66W, N146I, M153T,V163A, S72A, Y145A, H148D

CFP

Cerulean

wtGFP EGFP

YFP Citrine

Venus

DsRed2 HcRed PA-GFP Kaede KFP1 (kindling) DsRed timer a

F64L, S65T

S65G, V68L, S72A, T203Y S65G, V68L, Q69M, S72A, T203Y F46L, F64L, S65G, V68L, S72A

V163A, T203H

Excitation

Emission

References

384

448

Heim et al. (1994)

433

474

Heim et al. (1994), Ellenberg et al. (1998)

433

474

Rizzo et al. (2004)

397, (475) 489

504 508

514

527

516

529

Chalfie et al. (1994) Heim et al. (1995), Chiu et al. (1996), Cormack et al. (1996), Yang et al. (1996a) Ormö et al. (1996), Ellenberg et al. (1998) Griesbeck et al. (2001)

515

528

Nagai et al. (2002)

558 590 504

583 620 517

572 580 558

582 600 583

Bevis and Glick (2002) Gurskaya et al. (2001) Patterson and LippincottSchwartz (2002) Ando et al. (2002) Chudakov et al. (2003) Terskikh et al. (2000)

Wavelengths are given as the peak of the excitation or emission spectra in nanometers.

307

GFP AND GFP VARIANTS

A

ER

ssKDEL-GFP

ER exit sites

Sec13-GFP

Golgi

GalT-GFP

PM + Golgi

GPI-GFP

B

Figure 14.1. Examples of GFP chimeras and their subcellular localization. (A) Steady-state distribution of several proteins. (B) Confocal images of a cell expressing the secretory cargo protein VSVG-GFP imaged by time lapse as the protein leaves the Golgi apparatus. Eight images at 10-s intervals were overlaid. (Boxed areas) The route of a single post-Golgi carrier to the cell periphery. [Courtesy of Hirschberg et al. (1998).] See color insert.

efficiently used by mammalian cells, producing increased levels of intracellular protein expression. Finally, in all of the GFP mutants it is possible to make the additional mutations of Ala206 to Lys206, Leu221 to Lys221, or Phe223 to Arg223 to prevent GFP from dimerizing at high concentrations (Delagrave et al., 1995; Ehrig et al., 1995; Heim et al., 1995; Cormack et al., 1996; Crameri et al., 1996; Yang et al., 1996a; Zhang et al., 1996; Zolotukhin et al., 1996; Zacharias et al., 2002). The A. Victoria GFP variant known as enhanced GFP (EGFP), which contains the double mutant of Phe64 to Leu64 and Ser65 to Thr65 (F64L/S65T), has become the variant of choice for GFP expression in mammalian cells (Cormack et al., 1996; Yang et al., 1996a; Yang et al., 1996b; Zhang et al., 1996). It is more stable and fluoresces many-fold more intensely than wtGFP when excited at 488 nm, a wavelength of commonly used filter sets and the main emission wavelength of the argon ion laser used in fluorescence-activated cell sorter (FACS) machines as well as in the confocal scanning laser microscope. EGFP’s increased stability and brightness enables proteins labeled with this tag to be visualized in cells with low light intensities over many hours with minimal photobleaching (the photoinduced destruction of a fluorophore), permitting protein trafficking pathways and organelle dynamics to be analyzed in unprecedented detail. Development of other spectral variants of GFP has enabled multispectral imaging to be performed within living mammalian cells. GFP variants with blue emission spectra have

308

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

been used together with red-shifted variants to label two different protein species, which is useful in protein colocalization experiments (Rizzuto et al., 1996; Yang et al., 1996b). These variants also offer the potential for assessing differential gene expression by flow cytometry (Ropp et al., 1996) and for measuring protein–protein interactions through fluorescence resonance energy transfer analysis (FRET) (Heim and Tsien, 1996; Mitra et al., 1996). However, blue fluorescent protein (BFP) is dim and tends to photobleach readily, so alternative multicolor pairs have been developed. One such pair is cyan fluorescent protein (CFP) and yellow fluorescent protein (YFP), which have superseded BFP and GFP, respectively, as better dual imaging partners. CFP has spectra that are intermediate between BFP and EGFP due to a Tyr66 to Trp66 substitution (Heim and Tsien, 1996; Ellenberg et al., 1998), and it is brighter and displays more photostability and less photodamage under imaging than BFP. YFP, which is much brighter than EGFP, was rationally designed on the basis of the GFP crystal structure to red-shift the absorbance and emission spectra with respect to EGFP (Ormö et al., 1996; Ellenberg et al., 1998). This makes it more efficiently excited by the 514-nm line of an argon ion laser. Together, CFP and YFP are readily imaged as dual signals using the two auxiliary lines of the argon laser, permitting simultaneous analysis of the temporal and spatial behavior of two different proteins. In addition to dual-color imaging, the pairing of CFP and YFP has been instrumental in the study of protein–protein interactions in fluorescent resonance energy transfer (FRET) experiments (Wu and Brand, 1994; Clegg, 1995; Lippincott-Schwartz et al., 2001; Rizzo et al., 2004). The development of brighter variants, Cerulean for CFP (Rizzo et al., 2004) and Citrine and Venus for YFP (Griesbeck et al., 2001; Nagai et al., 2002), provides a potentially superior alternative to the CFP/YFP pairing in multispectral experiments. Efforts to further red-shift GFP excitation and emission spectra to produce additional partners has led to the identification of a number of candidate proteins, predominantly from reef coral, with emission peaks ranging from 576 to 645 nm (Zhang et al., 2002). However, they are still under improvement due to their propensity to oligomerize. DsRed from Discosoma striata forms tetramers, while HcRed from Heteractis crispa dimerizes. Mutations to derive monomeric forms have proved helpful for some chimeric constructs (Campbell et al., 2002), or to concatemerize the red fluorescent protein gene such that in a chimeric protein it is able to oligomerize within the fusion protein (Gerlich et al., 2003), but many constructs (particularly those for membrane-bound proteins) that work well with a GFP label are not proving replicable with a red fluorescent label (Hayes et al., 2004). Analysis of the temporal expression pattern and turnover of proteins has become feasible with the development of GFP variants whose spectral properties change with time or are photoactivatable. Examples of these variants are the fluorescent timer protein (Terskikh et al., 2000) and the photoactivatable proteins including photoactivatable GFP (PA-GFP) (Patterson and Lippincott-Schwartz, 2002), Kaede (Ando et al., 2002), and KFP1 (Chudakov et al., 2003). The fluorescent timer protein was generated by random mutagenesis of the red fluorescent protein drFP583 (Terskikh et al., 2000) to a variant that is initially similar to GFP in terms of emitted light, but which over several hours (~16 h) converts to a red-emitting fluorophore. By observing the ratio of green to red fluorescence, it is possible to determine the age of a protein tagged with the timer protein. The use of PA-GFP, developed by improving on wtGFP’s photoconversion from a neutral to anionic species (Elowitz et al., 1997; Patterson and Lippincott-Schwartz, 2002), offers an even better approach to studying protein turnover. PA-GFP displays little initial fluorescence under excitation at the imaging wavelength (~488 nm) but increases its fluorescence up to 100-fold after activation by irradiation at a different wavelength (~400 nm). This allows

FLUORESCENCE MICROSCOPY-BASED TECHNIQUES USING GFP

Figure 14.2. Examples of three photobleaching techniques—FRAP (A), FLIP (B), and photoactivation (C)—that are commonly used with GFP and GFP chimeras to monitor discrete populations of molecules within cells.

direct highlighting of distinct pools of molecules within cells (Fig. 14.2). Because only photoactivated molecules are fluorescent, the lifetime and behavior of molecules can be studied independently of newly synthesized proteins. Many of the above-mentioned GFP variants are readily expressed as fusion products with other proteins in most mammalian cell types, including primary cells such as neurons, hepatocytes, muscle cells, and hematopoietic cells. This property has allowed them to be used as tools in numerous applications, including as minimally invasive markers to track and quantify individual or multiple protein species, as probes to monitor protein–protein interactions, as photomodulatable proteins to highlight and follow the fate of specific protein populations within a cell, and as biosensors to describe biological events and signals. Below, we describe the fluorescence imaging methods that have been used with these GFP variants in mammalian cells, the types of applications these methods have been used for, and the new insights they have gleaned in the analysis of mammalian cell biology.

14.3

FLUORESCENCE MICROSCOPY-BASED TECHNIQUES USING GFP

Addition of GFP to proteins is usually benign with no apparent disruption of function, despite its relatively large size. Since no exogenously added substrate or cofactors are necessary for detecting GFP fluorescence, cells are exposed to minimal invasive treatment. Furthermore, the GFP fluorophore is relatively photostable, with little photodamage occurring during imaging. Not surprisingly, a wide variety of imaging methods have been developed to take advantage of these properties of GFP (see Table 14.2), which are now being used in mammalian cells.

309

310

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

TABLE 14.2. Uses of Different GFP Imaging Techniques in Mammalian Cells Imaging Method

Description

References

Time lapse

General visualization of processes and protein movement in cells Two or more chimeric proteins expressed together to compare localization and dynamics Quantitation of comparative amounts of two or more proteins in region of interest Measures diffusional mobility by selective photobleaching of a region of interest Visualization of exchange of proteins between compartments Visualization of the kinetics of one protein by photobleaching a selected pool while using another tagged with a different fluorescent protein as a marker, particularly useful for FRAP of a moving component FRAP of a region surrounding the area of interest to visualize movement of faint objects or exchange dynamics Repeated photobleaching to measure compartment connectivity Detection of protein-protein interactions by proximity of fluorophores Measures lifetime of fluorescence which can modulate with changes in oligomerization, pH, Ca2+, or degradation Measures molecular diffusion at low concentrations; sensitive to protein–protein interactions Visualization of processes close to, or at, the cell surface Low level expression of tagged cytoskeletal components to visualize turnover and movement of polymers Activation of selected pool of tagged protein enables pulse-chase of its subsequent dynamics Fluorescence changes color over time, thus monitoring populations of protein synthesized in response to stimuli (e.g., during development)

Presley et al. (1997), Hirschberg et al. (1998) Ellenberg et al. (1998)

Multispectral imaging Ratio imaging

FRAP

FLAP

iFRAP

FLIP FRET

FLIM

FCS

TIR-FM FSM

PA-GFP

Fluorescent Timer

White et al. (2001)

Edidin (1994), LippincottSchwartz et al. (2001)

Dunn et al. (2002)

Presley et al. (2002), Bubulya and Spector (2004)

Cole et al. (1996), White and Stelzer (1999) Stryer (1978), Pollok and Heim (1999), LippincottSchwartz et al. (2001) Lakowicz et al. (1992), Bastiaens and Squire (1999), Pepperkok et al. (1999) Krichevsky and Bonnet (2002), Weiss and Nilsson (2004) Axelrod (2001), Toomre and Manstein (2001) Waterman-Storer et al. (1998), Waterman-Storer and Danuser (2002) Elowitz et al. (1997), Patterson and LippincottSchwartz (2002) Terskikh et al. (2000)

FLUORESCENCE MICROSCOPY-BASED TECHNIQUES USING GFP

14.3.1

Time-Lapse, Multispectral, and Ratio Imaging

GFP can be used to label the steady-state distribution of a molecule within a cell (Fig. 14.1A). However, perhaps the most widely used imaging technique for GFP-based mammalian cell studies is now time-lapse imaging, in which a single focal plane of a live cell specimen is observed over time (Fig. 14.1B). This has allowed the localization and dynamics of GFP chimeras to be studied in real time, providing enormous insights into a protein’s distribution and transport pathways (including their response to cellular perturbations such as drug treatments and temperature shifts). Previous work examining these issues relied on static images or “snapshots” of large populations of cells, in which a specified cellular response is often difficult or impossible to piece together. In addition to time-lapse imaging, recent work has utilized 4D microscopy, which involves the collection of threedimensional datasets over time (Gerlich et al., 2001). This allows the behavior of a protein to be examined within the entire cell. To analyze the changes in a fluorescent protein’s spatial and temporal behavior in such experiments, researchers have used computer-based visualization programs that can quantify and discriminate fluorescent signals (Bergsma et al., 2001). Use of GFP mutants that fluoresce or are excited at different wavelengths offers the possibility of double labeling to compare the distribution and dynamics of two different populations of proteins simultaneously within cells (Ellenberg et al., 1998). In this method, cells are doubly transfected with proteins attached to different GFP variants that have different excitation or emission spectra and are imaged with alternative filter sets (Rizzuto et al., 1996). With the development of spectral imaging systems, it is now theoretically possible to resolve all six fluorescent protein colors (BFP, CFP, GFP, YFP, DsRed, HcRed) within the same cell. The limitation then becomes whether the cell can actually cope with the overexpression of all the constructs and whether their localization and behavior are affected as a result. Pairs of images can also be quantified using digital image processing techniques to see if the ratio of intensity of the two populations changes with time. Such ratio imaging approaches have already been standardized and used with rhodamine and fluorescein tags in the endosomal system (Mayor et al., 1993), and they promise to be an important application of GFP variants (White et al., 2001).

14.3.2

Photobleaching Techniques: FRAP, FLAP, iFRAP, and FLIP

The time-lapse and multispectral imaging techniques mentioned above can provide important information about the steady-state distribution of a protein over time, but they do not address the kinetic properties of a protein, such as whether the protein is free to diffuse through the cell or is attached to a matrix, or whether it is undergoing exchange between compartments or on/off a substrate. To obtain this type of information, a researcher must differentiate a selected pool of fluorescent proteins from other fluorescent molecules and then follow that pool as it equilibrates with other molecules over time. This differentiation can be accomplished using photobleaching techniques, in which an area of the cell is photobleached with a high-intensity laser pulse and the movement of unbleached molecules from neighboring areas into the bleached area is recorded by time-lapse microscopy (Fig. 14.2). Because photobleaching alters the fluorescence steady state in a cell, the dynamics of a GFP chimera (i.e., its diffusion rate, binding constant, or intracellular trafficking routes) can be unraveled in the absence of conditions that disrupt protein pathways or create protein gradients. Perhaps the most widely used photobleaching technique is f luorescence recovery after photobleaching (FRAP), in which fluorescent proteins in a small area are irreversibly

311

312

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

bleached by an intense laser flash and recovery is measured using an attentuated laser beam (Edidin, 1994). This technique can provide an estimation of the effective diffusion coefficient (Deff) and mobile fraction (Mf) of a protein. Whereas Deff reflects the meansquared displacement that an idealized protein moves by random walk over time, Mf represents the fraction of fluorescent proteins that can diffuse into a bleached region during the time course of the experiment. By comparing the observed Deff and Mf to idealized values, one can determine whether the GFP chimera under study undergoes interactions with other molecules, is bound to a scaffold, or can freely diffuse within cells [Fig. 14.3; for review see Lippincott-Schwartz et al. (2001)]. With the advent of two-color GFP imaging (Ellenberg et al., 1998), another application of photobleaching that has become feasible is the technique called fluorescence localization after photobleaching (FLAP) (Dunn et al., 2002), which uses photobleaching to visualize protein trafficking or flux through a two-color bleaching protocol (White et al., 2001; Dunn et al., 2002). In this approach, proteins to be visualized are tagged with either the CFP or YFP (producing two chimeras of the same protein) and coexpressed in the same cells, or alternatively a single chimera is introduced carrying a CFP–YFP concatamer. The YFP and CFP molecules are then excited together, using a multitracking mode of a confocal microscope with a 514-nm laser line to excite YFP and a 458-nm laser line (or 413nm laser line) to excite CFP. YFP is selectively photobleached using the 514-nm laser line at maximum power, leaving CFP fluorescence unaltered. Fluorescence recovery is then tracked through image differencing by subtracting the image of the bleached fluorochrome from that of the unbleached fluorochrome. This technique is a great advance in photobleaching technology, since it exploits CFP as a visual reference to follow the dynamics of a targeted protein or organelle and, simultaneously, to follow and quantify YFP fluorescence recovery. Another application of photobleaching is called inverse FRAP (iFRAP), which can be used to reduce fluorescence from background noise to reveal faint populations of fluorescent proteins (Presley et al., 2002). In this approach, fluorescence surrounding a particular region of the cell is photobleached to allow visualization of fluorescent protein movement from the unbleached to bleached areas. As an example, photobleaching of fluorescence associated with the plasma membrane will allow visualization of organelle behavior inside the cell, which otherwise is masked by the plasma membrane fluorescence (Nichols et al., 2001). This approach has also been used to investigate intranuclear dynamics [e.g., exchange between the nucleolus and the nucleoplasm (Dundr et al., 2002)]. Fluorescence loss in photobleaching (FLIP) investigates fluorophore mobility and the continuity of various intracellular environments or compartments (Cole et al., 1996; Fig. 14.2). FLIP is similar to FRAP in that a region of interest is photobleached with a highpower laser; however, unlike FRAP, the region is repeatedly bleached over time to deplete the entire fluorescent pool. If photobleaching of one region depletes the entire fluorescence of the other, then the fluorescent molecules are capable of freely diffusing between the two regions. Thus, by using this technique, it is possible to address whether a protein can diffuse uniformly across a compartment or whether there are regions of restricted mobility.

14.3.3

FRET and FCS

The two fluorescence-based techniques, fluorescence resonance energy transfer (FRET) and fluorescence correlation spectroscopy (FCS), enable protein–protein interactions to be spatially and temporally resolved in living cells. Whereas FRET detects the close

FLUORESCENCE MICROSCOPY-BASED TECHNIQUES USING GFP

Figure 14.3. Distribution and mobilities of a nuclear envelope membrane protein, lamin B receptor (LBR) tagged with GFP in interphase membranes. At steady state, LBR-GFP is found localized within the ER network and in the inner nuclear envelope (NE). Qualitative FRAP experiments in ER and NE membranes in interphase cells expressing LBR-GFP show (left) photobleach recovery in ER membranes, and (right) photobleach recovery in NE membranes. Note the complete recovery of fluorescence in the ER and the lack of recovery in the NE. [Courtesy of Ellenberg et al. (1997).]

313

314

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

proximity of interacting proteins, FCS detects either changes in the diffusion or the codiffusion of bound species. The use of GFP chimeras with these techniques is providing quantitative data on the physicochemical properties of molecules within cells. The type of data differs from that shown by biochemical approaches in that the microscopy approaches effectively measure the ability of molecules to cluster together (as detailed below). However, FCS can determine the absolute concentration of a species in vivo, which, in conjunction with the diffusion measurements, permits calculation of the Kd for protein–protein interactions. In contrast to biochemical methods, this may not be measurable for the Kd of a small enzyme and its substrate, but is measurable for a protein that is interacting with a large substrate such as a complex of proteins or a membrane, for example. Using FCS, the GTPase Arf1, which transiently binds and releases from Golgi membranes, has been shown to be freely soluble in mitotic cells, rather than bound to membranes (Altan-Bonnet et al., 2003). Furthermore, because FCS and FRET are measurable in real time in living cells, transient interactions over short periods of time can be captured, which may be missed by more traditional approaches (such as immunoprecipitation or chemical cross-linking) that are dependent on populations of cells exhibiting sufficient fractionable protein–protein interaction to be measurable [for example, FCS has demonstrated Gag–Gag interactions in the cytosol of Rous sarcoma virus-infected cells (Larson et al., 2003)]. FRET works by measuring the transfer of photon energy from one fluorophore to another molecule when both are located within a few nanometers of each other (Stryer, 1978; Uster and Pagano, 1986; Tsien et al., 1993). If the energy of the excited fluorophore coincides with the energy needed to excite the absorber, then energy is transferred. This transfer results in loss of fluorescence intensity of donor and fluorescence emission from the acceptor. The working scale of FRET is less than or equal to 100 Å, in contrast to conventional light microscopy, which is a few tenths of a micron (Stryer, 1978). The availability of several different mutants of GFP opens the possibility of using FRET to probe inter- and intramolecular distances in proteins, allowing the possibility of mapping protein–protein interactions within cells by fluorescence microscopy. An advantage of intramolecular FRET is that the stoichiometry between the donor and acceptor fluorophore is fixed, thereby enabling the ratio of acceptor to donor fluorescence to accurately measure FRET changes. An excellent example of intramolecular FRET is from the study by Heim and Tsien (1996). They attached the GFP mutants, Y66H/Y145F and S65C, to the same protein by a 25-residue cleavable spacer and used the first as donor and the second as acceptor in FRET experiments. Proteolytic cleavage of the spacer resulted in the two protein domains diffusing apart, causing loss of green emission by the acceptor S65C domain and enhancement of blue emission from the donor domain. Since this early seminal study, FRET using GFP chimeras has become a regular tool in cell biology. Intermolecular FRET can detect interactions between two proteins in real time. However, the involvement of mixed complexes between the endogenous protein and its labeled counterpart becomes an issue, and the ratio of donor to acceptor expression is no longer fixed. FRET is then better measured through acceptor photobleaching or by fluorescence lifetime imaging microscopy (FLIM), which measures the decay kinetics of the excited state. FLIM requires specialized equipment and complex mathematical analysis but can be used in living cells to measure changes in pH, Ca2+ concentration, protein–protein interactions (by FRET/FLIM, because lifetime greatly decreases when FRET is occurring), and proteolytic processing (Lakowicz et al., 1992; Bastiaens and Squire, 1999; Ng et al., 1999; Pepperkok et al., 1999; Calleja et al., 2003; Lin et al., 2003).

FLUORESCENCE MICROSCOPY-BASED TECHNIQUES USING GFP

In contrast to FRET, FCS measures the fluctuations in photons resulting from diffusion of fluorescently labeled molecules in and out of a small, defined volume (~1 femtoliter). Because the fluctuations reflect the average number of fluorescent molecules in the volume and the time of their diffusion, parameters such as the concentration of the fluorescent molecule and its diffusion constant can be derived using this technique (Krichevsky and Bonnet, 2002). It is also possible to look at protein–protein interactions using FCS, because binding to another protein alters the protein’s diffusional mobility. This technique was used to great effect in a study looking at the membrane association of the small GTPase Arf1 during mitosis (Altan-Bonnet et al., 2003). Inactive Arf1 is cytosolic (i.e., it has a diffusional coefficient D of a small cytoplasmic molecule), whereas if a fraction of Arf1-GFP were active and associated with mitotic membranes (e.g., vesicles), it would be detected by FCS as a species that diffused slower (with a D characteristic of vesicles). Because >98% Arf1 was found to diffuse as a small molecule, it could be inferred that Arf1 is persistently inactive during metaphase and does not associate with membranes. The degree of protein interaction or identity of the binding partner can be confirmed using fluorescence cross-correlation spectroscopy (FCCS), where two differently labeled proteins can be monitored together (Pyenta et al., 2001; Bacia et al., 2002; Weiss and Nilsson, 2004). FCS is currently less popular than FRET, but given its ability to measure both concentrations and diffusion constants, and the availability of confocal microscopes capable of sampling small volumes, FCS holds great promise for advancing our understanding of protein behavior in vivo (Elsner et al., 2003; Fradin et al., 2003; Weiss et al., 2003).

14.3.4 Total Internal Reflection Fluorescence and Fluorescent Speckle Microscopy Total internal reflection fluorescence microscopy (TIR-FM), or evanescent wave microscopy, allows processes to be imaged that only occur within very close proximity to the coverslip. This is accomplished by directing an excitatory laser beam through the coverslip at an angle steep enough so that it completely reflects off the water–coverslip interface. The result is the production of an evanescent field, in which a layer of ~100 nm is excited. Because only fluorescent molecules within this distance from the coverslip are excited, a high signal-to-background imaging of surface events is possible (Axelrod, 2001; Toomre and Manstein, 2001). GFP chimeras expressed in mammalian cells imaged using this technique have revealed important new insights into the mechanism(s) underlying fusion of (a) constitutive membrane transport carriers with the plasma membrane [Fig. 14.4; Schmoranzer et al., 2000; Toomre et al., 2000; Kreitzer et al., 2003; Schmoranzer and Simon, 2003) and (b) regulated secretory organelles (e.g., Weibel–Palade bodies (Manneville et al., 2003)]. In addition, interactions between microtubules and focal adhesion contacts at the cell surface have been analyzed with this approach (Krylyshkina et al., 2003). Also, dual-color TIR-FM has been used to image the interrelationship of clathrin-coated pits with actin and dynamin (Merrifield et al., 2002). Fluorescent speckle microscopy (FSM) is a technique that has been used for visualizing the movement, assembly, and turnover of macromolecular assemblies like the cytoskeleton in living cells. In this method, a fluorescently labeled protein is introduced into a cell at very low levels (0.1–0.5%) such that it can co-assemble as a small fraction of fluorescent subunits in a pool of unlabeled subunits. Because the labeled proteins are randomly incorporated into the polymer lattice, the individual molecules can be detected as a “fluorescent speckle” pattern. Movement of the speckles within cytoskeletal filaments

315

316

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

Figure 14.4. Visualizing fusion of post-Golgi carriers with the plasma membrane using total internal reflection fluorescence microscopy (TIR-FM). Selected frames from a sequence showing the transport, docking, and fusion of a tubular carrier labelled with VSVG-GFP. Times are marked relative to the start of fusion. [Courtesy of Schmoranzer et al. (2000); reproduced from The Journal of Cell Biology 149:23–32 by copyright permission of The Rockefeller University Press.]

can be detected as they move from the site of assembly to regions of depolymerization (Waterman-Storer et al., 1998). The technique works with either microinjection of fluorescently labeled subunits or expression of subunits ligated to GFP. However, sometimes multiple GFP sequences need to be attached to the protein-encoding gene to get enough fluorescent signal from a single protein molecule.

14.4

APPLICATIONS OF GFP IN MAMMALIAN CELLS

Given the numerous fluorescence-based techniques available for monitoring GFP, it is not surprising that an enormous variety of applications for GFP in mammalian cells have been developed (see Table 14.3). These range from the determination of a protein’s geography, movement, and molecular interactions to the development of gene therapy vectors and cell sorting protocols, as discussed below.

14.4.1 Determining a Protein’s Localization, Dynamics, and Concentration Within Cells The most widespread application of GFP in mammalian cells has been for characterizing the location and dynamics of proteins expressed as fusion partners with GFP. GFP chimeras provide a major advance over previous methods for studying the intracellular localization and dynamics of proteins (Fig. 14.1). Previous techniques required fixation and permeabilization methods to gain access within the cell to the protein of interest. Numerous problems in specimen preparations often arise as a result of fixation including the danger of extracting or damaging antigen and the possibility that labeling efficiencies within different cell structures will differ. With GFP chimeras, these problems are avoided because the protein of interest is viewed in a living, unperturbed cell. The GFP reporter,

APPLICATIONS OF GFP IN MAMMALIAN CELLS

TABLE 14.3. Applications for GFP in Mammalian Cells Applications

Example References

Protein localization

Presley et al. (1997), Chao et al. (1999), Zaal et al. (1999), Griffis et al. (2002) Sutherland et al. (2001), Conrad et al. (2004), Sineshchekova et al. (2004) Cole et al. (1996), Ellenberg et al. (1997), Nehls et al. (2000), Phair and Misteli (2000), Daigle et al. (2001), Stenoien et al. (2001), Elsner et al. (2003), Weiss et al. (2003), Kenworthy et al. (2004), Shav-Tal et al. (2004) Vasudevan et al. (1998), Gaidarov et al. (1999), Wu et al. (2001b), Presley et al. (2002), Salmon et al. (2002), Altan-Bonnet et al. (2003), Elsner et al. (2003), Weiss and Nilsson (2003), Dundr et al. (2004), Engqvist-Goldstein et al. (2004) Hirschberg et al. (1998), Zaal et al. (1999), Dahm et al. (2001), Nichols et al. (2001) Majoul et al. (2001), Zacharias et al. (2002), Hayes et al. (2004), Snapp et al. (2004) Cole et al. (1996), Zaal et al. (1999), Nehls et al. (2000), Nichols et al. (2001) Presley et al. (1997), Scales et al. (1997), Hirschberg et al. (1998), Toomre et al. (1999), Nichols et al. (2001) Moriyoshi et al. (1996), Mosser et al. (1997), Mancia et al. (2004) Bartlett et al. (1995), Dorsky et al. (1996) Sönnichsen et al. (2000), Keller et al. (2001), Nichols et al. (2001), Presley et al. (2002), Stephens and Pepperkok (2002, 2004), Kreitzer et al. (2003), Mironov et al. (2003), Polishchuk et al. (2004) Elliott and O’Hare (1999), Pelkmans et al. (2002), Meulenbroek et al. (2004) Mahajan et al. (1999), Wiegand et al. (2003), Cook and Hinkle (2004) Ellenberg et al. (1997), Zaal et al. (1999), Bergeland et al. (2001), Jokitalo et al. (2001), Ward et al. (2001), Beaudouin et al. (2002), Salina et al. (2002), Gerlich et al. (2003), Walter et al. (2003) Presley et al. (1997), Chao et al. (1999), Tvaruskó et al. (1999), Eils et al. (2000), Stephens et al. (2000), Dahm et al. (2001), Gerlich et al. (2001), Wu et al. (2001a), Schmoranzer and Simon (2003) Anderson et al. (1996), Ropp et al. (1996), Mosser et al. (1997), Espinet et al. (2000) Niswender et al. (1995), Terasaki et al. (1996), Hirschberg et al. (1998)

GFP gene trap Protein diffusion rates

Binding and dissociation constants

Rate constants for intracellular transport steps Protein–protein interactions Compartment connectivity Visualization of transport intermediates

Expression marker Viral gene marker Intracellular sorting

Monitoring viral infection Protein turnover Organelle dynamics and assembly

Protein tracking

FACS Protein concentration within cells or compartments

317

318

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

itself, usually does not interfere with the normal functioning, or targeting, of the tagged protein and can be added to either the C- or N-terminus of target proteins. Hence, use of GFP chimeras has become the primary means for identifying and studying a protein’s distribution within cells. In addition to offering a simple way to localize proteins within living cells, GFP fusion proteins have become an important tool for understanding protein dynamics, because distinct populations of GFP chimeras can be easily highlighted using photobleaching or photoactivation techniques. This has revolutionized our understanding of complex processes and structures within cells. In addition to describing a protein’s dynamics within cells, use of GFP chimeras allows the effective intracellular concentration of these molecules to be determined. This can be accomplished by comparing the intensity of a GFP chimera’s fluorescence with fluorescence of a known fluorescent protein standard (whose concentration in solution is known) using a sensitive camera system (e.g., a cooled CCD camera) or confocal microscope (Niswender et al., 1995; Terasaki et al., 1996). Knowledge of a protein’s concentration within cells is invaluable for the development and testing of mathematical models describing a protein’s biophysical parameters (including its binding and dissociation constants, or rate constants). As one example, Hirschberg et al. (1998) used a standardized recombinant GFP solution as a control to quantify the amount of VSVG-GFP protein in different subcellular compartments of mammalian tissue culture cells over time. This information was then used to model the kinetics by which a bolus of secretory cargo trafficked between compartments of the secretory pathway. Quantitative modeling can also be used to track protein residence times through organelle-specific photobleaching (Zaal et al., 1999; Dahm et al., 2001; Nichols et al., 2001; Presley et al., 2002; Weiss and Nilsson, 2003).

14.4.2

GFP as a Co-transfection or Expression Marker

GFP has been used as an expression marker in order to determine a protein’s expression level in the absence of direct tagging with GFP. This method is useful when a protein of interest cannot be directly tagged with GFP. Several approaches have been used to achieve this, including: transfection of two plasmids in tandem (White et al., 2001); transfection of two transcription initiation startpoints within the same vector, one transcribing GFP, the other the protein of interest (Mancia et al., 2004); or use of a polycistronic vector where the GFP is translated off the same RNA as, but not fused to, another protein through the use of an internal ribosome entry site [IRES (Mosser et al., 1997)].

14.4.3

Gene Targeting Using Viral Systems

The use of viral vector systems to express GFP has been used to monitor production and release of therapeutic molecules from cells and tissues. Bartlett et al. (1995) developed an adenovirus vector delivery system using GFP inserted downstream from the human muscle creatine kinase promoter and found efficient GFP expression in skeletal muscle injected with the vector. In contrast to traditional reporter methods including b-galactosidase, firefly luciferase, or chloramphenicol amino transferase (CAT) assays (which require cell lysis and introduction of the reporter enzyme substrate), GFP expression could be monitored consecutively over several days. In addition to the enormous potential of viral reporter genes carrying GFP in clinical studies for tracking the expression of gene products, such molecules are also extremely valuable for basic research. Moriyoshi et al. (1996), for example, used an adenovirus vector to transfer GFP into postmitotic neuronal cells in vivo

GFP REVELATIONS

to study cell migration and development of neuronal connections. Adeno-associated virus vectors expressing GFP were also used to target GFP to spinal neurons (Peel et al., 1997), allowing the fate of neurons to be followed and their response to various transducers analyzed.

14.4.4

Viral Infection and Pathogenesis

Viral vectors containing GFP have been used to monitor viral infection and pathogenesis with no need for processing of cells to detect infected cells. Using this approach, Dorsky et al. (1996) identified human immunodeficiency virus (HIV)-1-infected cells in tissue using GFP tagged HIV-1. GFP under the control of HIV-1 LTR promoter was readily detected in virally infected cells either by fluorescence microscopy or by fluorescenceactivated cell sorting. With the same goal in mind, Dhandayuthapani et al. (1995) used a mycobacterial shuttle-plasmid vector carrying GFP cDNA to assess mycobacterial interactions with macrophages. More recently, GFP has been used to comprehend the cytology of viral infection directly by visualizing cell uptake and viral factories in vivo (Elliott and O’Hare, 1999; Chen and Ahlquist, 2000; Ward and Moss, 2001; Potel et al., 2002; Taylor et al., 2003; La Boissière et al., 2004; Moradpour et al., 2004).

14.4.5

Flow Cytometry

Screening and selection of cells by flow cytometry has been greatly facilitated using GFP expression, since it provides an easy method for fluorescent labeling of viable cells. This method eliminates the task of characterizing cell lines through standard biochemical methods involving protein analysis. As an example, Mosser et al. (1997) generated a dicistronic mRNA encoding both a gene of interest and the gene for GFP. Clone selection involved the simple monitoring for GFP fluorescence using the fluorescence-activated cell sorter. Quantitative detection from two different genes within single mammalian cells has also been demonstrated using multiparameter flow cytometry with GFP and its red-shifted variant (Anderson et al., 1996) or EGFP and YFP (Espinet et al., 2000).

14.5

GFP REVELATIONS

Given the saturation of current cell biology publications with the use of GFP chimeras, this review is unable to cover the entire extent of the literature. The following sections, therefore, will focus on areas of mammalian cell biology that have been crucially changed by GFP-based experimental approaches.

14.5.1

Cytoskeleton

The cytoskeleton has been “illuminated” by the use of GFP. Not only can individual polymer subunits be independently labeled, but associated proteins including motor proteins and binding proteins can be tagged with GFP. Direct labeling of cytoskeletal subunits was initially used primarily as a noninvasive label of the cytoskeleton upon which to watch movements of vesicular intermediates or membranes (Robbins et al., 1999; Toomre et al., 1999) or of mitotic chromosomes (Haraguchi et al., 1999) in living cells (Ludin and Matus, 1998). With the introduction of fluorescence speckle microscopy (FSM) (see above), the dynamics of growth and shrinkage of GFP-tagged actin and microtubule

319

320

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

filaments has now been analyzed (Waterman-Storer et al., 1998; Watanabe and Mitchison, 2002). Furthermore, use of dual-wavelength FSM has enabled the relative dynamics of GFP-labeled f-actin and microtubules to be monitored in migrating cells. These studies have shown that the movement and organization of f-actin helps to coordinate the dynamic organization of microtubules, suggesting that these two cytoskeletal components dynamically bind and interact with each other in vivo (Salmon et al., 2002). GFP-labeling of auxiliary proteins such as those that tether actin filaments to focal contacts and fibrillar adhesions (Zamir et al., 2000) have also revealed that these molecules are dynamic, moving at an average rate of 19 mm/h. Other important new insights into the cytoskeleton using GFP chimeras have been obtained using TIR-FM (Krylyshkina et al., 2003). Using either GFP-tubulin to label microtubules or GFP-CLIP-170 to label microtubule ends, these studies showed that microtubules consistently track to adhesion complexes labeled with DsRed-zyxin, suggesting that microtubules can provide tracks from the cell interior to specific zones of the plasma membrane.

14.5.2 Nucleus Our understanding of nuclear architecture and dynamics has been dramatically improved due to the use of GFP in the study of nuclear behavior. Perhaps most significant has been the realization that interphase nuclear organization, once thought to be comprised of stably associated components assembled into rigid arrays, is extremely dynamic (both spatially and temporally) and is capable of self-organization (Misteli, 2001; Janicki and Spector, 2003; Bubulya and Spector, 2004). Using time-lapse imaging and photobleaching approaches to probe the behavior of different GFP-tagged nuclear components, researchers have found that many nuclear proteins are undergoing rapid movement within the nucleoplasm. The linker histone, H1, for example, was found to undergo rapid association and dissociation with chromatin (Misteli et al., 2000). Likewise, only a transient association of the transcription factor, glucocorticord receptor, with its promoter elements was observed (McNally et al., 2000). Moreover, RNA polymerase I and II components, DNA topoisomerase II, as well as DNA repair complexes were found to be highly mobile in the nucleoplasm (Houtsmuller et al., 1999; Becker et al., 2002; Christensen et al., 2002; Dundr et al., 2002; Essers et al., 2002; Kimura et al., 2002). An overriding theme from these studies is that many multiprotein complexes in the nucleus are very dynamic and do not form the stable holo-complexes previously predicted by biochemical methods. In the face of so many nuclear components being highly dynamic, the challenge has become one of determining what mechanisms allow the nucleus to establish and maintain its organization. This organization has been highlighted by the identification of nuclear subcompartments using a large-scale screen with a GFP gene trap technique, which found many discretely localized proteins (i.e., splice-rich speckles and nucleoli) (Sutherland et al., 2001). The current thinking is that some proteins act as structural components to create scaffolds onto which other components dynamically associate through transient interactions. This fits with photobleaching results such as those demonstrating that core histones labeled with GFP exhibit tight association with chromatin, whereas linker histones are only transiently associated (Lever et al., 2000; Misteli et al., 2000; Phair and Misteli, 2000; Kimura and Cook, 2001). An additional mechanism underlying nuclear organization could be through transient interactions between different components that are of variable strengths. This would allow different components to dynamically self-organize into steadystate assemblies within the nucleus, as has been suggested for Cajal bodies and speckles (Misteli, 2001; Lamond and Spector, 2003; Dundr et al., 2004).

GFP REVELATIONS

The regulation of chromosome position within the nucleus is another area in which GFP-based techniques is providing new insights. A study by Chubb et al. (2002) reported that chromosome dynamics in interphase correlated with nuclear positioning, with loci adjacent to nucleoli or to the nuclear periphery exhibiting more restricted movement than loci in nucleoplasmic locations. Two other studies revealed that during mitosis, metaphase chromosome positioning is inherited. One of these used the lac operator integrated into the genome to visualize a lac repressor-GFP fusion protein associated with chromosomes (Dietzel and Belmont, 2001), while the other used photobleaching of half the nucleus, either parallel to or perpendicular to the spindle axis, to address whether similar or different bleaching patterns reappeared after cell division (Gerlich et al., 2003). Results from both approaches were consistent with chromosome organization and behavior within mammalian cells being nonrandom. However, by visualizing a smaller population of GFPlabeled histone, a study by Walter et al. (2003) found that chromosome neighborhoods do not seem to be tightly maintained during mitosis. This highlights how different techniques applied to the same fusion proteins can reveal different conclusions. Another topic under active investigation is the dynamics of the nuclear envelope. GFP-based techniques have been used to study the organization and behavior of nuclear pore complexes. These studies have shown that nuclear pore complexes do not diffuse within the plane of the nuclear envelope and are composed of proteins that can be either stable [e.g., lamin B receptor, emerin, and lamins (Fig. 14.3; Ellenberg et al., 1997; Ostlund et al., 1999; Moir et al., 2000; Daigle et al., 2001)] or highly dynamic [nucleoporins (Daigle et al., 2001; Griffis et al., 2002)] proteins. They have further shown that nuclear lamina and nuclear pore complexes are interconnected, immobile two-dimensional networks that move synchronously during nuclear shape changes (Daigle et al., 2001). Live cell imaging and photobleaching studies of the nuclear envelope have also provided new insights into nuclear envelope breakdown and reassembly during mitosis (Ellenberg et al., 1997; Burke and Ellenberg, 2002). Such studies have shown that during mitosis, the nuclear envelope does not vesiculate, but is absorbed into the ER (Ellenberg et al., 1997). They have also demonstrated that initiation of nuclear envelope breakdown occurs by a microtubule-dependent tearing process combined with nuclear pore disassembly (Beaudouin et al., 2002; Salina et al., 2002), while nuclear envelope reassembly at the end of mitosis involves wrapping of ER membranes enriched in nuclear envelope components around chromatin (Ellenberg et al., 1997; Haraguchi et al., 2000; Moir et al., 2000; Holaska et al., 2002).

14.5.3

Membrane Trafficking and Organelle Dynamics

GFP-based studies are having an equally powerful impact on our understanding of the endomembrane of mammalian cells. Comprised of distinct organelles [including the endoplasmic reticulum (ER), Golgi apparatus, endosomes, lysosomes, and plasma membrane] and membrane-bound transport intermediates, the endomembrane system regulates the synthesis and sorting of all membrane-associated proteins within the cells. It is also responsible for protein secretion and the uptake of macromolecules from the extracellular environment. Among the major new findings obtained from GFP-based studies has been the recognition that most transport intermediates—long thought to be small, spherical vesicles of 60–90 nm—are, in fact, large, pleiomorphic structures capable of transforming into globular or tubular shapes depending on their interactions with cytoskeletal elements (Presley et al., 1997; Sciaky et al., 1997; Hirschberg et al., 1998; Shima et al., 1999; Toomre et al.,

321

322

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

1999; Blum et al., 2000; Polishchuk et al., 2000, 2003; Schmoranzer et al., 2000; Mironov et al., 2003; Schmoranzer and Simon, 2003). These structures have been demonstrated using GFP-tagged cargo proteins as they move within membrane-bound carriers to different intracellular sites, and correlative electron microscopy has been used to verify that the large carriers seen by light microscopy are indeed tubular structures (Polishchuk et al., 2000, 2003; Mironov et al., 2003). Double-labeling of different cargo proteins has revealed that different proteins, particularly those destined for different cell surface domains, are capable of segregating into spatially distinct domains within a single carrier (Shima et al., 1999; Keller et al., 2001; Kreitzer et al., 2003; Polishchuk et al., 2004). Transport carriers all seem to be capable of interacting with microtubules and of moving along them in a directional manner (either plus- or minus-end directed). In studies examining the movement of GFP-tagged transport carriers destined for the plasma membrane, for example, the carriers underwent switching between microtubule tracks highlighted with rhodamine tubulin (Toomre et al., 1999) and moved in a stop-and-go manner (Hirschberg et al., 1998; Toomre et al., 1999) that was dependent on a kinesin-like motor (Kreitzer et al., 2000). Dual-color TIR-FM furthermore showed that such carriers remained associated with microtubules all the way to within 100 nm of the plasma membrane (Schmoranzer and Simon, 2003). While transport carriers can still dock and fuse with acceptor membranes in the absence of microtubules (Hirschberg et al., 1998), in cell types in which carriers must traverse significant distances to reach their destinations, such as neurons, microtubules are necessary (Rudolf et al., 2001; Hume et al., 2001; Wu et al., 2001a; Nakata and Hirokawa, 2003). Many transport carriers use cytosolic coat proteins (i.e., clathrin, COPI, and COPII) to pinch off from a membrane surface (Bonifacino and Lippincott-Schwartz, 2003). Recent photobleaching experiments using GFP-tagged versions of these coat proteins have revealed that their membrane binding, polymerization, and release from membranes occur rapidly and by a process that is not dependent on the carrier pinching off the membrane as a coated vesicle (Wu et al., 2001b; Presley et al., 2002). The role of small GTPases (Arf1 and Sar1) in the regulation of coat protein dynamics on membranes has also been addressed using GFP (Vasudevan et al., 1998; Ward et al., 2001; Presley et al., 2002; AltanBonnet et al., 2003; Elsner et al., 2003; Weiss and Nilsson, 2003). In cells expressing mutants of these proteins held in their GTP-bound state, the GTPases and their corresponding coat and effector molecules became irreversibly bound to membranes, exhibiting no recovery in photobleaching experiments (Presley et al., 2002). This contrasted with the normal behavior of these GTPases and their effectors, in which rapid cycling between membrane-bound and cytoplasmic pools was observed (Vasudevan et al., 1998; Stephens et al., 2000; Ward et al., 2001; Presley et al., 2002; Altan-Bonnet et al., 2003; García-Mata et al., 2003). Just as GFP-based studies have helped clarify key properties of transport carriers, they have also revealed important characteristics of membrane-bound organelles. In particular, they have shown that all secretory and endocytic organelles continuously exchange components with each other and they can undergo extensive changes in their structural organization. As one example, the ER was shown to transform from a network of branching tubules into stacked membrane arrays [termed organized smooth ER (OSER)] in response to elevated levels of resident components containing a dimerizing form of GFP on their cytoplasmic domains (Fig. 14.5) (Snapp et al., 2003). Because GFP is known to be capable of dimerizing in an antiparallel orientation (Yang et al., 1996a) through a low-affinity mechanism (Zacharias et al., 2002), one explanation of the OSER phenomenon is that it is caused by dimerization of GFP on apposing ER membranes (Snapp et al., 2003). This

GFP REVELATIONS

Figure 14.5. Morphological perturbation of the ER that occurs in response to the overexpression of a dimerizing form of GFP attached to the cytoplasmic tail of cytochrome b(5), an ER resident protein. Note that at low levels of chimera expression the ER looks normal (left panel), whereas at high levels (right panel) the ER is converted into swirls and tightly compacted stacks of membranes (called organized smooth ER or OSER). No change in the reticular pattern of the ER observed in the left panel was observed when a monomeric form of GFP was used to generate the chimera (not shown). [Courtesy of Snapp et al. (2003).]

could cause these membranes to then stack into geometric shapes. Consistent with this, when the authors expressed a nondimerizing form of GFP, no OSER structures were generated in the cells. Another example of organelle dynamics relates to the Golgi apparatus, which contains hundreds of diverse protein components with roles in the processing and sorting of secretory cargo. Photobleaching studies examining the residency time of different Golgi proteins on Golgi membranes revealed that no class of protein persisted stably (Storrie et al., 1998; Zaal et al., 1999; Miles et al., 2001; Nichols et al., 2001; Ward et al., 2001). Golgi processing enzymes stayed on the Golgi for ~60 min, cargo proteins for ~30 min, and cargo receptors and peripheral proteins for ~1 min before moving to other intracellular locations within the cell. These findings, together with other GFP-based observations on the transformation of Golgi membranes in response to specific stimuli (Sciaky et al., 1997; Presley et al., 1998; Feng et al., 2003) and the Golgi’s disassembly and reassembly during mitosis (Zaal et al., 1999; Jokitalo et al., 2001; Shorter and Warren, 2002), have led to the view that the Golgi apparatus is a dynamic, steady-state membrane system in a constant state of growth and consumption. New insights into the characteristics of lysosomes, which receive and digest endocytic cargo by hydrolytic enzymes, and of endosomes, which transfer material between the plasma membrane and other organelles, have also been obtained using GFP-based imaging approaches. Studies using PA-GFP photoactivation revealed that lysosomes undergo rapid interlysosome protein exchange (Patterson and Lippincott-Schwartz, 2002). This exchange was demonstrated in experiments in which a small population of lysosomes labeled with the lysosomal membrane protein (lgp120) tagged with PA-GFP was photoactivated and monitored over time. Because nearly all lysosomes were fluorescent within 20 min, transfer of lysosomal proteins between lysosomes is a rapid and extensive process. To clarify the properties of endosomes, Zerial and colleagues (Sönnichsen et al., 2000) used GFP-tagged Rab proteins and found that endosomes are comprised of a mosaic of

323

324

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

domains enriched in Rab4, Rab5, and Rab11. These domains are dynamic yet do not significantly intermix, and they display differential pharmacological sensitivities. GFP-based studies have provided new revelations regarding the properties of the plasma membrane, including the mechanism(s) for recruitment of signaling complexes to receptors (Choy et al., 1999; Gillham et al., 1999; Varnai et al., 1999), the lifetime of activated receptors and their substrates on the plasma membrane (Wouters and Bastiaens, 1999; Smith et al., 2001), the complexity of endocytic uptake pathways (Benmerah et al., 1999; Gaidarov et al., 1999; Roberts et al., 1999; Nichols et al., 2001; Wu et al., 2001b; Bacia et al., 2002; Merrifield et al., 2002; Mundy et al., 2002; Nichols, 2002; Pelkmans et al., 2002; Thomsen et al., 2002; Engqvist-Goldstein et al., 2004), and the dynamics of lipid rafts (Zacharias et al., 2002; Glebov and Nichols, 2004; Kenworthy et al., 2004). Lipid rafts are thought to be small regions of membrane inhomogeneity enriched in cholesterol and glycosphingolipids. However, to what extent lipid raft domains concentrate membrane proteins under steady-state conditions is a controversial question (Kenworthy, 2002; Munro, 2003; Simons and Vaz, 2004). FRAP was used to systematically measure the diffusion coefficients of several types of GFP-tagged raft and nonraft components on the plasma membrane in response to raft perturbations (Kenworthy et al., 2004). Because different raft proteins (i.e., lipid-anchored, GPI-anchored, and transmembrane proteins) were found to freely diffuse over large distances (>4 mm) at completely different rates, the data ruled out models in which raft proteins undergo long-range diffusion as part of discrete, stable raft domains. Instead, they supported the view that raft proteins rapidly partition into and out of cholesterol-enriched membrane domains.

14.6

FUTURE DIRECTIONS

The hundreds of mammalian studies using GFP-based techniques that have been published to date (many of which are unmentioned in this review) attest to the revolutionary impact that GFP is now having on efforts to understand cellular processes and protein function in the complex environment within cells. Clearly, we are in a new era in which the continued developments in GFP techniques and applications, as well as in microscopy approaches, are having a tremendous impact in the study of protein dynamics and interactions within cells. Looking to the future, the engineering of new GFP-like fluorophores and reporter classes will be important given their potential for improving the detection limits and in vivo applicability of fluorescence-based reporters. Brighter and more red-shifted fluorescent proteins, for example, can provide probes for greater tissue penetration or as readouts for high-throughput approaches, as well as serve as additional tags for multispectral imaging and FRET-based methods. New developments in GFP-based indicators that are designed to respond to various biological events and signals will also be valuable. Currently, various biochemical parameters can be measured by the modulation of fluorescent spectra, such as pH using pHluorins, where fluorescence is reversibly quenched by low pH (Miesenböck et al., 1998), halide with the halide-sensitive YFP (Jayaraman et al., 2000), or indirect measurement of phosphorylation of tyrosine residues by recruitment of a YFP-tagged phosphotyrosinebinding SH2 domain construct (Kirchner et al., 2003). Indicators that rely on intramolecular FRET include those that measure protease activity (consisting of BFA and GFP with a protease-sensitive linker; Mahajan et al., 1999), direct phosphorylation using phocuses (Sato et al., 2002), or changes in calcium levels (i.e., cameleons, which consist of calmod-

REFERENCES

ulin and a CaM-binding protein sandwiched between YFP and CFP) (Miyawaki et al., 1997; Zaccolo et al., 2002, Zhang et al., 2002). By designing indicators for other cellular parameters, including those that are sensitive to metabolite concentrations or enzyme activity, or which show increased sensitivity for structural changes in a protein (Zhang et al., 2002), a set of powerful tools for probing the environment within a cell will become available. The incorporation of advanced microscopy techniques into everyday experiments will be needed to maximize the advantages of the new GFP-based reagents. One promising technique is single-molecule spectroscopy, which allows the visualization of specific molecular interactions such as EGF receptor dimerization (Sako et al., 2000) or conformational changes in voltage-gated ion channels (Harms et al., 2001). A different technique is correlative light-electron microscopy, which allows the distribution of molecules in a single fluorescent image to be analyzed at the electron microscopic level (Polishchuk et al., 2000). Fluorescence anisotropy microscopy is another emerging technique that elucidates the microenvironment of a protein by measuring the protein’s rotational diffusion (Rocheleau et al., 2003). Many other existing imaging techniques, including two-photon (Piston, 1999), TIR-FM, FCS, image correlation microscopy (ICM) (Petersen et al., 1993; Wiseman et al., 2004) and stimulated emission depletion (Dyba et al., 2003), should become more widely used given their ability, or potential, to visualize and quantify molecules and events at high spatial and temporal resolution. The enormous amount of data generated from these methods will necessitate the use of kinetic modeling and analysis tools in order to interpret the data. In summary, as newer versions of GFP and imaging techniques become available, the applications for GFP and GFP chimeras will continue to expand, from the analysis of small subcellular structures such as viruses, proteasomes, and mRNA, to movement in threedimensional matrices (Cukierman et al., 2001; Hegerfeldt et al., 2002; Petroll et al., 2003), imaging within living organs (Mempel et al., 2004), and whole-body in vivo imaging (Contag et al., 1998; Bacskai et al., 2003; Cook and Griffin, 2003; Zhang et al., 2004). In so doing, they will provide exciting new insights into the biology of mammalian cells.

REFERENCES Altan-Bonnet, N., Phair, R. D., Polishchuk, R. S., Weigert, R., and Lippincott-Schwartz, J. (2003). A role for Arf1 in mitotic Golgi disassembly, chromosome segregation, and cytokinesis. Proc. Natl. Acad. Sci. USA 100:13314–13319. Anderson, M. T., Tjioe, I. M., Lorincz, M. C., Parks, D. R., Herzenberg, L. A., Nolan, G. P., and Herzenberg, L. A. (1996). Simultaneous fluorescence-activated cell sorter analysis of two distinct transcriptional elements within a single cell using engineered green fluorescent proteins. Proc. Natl. Acad. Sci. USA 93:8508–8511. Ando, R., Hama, H., Yamamoto-Hino, M., Mizuno, H., and Miyawaki, A. (2002). An optical marker based on the UV-induced green-to-red photoconversion of a fluorescent protein. Proc. Natl. Acad. Sci. USA 99:12651–12656. Axelrod, D. (2001). Total internal reflection fluorescence microscopy in cell biology. Traffic 2:764–774. Bacia, K., Majoul, I. V., and Schwille, P. (2002). Probing the endocytic pathway in live cells using dual-color fluorescence cross-correlation analysis. Biophys. J. 83:1184–1193. Bacskai, B. J., Hickey, G. A., Skoch, J., Kajdasz, S. T., Wang, Y., Huang, G.-F., Mathis, C. A., Klunk, W. E., and Hyman, B. T. (2003). Four-dimensional multiphoton imaging of brain entry, amyloid

325

326

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

binding, and clearance of an amyloid-b ligand in transgenic mice. Proc. Natl. Acad. Sci. USA 100:12462–12467. Bartlett, R. J., Singer, J. T., Shean, M. K., Weindler, F. W., and Secore, S. L. (1995). Expression of a fluorescent reporter gene in skeletal muscle cells from a virus-based vector: Implications for diabetes treatment. Transplant. Proc. 27:3344. Bastiaens, P. I. H., and Squire, A. (1999). Fluorescence lifetime imaging microscopy: spatial resolution of biochemical processes in the cell. Trends Cell Biol. 9:48–52. Beaudouin, J., Gerlich, D., Daigle, N., Eils, R., and Ellenberg, J. (2002). Nuclear envelope breakdown proceeds by microtubule-induced tearing of the lamina. Cell 108:83–96. Becker, M., Baumann, C., John, S., Walker, D. A., Vigneron, M., McNally, J. G., and Hager, G. L. (2002). Dynamic behavior of transcription factors on a natural promoter in living cells. EMBO Rep. 3:1188–1194. Benmerah, A., Bayrou, M., Cerf-Bensussan, N., and Dautry-Varsat, A. (1999). Inhibition of clathrincoated pit assembly by an Eps15 mutant. J. Cell Sci. 112:1303–1311. Bergeland, T., Widerberg, J., Bakke, O., and Nordeng, T. W. (2001). Mitotic partitioning of endosomes and lysosomes. Curr. Biol. 11:644–651. Bergsma, C. B. J., Streekstra, G. J., Smeulders, A. W. M., and Manders, E. M. M. (2001). Velocity estimation of spots in three-dimensional confocal image sequences of living cells. Cytometry 43:261–272. Bevis, B. J., and Glick, B. S. (2002). Rapidly maturing variants of the Discosoma red fluorescent protein (DsRed). Nat. Biotechnol. 20:83–87. Blum, R., Stephens, D. J., and Schulz, I. (2000). Lumenal targeted GFP, used as a marker of soluble cargo, visualizes rapid ERGIC to Golgi traffic by a tubulo-vesicular network. J. Cell Sci. 113:3151–3159. Bonifacino, J. S., and Lippincott-Schwartz, J. (2003). Coat proteins: shaping membrane transport. Nat. Rev. Mol. Cell Biol. 4:409–414. Bubulya, P. A., and Spector, D. L. (2004). “On the movements of nuclear components in living cells.” Exp. Cell Res. 296:4–11. Burke, B., and Ellenberg, J. (2002). Remodelling the walls of the nucleus. Nat. Rev. Mol. Cell Biol. 3:487–497. Calleja, V., Ameer-Beg, S. M., Vojnovic, B., Woscholski, R., Downward, J., and Larijani, B. (2003). Monitoring conformational changes of proteins in cells by fluorescence lifetime imaging microscopy. Biochem. J. 372:33–40. Campbell, R. E., Tour, O., Palmer, A. E., Steinbach, P. A., Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2002). A monomeric red fluorescent protein. Proc. Natl. Acad. Sci. USA 99: 7877–7882. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263:802–805. Chao, D. S., Hay, J. C., Winnick, S., Prekeris, R., Klumperman, J., and Scheller, R. H. (1999). SNARE membrane trafficking dynamics in vivo. J. Cell Biol. 144:869–881. Chen, J., and Ahlquist, P. (2000). Brome mosaic virus polymerase-like protein 2a is directed to the endoplasmic reticulum by helicase-like viral protein 1a. J. Virol. 74:4310–4318. Chiu, W.-L., Niwa, Y., Zeng, W., Hirano, T., Kobayashi, H., and Sheen, J. (1996). Engineered GFP as a vital reporter in plants. Curr. Biol. 6:325–330. Choy, E., Chiu, V. K., Silletti, J., Feoktistov, M., Morimoto, T., Michaelson, D., Ivanov, I. E., and Philips, M. R. (1999). Endomembrane trafficking of Ras: The CAAX motif targets proteins to the ER and Golgi. Cell 98:69–80. Christensen, M. O., Larsen, M. K., Barthelmes, H. U., Hock, R., Andersen, C. L., Kjeldsen, E., Knudsen, B. R., Westergaard, O., Boege, F., and Mielke, C. (2002). Dynamics of human DNA topoisomerases IIa and IIb in living cells. J. Cell Biol. 157:31–44.

REFERENCES

Chubb, J. R., Boyle, S., Perry, P., and Bickmore, W. A. (2002). Chromatin motion is constrained by association with nuclear compartments in human cells. Curr. Biol. 12:439–445. Chudakov, D. M., Belousov, V. V., Zaraisky, A. G., Novoselov, V. V., Staroverov, D. B., Zorov, D. B., Lukyanov, S., and Lukyanov, K. A. (2003). Kindling fluorescent proteins for precise in vivo photolabeling. Nat. Biotechnol. 21:191–194. Clegg, R. M. (1995). Fluorescence resonance energy transfer. Curr. Opin. Biotechnol. 6:103–110. Cole, N. B., Smith, C. L., Sciaky, N., Terasaki, M., Edidin, M., and Lippincott-Schwartz, J. (1996). Diffusional mobility of Golgi proteins in membranes of living cells. Science 273: 797–801. Conrad, C., Erfle, H., Warnat, P., Daigle, N., Lörch, T., Ellenberg, J., Pepperkok, R., and Eils, R. (2004). Automatic identification of subcellular phenotypes on human cell arrays. Genome Res. 14:1130–1136. Contag, P. R., Olomu, I. N., Stevenson, D. K., and Contag, C. H. (1998). Bioluminescent indicators in living mammals. Nat. Med. 4:245–247. Cook, L. B., and Hinkle, P. M. (2004). Fate of internalized thyrotropin-releasing hormone receptors monitored with a timer fusion protein. Endocrinology 145:3095–3100. Cook, S. H., and Griffin, D. E. (2003). Luciferase imaging of a neurotropic viral infection in intact animals. J. Virol. 77:5333–5338. Cormack, B. P., Valdivia, R. H., and Falkow, S. (1996). FACS-optimized mutants of the green fluorescent protein (GFP). Gene 173:33–38. Crameri, A., Whitehorn, E. A., Tate, E., and Stemmer, W. P. (1996). Improved green fluorescent protein by molecular evolution using DNA shuffling. Nat. Biotechnol. 14:315–319. Cukierman, E., Pankov, R., Stevens, D. R., and Yamada, K. M. (2001). Taking cell-matrix adhesions to the third dimension. Science 294:1708–1712. Dahm, T., White, J., Grill, S., Füllekrug, J., and Stelzer, E. H. K. (2001). Quantitative ER ´ Golgi transport kinetics and protein separation upon Golgi exit revealed by vesicular integral membrane protein 36 dynamics in live cells. Mol. Biol. Cell 12:1481–1498. Daigle, N., Beaudouin, J., Hartnell, L., Imreh, G., Hallberg, E., Lippincott-Schwartz, J., and Ellenberg, J. (2001). Nuclear pore complexes form immobile networks and have a very low turnover in live mammalian cells. J. Cell Biol. 154:71–84. Delagrave, S., Hawtin, R. E., Silva, C. M., Yang, M. M., and Youvan, D. C. (1995). Red-shifted excitation mutants of the green fluorescent protein. Biotechnol. 13:151–154. Dhandayuthapani, S., Via, L. E., Thomas, C. A., Horowitz, P. M., Deretic, D., and Deretic, V. (1995). Green fluorescent protein as a marker for gene expression and cell biology of mycobacterial interactions with macrophages. Mol. Microbiol. 17:901–912. Dietzel, S., and Belmont, A. S. (2001). Reproducible but dynamic positioning of DNA in chromosomes during mitosis. Nat. Cell Biol. 3:767–770. Dorsky, D. I., Wells, M., and Harrington, R. D. (1996). Detection of HIV-1 infection with a green fluorescent protein reporter system. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 13:308–313. Dundr, M., McNally, J. G., Cohen, J., and Misteli, T. (2002). Quantitation of GFP-fusion proteins in single living cells. J. Struct. Biol. 140:92–99. Dundr, M., Hebert, M. D., Karpova, T. S., Stanek, D., Xu, H., Shpargel, K. B., Meier, U. T., Neugebauer, K. M., Matera, A. G., and Misteli, T. (2004). In vivo kinetics of Cajal body components. J. Cell Biol. 164:831–842. Dunn, G. A., Dobbie, I. M., Monypenny, J., Holt, M. R., and Zicha, D. (2002). Fluorescence localization after photobleaching (FLAP): a new method for studying protein dynamics in living cells. J. Microsc. 205:109–112. Dyba, M., Jakobs, S., and Hell, S. W. (2003). Immunofluorescence stimulated emission depletion microscopy. Nat. Biotechnol. 21:1303–1304.

327

328

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

Edidin, M. (1994). Fluorescence photobleaching and recovery, FPR, in the analysis of membrane structure and dynamics. In Mobility and Proximity in Biological Membranes, Damjanovich, S., Edidin, M., Szollosi, J., and Tron, L., Eds., CRC, Boca Raton, FL, pp. 109–135. Ehrig, T., O’Kane, D. J., and Prendergast, F. G. (1995). Green-fluorescent protein mutants with altered fluorescence excitation spectra. FEBS Lett. 367:163–166. Eils, R., Gerlich, D., Tvaruskó, W., Spector, D. L., and Misteli, T. (2000). Quantitative imaging of pre-mRNA splicing factors in living cells. Mol. Biol. Cell 11:413–418. Ellenberg, J., Siggia, E. D., Moreira, J. E., Smith, C. L., Presley, J. F., Worman, H. J., and Lippincott-Schwartz, J. (1997). Nuclear membrane dynamics and reassembly in living cells: targeting of an inner nuclear membrane protein in interphase and mitosis. J. Cell Biol. 138:1193– 1206. Ellenberg, J., Lippincott-Schwartz, J., and Presley, J. F. (1998). Two-color green fluorescent protein time-lapse imaging. Biotechniques 25:838–846. Elliott, G., and O’Hare, P. (1999). Live-cell analysis of a green fluorescent protein-tagged herpes simplex virus infection. J. Virol. 73:4110–4119. Elowitz, M. B., Surette, M. G., Wolf, P. E., Stock, J., and Leibler, S. (1997). Photoactivation turns green fluorescent protein red. Curr. Biol. 7:809–812. Elsner, M., Hashimoto, H., Simpson, J. C., Cassel, D., Nilsson, T., and Weiss, M. (2003). Spatiotemporal dynamics of the COPI vesicle machinery. EMBO Rep. 4:1000–1004. Engqvist-Goldstein, A. E. Y., Zhang, C. X., Carreno, S., Barroso, C., Heuser, J. E., and Drubin, D. G. (2004). RNAi-mediated Hip1R silencing results in stable association between the endocytic machinery and the actin assembly machinery. Mol. Biol. Cell 15:1666–1679. Espinet, C., Gómez-Arbonés, X., Egea, J., and Comella, J. X. (2000). Combined use of the green and yellow fluorescent proteins and fluorescence-activated cell sorting to select populations of transiently transfected PC12 cells. J. Neurosci. Methods 100:63–69. Essers, J., Houtsmuller, A. B., van Veelen, L., Paulusma, C., Nigg, A. L., Pastink, A., Vermeulen, W., Hoeijmakers, J. H. J., and Kanaar, R. (2002). Nuclear dynamics of RAD52 group homologous recombination proteins in response to DNA damage. EMBO J. 21:2030–2037. Feng, Y., Yu, S., Lasell, T. K. R., Jadhav, A. P., Macia, E., Chardin, P., Melancon, P., Roth, M., Mitchison, T., and Kirchhausen, T. (2003). Exo1: A new chemical inhibitor of the exocytic pathway. Proc. Natl. Acad. Sci. USA 100:6469–6474. Fradin, C., Abu-Arish, A., Granek, R., and Elbaum, M. (2003). Fluorescence correlation spectroscopy close to a fluctuating membrane. Biophys. J. 84:2005–2020. Gaidarov, I., Santini, F., Warren, R. A., and Keen, J. H. (1999). Spatial control of coated-pit dynamics in living cells. Nat. Cell Biol. 1:1–7. García-Mata, R., Szul, T., Alvarez, C., and Sztul, E. (2003). ADP-ribosylation factor/COPI-dependent events at the endoplasmic reticulum-Golgi interface are regulated by the guanine nucleotide exchange factor GBF1. Mol. Biol. Cell 14:2250–2261. Gerlich, D., Beaudouin, J., Gebhard, M., Ellenberg, J., and Eils, R. (2001). Four-dimensional imaging and quantitative reconstruction to analyse complex spatiotemporal processes in live cells. Nat. Cell Biol. 3:852–855. Gerlich, D., Beaudouin, J., Kalbfuss, B., Daigle, N., Eils, R., and Ellenberg, J. (2003). Global chromosome positions are transmitted through mitosis in mammalian cells. Cell 112:751–764. Gillham, H., Golding, M. C. H. M., Pepperkok, R., and Gullick, W. J. (1999). Intracellular movement of green fluorescent protein-tagged phosphatidylinositol 3-kinase in response to growth factor receptor signaling. J. Cell Biol. 146:869–880. Glebov, O. O., and Nichols, B. J. (2004). Lipid raft proteins have a random distribution during localized activation of the T-cell receptor. Nat. Cell Biol. 6:238–243.

REFERENCES

Griesbeck, O., Baird, G. S., Campbell, R. E., Zacharias, D. A., and Tsien, R. Y. (2001). Reducing the environmental sensitivity of yellow fluorescent protein. Mechanism and applications. J. Biol. Chem. 276:29188–29194. Griffis, E. R., Altan, N., Lippincott-Schwartz, J., and Powers, M. A. (2002). Nup98 is a mobile nucleoporin with transcription-dependent dynamics. Mol. Biol. Cell 13:1282–1297. Gurskaya, N. G., Fradkov, A. F., Terskikh, A., Matz, M. V., Labas, Y. A., Martynov, V. I., Yanushevich, Y. G., Lukyanov, K. A., and Lukyanov, S. A. (2001). GFP-like chromoproteins as a source of far-red fluorescent proteins. FEBS Lett. 507:16–20. Haraguchi, T., Ding, D. Q., Yamamoto, A., Kaneda, T., Koujin, T., and Hiraoka, Y. (1999). Multiple-color fluorescence imaging of chromosomes and microtubules in living cells. Cell Struct. Funct. 24:291–298. Haraguchi, T., Koujin, T., Hayakawa, T., Kaneda, T., Tsutsumi, C., Imamoto, N., Akazawa, C., Sukegawa, J., Yoneda, Y., and Hiraoka, Y. (2000). Live fluorescence imaging reveals early recruitment of emerin, LBR, RanBP2, and Nup153 to reforming functional nuclear envelopes. J. Cell Sci. 113:779–794. Harms, G. S., Cognet, L., Lommerse, P. H. M., Blab, G. A., Kahr, H., Gamsjäger, R., Spaink, H. P., Soldatov, N. M., Romanin, C., and Schmidt, T. (2001). Single-molecule imaging of L-type Ca2+ channels in live cells. Biophys. J. 81:2639–2646. Hayes, N., Howard-Cofield, E., and Gullick, W. (2004). Green fluorescent protein as a tool to study epidermal growth factor receptor function. Cancer Lett. 206:129–135. Hegerfeldt, Y., Tusch, M., Bröcker, E.-B., and Friedl, P. (2002). Collective cell movement in primary melanoma explants: Plasticity of cell–cell interaction, b1-integrin function, and migration strategies. Cancer Res. 62:2125–2130. Heim, R., and Tsien, R. Y. (1996). Engineering green fluorescent protein for improved brightness, longer wavelengths and fluorescence resonance energy transfer. Curr. Biol. 6:178–182. Heim, R., Prasher, D. C., and Tsien, R. Y. (1994). Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proc. Natl. Acad. Sci. USA 91:12501–12504. Heim, R., Cubitt, A. B., and Tsien, R. Y. (1995). Improved green fluorescence. Nature 373:663–664. Hirschberg, K., Miller, C. M., Ellenberg, J., Presley, J. F., Siggia, E. D., Phair, R. D., and Lippincott-Schwartz, J. (1998). Kinetic analysis of secretory protein traffic and characterization of Golgi to plasma membrane transport intermediates in living cells. J. Cell Biol. 143:1485– 1503. Holaska, J. M., Wilson, K. L., and Mansharamani, M. (2002). The nuclear envelope, lamins and nuclear assembly. Curr. Opin. Cell Biol. 14:357–364. Houtsmuller, A. B., Rademakers, S., Nigg, A. L., Hoogstraten, D., Hoeijmakers, J. H. J., and Vermeulen, W. (1999). Action of DNA repair endonuclease ERCC1/XPF in living cells. Science 284:958–961. Hume, A. N., Collinson, L. M., Rapak, A., Gomes, A. Q., Hopkins, C. R., and Seabra, M. C. (2001). Rab27a regulates the peripheral distribution of melanosomes in melanocytes. J. Cell Biol. 152:795–808. Inouye, S., and Tsuji, F. I. (1994). Aequorea green fluorescent protein. Expression of the gene and fluorescence characteristics of the recombinant protein. FEBS Lett. 341:277–280. Janicki, S. M., and Spector, D. L. (2003). Nuclear choreography: Interpretations from living cells. Curr. Opin. Cell Biol. 15:149–157. Jayaraman, S., Haggie, P., Wachter, R. M., Remington, S. J., and Verkman, A. S. (2000). Mechanism and cellular applications of a green fluorescent protein-based halide sensor. J. Biol. Chem. 275:6047–6050. Jokitalo, E., Cabrera-Poch, N., Warren, G., and Shima, D. T. (2001). Golgi clusters and vesicles mediate mitotic inheritance independently of the endoplasmic reticulum. J. Cell Biol. 154:317–330.

329

330

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

Kaether, C., and Gerdes, H.-H. (1995). Visualization of protein transport along the secretory pathway using green fluorescent protein. FEBS Lett. 369:267–271. Keller, P., Toomre, D., Diaz, E., White, J., and Simons, K. (2001). Multicolour imaging of postGolgi sorting and trafficking in live cells. Nat. Cell Biol. 3:140–149. Kenworthy, A. (2002). Peering inside lipid rafts and caveolae. Trends Biochem. Sci. 27:435–437. Kenworthy, A. K., Nichols, B. J., Remmert, C. L., Hendrix, G. M., Kumar, M., Zimmerberg, J., and Lippincott-Schwartz, J. (2004). Dynamics of putative raft-associated proteins at the cell surface. J. Cell Biol. 165:735–746. Kimura, H., and Cook, P. R. (2001). Kinetics of core histones in living human cells: Little exchange of H3 and H4 and some rapid exchange of H2B. J. Cell Biol. 153:1341–1353. Kimura, H., Sugaya, K., and Cook, P. R. (2002). The transcription cycle of RNA polymerase II in living cells. J. Cell Biol. 159:777–782. Kirchner, J., Kam, Z., Tzur, G., Bershadsky, A. D., and Geiger, B. (2003). Live-cell monitoring of tyrosine phosphorylation in focal adhesions following microtubule disruption. J. Cell Sci. 116:975–986. Kreitzer, G., Marmorstein, A., Okamoto, P., Vallee, R., and Rodriguez-Boulan, E. (2000). Kinesin and dynamin are required for post-Golgi transport of a plasma-membrane protein. Nat. Cell Biol. 2:125–127. Kreitzer, G., Schmoranzer, J., Low, S. H., Li, X., Gan, Y., Weimbs, T., Simon, S. M., and RodriguezBoulan, E. (2003). Three-dimensional analysis of post-Golgi carrier exocytosis in epithelial cells. Nat. Cell Biol. 5:126–136. Krichevsky, O., and Bonnet, G. (2002). Fluorescence correlation spectroscopy: the technique and its applications. Rep. Prog. Phys. 65:251–297. Krylyshkina, O., Anderson, K. I., Kaverina, I., Upmann, I., Manstein, D. J., Small, J. V., and Toomre, D. K. (2003). Nanometer targeting of microtubules to focal adhesions. J. Cell Biol. 161:853–859. La Boissière, S., Izeta, A., Malcomber, S., and O’Hare, P. (2004). Compartmentalization of VP16 in cells infected with recombinant herpes simplex virus expressing VP16-green fluorescent protein fusion proteins. J. Virol. 78:8002–8014. Lakowicz, J. R., Szmacinski, H., Nowaczyk, K., Berndt, K. W., and Johnson, M. (1992). Fluorescence lifetime imaging. Anal. Biochem. 202:316–330. Lamond, A. I., and Spector, D. L. (2003). Nuclear speckles: a model for nuclear organelles. Nat. Rev. Mol. Cell Biol. 4:605–612. Larson, D. R., Ma, Y. M., Vogt, V. M., and Webb, W. W. (2003). Direct measurement of Gag–Gag interaction during retrovirus assembly with FRET and fluorescence correlation spectroscopy. J. Cell Biol. 162:1233–1244. Lever, M. A., Th’ng, J. P., Sun, X., and Hendzel, M. J. (2000). Rapid exchange of histone H1.1 on chromatin in living human cells. Nature 408:873–876. Lin, H.-J., Herman, P., and Lakowicz, J. R. (2003). Fluorescence lifetime-resolved pH imaging of living cells. Cytometry 52A:77–89. Lippincott-Schwartz, J., Snapp, E., and Kenworthy, A. (2001). Studying protein dynamics in living cells. Nat. Rev. Mol. Cell Biol. 2:444–456. Ludin, B., and Matus, A. (1998). GFP illuminates the cytoskeleton. Trends Cell Biol. 8:72–77. Mahajan, N. P., Harrison-Shostak, D. C., Michaux, J., and Herman, B. (1999). Novel mutant green fluorescent protein protease substrates reveal the activation of specific caspases during apoptosis. Chem. Biol. 6:401–409. Majoul, I., Straub, M., Hell, S. W., Duden, R., and Söling, H. D. (2001). KDEL-cargo regulates interactions between proteins involved in COPI vesicle traffic: Measurements in living cells using FRET. Dev. Cell 1:139–153. Mancia, F., Patel, S. D., Rajala, M. W., Scherer, P. E., Nemes, A., Schieren, I., Hendrickson, W. A., and Shapiro, L. (2004). Optimization of protein production in mammalian cells with a coexpressed fluorescent marker. Structure 12:1355–1360.

REFERENCES

Manneville, J.-B., Etienne-Manneville, S., Skehel, P., Carter, T., Ogden, D., and Ferenczi, M. (2003). Interaction of the actin cytoskeleton with microtubules regulates secretory organelle movement near the plasma membrane in human endothelial cells. J. Cell Sci. 116:3927–3938. Marshall, J., Molloy, R., Moss, G., Howe, J. R., and Hughes, T. E. (1995). The jellyfish green fluorescent protein: A new tool for studying ion channel expression. Neuron 14:211–215. Mayor, S., Presley, J. F., and Maxfield, F. R. (1993). Sorting of membrane components from endosomes and subsequent recycling to the cell surface occurs by a bulk flow process. J. Cell Biol. 121:1257–1269. McNally, J. G., Müller, W. G., Walker, D., Wolford, R., and Hager, G. L. (2000). The glucocorticoid receptor: Rapid exchange with regulatory sites in living cells. Science 287:1262–1265. Mempel, T. R., Henrickson, S. E., and von Andrian, U. H. (2004). T-cell priming by dendritic cells in lymph nodes occurs in three distinct phases. Nature 427:154–159. Merrifield, C. J., Feldman, M. E., Wan, L., and Almers, W. (2002). Imaging actin and dynamin recruitment during invagination of single clathrin-coated pits. Nat. Cell Biol. 4:691–698. Meulenbroek, R. A., Sargent, K. L., Lunde, J., Jasmin, B. J., and Parks, R. J. (2004). Use of adenovirus protein IX (pIX) to display large polypeptides on the virion—Generation of fluorescent virus through the incorporation of pIX-GFP. Mol. Ther. 9:617–624. Miesenböck, G., De Angelis, D. A., and Rothman, J. E. (1998). Visualizing secretion and synaptic transmission with pH-sensitive green fluorescent proteins. Nature 394:192–195. Miles, S., McManus, H., Forsten, K. E., and Storrie, B. (2001). Evidence that the entire Golgi apparatus cycles in interphase HeLa cells: sensitivity of Golgi matrix proteins to an ER exit block. J. Cell Biol. 155:543–555. Mironov, A. A., Mironov, A. A., Jr., Beznoussenko, G. V., Trucco, A., Lupetti, P., Smith, J. D., Geerts, W. J. C., Koster, A. J., Burger, K. N. J., Martone, M. E., Deerinck, T. J., Ellisman, M. H., and Luini, A. (2003). ER-to-Golgi carriers arise through direct en bloc protrusion and multistage maturation of specialized ER exit domains. Dev. Cell 5:583–594. Misteli, T. (2001). The concept of self-organization in cellular architecture. J. Cell Biol. 155:181–185. Misteli, T., Gunjan, A., Hock, R., Bustin, M., and Brown, D. T. (2000). Dynamic binding of histone H1 to chromatin in living cells. Nature 408:877–881. Mitra, R. D., Silva, C. M., and Youvan, D. C. (1996). Fluorescence resonance energy transfer between blue-emitting and red-shifted excitation derivatives of the green fluorescent protein. Gene 173:13–17. Miyawaki, A., Llopis, J., Heim, R., McCaffery, J. M., Adams, J. A., Ikura, M., and Tsien, R. Y. (1997). Fluorescent indicators for Ca2+ based on green fluorescent proteins and calmodulin. Nature 388:882–887. Moir, R. D., Yoon, M., Khuon, S., and Goldman, R. D. (2000). Nuclear lamins A and B1: Different pathways of assembly during nuclear envelope formation in living cells. J. Cell Biol. 151:1155–1168. Moradpour, D., Evans, M. J., Gosert, R., Yuan, Z., Blum, H. E., Goff, S. P., Lindenbach, B. D., and Rice, C. M. (2004). Insertion of green fluorescent protein into non-structural protein 5A allows direct visualization of functional hepatitis C virus replication complexes. J. Virol. 78:7400–7409. Moriyoshi, K., Richards, L. J., Akazawa, C., O’Leary, D. D., and Nakanishi, S. (1996). Labeling neural cells using adenoviral gene transfer of membrane-targeted GFP. Neuron 16:255–260. Mosser, D. D., Caron, A. W., Bourget, L., Jolicoeur, P., and Massie, B. (1997). Use of a dicistronic expression cassette encoding the green fluorescent protein for the screening and selection of cells expressing inducible gene products. Biotechniques 22:150–154.

331

332

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

Mundy, D. I., Machleidt, T., Ying, Y. S., Anderson, R. G., and Bloom, G. S. (2002). Dual control of caveolar membrane traffic by microtubules and the actin cytoskeleton. J. Cell Sci. 115:4327– 4339. Munro, S. (2003). Lipid rafts: elusive or illusive? Cell 115:377–388. Nagai, T., Ibata, K., Park, E. S., Kubota, M., Mikoshiba, K., and Miyawaki, A. (2002). A variant of yellow fluorescent protein with fast and efficient maturation for cell-biological applications. Nat. Biotechnol. 20:87–90. Nakata, T., and Hirokawa, N. (2003). Microtubules provide directional cues for polarized axonal transport through interaction with kinesin motor head. J. Cell Biol. 162:1045–1055. Nehls, S., Snapp, E. L., Cole, N. B., Zaal, K. J., Kenworthy, A. K., Roberts, T. H., Ellenberg, J., Presley, J. F., Siggia, E., and Lippincott-Schwartz, J. (2000). Dynamics and retention of misfolded proteins in native ER membranes. Nat. Cell Biol. 2:288–295. Ng, T., Squire, A., Hansra, G., Bornancin, F., Prevostel, C., Hanby, A., Harris, W., Barnes, D., Schmidt, S., Mellor, H., Bastiaens, P. I., and Parker, P. J. (1999). Imaging protein kinase Ca activation in cells. Science 283:2085–2089. Nichols, B. J. (2002). A distinct class of endosome mediates clathrin-independent endocytosis to the Golgi complex. Nat. Cell Biol. 4:374–378. Nichols, B. J., Kenworthy, A. K., Polishchuk, R. S., Lodge, R., Roberts, T. H., Hirschberg, K., Phair, R. D., and Lippincott-Schwartz, J. (2001). Rapid cycling of lipid raft markers between the cell surface and Golgi complex. J. Cell Biol. 153:529–541. Niswender, K. D., Blackman, S. M., Rohde, L., Magnuson, M. A., and Piston, D. W. (1995). Quantitative imaging of green fluorescent protein in cultured cells: Comparison of microscopic techniques, use in fusion proteins and detection limits. J. Microsc. 180:109–116. Ogawa, H., Inouye, S., Tsuji, F. I., Yasuda, K., and Umesono, K. (1995). Localization, trafficking, and temperature-dependence of the Aequorea green fluorescent protein in cultured vertebrate cells. Proc. Natl. Acad. Sci. USA 92:11899–11903. Olson, K. R., McIntosh, J. R., and Olmsted, J. B. (1995). Analysis of MAP 4 function in living cells using green fluorescent protein (GFP) chimeras. J. Cell Biol. 130:639–650. Ormö, M., Cubitt, A. B., Kallio, K., Gross, L. A., Tsien, R. Y., and Remington, S. J. (1996). Crystal structure of the Aequorea victoria green fluorescent protein. Science 273:1392– 1395. Ostlund, C., Ellenberg, J., Hallberg, E., Lippincott-Schwartz, J., and Worman, H. J. (1999). Intracellular trafficking of emerin, the Emery–Dreifuss muscular dystrophy protein. J. Cell Sci. 112:1709–1719. Patterson, G. H., and Lippincott-Schwartz, J. (2002). A photoactivatable GFP for selective photolabeling of proteins and cells. Science 297:1873–1877. Peel, A. L., Zolotukhin, S., Schrimsher, G. W., Muzyczka, N., and Reier, P. J. (1997). Efficient transduction of green fluorescent protein in spinal cord neurons using adeno-associated virus vectors containing cell type-specific promoters. Gene Ther. 4:16–24. Pelkmans, L., Püntener, D., and Helenius, A. (2002). Local actin polymerization and dynamin recruitment in SV40-induced internalization of caveolae. Science 296:535–539. Pepperkok, R., Squire, A., Geley, S., and Bastiaens, P. I. (1999). Simultaneous detection of multiple green fluorescent proteins in live cells by fluorescence lifetime imaging microscopy. Curr. Biol. 9:269–272. Petersen, N. O., Hoddelius, P. L., Wiseman, P. W., Seger, O., and Magnusson, K. E. (1993). Quantitation of membrane receptor distributions by image correlation spectroscopy: concept and application. Biophys. J. 65:1135–1146. Petroll, W. M., Ma, L., and Jester, J. V. (2003). Direct correlation of collagen matrix deformation with focal adhesion dynamics in living corneal fibroblasts. J. Cell Sci. 116:1481–1491.

REFERENCES

Phair, R. D., and Misteli, T. (2000). High mobility of proteins in the mammalian cell nucleus. Nature 404:604–609. Piston, D. W. (1999). Imaging living cells and tissues by two-photon excitation microscopy. Trends Cell Biol. 9:66–69. Polishchuk, R. S., Polishchuk, E. V., Marra, P., Alberti, S., Buccione, R., Luini, A., and Mironov, A. A. (2000). Correlative light-electron microscopy reveals the tubular-saccular ultrastructure of carriers operating between Golgi apparatus and plasma membrane. J. Cell Biol. 148:45– 58. Polishchuk, E. V., Di Pentima, A., Luini, A., and Polishchuk, R. S. (2003). Mechanism of constitutive export from the Golgi: bulk flow via the formation, protrusion, and en bloc cleavage of large trans-Golgi network tubular domains. Mol. Biol. Cell 14:4470–4485. Polishchuk, R., Di Pentima, A., and Lippincott-Schwartz, J. (2004). Delivery of raft-associated, GPIanchored proteins to the apical surface of polarized MDCK cells by a transcytotic pathway. Nat. Cell Biol. 6:297–307. Pollok, B. A., and Heim, R. (1999). Using GFP in FRET-based applications. Trends Cell Biol. 9:57–60. Potel., C., Kaelin, K., Gautier, I., Lebon, P., Coppey, J., and Rozenberg, F. (2002). Incorporation of green fluorescent protein into the essential envelope glycoprotein B of herpes simplex virus type 1. J. Virol. Methods 105:13–23. Presley, J. F., Cole, N. B., Schroer, T. A., Hirschberg, K., Zaal, K. J. M., and Lippincott-Schwartz, J. (1997). ER-to-Golgi transport visualized in living cells. Nature 389:81–85. Presley, J. F., Smith, C., Hirschberg, K., Miller, C., Cole, N. B., Zaal, K. J. M., and LippincottSchwartz, J. (1998). Golgi membrane dynamics. Mol. Biol. Cell 9:1617–1626. Presley, J. F., Ward, T. H., Pfeifer, A. C., Siggia, E. D., Phair, R. D., and Lippincott-Schwartz, J. (2002). Dissection of COPI and Arf1 dynamics in vivo and role in Golgi membrane transport. Nature 417:187–193. Pyenta, P. S., Holowka, D., and Baird, B. (2001). Cross-correlation analysis of inner-leaflet-anchored green fluorescent protein co-redistributed with IgE receptors and outer leaflet lipid raft components. Biophys. J. 80:2120–2132. Rizzo, M. A., Springer, G. H., Granada, B., and Piston, D. W. (2004). An improved cyan fluorescent protein variant useful for FRET. Nat. Biotechnol. 22:445–449. Rizzuto, R., Brini, M., De Giorgi, F., Rossi, R., Heim, R., Tsien, R. Y., and Pozzan, T. (1996). Double labelling of subcellular structures with organelle-targeted GFP mutants in vivo. Curr. Biol. 6:183–188. Robbins, J. R., Barth, A. I., Marquis, H., de Hostos, E. L., Nelson, W. J., and Theriot, J. A. (1999). Listeria monocytogenes exploits normal host cell processes to spread from cell to cell. J. Cell Biol. 146:1333–1350. Roberts, R. L., Barbieri, M. A., Pryse, K. M., Chua, M., Morisaki, J. H., and Stahl, P. D. (1999). Endosome fusion in living cells overexpressing GFP-rab5. J. Cell Sci. 112:3667– 3675. Rocheleau, J. V., Edidin, M., and Piston, D. W. (2003). Intrasequence GFP in class I MHC molecules, a rigid probe for fluorescence anisotropy measurements of the membrane environment. Biophys. J. 84:4078–4086. Ropp, J. D., Donahue, C. J., Wolfgang-Kimball, D., Hooley, J. J., Chin, J. Y., Cuthbertson, R. A., and Bauer, K. D. (1996). Aequorea green fluorescent protein: simultaneous analysis of wild-type and blue-fluorescing mutant by flow cytometry. Cytometry 24:284–288. Rudolf, R., Salm, T., Rustom, A., and Gerdes, H.-H. (2001). Dynamics of immature secretory granules: Role of cytoskeletal elements during transport, cortical restriction, and F-actin-dependent tethering. Mol. Biol. Cell 12:1353–1365.

333

334

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

Sako, Y., Minoghchi, S., and Yanagida, T. (2000). Single-molecule imaging of EGFR signalling on the surface of living cells. Nat. Cell Biol. 2:168–172. Salina, D., Bodoor, K., Eckley, D. M., Schroer, T. A., Rattner, J. B., and Burke, B. (2002). Cytoplasmic dynein as a facilitator of nuclear envelope breakdown. Cell 108:97–107. Salmon, W. C., Adams, M. C., and Waterman-Storer, C. M. (2002). Dual-wavelength fluorescent speckle microscopy reveals coupling of microtubule and actin movements in migrating cells. J. Cell Biol. 158:31–37. Sato, M., Ozawa, T., Inukai, K., Asano, T., and Umezawa, Y. (2002). Fluorescent indicators for imaging protein phosphorylation in single living cells. Nat. Biotechnol. 20:287–294. Scales, S. J., Pepperkok, R., and Kreis, T. E. (1997). Visualization of ER-to-Golgi transport in living cells reveals a sequential mode of action for COPII and COPI. Cell 90:1137– 1148. Schmoranzer, J., and Simon, S. M. (2003). Role of microtubules in fusion of post-Golgi vesicles to the plasma membrane. Mol. Biol. Cell 14:1558–1569. Schmoranzer, J., Goulian, M., Axelrod, D., and Simon, S. M. (2000). Imaging constitutive exocytosis with total internal reflection fluorescence microscopy. J. Cell Biol. 149:23–32. Sciaky, N., Presley, J., Smith, C., Zaal, K. J. M., Cole, N., Moreira, J. E., Terasaki, M., Siggia, E., and Lippincott-Schwartz, J. (1997). Golgi tubule traffic and the effects of brefeldin A visualized in living cells. J. Cell Biol. 139:1137–1155. Shav-Tal, Y., Darzacq, X., Shenoy, S. M., Fusco, D., Janicki, S. M., Spector, D. L., and Singer, R. H. (2004). Dynamics of single mRNPs in nuclei of living cells. Science 304:1797–1800. Shima, D. T., Scales, S. J., Kreis, T. E., and Pepperkok, R. (1999). Segregation of COPI-rich and anterograde-cargo-rich domains in endoplasmic-reticulum-to-Golgi transport complexes. Curr. Biol. 9:821–824. Shorter, J., and Warren, G. (2002). Golgi architecture and inheritance. Annu. Rev. Cell Dev. Biol. 18:379–420. Simons, K., and Vaz, W. L. (2004). Model systems, lipid rafts, and cell membranes. Annu. Rev. Biophys. Biomol. Struct. 33:269–295. Sineshchekova, O. O., Kawate, T., Vdovychenko, O. V., and Sato, T. N. (2004). Protein-trap version 2.1: screening for expressed proteins in mammalian cells based on their localizations. BMC Cell Biol. 5:8. Smith, J., Yu, R., and Hinkle, P. M. (2001). Activation of MAPK by TRH requires clathrindependent endocytosis and PKC but not receptor interaction with b-arrestin or receptor endocytosis. Mol. Endocrinol. 15:1539–1548. Snapp, E. L., Hegde, R. S., Francolini, M., Lombardo, F., Colombo, S., Pedrazzini, E., Borgese, N., and Lippincott-Schwartz, J. (2003). Formation of stacked ER cisternae by low affinity protein interactions. J. Cell Biol. 163:257–269. Snapp, E. L., Reinhart, G. A., Bogert, B. A., Lippincott-Schwartz, J., and Hegde, R. S. (2004). The organization of engaged and quiescent translocons in the endoplasmic reticulum of mammalian cells. J. Cell Biol. 164:997–1007. Sönnichsen, B., De Renzis, S., Nielsen, E., Rietdorf, J., and Zerial, M. (2000). Distinct membrane domains on endosomes in the recycling pathway visualized by multicolor imaging of Rab4, Rab5, and Rab11. J. Cell Biol. 149:901–914. Stenoien, D. L., Nye, A. C., Mancini, M. G., Patel, K., Dutertre, M., O’Malley, B. W., Smith, C. L., Belmont, A. S., and Mancini, M. A. (2001). Ligand-mediated assembly and real-time cellular dynamics of estrogen receptor a-coactivator complexes in living cells. Mol. Cell. Biol. 21:4404–4412. Stephens, D. J., and Pepperkok, R. (2002). Imaging of procollagen transport reveals COPIdependent cargo sorting during ER-to-Golgi transport in mammalian cells. J. Cell Sci. 115:1149–1160.

REFERENCES

Stephens, D. J., and Pepperkok, R. (2004). Differential effects of a GTP-restricted mutant of Sar1p on segregation of cargo during export from the endoplasmic reticulum. J. Cell Sci. 117: 3635–3644. Stephens, D. J., Lin-Marq, N., Pagano, A., Pepperkok, R., and Paccaud, J. P. (2000). COPI-coated ER-to-Golgi transport complexes segregate from COPII in close proximity to ER exit sites. J. Cell Sci. 113:2177–2185. Storrie, B., White, J., Röttger, S., Stelzer, E. H., Suganuma, T., and Nilsson, T. (1998). Recycling of Golgi-resident glycosyltransferases through the ER reveals a novel pathway and provides an explanation for nocodazole-induced Golgi scattering. J. Cell Biol. 143:1505–1521. Stryer, L. (1978). Fluorescence energy transfer as a spectroscopic ruler. Annu. Rev. Biochem. 47:819–846. Sutherland, H. G., Mumford, G. K., Newton, K., Ford, L. V., Farrall, R., Dellaire, G., Caceres, J. F., and Bickmore, W. A. (2001). Large-scale identification of mammalian proteins localized to nuclear sub-compartments. Hum. Mol. Genet. 10:1995–2011. Taylor, T. J., McNamee, E. E., Day, C., and Knipe, D. M. (2003). Herpes simplex virus replication compartments can form by coalescence of smaller compartments. Virol. 309:232– 247. Terasaki, M., Jaffe, L. A., Hunnicutt, G. R., and Hammer, J. A., III (1996). Structural change of the endoplasmic reticulum during fertilization: Evidence for loss of membrane continuity using green fluorescent protein. Dev. Biol. 179:320–328. Terskikh, A., Fradkov, A., Ermakova, G., Zaraisky, A., Tan, P., Kajava, A. V., Zhao, X., Lukyanov, S., Matz, M., Kim, S., Weissman, I., and Siebert, P. (2000). “Fluorescent timer”: Protein that changes color with time. Science 290:1585–1588. Thomsen, P., Roepstorff, K., Stahlhut, M., and van Deurs, B. (2002). Caveolae are highly immobile plasma membrane microdomains, which are not involved in constitutive endocytic trafficking. Mol. Biol. Cell 13:238–250. Toomre, D., and Manstein, D. J. (2001). Lighting up the cell surface with evanescent wave microscopy. Trends Cell Biol. 11:298–303. Toomre, D., Keller, P., White, J., Olivo, J. C., and Simons, K. (1999). Dual-color visualization of trans-Golgi network to plasma membrane traffic along microtubules in living cells. J. Cell Sci. 112:21–33. Toomre, D., Steyer, J. A., Keller, P., Almers, W., and Simons, K. (2000). Fusion of constitutive membrane traffic with the cell surface observed by evanescent wave microscopy. J. Cell Biol. 149:33–40. Tsien, R. Y., Bacskai, B. J., and Adams, S. R. (1993). FRET for studying intracellular signalling. Trends Cell Biol. 3:242–245. Tvaruskó, W., Bentele, M., Misteli, T., Rudolf, R., Kaether, C., Spector, D. L., Gerdes, H. H., and Eils, R. (1999). Time-resolved analysis and visualization of dynamic processes in living cells. Proc. Natl. Acad. Sci. USA 96:7950–7955. Uster, P. S., and Pagano, R. E. (1986). Resonance energy transfer microscopy: observations of membrane-bound fluorescent probes in model membranes and in living cells. J. Cell Biol. 103:1221–1234. Varnai, P., Rother, K. I., and Balla, T. (1999). Phosphatidylinositol 3-kinase-dependent membrane association of the Bruton’s tyrosine kinase pleckstrin homology domain visualized in single living cells. J. Biol. Chem. 274:10983–10989. Vasudevan, C., Han, W., Tan, Y., Nie, Y., Li, D., Shome, K., Watkins, S. C., Levitan, E. S., and Romero, G. (1998). The distribution and translocation of the G protein ADP-ribosylation factor 1 in live cells is determined by its GTPase activity. J. Cell Sci. 111:1277–1285. Walter, J., Schermelleh, L., Cremer, M., Tashiro, S., and Cremer, T. (2003). Chromosome order in HeLa cells changes during mitosis and early G1, but is stably maintained during subsequent interphase stages. J. Cell Biol. 160:685–697.

335

336

THE USES OF GREEN FLUORESCENT PROTEIN IN MAMMALIAN CELLS

Wang, S., and Hazelrigg, T. (1994). Implications for bcd mRNA localization from spatial distribution of exu protein in Drosophila oogenesis. Nature 369:400–403. Ward, B. M., and Moss, B. (2001). Vaccinia virus intracellular movement is associated with microtubules and independent of actin tails. J. Virol. 75:11651–11663. Ward, T. H., Polishchuk, R. S., Caplan, S., Hirschberg, K., and Lippincott-Schwartz, J. (2001). Maintenance of Golgi structure and function depends on the integrity of ER export. J. Cell Biol. 155:557–570. Watanabe, N., and Mitchison, T. J. (2002). Single-molecule speckle analysis of actin filament turnover in lamellipodia. Science 295:1083–1086. Waterman-Storer, C. M., and Danuser, G. (2002). New directions for fluorescent speckle microscopy. Curr. Biol. 12:R633–R640. Waterman-Storer, C. M., Desai, A., Bulinski, J. C., and Salmon, E. D. (1998). Fluorescent speckle microscopy, a method to visualize the dynamics of protein assemblies in living cells. Curr. Biol. 8:1227–1230. Weiss, M., and Nilsson, T. (2003). A kinetic proof-reading mechanism for protein sorting. Traffic 4:65–73. Weiss, M., and Nilsson, T. (2004). In a mirror dimly: tracing the movements of molecules in living cells. Trends Cell Biol. 14:267–273. Weiss, M., Hashimoto, H., and Nilsson, T. (2003). Anomalous protein diffusion in living cells as seen by fluorescence correlation spectroscopy. Biophys. J. 84:4043–4052. White, J., Keller, P., and Stelzer, E. H. (2001). Spatial partitioning of secretory cargo from Golgi resident proteins in live cells. BMC Cell Biol. 2:19. White, J., and Stelzer, E. (1999). Photobleaching GFP reveals protein dynamics inside live cells. Trends Cell Biol. 9:61–65. Wiegand, U. K., Duncan, R. R., Greaves, J., Chow, R. H., Shipston, M. J., and Apps, D. K. (2003). Red, yellow, green go! A novel tool for microscopic segregation of secretory vesicle pools according to their age. Biochem. Soc. Trans. 31:851–856. Wiseman, P. W., Brown, C. M., Webb, D. J., Hebert, B., Johnson, N. L., Squier, J. A., Ellisman, M. H., and Horwitz, A. F. (2004). Spatial mapping of integrin interactions and dynamics during cell migration by Image Correlation Microscopy. J. Cell Sci. 117:5521–5534. Wouters, F. S., and Bastiaens, P. I. (1999). Fluorescence lifetime imaging of receptor tyrosine kinase activity in cells. Curr. Biol. 9:1127–1130. Wu, P., and Brand, L. (1994). Resonance energy transfer: Methods and applications. Anal. Biochem. 218:1–13. Wu, X., Rao, K., Bowers, M. B., Copeland, N. G., Jenkins, N. A., and Hammer, J. A., III (2001a). Rab27a enables myosin Va-dependent melanosome capture by recruiting the myosin to the organelle. J. Cell Sci. 114:1091–1100. Wu, X., Zhao, X., Baylor, L., Kaushal, S., Eisenberg, E., and Greene, L. E. (2001b). Clathrin exchange during clathrin-mediated endocytosis. J. Cell Biol. 155:291–300. Yang, T. T., Cheng, L., and Kain, S. R. (1996a). Optimized codon usage and chromophore mutations provide enhanced sensitivity with the green fluorescent protein. Nucleic Acids Res. 24:4592–4593. Yang, T. T., Kain, S. R., Kitts, P., Kondepudi, A., Yang, M. M., and Youvan, D. C. (1996b). Dual color microscopic imagery of cells expressing the green fluorescent protein and a red-shifted variant. Gene 173:19–23. Zaal, K. J. M., Smith, C. L., Polishchuk, R. S., Altan, N., Cole, N. B., Ellenberg, J., Hirschberg, K., Presley, J. F., Roberts, T. H., Siggia, E., Phair, R. D., and Lippincott-Schwartz, J. (1999). Golgi membranes are absorbed into and reemerge from the ER during mitosis. Cell 99:589–601. Zaccolo, M., Magalhães, P., and Pozzan, T. (2002). Compartmentalisation of cAMP and Ca2+ signals. Curr. Opin. Cell Biol. 14:160–166.

REFERENCES

Zacharias, D. A., Violin, J. D., Newton, A. C., and Tsien, R. Y. (2002). Partitioning of lipidmodified monomeric GFPs into membrane microdomains of live cells. Science 296:913–916. Zamir, E., Katz, M., Posen, Y., Erez, N., Yamada, K. M., Katz, B. Z., Lin, S., Lin, D. C., Bershadsky, A., Kam, Z., and Geiger, B. (2000). Dynamics and segregation of cell-matrix adhesions in cultured fibroblasts. Nat. Cell Biol. 2:191–196. Zhang, G., Gurtu, V., and Kain, S. R. (1996). An enhanced green fluorescent protein allows sensitive detection of gene transfer in mammalian cells. Biochem. Biophys. Res. Commun. 227:707–711. Zhang, J., Campbell, R. E., Ting, A. Y., and Tsien, R. Y. (2002). Creating new fluorescent probes for cell biology. Nat. Rev. Mol. Cell Biol. 3:906–918. Zhang, G.-J., Safran, M., Wei, W., Sorensen, E., Lassota, P., Zhelev, N., Neuberg, D. S., Shapiro, G., and Kaelin, W. G., Jr. (2004). Bioluminescent imaging of Cdk2 inhibition in vivo. Nat. Med. 10:643–648. Zolotukhin, S., Potter, M., Hauswirth, W. W., Guy, J., and Muzyczka, N. (1996). A “humanized” green fluorescent protein cDNA adapted for high-level expression in mammalian cells. J. Virol. 70:4646–4654.

337

15 PRACTICAL CONSIDERATIONS FOR USE OF REEF CORAL FLUORESCENT PROTEINS IN MAMMALIAN CELLS: APPLICATIONS IN FLUORESCENCE MICROSCOPY AND FLOW CYTOMETRY Yu Fang, Olivier Déry, Michael Haugwitz, and Pierre Turpin BD Biosciences Clontech, Palo Alto, CA

Steven R. Kain Agilent Technologies, Palo Alto, CA

The discovery of fluorescent proteins from nonbioluminescent reef corals [reef coral fluorescent proteins (RCFPs)] has greatly expanded the panel of emission wavelengths available for in vivo analyses of cellular events and has opened the door to many new multicolor applications (Matz et al., 1999). Indeed, the excitation maxima of the RCFPs range from 458 nm to 588 nm, and their emission maxima range from 489 nm to 618 nm. Whereas the color variants of Aequorea victoria green fluorescent protein GFP were generated by mutagenesis of a single parental gene, the RCFPs AmCyan1, ZsGreen1, ZsYellow1, DsRed2 (orange-red), AsRed2 (true red), and HcRed1 (far-red) are encoded by discrete genes and were isolated from distinct species (Table 15.1). The RCFP family members share at most 30% amino acid sequence identity with A. victoria GFP. However, the three-dimensional structure obtained from DsRed1 crystals show that at least one member of the RCFP family has a structure very similar to that of A. victoria GFP. These RCFP proteins differ not only in their excitation and emission spectra, but also in the time course of fluorescence development, relative brightness, and long-term expression. In this chapter, we describe the unique properties of each RCFP, their expression in mammalian cells, and their use and detection in multiplex applications.

Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

339

340

PRACTICAL CONSIDERATIONS FOR USE OF REEF CORAL FLUORESCENT PROTEINS IN MAMMALIAN CELLS

TABLE 15.1. Nomenclature of Reef Coral Fluorescent Proteins Commercial Name (Mutants)

Mutants (AA Sequence)

Species

Original Name (as published)

References

AmCyan1 ZsGreen1 ZsYellow1 DsRed2 DsRed-Express AsRed2 HcRed1

N34S, K68M N66M M129V R2A, K5E, K9T R2A, K5E, N6D F4L, K12R, F35L A2S, T36A, L122H

Anemonia Zoanthus Zoanthus Discosoma Discosoma Anemonia Heterectis

amFP486 zFP506 zFP540 drFP583 DsRed-T1 asFP595 HcRed2A

Matz et al. (1999) Matz et al. (1999) Matz et al. (1999) Matz et al. (1999) Bevis and Glick (2002) Lukyanov et al. (2000) Gurskaya et al. (2001)

Figure 15.1. Spectral properties of reef coral fluorescent proteins AmCyan1, ZsGreen1, ZsYellow1, DsRed2, AsRed2, and HcRed1. Except for the lower level of residual green emission, the spectrum for DsRed-Express (not shown) closely resembled that of DsRed2.

15.1

PROPERTIES OF RCFPs

With the exception of ZsYellow1 and ZsGreen1, which were both derived from the same Zoanthus gene, the genes encoding the RCFPs were isolated from different species: AmCyan1, DsRed2 (orange-red), AsRed2 (true red), and HcRed1 (far-red) were isolated in Anemonia majano, Discosoma, Anemonia sulcata, and Heteractis crispa, respectively. As can be seen in Fig. 15.1, RCFPs cover a broad range of excitation and emission wavelengths: Excitation maxima range between 458 nm and 588 nm, and emission maxima range from 489 nm to 618 nm. Also, the excitation spectrum of each RCFP is broad, whereas their emission spectra are quite narrow. Such properties are quite valuable for multiplexing applications. An important characteristic of RCFPs is their tendency to form homo-oligomers. Crystallographic analysis of the original DsRed1 revealed a tight tetrameric structure (Gross et al., 2000; Wall et al., 2000; Yarbrough at al., 2001). These structure analyses were supported by the findings of analytical ultracentrifugation experiments (Baird at al., 2000). The subunits of this tetramer are similar in structure to A. victoria GFP (Ormö et al., 1996). They form a beta-sheet barrel-like structure, containing an internal alpha helix that bears the chromophore-forming amino acids. It is likely that other RCFPs such

PROPERTIES OF RCFPs

as AmCyan1, ZsGreen1, ZsYellow1, and AsRed2 also form tetramers as suggested by pseudo-native gel electrophoresis (Yanushevich et al., 2002). Due to their tetrameric structure, it is difficult to predict the function and/or localization of proteins fused to RCFPs. It has also been shown that RCFPs tend to form higher-order structures (aggregates) in mammalian cells. The extent of aggregation varies depending on the cell type in which the RCFPs are expressed. Replacing positively charged amino acids at the extreme N-terminus by neutral or negatively charged amino acids via site-directed mutagenesis has increased the overall solubility of RCFPs (Yanushevich et al. 2002; Bevis and Glick, 2002). The tetrameric structure of RCFPs (dimeric in the case of HcRed1) might restrict their use as fusion tags with other cellular proteins and peptides. However, a variety of fusion proteins have been successfully expressed in mammalian cells using these RCFPs. For example, all RCFPs have been successfully targeted to intracellular organelles or compartments such as the nucleus (nuclear targeting signal fused to the C-terminus of RCFPs) and the mitochondria (mitochondrial targeting signal fused to the N-terminus of RCFPs). Fusion proteins of RCFPs (not tested with DsRed-Express) and protein kinase C (PKC) alpha translocate from the cytosol to the plasma membrane upon induction with PMA, as is the behavior of endogenous PKC alpha. Proper localization of RCFP fusions suggests that in many cases the biological function of the fusion partner is not adversely impacted by linkage to RCFPs. The biophysical properties of each fluorescent protein are summarized in Table 15.2.

15.1.1

AmCyan1

AmCyan1 is a mutant of the original wild-type amFP486 from Anemonia majano (Matz et al., 1999). Two amino acid substitutions, N34S and K68M, were introduced to enhance brightness of the expressed protein. AmCyan1 has an excitation maximum at 458 nm and an emission maximum at 489 nm, and it has greater fluorescence intensity than ECFP (Fig. 15.2A). When expressed in mammalian cells, AmCyan1 can be detected 8–12 h after transfection. In comparison to ECFP, AmCyan1 exhibits a higher relative fluorescence intensity (quantum yield 0.75, extinction coefficient 39,000). AmCyan1 is also more resilient than ECFP to destructive photobleaching, which is the fast loss of fluorescence upon an extended excitation period. Because AmCyan1 resists photobleaching, the fluorescence intensity remains stable during analysis with a spectrofluorometer or fluorescence microscope. AmCyan1 is rather insoluble and tends to aggregate in mammalian cells when it is overexpressed. AmCyan1 is well suited for genetic reporter assays, such as transcriptional reporting. However, AmCyan1 has also been successfully localized to intracellular organelles and compartments including the nucleus and the mitochondria using appropriate targeting sequences.

15.1.2

ZsGreen1

ZsGreen1 is a mutant of wild-type zFP506 from a species of Zoanthus (Matz et al., 1999). A single amino acid substitution (N66M) has been made to enhance the brightness of the expressed protein. ZsGreen1 has an excitation maximum at 493 nm and emission maximum at 505 nm, and it is much brighter than EGFP (Fig. 15.2B). The time required to detect the fluorescence in mammalian cells is about 8–12 h post transfection. In comparison with EGFP, ZsGreen1 is brighter (quantum yield 0.91, extinction coefficiency 43,000). In fact, ZsGreen1 is the brightest fluorescent protein of the family of RCFPs. However, its solubility is poor, and the protein tends to aggregate in mammalian cells.

341

579

557

563 576 588

DsRed-Express

DsRed2 AsRed2 HcRed1

a

489 505 539

Reef Coral Fluorescent Proteins AmCyan1 458 ZsGreen1 493 ZsYellow1 529

24 8–12 16

8–12

8–12 8–12 8–12

8–12 8–12 8–12

Time to detection (hr)*

Tetramer Tetramer Tetramer

Tetramer

Tetramer Tetramer Dimer

+++

+++ ++ +

Monomer Monomer Monomer

Structure

+++ ++++ ++

+ +++ ++

Brightness relative to EGFP

As measured by FACS analysis using transiently transfected mammalian cells cultures.

582 592 618

476 510 529

Emission Max (nm)

Aequorea victoria GFP Variants ECFP 439 EGFP 484 EYFP 512

Protein

Excitation Max (nm)

TABLE 15.2. Comparison of BD Living ColorTM Fluorescent Proteinsa

+++ +++ +

+++

+++ ++++ +++

+ +++ +++

Utility as a reporter

++ + +++

++

+ + +

++++ ++++ ++++

Utility in fusions

Far-red fluorescence ideal for multiplexing, no detectable aggregation

Preferred DsRed for FACS due to diminished green emission, faster maturation Low aggregation

Photostable alternative to ECFP Bright green True yellow emission; ideal for multicolor applications

Green / yellow

Not as photostable as EGFP, EYFP

Comments

342 PRACTICAL CONSIDERATIONS FOR USE OF REEF CORAL FLUORESCENT PROTEINS IN MAMMALIAN CELLS

PROPERTIES OF RCFPs

Figure 15.2. Spectral comparison using fluorometer. (A) 1 mM of AmCyan1 and ECFP excited at 280 nm. (B) 1 mM of ZsGreen1 and EGFP excited at 488 nm. (C) 1 mM of ZsYellow1 and EYFP excited at 280 nm.

Therefore, ZsGreen1 is recommended for reporter assays. Additionally, ZsGreen was successfully targeted for degradation by the proteasome by fusion with PEST sequences from mouse ornithine decarboxylase. This construct is the basis of an assay to monitor proteasome activity in live cells (pZsProSensor, the plasmid encoding this proteasome sensor, is commercially available from BD Biosciences Clontech).

15.1.3

ZsYellow1

ZsYellow1 is a mutant of wild-type zFP540 from a species of Zoanthus (Matz et al., 1999). A single amino acid substitution (M129V) has been made to enhance the brightness of the expressed protein. ZsYellow1 has its excitation maximum at 529 nm and an emission maximum at 539 nm. ZsYellow1 is a true yellow protein. With an emission maximum at 539 nm, its spectrum is red-shifted in comparison to EYFP, an EGFP variant with an emission maximum at 529 nm (see Table 15.2). The distinctive spectrum of ZsYellow1 allows the separation of three colors—AmCyan1, ZsGreen1, and ZsYellow1—by single laser excitation (488 nm) of cells expressing each protein by flow cytometry. The time required to detect ZsYellow1 fluorescence in mammalian cells is about 8–12 h. ZsYellow1 is quite bright (quantum yield 0.65, extinction coefficiency 20,000), but the fluorescent intensity is a little lower in comparison with EYFP (Fig. 15.2C). ZsYellow1 is not very soluble and tends to aggregate in mammalian cells. ZsYellow1 is very suitable for transcription reporter assays.

15.1.4

Red Fluorescent Proteins

The family of RCFPs includes three red fluorescent proteins, with emission spectra ranging from 582 nm to 618 nm. They are ideal for multiplexing and FRET applications (Kohl et al., 2002; Erickson et al., 2003). The first red fluorescent protein, DsRed1, was identified in a species of Discosoma and emits light in the orange-red range with a maximum at 582 nm (Matz et al., 1999). The second red fluorescent protein, AsRed2, which emits light in the true red range with an emission maximum at 592 nm, was generated by mutating a nonfluorescent chromoprotein of Anemonia sulcata (Lukyanov et al., 2000). More recently, the gene coding for a far-red-shifted fluorescent protein, HcRed1, with unique

343

344

PRACTICAL CONSIDERATIONS FOR USE OF REEF CORAL FLUORESCENT PROTEINS IN MAMMALIAN CELLS

far-red fluorescence emission at 618 nm was generated by a combination of site-directed and random mutagenesis of the gene for a nonfluorescent protein isolated from Heteractis crispa (Gurskaya et al., 2001).

15.1.5

DsRed2/DsRed-Express

DsRed2 contains six amino acid substitutions when compared with DsRed1. The mutations at V105A, I161T, and S197A result in a more rapid appearance of red fluorescence compared to DsRed1. The cause for the faster chromophore formation may be that the mutations create a more flexible and less “crowded” space around the chromophore, allowing faster folding. The N-terminal mutations R2A, K5E, and K9T reduce the tendency of the protein to form aggregates in mammalian cells. The time required to detect of DsRed2 fluorescence in mammalian cells is about 24 h post transfection. DsRed2 has an excitation maximum at 563 nm and an emission maximum at 582 nm. Although it likely forms the same tetrameric structure as DsRed1 (Yarbrough et al., 2001), DsRed2 is less prone to forming large insoluble aggregates, which often develop in mammalian cells expressing DsRed1. DsRed2 is well-suited for both fusion proteins and transcription reporter assays. DsRed2 has been expressed in mammalian cells as an N-terminal fusion protein with Bid, a proapoptotic “BH3-only” member of the Bcl-2 family. The Bid-DsRed2 fusion protein was proteolytically processed and translocated from the cytosol to the mitochondria as is characteristic of its endogenous Bid counterpart. In addition, DsRed2 has been used in a variety of different mammalian cell applications (Iida et al., 2003; Lu et al., 2003; Mathieu and El-Battari, 2003; Zhe et al., 2003). DsRed-Express was developed by a combination of random and site-directed mutagenesis of DsRed1 (Bevis and Glick, 2002). It contains nine amino acid substitutions. The N-terminal mutations R2A, K5E, and N6D enhance the solubility of DsRed-Express. Interestingly, two of these same mutations were found to increase the solubility of DsRed2 using separate mutagenesis procedures in different laboratories. The increase in solubility could be a result of decreased positive charge at the N-terminus of DsRed-Express and DsRed2 in comparison to DsRed1. Four additional mutations—T21S, H41T, N42Q, and V44A—were found to be crucial for a faster formation of a functional chromophore. The red fluorescence of DsRed-Express can be detected in mammalian cells about 8–12 h after transfection compared with 48 h for DsRed1. Another limitation of the DsRed1 and DsRed2 is a substantial residual green emission due to a chromophore maturation intermediate. This residual emission interferes with double and triple labeling applications using green fluorescent protein, especially in fluorescence microscopy and flow cytometry. Two mutations in DsRed-Express, C117S and T217A, yield a profound reduction of this residual green emission peak. However, there is no direct evidence that the mutations listed above changed the tendency of DsRed-Express to form a tetramer. The spectral properties of DsRed-Express are similar to those of DsRed2, with an excitation maximum at 557 nm and an emission maximum at 579 nm. DsRed-Express is bright (quantum yield 0.90, extinction coefficient 19,000). However, if compared to DsRed2, DsRed-Express has a lower extinction coefficient (43,800 versus 30,100) and a reduced relative brightness (0.68 for DsRed2 versus 0.36 for DsRed-Express; Bevis and Glick, 2002). DsRed-Express can be used as a fusion tag, if the presumed tetrameric structure is not problematic. DsRedExpress has been expressed in mammalian cells as an N-terminal fusion protein with Bid, a proapoptotic “BH3-only” member of the Bcl-2 family. The Bid-DsRed-Express fusion protein was proteolytically processed and translocated from the cytosol to the mitochondria as is characteristic of its endogenous Bid counterpart.

PROPERTIES OF RCFPs

15.1.6

The “Fluorescent Timer” (E5)

A set of two mutations (V105A; S197T) of the original DsRed1 protein gave rise to a fluorescent protein with a very unique property. This protein, called fluorescent timer, has a fluorescent spectrum that changes over time from an initial bright green to red (Terskikh et al., 2000). Purified recombinant “fluorescent timer” protein achieves maximum green fluorescence about 4 h after purification. Subsequent to this initial 4-h period, green fluorescence declines; simultaneously, red fluorescence starts to appear. This “color switch” of the fluorescent timer can be used to monitor activation as well as downregulation of a specific promoter of interest using the color switch as the timer for this event. The ratio of green to red fluorescence can be used to determine the time in the past at which a promoter was switched on and when it was switched off (Terskikh et al., 2000).

15.1.7

DsRed Monomer

Many efforts have been made to weaken the tetrameric structure of DsRed in order to obtain a monomeric red fluorescent protein. However, this goal has been challenging due to the very tight structure of DsRed. Recently Tsien et al. were successful in generating a monomeric DsRed by mutating DsRed T1 (DsRed-Express: Bevis and Glick, 2002). The monomeric mutant, mRFP1, contains 33 mutations in comparison to the original DsRed1. Three mutations are located at the A/B interface, and 10 mutations are located at the A/C interface of the original tetramer. These mutations seem to be essential to separate the DsRed subunits of the tetramer into monomers. The emission maximum of monomeric DsRed in comparison to the tetrameric DsRed2 is shifted from 582 nm to 607 nm. However, the extinction coefficient, quantum yield, and photostability is lower than in DsRed2 (Campbell et al., 2002).

15.1.8

AsRed2

AsRed2 is a mutant of AsRed1 from Anemonia sulcata (Lukyanov et al., 2000) and has been engineered for stronger fluorescence intensity. It contains eight amino acid substitutions: F4L, K12R, F35L, T68A, F84L, A143S, K163E, and M202L. The time to detect red fluorescence in mammalian cells is about 8–12 h. AsRed2 has an excitation maximum at 576 nm and an emission maximum at 592 nm. Thus, it is slightly red-shifted compared to DsRed2 and DsRed-Express. AsRed2 can be used as a fusion tag, although it likely forms the tetrameric structure proposed for all RCFPs. AsRed2 is well-suited for transcription reporter assays.

15.1.9

HcRed1

HcRed1 was generated by random and site-directed mutagenesis of a gene coding for a nonfluorescent chromoprotein in the reef coral Heteractis crispa (Gurskaya et al., 2001). Early rounds of random mutagenesis were used to produce variants with extreme far-red fluorescence and rapid maturation kinetics. After isolation of the brightest variant, investigators used site-directed mutagenesis to optimize the solubility of the protein. The final variant, HcRed1 (HcRed-2A; Gurskaya et al., 2001), was selected not just because of its bright far-red fluorescence, but also because this mutant forms a dimer rather than a tetramer due to the mutation L126H located at the interface of the originally tetrameric protein. In comparison to wild type HcRed, HcRed1 contains seven additional mutations:

345

346

PRACTICAL CONSIDERATIONS FOR USE OF REEF CORAL FLUORESCENT PROTEINS IN MAMMALIAN CELLS

A2S, T36A, L122H, C143S, R168H, L173H, and P201L. HcRed1 does not aggregate in mammalian cells, and it elutes in gel filtration column chromatography with the predicted size of a dimeric protein. The good solubility of HcRed1 makes it more suitable for fusion proteins and mammalian expression compared to other RCFPs with a predicted tetrameric structure. Numerous fusions proteins have been successfully fused to HcRed1 (Hori et al., 2003; Kogata et al., 2003; Mulholland et al., 2003; Rintoul et al., 2003; Valentijn et al., 2003; Wagner et al., 2003). The time required to detect red fluorescence of HcRed1 is about 16 h post transfection in mammalian cells. HcRed1 has an excitation maximum at 588 nm and an emission maximum at 618 nm. In comparison to DsRed-Express and AsRed2, HcRed1 is not as bright (quantum yield 0.03, extinction coefficient 20,000). However, if HcRed1 is expressed in mammalian cells, it can be easily detected by fluorescence microscopy as well as flow cytometry. It may be that the in vitro conditions used to determine the quantum yield and extinction coefficient of recombinant HcRed1 does not reflect the fluorescence intensity of HcRed1 that can be achieved when expressed in mammalian cells. The fluorescence of HcRed1 is easily detected 16 h after transfection. Due to the far-red emission spectral properties, it is very easy to separate HcRed1 from other fluorescent proteins in multiplexing applications.

15.2

EXPRESSION AND DETECTION OF RCFPs

RCFPs have been expressed in variety of hosts. We have expressed RCFPs in mammalian cell lines such as HEK 293, HeLa, 3T3, Jurkat, and HT1080 cells. From unpublished observations, RCFPs have been successfully expressed in a variety of organisms such as fungi, plants, yeast, E. coli, C. elegans, Drosophila, Xenopus, zebrafish, and mouse. Here we focus on the practical considerations for the use of RCFPs in mammalian cells. RCFPs can be expressed in all mammalian systems as long as a functional promoter is used to initiate transcription. A promoter can be constitutively active, meaning that it initiates a persisting transcription of the RCFP gene in the plasmid downstream of the promoter. The immediate early cytomegalovirus promoter (CMV IE) has been used frequently to drive the constitutive expression of exogenous genes in mammalian cells. However, it is also possible to express the RCFPs under the control of an inducible promoter. This option allows the researcher to monitor the activation or inactivation of a promoter via the appearance or disappearance of the respective fluorescent signal. The signal-to-noise ratio in this type of application can be increased dramatically by using fusion proteins consisting of RCFPs and protein degradation motifs such as that found in mouse ornithine decarboxylase. Indeed, such “destabilized” RCFPs are constitutively degraded by the proteasome complex, and their accumulation and degradation within the cells are better correlated with promoter functions than in the case of nondestabilized reporters. Several different promoterless vectors, lacking a functional promoter sequence for mammalian cells and encoding these destabilized RCFPs, are available from BD Biosciences Clontech (pZsGreen1-DR, pDsRed-Express-DR, and pHcRed1-DR) and can be used to monitor the activity of promoter/enhancer combinations of interest. When expressing RCFPs and other fluorescent proteins as a tag fused to a protein of interest, the behavior of the resulting fusion protein cannot always be predicted. Changes in the proper function and/or localization of the protein/peptide of interest upon fusion to RCFPs may occur in some cases. To minimize this risk, it is often helpful to consider alternative orientations of the protein/peptide of interest with respect to the RCFP. It is possible to fuse the protein/peptide of interest either to the N-terminus or to the C-terminus of

EXPRESSION AND DETECTION OF RCFPs

the respective RCFP. This option is of particular importance if the protein/peptide of interest has a functional domain at the extreme N- or C-terminus. The two fusion proteins that contain either a free N-terminus or a free C-terminus of the protein of interest can then be tested separately. When considering the expression level of an RCFP in mammalian cells, the reef coral species of origin must be considered. The optimal DNA codon usage between reef corals and mammalian cells is quite different. Therefore it may be necessary to optimize the codon usage of the RCFP for a specific expression system. Currently, all RCFPs are available in human codon optimized forms that have been tested in a variety of mammalian expression systems. However, the human codon usage might be suboptimal for specific expression systems, especially nonmammalian expression systems. In this case a possible series of silent base pair changes should be considered in order to increase the expression level of the RCFP protein. However, these steps may only be necessary in exceptional cases. The expression level in eukaryotic cells can also be increased by incorporating a Kozak consensus sequence (CGCCACCATGG) including the ATG start codon at the 5¢end of the RCFP gene. Commercially available mammalian expression vectors (from BD Biosciences Clontech) contain the human codon optimized versions of the respective RCFP genes. A high expression level using those mammalian expression vectors is ensured by the use of a 5¢ Kozak sequence. Multiple cloning sites, located at either the 5¢- or 3¢ end of the respective RCFP gene, allow the expression of either N- or C-terminal fusion proteins. It has often been a concern of many researchers that fluorescent proteins used as a research tool might have a toxic effect on cells, therefore altering their normal function, growth, or differentiation state. In order to evaluate cytotoxicity due to overexpression in mammalian cells, HEK 293 cells were transfected with cytoplasmic targeted RCFPs, and cells expressing high levels of the fluorescent protein were sorted by flow cyotmetry on a BD FACSVantage SE cell sorter. Fluorescent cells were collected and were further studied for their ability to grow normally when returned to culture, as well as to maintain their level of fluorescence. In all studies, fluorescence was evaluated by flow cytometry. EGFP was used as a control and a basis for comparison between the different tests. Cells expressing EGFP, AmCyan1, ZsGreen1, ZsYellow1, DsRed2, and AsRed2 were analyzed on a BD FACSCalibur flow cytometry system using a 488-nm laser. HcRed1 required a 568.2-nm laser line for proper excitation; cells expressing this far-red shifted variant were analyzed on a BD FACSVantage SE cell sorter. For a period of 14–16 weeks, we monitored both the percentage of cells that remained fluorescent as well as the mean fluorescence of these positive cells. Our results show that for most of the fluorescent proteins, the percentage of positive cells was stable between 85% and 100%. Only DsRed2 and ZsYellow1 exhibited a decline to 80% and 60%, respectively (Fig. 15.3). These data suggest that for most of the RCFPs, almost all cells were able to maintain high levels of expression. The difference of intensities between the different cell populations is mostly due to the fact that all fluorescent proteins are not detected with the same efficiency on a specific flow cytometer (e.g., AmCyan1 is very bright but poorly detected on a BD FACSCalibur flow cytometry system).

15.2.1

Detection of RCFP by Fluorescence Microcopy

Protocol 1: Preparation of Cells Expressing Fluorescent Proteins for Fluorescence Microscopy. Prior to transfecting cells with the specific plasmid, 1 ¥ 105

347

348

PRACTICAL CONSIDERATIONS FOR USE OF REEF CORAL FLUORESCENT PROTEINS IN MAMMALIAN CELLS

Figure 15.3. Cells tolerate long-term expression of RCFPs. HEK 293 cells were stably transfected with N1 vectors encoding the indicated protein, and they were analyzed over time using BD FACSCaliburTM and BD FACSVantageTM. (A) The percentage of positive cells remaining in the population. (B) The non-normalized medium fluorescence intensity of the positive population. AmCyan1, EGFP, ZsGreen1, and ZsYellow1 were excited with a 488-nm line; DsRed2, DsRedExpress, AsRed2, and HcRed1 with a 568.2-nm laser line. AmCyan1, EGFP, ZsGreen1, and ZsYellow1 were detected with a 530/30 filter; DsRed2, DsRed-Express, and AsRed2 with a 595/25 filter; HcRed with a 630/20 filter.

cells are plated onto glass coverslips in a 6-well plate. Twenty-four hours after plating, cells are transfected using 0.75 mg of plasmid DNA using either lipid-based transfection agents like FuGene6 (Roche) or BD CLONfectin transfection reagent (BD Biosciences Clontech), or by using the calcium phosphate method (BD CalPhos Mammalian Transfection Kit, BD Biosciences Clontech), following the standard protocol. Twenty-four hours post transfection, the coverslips are rinsed twice with prewarmed (37°C) PBS with Ca2+/Mg2+. After these washes, the cells are fixed using 4% paraformaldehyde in PBS for 15 min at RT. After fixation, the coverslips are rinsed three times using PBS before being mounted onto glass slides (e.g., Molecular Probe mounting medium). The fixed cells are then stored at 4°C for 24 h before they are analyzed on a fluorescence microscope (e.g., Zeiss Axioskop) using different filter sets (e.g., Chroma RCFP filter sets).

15.2.2

Single-Color Analysis—Using 1 RCFP

In order to obtain good-quality imaging in microscopy using any fluorophore, it is important to use optimized filter sets. The sensitivity and specificity of the signal will be affected by this choice. Although it is possible to achieve good results using standard filter sets (such as FITC filter sets to detect ZsGreen1 and rhodamine or propidium iodide filter sets to detect DsRed), it is best to use filter sets that have been developed specifically for each fluorescent protein. Optimized filter sets for detecting all RCFPs including AmCyan1, ZsGreen1, ZsYellow1, DsRed2, AsRed2, and HcRed1 have been developed by Chroma Technology Corporation (Table 15.3). Detailed information on detecting RCFPs can be found on Chroma’s website (www.chroma.com). These filter sets have been developed to maximize the detection by adjusting the bandpass filters across the peak on the excitation and emission spectra of a given fluorescent protein. In addition, the width of the bandpass as well as the shift between the two bandpasses of the two filters has been carefully chosen

349

EXPRESSION AND DETECTION OF RCFPs

TABLE 15.3. Filter Sets for One-Color Analysis of Reef Coral Fluorescent Proteins Protein AmCyan1 ZsGreen1 ZsYellow1 DsRed2 AsRed2 HcRed1

Excitation Filter

Dichroic Mirror

Emission Filter

D440/40x HQ470/35 HQ550/40 D540/40x D540/40x HQ575/50x

470dcxr Q490lp 530dclp 570dclp 570dclp 610dclp

D500/40m HQ520/40 HQ550/40m D600/50m HQ620/60m HQ640/50m

Note: DsRed and AsRed filter sets are very close and interchangeable. All filter sets are available from Chroma Technology Corp. (www.chroma.com).

to minimize background and maximize fluorescent signal. Depending on the application, the RCFPs are not all equivalent; for fusion proteins, it is always best to try fluorescent proteins that are monomeric such as EGFP or dimers like HcRed1. Indeed, oligomerization of the fluorescent protein can result in the loss of the function of the protein studied, or it can lead to improper localization inside the cell. By contrast, for application where the subcellular localization of the fluorescent protein is not as important as a variation of fluorescence intensity, it is recommended to use the brightest and most photostable fluorescent proteins such as AmCyan1, ZsGreen1, or DsRed2. As an example, when under the control of a promoter specific of a signal transduction pathway, ZsGreen1 works well as an expression fluorescent reporter.

15.2.3

Multicolor Analysis—Using 2 or More RCFPs

Researchers are increasingly interested in multicolor analyses. With the introduction of the red fluorescent proteins, DsRed2, AsRed2, and HcRed, combined with the development of optimized filter sets by Chroma, it is now possible to separate as many as three fluorescent reporters (cyan, yellow, and red) by fluorescence microscopy. Here we offer several recommendations for two-and three-color analyses (Table 15.4). As a general rule, it is very difficult to use two fluorescent proteins simultaneously if their spectral characteristics are too similar. This will result in bleeding of one fluorescent protein signal in the filter set to the other, rendering signal separation problematic. For two-color analyses, AmCyan1 can be used in combination with ZsYellow1 or any of the red fluorescent proteins. AmCyan1 is less prone to photobleaching than ECFP and is potentially a better partner to yellow and red fluorescent proteins when long exposure times are required due to multicolor imaging. ZsYellow1 can be separated from AmCyan1 as well as the far-red-shifted HcRed1, so it is possible to use it in combination with either of these two for dual detection. ZsGreen1 can be used with all the red fluorescent proteins but not with AmCyan1 and ZsYellow1, whose spectral characteristics are too similar. The combination of ZsGreen1 and DsRed2 or DsRed-Express is especially convenient since it can be visualized with standard FITC and rhodamine filter sets of any conventional microscope. As shown in Fig. 15.4, ZsGreen1 targeted to the mitochondria (ZsGreen-Mito) and a DsRed2-bid fusion were coexpressed in HeLa cells, Bid is a proapoptotic “BH3only” member of the Bcl-2 family. In nonapoptotic cells, uncleaved Bid is localized in the cytosol. However, upon induction of apoptosis with an appropriate stimulus, Bid can be cleaved by the enzyme Caspase 8. Truncated tBid translocates to the mitochondria and may induce the release of Cytochrome C from the mitochondria into the cytosol. The

350

PRACTICAL CONSIDERATIONS FOR USE OF REEF CORAL FLUORESCENT PROTEINS IN MAMMALIAN CELLS

TABLE 15.4. Recommended Combination for Multicolor Microscopy First Color

Second Color

Third Color

AmCyan1

ZsYellow1 DsRed2 DsRed-Express HcRed1 DsRed2 DsRed-Express HcRed1 AmCyan1 HcRed1 AmCyan1 ZsGreen1 AmCyan1 ZsGreen1 ZsYellow1

HcRed1

ZsGreen1

ZsYellow1 DsRed2 DsRed-Express HcRed1

A

ZsYellow1

HcRed1 AmCyan1

ZsYellow1 AmCyan1

B

Control

Staurosporin

Figure 15.4. Dual-color analysis for monitoring Bid activation with DsRed2. HeLa cells were transiently cotransfected with plasmids encoding the fusion protein Bid-DsRed2 and a mitochondriatargeted ZsGreen1 (ZsGreen1-Mito). (A) Before induction of apoptosis, Bid-DsRed2 is localized in the cytosol and ZsGreen1-Mito labels the mitochondria. (B) After induction of apoptosis with 1 mM staurosporine for 3 h, the relocalization of Bid-DsRed2 to mitochondria as revealed by the colocalization with the mitochondria marker ZsGreen1-Mito. Images were taken with a 100¥ objective using Chroma filter sets hq460/40x, 490dclp, and hq515/30m for ZsGreen1 and using hq545/50x, 580dcxr, and hq630/60m for DsRed2. See color insert.

translocation of cleaved tBid from the cytosol to the mitochondria upon apoptotic stimulus is a very important hallmark for the role of mitochondria in the apoptotic destruction of cells. The expression of a chimeric fusion of Bid and a fluorescent protein, like DsRed2, allows one to monitor the translocation event by observing the redistribution of the fluorescence signals. The colocalization of green fluorescent mitochondria (ZsGreen-Mito) and red fluorescent tBid can therefore only be observed upon induction of apoptosis.

EXPRESSION AND DETECTION OF RCFPs

Figure 15.5. Detection of three fluorescent proteins by fluorescent microscopy. HeLa cells were separately transfected with plasmids pAmCyan1-N1, pZsYellow1-N1, and pHcRed1-N1, mixed, and observed by microscopy using Chroma Technology Corp. filter sets d440/40x, 470dcxr, and d500/40m for AmCyan1, using hq500/40, 530dclp, and hq550/40m for ZsYellow1, and using hq575/50x, 610dclp, and hq640/50m for HcRed1. See color insert.

Combining ZsGreen1 and DsRed2 for dual color labeling and using specific filter sets for their detection usually provides high-quality images, but in some cases the fluorescent signal of one or both fluorescent proteins is very strong. This might result in some bleedthrough of the green fluorescence in the red filter set or vice versa. To overcome this problem, we recommend visualizing ZsGreen1 with the AmCyan1 filter set and DsRed2 with the HcRed1 filter set. This configuration conveys adequate fluorescent signal from ZsGreen1 and DsRed2 to obtain good-quality images and efficiently reduces the bleeding of nonspecific fluorescence. For three-color analyses, there is only one recommended combination for microscopy: AmCyan1, ZsYellow1, and HcRed1. As shown in Fig. 15.5, HeLa cells were separately transfected with the plasmids pAmCyan1-N1, pZsYellow1-N1 and pHcRed1-N1. After mixing the transfected cells, the fluorescence from the three cell subpopulations were correctly visualized using Chroma Technology Corp. filter sets (Table 15.3). In multiplex experiments, the choice of filter sets is critical to optimize separation of the fluorescence signals. Aside from the fact that these filter sets were designed for a high signal-to-background ratio, they were also chosen to minimize the bleeding of fluorescent signal of one fluorescent protein into the other filter set. In the case of triple labeling for example, the images obtained for ZsYellow1 must not contain much signal from AmCyan1 or HcRed1. Otherwise it is impossible to distinguish any of the three fluorescent signals.

15.2.4

Detection of RCFP by Flow Cytometry

Protocol 2: Preparation of Cells Expressing Fluorescent Proteins for Flow Cytometry 1. BD FACSTM Analysis of Unfixed Cells. Routinely, 3 ¥ 105 cells are plated into a six-well plate. Twenty-four hours after plating, cells are transfected with the plasmid DNA of interest using 3 mg of DNA. Cells are transfected using either the calcium phosphate method or a lipid-based transfection reagent like FuGene6 (Roche) or BD CLONfectin transfection reagent (BD Biosciences Clontech). Twenty-four hours post transfection, cells are collected. Adherent cells are removed from the tissue culture plate by rinsing the plate with PBS w/o Ca2+/Mg2+

351

352

PRACTICAL CONSIDERATIONS FOR USE OF REEF CORAL FLUORESCENT PROTEINS IN MAMMALIAN CELLS

before adding trypsin, followed by a 1- to 5-min incubation at 37°C. The cells are than rinsed from the plate using DMEM medium. The collected cells are pelleted and washed once using ice-cold PBS. After washing, the cells are resuspended in 1 ml of ice-cold PBS or sheat fluid. This cell suspension is then analyzed using either a BD FACSCaliburTM flow cytometry system or a BD FACSVantageTM SE cell sorter. 2. BD FACSTM Analysis of Fixed Cells. Routinely, 3 ¥ 105 cells are plated into a 6well plate. twenty-four hours after plating, cells are transfected with the plasmid DNA of interest using 3 mg of DNA. Cells are transfected using either the calcium phosphate method or a lipid-based transfection reagent like FuGene6 (Roche) or BD CLONfectin transfection reagent (BD Biosciences Clontech). Twenty-four hours post transfection, cells are collected. Adherent cells are removed from the tissue culture plate by rinsing the plate with PBS w/o Ca2+/Mg2+ before adding trypsin, followed by a 1- to 5-min incubation at 37°C. The cells are then rinsed from the plate using DMEM medium. The collected cells are pelleted and washed once using ice-cold PBS. The cells are than resuspended in 1 ml of 4% paraformaldehyde in PBS. This cell suspension is then incubated at RT for 30 min either by mixing the cell suspension slowly every 5 min or by using a slow shaker. After fixation, cells are pelleted by centrifugation and washed 3 times with 1 ml of ice-cold PBS. The cell pellet of the last washing step is then resuspended in 1 ml of PBS or sheat fluid and analyzed using a BD FACSCalibur flow cytometry system or a BD FACSVantage SE cell sorter. Protocol 3: Generation of Stable Cell Populations and Clones Expressing RCFPs. In order to establish cell populations that express RCFPs, 3 ¥ 105 HEK 293 cells are plated in 6-well plates and transfected with 2 mg of the desired plasmid (e.g., pAmCyan1-N1). After 48 h, transfected cells are selected using media containing 0.5 mg/ml G418 and grown for another 2 weeks. At this stage, all the cells are able to grow in selective media, and a majority of the cells express the fluorescent protein. However, the expression level varies widely among the stable cells, resulting in a wide range of fluorescence intensity in the mixed population. If cells with brighter fluorescence signals are necessary for an experiment, the mixed population can be enriched using flow cytometry. For example, we used this procedure to isolate cells for studying potential long-term cytotoxicity effects of RCFP expression. The mixed population was run on a BD FACSVantage SE cell sorter, and the brightest 30% of cells were separated to achieve a more homogeneous sampling. Then, each week the new population was harvested and analyzed for the percentage of remaining fluorescent cells and the average fluorescence intensity. The BD FACSVantage SE cell sorter was also used to establish clones of HEK 293 cells expressing RCFPs. As described above, the brightest 30% of the cells were isolated and sorted into 96-well plates with 1 cell per well. After 2–3 weeks, individual clones were screened by fluorescence microscopy and the 12 brightest clones were expanded into 24well plates, 6-well plates, and 100-mm dishes. Each clone was analyzed for homogeneity and fluorescence intensity. Among all the RCFPS, HcRed1 is the only fluorescent protein that has offered challenges in establishing stable cell populations and clonal lines. We noticed that the expression of HcRed1 tends to reduce the growth rate of the cells. To overcome this problem,

353

EXPRESSION AND DETECTION OF RCFPs

we transfected cells with a lower quantity of plasmid and selected the stable population with a lower concentration of antibiotic. Interestingly, though the cells grew slowly, the level of HcRed1 expression and the percentage of positive cells remained stationary.

15.2.5

Single-Color Analysis by Flow Cytometry—Using 1 RCFP

When only one RCFP or other fluorescent protein is expressed in cells to be analyzed by flow cytometry, the two most important considerations are (a) the use of an appropriate laser line for excitation of the protein and (b) the channel selected to collect the fluorescent signal. We describe here our experience using each RCFP expressed in HEK 293 cells. AmCyan1. With an excitation maxima of 458 nm and emission maxima of 489 nm, AmCyan1 can be detected on flow cytometers equipped with laser lines such as 407 nm (BD FACSAria cell sorter) or 458 nm (Table 15.5). We have also successfully detected AmCyan1 using a BD FACSCalibur instrument. We used a clone that exhibited high levels of fluorescence that was detectable using the 488-nm laser line and the FL1 channel (530/30 bandpass filter).

TABLE 15.5. Detection of RCFPs on BD’s Flow Cytometers Excitation Max

Excitation Required

AmCyan1

458

ZsGree ZsYellow1

493 529

DsRed2

563

DsRedExpress

557

AsRed2

576

HcRed1

588

405 407 413 458 458 488 488 488 514 531 488 514 531 568 488 514 531 568 488 514 531 568 568 633 635

Protein

Suitable Laser VioFlame Point source Krypton Argon Spectrum Argon Argon Argon Argon Spectrum Argon Argon Spectrum Spectrum Argon Argon Spectrum Spectrum Argon Argon Spectrum Spectrum Spectrum HeNe Red diode

Emission Max 489

505 539

582

Suitable Emission Filters 485/22 or 530/30

530/30 519/20 or 520/40 530/30 or 550/30 550/30 or 585/42 585/42 555/20, 580/30, 585/42, or 610/20

579

610/20 or 630/30 555/20, 580/30, 585/42, or 610/20

592

610/20 or 630/30 580/30, 585/42, 610/20, or 630/30

618

610/20 or 630/30 610/20 or 630/30 660/20

354

PRACTICAL CONSIDERATIONS FOR USE OF REEF CORAL FLUORESCENT PROTEINS IN MAMMALIAN CELLS

ZsGreen1. Like EGFP, ZsGreen1 is ideal for flow cytometry because it is efficiently activated with a 488-nm laser found on most machines and is easily detected on the FL1 channel of a BD FACSCalibur instrument. ZsGreen1 is the brightest fluorescent protein of all the RCFPs, and we needed to establish and select clones expressing medium levels of ZsGreen1 (Fig. 15.6B); otherwise the entire cell population would be out of the FL1 channel scale. Its brightness makes it the RCFP of choice when the sensitivity of detection might be an issue. For example, we have successfully monitored subtle variations in ZsGreen1 fluorescence when using the protein as a gene expression reporter. ZsYellow1. ZsYellow1 can be activated using a 488-nm laser as well as the 531-nm line of a spectrum laser (Table 15.5). With an emission maximum of 539 nm, it can be detected using either the FL1 (530/30) or FL2 (585/42) channels of a BD FACSCalibur instrument. DsRed2, DsRed-Express and AsRed2. Typically, DsRed2, DsRed-Express, and AsRed2 are detected using a 488-nm laser and the FL2 (485/42) channel of a BD FACSCalibur instrument. All the clones we established were easily detectable, though these settings are not optimal. DsRed-Express matures faster than DsRed2 (Fig. 15.6C) and is therefore a better fit for reporter studies using flow cytometry. In cases where the level of expression of DsRed is very low and difficult to detect on a BD FACSCalibur

Figure 15.6. Two-color analysis of ZsGreen1 and DsRed-Express by flow cytometry. HEK 293 cells were transiently transfected with either pZsGreen1 or pDsRed-Express, mixed, and then analyzed by flow cytometry on a BD FACSCaliburTM. (A) Mock. (B) ZsGreen1. (C) DsRed-Express. (D) Mixed with ZsGreen1 and DsRed-Express. (E) Cotransfected with ZsGreen1 and DsRed-Express. (F) Mixed and cotransfected with ZsGreen1 and DsRed-Express.

EXPRESSION AND DETECTION OF RCFPs

instrument, we use the 568-nm line of an argon/krypton laser on the BD FACSVantage cell sorter in order to excite DsRed with higher efficiency. HcRed1. HcRed1 is not detectable on a BD FACSCalibur instrument because it is not excited efficiently with a 488-nm laser. We use the 568-nm laser of a BD FACSVantage SE cell sorter to excite HcRed1 and the FL5 channel equipped with a 640 long-pass filter for detection. Cells expressing HcRed1 are easily detected with these settings, even when the level of expression and the overall fluorescence of HcRed1 are low.

15.2.6

Two Color Analysis by Flow Cytometry—Using 2 RCFPs

Flow cytometry offers the possibility to detect several fluorescent proteins simultaneously and is an ideal complement to microscopy where quantitative information is required without the need to visualize the spatial distribution of fluoresence. ZsGreen1 and DsRedExpress form the most appropriate pair of fluorescent proteins because both can be excited with a 488-nm laser and are easily detected on a BD FACSCalibur instrument using the FL1 and FL2 channels (Fig. 15.6D). Although some instrument compensation might be required, it is generally easy to separate both signals so that cells cotransfected with ZsGreen1 and DsRed-Express can be visualized and potentially sorted using a BD FACSVantage SE cell sorter (Figs. 15.6E and 15.6F). DsRed-Express, which exhibits lower green emission than DsRed2, is the best partner for ZsGreen1 because it requires minimum compensation. Using machines with other laser lines and detection channels, it is possible to separate almost every fluorescent protein in performing two-color analysis. AmCyan1 can be combined with ZsYellow1 or any red fluorescent protein. Furthermore, using the 568-nm line of a spectrum laser on a BD FACSVantage SE cell sorter, combined with a 595/25 filter on FL4 and a 640 long-pass filter on FL5 channels, we have successfully separated DsRed-Express and HcRed1 signals (Fig. 15.7A) in such a manner that cells cotransfected with both were also visualized (Fig. 15.7B).

Figure 15.7. Two- and three-color analysis by flow cytometry. (A) HEK 293 cells were transiently transfected separately with either pDsRed2-N1 or pHcRed1-N1, mixed, and then analyzed by FACSVantage. (B) Cells were also cotransfected with both vectors, mixed with the independently transfected cells, and analyzed. Cells were excited using a BD FACSVantageTM SE cell sorter using the 568.2-nm line of a coherent krypton/argon laser. DsRed2 was detected with a Chroma Technology Corp. 595/25 bandpass filter; HcRed1, with a 640 long-pass filter. (C) HEK 293 cells were stably transfected with either pZsGreen-N1, pZsYellow-N1, or pAsRed2-N1, mixed, and then analyzed on a BD FACSCaliburTM using the 488-nm laser line.

355

PRACTICAL CONSIDERATIONS FOR USE OF REEF CORAL FLUORESCENT PROTEINS IN MAMMALIAN CELLS

15.2.7

Three- and Four-Color Analysis by Flow Cytometry

Compared to microscopy, one advantage of flow cytometry is that it measures the fluorescence intensity of cells. It is therefore possible to visualize the fluorescence of several fluorescent proteins using two-dimensional scatter plots, even if the spectral properties of the proteins are similar. For example, it is difficult to distinguish AmCyan1, ZsYellow1, and DsRed2 by microscopy because of the degree of spectral overlap. By contrast, we were able to separate signals from these proteins using a BD FACSCalibur instrument (Fig. 15.7C). AmCyan1 is mostly detected in the FL1 channel, and cells expressing it appear close to the FL1 axis in a FL1 versus FL2 scatter plot. Similarly, cells expressing DsRed2 appear close to the FL2 axis. Cells expressing ZsYellow1, which is equally detected by the FL1 and FL2 channels, are visualized in the diagonal of the same plot. It is therefore possible to establish three distinct regions and measure the individual fluorescence intensity of a mixed population of cells each expressing a different RCFP. Using a BD FACSVantage SE cell sorter, it has been possible to analyze cells expressing ECFP, EGFP, EYFP, and DsRed1 (Hawley et al., 2001). We achieved similar results with a BD FACSAria cell sorter using the three available laser lines (407, 488, and 633 nm) to activate AmCyan1, ZsYellow1, DsRed-Express, and HcRed1. Figures 15.8A and 15.8B show two scatter plots of the same mixed population of cells expressing the four fluorescent proteins. While DsRed-Express and ZsYellow1 can be separated as previously described on the BD FACSCalibur instrument, it was also possible to plot the HcRed1 channel versus the AmCyan1 channel to also distinguish these two cell populations. Overall, the growing number of flow cytometry instruments available on the market combined with the entire panel of reef coral fluorescent proteins provide an excellent array of tools for multiplexing applications. This technology has now reached a point where the automation of large-scale experiments is practical.

15.2.8

Other Methods of Detection

Since the fluorescent proteins have been used in research, microscopy and flow cytometry have been the two predominant detection methods. It is also possible to detect fluoB

A 5

10

AmCyan1

10

DsRed-Express

356

104

103

5

104

103

102

102

102

103

104

ZsYellow1

105

102

103

104

105

HcRed1

Figure 15.8. Four-color separation of RCFP-expressing HEK 293 cells using flow cytometry. A mixed population of cells stably expressing either DsRed2, ZsYellow1, HcRed1, or AmCyan1 was separated by flow cytometry with a BD FACSAriaTM cell sorter using three separate laser lines: 407 nm to excite AmCyan1; 488 nm to excite DsRed2 and ZsYellow1; and 633 nm to excite HcRed1.

REFERENCES

rescent proteins using spectrophotometry, and a number of investigators have developed applications where fluorescence intensity is measured in cells in a 96-well plate reader. Recently, a novel type of instrument has been applied to the detection of RCFPs expressed in cells. Acumen Bioscience has developed a laser-scanning device, the Acumen ExplorerTM, which is able to measure the fluorescence of cells attached to 96-well plates. With a resolution of less than one micron, the Explorer is able to visualize the fluorescence profile of a cell expressing a fluorescent protein. With the convenience of a plate reader, sensitivity close to that of a flow cytometer, and potential subcellular fluorescence localization capabilities, the Acumen Explorer is becoming a very popular tool for the development of cell-based assays in drug discovery. In summary, convenient coexcitation of green, yellow, and red fluorescent proteins with the common 488-nm laser line allows for easy detection of these proteins on nearly all flow cytometers. With the addition of more laser lines and appropriate detection filters, at least four cell populations, each harboring a different fluorescent protein, can be distinguished by the proteins’ unique emission spectra. These characteristics provide great flexibility for multiplexing in various applications. Fluorescent proteins are a valuable resource when combined with flow cytometry for both preparative and analytical applications. Applications range from the very simple enrichment of cell populations after transient transfection and cell clone isolation to the more complex isolation of cells from transgenic animals or analyses in gene therapy research.

ACKNOWLEDGMENTS We are grateful to Eric Machleder of BD Biosciences Clontech for help in preparing this chapter.

REFERENCES Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (2000). Biochemistry, mutagenesis, and oligomerization of DsRed, a red fluorescent protein from coral. Proc. Natl. Acad. Sci. USA 97:11984–11989. Bevis, B. J., and Glick, B. S. (2002). Rapidly maturing variants of the Discosoma red fluorescent protein (DsRed). Nat. Biotechnol. 20:83–87. Campbell, R. E., Tour O., Palmer A. E., Steinbach P. A., Baird G. S., Zacharias D. A., and Tsien R. Y. (2002). A monomeric red fluorescent protein. Proc. Natl. Acad. Sci. USA 99:7877–7882. Erickson, M. G., Moon, D. L., and Yue, D. T. (2003). DsRed as a potential FRET partner with CFP and GFP. Biophys J. 85:599–611. Gross, L. A., Baird, G. S., Hoffman, R. C., Baldridge, K. K., and Tsien, R. Y. (2000). The structure of the chromophore within DsRed, a red fluorescent protein from coral. Natl. Acad. Sci. USA 97:11990–11995. Gurskaya, N. G., Fradkov, A. F., Terskikh, A., Matz, M. V., Labas, Y. A., Martynov, V. I., Yanushevich, Y. G., Lukyanov, K. A., and Lukyanov, S. A. (2001). GFP-like chromoproteins as a source of far-red fluorescent proteins. FEBS Lett. 507:16–20. Hawley, T. S., Telford, W. G., Ramezani, A., and Hawley, R. G. (2001). Four-color flow cytometric detection of retrovirally expressed red, yellow, green, and cyan fluorescent proteins. Biotechniques 30:1028–1034.

357

358

PRACTICAL CONSIDERATIONS FOR USE OF REEF CORAL FLUORESCENT PROTEINS IN MAMMALIAN CELLS

Hori, T., Haraguchi, T., Hiraoka, Y., Kimura, H., and Fukagawa, T. (2003). Dynamic behavior of Nuf2-Hec1 complex that localizes to the centrosome and centromere and is essential for mitotic progression in vertebrate cells. J. Cell Sci. 116:3347–3362. Iida, R., Yasuda, T., Tsubota, E., Takatsuka, H., Masuyama, M., Matsuki, T., and Kishi, K. (2003). M-LP, Mpv17-like protein, has a peroxisomal membrane targeting signal comprising a transmembrane domain and a positively charged loop and up-regulates expression of the manganese superoxide dismutase gene. J. Biol. Chem. 278:6301–6306. Kogata, N., Masuda M., Kamioka Y., Yamagishi A., Endo A., Okada M., and Mochizuki N. (2003). Identification of Fer tyrosine kinase localized on microtubules as a platelet endothelial cell adhesion molecule-1 phosphorylating kinase in vascular endothelial cells. Mol. Biol. Cell 14:3553–3564. Kohl, T., Heinze, K. G., Kuhlemann, R., Koltermann, A., and Schwille, P. (2002). A protease assay for two-photon crosscorrelation and FRET analysis based solely on fluorescent proteins. Proc. Natl. Acad. Sci. USA 19:12161–12166. Lu, J. Y., Chen, H. C., Chu, R. Y., Lin, T. C., Hsu, P. I., Huang, M. S., Tseng, C. J., and Hsiao, M. (2003). Establishment of red fluorescent protein-tagged HeLa tumor metastasis models: determination of DsRed2 insertion effects and comparison of metastatic patterns after subcutaneous, intraperitoneal, or intravenous injection. Clin. Exp. Metastasis 20:121–133. Lukyanov K. A., Fradkov, A. F., Gurskaya, N. G., Matz, M. V., Labas Y. A., Savitsky A. P., Markelov, M. L., Zaraisky A. G., Zhao X., Fang, Y., Tan, W., and Lukyanov, S. A. (2000). Natural animal coloration can be determined by a nonfluorescent green fluorescent protein homolog. J. Biol. Chem. 275:26879–25882. Mathieu, S., and El-Battari, A. (2003). Monitoring E-selectin-mediated adhesion using green and red fluorescent proteins. J. Immunol. Methods 272:81–92 Matz, M. V., Fradkov, A. F., Labas, Y. A., Savitsky, A. P., Zaraisky, A. G., Markelov, M. L., and Lukyanov, S. A. (1999). Fluorescent proteins from nonbioluminescent Anthozoa species. Nat. Biotechnol. 17:969–973. Mulholland, D. J., Read, J. T., Rennie, P. S, Cox, M. E., and Nelson, C. C. (2003). Functional localization and competition between the androgen receptor and T-cell factor for nuclear beta-catenin: A means for inhibition of the Tcf signaling axis. Oncogene 22:5602–5613. Ormö, M., Cubitt, A. B., Kallio, K., Gross, L. A., Tsien, R. Y., and Remington, S. J. (1996). Crystal structure of the Aequorea victoria green fluorescent protein. Science 273:1392– 1395. Rintoul, G. L., Filiano, A. J., Brocard, J. B., Kress, G. J., and Reynolds Y. J. (2003). Glutamate decreases mitochondrial size and movement in primary forebrain neurons. J. Neurosci. 23:7881–7888. Terskikh, A., Fradkov, A., Ermakova, G., Zaraisky, A., Tan, P., Kajava, A. V., Zhao, X., Lukyanov, S., Matz, M., Kim, S., Weissman, I., and Siebert, P. (2000). “Fluorescent timer”: Protein that changes color with time. Science 290:1585–1588. Valentijn, A. J., Metcalfe, A. D., Kott, J., Streuli, C. H., and Gilmore, A. P. (2003). Spatial and temporal changes in Bax subcellular localization during anoikis. J. Cell Biol. 162:599– 612. Wagner, L. E., Li, W., and Yule, D. I. (2003). Phosphorylation of type-1 inositol 1,4,5-trisphosphate receptors by cyclic nucleotide-dependent protein kinases: A mutational analysis of the functionally important sites in the S2+ and S2- splice variants. J. Biol. Chem., in press. Wall, M. A., Socolich, M., and Ranganathan, R. (2000). The structural basis for red fluorescence in the tetrameric GFP homolog DsRed. Nat. Struct. Biol. 7:1133–1138. Yanushevich, Y. G., Staroverov, D. B., Savitsky, A. P., Fradkov, A. F., Gurskaya, N. G., Bulina, M. E., Lukyanov, K. A., and Lukyanov, S. A. (2002). A strategy for the generation of nonaggregating mutants of Anthozoa fluorescent proteins. FEBS Lett. 511:11–14.

REFERENCES

Yarbrough, D., Wachter, R. M., Kallio, K., Matz, M. V., and Remington, S. J. (2001). Refined crystal structure of DsRed, a red fluorescent protein from coral, at 2.0-Å resolution. Proc. Natl. Acad. Sci. USA 98:462–467. Zhe, X., Yang, Y., Jakkaraju, S., and Schuger, L. (2003). Tissue inhibitor of metalloproteinase-3 downregulation in lymphangioleiomyomatosis: Potential consequence of abnormal serum response factor expression. Am. J. Respir. Cell. Mol. Biol. 28:504–511.

359

16 PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP Nicola Bevan Stephen Rees Screening and Compound Profiling, GlaxoSmithKline, Stevenage, Herts, United Kingdom

16.1

INTRODUCTION

The color variants of the Aequorea victorea green fluorescent protein (GFP) and the reef coral fluorescent proteins (RCFP) are in widespread use within the pharmaceutical industry. In this chapter we describe a number of applications of the use of these proteins in the early phase of drug discovery. Many novel drug screening assays have been developed through the use of fluorescent proteins. These assays are being applied both (a) for the initial discovery of molecules with activity at a target protein in a process termed highthroughput screening (HTS) and (b) in the more detailed characterization of compound efficacy in a number of assays collectively termed high-content, or high-information, screening assays (referred to as HCS assays in this chapter). Fluorescent protein technology has also been applied within the target validation phase of drug discovery. Target validation describes a collection of techniques used to generate information on the likely involvement of a new target in disease. The objective of such studies is to generate a hypothesis that small-molecule intervention at that target may have efficacy within that disease state. Fluorescent protein technology has been applied for both in vitro and in vivo target validation. We will present some examples of the use of GFPs and RCFPs in model organisms such as C. elegans, Drosophila, and transgenic mice.

Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

361

362

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

16.2

HIGH-CONTENT AND HIGH-THROUGHPUT SCREENING

Perhaps the widest application of fluorescent protein technology within the pharmaceutical industry has been for the identification and characterisation of hit and lead molecules in processes termed HTS (Walters and Namchuk, 2003) and HCS (Kapur, 2002). HTS describes the process whereby chemical libraries, of many hundred thousand or indeed million molecules, are screened against drug targets to identify molecules with modulatory activity at the target protein that may form the basis of a lead optimisation drug discovery program. HTS assays are typically run in 384- or 1536-well microtiter plates in assay volumes of between 5 and 50 ml at compound throughputs of between 10,000 and 100,000 test molecules/day, using complex laboratory automation and detection systems. In contrast to HTS, which demonstrates whether or not a compound has activity at a target protein, HCS provides additional information on how the compound affects cell physiology. HCS assays enable multiple measurements of cellular phenotype as a consequence of compound activity. Assay end-points include the intracellular location and translocation of a fluorescent protein fusion protein, a change in cell shape, or neurite outgrowth, through the use of measurements taken on a cell-by-cell basis (Liptrot, 2001; Kapur, 2002). Several fluorescent protein translocation assays have been developed (Table 16.1). Fluorescent protein translocation assays require the generation of a fusion protein between the protein under study and a fluorescent protein. Plasmid vectors containing fluorescent proteins are available from suppliers such as Clontech (Palo Alto, CA) (www.clontech.com). For membrane proteins, GFP is typically fused to the carboxylterminus of the protein such that GFP is not required to pass through the membrane; for soluble proteins there are examples of the generation of GFP fusions at either the amino or carboxyl termini of the protein. In either case, care has to be taken to ensure that the generation of the fusion protein does not affect the function of the protein under study. It is a remarkable observation that the fusion of GFP or RCFP to other proteins has little detectable effect on the function of the protein partner. The detection of a change in the cellular localization of a fluorescent protein fusion protein in an HCS assay is determined using a plate-based fluorescence imaging system. Such instruments enable the generation of quantitative data from both individual cells and cell populations. Specific informatic tools have been developed in order to generate numerical data from fluorescence images captured by these analysis systems. These algorithms typically rely upon the identification of a cell by staining the nucleus with a nuclear specific fluorescent dye. The analysis algorithm captures the distribution of the fluorescent protein within the cell in relation to the nucleus to generate quantitative data on the intensity and location of the fluorescent protein. Such data are typically generated from pre- and posttreatment of cells with the test compound in order to determine the effect of that compound on the physiology under study. HCS assays may provide new insights into compound mode of action. It is hoped that the increasing application of HCS screening assays will enable a more thorough characterization of the activity of lead molecules prior to more expensive and complex animal studies, which ultimately will lead to the development of more efficacious and selective drugs (Kapur, 2002). In the last five years a number of fluorescence imaging systems have become available including the Cellomics ArrayscanTM, the GE Healthcare InCell Analyser 3000TM, or the Evotec OperaTM. These systems allow the detection of the subcellular localization of a GFP fusion protein within cells plated into 96-, 384-, or 1536-well microtiter plate formats. In the fastest of these systems, data can be collected from a 1536-well plate in

363

FLUORESCENT PROTEIN DRUG SCREENING ASSAYS

TABLE 16.1. Examples of Fluorescent Protein Translocation Assays Assay

Application

Translocation

Reference

Nuclear translocation

Agonist/antagonist identification at nuclear receptors

Cytoplasm to nucleus

GPCR/GFP internalization

Agonist/antagonist identification at GPCR

Membrane to cytoplasm

b-Arrestin recruitment

Agonist/antagonist identification at GPCR

Cytoplasm to membrane

GLUT4/GFP

Reporter of insulin activation Agonist/antagonist identification at GPCR

Cytoplasm to membrane Cytoplasm to membrane

Giuliano et al. (1997) Koster and Hauser (1999) Zhu et al. (1998) Htun et al. (1996) Conway et al. (1999) Conway et al. (2001) Tarasova et al. (1997) Drmota et al. (1998) Kallal et al. (1998) Barak et al. (1997b) Awaji et al. (1998) Slice et al. (1998) Barlic et al. (1999) Lamb et al. (2001) Schlador and Nathanson (1997) Zhang et al. (1999) Luttrell et al. (1999) Ferguson et al. (1998) Vrecl et al. (1998) Groarke et al. (1999) Evans et al. (2001) Barak et al. (1997a) Matharu et al. (2001) Mundell et al. (2001) Richardson et al. (2003) Patki et al. (2001)

PKC/GFP or PLC/GFP recruitment

Richardson et al. (2003) Feng et al. (1998) Almholt et al. (1999) Wang et al. (2001)

approximately 45 minutes. Further information about the characteristics of these detection instruments, the informatic tools developed to complement these detection systems, and the type of assays enabled by these systems can be obtained from the corresponding manufacturers (Table 16.2).

16.3

FLUORESCENT PROTEIN DRUG SCREENING ASSAYS

A number of drug screening assays based on the use of fluorescent protein technology have been described, and the use of such assays for HTS and HCS is increasing. Reporter gene assays, fluorescence resonance energy transfer (FRET) assays, bioluminescence resonance energy transfer (BRET) assays, and fluorescent protein degradation assays have been applied to HTS. In contrast, a range of fluorescent protein cellular redistribution assays have been developed for HCS.

364

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

TABLE 16.2. Suppliers of Instrumentation and Assay Reagents for High Content Screening Company

Website

Summary

Accumen GE Healthcare BioImage Cellomics Evotec Imaging Research Molecular Devices Molecular Devices PerkinElmer

www.acumenbioscience.com www.amershambiosciences.com www.bioimage.com www.cellomics.com www.evotecoai.com www.imagingresearch.com www.moleculardevices.com www.moleculardevices.com www.perkinelmer.com

Instrumentation Reagents, instrumentation Reagents Instrumentation Instrumentation Instrumentation Instrumentation TransfluorTM Reagents, instrumentation

Fluorescent protein screening assays require the generation of a mammalian cell line transfected to express the target being screened together with the fluorescent reporter molecule. In many assays this involves the generation of fusion proteins between the protein being studied and the fluorescent protein. To generate a compound screening assay, mammalian cells in culture are either stably or transiently transfected with the fusion protein using either a plasmid or viral expression vector. For HTS assays or HCS assays in which many thousands of compounds are tested across many weeks or months, it is common practice to generate stable cell lines. HCS assays in which small numbers of compounds are tested are often generated using transient expression systems. A variety of methods have been applied to transfect immortalized mammalian cell lines in culture with a fluorescent protein expression plasmid. However, the efficiency of plasmid transfection in many immortalized cell lines, and most primary cell lines, is poor. Viral expression vectors, including those derived from adenovirus or retrovirus, are more complex to generate, and their use often requires stringent Biosafety conditions. However, such vectors are capable of transducing a wide variety of immortalised and primary cell lines. The use of viral vectors in HCS assays allows the study of compound activity in primary cell types which are often available in small numbers but which can generate a more physiologically relevant understanding of compound activity. HTS assays are typically performed on live cells. In contrast, HCS screening assays may be performed on fixed or live cells. The use of live cells allows the detection of translocation in real time and hence enables the study of the kinetics of protein translocation; however, this requires that the plate be imaged for the entire duration of the translocation event. The use of fixed cells allows many assay plates to be prepared and fixed at a predetermined point following compound stimulation. Images can then be collected for analysis. The use of fixed cells increases the daily capacity of the detection instrument as imaging times are generally shorter. In the following sections we review examples of such assays and their application to particular classes of drug targets to provide the reader with some insight into the application of fluorescent proteins for compound screening within the pharmaceutical industry.

16.3.1

Nuclear Translocation Assays

A number of HCS screening assays have been established that rely upon the detection of the translocation from the cytosol to the nucleus of a fusion protein between a transcription factor and a fluorescent protein. Following stable or transient expression in mam-

FLUORESCENT PROTEIN DRUG SCREENING ASSAYS

malian cells, translocation of the fusion protein is detected using a plate-based imaging system. A number of assays have been developed, including assays of translocation of the glucocorticoid receptor (Htun et al., 1996), NFkB (Schmid et al., 2000), ERK MAP kinase (Horgan and Stork, 2003), NFAT (Scott et al., 2001), and STAT (Koster and Hauser, 1999). Such assays have been configured to screen for direct modulators of transcription factor function or as reporters of compound activity at cell surface receptors known to regulate the signal transduction cascade under study. Within our laboratory we have developed a number of transcription factor and nuclear receptor translocation assays. In collaboration with GE Healthcare Biosciences (Cardiff, UK), we developed an assay for direct modulators of the glucocorticoid receptor (GR), through the generation of a fusion protein between human GR and GFP. This protein was expressed in HEK 293 cells, and a 96-well microtiter plate assay for GR modulators was established using the GE Healthcare InCell Anaylser 3000TM for signal detection. In unstimulated cells this fusion protein is found in the cytoplasm (Fig. 16.1). Following treatment with a steroid agonist such as dexamethasone, the fusion protein is seen to translocate to the nucleus within 30 minutes of drug treatment (Fig. 16.1). Using this assay, we have profiled a series of GR agonists and antagonists and demonstrated that these compounds are capable of regulating GR activity with the expected concentration dependence. This assay has the potential to be used in HTS to identify compounds capable of causing GR translocation. In combination with other nuclear receptor translocation assays, this assay has since been applied to profile the activity of GR agonists, for their activity against a panel of nuclear receptors and other transcription factors.

Figure 16.1. Confocal visualization of nuclear translocation of a glucocorticod receptor/GFP fusion protein. The glucocorticod receptor/GFP fusion protein was stably expressed in HEK 293 cells. Assays were performed on cells plated in a 96-well microtiter plate and images taken using the GE Healthcare InCell Analyser 3000TM. Basal fluorescence (A) and fluorescence distribution following 30 min incubation with dexamethasone (100 nM) (B) are shown. (Data provided by GE Healthcare, Cardiff, UK.)

365

366

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

16.3.2

Cell Adhesion Assays

In a novel application of fluorescent protein technology, Mathieu and El-Battari (2003) have developed an E-selectin adhesion assay using both GFP and DsRed2 RCFP. Adhesion assays are in widespread use within the pharmaceutical industry for the identification of inhibitors of integrin and selectin mediated cell adhesion (Loster and Horstkorte, 2000). Such assays have relied upon the use of fluorescent dyes, such as BCECF, to label cells expressing the integrin or selectin, which are then examined for their ability to adhere to a matrix. The inhibition of cell adhesion is detected by a decrease in the adhesion of fluorescently labeled cells to the matrix. To avoid the use of fluorescent dyes, Mathieu and El-Bathtari describe the generation of a stable CHO cell line expressing both E-selectin and GFP or DsRed2 RCFP. The ability of such cells to adhere to a variety of E-selectin ligands was examined. This assay provides an alternative to the use of fluorescent dyes which avoids the requirement of loading cells with dye prior to assay and the dye leaching and photobleaching issues associated with fluorescent dye technology.

16.3.3

Reporter Gene Assays

The application of reporter gene assays for compound screening has been extensively reviewed elsewhere (Hill et al., 2001; Rees et al., 1999). A reporter gene is a sequence of DNA whose easily measured product is synthesized in response to the activation of a specific signal transduction cascade. The DNA sequence consists of a promoter element (where transcription factors bind to control transcription), a reporter gene, and a transcription stop signal. The choice of promoter element dictates the sensitivity and specificity of the reporter. The reporter gene should offer a unique property to the cell system being studied, or at least be easily distinguishable from other cell products, have a short half-life to minimize basal accumulation of the reporter product, and be easily measurable in simple, cheap assays. Commonly used reporter genes include the enzymes firefly luciferase, secreted placental alkaline phosphatase and b-lactamase (Rees et al., 1999). These assays have been applied throughout academia and the biotechnology and pharmaceutical industries for compound screening in 384 and 1536 well format for many target classes including G-protein coupled receptors (GPCRs), cytokine receptors and growth factor receptors (Terstappen et al., 2000; Subbaramaiah et al., 2001). GFP offers an attractive alternative to the enzymatic reporter genes as GFP assays do not require cell lysis and reagent addition prior to signal detection. In contrast, these assays rely upon the detection of the accumulation of GFP in living cells. This offers a simple assay protocol, does not require the purchase of often expensive detection reagents, and allows the detection of reporter gene activity in real time in live cells. The use of GFP as an inducible reporter gene has been limited due to its poor brightness and its extremely long half-life (>36 h), which results in substantial basal accumulation of GFP and hence a low signal window in the assay. In recent years a number of GFP mutants have been described that address these issues. To facilitate expression in mammalian cells, GFP vectors have been developed in which the open reading frame has been optimized for preferred human codon usage. In parallel with codon optimization, specific point mutations have been introduced within the GFP chromphore to generate proteins that show a 35-fold increase in fluorescence compared to the native protein. These proteins have been designated enhanced-GFP (E-GFP) and are available from Clontech (Palo Alto, CA). Furthermore, to address the issue of GFP accumulation from an un-induced reporter gene, Li et al. (1998) have developed a series of destabilised GFPs with a half-life of 2 h. This protein,

FLUORESCENT PROTEIN DRUG SCREENING ASSAYS

designated d2GFP, contains the PEST domain from mouse ornithine decarboxylase, which targets GFP for rapid degradation, fused to the C-terminus of GFP. Several groups have now reported the use of E-GFP or d2GFP in reporter gene assays. As an example of the application of E-GFP in a reporter gene assay, Nagy et al. (2002) described the development of a screen for Ah receptor agonists. The Ah receptor is a ligand-dependent transcription factor that mediates the biological and toxic effects of polycyclic and halogenated aromatic hydrocarbons such as dioxin. A HepG2 cell line containing an integrated Ah receptor responsive E-GFP reporter gene was used to screen chemical libraries to identify novel activators of this receptor.

16.3.4 Fluorescence Resonance Energy Transfer (FRET) and Bioluminescence Resonance Energy Transfer (BRET) Assays FRET and BRET describe the process of nonradiative energy transfer between a fluorescent donor protein (FRET) or a bioluminescent donor enzyme (BRET) and a fluorescent acceptor protein (Fig. 16.2; Boute et al., 2002). In a FRET assay the fluorescent donor protein is usually the enhanced blue or cyan fluorescent protein (EBFP/ECFP) and the fluorescent acceptor protein, one of the derivatives of the Green/Yellow fluorescent protein class (EGFP or EYFP). The bioluminescent donor enzyme in a BRET assay is the coelenterate luciferase cloned from Renilla reniformis. This luciferase catalyzes the oxidative degradation of coelenterazine to generate blue light with a lmax of 460 nM; hence it is able to act as an energy donor to one of the derivatives of the green/yellow fluorescent protein class (EGFP or EYFP) (Fig. 16.2). FRET or BRET occurs when the donor and acceptor

Figure 16.2. Principles of bioluminescence resonance energy transfer (BRET) (A) and fluorescence resonance energy transfer (FRET) (B) assays. For BRET fusion, proteins are generated between the proteins under study (A, B) and Renilla luciferase (Rluc) and a fluorescent protein (YFP). In the presence of coelenterazine, light is generated by Rluc, which, if the luciferase is in close proximity to YFP, will excite the fluorescent protein to generate light at the emission maxima of this protein. In FRET, fusion proteins are generated between the proteins under study (A, B) and donor [cyan fluorescent protein (CFP)] and acceptor [yellow fluorescent protein (YFP)] fluorescent proteins. Excitation of CFP results in emission at 475 nM which, when in close proximity to the YFP, will excite this protein to generate light at the emission maxima of YFP (525 nM).

367

368

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

proteins are brought into close proximity. This can be achieved through the construction of a biosensor containing both donor and acceptor proteins within the same polypeptide, or following the fusion of the donor and acceptor proteins to two proteins known to interact. FRET and BRET have been used to establish a number of assays as described in the following sections. In each case, compound activity is determined as either an increase or a decrease in FRET/BRET, according to whether the compound promotes an association or dissociation of the FRET/BRET partners. 16.3.4.1 FRET Protease Assays. Fluorescent protein technology has been applied to develop FRET screening assays for protease inhibitors. Such assays rely upon the generation of a fusion protein between a cyan/blue fluorescent protein (ECFP) and a green/yellow fluorescent protein (EYFP) separated by a specific protease cleavage site. In this fusion protein, excitation of ECFP at 433 nM results in the generation of fluorescence emission from EYFP at an emission maxima of 527 nM due to FRET between the two closely associated fluorescent proteins. Cleavage of the fusion protein results in a loss of FRET because the CFP and YFP molecules are no longer in close proximity. This results in an increase in ECFP fluorescence emission at 475 nM and a decrease in EYFP emission at 527 nM as FRET is lost (Fig. 16.2). Thus FRET can be quantified according to a ratio of light emission at these two wavelengths. Instruments designed for the detection of FRET are capable of simultaneous excitation and signal detection at two emission wavelengths from the sample under study. Protease biosensors have been isolated and purified to generate in vitro assays for protease inhibitors as alternatives to the generation of peptide substrates labeled with small fluorescent probes. Protease biosensors have also been expressed in mammalian cell lines to generate whole-cell-based assays for protease inhibitors (Berdichevsky et al., 2003; Jones et al., 2000; Kohl et al., 2002). Caspase activity resulting in the proteolytic cleavage of target proteins is an integral step in the pathway leading to apoptotic cell death (Jones et al., 2000). Many pharmaceutical companies are attempting to identify caspase inhibitors for a variety of neurological disorders. To address this, FRET-based HTS and HCS assays for caspase inhibitors have been developed and applied in intact mammalian cells (Tawa et al., 2001; Jones et al., 2000; Mahajan et al., 1999). These assays rely upon the generation of FRET substrates containing ECFP and EGFP/EYFP linked by a specific caspase cleavage site. Such assays have been developed in 96- and 384-well assay plates. Caspase inhibitors are identified because they prevent degradation of the FRET substrate and hence prevent a loss of FRET. Caspase biosensors have been used in recombinant cell lines and also introduced into primary neurones to permit the study of caspase inhibitors in a more relevant disease cell type. 16.3.4.2 FRET Assays of Protein–Protein Interaction. There are many reports of the application of FRET to demonstrate the interaction of two proteins [for reviews see Sekar and Periasamy (2003) and Eidne et al. (2002)]. In our laboratory we have applied FRET technology to demonstrate the interaction of the two subunits of the GABA-B receptor. The GABA-B receptor is a heterodimer of two GPCRs, namely, the GABA-BR1 receptor and the GABA-BR2 receptor. The generation of a functional receptor at the cell surface requires the expression of both these receptors (White et al., 2002). To confirm that the receptor exists as a heterodimer at the cell surface we generated fusion proteins between the GABA-BR2 receptor and the cyan RCFP and the GABA-BR1 receptor and the yellow RCFP (Fig. 16.3). When expressed alone, the GABA-BR1 receptor does not reach the cell surface. When coexpressed with the GABA-BR2 receptor, both receptors

FLUORESCENT PROTEIN DRUG SCREENING ASSAYS

Figure 16.3. Confocal visualization of GABA-B receptor heterodimerization. HEK 293T cells were transfected with the fusion proteins, GABA-BR2/cyanRCFP and GABA-BR1/yellowRCFP. From left to right in the figure, the images show cellular expression of GABA-BR2/cyanRCFP (excitation at 433 nM, emission at 475 nM), GABA-BR1/yellowRCFP (excitation at 488 nM, emission at 525 nM; the overlay of the first two images demonstrate that both proteins are expressed at the same site), and the FRET signal (excitation at 433 nM; emission at 525 nM). The FRET event demonstrates that the proteins are in close proximity. See color insert.

can be visualized at the cell surface. That they exist in close proximity can be inferred from the observation of FRET between the two fusion proteins (Fig. 16.3). While many groups have applied FRET to identify and confirm protein–protein interactions, the application of FRET assays for compound screening has been limited. This is due to the relative weakness of the FRET signal in comparison to that obtained with BRET. However, such assays are being used within the pharmaceutical industry to identify and characterize novel protein–protein interactions during the target validation phase of drug discovery. 16.3.4.3 BRET Assays of Protein–Protein Interaction. G-protein-coupled receptors (GPCRs) are a family of approximately 750 membrane-spanning receptors that are known to have a fundamental role in physiology and pathophysiology. Members of this receptor family are responsible for the detection of visual and olfactory stimuli, and they play a pivotal role in cell signaling processes such as neurotransmission, chemotaxis, and inflammation. GPCRs are potential targets for therapeutic intervention in many diseases. Of the approximately 500 marketed drugs, around 30% are modulators of GPCR function having efficacy in diseases such as pain, psychiatric disorders, hypertension, asthma, cardiovascular disease, and peptic ulcers [Wise et al. (2002) and references therein]. The development of GPCR screening assays to enable the identification of novel GPCR drugs is a major challenge for the pharmaceutical and biotechnology industries. Agonist binding to a GPCR results in the activation of intracellular signal transduction cascades. To attenuate receptor signaling, serine and threonine residues within the Cterminal tail of many receptors become multiply phosphorylated by members of the GPCR kinases (GRK) family of protein kinases (Zhang et al., 1997). The phosphorylated Cterminal tail of the receptor is then able to bind members of the arrestin family. Arrestin binding promotes receptor internalization into the endosome compartment to attenuate receptor signaling. Within the acidic environment of the endosome, the ligand dissociates from the GPCR, the receptor C-terminal tail is dephosphorylated, and the receptor is recycled to the cell membrane (summarized in Fig. 16.6).

369

370

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

BRET assays for compound screening at GPCRs have been described in which compound activity is determined according to the degree of interaction between a Renilla luciferase tagged GPCR and a YFP tagged b-arrestin (Bertrand et al., 2002). Assays have been established to identify both agonist and antagonist ligands at the receptor. Agonist ligands cause an increase in BRET following association between the receptor and the barrestin, wheras antagonists can be identified through the inhibition of agonist mediated BRET. BRET assays have also been established to study GPCR homo- and heterodimerization (Mercier et al., 2002; Jensen et al., 2002) and to monitor insulin receptor activation (Issad et al., 2002). A BRET assay for the interaction of the insulin receptor with protein tyrosine phosphatase 1B has been established by generating fusion proteins between the two interacting partners (Boute et al., 2003). The application of BRET to establish screening assays for regulators of such protein/protein interactions has been reviewed by Boute et al. (2002). Three combinations of donor and acceptor protein have been used in BRET assays (Table 16.3). In each case the donor protein is Renilla luciferase; however, different coelenterazine substrates and different acceptor proteins have been used in the different versions of BRET. In BRET1 the A. victoria E-YFP was used as the acceptor protein in combination with native coelenterazine as the Renilla luciferase substrate. BRETZS-Yellow again uses native coelenterazine as the Renilla luciferase substrate and the RCFP Zoanthus sp. YFP as the acceptor protein. Experiments in our laboratory have demonstrated that the BRET ratio change in equivalent experiments is greater with BRETZS-Yellow than with BRET1 due to the higher quantum yield of the Zoanthus sp. yellow RCFP and the red-shifted emission wavelength of this fluorescent protein. However, in both assays the emission spectra of native coelenterazine overlaps the emission spectra of the fluorescent protein, resulting in a low signal-to-noise ratio in the assay. To increase the Stokes shift between the emission maxima of native coelenterazine and the available fluorescent proteins, Perkin Elmer Biosciences (Monteal, Canada) have developed a novel Renilla luciferase substrate, DeepBlueC coelenterazine, and a novel YFP mutant with an excitation maxima at 400 nM (Bertrand et al., 2002). This version of BRET, designated BRET2, allows a wider separation of the emission maxima of the luciferase and fluorescent protein, thus leading to a large signal-to-noise ratio in the assay. However, this coelenterazine is substantially less bright, leading to a weak signal in the assay. Detection instruments, such as the Perkin Elmer Biosciences Fusion, are now available that enable the detection of BRET in 96- and 384-well plate formats. In collaboration with Perkin Elmer Biosciences (Montreal, Canada), we have developed a BRET assay to detect modulators of the interaction of the human glucocorticoid

TABLE 16.3. Fluorescence Characteristics of Donor and Acceptor Proteins Used in Bioluminescence Resonance Energy Transfer (BRET) Substrate

Substrate lmax Emission

BRET1

WT coelenterazine

475 nM

BRET2

DeepBlueC coelenterazine

400 nM

BRETZS-Yellow

WT coelenterazine

475 nM

Acceptor E-YFP (BD Biosciences) GFP-Topaz (Perkin Elmer) ZsYellow-NFP (BD Biosciences)

Acceptor lmax Emission 527 nM 508 nM 540 nM

FLUORESCENT PROTEIN DRUG SCREENING ASSAYS

receptor (GR) and the transcription factor NFkB. Because it is difficult to predict how the generation of a fusion protein will affect the biology of the interaction under study, Renilla luciferase and GFP-topaz were fused in-frame to both the amino and carboxy termini of GR and NFkB. A panel of transient transfection assays were performed to determine the vector combination that generated the greatest increase in BRET following agonist stimulation using the Perkin Elmer Fusion for BRET detection. These studies demonstrated that the GR agonist dexamethasone promoted an increase in BRET following transient cotransfection of the following fusion protein combinations: Renilla luciferase fused to the C-terminus of GR coexpressed with a fusion protein in which GFP was fused to either the N- or C-terminus of NFkB. It is important to note that in these studies other combinations of fusion proteins failed to generate a signal in the BRET assay. The reason for the lack of signal is undetermined but could be due to insufficient levels of expression of some fusion partners, or the generation of fusion partners in which the distance between the GFP and luciferase is too great for BRET. This example illustrates the requirement to generate a range of fusion proteins during assay development in order to identify the combination that generates the greatest signal in a BRET assay. Within the pharmaceutical industry, compounds are often stored in solvents such as dimethylsulfoxide (DMSO) to prevent compound degradation. Many cell-based screening assays are intolerant to DMSO concentrations of greater than 1% due to cytotoxicity. In this assay we have determined that the BRET ratio change in response to treatment with 1 mM dexamethasone was unchanged at DMSO concentrations of up to 2.5%, demonstrating the assay to be of appropriate robustness for drug screening. Similar robustness has been seen in other BRET assays and is a consequence of the ratio method used to calculate BRET. Increasing DMSO concentrations result in a decrease in emission from both Renilla luciferase and YFP; however, the ratio of emission is unchanged. In this assay a range of GR agonists promoted GR/NFkB interaction including dexamethasone (EC50 = 9 nM), fluticasone propionate (EC50 = 3.3 nM), Ru486 (EC50 = 2.6 nM), and Ru24858 (EC50 = 13.7 nM) (Fig. 16.4). As expected, progesterone antagonized the interaction between GR and NFkB (data not shown). This assay has been applied to screen for the effect of GR agonists on the interaction of this receptor with NFkB. This example demonstrates the application of BRET, and indeed FRET, to study the effect of compound activity directly on protein interactions.

16.3.5

Dual Colour Spectroscopy Assays

As an alternative to FRET, Kohl et al. (2002) have generated fusion proteins between EGFP and DsRed RCFP to develop a dual color correlation spectroscopy assay for protease inhibitors. This is a single-molecule-based detection technology that selectively probes the movement in solution of two fluorescent groups. The measurement principle is based on a spectrally resolved detection of single fluorescent molecules diffusing in and out of a diffraction-limited laser focus. Movement of the uncleaved FRET substrate through the laser allows the detection of both red and green fluorescence. Following protease cleavage GFP and DsRed RCFP become free to move in solution to cause only single molecules to be detected as they move through the laser focus. In contrast to FRET assays, the dual color correlation assay is not limited to certain ranges of distance between the two flurophores in the substrate because detection of cleavage does not rely upon FRET; instead, it relies upon the generation of distinct GFP and RCFP molecules. This assay is potentially more versatile and sensitive than FRET-based protease assays.

371

372

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

Figure 16.4. BRET2 assay of ligand-mediated regulation of the interaction of the glucocorticoid receptor (GR) and NFkB. HEK 293 cells were transfected with the GR/Rluc and p65NFkB/GFP fusion proteins. Measurement of changes in BRET in response to various ligands was determined using the Packard BRETcount (Perkin Elmer Biosciences). The BRET signal was measured after stimulation with varying concentrations of the following agonists: RU24858 (), RU486 (D), fluticasone propionate (), and dexamethasone (). (Data were generated in collaboration with Perkin Elmer Biosciences.)

16.3.6

Fluorescent Protein Degradation Assays

In addition to monitoring cellular localization, GFP fusion proteins can be used to track the lifetime of an expressed protein. An example of where this has been applied in drug discovery is the development of a screening assay for modulators of the interaction between IkBa and NFkB (Li et al., 1999). NFkB activity is upregulated by many proinflammatory cytokines including TNFa and IL-1 and is thought to play a fundamental role in many pro-inflammatory diseases. For this reason, many pharmaceutical companies are running drug discovery programs to identify antagonists of this signal transduction cascade. Prior to activation, NFkB exists as a complex with IkBa in the cytoplasm. Upon activation of the NFkB signal cascade, IkBa is phosphorylated and degraded to release NFkB, which is then able to exert its biological effects by modulating the expression of several genes. Degradation of IkBa is rapid and can be used as a reporter of NFkB activation. To establish a screening assay for modulators of the NFkB pathway, we have generated a fusion protein between IkBa and GFP. Following expression in mammalian cells, the degradation of this protein following the activation of the NFkB cascade can be measured on a confocal microscope (Fig. 16.5) or on an HCS machine such as the ArrayScanTM (Cellomics).

16.3.7 b-Arrestin Recruitment Assay Agonist-mediated translocation of arrestin has been applied to develop a highly novel and specific assay for GPCR ligands. This assay relies upon the detection of ligand-mediated recruitment of a fusion protein between b-arrestin1 or b-arrestin2 and GFP from the cytosol to the plasma membrane (Claing et al, 2002; Fig. 16.6). In this assay, stable mam-

FLUORESCENT PROTEIN DRUG SCREENING ASSAYS

Figure 16.5. Fluorescent protein assay of NFkB activity. A fusion protein between IkBa and GFP was transiently expressed in HeLa cells. Confocal images were taken various times after stimulation with TNFa and the fluorescence signal quantified. The images show the time course of the degradation induced by TNF alpha (A), and the graph shows the fluorescence intensity measured from these images over time (B).

malian cell lines are generated expressing both the GPCR and a fusion protein between b-arrestin and EGFP or b-arrestin and the dsRedRCFP. In the unstimulated cell the barrestin/fluorescent protein is distributed in the cytoplasm. Following agonist stimulation of the GPCR, the b-arrestin/fluorescent protein fusion is recruited to the cell membrane as part of the GPCR desensitization and internalization process. For many receptors, this is followed by internalization of the GPCR/arrestin/fluorescent protein complex into the endosome compartment. When visualized using a fluorescence or confocal plate reader, arrestin translocation is visualized as a change in the distribution of GFP fluorescence from the cytosol to the membrane (Fig. 16.7), or to generate a “spotted” appearance following internalization of the complex into the endosome compartment (Fig. 16.7). Because many, if not the majority, of GPCRs are internalized through the interaction of the receptor and b-arrestin, this assay appears to offer a generic HCS, and perhaps HTS, screening assay for GPCR ligands. The assay developed by Xsira Pharmaceuticals (Durham, North Carolina) and marketed as TransfluorTM has been applied to in excess of 20 receptors spanning the GPCR target class. This assay is now available from Molecular Devices, Sunnyvale, CA. This includes the b2-adrenoceptor (Barak et al., 1997b), the neurokinin NK-1 receptor (Richardson et al., 2003), the adenosine A2b receptor (Matharu et al., 2001), the metabotropic glutamate mGluR1 receptor (Mundell et al., 2001), the bradykinin B1 and B2 receptors (Lamb et al., 2001), and the CXCR1 chemokine receptor (Barlic et al., 1999). The application of this technology for GPCR screening is reviewed extensively by Conway and Demarest (2002).

373

374

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

Figure 16.6. Schematic representation of b-arrestin-mediated GPCR internalization in response to agonist stimulation (see text for details). (Figure provided by Xsira Pharmaceuticals, North Carolina.)

We have also used the arrestin/GFP recruitment assay for GPCR HTS using the GE Healthcare InCell Analyser 3000TM to detect agonist-mediated redistribution of the arrestin/GFP fusion protein. In these studies a stable U2OS cell line was generated coexpressing the human b2-adrenoceptor together with the b-arrestin2/GFP fusion protein. A high-throughput screening assay was developed and applied to identify both agonists and antagonists of this receptor. This assay was demonstrated to be capable of identifying known b2-adrenoceptor ligands with the expected concentration dependence (Fig. 16.8). As mentioned previously, most pharmaceutical companies store their compound libraries in solvents such as DMSO. We have demonstrated that the performance of this assay is unaffected by DMSO concentrations of up to 2%. To demonstrate the application of this assay for HTS, a random set of 1280 compounds were screened in duplicate in 96-well plate format. Compounds were screened for both agonist and antagonist activity at this receptor. The screen was performed on two separate days to allow the reproducibility of the assay and the effect of compound interference to be assessed. Assay performance was determined according to Z¢, a statistical measure used to measure the quality of drug screening assays. A good HTS assay is defined as having a Z¢ factor of greater than 0.4 (Zhang et al., 1999). In the agonist and antagonist screens the Z¢ was consistently above 0.7 and 0.6, respectively. In the agonist screen, active molecules or “hits” were classified as all molecules with an activity of greater than 40% of the positive control agonist response. Using this measure, the agonist screen did not identify any active compounds.

FLUORESCENT PROTEIN DRUG SCREENING ASSAYS

Figure 16.7. Cytosol to membrane translocation of b-arrestin/fluorescent protein fusion proteins following ligand stimulation. (A) Cells expressing the human angiotensin AT1A receptor (top) and substance P receptor (bottom), together with a fusion protein between b-arrestin2 and GFP, were treated with the agonist ligands angiotensin and substance P, respectively. Images were taken at time 0 and after 30 min of agonist treatment. Data generated by Xsira Pharmaceuticals. (B) CHO cells stably expressing the human CCR2 chemokine receptor were transiently transfected with a fusion protein between b-arrestin2 and the Anemonia sulcata AsRed1 RCFP. Confocal images were taken at various times following stimulation with the CCR2 receptor ligand MCP-1 (10 mM). (Images were generated in collaboration with Professor G. Milligan, University of Glasgow.)

This was an expected observation because this compound set did not contain any b2adrenoceptor agonists. In the antagonist screen, active molecules were classified as all molecules that generated an inhibitory activity of greater than 40% of the control antagonist response. Using this measure, 1.2% of the compounds screened were determined to possess antagonist activity at the b2-adrenoceptor. As expected, the correlation of activity between compounds screened on different days illustrates that the majority of active compounds were identified on both days of assay. These assays are in use in many pharmaceutical companies for GPCR HTS in which chemical libraries of greater than half a million compounds have been screened to identify molecules of interest. Genome sequencing studies have identified approximately 160 so-called orphan GPCRs—that is, receptors with no known function and no known ligand (Wise et al., 2002). There is considerable activity both within the pharmaceutical industry and elsewhere aimed at the identification of the ligands and the physiological role of these new

375

376

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

Figure 16.8. Application of the TransfluorTM b-arrestin recruitment assay for drug screening. U2OS cells stably expressing the human b2-adrenoceptor and a fusion protein between b-arrestin2 and GFP were provided for this study by Xsira Pharmaceuticals (North Carolina). Assays were performed in the 96-well microtiter plate format, and images were collected on using the GE Healthcare InCell Analyser 3000TM (GE Healthcare, Cardiff, UK). (A) Concentration response curve to the standard agonist isoprenaline. (B) Agonist screening assay. A set of 1280 random compounds were screened for agonist activity at 5 mM final concentration on two separate days. Plot shows the activity on day 1 plotted against the activity on day 2 (axis = % of maximal stimulation with isoprenaline). (C) Antagonist screening assay. A set of 1280 random compounds were screened for antagonist activity at 5 mM final concentration on two separate days. Cells were preincubated for 30 min with compound before stimulation with an EC80 concentration of isoprenaline (20 nM). Plot shows the activity on day 1 plotted against the activity on day 2 (axis = % of maximal inhibition of isoprenaline response). (Data were generated in collaboration with Xsira Pharmaceuticals, North Carolina.)

receptors (Wise et al., 2002). In the absence of an activating ligand, little is known about the signal transduction capabilities of an orphan receptor. Because b-arrestin recruitment following agonist binding at a GPCR is a phenomenon common to most GPCRs, the arrestin recruitment assay offers utility for the identification of ligands at orphan GPCRs. As a proof of concept experiment to investigate the use of this assay for these studies, HEK293 cells stably expressing the b-arrestin/GFP fusion protein were transiently transfected with the Neuromedin NMUR-1 receptor for which the ligand has recently been identified. Eighty random compounds were then screened for agonist activity in this assay. Five of the random compounds were spiked with a maximal concentration of NMU, the ligand for this receptor. The ability of the assay to detect the spiked samples was determined. Only those spiked wells showed any activity in this assay (Fig. 16.9).

16.3.8

GPCR/GFP Fusion Assays

A second HCS assay for the identification of GPCR ligands relies upon the generation of fusion proteins between the GPCR under study and the fluorescent protein to create a fluorescent receptor that is expressed on the cell surface. GFP has been fused to the intracellular C-terminal tail of more than 20 GPCRs with little or no effect on receptor pharmacology [reviewed in Kallal and Benovic (2000)]. Agonist binding to such a receptor results in receptor internalization into the endosome compartment as part of the signal transduction desensitization process. For example, McLean et al. (1999) generated fusion proteins between GFP and the native and a constituitively active mutant of the b2 adrenoceptor and used imaging techniques to study ligand activity at both forms of the receptor.

FLUORESCENT PROTEIN DRUG SCREENING ASSAYS

Figure 16.9. Application of the TransfluorTM b-arrestin recruitment assay for orphan GPCR screening. The neuromedin U receptor (NMR) was transiently transfected into HEK 293 cells stably expressing the b-arrestin/GFP fusion protein. In a 96-well microtiter plate screening assay, 80 random compounds were screened for agonist activity. The five compound wells that were spiked with Neuromedin U were active in this assay. Data were collected on using the GE Healthcare InCell Anaylser 3000TM (GE Healthcare, Cardiff, UK). (Data were generated in collaboration with Xsira Pharmaceuticals, North Carolina.)

In these studies the authors were able to determine both ligand efficacy and the ability of these ligands to regulate receptor degradation. This movement of fluorescent receptor is visualized as a membrane-to-endosome movement using HCS detection apparatus. This assay has been applied for HTS (Conway et al., 1999; Conway and Demarest, 2002).

16.3.9

Protein Complementation Assays

A number of protein complementation assays have been developed for both compound screening and target validation. Such assays have relied upon splitting enzymes such as b-galactosidase or b-lactamase into distinct domains with no enzymatic activity (Graham et al., 2001; Galarneau et al., 2002). Fusion proteins are generated between the domains of these enzymes and two proteins predicted to interact. The interaction of the protein partners facilitates the interaction of the enzyme domains to reconstitute enzyme activity. As such, the reconstitution of enzyme activity can be used as an indicator of protein–protein interaction. Such assays have been used to (a) screen random compounds to identify inhibitors of a known protein–protein interaction and (b) characterize the interaction between proteins presumed to interact. However, these assays do not permit the visualization of protein–protein interaction in living cells. To address this, Ozawa et al. (2001) have split GFP into two domains that independently do not exhibit fluorescence. When fused to interacting proteins, the two “halves” of GFP are brought into close proximity to reconstitute the GFP fluorophore. While the only successful demonstration of split GFP

377

378

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

technology has been in E. coli, the application of this technology to mammalian cells could have great utility for drug discovery.

16.3.9

Fluorescence-Activated Cell Sorting

Fluorescence-activated cell sorting (FACS) is a long-established technology used to sort cells from within mixed cell populations and to determine the level of expression of proteins on the cell surface or, following permeabilization, within the cell (Herezenberg and de Rosa, 2000). Cells of different lineages express characteristic marker proteins. Through the use of fluorescently labeled antisera specific to these markers, it is possible to sort specific cells from within a mixed population. Similarly, it is possible to sort cells into different populations based on the level of expression of the marker protein. Cells expressing a greater level of the protein bind a greater level of the labeled antibody and hence can be sorted according to fluorescence intensity. GFP and RCFP have been widely applied for FACS as an alternative to the use of fluorescently labeled antibodies. In transient gene expression studies, cells are often cotransfected with the gene under study together with a GFP expression vector in order to sort transfected from untransfected cells by FACS. For example Guerra-Crespo et al. (2003) cotransfected primary rat hypothalamic neurones with a GFP reporter under the transcriptional control of the thyrotrophin-releasing hormone (TRH) promoter. FACS was applied to isolate TRH neurons for further study. Similar approaches have been developed to isolate specific cell lineages from transgenic animals. Tomomura et al. (2001) generated a transgenic mouse expressing GFP within Purkinje cells and applied FACS to isolate these cells. The use of GFP as a marker for FACS removes the requirement for the generation and labeling of specific antisera. The technique can be applied to sort cells on the basis of their expression of potentially any protein, and the technique requires no specific preparation of the cells prior to FACS. In addition to the use of FACS for monitoring the level of protein expression within a given cell type, FACS assays have been developed to detect the effect of drug molecules on the level of protein expression within given cells and the proportion of cells of a particular type within a tissue sample. Furthermore, GFP in combination with FACS has been used to isolate specific cell types from transgenic animals for the development of drug screening assays.

16.3.10

Fluorescent Protein Biosensors

GFP biosensors are genetically encoded probes, the fluorescent characteristics of which are modified by a change in the level of a cellular metabolite or second messenger protein. A GFP biosensor acts as a direct probe to enable the detection of a change in the level of expression of a second messenger metabolite in living mammalian cells in culture. The ability to genetically express reporter molecules in recombinant cell lines, primary cell lines, tissue slices, or indeed whole animals may be one of the most exciting applications of fluorescent protein technology (Conway and Demarest, 2002; Giuliano and Taylor, 1998). Such assays allow the direct visualization of signal transduction in conventional plate readers or confocal imaging systems without the need for assay reagent addition and cell lysis. Such assays are inexpensive to run and generate direct information regarding the level of the second messenger in the cell and, perhaps through the targeting of the biosensor, the level of the second messenger within a discrete cellular location. The use of GFP/RCFP as biosensors in drug screening assays for HTS offers a number of benefits compared to the use of chemical detection agents such as fluorescent dyes. Flu-

FLUORESCENT PROTEIN DRUG SCREENING ASSAYS

orescent dyes have been widely used to detect changes in cellular pH and changes in the intracellular concentration of sodium, calcium, potassium, and other ions, as indicators of cell number or cytotoxicity, and as detectors of the level of intracellular second messengers. Such dyes are expensive, the assays always involve a dye loading step that is often followed by a wash step to remove excess dye, the dyes are often subject to photobleaching, and they can be cytotoxic. In contrast, GFP/RCFP fluorescent indicator proteins are expressed within the cell; hence there is no need for dye loading and cell washing. The assays are inexpensive because no chemical detection agent is required, and there is little cytotoxicity. The absence of dye loading and wash steps in fluorescent protein assays is a particular advantage because this simplifies the assay protocol, leading to the generation of a more robust assay that is easier to transfer onto HTS laboratory automation. An early example of the use of GFP as a biosensor arose from the observation that the fluorescence characteristics of the YFP-H148Q mutant are pH-sensitive, leading to the application of this protein as a probe for changes in intracellular pH (Kneen et al., 1998). It was later observed that this protein exhibits halide sensitivity. At pH 7.5 the fluorescence emission of this protein decreases twofold (Jayaraman et al., 2000). In perhaps the first example of the use of a GFP biosensor for drug screening, a high-throughput screen has been run to identify activators of the cystic fibrosis transmembrane conductance regulator (CFTR). A stable cell line expressing both the CFTR and the YFP-H158Q protein was used to screen a 60,000 compound library in 96-well assay plates. CFTR activators caused an increase in intracellular chloride concentration to cause a decrease in YFPH148Q emission that was detected using a FluroStar fluorescence plate reader (BMG Lab Technologies). This screen resulted in the identification of a number of novel activators of CFTR and demonstrates the potential utility of this fluorescent protein for drug screening (Ma et al., 2002). Two of the most commonly used drug screening methods for GPCRs are the detection of drug-mediated changes in intracellular calcium or cAMP. Calcium assays are performed using the fluorescent indicator dyes Fluo-3 or Fluo-4 and the fluorescence imaging plate reader (FLIPR; Molecular Devices, Sunnyvale, CA). A variety of methods exist for the detection of changes in intracellular cAMP. Genetically encoded FRET indicators have been developed for the detection of changes in intracellular calcium (Miyawaki et al., 1997, 1999) and cAMP (Zaccolo et al., 2000). A calcium indicator, or chameleon, was developed through the construction of a fusion protein consisting of cyan fluorescent protein (CFP) and yellow fluorescent protein (YFP) separated by the calcium-binding protein calmodulin and the calmodulin-binding peptide M13. Binding of calcium to calmodulin facilitates the interaction between calmodulin and M13 to cause an intramolecular rearrangement to cause an increase in FRET between the flanking CFP and YFP. Such proteins have been used to monitor calcium levels in the cytosol, endoplasmic recticulum, and nucleus of living cells (Miyawaki et al., 1997). A similar fluorescent indicator for cAMP was developed through generating fusion proteins between the regulatory and catalytic subunits of protein kinase A and a blue fluorescent protein and GFP, respectively. Binding of cAMP to the regulatory subunit causes a dissociation of the PK-A complex to result in a decrease of FRET. In contrast to existing assays for changes in the intracellular concentration of cAMP, which involve cell lysis and the use of anti-cAMP antisera, fluorescent cAMP biosensors can be used non-invasively in living cells (Zuccolo et al., 2000). It will be of great interest to see whether the fluorescent protein biosensors will replace the use of these dye or biochemical assays for GPCR HTS.

379

380

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

A GFP biosensor designed to detect changes in cellular membrane potential as a result of ion channel activity has been developed through the generation of a fusion protein between the Drosophila shaker potassium channel and GFP. In this example, the fluorescence characteristics of the fusion protein change according to the membrane potential of the cell (Guerrero and Isacoff, 2001). This probe offers an alternative to the use of fluorescent membrane potential dyes for compound screening at ion channels and other targets that regulate cell membrane potential.

16.4

HIGH CONTENT SCREENING DETECTION INSTRUMENTATION

Throughout this chapter we have described the development of HCS assays that rely upon the translocation of fluorescent protein fusion proteins, or the gain or loss of fluorescence from a fluorescent protein, in live and fixed cells. Prior to the development of instruments capable of the detection of changes in the level and intracellular localization of fluorescence in single cells plated in 96- and 384-well microtiter plates, assays such as receptor internalisation or nuclear translocation could be visualized on a fluorescence microscope but could not be quantified in a high-throughput manner. Moreover, the development of automated image analysis software available on such instruments has been critical to convert virtual features into numerical results that can be used to assess the efficacy of a drug. A number of instruments are now available that can be used for this purpose. Each of these instruments has different features such as differences in excitation light source or detection methods, cellular resolution, speed of image acquisition and data processing and cost of instrumentation. A brief comparison of currently available instruments can be found in Table 16.4. Most, if not all, are able to detect membrane to cytosol translocations (receptor internalisation), cytosol to membrane translocations (b-arrestin recruitment), and membrane to nucleus translocation (nuclear translocation assays) and will detect changes in fluorescence intensity. The majority of these readers can also be integrated onto robot tracks to allow fully automated high content screens to be run as high-throughput screens. It is difficult to provide an exhaustive list of the different applications of the various instruments because few comparative studies have been reported. Furthermore, the development of such instrumentation is proceeding at a rapid pace (Table 16.2).

16.5 APPLICATION OF FLUORESCENT PROTEINS IN TARGET VALIDATION STUDIES Genome sequencing and bioinformatic studies indicate that there are approximately 28,000 genes in the human genome. The role of the vast majority of these genes in both normal physiology and disease processes remains unknown. About 5000 of these genes fall into the so-called tractable target classes—that is, classes of protein such as GPCRs, ion channels, proteases, kinases, and so on, for which there are examples of clinically marketed drugs. There is intense effort within the pharmaceutical industry to identify the function of novel genes—in particular, those that fall within the tractable target classes—and their possible involvement in disease in order to identify the drug targets of the future. The techniques used to achieve this have been loosely grouped under the label “target validation” and have been reviewed elsewhere (Lindsay 2003; Wise et al., 2002). Target validation studies rely upon the detection of the spatial and temporal nature of target expression, the study of whether target expression is altered in disease, the identification of interacting

APPLICATION OF FLUORESCENT PROTEINS IN TARGET VALIDATION STUDIES

TABLE 16.4. Key Features of Commonly Used High Content Screening Detection Apparatus Company

Reader

Attributes

Cellomics

ArrayScanTM II

Mercury arc lamp, nonconfocal Four excitation and four emission filters CCD camera

ArrayScanTM Kinetics

Mercury arc lamp, nonconfocal Eight excitation and eight emission filters CCD camera Liquid handling, robotics

IN Cell Anaylser 1000TM

Xenon lamp, line scanning confocal Multiple excitation and emission filters CCD camera Liquid handling

IN Cell Anaylser 3000TM

Laser (3 lines), line scanning confocal Multiple emission filters 3¥ CCD cameras Liquid handling

Evotec

Opera

Laser (3 lines), Nipkow disc, confocal Multiple emission filters 2¥ CCD cameras Plate formats up to 2080 Separate image capture and analysis

Acumen

Explorer

Laser (488 nm), nonconfocal 4 PMT detection Independent of plate type

Molecular Devices

Discovery-1

Arc lamp Multiple excitation and emission filters CCD camera

GE Healthcare

partners of proteins of interest, the generation of transgenic animals, and many other studies. The ultimate goal of target validation is the generation of data that indicates that therapeutic intervention at the protein under study will have efficacy in human disease. Aequorea victorea GFP and the RCFPs are becoming valuable tools within the target validation phase of drug discovery. Fluorescent proteins have been applied to determine the tissue and subcellular localization of a novel protein. The expression in mammalian cell lines of fusion proteins between a fluorescent protein and the protein under study can allow the detection of the site of expression of that protein, and can be used to determine if this is altered by known stimuli in order to provide some indication of function. Similar studies in transgenic animals can identify all sites of protein expression, and studies in disease models can be used to determine if gene expression is altered during disease progression (Hadjantonakis and Nagy, 2001). Fluorescent proteins are being extensively used in pathway mapping. The objective of such studies is to take a protein of unknown function and identify interacting partners in order to identify whether that protein is involved in a known cellular process. In the past techniques such as the yeast two-hybrid system have been used for this purpose (Causier and Davies, 2002). Many groups are now developing both FRET- and BRETbased assays in order to perform such studies in mammalian cells and other species. FRET

381

382

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

and BRET are also being applied to confirm the results of yeast two-hybrid studies in mammalian cells. One example of this approach is the demonstration of the interaction of the GABA-BR1 and GABA-BR2 receptors by FRET which confirmed earlier studies using the yeast two hybrid assay (White et al., 2002; Fig. 16.3). The investigation of the function of a target gene has been greatly facilitated through the development of overexpression and gene deletion techniques in model species such as the nematode worm Caenorhabditis elegans (David et al., 2003), the fruit fly Drosophila melanogaster (Misteli and Spector 1997; Brand 1999), and transgenic mice (Hadjantonakis and Nagy, 2001; Misteli and Spector, 1997; Zambrowicz and Sands, 2003). Such techniques are widely applied throughout academia and are in many pharmaceutical companies. Fluorescent protein technology has been applied to facilitate the detection of the site of expression of the gene of interest and to develop whole organism model systems to study the effects of novel drugs. The nematode worm C. elegans contains just 128 cells. The developmental fate and function of each of these cells has been well-characterized. As a result, this animal has been applied to study the effect of the deletion of genes on cell fate patterning during development and on the physiology of the mature worm. Fluorescent proteins have been applied as a reporter to enable the noninvasive generation of data from the live worm. As an example of the application of GFP in this species, David et al. (2003) have developed a strain of C. elegans expressing a hsp16-GFP-lacZ fusion protein for use in environmental monitoring. In this animal a chimeric GFP/b-galactosidase (lacZ) reporter gene was fused in frame into the C. elegans hsp16 gene such that expression of the reporter is under the transcriptional control of the hsp16 promoter. The reporter was modified to facilitate detection such that upon expression it accumulates in the cell nucleus. This worm has been used to examine the effects of environmental stressors such as heat and microwave radiation. In each case the authors were able to monitor the stress effect in the whole organism through the detection of the fluorescence intensity of the cell nuclei. GFP has been used extensively in transgenic mice as a reporter of gene expression. Such studies have been reviewed by Hadjantonakis and Nagy (2001) and Misteli and Spector (1997). Technologies have been developed that permit the targeted deletion of specific genes in mice. During the generation of such animals, the inclusion of a reporter gene in the gene deletion vector enables the creation of transgenic mice no longer expressing the gene under study and now expressing the reporter gene under the transcriptional control of the promoter of the deleted gene. Similarly, experiments have been performed to generate fusion proteins between the gene under study and a reporter gene. GFP has been widely applied as the reporter in transgenic animals because the sites of gene expression can be readily detected in cells and tissue slices by fluorescence microscopy. For example, a transgenic mouse expressing endothelial nitric oxide synthase (eNOS) fused to GFP was generated to identify the location and regulation of e-NOS expression (van Haperen et al, 2003). The expression of the fusion protein facilitated the isolation of eNOS expressing tissues from these animals through the detection of GFP fluorescence. These animals have been used to study the effects on eNOS expression of several vascular challenges. This application of GFP permits the study of the temporal and spatial expression of the targeted gene through imaging of GFP expression. In a second example a transgenic mouse was generated expressing GFP under the control of the melanocortin-4 (MC4) receptor promoter (Liu et al., 2003). This receptor is an important regulator of energy homeostasis, and antagonists of this receptor are proposed to have efficacy in obesity. These animals have been used to study the site of MC4 receptor expression and the role of this receptor on feeding behavior. In a final example, transgenic mice have been developed expressing

CONCLUDING REMARKS

GFP under the control of the mouse insulin I gene promoter (MIP). In such animals, GFP is expressed within the insulin-secreting beta cells of the pancreatic islet. These animals are being applied to study beta cell biology in normal and diabetic animals (Hara et al., 2003). Fluorescent proteins have been applied in transgenic mice to determine the effect of drug molecules on the level of expression of the deleted gene in the intact animal through monitoring changes in GFP expression. Lindsten et al. (2003) created transgenic mice expressing a GFP reporter carrying a constitutively active degradation signal to generate a model for the study of the ubiquitin/proteasome system. Impairment of this system has been proposed to play a role in many neurodegenerative disorders. Administration of proteosome inhibitors to these animals resulted in the accumulation of GFP in several tissues through the prevention of GFP degradation. This animal has been used as a model for the characterization of novel inhibitors of the ubiuquitin/proteosome system. Furthermore, primary neurones have been isolated from these animals and used in transfection studies to demonstrate that an aberrant ubiquitin found in Alzheimer’s disease patients causes the accumulation of the GFP reporter. These cells and animals could be applied to understand the role of ubiquitin in this disease and to screen for inhibitors of this complex. In a further novel application of GFP as a reporter in transgenic mice, Metzger et al. (2002) have expressed the pH- and halide-sensitive GFP described earlier (Jayaraman et al., 2000) in transgenic mice under the control of a potassium channel promoter. Neuronal tissue from these animals has been used to study drug effects by imaging changes in cellular fluorescence. The generation and characterization of transgenic animals expressing GFP reporters has facilitated the study of the effects of external stimuli, including novel small-molecule drugs, both within the intact animal and in tissue slices and primary cell lines generated from such animals. The advantages of GFP as a reporter for these purposes is that drug effects can be monitored noninvasively by simple imaging of the cell lines or tissue samples under study. In addition to the application of GFP for target validation, fluorescent protein technology may have application within the toxicology departments of pharmaceutical companies. It will be possible to develop novel noninvasive assays that permit the study of compound activity within both the whole animal and tissue samples by determining the effect of such compounds on the site and level of expression of GFP reporters. The challenge in this work will be the development and validation of assays suitable for this purpose.

16.6

CONCLUDING REMARKS

In the last five years, fluorescent protein technology has become integral to drug discovery. To date, the greatest impact of such proteins has been within the compound screening phase of drug discovery for which a plethora of fluorescent protein screening assays and fluorescent protein probes have been developed. In parallel, many instruments have become available that allow the detection of changes in the fluorescence characteristics of single and populations of cells within microtiter plate formats. It is now possible to determine the effect on the subcellular localization of a fluorescent protein fusion protein of 384 compounds within six minutes, a process which by confocal microscopy would have taken several days, if not weeks. In the coming years, with the development of more sensitive and faster detection apparatus, and improved fluorescent protein probes and biosensors, fluorescent protein screening assays will become integral to the high-throughput

383

384

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

screening and pharmacology departments in most pharmaceutical companies. These developments should permit drug screening to be performed on biologically relevant cell types, perhaps including primary human cells, at very high throughput. The application of fluorescent protein technology within the development organisations of most drug companies is in its infancy. In the coming years, fluorescent protein technology will be applied to generate novel drug developability and toxicology assays. The application of fluorescent protein assays for primary screening, developability, and toxicology should lead to the identification of molecules with improved efficacy, toxicity profiles, and pharmacokinetics and thus should contribute to the reduction of both both cycle time and attrition within the drug discovery process.

ACKNOWLEDGMENTS The authors would like to acknowledge the contribution of many colleagues to the work described in this chapter. In particular, we would like to thank Mike Allen, Peter Chalk, Katy Gearing, Debbie Graham, Brian Hayes, Peter Lowe, Barbara Maschera, Rebecca Milton, Alan Wise, and Julie White, all of GlaxoSmithKline, Stevenage. These studies have been performed through a number of collaborations. The BRET studies were performed in collaboration with Benoit Houle and Luc Menard of Perkin Elmer Life Sciences, Montreal, and Greame Milligan and Douglas Ramsay of the University of Glasgow. The GR translocation studies were performed in collaboration with Suzanne Hancock and Fergus Mckenzie of GE Healthcare Biosciences, Cardiff and the arrestin recruitment studies with Robert Oakly, Rachel Cruickshank, Shay Rhem, and Carson Loomis of Xsira Pharmaceuticals, North Carolina.

REFERENCES Almholt, K., Arkhammar, P. O., Thastrup, O., and Tullin, S. (1999). Simultaneous visualization of the translocation of protein kinase Ca-green fluorescent protein hybrids and intracellular calcium concentrations. Biochem. J. 337:211–218. Awaji, T., Hirasawa, A., Kataoka, M., Shinoura, H., Nakayama. Y., Sugawara, T., Izumi, S.-I., and Tsujimoto, G. (1998). Real-time optical monitoring of ligand-mediated internalization of a1badrenoceptor with green fluorescent protein. Mol. Endocrinol. 12:1099–1111. Barak, L. S., Ferguson, S. S., Zhang, J., and Caron, M. G. (1997a). A b-arrestin/green fluorescent protein biosensor for detecting G protein-coupled receptor activation. J. Biol. Chem. 272: 27497–27500. Barak, L. S., Ferguson, S. S. G., Zhang, J., Martenson, C., Meyer, T., and Caron, M. G. (1997b). Internal trafficking and surface mobility of a functionally intact b2-adrenergic receptor-green fluorescent protein conjugate. Mol. Pharm. 51:177–184. Barlic, J., Khandaker, M. H., Mahon, E., Andrews, J., DeVries, M. E., Mitchell, G. B., Rahimpour, R., Tan, C. M., Ferguson, S. S., and Kelvin, D. J. (1999). b-Arrestins regulate interleukin-8induced CXCR1 internalization. J. Biol. Chem. 274:16287–16294. Berdichevsky, Y., Zemel, R., Bachmatov, L., Abramovich, A., Koren, R., Sathiyamoorthy, P., GolanGoldhirsh, A., Tur-Kaspa, R., and Benhar, I. (2003). Novel high throughput screening assay for HCV NS3 serine protease inhibitors. J. Virol. Methods 107:245–255. Bertrand, L., Parent, S., Caron, M., Legault, M., Joly, E., Angers, S., Bouvier, M., Brown, M., Houle, B., and Menard, L. (2002). The BRET2/arrestin assay in stable recombinant cells: A platform to

REFERENCES

screen for compounds that interact with G protein-coupled receptors. J. Rec. Signal Transduction Res. 22:533–541. Boute, N., Jockers, R., and Issad, T. (2002). The use of resonance energy transfer in highthroughput screening: BRET versus FRET. Trends Pharmacol. Sci. 23:351–354. Boute, N., Boubekeur, S., Lacasa, D., and Issad, T. (2003). Dynamics of the interaction between the insulin receptor and protein tyrosine-phosphatase 1B in living cells. EMBO Rep. 4:313–319. Brand, A. (1999). GFP as a cell and developmental marker in the Drosophila central nervous system. Methods Cell Biol. 58:153–163. Causier, B., and Davies, B. (2002). Analysing protein–protein interactions with the yeast two-hybrid system. Plant Mol. Biol. 50:855–870. Claing, A., Laporte, S. A., Caron, M. G., and Lefkowitz, R. J. (2002). Endocytosis of G proteincoupled receptors: Roles of G protein-coupled receptor kinases and b-arrestin proteins. Prog. Neurobiol. 66:61–79. Conway, B. R., and Demarest, K. T. (2002). The use of biosensors to study GPCR function: Application for high-content screening. Receptors and Channels 8:331–341. Conway, B. R., Minor, L. K., Xu, J. Z., Gunnet, J. W., DeBiasio, R., D’Andrea, M. R., Rubin, R., DeBiasio, R. L., Giuliano, K., Zhou, L., and Demarest, K. T. (1999). Quantification of G-protein coupled receptor internalization using G-protein coupled receptor-green fluorescent protein conjugates with the arrayscan high-content screening system. J. Biomol. Scr. 4:75–86. Conway, B. R., Minor, L. K., Xu, J. Z., D’andrea, M. R., Ghosh, R. N., and Demarest, K. T. (2001). Quantitative analysis of agonist-dependent parathyroid hormone receptor trafficking in whole cells using a functional green fluorescent protein conjugate. J. Cell. Physiol. 189:341– 355. David, H. E., Dawe, A. S., de Pmoerai, D. I., Jones, D., Candido, E. P., and Daniells, C. (2003). Construction and evaluation of transgenic hsp16-GFP-lacZ Caenorhabditis elegans strain for environmental monitoring. Environ. Toxicol. Chem. 22:111–118. Drmota, T., Novotny, J., Kim, G.-D., Eidne, K. A., Milligan, G., and Svoboda, P. (1998). Agonistinduced internalization of the G protein G11a and thyrotropin-releasing hormone receptors proceed on different time scales. J. Biol. Chem. 273:21699–21707. Eidne, K. A., Kroeger, K. M., and Hanyaloglu, A. C. (2002). Applications of novel resonance energy transfer techniques to study dynamic hormone receptor interactions in living cells. Trends Endocrinol. Metabol. 13:415–421. Evans, N. A., Groarke, D. A., Warrack, J., Greenwood, C. J., Dodgson, K., Milligan, G., and Wilson, S. (2001). Visualizing differences in ligand-induced b-arrestin-GFP interactions and trafficking between three recently characterized G protein-coupled receptors. J. Neurochem. 77:476–485. Feng, X., Zhang, J., Barak, L. S., Meyer, T., Caron, M. G., and Hannun, Y. A. (1998). Visualization of dynamic trafficking of a protein kinase C bII/green fluorescent protein conjugate reveals differences in G protein-coupled receptor activation and desensitization. J. Biol. Chem. 273:10755–10762. Ferguson, S. S., Zhang, J., Barak, L. S., and Caron, M. G. (1998). Molecular mechanisms of G protein-coupled receptor desensitization and resensitization. Life Sci. 62:1561–1565. Galarneau, A., Primeau, M., Trudeau, L. E., and Michnick, S. W. (2002). b-Lactamase protein fragment complementation assays as in vivo and in vitro sensors of protein protein interactions. Nat. Biotechnol. 20:619–622. Giuliano, K. A., and Taylor, D. L. (1998). Fluorescent-protein biosensors: New tools for drug discovery. Trends Biotechnol. 16:135–140. Giuliano, K. A., DiBiasio, R. L., Dunlay, R. T., Gough, A., Volosky, J. M., Zock, J., Pavlakis, G. N., and Taylor, D. L. (1997). High-content screening: A new approach to easing key bottlenecks in the drug discovery process. J. Biomol. Scr. 2:249–259.

385

386

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

Graham, D. L., Bevan, N., Lowe, P. N., Palmer, M., and Rees, S. (2001). Application of bgalactosidase enzyme complementation technology as a high throughput screening format for antagonists of the epidermal growth factor receptor. J. Biomol. Scr. 6:401–411. Groarke, D. A., Wilson, S., Krasel, C., and Milligan, G. (1999). Visualization of agonist-induced association and trafficking of green fluorescent protein-tagged forms of both b-arrestin-1 and the thyrotropin-releasing hormone receptor-1. J. Biol. Chem. 274:23263–23269. Guerra-Crespo, M., Charli, J. L., Rosales-Garcia, V. H., Pedraza-Alva, G., and Perez-Martinez, L. (2003). Polyetheylenimine improved the transfection efficiency of primary cultures of postmitotic rat fetal hypothalamic neurons. J. Neurosci. Methods 127:179–192. Guerrero, G., and Isacoff, E. Y. (2001). Genetically encoded optical sensors of neuronal activity and cellular function. Curr. Opin. Neurobiol. 11:601–607. Hadjantonakis, A. K., and Nagy, A. (2001). The color of mice: In the light of GFP-variant reporters. Histochem. Cell Biol. 115:49–58. Hara, M., Wang, X., Kawamura, T., Bindokas, V. P., Dizon, R. F., Alcoser. S. Y., Magnuson, M. A., and Bell, G. I. (2003). Transgenic mice with green fluorescent protein-labelled pancreatic beta cells. Am. J. Physiol. 284:E177–E183. Herezenberg, L. A., and de Rosa, S. C. (2000). Moloclonal antibodies and FACS: Complementary tools for immunobiology and medicine. Immunol. Today 21:383–390. Hill, S. J., Baker, J. G., and Rees, S. (2001). Reporter gene systems for the study of G-protein coupled receptors. Curr. Opin. Pharmacol. 1:526–532. Horgan, A. M., and Stork, P. J. (2003). Examining the mechanism of Erk nuclear translocation using green fluorescent protein. Exp. Cell Res. 285:208–220. Htun, H., Barsony, J., Renyi, I., Gould, D. L., and Hager, G. L. (1996). Visualization of glucocorticoid receptor translocation and intranuclear organization in living cells with a green fluorescent protein chimera. Proc. Natl. Acad. Sci. USA 93:4845–4850. Issad, T., Boute, N., and Pernet, K. (2002). The activity of the insulin receptor assessed by bioluminescence resonance energy transfer. Ann. NY Acad. Sci. 973:120–123. Jayaraman, S., Haggie, P., Wachter, R. M., Remington, S. J., and Verkman, A. S. (2000). Mechanism and cellular applications of a green fluorescent protein-based halide sensor. J. Biol. Chem. 275:6047–6050. Jensen, A. A., Hansen, J. L., Sheikh, S. P., and Brauner-Osborne, H. (2002). Probing intermolecular protein–protein interactions in the calcium-sensing receptor homodimer using bioluminescence resonance energy transfer (BRET). Eur. J. Biochem. 269:5076–5087. Jones, J., Heim, R., Hare, E., Stack, J., and Pollok, B. A. (2000). Development and application of a GFP-FRET intracellular caspase assay for drug screening. J. Biomol. Scr. 5:307–318. Kallal, L., Gagnon, A. W., Penn, R. B., and Benovic, J. L. (1998). Visualization of agonist-induced sequestration and down-regulation of a green fluorescent protein-tagged b2-adrenergic receptor. J. Biol. Chem. 273:322–328. Kallal, L., and Benovic, J. L. (2000). Using green fluorescent proteins to study G-protein-coupled receptor localization and trafficking. Trends Pharmacol. Sci. 21:175–180. Kapur, R. (2002). Fluorescence imaging and engineered biosensors: functional and activity-based sensing using high content screening. Ann. NY Acad. Sci. 961:196–197. Kneen, M., Farinas, J., Li, Y., and Verkman, A. S. (1998). Green fluorescent protein as a noninvasive intracellular pH indicator. Biophys. J. 74:1591–1599. Kohl, T., Heinze, K. G., Kuhlemann, R., Koltermann, A., and Schwille, P. (2002). A protease assay for two-photon cross correlation and FRET analysis based solely on fluorescent proteins. Proc. Natl. Acad. Sci. USA 99:12161–12166. Koster, M., and Hauser, H. (1999). Dynamic redistribution of STAT1 protein in IFN signaling visualized by GFP fusion proteins. Eur. J. Biochem. 260:137–144.

REFERENCES

Lamb, M. E., De Weerd, W. F., and Leeb-Lundberg, L. M. (2001). Agonist-promoted trafficking of human bradykinin receptors: Arrestin- and dynamin-independent sequestration of the B2 receptor and bradykinin in HEK293 cells. Biochem. J. 355:741–750. Li, X., Zhao, X., Fang, Y., Jiang, X., Duong, T., Fan, C., Huang, C. C., and Kain, S. R. (1998). Generation of destabilized green fluorescent protein as a transcription reporter. J. Biol. Chem. 273:34970–34975. Li, X., Fang, Y., Zhao, X., Jiang, X., Duong, T., and Kain, S. R. (1999). Characterization of NFkB activation by detection of green fluorescent protein-tagged IkB degradation in living cells. J. Biol. Chem. 274:21244–21250. Lindstren, K., Menendez-Benito, V., Masucci, M. G., and Dantuma, N. P. (2003). A transgenic mouse model of the ubiquitin/proteasome system. Nat. Biotechnol. 21:897–902. Liptrot, C. (2001). High content screening—From cells to data to knowledge. Drug Disc. Today 6:832–834. Liu, H., Kishi, T., Roseberry, A. G., Cai, X., Lee, C. E., Montez, J. M., Friedman, J. M., and Elmquist. J. K. (2003). Transgenic mice expressing green fluorescent protein under the control of the melanocortin-4 receptor promoter. J. Neurosci. 23:7143–7154. Lindsay, M. (2003). Target discovery. Nat. Rev. Drug Disc. 2:831–838. Loster, K., and Horstkorte, R. (2000). Enzymatic quantification of cell–matrix and cell–cell adhesion. Micron 31:41–53. Luttrell, L. M., Ferguson, S. S. G., Daaka, Y., Miller, W. E., Maudsley, S., Della Rocca, G. J., Lin, F.-T., Kawakatsu, H., Owada, K., Luttrell, D. K., Caron, M. G., and Lefkowitz, R. J. (1999). barrestin-dependent formation of b2 adrenergic receptor-Src protein kinase complexes. Science 283:655–661. Ma, T., Vetrivel, L., Yang, H., Pedemonte, N., Zegarra-Moran, O., Galietta, L. J., and Verkman, A. S. (2002). High-affinity activators of cystic fibrosis transmembrane conductance regulator (CFTR) chloride conductance identified by high-throughput screening. J. Biol. Chem. 277: 37235–37241. Mahajan, N. P., Harrison-Shostak, D. C., Michaux, J., and Herman, B. (1999). Novel mutant green fluorescent protein protease substrates reveal the activation of specific caspases during apoptosis. Chem. Biol. 6:401–409. Matharu, A. L., Mundell, S. J., Benovic, J. L., and Kelly, E. (2001). Rapid agonist-induced desensitization and internalization of the A(2B) adenosine receptor is mediated by a serine residue close to the COOH terminus. J. Biol. Chem. 276:30199–30207. Mathieu, S., and El-Battari, A. (2003). Monitoring E-selectin-mediated adhesion using green and red fluorescent proteins. J. Immunol. Methods 272:81–92. McLean, A. J., Bevan, N., Rees, S., and Milligan, G. (1999). Visualizing differences in ligand regulation of wild-type and constitutively active mutant b2-adrenoceptor-green fluorescent protein fusion proteins. Mol. Pharm. 56:1182–1191. Mercier, J. F., Salahpour, A., Angers, S., Breit, A., and Bouvier, M. (2002). Quantitative assessment of b1- and b2-adrenergic receptor homo- and heterodimerization by bioluminescence resonance energy transfer. J. Biol. Chem. 277:44925–44931. Metzger, F., Repunte-Canaonigo, V., Matsushita, S., Akeman, W., Diez-Garcia, J., Ho, C. S., Iwasto, T., Grandes, P., Itohara, S., Joho, R. H., and Knopfel, T. (2002). Transgenic mice expressing a pH and Cl- sensing yellow fluorescent protein under the control of a potassium channel promoter. Eur. J. Neurosci. 15:40–50. Misteli, T., and Spector, D. L. (1997). Applications of green fluorescent protein in cell biology and biotechnology. Nat. Biotechnol. 15:961–964. Miyawaki, A., Llopis, J., Heim, R., McCaffery, J. M., Adams, J. A., Ikura, M., and Tsien, R. Y. (1997). Fluorescent indicators for Ca2+ based on green fluorescent proteins and calmodulin. Nature 388:882–887.

387

388

PHARMACEUTICAL APPLICATIONS OF GFP AND RCFP

Miyawaki, A., Griesbeck, O., Heim, R., and Tsien, R. Y. (1999). Dynamic and quantitative Ca2+ measurements using improved cameleons. Proc. Natl. Acad. Sci. USA 96:2135–2140. Mundell, S. J., Matharu, A. L., Pula, G., Roberts, P. J., and Kelly, E. (2001). Agonist-induced internalization of the metabotropic glutamate receptor 1a is arrestin- and dynamin-dependent. J. Neurochem. 78:546–551. Nagy, S. R., Liu, G., Lam, K. S., and Denison, M. S. (2002). Identification of novel Ah receptor agonists using a high-throughput green fluorescent protein-based recombinant cell bioassay. Biochemistry 41:861–868. Ozawa, T., Takeuchi, T. M., Kaihara, A., Sato, M., and Umezawa, Y. (2001). Protein splicing-based reconstitution of split green fluorescent protein for monitoring protein–protein interactions in bacteria: Improved sensitivity and reduced screening time. Anal. Chem. 73:5866–5874. Patki, V., Buxton, J., Chawla, A., Lifshitz, L., Fogarty, K., Carrington, W., Tuft, R., and Corvera, S. (2001). Insulin action on GLUT4 traffic visualized in single 3T3-l1 adipocytes by using ultrafast microscopy. Mol. Biol. Cell. 12:129–141. Rees, S., Brown, S., and Stables, J. (1999). Reporter gene systems for the study of G-protein coupled receptor signal transduction in mammalian cells. In Milligan, G., Eds. Signal Trandsduction: A Practical Approach, 2nd Ed., Oxford University Press, New York. pages 171–221. Richardson, M. D., Balius, A. M., Yamaguchi, K., Freilich, E. R., Barak, L. S., and Kwatra, M. M. (2003). Human substance P receptor lacking the C-terminal domain remains competent to desensitize and internalize. J. Neurochem. 84:854–863. Schlador, M. L., and Nathanson, N. M. (1997). Synergistic regulation of m2 muscarinic acetylcholine receptor desensitization and sequestration by G protein-coupled receptor kinase-2 and b-arrestin1. J. Biol. Chem. 272:18882–18890. Schmid, J. A., Birbach, A., Hofer-Warbinek, R., Pengg, M., Burner, U., Furtmuller, P. G., Binder, B. R., and de Martin, R. (2000). Dynamics of NFkB and IkBa studied with green fluorescent protein (GFP) fusion proteins. Investigation of GFP-p65 binding to DNA by fluorescence resonance energy transfer. J. Biol. Chem. 275:17035–17042. Scott, E. S., Malcomber, S., and O’Hare, P. (2001). Nuclear translocation and activation of the transcription factor NFAT is blocked by herpes simplex virus infection. J. Virol. 75:9955– 9965. Sekar, R. B., and Periasamy, A. (2003). Fluorescence resonance energy transfer (FRET) microscopy imaging of live cell protein localizations. J. Cell Biol. 160:629–633. Slice, L. W., Yee, H. F. Jr., and Walsh, J. H. (1998). Visualization of internalization and recycling of the gastrin releasing peptide receptor-green fluorescent protein chimera expressed in epithelial cells. Receptors and Channels 6:201–212. Subbaramaiah, K., Bulic, P., Lin, Y., Dannenberg, A. J., and Pasco, D. S. (2001). Development and use of a gene promoter-based screen to identify novel inhibitors of cyclooxygenase-2 transcription. J. Biomol. Scr. 6:101–110. Tarasova, N. I., Stauber, R. H., Choi, J. K., Hudsoni, E. A., Czerwinski, G., Miller, J. L., Pavlakis, G. N., Michejda, C. J., and Wank, S. A. (1997). Visualization of G protein-coupled receptor trafficking with the aid of the green fluorescent protein. J. Biol. Chem. 272:14817–14824. Tawa, P., Tam, J., Cassady, R., Nicholson, D. W., and Xanthoudakis, S. (2001). Quantitative analysis of fluorescent caspase substrate cleavage in intact cells and identification of novel inhibitors of apoptosis. Cell Death Differ. 8:30–37. Terstappen, G. C., Giacometti, A., Ballini, E., and Aldegheri, L. (2000). Development of a functional reporter gene HTS assay for the identification of mGluR7 modulators. J. Biomol. Scr. 5:255–262. Tomomura, M., Rice, D. S., Morgan, J. I., and Yuzaki, M. (2001). Purification of Purkinje cells by fluorescence activated cell sorting from transgenic mice that express green fluorescent protein. Eur. J. Neurosci. 14:57–63.

REFERENCES

van Haperen, R., Cheng, C., Mees, B. M., van Deel, E., de Waard, M., van Damme, L. C., van Gent, T., van Aken, T., Krams, R., Dunker, D. J., and de Crom, R. (2003). Functional expression of endothelial nitric oxide synthase fused to green fluorescent protein in transgenic mice. Am. J. Path. 163:1677–1686. Vrecl, M., Anderson, L., Hanyaloglu, A., McGregor, A. M., Groarke, A. D., Milligan, G., Taylor, P. L., and Eidne, K. A. (1998). Agonist-induced endocytosis and recycling of the gonadotropinreleasing hormone receptor: Effect of b-arrestin on internalization kinetics. Mol. Endocrinol. 12:1818–1829. Walters, W. P., and Namchuk, M. (2003). Designing screens: How to make your hits a hit. Nature Drug Disc. Today 2:259–266. Wang, X. J., Liao, H. J., Chattopadhyay, A., and Carpenter, G. (2001). EGF-dependent translocation of green fluorescent protein-tagged PLC-g1 to the plasma membrane and endosomes. Exp. Cell Res. 267:28–36. White, J. H., Wise, A., and Marshall, F. H. (2002). Heterodimerization of gamma-aminobutyric acid B receptor subunits as revealed by the yeast two-hybrid system. Methods 27:301–310. Wise, A., Gearing, K., and Rees, S. (2002). Target validation of G-protein coupled receptors. Drug Discov. Today 7:235–246. Zaccolo, M., De Giorgi, F., Cho, C. Y., Feng, L., Knapp, T., Negulescu, P. A., Taylor, S. S., Tsien, R. Y., and Pozzan, T. (2000). A genetically encoded, fluorescent indicator for cyclic AMP in living cells. Nat. Cell Biol. 2:25–29. Zambrowicz, B. P., and Sands, A. T. (2003). Knockouts model the 100 best-selling drugs—will they model the next 100? Nat. Rev. Drug Disc. 2:38–51. Zhang, J., Ferguson, S. S., Barak, L. S., Aber, M. J., Giros, B., Lefkowitz, R. J., and Caron, M. G. (1997). Molecular mechanisms of G protein-coupled receptor signaling: Role of G proteincoupled receptor kinases and arrestins in receptor desensitization and resensitization. Receptors and Channels 5:193–199. Zhang, J., Barak, L. S., Anborgh, P. H., Laporte, S. A., Caron, M. G., and Ferguson, S. S. G. (1999). Cellular trafficking of G protein-coupled receptor/b-arrestin endocytic complexes. J. Biol. Chem. 274:10999–11006. Zhu, X. G., Hanover, J. A., Hager, G. L., and Cheng, S. Y. (1998). Hormone-induced translocation of thyroid hormone receptors in living cells visualized using a receptor green fluorescent protein chimera. J. Biol. Chem. 273:27058–27063.

389

17 REASSEMBLED GFP: DETECTING PROTEIN–PROTEIN INTERACTIONS AND PROTEIN EXPRESSION PATTERNS Thomas J. Magliery* Department of Molecular Biophysics & Biochemistry, Yale University, New Haven, CT

Lynne Regan Department of Molecular Biophysics & Biochemistry and Department of Chemistry, Yale University, New Haven, CT Genomic research has resulted in the identification of tens of thousands of putative proteins from all three domains of life in recent years, many of which have no clear function. Key clues to the function of these proteins come from identifying their binding partners and expression patterns. Therefore, it is now of critical importance to develop robust, highthroughput methods to address these issues (Zhu et al., 2003). Immunoprecipitation and related methods like TAP-TAG (Puig et al., 2001) require purification of the analyte protein, demand relatively strong interactions between protein partners, and are not amenable to library approaches. Fusions to Aequorea victoria GFP and its variants have been used to examine expression patterns (Chalfie et al., 1994) and protein interactions through fluorescence resonance energy transfer (FRET) (Miyawaki et al., 1997), but these methods are limited by the photophysical properties of GFP variants and the promoters available to drive expression, particularly in whole organisms. Several combinatorial screens based on the reassembly of dissected proteins have been introduced in recent years to determine the identity of protein ligands, beginning with the yeast two-hybrid screen. In the last five years, experiments have demonstrated that GFP and its variants can be dissected and reassembled to yield fluorescent products. GFP reassembly can be used to demonstrate and identify protein–protein interactions and protein expression patterns in cells and whole organisms. * Present address: Department of Chemistry and Department of Biochemistry, The Ohio State University, Columbus, OH Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

391

392

REASSEMBLED GFP: DETECTING PROTEIN–PROTEIN INTERACTIONS AND PROTEIN EXPRESSION PATTERNS

17.1

PROTEIN DISSECTION

Cleavage of a single peptide bond in a protein can result in protein unfolding and inactivation, because protein folding is typically a highly cooperative process. Despite the dramatic energetic consequences of covalent bond cleavage, some proteins remain folded even after being cleaved into two pieces, held together by noncovalent interactions. Moreover, a mixture of the polypeptides corresponding to the proteolytic fragments can sometimes result in reassembly of active protein (Richards, 1958; Anfinsen, 1973). This observation has been exploited to make useful biological tools. For example, independent expression of the alpha and omega fragments of b-galactosidase results in spontaneous assembly of the active enzyme, capable of turning over chromogenic galactosides such as X-gal, which is the basis of “blue/white” screening in plasmid subcloning (Ullmann et al., 1967). Fusion to ribonuclease S, arising from the noncovalent reassembly of the RNase S-peptide and S-protein resulting from cleavage of RNase A by subtilisin, can be used as an indicator of protein expression and folding using a fluorogenic substrate (Kelemen et al., 1999). The system is marketed as FRETWorks by Novagen (Madison, WI). In the cases of LacZ and RNase S, the “dissected” protein fragments spontaneously reassemble, resulting in active protein. Fields and Song (1989) realized that if a reporter protein could be split such that the dissected fragments did not spontaneously reassemble, the interaction of proteins fused to those fragments might drive the reassembly and function of the reporter. In the yeast two-hybrid (Y2H) screen, the DNA-binding and activation domains of the GAL4 transcription factor do not spontaneously reassemble, but if they are brought together by the interaction of “bait” and “prey” proteins fused to the two domains, the reassembled GAL4 can drive the transcription of a reporter like bgalactosidase. Y2H has been tremendously useful for identifying protein interaction partners (Uetz et al., 2000). However, the method has considerable limitations: It must be done in yeast, it requires nuclear importation and function (excluding whole protein classes, such as membrane-associated proteins), it does not demand a direct interaction (i.e., interactions through complexes are sometimes detected), and it can be confounded by proteins that activate transcription in the absence of a binding partner. Although Y2H allows detection of fairly weak interactions, it is hampered by abundant false positives. Several assays have been introduced to circumvent some of the problems with Y2H, including bacterial two-hybrid systems (Karimova et al., 1998; Joung et al., 2000; Hays et al., 2000) and functional interaction traps based on fusion to dissected fragments of ubiquitin (Johnsson and Varshavsky, 1994), b-galactosidase (Rossi et al., 1997), dihydrofolate reductase (Pelletier et al., 1998; Pelletier et al., 1999), or b-lactamase (Galarneau et al., 2002). The reassembly of these dissected proteins must be detected by the addition of a chromogenic, fluorogenic or chemiluminescent reagent, or through survival selection. Recently, our group introduced a screen for protein–protein interactions based on the reassembly of dissected fragments of GFP (Ghosh et al., 2000). GFP is especially attractive because no exogenous reagent must be added to detect the reassembled protein and because GFP is known to express and mature in virtually every cell type and subcellular structure in which it has been tested (as demonstrated in other chapters in this volume). Moreover, since most cells do not have significant fluorescence background at the GFP emission/excitation wavelengths, virtually all the signal can be attributed to the reassembled GFP, and subcellular localization can be directly visualized. The method has successfully been used by our group and others to detect protein–protein interactions in bacteria and eukaryotes, to identify unknown interaction partners, and to visualize cell type and subcellular protein localization in multicellular organisms.

DISSECTION AND REASSEMBLY OF GFP

Figure 17.1. Schematic of GFP dissection. (A) The original system used by Ghosh et al. (2000) split GFP at 157–158. The reassembled GFP, fused to antiparallel leucine zipper peptides (blue), is depicted with the N- and C-terminal fragments are colored green and red, respectively. (B) The dissection points discussed in the text are highlighted. Those in bold have been the most generally successful. Created with PyMOL (http://www.pymol.org) from PDB entries 1EMA and 1SER. See color insert.

17.2 17.2.1

DISSECTION AND REASSEMBLY OF GFP Variants and Topology

Our original implementation of the GFP fragment reassembly involved dissection of the sg100 GFP variant between residues 157 and 158 (Ghosh et al., 2000). Antiparallel leucine zipper peptides were fused to the C-terminus of GFP(1–157) and the N-terminus of GFP(158–238), an arrangement that was designed to allow interaction of the peptides in the reassembled complex (Fig. 17.1A). Coexpression of these peptide-fragment fusions results in GFP reassembly and cellular fluorescence. When the peptide-fragment fusions are expressed separately, or when one or both of the peptides are not fused, no reassembly occurs. Therefore, the leucine zipper peptide–peptide interaction is required for GFP reassembly.1 We and others typically call the GFP fragments “NGFP” and “CGFP,” but we will use the nomenclature GFP(1–157) and GFP(158–238) here to make it clear what spectral variant and dissection point we are referring to. See Fig. 17.1B for the dissection points discussed in this section. The reassembly reaction is not extremely sensitive to linker length. We have tested linkers of 4–15 amino acids between the GFP fragments and the zipper peptides that are compatible with reassembly (Magliery et al., 2005). Moreover, the linkers do not have to be the same length. Reassembly occurs when the linker between the GFP(158–238) and the peptide is 7 amino acids longer than the linker between the GFP(1–157) and peptide. The sg100 variant (G. J. Palm, personal communication) has the following mutations from the original gfp10 wild-type sequence: F64L, S65C, Q80R, Y151L, I167T, and 1

The plasmid vectors pET11a-link-NGFP and pMRBAD-link-CGFP, which permit facile cloning of analyte proteins and independent maintenance in E. coli, are available upon request. See the Regan Lab webpage (http://www.csb.yale.edu/people/regan/publications.html) for sequences and information, in addition to Wilson et al. (2004).

393

394

REASSEMBLED GFP: DETECTING PROTEIN–PROTEIN INTERACTIONS AND PROTEIN EXPRESSION PATTERNS

K238N. The actual construct used originally had a proline at position 151, but we later demonstrated that this had no effect on the fluorescence phenotype relative to leucine (Magliery et al., 2005). Using the same antiparallel leucine zipper fusions, Chalfie and coworkers found that CFP and YFP can also reassemble, and the CFP(158–238) can reassemble with GFP(1–157) and YFP(1–157) to yield fluorescent proteins with unique spectral properties (Zhang et al., 2004). Michnick and co-workers have used virtually the same dissection point (158–159) and topology in EGFP, an F64L, S65T GFP mutant codonoptimized for mammalian cells, adding 10-amino-acid (GGGGS)2 linkers between the fused proteins and the dissection point (Remy and Michnick, 2004a). Other points of dissection have also been explored. Tsien and co-workers showed that large insertions and circular permutation were possible in EYFP between residues 144– 145 (Baird et al., 1999). Based on this, Miyawaki and colleagues fused calmodulin to the N-terminus of EYFP(1–144) and fused the tightly interacting M13 peptide to the Nterminus of EYFP(145–238), and they observed reassembly in HeLa cells (Nagai et al., 2001). This is both a different point of dissection and a different fusion topology, especially notable because it is at the end of the b-barrel distal to the native termini. Kerppola and co-workers tested various dissection points in EYFP (S65G S72A T203Y) using yet another topology and parallel leucine zippers from Fos and Jun (Hu et al., 2002). Here, both zipper peptides were fused to the C-termini of the two YFP fragments with short 5to 7-amino-acid linkers between the fragments and the zipper peptides. Loops at both ends of the b-barrel, proximal and distal to the C-terminus, were selected for dissection (38–39, 101–102, 144–145, 154–155, 168–169, 172–173, and 192–193). The 154–155 dissection point, in the same loop used by our group, gave the best results. Hu and Kerppola (2003) later tested fragments derived from EYFP, EGFP, ECFP, and EBFP dissected at 154–155 and 172–173 using the same fusion topology as before to the C-termini of both fragments. Surprisingly, the GFP fragments split at 154–155 did not reassemble, but YFP(1–154) did reassemble with CFP(155–238) to give a reassembled protein with spectral characteristics distinct from YFP or CFP. At a second dissection point at the opposite end of the b-barrel (172–173), it was found that YFP(173–238) reassembled with GFP(1–172), CFP(1–172) and YFP(1–172), but that B/C/GFP(173–238) did not reassemble with any 1–172 fragment. Perhaps even more remarkably, the 1–172 fragments of GFP, CFP, and YFP reassembled with the 155–238 fragments of CFP and YFP despite the resulting duplication of a b-strand (Fig. 17.2). Moreover, BFP(1–172) and CFP(155–238) also reassembled, even though no reassembly occurred with B/C/ GFP(173–238). Since the extra “strand” is likely just an unstructured linker between the N-terminal fluorescent protein fragment and the zipper peptide, the mixed dissection-point data actually suggest that a longer linker between the peptides and fragments is beneficial with this topology. Umezawa and colleagues developed a different means of reassembling GFP. Specifically, EGFP fragments are covalently reassembled when fused interacting proteins drive the association and splicing of an intein (Ozawa et al., 2000). Originally, the N-terminal domain of the yeast VMA1 intein was sandwiched between EGFP(1–128) and one analyte protein, and the C-terminal intein domain was sandwiched between the second analyte protein and EGFP(129–238). The analyte proteins (originally calmodulin and M13 peptide) were separated from the intein fragments by 9- to 10-amino-acid linkers. To achieve efficient splicing, the EGFP sequence between I124 and I129 was altered from IEKKGI to IILKGC, resulting in weak cellular fluorescence with the CaM/M13 fusions. To improve the efficiency of the system, the Umezawa group later replaced the VMA1 intein with the smaller, more soluble, bacterial dnaE intein, and other EGFP dissection

DISSECTION AND REASSEMBLY OF GFP

Figure 17.2. Multicolor reassembly of fluorescent proteins. Reassembly of CFP(155–238) with (A) YFP(1–172), (B) GFP(1–172), (C) BFP(1–172), and (D) CFP(1–172) results in yellow, green, blue, and cyan cells. [Adapted from Hu and Kerppola (2003) with permission.] See color insert.

points were examined (Ozawa et al., 2001). Dissection at residues 144–145 or 224–225 gave poor fluorescence, but dissection at 157–158 was successful, with either K156Y Q157C mutations or a KFAEYC insertion after Q157. Recently, Cabantous et al. (2005) dissected the so-called “superfolder” variant of GFP in the last loop, at residue 214. Both the GFP(1–214) and GFP(214–230) fragments were optimized by directed evolution for enhanced fluorescence and solubility. These optimized GFP fragments spontaneously reassemble, resulting in chromophore maturation and cellular fluorescence upon their co-expression without the mediation of fused, interacting proteins. These GFP tags are therefore useful for detecting protein expression and solubility in cells or lysates, but they are not useful for detecting protein-protein interactions. Alteration of the unfused termini of GFP sometimes prevents the reassembly reaction. In our implementation in which the fusions are made at the point of dissection, we found that N-terminal hexahistidine tagging of the GFP(1–157) fragment is not detrimental, but that tagging the C-terminus of GFP(158–238) with a biotinylation sequence and His6-tag prevented reassembly. The Kerppola group fused interacting proteins at the C-termini of both fluorescent protein fragments and tagged both N-termini with His6. Because the N- and C-termini of GFP are close in space at the same “end” of the barrel as the 157–158 loop, the spatially near termini might affect the reassembly reaction. Further experimentation is needed to see if modification of the termini or even a random mutagenic approach, as Stemmer took to engineer GFPuv (Crameri et al., 1996), can improve the reassembly. Therefore, several GFP variants are amenable to dissection and reassembly, and several points of dissection are useful. The most generally useful point of dissection appears to be in a surface loop near residues 157–158, spatially near the N- and C-termini, regardless of fusion topology. The striking fact that different fusion topologies and a wide range of linker lengths lead to reassembly is useful, since it does not require extensive

395

396

REASSEMBLED GFP: DETECTING PROTEIN–PROTEIN INTERACTIONS AND PROTEIN EXPRESSION PATTERNS

optimization for each protein pair. It suggests that precise alignment of the GFP fragments is not necessary to nucleate reassembly and that the persistent interaction of the fused proteins is not necessary to maintain the reassembled complex.

17.2.2

Kinetics and Mechanism

For the reassembly of GFP to be useful as a screen for protein–protein interactions, three key mechanistic parameters need to be determined: the affinity required to drive reassembly, the time required to observe cellular fluorescence, and the reversibility of the reassembly reaction. We addressed the first of these, the minimum affinity required for reassembly, by screening a library of antiparallel leucine zipper mutants with different affinities (Magliery et al., 2005). The reassembly was found to be remarkably sensitive, resulting in fluorescent cells from peptide–peptide interactions with dissociation constants as weak as 500 mM to 1 mM. In addition to peptide–peptide interactions, protein–peptide interactions in the range of 500 mM result in weak cellular fluorescence, but clearly more than negative controls. Moreover, we found that peptides with different affinities for a given peptide-binding domain can be distinguished by the level of cellular fluorescence. In Umezawa’s intein-based implementation, association of the intein fragments is required for GFP reassembly, but no information is available about the affinity required for this interaction. In addition, the use of these polypeptides results in considerable background fluorescence even in the absence of fused interacting proteins [a feature that has been exploited to examine mitochondrial localization (vide infra)]. In contrast, no cellular fluorescence is detected for negative controls in our implementation of the screen, and little to no fluorescence results from the various unfused fragment pairs tested by Kerppola (Magliery et al., 2005; Hu and Kerppola, 2003). The acquisition of cellular fluorescence is quite slow. Intact GFP alone usually requires several hours to mature in the cell, with rate-limiting oxidation of the chromophore (Tsien, 1998). We see optimum cellular fluorescence in E. coli with fused interacting proteins after overnight growth at 30°C followed by 1–2 days at room temperature, or with about 3 days all at room temperature. Kerppola and co-workers observe cellular fluorescence in COS-1 cells 8 h after transfection, but they show data from cells 36–48 h after transfection (Hu et al., 2002). More recently, they report cell growth at 37°C for 24 h followed by 30°C for 0–24 h (Hu and Kerppola, 2003). Michnick also reports screening COS-1 cells by FACS 48 h after transfection (Remy and Michnick, 2004a). Chalfie and co-workers see fluorescence in embryos and newly hatched nematode larvae, which probably corresponds to less than 8 h of expression time at 20°C (Zhang et al., 2004). Kerppola and co-workers estimated the t1/2 for the reassembly reaction (i.e., folding into the GFP conformation after association of the fused proteins), which is a pseudo-first-order process, to be about 60 s (Hu et al., 2002). (Rate-limiting chromophore maturation followed reassembly, in the proposed kinetic scheme.) This analysis is complicated by the fact that the individual, fused GFP fragments are almost entirely insoluble. Thus, the rate was estimated from dilution out of 6 M guanidine, and it is difficult to know the initial concentration, since presumably most of the protein precipitated upon dilution. The improved, dnaE intein-mediated version of the system results in rapid acquisition of cellular fluorescence. Fluorescence can be detected in the lysate 4 h after IPTG induction, and bacteria are typically grown 12–16 h on agar before observation, presumably in part to achieve sufficient colony size (Ozawa et al., 2001). The GFP reassembly reaction is essentially irreversible in vitro (Magliery et al., 2005). We purified fluorescent, soluble, reassembled GFP complex His6 tagged at the N-terminus

DISSECTION AND REASSEMBLY OF GFP

of GFP(1–157) over NiNTA-agarose. The urea-induced denaturation of the complex is slow and irreversible, with an estimated t1/2 in aqueous buffer of almost 10 years (when one of the peptides is pre-cleaved from the purified complex with protease). However, Chalfie and co-workers suggest that the reassembled GFP may have a much shorter half-life in C. elegans than GFP itself (Zhang et al., 2004). The irreversibility of the reassembly reaction could explain why it is possible to detect such weak interactions, and presumably would allow detection of transient interactions as well. The intein-mediated formation of EGFP is also obviously irreversible. Some caution must be exercised in applying the reassembly of GFP to investigating protein–protein interactions. Association of the fused proteins (nucleation), the critical first step of reassembly, is affected by the solubility and expression level of the fused GFP fragments. Thus, GFP reassembly can distinguish the binding of cognate from noncognate peptide to a particular peptide-binding domain, but the absolute amount of fluorescence does not correspond to differences in protein-peptide affinity between different peptidebinding domains (Magliery et al., 2005). Presumably, this lack of correspondence is due, at least in part, to different expression levels and solubilities of the GFP fragment fusions. Interrogation of a library, where the expression and solubility properties of the fusions might vary considerably, must therefore especially be scrutinized for the possibility of false negatives (i.e., authentic interactions that do not lead to cellular fluorescence). The tolerance of the reassembly reaction to topology and linker length suggests that precise positioning of the GFP fragments is not required; the most important factor appears to be effective concentration. Moreover, the interaction of the fused analyte proteins is not necessary to maintain the complex, since fluorescence persists after cleavage of one of the fused proteins (Magliery et al., 2005). However, we have also shown by CD spectroscopy that the antiparallel leucine zippers on which we originally tested the screen do associate in the reassembled complex and increase its already considerable kinetic stability. This tolerance is critical for the function of the screen, since many false negatives would result if exact positioning were required for either reassembly or maintenance of the reassembled complex.

17.2.3

Scope: Proteins and Cells

A variety of different peptides and proteins have been used successfully with GFP reassembly. The artificial antiparallel leucine zipper used with our approach and the parallel leucine zipper from the bZIP proteins Fos and Jun used by Kerppola work very well (Ghosh et al., 2000; Hu et al., 2002; Magliery et al., 2005). Fusion of TPR peptide-binding domains (see below) to the N-terminus of GFP(158–238), as well as fusion of peptide ligands to the C-terminus of GFP(1–157), was also successful (Magliery et al., 2005). Due to the insolubility of the GFP fragments, we are more wary of applications in which proteins are fused to the C-termini of the fragments, since the proteins almost certainly have to re-fold to initiate reassembly. However, several successes have been reported, including (a) fulllength Fos and Jun, and domains of Rel family proteins NF-kB and IkBa, fused to the Ctermini of YFP fragments split at 154–155 (Hu et al., 2002) and (b) fusion of proteins from a cDNA library to the C-terminus of EGFP(1–158) against N-terminally fused PNK/Akt to EGFP(159–239) (Remy and Michnick, 2004a). The intein-mediated system has been used exclusively with calmodulin and its ligand M13 peptide, although in both fusion orientations (Ozawa et al., 2001). The reassembly reaction has been tested extensively in E. coli, but it also functions well in mammalian COS-1 cells, both cytoplasmically and in the nucleus. It has been used

397

REASSEMBLED GFP: DETECTING PROTEIN–PROTEIN INTERACTIONS AND PROTEIN EXPRESSION PATTERNS

successfully in mammalian NIH3T3, HeLa, and HEK293T cells. Chalfie and co-workers have used the system successfully in the nematode C. elegans, with localization to the specific sets of neurons, as well as with subcellular localization to nuclei and presynaptic vesicles. The intein-mediated reassembly has been tested in E. coli, and has been used slightly differently (see below) in mammalian BNL1ME mitochondria.

17.3

APPLICATIONS OF THE GFP REASSEMBLY

17.3.1

Understanding Protein–Protein Interactions

Identification and characterization of protein–protein interactions from libraries are the most obvious and perhaps ambitious applications of this system. Two key benefits of the screen relative to other two-hybrid methods are that the screen can be carried out, in principle, in any organism or subcellular compartment and that no exogenous reagent is necessary to detect the reporter. To date, the reassembly has been used to detect or discriminate protein–protein interactions in bacteria and mammalian cells. We have used GFP reassembly to explore the structural determinants of antiparallel leucine zipper formation. Coiled coils are generally stabilized by the burial of hydrophobic residues and the interaction of oppositely charged “edge” residues (O’Shea et al., 1993). What controls the orientation (parallel versus antiparallel) is less clear, but favorable edge interactions and buried hydrogen bonds can be used to favor the antiparallel orientation (Oakley and Hollenbeck, 2001). We constructed a library in which the eight edge positions on one peptide were randomized between Glu and Lys, which would formally vary from 0 to 8 charge–charge mismatches in the antiparallel orientation with a constant peptide (Magliery et al., 2005) (Fig. 17.3). Only peptides with three or fewer charge-charge mismatches passed the screen, and the positions of the mismatches were not equivalent. Specifically, mismatches near the ends of the zipper were less significant. Comparison of

E5 E12 E19 E26

(A) K27 K20 K13 K6

L24 L17 L10 L3

(B)

L2 N9 L16 L23

c' f'

c

Z-NGFP b'

Z-CGFP

f

L7 L14 N21 L28

b E4 E11 E18 E25

L27 L20 L13 L6

K24 K17 K10 K3

Number of clones (scaled to N=100)

398

40 Positive Negative Normal

30

20

10

0

1

2 3 4 5 6 Number of mismatches

7

8

Figure 17.3. A library of antiparallel leucine zipper interactions. (A) A helical wheel diagram of the library, in which the e and g positions in one peptide (boxed) were randomized between Lys and Glu. (B) Positive clones and three or fewer mutation, while negative have three or more. The distributions are shown in contrast to the distribution that would have resulted if the screen did not select for tight binders. [Adapted from Magliery et al. (2005).]

APPLICATIONS OF THE GFP REASSEMBLY

the screening data and in vitro biophysical data suggested that binding in the parallel orientation was not possible in our model system even with favorable charge pairings. Tetratricopeptide repeat (TPR) domains are composed of sequentially arrayed 34amino-acid motifs, typically in groups of three, that sometimes act as a peptide-binding interface (D’Andrea and Regan, 2003). Hsp-organizing protein (HOP) has three TPR domains (TPR1, TPR2A, and TPR2B), the first two of which are known to bind to the Ctermini of chaperones Hsc70 and Hsp90 (Brinker et al., 2002). (HOP-TPR2B has no known binding partner.) When challenged with C-terminal peptides from Hsc70 and Hsp90, as well as an unrelated leucine zipper peptide, the TPR domains could distinguish their related cognate from noncognate ligands and could discriminate both from the unrelated negative control (Magliery et al., 2005) (Fig. 17.4). This type of experiment may be a route to identifying the targets of TPR2B and other TPR domains without known binding partners, using either random or cDNA-based peptide libraries (see below). The use of multicolor fluorescent proteins makes it possible to examine multiple interactions in a cell at one time, or to compare the relative efficiency of dimerization with a given protein to multiple interaction partners. For example, the relative efficiencies of dimerization of the leucine zipper peptide from Jun with the peptides from Fos and ATF2 were compared (Hu and Kerppola, 2003). Coexpression of YFP(1–154)-bFos, CFP(1–173)-bATF2, and CFP(155–238)-bJun resulted in yellow fluorescence in the nucleoli, consistent with preferential bFos/bJun dimerization. More recently, the competition for in vivo dimerization of Myc/Mad/Max proteins was examined (Grinberg et al., 2004),

Figure 17.4. TPR domains and peptide ligands. (A) Schematic of interactions between the TPR domains of HOP and the chaperones Hsc70 and Hsp90. (B) Screening of the three TPR domains from HOP against peptides from the C-terminus of Hsc70 and Hsp90, and an unrelated peptide (Z). The cognate TPR-peptide interactions result in stronger fluorescence than the noncognate interactions. [From Magliery et al. (2005).]

399

400

REASSEMBLED GFP: DETECTING PROTEIN–PROTEIN INTERACTIONS AND PROTEIN EXPRESSION PATTERNS

and the combinatorial interaction of crystallin-regulating factors in lens development was studied (Rajaram and Kerppola, 2004).

17.3.2

Identifying Unknown Protein–Protein Interactions

Remy and Michnick searched for binding partners of the protein kinase PKB/Akt by challenging it with a human brain cDNA-derived library fused to an EGFP fragment (Remy and Michnick, 2004a). Screening was done by FACS in transiently transfected mammalian COS-1 cells. The positive clones, after two rounds of selection, contained a substantial number that did not produce readable sequence or were false positives for various reasons, but 22 out of 100 contained identifiable genes. One of these proteins, a human homolog of mouse Ft1, was shown to interact in a serum- and insulin-dependent manner, consistent with participation in the PI3K signal transduction pathway. The direct interaction of hFT1 and PKB was later established by immunoprecipitation, and a role in increasing the activity of PKB was established, probably involving apoptosis regulation (Remy and Michnick, 2004b). Despite the relatively high rate of false positives, along with the potential for false negatives from fusion-insolubility or refolding problems, this is an exciting demonstration of the general applicability of the screen. Problems may arise, both in the discovery of new ligands and in the characterization competing interactions, because unfused, endogenous proteins may compete with the GFP fragment-fused analyte protein. For example, quantification of Ca2+ concentration was not reliable with fusion of calmodulin and M13 peptide to dissected GFP fragments, because of competition with endogenous calmodulin (Nagai et al., 2001). Genetically tractable systems like E. coli, yeast, Drosophila, and C. elegans allow replacement of the gene of interest and testing for complementation by the fused protein, which may be a general solution to this problem. This technology should be readily extended to the discovery of inhibitors and activators of protein–protein interactions. Because the GFP reassembly reaction is essentially irreversible, one cannot expect to disrupt the interaction of proteins fused to GFP fragments by adding inhibitors. However, if inhibitors are added at the inception of cell growth, they might prevent the reassembly reaction altogether. This approach is likely to provide a very stringent screen for inhibitors. On the other hand, the irreversibility of the reassembly reaction makes it particularly suited to the discovery of interaction agonists, or so-called chemical inducers of dimerization (CIDs). Similarly, the ability to screen in eukaryotic cells raises the possibility of identifying interactions requiring posttranslational modifications, perhaps in an environment- or stimulus-dependent manner.

17.3.3

Detection of Subcellular Localization

Since GFP fluorophore maturation is an autocatalytic process dependent upon protein folding but not accessory proteins or environment (except that folding and oxidation must be possible), GFP reassembly can be used to detect the subcellular localization of interacting proteins. Specifically, this means that protein–protein interactions can be examined in their native context, away from potentially complicating interactions that arise from expression in other compartments (like the nucleus). For example, Kerppola demonstrated in mammalian COS-1 cells that the leucine zippers of Fos and Jun alone localize to the nucleoli, but that the full-length bZIP dimers localize to the nucleoplasmic region and are excluded from the nucleoli (Hu et al., 2002) (Fig. 17.5). Similarly, the ATF2/Jun dimer

APPLICATIONS OF THE GFP REASSEMBLY

Figure 17.5. Subcellular localization of protein–protein interactions. The leucine zipper domains from Fos and Jun localize to the nucleoli of COS-1 cells (A), while the full-length proteins localize to the nucleoplasm but are excluded from the nucleoli (B). ATF2 and Jun localize to the perinuclear region (C), but are translocated to the nucleus when p38 is overexpressed (D). [Adapted from Hu et al. (2002) with permission.]

localizes to the perinuclear region, but is translocated into the nucleus when p38 is overexpressed. Umezawa and co-workers have adapted their intein-based GFP reassembly for identification of mitochondrial proteins (Ozawa et al., 2003). Essentially, the two-component system is used as a logical “AND” gate (Zhang et al., 2004), such that cells are fluorescent if and only if two different proteins are both localized to the mitochondria. The Cterminal intein/EGFP fragment was fused to a mitochondrial targeting signal (MTS), and the N-terminal fragment was fused to proteins from a cDNA library. If the proteins in the cDNA library contain an MTS of their own, the two intein/EGFP fragments are colocalized to the mitochondria, resulting in splicing and EGFP folding and maturation. Seventy different clones were identified, including known mitochondrial proteins, known proteins with no previous information about subcellular localization, and proteins of unknown function. It should be noted that splicing occurs here because of background intein selfassociation. This limits the utility of this version of the screen in detecting protein–protein interactions, but it is perfectly usable for detecting expression patterns.

17.3.4

Visualizing Protein Expression Patterns

Chalfie and co-workers have exploited the combinatorial features of the reassembly of GFP to detect cellular and subcellular expression patterns in whole animals (C. elegans)

401

402

REASSEMBLED GFP: DETECTING PROTEIN–PROTEIN INTERACTIONS AND PROTEIN EXPRESSION PATTERNS

Figure 17.6. Cellular colocalization in whole organisms. (A) Punc-24gfp is expressed in many adult cells. (B) Punc-24nzgfp and Pmec-2czgfp are coexpressed only in six touch receptor neurons. (C) Pmec-3nzgfp and Pegl-44czgfp are coexpressed only in the two FLP neurons. [From Zhang et al. (2004) with permission.]

(Zhang et al., 2004). Specifically, by driving the expression of the two GFP-leucine zipper fusions from two separate promoters, it was possible to (1) establish gene coexpression in the same cells, (2) label cells that uniquely coexpress genes from both promoters, and (3) determine the expression pattern of one promoter when the expression of the other promoter is known. For example, there is no known promoter that is only active in the two FLP neurons of the nematode. However, mec-3 and egl-44, which are expressed in various cell types, are uniquely coexpressed in FLP neurons. Expression of the GFP fragments from these two promoters resulted in unique labeling of the FLP cells (Fig. 17.6). The combinatorial aspect of GFP reassembly frees the geneticist from the limitations of the expression patterns of available promoters. The authors also showed that fusion of one of the fragments to a nuclear localization signal or synaptobrevin, which localizes to presynaptic vesicles, could be used to label subcellular components only in cells in which both fragments’ promoters are active.

17.4

FUTURE DIRECTIONS

GFP reassembly can be used to identify protein binding partners in bacteria and mammalian cells, but the potential of the system has not been stretched by the early studies described here. In particular, the identification of interactions that requisitely occur in a particular cell type or subcellular structure should be possible. Similarly, the identification of interactions requiring cell-specific modifications is not difficult to imagine. Adaptation of the GFP reassembly to yeast, in which virtually every genetic knockout is available,

REFERENCES

may be particularly useful in this regard. Also, the irreversibility of the reassembly suggests that it may be useful in capturing transiently interacting proteins, although the rate of degradation may limit this approach in some kinds of cells. A major area of interest in developing small-molecule therapeutics is the identification of inhibitors and activators of protein–protein interactions. Although mechanistic complications arise in adapting the screen for this purpose, they are likely to be surmountable. Again, the finding that the reassembly can be carried out in mammalian cells or whole organisms is particularly attractive for discovering therapeutics. Finally, understanding the expression and coexpression patterns of proteins is a major advance for studying organismal development and cellular differentiation. Such localization may be useful, for example, in the precise characterization of cancer cell lines and will clearly be useful for researchers wishing to study particular a cell type during animal development. GFP has already proven useful for (a) understanding cellular and subcellular expression patterns with fusions of GFP to proteins of interest (Chalfie et al., 1994), (b) visualizing protein–protein interactions with fused FRET pairs of GFP variants (Miyawaki et al., 1997), and (c) establishing protein expression or solubility by direct fusion of GFP to the C-terminus of the analyte protein (Waldo et al., 1999). The use of reassembled GFP not only remedies some of the technical complications with these elegant methods, but also allows uses that are not possible with intact GFP.

ACKNOWLEDGMENTS T.J.M. is an N.I.H. Postdoctoral Fellow (GM065750). Work on the reassembled GFP-based protein interaction trap was supported in part by NIH grants GM62413 and GM57265 (L.R.).

REFERENCES Anfinsen, C. B. (1973). Principles that govern the folding of protein chains. Science 181:223–230. Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (1999). Circular permutation and receptor insertion within green fluorescent proteins. Proc. Natl. Acad. Sci. USA 96:11241–11246. Brinker, A., Scheufler, C., Von Der Mulbe, F., Fleckenstein, B., Herrmann, C., Jung, G., Moarefi, I., and Hartl, F. U. (2002). Ligand discrimination by TPR domains. Relevance and selectivity of EEVD-recognition in Hsp70 ¥ Hop ¥ Hsp90 complexes. J. Biol. Chem. 277:19265–19275. Cabantous, S., Terwilliger, T. C., and Waldo, G. S. (2005). Protein tagging and detection with engineered self-assembling fragments of green fluorescent protein. Nat. Biotechnol. 23:102–107. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263:802–805. Crameri, A., Whitehorn, E. A., Tate, E., and Stemmer, W. P. (1996). Improved green fluorescent protein by molecular evolution using DNA shuffling. Nat. Biotechnol. 14:315–319. D’Andrea, L., and Regan, L. (2003). TPR proteins: The versatile helix. Trends Biochem. Sci. 28:655–662. Fields, S., and Song, O. (1989). A novel genetic system to detect protein–protein interactions. Nature 340:245–246. Galarneau, A., Primeau, M., Trudeau, L. E., and Michnick, S. W. (2002). beta-Lactamase protein fragment complementation assays as in vivo and in vitro sensors of protein–protein interactions. Nat. Biotechnol. 20:619–622.

403

404

REASSEMBLED GFP: DETECTING PROTEIN–PROTEIN INTERACTIONS AND PROTEIN EXPRESSION PATTERNS

Ghosh, I., Hamilton, A. D., and Regan, L. (2000). Antiparallel leucine zipper-directed protein reassembly: Application to the green fluorescent protein. J. Am. Chem. Soc. 122:5658–5659. Grinberg, A. V., Hu, C. D., and Kerppola, T. K. (2004). Visualization of Myc/Max/Mad family dimers and the competition for dimerization in living cells. Mol. Cell. Biol. 24:4294–4308. Hays, L. B., Chen, Y. S., and Hu, J. C. (2000). Two-hybrid system for characterization of protein–protein interactions in E. coli. Biotechniques 29:288–296. Hu, C. D., and Kerppola, T. K. (2003). Simultaneous visualization of multiple protein interactions in living cells using multicolor fluorescence complementation analysis. Nat. Biotechnol. 21:539–545. Hu, C. D., Chinenov, Y., and Kerppola, T. K. (2002). Visualization of interactions among bZip and Rel family proteins in living cells using bimolecular fluorescence complementation. Mol. Cell 9:789–798. Johnsson, N., and Varshavsky, A. (1994). Split ubiquitin as a sensor of protein interactions in vivo. Proc. Natl. Acad. Sci. USA 91:10340–10344. Joung, J. K., Ramm, E. I., and Pabo, C. O. (2000). A bacterial two-hybrid selection system for studying protein–DNA and protein–protein interactions. Proc. Natl. Acad. Sci. USA 97:7382– 7387. Karimova, G., Pidoux, J., Ullmann, A., and Ladant, D. (1998). A bacterial two-hybrid system based on a reconstituted signal transduction pathway. Proc. Natl. Acad. Sci. USA 95:5752–5756. Kelemen, B. R., Klink, T. A., Behlke, M. A., Eubanks, S. R., Leland, P. A., and Raines, R. T. (1999). Hypersensitive substrate for ribonucleases. Nucleic Acids Res. 27:3696–3701. Magliery, T. J., Wilson, C. G. M., Pan, W., Mishler, D., Ghosh, I., Hamilton, A. D., and Regan, L. (2005). Detecting protein–protein interactions with a GFP-fragment reassembly trap: Scope and mechanism. J. Am. Chem. Soc. 127:146–157. Miyawaki, A., Llopis, J., Heim, R., McCaffery, J. M., Adams, J. A., Ikura, M., and Tsien, R. Y. (1997). Fluorescent indicators for Ca2+ based on green fluorescent proteins and calmodulin. Nature 388:882–887. Nagai, T., Sawano, A., Park, E. S., and Miyawaki, A. (2001). Circularly permuted green fluorescent proteins engineered to sense Ca2+. Proc. Natl. Acad. Sci. USA 98:3197–3202. Oakley, M. G., and Hollenbeck, J. J. (2001). The design of antiparallel coiled coils. Curr. Opin. Struct. Biol. 11:450–457. O’Shea, E. K., Lumb, K. J., and Kim, P. S. (1993). Peptide velcro—Design of a heterodimeric coiled coil. Curr. Biol. 3:658–667. Ozawa, T., Nogami, S., Sato, M., Ohya, Y., and Umezawa, Y. (2000). A fluorescent indicator for detecting protein–protein interactions in vivo based on protein splicing. Anal. Chem. 72:5151–5157. Ozawa, T., Takeuchi, T. M., Kaihara, A., Sato, M., and Umezawa, Y. (2001). Protein splicing-based reconstitution of split green fluorescent protein for monitoring protein–protein interactions in bacteria: Improved sensitivity and reduced screening time. Anal. Chem. 73:5866–5874. Ozawa, T., Sako, Y., Sato, M., Kitamura, T., and Umezawa, Y. (2003). A genetic approach to identifying mitochondrial proteins. Nat. Biotechnol. 21:287–293. Pelletier, J. N., Campbell-Valois, F. X., and Michnick, S. W. (1998). Oligomerization domaindirected reassembly of active dihydrofolate reductase from rationally designed fragments. Proc. Natl. Acad. Sci. USA 95:12141–12146. Pelletier, J. N., Arndt, K. M., Pluckthun, A., and Michnick, S. W. (1999). An in vivo library-versuslibrary selection of optimized protein-protein interactions. Nat. Biotechnol. 17:683–690. Puig, O., Caspary, F., Rigaut, G., Rutz, B., Bouveret, E., Bragado-Nilsson, E., Wilm, M., and Seraphin, B. (2001). The tandem affinity purification (TAP) method: A general procedure of protein complex purification. Methods 24:218–229.

REFERENCES

Rajaram, N., and Kerppola, T. K. (2004). Synergistic transcription activiation by Maf and Sox and their subnuclear localization are disrupted by a mutation in Maf that causes cataracts. Mol. Cell. Biol. 24:5694–5709. Remy, I., and Michnick, S. W. (2004a). A cDNA library functional screening strategy based on fluorescent protein complementation assays to identify novel components of signaling pathways. Methods 32:381–388. Remy, I., and Michnick, S. W. (2004b). Regulation of apoptosis by the Ft1 protein, a new modulator of protein kinase B/Akt. Mol. Cell. Biol. 24:1493–1504. Richards, F. M. (1958). On the enzymatic activity of subtilisin-modified ribonuclease. Proc. Natl. Acad. Sci. USA 44:162–166. Rossi, F., Charlton, C. A., and Blau, H. M. (1997). Monitoring protein–protein interactions in intact eukaryotic cells by beta-galactosidase complementation. Proc. Natl. Acad. Sci. USA 94:8405–8410. Tsien, R. Y. (1998). The green fluorescent protein. Annu. Rev. Biochem. 67:509–544. Uetz, P., Giot, L., Cagney, G., Mansfield, T. A., Judson, R. S., Knight, J. R., Lockshon, D., Narayan, V., Srinivasan, M., Pochart, P., Qureshi-Emili, A., Li, Y., Godwin, B., Conover, D., Kalbfleisch, T., Vijayadamodar, G., Yang, M., Johnston, M., Fields, S., and Rothberg, J. M. (2000) A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae. Nature 403:623–627. Ullmann, A., Jacob, F., and Monod, J. (1967). Characterization by in vitro complementation of a peptide corresponding to an operator-proximal segment of the beta-galactosidase structural gene of Escherichia coli. J. Mol. Biol. 24:339–343. Waldo, G. S., Standish, B. M., Berendzen, J., and Terwilliger, T. C. (1999). Rapid protein-folding assay using green fluorescent protein. Nat. Biotechnol. 17:691–695. Wilson, C. G. M., Magliery, T. J., and Regan, L. (2004) Detecting protein–protein interactions with GFP-fragment reassembly. Nat. Methods 1:255–262. Zhang, S., Ma, C., and Chalfie, M. (2004). Combinatorial marking of cells and organelles with reconstituted fluorescent proteins. Cell 119:137–144. Zhu, H., Bilgin, M., and Snyder, M. (2003). Proteomics. Annu. Rev. Biochem. 72:783–812.

405

METHODS AND PROTOCOLS Steven R. Kain Agileut Technologies, Palo Alto, CA

Since the seminal work by Prasher et al. (1992) and Chalfie et al. (1994) to clone and express A. victoria GFP, this magnificent protein has moved from the spotlight of journal covers and article titles to join the ranks of a “reagent” in the methods section of most published material. As an editor of this text and a pioneer in the field of GFP application development, I feel that the more humble position now held by GFP is both sad and momentous. On the one hand, we miss the glory days of GFP experimentation, when reports of the use of GFP in any new species of cell or organism were newsworthy. Then again, the fact that GFP is now peripheral to most studies means that this biological marker has not only arrived, but is firmly established as a means to investigate fundamental questions in science. At the time of the first edition of this text, researchers were scrambling to understand how to use GFP. Methodological questions arose such as: Which GFP should I use? Will it work in my organism? How do I express the protein? How do I detect the signal? These questions have largely been answered through the more than 10,000 papers published concerning the application of GFP and its variants. Therefore, we have elected to include from the first volume only the broadest information, methods, and protocols that should be relevant for all applications that employ these biological markers.

PROTOCOL I: EXPRESSION OF GREEN FLUORESCENT PROTEIN I.C

Toxicity Due to GFP Expression

There have been published reports that overexpression of GFP may be toxic or interfere with regeneration required to generate transgenic plants (Haseloff and Amos, 1995; Chiu et al., 1996), as well as unpublished reports of toxicity associated with the overexpression of wild-type GFP in bacteria. Greatly overexpressed EGFP in bacteria—for example, from pUC-based vectors—can cause slower growth rates and osmosensitivity (Valdivia, Cormack, and Falkow, personal communication). Some bacterial species appear to tolerate high levels of GFP better than others. For example, high levels of GFP are tolerated by Yersinia species (Valdivia and Falkow, 1996) but not by Salmonella or Anabaena species (Valdivia and Falkow, 1996) (Buikema and Haselkorn, unpublished results). Toxicity due to high levels of GFP may also explain the difficulty that some workers have

Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

407

408

METHODS AND PROTOCOLS

had in obtaining stably transfected mammalian cell lines. The problems that have been encountered with toxic effects associated with GFP overexpression may be a general problem associated with protein overexpression, rather than specifically due to GFP.

I.D Folding and Temperature Sensitivity The ability of GFP to absorb blue light and emit green light is believed to depend on the formation of a chromophore by cyclization and oxidation of S-Y-G (residues 65–67) (Heim et al., 1994; Cody et al., 1993). The time constant of chromophore formation has been reported to be about 2–4 h for wild type and 0.45 h for S65T mutant GFP (Heim et al., 1994, 1995). The rate-limiting formation of the chromophore may limit the ability to visualize GFP fluorescence in transfected cells or transgenic organisms until a defined time after its expression. For Drosophila, the time required for the appearance of GFP fluorescence has been reported to be 3–5 h following expression (Davis et al., 1995). Wild-type GFP has been reported to be sensitive to temperature when expressed in mammalian cells, producing a brighter fluorescent signal at 33°C compared to 37°C (Pines, 1995). Temperature sensitivity of the fluorescence of a wild-type GFP fusion protein has also been observed in studies utilizing a GFP–human glucocorticooid receptor (GFP–hGR) fusion construct. These studies showed efficient transactivation of the mouse mammary tumor virus promoter in the presence of dexamethasone at 30°C but not at 37°C (Ogawa et al., 1995), a result demonstrating that the activity of the GFP–fusion protein, as well as its fluorescence, is greatly reduced at higher temperature. These effects may be due to the folding or redox state of GFP in the cell. The studies also showed that cells fluorescing at 30°C continued to fluoresce for at least 48 h upon shifting to 37°C. The time course of GFP–hGR movement from the cytoplasm into the nucleus after induction could be determined by addition of hormone to cells grown at 30°C, followed by incubation for various time periods at 37°C. Studies in yeast have shown that wild-type GFP and a wild-type GFP fusion protein expressed in S. cerevisiae showed markedly reduced fluorescence when cells were grown at 30°C, and that fluorescent cells grown at lower temperature retained their fluorescence after a shift to higher temperature (Lim et al., 1995). These observations allowed the workers to monitor relocalization of a GFP–nucleoplasmin fusion protein in a temperature-sensitive mutant of the nucleoporin gene by first culturing cells at 23°C to allow the fusion protein to accumulate, then shifting to 35°C. These studies illustrate the usefulness of the temperature sensitivity of wild-type GFP given appropriate experimental design. Such temperature sensitivity has not been reported for mutant forms of GFP expressed in mammalian cells. It should be noted, however, that the autocatalytic folding process may be less efficient under certain conditions, and that optimized expression protocols are likely to continue to be developed over the next few years.

I.E

Purification of GFP

The original purification of Aequorea GFP from photogenic organs of the jellyfish (Morise et al., 1974) is described in Chapter 1. The cloning of the gfp gene (Prasher et al., 1992) has permitted expression of GFP in bacteria for biochemical and biphysical studies. Bacterially expressed GFP has been characterized in clarified induced cell lysates without further purification (Heim et al., 1994) or following purification on a Ni-affinity column of a His6-tagged GFP containing a 34-residue peptide with six contiguous His residues

PROTOCOL I: EXPRESSION OF GREEN FLUORESCENT PROTEIN

fused to the N-terminus of GFP (Inouye and Tsuji, 1994). A fusion of GFP to GST (glutathione S-transferase) has also been made and purified by glutathione affinity chromatography (Niswender et al., 1995). Purification of bacterially-expressed GFP without an affinity tag has also been achieved and is detailed in Protocol 1.

Protocol 1: Purification of Recombinant GFP from Bacteria The following protocol gives high yields of purified GFP from a high-expression strain derived from TU#58 (Chalfie et al., 1994). The yield of purified protein has been as much as 150 mg/liter. This protocol has also been used to purify native GFP directly from the jellyfish, A. victoria, and several recombinant wild-type or mutant GFPs. One example is the mutant GFP expressed from the pBAD/gfp construct (Crameri et al., 1996). The mutant gfp in this vector is under the tight control of the arabinose promoter/repressor, araBAD, and can be induced continuously with a final concentration of 0.2% L(+) arabinose (w/v) in LB. We have transformed the E. coli strain DH5a with pBAD/gfp and have grown 10 liter of LB under continuous induction at 28°C for 24 h before harvesting. Ultimately, we were able to purify 50 mg of mutant GFP from the 10 liters. The most readily available high GFP expressing construct is TU#60 (Chalfie et al., 1994) sold by CLONTECH as pGFP. The gfp gene in pGFP is fused in-frame to the lacZ initiation codon from pUC19 (which adds an additional 24 amino acids to the N-terminus of GFP) that allows for high expression from the lac promoter. The pGFP vector also contains the bla gene for ampicillin selection and is a high copy number plasmid. Suitable E. coli strains which can be used to produce the protein include DH5a, JM109, and TB1 (New England BioLabs, Inc.). The GFP purified according to this protocol is suitable for biochemical and biophysical experiments, including crystallization trials. Materials Bacterial strain derived from TU#58 [pET3a/gfp, BL21 (DE3)] This strain expresses GFP as a nonfusion protein under the control of the T7 F 10-s10 promoter fragment. The transcription of the gfp gene is directly controlled by the T7 RNA polymerase. For transcription of gfp, TU#58 must be maintained in E. coli cells lysogenic for the l phage derivative, DE3 (Studier et al., 1990). The lDE3 lysogen carries the T7 RNA polymerase gene driven by the IPTG-inducible lacUV5 promoter. The pET3a plasmid also contains an ampR gene for selection. Cells are maintained continuously on ampicillin selection plates and selected on the basis of fluorescence prior to large-scale culture. LB (Miller, 1972) 10 g NaCl 10 g Bacto tryptone (Difco) 5 g Bacto yeast extract (Difco) Add DW to 1 liter, autoclave. Add ampicillin to 37 mg/ml for cell culture. LB + ampicillin plates 1 liter LB 14 g Bacto agar (Difco) 10 g Lactose Add ampicillin to 37 mg/ml prior to pouring plates. This concentration of ampicillin is suitable for the copy number of the plasmid. For pGFP, the optimal ampicillin concentration is 60 mg/ml.

409

410

METHODS AND PROTOCOLS

0.5 M IPTG (isopropyl-b-D-thiogalactopyranoside) in DW Extraction buffer 25 mM Tris-HCl, pH 8.0 1 mM b-mercaptoethanol (Eastman Kodak Co.) 0.1 M Phenylmethylsulfonylfluoride (PMSF) (Sigma) in 2-Propanol Low ionic strength buffer 10 mM Tris-HCl, pH 8.0 10 mM EDTA 50 mg/ml protamine sulfate (Sigma) in DW The protamine sulfate will not be in solution at room temperature. Warm the bottle under hot tap water immediately before dispensing. Ammonium sulfate (solid) Tris base (solid) Octyl agarose column buffer 10 mM Tris-HCl, pH 8.0 10 mM EDTA 1.0 M ammonium sulfate Sepharose column buffer 5 mM Tris-HCl, pH 8.0 0.02% NaN3 DEAE column buffer 5 mM Tris-HCl, pH 8.0 0.02% NaN3 Special Equipment Chromatography columns Octyl agarose HIC (Hydrophobic Interaction Chromatography) column (Pharmacia Biotech), 2.5 ¥ 12 cm Pre-equilibrate the column with octyl agarose column buffer. Sepharose CL-6B (Pharmacia Biotech), 3 ¥ 95 cm Pre-equilibrate the column with sepharose column buffer. DEAE Sepharose Fast Flow (Pharmacia Biotech), 2.5 ¥ 17 cm Pre-equilibrate the column with DEAE column buffer. 1. Select a brightly fluorescent colony from an LB + ampicillin plate and use it to inoculate 50 ml of LB + ampicillin. Grow overnight at 37°C. Visually select the most intensely green fluorescent colony from an ampicillin plate by placing the plate on a hand-held long-wave UV lamp (lmax = 365 nm). 2. Inoculate a flask containing 1 liter of LB + ampicillin with the 50 ml overnight culture. Grow at 37°C to OD660 = 0.8, then add IPTG to a final concentration of 0.5 mM. Induce cells at 37°C for 12 h.

PROTOCOL I: EXPRESSION OF GREEN FLUORESCENT PROTEIN

3.

4.

5.

6.

7.

8.

9.

No significantly higher production of GFP is obtained by longer incubation. The ideal temperature of expression is 37°C, but reasonable expression (up to 100 mg/l of LB) is achieved at 28°C. In our initial work with the TU#58-derived pET3a construct we obtained lower yields of 1–3 mg/l in LB. We increased the yields to 10 mg/l of LB broth by growing the E. coli cells at 28°C. With improvements in our colony selection technique, we now have cells that are considerably more productive and we are now able to grow these cells at 37°C with excellent GFP yields. Collect the GFP-producing E. coli cells by gentle centrifugation at 1000¥ g for 15 minutes at 4°C. Pellets can be stored at this step by freezing after harvesting. Extract the GFP from induced cells by repeated cycles of freezing and thawing. Slowly freeze (60 min at -20°C) and slowly thaw (60 min at room temperature) the packed pellets through three cycles (Johnson and Hecht, 1994). Follow this by two to four cold buffer washes (20¥ pellet volume) with the extraction buffer. Collect the wash supernatant by centrifugation (10,000¥ g, 15 min) and add PMSF to a final concentration of 1 mM. The PMSF will help prevent proteolytic cleavage of the protease-susceptible C-terminal “tail” of GFP. Usually it is necessary to freeze pellets between buffer washes to release all the GFP. While the freeze–thaw process is slower than other methods such as sonication or lysozyme treatment, the freeze–thaw extracts are remarkably clean (low viscosity, low DNA content, and high GFP content—up to 10% of total soluble protein). More than 90% of the GFP can be released from the cells by this method. Treat clarified extracts in low ionic strength buffer at 0–4°C with protamine sulfate to remove residual nucleic acids. Generally 1 mg protamine sulfate per 100 OD260 units is sufficient to precipitate most of the DNA, but not the GFP. Remove the precipitate by centrifugation (5000¥ g, 5 min). Add the protamine sulfate dropwise while stirring rapidly, so as not to precipitate GFP in localized regions of the extract. It is advisable to test each batch of GFP by small-scale titration in microfuge tubes to avoid “overshooting” the titration. It is not easy to recover GFP that is inadvertently precipitated by protamine sulfate. Precipitate the protamine-treated and clarified extract with ammonium sulfate (100% of saturation, 697 g/l extract) at 0°C. Add approximately 10 g of solid Tris base per liter of extract during the precipitation step to maintain a pH near 7.0. The pH of unbuffered saturated ammonium sulfate is close to 5.5, dangerously close to the low end of the GFP pH stability range. Generally, the precipitation of GFP is rapid. Collect the precipitate by centrifugation (10,000¥ g, 30 min) within an hour of precipitation. Expect near quantitative recovery if the GFP concentration in the crude extract is 0.2 mg/ml. Dissolve the GFP-containing pellet in a minimal volume (just sufficient to dissolve the GFP) of octyl agarose column buffer containing 1 mM PMSF. Clarify the dissolved pellet by centrifugation (15,000¥ g, 20 min). Pelleted GFP, which appears yellow, will take on the familiar bright green color as it goes into solution and the entire suspension becomes clear. Load the clarified solution, at room temperature, onto an octyl agarose HIC column preequilibrated with column buffer. Elute the column stepwise, first with 250 ml of low ionic strength buffer +0.5 M ammonium sulfate, and then with 250 ml of low ionic strength buffer without ammonium sulfate. Complete elution of GFP requires 100–200 ml of the second buffer solution. The column is capable of binding more than 1 g of total protein and can be eluted free of GFP in less than 2 h. Hydrophobicity of octyl agarose columns varies greatly with the length and chemical

411

412

METHODS AND PROTOCOLS

nature of the spacer arm. We prefer a three-carbon spacer with ether linkage to the agarose beads. 10. Concentrate the GFP sample to a volume of 2–5 ml by ultrafiltration or ammonium sulfate precipitation. 11. Chromatograph at room temperature on a column of Sepharose CL-6B, preequilibrated in column buffer, at a flow rate of 1.4 ml/min. GFP elutes from this column at an apparent molecular weight of about 40 kDa, indicating significant dimerization. a. Alternate two-step purification Reversible dimerization can be used in an alternative purification scheme following gel filtration on Sepharose to achieve ~95% purity. Additional Materials Bio-Gel column buffer 0 mM Tris-HCl, pH 8.0 10 mM EDTA 1 M ammonium sulfate 0.02% NaN3 Additional Special Equipment Two columns of Bio-Gel P-100 medium resin (Bio-Rad), 10 ¥ 120 cm (~8 liters) and 3 ¥ 120 cm (~0.75 liter) 1. Run the larger Bio-Gel P-100 column at room temperature with a dilute GFP sample (0.2 mg/ml). Partial hydrophobic interaction in high salt causes GFP to elute at an apparent molecular weight of 21 kDa. 2. Then run the second smaller column at room temperature with a very concentrated GFP sample (20–100 mg/ml) in the Bio-Gel column buffer, with or without ammonium sulfate. GFP dimerizes at high-protein concentrations and elutes at an apparent molecular weight of 44 kDa. Nearly all contaminants that co-elute with GFP on the first column are removed on the second. 12. Final polishing to achieve >95% purity is on a DEAE Sepharose Fast Flow column at room temperature. Load the sample in the column buffer by gravity at a flow rate of 2–4 ml/min and elute with a 2.0-liter gradient of salt (0 to 0.5 M NaCl) in the same buffer. Gravity-driven flow rates of 5 ml/min can be achieved with excellent resolution, equaling that achieved with a 4-h-long shallow salt gradient on Pharmacia’s Mono Q FPLC column. In fact, minor isoforms of GFP that differ by one charged amino acid are quantitatively removed on DEAE Fast Flow. This column is capable of purifying up to 1 g of GFP. 13. Judge the purity of GFP by the ratio of the absorbance of the chromophore at its lmax to that of the aromatic region of the protein at 280 nm. Note that, in the chromophore absorption band, the wild-type recombinant GFP, the so-called “Stemmer” mutant [cycle 3 mut (F99S, M153T, V163A)] (Crameri et al., 1996), and native Aequorea GFP all fail to follow Beer’s law. For these three types of GFP, as protein concentration increases, the absorbance at 395 nm increases disproportionately while the absorbance at 475 nm decreases disproportionately. The suppression of the 475 nm shoulder upon dimerization (as occurs in cells overexpressing GFP) can be as great as five-fold, making these forms of GFP exceedingly poor absorbers of blue light (molar extinction coefficient 3000 for the dimer). Thus, for accu-

413

PROTOCOL II: SPECIMEN PREPARATION

TABLE A.1.3. GFP Absorption Ratios Useful in Establishing Purity Chromophore lmax

e

l1/l2

Numerical Ratio

27,600 12,000 27,000 12,000 56,000

395/280

1.25

Recombinant wild-type GFP S65T

395 470 397 475 489

397/280

1.25

489/280

2.25

Stemmer mutant cycle 3 mut: F99S, M153T, V163A P4: Y66H

397 475 382

27,000 12,000 25,000

397/280

1.25

382/280

1.23

Protein Native Aequorea GFP

rate quantitation, it is necessary to measure absorbance at low protein concentration (0.05–0.20 mg/ml) in a concentration range that obeys Beer’s law. Table A.I.3 shows a partial list of GFP forms and spectral characteristics that may be used to judge purity. Contributed by Daniel G. González and William W. Ward

PROTOCOL II: SPECIMEN PREPARATION II.B

Fixed Specimens

Visualization of GFP in fixed cells or tissues has been successfully achieved for many types of cells after fixation with formaldehyde. In some cases, fixation of cells in sodium azide, methanol, ethanol or glutaraldehyde has also proven successful for subsequent visualization of GFP (some fixatives with sodium azide have been problematic). Although glutaraldehyde is a superior fixative to formaldehyde, it is not commonly used in fluorescence microscopy because it causes autofluorescence. At high concentrations, glutaraldehyde will destroy GFP fluorescence, but at low concentrations (e.g., 0.025%), this problem can be partially avoided. Denaturants such as 1% sodium dodecyl sulfate (SDS) or 8 M urea at room temperature can also be used in fixation procedures with the preservation of GFP fluorescence, but if GFP is fully denatured, or treated with 1% hydrogen perooxide or sulfhydryl reagents (Inouye and Tsuji, 1994), fluorescence is irreversibly destroyed. At high protein concentrations (above 5–10 mg/ml) or in high-salt solutions, GFP has been reported to dimerize, resulting in a fourfold reduction in absorption. GFP fluorescence has been reported to be sensitive to some nail polishes used to seal coverslips to slides (Chalfie et al., 1994; Wang and Hazelrigg, 1994). Molten agarose, rubber cement, or VALAP [1 : 1 : 1 Vaseline (petroleum jelly) : lanolin : paraffin chips, heated until clear] is recommended as a substitute for sealing coverslips. Alternatively, specimens can be viewed without applying a sealant by pressing down firmly on the coverslip to remove any excess mounting media between the slide and coverslip that might cause slippage of the coverslip.

414

METHODS AND PROTOCOLS

The use of slides marked with grids can aid in relocating cells of interest. Gridded slides available for use contain divisions of several millimeters in a printed Teflon coating (Cel-line Associates, Carlson Scientific), numbered divisions of 200–500 mm (“England finders”; Klarfield Rulings, Inc.), or squares of 55 mm (Eppendorf North America). The following methods have been successfully used to prepare fixed cells for GFP visualization.

Protocol 1: Bacteria (E. coli, EGFPmut1-3) Materials E. coli expressing EGFPmut1-3 5 mM Sodium azide in PBS (phosphate-buffered saline) Microscope slides Coverslips 1. Resuspend bacterial samples in 5 mM sodium azide in PBS to fix cells. 2. Mount cells on a glass slide with a glass coverslip. Fluorescence due to soluble EGFP is maintained as long as the cell surface integrity is not compromised. Fixing in sodium azide in PBS has no obvious effect on fluorescence. The cells can be fixed with a variety of crosslinking agents such as paraformaldehyde (2% w/v) and formaldehyde (1% v/v), and will retain some of the fluorescence from soluble EGFP. The GFP fluorescence in the fixed cells is also stable to photobleaching. Detergents and permeabilizing fixatives like methanol will destroy fluorescence. The effect of fixation on the ability to visualize GFP is most pronounced when imaging low levels of EGFP expression. CONTRIBUTED by RAPHAEL H. VALDIVIA, BRENDAN P. CORMACK, and STANLEY FALKOW

Protocol 2: Yeast (S. cerevisiae, TUB4-GFP) Materials Yeast expressing Tub4-GFP Formaldehyde (37%) Methanol Acetone DAPI (4¢,6-diamidino-2-phenylindole) Microscope slides Coverslips

415

PROTOCOL II: SPECIMEN PREPARATION

1. Fix cells by adding formaldehyde to a final concentration of 3.7%. 2. Incubate at RT or 30°C 1–2 h. 3. Treat with cold methanol for 5 min, then with cold acetone for 30 s. This procedure was performed to help flatten the cells, making the spindles easier to visualize in a single focal plane, rather than specifically for the GFP visualization. 4. To visualize DNA, incubate in 1 mg/ml DAPI for 1 min. 5. Mount cells on a glass slide with a glass coverslip. The ability to visualize GFP after fixation with formaldehyde may not pertain to all GFP fusion proteins in yeast. Although Tub4-GFP fluorescence could be observed following formaldehyde fixation, a-tubulin-GFP fusion proteins do not fluoresce visibly after fixation. CONTRIBUTED by TIM STEARNS (MARSHALL et al., 1996)

Protocol 3: Yeast (S. pombe, p93dis1-GFP) Materials Yeast expressing p93dis1-GFP Methanol or 2.5% glutaraldehyde Microscope slides Coverslips 1. Treat cells with methanol for 8 min at -80°C or with 2.5% glutaraldehyde at 33°C for 1 h. 2. Mount cells on a glass slide with a glass coverslip. TAKEN from NABESHIMA et al. (1995)

Protocol 4: Yeast (Schizosaccharomyces pombe, GFP) Protocol for preparation of fixed yeast cells for FACS analysis Materials S. pombe cells expressing GFP from an episomal expression vector EMM medium Sterile DDW 100% Ethanol 50 mM Sodium citrate 1. Grow cells in EMM to a density of 107 cells/ml. 2. Harvest 20 ml of cells by pelleting. 3. Suspend cells in sterile DDW to wash. Pellet again.

416

METHODS AND PROTOCOLS

4. Resuspend cells in 6 ml sterile DDW. 5. Add 14 ml 100% ethanol. The fixed cells can be stored indefinitely at 4°C. 6. Prior to analysis, harvest cells by centrifugation. 7. Wash with sterile DDW. 8. Resuspend in 5 volumes of 50 mM sodium citrate. 9. Sonicate briefly. 10. Analyze by FACS. Data was acquired for 20,000 cells for each sample and analyzed by plotting fluorescence against forward scatter. TAKEN from ATKINS and IZANT (1995)

Protocol 5: Drosophila (D. melanogaster, GFP-Exu in egg chambers) Materials exuSco2/exuSco2; P[Cas,NGE]3/+ females PBS (phosphate buffered saline) Fixative (Theurkauf and Hawley, 1992) 8% Formaldehyde 100 mM Potassium cacodylate, pH 7.2 100 mM Sucrose 40 mM Potassium acetate 10 mM Sodium acetate 10 mM EGTA Strips of Whatman filter paper 50% (v/v) Glycerol in PBS Microscope slides Coverslips 1. 2. 3. 4. 5. 6. 7. 8.

Collect 0- to 1-day-old females and place in well-yeasted vials with males. Keep 2 days prior to use, so females are 2–3 days old. Anesthetize females with CO2. Dissect ovaries in PBS. Place in fixative for 10 min. Wash 3 ¥ 10 min in PBS. Tease ovarioles apart in a drop of PBS on a glass slide. Remove excess PBS with strips of Whatman filter paper, then cover with a drop of 50% glycerol in PBS. 9. Cover with a glass coverslip. CONTRIBUTED by TULLE HAZELRIGG (WANG and HAZELRIGG, 1994)

417

PROTOCOL II: SPECIMEN PREPARATION

Protocol 6: Mammalian Cells (HeLa cells, GFP chimeras) Materials PBS (phosphate buffered saline) Fixative 2% Formaldehyde or 0.025% glutaraldehyde in PBS Fluoromount (Southern Biotechnology) Microscope slides Coverslips 1. Place cells in fixative for 10 min at room temperature. 2. Rinse twice in PBS. 3. Mount cells on a glass slide in Fluoromount with a glass coverslip. a. Staining of fixed cells with antibodies Additional Materials Primary antibody solution Primary antibody in PBS 10% Bovine serum 0.5% Saponin 10% Bovine serum in PBS Secondary antibody solution Rhodamine-labeled secondary antibody in PBS 10% Bovine serum 0.5% Saponin 1. Carry out fixation and PBS wash steps as described above. 2. Incubate cells for 1 h at room temperature in primary antibody solution. 3. Wash cells three times over 30 min in 10% bovine serum in PBS to remove unbound antibody. 4. Incubate cells for 1 h at room temperature in secondary antibody solution. 5. Wash cells three times over 30 min in 10% bovine serum in PBS to remove unbound secondary antibody. 6. Rinse cells quickly in PBS without serum. 7. Mount cells on a glass slide in Fluoromount with a glass coverslip. Observe using rhodamine and fluorescein filters to determine the distribution of antibody and GFP. CONTRIBUTED by JENNIFER LIPPINCOTT-SCHWARTZ

418

METHODS AND PROTOCOLS

III.B.2 Photobleaching, Photoactivation, Photodamage, and pH Dependence of GFP. Photobleaching of GFP has been reported to be slow (Chalfie et al., 1994; Niswender et al., 1995), certainly much slower than fluorescein under similar conditions. For instance, continuous observation in a confocal microscope for 20 min only reduced GFP intensity to one-half its original value (Niswender et al., 1995). Although GFP is resistant to photobleaching, it has proven very useful in fluorescence photobleaching recovery experiments (described below and in Chapter 12) because of its ability as a fusion protein to be targeted to specific organelles. Some problems with photobleaching of GFP have been reported. For instance, photobleaching with 395–440 nm light is accelerated by some agents used to anesthestize C. elegans, such as 10 mM NaN3, and another anesthetic agent, phenoxypropanol, has been reported to quench GFP fluorescence (Chalfie et al., 1994). An alternative anesthetic agent for use in visualizing GFP in live C. elegans is described in Protocol II. Some mutant forms of GFP may also be more susceptible to photobleaching. For imaging in cultured mammalian cells, 10 mM Trolox has been added to live cells when visualizing Y66H, Y145F (P4-3) GFP to reduce the rapid photobleaching of this mutant GFP (Rizzuto et al., 1996). Contrary to early reports, the photobleaching rate of wild-type GFP is close to the same whether it is excited in the UV or the blue, but the decrease in UV-excited fluorescence appears more rapid because of photoisomerization (Cubitt et al., 1995). Wild-type GFP has two absorption peaks, 395 and 475 nm, which are thought to be related via rotation about a bond (isomerization) within the chromophore structure. The isomerization can be induced by irradiation of wild-type GFP with either 395- or 490-nm light and the kinetics of this photoinduced reaction have recently been measured (Chattoraj et al., 1996). The photoisomerization affects the brightness of wild-type GFP during visualization, but does not occur in either the S65T or F64L, S65T (EGFPmut1) mutants of GFP. The recent solving of the crystal structures of wild-type GFP and the S65T mutant (described in Chapter 4) shows that the chromophore is tilted within the protein structure differently in the two proteins. This difference in tilting angle is likely the origin of the photoisomerization observed for wild-type GFP. The photoisomerization can also be used to photoactivate wild-type GFP. That is, the fluorescence properties of GFP change after irradiation. When wild-type GFP is irradiated with either UV (~395 nm) or blue light (488 nm), photoisomerization occurs and causes an increase in the 475-nm peak and a decrease in the 395-nm peak (Cubitt et al., 1995). In this manner, UV preexposure can be used to increase the blue excitation brightness of wild-type GFP (Chalfie et al., 1994). A detailed discussion of the chromophore structure and its photoinduced reactions is presented in Chapters 4 and 5. Since the photobleaching of GFP is low, cellular photodamage arising from GFP can also be expected to be low. Only a few extended studies have been reported to date, but photodamage associated with using GFP has not been identified as a problem. Anecdotal evidence suggests that photodamage associated with GFP fusion proteins is less than that found with labeling with other fluorophores. For example, fluorescence photobleaching recovery has been used to examine the diffusional mobilities of Golgi-targeted GFP chimeras (Cole et al., 1996), and in neither approach used was there evidence of cellular photodamage or photoinduced crosslinking of the GFP chimeras. Multiple bleaches of the same spot did not affect the measured diffusion coefficient, and staining with antibodies after photobleaching revealed intact membrane structures. Moreover, the mobile fraction did not decrease with successive bleaches, which would be expected if incomplete recovery was due to photoinduced immobilization of a fraction of the labeled molecules. Evi-

REFERENCES

dence such as this suggests that GFP will be excellent for time-lapse and four-dimensional imaging. The resistance of GFP to photobleaching may be due to the protection of the chromophore by a tightly packed barrel of b sheets [referred to as a b-can structure by Yang et al. (1996)], as revealed by the crystal structures (Ormö et al., 1966; Yang et al., 1996). Still, the intense light typically used for excitation of fluorescence can generate free radicals that in turn can damage cellular proteins. Excitation with UV light can also cause crosslinking and breakage of DNA, producing further detrimental effects. As with any fluorescence microscopy study of living samples, it is important to monitor sample viability during and after the experiment. One measure of viability is the ability of irradiated cells to divide further with the same doubling or mitotic cycle time as unirradiated cells. During time-lapse confocal imaging of mitotic spindles in live Drosophila embryos visualized using the Ncd motor protein fused to wild-type or S65T mutant GFP, photodamage caused the spindles in the irradiated region of the embryo to become delayed relative to the unirradiated region, resulting in asynchronous divisions (Endow, unpublished results). The cause of this damage was not determined (it could be caused by the GFP or by interactions between the incident light and endogenous absorbers), but lowering the excitation intensity to a level that reduced or eliminated the photodamage still gave a strong GFP signal. Delayed mitotic divisions or cell arrest has also been observed in Dictyostelium (Maniak et al., 1995) and yeast (Kahana and Silver, unpublished results) upon overirradiation of cells. In general, addition of free radical scavengers (Mikhailov and Gundersen, 1995) or antioxidants to the medium may aid in the imaging of GFP in live specimens with bright light for long periods of time. Oxyrase (Oxyrase Inc. P.O. Box 1345, Mansfield, OH 44901) (0.3 U/ml) and ascorbic acid (0.1–1.0 mg/ml) are two antioxidants that have been shown to reduce photodamage when added to the medium of living cells. Where possible, increasing gfp gene dose, or the use of brighter GFP variants is advantageous in imaging live specimens, permitting greatly reduced levels of exposure to fluorescent or laser light. Finally, the brightness of some GFP mutants appears to be sensitive to pH, although detailed studies have not yet appeared in the literature. As an example, wild-type GFP shows relatively even brightness from pH 5 to 10 (Ward, 1981), while the S65T mutant is 2-fold brighter at pH 7 than at pH 6 (Patterson and Piston, unpublished results). A similar fall-off in brightness at lower pH is exhibited by the F64L, S65T (EGFPmut1) double mutant.

REFERENCES Atkins, D., and Izant, J. G. (1995). Expression and analysis of the green fluorescent protein gene in the fission yeast Schizosaccharomyces pombe. Curr. Genet. 28:585–588. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W., and Prasher, D. C. (1994). Green fluorescent protein as a marker for gene expression. Science 263:802–805. Chattoraj, M., King, B. A., Bublitz, G. U., and Boxer, S. G. (1996). Ultra-fast excited state dynamics in green fluorescent protein: multiple states and proton transfer. Proc. Natl. Acad. Sci. USA 93:8362–8367. Chiu, W.-L., Niwa, Y., Zeng, W., Hirano, T., Kobayashi, H., and Sheen, J. (1996). Engineered GFP as a vital reporter in plants. Curr. Biol. 6:325–330.

419

420

METHODS AND PROTOCOLS

Cody, C. W., Prasher, D. C., Westler, W. M., Prendergast, F. G., and Ward, W. W. (1993). Chemical structure of the hexapeptide chromophore of the Aequorea green-fluorescent protein. Biochemistry 32:1212–1218. Cole, N. B., Smith, C. L., Sciaky, N., Terasaki, M., Edidin, M., and Lippincott-Schwartz, J. (1996). Diffusional mobility of Golgi proteins in membranes of living cells. Science 273:797–801. Crameri, A., Whitehorn, E. A., Tate, E., and Stemmer, W. P. C. (1996). Improved green fluorescent protein by molecular evolution using DNA shuffling. Nature Biotech. 14:315–319. Cubitt, A. B., Heim, R., Adams, S. R., Boyd, A. E., Gross, L. A., and Tsien, R. Y. (1995). Understanding, improving and using green fluorescent proteins. Trends Biochem. Sci. 20:448–455. Davis, I., Girdham, C. H., and O’Farrell, P. H. (1995). A nuclear GFP that marks nuclei in living Drosophila embryos; maternal supply overcomes a delay in the appearance of zygotic fluorescence. Dev. Biol. 170:726–729. Haseloff, J., and Amos, B. (1995). GFP in plants. Trends Genet. 11:328–329. Heim, R., Prasher, D. C., and Tsien, R. Y. (1994). Wavelength mutations and posttranslational autoxidation of green fluorescent protein. Proc. Natl. Acad. Sci. USA 91:12501–12504. Heim, R., Cubitt, A. B., and Tsien, R. Y. (1995). Improved green fluorescence. Nature (London) 373:663–664. Heim, R., and Tsien, R. Y. (1996). Engineering green fluorescent protein for improved brightness, longer wavelengths and fluorescence resonance energy transfer. Curr. Biol. 6:178–182. Inouye, S., and Tsuji, F. I. (1994). Aequorea green fluorescent protein Expression of the gene and fluorescence characteristics of the recombinant protein. FEBS Lett. 341:277–280. Johnson, B. H., and Hecht, M. H. (1994). Recombinant proteins can be isolated from E. coli cells by repeated cycles of freezing and thawing. Bio Technol. 12:1357–1360. Lim, C. R., Kimata, Y., Nomaguchi, K., and Kohno, K. (1995). Thermosensitivity of green fluorescent protein fluorescence utilized to reveal novel nuclear-like compartments in a mutant nucleoporin NSP1. J. Biochem. 118:13–17. Maniak, M., Rauchenberger, R., Albrecht, R., Murphy, J., and Gerisch, G. (1995). Coronin involved in phagocytosis: dynamics of particle-induced relocalization visualized by a green fluorescent protein tag. Cell 83:915–924. Marshall, L. G., Jeng, R. L., Mulholland, J., and Stearns, T. (1996). Analysis of Tub4p, a yeast gtubulin-like protein: implications for microtubule-organizing center function. J. Cell Biol. 134:443–454. Mikhailov, V. S., and Gunderson, G. G. (1995). Centripetal Transport of microtubules in motile cells. Cell Motil. Cytoskeleton 32:173–186. Morise, H., Shimomura, O., Johnson, F. H., and Winant, J. (1974). Intermolecular energy transfer in the bioluminescent system of Aequorea. Biochemistry 13:2656–2662. Nabeshima, K., Kurooka, H., Takeuchi, M., Kinoshita, K., Nakaseko, Y., and Yanagida, M. (1995). p93dis1, which is required for sister chromatid separation, is a novel microtubule and spindle pole body-associating protein phosphorylated at the Cdc2 target sites. Genes Dev. 9:1572–1585. Niswender, K. D., Blackman, S. M., Rohde, L., Magnuson, M. A., and Piston, D. W. (1995). Quantitative imaging of green fluorescent protein in cultured cells: comparison of microscopic techniques, use in fusion proteins and detection limits. J. Microscopy 180:109–116. Ogawa, H., Inouye, S., Tsuji, F. I., Yasuda, K., and Umesono, K. (1995). Localization, trafficking, and temperature-dependence of the Aequorea green fluorescent protein in cultured vertebrate cells. Proc. Natl. Acad. Sci. USA 92:11899–11903. Ormö, M., Cubitt, A. B., Kallio, K., Gross, L. A., Tsien, R. Y., and Remington, S. J. (1966). Crystal structure of the Aequorea victoria green fluorescent protein. Science 273:1392–1395. Pines, J. (1995). GFP in mammalian cells. Trends Genet. 11:326–327. Prasher, D. C., Eckenrode, V. K., Ward, W. W., Prendergast, F. G., and Cormier, M. J. (1992). Primary structure of the Aequorea victoria green-fluorescent protein. Gene 111:229–233.

REFERENCES

Rizzuto, R., Brini, M., De Giorgi, F., Rossi, R., Heim, R., Tsien, R. Y., and Pozzan, T. (1996). Double labelling of subcellular structures with organelle-targeted GFP mutants in vivo. Curr. Biol. 6:183–188. Theurkauf, W. E., and Hawley, R. S. (1992). Meiotic spindle assembly in Drosophila females: behavior of nonexchange chromosomes and the effects of mutations in the nod kinesin-like protein. J. Cell Biol. 116:1167–1180. Valdivia, R. H., and Falkow, S. (1996). Bacterial genetics by flow cytometry: Rapid isolation of Salmonella typhimurium acid-inducible promoters by differential fluorescence induction. Mol. Microbiol. 22:367–378. Wang, S., and Hazelrigg, T. (1994). Implications for bcd mRNA localization from spatial distribution of exu protein in Drosophila oogenesis. Nature (London) 369:400–403. Ward, W. W. (1981). Properties of coelenterate green fluorescent proteins. In Bioluminescence and Chemiluminescence, DeLuca, M. A. and McElroy, W. D., Eds., Academic, San Diego, pp. 225–234. Yang, T.-T., Kain, S. R., Kitts, P., Kondepudi, A., Yang, M. M., and Youvan, D. C. (1996). Dual color microscopic imagery of cells expressing the green fluorescent protein and a red-shifted variant. Gene 173:19–23.

421

INDEX

Absorption/extinction cross-section, green fluorescent protein, 57–59 Absorption of chromophores, suppression, molecular mechanism for, 59–61 Absorption ratios, GFP purification protocols, 413 Acid-inducible gene isolation, S. typhimurium GFP, 172–173 Actin cycling, Drosophila GFP, FRAP analysis, 243 Action potentials, in O. geniculata, 20–21 Activation control, green fluorescent protein fluorophore, 76 Acumen Explorer, reef coral fluorescent protein expression and detection, 357 Adenosine triphosphate (ATP), firefly luminescence, 31–33 ADH1 promoter, S. cerevisiae GFP gene expression, 181–182 adh1+ promoter, S. pombe GFP gene expression, 183 Aequorea aequorea: biochemistry, 19 green fluorescent protein, 1–4 biological function, 41 chromophore structure, 53–54 denaturation/renaturation, 49–50 mutant structures, 73–74 organic solvents, 51–52 physical characteristics, 42–52 protease resistance in, 50–51 light organs in, 6–8 protein concentration, and absorption suppression, 59–61 Aequorea coerulescens, green fluorescent protein mutant structures, 112–113 Aequorea forskalea, green fluorescent protein in, 10–11

Aequorea victoria: GFP-like proteins, 140–141 green fluorescent protein in, 10–11, 41–42 monomer/dimer equilibrium, 57–59 mutant structures, 83–84, 112–113 stability of, 48–49 stability properties, 49 Aequorin: biological function, 41 calcium channel buffer, 6–8 crystalline structure, 23–25 discovery of, 3–4 jellyfish species, 9–11 physical characteristics, 41–52 Aggregation, GPF mutation modulation, 102–105 AmCyan1 mutant: basic properties, 341 flow cytometry, single-color analysis, 353 multicolor analysis, 349–351 Amino acids: DsRed/DsRed-Express mutants, 344 fluorophore formation, 72–73 GFP dissection and reassembly, 393–396 in green fluorescent protein, 41–48 green fluorescent protein mutant structures, 73–75 heterogeneous residues, 84–85 nucleic acid changes, 112 phenolate anion with stacked p-electron system, 92–93 plant GFP, 270–272 predetermined stretch, randomization, 86 wavelength altered ratios, 97–100 nonbioluminescent anthozoa GFP-like proteins, sequence comparisons, 124–126 protein absorption suppression, 60–61 ZaGreen1 mutant, substitutions, 341, 343

Green Fluorescent Protein: Properties, Applications, and Protocols, Second Edition Edited by Martin Chalfie and Steven R. Kain Copyright © 2006 John Wiley & Sons, Inc.

423

424

INDEX

Anabaena sp., heterocyst formation, 167–168 Ancient diversity, GFP-like proteins, 154–155 “Ancient duplication” scenario, GFP-like proteins, 151 Anemonia majano, reef coral fluorescent proteins, 340–341 Anemonia sulcata, reef coral fluorescent proteins, 340–341 Animal hosts, GFP-host interactions, 169 Anthozoa, GFP-like proteins: FP modification, mutagenesis, 129–130 homolog color diversity, 122–124 mutagenesis color transitions, 126–128 oligomeric state, 126 photoactivatable probes, 130–134 Dronpa, 134 DsRed “greening,” 130–132 fluorescent protein kindling, 132–134 UV-induced green-to-red photoconversion, 132 phylogenetic relationships, 151–152 research background, 121–122 sequence comparisons, 124–126 Antibody staining, C. elegans GFP pattern analysis, 204–205 Antiparallel beta strands, green fluorescent protein folding, 78–79 Apoaequorin, isolation, 25 Arabidopsis, green fluorescent proteins in: cell imaging, 274–276 cell marking, 280–281 cryptic intron removal, 262–263 gene expression, 263–264 modification, 274 gfp mRNA cryptic splicing in, 260–262 localization, 264–265 maturation, 269–272 spectral modification, 272–273 subcellular dynamics visualization, 276–279 subcellular targeting, 265–269 b-Arrestin recruitment assay, fluorescent proteins, 372–376 ars1+ sequence, S. pombe GFP gene expression, 182–183 Artifacts, C. elegans GFP, 219–220 AsRed2 mutant: flow cytometry, single-color analysis, 354–355 reef coral fluorescent proteins, 345 asulCP protein, photoconversion, 132 Autofluorescence techniques, yeast GFP analysis, 196 Autoinduction, bacterial luciferases, 26–28

Autonomously replicating, centromeric (ARS/CEN) plasmids: Candida albicans GFP gene expression, 183–184 S. cerevisiae GFP gene expression, 181–182 S. cerevisiae GFP gene expression, extrachromosomal constructs, 194–195 S. pombe GFP gene expression, 182–183 B. subtilis, GFP isolation, 165 spore formation, 165–167 BAC injection, transgenic fish GFP, 292 Bacteria: biochemistry, 18–19 GFP purification protocols, 409–413 green fluorescent protein in: complex environments, 168–171 development and cell biology, 165–168 cell division, E. coli, 168 heterocyst formation, Anabaena, 167–168 spore formation, B. subtilis, 165–167 ecology and behavior, 170–171 fixed specimen preparation, 414 future research issues, 174–175 genetic applications, 171–174 acid-inducible gene isolation, S. typhimurium, 172–173 flow cytometry, 172 macrophage-inducible gene isolation, S. typhimurium, 173–174 host interactions, 169–170 research background, 163–165 luminescence in, 25–28 Balancers, Drosophila GFP, 247 BD FACS analysis protocol, reef coral fluorescent protein expression and detection: fixed cells, 352 unfixed cells, 351–352 BD FACSVantage SE cell sorter, reef coral fluorescent protein expression and detection, 352–353 multicolor analysis, 356 BD Living Color fluorescent proteins, comparison table, 342 b-can structure: green fluorescent protein, 69–71 nonbioluminescent anthozoa GFP-like proteins, sequence comparisons, 124–126 stability, 48–49

425

INDEX

Bioinformatic studies, fluorescent protein applications, 380–383 Biological functions, of green fluorescent protein, 40–41 Bioluminescence resonance energy transfer (BRET), drug assays, fluorescent proteins, 367–371 target validation studies, 381–383 Bioluminescent coelentrates: evolutionary function, 15–19 GFP-like proteins, 140–141 green fluorescent protein sources, 4–8 phylogenetic organization, 15–17 BioRad MRC-600 microscope, plant GFP imaging, 278–279 Bioremediation, bacterial GFP ecology and behavior, 170–171 Biosensors: fluorescent protein drug screening, 378–380 GFP mutant structures, 106–108 Blue fluorescent protein (BFP): biosensor mutations, 107–108 drug assays, 367–371 imidazole derivation from chromophores, 93–94 isolation of, 8–9 in mammalian cells, 308–309 plant GFP modification, 273 yeast GFP spectral variants, 185–186 Blue-shifted emissions, bacterial luciferases, 27–28 Brightness properties, green fluorescent protein, 54 Caenorhabditis elegans: green fluorescent protein in: artifacts, 219–220 cell biology processes, 213 cell culturing and sorting, 215 cell fate markers, 212 cellular anatomy visualization, 208–212 CFP/YFP variants, 216–217 co-labeling of expression patterns, 213–215 electrophysiological recording, 213 expression pattern analysis, 204–205 injection marker experiments, 215–216 laser ablation, 213 mosaic analysis, 216 neuronal function and plasticity measurement, 212–213

PCR fusion, 217 protein localization, 206–208 protein tagging, 215 recombination cloning, 219 regulatory sequence decoding, 205–206 reporter gene constructs, 217–219 research background, 203–204 transgenic cell lines, 219 visualization of, 219 in vivo recombination, 219 target validation studies, fluorescent proteins, 382–383 Calcium channels: aequorin discovery, 3–4 fluorescent protein biosensors, 379–380 green fluorescent photocyte dispersion, 22–25 Candida albicans, green fluorescent protein in, 180 gene expression system, 183–184 Carboxypeptidases, in green fluorescent proteins, 45–48 Cargo proteins, mammalian GFPs, organelle structure and membrane trafficking, 322–324 Cauliflower mosaic virus (CaMV), S. pombe GFP gene expression, 183 cDNA protein sequence: GFP-like proteins, gene coding, 142–143 green fluorescent protein: mutant structures, 83–85 durable photoisomerization, 99–100 prokaryotes, 164–165 Cell adhesion assays, fluorescent proteins, 366 Cell-based diagnostics, green fluorescent protein research, 61–62 Cell biology: bacterial GFP, 165–168 animal host interactions, 169 cell division, E. coli, 168 heterocyst formation, Anabaena, 167–168 spore formation, B. subtilis, 165–167 Drosophila GFP, 232–243 chromosome structure, 234–235 fusion protein FRAP analysis, 242–243 meiosis and mitosis, 235–238 morphogen gradients, 239–240 neuronal function and connectivity, 238–239 organelle function and transport, 234 original studies, 232–233 tagged RNAs, 241–242 tissue morphogenesis, 233–234

426

INDEX

Cell biology (continued) transcription factors, 241 mammalian cell GFP, 316–318 transgenic GFP markers, 296–297 Cell culturing and sorting: C. elegans GFP, 215 reef coral fluorescent protein expression and detection: flow cytometry analysis, 351–352 fluorescence microscopy, 347–348 Cell fate marker, C. elegans GFP, 212 Cellular anatomy, C. elegans GFP visualization, 208–212 Cerulean mutants: indole derivation from chromophores, 93 in mammalian cells, 308–309 Chaotropes, green fluorescent protein and, 52 Chemical mutagens, random mutagenesis, green fluorescent protein, 85 Chicks, transgenic GFP in, 294 Chromophore structure: green fluorescent protein, 8–9 absorption suppression, molecular mechanisms, 59–61 brightness properties, 54 molar extinction coefficient, 55–56 mutations, 83–85 prokaryotes, 164–165 quantum yield measurements, 54–55 spectral mutants, 89–94 imidazol derivation, Y66H, 93–94 indole derivation from Y66W, 93 neutral phenol, 92 phenolate anion, 89, 92 phenolate anionstacked pi-electron system, 92–93 phenyl derivation, Y66F, 94 wild-type neutral phenol/anionic phenolate mixture, 89 spectroscopic analysis, 53–56 plant GFP, maturation mechanisms, 270–272 Chromoproteins (CPs), color diversity, anthoza GFP homologs, 122–124 Chromosome dynamics: C. elegans GFP visualization, 213 Drosophila GFP, 234–235 mammalian cell GFP, 321 2m circle sequences, S. cerevisiae GFP gene expression, 182 Circular permutations, GFP mutations, 96–100 Cis-regulatory logic, C. elegans GFP decoding, regulatory sequences, 206

Citrine YFP variants: in mammalian cells, 308–309 in transgenic vertebrates, 288–289 Cln2p protein, yeast GFP gene expression, 193–194 Clonal analysis: Drosophila GFP, 247–248 reef coral fluorescent protein expression and detection, 352–353 Cloning mechanisms, in transgenic fish GFP, 293 Cnidarian species: biochemistry, 18–19 GFP-like proteins, 139–140 deep-level relationships, 149–151 luminescence, 19–25 phylogenetic classification, 18 Codon optimization, green fluorescent protein mutant structures, nucleic acid changes, 112 Coelentrates. See also Nonbioluminescent anthozoa green fluorescent protein in, 41–42 photoproteins in, 19–25 Coelentrazine, cnidarian luminescent systems, 22–25 Coexpression marker, transgenic GFP as, 296 Co-labeling techniques, C. elegans GFP gene expression, 213–214 Color diversity: anthoza GFP homologs, 122–124 GFP-like proteins, 144–146 phylogenetic relationships, 151–153 nonbioluminescent anthozoa GFP-like proteins, mutagenesis, 126, 128 Colorless nonfluorescent GFP (acGFPL), cloning of, 112–113 Complex environments, bacterial GFP in, 168–171 Confocal microscopy: green fluorescent protein brightness, 54 nuclear translocation assays, fluorescent proteins, 365 transgenic vertebrate GFP, 289–290 Conjugated double bonds, color diversity, anthoza GFP homologs, 124 Conjugative DNA transfer, bacterial GFP ecology and behavior, 170–171 Connectivity structures, Drosophila GFP cell biology, 238–239 Copepods, GFP-like proteins, 139–140 deep-level relationships, 149–151 functions, 147

427

INDEX

Co-transfection, mammalian cell GFP, 318 Courtship patterns, firefly luminescence, 30–33 Covalent mutant alteration, fluorophore structures, 94–96 Cryptic introns, in plant GFP, removal, 262–263 Cryptic splice sites: Arabidopsis gfp mRNA, 260–262 green fluorescent protein mutant structures, 112 Ctenophores, luminescence in, 19–25 C-termainal domains: in dinoflagellate luciferase, 29–30 in firefly luciferase, 31–33 GFP dissection and reassembly, 397–398 GFP truncation and fusion constructs, 76–77 in green fluorescent proteins, 45–52 CUG codons, Candida albicans GFP gene expression, 183–184 CUP1 promoter, yeast GFP gene expression, 193–194 Cyan fluorescent proteins (CFPs): C. elegans variants, 216–217 Candida albicans GFP gene expression, 183–184 cellular anatomy visualization, 210–212 color diversity, anthoza GFP homologs, 122–124 mutagenesis, 128 dissection and reassembly, 394–396 drug assays, 367–371 indole derivation from chromophores, 93 in mammalian cells, 308–309 neuronal function and plasticity studies, 212–213 phylogenetic relationships, 151–152 protease assays, 368 S. cerevisiae GFP gene expression, 181–182 S. pombe GFP gene expression, 183 yeast GFP localization, 189–190 fluorescence resonance energy transfer, 192–193 Cyclization, green fluorescent protein structures: aggregation modulation, GFP mutations, 103–105 fluorophore, activation control, 76 mutant structures, 84–85 truncation and fusion constructs, 76–77 Cypridina luciferin, green fluorescent protein chromophore, 8–9 Cytomegalovirus promoter, reef coral fluorescent protein expression and detection, 346

Cytoskeleton structure, mammalian cell GFP analysis, 319–320 Cytosolic coat proteins, mammalian GFPs, organelle structure and membrane trafficking, 322–324 Decapentaplegic (Dpp) gradient, Drosophila GFP cell biology, 239–240 Deep-level relations, GFP phylogeny, 149–151 Denaturation, in green fluorescent protein, 49–50 Dendronephtya, phylogenetic relationships, 152–153 Detergents, green fluorescent protein and, 52 Deterring luminescence, GFP-like proteins, 141 Diethylaminoethyl (DEAE) particles, aequorin, 6–8 Differential fluorescence induction (DFI), bacterial GFP: acid-inducible gene isolation, S. typhimurium, 173 genetic applications, 172 Dimeric structure: aggregation modulation, GFP mutations, 102–105 green fluorescent protein: dimerization control, 77–78 monomer/dimer equilibrium, 57–59 site-directed mutations, 86 in green fluorescent protein, 42–52 Dinoflagellates: biochemistry, 18–19 bioluminescence properties, 28–30 Directed evolution, green fluorescent protein mutant structure, 87 Discosoma sp., reef coral fluorescent proteins, 340–341 DNA-binding proteins, yeast GFP localization, 190–191 DNA injection, transgenic mice GFP, 294–295 DNA localization, yeast GFP, 190–191 DNA shuffling, green fluorescent protein mutation, 86–87 DNA transfer, GFP spore formation, B. subtilis, 167 Donor/acceptor proteins, bioluminescence resonance energy transfer assays, 370–371 Dronpa protein, photoactivation of, 134 Drosophila, green fluorescent protein applications in: cell biology and development, 232–243

428

INDEX

Drosophila, green fluorescent protein applications in (continued) chromosome structure, 234–235 fusion protein FRAP analysis, 242–243 meiosis and mitosis, 235–238 morphogen gradients, 239–240 neuronal function and connectivity, 238–239 organelle function and transport, 234 original studies, 232–233 tagged RNAs, 241–242 tissue morphogenesis, 233–234 transcription factors, 241 clonal analysis, 247–249 embryonic balancers, 247 fixed specimen preparation, 416 gene expression, 228–231 detection timing and sensitivity, 229–231 transformation vectors, 228–229 transient expression, 229 gene regulation studies, 244–247 research background, 227–228 screens, 250–251 Drug screening assays, fluorescent proteins, 363–380 b-arrestin recruitment assay, 372–376 BRET protein-protein interaction, 369–371 cell adhesion assays, 366 dual colour spectroscopy assays, 371–372 fluorescence-activated cell sorting, 378 fluorescent protein biosensors, 378–380 fluorescent protein degradation, 372 FRET/BRET assays, 367–371 FRET protease assays, 368 FRET protein-protein interaction, 368–369 GPCR/GFP fusion assays, 376–377 nuclear translocation assays, 364–365 protein complementation assays, 377–378 reporter gene assays, 366–367 DsRed-Express mutant: flow cytometry, single-color analysis, 354–355 reef coral fluorescent proteins, 344 DsRed mutants: C. elegans FP variants, 217 cell adhesion assays, 366 color diversity, anthoza GFP homologs, 124 flow cytometry, single-color analysis, 354–355 Fluorescent timer (E5) protein, 345 “greening” of, 130–132 monomer structure, 345 mutagenic modification, 129–130

nonbioluminescent anthozoa GFP-like proteins: E5 mutant, 128 oligomeric state, 126–127 reef coral fluorescent proteins, 344 dual color analysis, 351 dual color spectroscopy assays, 371–372 S. cerevisiae GFP gene expression, 181–182 S. pombe GFP gene expression, 183 in transgenic vertebrates, 289 yeast GFP spectral variants, 185–186 Dual-channel imaging, plant GFP imaging, 275–276 Dual protein labeling, C. elegans FP variants, 216–217 Durable photoisomerization, GFP mutations, 96–100 Ecological applications, bacterial GFP, 170–171 Electrophysiological recording, C. elegans GFP cell identification, 213 Electroporation, transgenic chick GFP expression, 294 Emission spectra: Drosophila GFP detection, 230 green fluorescent proteins, 41–42 Energy levels, green fluorescent protein mutant structures, 100–101 Enhanced green fluorescent protein (EGFP): AmCyan1 mutant comparison, 341 dissection and reassembly, 394–396 dual color spectroscopy assays, 371–372 in mammalian cells, 307–309 reef coral fluorescent protein expression and detection, 347 in transgenic vertebrates, 288–289 Enhancer analysis, transgenic GFP for, 297–298 Enhancer trap screens: Drosophila GFP, 250–251 plant GFP marking, 280–281 Entactin, G2F domains, 147–149 Error-prone polymerase chain reaction, green fluorescent protein mutants, 85–86 Escherichia coli: cell division, bacterial GFP and, 168 firefly luciferase in, 31–33 green fluorescent protein expression in, 72 Evolution, bioluminescense and, 15–19 Excitation spectra: Drosophila GFP detection, 230

429

INDEX

green fluorescent proteins, 41–42 mutant ratios, 96–100 plant GFP modification, 272–273 Extrachromosomal (plasmid) GFP constructs, S. cerevisiae GFP gene expression, 194–195 Exuperantia protein, Drosophila GFP, 232–243 fbp1+ promoter, S. pombe GFP gene expression, 183 Fireflies: biochemistry, 18–19 bioluminescence in, 30–33 Fish, transgenic green fluorescent protein in, 290–293 BAC injection, 292 meganuclease, 292 plasmid injection, 290–291 somatic nuclear transfer, 293 transposons, 292–293 Fission yeast, S. pombe GFP gene expression, 182–183 Fixatives, green fluorescent protein and, 52 Fixed specimen preparation protocols, GFP specimens, 413–418 Flashing mechanism, firefly luminescence, 30–33 Flavin mononucleotide (FMN): in bacteria, 26–28 in yellow fluorescent protein, 27–28 Flow cytometry techniques: bacterial GFP: animal host interactions, 169 genetic applications, 172 mammalian cell GFP analysis, 319 reef coral fluorescent proteins, 351–353 single-color analysis, 353–355 AmCyan1, 353 DsRed2, DsRed-Express and AsRed2, 354–355 HcRed1, 355 ZsGreen1, 354 ZsYellow1, 354 three- and four-color analysis, 356 two color analysis, 355 Fluorescence activated cell sorting (FACS): bacterial GFP: acid-inducible gene isolation, S. typhimurium GFP, 173 genetic applications, 172 fluorescent protein drug screening, 378 green fluorescent protein: brightness properties, 54

mutant structures, 88 transgenic GFP markers, 297 yeast GFP gene expression, 194 Fluorescence correlation spectroscopy (FCS), mammalian cell GFP, 312–315 Fluorescence cross-correlation spectroscopy (FCCS), mammalian cell GFP, 315 Fluorescence emission: covalent mutation of fluorophore, 95–96 green fluorescent protein, 6–8 spectroscopic analysis, 54–55 Fluorescence in situ hybridization (FISH), yeast GFP, RNA localization, 191–192 Fluorescence lifetime imaging microscopy (FLIM), mammalian cell GFP, 314–315 Fluorescence localization after photobleaching (FLAP), mammalian cell GFPs, 312 Fluorescence loss in photobleaching (FLIP), mammalian cell GFPs, 312 Fluorescence microscopy: green fluorescent protein brightness, 54 mammalian cell green fluorescent proteins, 309–316 FRET and FCS, 312–315 photobleaching techniques, 311–312 time-lapse multispectral, and ratio imaging, 311 total internal reflection fluorescence and fluorescent speckle microscopy, 315–316 plant GFP imaging, 274–279 reef coral fluorescent protein expression and detection, 347–348 yeast GFP, 179–180 Fluorescence recovery after photobleaching (FRAP): C. elegans GFP visualization, 213 Drosophila GFP, 242–243 mammalian GFPs, 311–312 yeast GFP localization, 189–190 Fluorescence resonance energy transfer (FRET): drug assays, fluorescent proteins, 367–371 dual color spectroscopy assays, 371–372 target validation studies, 381–383 DsRed “greening,” 130–132 GFP mutations: aggregation modulation, 104–105

430

INDEX

Fluorescence resonance energy transfer (FRET) (continued) biosensors, 106–108 tandem concatenations, 110–111 mammalian cell GFP, 308–309 protein-protein interactions, 312–315 yeast GFP, 180, 192–193 Fluorescence speckle microscopy (FSM): mammalian cell GFP, 315–316 cytoskeleton analysis, 319–320 yeast GFP localization, 189–190 Fluorescent dissecting microscopy, transgenic vertebrate GFP, 289 Fluorescent lipid-associated reporters (FLAREs), yeast GFP lipid localization, 192 Fluorescent protein degradataion assays, 372 Fluorescent proteins: color diversity, anthoza GFP homologs, 122–124 kindling, 132–133 mutagenesis modification, 129–130 Fluorescent timer protein: DsRed mutant, 128 in mammalian cells, 308–309 reef coral fluorescent proteins, 345 Fluorophore properties: covalent mutation, 94–96 green fluorescent protein structure, 71–73 activation control, 76 mammalian GFPs, 324–325 Folding mechanisms: GFP expression protocols, 408 green fluorescent protein structure, 78–79 mutant improvements, 101–102 plant GFP, 270–272 Forespore-specific fluorescence, GFP spore formation, B. subtilis, 166–167 Frozen sectioning, transgenic vertebrate GFP, 289–290 FtsA protein, E. coli cell division, bacterial GFP and, 168 FtsZ protein, E. coli cell division, bacterial GFP and, 168 Functionality studies, yeast green fluorescent protein localization, 184–185 Fusion constructs: C. elegans GFP, creation of, 217–219 Drosophila GFP, cell biology and development, 232–243 chromosome structure, 234–235 fusion protein FRAP analysis, 242–243 meiosis and mitosis, 235–238

morphogen gradients, 239–240 neuronal function and connectivity, 238–239 organelle function and transport, 234 original studies, 232–233 tagged RNAs, 241–242 tissue morphogenesis, 233–234 transcription factors, 241 Drosophila GFP gene expression, 229 fluorescent protein degradataion assays, 372 GFP reassembly, 397 green fluorescent protein, 76–77 genetic applications, 172 GFP spore formation, B. subtilis, 166–167 reef coral fluorescent protein expression and detection, 346–347 yeast GFP, 186–194 integration of, 195 organelle structure, 187–188 protein localization, 188–190 b-galactosidase (bGAL): Drosophila GFP detection, 229–230 GFP vs., 204 GAL4 gene expression, plant GFP, 280–281 GAL1 promoter, yeast GFP gene expression, 193–194 GAL1-10 promoter, S. cerevisiae GFP gene expression, extrachromosomal constructs, 194–195 Gal4 system, Drosophila GFP gene expression, 228 GATA promoters, transgenic fish GFP, plasmid injection, 290–291 Gene batteries, C. elegans GFP as cell fate marker, 212 Gene expression: C. elegans GFP: co-labeling techniques, 213–215 pattern analysis, 204–205 Drosophila GFP, 228–231 detection timing and sensitivity, 229–231 transformation vectors, 228–229 transient expression, 229 mammalian cell GFP markers, 318 plant GFP, 263–264 modified genes, 274 reassembled GFP visualization, 401–402 reef coral fluorescent proteins (RCFPs), 346–357 flow cytometry, 351–353 single-color analysis, 353–355 three- and four-color analysis, 356

431

INDEX

two color analysis, 355 fluorescence microscopy techniques, 347–348 laser scanning, 357 multicolor analysis, 349–351 single-color analysis, 348–349 spectrophotemetry, 357 in transgenic vertebrates: levels of, 287 markers for patterns, 297 yeast green fluorescent protein, 180–186 assessment of, 195–196 C. albicans, 183–184 functional and proper localization, 184–185 genetic applications, 193–194 S. cerevisiae, 181–182, 194–195 S. pombe, 182–183 spectral variants and red fluorescent proteins, 185–186 Gene regulation, Drosophila GFP reporters, 244–247 Gene targeting, mammalian cell GFP, 318–319 Genetic applications: bacterial green fluorescent protein, 171–174 acid-inducible gene isolation, S. typhimurium, 172–173 flow cytometry, 172 macrophage-inducible gene isolation, S. typhimurium, 173–174 yeast GFP gene expression, 193–194 Genome sequencing studies: b-arrestin recruitment assay, 375–376 fluorescent protein applications, 380–383 G2F domains, functions, 147–149 GFP400, neutral phenol composition, 92 GFPA mutant, plant GFP, 270–273 gfp 10 gene, green fluorescent protein brightness, 54 gfp genes, prokaryotic GFP, 164–165 GFP-like proteins: classification, 139–140 functions: bioluminescent organisms, 140–141 copepoda species, 147 G2F domains, 147–149 nonbioluminescent organisms, 141–147 anthozoa: FP modification, mutagenesis, 129–130 homolog color diversity, 122–124 mutagenesis color transitions, 126–128 oligomeric state, 126 photoactivatable probes, 130–134

Dronpa, 134 DsRed “greening,” 130–132 fluorescent protein kindling, 132–134 UV-induced green-to-red photoconversion, 132 research background, 121–122 sequence comparisons, 124–126 corals, gene encoding, 142–143 hydroid medusae, 141–142 oligomerization, 146–147 photoprotection hypothesis, 143–144 phylogeny, 149–155 deep-level relationships, 149–151 ancient diversity, 154–155 anthoza proteins, 151 color diversity, 151–154 GFP5 mutant, plant GFP modification, 272–273 b-Glucuronidase (guaA) gene, plant GFP, 259–260 Glucocorticoid receptor (GR), nuclear translocation assays, fluorescent proteins, 364–365 Glutamate receptors, C. elegans GFP protein tagging, 215 Golgi apparatus, mammalian GFPs, 323–324 GPCR kinases (GRK), protein-protein interaction, bioluminescence resonance energy transfer assays, 369–371 G-protein-coupled receptors (GPCRs): b-arrestin recruitment assay, 372–376 fluorescent protein biosensors, 379–380 GFP fusion assays, 376–377 protein-protein interaction, bioluminescence resonance energy transfer assays, 369–371 target validation studies, 380–383 Green fluorescent protein (GFP). See also GFP-like proteins absorption/extinction cross-section, 57–59 monomer/dimer equilibrium, 57–59 barrel structure, 20 b-can structure, 69–71 biological function, 40–41 future research issues, 61–62 chromophore absorption suppression, molecular mechanism, 59–61 chromophore structure, 8–9 brightness properties, 54 molar extinction coefficient, 55–56 quantum yield, 54–55 spectroscopic analysis, 53–56

432

INDEX

Green fluorescent protein (GFP) (continued) in cnidarians and ctenophores, 19–25 color diversity, anthoza GFP homologs, 122–124 dimerization control, 77–78 discovery of, 1–4 dissection and reassembly mechanisms: kinetics and mechanism, 396–397 proteins and cells, 397–398 variants and topology, 393–396 isolation and properties, 4–8 jellyfish species, 9–11 molecular biology and mutation: aggregation modulation, 102–105 biosensors, 106–108 covalent alteration, Y66FHW, 94–96 energy level modification, 100–101 excitation peak ratios, durable photoisomerizatai, 96–100 halides, 109–110 insertion of other proteins, 110 mutational strategies, 85–87 nucleic acid changes, 112 nucleotide sequencing, 83–85 pH levels, 108–109 screening methods, 87–88 silent and loss-of-function mutations, 111–112 spectral mutant classification by chromophore, 89–94 imidazol derivation, Y66H, 93–94 indole derivation from Y66W, 93 neutral phenol, 92 phenolate anion, 89, 92 phenolate anionstacked pi-electron system, 92–93 phenyl derivation, Y66F, 94 wild-type neutral phenol/anionic phenolate mixture, 89 tandem concentrations, 110–111 temperature-dependent folding, 101–102 truncations, 105–106 natural sources, 41 pharmaceutical applications: drug screening assays, 363–380 b-arrestin recruitment assay, 372–376 BRET protein-protein interaction, 369–371 cell adhesion assays, 366 dual colour spectroscopy assays, 371–372 fluorescence-activated cell sorting, 378 fluorescent protein biosensors, 378–380

fluorescent protein degradation, 372 FRET/BRET assays, 367–371 FRET protease assays, 368 FRET protein-protein interaction, 368–369 GPCR/GFP fusion assays, 376–377 nuclear translocation assays, 364–365 protein complementation assays, 377–378 reporter gene assays, 366–367 high-content/high-throughput screening, 362–363 high content screening detection instrumentation, 380 target validation studies, 380–383 physical characteristics of, 41–52 denaturation and renaturation, 49–50 detergents and chaotropes, 52 fixatives and preservatives, 52 organic solvents, 51–52 protease effects, 50–51 stability, 48–49 three-dimensional structure: activation control, 76 Aequorea mutant, 73–75 fluorophore properties, 71–73 folding mechanisms, 78–79 future research issues, 79 research background, 67–68 spectral and physical properties, 68–69 structure-based engineering, 76 truncation and fusion constructs, 76–77 “Greening” of DsRed, 130–132 Halides, green fluorescent protein mutant structures, 109–110 HcRed mutant: color transition to, 128 reef coral fluorescent proteins, 345–346 three-color analysis, 351 stable cell populations and clone expression, 352–353 Hermes element, Drosophila GFP gene expression, 228–229 Heteractis crispa: HcRed1 mutant, 345–346 reef coral fluorescent proteins, 340–341 ZaGreen1 mutant, 344 Heterocyst formation, Anabaena GFP, 167–168 Heterodimeric proteins, bacterial luciferases, 26–28 Heteroduplex recombination, green fluorescent protein mutant structure, 87

433

INDEX

hetR gene expression, heterocyst formation in Anabaena sp., 167–168 High-content screening (HCS): b-arrestin recruitment assay, 373–376 detection instrumentation, 380 GFP/RCFP pharmaceutical applications, 362–363 drug screening assays, 363–364 GPCR/GFP fusion assays, 376–377 High-throughput microscopy (HTM), green fluorescent protein mutation, 88 High-throughput screening (HTS): b-arrestin recruitment assay, 373–376 fluorescent protein biosensors, 378–380 GFP/RCFP pharmaceutical applications, 362–363 drug screening assays, 363–364 Homologous recombination: Candida albicans GFP gene expression, 184 S. pombe GFP gene expression, 182–183 Homo-oligomer formation, reef coral fluorescent proteins, 340–341 Human immunodeficiency virus (HIV-1), mammalian cell GFP and pathogenesis, 319 Hydrogen bonding, green fluorescent mutants, 75 Hydroid medusa, GFP-like proteins, 141–142 Hydrophobic interactions, protein absorption suppression, 59–61 Imidazole, chromophore derivation, 93–94 Immunity, Drosophila GFP reporters, 245–247 Indole, cyan fluorescent proteins, chromophore derivation, 93 Injection markers, C. elegans GFP experiments, 215–216 Insertion mechanisms, green fluorescent protein mutations, 110 In situ hybridization, transgenic GFP expression patterns, 297 Intein fragments, GFP reassembly, 396–397 subcellular localization, 401 Internal ribosome entry sequence (IRES), transgenic chick GFP expression, 294 Inverse fluorescence recovery after photobleaching (iFRAP), mammalian cell GFPs, 312 In vivo imaging: bacterial GFP applications, 174 Candida albicans GFP gene expression, 183–184

transgenic GFP for, 298 yeast GFP, 180 fusion constructs, 186–194 spectral variants and red fluorescent proteins, 185–186 In vivo recombination, C. elegans fusion gene constructs, 219 Isoform structure, in green fluorescent proteins, 45–52 Jellyfish species, aequorin and GFP sources, 9–11 “Kaede” protein, ultraviolet-induced green-tored photoconversion, 132 Kindling fluorescent proteins (KFP), photoactivation, 132–133 Kinetic parameters, GFP reassembly, 396–397 Knock-in gene expression, transgenic mice GFP, 295–296 Kruppel gene, Drosophila GFP balancers, 247 Laser ablation, C. elegans GFP cell identification, 213 Laser scanning confocal microscopy, plant GFP imaging, 275–279 Laser-scanning detection, reef coral fluorescent protein expression and detection, 357 Leptoseris fragilis, color diversity, 145–146 Leucine zipper peptides, GFP dissection and reassembly, 393–398 protein-protein interactions, 398–400 Lipids, yeast GFP localization, 192 Localization studies: plant GFP, 264–265 protein localization: absorption suppression and, 59–61 C. elegans GFP determination, 206–208 mammalian cell GFP, 316–318 yeast GFP, 188–190 yeast green fluorescent protein, 188–192 DNA, 190–191 lipids, 192 protein, 188–190 RNA, 191–192 Loss-of-function muations, green fluorescent proteins, 111–112 Luciferases: in bacteria, 26–28 crystalline structure, 23–25 in dinoflagellates, 28–30 in fireflies, 30–33

434

INDEX

Luciferin binding protein (LBP), in dinoflagellates, 28–30 Luciferins, biochemistry, 18–19 Lumazine protein (LUMP), blue-shifted emission, 27–28 Luminescent potentials, green fluorescent protein species, 21–25 Luminous systems, biochemistry, 18–19 lux I gene, bacterial luciferases, 26–28 Lysosomes, mammalian GFPs, 323–324 Macromolecular localization, yeast GFP, 186–187 Macrophage-inducible gene isolation, S. typhimurium GFP, 173–174 Mammalian cells: green fluorescent protein in: applications, 316–319 co-transfection or expression markers, 318 protein localization, dynamics, and concentration, 316–318 viral infection and pathogenesis, 319 viral system targeting, 318–319 characteristics, 305–309 cytoskeleton illumination, 319–320 fixed specimen preparation, 417 flow cytometry, 319 fluorescence microscopy-based techniques, 309–316 FRET and FCS, 312–315 photobleaching techniques, 311–312 time-lapse multispectral, and ratio imaging, 311 total internal reflection fluorescence and fluorescent speckle microscopy, 315–316 future research issues, 324–325 membrane trafficking and organelle dynamics, 321–324 nucleus revelation, 320–321 reef coral fluorescent proteins, fluorescence microscopy/flow cytometry applications, 339–357 Maturation mechanisms, plant GFP, 269–272 Meganuclease, in transgenic fish GFP, 292 Meiosis, Drosophila GFP cell biology, 235–238 Membrane trafficking, mammalian GFPs, 321–324 Messenger RNA (mRNA): Arabidopsis gfp, cryptic splicing, 260–262

dinoflagellate luciferases, 29–30 Drosophila GFP, tagged RNA, 241–242 yeast GFP RNA localization, 192 MET3 promoter, S. cerevisiae GFP gene expression, 182 mfgp4-ER gene, plant GFP: subcellular targeting, 266–269 visualization techniques, 276–279 mfgp4 gene, plant GFP: expression, 263–264 localization, 264–265 maturation mechanisms, 270–272 subcellular targeting, 265–269 Mice, transgenic GFP in, 294–296 target validation studies, fluorescent proteins, 382–383 Micro-RNA (miRNA) analysis, Drosophila GFP reporters, 245–247 Microscopic techniques, yeast GFP analysis, 196 Microtubule dynamics, yeast GFP localization, 189–190 Mini-white gene, Drosophila GFP gene expression, 228–229 Mitochondrial GFP: Drosophila GFP organelles, 234 organelle structure, 187–188 Mitosis, Drosophila GFP cell biology, 235–238 Mitotic spindle analysis, yeast GFP localization, 188–190 Moesin fusion, Drosophila GFP, tissue morphogenesis, 233–234 Molar extinction coefficient, green fluorescent protein, 55–56 Molecular function, GFP spore formation, B. subtilis, 166–167 Molecular weight analysis: fluorescent proteins, mutagenic modification, 129–130 green fluorescent proteins, 45, 47 Monomer/dimer equilibrium, green fluorescent protein absorption/extinction crosssection, 57–59 Montastraea cavernosa: GFP-like proteins, gene coding, 142–144 phylogenetic relationships, 151–152 Morphogen gradients, Drosophila GFP cell biology, 239–240 Mosaic analysis, C. elegans GFP experiments, 216 Mosaic analysis with a repressible cell marker (MARCM), Drosophila GFP, 248–249

435

INDEX

Mos1 gene, Drosophila GFP gene expression, 228–229 mRFP1 protein, mutagenic modification, 129–130 mtrA gene expression, S. typhimurium GFP gene expression, 174 Multicolor analysis: GFP dissection and reassembly, 394–396 protein-protein interactions, 399–400 reef coral fluorescent proteins, 349–351 flow cytometry, 356 Multiphoton microscopy, transgenic vertebrate GFP, 289–290 Multispectral imaging, mammalian cell green fluorescent proteins, 311 Mushroom bodies (MB), Drosophila GFP cell biology, 238–239 Mutagenesis: fluorescent protein modification, 129–130 kindling fluorescent proteins, 132–133 nonbioluminescent anthozoa GFP-like proteins, color transitions, 126, 128 Mutant structures, green fluorescent proteins, 73–75 aggregation modulation, 102–105 biosensors, 106–108 covalent alteration, Y66FHW, 94–96 energy level modification, 100–101 excitation peak ratios, durable photoisomerization, 96–100 halides, 109–110 insertion of other proteins, 110 mutational strategies, 85–87 nucleic acid changes, 112 nucleotide sequencing, 83–85 pH levels, 108–109 screening methods, 87–88 silent and loss-of-function mutations, 111–112 spectral mutant classification by chromophore, 89–94 imidazol derivation, Y66H, 93–94 indole derivation from Y66W, 93 neutral phenol, 92 phenolate anion, 89, 92 phenolate anionstacked pi-electron system, 92–93 phenyl derivation, Y66F, 94 wild-type neutral phenol/anionic phenolate mixture, 89 structure-based engineering, 76 tandem concentrations, 110–111 temperature-dependent folding, 101–102 truncations, 105–106

Mycobacterial promoters, bacterial green fluorescent protein, genetic applications, 171–174 Mycobacteria sp., GFP-host interactions, 169 Natural sources of GFP, 41–42 Neuronal function: C. elegans GFP measurement, 212–213 Drosophila GFP cell biology, 238–239 transgenic GFP analysis, 298 Neutral phenol: in chromophore, 92 wild-type green fluorescent proteins, 89 NFkB activity: b-arrestin recruitment assay, 372–376 fluorescent protein degradation assays, 372 Nidogen, G2F domains, 147–149 Nitrogen fixation, heterocyst formation in Anabaena sp., 167–168 nmt1+ promoter, S. pombe GFP gene expression, 183 Nomarski optical analysis, C. elegans GFP visualization, cellular anatomy, 208–212 Nonbioluminescent organisms, GFP-like proteins: anthozoa: FP modification, mutagenesis, 129–130 homolog color diversity, 122–124 mutagenesis color transitions, 126–128 oligomeric state, 126 photoactivatable probes, 130–134 Dronpa, 134 DsRed “greening,” 130–132 fluorescent protein kindling, 132–134 UV-induced green-to-red photoconversion, 132 research background, 121–122 sequence comparisons, 124–126 corals, gene encoding, 142–143 hydroid medusae, 141–142 oligomerization, 146–147 photoprotection hypothesis, 143–144 NOP1 promoter, S. cerevisiae GFP gene expression, 182 N-terminal domain: in dinoflagellate luciferase, 29–30 in firefly luciferase, 31–33 GFP truncation and fusion constructs, 76–77 mutant structures, 105–106 Nuclear architecture and dynamics, mammalian cell GFP, 320–321

436

INDEX

Nuclear-localized GFP (nlsGFP), Drosophila GFP detection, 230 Nuclear pore complexes (NPCs), yeast GFP gene expression, 193–194 Nuclear translocation assays, fluorescent proteins, 364–365 Nucleic acid sequences, green fluorescent protein mutations, 112 Nucleocytoplasmic transport, yeast GFP localization, 188–190 Obelia geniculata, green fluorescent protein in, 19–25 Oligomerization: GFP-like proteins, 146–147 GFP mutations, aggregation modulation, 104–105 nonbioluminescent anthozoa GFP-like proteins, 126 mutagenic modification, 129–130 “150% recovery,” organic solvents on green fluorescent proteins, 51–52 Open reading frame (ORF) technique: S. cerevisiae GFP gene expression, 181–182 yeast GFP gene expression, 195–196 Operons, bacterial luciferases, 26–28 Organelle structure: Drosophila GFP, 234 mammalian GFPs, 321–324 yeast GFP, 187–188 Organic solvents, green fluorescent protein, 51–52 absortpion/excitation spectral shifts, 60–61 Organism structure, in transgenic vertebrates, GFP expression, 287–288 Organized smooth endoplasmic reticulum (OSER), mammalian GFPs, 322–324 Orphan GPCRs, b-arrestin recruitment assay, 375–376 P. fluorescens: GFP-host interactions, 169–170 GFP isolation, 165 Pathway mapping, fluorescent proteins, 381–383 P-element transformation system, Drosophila GFP gene expression, 228–231 Pelican vectors, Drosophila GFP gene expression, 228–229 Peroxy flavin, lifetime, 26 pGALI-10 promoter, S. cerevisiae GFP gene expression, 182

Pharmaceutical applications, fluorescent proteins: drug screening assays, 363–380 b-arrestin recruitment assay, 372–376 BRET protein-protein interaction, 369–371 cell adhesion assays, 366 dual colour spectroscopy assays, 371–372 fluorescence-activated cell sorting, 378 fluorescent protein biosensors, 378–380 fluorescent protein degradation, 372 FRET/BRET assays, 367–371 FRET protease assays, 368 FRET protein-protein interaction, 368–369 GPCR/GFP fusion assays, 376–377 nuclear translocation assays, 364–365 protein complementation assays, 377–378 reporter gene assays, 366–367 high-content/high-throughput screening, 362–363 high content screening detection instrumentation, 380 target validation studies, 380–383 Phenolate anion: stacked p-electron system, 92–93 wild-type green fluorescent proteins: chromophore structures, 89, 92 neutral phenols, 89 Phenyl, chromophore derivation, 94 Phialidium sp.: GFP-like proteins, 141–142 green fluorescent protein structure, 45, 47–48 pH levels, green fluorescent protein structure, 45, 48 mutant structures, 108–109 specimen preparation, 418–419 Phosphatidylinositol (PtdIns), yeast GFP lipid localization, 192 Photoactivatable green fluorescent proteins: in mammalian cells, 308–309 nonbioluminescent anthozoa GFP-like proteins, 130–134 Dronpa, 134 DsRed “greening,” 130–132 fluorescent protein kindling, 132–134 UV-induced green-to-red photoconversion, 132 specimen preparation, 418–419 Photobleaching: Drosophila GFP, FRAP analysis, fusion proteins, 242–243

437

INDEX

GFP specimen preparation, 418–419 mammalian GFPs: kinetic analysis, 311–312 nuclear envelope and, 321 variant populations, 308–309 yeast GFP localization, 189–190 Photocyte dispersions: firefly luminescence, 30–33 green fluorescent protein species, 21–25 Photodamage, GFP specimen preparation, 418–419 Photoisomerization, green fluorescent protein mutant structures, 96–100 Photoprotection hypothesis, GFP-like proteins, 143–144 Photoprotein energy transfer, green fluorescent protein, 4–8 Photoreception, GFP-like proteins, 145–146 Phrixothrix, bioluminescence in, 30–33 Physiological processes, C. elegans GFP visualization, 213 PiggyBac element, Drosophila GFP gene expression, 228–229 Plants: GFP-host interactions, 169–170 green fluorescent proteins in: arabidopsis, cryptic splicing of gfp mRNA, 260–262 cell imaging, 274–276 cell marking, 280–281 cryptic intron removal, 262–263 gene expression, 263–264 modification, 274 localization, 264–265 maturation, 269–272 spectral modification, 272–273 subcellular dynamics visualization, 276–279 subcellular targeting, 265–269 Plasmid injection, transgenic fish GFP, 290–291 Plasmid-swap experiment, yeast green fluorescent protein localization, 188–192 Plasmid vectors, S. cerevisiae GFP gene expression, extrachromosomal constructs, 194–195 Plasticity studies, C. elegans GFP measurement, 212–213 Plextrin homology domain, yeast GFP lipid localization, 192 Pocillopora damicornis, GFP-like proteins, photoprotection hypothesis, 144

Polar follicle cells, Drosophila GFP, tissue morphogenesis, 233–234 Polymerase chain reaction (PCR): Arabidopsis gfp mRNA, cryptic splicing, 260–262 C. elegans fusion gene constructs, 217–218 Candida albicans GFP gene expression, 184 green fluorescent protein structure, 68 error-prone mutational strategies, 85–86 S. cerevisiae GFP gene expression, 181–182 Polymorphism, GFP-like proteins, gene coding, 142–143 Position effect variegation, Drosophila GFP reporters, 244–247 Potassium chloride, in Aequorea aequorea, 1–4 Preservatives, green fluorescent protein and, 52 Prokaryotes, green fluorescent protein in: bacterial development and cell biology, 165–168 cell division, E. coli, 168 heterocyst formation, Anabaena, 167–168 spore formation, B. subtilis, 165–167 bacterial ecology and behavior, 170–171 bacterial-host interactions, 169–170 complex environments, 168–171 future research issues, 174–175 genetic applications, 171–174 acid-inducible gene isolation, S. typhimurium, 172–173 flow cytometry, 172 macrophage-inducible gene isolation, S. typhimurium, 173–174 research background, 163–165 Protease assays: fluorescence resonance energy transfer assays, 368 in green fluorescent proteins, 45–48 physical effects of, 50 tandem concatenations, 111 Protein complementation assays, fluorescent protein screening, 377–378 Protein dissection, reassambled green fluorescent proteins, 392–393 Protein disulfide isomerase (PDI), Drosophila GFP, organelle fusion and transport, 234 Protein function, transgenic GFP analysis, 298 Protein localization: absorption suppression and, 59–61 C. elegans GFP determination, 206–208 mammalian cell GFP, 316–318 yeast GFP, 188–190

438

INDEX

Protein-protein interactions: bioluminescence resonance energy transfer assays, 369–371 dissected proteins, 392–393 fluorescence resonance energy transfer assays, 368–39 GFP reassembly, 398–400 unknown interactions, 400 mammalian cell GFP, FRET/FCS imaging, 312–315 Protein trap screens, Drosophila GFP, 250–251 Protocols for GFP expression: folding and temperature sensitivity, 408 purification, 408–413 specimen preparation, 413–419 fixed specimens, 413–418 photobleaching, photoactivation, photodamage, and pH dependence, 418–419 toxicity studies, 407–408 Proton transfer, green fluorescent protein mutant structures, 74–75 Purification protocols, GFP expression, 408–413 Pyrophorus plagiophthalamus, firefly luciferase, 32–33 Quantum mechanical modeling, green fluorescent protein structure, 68–69 Quantum yield of bioluminescence (Qbl), green fluorescent protein, 7–8 spectroscopic analysis, 54–55 Quorum sensing, bacterial luciferases, 27–28 Radiationless energy transfer, green fluorescent protein, 40–41 Random mutagenesis: DsRed/DsRed-Express mutants, 344 green fluorescent protein mutant structures, 85 amino acid predetermination, 86 nonbioluminescent anthozoa GFP-like protein color transitions, 126, 12800 Ratio imaging, mammalian cell green fluorescent proteins, 311 Reassambled green fluorescent proteins: applications, 398–402 protein expression patterns, 401–402 protein-protein interactions, 398–400 subcellular localization, 400–401 dissection and reassembly mechanisms: kinetics and mechanism, 396–397 proteins and cells, 397–398

variants and topology, 393–396 future research issues, 402–403 protein dissection, 392–393 Recombinant proteins: absorption suppression, 60–61 denaturation/renaturation, 49–50 dinoflagellate luciferases, 29–30 green fluorescent protein mutation, heteroduplex recombination, 87 monomer/dimer equilibrium, 58–59 S. cerevisiae GFP gene expression, 181–182 S. pombe GFP gene expression, 182–183 structural characteristics, 41–42, 44–52 Recombination cloning, C. elegans fusion gene constructs, 219 Red fluorescent proteins (RFPs): color diversity, anthoza GFP homologs, 122–124 mutagenesis, 128000 phylogenetic relationships, 152–153 yeast green fluorescent protein and, 185–186 Red-shifted emissions: bacterial luciferases, 27–28 green fluorescent protein mutant structures, wavelength altered ratios, 98–100 Red tides, dinoflagellates in, 28–30 Reef coral fluorescent proteins (RCFPs): basic properties of, 340–346 AmCyan 1 mutant, 341 AsRed2 mutant, 345 DsRed2/DsRed-express, 344 DsRed monomer, 345 fluorescent timer (E5), 345 HcRed1 mutant, 345–346 red fluorescent proteins, 343–344 ZsGreen 1 mutant, 341–343 ZsYellow1 mutant, 343 gene expression and detection, 346–357 flow cytometry, 351–353 single-color analysis, 353–355 three- and four-color analysis, 356 two color analysis, 355 fluorescence microscopy techniques, 347–348 laser scanning, 357 multicolor analysis, 349–351 single-color analysis, 348–349 spectrophotemetry, 357 GFP-like proteins, gene coding, 142–143 in mammalian cells, fluorescence microscopy/flow cytometry applications, 339–357 nomenclature, 340

439

INDEX

pharmaceutical applications: drug screening assays, 363–380 b-arrestin recruitment assay, 372–376 BRET protein-protein interaction, 369–371 cell adhesion assays, 366 dual colour spectroscopy assays, 371–372 fluorescence-activated cell sorting, 378 fluorescent protein biosensors, 378–380 fluorescent protein degradation, 372 FRET/BRET assays, 367–371 FRET protease assays, 368 FRET protein-protein interaction, 368–369 GPCR/GFP fusion assays, 376–377 nuclear translocation assays, 364–365 protein complementation assays, 377–378 reporter gene assays, 366–367 high-content/high-throughput screening, 362–363 high content screening detection instrumentation, 380 target validation studies, 380–383 phylogenetic relationships, 151–153 Regulatory sequences, C. elegans GFP decoding, 205–206 Renaturation, in green fluorescent protein, 49–50 Renilla: biochemistry, 19 GFP-like proteins, 140–1410 green fluorescent protein in, 7–8 biological function, 40–41 chromophore structure, 53–54 denaturation/renaturation, 49–50 fluorophore formation, 72–73 molar extinction coefficient, 56 organic solvents, 51–52 physical characteristics, 41–52 protease resistance in, 50–51 stability properties, 48–49 Reporter genes: C. elegans GFP decoding, regulatory sequences, 205–206 C. elegans GFP gene constructs, 217–219 C. elegans GFP pattern analysis, 204–205 C. elegans GFP protein localization, 206–208 cell fate markers, C. elegans GFP, 212

cellular anatomy visualization, C. elegans GFP, 211–212 Drosophila GFP, 244–247 fluorescent protein drug assays, 366–367 mammalian GFPs, 324–325 Retrotransposons, Drosophila GFP chromosomes, 234–235 Reverse transcriptase polymerase chain reaction (RT-PCR), Arabidopsis gfp mRNA, cryptic splicing, 260–262 Rhizobium sp., GFP-host interaction, 170 Ricordea proteins, phylogenetic relationships, 152–153 RNA-binding proteins, yeast GFP localization, 191–192 RNAi, Drosophila GFP reporters, 246–247 RNA localization, yeast GFP, 191–192 Rous sarcoma virus (RSV), transgenic chick GFP expression, 294 “RSGFP4” mutant, green fluorescent protein, 86 Saccharomyces cerevisiae: fixed specimen preparation, 414–415 green fluorescent protein in, 179–197 applications, 186–194 construct integration, 195 extrachromosomal (plasmid) constructs, 194–195 fluorescence resonance energy transfer, 192–193 gene expression systems, 180–186, 194–195 assessment of, 195–196 functional and proper localization, 184–185 S. cerevisiae, 181–182, 194–195 spectral variants and red fluorescent proteins, 185–186 genetic applications, 193–194 localization studies, 188–192 DNA, 190–191 lipids, 192 protein, 188–190 RNA, 191–192 microscopic analysis, 196 organelle structure, function, and inheritance, 187–188 Saccharomyces pombe: DNA localization, 191 fixed specimen preparation, 415 green fluorescent protein in, 179–180 gene expression systems, 182–183

440

INDEX

Salmonella typhimurium: acid-inducible gene isolation, 172–173 GFP-host interactions, 169 macrophage-inducible gene isolation, 173–174 Sapphire chromophores, neutral phenol, 92 Schizosaccharomyces pombe, fixed specimen preparation, 415–416 Scintillon fractionation, in dinoflagellates, 29–30 Screening methods: Drosophila GFP, 250–251 green fluorescent protein mutation, 87–88 Sequence comparison, nonbioluminescent anthozoa GFP-like proteins, 124–126 Ser-Tyr-Gly sequence, green fluorescent protein structure, 71–73 sg100 variant, GFP dissection and reassembly, 393–396 Sigma factors, GFP spore formation, B. subtilis, 165 Signature-tagged transposition, bacterial GFP, 175 Silent mutations, green fluorescent proteins, 111–112 Single-color analysis, reef coral fluorescent protein expression and detection, 348–349 Single-copy fusions, prokaryotic GFP, 165 Singlet oxygen, green fluorescent protein fluorophore formation, 72–73 Site-directed mutations: DsRed/DsRed-Express mutants, 344 green fluorescent protein structures, 86 nonbioluminescent anthozoa GFP-like protein color transitions, 126, 128 Sleeping Beauty transposon, in transgenic fish GFP, 292–293 Sodium dodecyl sulfate (SDS) gel electrophoresis, recombinant green fluorescent proteins, 45–46 Somatic nuclear transfer, in transgenic fish GFP, 293 Specimen preparation protocols, fixed GFP specimens, 413–418 Spectral emissions: green fluorescent protein structure, 68–69 photoproteins in coelentrates and, 20–25 plant GFP modification, 272–273 yeast green fluorescent protein, 185–186 Spectral mutants, green fluorescent protein, chromophore classification, 89–94

imidazol derivation, Y66H, 93–94 indole derivation from Y66W, 93 neutral phenol, 92 phenolate anion, 89, 92 phenolate anionstacked p-electron system, 92–93 phenyl derivation, Y66F, 94 wild-type neutral phenol/anionic phenolate mixture, 89 Spectrofluorimetric measurments, bacterial green fluorescent protein, genetic applications, 171–174 Spectroscopic analysis: green fluorescent protein chromophores, 53–56 brightness properties, 54 molar extinction coefficient, 55–56 quantum yield, 54–55 green fluorescent protein mutant structures, wavelength altered ratios, 98–100 Sperm nuclear transplantation, transgenic Xenopus GFP, 293–294 Spindle pole body (SPB): Drosophila GFP cell biology, 235–238 yeast GFP localization, 189 Spore formation, B. subtilis green fluorescent protein, 165–167 Stability properties, green fluorescent protein, 48–49 Stable cell populations, reef coral fluorescent protein expression and detection, 352–353 Stacked pi-electron system, phenolate anions, 92–93 Stress-response transcription factors, yeast GFP localization, 188–190 Structure-based engineering, green fluorescent protein, 76 Subcellular targeting: plant GFP, 265–269 visualization techniques, 276–279 reassambled GFPs, 400–401 “Superfolder” GFP variant, dissection and reassembly, 395–396 Synaptogyrin (SNG-1), C. elegans GFP protein localization, 207–208 Tagged proteins: C. elegans GFP, 215 Drosophila GFP, RNA, 241–242 Tandem concatenations, green fluoresent proteins, 110–111

441

INDEX

Tandem repeats, dinoflagellate luciferases, 29–30 Target validation studies, fluorescent protein applications, 380–383 Telomere formation, Drosophila GFP chromosomes, 234–235 Temperature loss (Tm): green fluorescent protein denaturation/renaturation, 49–50 protein folding, GFP mutants, 101–102 Temperature sensitivity: GFP expression protocols, 408 prokaryotic GFP, 164–165 Tetrameric structures, reef coral fluorescent proteins, 341 Tetratricopeptiderepeat (TPR) domains, GFP dissection and reassembly, 399–400 Three-dimensional structure, green fluorescent protein: activation control, 76 Aequorea mutant, 73–75 fluorophore properties, 71–73 folding mechanisms, 78–79 future research issues, 79 research background, 67–68 spectral and physical properties, 68–69 structure-based engineering, 76 truncation and fusion constructs, 76–77 Time-lapse imaging, mammalian cell green fluorescent proteins, 311 Timing issues, Drosophila GFP detection, 230 Tissue morphogenesis: Drosophila GFP, 233–234 in transgenic vertebrates, GFP expression, 287–288 Tobacco: firefly luciferase in, 31–33 GFP expression in, 263–264 Tol2 transposon, in transgenic fish GFP, 292–293 Topological analysis, GFP dissection and reassembly, 393–396 Total internal reflection fluorescence microscopy (TIR-FM), mammalian cell GFP, 315–316 Toxicity studies, GFP expression protocols, 407–408 Transcription activators, plant GFP marking, 280–281 Transcription factors: bioluminescence resonance energy transfer assays, 370–371 Drosophila GFP, 241

translocation assays, fluorescent proteins, 364–365 Transformation vectors, Drosophila GFP gene expression, 228–229 Transgenic lines, C. elegans fusion gene constructs, 219 Transgenic vertebrates, green fluorescent protein in: cell markers, 296–297 chicks, 294 coexpression markers, 296 definitions, 286 enhancer analysis, 297–298 expression levels, 287 fish species, 290–293 BAC injection, 292 meganuclease, 292 plasmid injection, 290–291 somatic nuclear transfer, 293 transposons, 292–293 gene expression markers, 297 mice, 294–296 DNA injection, 294–295 knock-in species, 295–296 techniques, 286–290 tissue and organism, 287–288 variants, 288–289 visualization methods, 289–290 in vivo protein, fusion constructs, 298 Xenopus species, 293–294 Transient gene expression, Drosophila GFP, 229 Transposons, in transgenic fish GFP, 292–293 Truncation constructs, green fluorescent protein, 76–77 mutant structures, 105–106 Tubulin binding, Drosophila GFP cell biology, 235–238 Turbo (T-sapphire) mutation, durable photoisomerization, 97–100 Two-color analysis: drug assays, fluorescent proteins, 371–372 reef coral fluorescent proteins, 349–351 flow cytometry, 355 T203Y GFP mutant, energy level modification, 100–101 Ultraviolet-induced green-to-red photoconversion, “Kaede” protein, 132 Upstream activator sequence (UAS), Drosophila GFP gene expression, 228

442

INDEX

vab-7 reporter gene, C. elegans GFP pattern analysis, 205 Variant green fluorescent proteins: mammalian cells, 305–309 in transgenic vertebrates, 288–289 Venus YFP variants: in mammalian cells, 308–309 in transgenic vertebrates, 288–289 Vibratome sectioning, transgenic vertebrate GFP, 289–290 Video imaging, green fluorescent protein mutation, 88 Viral infection: mammalian cell GFP and pathogenesis, 319 transgenic chick GFP, 294 Viral systems, mammalian cell GFP gene targeting, 318–319 Visualization techniques: C. elegans GFP, 219 plant GFP, 274–276 reassembled GFP expression patterns, 401–402 in transgenic vertebrate GFP, 289–290 Visual screening procedures, green fluorescent protein mutation, 87–88 Wavelength excitation: green fluorescent protein mutations, altered ratios, 96–100 organic solvents on green fluorescent proteins, 52 Wee-P trap, Drosophila GFP, 251 Wild-type neutral phenol/anionic phenolate, spectral properties, 89 Xenopus species, transgenic GFP in, 293–294 Y. pseudotuberculosis, GFP-host interactions, 169 Yeasts: green fluorescent protein in, 179–197 applications, 186–194 construct integration, 195 extrachromosomal (plasmid) S. cerevisiae constructs, 194–195 fixed specimen preparation, 414–416 fluorescence resonance energy transfer, 192–193 gene expression systems, 180–186 assessment of, 195–196 C. albicans, 183–184 functional and proper localization, 184–185

S. cerevisiae, 181–182, 194–195 S. pombe, 182–183 spectral variants and red fluorescent proteins, 185–186 genetic applications, 193–194 localization studies, 188–192 DNA, 190–191 lipids, 192 protein, 188–190 RNA, 191–192 microscopic analysis, 196 organelle structure, function, and inheritance, 187–188 transgenic fish GFP, 292 Yeast two-hybrid (Y2H) screening, protein dissection, 392–393 yEGFP variant, Candida albicans GFP gene expression, 183–184 Yellow fluorescent protein (YFP): aggregation modulation, 105 in bacteria, 25–28 C. elegans variants, 216–217 Candida albicans GFP gene expression, 183–184 color diversity, anthoza GFP homologs, 122–124 mutagenesis, 128 dissection and reassembly, 394–396 drug assays, 367–371 durable photoisomerization, 96–100 flavin mononucleotide (FMN), 27–28 halide mutations, 109–110 in mammalian cells, 308–309 neuronal function and plasticity studies, 212–213 phenolate anion with stacked pi-electron system, 92–93 protease assays, 368 S. cerevisiae GFP gene expression, 181–182 S. pombe GFP gene expression, 183 in transgenic vertebrates, 288–289 yeast GFP localization, 189–190 fluorescence resonance energy transfer, 192–193 Y66F chromophore, phenyl derivation, 94 Y66FHW, covalent mutations, 94–96 Y66H chromophore: covalent mutation of fluorophore, 95–96 imidazole derivation, 93–94 Y66W chromophore, indole derivation, 93

443

INDEX

Zoantharia orders: GFP-like proteins, 154–155 reef coral fluorescent proteins, 340–341 ZaGreen1 mutant, 341, 343 Zooxanthellae: color diversity, 144–145 GFP-like proteins, photoprotection hypothesis, 144

ZsGreen1 mutant: basic properties, 341, 343 flow cytometry, single-color analysis, 354 ZsYellow1 mutant: basic properties, 343 flow cytometry, single-color analysis, 354 multicolor analysis, 349–351