Reporter Genes: A Practical Guide

Reporter Genes A Practical Guide M E T H O D S I N M O L E C U L A R B I O L O G Y™ John M. Walker, SERIES EDITOR 43...

Author: Don Anson

40 downloads 1687 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Reporter Genes A Practical Guide

M E T H O D S I N M O L E C U L A R B I O L O G Y™

John M. Walker, SERIES EDITOR 436. Avian Influenza Virus, edited by Erica Spackman, 2008 435. Chromosomal Mutagenesis, edited by Greg Davis and Kevin J. Kayser, 2008 434. Gene Therapy Protocols: Vol. 2: Design and Characterization of Gene Transfer Vectors, edited by Joseph M. LeDoux, 2008 433. Gene Therapy Protocols: Vol. 1: Production and In Vivo Applications of Gene Transfer Vectors, edited by Joseph M. LeDoux, 2008 432. Organelle Proteomics, edited by Delphine Pflieger and Jean Rossier, 2008 431. Bacterial Pathogenesis: Methods and Protocols, edited by Frank DeLeo and Michael Otto, 2008 430. Hematopoietic Stem Cell Protocols, edited by Kevin D. Bunting, 2008 429. Molecular Beacons: Signalling Nucleic Acid Probes, Methods and Protocols, edited by Andreas Marx and Oliver Seitz, 2008 428. Clinical Proteomics: Methods and Protocols, edited by Antonio Vlahou, 2008 427. Plant Embryogenesis, edited by Maria Fernanda Suarez and Peter Bozhkov, 2008 426. Structural Proteomics: High-Throughput Methods, edited by Bostjan Kobe, Mitchell Guss, and Huber Thomas, 2008 425. 2D PAGE: Vol. 2: Applications and Protocols, edited by Anton Posch, 2008 424. 2D PAGE: Vol. 1: Sample Preparation and Pre-Fractionation, edited by Anton Posch, 2008 423. Electroporation Protocols, edited by Shulin Li, 2008 422. Phylogenomics, edited by William J. Murphy, 2008 421. Affinity Chromatography: Methods and Protocols, Second Edition, edited by Michael Zachariou, 2008 420. Drosophila: Methods and Protocols, edited by Christian Dahmann, 2008 419. Post-Transcriptional Gene Regulation, edited by Jeffrey Wilusz, 2008 418. Avidin–Biotin Interactions: Methods and Applications, edited by Robert J. McMahon, 2008 417. Tissue Engineering, Second Edition, edited by Hannsjörg Hauser and Martin Fussenegger, 2007 416. Gene Essentiality: Protocols and Bioinformatics, edited by Svetlana Gerdes and Andrei L. Osterman, 2008 415. Innate Immunity, edited by Jonathan Ewbank and Eric Vivier, 2007 414. Apoptosis in Cancer: Methods and Protocols, edited by Gil Mor and Ayesha Alvero, 2008

413. Protein Structure Prediction, Second Edition, edited by Mohammed Zaki and Chris Bystroff, 2008 412. Neutrophil Methods and Protocols, edited by Mark T. Quinn, Frank R. DeLeo, and Gary M. Bokoch, 2007 411. Reporter Genes: A Practical Guide, edited by Don Anson, 2007 410. Environmental Genomics, edited by Cristofre C. Martin, 2007 409. Immunoinformatics: Predicting Immunogenicity In Silico, edited by Darren R. Flower, 2007 408. Gene Function Analysis, edited by Michael Ochs, 2007 407. Stem Cell Assays, edited by Vemuri C. Mohan, 2007 406. Plant Bioinformatics: Methods and Protocols, edited by David Edwards, 2007 405. Telomerase Inhibition: Strategies and Protocols, edited by Lucy Andrews and Trygve O. Tollefsbol, 2007 404. Topics in Biostatistics, edited by Walter T. Ambrosius, 2007 403. Patch-Clamp Methods and Protocols, edited by Peter Molnar and James J. Hickman, 2007 402. PCR Primer Design, edited by Anton Yuryev, 2007 401. Neuroinformatics, edited by Chiquito J. Crasto, 2007 400. Methods in Membrane Lipids, edited by Alex Dopico, 2007 399. Neuroprotection Methods and Protocols, edited by Tiziana Borsello, 2007 398. Lipid Rafts, edited by Thomas J. McIntosh, 2007 397. Hedgehog Signaling Protocols, edited by Jamila I. Horabin, 2007 396. Comparative Genomics, Vol. 2, edited by Nicholas H. Bergman, 2007 395. Comparative Genomics, Vol. 1, edited by Nicholas H. Bergman, 2007 394. Salmonella: Methods and Protocols, edited by Heide Schatten and Abraham Eisenstark, 2007 393. Plant Secondary Metabolites, edited by Harinder P. S. Makkar, P. Siddhuraju, and Klaus Becker, 2007 392. Molecular Motors: Methods and Protocols, edited by Ann O. Sperry, 2007 391. Methicillin-Resistant Staphylococcus aureus (MRSA) Protocols, edited by Yinduo Ji, 2007 390. Protein Targeting Protocols, Second Edition, edited by Mark van der Giezen, 2007 389. Pichia Protocols, Second Edition, edited by James M. Cregg, 2007

M E T H O D S I N M O L E C U L A R B I O L O G Y™

Reporter Genes A Practical Guide

Edited by

Donald S. Anson Department of Genetic Medicine, Women’s and Children’s Hospital, Adelaide, Australia

HUMANA PRESS

TOTOWA, NEW JERSEY

© 2007 Humana Press Inc. 999 Riverview Drive, Suite 208 Totowa, New Jersey 07512 www.humanapress.com All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, or otherwise without written permission from the Publisher. Methods in Molecular BiologyTM is a trademark of The Humana Press Inc. All papers, comments, opinions, conclusions, or recommendations are those of the author(s), and do not necessarily reflect the views of the publisher. This publication is printed on acid-free paper. ∞ ANSI Z39.48-1984 (American Standards Institute) Permanence of Paper for Printed Library Materials. Cover illustration: Figure 1, Chapter 11, “Green Fluorescent Protein as a Tracer in Chimeric Tissues,” by Harold Jockusch and Daniel Eberhard. Fig. 1A from Chapter 11, "Human Solid Tumors," by Pietro Production Editor: Michele Seugling Cover design by Karen Schulz For additional copies, pricing for bulk purchases, and/or information about other Humana titles, contact Humana at the above address or at any of the following numbers: Tel.: 973-256-1699; Fax: 973-256-8341; E-mail: [email protected]; or visit our Website: www.humanapress.com Photocopy Authorization Policy: Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by Humana Press Inc., provided that the base fee of US $30.00 per copy is paid directly to the Copyright Clearance Center at 222 Rosewood Drive, Danvers, MA 01923. For those organizations that have been granted a photocopy license from the CCC, a separate system of payment has been arranged and is acceptable to Humana Press Inc. The fee code for users of the Transactional Reporting Service is: [978-1-58829-739-6/07 $30.00]. Printed in the United States of America. 10 9 8 7 6 5 4 3 2 1 e-ISBN: 978-1-60327-072-4 Library of Congress Cataloging-in-Publication Data Reporter genes : a practical guide / edited by Donald S. Anson. p. ; cm. — (Methods in molecular biology ; 411) Includes bibliographical references and index. ISBN-13: 978-1-58829-739-6 (alk. paper) 1. Reporter genes. 2. Mammals—Cytology. 3. Mammals— Genetics. I. Anson, Donald. II. Series: Methods in molecular biology (Clifton, N.J.); v.411. [DNLM: 1. Genes, Reporter. 2. Cytological Techniques. 3. Mammals—genetics. W1 ME9616J v.411 2007 / QU 470 R425 2007] QH447.8.R47R47 2007 572.8'19—dc22

2007004637

Preface Reporter genes have played, and continue to play, a vital role in many areas of biological research by providing a ready means for qualitative and quantitative assessment of the activity of genes and location of gene products in different environments. For example, reporter genes have played a major role in defining the activity of different genetic elements that control transcription of genes, both in vitro and in vivo (1,2), in the study of cell lineages (2,3) and in determining the effectiveness of different gene transfer technologies (4,5). While early reporter genes required fixation of cells for visualization, or the preparation of cell extracts for quantitative assays, there has been a move towards reporter genes that can be assayed quantitatively and/or qualitatively in live cells and animals. The two widely used examples of such markers are the fluorescent proteins (6–8) and luciferase (9,10). However, despite the development of these new reporter genes, one of the earliest, E. coli β-galactosidase (LacZ), is still widely used as a marker and offers significant advantages, especially for histological analysis (11). Most reporter genes originate from non-mammalian species, and most have subsequently been modified to enhance their expression in mammalian cells and/or to modify their characteristics, vastly increasing their usefulness. For example, codon optimisation for expression in mammalian cells has been applied to the fluorescent proteins (12), luciferase (13) and β-galactosidase (14). The characteristics of fluorescent proteins have been extensively modified, with variants of many different colours and characteristics now available (15,16 and see chapter by Patterson in this book). LacZ (14,17,18) and fluorescent proteins (18,19) can also be targeted to the nucleus of cells by addition of suitable trafficking signals. Fluorescent proteins are also widely used for making fusion proteins to allow analysis of sub-cellular trafficking and compartmentalisation of proteins of interest (2,20,21). Reporter genes have provided powerful tools for analysis of gene expression, either by localisation and/or by quantitative analysis. Examples of the former include the use of LacZ, where staining of microscopic sections gives information on gene expression at cellular resolution, and fluorescent proteins, where direct visualisation of tissue can give similar information. The use of luciferase as a marker gene in vivo also results in information regarding the localisation of gene expression in the living animal, although at a much lower resolution. In this system in vivo gene expression can be quantified by photonic imaging (see chapter by Ray and Gambhir). v

vi

Preface

Most marker genes can be used to provide quantitative data regarding gene expression. However, in some cases this requires the preparation of cell lysates. For example, β-galactosidase (LacZ) can be quantified by ELISA or enzyme activity and kits for this are available from a number of manufacturers, as are kits for quantitative analysis of several other enzymatic reporter genes such as alkaline phosphatase and luciferase. Reporter genes that can be quantitatively assayed without the preparation of a cell lysate include luciferase (via photonic imaging), secreted alkaline phosphatase (enzyme assay) and fluorescent proteins (via FACS analysis or microscopy). New marker/reporter gene systems are currently being developed. For example, systems based on the use of positron-emission tomography (PET) offer a means for non-invasive imaging of all tissues (22,23). While these are not covered in this book, the development of microPET machines suitable for imaging rodents means that this technology is likely to be rapidly developed over the next few years. As indicated above, molecular engineering is constantly being used to improve existing reporter systems. This book will describe practical protocols for experimentation with the most useful reporter genes for mammalian systems that are available and will concentrate on those marker genes that are currently most commonly used. Donald S. Anson

References 1. Mayer-Kuckuk, P., Menon, L. G., Blasberg, R. G., Bertino, J. R., and Banerjee, D. (2004) Role of reporter gene imaging in molecular and cellular biology. Biol. Chem. 385, 353–361. 2. Yu, Y. A., Szalay, A. A., Wang, G., and Oberg, K. (2003) Visualization of molecular and cellular events with green fluorescent proteins in developing embryos: a review. Luminescence 18, 1–18. 3. Trainor, P. A., Zhou, S. X., Parameswaran, M., et al. (1999) Application of lacZ transgenic mice to cell lineage studies. Methods Mol. Biol. 97, 183–200. 4. Bogdanov, A. Jr. (2003) In vivo imaging in the development of gene therapy vectors. Curr. Opin. Mol. Ther. 5, 594–602. 5. McCaffrey, A., Kay, M. A., and Contag, C. H. (2003) Advancing molecular therapies through in vivo bioluminescent imaging. Mol. Imaging 2, 75–86. 6. Hoffman, R. M. (2005) Advantages of multi-color fluorescent proteins for wholebody and in vivo cellular imaging. J. Biomed. Opt. 10, 41202-1–41202-10. 7. Passamaneck, Y. J., Di Gregorio, A., Papaioannou, V. E., and Hadjantonakis, A. K. (2006) Live imaging of fluorescent proteins in chordate embryos: from ascidians to mice. Microsc. Res. Tech. 69, 160–167.

Preface

vii

8. Wiedenmann, J. and Nienhaus, G. U. (2006) Live-cell imaging with EosFP and other photoactivatable marker proteins of the GFP family. Expert Rev. Proteomics 3, 361–374. 9. Sadikot, R. T. and Blackwell, T. S. (2005) Bioluminescence imaging. Proc. Am. Thorac. Soc. 2, 537–540. 10. Welsh, D. K. and Kay, S. A. (2005) Bioluminescence imaging in living organisms. Curr. Opin. Biotechnol. 16, 73–78. 11. Franco, D., de Boer, P. A., de Gier-de Vries, C., Lamersm W. H., and Moorman, A. F. (2001) Methods on in situ hybridization, immunohistochemistry and betagalactosidase reporter gene detection. Eur. J. Morphol. 39, 169–191. 12. Yang, T. T., Cheng, L., and Kain, S. R. (1996) Optimized codon usage and chromophore mutations provide enhanced sensitivity with the green fluorescent protein. Nucleic Acids Res. 24, 4592–4593. 13. Promega Inc sells synthetic luc2 (Photinus pyralis, see http://www.promega.com/ pnotes/89/12416_07/12416_07.pdf) and hRluc (Renilla reniformis, see http://www. promega.com/pnotes/79/9492_06/9492_06.pdf) genes that are codon-optimised for expression in mammalian cells. 14. Anson, D. S. and Limberis, M. (2004) An improved beta-galactosidase reporter gene. J. Biotechnol. 108, 17–30. 15. Chudakov, D. M., Lukyanov, S., and Lukyanov, K. A. (2005) Fluorescent proteins as a toolkit for in vivo imaging. Trends Biotechnol. 23, 605–613. 16. Miyawaki, A., Nagai, T., and Mizunom H. (2005) Engineering fluorescent proteins. Adv. Biochem. Eng. Biotechnol. 95, 1–15. 17. Bonnerot, C., Rocancourt, D., Briand, P., Grimber, G., and Nicolas, J. F. (1987) A beta-galactosidase hybrid protein targeted to nuclei as a marker for developmental studies. Proc. Natl. Acad. Sci. USA 84, 6795–6799. 18. Sorg, G. and Stamminger, T. (1999) Mapping of nuclear localization signals by simultaneous fusion to green fluorescent protein and to beta-galactosidase. Biotechniques 26, 858–862. 19. 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. 20. Miyawaki, A., Nagai, T., and Mizuno, H. (2005) Engineering fluorescent proteins. Adv. Biochem. Eng. Biotechnol. 95, 1–15. 21. van Roessel, P. and Brand, A. H. (2002) Imaging into the future: visualizing gene expression and protein interactions with fluorescent proteins. Nat. Cell. Biol. 4, E15–20. 22. Serganova, I. and Blasberg, R. (2005) Reporter gene imaging: potential impact on therapy. Nucl. Med. Biol. 32, 763–780. 23. Herschman, H. R. (2004) PET reporter genes for noninvasive imaging of gene therapy, cell tracking and transgenic analysis. Crit. Rev. Oncol. Hematol. 51, 191–204.

Contents Preface ......................................................................................................... 11v Contributors .................................................................................................. .1ix 1 Methodologies for Staining and Visualisation of β-Galactosidase in Mouse Embryos and Tissues Siobhan Loughna and Deborah Henderson ...................................... 111 2 Immunohistochemical Detection of β-Galactosidase or Green Fluorescent Protein on Tissue Sections Philip A. Seymour and Maike Sander ............................................... 113 3 Detection of Reporter Gene Expression in Murine Airways Maria Limberis, Peter Bell, and James M. Wilson ............................ 125 4 Three-Dimensional Analysis of Molecular Signals with Episcopic Imaging Techniques Wolfgang J. Weninger and Timothy J. Mohun .................................. 135 5 Fluorescent Proteins for Cell Biology George H. Patterson ......................................................................... 147 6 Detection of GFP During Nervous System Development in Drosophila melanogaster Karin Edoff, James S. Dods, and Andrea H. Brand ............................. 81 7 Autofluorescent Proteins for Flow Cytometry Charles G. Bailey and John E. J. Rasko ............................................... 99 8 Fluorescent Protein Reporter Systems for Single-Cell Measurements Steven K. Dower, Eva E. Qwarnstrom, and Endre Kiss-Toth ............ 111 9 Subcellular Imaging of Cancer Cells in Live Mice Robert M. Hoffman .......................................................................... 121 10 Noninvasive Imaging of Molecular Events with Bioluminescent Reporter Genes in Living Subjects Pritha Ray and Sanjiv Sam Gambhir ................................................ 131 11 Green Fluorescent Protein as a Tracer in Chimeric Tissues: The Power of Vapor Fixation Harald Jockusch and Daniel Eberhard ............................................. 145 Index ........................................................................................................... 155

viii

Contributors DONALD S. ANSON • Gene Technology Unit, Department of Genetic Medicine, Children, Youth and Women’s Health Service, Adelaide, Australia CHARLES G. BAILEY • Gene and Stem Cell Therapy Program, Centenary Institute of Cancer Medicine and Cell Biology, Australia PETER BELL • Gene Therapy Program, Department of Pathology and Laboratory Medicine, University of Pennsylvania, USA ANDREA H. BRAND • Wellcome Trust/Cancer Research UK Gurdon Institute and Department of Physiology, Development and Neuroscience, University of Cambridge, UK JAMES S. DODS • Wellcome Trust/Cancer Research UK Gurdon Institute and Department of Physiology, Development and Neuroscience, University of Cambridge, UK STEVEN K. DOWER • Section of Functional Genomics, School of Medicine and Biomedical Sciences, University of Sheffield, UK DANIEL EBERHARD • Department of Biology and Biochemistry, Centre for Regenerative Medicine, Bath University, UK KARIN EDOFF • Wellcome Trust/Cancer Research UK Gurdon Institute and Department of Physiology, Development and Neuroscience, University of Cambridge, UK SANJIV SAM GAMBHIR • Molecular Imaging Program at Stanford (MIPS), Departments of Radiology and Bioengineering, Bio-X Program, Stanford University, USA ROBERT M. HOFFMAN • AntiCancer, Inc., San Diego, and Department of Surgery, University of California at San Diego, USA HARALD JOCKUSCH • Developmental Biology and Molecular Pathology, Bielefeld University, Germany DR DEBORAH HENDERSON • Institute of Human Genetics, University of Newcastle Upon Tyne, UK ENDRE KISS-TOTH • Cardiovascular Research Unit, School of Medicine and Biomedical Sciences, University of Sheffield, UK MARIA LIMBERIS • Gene Therapy Program, Department of Pathology and Laboratory Medicine, University of Pennsylvania, USA DR SIOBHAN LOUGHNA • School of Biomedical Sciences, University of Nottingham, UK ix

x

Contributors

TIMOTHY J MOHUN • Developmental Biology Division, MRC National Institute for Medical Research, Mill Hill, UK GEORGE H. PATTERSON • Cell Biology and Metabolism Branch, National Institute of Child Health and Human Development, National Institutes of Health, USA EVA E. QWARNSTROM • Section of Cell Biology, School of Medicine and Biomedical Sciences, University of Sheffield, UK JOHN E.J. RASKO • Gene and Stem Cell Therapy Program, Centenary Institute of Cancer Medicine and Cell Biology, Australia and Cell and Molecular Therapies, Sydney Cancer Centre, Royal Prince Alfred Hospital, Australia PRITHA RAY • Molecular Imaging Program at Stanford (MIPS), Departments of Radiology and Bioengineering, Bio-X Program, Stanford University, USA MAIKE SANDER • Department of Developmental and Cell Biology, University of California, Irvine, USA PHILIP A. SEYMOUR • Department of Developmental and Cell Biology, University of California, Irvine, USA WOLFGANG J WENINGER • Integrative Morphology Group, Medical University of Vienna, Austria JAMES M. WILSON • Gene Therapy Program, Department of Pathology and Laboratory Medicine, University of Pennsylvania, USA

Detection of β-Galactosidase in Mouse Embryos

1

1 Methodologies for Staining and Visualisation of β -Galactosidase in Mouse Embryos and Tissues Siobhan Loughna and Deborah Henderson Summary This chapter provides information on β-galactosidase staining of whole mouse embryos, organs, tissue sections, and cultured cells, as well as double staining with horseradish peroxidase and use as a tool for genotyping. Using these protocols, localization of β-galactosidase can be visualized throughout development and in adult tissues. β-Galactosidase staining may be used purely as a marker of gene expression and also as a tracer in cell lineage studies. Key Words: β-Galactosidase; embryo; tissues; cultured cells; immunohistochemistry.

1. Introduction The E. coli lacZ gene is the first gene of the lac operon. The LacZ gene, encoding β-galactosidase (β-gal), is the classical histochemical reporter gene (1). β-Gal is a stable enzyme that may be expressed in cultured cells, in the fruit fly Drosophila, and in transgenic animals, with no apparent side effects. β-Gal hydrolyzes (2) the disaccharide lactose into two monosaccharide sugar groups, glucose and galactose. Enzyme activity can be detected using a variety of chromogenic substrates, such as the indole derivative 5-bromo-4-chloro-3indolyl-β-D-galactoside (X-gal). X-gal is cleaved by β-gal into an insoluble, stable, bright blue precipitate (3,4). The dimerization and oxidation reactions require transfer of an electron by the electron acceptors provided by the ferric and ferrous ions included in the LacZ staining solution (5,6). Protocols have arisen that allow gene expression to be monitored by linking up the promoter of a gene of interest to a marker gene, typically LacZ. An alternative strategy is to fuse a marker gene, in frame, to the coding sequence of a gene of interest. Both of these strategies allow the expression of a gene to be readily monitored, with the normal temporal pattern of expression maintained. From: Methods in Molecular Biology, vol. 411: Reporter Genes: A Practical Guide Edited by: D. Anson © Humana Press Inc., Totowa, NJ

1

2

Loughna and Henderson

This means that expression of LacZ will decrease when the gene would normally be downregulated, which allows a faithful pattern of gene expression to be obtained. However, for lineage tracing of specific cell types, another strategy is adopted. The ideal marker for lineage tracing of specific cell types in a developing embryo would completely and exclusively mark that cell and all daughter cells, with the minimum of background, while maintaining expression throughout the lifetime of the cell. This has been achieved over recent years in mammalian embryos by utilizing a two-component system, based on Cre/lox recombination, to mark specific cell lineages indelibly. The first component is a transgene expressing the enzyme Cre recombinase under the control of a cell type- or tissue-specific promoter. This allows Cre recombinase to be expressed only in those cells in which the promoter would normally drive expression. The second component is a conditional reporter gene that expresses a histological marker only upon Cre-mediated recombination. The most commonly used of these is the R26R gene, which expresses β-gal from the ROSA26 locus only upon Cre-mediated recombination (7). The R26R system has been shown to be ideal for these purposes, as ROSA26 is ubiquitously and uniformly expressed at all developmental and postnatal times, with no apparent sensitivity to genetic or environmental manipulation. In the absence of Cre recombinase, the ROSA26 gene is not expressed, but following recombination, a functional β-gal protein is produced. In this system, once the recombination event has occurred, it will be transferred to progeny of the original cells, even though the initial transgene is no longer expressed. Figure 1 shows LacZ staining when the ROSA26 β-gal gene is activated by expression of Cre from the Wnt1 promoter, which is specifically active in neural crest cells and their progeny (8,9). This chapter describes protocols for detecting β-gal expression in wholemount embryos, organs, or tissues, cryostat tissue sections, and cultured cells and for double staining for β-gal and horseradish peroxidase. Owing to endogenous β-gal activity, background staining can be a problem. However, this chapter will provide some useful tips on reducing background while maintaining high levels of positive staining. All protocols follow the same principles in terms of the initial fixation, washing, and staining, as outlined in the flowchart (Fig. 2). 2. Materials 2.1. Fixation 1. Phosphate-buffered saline (PBS), calcium free (Oxoid). Make up in deionized water and autoclave before use (see Note 1). 2. 25% glutaraldehyde (Sigma). May be freeze/thawed, but best to work from small aliquots. Caution: toxic.


3

Fig. 1. β-Galactosidase staining of an embryonic day 10.5 Wnt1CreRosa26R transgenic mouse embryo. Whole mount embryo (A) and tissue section (B) stained for β-galactosidase expression (the dotted line in A denotes the approximate line of cut for tissue section B). (A) LacZ staining can be seen in the craniofacial region, branchial arches, dorsal root ganglia (drg), and less intense staining can be seen in the outflow tract of the heart (oft). (B) Sectioning the same embryo confirms LacZ staining in the dorsal part of the neural tube (nt), surrounding the branchial arch arteries, and in the outflow tract of the heart (oft). a, aortic sac; l, limb; acv, anterior cardinal vein. See accompanying CD for color version.

3. Paraformaldehyde (PFA) comes as a powder (Sigma). We generally make up fresh each time, although it can be made up in aliquots and stored at −20°C. Avoid freeze/ thawing. Caution: toxic. a. Make up 4% PFA in water, and dissolve by adding 1 to 2 drops of 5 M NaOH and heating at about 60°C with gentle stirring. b. Add 10X PBS to make a 1X PBS solution. c. Cool and check pH with pH paper—must be pH 7.0. 4. LacZ fix: 1% PFA, 0.2% glutaraldehyde, 2 mM MgCl2, 5 mM EGTA (pH 8.0), 0.2% Nonidet-P40 (NP40) (stock: 10%, made in H2O stored at 4°C. We make fresh every few weeks). Make fresh before use. To make up the 0.5 M EGTA stock solution, add the required amount of EGTA to water, and dissolve by adding small amounts of 5 M NaOH and mixing until the EGTA has fully dissolved (see Note 2).

2.2. Washing and Staining 1. LacZ wash solution in 1X PBS: 2 mM MgCl2, 0.01% sodium deoxycholate (stock: 1%, made in H2O stored at 4°C), 0.02% NP40.

4


Fig. 2. Flowchart of the general procedure for LacZ staining.

2. LacZ stock substrate: 10% X-gal (Sigma or Melford Laboratories) dissolved in dimethylformamide. Caution: toxic. Store in the dark at 4°C. Solid X-gal should be stored at −20°C. Before opening bottle to remove any X-gal, allow to warm up to room temperature first to prevent moisture uptake by the solid, which will lead to gradual deterioration. 3. LacZ staining solution: a working staining solution is 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide (Caution: these two are toxic), 2 mM MgCl2, 0.01% sodium deoxycholate (stock: 1%, made in H2O stored at 4°C), 0.02% NP40 (stock: 10%, made in H2O stored at 4°C), 0.1% X-gal (from 10% stock solution). Store in the dark at 4°C; make fresh before use (see Note 3).

2.3. Tissue Embedding and Coating Slides 1. Tissue-Tek OCT compound (Sakura, available from Raymond Lamb). 2. TESPA (or 3-aminopropyltriethoxysilane [APES]; Sigma) coated slides are used (can be performed in house (see Note 4) or bought commercially (Marienfeld). 3. Paraffin wax (Fisher Scientific), and Histoclear (National Diagnostics).

2.4. β -Galactosidase and Horseradish Peroxidase Double Staining 1. Tris-buffered saline (TBS): 0.1 M Tris-HCl, pH 7.5, 0.15 M NaCl, in deionized water. Autoclave before use. 2. TBS-Tx: TBS with 0.5% Triton X-100 added. Do not autoclave after the addition of Triton X-100.


5

Table 1 Suggested Fixation and Staining Times for Embryonic and Fetal Mouse Specimens Agea E9.5 E10.5 E11.5 E12.5 E13.5–E14.5 E15.5–E18.5 aE

Whole or dissected embryo

Fixation time b (min)

Staining time

Whole Whole Whole Dissected Dissected Dissected

5 5 15 15 20 30–90

6h 6h Overnight Overnight Overnight Overnight

is embryonic day, where E0.5 is the morning of finding the copulation plug. time in wash buffer is equal to the time in fixing solution for each age.

b Total

3. Fetal calf serum (FCS): heat inactivate at 56°C for 30 min and then store as aliquots at −20°C. 4. AB Complex conjugated to horseradish peroxidase (HRP) (Dako) and made up according to the manufacturer’s instructions. 5. Diaminobenzidene (DAB) tablets (obtained from Sigma): Stored at −20°C. Caution: toxic. Make up according to the manufacturer’s instructions and filter through a 0.45-µm filter before use. 6. 100 mM Sodium citrate buffer, pH 6.0, is a 10X stock solution. Adjust the pH with NaOH. Autoclave before use. 7. Histoclear (National Diagnostics).

3. Methods 3.1. β -Galactosidase Staining of Whole Embryos (or Tissue Pieces/Whole Organs) β-Gal staining can be performed on whole embryos up to about mouse embryonic day 11.5, after which it is advisable to dissect the embryo into smaller pieces, isolate whole organs, or perform cryostat sectioning (see Subheading 3.2.) owing to decreased penetration of fix and staining reagents in large specimens. All steps take place at room temperature unless otherwise stated. 1. Wash embryos in PBS three times at room temperature to remove any media or serum. At this stage, larger embryos may need to be dissected to allow penetration of the staining solutions (see Note 5). 2. Fix embryos in cold LacZ fixative at 4°C for 5 to 90 min (see Note 2 and Table 1). 3. Wash embryos in LacZ wash solution three times for 20 min each at room temperature. 4. Stain the embryos in the LacZ staining solution for the required period at 37°C. This stage will need optimizing, but see Table 1 for guidelines. By the end of the

6


Table 2 Suggested Times for Fixation and Embedding of LacZ-Stained Embryos Embryo stage Treatment

E9.5

4% PFA postfixation O/N 1X PBS × 3 10 min Distilled H2O × 2 10 min 50% Ethanol 30 min 70% Ethanol × 2 30 min 95% Ethanol 30 min 100% Ethanol × 2 30 min Histoclear × 3 5 min Histoclear/wax 10 min Wax × 3 5–10 min

5. 6. 7.

8.

E10.5

E11.5

E12.5 E13.5–14.5 E15.5–18.5

O/N 10 min 10 min 30 min 30 min 30 min 30 min 10 min 20 min 20 min

O/N 2d 10 min 20 min 10 min 20 min 60 min 2h 60 min 2h 60 min 2h 60 min 2 h-O/N 10 min 15 min 20 min 30 min 20 min 30 min

2d 20 min 20 min 2h 2h 2h 2 h-O/N 15 min 60 min 60 min

3d 30 min 30 min 2h 2h 2h 2 h-O/N 20 min 90 min 90+ min

staining period the embryos/tissues should be stained dark blue at sites of β-gal expression, but with no staining elsewhere (i.e., minimal background; Fig. 1). Negative controls should always be included (see Note 6). Wash the embryos three times in PBS at room temperature. Postfix specimens in 4% PFA overnight or longer (Table 2). Embryos can be analyzed as whole mounts (Fig. 1A) or processed and embedded as for normal histology, although solvents can partially dissolve the LacZ precipitate. Therefore, stained tissues should be paraffin embedded and sectioned using minimum necessary times in solvents (Table 2). Sections (8–10 µm) should be cut for optimum visualization of β-gal staining (Fig. 1B). Tissue sections can then be counterstained (see Note 7).

3.2. β -Galactosidase Staining of Tissue Sections Perfusion fixation and cryostat sectioning is optimal for large pieces of tissue, as penetration of fix and staining reagents will be reduced in large specimens and paraffin embedding destroys β-gal activity. All steps take place at room temperature unless otherwise stated. 1. Wash embryos in PBS (see Note 1) three times at room temperature to remove any media or serum. 2. Fix embryos in cold LacZ fixative (see Note 2) at 4°C for 5 to 90 min (Table 1). 3. Wash embryos in PBS, three times for 20 min each at room temperature. 4. Equilibrate in 15% sucrose/PBS for 10 min to 1 h at 4°C. 5. Equilibrate in 30% sucrose/PBS until the tissue drops at 4°C (may take several hours). 6. Equilibrate further in Tissue-Tek (OCT) for 1 h at 4°C. 7. Embed in OCT (see Note 8). 8. Store blocks at −80°C in an air-tight container.


7

9. When ready to section, place block in cryostat chamber for approx 20 min. 10. Section cryostat blocks at 5 to 10 µm onto poly-lysine- or TESPA-coated slides (see Note 4). 11. Air-dry slides. 12. Store slides at −80°C in an air-tight slide box with desiccant. 13. To stain sections, take slide box out of freezer, and equilibrate to room temperature before opening box. Unused slides may be placed back in the freezer in an air-tight slide box with desiccant, although avoid freeze-thawing if possible. 14. Fix sections to slides in 0.2% glutaraldehyde in PBS for 5 to 10 min at 4°C. 15. Wash slides in LacZ wash solution, three times, for 5 min each at room temperature. 16. Stain the slides in the LacZ staining solution for the required period at the required temperature in the dark. This stage will need optimizing. Control (LacZ-negative) tissue should be used each time to control for endogenous β-gal (see Note 6). 17. Wash the slides in PBS, three times at room temperature. 18. Postfix in 4% PFA for 10 min, and wash with 1X PBS. Counterstaining may now be performed (see Note 7).

3.3. β -Galactosidase and Horseradish Peroxidase Double Staining This protocol can be used to carry out immunohistochemical staining for specific antigens on embryos that have already been stained for β-gal and then have been postfixed, embedded in paraffin wax, and sectioned onto slides at 8 to 10 µm (as in Subheading 3.1.). The indigo product of X-gal absorbs in the wavelengths emitted by the standard fluorescent-conjugated antibodies and is so dark that it can obscure the products from either horseradish peroxidase or alkaline phosphatase-conjugated antibodies. However, by carefully controlling the reaction time in LacZ staining solution so that only a small amount of indigo is produced (10), by having β-gal localized to the nucleus when the cellular antigen is nonnuclear (11), or by using antibodies to detect both β-gal and the cellular antigen, one can overcome these problems. All steps take place at room temperature unless otherwise stated. 1. 2. 3. 4. 5. 6. 7. 8. 9.

Place the slides in Histoclear for 5 min, twice, to dewax the sections. Remove the Histoclear with two rinses in 100% ethanol for 2 min each. Dehydrate through an alcohol series (90%, 70%, 50% ethanol) for 2 min each. Rinse in TBS for 5 min. Inactivate endogenous peroxidase in the tissue with 3% H2O2 in deionized water, for 5 min. Rinse in deionized water. If antigen retrieval is required for the unmasking of specific epitopes (see Note 9). Wash slides three times in TBS-Tx for 5 min each time. Block nonspecific antigens in the tissue sections in 10% FCS in TBS-Tx for 30 min. Place slides in a humid chamber. Add the primary antibody diluted in TBS-Tx containing 2% FCS. Add 80 to 100 µL per slide (to cover all the tissue sections), and cover with a Parafilm cover slip. Leave overnight at 4°C on a flat surface.

8


10. Equilibrate the slides back to room temperature for 15 min, and then wash three times in TBS-Tx for 5 min each. 11. Dilute the appropriate secondary antibody in TBS-Tx containing 2% FCS. Add 80 to 100 µL per slide, cover with a Parafilm cover slip, and leave at room temperature for 1 to 2 h. 12. Make AB Complex 30 min before use (according to the manufacturer’s instructions) and store at room temperature. 13. Wash slides three times in TBS-Tx for 5 min each. 14. Add AB Complex and leave for 30 min. Do not cover slides. 15. Rinse slides three times in TBS-Tx for 5 min each. Make up DAB solution according to the manufacturer’s instructions. 16. Add DAB solution and leave for up to 30 min until turns brown (see Note 10). 17. Stop reaction by rinsing well in PBS. 18. Counterstain as necessary. Choosing the counterstain can be problematic, as after LacZ and DAB staining, red in addition to blue stains can be confusing. We have had the best success with a light aqueous eosin stain in which the pale pink contrasts with the brick red/brown of the DAB staining (see Note 7).

3.4. β -Galactosidase Staining of Cultured Cells This protocol can be used on cell lines expressing LacZ or on primary cell lines isolated from LacZ-expressing tissues. For best results, cells should be grown on coated circular cover slips in 24-well plates. The cover slips may be coated using a variety of substances, such as fibronectin and collagen; the choice is cell type dependent. All steps are carried out in the 24-well plates using the same solutions as used for staining of cryosections and whole embryos. The cover slips are removed from the plates for photography. All steps take place at room temperature unless otherwise stated. 1. 2. 3. 4.

Remove the growth media from the cells and wash well with PBS at 4°C. Fix the cultured cells with in LacZ fixing solution for 2 min at 4°C. Wash the cells three times with wash buffer, for 2 min each. Replace the wash buffer with staining solution and wrap the plate in aluminum foil. Incubate at 37°C. The time of incubation varies greatly and should be optimized (see Note 6). 5. Once LacZ staining has developed, remove the staining solution, wash the cells twice with wash buffer, and carefully remove the cover slips from the 24-well plate using fine forceps or a hooked needle. 6. Mount the cover slip and photograph the staining using phase contrast microscopy.

3.5. Genotyping by β -Galactosidase Staining This is a quick and easy protocol for genotyping embryos that express LacZ. However, care must be taken to stain a piece of embryonic or adult tissue (for example, the yolk sac, an ear clip, or a tail tip) that expresses the LacZ product.


9

This can be a problem when tissue-specific promoters are used to drive LacZ expression. 1. 2. 3. 4. 5.

Place tissue piece in PBS in a 96-well plate. Fix for 15 min (0.2% glutaraldehyde in PBS) at room temperature. Wash three times in PBS for 5 min each. Stain with LacZ staining solution at 37°C in the dark (as in Subheading 3.1.). After 5 to 30 min, staining should be visible.

4. Notes 1. PBS must be calcium-free, pH 7.5. PBS tablets can be used. pH 7.5 is optimal, as endogenous β-gal (mammalian) is optimal at acidic pH, whereas the E. coli β-gal is optimal at neutral to slightly alkaline pH. 2. The authors have used two different types of fixatives. As well as the fixative described in Subheading 2.1., the same ingredients can also be made up in phosphate buffer to aid stability. Phosphate buffer: (0.1 M phosphate buffer, pH 7.5: 115 mL 0.1 M NaH2PO4 + 385 mL 0.1 M Na2HPO4). A mixture of glutaraldehyde/ PFA (Caution: both toxic) is often used for fixation, or 0.2% glutaraldehyde alone. Fixation with glutaraldehyde gives superior staining compared with PFA fixation alone. However, glutaraldehyde may also preserve endogenous enzyme activity, leading to nonspecific staining. 3. Although a stock solution can be made of potassium ferrocyanide and potassium ferricyanide and kept in the dark at 4°C, these solutions do go off with time (turn grayish yellow and grayish orange, respectively). The concentration used can vary, but usually 5 to 10 mM is used. Caution: these salts are also harmful if inhaled. The ferricyanide and ferrocyanide can form a blue precipitate (Prussian blue) upon reaction with free ferric ion. Therefore, do not use metal forceps to manipulate the tissue while it is in the LacZ staining solution. If a purple precipitate is preferred, nitroblue tetrazolium (NBT) salt can be added to the X-gal reaction in place of iron. This may be a faster and more sensitive reaction. A stock solution of NBT can be prepared by adding 50 mg NBT to 1 mL of 70% dimethylformamide and stored at −20°C. The final working concentrations of NBT are 0.25 to 1.0 mg/mL. Phenazine methosulfate (PMS) can also be added in conjunction with NBT to increase the reaction rate further. As PMS is very unstable, prepare a 100X stock of 2 mg/mL in H2O and use immediately. However, the authors have no personal experience with this variation to the X-gal reaction mix. 4. To coat slides with TESPA (also called APES), rack the slides and place overnight in 2% Decon (a detergent) to clean the slides. Rinse under hot tap water for 1 h, and then briefly wash in distilled water. Dry the slides at 60°C (overnight if convenient). Next day, place the slides in troughs for 5 min each with 2% TESPA in 100% EtOH, followed by two washes of 100% EtOH and one wash of 100% acetone. Once the slides are air-dried, they are ready to use. 5. Larger specimens may need dissection to allow full penetration of staining solutions. This may involve removal of tissues that are not of interest, or isolation of organs of interest. In older fetuses, removal of the skin will aid penetration.

10


6. The LacZ staining step can vary depending on the degree of expression. With embryos, low levels of expression may need an overnight step at 37°C. However, strong levels of expression may be seen in a few hours. Background staining can also be reduced if staining is performed at lower temperatures (20–30°C). With slides, staining can usually be seen after 3 to 6 h. The gut and kidney usually have particularly high levels of endogenous β-gal. However, to a degree all tissues have endogenous, lysosomal β-gal. The optimal pH for lysosomal β-gal is very low (acidic), and thus it is not very active at pH 7.5. However, some tissues also have a cytosolic form of β-gal, which may be active enough to give confounding results. Background may be reduced by varying the fixative type, length of time in fixative, pH of the buffer, and amount of time in the staining solution (12). If, after varying these conditions, background is still a problem, the LacZ enzyme can be detected by immunohistochemistry using commercially available antibodies. 7. Avoid counterstains that stain tissue blue, such as toluidine blue. Hematoxylin is also best avoided, as some tissues stain blueish. Counterstaining can be performed with stains such as 1% aqueous eosin (from seconds to minutes, depending on intensity required), which stains practically all cytoplasmic and intercellular substances a reddish pink color but does not stain the nucleus. Staining times may also vary depending on the tissue density and type. 8. There are several methods for embedding tissue samples in OCT. For larger specimens for which correct orientation is crucial, we usually place the sample in OCT in a plastic mold, orient the tissue/embryo with needles, and then freeze on dry ice (with the bucket lid on). For small samples or when orientation is not important, melting isopentane can be used. This involves a shallow dish with isopentane in the bottom (enough to surround a mold, but not to submerge). Place level on dry ice. Allow to freeze, and then remove from dry ice. As it starts to melt, place mold with tissue in OCT in isopentane, until OCT freezes. 9. For antigen retrieval: place slides into a plastic rack, place in a large beaker, and cover with 10 mM sodium citrate buffer (ix), pH 6.0. Cover the beaker with cling film and pierce for vent. Microwave on high for 10 min (check level of buffer during this step and top up if required). Allow to cool for 5 min and then wash in cool tap water for 10 min. Proceed to step 7 of Subheading 3.3. Other methods for antigen retrieval, such as trypsin treatment, may also be possible, although these have not been verified by the authors. 10. DAB can be precipitated very quickly in some cases (less than 1 min) but in other cases may take as long as 20 min. There is little advantage to leaving slides longer than this, as the enzymatic reaction is almost complete after this period. Staining of larger specimens can be seen with the naked eye. For smaller specimens, or when staining is expected in small areas or isolated cells, develop under a microscope.

References 1. Beckwith, J. R. (2005) Lac: The genetic system, in The Operon (Miller, J. H. and Reznikoff, W. S., eds.), Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.


11

2. Loughna, S., Bennett, P., Gau, G., Nicolaides, K., Blunt, S., and Moore, G. (1993) Overexpression of esterase D in kidney from trisomy 13 fetuses. Am. J. Hum. Genet. 53, 810–816. 3. Horwitz, J. P., Chua, J., Curby, R. J., et al. (1964) Substrates for cytochemical demonstration of enzyme activity. I. Some substituted 3-indolyl-beta-D-glycopyranosides. J. Med. Chem. 53, 574–575. 4. Davies, J. and Jacob, F. (1968) Genetic mapping of the regulator and operator genes of the lac operon. J. Mol. Biol. 36, 413–417. 5. Cotson, S. and Holt, S. J. (1958) Studies in enzyme cytochemistry. IV. Kinetics of aerial oxidation of indoxyl and some of its halogen derivatives. Proc. R. Soc. Lond B Biol. Sci. 148, 506–519. 6. Lojda, Z. (1970) Indigogenic methods for glycosidases. I. An improved method for beta-D-glucosidase and its application to localization studies on intestinal and renal enzymes. Histochemie 22, 347–361. 7. Soriano, P. (1999) Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat. Genet. 21, 70–71. 8. Danielian, P. S., Muccino, D., Rowitch, D. H., Michael, S. K., and McMahon, A. P. (1998) Modification of gene activity in mouse embryos in utero by a tamoxifeninducible form of Cre recombinase. Curr. Biol. 8, 1323–1326. 9. Jiang, X., Rowitch, D. H., Soriano, P., McMahon, A. P., and Sucov, H. M. (2000) Fate of the mammalian cardiac neural crest. Development 127, 1607–1616. 10. Vaysse, P. J. and Goldman, J. E. (1990) A clonal analysis of glial lineages in neonatal forebrain development in vitro. Neuron 5, 227–235. 11. Bonnerot, C., Rocancourt, D., Briand, P., Grimber, G., and Nicolas, J. F. (1987) A beta-galactosidase hybrid protein targeted to nuclei as a marker for developmental studies. Proc. Natl. Acad. Sci. USA 84, 6795–6799. 12. Rosenberg, W. S., Breakefield, X. O., deAntonio, C., and Isacson, O. (1992) Authentic and artifactual detection of the E. coli lacZ gene product in the rat brain by histochemical methods. Brain Res. Mol. Brain Res. 16, 311–315.

Immunofluorescent Staining Method for GFP and β-Gal

13

2 Immunohistochemical Detection of β -Galactosidase or Green Fluorescent Protein on Tissue Sections Philip A. Seymour and Maike Sander Summary With the recent advances in mouse genetics, it is now possible to mark specific cell types genetically in vivo and to study the fate of cells during development and adulthood. Cells are labeled and followed in vivo through the stable expression of reporter genes in particular cell types. The two most commonly used reporter genes are LacZ, which encodes the enzyme β-galactosidase (β-gal), and green fluorescent protein (GFP). β-Gal expression can be detected enzymatically, using 5-bromo-4-chloro-3-indolyl-β-Dgalactopyranoside (X-gal) as a substrate, and GFP can be directly visualized by fluorescence microscopy. However, with single detection of β-gal or GFP, it is often impossible to determine whether expression of the reporter protein is restricted to a particular cell type. To ascertain the identity of individual cells within a multicellular tissue, β-gal or GFP proteins must be visualized in conjunction with additional cellular markers. For such experiments, specific antibodies raised against β-gal or GFP can be used in coimmunofluorescence analyses. Such double-staining analyses on tissue sections are a powerful tool to study transgene expression or to trace cells in multicellular tissues. Key Words: Reporter gene; mouse; green fluorescent protein; LacZ; β-galactosidase; immunofluorescence; immunohistochemistry; fluorochrome; antibody.

1. Introduction Through the insertion of reporter genes into the genome, individual cells can be visualized in a given tissue. In mice, reporter genes are commonly used to monitor the expression of transgenes or to study the expression pattern of individual genes during development and in adulthood. In recent years, binary genetic systems, which are largely based on Cre/loxP-mediated recombination, have been developed to mark and trace cells genetically in vivo by activating expression of a reporter gene (1–5). Reporter genes that are commonly used to label cells genetically include the LacZ gene, alkaline phosphatase, luciferase, and From: Methods in Molecular Biology, vol. 411: Reporter Genes: A Practical Guide Edited by: D. Anson © Humana Press Inc., Totowa, NJ

13

14

Seymour and Sander

fluorochromes (6). The optimal choice of reporter gene depends on the intended application. The LacZ gene, which encodes β-galactosidase (β-gal), allows for detection of expressing cells through a simple tissue stain with 5-bromo-4chloro-3-indolyl-β-D-galactopyranoside (X-gal) as a substrate. X-Gal staining is highly sensitive and robust and can be performed either on whole-mount specimens or on tissue sections (7). The unique feature of fluorochromes is their ability to fluoresce when excited under a fluorescence microscope. Because of this feature, fluorochromes are widely used to visualize gene expression in living cells (8). A limitation to the use of fluorochromes in direct cell imaging is that their fluorescence is diminished by tissue fixation. To allow for the visualization of fluorochromes on fixed tissue, specific antibodies against green fluorescent protein (GFP) have been developed that can be used in indirect immunofluorescence or immunoperoxidase staining on tissue sections. A limitation of immunoperoxidase staining is that it does not allow for simultaneous detection of two antigens within one cell; therefore colocalization studies are precluded. Since staining for βgal or GFP in conjunction with additional antigens is commonly used to determine which cell type expresses the reporter gene, we have focused this chapter on immunofluorescence staining. We describe a double-immunofluorescence protocol that allows for simultaneous detection of β-gal or GFP with additional cellular markers on tissue sections. Additional information about the described protocols can be found in Current Protocols in Molecular Biology (9). 2. Materials 2.1. Fixation of Tissue 1. 1X phosphate-buffered saline (PBS): 137 mM NaCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, 2.7 mM KCl. Adjust to pH 7.4 with HCl. Store at room temperature. Can be prepared as a 10X stock. 2. 4% Paraformaldehyde in PBS: a. Dissolve paraformaldehyde (Fisher) in 1X PBS in a Pyrex container with a stir bar (4 g to 100 mL for 4% solution). b. Add a few drops of NaOH and heat in a hood (keep bottle cap loose) at 60°C to dissolve. c. Cool to room temperature on ice and adjust pH to 7.4. d. Filter to remove any undissolved powder. e. Make fresh prior to use; aliquots can be stored at −20°C.

2.2. Preparation of Cryoprotected Tissue Sections 1. Sucrose (Fisher) in PBS: prepare 10%, 15%, and 30% (w/v) solution in 1X PBS. Make fresh prior to use. Can be stored at 4°C for up to 1 d. 2. Tissue-Tek® O.C.T. (Optimal Cutting Temperature) Compound (Sakura Finetek, Torrance, CA).

Immunofluorescent Staining Method for GFP and β-Gal 3. 4. 5. 6.

15

Peel-A-Way® embedding molds (Polysciences, Warrington, PA). SuperFrost/Plus glass slides (Microm, Walldorf, Germany). Humidity Sponge™ silica desiccant sachets (Fisher). Polypropylene microscope slide box (Fisher).

2.3. Immunofluorescence Detection of GFP or β -Gal on Tissue Sections 1. Wheaton glass 20-slide staining dish with removable rack (Fisher). 2. Shandon Coverplate™ and Shandon Sequenza® Slide Rack (Thermo Electron, Waltham, MA). 3. Phosphate-buffered saline/Tween (PBST): prepare a 0.1% (v/v) solution of Tween20 (Sigma) in 1X PBS. Store at room temperature. 4. Blocking solution: 1% (v/v) normal goat serum (NGS) in PBST. Store 1 mL stock aliquots at −20°C; 1% NGS in PBST can be stored at 4°C for up to 1 wk (see Note 1). 5. Primary antibodies: rabbit anti-β-gal (MP Biomedicals, Irvine, CA; formerly ICN, cat. no. 55976, lot 03660) and rabbit anti-GFP (Invitrogen [Molecular Probes, cat. no. A6455], Carlsbad, CA) (see Note 2). Store small aliquots at −80°C. Once thawed, antibodies should be stored at 4°C. 6. Secondary antibody: cyanine (Cy3)-conjugated goat anti-rabbit IgG (H+L) (Jackson ImmunoResearch, West Grove, PA) (see Note 3). Store small (5–10 µL) aliquots in a 1:1 dilution with glycerol at −80°C. Once thawed, antibodies should be stored at 4°C. Avoid exposure of fluorophore-conjugated antibodies to direct overhead light, as this will induce photobleaching. 7. Nuclear stain: 300 nM DAPI (4.6-diamidino-2-phenylindole) in 1X PBS. 8. Mounting medium: Vectashield® (Vector, Burlingame, CA) or Aqua Poly/Mount (Polysciences). 9. Fisherbrand microscope coverslips (Fisher). 10. Nail polish. 11. Microscope slide folder (Fisher).

3. Methods Both β-gal and GFP can be readily detected by indirect immunofluorescence on frozen tissue sections. In our hands, detection on paraffin-embedded tissue is far less reliable. Unless directed to the nucleus through the addition of a nuclear localization signal, β-gal and GFP are normally localized to the cytoplasm. When localized to the cytoplasm, immunofluorescence detection of β-gal or GFP on frozen sections does not require additional procedures to unmask the epitopes. However, for nuclear detection of β-gal or GFP, epitope unmasking procedures often improve the intensity of the signal (see Note 4; 10–12). It should be recognized that any antigen unmasking procedure may introduce artifactual falsepositive staining. In all staining experiments, a negative control slide should be included, on which preimmune serum is applied instead of the primary antibody. If preimmune serum is not available, blocking solution can be used instead. It is also advisable to include a positive control side.

16

Seymour and Sander

In fluorescence microscopy of tissue sections, non-fluorophore-labeled cells are not readily visible under the microscope. A commonly used method to visualize all nuclei in a tissue section is to counterstain with DAPI, which is excited at 358 nm (blue emission). With the appropriate filters, DAPI can be observed on the same tissue section together with fluorochromes that emit in the red (e.g., Cy3) or green spectrum (e.g., fluorescein). Secondary antibodies with different emission spectra are commonly used to detect two different antigens simultaneously (see Notes 5 and 6). Double-labeling experiments are helpful for determining which cell populations express GFP or β-gal. The two different antigens are best distinguished by fluorochromes that emit in the green and red spectrum, respectively. Importantly, because of its more intense fluorescent emission, the Cy3-conjugated antibody should be used to stain the antigen that is more difficult to detect (see Note 7). For the same reason, Cy3 is the preferred fluorochrome for detection of the sole target antigen in single-labeling experiments. 3.1. Fixation of Tissue 1. Chill freshly prepared 4% paraformaldehyde in PBS on ice and transfer to appropriate sealable containers (e.g., screw-top glass or plastic tubes). 2. After the animal is sacrificed, transfer the tissue quickly into the fixative. Dissections of tissue under the microscope should be performed in ice-cold PBS prior to fixation. To ensure proper fixation, add at least 10 times the volume of fixative to the tissue sample. Fix the tissue under gentle agitation at 4°C. Fixation time depends on the size of the tissue (see Note 8).

3.2. Preparation of Cryoprotected Tissue Sections 1. After fixation, discard paraformaldehyde into a hazardous waste container, and wash the tissue three times for 15 min each in 1X PBS at 4°C, again with gentle agitation throughout. 2. Transfer samples into prechilled 10% sucrose in PBS and leave at 4°C until the tissue has sunk to the bottom of the tube. Larger samples can also be left in 10% sucrose in PBS at 4°C overnight. Subsequently transfer the tissue to ice-cold 15% sucrose in PBS and again leave at 4°C until the tissue has sunk to the bottom of the tube. Repeat the same steps in 30% sucrose in PBS. 3. Mix Tissue-Tek O.C.T. compound 1:1 (v/v) with 30% sucrose in PBS and leave tissue in the O.C.T./sucrose mixture for 30 min at room temperature under gentle agitation. 4. Transfer tissue to Tissue-Tek O.C.T. compound and gently agitate for 30 min at room temperature. 5. Place the embedding molds in an ethanol/dry ice bath, or place on a slab of dry ice dampened with ethanol to promote evaporation and further cooling. Fill the molds with Tissue-Tek O.C.T. compound and transfer tissue into the mold. Ensure that


6.

7. 8.

9.

10.

11.

17

no residual sucrose solution is surrounding the tissue. Leave the samples in the ethanol/dry ice bath until the O.C.T. compound is completely frozen. Then, transfer samples to the −80°C freezer for storage. Cut sections from the cryoblocks on a cryostat (e.g., Leica CM3050 S). Precool the chamber of the cryostat to the desired temperature for the tissue to be sectioned, and place the chuck (onto which the cryoblock will be mounted) inside. Cutting temperature is dependent on the fat content of the tissue sample; for most tissues, the temperature should be around −18°C to −20°C. Transfer the cryoblock(s) to the cryostat chamber on dry ice and allow 30 min to equilibrate to the chamber temperature. Cover the chuck with O.C.T. and place inside the chamber. Remove the cryoblock from the mold by splitting the mold at its corners, and orient it so that the desired section plane will be obtained. When the clear O.C.T. begins to turn opaque white in colour, press the cryoblock onto the chuck and apply pressure for 10 to 20 s to freeze the cryoblock securely onto the chuck. Allow a further 30 min for equilibration of the sample to the chamber temperature. Clamp the chuck into the specimen holder. Use the “trim” function of the cryostat to advance the block until the specimen can be seen through the O.C.T. With a new, precooled razor blade, trim the cryoblock face to give a 5- to 10-mm border around the specimen. Set the section thickness: 10-µm sections are typical, although thinner sections will give better resolution of cell layers when viewed microscopically. a. If a particular tissue needs to be identified within the sectioning block (e.g., within an embryo), view cut sections periodically under a binocular dissecting microscope until the desired tissue appears. b. Cut serial sections and mount on glass slides in “sets” of 5 or 10 slides as required for later staining. This method allows adjacent serial sections to be easily stained for different antigens by staining individual whole slides with distinct antisera. c. To prepare sets of five slides, for example with three sections per slide, hold the first slide at the frosted end and mount the first section at the “top,” or opposite end of the slide. d. Mount the successive four sections at the top ends of the next four slides in the set. e. Mount the next five sections centrally on the same five slides, followed by the successive five sections on the “bottom” position of each of those five slides. f. Repeat this process for further sets of five slides from the same specimen until the tissue is exhausted. g. A border of 5 mm should be left around each edge of the slide when mounting to allow for an unstained area owing to the interface with the staining coverplates. Leave freshly cut sections for between 1 and 6 h to air-dry at room temperature before storage. Slides can either be stained immediately or stored frozen. For storage, care should be taken to minimize water ingress, as this will lead to tissue damage. Place slides into a labeled polypropylene microscope slide box containing a

18

Seymour and Sander Humidity Sponge™ silica desiccant sachet (Fisher) and seal the box inside a Ziploc® bag for storage at −80°C.

3.3. Immunofluorescence Detection of GFP or β -Gal on Tissue Sections 1. When you are retrieving slides from the −80°C freezer for staining, temporarily place boxes on dry ice to avoid freeze-thaw damage to sections. Then, dry the removed slides at room temperature for approx 30 min. Subsequently, place slides in a slide holder and transfer to a glass staining dish with 1X PBS. Put a stir bar on the bottom of the glass dish and wash for 5 min at room temperature to dissolve the O.C.T. compound. 2. Apply 500 µL of 1X PBS to each Shandon Coverplate and slowly lower the slide, tissue-side down, onto the coverplate. Avoid the generation of air bubbles between the tissue and the coverplate. Transfer the coverplate and slide into the cassette slots in the Shandon Sequenza slide rack. The slides remain in the slide rack until they are mounted and coverslipped (see Note 9). Pipet 200 µL of fresh, prechilled blocking solution (1% NGS in PBST) into the cavity between the slide and the coverplate, and leave the samples at room temperature for a minimum of 30 min. 3. Dilute the primary antibody (anti-GFP 1:2000; anti-β-gal 1:500) in blocking solution. 4. Pipet 200 µL of primary antibody in 1% NGS/PBST into the coverplate cavity. Incubate overnight at 4°C. 5. Wash slides by applying 2 mL of 1X PBS into the coverplate cavity. Leave at room temperature for 5 min. Repeat the washing step twice for a total of three washes in 1X PBS. 6. Dilute the secondary antibody to a final dilution of 1:2000 in blocking solution (1% NGS in PBST). The secondary antibody is freshly prepared for each experiment. When exposed to light, photobleaching decreases the staining intensity of fluorochrome-conjugated antibodies. Therefore, samples should be protected from light exposure by placing the cassettes or slides in the dark during all following steps. Samples can be easily protected by covering the containers with aluminum foil. All subsequent steps should also be performed in the absence of direct overhead illumination. 7. Pipet 200 µL of diluted secondary antibody in 1% NGS/PBST into the coverplate cavity. Incubate for 1 h at room temperature. 8. Wash slides three times for 5 min each by applying 2 mL of 1X PBS as in step 5. 9. Once the final wash is completed, take each slide and its corresponding coverplate out of the slide rack together. To remove the coverplate, place the slide and coverplate gently into a 500-mL beaker containing 1X PBS, and allow the coverplate to float away from the slide. Forceful removal of the coverplate will damage the sections. 10. In preparation for reuse, first wash the coverplates under running double-distilled water and then immerse briefly in 70% ethanol before leaving them to air-dry. 11. The slides are then ready to be mounted in aqueous mounting medium. Although these can be prepared in the lab, it is often more efficient to purchase commercially available products (e.g. Vectashield or Aqua Poly/Mount) owing to the time-expense of preparation.


19

a. Apply a few drops of aqueous mounting medium onto the tissue sections and carefully invert the coverslip into the mounting medium. b. Remove air bubbles by gently pressing down on the coverslip. If a large number of air bubbles obscure any of the sections, then float the coverslip off the slide by immersing the newly mounted slide vertically in the beaker of 1X PBS and gently shaking until the coverslip slides off the end of the slide. The slide can then be remounted usually with no detriment to the sections. c. To seal the coverslip onto the slide, carefully apply nail polish to the corners of the coverslip. Mounting, as with step 6 and onward, should be performed in an area without direct overhead illumination. The samples can be viewed immediately after the nail polish has dried. d. For storage, lay the slides flat in a microscope slide folder or slide box and protect them from light. When slides are stored at 4°C, fluorescence is stable for several weeks. For longer term storage, slides should be placed at −20°C. 12. View the slides under a fluorescence microscope. Excitation at 543 nm induces Cy3 fluorescence, while excitation at 358 nm induces DAPI fluorescence. To gain better resolution or to analyze coexpression of two proteins in the same cell, imaging under a confocal microscope is superior to a fluorescence microscope. Images for each fluorochrome are taken separately and can be overlaid with the appropriate software (e.g., Photoshop, Adobe, San Jose, CA). Examples of immunofluorescence staining for β-gal and GFP are shown in Fig. 1.

4. Notes 1. For detection of GFP and β-gal with secondary antibodies raised in goat, blocking with 1% NGS in PBST produces high intensity staining with minimal background. In double-labeling experiments, detection of the second antigen sometimes requires use of an alternative blocking solution (e.g., when the primary antibody is produced in goat). A blocking solution that works very well with many antibodies is 1% IgGfree bovine serum albumin (BSA; lyophilized powder; Sigma) in PBST. When 1% IgG-free BSA in PBST is used for blocking, the primary and secondary antibodies should also be diluted in this solution. 2. We have found that these two antibodies work best in immunofluorescence on tissue sections. We tested numerous competitive reagents from other commercial sources but did not obtain comparable results. 3. In double-labeling experiments, the two antigens are commonly detected by use of a Cy3-conjugated secondary antibody in conjunction with a secondary antibody that emits in the green spectrum, such as Alexa Fluor 488-conjugated antibodies (Invitrogen [Molecular Probes], Carlsbad, CA). Like fluorescein, Alexa Fluor 488 fluorescent dye is excited at 495 nm but provides superior photostability. When you are choosing secondary antibodies for multiple labeling, both secondary antibodies should be derived from the same host species so they do not recognize one another. One should also ensure that the secondary antibodies do not cross-react with immunoglobulins from other species possibly present in the assay system, or with endogenous immunoglobulins possibly present in the tissues under investigation.

20

Seymour and Sander

Fig. 1. Immunofluorescence detection of β-galactosidase (β-gal) and green fluorescent protein (GFP) on pancreatic tissue sections from transgenic mice. (A–C) Transgene expression in pancreas from Pdx1-Nkx6.1-IRESLacZ transgenic embryos at embryonic day (E) 14.5 is identified by coimmunofluorescence staining with a rabbit anti-β-gal and a guinea pig anti-Pdx1 antibody. The anti-β-gal and anti-Pdx1 antibodies are detected with an Alexa 488-conjugated goat anti-rabbit IgG and a Cy3-conjugated goat anti-guinea pig IgG antibody, respectively. Since the LacZ transgene carries a nuclear localization signal, β-gal colocalizes with the transcription factor Pdx1 in the nucleus. (D–F) Transgene expression in pancreas from Ngn3-Nkx6.1-IRESeGFP transgenic embryos at E16.5 is identified by coimmunofluorescence staining with a rabbit anti-GFP and a guinea pig anti-Ngn3 antibody. The anti-GFP and anti-Ngn3 antibodies are detected with an Alexa 488-conjugated goat anti-rabbit IgG and a Cy3-conjugated goat anti-guinea pig IgG antibody, respectively. The transcription factor Ngn3 is localized to the nucleus, whereas GFP is detected in the cytoplasm. In coexpressing cells, a red nucleus is surrounded by green cytoplasm. See accompanying CD for color version. 4. Because antigens are frequently masked by the fixation methods used to prepare tissues for staining, detection of some antigens, in particular nuclear antigens, requires antigen retrieval or unmasking methods. Such methods usually rely on the use of slightly acidic solutions and heat treatments in various combinations to catalyze the hydrolysis of formaldehyde-induced crosslinks between neighboring


21

proteins. It should be noted that although the advantages of antigen unmasking on paraffin sections are well accepted, the efficacy of unmasking on frozen sections proves to be a matter of some contention. However, we find that antigen unmasking on frozen sections is beneficial to the staining of many nuclear antigens. Since each antibody can produce highly variable staining results contingent on the antigen retrieval procedure used, it is advisable to perform a pilot staining experiment using more than one method to optimize the protocol. Antigen retrieval is performed after the initial washing step with 1X PBS before application of the blocking solution. a. Leave slides in the glass dish, and incubate in citrate buffer (to make the 200 mL required to immerse the slides fully in the staining dish, add 98 mL doubledistilled water and 83 mL 0.1 M sodium citrate to 19 mL 0.1 M citric acid) for 1 h at 37°C. b. Seal the dish and lid with Parafilm to prevent evaporation. c. Wash in 1X PBS three times for 5 min in the glass dish. d. An alternative, harsher method is to microwave the slides in the glass dish containing 10 mM sodium citrate, pH 6.0, for 7 min at 30% power so the liquid is gently boiling. The dish should be completely filled with liquid to prevent the sections from drying out owing to evaporation. e. Slides should be left to cool at room temperature for 45 min before being washed in 1X PBS as described. f. Antigen unmasking can be used in conjunction with an additional permeabilization step to facilitate antibody access to the cell interior. g. To do this, transfer slides to Shandon Sequenza slide rack and add 1 mL of 0.15% Triton X-100 in 1X PBS to each coverplate. h. Incubate for 1 h at room temperature. i. Proceed directly with the blocking step. 5. In double-staining experiments, both primary and secondary antibodies are normally applied at the same time. In some instances, however, simultaneous application of both primary antibodies impairs the results of the immunostaining. In such cases, staining for both antigens should be performed sequentially, beginning with the antigen that is more difficult to detect. The Shandon Sequenza slide rack should be kept at 4°C following the first staining round until application of the second primary antibody. Usually, it is unnecessary to perform a second blocking step prior to the second staining round. 6. When mouse primary monoclonal antibodies are used to stain mouse tissue, high background staining often obscures the specific staining. This is because of the inability of the anti-mouse secondary antibody to distinguish between the mouse primary antibody and endogenous mouse immunoglobulins in the tissue. This background problem can be significantly reduced by using the Vector M.O.M. immunodetection kit. 7. For antigens that are difficult to detect, tyramide signal amplification™ (Invitrogen [Molecular Probes], Carlsbad, CA) can significantly enhance the signal. However, signal amplification often also increases background.

22

Seymour and Sander

8. Since antigens are sensitive to fixation, overfixation tends to impair immunodetection. Fixation times need to be optimized for each tissue, ranging from 45 min to overnight. 9. An alternative to working with the Shandon Coverplates and cassette system is the use of a hydrophobic PAP-Pen (DAKO, Carpinteria, CA) in conjunction with a humid chamber. Individual tissue sections are circled with the PAP-Pen, and a drop of 100 to 200 µL solution is applied to each section. The Shandon system has the advantages that the tissue quality is consistently maintained over the course of the immunostaining procedure and that less antibody is needed.

Acknowledgments The authors thank Dr. Shelley B. Nelson for contributing the images. We would also like to thank Dr. Christopher V.E. Wright for providing the anti-Pdx1 antibody. This work was supported by NIH/NIDDK-1R01-DK68471-01, NIH/ NIDDK-1U19-DK072495-01, the American Diabetes Association and the Juvenile Diabetes Research Foundation. References 1. Branda, C. S. and Dymecki, S. M. (2004) Talking about a revolution: the impact of site-specific recombinases on genetic analyses in mice. Dev. Cell 6, 7–28. 2. Soriano, P. (1999) Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat. Genet. 21, 70–71. 3. Lobe, C. G., Koop, K. E., Kreppner, W., Lomeli, H., Gertsenstein, M., and Nagy, A. (1999) Z/AP, a double reporter for cre-mediated recombination. Dev. Biol. 208, 281-292. 4. Novak, A., Guo, C., Yang, W., Nagy, A., and Lobe, C. G. (2000) Z/EG, a double reporter mouse line that expresses enhanced green fluorescent protein upon Cremediated excision. Genesis 28, 147–155. 5. Mao, X., Fujiwara, Y., Chapdelaine, A., Yang, H., and Orkin, S. H. (2001) Activation of EGFP expression by Cre-mediated excision in a new ROSA26 reporter mouse strain. Blood 97, 324–326. 6. Hadjantonakis, A. K., Dickinson, M. E., Fraser, S. E., and Papaioannou, V. E. (2003) Technicolour transgenics: imaging tools for functional genomics in the mouse. Nat. Rev. Genet. 4, 613–625. 7. Mombaerts, P., Wang, F., Dulac, C., et al. (1996) Visualizing an olfactory sensory map. Cell 87, 675–686. 8. Giepmans, B. N., Adams, S. R., Ellisman, M. H., and Tsien, R. Y. (2006) The fluorescent toolbox for assessing protein location and function. Science 312, 217–224. 9. Ausubel, F., Brent, R., Kingston, R. E., et al., eds. (1987) Current Protocols in Molecular Biology, vol. 3, John Wiley & Sons, New York, 14.1.1–14.6.8. 10. Brown, R. W. and Chirala, R. (1995) Utility of microwave-citrate antigen retrieval in diagnostic immunohistochemistry. Mod. Pathol. 8, 515–520.


23

11. Shi, S. R., Cote, R. J., and Taylor, C. R. (1997) Antigen retrieval immunohistochemistry: past, present, and future. J. Histochem. Cytochem. 45, 327–343. 12. Shi, S. R., Chaiwun, B., Young, L., Cote, R. J., and Taylor, C. R. (1993) Antigen retrieval technique utilizing citrate buffer or urea solution for immunohistochemical demonstration of androgen receptor in formalin-fixed paraffin sections. J. Histochem. Cytochem. 41, 1599–1604.

Detection of Reporter Gene Expression In Vivo

25

3 Detection of Reporter Gene Expression in Murine Airways Maria Limberis, Peter Bell, and James M. Wilson Summary We have shown that to overcome the low levels of expression from gene transfer vector-mediated β-galactosidase expression in lung, it is essential to replace the cytoplasmic β-galactosidase gene with a nuclear targeted β-galactosidase gene. We found that lung should be sectioned and fixed prior to staining for β-galactosidase expression and that en bloc staining of intact lung is inefficient at staining positively transduced cells located deeper in the lung spaces. For GFP fluorescence, it is important to inflate the lungs with fixative prior to freezing and subsequent sectioning. For processing of nasal tissues for β-galactosidase expression, we expand on a protocol used in previously reported gene transfer studies. Key Words: LacZ, β-galactosidase; GFP; placental alkaline phosphatase; airway; epithelium; nose; lung; gene expression.

1. Introduction Airway-directed gene transfer has emerged as a promising curative approach for the treatment of genetic lung diseases such as cystic fibrosis and α1-antitrypsin deficiency (1). In theory, delivery of a normal copy of the therapeutic gene to the defective airway epithelium will restore normal airway function. Gene transfer vectors for the treatment of lung diseases are initially evaluated in the airways of small animal models including mice. Therefore, it becomes important that detection of reporter transgene expression not be compromised by applying universal detection protocols that are suitable to other tissues. We found that sectioning of cryopreserved tissues followed by fixation with 0.5% glutaraldehyde/DPBS offers sensitive detection of β-galactosidase (βgal) gene expression (2). The detection efficiency of β-galactosidase-express-

From: Methods in Molecular Biology, vol. 411: Reporter Genes: A Practical Guide Edited by: D. Anson © Humana Press Inc., Totowa, NJ

25

26

Limberis, Bell, and Wilson

ing cells was greatly improved by using nuclear targeted β-gal as a transgene and maintaining a pH of 7.0 for the working 5-bromo-4-chloro-3-indoyl-β-Dgalactoside (X-gal) solution (2). En bloc staining of intact lung was inefficient at staining positively transduced cells located deeper in the lung (2), presumably owing to the poor diffusion of the staining solution throughout the tissue. To preserve green fluorescence protein (GFP) expression from the positively transduced cells the tissue had to be inflated with fixative followed by cryosectioning. Staining of nasal airway tissues for β-gal expression was performed based on previously reported protocols (3,4). 2. Materials 2.1. Collection of Airway Tissues 1. Dulbecco’s phosphate-buffered saline (without calcium and magnesium) (DPBS; Cellgro, Mediatech, Herndon, VA). Store at room temperature. 2. Optimal cutting temperature (OCT) freezing compound Tissue-Tek® (Fisher). Store at room temperature. 3. Inflation solution: prepare by shaking well equal parts of DPBS with OCT. Allow the solution to settle; the inflation solution should be free of air bubbles prior to use and kept at room temperature. 4. 2% (v/v) Paraformaldehyde (PFA) (16% [w/v] stock; Electron Microscopy Sciences, Hatfield, PA)/0.5% (v/v) glutaraldehyde (25% [w/v] stock; Electron Microscopy Sciences) in DPBS. Make fresh as required. 5. Formalin: 10% neutral buffered formalin (NBF) (Fisher). Store at room temperature. 6. Isopentane (methylbutane) (Fisher). Extremely volatile. 7. Cryomolds (Fisher). 8. Monoject Blunt needles (Sherwood Medical, St. Louis, MO). 9. Syringes (Fisher).

2.2. Sectioning Lung Tissues 1. Fisherbrand Superfrost Plus disposable microscope slides (Fisher). 2. Fisherbrand cover glasses (square). 3. Cryostat: used to section the cryorpreserved lung tissues.

2.3. Sectioning Nasal Tissues 1. Demineralizing solution (1X): 7% (v/v) HCl in 1.5% (w/v) EDTA. Carefully dissolve EDTA in H2O on a stirring hot plate in a fume hood (do not allow solution to exceed 60°C). When EDTA is completely dissolved, the solution should appear cloudy. Carefully add conc. HCl and continue to stir until solution turns clear. Caution: The demineralizing solution is highly toxic and should only be opened in a fume hood (see Note 1). Stable at room temperature for 1 mo. 2. Carnoy’s fixative: 3:1 of 95% ethanol to glacial acetic acid. Store at room temperature and use only in a fume hood. 3. Microtome: used to section paraffin-embedded tissues.


27

4. Fisherbrand Unisette Biopsy processing/embedding cassette with Lid (Fisher).

2.4. Histochemical Staining for Reporter Gene Expression in Lung 2.4.1. X-Gal Staining for β-Gal Expression 1. 1 mM MgCl2 in DPBS. Store at room temperature. 2. 0.5% (v/v) Glutaraldehyde in DPBS. Make fresh as required. 3. Pre X-gal solution: 5 mM potassium ferrocyanide (K4Fe(CN)6; Sigma, St. Louis, MO), 5 mM potassium ferricyanide (K3Fe(CN)6; Sigma) and 1 mM MgCl2 (Sigma) in 1X DPBS. Solution is light sensitive and is stable at room temperature for 1 mo. 4. X-gal (Fisher). Dissolve 40 mg/mL X-gal in dimethylformamide. Aliquot in 500-µL lots, and store at −20°C for 3 mo. Solution is light sensitive. 5. Working X-gal solution: make fresh as required by adding X-gal stock solution (1:40) to Pre-X-gal and mixing well. If necessary, adjust pH (with HCl) to 7.0 (see Note 2). Working X-gal solution is light sensitive and should be used within 1 h. 6. Glass staining dishes (Fisher).

2.4.2. NBT/BCIP Staining for Placental Alkaline Phosphatase Expression 1. 1 mM MgCl2 in DPBS. Store at room temperature. 2. 0.5% (v/v) glutaraldehyde in DPBS. Stable at 4°C for 1 mo. 3. NBT/BCIP (nitroblue tetrazolium chloride/5-bromo-4-chloro-3-indolyl-phosphate, toluidine salt) working solution: dissolve 1 ready-to-use tablet (Roche, Indianapolis, IN) in 10 mL of distilled H2O. Solution is light sensitive and should be made fresh as required. A dark blue color precipitate indicates sites of alkaline phosphatase enzyme activity. Incubation times vary; for highest sensitivity, the sections should be stained overnight. 4. Glass staining dishes (Fisher).

2.5. Counterstaining Tissues 1. 2. 3. 4. 5. 6. 7. 8. 9.

70% Ethanol (Fisher). Store at room temperature. 95% Ethanol (Fisher). Store at room temperature. 100% Ethanol (Fisher). Store at room temperature. Xylene (Fisher). Store at room temperature and handle in fume hood. Nuclear Fast Red (NFR pre-made solution, Vector). For long-term storage, keep at 4°C. NFR can be reused. Harris Hematoxylin #7211 (Richard Allen Scientific). Clarifier (Clarifier 2, Richard Allen Scientific). Eosin Y (Richard Allen Scientific). Permount mounting medium (Fisher).

2.6. Analysis of GFP 1. Vectashield mounting medium with DAPI (Vector).

28


3. Methods 3.1. Collection of Airway Tissues 3.1.1. Collection of Lungs 1. Anesthetize mice weighing approx 25 g by an intraperitoneal injection of a mixture of ketamine/xylazine in DPBS (70/7 mg/kg). 2. Bleed the heart transdiaphragmatically to minimize blood spillage into the airway lumen, as blood interferes with the histochemical staining (5). 3. Expose the ventrum of the thorax and neck. a. Cut along the diaphragm and then cut along both sides of the ribs and remove the rib cage. b. Dissect the cervical musculature to expose the trachea. c. Using a 27-g needle, make a small hole between the cartilage rings and insert a blunt needle attached to a 5-mL syringe that contains inflation solution. d. Inflate the lung while it is still inside the rib cage to avoid overinflation. Lungs are adequately inflated when the inflation solution reaches the margins of all lung lobes and the lungs fill the chest cavity. e. Immediately remove the needle, and grasp the trachea with forceps to keep the lumen closed. f. Cut the trachea proximal to the forceps using scissors. g. Holding the trachea, remove the entire pluck by cutting the esophagus and aorta at the diaphragm. h. Using fine-tip forceps and dissection scissors remove the entire heart. i. Place the lungs into a cryomold in the preferred orientation, and cover completely in OCT. 4. Use two containers, of which the outer one holds liquid nitrogen, and the inner one, preferably a metal beaker, is filled with isopentane. Place the container with isopentane into the liquid nitrogen until the isopentane starts to freeze on the wall of the inner container. 5. With a pair of long forceps, carefully drop the whole cryomold with the OCTembedded lungs into the cold isopentane. Leave for approx 10 to 30 s (until completely frozen), and then remove the sample and place on dry ice until all tissues are collected. 6. Store cryomolds at −80°C until sectioned.

3.1.2. Collection of Heads 1. Proceed with steps 1 and 2 in Subheading 3.1.1. 2. Separate the head from the carcass and remove the skin and eyes using fine-tip forceps and scissors. 3. Snip the soft portion of the nose tip with fine scissors and flush the head with 2% (v/v) PFA/0.5% (v/v) glutaraldehyde in DPBS via the tracheal remnant. Keep on ice for 2 h.


29

3.2. Sectioning Lung Tissues 1. Remove lung tissues from −80°C and equilibrate to −25°C. 2. Using the cryostat, section lung tissues at 8–10 µm. 3. Store slides at −80°C until time for staining. If slides are stored in a box, remove box from −80°C and do not open until box has reached room temperature.

3.3. Sectioning Nasal Tissues 1. Decalcify the heads in demineralizing solution for 22 h in a fume hood. Do not tighten caps, as the decalcification process releases toxic fumes. 2. Decant the demineralizing solution in the fume hood, and wash softened heads under running H2O for 30 min (see Note 3). 3. Place heads in 70% ethanol for long-term storage or until paraffin embedding. 4. Remove the jaw and tongue. 5. Mark the right side with a strip of black ink, which is made permanent by dabbing the marked head region in Carnoy’s fixative solution. 6. The softened heads are sectioned with disposable microtome blades in three standard sections (4). These sections are taken at level 6 (immediately posterior to the dorsal incisor), level 16 (where the two nasal airways coalesce into the nasopharyngeal duct), and level 24 (at the rear of the head) (6). 7. Place the three cross-sections in sectioning cassettes, anterior face down, and embed in paraffin.

3.4. Histochemical Staining for Reporter Gene Expression in Lung 3.4.1. X-Gal Staining for β-Gal Expression 3.4.1.1. LUNG 1. Slides should be at room temperature. 2. Fix lung sections in a glass staining dish that contains 0.5% (v/v) glutaraldehyde/ DPBS (see Note 4) (cooled to 4°C) for 10 min. 3. Remove slide and immediately blot excess fixative from the sides of the slide. (Do not allow slide to dry.) 4. Wash slide(s) in a glass staining dish containing 1 mM MgCl2 in DPBS for 15 min. 5. Repeat step 4. 6. Remove slide and immediately blot excess wash solution from the sides of the slide. (Do not allow slide to dry.) 7. Stain tissues for β-gal expression by placing them in a glass staining dish containing working X-gal solution (see Note 5). Protect from light and incubate at 37°C (see Note 6) from 30 min to 16 h. A blue color precipitate indicates sites of enzyme activity. Incubation times may vary; for highest sensitivity, the sections should be stained overnight (approx 16 h). 3.4.1.2. NOSE 1. Flush the head with 1 mM MgCl2/DPBS through the tracheal remnant and place in 1 mM MgCl2/DPBS. Incubate on ice for 15 min.

30


2. Repeat step 1. 3. Stain heads for β-gal expression by placing them in polystyrene containers containing working X-gal solution (approx 8 mL/head). Protect from light and incubate at 37°C for 8 to 16 h with intermittent flushing of the nasal cavity though the tracheal remnant. (Solution should come out of the nose.) 4. Wash the heads in 0.9% (w/v) NaCl/H2O for 30 min at room temperature. 5. Postfix the heads in 10% (v/v) NBF for 24 h at room temperature. (Keep in fume hood.)

3.4.2. NBT/BCIP Staining for Placental Alkaline Phosphatase Expression 1. Warm a glass staining dish containing 1 mM MgCl2/DPBS to 65°C (see Note 7). 2. Incubate the glass slide with the sectioned lung tissue in the heated 1 mM MgCl2/ DPBS solution for 30 min. 3. Remove the slide and immediately blot excess solution from the sides of the slide. (Do not allow slide to dry.) 4. Stain tissues for placental alkaline phosphatase gene expression by placing them in a glass staining dish containing working NBT/BCIP solution. Protect from light and incubate at room temperature. Incubation times may vary; for highest sensitivity, the sections should be stained overnight (approx 16 h).

3.5. Counterstaining Tissues 3.5.1. Counterstaining Cryopreserved Tissues with NFR 1. Remove the slide holder and immediately blot excess stain solution from the sides of the slide holder. (Do not allow slide to dry.) 2. Rinse slide holder in PBS (at room temperature). 3. Rinse slide holder in H2O (at room temperature). 4. Blot excess H2O from the sides of the slide holder, place in a glass staining dish that contains NFR (at room temperature), and incubate for 3 to 5 min (see Note 8). 5. Rinse slide holder in H2O (at room temperature). 6. Dehydrate lung tissues by placing the slide holder through a graded ethanol series starting once at 70% ethanol, twice at 95% ethanol, twice at 100% ethanol, and three times with xylene for 1 min each. (Blot excess solution from slide holder in between steps.) 7. Cover slip with Permount mounting medium and allow to dry overnight.

3.5.2. Counterstaining Cryopreserved Tissues with Hematoxylin and Eosin 1. Proceed with steps 1 to 3 of Subheading 3.5.1. 2. Blot excess H2O from the sides of the slide holder and place in a glass staining dish that contains hematoxylin (at room temperature) and incubate for 1 min. 3. Rinse slide holder in warm H2O (37°C) for 1 min. 4. Blot excess H2O from the sides of the slide holder and place in a glass staining dish that contains Clarifier solution for 10 s.


31

5. Rinse slide holder in H2O (at room temperature) for 10 s. 6. Blot excess H2O from the sides of the slide holder and place in a glass staining dish that contains Blueing Reagent for 10 s. 7. Rinse slide holder in H2O for 10 s. 8. Blot excess H2O from the sides of the slide holder and place in a glass staining dish that contains Eosin Y for 10 s. 9. Rinse slide holder in H2O for 3 s. 10. Dehydrate tissues by placing the slide holder through a graded ethanol series starting twice at 70% ethanol, twice at 95% ethanol, twice at 100% ethanol, and three times with xylene for 1 min each. 11. Cover slip with Permount mounting medium and allow to dry overnight.

3.5.3. Counterstaining Paraffin-Embedded Tissues with NFR 1. Place slide holder with tissue in xylene for 5 min at room temperature (in a fume hood). 2. Repeat step 1 with fresh xylene solution. 3. Remove slide holder and immediately blot excess xylene solution from the sides of the slide holder (do not allow slide to dry) and add to xylene:ethanol (1:1 [v/v]) solution for 1 min at room temperature. 4. Place the slide holder through a graded ethanol series starting twice at 100% ethanol, once at 95% ethanol, and once at 70% ethanol for 1 min each. 5. Rinse slide holder in H2O (at room temperature). 6. Repeat step 5. 7. Blot excess H2O from the sides of the slide holder and place in a glass staining dish that contains NFR (at room temperature) and incubate for 3 to 5 min. 8. Rinse slide holder in H2O for 10 s. 9. Proceed to steps 12 and 13 of Subheading 3.5.2. 10. Cover slip with Permount mounting medium and allow to dry overnight.

3.5.4. Counterstaining Paraffin-Embedded Tissue with Hematoxylin and Eosin 1. Place slide holder with tissue in xylene for 5 min at room temperature (in a fume hood). 2. Repeat step 1 with fresh xylene solution. 3. Remove slide holder and immediately blot excess xylene solution from the sides of the slide holder (do not allow slide to dry) and add to xylene:ethanol (1:1 [v/v]) solution for 1 min at room temperature. 4. Place the slide holder through a graded ethanol series starting twice at 100% ethanol, once at 95% ethanol, and once at 70% ethanol for 1 min each. 5. Rinse slide holder in H2O (at room temperature). 6. Repeat step 5. 7. Blot excess H2O from the sides of the slide holder, place in a glass staining dish that contains hematoxylin (at room temperature), and incubate for 2 min.

32


8. Rinse slide holder in warm H2O (37°C) for 1 min. 9. Blot excess H2O from the sides of the slide holder and place in a glass staining dish that contains Clarifier solution for 1 min. 10. Rinse slide holder in H2O for 10 s. 11. Blot excess H2O from the sides of the slide holder and place in a glass staining dish that contains Blueing Reagent for 1 min. 12. Rinse slide holder in H2O for 10 s. 13. Blot excess H2O from the sides of the slide holder and place in a glass staining dish that contains Eosin Y for 10 s. 14. Rinse slide holder in H2O for 3 s. 15. Proceed to steps 12 and 13 of Subheading 3.5.2. 16. Cover slip with Permount mounting medium and allow to dry overnight.

3.6. Cryopreserving Green Fluorescence Protein-Expressing Lung 1. Follow steps 1 and 2 of Subheading 3.1.1. 2. Expose the ventrum of thorax and neck. a. Cut along the diaphragm and then cut along both sides of the ribs and remove the rib cage. b. Dissect the cervical musculature to expose the trachea. c. Using a 27-g needle, make a small hole between the cartilage rings and insert a blunt needle attached to a 5-mL syringe that contains 10% NBF (see Note 9). d. Inflate the lung while it is still inside the rib cage to avoid over inflation. Lungs are adequately inflated when the 10% NBF solution reaches the margins of all lung lobes and the lungs fill the chest cavity. e. Immediately remove needle, and suture close the trachea. f. Cut the trachea proximal to the suture using scissors. g. Place the lungs into a polystyrene container filled with 10% NBF. h. Protect from light and incubate overnight on a shaker on a low setting. 3. Wash lung with DPBS for 2 h at room temperature. Protect from light. 4. Repeat step 3 twice. 5. Blot the lungs dry using 3MM Whatman paper (Fisher) and place into a cryomold. 6. Cover lung with OCT and freeze as described in Subheading 3.1.1., steps 4 and 5. 7. Prepare cryosections at 8–10 µm and collect on slides. 8. Cover slip immediately with Vectrashield (containing DAPI to visualize cell nuclei).

4. Notes 1. Crystals may form in the demineralizing solution. The solution can be filtered occasionally, although this step is not necessary. 2. The β-gal signal increased when the pH was lowered from 7.5 to 6.5 and decreased slightly at pH 6.0. However, from pH 6.0 to 6.5, nonspecific staining was obtained that was not present at (and higher than) pH 7.0. Therefore, a pH of 7.0 appears to provide a good tradeoff between staining sensitivity and the elimination of falsepositive signals.


33

3. Tape uncovered specimen pots with autoclave tape in a cross formation so the tissues stay in their correct pots. 4. Other aldehyde-based fixatives such as 10% NBF or PFA resulted in slightly weaker but still acceptable staining reactions with X-gal. However, Streck tissue fixative (S.T.F.) inactivated the enzyme activity completely. Solvents such as methanol or acetone, used at −20°C for 10 min, weakened β-gal activity, resulting in relatively weak and diffuse staining. 5. An alternative substrate for β-gal staining is Bluo-gal, which has been reported to have a lower diffusion rate and to provide a “crisper” color precipitate than X-gal (7). Owing to its higher electron density, Bluo-gal has also been used for the detection of β-gal activity by electron microscopy (e.g., refs. 8–10). We found that this substrate results in the generation of small speck-like precipitates that are nonspecific. 6. Avoid cell culture incubators, because the high CO2 concentration can cause a drop in pH. 7. This step eliminates endogenous alkaline phosphatase activities but does not affect the heat-resistant human placental alkaline phosphatase. 8. Counterstaining with NFR or hematoxylin and eosin should be monitored since the aim is to stain the lung tissue only lightly to allow the quantitation of β-gal-expressing cells. 9. It is important to fix the whole lungs before freezing and sectioning in order to preserve the GFP fluorescence. Sections from unfixed tissues, even when fixed after sectioning, show a highly diffuse fluorescence, which makes it almost impossible to localize GFP-expressing cells.

References 1. Crystal, R. G. (1992) Gene therapy strategies for pulmonary disease. Am. J. Med. 92, 44S–52S. 2. Bell, P., Limberis, M., Gao, G., et al. (2005) An optimized protocol for detection of E. coli beta-galactosidase in lung tissue following gene transfer. Histochem. Cell. Biol. 124, 77–85. 3. Parsons, D. W., Grubb, B. R., Johnson, L. G., and Boucher, R. C. (1998) Enhanced in vivo airway gene transfer via transient modification of host barrier properties with a surface-active agent. Hum. Gene Ther. 9, 2661–2672. 4. Limberis, M., Anson, D. S., Fuller, M., and Parsons, D. W. (2002) Recovery of airway cystic fibrosis transmembrane conductance regulator function in mice with cystic fibrosis after single-dose lentivirus-mediated gene transfer. Hum. Gene Ther. 13, 1961–1970. 5. Johnson, L. G., Olsen, J. C., Naldini, L., and Boucher, R. C. (2000) Pseudotyped human lentiviral vector-mediated gene transfer to airway epithelia in vivo. Gene Ther. 7, 568–574. 6. Mery, S., Gross, E. A., Joyner, D. R., Godo, M., and Morgan, K. T. (1994) Nasal diagrams: a tool for recording the distribution of nasal lesions in rats and mice. Toxicol. Pathol. 22, 353–372.

34


7. Weis, J., Fine, S. M., David, C., Savarirayan, S., and Sanes, J. R. (1991) Integration site-dependent expression of a transgene reveals specialized features of cells associated with neuromuscular junctions. J. Cell Biol. 113, 1385–1397. 8. Loewy, A. D., Bridgman, P. C., and Mettenleiter, T. C. (1991) Beta-galactosidase expressing recombinant pseudorabies virus for light and electron microscopic study of transneuronally labeled CNS neurons. Brain Res. 555, 346–352. 9. Blanks, J. C., Spee, C., Barron, E., Rich, K. A., and Schmidt, S. (1997) Lineage study of degenerating photoreceptor cells in the rd mouse retina. Curr. Eye Res. 16, 733–737. 10. Sekerkova, G., Katarova, Z., Joo, F., Wolff, J. R., Prodan, S., and Szabo, G. (1997) Visualization of beta-galactosidase by enzyme and immunohistochemistry in the olfactory bulb of transgenic mice carrying the LacZ transgene. J. Histochem. Cytochem. 45, 1147–1155.

3D Analysis of Molecular Signals

35

4 Three-Dimensional Analysis of Molecular Signals with Episcopic Imaging Techniques Wolfgang J. Weninger and Timothy J. Mohun Summary This chapter describes two episcopic imaging methods, episcopic fluorescence image capturing (EFIC) and high-resolution episcopic microscopy (HREM). These allow analysis of molecular signals in a wide variety of biological samples such as tissues or embryos, in their precise anatomical and histological context. Both methods are designed to work with histologically prepared and whole-mount stained material, and both provide highresolution data sets that lend themselves to 3D visualization and modeling. Specimens are embedded in wax (EFIC) or resin (HREM) and sectioned on a microtome. During the sectioning process, a series of digital images of each freshly cut block surface is captured, using a microscope and CCD camera aligned with the position at which the microtome block holder comes to rest after each cutting cycle. The resulting stacks of serial images retain virtually exact alignment and are readily converted to volume data sets. The two methods differ in how tissue architecture is visualized and hence how specific molecular signals are detected. EFIC uses endogenous, broad-range, tissue autofluorescence to reveal specimen structure. Addition of dyes to the wax embedding medium suppresses detection of any signal except that originating from the block surface. EFIC can be used to detect specific signals (such as LacZ) by virtue of their ability to suppress such fluorescence. In contrast, the plastic embedding medium used in HREM is strongly fluorescent, and tissue architecture is detected at the surface because of the ability of cellular and subcellular structures to suppress this signal. Specific signals generated as a result of chromogenic reactions can be visualized using band-pass filters that suppress the appearance of morphological data. In both methods, the digital volume data show high contrast; for HREM, such data achieve true cellular resolution. Their intrinsic alignment greatly facilitates their use for 3D analysis of transgene activity that can be visualized in the context of complex cellular and tissue morphology. Both methods are relatively simple and can be set up using common laboratory apparatuses. Together, they provide powerful tools for analyzing gene function in embryogenesis or tissue remodeling and for investigating developmental malformations. Key Words: Gene expression; gene activity; RNA pattern; imaging; 3D analysis; embryogenesis; development; 3D reconstruction; episcopic microscopy; remodeling. From: Methods in Molecular Biology, vol. 411: Reporter Genes: A Practical Guide Edited by: D. Anson © Humana Press Inc., Totowa, NJ

35

36

Weninger and Mohun

1. Introduction The shaping of an embryo is orchestrated by complex interactions of thousands of genes and their products that regulate changing patterns of cell type differentiation, cellular morphology, and tissue interactions throughout development. By unraveling this complex 4D developmental program in experimental models such as the mouse, we will better understand the genesis of human birth defects, improve prenatal diagnosis, and lay the basis for the rational development of therapies. A prerequisite is the acquisition of detailed knowledge about the function of proteins and the regulation of genes that encode them. Three main approaches are commonly taken for studying gene function in normal and abnormal embryogenesis. First, the roles played by individual genes can be analyzed by examining the consequences of gene mutation or ablation on normal embryo morphology. Such mutants are commonly obtained using genetic engineering to target the gene of interest but can also be obtained as individual genetic lines obtained from random mutagenesis screens. A broad range of methods can be used to analyze the phenotypes generated in such mutants, such as computed tomography (CT), magnetic resonance imaging (MRI), ultrasound (US), optical projection tomography (OPT), electron microscopy, confocal imaging, histological sections, and episcopic imaging (1–19). A second approach is to study the expression of individual genes and gene products in the developing embryo and correlate this spatiotemporal pattern with the changing morphologies of tissues and organs. As long as the gene transcript or protein product can be labeled in a specific manner, its distribution can be examined using a variety of imaging techniques (e.g., RNA and immunohistochemistry, OPT, confocal imaging, or episcopic imaging methods). In both of these approaches, gene function or expression is directly correlated with embryo morphology, allowing inferences to be made about the role of individual genes on tissue or organ development. A third and more systematic approach is to analyze changing patterns of gene activity using cDNA microarray techniques, which facilitate the simultaneous study of many thousands of genes (20–24). This approach cannot provide the spatial resolution possible with the imaging techniques available with the first two methods, since it depends on extraction of RNA from isolated cells or pieces of tissue. It cannot, for example, support the analysis of gene function or expression in individual cell lineages, nor can it examine the expression of gene products rather than gene transcripts themselves. In this chapter we will discuss two episcopic imaging methods, episcopic fluorescence image capturing (EFIC) and high-resolution episcopic microscopy (HREM). These methods provide an imaging procedure useful for comparing normal and abnormal embryo morphology as well as for studying the topology


37

of individual gene expression. Both provide highly detailed images of tissue and organ architecture upon which can be superimposed a similarly precise visualization of molecular signals reflecting gene expression. By providing precisely aligned, serial images captured during sample sectioning, the data readily provide spatial visualization of gene expression in its accurate histological and anatomical context (16,17). Of course, these methods are not restricted to embryological studies and can equally be used to study growth, maintenance, or remodeling in postnatal organs or adult tissues. EFIC and HREM are both destructive imaging techniques, utilizing fixed tissue specimens progressively sectioned on a microtome. Sample fixation is therefore common. In each method, successive episcopic images (i.e., images of the block surface) are captured using similar data acquisition procedures. However, the two methods differ in the embedding medium used, necessitating individual protocols for each, and the characteristics of the images they yield are radically different. As a result, each is best suited for a distinct method of data processing. EFIC produces “negative images” since it detects tissue autofluorescence, and background fluorescence from the embedding medium is suppressed. Its ability to detect molecular signals is also a negative contrast technique, since it relies on the extinction of autofluorescence by the LacZ chromogenic precipitate. In contrast, HREM provides “positive images,” since tissue is visualized by its ability to reduce the high level of fluorescence from the plastic embedding medium. Its ability to detect specific signals rests on the use of filters to reduce the signal:background ratio for tissue without a comparable reduction in the ratio obtained with a chromogenic reaction product. HREM is therefore a positive contrast technique, which provides information on molecular signals by detecting color reactions. Both techniques can be used with a broad range of biological specimens, ranging from embryos to adult tissues to biopsy material in a variety of species. In principle, imaging can be carried out in any resolution, this being limited only by the optics used for image capture and the resolution of the CCD camera used. In practice, the different methods by which tissue is visualized affect the effective resolution and the size of the specimen that is appropriate. In particular, since EFIC depends on endogenous autofluorescence of tissue, its ability to resolve different tissues and reveal architecture within tissues or organs depends on intrinsic variations in autofluorescence between cell types. For example, late-gestation mouse embryos show a wide range of autofluorescence levels in different tissues and yield complex images revealing considerable morphology. Early gestation embryos, however, show far more uniform and much lower levels of autofluorescence, presumably reflecting the much lower degree of

38

Weninger and Mohun

tissue specialization and tissue density in early embryos. Some individual cells can readily be detected by EFIC (for example, red blood cells, which are highly fluorescent); however, in general, the dispersion of fluorescent signal precludes effective imaging at cellular resolution within most tissues. Despite these limitations, the negative image character and the broad range of autofluorescence signals detected in EFIC data sets make them ideally suited for 3D volume rendering. EFIC therefore provides the simplest way to obtain 3D models of morphology. Unlike EFIC, HREM is capable of consistently providing cellular resolution, and indeed subcellular structures such as nuclei are the most prominent feature of HREM images. HREM imaging shows none of the signal dispersion found with EFIC at higher magnifications, and useful data can readily be obtained over a range from 1- to 200-fold magnification. Paradoxically, the wealth of structural information available from HREM limits the ease with which data sets can be used for 3D modeling, Volume rendering is possible but generally less effective, both because of the narrower range of grayscale values used to represent tissue structure and also because of the inherent complexity of cellular resolution data. However, HREM data are ideal for the more laborious methods of isosurface rendering after manual or semiautomatic segmentation. In each imaging method, embryo or tissue specimens are first used for labeling in a whole mount (RNA in situ hybridization or immunohistochemistry) and then embedded in an embedding medium of wax (EFIC) or resin (HREM); the resulting blocks are mounted and sectioned on a microtome. In order to capture perfectly aligned images of the block face, microscope optics and digital camera are positioned perpendicular to the block face at the stopping position of the block holder (Fig. 1). Appropriate emission and detection filter sets can be introduced into the light path, and sequential cycles of block cutting, filter changing, and image acquisition can be performed manually or automatically depending on the precise equipment. In either case, stacks comprising hundreds of precisely aligned digital images of freshly cut block surface can be generated within a few hours. Figure 2 outlines the steps and their order during episcopic data generation. Depending on the optical magnification and the CCD camera chip size, digital images with pixel sizes between 0.4 and 50 µm can be achieved in the image data. Because of their inherent alignment, captured image stacks can immediately be converted into volume data sets and used for precise 2D and 3D analysis. With minimal voxel sizes as small as 0.4 × 0.4 × 1.0 µm, cellular features, tissue architecture, and organ morphology can be visualized extremely precisely along with equally precise representations of specific gene or transgene expression (see, for example, http://www.univie.ac.at/efic and http://www.meduniwien. ac.at/3D-Rekonstr/HREM).


39

Fig. 1. Data capturing apparatuses. (A) Based on a sliding microtome. (B) Based on a rotary microtome.

2. Materials 2.1. Specimen Preparation 2.1.1. EFIC 1. 4% Paraformaldehyde in phosphate-buffered saline (PBS) or 10% phosphate-buffered formalin (Sigma Aldrich) (see Note 1). 2. Ethanols: 30%, 50%, 70%, and 100% (see Note 2).

40

Weninger and Mohun

Fig. 2. Steps in data generation and data analysis. 3. Histoclear. 4. Embedding wax comprising 20% Vybar, 76.5% paraffin wax, 3.5% stearic acid, and 0.35% Sudan IV aniline dye (Sigma Aldrich). The mixture is melted at 65°C and left overnight. Undissolved wax is then removed by filtration through paper at 65°C overnight. 5. Disposable embedding molds (Leica Microsystems) (see Note 3).

2.1.2. HREM 1. 2. 3. 4. 5. 6.

4% Paraformaldehyde in PBS. PBS, 0.1% (v/v) Tween 20: methanol mixes, 25% (v/v), 50% (v/v), and 75% (v/v). 100% Methanol. In situ hybridization solutions (e.g., see refs. 25 and 26). Ethanols: 30%, 50%, and 70%. Infiltration solution: 100 mL JB-4 Solution A (Polysciences, www.polysciences. com), mixed with 1.25 g Benzoyl Peroxide Plasticized (Catalyst) and 0.4 g eosin (Eosin, spritlöslich, Waldeck), stirred at 4°C. 7. Embedding molds and block holders (Leica Microsystems). 8. Embedding solution: 25 mL infiltration solution mixed with 1 mL of JB-4 Solution B.

2.2. Data Capturing 2.2.1. EFIC 1. Sliding microtome (e.g., Leica SM2500), equipped with a photo-stop position (see Note 4).


41

2. Dissecting microscope (MZ16 F/FA, Leica Microsystems), equipped with a GFP1 barrier filter set (peak excitation 425 nm, emission barrier 480 nm), mounted on an x-y micrometer stage above the photo-position of the microtome (see Note 5). 3. Cooled, grayscale CCD camera capable of detecting low-intensity fluorescence (e.g., Hamamatsu Orca ER). 4. Image acquisition software with appropriate camera driver and computer (e.g., Power Mac G5 and OpenLab [Improvision] or IPLab [Scanalytics] software) (see Note 6).

2.2.2. HREM 1. Rotary microtome (CUT 4060 E, microTec), specially adapted by the company to increase the accuracy of the block holder resting position. The microtome is placed on a motor-driven x-y stage (Walter Uhl, technische Mikroskopie). 2. Optics of a compound microscope (DM LM Head, Leica Microsystems), equipped with band-pass GFP (excitation 470/40 nm, emission 525/50 nm) and TX2 (excitation 560/40 nm, emission 645/75 nm) filter cubes. The optics are placed on a solid, z-axis adjustable stage (components from Stahlbau Reumüller and MikroscopieService). 3. Digital color video camera (Leica DFC 480) with a target size of 2560 × 1920 pixels. 4. PC, equipped with camera driver and data capturing routine (software assembling by M. Donoser, TU-Graz, A).

2.3. Data Processing and Visualization Any 32- or 64-bit computer with image manipulation and rendering software, for example, Power Mac G5 or Windows PC with Openlab (Improvision), Photoshop (Adobe Systems), and “Volocity” (Improvision) packages, or alternatively, a Linux-based workstation with Amira (Mercury Computer Systems) software. Data stacks are large, and 2 to 4 GB of RAM are essential. 3. Methods As with all histological methods, the optimum processing time depends on specimen size and tissue density. The protocols below are appropriate for “average” specimens (for EFIC: midgestation mouse embryos; for HREM specimens of 1 to 2 mm). Up to 10 specimens can be comfortably processed simultaneously. Small specimens such as young embryos can be processed more quickly; larger or denser specimens will require considerable increase in preparation and processing times. As a guide, embedding a single sample requires approximately 10 mL of wax (EFIC) or 2 mL of resin (HREM). 3.1. Specimen Preparation 1. Harvest specimens, transfer them into PBS, if necessary dissect them, and fix them overnight in 4% formaldehyde, buffered with PBS. Wash at 4°C in PBS, 0.1% (v/v) Tween 20 twice for 5 min.

42

Weninger and Mohun

2. Perform whole-mount staining. If working with specimens carrying a LacZ transgene, use X-gal staining for detection of transgene expression patterns. Use digoxigenin (DIG)-labeled RNA probes and the NBT/BCIP detection system for visualizing specific RNA expression patterns following a standard protocol appropriate for the sample (e.g., refs. 25 and 26). Stained specimens can be stored in PBS, 0.1% (v/v) Tween 20 for several months at 4°C. 3. Dehydrate specimens through a graded series of alcohols, e.g., 30%, 50%, and 70% ethanol using wash times appropriate for sample size and density. (One to 2 h each is generally adequate.) Samples are now ready for infiltration with embedding medium. They can be stored in 70% ethanol at 4°C prior to embedding, although prolonged storage prior to EFIC is not recommended.

3.2. Infiltration and Embedding 3.2.1. EFIC 1. Complete the dehydration of samples using sequential washes with 80%, 85%, 90%, 95%, and 100% ethanol (20–30 min each). a. Wash with Histoclear/ethanol (1:1 mix) briefly (5–20 min depending on specimen size) followed by Histoclear for a similar period and immediately transfer into wax dye mix maintained at 65°C. b. Wash twice for a minimum of 15 min with fresh wax and then infiltrate with wax for a minimum of 60 min (twice). c. For larger samples, infiltration times may need to be extended several fold, although vacuum embedding can increase infiltration rates. 2. Transfer samples into fresh wax mix in a disposable mold, and carefully alter specimen orientation at room temperature as the wax begins to set. Samples can be photographed at this stage to record their precise location within the block (see Note 3).

3.2.2. HREM 1. Use infiltration solution at 4°C under continuous rocking. Infiltrate the specimen for 1 h, change, and infiltrate overnight; change and infiltrate for an additional 1 to 2 h (see Note 7). 2. Prepare fresh embedding solution and embed the specimens (see Notes 8–10). a. Put the specimens into the embedding molds, fill them with embedding solution, and orient the specimens as the medium becomes “sticky”. b. Immediately put the block holder into the mold and fill the rest of the mold with embedding solution. c. Cover the mold tightly and store at 18°C overnight. d. Remove the blocks from the molds and keep them for 4 to 24 h at room temperature prior to data capture.

3.3. Data Capturing 1. Prepare the block for sectioning. For wax blocks (EFIC), trim using the grid lines on the image captured during embedding to ensure that the specimen is centered


2. 3. 4. 5. 6. 7.

8.

43

in the remaining wax. For plastic blocks (HREM), illuminate from the side and draw in the future field of view by projecting the position of the specimen inside the block onto the block surface (see Notes 12 and 13). Mount the blocks on the microtome and move it near the knife (see Notes 15 and 16). Choose an objective and the x-y position of the optics according to the position and the size of the desired field of view. Section until the entire block surface is freshly cut. Focus the optics on the block surface using the camera images. Select appropriate filter (for EFIC: GFP3; for HREM: GFP or TX2) (see Note 11). Select appropriate section thickness, adequate exposure times, number of images to be captured from each freshly cut block surface, and total number of images to be captured in this session, and commence image capturing. For HREM, switch between GFP and TX2 filter sets during each image capture cycle.

3.4. Data Processing and Visualization With either imaging procedure, a single data set obtained with the GFP filter captures morphology. For EFIC, this also contains the specific signal data since this is represented by localized regions of signal suppression. To assist in identifying such regions, it is helpful to compare data with an equivalent, unstained sample. For HREM, the specific signal is obtained from a second data set, captured with the TX2 filters. Further processing and analysis of data sets depends on the choice of software used and the visualization method required. 1. Commercial software tools (e.g., Amira or Volocity) or public domain software (e.g., ImageJ or EMAP software) permit rapid review of image stacks either in the original section plane or as virtual, resectioned image stacks. 2. For 3D modeling by volume rendering, HREM data sets require inversion of the gray LUT to transform images from “positive” to “negative”. Grayscale mapping within each data set can then be adjusted in order to obtain suitable contrast and background black levels in volume-rendered models. Visualization “inside” regions of the model or through appropriately positioned “windows” is easily achieved either by modeling of partial stacks or by modeling entire stacks in which an area of data has been removed on successive images. 3. Image editing software can be used to extract structures of interest, generate surfaced-rendered virtual 3D models, and visualize them simultaneously. This requires segmentation of data. The method chosen to do this depends on both the structure under analysis and the type of data set. Where the structure boundary is accurately defined by a clear transition in grayscale value, segmentation can be achieved using automated thresholding algorithms. This approach can be used with HREM data sets of molecular signals captured with the TX2 filter, since morphology data are suppressed under these conditions. Most morphological structures, however,

44

Weninger and Mohun must be extracted from GFP data sets set using semiautomatic thresholding or purely interactive (manual) contour finding.

4. Notes 1. The choice of fixative can affect the level of autofluorescent signal and the range of contrast between different tissues. Formaldehyde-based fixatives yield the strongest signal with the greatest contrast. They also show the brightest signal from blood cells. Bouin’s fixative will yield much less intense signals from blood, but the overall range of contrast between tissues is significantly reduced. 2. There is some evidence that prolonged storage of samples in ethanol prior to analysis leads to a loss of autofluorescent signal. Samples are probably better stored in fixative. 3. In order to locate samples within the wax block and to align the optics appropriately, it can be helpful to photograph the sample in the embedding mold while the wax remains molten. This is possible using transmitted light and a reference grid placed either below or within the base of the mold. 4. An economical alternative to the motorized SM2500 microtome is a manual sliding blade microtome (e.g., Leica SM2000R), since the position of the block face remains fixed in all three dimensions. Data acquisition is more laborious, since sectioning is manual. 5. For the motorized SM2500 microtome, the optics and x-y base stage must be mounted directly onto the blade holder assembly to ensure that the block face remains in focus after each section cycle. 6. Complete automation of sectioning and image capture can be achieved using the camera trigger signal from the SM2500 microtome to initiate image capture within the software package. 7. Because of eosin, the infiltration solution soon becomes viscous and hardens within a few days, even without resin solution B. It is therefore essential to keep the eosinstained solution A in the fridge and use freshly mixed solutions for infiltration and embedding. Penetration of the specimen by the infiltration solution is often inadequate if old solutions are used. 8. Cool the embedding solution outside the fridge during embedding. 9. Tightly cover the filled embedding molds with paraffin. This is essential for adequate hardening of the embedding solution. 10. Before sectioning, store the blocks for a few hours in the room in which it is sectioned. The embedding medium can then adapt to the humidity of the room, which prevents the block from breaking during sectioning. 11. As an alternative to the GFP filter set, the Leica YFP filter set (excitation filter 500 nm, emission filter 535 nm) can be used. This reduces excitation times. 12. Keep the light in the sectioning room dark and at constant intensity. This enhances the quality of the single images and ensures homogeneity of the image series. 13. Keep the temperature in the sectioning room low and constant. This avoids artifacts caused by heating of the block.


45

14. Avoid vibrations near the data capturing apparatus. 15. Test the orientation of the digital images. Some optical arrangements generate mirrored images, which must be flipped prior to reconstruction. 16. Use D-faceted tungsten knifes for sectioning. This minimizes scratching artifacts.

References 1. Schneider, J. E., Bamforth, S. D., Farthing, C. R., Clarke, K., Neubauer, S., and Bhattacharya, S. (2003) Rapid identification and 3D reconstruction of complex cardiac malformations in transgenic mouse embryos using fast gradient echo sequence magnetic resonance imaging. J. Mol. Cell. Cardiol. 35, 217–222. 2. Sharpe, J., Ahlgren, U., Perry, P., et al. (2002) Optical projection tomography as a tool for 3D microscopy and gene expression studies. Science 296, 541–545. 3. Denk, W. and Horstmann, H. (2004) Serial block-face scanning electron microscopy to reconstruct three-dimensional tissue nanostructure. PLoS Biol. 2, e329. 4. Effmann, E. L., Johnson, G. A., Smith, B. R., Talbott, G. A., and Cofer, G. (1988) Magnetic resonance microscopy of chick embryos in ovo. Teratology 38, 59–65. 5. Yu, Q., Shen, Y., Chatterjee, B., et al. (2004) ENU induced mutations causing congenital cardiovascular anomalies. Development 131, 6211–6223. 6. Shen, Y., Leatherbury, L., Rosenthal, J., et al. (2005) Cardiovascular phenotyping of fetal mice by noninvasive high-frequency ultrasound facilitates recovery of ENUinduced mutations causing congenital cardiac and extracardiac defects. Physiol. Genomics 24, 23–36. 7. Streicher, J., Weninger, W. J., and Muller, G. B. (1997) External marker-based automatic congruencing: a new method of 3D reconstruction from serial sections. Anat. Rec. 248, 583–602. 8. Schierlitz, L., Dumanli, H., Robinson, J. N., et al. (2001) Three-dimensional magnetic resonance imaging of fetal brains. Lancet 357, 1177–1178. 9. Ruijter, J. M., Soufan, A. T., Hagoort, J., and Moorman, A. F. (2004) Molecular imaging of the embryonic heart: fables and facts on 3D imaging of gene expression patterns. Birth Defects Res. C Embryo Today 72, 224–240. 10. Soufan, A. T., Ruijter, J. M., van den Hoff, M. J., de Boer, P. A., Hagoort, J., and Moorman, A. F. (2003) Three-dimensional reconstruction of gene expression patterns during cardiac development. Physiol. Genomics 13, 187–195. 11. Weninger, W. J., Streicher, J., and Müller, G. B. (1996) [3-Dimensional reconstruction of histological serial sections using a computer]. Wien Klin. Wochenschr. 108, 515–520. 12. Kaufman, M. H. and Richardson, L. (2005) 3D reconstruction of the vessels that enter the right atrium of the mouse heart at Theiler stage 20. Clin. Anat. 18, 27–38. 13. Louie, A. Y., Huber, M. M., Ahrens, E. T., et al. (2000) In vivo visualization of gene expression using magnetic resonance imaging. Nat. Biotechnol. 18, 321–325. 14. Ewald, A. J., McBride, H., Reddington, M., Fraser, S. E., and Kerschmann, R. (2002) Surface imaging microscopy, an automated method for visualizing whole embryo samples in three dimensions at high resolution. Dev. Dyn. 225, 369–375.

46

Weninger and Mohun

15. Weninger, W. J., Meng, S., Streicher, J., and Müller, G. B. (1998) A new episcopic method for rapid 3-D reconstruction: applications in anatomy and embryology. Anat. Embryol. (Berl.) 197, 341–348. 16. Weninger, W. J. and Mohun, T. (2002) Phenotyping transgenic embryos: a rapid 3-D screening method based on episcopic fluorescence image capturing. Nat. Genet. 30, 59–65. 17. Weninger, W. J., Geyer, S. H., Mohun, T. J., et al. (2006) High-resolution episcopic microscopy: a rapid technique for high detailed 3D analysis of gene activity in the context of tissue architecture and morphology. Anat. Embryol. (Berl.) 211, 213–221. 18. Kerwin, J., Scott, M., Sharpe, J., et al. (2004) 3-Dimensional modelling of early human brain development using optical projection tomography. BMC Neurosci. 5, 27. 19. Rosenthal, J., Mangal, V., Walker, D., Bennett, M., Mohun, T. J., and Lo, C. W. (2004) Rapid high resolution three dimensional reconstruction of embryos with episcopic fluorescence image capture. Birth Defects Res. C Embryo Today 72, 213–223. 20. Chen, H. W., Yu, S. L., Chen, W. J., et al. (2004) Dynamic changes of gene expression profiles during postnatal development of the heart in mice. Heart 90, 927– 934. 21. Kaynak, B., von Heydebreck, A., Mebus, S., et al. (2003) Genome-wide array analysis of normal and malformed human hearts. Circulation 107, 2467–2474. 22. Balza, R. O. Jr. and Misra, R. P. (2005) The role of serum response factor in regulating contractile apparatus gene expression and sarcomeric integrity in cardiomyocytes. J. Biol. Chem. 281, 6498–6510. 23. Napoli, C., Lerman, L. O., Sica, V., Lerman, A., Tajana, G., and de Nigris, F. (2003) Microarray analysis: a novel research tool for cardiovascular scientists and physicians. Heart 89, 597–604. 24. Alvarez, E., Zhou, W., Witta, S. E., and Freed, C. R. (2005) Characterization of the Bex gene family in humans, mice, and rats. Gene 357, 18–28. 25 Wilkinson, D. G. (1998) In Situ Hybridisation: A Practical Approach. Oxford University Press, Oxford. 26. Streit, A. and Stern, C.D. (2001) Combined whole-mount in situ hybridization and immunohistochemistry in avian embryos. Methods 23, 339–344.

Fluorescent Proteins for Cell Biology

47

5 Fluorescent Proteins for Cell Biology George H. Patterson Summary Through the use of exogenous labels, such as antibodies and synthetic fluorophores, experimenters have been able to readily observe the localization of proteins and organelles within a cell by fluorescence microscopy. The discovery and application of fluorescent proteins spanning a large wavelength range have revolutionized these studies. This chapter attempts to introduce the vast array of these molecules, discuss their characteristics, and assess the advantages and disadvantages that each displays for use in imaging. Key Words: GFP; DsRed; photoactivatation; imaging.

1. Why Use a Fluorescent Protein? With the number of tools available to the biologist for the study of his/her favorite gene or protein, why choose to work with a fluorescent protein? First, fluorescent proteins can be specifically fused to a protein of interest without the nonspecific labeling associated with many other techniques. This is made possible with the straightforward methods of genetically tagging proteins of interest and the more than 50 variations of fluorescent protein coding sequences contributed to GENBANK. Second, their fluorescence forms without the requirements of factors from native organisms or any additional exogenous agents other than molecular oxygen, which allows expression in a diverse range of organisms, tissues, and cell types. Finally, the fluorescence from these molecules can be observed directly without the requirement of fixation and addition of a secondary label or reactant, which makes observations possible in the living specimen. 2. Aequorea victoria GFP Advances in fluorescent protein technology in imaging began with the cloning of the Aequorea victoria green fluorescent protein (GFP) gene (1) and its From: Methods in Molecular Biology, vol. 411: Reporter Genes: A Practical Guide Edited by: D. Anson © Humana Press Inc., Totowa, NJ

47

48

Patterson

functional expression in heterologous systems (2). The impact of these two studies on cell biology should not be underestimated since they promoted GFP from an interesting protein isolated from a marine organism to a cell biology tool cited in almost 20,000 publications (PubMed search for ‘GFP’; June 2006). Prior to 1994, extensive biochemical and spectroscopic data were collected on the bioluminescent protein extract isolated from the jellyfish (3–7). Evidence from these early investigations demonstrated that the protein held promise as a useful tool for monitoring cell and organism processes if its fluorescence properties could be transferred into other species. The nuclear magnetic resonance (NMR) structure of proteolytic peptides containing the GFP chromophore indicated that residues serine 65, tyrosine 66, and glycine 67 undergo an uncommon cyclization reaction (8). Since such a structure had been previously unknown in protein structural biology, the novel chromophore suggested a requirement for endogenous jellyfish factor(s). Nevertheless, upon expression of the cDNA in heterologous systems, it was evident that the formation of this structure occurs in the absence of specific jellyfish factors (2,9). It was later shown that oxygen is required for production of the functional chromophore (10); however, no other exogenous factors have been reported as being necessary for the formation of the fluorescent structure. 2.1. Aequorea victoria Fluorescent Protein Structure Two independent crystal structure determinations showed that GFP is an 11-strand β-barrel containing a short segment of α-helix within the interior of the barrel (11,12) (Fig. 1). The light-producing p-hydroxybenzylidene-imidozolidinone chromophore consists of a cyclized tripeptide, composed of residues serine 65, tyrosine 66, and glycine 67, located in the central portion of the αhelix. The formation of a fluorescing protein requires proper folding into the barrel structure, followed by cyclization of the three amino acids making up the chromophore and oxidation/dehydration reactions to produce a functional fluorescent molecule (13). 2.2. Aequorea victoria Fluorescent Protein Folding Fluorescent proteins provide a good self-indicator for their folding, since the unfolded proteins generally do not fluoresce. However, for a newly synthesized fluorescent protein, it is often difficult to separate the kinetics of folding into the barrel structure from the formation of the chromophore. A study addressing these two aspects of GFP folding kinetics was performed on the S65T mutant (discussed later) (14). After denaturation in 8 M urea, refolded S65T protein recovers fluorescence with biexponential kinetics (fast phase t1/2 approx 28 s and slow phase t1/2 approx 284 s), which are much faster than the formation of a fluorescent protein from misfolded S65T protein purified from inclusion bodies


49

Fig. 1. The structure of the green fluorescent protein is an 11-strand β-barrel (peptide backbone rendered as a tube) with an α-helix located in the center containing the chromophore (rendered in ball and stick model). This figure was produced using Cn3D 4.1 (National Center for Biotechnology Information, National Institutes of Health, Bethesda, MD) using the protein data bank coordinates 1GFL submitted by the authors of ref. 11. See accompanying CD for color version.

(14). In the latter experiments, the folding into the barrel is best described by a t1/2 of approx 10 min, whereas the cyclization step (Fig. 2) is faster (t1/2 approx 180 s); the final oxidized and fluorescent chromophore (Fig. 2) requires much longer (t1/2 approx 76 min) to form. These slow time constants are not thought to be owing to the starting material, since proteolysis resistance indicates that folding is similar between the insoluble nonfluorescent proteins purified from bacterial inclusion bodies and soluble fluorescent proteins purified from bacteria (14). 2.3. Aequorea victoria Fluorescent Chromophore Formation The proposed mechanism for formation of the GFP chromophore (Fig. 2) was based largely on chemical logic (13,15) using the chromophore structure derived from an NMR structure of a proteolytic peptide (8) and eventually the crystal structures of the wild-type (11) and S65T (12) Aequorea victoria (avGFP) proteins. The first step in the chromophore formation after protein folding is a cyclization step in which the peptide amido nitrogen at amino acid position G67 undergoes a nucleophilic attack on the carbonyl of the serine at position S65 to produce a five-membered imidazolinone ring (13) (Fig. 2). In recent years,

50

Patterson

Fig. 2. Formation of the Aequorea victoria GFP chromophore. The chromophore of GFP consists of S65, Y66, and G67. These residues undergo a cyclization reaction between S65 and G67 to form a five-membered ring. Oxidation and dehydration reactions follow to extend the π-bonding system of the aromatic Y66 and produce a fluorescent molecule.

molecular modeling and crystal structures of GFP mutants trapped in intermediate stages of the chromophore formation have confirmed the basic mechanism, but the role of the surrounding secondary amino acid residues and the sequence of the oxidation and dehydration steps remain unresolved (16–19). 2.4. Aequorea victoria Fluorescent Chromophore Requirements Whereas the secondary residues involved in chromophore formation, as well as their roles, are still under debate, more is known about the residues within the main chromophore. Position 65 in avGFP can be altered to a number of amino acids without deterring the development of a fluorescent molecule. Furthermore, the equivalent position in other fluorescent proteins can be any of several amino acids (Tables 1–3). The amino acid in position Y66 in the cyclized peptide is usually a tyrosine, and it can also be altered but generally requires a phenylalanine, histidine, or tryptophan to produce a protein that fluoresces (13). However, the glycine at position 67 appears to be required in most, if not all, of the GFP-like molecules discovered to date. 2.5. Improvements in Aequorea victoria GFP Although the expression of avGFP in the bacterial and nematode systems (2) demonstrated the potential of this marker, the initial transition to other cell systems was problematic.


51

Table 1 Selected Fluorescent Protein Variants Developed from Aequorea victoria GFP Wavelengths (nm) a Protein

Amino acid substitutions

wtGFP Sapphire (H9-40) BFP CFP Cerulean

CGFP GFP Emerald YFP (10C) Topaz Citrine Venus

T203I F64L, Y66H, Y145F F64L, S65T, Y66W, N146I, M153T, V163A F64L, S65T, Y66W, S72A, N146I, Y145A, H148D, M153T, V163A, A206K F64L, S65T, Y66W, N146I, M153T, V163A, T203Y F64L, S65T S65T, S72A, N149K, M153T, I167T S65G, V68L, S72A, T203Y S65G, S72A, K79R, T203Y S65G, V68L, Q69M, S72A, T203Y F46L, F64L, M153T, V163A, S175G, T203Y

λ ex

λ em

Reference

393–400 (473–475) 399

504–509

2,3,5

511

13,35

380–383 434 (452)

440–447 476 (505)

13,87 13,87

433

475

39

463

506

88

488–489 487

507–509 509

514 514 516

527 527 529

12,34 34 41

515

528

42

10,32,87 34

a Wavelengths in parentheses represent minor peaks. Where various peaks have been reported,

the ranges of reported wavelengths are indicated.

The first obstacle was that wild-type avGFP has a tendency to misfold at approx 37°C. Several mutations have been reported to improve folding and/or chromophore formation (20–22) and are helpful when the protein is expressed at 37°C. In addition, folding reporter and superfolder GFPs, which develop fluorescence when tagged to poorly folded proteins, were produced (23). Some such mutations, such as F64L, V68L, S72A, Y145F, and I167T, are located in close proximity to the chromophore, and it is generally assumed that these assist in positioning residues for the peptide cyclization or the chromophore oxidation. However, others, such as F99S, M153T, V163A, and S175G, are located many Angstroms from the chromophore within the exterior barrel structure, and we have no knowledge of the mechanism by which these mutations improve folding (22).

52

Patterson

Table 2 Selected Fluorescent Protein Variants Developed from Discosoma drFP583 Wavelengths (nm) Protein DsRed (drFP583) DsRed1 DsRed2


R2A,K5E, K9T, V105A, I167T, S197A DsRed1.T1 R2A, K5E, N6D, T21S, H41T, N42Q, V44A, C117S, T217A dimer2 R2A, K5E, N6D, T21S, H41T, N42Q, V44A, V71A, C117T, F118L, I125R, V127T, S131P, K163Q, S179T, S197T, T217S mRFP1 R2A, K5E, N6D, T21S, H41T, N42Q, V44A, V71A, K83L, C117E, F124L, I125R, V127T, L150M, R153E, V156A, H162K, K163M, A164R, L174D, V175A, F177V, S179T, I180T, Y192A, Y194K, V195T, S197I, T217A, H222S, L223T, F224G, L225A mHoneydew Q66M, Y67W, T147Sa mBanana b Q66C, T147S, Q213L, I197Ea b mOrange V7I, T41F, L83F, Q66T, T147S, M182K, T195V Q66Ma dTomato b mTangerine Q66C, T147S, Q213La mStrawberry b V7I, S62T, Q64N, Q66T, T147S, M182K, T195V, Q213La b mCherry V7I, Q66M, T147S, M163Q, T195Va mRaspberry F65C, A71G, I161Ma mPlum V16E, R17H, K45R, F65I, L124V, I161M, K166Ra a Mutations

λ ex

λ em

Reference

558

583

46

558 561

583 587

554

586

54

552

579

51

584

607

51

487 (504) 537 (562) 540 553 548 562

55 55 55

554 568 574

581 585 596

55 55 55

587

610

55

598 590

625 649

56 56

in mRFP1 coding sequence. the first seven N-terminal amino acids of mRFP1 replaced with the N-terminal amino acids MVSKGEE of avGFP and has the addition of the C-terminal seven amino acids GMDELYK of EGFP. b Has

53

asulGFP (asFP499) asFP522 asCP562 mcGFP mmGFP MiCy Kusabira-Orange 522

511 562 432 398 472 (380) 548 477 505 495 561

597 620 612 612 600 499

572 583 590 590 592 480

595

572 A148S S165V S165A S165C S165T

645 538

598 528

HcRed zoanYFP (zFP538) asCP (asculCP, asFP595)

509

λem

498

λex

Wavelengths (nm)

Renilla GFP

Protein


M65, Y66, G67 Q62, Y63, G64 T60, Y61, G62 Q69, Y70, G71 C64, Y65, G66

M65, Y66, G67 M65, Y66, G67 M65, Y66, G67 M65, Y66, G67 M65, Y66, G67 Q65, Y66, G67

M65, Y66, G67

E64, Y65, G66 K66, Y67, G68

S66, Y67, G68

Chromophore peptide

Anemonia sulcata Anemonia sulcata Montastrea cavernosa Meandrina meandrites Acropara sp. Fungia concinna

Anemonia sulcata Anemonia sulcata Anemonia sulcata Anemonia sulcata Anemonia sulcata Anemonia sulcata

Anemonia sulcata

Heteractis crispa Zoanthus sp.

Renilla reniformis

Organism

Table 3 Selected Fluorescent Protein Variants Discovered and Developed from Other Marine Organisms

(continued)

91 91 92 92 60 60

90 90 90 90 90 91

80

5 Patent: US 6232107-B 15-MAY-2001 89 46

Reference

Fluorescent Proteins for Cell Biology 53

54 S3P, K6T, K7E, V19I, Y101S, C143S, M146I, S158A, N168D

V123T, Y188A, F190K

F102S, A104S, V123T, C151S, F162Y, F193Y, G195S, K11R, V25I, K32R, S55A, T62V, Q96E, E117Y, V133I, S139V, T150A, A166E, Q190G, F13Y, C115T, C217S

eqFP611 zoanGFP(zFP506) amajGFP(amFP486) dstrGFP(dsFP483)

gtCP aeCP597 AQ143

Azami-Green (AG) mAG

mKO

Protein


Table 3 (Continued)

559 496 458 456

580 abs 597 abs 595

492

492

548

λex

611 506 486 484

655

505

505

559

λem

Wavelengths (nm)

M63, Y64, G65 N66, Y67, G68 K68, Y69, G70 Q66, Y67, G68

Q65, Y66, G67 M63, Y64, G65 M63, Y64, G65

Q62, Y63, G64

Q62, Y63, G64

C64, Y65, G66

Chromophore peptide

Entacmaea quadricolor Zoanthus sp. Anemonia majano Discosoma striata

93 46 46 46

59 57 57

59 Galaxeidae (stony coral) Goniopora tenuidens Actinia equina Actinia equina

59

60

Reference

Galaxeidae (stony coral)

Fungia concinna

Organism

54 Patterson

dis3GFP dendGFP mcavGFP rfloGFP scubGFP1 scubGFP2 dis2RFP (dsFP593) zoan2RFP mcavRFP rfloRFP anm1GFP1 anm1GFP2 phiYFP anm2CP ppluGFP1 ppluGFP2 laesGFP pmeaGFP1

rmueGFP

clavGFP(cFP484) cgigGFP hcriGFP ptilGFP

510

483 496 500 508

503 512 494 508 506 516 508 518 497 506 497 506 573 593 552 576 508 (572) 520 (580) 506 (566) 517 (574) 475 495 490 504 525 537 572 597 480 500 482 502 491 506 489 504

498

443 399 (482) 405 (481) 500

Q62, Y63, G64 H62, Y63, G64 D69, Y70, G71 Q62, Y63, G64 Q66, Y67, G68 Q68, Y69, G70 Q66, Y67, G68 D66, Y67, G68 H62, Y63, G64 H62, Y63, G64 S99, Y100, G101 S62, Y63, G64 T65, Y66, G67 Q65, Y66, G67 G57, Y58, G59 G57, Y58, G59 G57, Y58, G59 G57, Y58, G59

Q69, Y70, G71

Q104, Y105, G106 Q63, Y64, G65 R63, Y64, G65 Q69, Y70, G71

55

(continued)

46 45 45 Patent: US 6232107-B 15 MAY 2001 Renilla muelleri Patent: US 6232107-B 15 MAY 2001 45 Discosoma sp. 3 45 Dendronephthya sp. Montastracea cavernosa 45 Ricordea florida 45 Scolymia cubensis 45 Scolymia cubensis 45 94 Discosoma sp. 2 45 Zoanthus sp. 2 Montastracea cavernosa 45 Ricordea florida 45 Anthomedusae sp. SL-2003 58 Anthomedusae sp. SL-2003 58 Phialidium sp. 58 Anthomedusae sp. SL-2003 58 Pontellina plumata 58 Pontellina plumata 58 Labidocera aestiva 58 Pontella meadi 58 Clavularia sp. Condylactis gigantea Heteractis crispa Ptilosarcus sp.

Fluorescent Proteins for Cell Biology 55

56

a Proteins

487 491 V11I, F64L, K101E, 480 T206A, E222G 578 abs 571 abs 571 abs 580 abs

pmeaGFP2 pdae1GFP acGFPL (aceGFP) hcriCP a (hcCP) cgigCP a (cgCP) cpasCP a (cpCP) gtenCP a (gtCP) 502 511 505

λem

58 58 78 89 89 89 89

Heteractis crispa Condylactis gigantea Condylactis passiflora Goniopora tenuidens

E63, Y64, G65 A63, Y64, G65 A63, Y64, G65 Q62, Y63, G64

Reference

Pontella meadi Pontellidae sp. SL-2003 Aequorea coerulescens

Organism

G57, Y58, G59 G57, Y58, G59 S65, Y66, G67

Chromophore peptide

with “CP” in their name are referred to as chromoproteins and are not fluorescent.

λ ex

Protein

Wavelengths (nm)


Table 3 (Continued)

56 Patterson


57

Second, the Aequorea victoria specific codon usage of wtGFP rendered protein expression in other systems inefficient. Mutagenesis improved these properties by converting the coding sequence to versions more efficiently used by plant (24,25), yeast (26), and mammalian systems (27). Although these mutations may have resulted in small changes in the intrinsic brightness of the protein, their major influence was to produce more observable molecules. Because the fluorescent protein requires synthesis by the cell, a time delay is inevitable between the formation of the translated protein and the maturation into a fluorescing protein. Work on the avGFP protein revealed that the wildtype GFP formed fluorescence with a time constant of approx 2 h (10). In contrast, the S65T mutant was observed to develop fluorescence with a time constant of approx 0.45 h (10). Experiments with fluorescent protein fusion proteins can be complicated by any influence that the fluorescent protein has over the protein of interest, such as self-association or oligomerization of the fluorescent protein. Although the wtGFP and its derivatives dimerize at high concentrations (Kd approx 0.11 mM) (22), this potential artifact was not generally considered a problem until a study of plasma membrane raft molecules was found to be complicated by significant oligomerization (28). In this study, localization of molecules in lipid rafts was analyzed using a technique called Förster resonance energy transfer (FRET), and an artifactual FRET signal was attributed to the interaction of the fluorescent proteins used (28). This could be disrupted by mutation of one of three hydrophobic residues on the exterior of the GFP barrel to charged residues (A206K, L221K, or F223R) (28). At low expression levels, self-association may show negligible interference with the proper localization and dynamics of many chimeras, but the use of fluorescent proteins with little affinity for each other is generally encouraged to avoid this complication (29). 2.6. Variations in Aequorea victoria GFP Spectra Further refinements of the Aequorea victoria GFP concentrated on its spectral properties. One of the first improvements dealt with the major (approx 400 nm) and the minor (approx 475 nm) excitation peaks (Fig. 3A). Excitation of either peak results in green emission at approx 508 nm, but this is not optimal for imaging with common laser lines and/or filter sets. Second, the two excitation peaks of wtGFP exhibit a photo-induced phenomenon, referred to as photoconversion or photoisomerization, in which the peaks interconvert during excitation (2,30,31). A single substitution of the serine at the 65 position with either a threonine, alanine, glycine, cysteine, or leucine collapses the major and minor peaks to a single peak at approx 489 nm while maintaining green fluorescence (10,32) (Fig. 3B), thus alleviating the effects of photoconversion. In addition, the excitation peak at 489 nm makes GFP brighter under excitation at

58

Patterson

Fig. 3. Excitation spectra, emission spectra, and chromophores of selected Aequorea victoria variants. The excitation spectra (solid lines) and emission spectra (dashed lines) for (A) wtGFP, (B) EGFP, (C) EBFP, (D) ECFP, (E) T203I, and (F) EYFP are shown with drawings of their respective chromophores and surrounding residues that result in their phenotypic spectra.

488 nm (reported extinction coefficients between 52,000 and 58,000) (22). Codon-optimized versions with this mutation, EGFP (33) and Emerald (34), are the common choices for imaging in mammalian systems. The next major advance in GFP imaging came with the ability to image more than one tagged molecule within the same cell. For this type of multicolor experiment, fluorescent proteins with differing spectra properties were required. The necessary spectra changes were accomplished by alterations of amino acids within and around the GFP chromophore and were discovered through random


59

mutagenesis experiments of GFP (10,13,35). Substitution of the tyrosine at position Y66 with a histidine shifts the absorbance spectrum to a peak at approx 380 nm with emission at approx 445 nm to produce the blue fluorescent proteins (BFPs) (13,22) (Fig. 3C). The BFPs (Y66H mutants) and the GFPs (S65 mutants) were breakthroughs in allowing the simultaneous imaging of multiple fluorescent proteins. The excitation and emission spectra of BFPs differ enough from these of GFPs to allow separation of the two signals for dualcolor imaging, even though the BFP emission overlaps sufficiently with the GFP absorbance spectrum to allow energy transfer (36). Researchers were able to take advantage of these properties in several studies, although the application of BFPs is severely limited by modest fluorescence (31,36), near ultraviolet excitation (13), and a tendency to photobleach readily (31,37). Substitution with a tryptophan in the 66 position produced W7, one of the first cyan fluorescent proteins (CFPs), which was identified in the same mutagenesis screen (13). The CFPs feature moderate photostability (37,38) and have peaks in the 435- to 450-nm (absorbance) and 475- to 505-nm (emission) regions (13,36), which places them between the BFP mutants and the green S65 mutants in the fluorescence spectrum (Fig. 3D). However, Y66W mutants have multiple fluorescent lifetimes, which complicates their use in a specialized imaging technique known as fluorescence lifetime imaging (FLIM). Structural evidence indicated that CFP molecules probably existed as a mixed population, in which the histidine at 148 and tyrosine at 145 were in alternate orientations, and their effect on the chromophore led to the multiple fluorescent lifetimes. Since these fluorescent proteins have side chains within a few Angstroms of the chromophore, Rizzo et. al. (39) hypothesized that converting the H148 into a hydrophilic amino acid (aspartic acid) would stabilize one position. The H148D mutation Cerulean did indeed result in behavior better suited for fluorescence lifetime work, but more generally, Cerulean has increased brightness by approx 2.5-fold and increased photostability by approx 30% compared with ECFP (39). An additional mutant of note from the initial mutagenesis screen of Tsien and colleagues (13), which was also independently isolated by Prendergast and co-workers (35), is referred to as Sapphire (34). Similar to wild-type GFP, Sapphire has a tyrosine in the 66 position but has an isoleucine substituted for the threonine in the 203 position. The T203I mutant has a single major peak at approx 400 nm and a green emission peak at 511 nm (Fig. 3E). This mutant has been used as a near-UV excitable fluorophore. However, Sapphire probably lacks the stability necessary for any long-term imaging and has a characteristic similar to wild-type GFP in that it undergoes a photoconversion under approximately 400 nm excitation, in which the 400-nm absorbance peak decreases, and a peak at approximately 504 nm increases (40).

60

Patterson

The T203 position was later exploited in mutants rationally designed from the crystal structure of the S65T mutant of GFP (12). These are generally referred to as yellow fluorescent proteins (YFPs) and were developed by coupling an S65G mutation with a substitution at the neighboring T203 position with tyrosine, phenylalanine, tryptophan, or histidine (12). The rationale behind targeting this residue was to place histidine or one of the aromatic amino acids in a position such that a π-orbital stacking system was produced. This was proposed to lower the energy required for excitation of the molecule and thus shift the excitation and emission wavelengths toward the red part of the spectrum. These mutations indeed led to spectra that are red-shifted by approx 20 nm. The most common version of YFP contains S65G and T203Y substitutions, which result in excitation and emission peaks at 513 and 527 nm, respectively (Fig. 3F). The shifted excitation spectrum of YFP allows excitation with the 514-nm line of an argon ion laser, and it is easily distinguished from the CFPs. The YFPs have also undergone significant improvement. Two notable versions are Citrine (41) and Venus (42). The original versions of YFP displayed pH sensitivity (pKa approx 6.9), sensitivity to chloride ions (43,44), and decreased maturation to the fluorescent form at 37°C compared with GFP (41). YFPs are also generally thought to have lower photostability compared with GFP (41), although single-molecule work addressing this issue found that YFP was more stable (38). Nevertheless, Citrine and Venus have each improved on several YFP sensitivities. Citrine resulted from a single mutation, Q69M, and was developed in the context of a Ca2+ indicator construct, camgaroo-2. The Q66M mutation largely alleviates both the halide (principally chloride) sensitivity and the pH dependence, producing a molecule with a pKa of approx 5.7. In addition, the photostability of Citrine is increased by about twofold compared with the older YFP variants and appears brighter when expressed at 37°C (41). Venus, on the other hand, relies on a mutation, F46L, which increases the efficiency of development of a fluorescent molecule at 37°C by 20- to 30-fold. This mutation, coupled with other mutations (Table 1), produces a molecule with decreased sensitivity to pH (pKa approx 6.0) and decreased chloride sensitivity (42). Venus matures more rapidly and is brighter when expressed in cells compared with Citrine (29). However, it lacks the photostability of Citrine, since it photobleaches with similar kinetics as the previous YFPs. This property is not improved by the Q69M mutation (42). 3. DsRed Fluorescent Protein Up to this point, the discussion has centered on fluorescent proteins developed from the original Aequorea victoria GFP (avGFP). Discoveries of similar proteins from other marine organisms have extended the bank of fluorescent


61

Fig. 4. Structure of the DsRed fluorescent protein. DsRed is also a β-barrel (peptide backbone rendered as a tube) with an α-helix located in the center containing the chromophore (rendered in ball and stick model). Wild-type DsRed is normally an (A) tetramer made up of (B) monomer subunits that are structurally similar to avGFP. This figure was produced using Cn3D 4.1 (National Center for Biotechnology Information, National Institutes of Health, Bethesda, MD, USA) using the protein data bank coordinates 1G7K submitted by the authors of ref. 48. See accompanying CD for color version.

proteins for imaging to the red wavelength range (45). The first and best characterized of the red fluorescent proteins is the DsRed protein cloned from the Discosoma coral (Fig. 4). The wild-type DsRed (drFP583) has a major absor-

62

Patterson

bance peak at 558 nm and minor peaks at 530 and 487 nm, with an emission peak at 583 nm (46), and has led to the production of a plethora of new variants with diverse characteristics. 3.1. DsRed Fluorescent Protein Structure and Chromophore Formation The overall protein structure of DsRed is similar to that of avGFP in that it consists of a β-barrel with an α-helix in the center containing the chromophore (47,48). The formation of the DsRed fluorescent chromophore is proposed to proceed similarly to that of the avGFPs by a mechanism whereby a chromophore that produces green fluorescence is initially formed but matures via an additional oxidation step into the red species (49). The cyclized tripeptide is made of Q66, Y67, and G68, with the π-bonding system of the chromophore extended along the Cα–N bond of the glutamine at position Q66 (47,48) resulting in a shift of the excitation/emission spectra to longer wavelengths (49,50). 3.2. DsRed Improvements As with the wild-type avGFP, the native DsRed protein has many undesirable characteristics, including several that rendered it almost unusable in many fusion constructs. The reaction for forming the mature DsRed chromophore requires molecular oxygen and up to 48 h (50,51). This time lag between protein expression and development of the red fluorescence made use of DsRed as a secondary red marker difficult because of cross-talk with green fluorescent signals. A mutant referred to as E5 (fluorescent timer), which contains V105A and S197T substitutions, changes DsRed fluorescence from a predominantly green to a red signal with a t1/2 of approx 10 h (52). The addition of an I161T substitution to E5, referred to as E57, further reduces the t1/2 for formation of the mature red chromophore to 3 to 4 h (53). Finally, a mutagenesis screen searching for mutants that rapidly form red fluorescence eventually reduced maturation to less than 1 h in a variant named DsRed1.T1 (54). Many of the DsRed1.T1 mutations listed in Table 2 are scattered about the DsRed barrel structure and provide no clear indication of their influence on protein folding and/or chromophore formation, illustrating the fact that there is much left to learn about the folding mechanism of these proteins. Notably, many of the currently used DsRed variants listed in Table 2 contain mutations similar to DsRed1.T1. Another major problem with DsRed is that it forms obligatory tetramers, even at low concentrations (50). Therefore, any intracellular trafficking studies have to contend with potential artifacts from the self-interactions of the DsRed attached to the protein of interest, particularly with fusions that naturally form oligomers or biopolymers. However, combinations of mutations were found that result in the dimeric red fluorescent protein dimer2 and the monomeric red fluorescent protein mRFP1 (Fig. 5) (51). Unfortunately, mRFP1 has decreased


63

Fig. 5. Excitation spectrum (solid lines), emission spectrum (dashed lines), and chromophore (drawing) of mRFP1.

brightness and reduced photostability relative to DsRed, both of which limit its utility. Rather than undergo the effort to monomerize other fluorescent molecules, oligomerization problems were avoided by using mRFP1 as starting material to improve on and further develop new spectral variants (55). In addition to random mutagenesis experiments, the similarity between the overall structure of avGFP (11,12) and the DsRed structure (47,48) led to sitedirected mutagenesis of mRFP1 to produce molecules with the desired characteristics. For example, mRFP1 has decreased fluorescence when fused to the C terminus of a protein of interest (55). Since the avGFP does not seem to be similarly affected, the N terminus of avGFP was used to replace that of mRFP1, and the C terminus of avGFP was added to the C terminus of mRFP1. The result was a molecule that more efficiently develops red fluorescence when tagged to a polymerizing molecule, such as tubulin (55). In addition, because it was established that mutations at the S65 position in avGFP improve its maturation into the fluorescent form (10), the equivalent residue (Q66) was targeted within the mRFP1 chromophore. This approach produced molecules containing a Q66M mutation that exhibited more efficient chromophore maturation and provided the starting material for subsequent rounds of random and directed mutagenesis (55). One of the final results of this approach was the discovery of mCherry, one of the most advanced red fluorescent proteins yet developed. mCherry is only slightly brighter and slightly red-shifted compared with mRFP1, but it is 10-fold more photostable under imaging. 3.3. Spectral Variations in mRFP1 The knowledge and experience gained with avGFP proved most useful for altering the spectra of mRFP1. Residue Y67 of mRFP1 (equivalent to Y66 in

64

Patterson

avGFP) is the aromatic amino acid that is one of the key residues in defining a fluorescent molecule’s spectral characteristics. Changing this amino acid to histidine or tryptophan was known to blue-shift the GFP spectra to produce the BFP and CFP variants, respectively (13). A similar substitution in mRFP1, Y67W, produced the mHoneydew variant, a molecule with the expected blueshifted spectra (55) (Table 2). Other blue-shifted variants were produced by targeting another key amino acid within the chromophore, residue 66. In this case, the M66 mutant produced as a precursor to mCherry was converted into a cysteine or threonine and further mutated to eventually produce the mTangerine and mBanana or the mOrange and mStrawberry derivatives, respectively (55) (Table 2). The other residues modified in this process are predicted to surround the chromophore and change its local environment. The mRFP1 family was red-shifted even more toward the infrared region of the spectrum with the introduction of mPlum and mRaspberry (56). Rather than relying on random mutagenesis and screening bacteria colonies, the coding region for the mRFP1.2 variant was expressed in B lymphocytes, which use somatic hypermutation to alter and improve immunoglobins. The mRFP1 gene introduced into these cells was mutated by somatic hypermutation with over 23 rounds of cell selection employed to achieve far-red-shifted fluorescence. The mRaspberry mutant has excitation at 598 nm and emission at 625 nm and a maturation half-time of approx 55 min. mPlum has an excitation peak at 590 nm and emission at 649 nm, which is the most red-shifted emission peak reported for a fluorescent protein with the exception of the tetrameric AQ143 at 655 nm (57) (Table 3). mPlum matures into a fluorescent molecule with a half-time of approx 100 min and is 30-fold more photostable than mRFP1 (56). Based entirely on the extinction coefficients and quantum yields, the mHoneydew, mBanana, mTangerine, and mPlum variants are less bright than mRFP1. The mOrange derivative has a quantum efficiency and extinction coefficient comparable to those of the avGFP fluorescent proteins that are commonly used, such as EGFP, EYFP, Citrine (41), and Venus (42). However, mOrange displays a moderate sensitivity to pH (pKa approx 6.5) and matures slowly (t1/2 approx 2.5 h). The rest of the fruit-named molecules mature within a reasonable period (t1/2 ≤ 1 h) but display varying degrees of photostability. Of the monomeric forms, mCherry (approx 10-fold better than mRFP1) and mPlum (approx 30-fold better than mRFP1.2) are the most photostable. 4. Fluorescent Proteins from Other Organisms The remarkably diverse array of proteins discussed above has been derived from just two sources, the Aequorea victoria jellyfish and the Discosoma coral; they represent only the beginning of fluorescent protein technology. The past 5 years have witnessed a marvelous expansion of new fluorescent proteins derived


65

from an array of marine species representing three different classes, Copepoda, Hydrozoa, and Anthozoa. The evolution and diversity of fluorescent proteins have been previously reviewed (45,58). A listing of molecules from more than 20 species is included in Table 3, and these span almost the entire visible spectral region, with emission wavelengths ranging from 477 to 655 nm. Unfortunately, published accounts of their use are limited, but given the broad diversity of spectral characteristics, these molecules offer much potential in cell and developmental biology experiments. Readers should be aware that many of these fluorescent proteins have very low fluorescence (quantum yields < 0.1) and/or tend to oligomerize. Both characteristics decrease the utility of these molecules for trafficking and/or localization experiments. On the other hand, several reported monomeric molecules such as mAG (59), phiYFP (58), and mKO (60), provide alternatives for green, yellow, and red fluorescent proteins. 5. Photoactivatable Fluorescent Proteins In most cases, the battery of fluorescent proteins so far discovered has proved quite suitable for studying gene expression, for observing localization of proteins, and for monitoring the dynamics of organelles and cells; however, the dynamics of localized proteins within a population may remain unresolved (61). For this, the steady-state fluorescence profile must be altered such that a subpopulation of the fluorescent protein chimeras is highlighted. Photobleaching (excitation-induced photodestruction) is generally used to highlight a population of molecules indirectly, but advances and discoveries have been made for directly highlighting a pool of molecules using “photoactivatable” or “optical highlighter” fluorescent proteins (62) (Table 4). Molecules placed in the photoactivatable category have the general characteristic that they are initially dark at the activated fluorescence wavelength but upon activation display an increase in fluorescence and are “highlighted” over a darker background. At the writing of this chapter, proteins derived from nine species have been reported (Table 4). The optical highlighters have been subclassed based on the spectral characteristics of their activated and nonactivated states (green vs red fluorescence) and on the reversibility of the activated state (irreversible vs reversible). An overview of these molecules is given below, and interested readers are directed to a review concentrating on this category of fluorescent proteins for further information (62). 5.1. Aequorea victoria GFP One early approach to develop a photoactivatable fluorescent protein relied on the fact that several Aequorea victoria variants, wtGFP, GFPmut1, -2, and -3 (20), S65T (10), I167T (13), and GFPuv (63), convert into red fluorescent

66 V123T/T158H

mEosFP

L40A, L61V, Y95F, N121K, M123T, Y188A, S199G, G213A

T158H

d2EosFP

DRONPA Dendra

V123T

d1EosFP

Kikume Green-Red (KikGR) PAmRFP1-1 PAmRFP1-2 PAmRFP1-3 EosFP S146H, I161V, I197H S146H, I161C, I197H S146H, I161S, I197H

V163A, T203H

PA-GFP

Kaede


Protein 400 (Pre) 504 (Post) 508 (Pre) 572 (Post) 507 (Pre) 583 (Post) 578 (Post) 578 (Post) 578 (Post) 506 (Pre) 571 (Post) 505 (Pre) 571 (Post) 506 (Pre) 569 (Post) 505 (Pre) 569 (Post) 503 (Post) 486 (Pre) 558 (Post)

λ ex 515 517 518 582 517 593 605 605 605 516 581 516 581 516 581 516 581 518 505 575

λ em

Wavelengths (nm) a

Table 4 Selected Optical Highlighter Proteins (Photoactivatable Fluorescent Proteins)

75 83 76 Pectiniidae spp. Dendronephthya sp.

75 Lobophyllia hemprichii

Lobophyllia hemprichii

75

72 72 72 75 Discosoma sp. Discosoma sp. Discosoma sp. Lobophyllia hemprichii Lobophyllia hemprichii

74

73

40

Reference

Favia favus

Trachyphyllia geoffroyi

Aequorea victoria

Organism

66 Patterson

a(Pre)

A148G G222E T62A, N121S, H148T, K158R, I167V, E172K, F221L, G222E, K238Q T62A, S108T, N121S, H148T, M153V, T154A, K158R, I167V, E172K, F221L, G222E, K238Q

L40A, L61V, Y95F, N121K, M123T, Y188A, S199G, G213A, A224V

400 (Pre) 490 (Post)

580 (Post) 480 402 (Pre) 490 (Post)

486 (Pre) 558 (Post)

470 511

600 505 468 511

505 575

Aequorea coerulescens

Anemonia sulcata Anemonia sulcata Aequorea coerulescens Aequorea coerulescens

Dendronephthya sp.

represents the major peak prior to photoactivation, and (Post) represents the major peak after photoactivation.

PS-CFP2

asFP595 KFP1 aceGFP PS-CFP

Dendra2

80 82 78 77


67

67

68

Patterson

species upon irradiation with 488-nm light (64,65). This “photoactivation” produced high contrast with background fluorescence and worked well for monitoring protein diffusion in bacteria (64). However, the widespread use of this phenomenon in cell biology has been limited by the low oxygen conditions required. An approach suitable for aerobic conditions exploited the photo-induced conversion (photoconversion) that occurs within the wild-type avGFP (66). The wtGFP chromophore population is thought to exist as a mixture of neutral phenols (Y66 is protonated) and anionic phenolates (Y66 is deprotonated), giving rise to a major absorbance peak at approx 397 nm and a minor absorbance peak at 475 nm, respectively. Upon irradiation, the chromophore undergoes a proton transfer and photoconverts into the anionic form (22,30,67). The threonine at the 203 position rotates, and the hydroxyl group could help stabilize the anionic chromophore (68–70). In addition, the glutamic acid at position 222 (E222), which is in close proximity to the chromophore, is decarboxylated (70). Either or both of these structural changes may play key roles in stabilizing the shift of the chromophore population from a predominantly neutral species to an anionic species. The resulting absorbance increase at the minor peak leads to an increase in the fluorescence intensity when excited at this wavelength. Unfortunately, owing to the high initial background, photoconversion of the wild-type avGFP results in only a modest (about threefold) increase in fluorescence (22,40,66). The T203I mutant of avGFP produces a variant that has predominantly neutral phenol chromophore, which reduces the minor absorbance peak while maintaining the major peak (13,35). Further mutagenesis of the T203 position uncovered several other substitutions that also reduce the minor peak and retain the major peak at approx 400 nm (40). Surprisingly, many of these mutants, including T203I, maintain the ability to undergo photoconversion. However, the T203H mutant (denoted PA-GFP) was found to be exceptional, with the minor absorbance peak drastically decreased, giving more than 60- and more than 100-fold fluorescence increases after photoactivation in cells and in vitro, respectively (40). 5.2. DsRed Fluorescent Protein The DsRed protein has been used as a photoactivatable fluorescent protein in two different approaches. In the first case, the molecules of the DsRed tetramer exist as mixed populations of immature green fluorescent molecules and mature red fluorescent molecules. Since the protomers are in such close proximity and the emission spectrum of the green species overlaps with that of the red species, the energy of the excited green molecules can transfer nonradiatively to the red molecules. Selective photobleaching of the red species dequenches


69

the green fluorescence and leads to an approx 2.4-fold increase in green fluorescence emission (71). The second approach involves conversion of the monomeric version of DsRed, mRFP1, into a series of photoactivatable fluorescent proteins, PAmRFP1-1, PAmRFP1-2, and PAmRFP1-3 (72). The brightest of these, PAmRFP1-1, has a quantum yield of only 0.08, but produces an approx 70-fold increase in red fluorescence upon UV light excitation (72). Although this molecule is significantly less bright than other red photoactivatable fluorescent proteins, its derivation from mRFP1 suggests it may be used as a less perturbative protein marker. In addition, the lack of a green fluorescent component before or after activation renders it more amenable for use in multifluorescent protein experiments. 5.3. Green-to-Red Photoconversions The stony coral, Trachyphyllia geoffroyi, produces the protein Kaede, which can be photoactivated by irradiation at approx 400 nm (73). Prior to photoactivation, Kaede has a major absorbance peak at 508 nm and emission at 518 nm. After photoactivation, Kaede exhibits a new red-shifted absorbance peak at 572 nm, which upon excitation fluoresces with a new emission peak at 582 nm. This shift in both the excitation and emission peaks results in a more than 2000 fold increase in the red-to-green fluorescence ratio. Unfortunately, Kaede forms tetramers (73), which limit its usefulness as a protein trafficking tool. Nevertheless, as with the DsRed protein, the self-association properties can perhaps be suppressed (51), and Kaede’s large contrast with background after photoactivation gives it enormous potential as a photoactivatable fluorescent protein marker. Another coral fluorescent protein, KikG from Favia favus (74), was developed as a photoactivatable marker. Although the original KikG from did not exhibit photoactivatable properties, using information learned from their structural characterizations of Kaede, Miyawaki and colleagues engineered KikG into the KikGR variant, which produced a more than 2000-fold increase in fluorescence contrast during ratio imaging of the red and green components after photoactivation. The authors found that the purified KikGR protein lacked the brightness of Kaede in vitro but exhibited several fold more fluorescence than Kaede when expressed in cells. It was unclear whether this difference reflected protein expression levels or protein folding efficiency. As with Kaede, KikGR was also found to be a tetramer (74). The fluorescent protein EosFP from the stony coral Lobophyllia hemprichii also exhibits a green-to-red fluorescence photoconversion upon UV or near UV light irradiation (75). Similar to Kaede, EosFP has a preactivated excitation maximum at 506 nm with emission at 516 nm. Upon activation at 405 nm, the photoactivated excitation peak is located at 571 nm with emission at 581 nm. Initially

70

Patterson

determined to be a tetramer, EosFP was engineered into two dimeric forms, d1EosFP and d2EosFP. The combination of these mutations produced a monomeric molecule with a Kd of approx 0.1 mM, which has been named mEosFP. The emission maxima of these mutants remain constant whereas the excitation maxima and brightness change slightly (Table 4). Thus it would seem obvious that the mEosFP, owing to its decreased self-association, is the protein of choice in this subclass. However, mEosFP inefficiently forms a fluorescent molecule when expressed at 37°C (75). A recent addition to this category of optical highlighters is Dendra, which exhibits up to 4500-fold photoconversion from its green-to-red fluorescent forms (76). The wild-type form, isolated from Dendronephthya sp. and named dendGFP (45), was engineered into the monomeric Dendra with its photoactivation properties. Dendra can be activated with approx 400 nm light, which is the wavelength required by most of the other photoactivatable fluorescent proteins, but allows the option of activation with high levels (>0.5 W/cm2) of potentially less phototoxic wavelengths (approx 488 nm). It was noted that excitation of the unactivated green state of Dendra with lower levels of 488-nm light (

Reporter Genes: A Practical Guide

Reporter Genes. A Practical Guide

Reporter

Reporter

Reporter

Reporter

Reporter

Reporter

Mediasmart: How to Handle a Reporter by a Reporter

Molecular Imaging with Reporter Genes (Cambridge Molecular Imaging Series)

Lecturing: A Practical Guide

Beekeeping: A Practical Guide

Playwriting: A Practical Guide

Chemometrics: A Practical Guide

HPLC. A Practical Guide

Interplanetary Reporter

Interplanetary Reporter

A Practical Guide to Information Architecture (Practical Guide Series)

Parsing Techniques: A Practical Guide

Materials Syntheses: A Practical Guide

Surviving Prescribing: A Practical Guide

A Practical Guide to Racism

Constructive Psychotherapy: A Practical Guide

Constructive Psychotherapy: A Practical Guide

HPLC: A Practical User's Guide

Parsing techniques: a practical guide

Parsing Techniques: A Practical Guide

A Practical Guide To Asp

Agricultural Valuations: A Practical Guide

Ear Acupuncture: A Practical Guide

A Practical Guide to Splines

Reporter Genes: A Practical Guide