Light Microscopy: Methods and Protocols

METHODS IN MOLECULAR BIOLOGYTM

Series EditorJohn M. Walker

School of Life SciencesUniversity of Hertfordshire

Hatfield, Hertfordshire, AL10 9AB, UK

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

Light Microscopy

Methods and Protocols

Edited by

Hélio Chiarini-Garcia

Laboratory of Structural Biology and Reproduction, Department of Morphology,ICB, Federal University of Minas Gerais,

Belo Horizonte, MG, Brazil

Rossana C.N. Melo

Laboratory of Cellular Biology, Department of Biology, ICB,Federal University of Juiz de Fora, Juiz de Fora,

MG, Brazil

EditorsHélio Chiarini-GarciaDepartment of MorphologyFederal University of Minas GeraisBelo Horizonte, MG 31270-901, [email protected]

Rossana C.N. MeloDepartment of BiologyFederal University of Juiz de ForaJuiz de Fora, MG 36036-900, [email protected]

ISSN 1064-3745 e-ISSN 1940-6029ISBN 978-1-60761-949-9 e-ISBN 978-1-60761-950-5DOI 10.1007/978-1-60761-950-5Springer New York Dordrecht Heidelberg London

Library of Congress Control Number: 2010936902

© Springer Science+Business Media, LLC 2011All rights reserved. This work may not be translated or copied in whole or in part without the written permission ofthe publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013,USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form ofinformation storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodologynow known or hereafter developed is forbidden.The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identifiedas such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.

Printed on acid-free paper

Humana Press is part of Springer Science+Business Media (www.springer.com)

Preface

Of all scientific instruments, probably none has had more applications in the life sciencesthan the light microscope. Advances in microscope instrumentation, sample preparationand imaging techniques have been producing fundamental insights into the functions ofcells and tissues.

The protocols in Light Microscopy: Methods and Protocols cover a variety of bright-field and fluorescence microscopy-based approaches central to the study of a range ofbiological questions. The book provides information on how to prepare cells and tissuesfor microscopic investigations, including detailed staining procedures and how to analyzeimages and interpret results accurately. Techniques are presented in a friendly, step-by-stepfashion with helpful information and useful tips. Section I covers selected applications ofbright-field microscopy to the study of animal and plant biology. Section II covers the fun-damental principles of fluorescence microscopy as well as its applications to multiple fieldsincluding immunology, ecology, cancer biology and cell signaling. Light Microscopy: Meth-ods and Protocols addresses different needs of researchers, who are exploring the micro-scopic and intriguing world of the cell.

We thank Prof. John M. Walker and the staff at Humana Press for their invitation,editorial guidance, and assistance throughout the preparation of this book for publication.We also would like to express our sincere appreciation and gratitude to the contributorsfor sharing their precious laboratory expertise with the microscopy research community.

Hélio Chiarini-GarciaRossana C.N. Melo

v

Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

SECTION I BRIGHT-FIELD MICROSCOPY APPLICATIONS

1. Glycol Methacrylate Embedding for Improved Morphological,Morphometrical, and Immunohistochemical Investigations Under LightMicroscopy: Testes as a Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Hélio Chiarini-Garcia, Gleydes Gambogi Parreira,and Fernanda R.C.L. Almeida

2. Histological Processing of Teeth and Periodontal Tissuesfor Light Microscopy Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 19Gerluza Aparecida Borges Silva, Adriana Moreira, and José Bento Alves

3. Large Plant Samples: How to Process for GMA Embedding? . . . . . . . . . . . 37Élder Antônio Sousa Paiva, Sheila Zambello de Pinho,and Denise Maria Trombert Oliveira

4. Image Cytometry: Nuclear and Chromosomal DNA Quantification . . . . . . . 51Carlos Roberto Carvalho, Wellington Ronildo Clarindo,and Isabella Santiago Abreu

5. Histological Approaches to Study Tissue Parasitism During theExperimental Trypanosoma cruzi Infection . . . . . . . . . . . . . . . . . . . . 69Daniela L. Fabrino, Grazielle A. Ribeiro, Lívia Teixeira,and Rossana C.N. Melo

6. Intravital Microscopy to Study Leukocyte Recruitment In Vivo . . . . . . . . . . 81Vanessa Pinho, Fernanda Matos Coelho, Gustavo Batista Menezes,and Denise Carmona Cara

SECTION II FLUORESCENCE MICROSCOPY APPLICATIONS

7. Introduction to Fluorescence Microscopy . . . . . . . . . . . . . . . . . . . . . 93Ionita C. Ghiran

8. Using the Fluorescent Styryl Dye FM1-43 to Visualize Synaptic VesiclesExocytosis and Endocytosis in Motor Nerve Terminals . . . . . . . . . . . . . . 137Ernani Amaral, Silvia Guatimosim, and Cristina Guatimosim

9. Imaging Lipid Bodies Within Leukocytes with Different LightMicroscopy Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149Rossana C.N. Melo, Heloisa D’Ávila, Patricia T. Bozza,and Peter F. Weller

vii

viii Contents

10. EicosaCell – An Immunofluorescent-Based Assay to Localize NewlySynthesized Eicosanoid Lipid Mediators at Intracellular Sites . . . . . . . . . . . 163Christianne Bandeira-Melo, Peter F. Weller, and Patricia T. Bozza

11. Nestin-Driven Green Fluorescent Protein as an Imaging Marker forNascent Blood Vessels in Mouse Models of Cancer . . . . . . . . . . . . . . . . 183Robert M. Hoffman

12. Imaging Calcium Sparks in Cardiac Myocytes . . . . . . . . . . . . . . . . . . . 205Silvia Guatimosim, Cristina Guatimosim, and Long-Sheng Song

13. Light Microscopy in Aquatic Ecology: Methods for PlanktonCommunities Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215Maria Carolina S. Soares, Lúcia M. Lobão, Luciana O. Vidal,Natália P. Noyma, Nathan O. Barros, Simone J. Cardoso,and Fábio Roland

14. Fluorescence Immunohistochemistry in Combination with DifferentialInterference Contrast Microscopy for Studies of Semi-ultrathinSpecimens of Epoxy Resin-Embedded Samples . . . . . . . . . . . . . . . . . . 229Shin-ichi Iwasaki and Hidekazu Aoyagi

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241

Contributors

ISABELLA SANTIAGO ABREU • Laboratory of Cytogenetics and Cytometry, Department ofGeneral Biology, Federal University of Viçosa, Viçosa, MG, Brazil

FERNANDA R.C.L. ALMEIDA • Laboratory of Structural Biology and Reproduction,Department of Morphology, ICB, Federal University of Minas Gerais, Belo Horizonte,MG, Brazil

JOSÉ BENTO ALVES • University of Uberaba, Uberaba, MG, BrazilERNANI AMARAL • Department of Morphology, ICB, Federal University of Minas Gerais,

Belo Horizonte, MG, BrazilHIDEKAZU AOYAGI • Advanced Research Center, School of Life Dentistry at Niigata, The

Nippon Dental University, Niigata, JapanCHRISTIANNE BANDEIRA-MELO • Laboratory of Inflammation, Carlos Chagas Filho

Institute of Biophysics, Federal University of Rio de Janeiro, Rio de Janeiro, RJ, BrazilNATHAN O. BARROS • Laboratory of Aquatic Ecology, Department of Biology, ICB,

Federal University of Juiz de Fora, Juiz de Fora, MG, BrazilPATRICIA T. BOZZA • Laboratory of Immunopharmacology, IOC, Oswaldo Cruz Foun-

dation, Rio de Janeiro, RJ, BrazilSIMONE J. CARDOSO • Laboratory of Aquatic Ecology, Department of Biology, ICB,

Federal University of Juiz de Fora, Juiz de Fora, MG, BrazilCARLOS ROBERTO CARVALHO • Laboratory of Cytogenetics and Cytometry, Department

of General Biology, Federal University of Viçosa, Viçosa, MG, BrazilHÉLIO CHIARINI-GARCIA • Laboratory of Structural Biology and Reproduction, Depart-

ment of Morphology, ICB, Federal University of Minas Gerais, Belo Horizonte, MG,Brazil

WELLINGTON RONILDO CLARINDO • Laboratory of Cytogenetics and Cytometry,Department of General Biology, Federal University of Viçosa, Viçosa, MG, Brazil

FERNANDA MATOS COELHO • Laboratory of Immunopharmacology, Department ofBiochemistry and Immunology, ICB, Federal University of Minas Gerais, Belo Horizonte,MG, Brazil

DENISE CARMONA CARA • Department of Morphology, ICB, Federal University of MinasGerais, Belo Horizonte, MG, Brazil

HELOÍSA D’ÁVILA • Laboratory of Cellular Biology, Department of Biology, ICB, FederalUniversity of Juiz de Fora, Juiz de Fora, MG, Brazil

DANIELA L. FABRINO • Laboratory of Cellular Biology, Department of Biology, ICB,Federal University of Juiz de Fora, Juiz de Fora, MG, Brazil

IONITA C. GHIRAN • Department of Medicine, Beth Israel Deaconess Medical Center,Harvard Medical School, Boston, MA, USA

CRISTINA GUATIMOSIM • Department of Morphology, ICB, Federal University of MinasGerais, Belo Horizonte, MG, Brazil

ix

x Contributors

SILVIA GUATIMOSIM • Department of Physiology, ICB, Federal University of MinasGerais, Belo Horizonte, MG, Brazil

ROBERT M. HOFFMAN • AntiCancer Inc and Department of Surgery, University ofCalifornia, San Diego, CA, USA

SHIN-ICHI IWASAKI • Advanced Research Center, School of Life Dentistry at Niigata, TheNippon Dental University, Niigata, Japan

LÚCIA M. LOBÃO • Laboratory of Aquatic Ecology, Department of Biology, ICB, FederalUniversity of Juiz de Fora, Juiz de Fora, MG, Brazil

ROSSANA C.N. MELO • Laboratory of Cellular Biology, Department of Biology, ICB,Federal University of Juiz de Fora, Juiz de Fora, MG, Brazil

GUSTAVO BATISTA MENEZES • Laboratory of Immunopharmacology, Department ofMorphology, ICB, Federal University of Minas Gerais, Belo Horizonte, MG, Brazil

ADRIANA MOREIRA • Department of Morphology, ICB, Federal University of MinasGerais, Belo Horizonte, MG, Brazil

NATÁLIA P. NOYMA • Laboratory of Aquatic Ecology, Department of Biology, ICB,Federal University of Juiz de Fora, Juiz de Fora, MG, Brazil

DENISE M.T. OLIVEIRA • Department of Botanic, ICB, Federal University of MinasGerais, Belo Horizonte, MG, Brazil

ÉLDER ANTÔNIO SOUSA PAIVA • Department of Botanic, ICB, Federal University ofMinas Gerais, Belo Horizonte, MG, Brazil

GLEYDES GAMBOGI PARREIRA • Laboratory of Structural Biology and Reproduction,Department of Morphology, ICB, Federal University of Minas Gerais, Belo Horizonte,MG, Brazil

SHEILA ZAMBELLO DEPINHO • Department of Biostatistics, Institute of Biosciences,UNESP – Universidade Estadual Paulista, Botucatu, SP, Brazil

VANESSA PINHO • Department of Morphology, ICB, Federal University of Minas Gerais,Belo Horizonte, MG, Brazil

GRAZIELLE A. RIBEIRO • Laboratory of Cellular Biology, Department of Biology, ICB,Federal University of Juiz de Fora, Juiz de Fora, MG, Brazil

FÁBIO ROLAND • Laboratory of Aquatic Ecology, Department of Biology, ICB, FederalUniversity of Juiz de Fora, Juiz de Fora, MG, Brazil

GERLUZA APARECIDA BORGES SILVA • Department of Morphology, ICB, FederalUniversity of Minas Gerais, Belo Horizonte, MG, Brazil

MARIA CAROLINA S. SOARES • Laboratory of Aquatic Ecology, Department of Biology,ICB, Federal University of Juiz de Fora, Juiz de Fora, MG, Brazil

LONG-SHENG SONG • Division of Cardiovascular Medicine, Department of InternalMedicine, University of Iowa Carver College of Medicine, Iowa City, IA, USA

LÍVIA TEIXEIRA • Laboratory of Cellular Biology, Department of Biology, ICB, FederalUniversity of Juiz de Fora, Juiz de Fora, MG, Brazil

LUCIANA O. VIDAL • Laboratory of Cellular Biology, Department of Biology, ICB,Federal University of Juiz de Fora, Juiz de Fora, MG, Brazil

PETER F. WELLER • Department of Medicine, Beth Israel Deaconess Medical Center,Harvard Medical School, Boston, MA, USA

Section I

Bright-Field Microscopy Applications

Chapter 1

Glycol Methacrylate Embedding for ImprovedMorphological, Morphometrical, and ImmunohistochemicalInvestigations Under Light Microscopy: Testes as a Model

Hélio Chiarini-Garcia, Gleydes Gambogi Parreira, and FernandaR.C.L. Almeida

Abstract

Glycol methacrylate (GMA), a water and ethanol miscible plastic resin, is a medium handy to use forlight microscopy embedding that has a number of advantages than paraffin embedding. The GMAimproves the histological, morphometrical, and immunohistochemical evaluations, mainly due to theaccurate assessment of cytological details. This chapter focuses on our experience in the GMA processingand describes in detail the fixation, embedding, and staining methods that we have been using for testesevaluations.

Key words: Glycol methacrylate, fixation, embedding, light microscopy, testes.

1. Introduction

The first studies that described the seminiferous epithelium struc-ture as we know today emerged at the end of 1950 and start-ing of 1960 (1–3). These studies demonstrated that germ cellsin different steps of the spermatogenic process – spermatogo-nial, spermatocytary, and spermiogenic – are distributed in a well-organized way along the seminiferous tubules. These initial stud-ies were developed mainly in testes fixed in Zenker or Bouinsolutions and embedded in paraffin. Although these histologi-cal processing were adequate to describe the development of theacrossomal system (stained in purple by periodic acid-Schiff), to

H. Chiarini-Garcia, R.C.N. Melo (eds.), Light Microscopy, Methods in Molecular Biology 689,DOI 10.1007/978-1-60761-950-5_1, © Springer Science+Business Media, LLC 2011

3

4 Chiarini-Garcia, Parreira, and Almeida

differentiate steps of the spermatocyte and able to distinguishsome spermatogonia, accurate morphological details of all pro-cess, mainly those related to spermatogonial subtypes, were notclearly demonstrated. At the same time, another morphologicalmethod was also standardized; this was very important for thespermatogonial biology studies. It was a whole mount methodthat analyzed the testis in toto and classified the spermatogonialsubtypes by their grouping, that is, if they were single, in pairor aligned in 4 up to 32 spermatogonia together in the sameclone (4). However, this method was not accurate to distinguishthe spermatogonial subtypes by their morphology. More recently,it was demonstrated that a morphological method used in thepast for important male reproduction researches (5), applyingglutaraldehyde fixation, araldite embedding, and semi-thin sec-tions, could be used for high-resolution light microscopic evalu-ations of the spermatogonial cell in different morphological andmorphometrical approaches (6–13). However, as this method isa pre-preparation for transmission electron microscopy process,fragments have to be very small (2 mm2), allowing adequate pen-etration of fixatives and resins into tissues.

A method that combines different advantages of the methodexposed above and allows morphological, morphometrical, andimmunohistochemical studies in semi-thin or thick sections, inlarge fragments and with satisfactory morphology, is the one thatuses fixation with paraformaldehyde and/or glutaraldehyde andembedding in plastic resin based in glycol methacrylate (GMA).The GMA embedding has been used to present some advantagesover the usual methods (14–16), namely (a) fast processing, (b)hydrosoluble, (c) easy handling, (d) infiltration and polymeriza-tion at room temperature, (e) possible to obtain semi-thin section(0.5 μm), (f) less distortion and artifacts, and (g) better resolutionover light microscopy.

We have been using GMA embedding since 1990 for differentstudies, such as mast cells (17–22), male reproduction (23, 24),equine endometrium (25), and aquatic organisms (26). Here,we present, in detail, our experience in GMA processing, andits tricks, focusing on testis preparation for its high performancestudies.

2. Materials

1. Heparin (Liquemine, Roche).2. Sodium thiopental (Thiopentax, Cristalia) intravenous

bottle.3. Three-way stopcock.

GMA for Improved Investigations Under Light Microscopy 5

4. Catheter (Angiocath, BD).5. Saline (sodium chloride at 0.9%).6. Phosphate buffered solution of 0.1 M and pH 7.4: Dis-

solve 1.38 g NaH2HPO4.H2O (0.1 M) in 100 mL distilledwater (solution A) and 1.42 g Na2H.HPO4 (0.1 M) in100 mL distilled water (solution B). To prepare the bufferat pH 7.4, mix 19 mL of solution A in 81 mL of solutionB. Adjust pH with the same solutions.

7. 8% Paraformaldehyde solution: Heat 70 mL distilled waterat 60–70◦C and add 8 g of paraformaldehyde. Mix welland add drops of NaOH (0.1 M) until a clean solutionis obtained. Wait to get to room temperature before using.

8. 4% paraformaldehyde in phosphate buffer 0.05 M pH7.4: Prepare 100 mL solution by mixing 50 mL of 8%paraformaldehyde, freshly prepared in 50 mL of phosphatebuffer at 0.1 M and pH 7.4.

9. 5% glutaraldehyde in phosphate buffer 0.05 M pH 7.4:Mix 10 mL of glutaraldehyde (biological grade at 50%) in50 mL of 0.1 M phosphate buffer at pH 7.4 and completethe volume to 100 mL with distilled water.

10. Karnovsky’s fixative − 2% paraformaldehyde, 2.5% glu-taraldehyde in phosphate buffer 0.05 M pH 7.4: We haveused the original formula diluted in a half as follows: mix50 mL phosphate buffer (0.1 M), 5 mL glutaraldehyde(50%), 20 mL paraformaldehyde (10%) and complete thevolume to 100 mL with distilled water.

11. Alcohol (from 70 to 100% in distilled water).12. GMA kit (Historesin, Leica).13. Plastic mold for embedding.14. Wooden pin holder.15. Dentist acrylic resin kit: Mix the powder and the liquid to

prepare a viscous medium. Just after this, pour the mixtureinto the mold holes and immediately put a wooden pin.The resin takes only a couple of minutes to polymerize.

16. Razor blade.17. Glass knife: prepared with glass strips (400 × 25 × 6.4 mm)

from Leica in an LKB Knifemaker, model 7880B.18. Toluidine blue-borate: 1 g toluidine blue O (Allied Chem-

ical) and 1 g of sodium borate (Na2B4O7 anidrous) dis-solved in 100 mL distilled water.

19. Erythrosine-Orange-Toluidine: solution A (0.2 g erythro-sine, 1 g orange G in 100 mL distilled water); solution B(toluidine blue-borate, detailed above at item 18).


20. Harris’ hematoxylin: mix 1 g hematoxylin, previously dis-solved in 10 mL ethanol, with 20 g aluminum potassiumsulfate (AlKO8S2 12H2O) previously dissolved in 200 mLof heated distilled water. Immediately, add 0.5 g mercuryoxide (HgO). Take out from the heater and cool the solu-tion by immersing it cold water. To increase the nuclearcontrast, 4% acetic acid can be added to the solution.

21. Mordent solution: add 2% solution of ammonium iron sul-fate (FeH8N2O8S2.6H2O) in distilled water.

22. Eosin solution: mix 1 g yellow eosin dissolved in 10 mLabsolute ethanol in 0.5 potassium dichromate (K2Cr2O7)dissolved in 80 mL distilled water, followed by 10 mL ofsaturated solution of picric acid (for saturation, add 1.4 gpicric acid in 100 mL distilled water).

23. Period acid: periodic acid at 0.5% in distilled water.24. Differentiator solution: mix 6 mL of 10% sodium

metabisulfite (Na2O5S2) and 5 mL of chloride acid 1 N(8.35 mL HCl up to 100 mL distilled water) in distilledwater and complete the volume up to 100 mL.

25. Schiff reactive: dissolve 1 g basic fuchsin in hot water butwithout boiling it. Wait to cool down to 50◦C and thenadd 10 mL of chloride acid (1 N). Wait to cool down to25◦C and add 1 g of sodium metabisulfite. Mix for 1 h andkeep it in a dark place for 24 h at room temperature. Keepthe final solution at 4◦C.

26. 5-Bromo-2-deoxyuridine (Sigma) – diluted 6 mg/mL inphophate buffer solution PBS.

27. Colorfrost Plus Microscope Slides (Fisher-Scientific).28. 0.6% Hydrogen Peroxide (H2O2): Dilute 1 mL of 30%

H2O2 (Sigma-Aldrich) in 49 mL of distilled water.29. 0.1% Protease (Sigma-Aldrich) diluted in PBS.30. 10× phosphate-buffered saline: Dissolve 76.0 g NaCl,

3.6 g NaH2PO4 and 9.94 g Na2HPO4 in 1000 mL of dis-tilled water. To make 1× PBS, dilute 10× PBS at a 1:9 ratioin distilled water. Adjust to pH 7.4 with HCl. Store in glassbottles at room temperature.

31. 2N hydrogen chloride (HCl): Dilute 73 mL of HCl to 1 Lof solution in distilled water. Store the solution in a glassbottle at room temperature.

32. 0.1M sodium borate (Na2B4O7): Dissolve 38.14 g ofNa2B4O7.10H2O (decahydrate) in distilled water, up to1 L. Mix it in a hot plate, until the salt is completelydissolved. Store the solution in a glass bottle at roomtemperature.


33. Triton X-100 (Sigma-Aldrich).34. PBST solution: 0.2% Triton X-100 in 1× PBS. Add 200 μL

of Triton X-100 to 1 L of PBS. Store the solution in a glassbottle at room temperature.

35. Normal horse serum (Sigma-Aldrich) diluted in PBST.36. Primary antibody anti-BrdU B44 (BD Biosciences).37. ABC kit, Mouse IgG (Vector Labs): This kit contains sec-

ondary biotinylated antibody and reagents A and B.38. Normal goat serum (Sigma-Aldrich), diluted in PBST.

39. DAB (3,3′-diaminobenzidine) kit (Vector Labs).40. Xylene.41. Entelan medium (Merck).42. Coverslip.

3. Methods

To obtain a good tissue preparation, all the processing steps haveto be carefully developed. The sum of minimal defect in somesteps can impair the final results. Thus, care has to be takenregarding fixation, embedding, sectioning, and staining. Below,details about each of these steps will be described considering dif-ferent experimental possibilities.

3.1. Fixationand Storage

Glutaraldehyde is a non-coagulant fixative known to cross-linkprotein, preserving very well cellular structures. As protein is auniversal component of cells, found in membranes and in thecytosol, the glutaraldehyde can fix it as a whole, making the cellas a single interlocking structure (27). Besides, this aldehyde reac-tion is not limited to protein. It can also react in a lower degreewith lipids, carbohydrates, and nucleic acids. Formaldehyde is alsoanother aldehyde used for cell fixation. However, it makes lesscross-linking, reducing the tissue meshwork. Considering that therate of penetration of glutaraldehyde into tissues is very low andthe formaldehyde penetration is about five times faster than glu-taraldehyde, a combination of both is frequently used, mainly intissues of difficult penetration and/or in fixation by immersion.This method was described by Karnovsky (28). Fixatives usingpicric acid and alcohol are coagulant. They can denaturate pro-teins, permanently modify their structure and affect the tissueresolution. Thus, for morphological evaluation of testes underlight microscopy, aldehydes are normally chosen. To avoid lowerpH reduction during the fixation procedure and the introduction


of artifacts, a buffering system should be used with the fixative.The main buffers used are phosphate and cacodylate buffer atpH 7.2–7.4.

3.1.1. Testes Fixation byPerfusion of Whole Body

When small animals are used, like rodents, marmosets, opossumsor cats, the best method to preserve whole testes is fixing themby the injection of the fixative into the whole body through thecirculatory system (11). This method of fixation by perfusion isreached by introducing a catheter into the left ventricle, in thedirection of the aorta (Note 1). This needle is also connected totwo vials, one with saline and the other with fixative, througha triway plastic pipe. These vials are positioned ∼1.2 m abovethe heart, reaching a liquid pressure of ∼80 mmHg inside thevascular system. Just before the beginning of the perfusion pro-cess, the right atrium is cut for draining the blood and solutions(saline and fixatives) that will be injected. First, the circulatory sys-tem is perfused with saline for 5–10 min, cleaning of blood cells(Note 2). Immediately after, it initiates the perfusion with thefixative for about 25–30 min (Note 3). Fifteen minutes beforeperfusion, heparin is injected intraperitonealy in the proportionof 125 IU/kg of body weight (Note 4). After perfusion, testesshould be cut in thin slabs and fixed by immersion for 12–24 h at4◦C.

3.1.2. Fixation byPerfusion of IsolatedTestis

When testes of large animals are used, like bull, boar, and ram,the whole body perfusion method becomes very expensive. Inthis way, the orchiectomy is made and only the testes are per-fused. The perfusion is made by introducing the catheter into thetesticular artery, once this artery is easily identified (Note 5). Toavoid blood clumps during the perfusion process, heparin shouldbe added to the saline solution in a proportion of 10 IU/L. Thetime of saline perfusion should be enough to clean the testis fromblood cells (∼5 min) and can be verified visually by the clear liq-uid running from the testicular vein. The fixation time shouldbe between 20 and 30 min (Note 3). Both, saline and fixative,should be perfused at a pressure of ∼80 mmHg. After perfusion,testes should be cut in thin slabs and fixed by immersion for12–24 h at 4◦C.

3.1.3. Testis Fixationby Immersion

If for some reason it is not possible to fix the testis by perfu-sion, they can be fixed by immersion. However, for good mor-phological preservation, some cares have to be taken. The use ofa fixative solution with combined aldehyde (glutaraldehyde andformaldehyde) can be an alternative solution, once formaldehydepenetrates faster and temporarily stabilizes cellular structures.Glutaraldehyde, which penetrates slowly, arrives later and perma-nently stabilizes cellular components. In case of the use of glu-taraldehyde only as a fixative, only thin slabs (∼1 mm thickness)


should be cut from the fragment surface (Note 6). The fixativetime by immersion should be for 12–24 h at 4◦C.

3.1.4. Comments Regardless of the method used, testes fragments can be kept inbuffer after fixation at 4◦C for a long time (Note 7). Testes fixedby perfusion are used to keep the seminiferous tubules together,preserving the interstitial tissues and possibly making more accu-rate morphometrical evaluation of testes compounds. Otherwise,testes fixed by immersion normally show seminiferous tubules dis-persed. The artifactual spaces among them are provoked by thepressure in a soft tissue during the cut processing in small slabs.For immunohistochemistry evaluations, fixatives with glutaralde-hyde are avoided once the resulted cross-link into tissues couldblock the antibody receptors. The most common fixatives usedfor immunohistochemical is the formaldehyde, which in spite ofnot preserving very well morphological details, keeps the tissuereceptors exposed for antibodies.

3.2. Embedding Testes fragments should be progressively dehydrated with alcoholat crescent concentrations (see Note 8), infiltrated and embeddedin GMA, as described below:

Ethanol 70% 30 minEthanol 85% 30 minEthanol 95% 30 minAbsolute ethanol 30 min (2×)Infiltration resin overnightPure resin overnightEmbedding (see Note 9, Fig. 1.1)

3.3. Sectioning The tissue embedded in GMA blocks can be cut from 0.25 μmup to 15 μm thickness (see Note 10). For the best sectioning,excessive resin should be trimmed with a razor blade, keeping aborder of approximately 1–2 mm around all tissue. To obtain anice section, glass knife can be used in a microtome. For disten-tion, sections are floated in distilled water just after the collectionat room temperature (see Note 11). The section should be pickedup with a slide and transferred to a hot plate (70–80◦C) until thewater droplet over the section evaporates (Fig. 1.1).

3.4. Staining

3.4.1. For Morphologicaland MorphometricalStudies

For different histological evaluations, testes embedded in GMAcan be stained by different methods, namely, toluidine blue-borate, erythrosine-orange-toluidine, hematoxylin-eosin, andPAS-hematoxylin. As tissues embedding in GMA present weakstaining, when compared with those embedded in paraffin


7

D

A

C

B

HP

14

WB

Mi

Re

Mo

BL

GK

BL

T

WP

Fig. 1.1. In a, microtome (Mi, Reichert Jung, model 1140/Autocut) setup with GMAblock (BL) and glass knife (GK) for sectioning, evident in the circle (detail in B). More-over, a room temperature water bath (WB) for section distension and a hot plate (HP)for drying and section adherence on the slide. In b, detail of the circle in a, showingalso the tissue (T) facing the glass knife. In c, a slide with histological sections of 7,4, and 1 μm thickness, showing the different staining intensity. d shows componentsused for GMA embedding: plastic mold (Mo), wooden pin (WP), dentist resin (Re), andtissue into the molds under polymerization (black arrows) and block with tissue (BL)after polymerization and with wooden pin adherence.

(13–15), the methods described below were modified from thoseoriginally used to stain testes embedded in paraffin, with the pur-pose of obtaining good staining and contrast and, consequently,


A

s

CB

s s

D E F

s

ss

IHG

sss

J LK

ss s

7µm 1µm4µm

AT

PAS

EOGAT

HE

Fig. 1.2. Photomicrographies of the human seminiferous epithelium stained with toluidine blue-borate (a–c),hematoxilin-eosin (d–f), erythrosine-orange G-toluidine (g–i), and PAS-hematoxylin (j–l). These pictures show the rela-tionship between resolution and section thickness. In the first column (a, d, g, j), sections were obtained at 7 μm andthe spermatocyte (S) heterochromatin were not easily visible. In the second column (b, e, h, k), sections at 4 μm thick-ness allowed the observation of details from the spermatocyte (S) heterochromatin. One micrometer is the thickness ofsections in the third column (c, f, i, l) and more details can be observed in the spermatocyte (S) nuclei. The PAS stainedin purple grains in the cytoplasm of the germ cells, which were differently observed depending on the section thickness.While in J they were compactly observed (arrows), in k some grains can be seen and in L the PAS grains were individuallyobserved. Bar: 10 μm.

best resolution under light microscopy (Fig. 1.2). All the stain-ing methods presented below can be changed depending on thetissue, fixative applied, and the section thickness. Tests shouldbe made to standardize them for each specific experimentalapproach.

3.4.1.1. ToluidineBlue-Borate

Although it presents just one color, exceptions for metachromaticcells/tissues (mast cells, goblet cells, mucus), it is elected as one ofthe best staining for tissues embedded in GMA for good contrastand resolution, mainly in black & white pictures (see Note 12,Fig. 1.2).

1. Put several drops of toluidine blue-borate over the sectionson a slide for 1 min.


2. Rinse the slide with running water to clean the excessivestaining (see Note 13).

3. Remove excess water by gently pressing the slide, with thetissue facing down, over a piece of filter paper.

4. Allow the slide to dry completely at room temperature andmount with a coverslip using any conventional medium.

3.4.1.2.Hematoxylin-Eosin

Even after some methodological modification, the H&E stainingdoes not present very nice contrast in GMA-embedded sections.The contrast can be increased on thicker sections; however, theresolution can be impaired. In order to intensify the H&E stain-ing, the use of a mordant solution (see Note 14) and the increaseof the staining time (Fig. 1.2) are recommended.

1. Put several drops of mordent solution over sections on aslide for 10 min.

2. Rinse the slide with running water for 5 min.3. Put several drops of hematoxilin solution over sections for

15 min.4. Rinse the slide with running water for 5 min.5. Put several drops of eosin solution over sections for 30 s.6. Rinse the slide with running water to clean the excessive

staining (see Note 13).7. Remove excess water by gently pressing the slide, with the

tissue facing down, over a piece of filter paper.8. Allow the slide to dry completely at room temperature and

mount with a coverslip using any conventional medium.

3.4.1.3. Erythrosine-Orange-Toluidine

In paraffin, the trichrome stains are used for cytoplasmic stainscombined with nuclear stains. However, these methods do notwork in GMA. We have applied an alternative method which hasbeen used with relative success for tissues embedded in GMA(Fig. 1.2).

1. Put several drops of solution A (erythrosine-orange) oversections on a slide for 10 min.

2. Rinse the slide with running water to clean the excessivestaining.

3. Put several drops of solution B (toluidine blue-borate) oversections on a slide for 1 min.

4. Rinse the slide with running water to clean the excessivestaining (see Note 13).

5. Remove excess water by gently pressing the slide, with thetissue facing down, over a piece of filter paper.

6. Allow the slide to dry completely at room temperature andmount with a coverslip using any conventional medium.


3.4.1.4.PAS-Hematoxylin

Used for neutral amino-glycol localization in tissues. Althoughthis method has been commonly used for acrosomal identificationin the testis, here it was used to show PAS-positive granules intothe cytoplasm of germ cells (Fig. 1.2).

1. Put several drops of periodic acid solution over sections for20 min.

2. Rinse the slide with distilled water for 5 min.3. Put several drops of Schiff reactive over sections for 60 min.4. Rinse in three baths of differentiator solution in a total of

3 min.5. Rinse the slide with running water for 30 min.6. Put several drops of hematoxylin for 10 min.7. Rinse the slide with running water for 30 min.8. Remove excess water by gently pressing the slide, with the

tissue facing down, over a piece of filter paper.9. Allow the slide to dry completely at room temperature and

mount with a coverslip using any conventional medium.

3.4.2. ForImmunohistochemicalStudies

Some immunohistochemical studies can be performed usingGMA. As an example, 5-bromo-2-deoxyuridine (BrdU) methodhas been used to study cellular cycle in the testis (29). During cel-lular division, BrdU incorporates into the DNA chain and can befollowed during the cellular cycle using antibody against BrdU.BrdU was injected intraperitonealy in a dose of 60 mg/kg ofbody weight one hour before killing the mice. The spermatogo-nia that divided during this time incorporated the BrdU in theirnuclei. After one hour, the mice were fixed by perfusion and thetestes embedded in GMA as described above. Testes sections of5 μm thickness were used for the present evaluation. Examples ofimmunohistological staining of BrdU in spermatogonia are pre-sented in Fig. 1.3.

BA

Fig. 1.3. a and b: Sections of the seminiferous epithelium of a mouse showing immuno-histochemical staining for BrdU in spermatogonia (arrows). In b, note the staining con-centrated in the inner border of the envelope nuclear. Bar: 10 μm.


To perform the BrdU immunostaining in GMA, the follow-ing steps must be taken:

1. Put slides in distillated water for 1 min.2. Prepare 0.6% H2O2 in distilled water and immerse slides in

a 50 mL Coplin jar, at room temperature, for 5 min.3. Wash slides in distilled water for 1 min.4. In a moisture chamber, incubate sections with 0.1%

protease in 1× PBS, for 60 min, at room temperature.5. Rinse slides two times with 1× PBS for 5 min each time.6. Denature sections with 2 N HCl for 50 min, in a Coplin

jar, at room temperature.7. Neutralize the acid with 0.1 M Na2B4O7 for 2 min.8. Wash slides two times in PBST, for 5 min each. Prepare the

blocking solution while washing slides. 10% NHS (normalhorse serum) in PBST is used as blocking solution, beforethe primary antibody incubation step.

9. Incubate slides with the blocking solution, in a moisturechamber, during 30 min at 37◦C. While incubating withthe NHS, prepare the 1◦ antibody (B44) in 10%NHS inPBST, following 1:200 dilution.

10. Pipet off the blocking solution from sections to be tested,keeping one section as a negative control. Add 100 μL ofthe antibody prepared to each section on the slide. Incu-bate slides with the 1◦ antibody for 60 min, at 37◦C, in amoisture chamber.

11. After incubation with the B44, wash slides twice withPBST, for 5 min each. While washing slides, prepare thesecondary biotinylated antibody in 10% NGS (normal goatserum) with PBST, following 1:1000 dilution.

12. Add 100 μL of secondary antibody to each section andkeep it at 37◦C, for 30 min, in a moisture chamber. Pre-pare the ABC reagent while incubating with the 2◦ anti-body. Add 20 μL of Reagent A plus 20 μL of Reagent B in800 μL 1× PBS. Keep the same proportion for all reagentswhen preparing more than 1000 μL.

13. Wash slides twice with PBST, for 5 min each.14. Add 100 μL of the prepared Reagent ABC to sections and

incubate at 37◦C, for 30 min, in a moisture chamber.15. Wash slides twice with 1× PBS, for 5 min each. Prepare

the DAB while washing slides for the last time in 1× PBS.The DAB should be prepared following the manufacter’sinstructions. When using Vector Labs DAB kit, add twodrops of buffer, plus four drops of DAB, plus two drops ofH2O2 to 5 mL distilled water.


16. Add one drop of DAB to each section, for 30 s, thenwash in distilled water. Keep slides in distilled water whilestaining the other slides with DAB.

17. Stain the section with hematoxylin, diluted 1:1 in distilledwater, for 1 min. Wash with tap water for 5 min.

18. Dehydrate slides using 95% ethanol and 100% ethanol for3 min each and emerge slides in xylene for 5 min.

19. Mount slides with conventional mounting media andcoverslip.

3.5.Photomicrography

Photomicrographies were obtained using a microscopy BX-51 inwhich a Q-Color 3 digital camera from Olympus was connected.The obtained images were transferred to a computer through theImage-Pro Express (Media Cybernetics) software and adjusted forresolution (1000 dpi), sharpness, contrast, brightness, and graylevels using Photoshop (Adobe System, Inc., Mountain View,CA). Plates were organized and characters added using the AdobeIllustrator software.

4. Notes

1. The circulatory system should be closed and under pressurefor adequate entrance of the fixative in the aorta, reachingtestes afterwards and exiting by the right atrium. The mostcommon mistake during the perfusion is the needle plac-ing. If a hole is made in the interventricular septum – athin wall that divides the two ventricular chambers – thepressure comes down and the fixative will enter the rightventricle reaching the pulmonary circulatory system. Thismistake will decrease the pressure inside the general circu-latory system and impair testes fixation.

2. The effective time for saline perfusion is the one necessaryfor cleaning the blood vessels. When the saline that comesout of the right atrium is clean, it is time to stop the salineand start the fixative perfusion.

3. As the fixative penetrates the testis by the blood vessels, itreaches all testicular compartments and cells by diffusion.These processes take a long time and require a fixation timeof at least 25 min, even if the animal body is apparently wellperfused.

4. The use of heparin is very important for a successful testesperfusion success (30), as it avoids blood clumps into bloodvessels during the fixative perfusion.


5. For testes fixation directly through the testicular artery, toavoid reflux of solutions the testicular artery has to be tiedwith a line, but not too strong to avoid cutting the vesselwall.

6. For testes fixation by immersion, the albuginea capsule hasto be taken out once it is a dense connective tissue thatavoids fixative penetration by diffusion. If it is not possibleto take the albuginea completely out, small holes or cutsshould be done on it, around the entire testis. Anotheralternative is to cut the testis in large slabs and put themin the fixative by immersion. After 12–24 h, only thin slabsfrom the surface (1–2 mm thickness) of the big slabs shouldbe taken. The rest has to be discharged. Fixation by immer-sion requires a fixative volume of at least 30 times greaterthan the tissue volume, for adequate fixation.

7. If it is necessary to keep tissues in phosphate buffer for along time, even under 4◦C, add one drop of glutaraldehydein the flask to avoid fungi growth.

8. If for any reason it is not possible to dehydrate with alcohol,the water can be eliminated by crescent concentration ofresin in water, such as 50, 70, 80, 90% and finally pureresin.

9. Testes fragments should be smaller than the cutting surfaceof the block and centrally positioned (Fig. 1.1). During thecutting procedure, the border of blocks could be damaged,impairing the histological analyses of the whole tissues.After resin polymerization, a support should be attachedto each resin block to firmly attach them in the microtomefor sectioning. Originally aluminum support has been used.However, in our laboratory we have used wooden pins assupport, which are attached to the resin block using dentistresin (Fig. 1.1).

10. The best section thickness depends on the type of the tissueand the researcher’s interest. When the research requireslow magnification under microscopy (×2 to ×10 objec-tives), the tissue should be sectioned at a range thicknessof 4 to 8 μm, once the stain intensity of the biological tis-sues is proportional to the section thickness. Otherwise,sections of 0.5–3 μm are adequate for high magnifications(×40 to ×100 objectives), once there is less over super-position of the cellular structures. As the correct thicknessdepends on the tissue, if it has more cells or connective tis-sues, we previously define the correct thickness by cuttingsections of the same block from 1 up to 6 μm and collectthem in a same slide. Afterwards, two thicknesses are cho-sen from them to develop the project, one thicker and theother thinner. We have frequently used slides with sectionsof 2 and 4 or 3 and 5 μm.


11. During the sectioning procedure, the section should betaken individually with a forceps and laid down in a waterwith distilled water at room temperature. We should wait acouple of minutes for the section distension and collect itover a clean slide.

12. The green filter has been used in microscopy to increasethe contrast of black and white micrographies.

13. When the staining is excessive or if it is necessary to takethe stain out of the tissue, slides can be immersed severaltimes and quickly in acid–water to clean tissues. Acid–watersolutions are made with chloride or acetic acid at the pro-portion of 0.5–2%. Acid–alcohol solution should not beused once the alcohol wrinkles the GMA.

14. The mordant acts by increasing the electrostatic forces oftissue macromolecules, intensifying stains attachment.

Acknowledgments

These methods were standardized during the development of dif-ferent projects that were partially supported by Brazilian finan-cial foundations (CAPES, CNPq, FAPEMIG, PRPq-UFMG). Wethank Ana Luiza Drumond for helping in the immunohistochem-istry processing.

References

1. Clermont, Y., Perey, B. (1957) Quantitativestudy of the cell population of the seminif-erous tubules of immature rats. Am J Anat100, 241–268.

2. Clermont, Y., Perey, B. (1957) The stagesof the cycle of the seminiferous epitheliumof the rat: practical definitions in PA-Schiff-hematoxylin stained sections. Rev Can Biol16, 451–462.

3. Clermont, Y. (1962) Quantitative analysis ofspermatogenesis of the rat: a revised modelfor the renewal of spermatogonia. Am J Anat111, 111–129.

4. Clermont, Y., Bustos-Obregon, E. (1968)Re-examinations spermatogonial renewal inthe rat by means of seminiferous tubulesmounted ‘in toto’. Am J Anat 122,237–247.

5. Russell, L. D., Clermont, Y. (1977) Degen-eration of germ cells in normal, hypophysec-

tomized and hormone treated hypophysec-tomized rats. Anat Rec 187, 347–366.

6. Chiarini-Garcia, H., Russell, L. D. (2001)High-resolution light microscopic character-ization of mouse spermatogonia. Biol Reprod65, 1170–1178.

7. Chiarini-Garcia, H., Hornick, J. R., Gris-wold, M. D., Russell, L. D. (2001) Dis-tribution of type-A spermatogonia in themouse is not random. Biol Reprod 65,1179–1185.

8. Russell, L. D., Chiarini-Garcia, H.,Korsmeyer, S. J., Knudson, C. M. (2002)Bax-dependent spermatogonia apopto-sis is required for testicular developmentand spermatogenesis. Biol Reprod 66,950–958.

9. Chiarini-Garcia, H., Raymer, A. M., Russell,L. D. (2003) Non-random distribution ofspermatogonia in rats: evidence of niches in


the seminiferous tubules. Reproduction 126,669–680.

10. Bolden-Tiller, O. U., Chiarini-Garcia, H.,Poirier, C., Alves-Freitas, D., Weng, C. C.,Shetty, G., Meistrich, M. L. (2007) Geneticfactors contributing to defective spermatogo-nial differentiation in juvenile spermatogonialdepletion (Utp14b jsd) mice. Biol Reprod 77,237–246.

11. Nascimento, H. F., Drumond, A. L., França,L. R., Chiarini-Garcia, H. (2008) Sper-matogonial morphology, kinetics and nichesin hamsters exposed to short- and long-photoperiod. Int J Androl, 32, 486–497.Doi:10.1111/j.1365–2605.2008.00884.x.

12. Chiarini-Garcia, H., Meistrich, M. L. (2008)High-resolution light microscopic character-ization of spermatogonia, in (Hou, S. X.,Singh, S. R., eds.), Germline Stem Cells, vol450. Humana Press, Totowa, NJ, Methodsin molecular biology, pp. 95–107.

13. Chiarini-Garcia, H., Alves-Freitas, D.,Barbosa, I. S., Almeida, F. R. L. C.(2009) Evaluation of the seminiferousepithelial cycle, spermatogonial kinet-ics and niche in donkeys (Equus asi-nus). Anim Reprod Sci, 116, 139–154.Doi:10.1016/j.anireprosci.2008.12.019.

14. Bennett, H. S., Wyrick, A. D., Lee, S. W.,McNeil, J. H. (1976) Science and art inpreparing tissues embedded in plastic forlight microscopy, with special reference toglycol methacrylate, glass knives and simplestains. Stain Technol 51, 71–97.

15. Cole, M. B., Jr., Sykes, S. M. (1974)Glycol methacrylate in microscopy: a rou-tine method for embedding and sec-tioning animal tissues. Stain Technol 49,387–400.

16. Woodruff, J. M., Greenfield, S. A. (1979)Advantages of glycol methacrylate embed-ding systems for light microscopy. J His-totechnol 2, 164–167.

17. Chiarini-Garcia, H., Machado, C. R. S.(1992) Mast cell types in the lymph nodesof the opossum Didelphis albiventris (Mar-supialia, Didelphidae). Cell Tissue Res 268,571–574.

18. Chiarini-Garcia, H., Ferreira, R. M. A.(1992) Histochemical evidence of heparin ingranular cells of Hoplias malabaricus Bloch.J Fish Biol 41, 155–157.

19. Chiarini-Garcia, H., Pereira, F. M. A. (1999)Comparative studies of lymph nodes mastcell populations form five different marsupi-als species. Tissue Cell 31, 318–326.

20. Chiarini-Garcia, H., Santos, A. A. D.,Machado, C. R. S. (2000) Mast cell types andcell-to-cell interactions in lymph nodes of theopossum Didelphis albiventris. Anat Embryol201, 197–206.

21. Santos, A. A. D., Chiarini-Garcia, H.,Oliveira, K. R., Machado, C. R. S. (2003)Development of different mast cell typesin the opossum Didelphis albiventris. AnatEmbryol 206, 239–245.

22. Rocha, J. S., Chiarini-Garcia, H. (2007)Mast cell heterogeneity between two dif-ferent species of Hoplias sp (Characiformes:Erythrinidae): response to fixatives, anatom-ical distribution, histochemical contents andultrastructural features. Fish Shellfish Immun22, 218–229.

23. Paula, T. A. R., Chiarini-Garcia, H., França,L. R. (1999) Seminiferous epithelium cycleand its duration in capybaras (Hydrochoerushydrochaeris). Tissue Cell 31, 327–334.

24. Neves, E. S., Chiarini-Garcia, H., França, L.R. (2002) Comparative testis morphometryand seminiferous epithelium cycle length indonkey and mules. Biol Reprod 67, 247–255.

25. Amaral, D., Chiarini-Garcia, H., Vale Filho,V. R., Allen, W. R. (2004) Effects of formalinand bouin fixation upon the mare’s endome-trial biopsies embedded in plastic resin. BrazJ Vet Anim Sci 56, 7–12.

26. Melo, R. C. N., Rosa, P. G., Noyma, N.P., Pereira, W. F., Tavares, L. E. R., Par-reira, G. G., Chiarini-Garcia, H., Roland,F. (2007) Histological approaches for high-quality imaging of zooplanktonic organisms.Micron (Oxford) 38, 714–721.

27. Bozzola, J. J., Russell, L. D. (1999) ElectronMicroscopy: Principles and Techniques for Biol-ogists, 2nd edn. Jones and Bartlett, Sudbury,MA.

28. Karnovsky, M. J. A. (1965) A formaldehyde-glutaraldehyde fixative of high osmolarity foruse in electron microscopy. J Cell Biol 27,137A–138A.

29. Shuttlesworth, G. A., de Rooij, D.G.,Huhtaniemi, I., Reissmann, T., Russell, L.D.,Shetty, G., Wilson, G., Meistrich, M.L..(2000) Enhancement of a spermatogonialproliferation and differentiation in irradi-ated rats by gonadotropin-releasing hormoneantagonist administration. Endocrinol 141,37–49.

30. Russell, L. D., Ettlin, R.A., Sinha Hikim, A.P., Clegg, E. D. (eds.) (1990) Histologicaland Histopathological Evaluation of the Testis.Cache River Press, Clearwater, IL.

Chapter 2

Histological Processing of Teeth and Periodontal Tissuesfor Light Microscopy Analysis

Gerluza Aparecida Borges Silva, Adriana Moreira, and José BentoAlves

Abstract

It is possible to obtain histological preparation of teeth and periodontium with satisfactory levels of qual-ity by means of routine histological techniques, since specific cares are implemented during the sampleprocessing. The formation of access ducts for the quick penetration of the fixative solution, the com-plete removal of the demineralizing agent and the increase of the time of dehydration, clearing, andparaffin embedding are some of these cares. A variety of fixing and demineralizing solutions have beenproposed in the literature for teeth and periodontium processing. The author’s’ experience along theyears demonstrated the possibility of satisfactory results with 10% buffered neutral formalin as fixativesolution and 10% pH 7.3 EDTA as demineralizing solution. Sections of 6 μm in thickness obtainedfrom paraffin-embedded samples, stained with hematoxylin and eosin, comply with the most morpho-logical and morphometric evaluations. Besides, this routine protocol allows the use of serial sectioningfor more specific techniques such as histochemical and immunohistochemical analyses, which are suitablefor cellular constituent and extracellular matrix evaluation of teeth and periodontium. For the study ofmineralized phases of isolated human teeth, ground sections can be obtained by the cutting–grindingtechnique. Though it is a recognized method of study, there are some technical difficulties involved,which are little exploited in the literature. This chapter presents a detailed cutting–grinding protocolfor the histological evaluation of undecalcified isolated teeth and routine histology, which can be easilyreproduced in any research or teaching support laboratory.

Key words: Teeth, periodontium, histological processing, buffered neutral formalin, EDTA,cutting–grinding technique.

1. Introduction

The light microscopy evaluation of teeth and periodontiumis often performed by means of two basic histological tech-niques: 1. Cutting–grinding technique: a method for the study


19

20 Silva, Moreira, and Alves

of undemineralized samples and 2. Routine histological tech-nique for the study of demineralized teeth and periodontium. Thecutting–grinding technique is an appropriate method in the evalu-ation of undecalcified isolated teeth, bone, and other hard tissuesthrough macroscopic and light microscopic investigations. Thefact that teeth and periodontium have both tissues of very differ-ent consistencies, constituted by mineralized connective tissuesjuxtaposed to the loose connective tissues, makes more complexthe histological processing of these samples. Artifacts, as displace-ments, can be easily produced and may interfere in samples’ eval-uation. Some researchers have been using the cutting–grindingtechnique for simultaneous evaluation of soft and hard tissuesof teeth, periodontium, bone, and bone-anchored implants withgood histological results (1–5). However, when the study refersonly to evaluation of mineralized portion of isolated teeth, thecutting–grinding technique is a simple method that allows resultswith excellent quality for histological and morphometrical anal-ysis. Though the cutting–grinding technique is a recognizedmethod of study, there are some technical difficulties involved,which are little exploited in the literature. This chapter presents adetailed cutting–grinding protocol for isolated human teeth thatcan be easily reproduced in any research or teaching support lab-oratory.

The soft tissues and organic matrix present in the teeth andperiodontal structures can be evaluated from demineralized sec-tions, prepared according to routine histological technique. Alarge variety of histological methods for demineralized sampleshave been suggested in the literature, with use of different associ-ations between fixing and demineralizing solutions (6–14). How-ever, their protocols are poorly described, making its repeatabilityas well as their comparative analysis difficult.

The quality of histological sections is conditioned to the tech-nique selected and the incorporation of some specific cares dur-ing histological processing of samples, aiming at the simultaneouspreservation of tissues that constitute both the teeth and peri-odontium. Observations related to several cellular events, suchas presence of inflammatory infiltrate and resorption processes,require material of excellent histological quality obtained by stan-dardized and reproducible technique. At least four factors mustbe considered for the selection of the histological method: (a) thecase urgency, (b) the tissue mineralization stage, (c) the purposeof the research, and (d) the staining technique that will be used(14). For example, the more rapid the decalcifier, the more inju-rious are its effects on subsequent staining likely to be. The effectis most noticeable in nucleic acids and manifests itself chiefly inthe failure of nuclear chromatin to take up hematoxylin and basicdyes as readily as undecalcified soft tissues (15).

Histological Processing of Teeth and Periodontal Tissues for Light Microscopy Analysis 21

Our laboratory experience along the years demonstrated thathistological technique considered as routine (use of 10% bufferedneutral formalin as fixative solution, 10% EDTA as demineraliz-ing solution, paraffin embedding and hematoxylin–eosin staining)allow to obtain satisfactory histological sections to the most qual-itative and morphometric evaluations of teeth and periodontium.

The staining with hematoxylin and eosin is the most com-monly used for general purpose in the histological laboratory.Hematoxylin can be thought of being a basic dye when combinedto aluminum, iron, copper, and tungsten salts, having an affinityto nucleic acids of the cell nucleus (16). It binds to acidic struc-tures, structures yielding a blue-purple color. As such, the nucleusstains blue. Eosin is an acidic dye, which stains basic structuresresulting from electrostatic combinations with tissues. The cyto-plasm, collagen, and muscle are usually stained in red or pinkishred by eosin (17, 18).

The staining with Gomori’s trichrome can be an alternativefor the analysis of teeth and periodontium processes of repair,since it shows up the type I collagen – the main component ofthe organic matrix of these tissues. The expected results are col-lagen in blue and nuclei in blue to black (19, 20). Besides, theGomori’s trichrome can be also indicated for the visualization ofvascular alterations, such as hemorrhagic areas, hyperemia, andprocesses of vascular neoformations, since it enhances vessels anderythrocytes (cytoplasm in red), with a higher contrast than thatof hematoxylin and eosin staining.

This chapter presents, in details, protocols of cutting–grinding technique for undemineralized human teeth and routinehistology for the analysis of demineralized samples, both stan-dardized in our laboratory, with notes and orientations, whichmake possible its reproduction with predictable results.

2. Materials

2.1. Cutting–GrindingTechnique:Mineralized Samples

1. Wax blocks for dental carving measuring 1.5 cm × 4 cm ×1.5 cm.

2. Carton cuts, scotch tape and petroleum jelly.3. Synthetic rubber of white silicone 8001; room temperature

vulcanizing (base and HS II catalyser) – Prepare a mixturewith 3% of the catalyser. In the pattern, e.g., we used 100 gof silicone and 3 g of catalyzer, mixed with a wooden spat-ula to form a homogeneous mixture. Silicone 8001 has amold durability of 2 years.


4. Crystal polyester resin (3061) prepared with 10% of styreneand one drop of hardener liquid (MEKP) for each 5 mLof solution (see Note 1). As a pattern, for three blocks ofresin, we use 27 mL of resin + 3 mL of styrene + 6 dropsof hardener liquid.

5. Pins for disposition of teeth during resin embedding. Thesepins can be made either of wood, plastic or acrylic. Exam-ples: toothpicks cuts or cottontail rods.

6. Rapid glue – Super Bonder R©.7. Apparatus:

(a) IsoMet 1000 – Precision saw with diamond waferingblade series 15 LC diamond (6′′ diameter × 0.220 –152 mm × 0.5 mm)/Buehler-USA.

(b) Motored polisher sanding machine.8. Wet sand papers (600, 1000, and 1200 granulation/mm2).9. Wood/aluminum blocks for supporting of dental sections

during the final grinding.10. Xylene. Inflammable: exhibits neurological effects. Can

also cause irritation of the skin, eyes, nose, and throat.Requires fume hood for safe usage.

11. Glass slides and histological cover glass (24 × 50 mm).12. Mounting medium: Entellan R©.

2.2. HistologicalProcessing forDemineralizedSamples

1. Double-face diamond disk adapted in a low rotation equip-ment (microrectifier with flexible axle).

2. Discardable steel blades or surgical knifes.3. 10% Neutral buffered formalin solution:

(a) 36% Formaldehyde solution;Carcinogenic: toxic by inhalation, in contact with skinand if swallowed. Cause burns. May cause sensitizationby skin contact. Use gloves to avoid bare hand contact.May cause heritable genetic damage. Use only in wellventilated areas.

(b) Disodium hydrogen orthophosphate anhydrous –Na2HPO4 – 0.65% w/v;

(c) Sodium dihydrogen orthophosphate monohydrate –NaH2PO4.H2O – 0.4% w/v.Prepare solution 0.65% w/v disodium hydrogen

orthophosphate anhydrous and 0.4% w/v sodium dihy-drogen orthophosphate monohydrate in distilled water.Stir up until obtaining a homogeneous solution. Add 36%formaldehyde solution – 10% v/v. Store it in a dark vial (seeNote 2).


4. Ethylenediaminetetraacetic acid disodium salt 2-hydrate(EDTA). Prepare 10% w/v EDTA aqueous solution pH7.2–7.4.Dissolve EDTA under stir in distilled water at 59◦C. Waituntil the solution reaches room temperature and adjust pHusing sodium hydroxide pastilles. Complete with distilledwater to final volume of solution.

5. Absolute ethanol – The preparation of alcohol solutions 70,80, 90, and 95% by dilution in distilled water is performedby using a Gay-Lussac alcoholmeter.

6. Xylene.7. Embedding agent for histology – pastilles-solidification

point 56–58◦C.8. Microtome blades PTFE coated – Low profile blades.9. Glass slides and cover glasses (24 × 50 mm).

10. Mayer’s albumin: Two white eggs, glycerin 87% and thy-mol. Beat the white eggs to resemble firm snow, thenreserve for a 24 h period at 4◦C, filter through filter paperand add glycerin 1:1 v/v. Add thymol crystals 0.1% w/v(see Note 3).

11. Staining methods:(a) Routine staining with hematoxylin and eosin:

i. Harris hematoxylin solution – ready for use.ii. Putt’s eosin: eosin yellowish p.a.; potassium dichro-

mate p.a.; ethanol p.a. and saturated picric acidsolution (1.2%). Dissolve 1 g of eosin yellowishin 10 mL of ethanol. Dissolve 0.5 g of potassiumdichromate in 80 mL of distilled water and add theeosin solution. Add 10 mL of saturated picric acidsolution (see Note 4).

(b) Staining for collagen fibers: Gomori’s trichromei. Harris hematoxylin solution – ready for use.ii. Gomori’s trichrome solution: chromotrope 2R,

fast green, phosphotungstic acid p.a., acetic acidp.a. Prepare the solution by dissolving 1.8 g ofchromotrope 2R + 0.9 g of fast green + 1.8 gof phosphotungstic acid in 300 mL of distilledwater. Heat up to dissolve. Wait until the solutionreaches room temperature and add 3 mL of aceticacid.

12. Quantitative filter paper – rapid filtration/white band(15 cm diameter)

13. Mounting medium: Entellan R©.


3. Methods

3.1. Cutting–GrindingTechnique:Mineralized Samples

1. Preparation of the mold for silicone (see Note 5): Draw amold in a carton sheet, in a proportional size to the numberof resin blocks (see Fig. 2.1a–c). After drawing the box, thecarton is cut with a scissor and walls fitted with adhesivetape (see Fig. 2.1d).

2. Glue one (or more) block of wax for dental carving insideof the carton box (see Note 6). Fix a little support (about1.0 cm) on one of the surfaces of the block. This fixation ismechanically performed, only with a small pressure of the

A B C

D F

G H I

J K L

E

Fig. 2.1. Sequence of stages for cutting–grinding technique: a histological method for mineralized dental samples. a–hPreparation of silicone mold for attainment of blocks in resin. i–k Preparation of teeth for embedding. l Dental sampleembedded in crystal polyester resin.


pin on the wax surface. Such block(s) will act as mold forthe cavity to be filled with resin in other step. The cou-pled pin generates an orifice that will work as a guide forthe placement of teeth (see Fig. 2.1 h and j). Lubricatethe bottom of the carton box with petroleum jelly beforeapplying the silicone.

3. Wearing gloves, handle the silicone and throw it inside ofthe fitted box (see Fig. 2.1e), covering about 1 cm abovethe block of wax. Wait 24 h for rubber vulcanization (hard-ening). Dismount the box of paper (see Fig. 2.1f). Removeblocks of wax.

4. Preparation of teeth for embedding:(a) Glue a support with a measure such like that used in

blocks of wax, on the palatine/lingual surface of teeth(see Fig. 2.1i). Such procedure is performed with rapidglue (Super Bonder R©) and aims at standardizing theembedding position (see Fig. 2.1 k).

(b) Using a paint brush, overlay the tooth surface with theresin mixture before the embedding so as to reduce thesurface tension.

5. Locate the tooth (teeth) on the center of cavity(ies) forembedding (see Fig. 2.1j, k). Prepare the solution of resin–styrene (see Note 1) and pour it into the silicone mold, incavities molded by blocks of wax.

6. After a 24 h period of resin cure, withdraw the set resin +tooth from the silicone mold (see Fig. 2.1 l).

7. Make a guide for the block cutting, by drawing a dot-ted line on the resin surface passing along the axis of theembedded tooth (detail in Fig. 2.2c). With the aid of thecutting set IsoMet (slicing machine fitted with a diamond-impregnated cutting disc of 0.5 mm thick (see Fig. 2.2a,c)), regulated to the speed of 300 rpm with a loading of100 g (see Note 7), one can section the block, under refrig-eration (see Note 8).

8. Using a motorized polisher sanding machine (seeFig. 2.2b) – polish the halves attained, with the surfaceof teeth turned to the wet sand paper sheet (see Fig. 2.2d),using all the sequence of granulation (600, 1000, and 1200granulation/mm2) (see Note 9).

9. Return with the block to IsoMet cutting set for a newcutting, thus obtaining a slice of about 1.0 mm thick (seeFig. 2.2e).

10. Glue the tooth slice onto a piece of aluminum or planewood (support for gripping), with a mounting mediumEntellan R© (see Fig. 2.2f). The polished side, alreadytreated, must be turned to the support, leaving free the


A B

C D

E F G

H IFig. 2.2. a IsoMet 1000. b APL-04 motorized sand polisher machine. c First section of resin block and detaching of thetooth in two halves. d First finish stage of dental face in refrigerated wet sand paper sheets. e Second section of resinblock for obtaining slices of 1 mm thick. f Gluing of slice onto an aluminum support. g Trimming of the slice and finishingin sand paper sheets 600, 1000, 1200. h Image of slice of the tooth imbedded in resin, after trimming in the polisher.i Ground section prepared from half of the tooth, using a cutting–grinding system. Section ready for mounting andanalysis at light microscope.


side not yet submitted to the polishing. Keep inside ofa drying kiln at 35–37◦C during a 24-h period for resinpolymerization.

11. Perform the same treatment described in item 8 (seeFig. 2.2 g), until obtaining a smooth surface and a thick-ness of 30 μm (see Fig. 2.2 h) that allows the microscopelight to pass through the specimen (see Note 10).

12. Set free the cut obtained, by immersing in xylene the sur-face glued onto the wood.

13. Wash dental sections in two baths of absolute ethanol,under stir or ultrasonic cleaning to remove debris. Imme-diately after, immerse in two baths of xylene (5 min each).

14. Mount the dental cuts on glass slides, with mount-ing medium Entellan R©, overlapping a cover glass (seeFig. 2.2i). Transfer slides to a drying kiln (35–37◦C)during a 24–48 h period.

15. Evaluate at light microscope with partial closing ofcondenser diaphragm. Results are shown in Fig. 2.3.

3.2. RoutineHistologicalProcessing ofDemineralizedSamples

3.2.1. SamplesCollection

1. Rats maxillae fragments: After decapitation of animals byguillotine, heads must be rapidly dissected for skin removal,separation of maxillae and brain removal in order to facilitatethe access of the fixative solution to the remaining tissues.

2. Isolated human teeth: Immediately after exodontics, theradicular apex (apices) must be removed aiming at enlargingthe apical orifice to facilitate the access of the fixative solu-tion to pulpal tissue. Apicectomy must be performed underrefrigeration with new double-face diamond disk.

3.2.2. Sample Fixationand Demineralization

1. Keep specimens in 10% buffered neutral formalin, into darkvials, of large mouth, at 4◦C, for 24 h (see Note 11). Afterthis period of time, replace the fixative solution and keepsamples at room temperature for 24 h (see Note 12). Thevolume of the fixative solution must be at least 20 timeslarger than the tissue to be fixed (21, 22).

2. Wash in running water (3 × 10 min).3. Demineralization with 10% EDTA aqueous solution (see

Note 13) in a shaker at room temperature. The decalcify-ing solution is changed every two days. Demineralization iscontrolled by means of superficial cuttings performed with a


Fig. 2.3. Light photomicrographs of human permanent teeth sections obtained by the cutting–grinding technique. aGround section of a canine. b Image of the tooth crown showing the Enamel (E); Mantle Dentin (MD); Dentinal tubules(DT), and Interglobular dentin (arrows). c Dentin (D); Incremental lines (arrows). d Dentin (D); Enamel (E); Dentin-enamelJunction (DEJ). e Pulp chamber (PC); Tertiary dentin (arrow). f Secundary dentin (arrow). g Dentinal tubules (thin arrows);Enamel lamella (thick arrows); h Dentinal canalicules (tubules) in the radicular wall (RC); Sharpey’s fiber (arrows); Acel-lular cementum (∗). i Root dentin (RD); Cellular cementum (∗). j Granular layer of Tomes (arrows); Cellular cementum (∗).Bars = 400 μm.

steel blade in fragments edges. The attainment of cuts of firmtexture, but without resistance to the steel blade indicatesthe ideal point of decalcification. After 15 days of deminer-alization, it is possible to perform the trimming of samples –reduction of fragments locating the area of interest, trim-ming fragments so as to standardize the embedding position


Fig. 2.4. Schematic drawing showing the preparation of maxillae of rats for histologicalevaluation of demineralized sections of molars and periodontium in longitudinal plane.a After initial demineralization stage in EDTA, maxillae are reduced by means of twocuts: Section 1. Crosscut with discard of pre-maxilla. Section 2. Reduction of the palatalfaces in a plane parallel to the imaginary line traced over molars occlusal surfaces. b Theinclusion of fragments, with palatal surface turned to the plane of microtomy (arrows),allows the simultaneous visualization of teeth (crown and roots) and periodontium. Dueto tipping of first molar mesial root of rodents, it is recommended to maintain part of thediastema region (∗), so as to make possible the visualization of this root and associatedperiodontal structures.

(Fig. 2.4). Samples of human teeth also can be reduced, cutor divided after 30 or 45 days of demineralization, depend-ing on dental type. After this preparation, it is advisable tokeep fragments in the demineralizing solution for a weekmore, in case of maxillae, and for two more weeks in case ofa human teeth, in order to assure the complete demineral-ization of samples.

3.2.3. Tissue Processingfor Paraffin Embeddingand Microtomy

1. Removal of EDTA: wash in running tap water (3 × 30 min+ 1× 14–18 h).

2. Dehydration.3. Clearing.4. Paraffin infiltration: see Table 2.1.5. Embedding in paraffin.6. Microtomy: Sections of 6 μm collected in glass slides cov-

ered by Mayer’s albumin. Allow it to dry inside a kiln(35–37◦C) for 24 h before staining.

3.2.4. Staining It is possible to perform staining in serial sections with hema-toxylin and eosin (H&E) and Gomori’s Trichrome, according toprotocols described in Table 2.2.

Figure 2.5 shows results obtained with histological rou-tine staining technique (H&E) and Gomori’s trichrome staining.


Table 2.1Protocol for dehydration, clearing and paraffin infiltration ofmaxillae of rat and isolated human teeth

StepMaxillaeof rat

Isolatedhuman teeth

2. Dehydration – series of alcohol solutions50% ethanol – 1 × 30 min70%; 80%; 90% ethanol 2 × 30 min 2 × 30 minAbsolute ethanol 3 × 30 min 3 × 60 min

3. ClearingXylene (In exhaustion hood, with

gloves)1 × 20 min 3 × 20 min

2 × 10 min (see Note 14)4. Infiltration in paraffin (inside kiln 58◦C)

Xylene: paraffin (1:1) – 1 × 30 minParaffin 3 × 30 min 3 × 60 min

For evaluation of processing quality, some parameters can betaken as reference: (a) preservation of odontoblasts layers with-out displacement of predentin (Fig. 2.5a–c); (b) preservation ofendothelial wall of blood vessels (Fig. 2.5a); (c) preservation ofcollagen fibers and extracellular matrix (Fig. 2.5a–c, h); and (d)bone trabeculae with osteocytes and endosteum layer wellpreserved (Fig. 2.5i).

4. Notes

1. Formation of small blisters in the preparation of resin iscommon due to its viscosity. The incorporation of styreneto the resin formula induces to lower viscosity and, as aresult, smaller quantity of blisters in the mixture. Styreneprolongs the resin cure stage a little, allowing blisters toflow together and reach the surface so as to be eliminatedbefore resin hardening.

2. The formaldehyde decomposes itself easily in formic acidby light action and may negatively interfere in the stain-ing of sections. The use of buffers neutralizes the action ofthese acids (23). For this reason, 10% neutral buffered for-malin solution is preferably used and must be stocked intoamber vials, for protection against luminosity that can favoracid formic formation.


Table 2.2Protocols for routine staining (hematoxylin and eosin) and for type I collagen fibers(Gomori’s trichrome)

Step H&E Gomori’s trichrome

1. DeparaffinizationXyleneXylene

1 × 30 min2 × 15 min

1 × 30 min2 × 15 min

2. HydrationAbsolute ethanol90% ethanol80% ethanol70% ethanolRunning tap water

3 × 2 min1 × 2 min1 × 2 min1 × 2 min1 × 20 min

3 × 2 min1 × 2 min1 × 2 min1 × 2 min1 × 20 min

3. 1ª staining solution Filtered Harris hematoxylinsolution

1 × 1 min

Filtered Harris hematoxylinsolution

1 × 1 min

4. Running tap water 1 × 20 min 1 × 20 min5. 2ª staining solution Putt’s eosin solution

40 s to 1 min(see Note 15)

Gomori’s Trichrome solution1 × 15 min(see Note 16)

6. Stain washingRunning tap water (10–20 s) (10–20 s)

7. Dehydration70% ethanol80% ethanol95% ethanolAbsolute ethanol

1 × 10 s1 × 10 s1 × 30 s3 × 2 min

––1 × 30 s3 × 2 min

8. ClearingXyleneXylene

2 × 2 min1 × 10 min

2 × 2 min1 × 10 min

9. Mounting (see Note 17) Cover slip using Entellan R©mounting medium

Cover slip using Entellan R©mounting medium

3. The glycerin prevents the slide from completely dryingwhile the egg albumin, a protein, is denatured whenimmersed in 70% ethanol and fixes sections onto the glassslide. The addition of thymol crystals allows the solutionpreservation, without contamination by fungi, at 4◦C evenfor 12 months.

4. Putt’s eosin solution intensely stains the acidophilic struc-tures when recently prepared. Dilution 1:1 in distilledwater is recommended to reduce the staining intensity.

5. The silicone form can be prepared either for one detachedblock or can be amplified with several cavities for simulta-neous embedding of several teeth. Preparation of the resin


Fig. 2.5. a–e Dentin-pulp complex of human tooth. a Panoramic view. HE. Bar = 140 μm. b Layers of pulp. HE. Bar = 50μm. c Dentin-pulp complex stained with Gomori’s trichrome. Bar = 160 μm. d Collagen of dentin bridge (arrowheads)induced by direct pulp capping with calcium hydroxide. Bar = 140 μm. e Presence of extravasated erythrocytes (arrow)in inflammated dental pulp. Gomori’s trichrome. Bar = 160 μm. f Components of periodontium stained with Gomori’strichrome. Bar = 350 μm. g Panoramic view of the first and second molars region in a longitudinal section of rat maxilla.HE. Bar = 600 μm. h Higher magnification of rat periodontium. HE. Bar = 80 μm. i Resorption area of alveolar bonein higher magnification. Osteoclasts (arrows). HE. Bar = 80 μm. Dentin (D); Predentin (Pd.); Vessel (V); Odontoblastlayer (Od); Cell-free zone (CFZ); Cell-rich zone (CRZ); Central layer (CL); Pulp (P); Root dentin (RD); Alveolar bone (AB);Periodontal Membrane (PM) or (∗); Interradicular septum (IS).


for the embedding of several teeth is desirable, since theproportion of the resin components for a larger volume ismore reliable.

6. The wax blocks can be replaced by wood blocks prepared inthe same dimensions or larger than the wax blocks. In thiscase, they must be lubricated with petroleum jelly beforeapplying the silicone.

7. Faster speed as well as heavier weight tends to damage thespecimen surface. With such careful management (200 or300 rpm and 100 g), the detailed structure of specimenswill be preserved from damage.

8. The cut of resin blocks can be handmade by using a double-face diamond disc, in a mandrel, adapted either to an elec-tric micromotor or to a microrectifier with flexible axle.

9. Wear and finishing process of dental sections embeddedin resin can be either manually performed or by meansof other types of electric polisher sand machines (orbital),since it allows the replacement of sand paper sheets. In thecase of manual processing, it is important to fix the wetsand paper sheets onto a plane surface and moisten themconstantly.

10. The thickness of sections is variable. It is important toobtain a homogeneous, semi-transparent, whitish texture.It has been reported that 100 μm thick sections are satis-factory for light microscopy (24).

11. The buffered neutral formalin is considered a universal fix-ative, with good results for most tissues (25). Formalin isnot significantly harmful to any tissue types and for thisreason it allows viewers to observe the tissue in its almostnatural state.

12. Samples can be stocked in buffered neutral formalin for sev-eral days without significant tissue alterations for the rou-tine histological analyses. However, for both histochemi-cal and immunohistochemical techniques, the fixation ofspecimens for at the most 72 h is recommended. We haveobserved that a 48 h period is sufficient for a good fixa-tion and this information is in agreement with the literature(14).

13. A variety of methods has been recommended for decalci-fying the hard tissues. Decalcification can be performedthrough immersion into acids (weak or strong solutions)or into chelant components. All acids, in some way, inter-fere with the tissue stability, in reliance on the acidity ofsolution and the time of sample demineralization. In wholetooth specimens, the distortion of the pulp tissue can beeasily demonstrated if a stronger decalcifying agent is used.


The removal of calcium by the acid produces carbon diox-ide. The pressure of this gas and its movement through thetissue can be one factor causing the separation of connec-tive tissues observed after decalcification (26). The EDTAis a chelant substance that captures metallic ions, amongwhich is calcium, removing them slowly with excellentpreservation of histological details (16). Hard tissues, par-ticularly dental tissues, decalcified by chelators (e.g. 10%EDTA neutralized with sodium hydroxide, pH 7.2), showa minimum of artifact. Staining is of an acceptable stan-dard after using this chelating agent and good results canbe obtained (17). Although these chemical agents are theslowest, it preserves cells and matrix constituents, partic-ularly mucopolysaccharides (27), and gives the best anti-genic/enzymatic preservation. In case of urgent dentaland periodontal analysis, we suggest the utilization of: (a)Planck Rychol’s solution (mixture composed by 7% alu-minum chloride (w/v); 8.5% hydrochloric acid (v/v); and5% formic acid (v/v) or (b) Anna Morse’s solution (mix-ture composed by 50% formic acid solution (v/v) and 20%sodium citrate (w/v) aqueous solution, used 1:1 (14, 28).Good results can be reached, but samples must be moni-tored every day, because these acid solutions demineralizein a short time, depending on the size of samples. Thereare risks of irreparable degradation of the organic matrix,in case the acid solution is kept in a higher time than thatrequired for demineralization of hard tissues of tooth andperiodontium.

14. The third bath of xylene, particularly for dental samples,can extend for even 2 h, for total clarifying of samples. Den-tal tissues clarify very well thus coming to the diaphanousaspect; therefore, any whitish spot inside the sample isa sign of the requirement of extending the clarificationperiod. Nevertheless, the monitoring of the third bath isextremely significant, because the excessive time in xylenecan harden the piece and make the microtomy stage diffi-cult.

15. After the eosin staining and rapid washing in running tapwater, one can proceed the regressive staining (differen-tiation), when the tissue presents itself extremely stained.Regressive staining consists in dipping slices rapidly (forsome seconds) in 1% chloridric acid solution in 70% ethanoland washing in running tap water for 2–3 min.

16. Formalin-fixed tissues may benefit from secondary fixationof sections in Bouin’s fluid, which enhances the red color ofmuscle fibers and epithelial cells in staining with trichromics(29). The Bouin’s fluid considerably enhances trichrome


intensity and brilliance due the action of picric acid in thetissue. After hydration step, plunge sections into Bouin’solution heated at 56◦C (saturated picric acid 750 mL;formaldehyde 250 mL; glacial acetic acid 50 mL) for 1 h.Wash slides in running tap water for 10 min before applica-tion of the advocated stains for Gomori’s technique (20).

17. The Entellan R© is a rapid cure resin and polymerizes in amaximum of 24 h. Alternatively, one can opt for syntheticCanada balsam due to the significantly lower cost whencompared to Entellan R©; however, this medium requiresabout 4–5 days for polymerization.

References

1. Donath, K., Breuner, G. (1982) A methodfor the study of undecalcified bones andteeth with attached soft tissues. The Säge-Schliff (sawing and grinding) technique. JOral Pathol 11(4), 318–326.

2. Lan, W. H., Kwan, H. W., Sunada, I. (1986)Slicing technique for tooth specimens in his-tological preparation. Bull Tokyo Med DentUniv 33(4), 129–136.

3. Rohrer, M. D., Schubert, C. C. (1992)The cutting-grinding technique for histo-logic preparation of undecalcified bone andbone-anchored implants. Improvements ininstrumentation and procedures. Oral SurgOral Med Oral Pathol 74(1), 73–78.

4. Günhan, M., Günhan, O., Celasun, B., Safali,M. (1996) Examination of periodontal tis-sues by a cutting-grinding technique. AustDent J 41(3), 173–175.

5. Cano-Sánchez, J., Campo-Trapero, J.,Gonzalo-Lafuente, J. C., Moreno-Lopes, L.A., Bascones-Martínez, A. (2005) Undecal-cified bone samples: a description of the tech-nique and its utility based on the literature.Med Oral Pathol Oral Cir Bucal 10 Suppl 1,E74–E87.

6. Abreu, E. M. (1990) Processo de reparaçãode feridas de extração dentária, após implantede colágeno microcristalino: estudo his-tológico em ratos. Rev Port de Est e Cir Max-ilofac 31(2), 95–102.

7. D’Souza, N. R., Bachman, T., Baumgardner,K. R., Butler, L. M. (1995) Characteriza-tion of cellular responses involved in repar-ative dentinogenesis in rat molars. J Dent Res74(2), 702–709.

8. Lekic, P., Sodek, J., Mcculloch, C. A. (1996)Osteopontin and bone sialoprotein expres-sion in regenerating rat periodontal ligamentand alveolar bone. Anat Rec 244(1), 50–58.

9. Lamano Carvalho, T. L., Bombonato, K. F.,Brentegani, L. G. (1997) Histometric analy-sis of rat alveolar wound healing. Braz Dent J8(1), 9–12.

10. Terai, K., Takano-Yamamoto, T., Ohba,Y., Hiura, K., Sugimoto, M., Sato, M.,Kawahata, H., Inaguma, N., Kitamura,Y., Nomura, S. (1999) Role of osteo-pontin in bone remodeling caused bymechanical stress. J Bone Miner Res 14(6),839–849.

11. Gonçalves, E. L., Pavan, A. J., Tavano, O.,Guimarães, S. A. C. (2002) Morphogeneticactivity of demineralized dentin matrix: astudy in dogs. Rev Fac Odontol Bauru 10(1),51–56.

12. Silva, G. A. B., Lanza, L. D., Lopes-Júnior,N., Moreira, A., Alves, J. B. (2006) Directpulp capping with a dentin bonding system inhuman teeth: a clinical and histological eval-uation. Oper Dent 31(3), 297–307.

13. Sawada, T., Sugawara, Y., Asai, T., Aida,N., Yanagisawa, T., Ohta, K., Inoue, S.(2006) Immunohistochemical characteriza-tion of elastic system fibers in rat molarperiodontal ligament. J Histochem Cytochem54(10), 1095–1103.

14. Fernandes, M. I., Gaio, E. J., Rosing, C.K., Oppermann, R. V., Rado, P. V. (2007)Microscopic qualitative evaluation of fixationtime and decalcification media in rat max-illary periodontium. Braz Oral Res 21(2),134–139.

15. Bancroft, J. D., Stevens, A., Turner, D. R.(1990) Bone, in Theory and Practice of His-tological Techniques, 3rd edn. Ed. ChurchillLivingstone, New York, NY, pp. 309–342.

16. Tolosa, E. M. C., Rodrigues, C. J., Behmer,O. A., Freitas-Neto, A. G. (2003) Téc-nica histológica, in Manual de Técnicas para


Histologia Normal e Patológica, 2nd edn. Ed.Manole, Barueri, SP, pp. 19–86.

17. Bancroft, J. D., Cook, H. C. (1994) Princi-ples of tissue demonstration and routine mor-phological staining, in Manual of Histologi-cal Techniques and Their Diagnostic Applica-tion. Ed. Churchill Livingstone, New York,NY, pp. 425–428.

18. Burkit, H. G., Stevens, A., Lowe, J. S.,Young, B. (1996) Notes on staining tech-niques, in Wheater’s Basic Histopathology: AColour Atlas and Text. Ed. Churchill Living-stone, New York, NY, p. 277.

19. Gomori, G. (1950) A rapid one-steptrichrome stain. Am J Clin Pathol 20(7),661–664.

20. Behmer, O. A., Tolosa, E. M. C., Freitas-Neto, A. G. (1976) Coloração do tecido con-juntivo, in Manual de Tecnicas para Histolo-gia Normal e Patologica. Ed. EDART – USP,São Paulo, pp. 109–131.

21. Michalany, J. (1980) Fixação, in TecnicaHistologica em Anatomia Patologica: comInstrucoes para o Cirurgiao, Enfermeira eCitotecnico. Ed. EPUP, São Paulo, pp. 40–51.

22. Junqueira, L. C. U., Junqueira, L. M. M. S.(1983) Fixação e descalcificação, in Tecnicasbasicas de Citologia e Histologia, Instituto de

Ciencias Biomedicas e Faculdades de Medicinada USP, Ed. Santos, São Paulo, pp. 13–20.

23. Burck, H. C., Carilla, P. C. (Transl.) (1969)Técnica Histológica: Manual para realizarpreparaciones microscópicas en el laboratorio,Ed. Paz Montalvo, Madrid.

24. Souza, E. M. D., Stott, G. G., Alves, J. B.(1999) Determination of age from cemen-tal incremental lines for forensic dentistry.Biotech Histochem 74(4), 185–193.

25. Grimaldi-Filho, G. (1981) Manual de Téc-nica Histológica. Fiocruz–Centro de Micro-scopia Eletrônica, Rio de Janeiro, 70p.

26. Brain, B. E. (1966) The Preparation of Decal-cified Sections. Charles C Thomas, Spring-field, IL, pp. 69–135.

27. Bélanger, L. F., Copp, D. H., Morton, M. A.(1965) Demineralization with EDTA by con-stant replacement. Anat Rec 153(1), 41–47.

28. Morse, A. (1945) Formic acid-sodium citratedecalcification and butyl alcohol dehydrationof teeth and bones for sectioning in paraffin.J Dent Res 24, 143–153.

29. Tolosa, E. M. C., Rodrigues, C. J., Behmer,O. A., Freitas-Neto, A. G. (2003) Col-oração para tecido conjuntivo, in Manualde Tecnicas para Histologia Normal e Patolo-gica, 2nd edn. Ed. Manole, Barueri, SP,pp. 111–142.

Chapter 3

Large Plant Samples: How to Process for GMA Embedding?

Élder Antônio Sousa Paiva, Sheila Zambello de Pinho,and Denise Maria Trombert Oliveira

Abstract

It is often necessary to process large plant samples for light microscopy studies, but due to structuralcharacteristics of plant tissues, especially intercellular spaces, large vacuoles, and phenolic substances,results are often unsatisfactory. When large samples are embedded in glycol methacrylate (GMA), theircore may not polymerize, remaining soft and moist and making it difficult to cut microtome sections.This situation has been erroneously interpreted as the result of poor infiltration, when the soft core ofthese samples is actually the result of incomplete polymerization. While GMA is in fact present insidesamples, unsatisfactory polymerization results from rapid external polymerization that does not allowsufficient hardener to reach the sample core, while the relatively large volume of GMA inside the tissueblock also dilutes the hardener. In this chapter we propose a new method for processing large plantspecimens that avoids these problems by: (1) slowing the polymerization process through cooling inorder to permit the penetration of hardener into the sample core and (2) increasing the hardener:GMAratio to aid polymerization of the sample core.

Key words: (2-Hydroxyethyl)-methacrylate, historesin, embedding, GMA, HEMA, plant anatomy,plant tissue, resin polymerization.

1. Introduction

Paraffin has traditionally been used in routine histological workwith animal and plant tissue samples, but resin embedding mediacommonly used in electron microscopy has been increasinglyemployed in light microscopy in the last few decades because itpreserves morphological features far better than paraffin mixtures(1). This resin embedding media was developed to resist damage


37

38 Paiva, Pinho, and Oliveira

by the electron beam and high vacuum, and to preserve cellularand subcellular details (1). The glycol methacrylate (GMA, syn-onymous with (2-hydroxyethyl)-methacrylate [HEMA]) (1) iscurrently employed for light microscopy purposes throughout theworld.

Among the advantages of the use of GMA as an embeddingmedium is the ease of preparing heterogeneous tissue sections (1).Glycol methacrylate embedding also has disadvantages, includingdifficulties encountered in embedding large samples.

Instructions provided with the historesin kits are sufficientfor standard samples, usually animal tissue that are relativelyhomogeneous, and when plant tissue samples with similar char-acteristics are used embedding is largely successful. Plant tis-sues generally have greater structural diversity; however, withunique physical and chemical properties they demand spe-cific procedures and methodological adaptations to assure goodresults.

Several tissues are found in plant organs, making them amosaic of cells with different physical–chemical properties. Plantsamples used for anatomical studies vary widely in their propor-tions of intercellular spaces, dimensions, the chemical contentsof their cell vacuoles, and the mechanical properties of their cellwalls. These characteristics can negatively affect procedures ofdehydration, infiltration, and embedding.

The most common problems encountered in processing plantsamples are principally related to difficulties with GMA infiltra-tion, although these can usually be minimized (or solved) by sim-ply reducing the sample size. Sample size reduction, however, canmake the posterior analysis of the material more difficult, andmust often be avoided.

In the specific case of plant tissues, the use of a vacuum inat least one of the processing steps will contribute to minimizinginfiltration problems; the vacuum should preferably be appliedduring fixing as this procedure removes air from the intercellularspaces and quickly replaces it with fixative.

When studies demand whole-organ interpretation (e.g. stud-ies of floral vasculature, fruit and seed structural organization,and nodal anatomy), large samples are essential. These large sam-ples require more time at each dehydration step and accordingto Feder and O’Brien (2), the specific dehydration time inter-vals will depend on the size and permeability of the specimen. Inthe case of infiltration and polymerization of the GMA, however,simple time adjustments alone may not be sufficient to solve theproblem.

This chapter presents solutions to some of the problemsencountered in processing large plant samples that can signifi-cantly increase GMA polymerization efficacy.

Large Plant Samples: How to Process for GMA Embedding? 39

2. Materials

1. Ginger rhizomes (Zingiber officinale Roscoe): cubes(0.8 cm on each side) without periderm and cortical por-tions were prepared using a razor blade.

2. Formalin-aceto-alcohol (FAA): 50 mL of 37%paraformaldehyde, 50 mL acetic acid, and 900 mLof 50% aqueous ethanol.

3. Ethanol.4. Histomolds (1 cm3 cells).5. HEMA (LeicaTM Historesin kit).6. Wood stubs and adhesive (we used epoxy-based glue –

AralditeTM).7. Steel knife: Stainless Steel C-Profile Knife.8. 0.1 M pH 6.8 Phosphate buffer (1.38 g NaH2HPO4.H2O

[0.1 M] in 100 mL of distilled water [solution A] and1.42 g Na2H.HPO4 [0.1 M] in 100 mL of distilled water[solution B]). To prepare the pH 6.8 buffer, 51 mL of solu-tion A was mixed with 49 mL of solution B and the pH wasadjusted with the same solutions.

9. Toluidine Blue O: (to prepare a 100 mL solution, add0.05 g of Toluidine Blue O [Allied Chemical] to 0.1 MpH 6.8 100 mL phosphate buffer).

10. Glass slides.11. Coverslips.

3. Methods

3.1. Fixation Samples were obtained from mature ginger rhizomes. The perid-erm and the cortical portion were removed with a razor blade inorder to prepare tissue cubes (0.8 × 0.8 × 0.8 cm) of the centralcylinder region. This region presents fairly homogeneous struc-tural characteristics, so sample size is a major obstacle to goodembedding.

Fifteen samples were employed in this experimental assay andthese were fixed by immersion in FAA 50 (3) and submitted tomoderate vacuum (180 mmHg) using a vacuum pump for 10 minin order to remove all air from the intercellular spaces and vessellumens (see Note 1). Samples remained in the fixative for 48 h(see Note 2).


3.2. Dehydration Samples were dehydrated in a 50–95% graded ethanol series.The dehydration stages were conducted at room temperature,as detailed below. It is important to emphasize that the timesrequired for dehydration vary according to samples’ sizes andtheir densities. Due to difficulties involved in evaluating param-eters related to absorption rates, we recommend at least 2 h foreach dehydration step for large samples (∼1 cm3).

All alcohol changes must be performed with care to avoidexposure to the open air and minimize air penetration intosamples.

Summary of the dehydration steps:

Step Duration Conditions

50% ethanol 2 h Room temperature

70% ethanol (see Note 2) 2 h Room temperature85% ethanol 2 h Room temperature

95% ethanol (see Note 3) 2 h Room temperature

3.3. Pre-infiltration,Infiltration, andEmbeddingTreatments

After dehydration, samples were submitted to infiltration withGMA historesin; we only recommend the use of GMA activatedby dibenzoylperoxide, as indicated in the historesin embeddingkit. The alcohol is first substituted by pre-infiltration of a mix-ture of 95% ethanol: GMA (1:1) for several hours. This use of thealcohol: GMA solution as an intermediate step with plant sam-ples facilitates infiltration and reduces problems related to resinviscosity (1).

During the embedding stage when the hardener (dimethylsulfoxide) is added to the GMA, resin polymerization time is shortand positioning samples into histomolds must be done efficientlyand quickly. When several samples are to be embedded, we sug-gest putting the histomold in an ice bath, which will delay poly-merization and provide enough time to adjust positions of sam-ples (see Note 4).

Steps referred to here are crucial to processing large sam-ples that frequently show embedding problems due to incom-plete resin polymerization. We experimented with different pre-infiltration/infiltration/embedding repertoires, altering times,temperatures, and hardener concentrations, with five repetitions(five samples) each:

1. Treatment 1 (T1) – The manufacturer’s instructions thataccompany the resin kit (Leica Historesin Embedding Kit)suggested longer infiltration times for plant specimens, butdetails of each step are not specified. According to the man-ufacturer, an intermediate infiltration solution (95% ethanol:GMA, 1:1) is recommended to ensure even sample penetra-


tion. Two hours of intermediate infiltration solution (pre-infiltration) were employed here and samples were thentransferred to a full-strength infiltration solution (GMA) for24 h. The embedding solution used the GMA: hardener pro-portion suggested by the manufacturer (15 mL:1 mL) and itwas placed into histomolds together with the tissue samples.All steps were conduced at room temperature.

This treatment sought to establish the effectiveness ofthe standard protocol with large specimens.

T1 Protocol:


Pre-infiltration (ethanol:GMA, 1:1)

2 h Room temperature

Infiltration (GMA) 24 h Room temperatureEmbedding (GMAa:

hardener, 15:1)∼30 min Room temperature

aThe resin volume must be sufficient to assure complete sample submersion in thehistomold cell.

2. Treatment 2 (T2) – Increasing the pre-infiltration exposuretime to 48 h, followed by extended cold infiltration in GMA(24 h in a refrigerator and 48 h in a freezer). Samples werekept for 48 h in a freezer to increase infiltration time underlow temperatures. After this time period, embedding wasproceeded using the GMA: hardener proportions suggestedby the manufacturer (15 mL:1 mL). Samples were placedinto histomolds and transferred to room temperature forpolymerization.

This treatment sought to determine effects of increasinginfiltration time under cold conditions.

T2 Protocol:


Pre-infiltration (ethanol: GMA,1:1)

48 h Room temperature

Infiltration (GMA) 24 h Refrigerator (±5◦C)Infiltration (same solution as in

previous step)48 h Freezer (±18◦C)

Embedding (GMAa: hardener,15:1)

∼30 min Room temperature

aThe resin volume must be sufficient to assure complete sample submersion in thehistomold cell.


3. Treatment 3 (T3) – Increasing the pre-infiltration expo-sure time to 48 h followed by cold infiltration in GMA(24 h under refrigeration) as in T2, plus cold infiltration ofthe embedding solution (48 h). The T3 protocol examinedeffects of prolonged cold exposure (more than 48 h) to theembedding solution (GMA + hardener). After this cold infil-tration period, histomolds were transferred to room temper-ature for polymerization. It is important to carefully arrangesamples in histomolds, as polymerization can be started dur-ing this stage since the hardener has already been added.

This treatment sought to determine effects of correctinghardener volume by considering the amount of resin presentin the tissue samples and extending GMA + hardener pene-tration time.

T3 Protocol:


Pre-infiltration (ethanol: GMA, 1:1) 48 h Room temperature

Infiltration (GMA) 24 h Refrigerator (±5◦C)Infiltration: Embedding (GMAa:

hardenerb)48 h Freezer (±18◦C)

Embedding (same solution fromprevious step)

∼ 30 min Room temperature

aThe resin volume must be sufficient to assure complete sample submersion in thehistomold cell.bThe proportion of hardener must be calculated considering the resin volume outsidethe tissue sample plus 80% of the tissue sample volume (assuming that the sample isfully infiltrated with resin). In the present treatment, the sample has 0.512 cm3 (0.8× 0.8 × 0.8 cm), which corresponds to a volume of 0.410 mL (80% of 0.512 mL) ofresin inside each block (i.e. 2.05 mL of resin in the five blocks). This value must beconsidered in calculating the hardener volume.

3.4. Microtomyand Staining

The following methodology was employed:1. After 24 h of polymerization in histomolds, the resin blocks

were glued to wooden stubs (Fig. 3.1) using commercialepoxy glue (AralditeTM). These blocks must be kept in alow-humidity environment; we recommend storage undersilica gel. Using this procedure, samples can be maintainedindefinitely before sectioning (see Note 5).

2. Transverse sections (10 μm thick) were obtained using aLeicaTM rotary microtome equipped with a steel knife (C-profile). This microtome has a section counter and we dis-carded the first 400 sections, collecting four sections fromthe median region of each block. Twenty sections weretherefore obtained from each treatment (five blocks, withfour sections from each block).


Fig. 3.1. Blocks obtained from the three treatments; from front to back, T1–T3, each row shows the five repetitions pertreatment.

3. Each section was carefully transferred to a droplet of distilledwater on a glass slide.

4. Slides were dried on a hot plate (50◦C–60◦C) until the waterdroplet completely evaporated (see Note 6).

5. Sections were stained with 0.05% Toluidine Blue, pH 4.7 (4)and mounted in distilled water under a clean coverslip.

3.5. Analysis The 60 sections obtained (20 per treatment) were examined witha light microscope coupled to a camera lucida; the external out-lines of sections were drawn, as were their internal outlines, whennon-polymerized areas were present. Illustrations of the polymer-ized areas were analyzed using the Planimetry System softwarepackage (SPLAN) developed by CINAG, São Paulo State Uni-versity, Botucatu, São Paulo, Brazil, to quantify the polymerizedareas. For statistical analysis, we considered the average polymer-ized area per treatment. Averages were compared using the Tukeytest (p < 0.05).

3.6. Main Resultsand Discussion

The tissue samples processed according to the manufacturer’s rec-ommendations (T1) demonstrated non-polymerized areas in thecore portion of blocks corresponding to 49.36% of their cross-section (Fig. 3.2a). During microtomy, these incompletely poly-merized samples tended to fragment, making it very difficult toobtain whole sections. Additionally, these sections did not spreadwell on the glass slides and sections were only rarely suitablefor microscopic analysis. This situation is often erroneously inter-preted as being due to poor infiltration.

Infiltration proved to be satisfactory and resin could beseen inside samples when using the standard manufacturer’stechniques. On the other hand, although resin penetrated


Fig. 3.2. Blocks and sections obtained from the three treatments. a–c Treatment 1. a General aspect of the block (arrowindicates the non-polymerized area); this is the first repetition of T1, as indicated on the side of the wood stub (1/1).b General view of the section; note limits of hardened resin around the section (arrowhead). (c) Detail of the section,showing the hollow core (asterisk). d–f Treatment 2. d General aspect of the block (arrow indicates the non-polymerizedarea); this is the third repetition of T2, as indicated on the side of the wood stub (2/3). e General view of the section;note limits of hardened resin around the section (arrowhead). f Detail of the section, showing the hollow core (asterisk).g–i Treatment 3. g General aspect of the block without any non-polymerized areas; this is the second repetition of T3,as indicated on the side of the wood stub (3/2). h General view of the section; note limits of hardened resin around thesection (arrowhead). i Detail of the perfectly extended section. Dotted lines in a, d, and g indicate the sample border.

to the sample core, polymerization was not observed in thisregion. If infiltration was therefore adequate, why did poly-merization not occur in the sample core? Apparently, althoughthe sample was well infiltrated, there was not enough hard-ener present for polymerization to occur rapidly and homo-geneously. This is a crucial point, for once hardener is addedto the activated resin and poured into histomolds, the poly-merization occurs quite rapidly, completing in about 30 minat room temperature. This period apparently is not suffi-cient to permit hardener penetration into the tissue sampleand polymerization only occurs in the external resin pool andon a small superficial portion of the sample, generating adecreasing gradient of polymerization towards the sample core(Fig. 3.2b, c).


As explained below, we do not recommend using the manu-facturer’s protocols for processing large samples.

In T2, the pre-infiltration stage was extended from 2 to 48 hand the infiltration stage itself was conducted at low temperaturesand divided into two stages: the first under refrigeration (5◦C)for 24 h and the second in a freezer (−18◦C) for 48 h. Afterinfiltration, the embedding and hardening media was preparedfollowing the manufacturer’s recommendations (15 mL GMA:1mL hardener) and samples were placed into histomolds at roomtemperature for polymerization.

Results of this procedure were similar to those of T1, withpolymerization being restricted to the superficial portion of thetissue blocks; the non-polymerized area occupied 44.61% ofthe sample core (Fig. 3.2d). In addition to poor polymeriza-tion, sections demonstrated problems during mounting and stain-ing, resulting in low quality material for microscopic analysis(Fig. 3.2e, f).

We therefore conclude that increasing the infiltration stageunder cooling improves infiltration but does not facilitate resinpolymerization in the sample core upon addition of the hardener.

In T3, the times of pre-infiltration and infiltration were iden-tical to T2, but the hardener was added in the beginning of thesecond infiltration step when samples were kept in freezer, slow-ing the polymerization reaction. The hardener volume was alsomodified to compensate for the fact that large samples containsignificant resin volumes (∼80% for the ginger samples in thepresent work). Thus the calculation of the amount of hardenerto be added must consider the volume of resin used to composethe block around the sample plus 80% of the volume of the sam-ple. For details of the hardener calculation, see Note 7.

The T3 procedures yielded 100% polymerization of the fivesamples (Fig. 3.2g), resulting in sections with excellent qualityfor histological analysis (Fig. 3.2h, i). What caused these largedifferences in polymerization with this treatment? Two factorswere critical: the addition of extra hardener in proportions appro-priated for resin polymerization inside the sample and the exten-sion of the infiltration/embedding times by maintaining histo-molds with complete embedding media in a freezer. The additionof extra hardener would normally produce even faster polymer-ization than the conventional approach if maintained at roomtemperature, but maintaining a very low temperature allowedthe GMA: hardener solution to penetrate into the tissue mate-rial without polymerizing. For polymerization, see Note 8.

Statistical analyses presented in the following table confirmobservations discussed here, showing the average polymerizedarea in each treatment, compared by Tukey test (averages fol-lowed by same letter do not differ statistically at a 5% level ofprobability):


Treatment Polymerized area (%)

T1 50.64b

T2 55.39b

T3 100.00a

a,bThe average values followed by different letters, in the same column,differ at 5% level of probability in the Tukey test.

It is important to note that this T3 procedure is only nec-essary when bulky or heavy materials are used, for the GMA:hardener solution penetrates very rapidly into small and porousmaterials, ensuring good results with this material while usingstandard procedures.

4. Notes

1. Several authors have noted fixation problems and the occur-rence of artifacts due to the air present in the intercellu-lar spaces in plant tissues. Unfortunately, air removal is nota common practice in plant anatomy laboratories, due inlarge part to the lack of vacuum pumps in plant anatomylaboratories.

We were able to develop an alternative, very inexpensiveand portable method that can substitute the use of a vac-uum pump – a simple disposable syringe. The tissue sampleswere immersed in fixative solution inside a syringe (we foundthe 20 mL variety, without a needle, to be most useful) andthe plunger was then pushed in to remove any air remain-ing in the syringe itself. When a small amount of fixative canbe seen emerging from the syringe opening, seal that aper-ture with your forefinger (do not forget to wear gloves!) andthen pull the plunger back until the inner syringe volumedoubles; maintain the plunger in this position and observethe air bubbles exiting the sample. Repeat this process twiceto ensure complete air removal. Samples are now air-free andthe fixative is in contact with the innermost cells. Addition-ally, air removal facilitates resin infiltration.

In addition to being extremely economical, air removalusing syringes (our portable ‘vacuum pump’) allows ade-quate fixation of plant samples in the field, immediately aftertheir collection.

2. According to Johansen (3), the minimum fixation time forFAA is 18 h and the material can then be left in this solu-tion almost indefinitely without appreciable damage. Thisauthor also remarked that ‘this property of nearly perfect


preservation makes FAA the ideal fluid to take on collectingtrips’. Jensen (5) noted that tissues fixed in FAA for a mini-mum of 4 h can be stored indefinitely, but considered thissolution to be only a moderately good fixative for mostplant tissues: ‘In most tissues, it causes considerable shrink-age, hardens cells and occasionally makes sectioning diffi-cult’. Formalin-aceto-alcohol is the most popular fixative inplant anatomy because it readily penetrates most tissues andis very cheap and easy to use in the field. To avoid problemsdetected by Jensen, samples can be conserved in 70% ethanolafter FAA fixation. Experience has shown that fixation timesshould not be longer than 48 h and that samples should thenbe washed with 50% ethanol (which has the same concen-tration as the fixative and helps remove its residues) beforestoring in 70% ethanol.

If samples will be processed quickly, they can be leftovernight in 60 or 70% ethanol.

3. As GMA is water soluble, complete sample dehydration isnot necessary (as stated in the Leica Historesin Embed-ding Kit) and the ethanol series was run only to commercialethanol levels (92–98%). Do not forget to immediately wash(with ethanol) all the glasswares that come into contact withGMA + hardener.

4. Important: all materials employed during embedding(including hardener, histomolds and GMA, and most glass-ware) must be kept at low temperatures (∼5◦C). Pipettesmust be used at room temperature to avoid imprecision.

If the glassware or histomold are warm, polymerizationoccurs more rapidly, making it difficult for the hardener topenetrate deep inside samples, which will result in unsatis-factory polymerization in cores of large tissue samples. Addi-tionally, rapid polymerization makes it difficult to correctlyposition samples inside the histomold cells.

Maintaining histomolds over ice will guarantee enoughtime to adjust the tissue sample positions.

5. The sample blocks embedded in GMA must be kept undervery low-humidity conditions in a sealed recipient with silicagel crystals because GMA easily hydrates, becoming soft andvery difficult to section.

6. We recommend temperatures of between 50 and 60◦C fordrying the glass slides on hot plates. At higher temperature(>60◦C), air present in the water can form bubbles betweenthe slide and the thin-section, in detriment to staining andmicroscopic analysis.

7. To calculate the amount of additional hardener needed forblocks with large specimens, assume that about 80% of thesample volume is resin. The sample volume must therefore


be multiplied by 0.8 to calculate the amount of resin it con-tains (volume A). Add to volume A the amount of resinrequired to enclose the tissue sample in the histomold (vol-ume B) and then calculate the quantity of hardener neededat the rate of 15 mL of GMA to 1 mL of hardener (volumeC). Quickly mix the volume of resin for the block (volumeB) with the hardener (volume C) in a beaker and quicklytransfer it to the histomold with the sample; then move themold immediately to the freezer. This operation must be per-formed quickly to avoid the beginning of polymerization.Cover histomolds with plastic film in order to protect thefreezer from the volatile historesin, especially after addingthe hardener.

A sample calculation: a tissue sample that is 0.8 × 0.8 ×0.8 cm (i.e. 0.512 cm3) that is properly infiltrated with resinwill have ∼80% of its volume filled by GMA. Volume Ais therefore 0.512 (sample size) times 0.8 (to reach 80%),which equals 0.4096 mL. If histomolds hold 1.0 cm3, then0.488 mL of resin (volume of histomold minus the volumeof the sample, i.e. 1.0–0.512) will be needed to top-off eachcell (this is the volume B). Adding volume A (0.4096) to vol-ume B (0.488) yields the total volume of resin (0.8976 mL),to be used to calculate the hardener volume in a proportionof 15:1. The volume of hardener (volume C) is therefore0.05984 mL (the total volume of resin [0.8976 mL] dividedby 15). Remember that you will use only 0.488 mL of newresin in each histomold because the rest is already inside thetissue sample. Thus for the embedding media you will mix0.488 mL of GMA to 0.05984 mL of hardener per histo-mold cell. Remember, also, that these values are for eachcell, and must be multiplied by the number of blocks to bemade.

If it is difficult to estimate the tissue sample volumebecause it has a very irregular contour, increase the volumeof hardener by 40–80% in tests to determine the most effi-cient proportions.

8. When the embedding mixture becomes viscous at roomtemperature (25◦C), the polymerization process should befinished in an oven at 40◦C until final hardening (∼30 min)as suggested by Igersheim and Cichocki (6).

Acknowledgments

Denise M.T. Oliveira and Elder A.S. Paiva thank CNPq for theirresearch grants. This method was standardized during the devel-opment of projects that were partially supported by Brazilianfoundations (CNPq, CAPES, FAPEMIG, and FAPESP).


References

1. Gerrits, P. O. (1991) The Application ofGlycol Methacrylate in Histotechnology: SomeFundamental Principles. Heidelberg, Ger-man: Leica Gmgh, 80p.

2. Feder, N., O’Brien, T. P. (1968) Plantmicrotechnique: some principles and newmethods. Am J Bot 55, 123–142.

3. Johansen, D. A. (1940) Plant Microtech-nique. McGraw-Hill, New York, NY, 523p.

4. O’Brien, T. P., Feder, N., McCully, M. E.(1964) Polychromatic staining of plant cell

walls by Toluidine Blue O. Protoplasma 59,368–373.

5. Jensen, W. (1962) Botanical Histochemistry:Principles and Practice. W.H. Freeman, SanFrancisco, CA, 408p.

6. Igersheim, A., Cichocki, O. (1996) A sim-ple method for microtome sectioning of pre-historic charcoal specimens, embedded in2-hydroxyethyl methacrylate (HEMA). RevPalaeobot Palynol 92, 389–393.

Chapter 4

Image Cytometry: Nuclear and Chromosomal DNAQuantification

Carlos Roberto Carvalho, Wellington Ronildo Clarindo,and Isabella Santiago Abreu

Abstract

Image cytometry (ICM) associates microscopy, digital image and software technologies, and has beenparticularly useful in spatial and densitometric cytological analyses, such as DNA ploidy and DNA con-tent measurements. Basically, ICM integrates methodologies of optical microscopy calibration, standarddensity filters, digital CCD camera, and image analysis softwares for quantitative applications. Apart fromall system calibration and setup, cytological protocols must provide good slide preparations for efficientand reliable ICM analysis. In this chapter, procedures for ICM applications employed in our laboratoryare described. Protocols shown here for human DNA ploidy determination and quantification of nuclearand chromosomal DNA content in plants could be used as described, or adapted for other studies.

Key words: Image cytometry, C value, nuclear DNA content, chromosomal DNA content,picograms, quantitative microscopy, integrated optical density, Feulgen reaction, genome size.

1. Introduction

Image cytometry (ICM) is an association of microscopy anddigital-image-based techniques (1), particularly useful for spa-tial and densitometric measurements in the cytological area (2,3). This tool has been widely used in quantitative applicationsfor DNA ploidy studies in cancer pathology (4–6) and in abso-lute C value determination in plant (1, 7–9) and animal species(10, 11). Considering these approaches, the microscope, cam-era, and software industries have developed new integrated tech-nologies in order to simplify the ICM laboratorial routine. With


51

52 Carvalho, Clarindo, and Abreu

the evolution in digital cameras and system stations for high-end laser analysis, ICM can be performed in almost all cytol-ogy laboratories using a good digital camera coupled to a lightmicroscope (12).

As main advantages in comparison with others methodolo-gies, like flow cytometry (FCM), ICM carries out analysis in sam-ples with relatively small number of nuclei (13), allowing visualrecognition, in a semi-interactive way, in order to monitor struc-tures of interest (14) and is cost-accessible for most specializedlaboratories.

Various examples of ICM studies in qualitative and quanti-tative DNA analysis have been reported in the following areas:(i) human pathology, mainly in tumorigenesis screening basedon aneuploidy and genome size instability, such as diagnosis andmonitoring of malignant melanomas and squamous cell carci-noma (5, 15), neoplasia (15, 16), cervical intraepithelial lesionsand invasive carcinoma (5), and primary achalasia (17); (ii) ani-mal genome size quantification, especially applied in taxonomicand evolution studies (10, 11); and (iii) plant science, for deter-mination of absolute nuclear DNA values and ploidy level inagronomical (1, 7–9) and ecological species (18). Besides nucleargenome size, chromosomal DNA content has also been quantifiedby ICM, such as in Zea mays (19) and Capsicum annuum (20),resolving the genome size at the chromosomal level. The nuclearand chromosomal DNA contents, reported in picograms (pg)and/or base pairs (bp) have become useful in genome sequencingprojects and genetic mapping (21).

For nuclear and chromosomal DNA densitometric measure-ments by ICM, the system components basically include a com-puter with image analysis software, linked to a digital video cam-era coupled to a microscope (2, 14, 22, 23). Digital images arecaptured from slides subjected to Feulgen reaction (24) and thenstored on hard disk. As images are recorded in pixel matrix, thesystem needs to be calibrated so as to grab images in known spa-tial and densitometric parameters (2). Since pixels do not havean intrinsic value, which depends on the system resolution, spa-tial and optical density (OD) values from micrometric and grayreference scales are attributed to them, respectively. To find themeasurement of interest, the nucleus or chromosome area (μm2)is multiplied by their average OD, resulting in the integrated opti-cal density (IOD). Generally, IOD values are automatically esti-mated by software algorithms (2, 23). Seeing that the IOD ofthe sample is equivalent to the DNA content, this value can beconverted to pg or bp (1). As this methodology combines thehigh technology of optical, electronic, and digital instruments,additional technical details may be found in a wide variety of spe-cialized references. Good theoretical references on digital cam-

Image Cytometry: Nuclear and Chromosomal DNA Quantification 53

era technology and optical microscopy have been overviewed(2, 12, 25).

In this chapter, we describe our ICM protocols for system cal-ibration and setup and for DNA content quantification in the fol-lowing materials: (a) human and plant nuclei and (b) plant chro-mosomes. Protocols presented here are reproducible in our rou-tine and were transcribed to contribute as a useful guideline forcytogenetic and cytometric laboratories. Given the need to assureaccurate measurements, mainly in clinical practice, the applica-tion of ICM methodology requires important quality control ofinstrumentation and procedures (26–32).

2. Materials

The equipments, reagents, labware trademarks and models citedin protocols correspond to those used in our laboratory routine,but similar ones can also be employed. We strongly recommendobserving the hazard classification on the label and the safety datasheet of reagents used, according to the manufacturer’s warnings.All laboratorial precautions should be taken to avoid accidents.

2.1. Calibrationand System Setup

2.1.1. Slide Preparation 1. A 5 mL tube with sodium heparin (Vacutainer R©) for takingchicken red blood cells (CRBCs).

2. CRBCs are obtained from healthy male and female individ-uals (Gallus domesticus).

3. Micropipette and tips.4. Glass microscope slides (26 × 76 × 1 mm), pre-cleaned with

ethanol to remove dirt and oil.5. Hot plate (surface temperature of 50◦C).6. Coplin jar with screw cap.7. Freezer (−20◦C).8. Fixative solution 1: 70% ethanol (Merck R©).9. Fixative solution 2: 4% phosphate-buffered formaldehyde,

freshly prepared from 37% formaldehyde and storedat −20◦C.

2.1.2. Feulgen Solution 1. Distilled water (dH2O).2. 5 M HCl: 42.5% (v/v) 37% HCl. Caution in the prepa-

ration: first add the dH2O, then the acid. Store at roomtemperature.


3. Schiff’s reagent (see Section 2.2).4. 0.5% SO2 water: 5% (v/v) 1 N HCl, 5% (v/v) 10%

metabisulphite. Store at room temperature.5. Immersion oils. Use a refractive index oil (nD = 1.525 −

1.540) between the slide and the coverslip, and a refractiveindex oil (nD = 1.515) between the coverslip and the micro-scope objective lens.

6. Glass coverslips (24 × 50 mm).7. Nail polish.

2.1.3. System Setup 1. Trinocular Photomicroscope (OlympusTM BX-60) with: (a)stabilized light, (b) UPlanFI objective magnification ×40with 0.75 numeric aperture, (c) PlanApo ×60 oil immersionobjective with 1.40 numeric aperture, (d) PlanApo ×100 oilimmersion objective with 1.40 numeric aperture, (e) AplanatAchromat condenser with aperture 1.4, and (f) neutraldensity filter (ND6).

2. Monochromatic charge-coupled device (CCD) digital videocamera of 12 bits gray and frame grabber card (PhotometricsCoolSNAP Pro R© – Roper Scientific, Tucson, AZ, USA) (seeNote 1).

3. Computer Pentium 4 HT, CPU 3.2 GHz, 1 GB RAM,80 GB ROM, Microsoft R© Windows XP Professional V 2002Operational System.

4. Image Pro R© – Plus 6.1 software (Media Cybernetics R©).5. Slide Micrometer (1000 μm, OlympusTM)6. Interference filter (green color, 550–570 nm).7. Neutral density filters (ND6): 0.15, 0.30, 0.40, 0.60, 0.90,

and 2.50 (Edmund Industrial Optics R©, Barrington, NJ,USA).

8. Linear 11 stepped density filter (Edmund IndustrialOptics R©, Barrington, NJ, USA).

2.2. Preparationof Schiff’s Reagent

To prepare 100 mL:1. Dissolve the reagent: in a glass bottle, dissolve 0.5 g basic

fuchsin in 85 mL boiling dH2O, with the help of a magneticshaker (see Note 2).

2. Decolorize with bisulfite: when the solution temperaturereaches 50◦C, add 15 mL 1 N HCl and 2.23 g K2S2O5and shake until dissolving.

3. Allow fuchsin to decolorize: after it cools down, store at4◦C in a dark glass bottle for 24 h, shaking occasionally todissolve any pink precipitate.

4. Remove organic impurities: add 0.703 g (up to 1.0 g/100mL is acceptable if demanded) activated charcoal and shake.


Then, pour through filter paper using a Büchner funnel aslong as needed. Use only enough charcoal to decolorize thestain solution, which should be colorless or champagne color(23, 33).

5. Storage of Schiff’s reagent: store in a tightly capped darkbottle in a refrigerator (4◦C), and use exposing as little aspossible to light (see Note 3).

2.3. Slide Preparationfor PloidyDetermination ofHuman Nuclear DNAby ICM

Items 3–9 under Section 2.1.1, items 1–7 under Section 2.1.2and the following additional materials are needed:

1. Endocervical swab and tube kit for collection of cellularmaterial.

2. Cervical cellular material (sample), collected by an autho-rized professional.

3. Cellular material of female buccal mucosa (standard).4. Refrigerator (4◦C).5. Vortex.6. Microcentrifuge tubes of 2 mL.7. Microcentrifuge.8. Diamond pencil.9. Cytocentrifuge.

10. Filter paper.

2.4. Slide Preparationfor Plant NuclearDNA Quantificationby ICM

Items 4–7, 9 under Section 2.1.1, 1–7 under Section 2.1.2,6, 8 under Section 2.3, and the following additional materialsare needed:

1. Plant nuclei material: root tips from germinated seeds of thestandard (see Note 4) and sample species.

2. Petri dishes.3. B.O.D incubator, with temperature ranging from 28 to

30◦C.4. Enzymatic solution (see Note 5).5. Water bath 34◦C.6. Stereoscopic microscope.7. Scalpel.8. Fixative solution 3: methanol and glacial acetic acid in pro-

portion of 3:1 (v/v), freshly prepared and stored at −20◦C.9. Fixative solution 4: 95% ethanol.

2.5. Slide Preparationfor PlantChromosomal DNAQuantification by ICM

Items 4–7 under Section 2.1.1, 1–4 under Section 2.1.2, 6under Section 2.3, 2–9 under Section 2.4, and the followingadditional materials are needed:

1. Plant chromosomal material: root tips from germinatedseeds.


2. Metaphase blocking solution: a chemical inhibitor formetaphase arrest, such as colchicine, oryzalin, trifluralin,8-hydroxyquinoline and amiprophos-methyl.

3. Fixative solution 5: methanol, 37% formaldehyde and glacialacetic acid in proportion of 17:5:1 (v/v/v), freshly prepared.

3. Methods

3.1. Calibration andSetup of System

3.1.1. Slide Preparation 1. Prepare the slide using the smear technique: drop 5 μL maleor female CRBC near one end of a slide and smear usinganother slide inclined at an angle of approximately 45◦ (seeNote 6).

2. Air-dry the slide through fast arm waving and place it on ahot plate.

3. Store the slide in the fixative solution 1 in a Coplin jarat −20◦C for at least 12 h, then fix it in solution 2 for 1 hat 25◦C.

3.1.2. Feulgen Reaction(see Note 7)

1. After fixation, wash the slide in running water for 10 minand air-dry.

2. Hydrolyze in 5 M HCl at 25◦C. The time should be adjustedfrom 10 to 60 min, according to cell type. Wash the slidethree times (for 3 min each time) in dH2O.

3. Stain the slide with Schiff’s reagent for 12 h at 4◦C, in thedark. Subsequently, wash it three times (for 3 min each time)in 0.5% SO2 water and three times (for 1 min each time) indH2O and air-dry.

4. Drop 10 μL of immersion oil on the slide, add a coverslip,seal with nail polish and store it in the dark until image anal-ysis, but no longer than two days.

3.1.3. Microscopy andDigital System Setup

1. All optical parts (filters, objectives, condenser and eyepieces)should be extremely clean.

2. Turn on the camera, computer and microscope, and wait for15 min (see item 3 under Section 3.1.6) for light stabiliza-tion.

3. Adjust the microscope prior to each slide capture session byapplying the Köhler method to obtain an optimal light pathand, consequently, to reduce stray light. Attention! Do thesetup for each objective magnification.


3.1.4. SpatialCalibration – From Pixelto Micrometer

1. Check your camera and software performance according toinstruction manuals and make sure that particular functionsare correct, especially the black and incident (white) levelsettings.

2. Use the slide micrometer scale and the software spatialcalibration tools to establish the unit of measurement ofthe image from pixel to micrometer and record it in thepreferences.

3.1.5. DensityCalibration – From ODto IOD

1. Couple interference filter to field diaphragm.2. Mount a “blank” slide, without cell material, add oil and lay

a coverslip. Place it on the stage and focus at slide level.3. Close field of the iris diaphragm to a slightly larger size than

the image size.4. Determine OD range (minimum–maximum) by opening the

histogram live-window of the software and adjusting themicroscope light intensity knob, so that the highest graylevel (peak moves to the right) is slightly lower than the max-imum value on the gray scale (Fig. 4.1).

5. Set 1/125 (8 ms) of exposure time for capture, or determineanother value (i.e. 5–30 ms).

6. Use eight accumulated frames and divide by 8, in the accu-mulate settings command.

Fig. 4.1. Histogram showing the 12 bits gray range (0 to 4095 – coordinate x) and thelight intensity peak at the position slightly lower than maximum saturation value on thegray scale.


7. Mount a step tablet with the standard neutral density filtersset to calibrate the OD scale. For that, capture a series ofempty images, interposing each one of these filters (or incombination to generate other OD values) over the interfer-ence filter in the light path (Fig. 4.2).

8. Use these step tablets and the density calibration tool of thesoftware to fit the curve of intensity to OD and save it inpreferences. This calibration process should be carried outfor each objective (Fig. 4.2).

9. Remake the illumination settings before the capture sessionof every new slide.

Fig. 4.2. Plot (left) showing a linear density before calibration. Plot (right) of the calibrated intensity curve obtainedfrom stepped tablet values mounted with standard neutral density filters (ND6 – 0.15, 0.30, 0.40, 0.60, 0.90, and 2.50,Edmund Industrial Optics R©, Barrington, NJ, USA).

3.1.6. Calibration Tests Calibration and evaluation of the image analysis system is per-formed based on three tests: stability (34), linearity (2, 23, 34),and uniformity (22).

1. Stability test: use the frame coordinates x, y on the center ofthe field to measure gray level of one pixel. Capture imagesevery 3 min during 1 h. Plot the data on a spreadsheet toevaluate light source variations. Attention! In capture rou-tine, use the analysis system only after its stabilization time(Fig. 4.3).

2. Linearity test: place the linear 11 stepped density filter onthe microscope stage and capture each step. Use the densitycommand to convert the captured image into a calibratedone and save. Apply the appropriate tool to segment each


Fig. 4.3. Stability test obtained from OD mean values (coordinate y) of one pixel (e.g.frame position: x = 150, y = 150) monitored every 3 min during 1 h (coordinate x). Notethat this analysis system stabilized after 12 min.

step region and calculate the OD. Use the spreadsheet tocompare, by linear regression, the obtained measurementswith the known OD certified for each step (see Note 8 andFig. 4.4).

3. Uniformity test: grab images, using the ×40 objective(×60 or ×100) of one single CRBC nucleus at distinct andwell-distributed visual field positions (e.g. 36). Attention!Be careful to maintain the moved nucleus at the same focus.Calculate its OD values and coefficient of variation (CV)(see Note 9).

3.2. Slide Preparationfor PloidyDetermination ofHuman Nuclear DNAby ICM

3.2.1. ProtocolEmploying Air-DryingTechnique

1. Transfer 1 mL of cellular material (sample or standard) in amicrocentrifuge tube and fill up to 2 mL with fixative solu-tion 1.

2. Centrifuge the material at 100g for 5 min. Discard the super-natant and add 2 mL of fixative solution 1. Repeat this stepthree times, with intervals of 10 min and store at −20◦C forat least 24 h.


Fig. 4.4. Linear 11 stepped density filter (above), showing 11 discrete density stepswith increments of 0.1 (Edmund Industrial Optics R©, Barrington, NJ, USA), used in thelinearity test. Stepped tablet (below) exhibiting the 11 square images captured fromeach step of the filter and showing the 11 OD calibrated. The OD calculated (0.04–0.95)was compared with the certified for each step (0.04–1.00) by linear regression, usingthe spreadsheet. The R2 and SE obtained were 0.999 and 1.14, respectively.

3. Centrifuge the cellular suspension at 100g for 5 min. Discardthe supernatant, add 500 μL of fixative solution 1 and storeat 4◦C.

4. Vortex at medium speed for 40–60 s.5. Carefully drop 50 μL of sample cellular suspension near the

left slide end; air-dry it through fast arm waving and placeit on a hot plate. Repeat this step for the standard materialnear the right slide end (see Note 10).

6. Place the slide in fixative solution 2, in a Coplin jar, at 25◦Cfor 1 h.

7. Repeat steps 1–4 of Section 3.1.2.

3.2.2. ProtocolEmployingCytocentrifugeTechnique

1. Repeat Steps 1–4 of Section 3.2.1.2. Cytocentrifuge 200 μL of sample cellular suspension, so that

the centrifuged material is imprinted centrally on the slide(on the right half), at 1800 rpm for 5 min.

3. Air-dry the slide by fast arm waving and place it on a hotplate.

4. Take the same slide and mount the centrifuge apparatus forstandard cellular suspension, in a way that the centrifugedmaterial is imprinted centrally on the slide, on the oppositeend (i.e. on the left half). Cytocentrifuge at 1800 rpm for5 min.


5. Repeat Step 3.6. Repeat Steps 6 and 7 of Section 3.2.1.

3.3. Slide Preparationfor Plant NuclearDNA ContentQuantification by ICM

1. Germinate seeds (sample and standard) in Petri dishes withdH2O in a B.O.D., at 28–30◦C.

2. Place the root tips in fixative solution 3, making threechanges with 10 min intervals. Store at −20◦C for 24 h.

3. Replace this solution with fixative solution 4, change threetimes and store at −20◦C for 24 h.

4. Put the root tips in a microcentrifuge tubes and maceratethem in an enzymatic solution (see Note 5) at 34◦C in awater bath.

5. Wash the root tips for 10 min in dH2O, place them inthe fixative solution 4, changing it three times and storeat −20◦C for at least 24 h.

6. Using the diamond pencil, trace a soft transversal lineacross the middle of the microscope slide and mark “a” forsample and “b” for standard material, on the end corners.

7. Place the slide inclined at an angle of 25–30◦ under astereoscopic microscope, keeping the slide center in focus.

8. Place the root tip (e.g. sample) right below the traced line.9. Apply the cellular dissociation and air-drying techniques

(35): quickly dissociate the meristem with a scalpel, whiledripping one to three drops of fixative solution 4. Air-drythe slide through fast arm waving and place on a hot plateto completely dry up (see Note 11 and Fig. 4.5).

10. Incline the same slide at 25–30◦ under the stereoscopicmicroscope, this time with the other (empty) slide half infocus.

11. Place another root tip (e.g. standard) left below the tracedline (see Note 10).

12. Repeat Step 9.13. Place the slide immediately in fixative solution 2, in a

Coplin jar, for 12 h at 25◦C.14. Repeat Steps 1–4 of Section 3.1.2.

3.4. Slide Preparationfor PlantChromosomal DNAQuantification by ICM(see Note 12)

1. Germinate the sample seeds in Petri dishes with dH2O in aB.O.D, at 28–30◦C.

2. Treat the root tip meristems with a blocking solution, withtreatment time and concentration being established for eachspecies.

3. Wash roots for 20 min in dH2O and place them in fixa-tive solution 3, changing it three times and store at −20◦Cfor 24 h.


Fig. 4.5. Start position for cellular dissociation procedure of the plant nuclei and chro-mosomes slide preparation. Note that the root tip (white draw) is placed on the slideinclined at 25–30◦.

4. Replace it with fixative solution 4, changing it three times andstore at −20◦C for at least 24 h (20).

5. Wash the root tips for 10 min in dH2O, transfer to microcen-trifuge tubes containing enzymatic solution (see Note 5) andincubate at 34◦C in water bath.

6. Wash roots for 10 min in dH2O, place them in fixative solu-tion 4 and store at −20◦C.

7. Prepare slides using the cellular dissociation and air-dryingtechniques (35): place each root tip on a slide inclined at25–30◦ under a stereoscopic microscope. Quickly dissociateit with a scalpel, while dripping one to three drops of fixativesolution 4. Air-dry the slide by fast arm waving movementsand place it on a hot plate (see Note 11 and Fig. 4.5).

8. Place the slide in a fixative solution 5, in a Coplin jar, at 25◦Cfor 12 h.

9. Repeat Steps 1–4 of Section 3.1.2.

3.5. Image Cytometry

3.5.1. Human NuclearICM

1. Place the slide on a microscope, recheck the microscopesetup and recall the spatial and OD calibration settings.

2. Capture (×40 or ×60 objective) 20–40 isolated and well-preserved nuclei of the standard.


3. Move to the other slide half (without changing the micro-scope setup) and capture over 200 isolated and well-preserved target nuclei of the sample.

4. Make sure that the captured images are saved as calibratedones. When not, open the customized spatial and densitycommands to convert the captured images to calibratedones and save them.

5. Mount a frame with a collection of the standard nuclei as a“family portrait”, using copy and paste, macro-customizedfunctions or programming sources. Repeat the procedurefor the sample nuclei, placing them below the standard col-lection.

6. Select the IOD in the measurement parameters list. Ifthe software does not have this parameter, alternativelyselect AREA and OD and multiply these values to obtainthe IOD.

7. Apply the selection tool around the standard nuclei collec-tion, adjust the appropriate segmentation level and applythe count/size tool to measure the IOD. Repeat this stepfor the sample nuclei collection.

8. Export the data to a spreadsheet for statistical analyses.9. Find the IOD values MODE for the standard nuclei, divide

each one of these values by the MODE value and multiplyby two, in order to convert them into C value (36).

10. Open a histogram window and customize its X coordinatefor 0, 2, 4, 6, 8 units and so on (2 corresponds to G1diploid nuclei).

11. Plot these C values on the X coordinate of the histogramand use as reference.

12. Repeat Step 9 for sample nuclei, plot these C values onthe X coordinate of the histogram and compare with thereference data (see Note 13 and Fig. 4.6).

3.5.2. Plant Nuclear ICM 1. Place the slide under the microscope, recheck the micro-scope setup and recall the customized spatial and OD cal-ibration settings.

2. Capture (×40 or ×60 objective) ten isolated and well-preserved early prophase nuclei of the standard and repeatthe session for late telophase.

3. Move to the other slide end (without changing the micro-scope setup) and repeat the capture sessions for the sample.

4. Repeat Steps 4–8 of Section 3.5.1 (Fig. 4.7).5. Measure the nuclear DNA content, using the formula:


Fig. 4.6. Histogram obtained from DNA ploidy values measurement of cervical nuclei.Note that images of nucleus (above) were selected to show some DNA ploidy level(C = 1.56–7.91).

Fig. 4.7 Collection of nuclei at the early prophase (P) and late telophase (T) captured with ×40 objective. a Nuclei ofPisum sativum L. “Ctirad” used as internal standard (2C = 9.09 pg). b Nuclei of Capsicum annuum L. “Fortuna Super”(2C = 8.10 pg). The CV and 4C/2C ratio obtained were 3.45 and 2.0, respectively. Bar 5 μm.

(I) 1 Cs = (IODs × ICp)/IODp

where

1Cs = 1C nuclear DNA content of the sample;1Cp = 1C nuclear DNA content of the standard;IODs = nuclear IOD value of the sample;IODp = nuclear IOD value of the standard;

3.5.3. PlantChromosomal ICM

1. Place the slide under the microscope, recheck the micro-scope setup and recall the customized spatial and ODcalibration settings.


2. Capture (×100 objective) the ten metaphases with allnon-overlapping and well-preserved chromosomes.

3. Repeat Step 4 of Section 3.5.1.4. Open a metaphase image file and use the software tools to

assemble the karyogram, in the same frame, according tocytogenetic rules.

5. Repeat Step 6 of Section 3.5.1.6. Select chromosomes, adjust the appropriate segmentation

level and apply the count/size tool to measure the IOD ofeach of them.

7. Export the data to a spreadsheet and do the statisticalanalyses.

8. Measure the mean DNA content of each chromosome, usingformulas:

(I) 1Cn = (�2Cn/r)/2(II) IODc = (�IODpc × n)/4

(III) IODt = �IODc

(IV) 1Cc = (1Cn × IODc)/IODt

where

1Cn = mean 1C nuclear DNA content;2Cn = 2C nuclear DNA content;r = number of FCM replications;IODc = mean IOD value of the chromosome with one

chromatid (1C);IODpc = IOD of the chromosome with two chromatids

(2C);n = number of metaphases;IODt = IOD of all chromosomes;1Cc = mean 1C chromosomal DNA content.

4. Notes

1. Besides the PhotometricTM camera, other high-perfor-mance CCD digital cameras can be used (i.e. SonyTM,OlympusTM, AndorTM, and HamamatsuTM). A good tech-nical reference for a choice of the most appropriate digitalcamera for a particular interest can be found at the Scien-tific Digital Camera solutions page (AndorTM technologycatalog – www.andor.com).

2. Alternatively, another reagent can be used: pararosaniline(prepare the solution as recommended by the Merck R©microscopy support).


3. When a white precipitate begins to form, the stain solutionshould be discarded. Schiff’s reagent is commercially avail-able, but sometimes it is not as reliable as freshly preparedsolutions and, therefore, it is not recommended (23, 33).

4. An ideal DNA reference standard should be known inabsolute value, genetically stable, easy to manipulate, avail-able in sufficient quantities for analyses, and should havea genome size close to that of the target species. Differ-ent plant standards have been used, including Arabidopsisthaliana, Raphanus sativus, Pisum sativum (37, 38).

5. Carry out several test sessions to find the optimal time andconcentration of the enzymatic treatment, according to celltype and species. Test the enzymatic solution prepared withpectinase, hemicellulase, cellulase, driselase, separately orcombined.

6. This method is recommended for providing undamagedmonolayers of blood cells and greatly preferred over “push-ing” the blood, which can damage cells (23).

7. The Feulgen reaction principle basically includes: onehydrolysis step, using strong acid to generate free aldehydegroups in the DNA molecule, specifically by splitting up thepurine bases A and G, thus producing “apurinic acid”, andanother step, in which the fuchsin molecule decolorizedwith SO2 can bind and recover its pink color (24, 33).

8. In our laboratory, the software of the image analysis sys-tem automatically calculated R2 = 0.999 in the linearitytest. The R2 and standard deviation considered adequatein the medical area are above 0.99 and below 5%, respec-tively (34).

9. Coefficient of variation below 3% is recommended for ICManalysis in cancer diagnosis (23, 34). Our CV value is lowerthan others of similar tests described in the literature.

10. In the internal standardization procedure, sample and stan-dard nuclei are strictly processed on the same slide (7, 39,40), in order to avoid random instrument drift and varia-tion in the preparation and staining.

11. Alternatively, slides can be prepared using the squashingtechnique: meristems are squashed onto glass slides, cover-slips removed over a cold plate, and slides air-dried (1, 7–9,40). However, the cellular dissociation technique providestarget material preparations flattened on the slide, wellspread and showing little cytoplasmic background, overlapsor structural deformations of the chromatin (20, 35).

12. In this case, there is no direct reference standard. Thenuclear DNA content of the same target species is used,


whose value can be established through nuclear ICM or byFCM, which also uses an internal standard (19, 20).

13. Basic performance standards for sampling, measurement,and scaling in ICM analyses must follow guidelines, con-sensuses, and reports of the European Society for Analyt-ical Cellular Pathology – ESACP (27–32) and algorithmsdescribed by (26).

Acknowledgments

We thank CNPq – Conselho Nacional de Desenvolvimento Cien-tífico e Tecnológico and CBP&D/Café – Consórcio Brasileiro dePesquisa e Desenvolvimento do Café, Brazil, for their financialsupport.

References

1. Vilhar, B., Greilhuber, J., Koce, J. D., Tem-sch, E. M., Dermastia, M. (2001) Plantgenome size measurement with DNA imagecytometry. Ann Bot Fenn 87, 719–728.

2. Chieco, P., Jonker, A., Van Noorden, C. J. F.(2001) Image Cytometry. Microscopy Hand-books 46. Springer, New York, NY, p. 116.

3. Pektas, Z. O., Keskin, A., Ömer, G.,Karslioglu, Y. (2006) Evaluation of nuclearmorphometry and DNA ploidy status fordetection of malignant and premalignant orallesions: quantitative cytologic assessment andreview of methods for cytomorphometricmeasurements. J Oral Maxillofac Surg 64,628–635.

4. Biesterfeld, S., Reus, K., Bayer-Pietsch, E.,Mihalcea, A. M., Böcking, A. (2001) DNAimage cytometry in the differential diagno-sis of endocervical adenocarcinoma. CancerCytopathol 93, 160–164.

5. Böcking, A., Nguyen, V. Q. H. (2003) Diag-nostic and prognostic use of DNA imagecytometry in cervical squamous intraepithe-lial lesions and invasive carcinoma. CancerCytopathol 102, 41–54.

6. Grote, H. J., Nguyen, H. V. Q., Leick, A. G.,Böcking, A. (2004) Identification of progres-sive cervical epithelial cell abnormalities usingDNA image cytometry. Cancer Cytopathol102, 373–379.

7. Greilhuber, J., Ebert, I. (1994) Genomesize variation in Pisum sativum. Genome 37,646–655.

8. Greilhuber, J., Obermayer, R. (1997)Genome size and maturity group in Glycinemax (soybean). Heredity 78, 547–551.

9. Baranyi, M., Greilhuber, J. (1999) Genomesize in Allium: in quest of reproducible data.Ann Bot Fenn 83, 687–695.

10. Gregory, T. R. (2001) The bigger theC-value, the larger the cell: genome size andred blood cell size in vertebrates. Blood CellMol Dis 27, 830–843.

11. Gregory, T. R. (2003) Genome size estimatesfor two important freshwater molluscs, thezebra mussel (Dreissena polymorpha) and theschistosomiasis vector snail (Biomphalariaglabrata). Genome 46, 841–844.

12. Russ, J. C., Russ, J. C. (ed.) (2008) Introduc-tion to Image Processing and Analysis. CRCPress – Taylor & Francis Group, Upper Sad-dle River, NJ, p. 355.

13. Greilhuber, J. (2008) Cytochemistry andC-values: the less-well-known world ofnuclear DNA amounts. Ann Bot Fenn 101,791–804.

14. Rodenacker, K., Bengtsson, E. (2003) Afeature set for cytometry on digitizedmicroscopic images. Anal Cell Pathol 25,1–36.

15. Bollmann, R., Méhes, G., Torka, R., Spe-ich, N., Schmitt, C., Bollmann, M. (2003)Human papillomavirus typing and DNAploidy determination of squamous intraep-ithelial lesions in liquid-based cytologic sam-ples. Cancer Cytopathol 99, 57–62.


16. Motherby, H., Pomjanski, N., Kube, M.,Boros, A., Heiden, T., Tribukait, B., Böck-ing, A. (2002) Diagnostic DNA-flow vs.image-cytometry in effusion cytology. AnalCell Pathol 24, 5–15.

17. Gockel, I., Kämmerer, P., Brieger, J., Hein-rich, U. R., Mann, W. J., Bittinger, F.,Eckardt, V. F., Junginger, T. (2006) Imagecytometric DNA analysis of mucosal biopsiesin patients with primary achalasia. World JGastroenterol 12, 3020–3025.

18. Temsch, E. M., Greilhuber, J., Krisai, R.(1998) Genome size in Sphagnum (PeatMoss). Bot Acta 111, 325–330.

19. Rosado, T. B., Carvalho, C. R, Saraiva, L.S. (2005) DNA content of maize metaphasicA and B chromosomes determined by imagecytometry. Maize Genet Cooperation Newsl79, 48–49.

20. Abreu, I. S., Carvalho, C. R., Clarindo, W. R.(2008) Chromosomal DNA content of sweetpepper determined by association of cytoge-netic and cytometric tools. Plant Cell Rep 27,1227–1233.

21. Doležel, J., Bartoš, J. (2005) Plant DNA flowcytometry and estimation of nuclear genomesize. Ann Bot Fenn 95, 99–100.

22. Puech, M., Giroud, F. (1999) Standardiza-tion of DNA quantitation by image analysis:quality control of instrumentation. Cytometry36, 11–17.

23. Hardie, D. C., Gregory, T. R., Hebert,P. D. N. (2002) From pixels to picograms: abeginners’ guide to genome quantification byFeulgen image analysis densitometry. J His-tochem Cytochem 50, 735–749.

24. Feulgen, R., Rossenbeck, H. (1924)Mikroskopisch-chemischer Nachweis einerNukleinasäure von Typus der Thymonuk-leinsäure und die darauf beruhende selec-tive Färbung von Zellkernen in mikroskopis-chen Präparaten. Hoppe Seylers Z Physiol Chem135, 203–248.

25. Andor TM Technology (2006) ScientificDigital Camera Solutions Catalog, p. 316.

26. Böcking, A., Adler, C. P., Common, H.H., Hilgarth, M., Granzen, B., Auffer-mann, W. (1984) Algorithm for a DNA-cytophotometric diagnosis and grading ofmalignancy. Anal Quant Cytol 6, 1–8.

27. Chieco, P., Jonker, A., Melchiorri, C., Vanni,G., Van Noorden, C. J. F. (1994) Anuser’s guide for avoiding errors in absorbanceimage cytometry: a review with originalexperimental observations. Histochem J 26,1–19.

28. Böcking, A., Giroud, F., Reith, A. (1995)Consensus report of the ESACP task force

on standardization of diagnostic DNA imagecytometry. Anal Cell Pathol 8, 67–74.

29. Haroske, G., Dimmer, V., Meyer, W., Kunze,K. D. (1997) DNA histogram interpretationbased on statistical approaches. Anal CellPathol 15, 157–173.

30. Giroud, F., Haroske, G., Reith, A., Böcking,A. (1998) ESACP consensus report on diag-nostic DNA image cytometry. Part II: specificrecommendations for quality assurance. AnalCell Pathol 17, 201–208.

31. Haroske, G., Giroud, F., Reith, A., Böck-ing, A. (1998) ESACP consensus report ondiagnostic DNA image cytometry. Part I:basic considerations and recommendationsfor preparation, measurements and interpre-tation. Anal Cell Pathol 17, 189–200.

32. Haroske, G., Baak, J. P. A., Danielsen,H. et al, Giroud, F., Gschwendtner, A.,Oberholzer, M., Reith, A., Spieler, P.,and Böcking, A. (2001) Fourth updatedESACP consensus report on diagnostic DNAimage cytometry. Anal Cell Pathol 23,89–95.

33. Chieco, P., Derenzini, M. (1999) The Feul-gen reaction 75 years on. Hystochem Cell Biol111, 345–358.

34. Vilhar, B., Dermastia, M. (2002) Standard-ization of instrumentation in plant DNAimage cytometry. Acta Bot Croat 61, 11–26.

35. Carvalho, C. R., Clarindo, W. R., Almeida,P. M. (2007) Plant cytogenetics: still look-ing for the perfect mitotic chromosomes.Nucleus 50, 453–463.

36. Gonçalves, S., Haas, P., Spada, C., Fontana,C. S., Rangel, L., Mendonça, M. A. C., Car-valho, C. R. (2007) Citometria de imagemdo conteúdo de DNA nuclear de célulasepiteliais do colo uterino. Rev Bras de AnalClın 39, 71–78.

37. Doležel, J., Sgorbati, S., Lucretti, S. (1992)Comparison of three DNA fluorochromes forflow cytometric estimation of nuclear DNAcontent in plants. Plant Physiol 85, 625–631.

38. Johnston, J. S., Bennett, M. D., Rayburn,A. L., Galbraith, D. W., Price, H. J. (1999)Reference standards for determination ofDNA content of plant nuclei. Am J Bot 86,609–613.

39. Thunnissen, F. B. J. M., Perdaen, H., For-rest, J. (1992) Influence of different cellextraction methods on cytometric features.Cytometry 13, 485–489.

40. Greilhuber, J., Temsch, E. M. (2001) Feul-gen densitometry: some observations rele-vant to best practice in quantitative nuclearDNA content determination. Acta Bot Croat60, 285–298.

Chapter 5

Histological Approaches to Study Tissue Parasitism Duringthe Experimental Trypanosoma cruzi Infection

Daniela L. Fabrino, Grazielle A. Ribeiro, Lívia Teixeira,and Rossana C.N. Melo

Abstract

During acute infection with the parasite Trypanosoma cruzi, the causal agent of Chagas’ disease, tissuedamage is related to intense tissue parasitism. Here we discuss histological approaches for an optimalvisualization and quantification of T. cruzi nests in the heart, the main target organ of the parasite. Theseanalyses are important to evaluate the course of the infection in different experimental models and alsocan be used to investigate parasite colonization and inflammatory processes in other infected tissues andbiopsies.

Key words: Chagas’ disease, Trypanosoma cruzi, histology, heart, inflammation, glycol methacry-late, animal models.

1. Introduction

The flagellated protozoa T. cruzi is the causal agent of Chagas’disease (also known as American trypanosomiasis), discoveredaround a century ago by the Brazilian physician Carlos Chagas(1). This disease remains a major problem with a great impacton public health in Latin America. Acute infections are usuallyasymptomatic, but the resulting chronic T. cruzi infections leadto high ratios of mortality, associated mainly with heart lesions(2, 3).

This parasite has an obligate intracellular, proliferative, non-flagellate form, termed amastigote. Once the infection has estab-lished in vertebrate hosts, intracellular parasite replication occurs


69

70 Fabrino et al.

as amastigotes, followed by the release of flagellate forms (termedtrypomastigotes) that can be carried by the bloodstream to infectdifferent organs, especially the heart (reviewed in (2, 4, 5)).

Because of the intense parasitism in the myocardium inducedby the acute Chagas’ disease, histological analyses of the heartare frequently performed during experimental studies to evaluatethe course of the infection. However, the value of the histologi-cal techniques in enabling high-quality imaging of parasite nestscan be limited by inadequate fixation, handling and/or process-ing/analysis of the tissue of interest. Here we report on the appli-cation of appropriate histological approaches for improved tissuemorphology and optimal visualization and quantification of para-site nests in the heart. These approaches include the right choiceand preparation of the fixative, embedding with a plastic resin forimproved tissue resolution and adequate analysis of the semi-serialsections. These basic methodologies can also be used to inves-tigate parasitism in other infected organs as well as to evaluatethe presence of inflammatory infiltrates, which generally occurs inparallel to the tissue colonization by the parasite.

2. Materials

2.1. For Fixation 1. Phosphate buffer, 0.2 M, pH 7.3 (stock solution). For bufferpreparation, dissolve 2.75 g of monobasic sodium phosphate(Na H2PO4.H2O, MW = 137.99) into 50mL of distilledwater and complete the volume up to 100 mL (final vol-ume). Prepare another solution with dibasic sodium phos-phate (Na2 HPO4.7H2O, MW = 268.14): dissolve 5.36 gof this salt into 50 mL of distilled water and complete thevolume to 100 mL (final volume). To obtain the buffer atthe pH 7.3 mix 20 mL of phosphate monobasic solutionwith 77 mL of phosphate dibasic solution. Adjust to pH 7.3(see Note 1).

2. Paraformaldehyde (4 g) is dissolved in 40 mL of distilledwater. Dilutions should be made in fume hood (see Note 2)under heating at a hot plate. Add 3–5 drops of 0.1 N sodiumhydroxide and heat the solution until 60◦C (do not allowgoing over this temperature). When the paraformaldehydesolution is completely mixed, cool to room temperature andcomplete the volume to 50 mL (final volume). Mix 50 mLof the paraformaldehyde solution with 50 mL of phosphatebuffer 0.2 M. Adjust to pH 7.3. The final concentration ofthe fixative will be 0.1 M. Fresh solutions of paraformalde-hyde should be used in each experiment for optimal fixation.

Histological Approaches to Study Tissue Parasitism 71

2.2. For Processingand Embedding

1. Glycol methacrylate (GMA) embedding kit preparedaccording to the manufacturer’s instructions. The GMA kitis composed of basic resin (solution of GMA monomer:2-hydroxyethyl methacrylate) and other solutions such aspolyethylene glycol and benzoyl peroxide.

2. Ethanol (from 70 to 100% in distilled water).3. Plastic molds for embedding.

2.3. For Sectioning 1. Microtome.2. Knife maker.3. Glass knives − 400 × 25 × 6.4 mm.4. Histological bath or large beaker.5. Slides and coverslips.

2.4. For Stainingand Mounting

1. Hematoxylin solution (Harris’ hematoxylin) can be pur-chased as a ready solution or alternatively be prepared. Forpreparation, dissolve 1 g of hematoxylin powder in 10 mLof 95% ethanol (solution I). Prepare the solution II: in aflask, to 200 mL of distilled water, add 20 g of potassiumalum. Place the flask on a heater/stirrer, turn on the heaterand allow mixing until the alum dissolves – this takes about15 min. Mix solutions I and II quickly and allow boilingfor 1 min. Add 0.5 g of red mercury oxide. The solutionbecomes dark purple. Take out the flask from the heater andallow cooling by immersing the flask in a container with coldwater. After that, the solution must be stored for at least 48 hbefore use. The indicative sign that the solution is ready touse is the uprising of a metallic layer on the surface. Hema-toxylin solution lasts for years, but eventually can deterio-rate. The hematoxylin solution can be used as concentratedor diluted 1% in distilled water. To increase the nuclear con-trast, 4% acetic acid can be added to the solution.

2. Two percent ammonium iron sulfate (also known as ferricammonium sulfate or iron alum) dissolved in distilled water.

3. Yellow eosin solution is prepared by dissolving 1 g yelloweosin in 10 mL absolute ethanol and mixing this solution in0.5 potassium dichromate dissolved in 80 mL distilled water.Just after, add 10 mL of saturated solution of picric acid(for saturation, add 1.4 g of picric acid in 100 mL distilledwater). Eosin preparation should be made in fume hood.Alternatively, eosin solution can be purchased as a readysolution.

4. Acid–alcohol solution is prepared by dissolving 1 mLhydrochloric acid (HCl) (0.5–2%) plus 99 mL of 70%

72 Fabrino et al.

alcohol. Alternatively, an acid–water solution prepared withHCl or acetic acid (0.5–2%) can be used.

5. Toluidine blue-borate solution is prepared by dissolving 1 gsodium tetraborate (Na2B4O7) and 1 g toluidine blue O in100 mL of distilled water.

6. Basic fuchsin solution (stock solution) prepared by dissolv-ing 1 g basic fuchsin powder in 100 mL of 50% ethanol. Theworking solution is prepared by adding 3 mL of this solutionto 60 mL of distilled water.

7. Mounting medium for histology.

3. Methods

3.1. Fixation,Processing, andEmbedding

Fixation is a crucial step during sample preparation. This process isintended to preserve cell structure by avoiding tissue autolysis. Inaddition, fixation inhibits bacterial and fungal growth, and makesthe tissue resistant to damage during the subsequent processing.Aldehyde solutions fix tissues by introducing cross-links betweendifferent tissue components (proteins, nucleic acids, and lipids).Generally, tissue fragments are fixed at least for 12 h. However, itis not recommended to fix more than 24 h.

Sample collection for fixation requires a special care. Tissuesmust be collected as fast as possible, cut into small pieces (toenable optimal fixation) and promptly transferred to the fixative.If necessary, samples may be quickly washed in saline before cut-ting to clean excessive blood. It is very important to handle sam-ples very gently to prevent mechanic damage that can lead to mor-phological alterations. For histological studies of the heart, it isimportant to define the anatomical region of interest (if atriumor ventricle, for example) before organ fragmentation. After thisprocedure, organ pieces are then carefully transferred to previ-ously and adequately labeled vials containing the fixative. Fixa-tion is followed by sample dehydration, a mandatory step beforeembedding.

Embedding is performed by using a resin (glycol methacry-late) instead of paraffin (6). Glycol methacrylate is a water andethanol-miscible plastic resin. One advantage of this resin is toavoid tissue damage induced by heating required for the embed-ding step with paraffin. Another critical advantage of plasticresin is that it permits increased tissue resolution, which is cru-cial to visualize fine cellular morphological details. Better reso-lution is due to plastic polymerization that causes less shrinkageand retraction compared to conventional techniques. Embeddingwith GMA has other important advantages when compared with


usual methods. This plastic resin is hydro-soluble, easily handled,and enables faster processing compared to paraffin embedding.Yet, it allows microtomy with section thickness varying from 0.5to 12 μm (7, 8). The use of GMA embedding for improved his-tological, morphometrical, and immunohistochemical evaluationsis also discussed in Chapter 1.

Fix samples overnight at 4◦C (see Note 3). Refer to Section2.1 for fixative preparation.

1. Transfer samples to vials with cold phosphate buffer 0.1 Mand, if necessary, keep at 4◦C. Refer to Note 4 for samplestorage.

2. Dehydrate samples. Dehydration step consists of successivebaths in graded ethanol (70, 80, 90% – 20 min in each solu-tion – followed by 100% ethanol (two changes of 20 min).Samples are kept in the same vial and the alcohol solutionquickly replaced. Keep vials closed during the dehydrationtime.

3. Infiltrate samples in a first bath of resin (infiltrating resin) atroom temperature, overnight.

4. Discard the resin (since it will contain alcohol) as the manu-facturer’s instructions.

5. Infiltrate samples in a second bath of resin (pure resin), atroom temperature, overnight. Use an amount of resin justto cover the sample. Refer to Note 5 for resin re-use.

6. After labeling molds with an identification code of yourmaterial, carefully transfer samples to within them and fillout with the embedding resin.

7. Keep molds inside a hood on a steady and plane surface for24 h at room temperature.

8. Unblock and store blocks in a dry place.

3.2. Sectioning For parasitism analyses, it is strongly recommended preparingsemi-serial sections from the organ of interest to enable a reli-able evaluation of the parasite nests within the organ. The indi-cated section thickness is usually 5 μm and it is important to keepa 70 μm-interval between sections to avoid parasite recounting(9, 10).

1. Set the microtome with the appropriate glass knife and knifeholder for plastic resin.

2. Carefully cut the sample block. Sections come out as singlesections. If you are performing semi-serial sections, collectjust sections between intervals of interest.

3. Carefully transfer the selected sections to a recipient withtap water (histological bath or large becker) at room tem-perature. Use preferentially an antistatic tweezer to transfer

74 Fabrino et al.

sections to the water surface. Sections will stretch on thewater and stay floating.

4. Collect sections with a clean slide and transfer it to a hotplate (70–80◦C) until the water droplet evaporates (approx-imately 2 min).

3.3. Stainingand Mounting

Tissues embedded in GMA usually stain weakly with conventionalmethods when compared to paraffin-embedded tissues, and forthis reason, some minor staining changes are required (11). Herewe describe two useful and basic staining methods for tissue par-asitism evaluations: hematoxylin/eosin and toluidine blue/basicfuchsin (12). Staining patterns obtained with these techniques aresimilar (Fig. 5.1). Both methods stain cell nuclei in dark-blue andcytoplasm and connective tissue fibers in various shades and inten-sities of pink (Fig. 5.1). However, different degrees of contrastcan be obtained when toluidine blue/basic fuchsin is used (13).

3.3.1.Hematoxylin–Eosin

1. Wash slides with sections in tap water for 5 min.2. Immerse slides in the solution of iron–ammonium sulfate

for 10 min. This solution acts as a mordant to increase theintensity of the hematoxylin staining.

3. Wash in tap water for 5 min.4. Stain with Hematoxylin for 15–25 min (Note 6).5. Rinse slides for 10 min in running tap water.6. Immerse slides several times and quickly in an acid–alcohol

or acid–water solution to enable nuclear differentiation (30s to 1 min). Refer to Section 2.4, Section 4, for prepara-tion of this solution (Note 7).

7. Rinse in tap water for 5 min.8. Stain with eosin for 1–2 min.9. Rinse in tap water for 2–5 min (Note 8).

10. Blot excess water from the slide by gently pressing the slidewith a filter paper. Be careful not to touch sections.

11. Dry slides at room temperature (Note 9).12. Mount using a conventional mounting medium for histol-

ogy and let dry at room temperature or 37◦C. (see Fig. 5.2for the slide-mounting technique). Refer to Note 10 forremoving excess mounting medium.

3.3.2. ToluidineBlue-Basic Fuchsin

1. Cover sections with a few drops of toluidine blue solutionfor 3–5 min on a hot plate at 55–60◦C (see Note 11). Referto Section 2.4, Section 5, for toluidine blue preparation.Wash in running tap water to remove excess stain.

2. Quickly dry on a hot plate (see Note 11).


Fig. 5.1. Light micrographs of the heart stained with hematoxylin-eosin (a, c) and basic fuchsin-toluidine blue (b, d)from uninfected (a, b) and infected (c, d) rats at day 12 of the acute infection with Trypanosoma cruzi. Histopathologicalanalyses show parasite nests (arrows) and inflammatory processes (encircled in d). The boxed area in c shows a nestwith amastigote forms of the parasite in higher magnification. Note in d a clear visualization of the cross striations in thecytoplasm of cardiomyocytes. Scale bar, 15 μm (a–c), 10 μm (d).

76 Fabrino et al.

Fig. 5.2. Procedure for slide-mounting. (1) Place the slide on a plain surface and apply a drop of a mounting mediumin the center of the slide using an eyedropper. (2) Hold the cover slip at an angle of 45◦ so that it rests on the slide andtouches one side of the medium drop. (3) Once the medium spreads along the coverslip edge, carefully lower the coverslip over the section and medium. Make sure there are no bubbles under the cover slip carefully. (4) Wait drying. Fulldrying at room temperature may require up to 48 h; however, drying can be accelerated at 37◦C.

3. Cover sections with few drops of basic fuchsin solution(working solution) for 1 min at room temperature. Referto Section 2.4, Section 6, for basic fuchsin solutionpreparation.

4. Wash well in running tap water.5. Blot excess water from the slide by gently pressing the slide

with a filter paper. Be careful not to touch sections.6. Dry slides completely at room temperature.7. Mount using a conventional mounting medium for histol-

ogy and let dry at room temperature (see Fig. 5.2 for theslide-mounting technique). Refer to Note 10 for removingexcess mounting medium.

3.4. Parasite Analysisand Quantification

Studies of tissue parasitism are dependent on histopathologicalanalyses of a considerable number of sections of the organ ofinterest so as to have a reliable picture of the infection. In exper-imental research, it is also crucial to evaluate material obtainedfrom at least three animals per group for optimal sampling.

Parasite quantification can be performed by different meth-ods. Our group has been using two methodologies for this pur-pose: enumeration of parasite nests and/or quantification of tissueareas occupied by the parasite. For both methods, we use 5 μm-thickness sections. Moreover, considering that a single T. cruzinest can reach up to 75 μm depth (9), enumeration of parasitenests must be performed on semi-serial sections with an interval


of 70 μm between them to avoid parasite recounting. By keepingthis interval and using rats as experimental models, we found that20–40 sections of the heart per animal (around 10–15 obtainedfrom atrium and 15–20 obtained from ventricle) provide a con-sistent analysis of the organ.

Trypanosoma cruzi nests can greatly vary in size. Enumera-tion of these nests is performed by analyzing successive fields (seeNote 12) and counting the numbers of nests on a light micro-scope at magnification of 400× (Fig. 5.3). To compare differentexperimental groups, the same number of fields must be analyzedin each group. For example, in our experimental studies in rats,we generally evaluate a total of 1600 fields (800 from atria and800 from ventricles) for each group of animals (3–4 animals).

The tissue area occupied by parasites is obtained by using amicroscope equipped with an integrating eyepiece usually with100 squares at a magnification of 400× (14). The numbers ofsquares partially or totally occupied by the parasite are thenenumerated and the proportion is calculated based on the totalnumber of counted squares (Fig. 5.4). For example, by ana-lyzing 145 heart sections from infected rats we found 1098squares occupied by the parasite after counting a total of 43,000squares. This represents an area of 2.20% occupied by theparasite. Using this methodology, it is also possible to deter-mine the area occupied by cardiomyocytes and inflammatoryprocesses.

After enumeration, statistical analysis is usually done with anon-parametric test such as the Mann-Whitney Test (also knownas U test or Wilcoxon rank-sum test) that is used to compare twoindependent groups of sampled data (15).

Inflammation analyses can also be performed using the samesections prepared for parasitism studies. However, the quantifica-tion of inflammatory infiltrates is usually done at the magnifica-tion of 200×, which enables the observation of a higher area oftissue and a better identification of the intensity of the infiltrateand if it is focal or diffuse. On the other hand, cell features of theinflammatory and other cells are observed at 400× (Fig. 5.1).

4. Notes

1. To raise pH, add more phosphate dibasic solution. To lowdown pH, add more phosphate monobasic solution, alwaysunder magnetic stirring.

2. Paraformaldehyde is volatile and its fumes are very toxic(causes severe irritation to eyes, skin, and respiratory tract).

78 Fabrino et al.

Fig. 5.3. Schematic representation of the parasitism analysis on a light microscope. Correct enumeration (a) is achievedby counting parasite nests on successive fields at 400×. An incorrect analysis of fields (b) may provide a deficient pictureof the infection.

Thus, any manipulation involving this chemical must beperformed in a fume hood and wearing gloves.

3. To fix appropriately, aldehyde-based fixatives mustbe freshly prepared to avoid formic acid formation.


Fig. 5.4. Schematic representation of the parasitism evaluation using an integratingeyepiece on a light microscope. The numbers of squares partially or totally occupiedby the parasite (marked with an “x” in the figure) are enumerated at 400× and theproportion is calculated based on the total number of counted squares.

Alternatively, these fixatives may be stored for a couple ofmonths under minus 20◦C.

4. After fixation, the tissue samples are transferred to vialscontaining phosphate buffer 0.1 M and can be stored inthis solution at 4◦C until processing. It is not recom-mended storing for more than 2 months.

5. The second resin can be re-used. Store at 4◦C and re-usefor the first bath of resin in another embedding procedure.

6. Filter the staining solutions immediately before their use.This procedure will avoid precipitation of sediments onsections.

7. Differentiation with acid solutions requires some practicalexperience to ascertain the correct end-point, since thesesolutions are used to bleach or decolor samples and thecolor of the tissue can be modified to red. The correct end-point is when, after bluing up, the background is almostcolorless.

8. Eosin is highly soluble in water. Over-staining is removedby washing in running water.

9. After staining, dehydration of sections is done just byletting sections dry at room temperature. Alcohol dehy-

80 Fabrino et al.

dration must be avoided since this solvent can damagesections.

10. After drying, any excess medium can be removed with acotton swab moistened with a bit of xylene.

11. Slide drying on a hot plate can sometimes induce a lightblue background on the resin. In this case, drying can bedone at room temperature.

12. To avoid an incorrect enumeration of parasite nests, becareful not to overlap fields during the histopathologicalanalysis on the light microscope (Fig. 5.3).

References

1. Chagas, C. (1909) New human trypanoso-miasis. Morphology and life cycle ofSchysotrypanum cruzi, the course of a newhuman disease. Mem I Oswaldo Cruz 1,159–218.

2. Melo, R. C. N. (2009) Acute heart inflamma-tion: ultrastructural and functional aspects ofmacrophages elicited by Trypanosoma cruziinfection. J Cell Mol Med 13, 279–294.

3. Teixeira, A. R., Nitz, N., Guimaro, M. C.,Gomes, C., Santos-Buch, C. A. (2006) Cha-gas’ disease. Postgrad Med J 82, 788–798.

4. Ropert, C., Ferreira, L. R., Campos, M. A.,Procopio, D. O., Travassos, L. R., Ferguson,M. A., Reis, L. F., Teixeira, M. M., Almeida,I. C., Gazzinelli, R. T. (2002) Macrophagesignaling by glycosylphosphatidylinositol-anchored mucin-like glycoproteins derivedfrom Trypanosoma cruzi trypomastigotes.Microbes Infect 4, 1015–1025.

5. Teixeira, A. R., Nascimento, R. J, Sturm, N.R. (2006) Evolution and pathology in Cha-gas’ disease – a review. Mem I Oswaldo Cruz101, 463–491.

6. Melo, R. C. N., Rosa, P. G., Noyma, N. P.,Pereira, W. F., Tavares, L. E., Parreira, G. G.,Chiarini-Garcia, H., Roland, F. (2007) His-tological approaches for high-quality imag-ing of zooplanktonic organisms. Micron 38,714–721.

7. Bennett, H. S., Wyrick, A. D., Lee, S. W.,McNeil, J. H. (1976) Science and art inpreparing tissues embedded in plastic forlight microscopy, with special reference toglycol methacrylate, glass knives and simplestains. Stain Technol 51, 71–97.

8. Cole, M. B., Jr., Sykes, S. M. (1974) Glycolmethacrylate in light microscopy: a routine

method for embedding and sectioning ani-mal tissues. Stain Technol 49, 387–400.

9. Hanson, W. L., Roberson, E. L. (1974) Den-sity of parasites in various organs and therelation to numbers of trypomastigotes inthe blood during acute infections of Try-panosoma cruzi in mice. J Protozool 21,512–517.

10. Melo, R. C. N., Machado, C. R. S. (2001)Trypanosoma cruzi: peripheral blood mono-cytes and heart macrophages in the resistanceto acute experimental infection in rats. ExpParasitol 97, 15–23.

11. Cerri, P. S., Sasso-Cerri, E. (2003) Stain-ing methods applied to glycol methacry-late embedded tissue sections. Micron 34,365–372.

12. Fabrino, D. L., Leon, L. L., Parreira, G.G., Genestra, M., Almeida, P. E., Melo,R. C. N. (2004) Peripheral blood mono-cytes show morphological pattern of activa-tion and decreased nitric oxide productionduring acute Chagas’ disease in rats. NitricOxide 11, 166–174.

13. Abreu, M., Baroza, L., Rossi, M. (1993)Toluidine blue-basic fuchsin stain for glycol-methacrylate embedded tissue. J Histotechnol16, 139–140.

14. Melo, R. C. N., Machado, C. R. S.(1998) Depletion of radiosensitive leuko-cytes exacerbates the heart sympathetic den-ervation and parasitism in experimental Cha-gas’ disease in rats. J Neuroimmunol 84,151–157.

15. Mann, H. B., Whitney, D. R. (1947) On atest of whether one of two random variablesis stochastically larger than the other. AnnMath Statistics 18, 50–60.

Chapter 6

Intravital Microscopy to Study Leukocyte RecruitmentIn Vivo

Vanessa Pinho, Fernanda Matos Coelho, Gustavo BatistaMenezes, and Denise Carmona Cara

Abstract

The intravital microscopy is a valuable tool to capture images of cells in living organisms and to makestudies of molecular determinants of leukocyte trafficking easier. Using this technique, we can directlyvisualize and measure each step of the leukocyte recruitment paradigm, including leukocyte rolling flux,rolling velocity, adhesion, and emigration. Thus, it is possible to understand the process involved inleukocyte homing as well as the cell recruitment to inflammatory tissues. Nowadays, two types of intravitalmicroscopy are used routinely. The light microscopy is used to assess migration of intravascular cells inthin, tissues which must be sufficiently translucent. Epifluorescence microscopy allows the visualizationof the microcirculation while permitting the distinction of leukocyte subpopulations in solid organs.

Key words: Intravital microscopy, leukocyte, recruitment, inflammation, light microscopy,epifluorescence microscopy.

1. Introduction

Intravital microscopy is an extremely useful tool used as a quali-tative and quantitative method to observe leukocyte–endothelialcell interactions in vivo. For many years, the involvement ofmicrocirculation in inflammation has been a fascinating andwidely studied subject. The major point of concern at the out-set was the now known involvement of the microcirculation inthe inflammatory insult, but as time and techniques have pro-gressed, the interest has moved onto leukocyte–endothelial inter-actions. One of the best techniques that have enabled scientiststo understand these processes is intravital microscopy. The study


81

82 Pinho et al.

of the microcirculation using intravital microscopy represents asophisticated research tool to analyze complex biological interac-tions and disease mechanisms. This technique is also importantin studies that aim to develop novel prophylactic and therapeu-tic approaches intending to modify microvascular disorders andcellular dysfunction associated with inflammatory diseases.

Leukocyte recruitment is a multistep recruitment processthat is optimally visualized and studied in vivo using intravitalmicroscopy. Making use of this approach, we directly visualizeand measure each of the steps of leukocyte recruitment paradigm,including leukocyte rolling flux, rolling velocity, adhesion, andemigration (1).

The initial use of intravital microscopy date back century byDutrochet in 1824 (reviewed in (1)). This researcher observedthat leukocytes emigrated across small blood vessel walls. Thisfinding that leukocyte emigration occurred due to tissue injurywas finally reported in 1843 by Addison (reviewed in (1)).

Several animals can be used for theses studies; however,rodents, especially mice due to genetic manipulations, have beenmore used. Using light microscopy, it is possible to assess thevasculature and leukocytes within some thin and transparent tis-sues without contrast-enhancing methods, such as hamster cheekpouch (2), mesentery (3), and cremaster muscle (4). In thesepreparations we can see leukocyte rolling, adhesion, emigration,and measure blood flux. Also, microcirculation of nontranspar-ent organs (called generically as solid) and tissues can be stud-ied using epifluorescence microscopy, including liver (5), spleen(6), brain (7), lymph node (8), intestinal wall (9), knee joint(10), and others. Epifluorescence microscopy can be useful forstudies of leukocyte rolling and adhesion and vascular permeabil-ity (11). The intravenous injection of FITC-labeled BSA (fluo-rescein isothiocyanate-labeled bovine serum albumin) allows theevaluation of vascular albumin leakage from the microcirculationto the extravascular space as a parameter of vascular permeabil-ity in several tissues (12, 13). To study leukocyte recruitment,leukocytes can be labeled with dyes such as rhodamine (14) orwith fluorochromes-conjugated antibodies (15). These strategiesmake possible the visualization of microcirculation while permit-ting the distinction of leukocyte subpopulations. Also, it is possi-ble to visualize transfected cells that express the green fluorescentprotein derived from transgenic mice (16).

Concerning the intravital studies developed in leukocyterecruitment field, several cell structures and movements wereclarified. A recent illustration of this situation is a new stepon the leukocyte recruitment cascade, when cells “crawl” intoendothelial layer, looking for a more adequate way to emigrate toextravascular spaces. Strikingly, the inhibition of certain subtypesof adhesion molecules can modify the way used by leukocytesto emigrate, especially, increasing the rate of transcellular

Intravital Microscopy to Study Leukocyte Recruitment In Vivo 83

emigration (through the endothelial cell) instead of the conven-tional intercellular extravasation (between the endothelial cells),forming indeed “dome-like” structures expressed by the endothe-lial cell that never had been visualized before (17, 18).

The most recent advanced observation of microcircula-tion includes the spinning disk and the multi-photon confo-cal microscopy. The confocal laser scanning microscopy was anoption to get rid of the out-of-focus haze of objects that can emitfluorescence. In order to improve the scanning velocity of imageacquisition, a new different type of scanning method, called ini-tially as Nipkow disk was developed and it is now currently knownas spinning disk confocal microscopy. As a consequence of thisimprovement, a fast phenomenon would be more clearly observedas leukocyte movements under a blood flow. Indeed, lasers canhave different wavelength, culminating in different fluorescenceproduction by the stained cell or molecule, allowing different col-ors and co-localization of structures in a merged picture.

The optical microscopy is still the only way to examine thefour dimensions of a live phenomenon (the three conventionaldimensions − x, y, z – allied with time course evaluation), whichis very close to the realistic conditions found under physiologi-cal states. In addition to the spinning disk confocal technology,it is widely used nowadays in the acquisition of images usingmore than one laser wavelength simultaneously. Nowadays a two-photon microscopy has been used to study several physiologicalphenomena, such as for imaging electrical activity in deep tissues,for assessing the rate of blood flow, and for tracking immune-cellmotility and morphology (19). The outlook of confocal intravitalmicroscopy is very promising for physiological studies and manyimportant advances are currently underway, like portable two-photon micro-endoscope for tridimensional imaging of mousebrains and lungs. Additional developments have been achieved onfluorescent probes and new fluorophores designed to provide amore intense bright using a lower time and a lower power of laserincidence, and improving the protection of the organ or tissue.In this chapter, we outlined techniques using a cremaster musclepreparation, as an example of light intravital microscopy, and ajoint knee preparation, as an example of epifluorescence intravitalmicroscopy.

2. Materials

2.1. Materialsfor BothVideomicroscopy

1. Ketamine (200 mg/kg).2. Xylazine hydrochloride (10 mg/kg).3. Bicarbonate-buffered saline (131.9 mM NaCl; 4.7 mM

KCl; 1.2 mM MgSO4; 20 mM NaHCO3; pH 7.4).

84 Pinho et al.

4. P10 canule (Polyethylene tube).5. Board with an optically clear viewing pedestal for cremaster

(20) or knee joint preparation (21).6. Surgery tools.7. Cautery.8. 4-0 suture thread.9. Rhodamine 6G chloride and/or coupled antibodies.

10. LPS (lipopolysaccharide; sigma; B4:0111).

2.2. Light MicroscopySet

1. Light microscope with ×25 objective lens and ×10 eyepiece.2. Video camera.3. TV monitor.4. DVD.5. Optical Doppler velocimeter (Microcirculation Research

Institute, Texas A&M University, College Station, TX,USA).

2.3. EpifluorescenceMicroscopy Set

1. Epifluorescence microscope.2. Video camera.3. Filter 45 MM NCP11.4. Computer 2 DUO 8 GB.5. Software NIS ELEMENT.

3. Methods

Methods described below outline the preparation of an experi-mental model (mouse) for intravital microscopy (1), the surgeryfor cremaster preparation (2), and preparation of knee joint(3). The basic set up of intravital equipment for light intrav-ital microscopy (see Fig. 6.1a) and epifluorescence intravitalmicroscopy (see Fig. 6.1c) is described.

3.1. MousePreparation forIntravital Microscopy

Male mice are used for intravital observations. The process ofleukocyte recruitment in cremaster muscle and knee joint (andalso in other tissues) may be induced by administration of differ-ent inflammatory agents. Here we use LPS or specific antigens(21).

1. Anesthetize the animal with an intraperitoneal injection of amixture of xylazine and ketamine hydrochloride. Althoughother anesthetics may be used, this approach generates verysteady hemodynamic parameters, which is absolutely neces-sary to avoid blood flow effects on leukocyte recruitment.


A C

DB

Fig. 6.1. Schematic representation of an intravital setup. a and c show the setup for light and epifluorescence intravitalmicroscopy, respectively. In b and d is observed a cremaster muscle preparation (b) and the intra-articular synovial tissue(d) of the knee joint. Arrow indicates the synovial microcirculation. Scale bar = 0.5 cm.

2. Cannulate the left jugular vein to administer additional anes-thetic or drug if necessary.

3. Administrate the inflammatory agent. For intravitalmicroscopy in cremaster muscle, LPS is administeredby subcutaneous injection beneath the left scrotal skin.For intravital microscopy in knee joint, an intra-articularinjection with antigen (model of antigen-induced arthritis)is administered (21).

3.2. IntravitalMicroscopy ofCremaster Muscle

1. Cut the scrotal skin to expose the left cremaster muscle.2. Carefully dissect this muscle free of the associated fascia.3. Cut the cremaster muscle longitudinally with a cautery.4. Separate the right testicle and the epididymis from the

underlying muscle.5. Move them back into the abdominal cavity. The muscle is

held flat on an optically clear viewing pedestal and is securedalong edges with 4-0 suture (Fig. 6.1b).

6. Superfuse the exposed tissue with 37◦C warmed-bicarbonate-buffered saline (0.15 M, pH 7.4) (see Note 1).

7. Examine the cremasteric microcirculation by using a lightmicroscope equipped with ×10 or ×20 objective and a ×10eyepiece. A video camera is used to project images onto a

86 Pinho et al.

monitor, and images are recorded for playback analysis usinga DVD recorder. Single unbranched cremasteric venules(25–40 μm in diameter) are preferred and to minimize vari-ability, the same section of cremasteric venule should beobserved throughout the experiment (Fig. 6.2a).

BBAA

E F

DC

Fig. 6.2. Evaluation of the interaction between leukocytes and endothelial cells in the cremaster muscle microvascula-tion. a, b Images captured from the cremaster muscle microcirculation of mice, 4 h after intrascrotal injection with saline(a) or E. coli lipopolysaccharide (LPS, 0.05 μg/kg) (b). LPS induces a decrease of leukocyte velocity (e) and an increaseof the number of rolling cells (c), adherent cells (d), and emigrated cells in the extravascular space (f). ∗indicates astatistical difference in relation to saline group. p < 0.05, Student T test, n = 5/group. Scale bar = 50 μm.

3.3. IntravitalMicroscopyof Knee Joint

Unlike the cremaster muscle, the intravital microscopy is per-formed in the synovial microcirculation of the mouse knee.

1. Anesthetize the mouse and place the hind limb on a stage,with the knee slightly flexed and the patellar tendon mobi-lized and partly resected (see Fig. 6.1d). The intra-articularsynovial tissue of the knee joint is then visualized forthe determination of leukocyte rolling and adhesion (seeFig. 6.1d).


2. Select 2–4 regions of interest (unbranched venules with25–40 μm in diameter) in each mouse by using an ×20objective and a ×10 eyepiece.

3. To measure the leukocyte–endothelial cell interactions,inject the fluorescent marker rhodamine 6G intravenouslythroughout caudal vein as a single bolus of 0.15 mg/kgimmediately before measurements. Rhodamine epilumina-tion is achieved with a 103 W/2 variable HBO mercurylamp in conjunction with a filter set (see Note 2).

4. Capture the microscopic images with a video camera andrecord them on a computer. Data analysis is performedoffline using the imaging software NIS ELEMENT.

3.4. Evaluationof LeukocyteRecruitment

The number of rolling, adherent, and emigrated leukocytes isdetermined offline as cited above. Rolling leukocytes is definedas those cells moving at a velocity less than that of erythrocyteswithin a given vessel. The flux of rolling cells is measured as thenumber of rolling leukocytes passing a given point in the venuleper minute. Leukocyte rolling velocity is measured for the first 20leukocytes entering the field of view at the time of recording andcalculated from the time required for a leukocyte to roll along a

0

18

36

54

72

saline

*

*

antigen

Rolling

cells

/min

0.0

2.5

5.0

7.5

10.0

saline

*

antigen

Adhesion

cells

/100

µm

C D

A B

Fig. 6.3. Evaluation of the interaction between leukocytes and endothelial cells in the synovial microvasculature. a, bImages captured after intra-articular administration of saline (a) and antigen (b) in mice. Rolling cells (c) and adherentleukocytes (d) to the synovial endothelium were assessed after injection of antigen or sterile saline (control) into the kneejoint of mice. ∗indicates a statistical difference in relation to saline group. p < 0.05, Student T test, n = 5/group. Scalebar =50 mm.

88 Pinho et al.

100 μm length of a venule. A leukocyte is considered to be adher-ent if it remained stationary for at least 30 s, and total leukocyteadhesion is quantified as the number of adherent cells within a100 μm length of venule in 5 min (see Figs. 6.2 and 6.3). Forlight intravital microscopy, it is possible also to count leukocyteemigration (see Note 3 and Fig. 6.2). Leukocyte emigration isquantified as the number of cells in the extravascular space on thevisible field adjacent to the observed venule. Only cells adjacentto and clearly outside the vessel under study should be countedas an emigrated leucocyte. All these parameters can be altered,for example, 4 h after the intrascrotal injection of LPS, as shownin Fig. 6.2. Mean red blood cell velocity can be measured usingan optical Doppler velocimeter, which uses a pair of photodiodesto generate a voltage from an image of moving red cells that is alinear representation of red cell velocity. Wall shear rate is calcu-lated based on the Newtonian definition as (mean red blood cellvelocity/diameter) × 8 (s−1) (22).

4. Notes

1. The prevention of bleeding is crucial. Excessive bleedingwill affect mainly the number of circulating leukocytes, andso it is crucial to assess the total leukocyte count afterevery experiment. During the whole experimental proto-col, the animal should be kept under physiological con-ditions. So, dehydration, volume lost, hypothermia shouldalways be avoided, especially on those preparations wherethe peritoneum cavity is exposed (liver, intestine, mesen-tery, etc.) and an important modification on animal parame-ters is induced. The cover of exposed organs that will notbe observed with a soaked gaze, napkin of a PVC-wrapwill minimize the water and temperature loss, and evenreduce the movement of organs. Importantly, frequent andstrong movement caused by animal breathing or instabilityof the microscope components will make the correct analy-ses of the data difficult. A warmer lamp (infrared) is stronglyrecommended in rooms where the temperature is lowerthan the mice (about 25–30◦C) and in peritoneum-openedprocedures.

2. Especially related to fluorescence microscopy, the contin-uous exposure of the same field for many minutes cancause “bleaching” of the fluorescence and incorrect anal-yses. The stability of the fluorophore should be evaluated


in accordance with exigencies of the experimental protocol.Finally, as the leukocyte recruitment will begin since the firstcells are injured, the improvement of the operatory skillswill avoid unexpected leukocyte recruitment and will affordmainly the reproducibly of the data. An exceedingly trau-matic operatory approach and even an incorrect administra-tion of the pro-inflammatory mediators (dose and vial) willinterfere in the local blood flow, which is a sine qua non-condition to leukocytes to get vessels on the field of view.

3. Light microscopy illuminates a translucent tissue and thisfacilitates the emigration enumeration. However, in the epi-fluorescence evaluation of solid organs, it is hard to clearlysee cells outside microcirculation. Additionally, the leakageof rhodamine from the microcirculation to the extravascularspace makes it difficult to track cells present in extravascularspace.

References

1. Springer, T. A. (1994) Traffic signals of lym-phocyte recirculation and leukocyte emigra-tion: the multistep paradigm. Cell 76(2),301–314.

2. Duling, B. R. (1973) The preparation anduse of the hamster cheek pouch for studiesof the microcirculation. Microvasc Res 5(3),423–429.

3. Kvietys, P. R., Perry, M. A., Gaginella,T. S., Granger, D. N. (1990) Ethanolenhances leukocyte–endothelial cell interac-tions in mesenteric venules. Am J Physiol259, 578–583.

4. Goldberg, M., Serafin, D., Klitzman, B.(1990) Quantification of neutrophil adhesionto skeletal muscle venules following ischemia-reperfusion. J Reconstr Microsurg 6(3),267–270.

5. Vollmar, B., Glasz, J., Menger, M. D.,Messmer, K. (1995) Leukocytes con-tribute to hepatic ischemia/reperfusioninjury via intercellular adhesion molecule-1-mediated venular adherence. Surgery 117(2),195–200.

6. Schmidt, E. E., MacDonald, I. C., Groom,A. C. (1990) Interactions of leukocytes withvessel walls and with other blood cells, stud-ied by high-resolution intravital videomi-croscopy of spleen. Microvasc Res 40(1),99–117.

7. Mooradian, A. D., McCuskey, R. S. (1992)In vivo microscopic studies of age-relatedchanges in the structure and the reactivityof cerebral microvessels. Mech Ageing Dev64(3), 247–254.

8. Von Andrian, U. H. (1996) Intravitalmicroscopy of the peripheral lymph nodemicrocirculation in mice. Microcirculation3(3), 287–300.

9. Sekizuka, E., Benoit, J. N., Grisham, M. B.,Granger, D. N. (1989) Dimethylsulfoxideprevents chemoattractant-induced leukocyteadherence. Am J Physiol 256, 594–597.

10. Veihelmann, A., Szczesny, G., Nolte, D.,Krombach, F., Refior, H. J., Messmer, K.(1998) A novel model for the study of syn-ovial microcirculation in the mouse kneejoint in vivo. Res Exp Med 198(1), 43–54.

11. Hulström, D., Svensjö, E. (1979) Intrav-ital and electron microscopic study ofbradykinin-induced vascular permeabilitychanges using FITC-dextran as a tracer. JPathol 129(3), 125–133.

12. Kubes, P., Gaboury, J. P. (1996) Rapid mastcell activation causes leukocyte dependentand -independent permeability alterations.Am J Physiol 271, 2438–2446.

13. Cara, D. C., Ebbert, K. V., McCafferty, D.M. (2004) Mast cell-independent mecha-nisms of immediate hypersensitivity: a role forplatelets. J Immunol 15, 4964–4971.

14. Baatz, H., Steinbauer, M., Harris, A. G.,Krombach, F. (1995) Kinetics of white bloodcell staining by intravascular administrationof rhodamine 6G. Int J Microcirc Clin Exp15(2), 85–91.

15. McDonald, B., McAvoy, E. F., Lam, F., Gill,V., de la Motte, C., Savani, R. C., Kubes, P.(2008) Interaction of CD44 and hyaluronanis the dominant mechanism for neutrophil

90 Pinho et al.

sequestration in inflamed liver sinusoids. JExp Med 205(4), 915–927.

16. Stein, J. V., Rot, A., Luo, Y.,Narasimhaswamy, M., Nakano, H., Gunn,M. D., Matsuzawa, A., Quackenbush, E. J.,Dorf, M. E., von Andrian, U. H. (2000)The CC chemokine thymus-derived chemo-tactic agent 4 (TCA-4, secondary lymphoidtissue chemokine, 6Ckine, exodus-2) trig-gers lymphocyte function-associated antigen1-mediated arrest of rolling T lymphocytesin peripheral lymph node high endothelialvenules. J Exp Med 191(1), 61–76.

17. Phillipson, M., Kaur, J., Colarusso, P., Bal-lantyne, C. M., Kubes, P. (2008) Endothelialdomes encapsulate adherent neutrophils andminimize increases in vascular permeabilityin paracellular and transcellular emigration.PLoS One 3(2), 1649.

18. Phillipson, M., Heit, B., Colarusso, P., Liu,L., Ballantyne, C. M., Kubes, P. (2006)Intraluminal crawling of neutrophils to emi-gration sites: a molecularly distinct process

from adhesion in the recruitment cascade. JExp Med 203(12), 2569–2575.

19. Benninger, R. K. P., Hao, M., Piston, D. W.(2008) Multi-photon excitation imaging ofdynamic processes in living cells and tissues.Rev Physiol Biochem Pharmacol 160, 71–92.

20. Cara, D. C., Kubes, P. (2004) Intravitalmicroscopy as a tool for studying recruit-ment and chemotaxis. Methods Mol Biol 239,123–132.

21. Coelho, F. M., Pinh, V., Amaral, F. A., Sachs,D., Costa, V. V., Rodrigues, D. H., Vieira,A. T., Silva, T. A., Souza, D. G., Bertini,R., Teixeira, A. L., Teixeira, M. M. (2008)The chemokine receptors CXCR1/CXCR2modulate antigen-induced arthritis by regu-lating adhesion of neutrophils to the synovialmicrovasculature. Arthritis Rheum 58(8),2329–2337.

22. Borders, J. L., Granger, H. J. (1984) An opti-cal doppler intravital velocimeter. MicrovascRes 27(1), 117–127.

Section II

Fluorescence Microscopy Applications

Chapter 7

Introduction to Fluorescence Microscopy

Ionita C. Ghiran

Abstract

This chapter is an overview of basic principles of fluorescence microscopy, including a brief history onthe invention of this type of microscopy. The chapter highlights important points related to propertiesof fluorochromes, resolution in fluorescence microscopy, phase contrast and fluorescence, fluorescencefilters, construction of a fluorescence microscope, and tips on the correct use of this equipment.

Key words: Fluorescence microscopy, fluorochromes, fluorescence filters.

1. A Brief Historyof FluorescenceMicroscopy

The invention of fluorescence microscopy was the result of a seriesof fortunate but largely unrelated discoveries that culminated atthe beginning of the 20th century. Long before, the curious prop-erties of various minerals and especially plant extracts to eitheremit light when kept in the dark (phosphorescence) or reflect a

Disclaimer: When preparing the manuscript I had to use examples that werepertinent to fluorescence microscopy and, therefore, to use actual fluorescencemicroscopes, objectives, cameras and image acquisition and analysis software.With one exception (the Nikon 63 × 1.49 objective, which was borrowed fromPerkin-Elmer), all the other instruments and accessories were already present inthe laboratory. The fact that I used those particular instruments to record imagespresent in this chapter does not constitute a quality judgment, either positive ornegative, about their optical and mechanical performances. The decision aboutpurchasing a certain brand over another is a complex process that has to balancequality, price, previous experience, technical support and idiosyncratic personalpreference.


93

94 Ghiran

certain color and transmit a different one when illuminated withsun light (fluorescence) were well known. A special place in thehistory of fluorescence is reserved for Nicolas Monardes, a 16thcentury Spanish botanist who wrote in “La historia medicinal delas cosas que se traen de nuestras Indias Occidentals” (“The med-ical history of the objects that are brought from our Occiden-tal Indies”) about the peculiar properties of the yellow aqueousextract of the kidney wood (Lignum nephriticum) that had a dis-tinct blue tint (fluorescence) in reflected light. However, the per-son most responsible for a scientific approach to the fluorescencephenomenon was George Stokes, an Irish physicist who noticedin 1852, while working at Cambridge University that the mineralfluorspar (or fluorite, CaF2) when illuminated with blue light,emitted red light. To describe this phenomenon he coined theterm “fluorescence,” 1 based on the name of the mineral he usedfor his experiments. He was also the first to notice that fluorescentsubstances emitted light at a longer wavelength than the excita-tion light, describing for the first time the shift in wavelength thatnow bears his name.

1.1. Inventionof FluorescenceMicroscopy

At the same time physicists were trying to understand the phe-nomenon of fluorescence, a race for microscope objectives withimproved resolution was underway in England and Germany.Before long (around 1886), the highest practical numerical aper-ture (NA) of objectives was reached based on Abbe’s (1872) pub-lished relationship between resolution and wavelength. It thenbecame clear that one could produce higher resolution imageswith a given objective if light with a shorter wavelength (i.e. blue,violet and ultraviolet) was used to illuminate the object. Althoughinvisible to the human eye, the existence of UV light had beenknown at that time for over 70 years due to research conductedin 1801 by the young German physicist, Johann Wilhelm Ritter(1776–1810). He noticed while studying the response of silverchloride (AgCl) to light that the intensity of the reaction increasedas the interacting color was closer and even beyond the left endof the visible spectrum (blue–violet–ultraviolet) and decreasedtowards the right end (red and infra-red). Using blue and UVlight appeared to be a good method to increase microscope res-olution, but there were several significant drawbacks that madeimplementation difficult. First, ordinary glass is not very trans-parent beyond 380 nm; therefore, new quartz lenses had to bedesigned and employed for UV microscopy. Next, because theeye is not sensitive to UV light, images had to be focused using

1 The name fluorite describes the crystal’s low melting point, from the latin“fluo”, which means to flow that is also the origin of the word fluid.

Introduction to Fluorescence Microscopy 95

visible light. Due to uncorrected axial chromatic aberrations ofobjectives, out-of-focus images were recorded when structures ofinterest were photographed using UV light results. Finally, exper-imenting with UV imaging was potentially hazardous because allmicroscopes at that time used trans-illumination that exposed theobserver directly to dangerous UV light.

In 1904, while experimenting with blue and UV light, AugustKohler (a German physicist who 10 years earlier, at only 27,described a new illumination method that became the gold stan-dard for all microscopes produced since) noticed that certaintissues fluoresce when illuminated with UV light. He was thefirst to describe primary fluorescence (auto-fluorescence) of tis-sues, although he reported it initially as a nuisance rather thana discovery with possible practical applications. In 1913, a ded-icated fluorescence microscope was built by Heinrich Lehmannand Stanislaus Josef Mathias von Prowazek that could observeand also measure fluorescence signals. However, the break-through that first demonstrated the exceptional potential of fluo-rescence microscopy in biology had to wait until 1941 when anAmerican immunologist, Albert Hewett Coons, successfully usedanthracene–isocyanate and later fluorescein to directly label pneu-mococcal anti-serum. Results obtained, although exceptional,were not well received by the scientific community due to lowspecificity (especially when using whole serum) and poor signalamplification when using monoclonal antibodies labeled with therapidly fading fluorescein (1). A second significant breakthroughhappened in 1954 when Thomas Weller (who received in thesame year the Nobel prize in Medicine), working together withAlbert Coons, developed and used succesfully an indirect methodof staining, comprised of a non-labeled primary and a fluorescein-labeled secondary antibody, to identify viral particles in culture(1). It is difficult to name another method that shaped cell biol-ogy, molecular biology and immunology more than the use of flu-orescently labeled antibodies. Whether fluorescence microscopy,flow cytometry, real-time PCR or gene array, all these methodsrely on the use of a fluorescently labeled probe (antibody, DNA,or RNA fragment) that specifically and quantitatively recognizesan epitope. Even the use of GFP-tagged proteins stems from thesame basic idea.

2. What IsFluorescence?

Fluorescence is the property of atoms and molecules to emit lightfollowing excitation by an outside source of energy. A moleculecapable of fluorescence is called a fluorochrome. Fluorescence

96 Ghiran

Fig. 7.1. Jablonski diagram of a theoretical fluorescent molecule showing most of thepossible electronic transitions during excitation and emission (see text for details).

occurs when excess energy released by electrons returning to theground state is released as photons (2). The illustration describ-ing this process is known as the Jablonski diagram (Fig. 7.1),after the Polish physicist Aleksander Jablonski who used it firstin 1935 to describe the transition of electrons during absorptionand emission of light.

Because fluorescence microscopy is based on fluorescence oforganic molecules rather than atoms, we will use molecular flu-orescence to discuss the physical basis of fluorescence. In steady-state conditions, electrons are found in the lowest energy level, S0,or the ground state. When an organic substance absorbs photonswith sufficient and appropriate energy, electrons from S0 transit tohigher energy states called excited single states, S1, S2 or Sn. Thetransition phase from S0 to S1 or S2 is very fast and takes about10−15 s. The excited single states of organic molecules are notas discreet as those of atoms, so the energy levels become energybands. The direct consequence of this is the very broad excitationand emission spectra of organic fluorochromes (Fig. 7.2) com-pared to the sharp peaks of the emission spectrum characteristicof individual atoms (see below). The ground or excited state iscalled single because all electrons on these bands are spin-paired.When the excited electron returns to the ground state it can takeseveral paths depending on the molecular configuration of themolecule, and the presence of other species of molecules in themedia, can result in

1. fluorescence2. phosphorescence3. radiation-less conversion (heat)


Fig. 7.2. Excitation and emission spectra of a hypothetical fluorescent molecule illus-trating the red shift (Stokes’s shift) of emission wavelength compared to excitation.The amount of the shift depends on the molecular composition of the molecule. Flu-orochromes with larger shifts are desirable for fluorescence microscopy.

When returning from an excited state, electrons usually gofrom one energy level to the next energy level down, releasingexcess energy as heat in a process known as internal conversion.Because the energy levels S2 and S1 are very close, the probabil-ity that an electron from S2 will bypass S1 and go directly to S0is significantly lower than going from S2 to S1. Therefore, mostfluorescence happens when electrons return from S1 to S0. Thedifference in energy between S1 and S0 is the energy of the emit-ted photon (Kasha’s rule). In other cases the energy lost by anelectron going from S1 to S0 can be given up as heat. In this situ-ation there will be a fluorescence-less process, where no photonsare emitted by the organic molecule.

The energy state of electrons depicted on the right side of theJablonski diagram is labeled “triplet state,” denoting that one setof electron spins is unpaired. This happens when spin of an elec-tron from one of the higher vibrational energy levels (S1 or S2) isreversed, switching the electron to the triplet state through a pro-cess called intersystem crossing. The triplet state is depicted belowthe corresponding singlet state with the same electronic configu-ration because it has a slightly lower energy level. Therefore, thelowest excited state of a molecule is often a triplet state. Althoughthe return from this state to the ground state does not resultin fluorescence but rather phosphorescence or heat, the tripletstate is important because in this state electrons are quite reac-tive, leading to intermolecular reactions that cause a decrease influorescence efficiency (photobleaching) and to interactions withoxygen that produce cell-toxic, reactive intermediate species (3).

2.1. The ChemicalBasis ofFluorescence

Certain organic molecules fluoresce whereas others do not. Whatmakes molecules fluorescent is the presence of conjugated, alter-nate double bonds that allow electrons to be displaced through

98 Ghiran

the rigid and planar molecule. Therefore, the benzene ring(C6H6) is a very common group among fluorescent molecules.Usually there is a correlation between the size of the molecule(the number of double bonds) and wavelengths of the excitationand emission states, specific for those particular molecules.

3. Properties ofFluorochromes

3.1. FluorescenceQuantum Yield

Fluorescence Quantum Yield (quantum efficiency, QE) is a mea-sure of the efficiency of the fluorescence process. It is defined asthe ratio between the number of emitted photons and the num-ber of absorbed photons. This does not mean that in the case offluorescein that has a quantum yield of 90%, the energy of theemitted photons is 90% of the energy of the absorbed photons. Itmeans that number-wise, there is a 10 to 9 relationship betweenexcitation and emission. The loss in energy is due to non-radiativeprocesses (Stokes’s losses) that render the emitted photon to bered-shifted compared to the excitation photon. However, the QEof fluorescein varies greatly under different conditions, especiallypH. At a pH of 8, fluorescein has the highest QE; whereas asthe pH drops towards 6, the QE reaches 0.3. This is importantbecause the same amount of fluorescein will fluoresce brightlyoutside the cell and be very dim once inside the cell in an acidiccompartment (i.e. phagosome). Therefore, one should avoid flu-orescein alone to track phagocytosed particles (or any biologicalprocesses where the pH varies) and choose a pH-insensitive dye.On the other hand, tagging a particle with both fluorescein and apH-insensitive dye represents a very powerful method that allowsone to track particles intracellularly and simultaneously monitorpH changes.

3.2. Quenching Quenching is defined as a decrease in the QE due to interactionswith molecular species nearby such as proteins or fluorochromes.The energy transfer is non-radiative and results in less efficientfluorescence. The emitted light of a quenched fluorochrome dis-plays identical spectral characteristics to the non-quenched flu-orochrome. Commonly, fluorochromes lose some of their QEfollowing conjugation to proteins such as BSA (bovine serumalbumin) or IgG (immune globulins) if the conjugation densityof the fluorochrome is too high. Therefore, it is important tokeep the molar ratio fluorochrome-to-protein low rather thanhigh to prevent quenching due to fluorochrome–fluorochromeinteractions. (Each company that sells fluorescence labeling kitshas an optimal range depending on the size of the fluorochromeand the size of the protein.) Other molecules such as ANS


(8-anilino-1-naphthalene sulfonate) behave differently, display-ing low fluorescence in polar solvents and increased fluorescenceupon interaction with various proteins. Quenching is not alwaysdetrimental; it may offer useful information about the nearbyspecies as well as the distance between them, which is always atleast an order of magnitude smaller than the resolution of a stan-dard fluorescence method.

3.3. Photobleaching Unlike quenching, which can be reversible, photobleachingis generally irreversible. This process is due to electronstransitioning from single to the triplet state; electrons willinteract efficiently with other molecular species with tripletground states such as the ubiquitously present oxygen. Thenet result is an irreversible change in the molecular struc-ture of the fluorochrome, which results in a permanent lossof fluorescence. Another side effect of this interaction is theformation of reactive oxygen species in or around the fluo-rescently labeled cell that will chemically interact and dam-age cellular structures. In our experience, very low con-centrations of ascorbic acid (nanomolar range) offer a cellfriendly way to prevent photobleaching and oxygen-mediated celldamage.

3.4. Molar Extinction Molar extinction (ε) (synonyms: molar absorption or molarabsorptivity), as defined by the Beer-Lambert law, measures theefficiency of a fluorochrome (or any substance) to absorb light ata certain wavelength (λ). Importantly, the absorption depends onthe molecular structure of the molecule, the wavelength of theabsorbed light, pH value and temperature.

Formula:

A = εcl

where

A=absorbance of the homogenous, isotropic sampleε = molar extinctionc = concentration of fluorescent dye (mol/L)l = thickness of the fluorescent solution, expressed in cm(although theoretically it should be expressed in meters)

3.5. FluorescenceLifetime

Fluorescence lifetime represents the average time a moleculespends in the excited state before returning to the ground state.In Fig. 7.1, we saw that following excitation, molecules transitfrom the ground state to excited states (Sn). The deactivationphase is a two-step process that includes a non-radiative (internalconversion) and a radiative (fluorescence) process that depopu-late the excited phase. If there is no energy transfer to the envi-ronment (no acceptors), the fluorescence lifetime is constant for

100 Ghiran

a given molecule in a given solvent and is usually in the order ofnanoseconds. However, when acceptors such as oxygen, calcium,magnesium or hydrogen are present, the lifetime of fluo-rochrome is significantly decreased. Measuring fluorescence life-time (not general fluorescence) can offer precise informationabout the environment surrounding the fluorochrome that allowresearchers to map cells with unsurpassed space and time resolu-tion for parameters such as: calcium, magnesium and pH, as wellas for the presence of certain quenching molecules. There is lit-tle doubt that fluorescence lifetime imaging will be a widespreadfluorescence microscopy method in the years to come.

3.6. Stokes Shift Stokes shift represents the shifted emission of a fluorochromecompared to excitation (Fig. 7.2). When the excitation and emis-sion spectra are displayed together the Stokes shift is representedby the distance between the two peaks. The curve on the leftside of graph is called the excitation spectrum and in general(but not always) is identical with the absorption spectrum of thefluorochrome. When the spectrum of excitation is very differ-ent than that of emission, it means that non-fluorescent species,likely dimers, are present in the solution (4). A better represen-tation of the excitation curve would be against the molar extinc-tion coefficient, which is the wavelength-dependent absorptivity(see above) of that particular fluorochrome in the solvent usedto generate the spectra. However, because excitation and emis-sion spectra are generally displayed on a single graph, fluorescenceexcitation is used against various wavelengths. The emission spec-trum is a plot of the fluorescence intensity versus various wave-lengths. The shape of the emission plot does not depend on theexcitation wavelength, but only on the molecular identity of thefluorochrome due to Kasha’s rule (see above).

4. Resolutionin FluorescenceMicroscopy

Classically, resolution of an optical microscope is defined as theminimal distance (expressed in μm) at which two points in thesample are seen separately. Ernst Abbe showed in 1872 that theminimal distance between two points (d) that are imaged usingtransmitted light (a regular bright-field microscope) depends onthe half-angle of the aperture of the objective (sinα), the wave-length (λ) of the light used to form the image and the refractiveindex (n) of the media between the frontal lens and the sam-ple (or coverglass). While the aperture (opening) of an objec-tive is measured in degrees, the NA is an adimensional valuethat depends on the aperture of the objective and the refractive


index of the media between the sample and the frontal lens.The smallest detail resolved with an objective with a NA of 1.40(the highest possible number until a few years ago without usingspecial immersion oil and special cover glasses; the value now is1.49, used with green light (λ = 550 nm) and with an immer-sion condenser (NA = 1.40), so called double immersion) on aperfectly aligned microscope would be about 220 nm.

d = λ

NAcond + NAobj= λ

2NA

Anyone who has ever used a fluorescence microscope knowsthat when using fluorescence, it is possible to see light originat-ing from structures that are about 10 times smaller than the the-oretical resolution of the best immersion lens. Even with a dryobjective (40×), it is possible to see quantum dots, which are flu-orescent crystals the size of a ribosome (about 20 nm). So what isthe explanation? In fluorescence microscopy, the resolution doesnot measure the size of the smallest self-luminous point it candetect but rather the radius of the image formed by the objectivethat represents that self-luminous point. If one images three parti-cles having 20, 100 and 180 nm in diameter, the microscope willsee them as having the same size. The intensity associated withparticles may be different depending on their actual size but as faras the final size the microscope will show, all particles will be thesame. In the same way, we can see on a dark night, a lit candlemiles away or stars that are hundreds of light years away from usalthough our eyes do not have the angular resolution needed toresolve neither the flame nor the disk of the stars at that distance.What we see is the light coming from the flame or stars because ofthe dark background. Notably, the light seen with unaided eyeswill contain no details regarding the shape or original size of theluminous points. Similarly, fluorescence microscopy allows us todetect the presence of fluorescent particles that are significantlybelow the resolution limit of the objective without any indicationof structure, shape or actual size. The initial work regarding theresolution of optical instruments imaging self-luminous objectswas done using astronomical instruments (stars are perfect self-luminous points). Therefore, not surprisingly, the theory explain-ing the resolution of fluorescent microscopy is based on the workof a British astronomer George Airy. He was the first to showthat when in focus, the image of a star seen in a telescope is nota point but rather a small bright disk that is surrounded by largerand ever fainter bright disks alternating with dark disks. Based onAiry’s work, Lord John Rayleigh (who received the Nobel Prizein 1904) derived an empirical formula that shows that in fluo-rescence microscopy the smallest fluorescent structure will form adisk with a radius of

102 Ghiran

r = 1.22λ

NAcond + NAobj= 0.61λ

NA= 0.61λ

n sin α

In this formula, λ is the wavelength of the fluorescence (emit-ted light), α is the half-angle of the objective aperture and n isthe refractive index of the media between the frontal lens andthe sample. In fluorescence microscopy, the objective is the con-denser; therefore, the NA of the condenser will be equal to theNA of the objective. Rayleigh resolution criteria also states thatin order for two self-luminous points to be resolved, the centerof one object has to be on the first minimum (the first dark ringthat surrounds the airy disk) of the adjacent point. Although thisdepends significantly on one’s ability to resolve different shades ofgray, novel digital methods used to record and process microscopyimages help standardize and even exceed this criterion (5).

5. TheConstructionof a FluorescenceMicroscope Current fluorescence microscopes are based on the design used

by Johan Ploem that was developed while working togetherwith Leitz in the early 1960s.2 Today, virtually all fluorescencemicroscopes are based on Ploem’s design, and, therefore, we willpresent it in the following pages. The simplest way to representthe basic fluorescence microscope design is shown in Fig. 7.3a.The light emitted by the fluorescence light source (excitation)bounces off of a dichromatic mirror, passes through the objectiveand excites fluorochromes in the sample. Due to the nature ofthe dichromatic mirrors that behave like true mirrors for shorterwavelengths and are transparent to longer wavelengths∗, the flu-orescent light (emission) originating in the sample, passes un-reflected through the dichromatic mirror towards the detector,which can be the eye or an electronic device (i.e. CDD camera).Some spinning disk confocal microscopes have dichromatic mir-rors that allow shorter wavelengths to pass through and reflect thefluorescent light to the detector. Figure 7.3b shows the fluores-cence path of a commercial fluorescence microscope along withthe bright-field in the lower part of the microscope. Although the

2 Due to poor circulation of scientific journals at that time and lack of translationof most scientific articles written in languages other than English, neither Ploemnor researchers from Leitz were aware that more than 5 years before, two Rus-sian scientists, Brumberg and Krylova, did not only described the advantages ofemploying “dividing mirrors” in fluorescence microscopy but actually used them.


Fig. 7.3. Diagram illustrating the basic principle of epi-fluorescence microscopy. Thelight source emits light that is reflected by the dichromatic mirror through the objectiveinto the sample. The fluorochromes present in the sample following excitation will emitlight of longer wavelength (fluorescence) that will go back through the objective anddichromatic mirror and form the final image on a detector that can be either the eye ora CCD camera.

presence of a series of lenses, diaphragms and filters in the flu-orescence path make the design of the fluorescence microscopefluorescence path seem more complicated than the schematic, thebasic principles apply. Therefore, to understand how a fluores-cence microscope works, and more importantly be able to get themost out of it and troubleshoot potential problems, we shall startby describing components of a fluorescence microscope followingthe light path, as shown in Fig. 7.4a.

Although it is not obvious, a fluorescence microscope is basi-cally a folded bright-field microscope (see Fig. 7.4a), where theobjective lens play the role of both the condenser and the objec-tive. As a rule, in epi-fluorescence microscopy everything involvedin image formation is located above the microscope slide. Noneof the controls that are located under the microscope stage, suchas the ones that modify the condenser height, the size of the fieldand aperture diaphragms, the position of neutral density, or greeninterference filters affect the fluorescence image. In some partic-ular instances, the height of the condenser may have a negativeimpact on the image (see below), but in general all of the adjust-ments for the fluorescence microscope are located on the fluores-cence illuminator.

On a bright-field microscope, the light that forms the imageoriginates from the halogen light bulb located for uprightmicroscopes in the base of the microscope. Less intuitive, influorescence microscopy the light that forms the final image isgenerated and emitted by fluorochromes present in the sampleand not by the light source located at the back-end of the fluo-rescence illuminator. In a functional epi-fluorescent microscope,

104 Ghiran

Fig. 7.4. a Diagram of the light paths (epi-fluorescence and trans-illumination) of a modern Olympus microscope. Allthe optical and mechanical components involved in fluorescence image formation are located above the stage. The lightsource can be either attached directly to the illuminator or be connected to the microscope through a liquid light guide inwhich case it will be away from the microscope body (see text). b Comparison between the apertures of a 10× and 40×objective at the imaging distance. While a 10× objective has a significantly larger frontal lens than a 40× objective, itis also located further away from the sample, which translates into a smaller actual imaging aperture. An objective withsmaller aperture will be less efficient in gathering photons and thus will render images with less resolution.

none of the light that originates from the light source will everreach the detector. Because a fluorescent structure is formed bynumerous self-luminous points, image formation in fluorescencemicroscopy have, in some respects, more in common with imagesformed by an astronomical instrument than with images formedby a bright-field microscope. Each fluorescence image is gener-ated by a large number of individual self-luminous points of lightor, more correctly, disks of light (see above) that scatter lightin all directions around them. Only a fraction of this scatteredlight will be captured by the objective lens. The amount of lightgathered by a lens is directly proportional to the opening (aper-ture) of the objective at the imaging distance of the lens. Thisis important because although a 10 × 0.25 objective has a sig-nificantly larger frontal lens than a 40 × 0.95 lens, the anglesubtended by the lens at the working distance is only about 15◦for the 10× lens and around 71◦ for the 40× lens (Fig. 7.4b).This is possible because the working distance of the 10 × 0.25objective is about 6 mm, whereas for the 40 × 0.95 it is about0.14 mm. In other words, the larger the angle of the objective


frontal lens at the imaging distance, the more light will enter theobjective and the more photons will contribute to the image (seeSection 4).

5.1. Light Sourcesor Illuminators

A light source suitable for fluorescence imaging has to delivera beam of light that is homogenous (uniform output), highpower, constant power output in time (no flickering) and con-tains all wavelengths that are needed to excite fluorochromespresent in a given specimen. Today, most fluorescence micro-scopes are equipped with one of the three main types of illumina-tors: mercury, xenon or metal halide lamp. Lamps are made outof fused silica quartz round or elliptical envelopes that are strongenough to sustain high temperature and extreme pressures. Mer-cury lamps and xenon lamps are very common with most micro-scopes 5–10 years and older, with their lamp housing physicallyattached to the microscope (see Fig. 7.3b). Newer microscopesare fitted with metal halide illuminators or LED light sources sep-arated from the microscope stand using liquid light guides (3–5ft long) to deliver the light to the fluorescence illuminator.

Mercury and xenon lamps deliver a very bright and continu-ous spectrum from UV to infrared (Fig. 7.5a and b). The powerof the lamps varies usually between 50 and 200 W for mercurylamps and 75 and 150 W for xenon lamps. A closer inspection ofthe mercury bulb output graph shows that at certain wavelengthsthe emission is significantly greater than others, such as 365, 405,436, 546 and 579 nm. Most of the fluorochromes that were syn-thesized in the past and are still in use today were centered onthese wavelengths to maximize the efficiency of the mercury lightsource. Additionally, most of the objectives used for wide-fieldfluorescence or confocal microscopy were also designed to offermaximum correction for these particular wavelengths. Interest-ingly, the most used fluorochrome, fluorescein, has its excitationmaxima at 494 nm where the mercury lamp has no major emis-sion lines. However, due to its high quantum efficiency and broadexcitation spectra, fluorescein is bright even when excited with amercury lamp. Fluorescein has several disadvantages, such as fastphotobleaching and high pH sensitivity, which makes it less suit-able for fluorescence microscopy but more appropriate for FACS(fluorescence-activated cell sorting), where the interaction of thelaser beam with the fluorescein is brief enough such that bleach-ing is not an issue.

Due to its characteristic larger distance between electrodes,mercury bulbs fill up the back focal plane of the objective sig-nificantly more compared to xenon sources, allowing a uniformillumination of the specimen even at small magnifications. On theother hand, mercury lamps have short life spans (about 200 h)and exhibit a steady decrease of power output due to carbondeposits that begin to appear after the first hours of use.

106 Ghiran

Fig. 7.5. Spectral characteristics of various light sources. a Mercury lamp, b xenonlamp and c metal halide.


Xenon lamps have a more homogenous output throughoutthe spectrum without any obvious emission peaks (Fig. 7.5b).The emission in the visual range is greater than that of themercury lamp, which is an advantage in certain fluorescenceapplications or when using particular fluorochromes. However,the output in the red and infra-red part of the spectrum is alsoconsiderable (almost half of the entire emission spectrum); there-fore, the heat generated by these sources is also significant andcreates problems during long time-lapse experiments. The heatproduced by the light source (xenon or mercury) is transmittedby the microscope frame to the stage and is reflected as an “x−y”drift of the image. To minimize this problem, certain companies(i.e. Olympus) manufacture microscope stands out of a metal–ceramic alloy that has a minimal thermal expansion coefficient andallows for long time experiments with a minimal heat-dependentdrift. Another efficient approach to correct for this problem is touse a liquid light guide that allows a complete separation betweenthe heat source (halogen or mercury/xenon lamp) and the micro-scope body.

A common problem for most fluorescence light sources is thevariability in the power output (seen as flickering of the excita-tion and emission light) due to either variation in the output ofthe power source or lack of stable connections between electrodeswhen the lamp is on. While correcting for the latter is more diffi-cult and requires the use of stabilized power sources (more expen-sive), the former is easier to prevent (never to correct) by simplyallowing a new bulb to “burn in” when turned on for the veryfirst time for at least one hour. By doing so, the discharge arcwill etch stable contact points (small pits) on the electrodes. Onthese two points (one on each electrode), the resistance will belower than on any other area of electrodes. Therefore, the nexttime the lamp is turned on, the discharge arc will form quicklyand stably between the two points generating a steady output oflight. Subsequently, if the lamp is turned off before “burn in,”the next time the lamp is turned on (and for the remaining of the200 h) the arc will “look” for areas with lower resistance on theelectrodes, producing quick drops in output as it jumps from onepoint to another. These drops in output are seen by the user as“flickering” of the fluorescence.

The main disadvantage of flickering is the inability to gener-ate any quantitative data using fluorescence microscopy, especiallyduring time-lapse experiments, as flickering of the lamp is com-pletely random.

Metal halide lamps are becoming more and more popularbecause they have all of the advantages of mercury lamps andfew, if any, of their disadvantages. The emission spectra is almostidentical but appreciably brighter, lamps last significantly longer(2000 has opposed to 200 h) and are almost flicker-free due to

108 Ghiran

a stabilized power supply. Most lamps are sold with a liquid lightguide that allows the power supply assembly containing the lampto be stored safely away from the microscope, preventing any heattransfer to the stand or stage. Also, using a liquid light guide takesaway the laborious task of centering the lamp after each replace-ment and provides a uniform light beam that will fill the backfocal plane of each objective. Disadvantages of liquid light guidesare the high susceptibility to mechanical strains that renders theminoperable seconds after stress (bending) and to the formationof air bubbles inside the guide after prolonged use that resultsin non-homogenous illumination. Taping the guide on the wallto prevent any accidental handling and making sure that the liq-uid guide does not form any sharp angles (especially behind thepower source) will prolong its use. Every 2000 h or so (each timethe lamp is replaced), unplugging the guide from the fluorescentilluminator and projecting it on the wall is a good practice tocheck the integrity of the light guide. If the resulting circle oflight (very bright white) is not uniformly illuminated, it may betime to purchase a replacement. Meanwhile, performing the flat-filed correction again for each objective is desirable (see below).

Never point the liquid light guide toward a person when thelamp is on and never point it toward reflective surfaces, as itis impossible to predict where the light will end up. Alwaysprevent people from entering the room while checking theintegrity of the liquid light guide by displaying a clear sign onthe door.

LED (light emitting diode) technology was adopted by themicroscopy industry only relatively recently, as in the past thelow power output of these light sources made them best suitedmostly for machine vision and the auto industry. Today, it isestimated that due to advances in LED technology the bright-ness of these devices doubles every couple of years, which makesLEDs the most promising illumination technology for fluores-cence microscopy. The output wavelengths of LEDs availabletoday cover most if not all of the fluorochromes that are currentlyin use. While the full width at half maximum (see above) is not asnarrow as the laser light, ultimately this is an advantage becausethe use of certain fluorochromes is no longer prevented by themismatch between the emission of the light source and the exci-tation of the fluorochrome. There are several advantages of theLED that make this technology superior to current illuminationmethods: LEDs are small, compact and have low power usagewith little heat generated during “on” times. LEDs also requireno warm-up time, last for 10,000 h or more with outstanding


stability (experiencing less than 1% decrease in the power outputover their lifetime). Another notable advantage over traditionalfluorescence sources is their ability to be turned on and off withinmilliseconds, circumventing the need for mechanical shutters, aknown source of vibration that becomes a serious imaging prob-lem especially during fast time-lapse experiments.

The downside of LED technology is the low power outputthat is equivalent (and that is true only for certain wavelengths)to a 75 W xenon lamp. However, it is just a matter of time beforethe LED will become bright enough to compete and likely replacemost standard mercury and metal halide fluorescence sources.

5.1.1. Operating Rulesof Fluorescence SourcePower Supplies

When turning on a fluorescence light source, there is a short (lessthan a couple of seconds) discharge of about 50,000 V that willcreate a powerful magnetic and electric field that potentially coulddestroy electronic components found nearby (computer-relatedequipment). Even though all of the cables and the power sourcesare (theoretically) shielded, the possibility remains. Therefore, ina laboratory with multiple users, it is always useful to post a warn-ing regarding the sequence in which components should be tunedon and tag each piece of equipment next to its power switchwith a label that specify the order and the total number of powerswitches that have to be on for all the equipment to work. Thisis necessary because older pieces of equipment have the on/offswitches hidden on the side or on the backside and are easy tomiss. Therefore labels should read: 1/7, 2/7 etc. In any case, asa rule, the first component that has to be turned on and the lastto be turned off is the fluorescence light source.

5.2. FluorescenceIlluminator

Fluorescence illuminator or epi-illuminator classically consists ofthe lamp housing located at the opposite part of the nosepiece,the collector lens assembly and the intermediate part that containsslots for balancer, neutral density filters, field diaphragm and theaperture diaphragm (see Fig. 7.6).

The collector lens assembly is located between the light sourceand the balancer. The lens system of the assembly (which islocated near the light source) can be adjusted i.e. focused, to col-lect (hence the name of the lens system) most of the light emittedby the source and send it to the filter cube turret. A second lenssystem (or several on newer microscopes) focuses the light orig-inating from the collector lens into and fills the back focal planeof the condenser (represented by the objective in epi-fluorescencemicroscopy). High optical quality of the collector lens assembly isa must especially as new devices that provide confocal-like imagesby creating virtual pinholes such as DSU (disk scanning unit,Olympus) or Zeiss (Apotome) are used with increasing frequency.These devices create images that are formed by using slits thatfocus light with different wavelengths on the sample. Therefore,

110 Ghiran

Fig. 7.6. Picture of an epi-fluorescence illuminator with its main parts labeled: CL, col-lector lens assembly; ND, neutral density filters; AD, aperture diaphragm – or sometimescalled aperture sto-p; FD, field diaphragm; FC, filter cubes. Before any imaging session,one has to check the status of all these accessories.

apochromatic corrected illuminators are essential. Each time thelamp is replaced it must be centered and adjusted to meet thecriteria for Koehler illumination. Most new microscopes have inthe lamp house either a concave or parabolic mirror. Therefore,the image of the lamp will be double when imaged without anobjective in the optical path. It is important to know that thepurpose of centering the lamp is not to overlap the two imagesof the lamp (one direct and the other one mirrored) but ratherto align them beside each other. Once this is accomplished, theimage must be slightly defocused by adjusting the collector lampto even out the illumination of the field of view. Although thisstep may seem unimportant, it is crucial when using fluores-cence images for quantitative purposes. When samples are imagedwith low-to-intermediate magnification objectives (10× to 60×)and recorded using medium-to-large format CCD cameras (1.2–3/4′′), an uneven illumination will cause unequal excitation offluorochromes in the sample, rendering different readouts fromvirtually identical fluorescent structures. To make the situationworse, because of the nature of fluorescence imaging where mostof the field of view is black, even significant variations in illumi-nation are not readily identifiable. Therefore, flat-field correctionmust be performed before any quantitative analysis for each objec-tive separately and, of course, after each bulb replacement. Thisoperation will correct not only for uneven illumination due to thelight source but also for dust particles located on optical surfacesbetween the objective and the CCD chip. Therefore, in theory,


each filter cube should be flat-field corrected individually. Evenon a homogenous illuminated field, the fluorescence originatingfrom the sample will be absorbed by the dust particles located onthe optical path, and the resulting fluorescence will be lower. Ifone quantifies changes in fluorescence of moving cells for a periodof time, an artifactual, biologically irrelevant, increase or decreasein fluorescence could be recorded when cells start in or cross suchareas.

Only images that are corrected in such a way are suitablefor any kind of image analysis when quantitative information isrequired. If parameters such as length, width, area etc., are ofinterest, flat-field correction may not be critical, but in any caseone should always be aware of it. If flat-field correction cannotbe done for various reasons and quantitative analysis must be per-formed, then comparing images that were acquired using the verysame area of the field of view could bypass the need for flat-fieldcorrection. This can be done by using an eyepiece reticule andacquire images that were framed in the same quadrant of the retic-ule. Nonetheless, results generated by this method are an approx-imation at best and should not be used routinely.

To perform flat-field correction one needs to acquire threeimages: (1) the image of interest, (2) a dark frame and (3) aflat-field image. The first image (I ) will contain structures ofinterest plus any artifacts induced by imperfect illumination, dustand electronic noise. Importantly, the exposure time should berecorded and used for the flat-field image and the dark frame.The second image, the dark frame (Df), has to be acquired withno light reaching the detector (Fig. 7.7a) and is needed to sub-tract the contribution of the electronic noise from both the flat-field image and from the image of interest. The third image, theflat-field (Ff) (Fig. 7.7b) can be the image of a concentrated solu-tion of a fluorescent dye (fluorescein) that lacks any structures.It may help to slightly defocus the image to obtain a homoge-nous field although it would be ideal to acquire all images at

Fig. 7.7. Surface representation of the dark frame and uncorrected uniformly fluorescent surface. a The noise levels areless than 5% of the maximum value. b Surface representation of a homogenous fluorescein solution showing significantvariation in illumination throughout the recorded field.

112 Ghiran

the same focus. It is also recommended that the exposure timeof the flat-field image be long enough so that at least half ofthe maximum value of the pixel is reached (around 2400 for a12-bit CCD). Because the exposure time is dictated by the imageof interest, using neutral density filters and changing the concen-tration of the fluorescein may be needed. Also, acquiring severalimages (10–15) that will be averaged in the end are needed tolimit the influence of noise in the flat-field image. Although theimage will look rather uniform, any image analysis program willshow significant differences in intensity throughout the image dueto imperfect illumination and dust on the optical components.Pseudo-coloring the image or displaying it in 3D-mode basedon the pixel value would make the unevenness of the field veryapparent (Fig. 7.7b). The exposure time should match that ofthe image of interest as the electronic noise increases with expo-sure time. Electronic noise is present in every image and has moreor less a constant value, which is about 5% (a pixel value of 180–220) of the maximum pixel value (4095 for a 12-bit camera).Scientific-grade CCD chips found in virtually all cameras suitablefor fluorescence microscopy will display, when imaging a perfectlyuniform bright surface, a range of values that are within 5% of themean pixel intensity. Consequently, when acquiring a backgroundimage (no illumination), the mean, lowest and highest pixel value,will be almost the same and around 5% of the maximum pixelvalue (see the “Y” axis in Fig. 7.7b).3 Therefore, when acquiringfluorescence images, one should aim for obtaining a pixel valueof structures of interest between 3600 and 3900 to increase thesignal-to-noise ratio. Saturated pixels, having a value of 4095, areuseless when the image is needed for quantitative analysis becausethe actual value of the pixel, whether it is 4095 or any other valueabove that, would still be displayed as 4095, making further anal-ysis inaccurate. Once all three images are available, most if not allof the image analysis programs have built-in capabilities that willuse the following formula (or a variation of it) to generate theflat-field corrected image, Ic(x,y):

Ic(x,y) = M (I(x,y) − Df(x,y))/Ff(x,y) − Df(x,y)

3 This is true for a certain exposure range that depends directly on the temper-ature of the chip. The lower the temperatures of the CCD chip, the lower thenoise levels during longer exposures times. In general, cameras used for astro-nomical imaging require deep cooling because it is not uncommon to expose anobject for hours at the time. Luckily, biological applications that require exposuretime longer than several minutes are rather rare. CCD cameras that are cooledat −40 to −65◦C are needed for applications where the fluorescence associatedwith biological structures or phenomena are very dim and short-lived. In theseinstances, it is crucial to lower the noise levels as much as possible so that faintfluorescence signal would become apparent.


This operation will subtract the value of each pixel of the darkframe (Df(x,y)) from the corresponding pixel in the actual image(I(x,y)) and flat-field image (Ff(x,y)) and divide the results. (M) rep-resents the mean value of the corrected flat-field frame. Becausethe value of pixels from the original image will be changed irre-versibly, this operation should be performed on a duplicate of theoriginal image.

The illuminator also houses the fluorescence shutter, as well asseveral important and often overlooked devices, such as the heatand neutral density filters, the excitation balancer, the apertureand the field diaphragms. Improper settings of any of these acces-sories could result in low-resolution, low-contrast images that areindependent of the sample and the optical components used togenerate images.

The heat filter is located between the collector lens andthe optical component of the epi-illuminator of all microscopes(Fig. 7.6) and is used to block wavelengths above 600 nm thatcould potentially overheat and shorten the lifespan of the coatedoptical elements. If near-IR or IR imaging is used, it may be use-ful to remove the filter to increase the excitation efficiency of IRfluorescent probes. Also, most cameras used for microscopy havea IR cut-off filter in front of the CCD/CMOS that could alsodecrease the efficiency of near-IR or IR imaging. One shouldalways check with the microscope/camera manufacturer beforeattempting to remove any built-in optical or mechanical com-ponents from the microscope or camera head. It is importantto remember that for 10–15 s or longer after the bright-fieldsource (likely a 100 W halogen lamp) is turned off, the bulb willcontinue to transmit infrared radiation (heat) that could poten-tially affect the CCD and fog the resulting image. If a bright-field method is used to locate the cell or tissue of interest andthe fluorescence imaging requires long exposure times, the IR-induced fogging could become a problem. There are several waysto avoid image fogging: waiting for 20–30 s once the halogenlamp is turned off to image in fluorescence mode, using a heat fil-ter or using a mechanical shutter in the bright-field path. A shuttermay be a better choice because all incandescent sources turn-onand die-off slowly. Without a shutter during time-lapse acquisi-tion of alternating fluorescence and bright-field images, such asphase contrast or DIC (see below for analyzer-induced artifacts),there will be a bleed-through of the bright-field light into thenext fluorescence image, especially during fast acquisition. Hav-ing a mechanical shutter will completely circumvent the problem,although it may transmit significant vibrations in the microscopestand that would blur the next fluorescence image. Most soft-ware dedicated to image acquisition have the option to delay thestep that follows the closing or opening of a mechanical shutter(either on fluorescence or bright-field path) or change in X−Y or

114 Ghiran

Z-axis of the microscope stage, to prevent blurring of the imagedue to mechanical vibrations. This may not be a viable option iffast acquisition is required. Having liquid light guides for bothlight sources (bright-field and fluorescence) completely circum-vents this problem, as the mechanical shutters are located awayfrom the microscope stand.

The excitation balancer (EB) slider (Fig. 7.6) is usually foundin the fluorescence illuminator between the collector lens assem-bly and the neutral density filters and is useful only when usedwith a double or triple filter cube in the turret. It functions asa variable excitation filter, shifting the excitation wavelengths byeither changing the angle of a dichromatic mirror from 0 to 45◦(system employed by Olympus) or changing between a long anda short-pass excitation filter (Nikon). The balancer is very usefulwhen an image with two or more fluorochromes is analyzed andcertain features are stained very brightly with one fluorochromewhereas others stained with another fluorochrome are very dim.By gradually decreasing the excitation of the bright fluorochromeand increasing the brightness of the dim fluorochrome, the topo-logical relationship between the two structures can easily be rec-ognized without the need to repeat the experiment to correctfor the unbalanced signal strength. Also, colocalization (if any)between two colors is easier to record because there is no “pixelshift” artifact, a common occurrence when using two or moresingle band cubes (see below). The disadvantage is that in orderto record images rendered by a balancer, a color camera is neededand most of the scientific-grade cameras use monochrome CCDchips. Using a color filter in front of the camera can easily fixthis problem. The downside of having the balancer installed on amicroscope that is shared among several users is that more oftenthan not, after they are done using it, most users leave the bal-ancer engaged in the optical path. So when the next user imagesa sample using a single band filter cube, a significantly dimmerimage than normal (or sometime no image at all) is obtained forreasons that have nothing to do with the sample itself. Again, agood practice is to always check the position of all accessories thatcould be in the optical path before any imaging session.

Neutral density (ND) interference filters are used to decreasethe intensity (amplitude) of the excitation light (Fig. 7.6) byreducing the intensity uniformly across the spectrum (from 400to 700 nm). In fluorescence microscopy, filters are used mainly toprotect the sample from excessive bleaching and, when imaginglive samples, to diminish the generation of reactive oxygen speciesthat depend directly on the amount of UV illumination. Mostmodern fluorescence microscopes have either two neutral filtersin the optical path or a wheel with several ND filters of increas-ing absorbency from 20 to 90%. Depending on the mechanismemployed by the filter to attenuate the incoming light, there are


two types of ND filters: absorbing ND filters that have a smokedappearance or reflecting ND filters that look like a mirror on thereflecting side (which must face the light source).

The aperture diaphragm (Fig. 7.6, AD) can usually be cen-tered on the X−Y axis and is located in a conjugated plane withthe rear aperture of the objective and, therefore, controls theintensity of excitation without affecting the size of the area thatis illuminated. Closing the aperture diaphragm will reduce theglare of bright fluorescent images, although in our experience,it is less useful than the iris diaphragms that are fitted on somehigh NA, immersion objectives. A semi-closed or closed aperturediaphragm will render dim images, and, because it does not affectthe size of the field of view but only the amount of fluorescencelight that illuminates the sample, the problem is not obvious fromthe beginning. The status of this diaphragm should always bechecked before any imaging session. The field diaphragm (Fig.7.6, FD) is also centerable and is located between the aperturediaphragm and the filter cubes and can be either round or rect-angular. It provides very useful means to control the size of theilluminated area to prevent unnecessary bleaching of the sampleand to increase the contrast in samples with high background. Formaximum contrast, the illuminated field should be as large as thestructured imaged even if it means closing the diaphragm downall the way. However, this can also be a means to block-out theexcitation of structures deemed “non-representative”; therefore,it should be used appropriately.

Filter cubes (Fig. 7.6, FC) are located above the objective ineither a manual or motorized turret (newer microscopes). Oldermicroscopes used separate sliders for excitation and emission fil-ters. Very few, if any, such microscopes are found today in researchlaboratories and, therefore, we will not discuss them any further.A fluorescence filter cube has three main components: an exci-tation filter, a dichromatic filter (dichroic mirror or beam split-ter) and an emission filter (Fig. 7.8). Please note that arrowson the emission and excitation filters point in different directionsdepending on the manufacturer of filters.

There are three main types of interference filters (Fig. 7.9)that are classified according to the range of wavelengths they allowto pass. The blocking range is usually not mentioned in the nameof a filter. The three types are short pass (SP) filters that allowlight with short wavelengths to pass while blocking everythingelse, long pass (LP) filters that allow light with long wavelengthsto pass while blocking short wavelengths and band-pass (BP) fil-ters. Band-pass filters allow only a range of wavelengths fromthe spectrum (a band or a window) for which the filter is per-fectly transparent. Figure 7.9 shows the measured transmissionprofiles of the three main filters that are characterized by centerwavelength (CWL, band filters), peak (P %T) and average

116 Ghiran

Fig. 7.8. The basic configuration of a fluorescent filter. Parts are shown before assemblyfor a better representation of the relationships between various components. Note theangled emission filter. The orientation of filters as shown by arrows on the emission andexcitation filter are not universal. See text for details.

Fig. 7.9. Representation of the three simple filters used for fluorescence microscopyand their defining parameters. Long pass (LP) and short pass (SP) filters are definedby their cut-on and cut-off wavelengths. The bandpass filters (Bandpass) are defined bytheir peak transmission, center wavelengths (CWL) and full-width half maximum (FWHM)(Image courtesy of Chroma Inc.).

transmission (Avg. %T), or for bandpass filters, the full-width halfmaximum (FWHM), which is a measure of the filter slope (thesteeper the better). Combinations of these three filter types arefound in virtually all filter cubes. Figure 7.10 shows the diagramsof a high-quality simple-band DAPI filter. The novel methodsemployed to manufacture filters have reached almost the lowestlimit for auto-fluorescence of the glass and the highest limit for


Fig. 7.10. Diagram showing a high-quality single-band DAPI filter. Note the almost rect-angular bandpass of the excitation and emission filters that have the 50% cut-on andcut-off values almost identical with the blocking range. The dichromatic mirror has analmost vertical slope and close to 0% transmittance for the excitation range and over90% transmittance for the emission range (Diagram courtesy of Semrock, Inc.).

transmittance, barrier and slope of filters. Note the almost per-fect wavelengths selection of the interference filters that ensuresbright signals and black background.

Excitation filters: In the past, selecting the desired wave-lengths was realized by using colored solutions of specific dyesthat were placed in glass containers in the light path. Coloredgelatin sheets and colored glass filters later replaced containers.These approaches for selecting the excitation wavelengths workedwell with trans-fluorescence methods where samples were illumi-nated through an immersion dark field condenser, which due toits specific design prevented most of the high-energy excitationlight to enter the objective and allowed only the emitted fluores-cence to form the image. While these filters were very affordable,they suffered from poor optical performance, low transmittanceand mediocre wavelength discrimination. A significant problemwas the high level of auto-fluorescence of dyes that were usedin these filters, which considerably decreased the signal-to-noiseratio of most samples. In recent years, the majority of filters usedin fluorescence microscopy are interference filters. The main tech-nologies used today for constructing an interference filter useeither soft or hard coating. Soft coating is based on deposition onseveral optically flat pieces of glass glued together by adhesive ofmultiple thin films of several different dielectric substances (metal

118 Ghiran

salts). The alternative, hard coating method, deposits dielectriclayers on a glass disc using a special technology (ion-beam sput-tering) that creates several layers on just one optically flat piece ofglass avoiding the use of any adhesives, which are responsible formost of the auto-fluorescence of the soft-coated filters. Aside frombeing more resistant to light damage and moisture, the later filtersare easier to manufacture with perfectly flat surfaces, a paramountcharacteristic when co-localization studies between two, three ormore colors are performed.

5.2.1. RegistrationArtifacts When UsingTwo or More Filters

In Fig. 7.11a is shown the overlap image of a triple-labeled10 μm-bead imaged by two microscopes, one using four stan-dard DAPI, FITC, TRITC and triple-band filters and the otherusing filters of the same type but of “no image-shift” quality. Inthe first image, all images are shifted in respect with the opticalaxis creating the false impression of no or partial co-localization,although all images originate from the same structure. The “noimage-shift” grade filters do not suffer from such an effect (Fig.7.11c). To better represent the lack of co-localization, imagesin the left were processed using the Sorbell filter (ImageProPlus,Media Cybernetics) to show only the outline of the imaged bead(Fig. 7.11b). The origin of the problem can be the “wedge”shape of the emission filter but also the “wedge” shape of thedichromatic filter (dichroic mirror) that would create the sameeffect regardless of the quality of the emission filter. In some sit-uations both filters suffer from the same defect. Therefore, it isnot indicated to “mix and match” dichromatic and emission fil-ters of different qualities if co-localization studies are needed. Ifstandard filters are the only ones available, co-localization stud-ies still can be accurately performed if the amount of shifting fortwo channels is calculated, assuming the third one to be perfectlyaligned with the optical axis of the microscope and a back-shift

Fig. 7.11. Filter cube-induced artifact during co-localization. A triple-labeled 10 μmlatex bead was imaged either through regular (a) or co-localization-quality (c) filtercubes. The white image of the bead (a, left) was recorded using a triple filter. The raw (a)and the processed image (b) show significant misalignment among filters, although thesame structure contained all fluorochromes and there was no shift of the bead duringacquisition. Co-localization-quality (c) filter cubes rendered a perfect overlapping image.The difference in bead size between the two images is due to the different pixel size ofcameras used for the acquisition.


correction is applied to the first two channels. All major imageacquisition and processing pieces of software, as well as severalfreeware programs, have built-in modules for this option. Hav-ing triple-labeled beads on the slide along with cells of interestduring acquisition can considerably simplify the task. The smallerthe beads, the more precise the correction. If the amount ofshifting between channels changes overtime, it could indicatea loose component (likely the dichroic mirror) in one of thefilters.

Types of interference filters: Depending on which range ofwavelengths a filter allows to pass, there can be band-pass filtersthat allow only certain range of wavelengths to pass and can beeither narrow or broad band-pass, short pass edge filters that allowonly short wavelengths to pass while blocking long wavelengthsor long pass edge filters that block short wavelengths and onlyallow long wavelengths to pass (see Fig. 7.8). Understanding theproperties of a filter and the way it does or does not match theexcitation and emission spectra of a certain fluorochrome is essen-tial when buying a filter or, more importantly, when assemblinga new filter cube from components available in the laboratory.Most filters must be mounted in a certain way and the directionis usually shown with an arrow that has to point to the dichro-matic mirror. Also, it is important to know that most filters (socalled soft-coated) are considered consumables with a limited lifes-pan that, depending on how much the microscope is used, haveto be replaced every 1–2 years, as the transmittance will decreasebelow 20–30% (so called “burnout” filters). Figure 7.12a showsthe picture of a soft-coated filter 9 months after installation fol-lowed by average to high usage. Figure 7.12b shows in detailthe moisture damage centered on impurities found between thethin layers of coating. Using the microscope in areas with highhumidity also favors the growth of fungus on all optical surfaces

Fig. 7.12. Damage of a soft-coated emission filter after less than a year of moder-ate to heavy use. a Image on the left shows the damage that started at the periphery(discolored areas) and then progressively moved toward the center. The first indicationof the damage was the out-of-focus round shapes found mostly at the edges of thefield of view noticed first during bright-field imaging. b Image on the right shows a lowmagnification detail of the affected area (see text for details).

120 Ghiran

(including filters) that drastically alters their optical properties. Avisual inspection would show structures that look like hairs orvines branching from a center if there is fungal growth presentor round darker areas if there is any delamination due to heator impurities in the filter coat. All these problems could eventu-ally lead to photo-darkening. Gradual decreases in light efficiencywhile imaging in fluorescence and the appearance of out-of-focuspatterns when the filter is used in bright-field are signs that the fil-ter has to be changed. It may be wise investing in hard-coated fil-ters that, although are more expensive than the regular filters, donot “burn-out.” However, they are not impervious to moisture-related damage (fungal growth). Most microscope manufacturesuse an antifungal treatment of all optical components (lenses andprisms) to slow down the growth of fungus that ultimately wouldetch the glass surface underneath rendering it useless.

Dichromatic beam-splitters (dichroic mirrors) are special longor band-pass filters that are mounted at 45◦ degrees in the fil-ter cube. Any variation of this angle will significantly change theintended optical properties of the filter. This property is used tochange the range of wavelengths that will pass through the fil-ter (see above, excitation balancers). Once the excitation filterhas selected a certain range of wavelengths, the dichromatic fil-ter performs like a mirror for the incoming shorter wavelengths,characterized by a percent reflection (close to 100% for certainwavelength range), reflecting the light through the objective intothe sample. The fluorescence that originates from the sample hasa longer wavelength than the excitation light (Stokes shift, seeabove) and, therefore, the dichromatic mirror allows the light topass un-reflected (therefore characterized by percent transmissionclose to 100% for a certain wavelength range). Depending on thetype of the filter cube that it is used, some dichromatic filters canbe long pass filters (single excitation filter cubes) or band-pass fil-ters (double and triple filter cubes). If the dichromatic mirror sur-faces are angled (edge-shaped) instead of being parallel, the imageformed by that particular filter cube will be shifted from the opti-cal axis of the instrument. The same is true for the emission filters(see above). Figure 7.13 shows the diagram of a triple-band fil-ters that allows simultaneous and specific excitation and emissionof several (up to three) fluorochromes. Note that in a triple-bandfilter the dichromatic mirror and both the emission and excitationfilters are band-pass type filters.

Emission filters can be long pass (allows light with longerwavelength to pass), increasing the overall brightness of the sig-nal but also increasing the background or band-pass type, inwhich case the brightness decreases slightly but the backgroundis darker. They are easy to identify in a filter cube because theyare mounted under a 5◦ angle to decrease the brightness of thebackground and prevent back-reflection.


Fig. 7.13. Diagram of a high-quality triple-band-pass filter. The filter is designed toallow simultaneous imaging of up to three fluorochromes. Note that all the filters foundin this cube are band-pass filters (Diagram courtesy of Semrock, Inc.).

5.2.2. Orientation of theFluorescence Filtersin a Filter Cube

The direction a filter has to be mounted in a filter cube is usuallyindicated with an arrow, but different companies use the arrowto point to different parts of the filter cube. In most cases, thearrow on the excitation filter points toward the dichromatic mir-ror (center of the cube), whereas the arrow on the emission fil-ter points toward the dichromatic filter (Chroma) or towardseyepieces (Semrock). The situation is more complicated for thedichromatic filter, which has a coated surface that always has toface the light source for optimal performance. Because the dichro-matic mirror is very thin, there are no arrows to show the properorientation of the mirror. One can find the coated surface if oneuses a light source (a desk lamp works well) to reflect it off fromthe dichromatic mirror; if the coated surface is pointing up, therewill be only one image of the light source; if the coated sur-face points down, the image of the light source will be doubledand slightly shifted. If the performance of a filter set is less thanexpected, especially if the filter was custom built from preexist-ing parts in the laboratory, it is always a good idea to check theorientation of filters and consult the manufacturer.

5.2.3. Methods toPrevent Stray Light fromExiting the Filter Cube

Because there are no perfect filters, a certain percentage of thelight that traverses through the excitation filter, instead of beingredirected by the dichromatic mirror into the objective, passesthrough it and bounces off the back wall of the cube and thenis reflected by the dichromatic mirror though the emission fil-ter into the eyepiece or detector, increasing the brightness ofthe background and decreasing the dynamic range of the actual

122 Ghiran

image. Various methods have been devised over time to preventthis from happening, starting with coating the interior surface ofthe filter cube with a black, non-reflective paint (black baffled)or mounting the emission filter under an angle of 5◦. In addi-tion, increasing the efficiency of the filter also prevents reflectionof the back-reflected light by the telan (tube) lens. Some micro-scope manufactures, in addition to an angled emission filter, usea slightly modified filter cube that contains either an absorbentneutral density filter on the opposite side of the excitation filter oran opening (Fig. 7.14) that allows stray photons to completelyescape the filter cube (“Light-trap” first employed by Zeiss andBackground terminator, employed by Nikon).

The next component in the light path is the objective, whichin all epi-illumination techniques plays the role of the condenser.

Fig. 7.14. Comparison between a standard fluorescence filter cube and light-trap(Zeiss). a In a standard fluorescence filter cube, the light that passes through the dichro-matic filter (thin green arrows) is reflected between the back wall of the filter cube andthe upper surface of the dichromatic filter, and crosses the emission filter increasing thebackground level of the final fluorescence image. b The light-trap filter cube (1) has theback wall removed so after passing through the dichromatic mirror (2), the stray light(3) is directed by the tapered wall of the fluorescence illuminator (4) outside the cube(5) thus improving significantly the signal-to-noise ratio (Diagram courtesy of Carl ZeissMicroImaging, Inc.).

5.3. Objectives Maybe the most important and specialized optical componentof a microscope is the objective, which points to and forms theimage of the examined object. In epi-fluorescence microscopy,the objective is also the condenser in the optical path, focusingthe excitation light into the specimen. There is no other micro-scope component with a greater role in the image formation thanthe objective. In the diagram shown in Fig. 7.15, a cross-sectionthrough a microscope objective illustrates the mechanical andoptical complexity of a fully corrected plan-apochromatic objec-tive that can contain between 13 and 15 lenses.

Therefore, the quality (or the lack thereof) of the objectivewill be crucial in determining the amount of information in thefinal image. That is not to say that a low-quality filter cube oran imaging device that is unsuitable for a certain application (due


Fig. 7.15. Cross-section through a high-magnification planapochromatic objective. Theimage shows the mechanical and optical complexity of high-quality objectives that cancontain up to 15 carefully polished and aligned lenses. Any mechanical stress (due tohitting or dropping of the objective) could potentially shift lenses from the optical axisof the objective significantly decreasing its optical qualities (Image courtesy of LeicaMicrosystems).

to speed, sensitivity, dynamic range or resolution) will not have anegative impact on the final image. However, irrespective of thequality of the other components involved in the image forma-tion or acquisition, if the objective does not have the requiredNA or the appropriate chromatic and spherical aberration correc-tions for a particular application, no amount of correction or post-acquisition processing will add the information that was missedby the objective. Therefore, for demanding specialized applica-tions (calcium imaging etc.), building a new system starts withthe objective followed by the camera, filters etc. If the resolutionrequired for an application is known, for instance 0.25 μm, it isalways a good idea to buy only that particular objective of thehighest quality (such as PlanApo 60 × 1.40, 1.42 or 1.49) andthe rest of the objectives (4×, 10×, 20× and maybe 100×) ofstandard quality (planacromats or planfluorite, see below). Thepremium one payments for fully corrected lenses are in the rangeof thousands of dollars and the money could be better usedtowards other components (camera, filters, image acquisition and

124 Ghiran

analysis software). If for most of the bright-field applications thetype of objective (correction-wise) is not critical, for fluorescencemicroscopy, and especially for application such as co-localizationof two or more colors in three dimensions (3D), the choice ofobjectives is paramount, due to the impact of the chromatic andspherical aberration on image formation (Fig. 7.16).

Fig. 7.16. Diagram showing the two main aberrations of microscope objectives. aSpherical aberrations are due to fact that light passing through the edge of a lens isrefracted closer to the lens than the light going through the center. Therefore, a simplelens will not have a single focus but rather the focus will be present on a certain dis-tance (area) along the optical axis of the lens between the two extremes formed by theperiphery and central rays. The place along this area where focus is the “best” is calledcircle of least confusion and it was first described for astronomical refractors and pho-tographic cameras. Its size depends on the focal distance of the lens and its diameter. bChromatic aberrations are a direct consequence of the variability of the refractive indexof a transparent material with the wavelength. Therefore, blue light will be refractedmore (closer to the lens) and red light less (farther from the lens) inducing a halo aroundthe imaged objects that changes color with focus.

There are several major optical aberrations that affect, in var-ious degrees, all microscope objectives due to the geometry oflens and the wide variation of the refractive index of a particularmedia (glass or water) with wavelength and temperature: sphericalaberration, chromatic aberration, astigmatism, field curvature andcoma. We will address here only the first two as they are directlyrelevant for fluorescence imaging.

Spherical aberrations are due to the fact that the thickness ofa lens varies significantly between the center and the periphery.As a consequence, the light is focused in different points alongthe optical axis depending on the place where the incident lightentered the lens. The light that is closer to the optical axis willpass through the lens with minimal refraction focusing fartherfrom the lens, whereas the light entering the lens near the periph-ery will undergo a more significant change in a direction focusingcloser to the lens (Fig. 7.16). The larger the frontal lens (aper-ture), the worst the aberrations. Images that result from an insuf-ficiently spherically corrected objective are soft, lacking contrastand sharpness. It is important to remember that corrections thatare built into any objective are calculated for specific conditions:


certain refractive index and temperature of the media between thelens and the coverslip, specific distance between the top surface ofthe coverslip and the imaged objects, and the presence or absenceof the coverslip. Any departures from such conditions would sig-nificantly degrade the final image. Efforts to decrease sphericalaberration started, surprisingly, over 300 million years ago whentrilobites developed as the first animals that had crystalline lensmade out of two lenses with different refractive indexes. Newmicroscope objectives have lenses manufactured out of specialtypes of glasses that are polished at precisely calculated angles∗that significantly minimize these aberrations.

Chromatic aberration is particularly important in fluorescencemicroscopy and, therefore, we will discuss it in more detail. Thisaberration is not due to the geometry of the lens but rather isa direct result of light interacting with glass (or any transpar-ent media). Depending on the wavelength of light, the refrac-tive index of the glass changes, increasing as the energy of thelight increases and the wavelength decreases. This means thatif white light passes through a lens, its blue component will berefracted closer to the lens, and the red component further fromthe lens. Green light will be focused in between (Fig. 7.16). Onan X−Y plane, this results in images of objects that have col-ored rings around them. On the X−Z and Y−Z planes this meansthat in order to focus properly using the three colors one has tochange the height of the stage to accommodate the three focalpoints from the optical axis. The danger of using an insufficientlychromatically corrected objective for 3D co-localization studiesbecomes obvious. A Z-stack of the same very object recorded inblue, green and red channels will be displayed at three differ-ent heights when a 3D rendering of the stack is performed. Thiswould suggest that depending on the size of the object, there areeither adjacent or even separated structures when, in fact, there isjust one structure. The further away the fluorochromes used forlabeling (blue, red or far red), the more likely the objective willgenerate an artifactual image.

Based on the degree of chromatic correction, there are threemain types of objectives: achromat, fluorite and apochromatic.Achromatic objectives are usually found in low-end microscopesand are corrected for two wavelengths, blue (490 nm) and red(660 nm), which are brought in a single focus. From a spher-ical aberration point of view, acromats are corrected for greenlight. These lenses are not particularly suitable for fluorescencemicroscopy and should be avoided whenever possible. Fluoriteobjectives were designed to correct for two or three wavelengthsby using a glass containing a special mineral, fluorite (hence thename of the lens). Because the mineral is very transparent for blueand UV light, fluorite objectives were used primarily for fluores-cence microscopy. In addition to increased UV transparency, the

126 Ghiran

optical properties of the fluorite lens allows for correction of anobjective for spherical aberrations for up to three wavelengths.This translates to objectives with a higher NA for the same mag-nification compared to acromats that significantly increases theirresolution and brightness. While acromats may not have a specificsymbol on the barrel, fluorite objectives are usually labeled as FL,Fluo or Fluorite. The third type of lens, the apochromatic objec-tives, first built by Zeiss following the specification of Ernst Abbein 1886, offers the highest level of correction for both spher-ical and chromatic aberrations. However, all these correctionscome with a very high price tag. Apochromatic lenses are cor-rected for four colors (some super-apochromatic objectives evenfor five wavelengths) for both chromatic and spherical aberra-tion. These lenses have the highest NA and usually provide thebrightest image. Due to their NA, apochromats are ideal for flu-orescence work, although for some specialized methods, such ascalcium imaging with ratiometric dyes excited in the UV range,major microscope manufacturers sell particular fluorite lenses thatsurpass apochromats. The correction type for apochromatic lensesis written always on the barrel as Apo, or Apochromatic.

On the objective barrel, in addition to the extent ofspherical and chromatic corrections, there is always additionalinformation such as magnification (both as number and colorcode), numerical aperture (as a number or, in the case ofobjectives with a built-in iris diaphragm, as a range or num-bers), the type of tube length either finite (160 mm), “160”(Fig. 7.17, right) or infinity tube length microscope systems“∞” (Fig. 7.17, left) application suitability (such as, phase

Fig. 7.17. Comparison between a new (infinity corrected) and old-style (finite corrected)microscope objectives. The type of geometrical (plan) and chromatic (SApo or Apo) cor-rections are clearly indicated on the barrels. Also, the correction collar range, tube lengthand the eyepiece field of view (26.5, left) are shown.


contrast, differential interference contrast, polarization, darkfiled), requirement for coverslip (“–” for indifferent, “0” for nocoverslip and “0.17” for regular coverslip (170 μm thick), andcorrection range for coverslips’ thicknesses common for high NAdry objectives. On newer objectives, additional information isadded by some manufacturers but not others, such as the max-imum field of view of the eyepiece expressed in mm (FN = 26.5),the type of DIC prism needed for the objective and the locationwith respect to the back focal plane of the objective, and work-ing distance (WD) of the objective, which is measured from thefrontal lens to the top of the coverglass (also expressed in mm).Aside for the color codes for linear magnification of objectives andtype of immersion required (see below), which are universal acrossall microscope manufacturers, each major brand has its own set ofcodes both letters and colors for extra corrections or suitabilityfor a particular method that makes reading (and understanding)the full potential of an objective a bit of a challenge even for anexperienced microscopist.

1/2x no color assigned1x, 1.25x, 1.5x Black2x, 2.5x Brown 4x, 5x Red10x, Yellow 16x, 20x Green 25x, 32x Turquoise 40x, 50x Light Blue 60x, 63x Cobalt Blue 100x, 150x, 250x White

Correcting spherical and chromatic aberrations for oil-immersion objectives is somewhat easier because standard immer-sion media has a refractive index close or identical with that ofthe coverglass (1.515) and the frontal lens of the objective. Fordry objectives, the air in between the frontal lens and the cover-glass along with a variable amount of mounting media betweenthe bottom of the coverglass and the actual imaged plane in thespecimen create additional problems that become significant asthe NA increases. The vast majority of dry objectives are designedto image samples covered with a 0.17 mm (170 μm) coverglasswith a refractive index of n = 1.515 and with minimal amount ofmedia between the coverglass and the specimen. It is importantto know that 170 μm is the distance between the top surface of thecoverglass and the sample to be analyzed for which the objective wasspherically and chromatically corrected. This value includes all themedia that is between the coverglass and the object (6).

An objective will only form a perfect image of the sample thatis adjacent to the coverslip, at 170 μm of distance from the topsurface of the coverglass. If the coverglass is thinner than 170 μm,

128 Ghiran

then the combined thickness of the coverglass and the media thatis on top of the sample (which is expected to have a refractiveindex of n = 1.515) has to be 170 μm. For no other depth willthe objective perform as expected. Having structures in the imagesurrounded by a halo is a certain sign of significant spherical aber-ration, likely due to excessive mounting media (assuming thatthe frontal lens is not dirty). This effect is less obvious for lowNA objectives, which are less sensitive to variations in coverglassthickness. Figure 7.18 shows two images recorded on the sameslide that was mounted with optimal amount of mounting mediaon the left side of the slide and excess in the right. The nucleusshowed on the left was right below the coverglass, whereas thenucleus on the right, although next to the one on the right, hada significant amount of mounting media present between its toppart and the coverglass. Note the increased background levels anddecreased resolution on the right image. Images were recordedunder identical conditions using an immersion objective (60 ×1.42 UPlanApo).

As a rule of thumb, all specimens must be mounted (not justplacing a coverslip over a dry specimen) in media with a coverslipif an objective with an NA higher than 0.3 is used for examina-tion. A few high magnification objectives (90 and 100×) requirespecimens, usually blood smears, to be examined dry withoutcoverslip. In this case, mounting the specimen with a coverslipwould significantly degrade the image because the objective isdesigned to form an image without coverslip. The lack of blackring (which signifies oil immersion) on a high-magnificationobjective and the inscription “0” on the barrel (160/0 for finiteor ∞/0 for infinity microscope systems) should help identifysuch objectives. On the positive side, these objectives are rarelyfound in research laboratories.

Fig. 7.18. The effect of spherical aberration induced by increased distance between thecoverslip and the imaging sample. Epithelial cells stained with DAPI, present either onthe coverglass or on the microscope slide, were mounted together (coverglass + slide)and imaged under identical conditions. The nucleus of a cell adherent on the coverglassis shown on the left. The nucleus on the right was fewer than 500 μm away from thenucleus on the left but more than 90 μm below. Note the increase background andreduced resolution on the left nucleus. Scale represents 10 μm.


In general, for objectives with NAs between 0.30 and 0.7,optical aberrations introduced by coverslips with thicknesses otherthan 0.17 mm does not degrade the image significantly. Forhigher NAs, dry objectives (0.7–0.95), minimal variations of thecoverslip thickness, increased distance (over 10–20 μm) betweencoverslip and specimen, and differences in the refractive index ofthe mounting media and coverglass are major sources of spheri-cal and chromatic aberrations that significantly alter the quality ofthe final image. To compensate for these aberrations, all high NA,dry objectives or immersion objectives that use media other thanoil (such as glycerol or water) have a built-in correction collar thatwill compensate for a range of coverslip thicknesses that is writtenon the objective barrel. For most of the objectives that are usedin upright microscopes, the range is typically between 0.11–0.23,whereas for most objectives that are dedicated for work with tis-sue culture dishes on inverted microscopes (phase contrast, DICNomarski and fluorescence), the range is wider, between 0 and2 mm. If such objectives are used to examine blood smear prepa-rations, the correction collar has to be set at 0, which signifies nocoverslip. Also, the side with the sample should face the objective.

Since the introduction of fluorescent tags that allow in vivotracking of intracellular proteins for long periods of time, imag-ing cells in physiological conditions (temperature, CO2 and O2pressure) has become the norm in many laboratories that aregeared towards microscopy (7). These new approaches requirenot just environmental chambers but also newer optics to accom-modate imaging in new conditions. Because the refractive indexof a media decreases as the temperature increases, newer immer-sion objectives also incorporate a correction collar to compensatefor the aberration induced by variations in the refractive index ofthe immersion media when the objective is used at room temper-ature (Fig. 7.19).

When adjusting the correction collar, a group of lenses insidethe objective barrel moves closer to the specimen when correctingfor a coverslip that is thicker than 0.17 mm coverslip or awaywhen correcting for thinner coverslips. The correction collar isa very valuable tool if used correctly but also can render images(especially fluorescence images) almost useless if used improperly.When imaging fluorescent samples, improper adjustment of thecorrection collar renders hazy images (due to presence of highlevels of spherical aberrations, Fig. 7.20, left), which decreasesthe dynamic range of the final image and makes focusing difficult.Although it may seem a chore to adjust the correction collar eachtime one changes the slide (or even on the same slide if there is anuneven distribution of mounting media), it is a worthy endeavor(Fig. 7.20, right). Correction collars are misused (or not usedat all) to such a degree that some microscope dealers refuse toquote them to avoid having users complaining about poor qualityof images when imaging with these pricey objectives. Also, it is

130 Ghiran

Fig. 7.19. The effect of spherical aberration induced by increased distance between thecoverslip and the imaging sample. Epithelial cells stained with DAPI, present either onthe cover glass or on the microscope slide, were mounted together (coverglass + slide)and imaged under identical conditions. The nucleus of a cell adherent on the coverglassis shown on the left. The nucleus on the right was fewer than 500 μm away from thenucleus on the left but more than 90 μm below. Note the increase background andreduced resolution on the left nucleus. Scale represents 10 μm. The highest NA objec-tive currently available that does not require special oil or coverglass. Only a decade agothe highest NA available for 60× and 10× objectives was 1.40. Note the correction col-lar for temperature. Temperature variations of the immersion oil induce small variationin its refractive index that, due to high NA of this objective, would significantly impactthe image quality (Courtesy of Davis Warren, Perkin-Elmer).

important to keep in mind that when adjusting the correctioncollar, the parafocality of the objective will change; hence, whenswitching back to another objective, one will have to refocus theimage, although all objectives are theoretically parafocal.

5.3.1. Adjustingthe Correction Collar

1. Set the collar so that the vertical mark points to the 0.17writing on the objective barrel, which is the thickness,expressed in millimeters, of most coverglasses used for bio-logical applications.

2. Focus on a small high contrast detail on the specimen3. Rotate the correction collar slightly to the left (0.18, 0.19),

because most of the samples are mounted with an excess ofmedia that create a thicker than optimal coverslip plus mediastack.


Fig. 7.20. Comparison between images of the same sample recorded with two differentcorrection collars settings. The image on the left was taken with the correction collarset at 0.17, whereas the image on the right was recorded with the collar set at 0.20.Both images were recorded with the same objective (SPlanApo 40 × 0.95), using thesame exposure time, dark level, gain and aperture diaphragm setting. The only differ-ence between the two images is the correction collar setting. The image on the rightappears blurry and out-of-focus with high background (Images courtesy of Sorina Ghi-ran, BIDMC.).

4. Refocus; if the image has not improved, continue to rotatethe collar to the left while refocusing.

5. If the image keeps deteriorating, reverse the direction, rotat-ing the correction collar to the right slightly and refocus.

6. Repeat the operation until the image becomes crisp.

6. Phase Contrastand Fluorescence

Since fluorescence images show only the tagged structures in cellsor tissues, it is sometimes useful to show the same image bothusing fluorescence imaging and a contrast enhancement tech-nique, such as phase contrast or DIC Nomarski, that would showthe whole cell or tissue, allowing a better orientation in the sam-ple. Using a phase contrast objective for imaging fluorescent sam-ples will significantly decrease the brightness of the image due tothe presence of the phase annulus near or in the back focal plane ofthe objective. If the analyzed sample is fixed, longer exposure willrender the desired brightness levels. If the sample is alive, increas-ing the binning or the gain of the camera will allow shorteningthe exposure time enough that even if images are recorded succes-sively, there will be minimal or no shift of the imaged structuresbetween the fluorescence and phase contrast image. Of course,this may not be true in all instances. However, if the same phasecontrast objective is used to acquire “z”-stacks of images that areused for deconvolution, the “ghost” of the phase ring will induceartifacts in the final restored stack that have no real correspondent

132 Ghiran

Fig. 7.21. Ghost image of the phase ring during fluorescence imaging. The out-of-focus image of a 10 μm fluorescent bead acquired using a regular objective suitablefor fluorescence work (40 × 1.30 UPlanFl) is shown on the left and the same imagerecorded with a 40 × 1.00 UPlanApo phase is shown on the right. Note the dark ringthat is overlapped with the actual (out-of-focus) image of the bead. The ring representsthe “ghost” of the phase ring, which is located on the back focal plane of the objective.The “ghost” will induce an artifact in the final image if a stack of images acquired witha phase contrast objective is used for deconvolution. When the sample is in focus thering will induce a significant drop in fluorescence and slight decrease in resolution.

in the original stack (Fig. 7.21). Therefore phase contrast objec-tives are not suitable for acquiring images used for deconvolutionand should be avoided if 3D-acquisition followed by deconvolu-tion is needed.

7. DIC andFluorescence

Objectives that are used to generate DIC Nomarski images donot contain any phase rings that would negatively impact the finalimage, but in order to generate the final 3D-like image, they dorequire, among other components, the presence of an analyzer(a polarization filter) in the light path that will decrease the inten-sity of the image about 60% or more. Because the analyzer islocated above the filter cube, it will not decrease the light usedto excite the fluorochromes in the sample; thus imaging with theanalyzer in the path will lead to unnecessary bleaching of the sam-ple. Therefore, it is important to check the status of the analyzerand remove it if it is engaged to prevent damaging the sample.Along with the analyzer, formation of a DIC Nomarski image alsorequires the presence of a modified Wollaston (Nomarski) prismabove the objective.

Depending on the type of DIC employed, each objective canhave its own small DIC prism mounted right above (or below inthe case of inverted microscopes) the objective (Zeiss and NikonFig. 7.22a) or there can be a prism located higher above the


Fig. 7.22. Various locations of DIC prisms on microscopes. a On Nikon and Zeiss micro-scopes, Wollaston prisms are located above each objective (arrows) either in or close totheir back focal plane (Image courtesy of Scott Berger, BIDMC). b Olympus microscopeshave only one slot for the Wollaston prisms (arrow), which is shared among severalobjectives. The height of the prism can be adjusted, depending on the objective used forimaging using a lever located on the DIC slider.

objective in the nosepiece (Olympus and Leica) (Fig. 7.22b).When using bright-field imaging, if the Wollaston prism is leftunintentionally in the path, the splitting effect of the prism in thefinal image will depend on the magnification of the objective (thehigher the magnification the worse the effect) and the type ofcontrast (high contrast prisms have a more significant effect com-pared to low contrast ones). In the case of fluorescent imaging,the effect is significant and becomes very obvious when using highmagnification, high-resolution objectives to image small struc-tures (Fig. 7.23). To record these images, red cell complementreceptor 1, (CD35 or CR1) was indirectly labeled with Alexa 594and imaged either without (left) or with (right) the Wollastonprism engaged using a 60 × 1.42 UPlanApo objective on anOlympus BX 62 microscope fitted with a high-contrast Wollas-ton prism (Fig. 7.23a). In the left image, receptors were split intwo due to the presence of the Wollaston prism in the optical path(false “dimerization” of the receptors). Importantly, the size and

Fig. 7.23. The effect of Wollaston prism during fluorescence imaging. Red cells labeledfor CR1 were imaged without (A) or with (B) a high contrast Wollaston prism in the path.Note the doubling effect of the prism on all fluorescent structures present in the sample.The direction of shear is 11-5.

134 Ghiran

the intensity of each pair were lower than the original and thewhole image shifted with respect to the optical axis of the micro-scope. The “tell-tale” signs of this type of artifact are two-fold: thedoubling effect is found in all structures present in the image andthe splitting of all structures happens in the same direction. Tothe best of the author’s knowledge, only motorized Leica micro-scopes automatically remove the Wollaston prism from the opticalpathway when the imaging method is changed from DIC to another method, because the prism is located on a motorized turretinstead of a slider.

7.1. Condensers Regular sub-stage condensers are not needed for imaging fluo-rescence samples, as the objective plays the role of condenserin epi-fluorescence microscopy (see above). However, even if itdoes not participate directly in image formation, it can certainlyinduce artifacts that can range from increasing the backgroundof the image to duplicating bright fluorescence structures presentin the field of view. Most motorized microscopes are equippedwith swing-out top lens sub-stage condensers that automaticallyswing out the top lens when the fluorescence shutter is open.And microscopes are designed to do so for good reasons. Figure7.24 shows the same image acquired with and without the toplens in the optical path (swung out). Aside from increasing thebackground throughout the image, the top lens also creates anoverlapping image that shows two “ghosts” that are not actuallypresent in the sample but represent the slightly shifted mirroredimage of the actual cells (with only a fraction of the original inten-sity); the brighter the objects in the field of view, the brighter theghosts. Any automatic script (macro) used for post-acquisitionimage analysis could potentially count the ghosts as actual cellswith different fluorescence characteristics. If a microscope is notequipped with a swing out top-lens condenser, simply loweringthe condenser will prevent the top lens of the condenser to act as a

Fig. 7.24. The effect of transmitted light condenser on fluorescence imaging. Imageon the left was acquired with the top lens of the condenser in place and at the heightrequired by the Kohler illumination for that particular objective when used for bright-fieldapplications. The image on the right was recorded with the top lens swung out. Note theincreased background and the apparition of the reflections (arrow) of the actual cells asdimmer counterparts. Both images were recorded using the same exposure time.


mirror for the incoming excitation light. In addition, most micro-scopes are sold with a thin non-reflective black metal plate that ismeant to be placed under the microscope slide when imaging influorescence to prevent the mirroring effect of the condenser toplens.

8. TrinocularHeads

In very few instances and settings, the trinocular head can createproblems, interestingly only when imaging and not when observ-ing samples directly. For that to happen, several factors have tocome together and set the stage for an artifact that it is not atall easy to identify and it is even harder to reproduce. First, thetrinocular head has to have three positions for the splitting prism(100% to eyepieces, 20–80% eyepieces–camera and 100% camera)and not two. Secondly, the room has to be lit during image acqui-sition and thirdly, the lever has to be pulled only half way (to 20–80% setting) while recording the fluorescent image. If all theseconditions are fulfilled then something similar to images showedin Fig. 7.25 can be obtained. The light that enters the eyepiecewill contribute to the final image and form a bright haze that willoverlap the actual fluorescent image (Fig. 7.25a). If a light bulbis present in direct optical path of the microscope (across fromthe microscope on a self or even fluorescent ceiling light), thenits shape will be overlapped with the fluorescence (or whatever is

Fig. 7.25. The role of ambient light and trinocular settings for fluorescence imaging. a Red cells labeled with a lipophiliccell tracker and for CR1 expression were imaged with the trinocular head lever half way out allowing a 20–80 split ofthe optical path (20% of the original intensity of the image to the eyepieces and 80% to the CCD camera). This settingallowed ambient light to reach the detector contaminating the final image, showed on the left. b The image was recordedusing the same exposure time but with fluorescence off. Note the presence of the light bulb image in the center of thefirst two images. The image of the bulb is doubled because the light entered through both eyepieces, as the lamp waslocated on a shelf across from the microscope. c The image on the right was recorded using the same settings as in (a),but with the ambient light off.

136 Ghiran

left of it) image. Turning the light off and using illuminated key-boards and pulling the lever to the very last setting (100% of thelight to the camera) will help one avoid this artifact.

Acknowledgments

I acknowledge Olympus America Inc., Leica MicroSystems, CarlZeiss MicroImaging Inc., and Chroma Technology Corp. forallowing the use of several of their images in this chapter.

References

1. Coons, A. H. (1951) Fluorescent antibod-ies as histochemical tools. Fed Proc 10,558–559.

2. Weller, T. H., Coons, A. H. (1954) Fluores-cent antibody studies with agents of varicellaand herpes zoster propagated in vitro. ProcSoc Exp Biol Med 86, 789–794.

3. Herman, B. (1999) in (Slavik, J., ed.)1998, Fluorescence Microscopy and FluorescentProbes, vol 2. Plenum Press, New York andLondon, p 292.

4. Murphy, D. B. (2001) Fundamentals of LightMicroscopy and Electronic Imaging. Wiley-Liss, New York, NY.

5. Wolf, D. E. (2007) Fundamentals of fluores-cence and fluorescence microscopy. MethodsCell Biol 81, 63–91.

6. Pawley, J. B. (2006) Handbook of Biologi-cal Confocal Microscopy. Springer, New York,NY.

7. Cox, G. (2007) Optical Imaging Techniquesin Cell Biology. CRC/Taylor & Francis, BocaRaton, FL, USA

8. Goldman, R. D., Spector, D. L. (2005)Live Cell Imaging: A Laboratory Manual.Cold Spring Harbor Laboratory Press, ColdSpring Harbor, NY

Chapter 8

Using the Fluorescent Styryl Dye FM1-43 to VisualizeSynaptic Vesicles Exocytosis and Endocytosis in MotorNerve Terminals

Ernani Amaral, Silvia Guatimosim, and Cristina Guatimosim

Abstract

The styryl dye FM1-43 is a powerful tool to track exocytosis, endocytosis and recycling of secretorygranules or vesicles. Due to its unique structure, dye molecules reversibly partition into the outer leafletof surface membrane without permeating due to two cationic charges located in their headgroup. When asecretory cell is stimulated to evoke exocytosis, FM1-43 molecules that were inserted in the membrane areinternalized during compensatory endocytosis and newly formed secretory granules or vesicles becomestained with dye (staining/endocytosis). If stained secretory granules or vesicles undergo exocytosis indye-free medium, due to concentration gradient, FM1-43 molecules dissociate from the membrane andloose fluorescence (destaining/exocytosis). Using a fluorescence microscope attached to a CCD cameraor a confocal, it is possible to follow secretion in live cell or tissue preparations and in this chapter, wewill make a description of FM1-43 staining and destaining protocol using the neuromuscular junctionas experimental model. This technique has allowed answering important questions concerning synapticvesicle recycling, which is a key step for neuronal communication. In addition, FM1-43 has proven to bean excellent tool for investigating membrane internalization and endosome recycling in a variety of celltypes.

Key words: FM1-43, neuromuscular junction, fluorescence microscopy, exocytosis, endocytosis,synaptic vesicles.

1. Introduction

Neurons have an amazing capacity to communicate and this ismediated by the release of chemical messengers (neurotransmit-ters) during the fusion of synaptic vesicles (exocytosis). Opticalmonitoring of neuronal activity is an approach that has yielded


137

138 Amaral, Guatimosim, and Guatimosim

new information about neuronal communication. In the past,the use of microelectrodes was the only approach with both thespatial and the temporal resolution required to investigate thereal-time function of individual neurons or neuronal networks.However, microelectrode recordings have some limitations thatmotivated researchers to seek for alternative approaches. Styryldyes are a subgroup of the cyanine dyes that were first designedas voltage-sensitive probes (1) and were successfully used in thepast to perform optical detection of neuronal activity. The firstoptical detection of neuronal activity from a preparation stainedwith an extrinsic voltage-sensitive dye was reported by Tasaki andcollaborators (1968) (2). Voltage-sensitive probe molecules areused to vitally stain a preparation. These dye molecules bind tothe external surface of excitable membranes and act as molec-ular transducers that transform changes in membrane potentialinto optical signals (reviewed by (2)). For many years, styryl dyeswere employed as optical probes for the rapid changes in mem-brane potential in a wide variety of preparations (3–5). Lichtmanand colleagues (1985) (6) visualized the potential of those dyesto stain multiple innervations from snake adult and embryonicneuromuscular junctions and performed pioneering studies ofactivity-dependent uptake of styryl pyridinium dyes (see also (7)).However, dyes used in their study could not be expanded to non-reptile preparation, which was a major limitation of the technique.This was overcome by William Betz’s group, who introduced anew styryl dye synthesized by Fei Mao, from Molecular Probesat that time, called FM1-43. This dye, which allowed direct visu-alization of synaptic vesicle recycling in preparations extra-vivo,presented a hydrophobic tail that reversibly binds to membranes;a polar dicationic head that prevents membrane permeation anda body containing two aromatic rings and a double bond thatdetermine spectral fluorescence properties ((8, 9); reviewed by(10)). Therefore, FM1-43 binds to synaptic membrane and whenthe nerve terminal is submitted to a stimulus that results in exo-cytosis, and consequently compensatory endocytosis, the fluores-cent dye is incorporated, resulting in a typical pattern of staining(9). Another relevant feature is that FM1-43 is weakly fluores-cent in water, presenting a quantum yield that increases in twoorders of magnitude in lipid environments. So, when a previ-ously labeled terminal is submitted to a new round of stimula-tion, in the absence of FM1-43 in the external medium, the dyeis released to the hydrophilic medium resulting in a decrease offluorescence intensity, which reflects synaptic vesicles exocytosis((8, 9, 11)). The hypothesized mechanism of action for FM1-43was confirmed by ultrastructural evidence that the dye is con-fined to synaptic vesicles and by fluirometric measurements show-ing that it is released from the nerve terminals during stimula-tion (12). Therefore, during the last decades, FM1-43 has been

Use of Styryl Dye FM1-43 to Track Exocytosis and Endocytosis 139

a powerful tool for visualizing exocytosis and endocytosis in avariety of neuronal and non-neuronal cell types, such as endothe-lial, epithelial, pancreatic beta cells, T-lymphocytes, lung pneumo-cytes, adipocytes, parathyroid cells and spermatozoa. Consider-ing that original experiments using FM1-43 to visualized synapticvesicle recycling were performed using nerve-muscle preparationsfrom frog and mice, we will next describe detailed protocols forFM1-43 staining and destaining in both preparations, which werepreviously described (13, 14).

2. Materials

1. Deionized, distilled water.2. FM1-43 R© (1 mg) and FM1-43FX R© (10× 1 μg) purchased

from InvitrogenTM.3. Frog Ringer with the following composition (mM): 115

NaCl, 2.5 KCl, 1.8 CaCl2, 5 Hepes; pH 7.2 (13).4. Mouse Ringer with the following composition (mM):

135 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 12 NaHCO3, 1NaH2PO4 and 11D-glucose. This solution has to be aer-ated with 95%CO2−5%O2 and the pH has to be correctedto 7,4 (15).

5. PBS 1× (g/L): 4 NaCl, 2,34 NaH2PO4 H2O and 10,32Na2HPO4.

6. All salts and buffers used in frog and mouse Ringer or PBScan be purchased from Sigma-Aldrich R©.

7. Sylgard R© Silicone Dielectric Gel can be purchased fromDow Corning Corporation and mounted in plastic Petridishes.

8. High potassium solution (60 mM). Qualitatively, it hasthe same components of frog or mouse Ringer; however,external [Na+] should be reduced to compensate the [K+]increase.

9. An electrical stimulator which fires tetanic pulses and a suc-tion electrode will be necessary in protocols with electricalstimulation.

10. D-tubocurarine chloride (Sigma-Aldrich R©) diluted indeionized water to make stock solutions of 16 mM storedat −20◦C. Aliquots will be dissolved in frog or mouseRinger to a final concentration of 16 μM.

11. Advasep-7 (Sigma-Aldrich R©) diluted in frog or mouseRinger solution to obtain a final concentration of 1 mM.


12. Paraformaldehyde (Sigma-Aldrich R©) solution: prepared 4%in PBS and kept at 4◦C.

13. Glycine (Sigma-Aldrich R©) solution: 1.5 mg/mL of PBS.14. Mounting Media: ProLong Antifade Kit (InvitrogenTM) or

HydramountTM (National Diagnostic, Atlanta, GA)15. Glass slides and coverslips (VWR International, West

Chester, PA)16. Fluorescence microscope equipped with FITC optics.17. Confocal microscope equipped with a 488 nm laser for

specimen excitation and emission at 510 nm.

3. Methods

3.1. Making StockSolutions of FM1-43

1. FM1-43 is a water-soluble probe commercialized aslyophilized material that can be dissolved in either deionizedwater or DMSO (Dimethyl sulfoxide) to make stock solu-tions. For neuromuscular junctions staining, it is typicallyused stock solutions of 4 mM (see Note 1).

2. Aliquots should be stored at −20◦C to preserve the chemicalproperties of the probe and avoid solvent evaporation thatcan occur with stock solutions in deionized water, stored at2–8◦C for prolonged time.

3. During storage or handling, protect aliquots from light.

3.2. MonitoringSynaptic VesicleCycling in FrogNeuromuscularJunctions

3.2.1. Staining FrogNeuromuscularJunctions SynapticVesicles with FM1-43

1. Frog cutaneous pectoris nerve-muscle preparation is an excel-lent model that can be used to stain recycling vesicles withFM1-43. This model provides a very thin muscle associatedto a large number of motor terminals.

2. Dissected muscles should be mounted in a Sylgard R©-linedPetri dish containing the appropriate saline solution (seeNote 2). Entomological pins could be used to fix specimensto the Sylgard R© (see Note 3).

3. Remove the excess of connective tissue over the specimen.The preparation has to be free of debris, unwanted or dam-aged tissue that could retain FM molecules and hinder effi-cient staining.


4. Add working samples of FM1-43 to the Ringer, which willbe used to bath the specimen until a final concentration of 4μM.

5. FM dyes are internalized by recycling vesicles, so a stimulishould be applied to induce exocytosis (vesicle release) andFM uptake during compensatory endocytosis (see Note 4).In frog cutaneous pectoris nerve-muscle preparations, depo-larizing stimuli with KCl (60 mM for 10 min) or electri-cal pulses (20 Hz, 0.5 ms, square wave pulses, 4 V dur-ing 10 min), fired by a suction electrode through the nervefragment, are efficient means to load recycling vesicles withFM1-43 (Fig. 8.1).

6. After stimulation, the dissected muscles should rest in salinesolution with FM1-43 (4 μM) for 15 min to assure completestaining of recycling vesicles (see Note 5).

Fig. 8.1. Illustrative destaining of a frog motor nerve terminal labeled with FM1-43. a Representative frog neuromuscularjunction nerve terminal stained with FM1-43 during electrical stimulation (20 Hz, 10 min). Note the punctuate pattern ofsynaptic vesicles clusters labeled with the fluorescent dye (Scale bar: 10 μm). b–f When submitted to a second roundof stimulation with high potassium solution (KCl 60 mM), the same terminal presented in panel “a” shows destaining assynaptic vesicles are exocytosed. b, c, d, e and f represent images acquired after 1, 3, 5, 7 and 10 min of stimulationwith high potassium, respectively. g Time-course curve of the illustrative destaining presented from “a” to “f” (Sevenfluorescence spots were considered for analysis. Error bars: S.E.M).

3.2.2. Removing theNon-internalizedFM1-43 from FrogNeuromuscular JunctionPreparations

1. The excess of non-internalized FM adhered to the mem-brane of muscle cells or to the myelin of nerves can beremoved during a washing period in saline solution withoutthe probe for at least one hour (see Notes 6 and 7). How-ever, if longer washing is possible (e.g. overnight at 4–8◦C),it will be more effective to reduce the background fluores-cence.


2. Advasep-7 (1 mM), a sulfonated β-cyclodextrin, which has ahigher affinity for FM1-43 than the plasma membrane (16),can be used to remove FM1-43 nonspecifically bound tothe outer leaflet of the plasma membrane or extracellularmolecules, significantly reducing background staining andthe washing time after FM1-43 labeling (see Note 8).

3. After washing, images should be acquired with the appro-priate optical parameters described in the following sections.When fixable form of FM1-43 is used, tissue fixation shouldbe carried after the washing time.

3.2.3. Tissue Fixation 1. Immerse the frog cutaneous pectoris muscle in a 4◦Csolution of paraformaldehyde (4% in PBS, pH: 7.2) for20–40 min.

2. After fixation, wash muscles in a solution of glycine(1.5 mg/mL of PBS) for 15 min in order to quench thefluorescence of paraformaldehyde.

3. Cut insertions of the cutaneous pectoris and remove the skinand sternum fragments used to fix the muscle to the Sylgard.

4. Put the muscle on a slide and then dry the excess ofparaformaldehyde using a piece of paper.

5. Cover the specimen with a mounting medium likehydramount or ProLong Antifade (see Note 9).

6. Put a coverslip over the specimen and wait until the mount-ing medium had dried. The lamina should be maintained at4–8◦C and it should be protected from light exposure beforeimage acquisition.

3.2.4. Image Acquisition 1. Images can be collected in “ex-vivo” preparations using flu-orescence microscopes equipped with CCD cameras andwater immersion objectives (40× or 63×). The excitationlight proceeding from Hg lamps passes through filters toselect the fluorescein or FITC spectra.

2. In confocal microscope, preparations stained with FM1-43 should be excited using a 488 nm laser and the emis-sion spectra should be collected from 510 to 626 nm (seeNote 10).

3. It is essential to maintain parameters of image acquisitionexactly the same when comparing motor terminals in controland test conditions. The ideal image parameters will dependfrom the quality of dissection and optical resources of micro-scopes employed on image acquisition.

4. Frog neuromuscular junctions loaded with FM1-43 willappear like a row of fluorescent spots over the muscle fiberslength.


3.2.5. Destaining ofNon-fixed Preparationsof Frog NeuromuscularJunctions

1. Non-fixed motor nerve terminals loaded with FM1-43 willdestain during a new stimuli round for exocytosis. Highpotassium solution (60 mM) or tetanic pulses (20 Hz) willstimulate synaptic vesicles release and FM1-43 unloadinginto saline solution. During destaining, images acquired willreflect the rate of exocytosis (Fig. 8.1)

2. To prevent muscle contractions during destaining, thepreparation should be preincubated with curare (16 μM)added to the bath during the washing time. Twitches dur-ing destaining can make the experiment impracticable anddetermine the loss of data.

3.2.6. Image Analysis 1. Softwares like Metamorph Imaging System 7.5 and ImageJ allow to draw lines around regions of interest and can beused to measure the brightness levels from each fluorescentspot.

2. The fluorescence signal emitted by clusters of synaptic vesi-cles loaded with FM1-43 represents an estimate of endocy-tosis. The examiner can compare the brightness of motorterminals stained in a control condition with the bright-ness of motor terminals stained in a test condition, whichmight influence the recycling of synaptic vesicle and FM1-43 uptake. The data obtained can be plotted in histogramsusing softwares like Microsoft Excel, Sigma Plot 10.0 orGraph Pad Prisma 5.

3. To monitor exocytosis, the brightness of each spot can bequantified during all destaining or at specific time inter-vals (for example, at every one or five minutes). The dataregistered can be plotted as time-course curves. For FM1-43 destaining experiments, it is important to make controlcurves that indicate the photobleaching levels. The photo-bleaching curves should be used for comparative analysiswith experimental conditions that stimulate or inhibit vesiclerelease.

3.3. MonitoringSynaptic VesicleCycle in MouseNeuromuscularJunctions

3.3.1. Staining MouseNeuromuscularJunctions SynapticVesicles with FM1-43

1. The protocol used to stain/destain mouse neuromuscularjunctions is quite similar to that described for frog neuro-muscular junctions. However, some important aspects haveto be stood out.

2. In preparations of mouse neuromuscular junctions, thediaphragm can be easily dissected associated to a fragment


of phrenic nerve. During dissection, remove any excess ofconnective tissue and avoid accumulation of blood over themuscle (see Note 11).

3. Dissected muscles should be mounted in a Sylgard R©-linedPetri dish containing the appropriate mouse Ringer solution.Entomological pins could be used to fix specimens to theSylgard R© (see Note 12).

4. Add working samples of FM1-43 to the mouse Ringer,which will be used to bath the specimen until a final con-centration of 4 μM (see Note 13).

5. To promote synaptic vesicle recycling and FM1-43 uptake,stimulate the muscle nerve preparation with high K+ solu-tion (60 mM for 10 min) or electrical pulses (20 Hz, 0.5 ms,square wave pulses, 4 V for 10 min) fired by a suction elec-trode through the fragment of phrenic nerve (see Note 14).

6. After stimulation, hemidiaphragms should rest in mouseRinger solution with FM1-43 (4 μM) for 15 min to assurecomplete staining of recycling vesicles (Fig. 8.2).

3.3.2. Removingthe Non-internalizedFM1-43 in Preparationsof MouseNeuromuscularJunctions

1. The excess of non-internalized FM adhered to the mem-brane is removed during a washing period in saline solu-tion without the probe. However, preparations of mouseneuromuscular junctions should be washed between 20 and40 min in mouse Ringer at room temperature. The Ringershould be aerated with carbogenic mixture during the wash-ing time.

2. Advasep-7 (1 mM) can be used to reduce background fluo-rescence (see Note 15).

3. After the washing time, images should be acquired withthe appropriate optical parameters for FM1-43 fluorescence.When fixable form of FM1-43 is used, tissue fixation shouldbe carried after the washing time.

3.3.3. Tissue Fixationand Mounting of Slides

1. Immerse the mouse hemidiaphragms in a 4◦C solution ofparaformaldehyde (4% in PBS; pH: 7.4) for 20–40 min.

2. After fixation, wash muscles in a solution of glycine(1.5 mg/mL of PBS) for 15 min to quench the fluorescencefrom paraformaldehyde.

3. Detach hemidiaphragms from its insertions on the ribs.4. Put the muscle on a slide and then dry the excess of

paraformaldehyde using a piece of paper.5. Cover specimen with mounting medium such as

hydramount or ProLong Antifade as was done for frogneuromuscular junction preparation.


Fig. 8.2. Destaining of a mouse motor nerve terminal labeled with FM1-43. a Illustrative image of mouse neuromuscularjunction stained with FM1-43FX, a fixable analog of FM1-43 (Scale bar: 10 μm). b Mouse neuromuscular junction labeledwith FM1-43 during a high potassium stimulus (60 mM, 10 min – Scale bar: 10 μm). c–e The same terminal presentedin “b” shows significant destaining during a second stimulus with high potassium solution. Images in c, d and e wereacquired after 1, 3 and 5 min of stimulation. f Time-course curve of the destaining observed in images b to e (Sixfluorescence spots were considered for analysis. Error bars: S.E.M).

6. Keep slides at low temperature (2–8◦C) and protect themfrom light exposure.

3.3.4. Destainingof Non-fixed MouseNeuromuscularJunctions

1. Non-fixed mouse neuromuscular junctions labeled withFM1-43 can be destained during a second round of stimula-tion with tetanic pulses (20 Hz) or high potassium solution(KCl 60 mM) (Fig. 8.2).

2. Hemidiaphragms should be preincubated with curare (16μM) to avoid muscle contractions, which can cause distur-bance during image acquisition.

3.3.5. Image Acquisitionand Analysis

1. Images can be collected in “ex-vivo” or fixed preparationsof mouse neuromuscular junctions according to the sameparameters described for frog motor end plates stained withFM1-43.

2. Image analysis can be processed using the same soft-wares employed on image analysis of frog neuromuscularjunctions.


4. Notes

1. Vials containing FM1-43 should be stored at −20◦C andprepared preferentially in deionized water. Dye dilution indeionized water can prevent any interference of DMSO inthe experimental data. It is important to homogenate solu-tions to guarantee the correct concentration of the probein each working aliquot.

2. Frog cutaneous pectoris nerve-muscle preparation can bedissected from animals (e.g. Rana catesbeiana) withapproximately 60 g. The muscle should be maintained infrog Ringer during experiment.

3. When attaching muscles to the Sylgard R©, do not insert pinsover the muscle. Use the skin and sternum fragments asso-ciated to frog cutaneous pectoris muscle to insert pins inorder to avoid muscle damage and increase in backgroundfluorescence.

4. Before applying the stimulus, it is recommended to loosenmuscles in order to prevent damage to the muscular fibersduring tetanic contraction. Any lesion to the muscle couldresult in FM1-43 retention in the sarcolema. Curare (16μM) should be used to avoid harmful contractions of mus-cles.

5. Staining with FM1-43 should be conducted at room tem-perature. Low temperature inhibits endocytosis and com-promises FM1-43 uptake.

6. Since FM1-43 partitioning into membranes is reversible,during the washing time, it will spread out to the bath.These FM molecules free on the extracellular medium emitno significant fluorescence when compared to moleculeslinked to the membrane and they will not interfere withthe fluorescent signal from vesicular pools labeled with thedye. Washing period after staining represents a crucial stepto reduce background fluorescence.

7. To inhibit spontaneous vesicle release and consequent lossof fluorescent signal during the washing time, preparationsshould be maintained at low temperature (4–8◦C).

8. A well-dissected nerve-muscle preparation of frog cuta-neous pectoris stained with FM1-43 and then submitted toa sufficient washing time will not present significant back-ground fluorescence. Therefore, the use of Advasep-7 isunnecessary in this case.

9. ProLong Antifade can be more useful since it provides fluo-rescence stabilization and reduces photobleaching with noquenching of the fluorescence signal.


10. The use of a confocal microscope may increase imagequality since they can eliminate most of the out-of-focuslight and removes background fluorescence. Moreover,confocal microscopes allow optical sectioning and 3Dreconstructions.

11. Mice diaphragms can be dissected since fetal until adult-hood (e.g. Swiss or C57). The muscle associated to thefragment of phrenic nerve should be maintained in aer-ated mouse Ringer (95%CO2–5%O2) during the experi-ment. Blood accumulation over the muscle, excess of con-nective tissue or damaged areas will turn image acquisitioninto a quiet difficult task. The diaphragm can be sectionedin hemidiaphragms in order to offer a control and a testspecimen for paired experiments.

12. When attaching muscles to the Sylgard R©, insert pins on ribsedges associated to mouse diaphragm and in its tendinouscentre.

13. 4 μM of FM1-43 is sufficient to label mouse neuromus-cular junctions. If it is necessary to improve the staining,the final concentration could be adjusted to 8 μM, whichis used to label fetal mouse neuromuscular junctions.

14. Before stimulus application, loosen muscles in order to pre-vent damage to the muscular fibers during tetanic stimula-tion and add curare (16 μM) to avoid any harmful contrac-tion of hemidiaphragms.

15. As mentioned previously, a well-dissected preparation ofneuromuscular junction usually presents low backgroundfluorescence, making the use of Advasep-7 not essential.

Acknowledgments

The authors would like to thank Professor William J. Betz fortransforming FM dyes into a powerful tool that allowed the visu-alization of the synaptic vesicles’ life cycle. This work was sup-ported by CNPq, FAPEMIG and CAPES.

References

1. Grinvald, A., Hildesheim, R., Farber, I.C., Anglister, L. (1982) Improved fluores-cent probes for the measurement of rapidchanges in membrane potential. Biophys J 39,301–308.

2. Tasaki, I., Watanabe, A., Sandlin, R., Carnay,L. (1968) Changes in fluorescence, turbid-ity, and birefringence associated with nerveexcitation. Proc Natl Acad Sci USA 61,883–888.


3. Cohen, L. B., Salzberg, B. M., Grinvald, A.(1978) Optical methods for monitoring neu-ron activity. Annu Rev Neurosci 1, 171–182.

4. Cohen, L. B., Salzberg, B. M. (1978) Opti-cal measurement of membrane potential. RevPhysiol Biochem Pharmacol 83, 35–88.

5. Waggoner, A. S. (1979) The use of cya-nine dyes for the determination of membranepotentials in cells, organelles, and vesicles.Methods Enzymol 55, 689–695.

6. Lichtman, J. W., Wilkinson, R. S., Rich,M. M. (1985) Multiple innervation of tonicendplates revealed by activity-dependentuptake of fluorescent probes. Nature 314,357–359.

7. Lichtman, J. W., Wilkinson, R. S. (1987)Properties of motor units in the transversusabdominis muscle of the garter snake. J Phys-iol 393, 355–374.

8. Betz, W. J., Bewick, G. S. (1992) Opticalanalysis of synaptic vesicle recycling at thefrog neuromuscular junction. Science 255,200–203.

9. Betz, W. J., Mao, F., Bewick, G. S. (1992)Activity-dependent fluorescent staining anddestaining of living vertebrate motor nerveterminals. J Neurosci 12, 363–375.

10. Betz, W. J., Mao, F., Smith, C. B. (1996)Imaging exocytosis and endocytosis. CurrOpin Neurobiol 6, 365–371.

11. Gaffield, M. A., Betz, W. J. (2006) Imag-ing synaptic vesicle exocytosis and endo-cytosis with FM dyes. Nat Protoc 1,2916–2921.

12. Henkel, A. W., Lübke, J., Betz, W. J.(1996) FM1-43 dye ultrastructural localiza-tion in and release from frog motor nerveterminals. Proc Natl Acad Sci USA 93,1918–1923.

13. Guatimosim, C., Romano-Silva, M. A.,Gomez, M. V., Prado, M. A. (1998) Recy-cling of synaptic vesicles at the frog neuro-muscular junction in the presence of stron-tium. J Neurochem 70, 2477–2483.

14. Prado, V. F., Martins-Silva, C., de Castro, B.M., et al. (2006) Mice deficient for the vesic-ular acetylcholine transporter are myasthenicand have deficits in object and social recogni-tion. Neuron 51, 601–612.

15. Xu, Y. F., Atchison, W. D. (1996) Effects ofomega-agatoxin-IVA and omega-conotoxin-MVIIC on perineurial Ca++ and Ca(++)-activated K+ currents of mouse motornerve terminals. J Pharmacol Exp Ther 279,1229–1236.

16. Kay, A. R., Alfonso, A., Alford, S.,et al. (1999) Imaging synaptic activityin intact brain and slices with FM1-43inC elegans, lamprey, and rat. Neuron 24,809–817.

Chapter 9

Imaging Lipid Bodies Within Leukocytes with Different LightMicroscopy Techniques

Rossana C.N. Melo, Heloisa D’Ávila, Patricia T. Bozza,and Peter F. Weller

Abstract

Lipid bodies, also known as lipid droplets, are present in most eukaryotic cells. In leukocytes, lipid bodiesare functionally active organelles with central roles in inflammation and are considered structural markersof inflammatory cells in a range of diseases. The identification of lipid bodies has methodological limita-tions because lipid bodies dissipate upon drying or dissolve upon fixation and staining with alcohol-basedreagents. Here we discuss several techniques to detect and visualize lipid bodies within leukocytes by lightmicroscopy. These techniques include staining with osmium or use of different fluorescent probes suchas Nile red, BODIPY, Oil red, P96 and immunofluorescence labeling for adipose differentiation-relatedprotein (ADRP).

Key words: Lipid bodies, lipid droplets, leukocytes, bright field and fluorescence microscopy,osmium staining, nile red, oil red O, BODIPY, 1-pyrenedodecanoic acid, adipose differentiation-related protein (ADRP).

1. Introduction

Lipid bodies, also named lipid droplets or adiposomes, arenow recognized as key organelles involved in lipid storage andmetabolism, cell signaling and inflammation (1, 2). Lipid bodiesare lipid-rich organelles distributed in the cytoplasm as roughlyspherical organelles lacking a delimiting classical bilayer mem-brane (3–6), but surrounded by an outer monolayer of phos-pholipids, which at least in some cells may have a unique fattyacid composition (4, 7). The internal core of lipid bodies is rich


149

150 Melo et al.

in neutral lipids; and it is likely that more complex membranousdomains, often obscured by overlying neutral lipids, are presentwithin lipid bodies (8, 9). Indeed, studies of lipid bodies are pro-viding functional, ultrastructural and protein compositional evi-dences that lipid bodies are not inert depots of neutral lipid.Rather, it has become evident that lipid bodies are highly reg-ulated, dynamic and functional organelles. Over the past yearssubstantial progresses have been made to demonstrate that lipidbody biogenesis is a highly regulated process, which culminatesin the compartmentalization of a specific set of proteins and lipids(reviewed in (1, 2)).

Lipid body accumulation within cells is observed in both clin-ical and experimental metabolic, infectious, neoplasic and otherinflammatory conditions. Because lipid bodies can be destroyedby drying or fixation and staining with alcohol-based reagents,there are consequently some methodological limitations to theirstudy (2, 10). Indeed, routinely used hematological staining asMay–Grunwald–Giemsa staining lead to dissolution of lipid bod-ies commonly precluding their identification. However, usingappropriate fixation procedures followed by methods of identi-fication of lipids and/or of lipid body-specific proteins, lipid bod-ies can be readily identified within cells. In this chapter, we detaildifferent techniques to visualize lipid bodies in different cell sus-pensions such as leukocytes isolated from the blood, cell lineagesand peritoneal, pleural or bronchoalveolar cells.

2. Materials

2.1. Osmium Staining 1. Sodium cacodylate (cacodylic acid – sodium salt) is dissolved(4.28 g) in 180 mL of distilled water. Adjust pH to 7.4 withHCl and then make up to 200 mL with distilled water for0.1 M final concentration.

2. Osmium tetroxide (see Note 1): to prepare a stock solution(1.5%), dissolve 1.5 g of osmium tetroxide in 100 mL of0.1 M sodium cacodylate buffer. Aliquot in small glass tubes(∼2 mL per tube) and store at 4–8◦C. Protect from light.

3. Paraformaldehyde or formaldehyde solution (formalin)(see Note 2). For paraformaldehyde preparation, diluteparaformaldehyde to 2% in Hanks-buffered salt solu-tion without calcium chloride and magnesium chloride(HBSS−/−) or phosphate-buffered saline (PBS). Dilutionsshould be made in fume hood and fresh dilutions ofparaformaldehyde should be used in each experiment. Pro-tect from light. For formalin preparation, dilute formalin(saturated solution of formaldehyde 37%) to 3.7% in PBSor HBSS−/−. Adjust to pH 7.4.

Imaging Lipid Bodies Within Leukocytes 151

4. Thiocarbohydrazide solution 1% should be freshly preparedby dissolving 50 mg of thiocarbohydrazide in 5 mL of dis-tilled water. The solution should be heated at a hot plate ormicrowave (30 s), followed by cooling to room temperaturejust prior to use.

5. Aqueous mounting medium.6. Liquid blocker pen.7. Glass microscope slides and coverslips.

2.2. Nile Red 1. Nile red: 9-diethylamino-5H-benzo [α] phenoxazine-5-oneis a phenoxazone dye 1-Acho que é phenoxazine dye, 2-incluir espaço entre as palavras poorly soluble in water but itdoes dissolve in a wide variety of organic solvents (11). Stocksolution: dissolve Nile red in acetone (1 mg/mL). Aliquotin small test tubes and store at −20◦C. Working solution(prepare fresh): dilute at 1:10,000 in HBSS−/− or PBS fromthe stock solution. Keep protected from light (see Note 3).

2. Paraformaldehyde or formaldehyde solution (see Note 2).Refer to Section 2.1 (Item 3) for fixative preparation.

3. Anti-fading mounting medium for fluorescence microscopy.4. Glass microscope slides and coverslips.

2.3. Oil Red O 1. Oil Red O: 1-([4-(Xylylazo)xylyl]azo)-2-naphthol, MW408.49 is prepared at 0.5%: add 5 mL of propylene glycol(100%) to 0.5 g of oil red O with stirring and gradually com-plete the volume with propylene glycol to 100 mL. Heat thesolution until 95◦C, but do not allow going over 100◦C. Fil-ter through paper filter. The solution can be stored at roomtemperature.

2. Hematoxylin solution.3. Aqueous mounting medium.4. Paper filter.

2.4. BODIPY 1. BODIPY R© 493/503: 4,4-difluoro-1,3,5,7,8-pentamethyl-4-bora-3a,4a-diaza-s-indacene, (molecular weight: 262;Molecular Probes, cat no. D-3922) is stored at−20◦C,protected from light (see Note 3). Stock solution: dis-solve BODIPY in dimethyl sulfoxide (DMSO) at 1 mM.Aliquot in small test tubes (∼10 μL per tube) and store at−20◦C. Working solution (prepare fresh): dilute 1000× inHBSS−/−. All solutions must be protected from light (seeNote 3).

2. Paraformaldehyde or formaldehyde solution (see Note 2).Refer to Section 2.1 (Item 3) for fixative preparation.

3. Anti-fading mounting medium for fluorescence microscopy.4. Glass microscope slides and coverslips.

152 Melo et al.

2.5.1-PyrenedodecanoicAcid

1. 1-pyrenedodecanoic acid (molecular weight 400.56.2;Molecular Probes, cat. no. P-96). It is poorly soluble inwater but it does dissolve in a variety of organic solventsincluding DMSO. Stock solution: dissolve P96 in DMSO(10 mM). Aliquot in small test tubes (∼10 μL per tube)and store at −20◦C, protected from light. Working solu-tion (prepare fresh): dilute 1000× in HBSS−/− or Hanks-buffered salt solution with calcium chloride and magnesiumchloride (HBSS−/−) (see Note 4). Keep protected from light(see Note 3).

2. Formaldehyde solution (formalin) prepared as Section 2.1(Item 3).

3. Glass microscope slides and coverslips.4. Anti-fading mounting medium for fluorescence microscopy.

2.6. AdiposeDifferentiation-Related Protein(ADRP, Adipophilin)

1. Monoclonal or polyclonal antibody to ADRP.2. Fluorescent-labeled secondary antibodies.3. Formaldehyde solution (formalin) prepared as Section 2.1

(Item 3).4. Triton R© X-100 (t-Octylphenoxypoly-ethoxyethanol).5. Glass microscope slides and coverslips.6. Anti-fading mounting medium for fluorescence.

3. Methods

3.1. SamplePreparation ontoSlides

After obtaining a cell suspension (∼0.5–1.0 × 106 cells/mL ofmedium), preparation of samples (cell suspensions) onto slidesfor lipid body staining can be done in two ways: using a cyto-centrifuge or by spreading a mixture of cells with melted agarosematrix onto a slide. For comparison of cell morphology observedwith these two techniques, refer to Note 5. For cytospin prepa-rations: label slides and cytocentrifuge a volume of 100 μL(∼0.5–1.0 × 105 cells) of a cell suspension sample, at 18–23 gfor 5 min. For agarose preparations: prepare first an agarosematrix. Weigh 0.125 g of agarose (low-melting point agarose; mp65.5◦C, gelling point 24◦C) into a 125-mL Erlenmeyer flask anddilute to 2.5% by adding 5 mL of distilled water. Cover with alu-minum foil. Mix well, but avoid swirling to prevent agarose bind-ing to flask wall. Solubilize agarose in 70◦C water bath for 15 minwith gentle agitation. Aliquot in small test tubes (∼1.5 mL) andstore at 4◦C. For slide preparation, resolubilize a tube at 70◦C,


mix cells with the liquid matrix and gently spread the cell–agarosemixture (∼20 μL) onto microscope slides using a microtip. Usecovered surface slides for cell adhesion. For spreading the cells,use surface tension to move agarose mixture throughout the slide.Cover gently this thin layer of agarose/cell mixture with a perfu-sion chamber (CoverWellTM).

3.2. Osmium Staining Properties: Osmium tetroxide binds to unsaturated lipids, and isreduced by organic materials to elemental osmium, an easily vis-ible and permanent black substance (Fig. 9.1a). Reduction ofosmium by thiocarbohydrazide highly enhances lipid labeling.

1. Prepare slides with samples using a cytocentrifuge (seeSection 3.1).

2. Fix samples, while still moist, with paraformaldehyde orformalin for 10 min. Refer to Note 2 for cell fixation andNote 6 for “moist cells”.

3. Rinse slides in distilled water.4. Circumscribe the adhered cells with liquid blocker pen to

facilitate the staining procedure.5. Stain the adhered cells by adding one drop of 0.1 M

cacodylate buffer and one drop of 1.5% osmium tetroxidefor 30 min. Refer to Note 1 for osmium manipulation.

6. Rinse slides in distilled water.7. Immerse in thiocarbohydrazide solution for 5 min at room

temperature.8. Rinse the adhered cells twice with distilled water.9. Re-stain by adding one drop of 0.1 M cacodylate buffer and

one drop of 1.5% osmium tetroxide for 3 min.10. Rinse slides in distilled water.11. Let the slides dry.12. Mount with aqueous mounting medium.

Alternatively, osmium staining can be performed on agarosepreparations.

1. Spread a mixture of cells/agarose onto a microscope slideand cover with a perfusion chamber (see Section 3.1).

2. Carefully pipet 400 μL of the fixative (2% paraformalde-hyde) over sample through the chamber access port, ensur-ing that chamber area is uniformly saturated and let for10 min.

3. Rinse slides in distilled water through the chamber accessport.

4. Rinse slides in 0.1 M cacodylate buffer through the cham-ber access port.

154 Melo et al.

Fig. 9.1. Lipid bodies within leukocytes imaged by light microscopy after staining withosmium (a), Nile red (b), BODIPY (c), oil red O (ORO) (d, e) or double labeled with1-pyrenedodecanoic acid (P 96) and anti-adipose differentiation-related protein (ADRP)(f). Lipid bodies appear as round, dark (a, arrowheads), fluorescent red (b, e) or green(c) organelles distributed throughout the cytoplasm. ORO-stained lipid bodies appear asround red organelles at both bright field (d) and fluorescence (e) microscopy, while thenucleus is imaged in light blue after counterstaining with hematoxylin (d). In f, mergedimages show P 96-labeled lipid bodies in blue at UV filter and ADRP immunolabelingas a ring in the periphery of the lipid body. In a, cells were counterstained with Hema3 R© kit (Fisher Scientific). a and b show eosinophils isolated from the blood of normalhuman donors and stimulated with eotaxin (a) or calcium ionophore (b) as before (20,21). c and d show murine peritoneal leukocytes and macrophages, respectively. Bars,6 μm (A, B, D, E and F), 10 μm (c).


5. Stain by adding 1.5% osmium tetroxide through the cham-ber access port for 30 min. Refer to Note 1 for osmiummanipulation.

6. Rinse slides in 0.1 M cacodylate buffer through the cham-ber access port.

7. Carefully remove the chamber.8. Immerse in thiocarbohydrazide for 5 min at room

temperature.9. Rinse slides in 0.1 M cacodylate buffer.

10. Re-stain with 1.5% osmium tetroxide for 3 min.11. Let the slides dry.12. Mount with aqueous mounting medium.

3.3. Nile Red Staining Properties: Nile red is intensely fluorescent, and can serve as a sen-sitive stain for the detection of cytoplasmic lipid bodies (11, 12).

1. Incubate a cell suspension (1.0 × 106 cells/mL) with aworking solution of Nile red (see Section 2.2, Item 1) for5 min at room temperature and protected from light. Cellsare incubated in a test tube.

2. Centrifuge (120 g/5 min) and resuspend in HBSS−/− orPBS to wash cells. Repeat this step once.

3. Cytospin onto slides using 100 μL of cell suspension at18–23 g for 5 min.

4. Fix with paraformaldehyde or formalin (see Note 2 and Sec-tion 2.1, Item 3, for fixative preparation) for 5 min at roomtemperature.

5. Wash twice in HBSS−/− or PBS.6. Mount while wet using anti-fading mounting medium. Keep

slides in the dark (see Notes 3 and 7).Alternatively, Nile Red staining can be done on agarose

preparations.1. Spread a mixture of cells/agarose onto a microscope slide

and cover with a perfusion chamber (see Section 3.1).2. Carefully pipet 400 μL of the fixative (2% paraformalde-

hyde) over sample through the chamber access port, ensur-ing that chamber area is uniformly saturated and let for5 min. This step can be performed after incubation with NileRed. Refer to Note 8 for Nile red staining in fixed/unfixedcells.

3. Wash twice with HBSS−/− (2× 400 μL) adding bufferthrough the chamber access port.

156 Melo et al.

4. Incubate with a working solution of Nile Red (400 μL) for5–10 min at room temperature. Keep slides protected fromlight (see Note 3).

5. Wash twice with HBSS−/−.6. Carefully remove the chamber.7. Mount while moist with HBSS−/− or with anti-fading

mounting medium after drying. Keep slides in the dark (seeNote 3).

3.4. Oil Red OStaining

Properties: Oil red O belongs to the polyazo group of dyes whichalso includes the Sudan series of dyes. The principle of the lipidstaining is based on the physical properties of the dye that pref-erentially divide into lipid-rich compartments. Oil red O stain-ing can be readily visualized in both bright field and fluorescentmicroscopy (Fig. 9.1d, f) (13).

1. For slide preparation, cytocentrifuge 100 μL of a samplecell suspension, at 18–23 g for 5 min.

2. Fix cells with 3.7% formalin in HBSS−/− (see Section 2.1,Item 3).

3. Wash twice in distilled water.4. Place slides in absolute propylene glycol for 5 min.5. Stain in 0.5% oil red O solution (see Section 2.4) for

10 min in the incubator at 60◦C.6. Rinse cells in 85% propylene glycol solution for 5 min.7. Wash twice in distilled water.8. Counterstain with hematoxylin solution for 30 s

(see Note 9).9. Wash thoroughly in tap water.

10. Mount with aqueous mounting medium.

3.5. BODIPY Staining Properties: BODIPY R© lipid probe is an effective dye for stain-ing neutral lipids and, for this reason it is very efficient for lipidbody staining (Fig. 9.1c). The fluorescence quantum yield of theBODIPY dyes is not diminished in water and this method can beused in conjugation with immunofluorescence (see Note 10 andChapter 11).

1. Incubate the cell suspension with 1 μM BODIPY for 1 h at37◦C. Cells are incubated in a test tube inside a water bath.

2. Pellet the cells (120 g/5 min) and resuspend in HBSS−/−or PBS to wash cells. Repeat this step once.



4. Fix with paraformaldehyde or formalin in HBSS−/− orPBS for 5–10 min (see Section 2.1, Item 3, for fixativepreparation).

5. Wash twice in HBSS−/− or PBS.6. Mount while wet using anti-fading mounting medium.

Slides are stored at room temperature in the dark until anal-ysis (see Notes 3 and 7).

Alternatively, the BODIPY staining can be done using anagarose preparation.

1. Spread a mixture of cells/agarose onto a microscope slideand cover with a perfusion chamber (see Section 3.1).


3. Carefully pipet 400 μL of BODIPY solution over sam-ple through the chamber access port, ensuring that cham-ber area is uniformly saturated. Place slides on a tray atophydrated pad. Place tray in humidified incubator (37◦C, 5%CO2) for 1 h.


5. Pipet 400 μL of the fixative (2% paraformaldehyde) oversample through the chamber access port, ensuring thatchamber area is uniformly saturated and let for 5–10 min.

6. Wash twice in HBSS−/− or PBS.7. Carefully remove the chamber.8. Mount while wet with HBSS−/− or anti-fading mounting

medium. Keep slides protected from light (see Note 3).

3.6.1-PyrenedodecanoicAcid Staining

Properties: 1-Pyrenedodecanoic acid (P-96) is a fluorescent fattyacid analog with the environment sensitive pyrene attached tothe terminal carbon atom that is furthest from the carboxylatemoiety. P96 is readily incorporated into lipid bodies and P96fluorescent-labeled lipid bodies are visualized under the UV (exci-tation/emission 340/376 nm) (Fig. 9.1e) (14, 15).

1. Incubate a cell suspension with a working solution of P96(see Section 2.5, Item 1). for 1 h at room temperature andprotected from light (see Note 3).

2. Pellet the cells (120 g/5 min) and resuspend in HBSS−/−or PBS to wash cells. Repeat this step once.


4. Fix with 3.7% formalin in HBSS−/− or PBS for 5 minat room temperature (see Section 2.1, Item 3, for fixativepreparation).

158 Melo et al.

5. Wash twice in distilled water.6. Mount while wet using anti-fading medium for fluorescence

microscopy (see Note 7).

3.7. ADRP Staining Properties: Adipose differentiation-related protein (ADRP) is astructural protein of lipid bodies considered essential for lipidstorage and metabolism (reviewed in (16)). ADRP is ubiqui-tously associated with cytoplasmic lipid bodies in different typesof cells and is described as a specific protein marker for lipid bodies(Fig. 9.1f) (17–19). This method can be conjugated withimmuno-labeling for other proteins (see Note 11).

1. For slide preparation, cytocentrifuge 100 μL of a samplecell suspension, at 18–23 g for 5 min.

2. Fix slides with 3.7% formalin in HBSS−/− or PBS (see Sec-tion 2.1, Item 3, for fixative preparation).

3. Wash once in HBSS−/− or PBS.4. Permeabilize the cells with 0.1% Triton X-100 in HBSS−/−

for 10 min.5. Circumscribe the adhered cells with liquid blocker pen.6. a. For human cells incubate with mouse anti-human ADRP

at dilution of 1:20 (2.5 μg/mL, final concentration) for 1 hat room temperature.b. For mouse, rat, human or bovine cells incubate withguinea pig anti-human ADRP polyclonal antibody at dilu-tion of 1:300 (final dilution) for 1 h at room temperature.

7. Wash three times in HBSS−/− or PBS.8. Incubate with fluorescent-labeled secondary antibody for

1 h room temperature.9. Wash three times in HBSS−/− or PBS.

10. Mount in mounting medium for fluorescence microscopy(see Note 7).

3.8. Lipid BodyAnalysis andQuantification

Lipid body analysis is performed on a bright field (osmium and oilred O staining) or fluorescence microscope (oil red O and fluores-cent probes) at 1000×. For example, analyses and image acqui-sition can be obtained using an Olympus BX-FLA fluorescencemicroscope equipped with a Plan Apo 100 × 1.4 Ph3 objective(Olympus) and CoolSNAP-Pro CF digital camera in conjunctionwith Image Pro Plus R© software (Media Cybernetics). The detec-tion of lipid bodies using different techniques will appear as rounddark (osmium staining, Fig. 9.1a) or fluorescent red (Nile red orOil red O) (Fig. 9.1b and e, respectively) or green (BODIPY)(Fig. 9.1c) organelles. Of note, Nile red can be observed throughboth green (fluorescein) and red (rhodamine) channels and Oilred O can be also observed at bright field microscopy (Fig. 9.1d).


ADRP immuno-labeling has a characteristic ring-shape appear-ance as this protein localizes preferentially at the periphery of thelipid body (Fig. 9.1f). Lipid bodies are usually enumerated usinga 100× objective lens in 50 consecutively scanned cells (10).

Alternatively, lipid bodies can be quantified by the measure-ment of oil red O (ORO) or BODIPY fluorescent area. The mea-surement of the area of lipid bodies is obtained with a 60 objective(at least four fields per slide). Images are transformed into blackand white pictures and analyzed with the Image 2D software (GEHealthcare). Spots are determined by automatic spot detectionand the total area of fluorescent lipid bodies is obtained for eachfield and divided by the number of cells in the respective field.

4. Notes

1. Osmium tetroxide is volatile and its fumes are very toxic(causes severe irritation to eyes, skin and respiratory tract).Thus, any manipulation involving this chemical must beperformed in a fume hood and wearing gloves.

2. Fixation of cells before osmium, Nile Red or BODIPYstaining can be done using either 2% paraformaldehyde or3.7% formalin.

3. When staining with fluorescent probes such as Nile red,P 96 and BODIPY, fluorescence is usually not stable fora long period and fluorescence bleaching will occur after acertain time. Keep the cell preparations in the dark to avoidfluorescence loss.

4. For some type of cells, the P96 staining can have betterresults using HBSS+/+.

5. In general, cells kept in agarose show better morphologycompared with cells from cytospin preparations becausecells are kept in a hydrated system. In addition, shapechanges, a feature of activated leukocytes, can be observedwhen cells are embedded in an agarose matrix. On theother hand, cytospin slides are fast prepared. In bothcytospin and agarose preparations, lipid bodies are well pre-served and can be easily detected.

6. It is very important to keep cells moist during the stain-ing with osmium. Dried cells will appear with very badmorphology.

7. For fluorescence microscopy, it is important to use amounting medium that prevents rapid loss of fluorescence

160 Melo et al.

during microscopic examination and retains its anti-fadingability during long-term storage.

8. Staining with Nile red can be carried out on either fixed orunfixed cells with no apparent difference in distribution orintensity of fluorescence.

9. Counterstaining with hematoxylin is important to shownuclear aspects. This step is not mandatory if visualizationof nuclei is not necessary.

10. BODIPY staining can be combined with an immunofluo-rescence protocol (for example, ADRP immuno-labeling).In this case, BODIPY can be mixed with the secondaryantibody (Chapter 10).

11. Immuno-labeling for ADRP can be conjugated withimmuno-labeling for other proteins by simultaneous incu-bation with two primary antibodies (raised in distincthosts) followed by incubation with two secondary antibod-ies (with distinct ranges of excitation/emission).

Acknowledgments

Supported by CNPq and FAPEMIG (Brazil) to RCNM; CNPq,FAPERJ and PRONEX (Brazil) to PTB and NIH grants (USA)AI020241, AI051645 and AI022571 to PFW. We acknowledgeClarissa M. Maya-Monteiro for the Fig. 9.1c used in this chapter.

References

1. Martin, S., Parton, R. G. (2006) Lipiddroplets: a unified view of a dynamicorganelle. Nat Rev Mol Cell Biol 7, 373–378.

2. Bozza, P. T., Melo, R. C. N., Bandeira-Melo, C. (2007) Leukocyte lipid bodies reg-ulation and function: contribution to allergyand host defense. Pharmacol Ther 113,30–49.

3. Dvorak, A. M., Dvorak, H. F., Peters, S.P., Shulman, E. S., MacGlashan, D. W., Jr.,Pyne, K., Harvey, V. S., Galli, S. J., Licht-enstein, L. M. (1983) Lipid bodies: cyto-plasmic organelles important to arachidonatemetabolism in macrophages and mast cells. JImmunol 131, 2965–2976.

4. Tauchi-Sato, K., Ozeki, S., Houjou, T.,Taguchi, R., Fujimoto, T. (2002) The surfaceof lipid droplets is a phospholipid monolayer

with a unique fatty acid composition. J BiolChem 277, 44507–44512.

5. Murphy, D. J. (2001) The biogenesis andfunctions of lipid bodies in animals, plantsand microorganisms. Prog Lipid Res 40,325–438.

6. Weller, P. F., Monahan-Earley, R. A., Dvo-rak, H. F., Dvorak, A. M. (1991) Cyto-plasmic lipid bodies of human eosinophils.Subcellular isolation and analysis of arachi-donate incorporation. Am J Pathol 138,141–148.

7. Bartz, R., Li, W. H., Venables, B., Zehmer, J.K., Roth, M. R., Welti, R., Anderson, R. G.,Liu, P., Chapman, K. D. (2007) Lipidomicsreveals that adiposomes store ether lipids andmediate phospholipid traffic. J Lipid Res 48,837–847.


8. Wan, H. C., Melo, R. C., Jin, Z., Dvo-rak, A. M., Weller, P. F. (2007) Roles andorigins of leukocyte lipid bodies: proteomicand ultrastructural studies. FASEB J 21,167–178.

9. Bozza, P. T., Magalhaes, K., Weller, P. F.(2009) Leukocyte lipid bodies–biogenesisand functions in inflammation. BiochimBiophys Acta doi:10.1016/j.bbalip_2009_01_005.

10. Melo, R. C. N., Sabban, A., Weller,P. F. (2006) Leukocyte lipid bod-ies: inflammation-related organelles arerapidly detected by wet scanning electronmicroscopy. J Lipid Res 47, 2589–2594.

11. Greenspan, P., Mayer, E. P., Fowler, S. D.(1985) Nile red: a selective fluorescent stainfor intracellular lipid droplets. J Cell Biol100, 965–973.

12. Fukumoto, S., Fujimoto, T. (2002) Defor-mation of lipid droplets in fixed samples. His-tochem Cell Biol 118, 423–428.

13. Koopman, R., Schaart, G., Hesselink, M. K.(2001) Optimisation of oil red O stainingpermits combination with immunofluores-cence and automated quantification of lipids.Histochem Cell Biol 116, 63–68.

14. Radom, J., Salvayre, R., Maret, A., Negre,A., Douste-Blazy, L. (1987) Metabolismof 1-pyrenedecanoic acid and accumula-tion of neutral fluorescent lipids in cul-tured fibroblasts of multisystemic lipid stor-age myopathy. Biochim Biophys Acta 920,131–139.

15. Yu, W., Bozza, P. T., Tzizik, D. M., Gray,J. P., Cassara, J., Dvorak, A. M., Weller, P.F. (1998) Co-compartmentalization of MAP

kinases and cytosolic phospholipase A2 atcytoplasmic arachidonate-rich lipid bodies.Am J Pathol 152, 759–769.

16. Brasaemle, D. L. (2007) Thematic reviewseries: adipocyte biology. The perilipin familyof structural lipid droplet proteins: Stabiliza-tion of lipid droplets and control of lipolysis.J Lipid Res 48, 2547–2559.

17. Brasaemle, D. L., Barber, T., Wolins, N.E., Serrero, G., Blanchette-Mackie, E. J.,Londos, C. (1997) Adipose differentiation-related protein is an ubiquitously expressedlipid storage droplet-associated protein. JLipid Res 38, 2249–2263.

18. Heid, H. W., Moll, R., Schwetlick, I.,Rackwitz, H. R., Keenan, T. W. (1998)Adipophilin is a specific marker of lipid accu-mulation in diverse cell types and diseases.Cell Tissue Res 294, 309–321.

19. D’Avila, H., Melo, R. C. N., Parreira, G. G.,Werneck-Barroso, E., Castro-Faria-Neto, H.C., Bozza, P. T. (2006) Mycobacterium bovisbacillus Calmette-Guerin induces TLR2-mediated formation of lipid bodies: intracel-lular domains for eicosanoid synthesis in vivo.J Immunol 176, 3087–3097.

20. Melo, R. C. N., Perez, S. A. C., Spencer,L. A., Dvorak, A. M., Weller, P. F. (2005)Intragranular vesiculotubular compartmentsare involved in piecemeal degranulationby activated human eosinophils. Traffic 6,866–879.

21. Bandeira-Melo, C., Perez, S. A. C., Melo, R.C. N., Ghiran, I., Weller, P. F. (2003) EliCellassay for the detection of released cytokinesfrom eosinophils. J Immunol Methods 276,227–237.

Chapter 10

EicosaCell – An Immunofluorescent-Based Assay to LocalizeNewly Synthesized Eicosanoid Lipid Mediators atIntracellular Sites

Christianne Bandeira-Melo, Peter F. Weller, and Patricia T. Bozza

Abstract

Eicosanoids (prostaglandins, leukotrienes and lipoxins) are a family of signaling lipids derived from arachi-donic acid that have important roles in physiological and pathological processes. Over the past years,it has been established that successful eicosanoid production is not merely determined by arachidonicacid and eicosanoid-forming enzymes availability, but requires sequential interactions between specificbiosynthetic proteins acting in cascade and may involve very unique spatial interactions. Direct assess-ment of specific subcellular locales of eicosanoid synthesis has been elusive, as those lipid mediators arenewly formed, not stored and often rapidly released upon cell stimulation. In this chapter, we discuss theEicosaCell protocol for intracellular detection of eicosanoid-synthesizing compartments by means of astrategy to covalently cross-link and immobilize the lipid mediators at their sites of synthesis followed byimmunofluorescent-based localization of the targeted eicosanoid.

Key words: Eicosanoids, prostaglandin, leukotriene, biosynthesis, compartmentalization,carbodiimide, EDAC (1-ethyl-3-(3-dimethylamino-propyl) carbodiimide), lipid droplets, phago-somes, perinuclear.

1. Introduction

Eicosanoids – including leukotrienes and prostaglandins – are afamily of signaling lipids derived from the enzymatic oxygena-tion of arachidonic acid (AA) that control key processes involvingcell–cell communication, including cell activation, proliferation,apoptosis, metabolism and migration (1, 2). Thus, eicosanoids


163

164 Bandeira-Melo, Weller, and Bozza

have important roles in physiological and pathological conditionssuch as tissue homeostasis, host defense, inflammation and can-cer. In view of the magnitude of eicosanoid actions, great effortshave been aimed at understanding the biochemical, cellular andmolecular aspects of their biosynthetic pathway.

In all cells, the highly regulated generation of eicosanoids isdependent on activation of specific phospholipases and specificeicosanoid-synthesizing enzymes and involves small molecules(e.g. Ca2+) and activation-dependent localization of enzymesat specific compartments within cells (3–8). Intracellular com-partmentalization of eicosanoid synthesis within leukocytes hasemerged as a key feature that regulates the amount and mayalso regulate the eicosanoid produced. Such intracellular sites ofeicosanoid formation in any cell have been inferred based onthe permanent or temporary localization of specific eicosanoid-forming enzymes under proper cell activation, since the directobservation of sites of eicosanoid synthesis has been hard to defineas those lipid mediators are newly formed, non-storable and oftenrapidly released upon cell stimulation. It was recently establishedthat successful eicosanoid production is not merely determinedby AA and eicosanoid-forming enzymes availability, but requiressequential interactions between specific biosynthetic proteins act-ing in cascade, and may involve very unique spatial interactions.Therefore, just by detecting eicosanoid-forming enzymes withindiscrete subcellular structures, one cannot assure that those sitesare indeed accountable for the efficient and enhanced eicosanoidsynthesis observed during inflammatory responses. The immuno-localization of eicosanoid-forming proteins does not necessarilyascertain that the localized protein is functional and activatedto synthesize a specific eicosanoid lipid at an intracellular site.We previously developed a method to capture and localize theeicosanoid, prostaglandin E2 (PG E2), released extracellularly by anematode parasite (9). By means of a strategy to covalently cross-link, capture and localize newly formed eicosanoids at their sitesof synthesis, we developed a more direct approach to detect theintracellular sites of arachidonic acid (AA)-derived lipid mediatorformation in leukocytes and other cell types.

To develop our new strategy for in situ immuno-localizationof newly formed eicosanoids to ascertain the intracellular com-partmentalization of their synthesis – the EicosaCell assay – modi-fications of a prior technique was used (9). The EicosaCell rationalrelies on the specific features of the heterobifunctional cross-linker1-ethyl-3-(3-dimethylamino-propyl) carbodiimide (C8H17N3-HCl; EDAC) used. EDAC immobilizes newly synthesizedeicosanoids by cross-linking the eicosanoid carboxyl groups to theamines of adjacent proteins localized at eicosanoid-synthesizingcompartment. Such EDAC-mediated reaction forms a bond with-out any spacer length between the two molecules, favoring an

EicosaCell 165

accurate positioning of the newly synthesized eicosanoid withinthe cell. In addition, while other cross-linkers formed bonds thatoften generate foreign molecules, EDAC-driven eicosanoid-bondis homologous to native eicosanoid that allows immunoassayslike EicosaCell. Besides the precise positioned coupling of animmuno-detectable eicosanoid at its sites of formation, EDACenables: (I) the ending of cell stimulation step; (II) cell fix-ation; (III) cell permeabilization, allowing the penetration ofboth anti-eicosanoid and the detecting fluorochrome-conjugatedantibodies into cells; and, importantly, (IV) the relative preser-vation of lipid domains, such as membranes and droplets,which dissipate with air drying or commonly used alcoholfixation.

2. Materials

2.1. ConventionalEicosaCell

1. EDAC (1-ethyl-3-(3-dimethylamino-propyl) carbodiimidehydrochloride) is diluted in Hanks-buffered salt solu-tion without calcium chloride and magnesium chloride(HBSS−/−). Refer to Note 1 for EDAC solution handling.EDAC final concentration with cells varies according to celltype and protocol used (see next subheadings). The work-ing solution should have twice concentration of the finalconcentration with cells. For instance, specifically regardingpurified human eosinophils stimulated as a cell suspension,EDAC final concentration with eosinophils should be 0.1%in HBSS−/−, therefore the EDAC working solution shouldbe diluted to 0.2%. Alternatively, with adherent macrophagesstimulated in 6 wells plate, EDAC final concentration shouldbe 0.5% in HBSS−/−, therefore the EDAC working solutionshould be diluted to 1.0%.

2. Primary antibody to the eicosanoid of interest.3. Fluorescent-labeled secondary antibodies.4. Glass microscope slides and coverslips.5. Anti-fading mounting medium for fluorescence.

2.2. Double-LabelingPurposes

1. DAPI (4′, 6′-Diamidino-2-phenylindole dihydrochloride)stock solution is prepared by dissolving 1 mg/mL of pow-der in distilled water. Aliquots should be stored at −20◦ pro-tected from light.

2. Monoclonal antibody against lysosome-associatedmembrane protein (LAMP) 1.


3. BODIPY R© 493/503 (4,4-difluoro-1,3,5,7,8-pentamethyl-4-bora-3a,4a-diaza-s-indacene) (Molecular Probes; cat no.D-3922, molecular weight: 262). To prepare BODIPY stocksolution, BODIPY should be dissolved in DMSO (1 mM),aliquoted in small Eppendorf tubes (∼10 μL per tube) andstored at −20◦C protected from light. BODIPY workingsolution should be diluted fresh 1000× in HBSS−/− andkept from light.

4. Monoclonal or polyclonal antibody to adiposedifferentiation-related protein (ADRP).

3. Methods

3.1. EicosaCell withCells in Suspension

EicosaCell can be easily performed with a varied of cell types insuspension, such as purified human blood leukocytes, cell lin-eages, as well as, peritoneal, pleural or bronchoalveolar animalcells. After in vivo or in vitro stimulation of these cell popu-lations, incubation with EDAC should instantaneously guaran-tee the immobilization of eicosanoids at their synthesizing spotwithin the cell, just before cytospin slides are prepared to allowmicroscopic analysis. As schematically illustrated in Fig. 10.1a,after preparing a cell suspension, EDAC working solution shouldbe added to cell suspension and incubated for a period of timeto ensure cell fixation, immobilization of eicosanoid and cellpermeabilization.

1. Prepare a cell suspension of 2 × 106/mL. Gently andimmediately add an equal volume of EDAC solution, pre-pared as described in Section 2.1, Step 1 (refer to Notes 1and 2 for details), to the cell suspension.

2. Incubate the cell suspension with EDAC for 30 min to 1 hat 37◦C.

3. Cytospin the cells onto slides using 100 μL of the cell sus-pension at 23 g for 5 min.

4. Wash twice in HBSS−/−.

5. Labeling of newly formed eicosanoids can be done witha variety of already tested antibodies, as already publishedelsewhere (10–13). Incubate cells with the primary anti-body to the eicosanoid of interest for 1 h at room tem-perature. The non-immune serum from the animal wherethe secondary antibody was produced may be added to theprimary antibody so as to decrease unspecific labeling.

EicosaCell 167

B

A

o

o

cells/agarose mixture

± stimuli

- eicosanoid-protein cross-linking

Phase-Contrast andFluorescence Microscopy

3. Detection

EDAC:

37oC37oC15 min 1h

A23187

LTC4

Lipid bodies

** *

2. Fixation

- cell permeabilization

1. Induction

Y

YY

- anti-eicosanoid- fluorochrome-labeled secundaryAb

1.

harvested leukocytes

- eicosanoid-protein cross-linking

Fluorescence Microscopy

4.Detection

- anti-eicosanoid- fluorochrome-labeled secundary Ab

EDAC:

** *

2. Fixation- cell permeabilization

Y

YY

CO2

1.In vivo stimulation

3. Cystospin

Fig. 10.1. Schematic illustration of EicosaAssay method. EicosaCell preparations, whichundergo EDAC-dependent capturing and fixation of newly formed-eicosanoids at theirsites of synthesis, are analyzed by phase-contrast and fluorescence microscopy and canemploy cytospun cells (a), adherent cells or cells embedded in a gel matrix (b).


6. Wash 2–3 times in HBSS−/−.

7. Incubate with the fluorescent-labeled secondary antibodyfor 1 h at room temperature.

8. At the end of the staining procedure, cytospun cells shouldbe always extensively washed with HBSS−/−, at least 3times for 5 min each.

9. Slides should be mounted using an aqueous mountingmedium, preferentially with anti-fading.

10. Analysis is performed on phase contrast to observe cellmorphology and fluorescence microscope or confocal scan-ning laser microscope to identify the eicosanoid label-ing. For example, analyses and image acquisition can beobtained using an Olympus BX-FLA fluorescence micro-scope equipped with a Plan Apo 100× 1.4 Ph3 objectiveand CoolSNAP-Pro CF digital camera in conjunction withImage Pro Plus R© software (Media Cybernetics) (see Note3 for details).

11. The specificity of the eicosanoid immuno-labeling usingEicosaCell system should be always ascertained by includ-ing some mandatory control conditions as detailed inNote 4.

As shown in Fig. 10.2, EicosaCell system was success-fully employed on macrophages recovered from pleuralcavities of BCG-infected or control mice (11). Briefly,cells obtained 24 h after infection with BCG and con-trols were recovered from the pleural cavity with 500μL of HBSS−/− and immediately mixed with 500 μLof EDAC (1% in HBSS−/−). After 30 min incubation at37◦C with EDAC, pleural leukocytes were then washedwith HBSS−/−, cytospun onto glass slides and incubatedwith mouse anti-PGE2 in 0.1% normal goat serum andguinea pig polyclonal anti-mouse ADRP (see Section 2.2)in 0.1% normal donkey serum simultaneously for 1 h atroom temperature. Isotyping matching antibodies (murineIgG1) were used as controls (Fig. 10.2). Cells were washedtwice and incubated with secondary antibodies, goat anti-mouse conjugated with AlexaFluor-488 (1/1000, Molec-ular Probes) and CY3-conjugated donkey anti-guinea pig(1/1000). Slides were washed (three times, 10 min each)and mounted with aqueous mounting medium. Cellswere analyzed by both phase-contrast and fluorescencemicroscopy. As a control for PGE2 specificity of detec-tion, one group of BCG-infected animals was treated withindomethacin (4 mg/kg), 4 h before sacrificing animals forcell recovery (not shown).

EicosaCell 169

Fig. 10.2. EicosaCell for PGE2 immuno-localization within BCG-infected cytospun macrophages. In upper panels,macrophages from BCG-infected animals were labeled for ADRP-associated lipid bodies (red staining) and for newlyformed PGE2 (green staining). Merged image showed co-localization of PGE2 in ADRP-associated lipid bodies (yel-low staining). In bottom panels, IgG1 irrelevant isotype (MOPC) was used as control for PGE2 labeling. Briefly, pleuralmacrophages obtained 24 h after infection with BCG were recovered from the thoracic cavity with 500 μL of HBSS,immediately mixed with 500 μL of EDAC (1% in HBSS) and incubated for 30 min at 37◦C. Cells were then washed withHBSS, cytospun onto glass slides and incubated with mouse anti-PGE2 (1/100) or MOPC 21 in 0.1% normal goat serumand guinea pig polyclonal anti-mouse ADRP (1/1000) in 0.1% normal donkey serum simultaneously for 1 h at RT. Cellswere washed twice and incubated with secondary Abs, goat anti-mouse conjugated with AlexaFluor-488 (1/1000) andCY3-conjugated donkey anti-guinea pig (1/1000). The slides were washed (three times, 10 min each) and mounted withaqueous mounting medium.

3.2. EicosaCell withAdherent Cells

To study the intracellular compartmentalization of eicosanoidsynthesis by EicosaCell in adherent cells, extra care should betaken to ensure the conservation of cell adherence and morphol-ogy during EDAC step. EicosaCell have succeeded to immuno-localize PGE2 within at least three distinct cell types: platedmurine macrophages (D’Avila et al., unpublished) and two lin-eages of intestinal cells, CACO-2 (a human colon adenocarci-noma cell line) (Fig. 10.3) (14) and IEC-6 (a rat epithelial cellline (15)).

1. While adherent on glass coverslips, cells can be incubated for30 min to 1 h at 37◦C with EDAC at 0.5% in HBSS−/− tocross-link the lipid mediator of interest to carboxyl groupsto amines in adjacent proteins (Fig. 10.3) without affecting


Fig. 10.3. EicosaCell for PGE2 immuno-localization within adherent CACO-2 cells. The largest panel shows fluores-cent microscopy of CACO-2 cells labeled for newly formed PGE2 (red staining). Bottom three images panel showedimmunofluorescent PGE2 (red staining), BODIPY-associated lipid bodies (green staining) and a merged image showingco-localization of PGE2 in lipid bodies (yellow staining). Insert panel showed lack of PGE2 immuno-labeling within lipid-body-enriched CACO-2 cells, which were treated with indomethacin (4 mg/kg) 1 h before EDAC. Briefly, CACO-2 cellswere fixed and permeabilized during 1 h at 37◦C with EDAC (0.5% in HBSS–/–). Then, cells were washed with HBSS andblocked with 2% donkey serum for 15 min before incubation with anti-PGE2 monoclonal antibody (Cayman Chemicals)for 45 min. Cells were washed with HBSS and incubated with fluorescent secondary antibody Cy3-conjugated affiniPureF(ab’) fragment donkey anti-mouse and BODIPY 493/503 (Molecular Probes, CA) for 45 min.

EicosaCell 171

cell morphology (refer to Note 5). Alternatively, cells grownin Lab-Tek chambers can be used.

2. Gently wash cells 2–3 times in HBSS−/−.3. Incubate cells with the primary antibody to the eicosanoid

of interest for 1 h at room temperature. The non-immuneserum from the animal where the secondary antibody wasproduced may be added to the primary antibody so as todecrease unspecific labeling.

4. Gently wash 2–3 times in HBSS−/−.5. Incubate with the fluorescent-labeled secondary antibody

for 1 h at room temperature.6. At the end of the staining procedure, cells should be gently

washed with HBSS−/−, at least 3 times for 5 min each.7. Cell-containing coverslips should be carefully glued to the

slide and mounted using an aqueous mounting medium,preferentially with anti-fading.

8. Analysis is performed on phase contrast to observe cellmorphology and fluorescence microscope or confocal scan-ning laser microscope to identify the eicosanoid label-ing. For example, analyses and image acquisition can beobtained using an Olympus BX-FLA fluorescence micro-scope equipped with a Plan Apo 100 × 1.4 Ph3 objective(Olympus) and CoolSNAP-Pro CF digital camera in con-junction with Image Pro Plus R© software (Media Cybernet-ics) (see Note 3 for details).

9. As for cytospun cells, the specificity of the eicosanoidimmuno-labeling using EicosaCell system should be ascer-tained by including mandatory controls listed in Note 4.

3.3. EicosaCell withCells Embeddedin a Gel Matrix

In contrast to analyzing cytospun cells which do not preservein situ morphology, cells embedded in an agarose matrix, thatare kept in a hydrated system with a substrate where they cancrawl, display tissue-like cell morphology exhibiting polariza-tion and other characteristics of activated leukocyte, for instance.Therefore, by immuno-localizing eicosanoids at its formation siteswithin agarose-embedded cells (as schematically illustrated inFig. 10.1b), generated products may be microscopically localizedat cell structures assembled during stimulation and preserved incells that are not cytospun into slides (10, 16).

1. To prepare the agarose matrix, 2.5% agarose (24 C gellingpoint) (Promega) in sterile distilled H2O, is melted at70◦C; and while liquid at 37◦C, 9 volumes of agarose aremixed with 1 volume of 10× concentrated RPMI 1640medium.


2. One volume of this medium-supplemented agarose is mixedwith one volume of RPMI 1640 medium containing 2% fattyacid free-albumin at 37◦C and with three volumes of thestudied cell, exemplified here as human eosinophils (Fig.10.1b), which should be at 15 × 106 cells/mL in RPMI1640 medium containing 1% fatty acid free-human albumin.

3. Stimuli are then added in 0.1 volumes to agarose/eosinophilmixtures. As schematically illustrated in Fig. 10.1b,immediately thereafter, 20 μL samples are gently spreadonto microscope slides and covered with perfusion chamber(CoverWellTM).

4. Each slide is overlaid with RPMI 1640 medium containing1% albumin and an identical concentration of the stimuluspresent in the agarose/eosinophil mixture.

5. Slides can be incubated (37◦C, humidified 5% CO2) forvarying periods of time.

6. Overlying medium should be removed and replaced withRPMI 1640, 1% albumin medium, that may contain or not0.1 μM calcium ionophore (A23187) and incubated forextra 15 min (37◦C; 5% CO2).

7. Stimulations are stopped by removing chambers and addingEDAC. Fixation and permeabilization of cells with properimmobilization of newly formed eicosanoids at its intracel-lular sites of synthesis are achieved by immersing the slidescontaining stimulated cells in 0.5% EDAC (in HBSS−/−) for30 min.

8. After three washes (5 min each) with HBSS−/−,the fluorochrome-labeled anti-eicosanoid, for instanceAlexa488-labeled rat anti-cysteinyl leukotriene (LT) detec-tion mAb (Sigma) (AlexaTM488 protein labeling using a kitfrom Molecular Probes) should be added (400 μL of 10μg/mL) for 1 h.1. Slides need to be extensively washed with HBSS−/−, at

least 3 times for 5 min each.2. Aqueous mounting medium should be applied to each

slide before coverslip attachment.3. Slides can be viewed by both phase-contrast and fluores-

cence microscopy as detailed above – Section 3.1 (Step11).

4. Mandatory control conditions, as listed in Note 4,should be always included as for cytospun and adher-ent cells to ascertain the specificity of the eicosanoidimmuno-labeling using EicosaCell system.

EicosaCell 173

3.4. Double-LabelingProcedures toIdentify Eicosanoid-SynthesizingIntracellular Sites

3.4.1. NuclearLocalization

To better visualize perinuclear eicosanoid synthesis by EicosaCell,a double labeling with DAPI is advised.

1. After EDAC and antibody incubation steps, EicosaCell slidespreparations should be extensively washed in HBSS−/−and then incubated with DAPI (DAPI working solution100 ng/mL or 300 nM, see Section 2.2, Step 1) for 5 minbefore aqueous mounting medium application.

2. The morphology of the cells’ nuclei is observed using a flu-orescence microscope at excitation wavelength 350 nm.

3.4.2. PhagosomalLocalization

As performed by Balestrieri and coworkers (17), phagosomeinvolvement in eicosanoid synthesis can be ascertained by co-localizing the phagosomal protein marker LAMP-1 in EicosaCellpreparations.

1. After incubation at 37◦C with EDAC, cells should bewashed with HBSS−/−, cytospun onto glass slides and incu-bated with the anti-eicosanoid of interest and the primaryantibody against LAMP-1 (2.5 μg/mL) in blocking buffer(5% normal donkey serum) for 2 h at room temperature.

2. Negative control cells are instead incubated for 2 h withappropriate IgG.

3. After 2 h, the cells are washed extensively with HBSS−/−and incubated for 1 h at room temperature with fluorescent-labeled secondary antibody to detect the primary antibodyagainst the eicosanoid of interest and with fluorescent-labeled secondary antibody to detect the primary antibodyagainst LAMP-1(1:200).

4. The cells should be washed five times with HBSS−/− andthen mounted in aqueous mounting medium.

3.4.3. Lipid BodyLocalization

To investigate lipid body role in eicosanoid synthesis by Eicosa-Cell assay, two double-labeling strategies can be employed:BODIPY or anti-ADRP immunostaining (for further informationon lipid body labeling refer to Chapter 9). Both approaches canbe used for adherent, suspension or agarose-embedded cells.

3.4.3.1. BODIPY R©493/503 LipidBody-Labeling

1. To employ BODIPY R© 493/503 strategy, incubate Eicosa-Cell preparations (coverslips or slides) with 1 μm BODIPY(working solution) simultaneously to secondary antibodyincubation for 45–60 min at 37◦C (water bath).


2. To remove free BODIPY after incubation, EicosaCell prepa-rations should be washed at least twice in HBSS−/− beforeaqueous mounting medium application and coverslip attach-ment to slides.

3.4.3.2. ADRP LipidBody-Labeling

Alternatively, to visualize lipid bodies, anti-ADRP immuno-labeling may be performed as detailed in Chapter 9.

1. Add anti-ADRP antibodies together with the anti-eicosanoid antibody of interest for 1 h at room temperature.(a) For human cells incubate with mouse anti-human ADRPat dilution of 1:20 (2.5 μg/mL, final concentration); (b) Formouse, rat, human or bovine cells, incubate with guinea piganti-human ADRP polyclonal antibody at dilution of 1:300(final dilution).

2. Wash three times in HBSS−/− or PBS.3. Incubate with fluorescent-labeled secondary antibody for

1 h room temperature (together with the secondary anti-body for the EicosaCell labeling.

4. Wash three times in HBSS−/− or PBS.5. Mount in mounting medium for fluorescence microscopy.

Common problems and non-obvious features found inimmunofluorescent-detection of eicosanoids in EDACpreparations by EicosaCell with their possible explanationsand potential solutions are described in Note 6.

4. Notes

1. EDAC working solution should be prepared fresh, kept pro-tected from light and discarded after each experiment.

2. Incubation of cells with EDAC can be carried out on eithercell in suspension or with the cells already cytospun ontoslides by dropping EDAC on top of the cells. Even thoughthe latter method is less costly, some differences in preserva-tion of cell morphology, cell permeabilization and eicosanoiddetection may occur and should be analyzed with care.

3. Analysis of EicosaCell preparations should be performedas soon as slides are mounted, inasmuch as immuno-fluorescent labeling is usually not stable for a long periodbleaching after a certain time. Even though freezingmay preserve fluorescence overnight, EDAC-treated cellsmay display altered cell appearance after freezing–thawingcycle.

EicosaCell 175

4. The specificity of the eicosanoid immuno-labeling usingEicosaCell system should be always ascertained by includ-ing some mandatory control conditions: (i) non-stimulatedEDAC-treated cells labeled with the proper anti-eicosanoidantibody; (ii) the incubation (1 h min before EDAC)with the eicosanoid synthesis inhibitors, such as cPLA2-αinhibitor (e.g. pyrrolidine-2; 1 μM), COX inhibitor (e.g.indomethacin; 1 μg/mL), FLAP inhibitor (e.g. MK886;50 μg/animal or 10 μM for in vitro incubations) or 5-LOinhibitor (e.g. zileuton; 50 μg/animal or 10 μM for in vitroincubations) to avoid the eicosanoid synthesis and (iii) theuse of an irrelevant antibody control. Optionally, other suit-able controls to check specificity and performance of Eicosa-Cell are (i) to use, instead of EDAC, paraformaldehyde,which will not immobilize the newly synthesized eicosanoidwithin cells; (ii) to, in parallel, carry out the EicosaCell in adifferent cell type that lacks the ability to synthesize the tar-geted eicosanoid (for instance, to use neutrophils to checkspecificity of LTC4 immunodetection by EicosaCell) or (iii)to analyze mixed populations of responsive plus unrespon-sive cells to a specific stimulus, so you can reassure that thetargeted eicosanoid is specifically detected only within stim-ulated cells.

5. While adherent CACO 2 cells can be incubated with EDACfor 1 h, IEC-6 cells can be incubated for at most 30 min (atsame concentration; 0.5% in HBSS−/−) to retain reasonablecell morphology and PGE2 immuno-detection at synthesiz-ing compartments (refer to the original articles (14, 15) fordetails of blocking and staining conditions with anti-PGE2monoclonal antibody (Cayman Chemicals) and proper sec-ondary antibodies.

6. Common problems and non-obvious features found inimmunofluorescent-detection of eicosanoids in EDACpreparations by EicosaCell:

Lack of eicosanoid detection: When few or no eicosanoid spe-cific immunostaining is observed (but expected), the prob-lem usually lies in the improper fixation (e.g. EDAC-drivencross-linking) of targeted eicosanoid at its sites of synthesis.Thus, the newly formed eicosanoid would be washed-outfrom the EicosaCell preparation turning detection impos-sible. Resolution of this problem is normally achieved byadjusting (slight increase) concentration and/or time ofincubation of EDAC. Alternatively, the lack of immuno-detection of newly formed eicosanoids can be due to inef-ficient stimulation; a positive control with a known agonistshould be always included in experiments.


Eicosanoid detection within non-stimulated cells: Eicosanoidsare lipid mediators non-storable in the cell and newlyformed upon stimulation, therefore non-stimulated cellsshould not show any immunostaining for the targetedeicosanoid. Thus, non-stimulated cells should always beincluded as an important negative control. However, cellactivation during procedures including cell incubation at37◦C or cell fixation/permeabilization with EDAC can leadto spontaneous, stimulus-independent eicosanoid synthesis.Throughout cell preparation, care is needed to ensure thatcells are not mechanically, chemically or immunologicallystimulated. Unexpected eicosanoid detection within Eicosa-Cell preparations can also result from non-specific detection(discussed below).

Non-specific detection: Fluorescent detection antibodies maynon-specifically bind to other lipids found within cells orbind to other cellular structures. The crosslinking prop-erties of EDAC may favor the tendency for cells to besticky; therefore antibodies could interact through low-affinity non-antigen binding site. To investigate non-specificbinding in EicosaCell preparations, a proper control usinghost/isotype-matched irrelevant antibodies, must be alwaysincluded. An additional mandatory control that needs tobe always included in the experimental design to rule outnon-specific immuno-staining is the condition with a synthe-sis inhibitor of the targeted eicosanoid. Synthesis inhibitor-treated controls should show no immune-labeling confirm-ing specific detection of targeted eicosanoid. If non-specificstaining is too high (>10% positive), there are several pos-sible remedies. The detecting antibody may be diluted fur-ther, or a different one from a different host may be tried.Also, it is possible to try an adsorbing reagent that effectivelyblocks out non-specific sites, such as a normal serum (samehost of the detecting antibody). Non-specific fluorescencecan also be detected when the solution of detecting antibodycontains a high degree of aggregated antibody; therefore,it is important to centrifuge the detecting antibody beforeadding to cell preparations.

Poor preservation of cell morphology: During EDAC incuba-tion step of EicosaCell assays, cell appearance may changefrom unimportant to severe modification of typical cell mor-phology. This undesirable effect of EDAC on cells canbe avoided by adjusting both EDAC concentration andincubation time.

Losing cell adherence with EDAC: Similar to unwantedEDAC effect on cell morphology, the ability of cells to stay

EicosaCell 177

adhered to coverslips or other substrates can be affected byEDAC incubation. Again, previous careful setting of EDACincubation step is obligatory and should be adjusted for eachcell type.

Lost of cell integrity: Eicosanoid localization within cellsby EicosaAssay may be tricky sometimes since somecell types are destroyed during EDAC-driven cross-linking/permeabilization step. For instance, even thoughlipid bodies of human neutrophils and basophils are sites of5-LO localization (10, 18), EicosaCell assays with agarose-embedded neutrophils and basophils were not feasiblesince these cells did not endure to EDAC-driven fixa-tion/permeabilization process, which precedes eicosanoidimmuno-detection by EicosaCell, indicating that the com-bination of gel matrix with EDAC step may be useful tostudy only a small group of tough cells, like eosinophils.Compartmentalization studies of eicosanoid synthesis withinmore fragile cells like neutrophils and basophils, how-ever, can be carried out with EicosaCell system in non-gelsolutions.

5. Applications

Over the past decade, intracellular compartmentalization ofeicosanoid-synthetic machinery has emerged as a key componentof the regulation of eicosanoid synthesis (reviewed in (4–6, 19).However, the direct evaluation of specific subcellular locales ofeicosanoid synthesis has been elusive, as those lipid mediators arenewly formed, not stored and often rapidly released upon cellstimulation. Thus, in the majority of studies, intracellular sites ofeicosanoid synthesis have been inferred based on the identifica-tion of eicosanoid-forming enzymes localization.

The EicosaCell technique described herein enables to directlypinpoint the intracellular locales of eicosanoid synthesis by detect-ing the newly formed lipids and has been successfully ableto confirm the dynamic aspect involved in the localization ofeicosanoid synthesis, providing direct evidence of compartmen-talization within perinuclear envelope (10, 14, 15, 20), phago-somes (17) or lipid bodies in accord to cell type and stimu-latory condition (10–12, 15, 21); (Figs. 10.4 and 10.5). Sofar, the EicosaCell assay has been used to identify the produc-tion of leukotriene C4 (LTC4) (10, 13, 17, 22), leukotrieneB4 (LTB4) (12, 23), prostaglandin E2 (PGE2) (11, 14, 15)


Fig. 10.4. EicosaCell for LTC4 immuno-labeling within gel-immersed humaneosinophils. Fluorescent microscopy of agarose-embedded eosinophils, fixed with EDAC(0.5% in HBSS–/–) and stained with Alexa488-labeled anti-cysteinyl LT mAb. To facilitateintracellular localization of anti-LTC4 immunoreactive sites (green staining) within repre-sentative eosinophils, blue and white dotted lines were drawn to delineate, respectively,the nuclear and cellular perimeters. As indicated, A23187-, eotaxin- or eotaxin plusA23187-stimulated eosinophils display fluorescent LTC4 immunostaining.

EicosaCell 179

3.Phagosomes

2. Lipid Bodies

1. Perinuclear Membrane

Cellular compartments of eicosanoid synthesis identified by EicosaCell

Bandeira-Melo et al., 2001Tedla et al., 2003Accioly et al., 2008Moreira et al., 20091 2

3

Balestrieri et al., 2006

Bandeira-Melo et al., 2001Bandeira-Melo et al., 2002Vieira-de-Abreu et al., 2005Mesquita-Santos et al., 2006 D’Avila et al., 2006Pacheco et al., 2007 Accioly et al., 2008Moreira et al., 2009

eicosanoid imunnofluorescence

Fig. 10.5. Schematic summary of EiocosaCell-derived reports identifying three distinct intracellular compartments ofeicosanoid synthesis: the nuclear envelope, cytoplasmic lipid bodies and zymozan-driven phagosomes.

and prostaglandin D2 (PGD2) (unpublished observations) indifferent cell types and under different stimulatory conditions.Moreover, it could in principle be adapted to intracellular detec-tion of other lipid mediators as long as specific antibodies areavailable.

Of note, the EicosaCell Assay has high sensitivity,enabling the detection of low levels of intracellular gener-ated eicosanoids even when extracellular released eicosanoidscould not be detected by conventional eicosanoid enzymeimmune assay (10, 15). Indeed, it has been shown that besidesparacrine/autocrine activities, eicosanoids may display intracrinefunctions (22, 24). For instance, by employing EicosaCell tech-nique, it has been uncovered that a lipid-body-derived LTC4have intracellular functions in controlling cytokine release fromeosinophils (25). Therefore, by identifying compartmentalizedlevels of eicosanoids, besides providing new insights of regulationof eicosanoid biosynthesis, EicosaCell assay may contribute toidentification of likely intracellular functions of newly synthesizedeicosanoids.

�Fig. 10.4. (continued) While, eosinophils stimulated with A23187 (0.1 μM) for 15 minexhibited exclusively perinuclear (stars) immunoreactive LTC4, eotaxin (100 ng/mL)-stimulated eosinophils showed punctate cytoplasmic lipid body-comprised LTC4(arrows). Differently, eosinophils stimulated with eotaxin for 1 h and activated forextra 15 min with A23187 exhibit perinuclear (stars) and punctate cytoplasmic (arrows)immunoreactive LTC4.


Acknowledgments

The work of authors is supported by PRONEX-MCT, ConselhoNacional de Desenvolvimento Cientifico e Tecnológico (CNPq,Brazil), PAPES-FIOCRUZ, Fundação de Amparo à Pesquisa doRio de Janeiro (FAPERJ, Brazil) and NIH grants (AI022571,AI020241, AI051645). Authors are indebted with Dr. HeloisaD’Avila and Dr. Patricia Pacheco for the contributions to figuresused in the manuscript.

References

1. Yaqoob, P. (2003) Fatty acids as gatekeepersof immune cell regulation. Trends Immunol24, 639–645.

2. Wymann, M. P., Schneiter, R. (2008) Lipidsignaling in disease. Nat Rev Mol Cell Biol 9,162–176.

3. Smith, W. L., DeWitt, D. L., Garavito, R. M.(2000) Cyclooxygenases: structural, cellular,and molecular biology. Annu Rev Biochem69, 145–182.

4. Bozza, P. T., Magalhaes, K., Weller, P. F.(2009) Leukocyte lipid bodies – biogenesisand functions in inflammation. Biochim Bio-phys Acta 1791, 540–51.

5. Mandal, A. K., Skoch, J., Bacskai, B. J.,Hyman, B. T., Christmas, P., Miller, D.,Yamin, T. T., Xu, S., Wisniewski, D., Evans,J. F., Soberman, R. J. (2004) The mem-brane organization of leukotriene synthesis.Proc Natl Acad Sci USA 101, 6587–6592.

6. Peters-Golden, M., Brock, T. G. (2001)Intracellular compartmentalization ofleukotriene synthesis: unexpected nuclearsecrets. FEBS Lett 487, 323–326.

7 Diaz, B. L., Arm, J. P. (2003) Phos-pholipase A(2). Prostaglandins leukotrienesessent. Fatty Acids 69, 87–97.

8. Bandeira-Melo, C., Weller, P. F. (2003)Eosinophils and cysteinyl leukotrienes.Prostaglandins Leukot Essent Fatty Acids 69,135–143.

9. Liu, L. X., Buhlmann, J. E., Weller, P.F. (1992) Release of prostaglandin E2 bymicrofilariae of Wuchereria bancrofti andBrugia malayi. Am J Trop Med Hyg 46,520–523.

10. Bandeira-Melo, C., Phoofolo, M., Weller, P.F. (2001) Extranuclear lipid bodies, elicitedby CCR3-mediated signaling pathways, are

the sites of chemokine-enhanced leukotrieneC4 production in eosinophils and basophils.J Biol Chem 276, 22779–22787.

11. D’Avila, H., Melo, R. C. N., Parreira, G. G.,Werneck-Barroso, E., Castro-Faria-Neto, H.C., Bozza, P. T. (2006) Mycobacterium bovisbacillus Calmette-Guerin induces TLR2-mediated formation of lipid bodies: intracel-lular domains for eicosanoid synthesis in vivo.J Immunol 176, 3087–3097.

12. Pacheco, P., Vieira-de-Abreu, A., Gomes, R.N., Barbosa-Lima, G., Wermelinger, L. B.,Maya-Monteiro, C. M., Silva, A. R., Bozza,M. T., Castro-Faria-Neto, H. C., Bandeira-Melo, C., Bozza, P. T. (2007) Monocytechemoattractant protein-1/CC chemokineligand 2 controls microtubule-driven biogen-esis and leukotriene B4-synthesizing func-tion of macrophage lipid bodies elicited byinnate immune response. J Immunol 179,8500–8508.

13. Vieira-de-Abreu, A., Assis, E. F., Gomes,G. S., Castro-Faria-Neto, H. C., Weller, P.F., Bandeira-Melo, C., Bozza, P. T. (2005)Allergic challenge-elicited lipid bodies com-partmentalize in vivo leukotriene C4 synthe-sis within eosinophils. Am J Respir Cell Mol33, 254–261.

14. Accioly, M. T., Pacheco, P., Maya-Monteiro,C. M., Carrossini, N., Robbs, B. K., Oliveira,S. S., Kaufmann, C., Morgado-Diaz, J. A.,Bozza, P. T., Viola, J. P. (2008) Lipid bodiesare reservoirs of cyclooxygenase-2 and sitesof prostaglandin-E2 synthesis in colon cancercells. Cancer Res 68, 1732–1740.

15. Moreira, L. S., Piva, B., Gentile, L. B.,Mesquita-Santos, F. P., D’Avila, H., Maya-Monteiro, C. M., Bozza, P. T., Bandeira-Melo, C., Diaz, B. L. (2009) Cytosolic phos-

EicosaCell 181

pholipase A2-driven PGE2 synthesis withinunsaturated fatty acids-induced lipid bodiesof epithelial cells. Biochim Biophys Acta 1791,156–165.

16. Bandeira-Melo, C., Sugiyama, K., Woods, L.J., Phoofolo, M., Center, D. M., Cruikshank,W. W., Weller, P. F. (2002) IL-16 pro-motes leukotriene C(4) and IL-4 releasefrom human eosinophils via CD4- andautocrine CCR3-chemokine-mediated sig-naling. J Immunol 168, 4756–4763.

17. Balestrieri, B., Hsu, V. W., Gilbert, H.,Leslie, C. C., Han, W. K., Bonventre, J. V.,Arm, J. P. (2006) Group V secretory phos-pholipase A2 translocates to the phagosomeafter zymosan stimulation of mouse peri-toneal macrophages and regulates phagocy-tosis. J Biol Chem 281, 6691–6698.

18. Pacheco, P., Bozza, F. A., Gomes, R.N., Bozza, M., Weller, P. F., Castro-Faria-Neto, H. C., Bozza, P. T. (2002)Lipopolysaccharide-induced leukocyte lipidbody formation in vivo: innate immu-nity elicited intracellular loci involved ineicosanoid metabolism. J Immunol 169,6498–6506.

19. Bandeira-Melo, C., Bozza, P. T., Weller, P.F. (2002) The cellular biology of eosinophileicosanoid formation and function. J AllergyClin Immunol 109, 393–400.

20. Tedla, N., Bandeira-Melo, C., Tassinari,P., Sloane, D. E., Samplaski, M., Cosman,D., Borges, L., Weller, P. F., Arm, J. P.(2003) Activation of human eosinophilsthrough leukocyte immunoglobulin-like

receptor 7. Proc Natl Acad Sci USA 100,1174–1179.

21. Mesquita-Santos, F. P., Vieira-de-Abreu,A., Calheiros, A. S., Figueiredo, I. H.,Castro-Faria-Neto, H. C., Weller, P. F.,Bozza, P. T., Diaz, B. L., Bandeira-Melo,C. (2006) Cutting edge: prostaglandinD2 enhances leukotriene C4 synthesis byeosinophils during allergic inflammation:synergistic in vivo role of endogenouseotaxin. J Immunol 176, 1326–1330.

22. Devchand, P. R., Keller, H., Peters, J. M.,Vazquez, M., Gonzalez, F. J., Wahli, W.(1996) The PPARalpha-leukotriene B4 path-way to inflammation control. Nature 384,39–43.

23. Silva, A. R., Pacheco, P., Vieira-De-Abreu,A., Dalegria, B., Bandeira-Melo, C., Castro-Faria-Neto, H. C., Bozza, P. T. (2009)Lipid bodies in oxidized LDL-induced foamcells are leukotriene-synthesizing organelles:a MCP-1/CCL2 regulated phenomenon.Biochim Biophys Acta 1791, 1066–75.

24. Kliewer, S. A., Lenhard, J. M., Willson,T. M., Patel, I., Morris, D. C., Lehmann,J. M. (1995) A prostaglandin J2 metabo-lite binds peroxisome proliferator-activatedreceptor gamma and promotes adipocyte dif-ferentiation. Cell 83, 813–819.

25. Bandeira-Melo, C., Woods, L. J., Phoofolo,M., Weller, P. F. (2002) Intracrine cys-teinyl leukotriene receptor-mediated sig-naling of eosinophil vesicular transport-mediated interleukin-4 secretion. J Exp Med196, 841–850.

Chapter 11

Nestin-Driven Green Fluorescent Protein as an ImagingMarker for Nascent Blood Vessels in Mouse Modelsof Cancer

Robert M. Hoffman

Abstract

A transgenic mouse, in which the regulatory elements of the stem cell marker, nestin drive green flu-orescent protein (ND-GFP), expresses GFP in nascent blood vessels. Red fluorescent protein (RFP)-expressing tumors transplanted to nestin-GFP mice enable specific visualization of nascent vessels in thegrowing tumors. The ND-GFP mouse was also utilized to develop a rapid in vivo/ex vivo fluorescentangiogenesis assay by implanting Gelfoam R©, a surgical sponge derived from pigskin, which was rapidlyvascularized by fluorescent nascent blood vessels. Angiogenesis could be imaged and quantified whenstimulated or inhibited by specific compounds in both tumors and Gelfoam R©. These fluorescent modelscan be used to study the early events of angiogenesis and to quantitatively determine efficacy of antian-giogenesis compounds.

Key words: Green fluorescent protein, red fluorescent protein, nude mouse, human tumors,color-coded, imaging, nascent blood vessels.

1. Introduction

1.1. Previous ModelsUsed to DetermineAngiogenesis

The discovery and evaluation of antiangiogenic substances ini-tially relied on methods such as the chorioallantoic membraneassay (1, 2), the monkey iris neovascularization model (3), thedisk angiogenesis assay (4) and various models that use the corneato assess blood vessel growth (5–10). Although they are impor-tant for understanding mechanisms of blood vessel induction,these models do not represent tumor angiogenesis.


183

184 Hoffman

Subcutaneous tumor xenograft mouse models have beendeveloped to study tumor angiogenesis, but these require cum-bersome pathological examination procedures, such as histol-ogy and immunohistochemistry. Measurements require animalsacrifice and therefore preclude ongoing angiogenesis studies inindividual, live, tumor-bearing animals. Moreover, subcutaneoustumor xenografts are not representative models of metastaticcancer.

Tumors transplanted in the cornea of rodents (11–13) androdent skin-fold window chambers have also been used for angio-genesis studies (14–20). The cornea and skin-fold chamber mod-els provide a means for studying tumor angiogenesis in livinganimals. However, quantification requires specialized proceduresand the sites do not represent natural environments for tumorgrowth. The cornea and skin-fold window chamber tumor mod-els do not allow metastasis to occur, which may involve mech-anisms of angiogenesis (21) that are qualitatively different fromthose occurring in ectopic models that do not metastasize.

We describe here the clinically-relevant imageable mousemodels of cancer to visualize and quantify tumor angiogenesisand efficacy of inhibitors.

1.2. FluorescentProteins to ImageAngiogenesis

For in vivo imaging, a strong signal and high resolution are neces-sary. The green fluorescent protein (GFP) gene, cloned from thebioluminescent jellyfish Aequorea victoria (22), was chosen to sat-isfy these conditions because it has great potential for use as a cel-lular marker (23, 24). Green fluorescent protein cDNA encodesa 283-amino acid monomeric polypeptide with Mr = 27,000(25, 26) that requires no other A. victoria proteins, substratesor cofactors to fluoresce (27). Gain-of-function bright mutantsexpressing the GFP gene have been generated by various tech-niques (28–30) and have been humanized for high expression andsignal (31). Red fluorescent proteins (RFP) from the Discosomacoral have similar features as well as the advantage of longer wave-length emission (32–34). Our laboratory has pioneered the use ofGFP for in vivo imaging (35) including non-invasive whole-bodyimaging (36, 37). The Nobel Prize in 2008 was awarded for thediscovery and first practical uses of GFP.

Fluorescent proteins have been shown by our laboratory tobe very useful for imaging tumor angiogenesis. We have devel-oped unique mouse models to image tumor angiogenesis withfluorescent proteins, which are described in this review.

1.3. Orthotopic TumorModels ExpressingFluorescent Proteinsto Visualize TumorAngiogenesis

For realistic and real-time imageable tumor angiogenesis mod-els, we have developed surgical orthotopic implantation (SOI)metastatic models of human cancer (38). These models placetumors in natural microenvironments and replicate clinical tumorbehavior more closely than do ectopic implantation models (38).

Nestin-Driven Green Fluorescent Protein as an Imaging Marker 185

The orthotopically-growing tumors, in contrast to most othermodels, give rise to spontaneous metastases that resemble, both intarget tissues and in frequency of occurrence, the clinical behav-ior of the original human tumor (38). Tumors implanted in theorthotopic model have been transduced and selected to stronglyexpress GFP or RFP in vivo (37). Orthotopically-implanted GFP-or RFP-labeled tumors enable the visualization of the role ofangiogenesis in metastasis. As Li et al. (18) point out, angiogene-sis initiation in metastatic tumors may be very different from thatof primary tumors and require different interventions. Moreover,the extreme detection sensitivity afforded by the strong GFP orRFP fluorescence allows imaging of very early events in bloodvessel induction.

GFP or RFP expression in primary tumors and in their metas-tases in the mouse models can be detected by an intense fluo-rescence seen by intravital or by whole-body imaging. The non-luminous angiogenic blood vessels appear in contrast as sharplydefined dark networks against this bright background. The high-image resolution permits quantitative measurements of total ves-sel length. These genetically fluorescent tumor models therebyallow quantitative optical imaging of angiogenesis in vivo. Tumorgrowth, vascularization and metastasis could be followed in realtime (39).

1.4. Intravital Imagesof Angiogenesis ofOrthotopic PancreasCancer

The clarity of angiogenic blood vessel imaging was initially illus-trated by intravital examination of the orthotopic growth of theBxPC3-GFP pancreatic tumor. The non-luminous blood vesselswere clearly visible in contrast against the GFP fluorescence of theprimary tumor. Angiogenesis associated with metastatic growthswas also readily imaged by intravital examination (39) (seeNote 1).

Because angiogenesis could be measured without animal sac-rifice, it was possible to determine a time course for individualanimals. Sequential intravital images of angiogenesis for the PC-3human prostate tumor expressing GFP and growing orthotopi-cally in a single nude mouse were acquired. The tumor-associatedblood vessels were clearly visible by day 7 and continued toincrease at least until day 20 (39).

1.5. Whole-BodyImaging ofAngiogenesis inOrthotopic BreastCancer

We have demonstrated whole-body images and quantitation ofthe time course of angiogenesis of the MDA-MB-435-GFPhuman breast cancer growing orthotopically in the mammary fatpad in a nude mouse. The development of the tumor and itsangiogenesis could be imaged in a completely noninvasive man-ner (39). The mouse mammary fat pad is the orthotopic environ-ment for the implanted MDA-MB-435-GFP breast cancer andallows noninvasive, whole-body imaging of tumor angiogenesis.The quantitative angiogenesis data show that microvessel density

186 Hoffman

increased over 20 weeks (39). Similar results were obtained bynoninvasive imaging of angiogenesis of the Lewis lung cancer-GFP growing in the footpad of a nude mouse (39).

1.5.1. Non-invasiveImaging of Tumor BloodFlow

To noninvasively image cancer cell/stromal cell interaction in thetumor microenvironment and drug response at the cellular levelin live animals in real time, we developed a new imageable three-color animal model. The model consists of GFP-expressing micetransplanted with dual-color cancer cells labeled with GFP in thenucleus and RFP in the cytoplasm. The Olympus IV100 LaserScanning Microscope, with ultra-narrow microscope objectives(‘stick objectives’), was used for three-color whole-body imagingof the two-color cancer cells interacting with the GFP-expressingstromal cells. In this model, drug response of both cancer andstromal cells in the intact live animal is also imaged in real time.Various in vivo phenomena of tumor-host interaction and cellulardynamics were non-invasively imaged, including tumor vascula-ture and tumor blood flow (40).

1.6. Skin FlapsEnable Ultra-highResolution ExternalImaging of TumorAngiogenesis

Opening a reversible skin flap in the light path markedly reducedsignal attenuation, increasing detection sensitivity many-fold. Theobservable depth of tissue is thereby greatly increased (41).The brilliance of the tumor GFP fluorescence, facilitated bythe lucidity of the skin-flap window, allowed imaging of theinduced microvessels by their dark contrast against a brightbackground. The orthotopically-growing BxPC3-GFP humanpancreatic tumor was externally visualized under fluorescencemicroscopy to be surrounded by its microvessels visible by theirdark contrast (41).

1.7. Imaging ofNascentAngiogenesis UsingNestin-Driven GFPTransgenic Mice

We initially reported that in mice in which the gene for the stemcell marker, nestin drives GFP (ND-GFP) (42), that ND-GFPalso labels developing skin blood vessels. The ND-GFP labeledvessels appear to originate from hair follicles and form a follicle-linking network. This was seen most clearly by transplanting ND-GFP-labeled vibrissa (whisker) hair follicles to unlabeled nudemice. New vessels grew from the transplanted follicle and thesevessels increased when the local recipient skin was wounded. TheND-GFP-expressing structures are blood vessels, because theydisplay the characteristic endothelial-cell-specific markers CD31and von Willebrand factor. This model displays very early eventsin angiogenesis and can serve for rapid antiangiogenesis drugscreening (43).

1.7.1. Dual-ColorImaging of TumorAngiogenesis

We visualized tumor angiogenesis by dual-color fluorescenceimaging in ND-GFP transgenic mice after transplantation ofthe murine melanoma cell line B16F10 expressing RFP. ND-GFP was highly expressed in proliferating endothelial cells andnascent blood vessels in the growing tumor (Fig. 11.1). Results


100 µm

ND-GFP-expressing blood vessels growing into tumor mass

Fig. 11.1. Visualization of ND-GFP vessels in an RFP-expressing tumor. On day 14 afterimplantation of RFP-expressing B16 mouse melanoma cells subcutaneously in ND-GFPmice, ND-GFP-expressing blood vessels (white arrows) could be seen in the growingtumor. Nascent ND-GFP blood vessels (white arrows) were forming a network in thegrowing tumor. Bar, 100 μm (44).

of immunohistochemical staining showed that the blood vessel-specific antigen CD31 was expressed in ND-GFP-expressingnascent blood vessels. ND-GFP expression was diminished in ves-sels with increased blood flow. Progressive angiogenesis duringtumor growth was readily visualized by GFP expression. Doxoru-bicin inhibited the nascent tumor angiogenesis as well as tumorgrowth in the ND-GFP mice transplanted with B16F10-RFP (44)(Fig. 11.2) (see Note 2).

1.8. Nestin-DrivenGFP Transgenic NudeMice

The ND-GFP gene was crossed into nude mice on the C57/B6background to obtain ND-GFP nude mice. ND-GFP wasexpressed in the brain, spinal cord, pancreas, stomach, esoph-agus, heart, lung, blood vessels of glomeruli, blood vessels ofskeletal muscle, testis, hair follicles and blood vessel network inthe skin of ND-GFP nude mice. Human lung cancer, pancre-atic cancer, breast and colon cancer cell lines as well as a murinemelanoma cell line expressing RFP were implanted orthotopi-cally and an RFP-expressing human fibrosarcoma was implanteds.c. in the ND-GFP nude mice. These tumors grew extensivelyin the ND-GFP mice. ND-GFP was highly expressed in prolife-rating endothelial cells and nascent blood vessels in the growingtumors, visualized by dual-color fluorescence imaging (Fig.11.3).The ND-GFP transgenic nude mouse model thus enables thevisualization of nascent angiogenesis in human and mouse tumorprogression (45).

188 Hoffman

Fig. 11.2.


1.9. Imagingof NascentAngiogenesisin Tumors in ND-GFPNude Mice

Angiogenesis in experimental lung and liver metastases ofmelanoma was imaged in the ND-GFP transgenic mice. Themurine melanoma cell line, B16F10 expressing RFP, was injectedi.v. in ND-GFP mice. ND-GFP was highly expressed in prolif-erating nascent blood vessels in tumors that developed in thelung after tail vein injection and in tumors that developed inthe liver after portal vein injection of RFP-expressing melanomacells. Liver metastasis and angiogenesis were imaged intravi-tally. Doxorubicin significantly decreased metastatic angiogenesis(Fig. 11.3) (46).

Dual-color fluorescence imaging visualized tumor angio-genesis in the ND-GFP transgenic nude mice after ortho-topic transplantation of the MIA PaCa-2 human pancreatic can-cer line expressing RFP. Mice were treated with gemcitabineat 150 mg/kg/dose on days 3, 6, 10 and 13 after tumorimplantation. At day 14, mice were sacrificed and mean nascentblood vessel density and tumor volume were calculated andcompared to control mice. The density of nascent blood ves-sels in the tumor was readily quantitated. Gemcitabine sig-nificantly decreased the mean nascent blood vessel density inthe tumor as well as decreased tumor volume. The dual-colormodel of the ND-GFP nude mouse orthotopically implantedwith RFP-expressing pancreatic tumor cells enabled the simul-taneous visualization and quantitation of tumor angiogenesisand tumor volume. These results demonstrated for the firsttime that gemcitabine is an inhibitor of angiogenesis as well astumor growth in pancreatic cancer. The results have importantimplications for the clinical application of gemcitabine in thisdisease (47).

Nascent angiogenesis was imaged in pancreatic cancer livermetastasis in the ND-GFP transgenic nude mice, formed afterintra-splenic injection of XPA-1 human pancreatic cancer cellsexpressing RFP, using color-coded fluorescence imaging. ND-GFP was highly expressed in proliferating endothelial cells andnascent blood vessels in the growing liver metastasis. The den-sity of nascent blood vessels in the tumor was readily quantitated.Gemcitabine significantly decreased the mean nascent blood ves-sel density in the pancreatic liver metastases (48).

�Fig. 11.2. (continued) Effect of doxorubicin on tumor angiogenesis. a On day 10 after implantation of tumor cells, theND-GFP nascent blood vessels (white arrows) were forming a network in the central tumor. b In the marginal area ofthe tumor, many newly formed nascent ND-GFP blood vessels were growing. The nascent ND-GFP blood vessels (whitearrows) had many branches and were connected to each other. c and d The mice were given daily i.p. injections of5 μg/g of doxorubicin at days 0, 1 and 2 after implantation of tumor cells. C, by day 10 after implantation of tumor cells,the nascent ND-GFP blood vessels were not seen in the central area of the tumor. d In the marginal area of the tumor,ND-GFP blood vessels (white arrows) were growing slightly. e Number of nascent blood vessels per tumor volume in thedoxorubicin-treated mice was less than NaCl solution- injected mice (p < 0.05). Bars, 100 μm (44).

190 Hoffman

Fig. 11.3. Fluorescence imaging of tumor angiogenesis in transgenic ND-GFP nudemice. Human HT1080 fibrosarcoma on day 14 after s.c. injection. Dual-color HT1080cells expressing GFP in the nucleus and RFP in the cytoplasm are polarized towardsND-GFP-expressing blood vessels (white arrows) growing in the tumor mass. Bar, 100μm (45).

The antiangiogenic efficacy of CPT-11 was evaluated in ahuman colon tumor growing in ND-GFP nude mice using color-coded fluorescence imaging. We orthotopically implanted ND-GFP nude mice with the human colon cancer cell line HCT-116expressing RFP. The mice were treated with CPT-11 at 40 mg/kgon days 7, 10 and 14. Tumor angiogenesis was imaged and visu-alized by color-coded fluorescence imaging on day 17, three daysafter the last CPT-11 treatment. Tumor volume and the meannascent blood vessel density were determined and compared tothe control mice. The nascent blood vessels were highly fluo-rescent and their density was determined. ND-GFP nude micethat were administered CPT-11 showed significant reduction inthe mean nascent blood vessel density and tumor volume. Theresults showed that CPT-11 is an effective inhibitor of angiogen-esis and provided strong implications for wider clinical applicationof CPT-11 for colon cancer (49).

Angiogenesis of the HT-1080 human fibrosarcoma cell line,expressing RFP, was imaged in the ND-GFP nude mice. Cancercells were injected into either the muscle or the bone. Nestinwas highly expressed in proliferating endothelial cells and nascentblood vessels in the growing tumors, including the surround-ing tissues. CD31 co-localized in ND-GFP-expressing nascent


blood vessels. The density of nascent blood vessels in the tumorwas readily quantitated. The mice were given daily i.p. injec-tions of 5 mg/kg doxorubicin after implantation of cancer cells.Doxorubicin significantly decreased the mean nascent blood ves-sel density in tumors as well as decreased tumor volume. Thesedata suggest targeting angiogenesis of sarcomas as a promisingclinical approach (50).

1.10. RapidIn Vivo/Ex VivoNascentAngiogenesis Assay

We developed a very convenient imageable in vivo angiogenesisassay after transplantation of Gelfoam R© (Pharmacia & UpjohnCompany, Kalamazoo, MI, USA) in the ND-GFP mice (51).Gelfoam R© is rapidly vascularized with GFP-expressing vesselsin the presence of an angiogenesis stimulator. Anti-angiogenesisagents inhibit this process. Thus, this rapid and simple newin vivo assay can rapidly identify angiogenesis stimulators andinhibitors. Gelfoam R© was treated with β-fibroblast growth fac-tor (bFGF). The treated Gelfoam R© was then transplanted intothe subcutis on both flanks of the ND-GFP transgenic mice. Themice were given daily intraperitoneal (i.p.) injections of doxoru-bicin or NaCl solution at day 0, 1 and 2 after transplantationof Gelfoam R©. Skin flaps were made at day 7 after transplanta-tion of Gelfoam R© under anesthesia. Angiogenesis was quantifiedby measuring the length of ND-GFP-expressing nascent bloodvessels in the Gelfoam R© in the skin flap by in vivo fluorescencemicroscopy imaging. The vessels on the surface were countedunder fluorescence microscopy. Each experimental group con-sisted of five mice. Gelfoam R©, treated with bFGF, implantedin the ND-GFP mice was rapidly vascularized with ND-GFP-expressing blood vessels. At day 7 after transplantation, the ND-GFP-expressing nascent blood vessels were observed forminga network on the surface of the bFGF-treated Gelfoam R© inthe skin flap (Fig. 11.4). Implanted Gelfoam R© that was nottreated with bFGF was not vascularized. The ND-GFP vesselsin the Gelfoam R© stained positively for CD31, demonstratingthe presence of endothelial cells. Day 7 was chosen as an arbi-trary time point to measure the GFP vessels in the implantedGelfoam R©. The Gelfoam R© can be analyzed at any time pointand an optimal time for measurement would depend on theangiogenesis drug being tested. ND-GFP mice that receivedi.p. injections of doxorubicin (5 μg/g) at day 0, 1 and 2 aftertransplantation of Gelfoam R©, with or without bFGF, had fewerND-GFP-expressing nascent blood vessels than NaCl-treatedmice (57). Future experiments will address the destruction ofpreformed vessels in Gelfoam R© by vascular disrupting agents(VDAs) (52).

192 Hoffman

Fig. 11.4. Angiogenesis of implanted Gelfoam R© with GFP-expressing vessels. ND-GFP mice were given daily intraperi-toneal (i.p.) injections of 0.9% NaCl solution at day 0, 1, and 2 after transplantation of Gelfoam R© with or without β

fibroblast growth factor (bFGF). a At day 7 after transplantation of Gelfoam R© with bFGF, the ND-GFP-expressing nascentblood vessels formed a network on the surface of Gelfoam R© visualized in a skin flap. The ND-GFP-expressing nascentblood vessels had many branches that were connected to each other. b The Gelfoam R© transplanted ND-GFP mice weretreated with 5 μg/g doxorubicin (DOX) at day 0, 1 and 2 after transplantation. Doxorubicin significantly decreased theblood-vessel density in the presence of bFGF at day 7. Scale bar, 500 μm (51).

1.11. A Brain-MetastaticParalyzing SpinalCord Glioma ThatInducesAngiogenesis andNeurogenesis

Cancer of the spinal cord is a highly malignant disease that oftenleads to paralysis and death. To develop an imageable, patient-like model of this disease, U87 human glioma tumor fragments,expressing RFP, were transplanted by surgical orthotopic implan-tation (SOI) into the spinal cord in non-transgenic nude miceor ND-GFP transgenic nude mice. In the ND-GFP mice, GFPis expressed in nascent blood vessels and neural stem cells. Ani-mals were treated with temozolomide or vehicle control. Theintramedullary spinal cord tumor (IMSCT) grew at the primarysite, causing hind-limb paralysis and also metastasized to thebrain. Temozolomide inhibited tumor size and prevented metas-tasis as well as prevented paralysis or delayed paralysis. The tumorstimulated both neurogenesis and angiogenesis (53).

1.12. Growth andMetastasis of SpinalCord Glioma

Four weeks after transplantation, the spinal cord was exposedby laminectomy. The RFP-expressing glioma in the spinal cordwas observed by fluorescence imaging. The skull was opened andthe brain was excised en bloc. Brain metastases were detectedeven though they were small. Metastases were found mainlyaround the brain stem and in leptomeninges at the basilarsulcus (53).

1.13. Efficacy ofChemotherapy onSpinal Cord Growthand Metastasis

The primary tumor size was significantly reduced by temozolo-mide compared to untreated controls (0.55 ± 1.1 vs. 9.7 ±4.1 mm2, p < 0.01, n = 5 for each group). Brain metastaseswere found only in the control group (60%). Histological anal-ysis of the control group showed aggressive tumor invasion in


the spinal cord. In contrast, the temozolomide-treated animalsshowed mostly scar tissue after tumor transplantation (53).

1.14. Efficacy ofChemotherapy onHind-Limb ParalysisDue to Spinal CordGlioma

The untreated mice showed progressive paralysis beginning at14 days after U87 glioma transplantation. The untreated controlgroup developed complete paralysis (BBB score = 0) between18 and 31 days after tumor transplantation. Some of thetemozolomide-treated mice started to show paralysis at approx-imately 35 days after transplantation and four mice were still notparalyzed at 60 days. The temozolomide-treated mice survivedwithout complete paralysis for at least 45 days. There was a sig-nificant delay to onset or else complete inhibition of paralysis inthe treated animals (log rank statistic 8.37, p < 0.005) (53).

1.15. Angiogenesisand Neurogenesis ofSpinal Cord Glioma

In frozen sections of the normal spinal cord, ND-GFP-expressingcells were mainly seen around the central canal. Ten daysafter U87-RFP glioma transplantation, ND-GFP-expressing cellsappeared stimulated by the tumor and started to surround it.The main stem of the ND-GFP-expressing cells had many smallbranches and ND-GFP-expressing cells appeared to originatefrom cells around the central spinal cord canal. In young (6weeks) mice, more ND-GFP-expressing cells were observed sur-rounding the tumor than in old (16 weeks) mice. The mean GFPintensity around the tumor in young mice was significantly higherthan in old mice (p < 0.05) (52).

Neuronal class III-β-tubulin is a marker of neuronal cells.Some of the ND-GFP-expressing cells surrounding the spinalcord glioma also expressed III-β-tubulin, indicating that someof the ND-GFP-expressing cells accumulating around the tumorare of neural origin. Some of the ND-GFP-expressing cells alsoappeared to be endothelial cells since they expressed CD31.Frozen sections showing the ND-GFP host cells and RFP-expressing U87 cells under fluorescence microscopy were com-pared with sister sections immunohistochemically stained forCD31. This comparison demonstrated co-localization of ND-GFP and CD-31. These results indicate that the ND-GFP-expressing cells surrounding the tumor contained both neural andendothelial types (53).

There have been only few reports of human tumors metasta-sizing to the brain in mouse models (54–57) making the presentmodel very valuable to study brain metastasis. Temozolomide iseffective in this model against primary tumor growth, in the spinalcord, paralysis and brain metastasis. Differential labeling of thetumor and host enabled the observation that the primary tumorstimulates both blood vessel and nerve growth. This novel mousemodel will enable a deeper understanding of spinal cord cancerand provide a clinically-relevant system for drug discovery andevaluation.

194 Hoffman

2. Materials

2.1. Reagents 1. Restriction enzymes HindIII and NotI2. RFP cDNA (pDsRed2; Clontech)3. Plasmid pLNCX24. PT67 packaging cells (Clontech); 3T3 cells for viral titer-

ing; human and mouse cell lines to be transfected withgenes encoding fluorescent proteins

5. Growth medium (normal and selective) appropriate for cellculture, such as DMEM (Invitrogen; Irvine Scientific)

6. Fetal bovine serum (FBS; Gemini Biological Products)7. Lipofectamine PLUS (Invitrogen)8. G418 neomycin (Invitrogen)9. Polysulfonic filter, 4.5 μm

10. Polybrene11. Trypsin-EDTA (Fisher Scientific) and trypsin12. Mice expressing GFP (‘GFP mice’; Jackson Laboratories;

Japan SLC, Hamamatsu, Japan)13. Immunocompetent and immunodeficient mice (Charles

River; Taconic; Harlan Teklad). Mice are fed an auto-claved laboratory rodent diet (Tecklad LM-485, WesternResearch Products).

14. Anesthetic reagents: ketamine mixture (10 μL ketamineHCl, 7.6 μL xylazine and 2.4 μL acepromazine maleate,injected s.c.).

15. Nair (Carter-Wallace)16. Doxorubicin17. NaCl, 0.9%18. Optimum cutting temperature blocks19. Antibody to rat immunoglobulin (anti-rat immunoglobu-

lin) and anti-mouse immunoglobulin horseradish peroxi-dase detection kits (BD PharMingen, San Diego, CA)

20. Monoclonal anti-CD31 (CBL1337; Chemicon, Temecula,CA)

21. Monoclonal anti-nestin (rat 401; BD PharMingen, SanDiego, CA)

22. Monoclonal anti-III-β-tubulin (Covance Research Prod-ucts, Berkeley, CA)

23. Substrate-chromogen 3,3′-diaminobenzidine


24. Doxorubicin (5 μg/g body weight in a 2 mg/mL solutionof 0.9% NaCl)

2.2. Equipment 1. Culture dishes, 60 mm (Fisher Scientific); flask, 25 mm;plates, 96-well

2. Humidified incubator at 37◦C and 5% CO2

3. Cloning cylinders (Bel-Art Products)4. 27G2 latex-free syringe, 1 mL (Becton Dickinson)5. 8-0 surgical suture6. Leica fluorescence stereo microscope, model LZ12, with a

mercury 50-W power supply7. D425/60 bandpass filter and 470 DCXR dichroic mirror8. D470/40 emission filter and GG475 emission filter

(Chroma Technology)9. C5810 three-chip cooled color charge-coupled device

(CCD) camera (Hamamatsu Photonics Systems) or DP70CCD camera (Olympus) for high-resolution capture (1024× 724 pixels)

10. Image-Pro Plus 3.1 or 4.0 software (Media Cybernetics)11. Personal computer (PC; IBM or Fujitsu-Siemens)12. VCR (Sony, model SLV-R1000)13. Blue LED flashlight (LDP LLC)14. Coolpix camera (Nikon)15. Fluorescent light box with fiberoptic lighting at 470 nm

(Lightools Fluorescent Imaging System; LightoolsResearch)

16. OV100 Small Animal Imaging System (Olympus) with anM20 light source (Olympus Biosystems) and 470-nm exci-tation light

17. IV100 Laser Scanning Microscope (Olympus)18. Paint Shop Pro 8 (Corel) and cellR (Olympus Biosystems)19. Olympus BH 2-RFCA fluorescence microscope equipped

with a mercury 100-W lamp power supply20. Leica CM1850 cryostat

3. Methods

Use one of the following options to establish a mouse tumormodel of fluorescent protein–expressing tumor cells: i.v. cell injec-tion or SOI.

196 Hoffman

3.1. Cell Injectionto Establish anExperimentalMetastasis Model

1. Collect fluorescent protein-expressing cancer cells bytrypsinization for 3 min at 37◦C with 0.25% trypsin.

2. Wash cells three times with cold serum-free medium using atabletop centrifuge at 500g.

3. Resuspend cells in approximately 0.2 mL serum-freemedium.

4. Within 30 min of collecting cells, inject 1 × 106 tumor cellsin a total volume of 0.2 mL into 6-week-old C57BL/6 GFPmice or nude (nu/nu) GFP mice, or ND-GFP C57BL/6immunocompetent or nu/nu mice, in the lateral tail veinor subcutaneously using a 1-mL 27G2 latex-free syringe.Cells in suspension may lose viability over time and there-fore should be injected as soon as possible.

5. For liver colonization, inject fluorescent protein-expressingcells directly into the portal vein in anesthetized mice (detailson inducing anesthesia are presented below).

3.2. SurgicalOrthotopicImplantationto Establisha SpontaneousMetastasis Model(IND)

1. Induce anesthesia with a ketamine mixture2. Use a microscope with magnification of ×6 to ×40 for all

procedures of the operation.3. Isolate fluorescent protein-expressing tumor fragments

(1 mm3) from subcutaneously growing tumors, formed byinjection of RFP-expressing cancer cells, by mincing tumortissue into 1-mm3 fragments. After proper exposure of thetarget organ, implant three tumor fragments per mouse.

4. With 8-0 surgical suture, penetrate the tumor fragments andsuture fragments onto the target organ. Orthotopic implan-tation of tumor fragments results in higher spontaneousmetastatic rates than injection of a cell suspension.

5. Keep mice in a barrier facility under high-efficiency particu-late air filtration.

3.3. Skin-FlapWindows

Tumor cells on the various internal organs are visualized throughthe body wall through a skin-flap window over the abdomen (41).

1. Animals are anesthetized with a ketamine mixture.2. An arc-shaped incision is made in the skin and s.c. connective

tissue is separated to free the skin flap.3. The skin flap can be opened repeatedly to image cancer

cells on the internal organs through the nearly transpar-ent body wall. The skin is closed with a 6–0 suture.This procedure greatly reduces the scatter of fluorescentphotons.

3.4. Tumor TissueSampling

1. Obtain tumor tissue biopsies from 3 days to 4 weeks afterinoculation of cancer cells. Biopsies of tumor tissue can be


obtained from anesthetized mice by removal of a small pieceof tumor tissue (1 mm3 or less) with a scalpel. Staunchbleeding by pressing the wound with sterile gauze. Alterna-tively, the mouse can be killed and the tissue can be collectedand processed for analysis.

2. Cut fresh tissue into pieces of about 1 mm3 and gently pressonto slides for fluorescence microscopy. This procedure isdone manually on normal slides.

3. To analyze tumor angiogenesis, digest tissues with trypsin-EDTA for 5 min at 37◦C before examination.

4. After trypsinization, put tissues on pre-cleaned microscopeslides and cover with another microscope slide.

3.5. FluorescenceMicroscopy

1. Use a fluorescence microscope equipped with a mercury100-W lamp power supply.

2. To visualize both GFP and RFP fluorescence at the sametime, produce excitation light via a D425/60 filter and a470 DCXR mirror.

3. Collect emitted fluorescence light through a GG475 filter.4. Capture high-resolution images and store directly on a PC.5. Process images for contrast and brightness using Image-Pro

software or its equivalent.

3.6. Methods forColor-Coded Imagingof Tumor BloodVessels of Mice

Use one of the following methods for whole-body imaging ofmice: microscopy, flashlight imaging, light-box imaging or cham-ber imaging.

3.6.1. Microscopy 1. Use a fluorescence stereo microscope equipped with a mer-cury lamp power supply.

2. Produce selective excitation of GFP via a D425/60 filter and470 DCXR mirror.

3. Collect emitted fluorescence through a long-pass filter(GG475) on a CCD camera.

4. Process images for contrast and brightness with the use ofImage-Pro software or its equivalent.

5. Capture high-resolution images directly on a PC or contin-uously through video output on a high-resolution VCR.

6. For C57BL/6 mice, remove hair with Nair or by shav-ing before images are obtained. Hair is highly autofluores-cent, so improper removal of hair will result in low-qualityimages.

3.6.2. Flashlight Imaging 1. Use a blue LED flashlight with an excitation filter (midpointwavelength peak of 470 nm) and a D470/40 emission filter

198 Hoffman

for whole-body imaging of GFP mice with RFP-expressingtumors growing in or on internal organs. Correct filters arenecessary to eliminate tissue autofluorescence (58).

2. Acquire images with a digital camera and store on a PC asdescribed above.

3. For C57BL/6 mice, remove hair with Nair or by shavingbefore images are obtained. Hair is highly autofluorescent,so improper removal of hair will result in low-quality images.

3.6.3. Light-Box Imaging 1. Do whole-body imaging in a fluorescent light box illumi-nated by fiberoptic lighting at 470 nm (36).

2. Collect emitted fluorescence through a GG475 filter on aCCD camera. (Use of separate band-pass filters for RFP orGFP emission allows a monochrome camera to be used.)

3. Capture high-resolution images directly on a PC.4. Process images for contrast and brightness with the use of

Image-Pro software or its equivalent.5. For C57BL/6 mice, remove hair with Nair or by shav-

ing before images are obtained. Hair is highly autofluores-cent, so improper removal of hair will result in low-qualityimages.

3.6.4. Chamber Imaging 1. Do whole-body imaging with an Olympus OV100 imagingsystem (59).

2. Collect emitted fluorescence through appropriate filters con-figured on a filter wheel with a CCD camera. Variable magni-fication imaging can be done with a series of objective lenses.

3. Capture images on a PC and process images for contrast andbrightness with Paint Shop Pro.

4. For C57BL/6 mice, remove hair with Nair or by shavingbefore images are obtained. Hair is highly autofluorescent,so improper removal of hair will result in low-quality images.

3.6.5. Imaging with theOlympus IV-100 LaserScanning MicroscopeSystem

1. The tissue to be imaged using the Olympus IV-100 maybe imaged either ex vivo or in a deeply anesthetized animalwhile secured. It is critical for optimal image resolution thatthe tissue being imaged does not move with the respiratoryand cardiac variation in the animal (40).

2. Ex vivo tissue can be simply placed on a dark surface underthe IV-100 objective with frequent application of PBS tokeep the tissue moist during imaging.

3. Imaging of blood vessels, lymphatics and tumor tissue in askin flap requires stabilization of the skin flap itself away fromthe body of the animal.


4. Likewise, imaging of tumor blood levels, lymphatics orblood vessels in the leg can be achieved by stabilization of theextremity such that the animal’s respiratory variation doesnot cause movement artifact in the tissue being imaged.

5. Intravital imaging of deeper organs in living animals requiresstabilization of the organ and tumor tissue in question. Thiscan be achieved in some organs, such as the pancreatictail, which can be moved and stabilized without sacrifice ofthe animal, provided that the mouse remains deeply anes-thetized throughout the duration of the imaging procedure.

6. Variable magnification down to the subcellular level can beimaged using the full range of objectives. Differential exci-tation of fluorophores can be achieved in this system by theuse of three different lasers for excitation at 488, 561 and633 nm.

3.6.6. Imaging UsingSpectral Separation

In general, spectral separation imaging systems can providegreater sensitivity for specific fluorophore emission, although notall systems are equipped for high-resolution imaging (60).

1. The standard fluorescence imaging system previouslydescribed is replaced with a cooled monochrome cam-era and liquid-crystal tunable filter (CRI, Inc., Woburn,MA or equivalent) positioned in front of a conventionalmacro-lens.

2. A series of images is typically acquired every 10 nm from 500to 650 nm and assembled into a spectral ‘stack’.

3. Using the predefined GFP and RFP emission spectra, thecollected spectral ‘stack’ can be resolved into various imagescorresponding to specific wavelengths of interest that repre-sent autofluorescence, GFP and RFP signals.

4. This method allows for maximal signal-to-noise ratio acqui-sition by virtue of its ability to separate out the competingautofluorescence or other fluorescence signals.

5. It is critical for this image acquisition that there be no move-ment in the tissue image when overlay images of multiplefluorescence signals are to be created (61).

3.7. Methodsof AngiogenesisAnalysis in GFPModels

3.7.1. FluorescenceContrast or Color-CodedImaging

1. Whole-body, skin-flap or intravital imaging can be per-formed (39, 40, 62).

2. Selective excitation of GFP is produced through a D425/60filter and a 470 DCXR mirror.

200 Hoffman

3. Emitted fluorescence is collected through a GG475 filter ona CCD camera.

4. Images are processed for contrast and brightness and ana-lyzed with the use of Image-Pro Plus software.

5. High-resolution images are captured directly on a PC orcontinuously through video output on a high-resolutionVCR.

3.7.2. QuantitativeAnalysis of Angiogenesis

Periodically, the tumor-bearing mice are examined by intravital orwhole-body fluorescence imaging (39, 40, 62).

1. The extent of blood vessel development in a tumor is evalu-ated based on the total length of blood vessels (L) in chosenareas: areas containing the highest number of vessels wereidentified by scanning tumors using intravital or whole-bodyimaging.

2. To compare the level of vascularization during tumorgrowth, the ‘hot’ areas with the maximum developmentof vessels per unit area are quantitated for L expressed inpixels. Captured images were corrected for unevenness inillumination.

3. Then the total number of pixels derived from the blood ves-sels is quantified with IMAGE PRO PLUS software.

3.8. Evaluation ofAnti-angiogeneticAgents

1. Give mice daily i.p. injections of doxorubicin or other drugsor 0.9% NaCl solution (vehicle controls) on days 0, 1 and 2after implantation of tumor cells (44, 45).

2. Anesthetize mice with the ketamine mixture and obtainbiopsies on days 10, 14, 21 and 28 after implantation.

3. Gently flatten the tumor tissue between the slide and cover-slip.

4. Quantify angiogenesis in the tumor tissue by measuring thelength of GFP-expressing blood vessels in all fields using flu-orescence microscopy.

5. Obtain measurements in all fields at ×40 or ×100 magnifi-cation to calculate the total length of GFP-expressing bloodvessels.

6. Calculate the vessel density by dividing the total length ofGFP-expressing blood vessels (in mm) by the tumor volume(in mm3).

3.9. Immunohisto-chemicalStaining

1. ‘Snap-freeze’ fresh tissue with liquid nitrogen, then orientand embed the frozen tissue in optimum cutting tempera-ture blocks and store at −80◦C. Cut the frozen sections to athickness of 5 μm with a cryostat (51).


2. Detect co-localization of GFP fluorescence, CD31, III-β-tubulin and nestin in the frozen skin sections ofmice transgenic for ND-GFP expression using the anti-rat immunoglobulin and anti-mouse immunoglobulinhorseradish peroxidase detection kits following the manu-facturer’s instructions.

3. Use monoclonal anti-CD31 (1:50 dilution), monoclonalanti-III-β-tubulin (1:500 dilution) and monoclonal anti-nestin (1:80 dilution) as primary antibodies.

4. Use staining with substrate-chromogen 3,3′-diaminobenzidine for antigen detection (see Note 3).

4. Notes

1. Advantages of visualizing blood vessels by their contrast toGFP-expressing tumors are the simplicity of the methodand possibility of whole-body or external imaging. Thedisadvantage is that very small capillaries may not bevisible.

2. The advantage of the dual-color model is the great resolu-tion it affords to visualize very fine vessels. The disadvan-tage is that for highest resolution, tissue preparation may beneeded.

3. The fluorescent models of angiogenesis will enable the pro-cess to be visualized at unprecedented resolution in real time(63).

Acknowledgments

These studies were supported in part by grants CA099258 andCA103563 from the National Cancer Institute.

References

1. Auerbach, R., Kubai, L., Knighton, D.,Folkman, J. (1974) A simple procedurefor the long-term cultivation of chickenembryos. Dev Biol 41, 391–394.

2. Crum, R., Szabo, S., Folkman, J. (1985) Anew class of steroids inhibits angiogenesis inthe presence of heparin or a heparin frag-ment. Science 230, 1375–1378.

3. Miller, J. W., Stinson, W. G., Folkman, J.(1993) Regression of experimental irisneovascularization with systemic alpha-interferon. Ophthalmology 100, 9–14.

4. Passaniti, A., Taylor, R. M., Pili, R., Guo,Y., Long, P. V., Haney, J. A., et al. (1992)A simple, quantitative method for assessingangiogenesis and antiangiogenic agents using

202 Hoffman

reconstituted basement membrane, heparin,and fibroblast growth factor. Lab Invest 67,519–528.

5. Alessandri, G., Raju, K., Gullino, P. M.(1983) Mobilization of capillary endothe-lium in vitro induced by effectors of angio-genesis in vivo. Cancer Res 43, 1790–1797.

6. Deutsch, T. A., Hughes, W. F. (1979) Sup-pressive effects of indomethacin on thermallyinduced neovascularization of rabbit corneas.Am J Ophthalmol 87, 536–540.

7. Korey, M., Peyman, G. A., Berkowitz, R.(1977) The effect of hypertonic ointmentson corneal alkali burns. Ann Ophthalmol 9,1383–1387.

8. Mahoney, J. M., Waterbury, L. D. (1985)Drug effects on the neovascularizationresponse to silver nitrate cauterization of therat cornea. Curr Eye Res 4, 531–535.

9. Li, W. W., Grayson, G., Folkman, J.,D’Amore, P. A. (1991) Sustained-releaseendotoxin. A model for inducing cornealneovascularization. Invest Ophthalmol Vis Sci32, 2906–2911.

10. Epstein, R. J., Hendricks, R. L., Stulting,R. D. (1990) Interleukin-2 induces cornealneovascularization in A/J mice. Cornea 9,318–323.

11. Gimbrone, M. A., Jr., Cotran, R. S., Leap-man, S. B., Folkman, J. (1974) Tumorgrowth and neovascularization: An experi-mental model using the rabbit cornea. J NatlCancer Inst 52, 413–427.

12. Fournier, G. A., Lutty, G. A., Watt, S.,Fenselau, A., Patz, A. (1981) A cornealmicropocket assay for angiogenesis in therat eye. Invest Ophthalmol Vis Sci 21,351–354.

13. Muthukkaruppan, V., Auerbach, R. (1979)Angiogenesis in the mouse cornea. Science205, 1416–1418.

14. Papenfuss, H. D., Gross, J. F., Intaglietta,M., Treese, F. A. (1979) A transparentaccess chamber for the rat dorsal skin fold.Microvasc Res 18, 311–318.

15. Shan, S., Robson, N. D., Cai, Y., Qiao, T., Li,C. Y., Kontos, C. D., et al. (2004) Responsesof vascular endothelial cells to angiogenic sig-naling are important for tumor cell survival.FASEB J 18, 326–328.

16. Dewhirst, M., Gross, J., Sim, D., Arnold, P.,Boyer, D. (1984) The effect of rate of heat-ing or cooling prior to heating on tumor andnormal tissue microcirculatory blood flow.Biorheol 21, 539–558.

17. Fukumura, D., Xavier, R., Sugiura, T., Chen,Y., Park, E. C., Lu, N., et al. (1998) Tumorinduction of VEGF promoter activity in stro-mal cells. Cell 94, 715–725.

18. Li, C. Y., Shan, S., Huang, Q., Braun, R.D., Lanzen, J., Hu, K., et al. (2000) Ini-tial stages of tumor cell-induced angiogen-esis: evaluation via skin window chambersin rodent models. J Natl Cancer Inst 92,143–147.

19. Al-Mehdi, A. B., Tozawa, K., Fisher, A. B.,Shientag, L., Lee, A., Muschel, R. J. (2000)Intravascular origin of metastasis from theproliferation of endothelium-attached tumorcells: a new model for metastasis. Nat Med 6,100–102.

20. Huang, Q., Shan, S., Braun, R. D., Lanzen,J., Anyrhambatla, G., Kong, G., et al. (1999)Noninvasive visualization of tumors in rodentdorsal skin window chambers. Nat Biotechnol17, 1033–1035.

21. Cowen, S. E., Bibby, M. C., Double, J.A. (1995) Characterisation of the vascula-ture within a murine adenocarcinoma grow-ing in different sites to evaluate the poten-tial of vascular therapies. Acta Oncol 34,357–360.

22. Prasher, D. C., Eckenrode, V. K., Ward,W. W., Prendergast, F. G., Cormier, M. J.(1992) Primary structure of the Aequoreavictoria green-fluorescent protein. Gene 111,229–233.

23. Chalfie, M., Tu, Y., Euskirchen, G., Ward, W.W., Prasher, D. C. (1994) Green fluorescentprotein as a marker for gene expression. Sci-ence 263, 802–805.

24. Cheng, L., Fu, J., Tsukamoto, A., Hawley, R.G. (1996) Use of green fluorescent proteinvariants to monitor gene transfer and expres-sion in mammalian cells. Nat Biotechnol 14,606–609.

25. Cody, C. W., Prasher, D. C., West-ler, W. M., Prendergast, F. G., Ward,W. W. (1993) Chemical structure of thehexapeptide chromophore of the Aequoreagreen fluorescent protein. Biochemistry 32,1212–1218.

26. Yang, F., Moss, L. G., Phillips, G. N.,Jr. (1996) The molecular structure ofgreen fluorescent protein. Nat Biotechnol 14,1246–1251.

27. Morin, J., Hastings, J. (1971) Energy trans-fer in a bioluminescent system. J Cell Physiol77, 313–318.

28. Cormack, B., Valdivia, R., Falkow, S.(1996) FACS-optimized mutants of thegreen fluorescent protein (GFP). Gene 173,33–38.

29. Crameri, A., Whitehorn, E. A., Tate, E.,Stemmer, W. P. (1996) Improved greenfluorescent protein by molecular evolutionusing DNA shuffling. Nat Biotechnol 14,315–319.


30. Delagrave, S., Hawtin, R. E., Silva, C.M., Yang, M. M., Youvan, D. C. (1995)Red-shifted excitation mutants of thegreen fluorescent protein. Biotechnology 13,151–154.

31. Heim, R., Cubitt, A. B., Tsien, R. Y. (1995)Improved green fluorescence. Nature 373,663–664.

32. Zolotukhin, S., Potter, M., Hauswirth, W.W., Guy, J., Muzyczka, N. (1996) A‘humanized’ green fluorescent protein cDNAadapted for high-level expression in mam-malian cells. J Virol 70, 4646–4654.

33. Gross, L. A., Baird, G. S., Hoffman, R. C.,Baldridge, K. K., Tsien, R. Y. (2000) Thestructure of the chromophore within DsRed,a red fluorescent protein from coral. ProcNatl Acad Sci USA 97, 11990–11995.

34. Fradkov, A. F., Chen, Y., Ding, L., Barsova,E. V., Matz, M. V., Lukyanov, S. A. (2000)Novel fluorescent protein from Discosomacoral and its mutants possesses a unique far-red fluorescence. FEBS Lett 479,127–130.

35. Chishima, T., Miyagi, Y., Wang, X.,Yamaoka, H., Shimada, H., Moossa, A.R. et al. (1997) Cancer invasion andmicrometastasis visualized in live tissue bygreen fluorescent protein expression. CancerRes 57,2042–2047.

36. Yang, M., Baranov, E., Jiang, P., Sun,F. X., Li, X. M., Li, L., et al. (2000)Whole-body optical imaging of green fluo-rescent protein-expressing tumors and metas-tases. Proc Natl Acad Sci USA 97,1206–1211.

37. Hoffman, R. M. (2002) Green fluorescentprotein imaging of tumour growth, metas-tasis, and angiogenesis in mouse models.Lancet Oncol 3, 546–556.

38. Hoffman, R. M. (1999) Orthotopicmetastatic mouse models for anticancer drugdiscovery and evaluation: a bridge to theclinic. Invest New Drugs 17, 343–359.

39. Yang, M., Baranov, E., Li, X. M., Wang, J.W., Jiang, P., Li, L., et al. (2001) Whole-bodyand intravital optical imaging of angiogene-sis in orthotopically implanted tumors. ProcNatl Acad Sci USA 98, 2616–2621.

40. Yang, M., Jiang, P., Hoffman, R. M. (2007)Whole-body subcellular multicolor imagingof tumor-host interaction and drug responsein real time. Cancer Res 67, 5195–5200.

41. Yang, M., Baranov, E., Wang, J. W., Jiang,P., Wang, X., Sun, F. X. et al. (2002) Directexternal imaging of nascent cancer, tumorprogression, angiogenesis, and metastasis oninternal organs in the fluorescent ortho-topic model. Proc Natl Acad Sci USA 99,3824–3829.

42. Li, L., Mignone, J., Yang, M., Matic, M.,Penman, S., Enikolopov, G., Hoffman, R.M. (2003) Nestin expression in hair folliclesheath progenitor cells. Proc Natl Acad SciUSA 100, 9958–9961.

43. Amoh, Y., Li, L., Yang, M., Moossa,A. R., Katsuoka, K., Penman, S., et al.(2004) Nascent blood vessels in the skin arisefrom nestin-expressing hair follicle cells. ProcNatl Acad Sci USA 101, 13291–13295.

44. Amoh, Y., Li, L., Yang, M., Jiang, P.,Moossa, A. R., Katsuoka, K., et al. (2005)Hair follicle-derived blood vessels vas-cularize tumors in skin and are inhib-ited by doxorubicin. Cancer Res 65,2337–2343.

45. Amoh, Y., Yang, M., Li, L., Reynoso,J., Bouvet, M., Moossa, A. R., et al.(2005) Nestin-linked green fluorescent pro-tein transgenic nude mouse for imaginghuman tumor angiogenesis. Cancer Res 65,5352–5357.

46. Amoh, Y., Bouvet, M., Li, L., Tsuji, K.,Moossa, A. R., Katsuoka, K., et al. (2006)Visualization of nascent tumor angiogen-esis in lung and liver metastasis by dif-ferential dual-color fluorescence imaging innestin-linked-GFP mice. Clin Exp Metastasis23, 315–322.

47. Amoh, Y., Li, L., Tsuji, K., Moossa, A. R.,Katsuoka, K., Hoffman, R. M., et al. (2006)Dual-color imaging of nascent blood vesselsvascularizing pancreatic cancer in an ortho-topic model demonstrates antiangiogenesisefficacy of gemcitabine. J Surg Res 132,164–169.

48. Amoh, Y., Nagakura, C., Maitra, A., Moossa,A. R., Katsuoka, K., Hoffman, R. M., et al.(2006) Dual-color imaging of nascent angio-genesis and its inhibition in liver metas-tases of pancreatic cancer. Anticancer Res 26,3237–3242.

49. Ji, Y., Hayashi, K., Amoh, Y., Tsuji, K.,Yamauchi, K., Yamamoto, N., et al. (2007)The camptothecin derivative CPT-11 inhibitsangiogenesis in a dual-color imageable ortho-topic metastatic nude mouse model ofhuman colon cancer. Anticancer Res 27,713–718.

50. Hayashi, K., Yamauchi, K., Yamamoto, N.,Tsuchiya, H., Tomita, K., Amoh, Y., et al.(2007) Dual-color imaging of angiogenesisand its inhibition in bone and soft tissue sar-coma. J Surg Res 140, 165–170.

51. Amoh, Y., Li, L., Katsuoka, K., Bouvet,M., Hoffman, R. M. (2007) GFP-expressingvascularization of Gelfoam R© as a rapid invivo assay of angiogenesis stimulators andinhibitors. Biotechniques 42, 294–298.

204 Hoffman

52. Hayashi, K., Yamauchi, K., Yamamoto, N.,Tsuchiya, H., Tomita, K., Bouvet, M.,Wessels, J., Hoffman, R. M. (2009) Acolor-coded orthotopic nude-mouse treat-ment model of brain-metastatic paralyzingspinal cord cancer that induces angiogen-esis and neurogenesis. Cell Proliferat 42,75–82.

53. Shaked, Y., Ciarrocchi, A., Franco, M., Lee,C. R., Man, S., Cheung, A. M., et al. (2006)Therapy-induced acute recruitment of circu-lating endothelial progenitor cells to tumors.Science 313, 1785–1787.

54. Yang, M., Jiang, P., Sun, F. X., Hasegawa,S., Baranov, E., Chishima, T., Shimada, H.,Moossa, A. R., Hoffman, R. M. (1999) Afluorescent orthotopic bone metastasis modelof human prostate cancer. Cancer Res 59,781–786.

55. Yang, M., Jiang, P., An, Z., Baranov, E., Li,L., Hasegawa, S., Al-Tuwaijri, M., Chishima,T., Shimada, H., Moossa, A. R., Hoffman, R.M. (1999) Genetically fluorescent melanomabone and organ metastasis models. Clin Can-cer Res 5, 3549–3559.

56. Cruz-Munoz, W., Man, S., Xu, P., Kerbel,R. S. (2008) Development of a preclinicalmodel of spontaneous human melanoma cen-tral nervous system metastasis. Cancer Res68, 4500–4505.

57. Hoffman, R. M. (2009) CommentRe: Preclinical model of spontaneousmelanoma metastasis. Cancer Res 69,719.

58. Yang, M., Luiken, G., Baranov, E., Hoff-man, R. M. (2005) Facile whole-body imag-ing of internal fluorescent tumors in micewith an LED flashlight. Biotechniques 39,170–172.

59. Yamauchi, K., Yang, M., Jiang, P., Xu, M.,Yamamoto, N., Tsuchiya, H., Tomita, K.,Moossa, A. R., Bouvet, M., Hoffman, R.M. (2006) Development of real-time subcel-lular dynamic multicolor imaging of cancercell trafficking in live mice with a variable-magnification whole-mouse imaging system.Cancer Res 66, 4208–4214.

60. Mayes, P. A., Dicker, D. T., Liu, Y., El-Deiry, W. S. (2008) Noninvasive vascu-lar imaging in fluorescent tumors usingmultispectral unmixing. Biotechniques 45,459–464.

61. Mansfield, J. R., Gossage, K. W., Hoyt, C. C.,Levenson, R. M. (2005) Autofluorescenceremoval, multiplexing, and automated analy-sis methods for in-vivo fluorescence imaging.J Biomed Opt 10, 41207.

62. Yang, M., Li, L., Jiang, P., Moossa, A. R.,Penman, S., Hoffman, R. M. (2003) Dual-color fluorescence imaging distinguishestumor cells from induced host angiogenicvessels and stromal cells. Proc Natl Acad SciUSA 100, 14259–14262.

63. Amoh, Y., Katsuoka, K., Hoffman,R. M. (2008) Color-coded fluorescentprotein imaging of angiogenesis: theAngioMouse models. Curr Pharm Des 14,3810–3819.

Chapter 12

Imaging Calcium Sparks in Cardiac Myocytes

Silvia Guatimosim, Cristina Guatimosim, and Long-Sheng Song

Abstract

Calcium ions play fundamental roles in many cellular processes in virtually all type of cells. The useof Ca2+ sensitive fluorescent indicators has proven to be an indispensable tool for studying the spatio-temporal dynamics of intracellular calcium ([Ca2+]i). With the aid of laser scanning confocal microscopyand new generation of Ca2+ indicators, highly localized, short-lived Ca2+ signals, namely Ca2+ sparks,were revealed as elementary Ca2+ release events during excitation–contraction coupling in cardiomy-ocytes. Since the discovery of Ca2+ sparks in 1993, the demonstration of dynamic Ca2+ micro-domainsin living cardiomyocytes has revolutionized our understanding of Ca2+-mediated signal transduction innormal and diseased hearts. In this chapter, we have described a commonly used method for record-ing local and global Ca2+ signals in cardiomyocytes using the fluorescent indicator fluo-4 acetoxymethyl(AM) and laser scanning confocal microscopy.

Key words: Calcium sparks, confocal microscopy, ventricular myocytes, fluorescence, calciumindicators.

1. Introduction

Cytosolic free Ca2+ ([Ca2+]i) is a versatile second messengerthat can simultaneously regulate multiple processes within anindividual cell. In the cardiac cell, cytosolic [Ca2+] ([Ca2+]i) isactively maintained at a very low level of around 100 nM byCa2+ homeostatic mechanisms, including the SR Ca2+-ATPase(SERCA) and the plasmalemmal Na+/Ca2+ exchanger (NCX)and Ca2+-ATPase and by a number of Ca2+ buffering molecules(1). During each heart beat, a time-dependent transient increasein intracellular Ca2+ concentration (“[Ca2+]i transient”) occursand is responsible for activating contraction in a process called


205

206 Guatimosim, Guatimosim, and Song

excitation–contraction coupling. The [Ca2+]i transient is trig-gered by the cardiac action potential (AP) and spreads throughthe heart as the AP is propagated (2).

[Ca2+]i transients during cardiac excitation–contraction (EC)coupling were first described as aequorin luminescence in frogcardiac muscle by David G. Allen (3) and in canine Pukinje fibersby Gil Wier (4) using microinjection of photoprotein into musclecells. The development of new fluorescent Ca2+ indicators, suchas fura-2 and indo-1 by Tsien and co-workers (5), had improved[Ca2+] measurements in single cardiac muscle cells. There aretwo advantages of these dyes. First, the acetoxymethyl (AM) esterderivatives of fura-2 and indo-1 can permeate the cell membrane,which makes the application of these Ca2+ indicators much easiercomparing to aequorin microinjection. Second, their unique exci-tation/emission features allow one to be able to make accurate,ratio measurement of intracellular Ca2+ concentrations. In con-junction with digital imaging technique, Wier et al. were the firstto document spontaneous [Ca2+] waves and different patterns ofsubcellular Ca2+ concentration in quiescent, spontaneously activeor hyper-contracting cardiomyocytes (6). However, the spatial-temporal resolution was not high enough for them to be able todetect the highly localized Ca2+ release events in cardiac myocytesas we can observe routinely nowadays.

Advances in Ca2+ fluorescence technology (still driven byTsien and his colleagues) combined with the advent of the laserscanning confocal microscope made it possible for the discov-ery of Ca2+ sparks in cardiac myocytes. Ca2+ sparks were firstlyreported in quiescent ventricular myocytes by Cheng et al. (7).Since then Ca2+ sparks or local Ca2+ release events with sparkcharacteristics have been recorded in skeletal muscle (8–10),smooth muscle (11), neurons (12) and more recently in fibrob-lasts (13). In heart muscle, Ca2+ sparks are now well accepted asthe elementary events of SR Ca2+ release underlying EC coupling,originated from the opening of a cluster of sarcoplasmic reticulum(SR) Ca2+ release channels or ryanodine receptors (RyRs). Ca2+

spark observed in unstimulated resting single cardiac myocytesrepresents a local transient increase in intracellular [Ca2+]i. It hasa rapid rise (∼10 ms, time to peak) and a moderately quick decaykinetics (∼20 ms, half-time of decay) and is confined to an areaof ∼2.0 μm in diameter or ∼8 fl by volume (14). At diastolic[Ca2+]i of about 100 nM, Ca2+ sparks are spontaneously firingat a very low rate (∼100 per second per cell). The occurrenceof spontaneous Ca2+ sparks does not require Ca2+ entry into thecardiomyocyte through L-type Ca2+ channels (LCCs) or by otherCa2+ pathways across the sarcolemma. However, during cardiacEC coupling, Ca2+ influx through LCCs evokes synchronousactivation of tens of thousands of Ca2+ sparks by a mechanism

Imaging Calcium Sparks in Cardiac Myocytes 207

called Ca2+-induced Ca2+ release (CICR) (15, 16). This processis locally controlled within a 12 nm junctional subspace betweent-tubular and SR membrane (17). It has been elegantly shownthat Ca2+ sparks can be triggered by adjacent single LCC open-ings (16, 18, 19). The summation of numerous Ca2+ sparks acti-vated simultaneously all over the myocyte compose a uniformCa2+ transient (see Fig. 12.1).

Fig. 12.1. a A train of steady-state Ca2+ transients elicited by 1-Hz field stimulation. b Spatially averaged Ca2+ profileshowing the dynamic change of Ca2+ signals with time. This panel also depicts the analysis of Ca2+ transient amplitudeand kinetics.

Fluo-3 and fluo-4 have been the indicators of choice in Ca2+

spark experiments, because of their unique properties that confera high signal-to-noise ratio, fast “on” and “off” kinetics and highsensitivity when the indicator responds to [Ca2+] gradients. Sinceits introduction in 1989, fluo-3 confocal Ca2+ imaging has madesignificant contribution to our understanding of spatial dynam-ics of many elementary process of Ca2+ signaling in different celltypes. Fluo-4, an analog of fluo-3, with higher quantum yieldwhen excited at 488 nm, provides brighter emission signals inresponse to Ca2+ binding when compared to fluo-3. When esti-mating [Ca2+]i from the observed fluorescence signal (F), a com-mon practice is to express the data as the ratio: R = F/F0, whereF0 refers to the baseline fluorescence at resting [Ca2+]i.

The major disadvantage of fluo-family of dyes (fluo-3/fluo-4), however, is that, upon binding of Ca2+ ions, there is little orno shift in its excitation or emission spectrum, which makes itimpossible to perform ratiometric measurements of [Ca2+] (20).Most chemical fluorescent indicators are cell impermeant, there-fore many of the fluorescent Ca2+ indicators are derivatized withAM ester groups. The AM form of the indicator can diffuseacross cell membranes, and once inside the cell, esterases cleave


the AM group off the probe leading to a cell-impermeant indica-tor. Because the AM derivative has low aqueous solubility, somedispersing agents such as Pluronic F-127 are often used to helpsolubilize large dye molecules in physiological media and facilitatecell loading (20).

2. Materials

1. Fluo-4 AM (10× 50 μg, F14201, Invitrogen). Store at−20◦C.

2. 20% Pluronic F127 in DMSO solution (P-3000MP, Invitro-gen). Store at room temperature.

3. Fluo-4 AM loading stock solution: dissolve 50 μg fluo-4AM with 50 μL 20% Pluronic F-127 DMSO solution. Storestock solution at −20◦C.

4. Tyrode’s solution with the following composition (mM):140 NaCl, 5 KCl, 5 HEPES, 1 NaH2PO4, 1 MgCl2, 1.8CaCl2 and 10 glucose (pH 7.4) adjusted with NaOH. Allsalts and buffers used for the preparation of normal Tyrode’ssolution can be purchased from Sigma-Aldrich R©. Store at4◦C.

5. Modified Dulbecco’s Modified Eagle Medium (DMEM):The basic medium routinely used to keep the isolatedadult myocytes is supplemented DMEM (powder purchasedfrom Sigma, catalog #D1152). To make up 50 mL ofmedia for incubating the cells, add 5 mL of inactivatedfetal bovine serum, 5 μL insulin (3.66 mg/mL) and 550μL NaCl (4 M) to 40 mL DMEM solution (made from0.87 g DMEM powder in 40 mL Milli-Q grade water).Adjust the pH with NaOH to 7.2 and complete the vol-ume. Then keep at room temperature for use on the sameday.

6. Electrical stimulator for field stimulation of myocytes, withthe capacity to deliver at least 20 V square pulses.

7. Perfusion chamber with attached platinum wires, mountedon the stage of a confocal microscope (see Note 1).

8. Confocal microscope equipped with an Argon laser of488 nm line for fluo-4 excitation and appropriate filters foracquiring emission signals at certain wavelength range (forexample, long pass filter that passes emission signals of wave-length >505 nm or band-pass filter that passes emission sig-nals of wavelength between 505 and 550 nm).


3. Methods

3.1. IsolatingVentricular Myocytesfrom Adult Rat

Adult rat ventricular cells will be prepared by standard meth-ods as previously described in the literature (21). Briefly, malerats weighing between 200 and 300 g will be sacrificed by lethalintraperitoneal injection of pentobarbital sodium (100 mg/kg).The hearts will be rapidly removed and perfused via the Langen-dorff apparatus with Ca2+-free modified Tyrode solution until theblood is washed out. Hearts will then be perfused with Tyrodesolution containing 50 μM CaCl2 along with 1.4 mg/mL colla-genase (type 2) and 0.04 mg/mL protease (type XIV) until theyare soft (approximately 10 min). The hearts will then be removedfrom the perfusion apparatus, minced into ~1-mm chunks andstirred for 4 min in Tyrode solution containing 50 μM CaCl2,0.7 mg/mL collagenase and 0.02 mg/mL protease. Cells willbe filtered through a 200 μm mesh to remove tissue chunksand extracellular Ca2+ concentration is raised to 0.5 mM over10 min through three centrifuge cycles (0.1 mM Ca2+, 0.2 mMCa2+, 0.5 mM Ca2+). Finally myocytes will be harvested andstored in modified DMEM until they are used (within 5 h)(22, 23).

3.2. Fluo-4 AMLoading

Add 10 μL fluo-4 AM stock solution to 1 mL of cell suspen-sion (final fluo-4 AM concentration = 10 μM). Cells shouldremain in the dark at room temperature for 20 min (see Notes2 and 3). Centrifuge the cells (2 min at 200–300 rpm centrifuga-tion), remove the supernatant and gently re-suspend the pellet inindicator-free Tyrode solution. Wait for 20 min to allow for com-plete de-esterification of AM esters. Then, cells will be ready forCa2+ spark imaging with confocal microscope for up to 2 h (seeNote 4). The anion-transport inhibitor probenecid (2 mM) maybe added to the cell solution to reduce leakage of the de-esterifiedindicator.

3.3. Ca2+ Imaging inVentricular Myocytes

Rod-shaped myocytes with clear striations and without activelyspontaneous contraction (less than one per minute) are consid-ered healthy Ca2+ tolerant cells and will be chosen for Ca2+ imag-ing (see Note 5). Ca2+ transients will be elicited by field stimu-lation through a pair of platinum electrodes, with a 2 ms supra-threshold square voltage pulse delivered by a commercially avail-able electrical stimulator (such as Myopacer 100, IonOptix Inc.).Cells are normally stimulated at 1 Hz for 15 s to let reach a steady-state condition before recording. An LSM 510 scanning system(Zeiss GmbH, Jena, Germany) equipped with a ×63 oil immer-sion objective (numerical aperture (NA) = 1.4) will be used forconfocal imaging of Ca2+ fluorescence (see Note 6). Fluo-4 will


be excited by 488 nm line of an Argon laser and emission signalsover 505 nm will be collected. The brightness of the fluorescentsignals represents the relative level of intracellular [Ca2+]i. Forrecording Ca2+ transients/sparks, a line scan mode is normallyutilized. The confocal pinhole is set to render spatial resolutionsof 0·4 μm in the horizontal plane and 0·9 μm in the axial direc-tion (see Note 7). Ideally, the detector gain is set at around 700(no digital gain). Line-scan images are acquired at sampling rateof 1.54 or 1.92 ms per line, along the longitudinal axis of the cell.Each line comprises 512 pixels spaced at 0.14 μm intervals. Aftera sequential scanning, a two-dimensional (2D) image of 512 ×1000 lines or 512 × 2000 lines will be generated and stored foroffline analysis (see below). It is not recommended to scan a cellin the same line region for prolonged time (see Note 8).

3.4. Recording Ca2+

Transients inVentricular Myocytes

Fluo-4 AM-loaded cells will be allowed to settle on a coatedglass coverslip by gravity (see Note 9). Wait for 5–10 min andthen turn the perfusion solution on; the cells will be bathed inTyrode’s solution. Cells can be paced with parallel platinum wiresconnected to an electrical stimulator (Myopacer 100, IonOptixInc.). Stimulation settings should be as follows: duration: 2 ms;continuous biphasic pulse stimulation; voltage: adjusted to 120%of the threshold voltage that induces Ca2+ transients. When cellsare stimulated at 1 Hz, the spatially averaged [Ca2+]i transientobtained by integrating the line-scan image should be similar tothat presented in Fig. 12.1 (see Note 10).

3.5. Recording Ca2+

Sparks in VentricularMyocytes

Spontaneous Ca2+ sparks may be recorded using the same con-focal settings used for Ca2+ transient imaging (e.g. line scanmode, laser power, pinhole size, detector gain, etc.). During sparkrecording, cells are kept in Tyrode’s solution under resting condi-tions (non-stimulated). For guinea pig, rabbit, canine and otherlarge mammalian heart cells, a 15-s field stimulation (1 Hz) isrequired to load the SR prior to spark recording (see Note 11).Soon after the halt of field stimulation, a series of line scan images(e.g. 6 sweeps; each sweep image can be 512 pixels × 1000 lines)will be acquired at a rate of 1.54 or 1.92 ms per line. For ratand mouse heart cells, spontaneous Ca2+ sparks may be visual-ized with or without pre-stimulation. However, it is more accu-rate to compare Ca2+ sparks recorded at steady-state conditions.Figure 12.2 shows a typical confocal line scan image of a Ca2+

spark recorded in a control rat ventricular myocyte.

3.6. Image Analyses Digital image processing will be performed by using custom-devised routines created with IDL programming language(Research Systems, Boulder, CO) (24). The Ca2+ level is reportedas F/F0 (or as �F/F0), where F0 is the resting Ca2+ fluorescence.


Fig. 12.2. Typical Ca2+ spark recorded from a control rat ventricular myocyte at resting condition, loaded with fluo-4AM. a The analysis of basic Ca2+ spark characteristics (amplitude, FDHM and FWHM). b A surface plot of the Ca2+ sparkshown in Panel a.

By using the following equation, we can convert the Ca2+ fluo-rescence ratio to a Ca2+ concentration (7):

[Ca2+]i = KR/{K/[Ca2+]rest + 1) − R}

where K is the dissociation constant of the Ca2+ indicator used,R is the fluorescence ratio (F/F0), [Ca2+]rest is the resting Ca2+

concentration. Assuming the dissociation constant (K) of fluo-4 AM is 400 nM (see Invitrogen Inc.), the resting [Ca2+]i of acardiac myocyte is 100 nM, the amplitude of a typical Ca2+ sparkis 2; we can then estimate that the peak Ca2+ concentration ofCa2+ spark is around 270 nM.

Figure 12.1 shows a typical train of Ca2+ transients from anormal rat ventricular myocyte. With the aid of computer pro-gramming analysis, the rising phase (time to peak, tpeak), theamplitude (F/F0) and the decay kinetics (t50, t75, t90) may beextracted from original Ca2+ images (25, 26).

Figure 12.2 displays a typical Ca2+ spark and illustrates theanalyses of key Ca2+ spark parameters: amplitude (F/F0), dura-tion (FDHM, full duration at half-maximal amplitude) and spa-tial width (FWHM, full-width at half-maximal amplitude). Theseparameters represent the basic gating properties of RyRs Ca2+


release channels: the release flux (F/F0) and the gating kinetics ofRyRs (FDHM). Ca2+ sparks of ventricular myocytes, on average,are about 1.8F/F0 in amplitude, 2 μm wide and 25 ms long (27).

4. Notes

1. The chamber bottom glass has to be no more than 170 μmthick (#1.5 glass coverslip). High numerical aperture lens(e.g. Zeiss Plan-Apochromat 63× oil immersion) have ashort working distance of 190 μm. We usually use 120 μmthick #1 glass coverslip).

2. Fluo-4 AM dye loading can vary, for example, rabbit ordog myocytes may require longer loading time. Myocyteswith an optimal fluo-4 AM loading shall report a baselinefluorescence of 30–40. Too high or too low baseline fluo-rescence levels indicate overloading or underloading of thefluorescent indicator into the cell.

3. Loading fluo-4 AM may be done at room temperature,rather than at 37◦C. Acetoxymethyl ester loading at hightemperature may often cause severe subcellular compart-mentalization of the indicator and may interfere with themeasurement of cytosolic Ca2+ concentration.

4. Cardiomyocytes can last for hours after Fluo-4 AM load-ing and still provide Ca2+ transient data; however, becauseof the leakage of esterified indicator (although slow),myocytes loaded with fluo-4 AM will exhibit dim fluo-rescence with time. Routinely, we examined cells for 2 hfollowing loading without marked deterioration of Ca2+

signals.5. Cardiomyocytes suitable for loading and measurement

should appear rod shaped, with clear striations and with-out cytoplasmic protrusions or blebs under phase-contrastlight microscopy.

6. Water immersion objective lens can be used. These objec-tive lens are designed for use with aqueous specimensand immersion medium and can be “corrected” for theunavoidable refractive index mismatch produced by the useof a glass coverslip (h = 1.51). Practically, oil immersionlens are often used for the much higher expense of waterimmersion lens.

7. The optical resolution of the confocal microscope is 0·4 μmin the horizontal plane and 0.9 μm in the axial direction,as determined by measuring the point spread function of


0·1 μm fluorescent bead (Molecular Probes – InvitrogenInc.).

8. Prolonged scanning in the same region may cause photo-bleaching of fluorescent molecules and experimental arti-facts induced by photo-damage.

9. Glass coverslip can be coated with laminin. Laminin pro-motes cell attachment, preventing stimulated contractionout of the focal plane during imaging. Laminin is dilutedto a final concentration of between 1 and 5 μg/mL in avolume of phosphate-buffered saline or culture medium,which adequately covers the culture surface. Lamininshould be applied to the coverslip at least 30 min beforeplating out cells.

10. Cells should be field-stimulated at 1.0 Hz for 15 s to reachsteady-state Ca2+ dynamics before image acquisition.

11. Myocytes of large mammals tend to unload their SR Ca2+

content, which makes Ca2+ sparks hard to be detected atresting conditions. One way to circumvent this problem isto use a conditioning protocol to upload Ca2+ into the SRbefore Ca2+ spark recording (28).

References

1. Bers, D. M., (2002) Cardiac excitation-contraction coupling. Nature 415, 198–205.

2. Guatimosim, S., Dilly, K., Santana, L. F.,Saleet, J. M., Sobie, E. A., Lederer, W. J.(2002) Local Ca(2+) signaling and EC cou-pling in heart: Ca(2+) sparks and the regula-tion of the [Ca(2+)](i) transient. J Mol CellCardiol 34, 941–950.

3. Allen, D. G., Blinks, J. R. (1978) Calciumtransients in aequorin-injected frog cardiacmuscle. Nature 273, 509–513.

4. Wier, W. G. (1980) Calcium transients dur-ing excitation-contraction coupling in mam-malian heart: aequorin signals of canine Purk-inje fibers. Science 207, 1085–1087.

5. Minta, A., Kao, J. P., Tsien, R. Y. (1989)Fluorescent indicators for cytosolic calciumbased on rhodamine and fluorescein chro-mophores. J Biol Chem 264, 8171–8178.

6. Wier, W. G., Cannell, M. B., Berlin, J. R,Marban, E., Lederer, W. J. (1987) Cellularand subcellular heterogeneity of [Ca2+]i insingle heart cells revealed by fura-2. Science235, 325–328.

7. Cheng, H., Lederer, W. J., Cannell, M. B.(1993) Calcium sparks: elementary eventsunderlying excitation-contraction couplingin heart muscle. Science 262, 740–744.

8. Klein, M. G., Cheng, H., Santana, L. F.,Jiang, Y. H., Lederer, W. J., Schneider, M.F. (1996) Two mechanisms of quantized cal-cium release in skeletal muscle. Nature 379,455–458.

9. Tsugorka, A., Rios, E., Blatter, L. A.(1995) Imaging elementary events of calciumrelease in skeletal muscle cells. Science 269,1723–1726.

10. Wang, X., Weisleder, N., Collet, C.,et al. (2005) Uncontrolled calcium sparksact as a dystrophic signal for mam-malian skeletal muscle. Nat Cell Biol 7,525–530.

11. Nelson, M. T., Cheng, H., Rubart, M.,et al. (1995) Relaxation of arterial smoothmuscle by calcium sparks. Science 270,633–637.

12. Ouyang, K., Zheng, H., Qin, X., et al.(2005) Ca2+ sparks and secretion in dorsalroot ganglion neurons. Proc Natl Acad SciUSA 102, 12259–12264.

13. Wei, C., Wang, X., Chen, M., Ouyang, K.,Song, L. S., Cheng, H. (2009) Calciumflickers steer cell migration. Nature 457,901–905.

14. Cheng, H., Lederer, W. J. (2008) Calciumsparks. Physiol Rev 88, 1491–1545.


15. Cannell, M. B., Cheng, H., Lederer, W. J.(1994) Spatial non-uniformities in [Ca2+]iduring excitation-contraction coupling incardiac myocytes. Biophys J 67, 1942–1956.

16. Cannell, M. B., Cheng, H., Lederer, W. J.(1995) The control of calcium release inheart muscle. Science 268, 1045–1049.

17. Franzini-Armstrong, C., Protasi, F., Ramesh,V. (1999) Shape, size, and distributionof Ca(2+) release units and couplons inskeletal and cardiac muscles. Biophys J 77,1528–1539.

18. Lopez-Lopez, J. R., Shacklock, P. S., Balke,C. W., Wier, W. G. (1995) Local calciumtransients triggered by single L-type calciumchannel currents in cardiac cells. Science 268,1042–1045.

19. Wang, S. Q., Song, L. S., Lakatta, E. G.,Cheng, H. (2001) Ca2+ signalling betweensingle L-type Ca2+ channels and ryan-odine receptors in heart cells. Nature 410,592–596.

20. Takahashi, A., Camacho, P., Lechleiter, J. D.,Herman, B. (1999) Measurement of intracel-lular calcium. Physiol Rev 79, 1089–1125.

21. Guatimosim, S., Sobie, E. A., dos Santos,C. J., Martin, L. A., Lederer, W. J. (2001)Molecular identification of a TTX-sensitiveCa(2+) current. Am J Physiol Cell Physiol 280,C1327–C1339.

22. Mitcheson, J. S., Hancox, J. C., Levi, A.J. (1998) Cultured adult cardiac myocytes:

future applications, culture methods, mor-phological and electrophysiological proper-ties. Cardiovasc Res 39, 280–300.

23. Mitra, R., Morad, M. (1985) A uni-form enzymatic method for dissocia-tion of myocytes from hearts and stom-achs of vertebrates. Am J Physiol 249,H1056–H1060.

24. Cheng, H., Song, L. S., Shirokova, N., et al.(1999) Amplitude distribution of calciumsparks in confocal images: theory and studieswith an automatic detection method. BiophysJ 76, 606–617.

25. Dias-Peixoto, M. F., Santos, R. A., Gomes,E. R., et al. (2008) Molecular mechanismsinvolved in the angiotensin-(1–7)/Mas sig-naling pathway in cardiomyocytes. Hyperten-sion 52, 542–548.

26. Song, L. S., Sobie, E. A., McCulle, S., Led-erer, W. J., Balke, C. W., Cheng, H. (2006)Orphaned ryanodine receptors in the fail-ing heart. Proc Natl Acad Sci USA 103,4305–4310.

27. Song, L. S., Guia, A., Muth, J. N.,et al. (2002) Ca(2+) signaling in cardiacmyocytes overexpressing the alpha(1) sub-unit of L-type Ca(2+) channel. Circ Res 90,174–181.

28. Song, L. S., Pi, Y., Kim, S. J., et al.(2005) Paradoxical cellular Ca2+ signaling insevere but compensated canine left ventricu-lar hypertrophy. Circ Res 97, 457–464.

Chapter 13

Light Microscopy in Aquatic Ecology: Methods for PlanktonCommunities Studies

Maria Carolina S. Soares, Lúcia M. Lobão, Luciana O. Vidal,Natália P. Noyma, Nathan O. Barros, Simone J. Cardoso,and Fábio Roland

Abstract

Planktonic organisms dominate waters in ponds, lakes and oceans. Because of their short life cycles,plankters respond quickly to environmental changes and the variability in their density and compositionare more likely to indicate the quality of the water mass in which they are found. Planktonic communityis formed by numerous organisms from distinct taxonomic position, ranging from 0.2 μm up to 2 mm.Despite others, the light microscopy is the most used apparatus to enumerate these organisms and differ-ent techniques are necessary to cover differences in morphology and size. Here we present some of themain light microscopy methods used to quantify different components of planktonic communities, suchas virus, bacteria, algae and animals.

Key words: Bacterioplankton, phytoplankton, limnology, enumeration, virioplankton,zooplankton.

1. Introduction

The general understanding of the word plankton is referred toany drifting organism (virus, bacteria, plants or animals) thatinhabit the pelagic zone of oceans, seas or bodies of freshwater.Planktonic community comprises organisms that are taxonomi-cally diverse and more defined by their ecological niche ratherthan their genetic classification. It includes:

• Virioplankton: the most diverse and abundant component ofplankton, infecting a wide range of planktonic organisms.


215

216 Soares et al.

• Bacterioplankton: morphological diverse group of prokary-otic organisms, which play an important role in re-mineralizing organic material down the water column.

• Phytoplankton: autotrophic, prokaryotic or eukaryotic organ-isms that live near the water surface where there is sufficientlight to support photosynthesis.

• Zooplankton: small protozoans or metazoans (e.g. crus-taceans and other animals) that feed on other plankton. Someof the eggs and larvae of larger animals, such as fish, crus-taceans and annelids, are included here.

These organisms are distributed along a wide range of size,varying from <0.2 μm, like some viruses (femtoplankton) toorganisms up to 20 mm (megaplankton) (Fig. 13.1). The knowl-edge related to abundance and composition of these communi-ties as well as the way these organisms interact is extremely rel-evant to understand important processes in aquatic ecosystems,such as biogeochemical cycles, primary production, harmful algaeblooms and global carbon budget. Studies in these communities

Fig. 13.1. Schematic plankton classification based on ecological niche (a) and size scale (b).

Light Microscopy in Aquatic Ecology 217

evolved together with microscopic techniques. The vast majorityof planktonic bacteria started to be revealed only with the use offluorescence microscopy and more recently, the discovery of theabundance of viruses in natural waters reflects the development ofdirect counting methods for bacterial enumeration. Other smallercomponents of plankton such as nanoflagellates and picoplanktonare still overlooked in many studies. On the other hand, due to itssize, techniques for phytoplankton and zooplankton estimationsare widely known and used, and these are probably the most well-known components of planktonic community. Although manyother techniques including molecular biology and transmissionelectronic microscopy have been widely used nowadays, the lightmicroscopy is still the primary source for qualitative studies inplanktonic organisms. Thus, this paper aims to group and describethe most used and available light microscopy techniques to studythe broad range of organisms from virioplankon to zooplankton.

2. Materials

2.1. Virus 1. Anodisc filters, 0.02 μm pore size, 25 mm diameter(aluminum oxide).

2. AA Milipore mixed-ester membrane filters, 0.8 μm poresize, 25 mm diameter.

3. Glass 22 mm diameter filter holder with 15 mL funnel,Millipore-type.

4. Vacuum source, suitable for at least 25 cm Hg vacuum.5. Glass microscope slides, 25 × 75 mm, standard thickness,

frosted at one end.6. Glass coverslips (25 mm × 25 mm).7. SYBR Green I nucleic-acid gel stain 10,000× concentrate

in anhydrous DMSO.8. p-phenylenediamine dihydrochloride or 1,4-

phenylenediamine dihydrochloride (Sigma).9. Glycerol.

10. Phosphate buffer saline (PBS, 0.05 M Na2HPO4, 0.85%(w/v) NaCl, pH 7.5).

11. 0.02 μm filtered-autoclaved ultra pure water.12. 0.02 μm filtered formalin (37–39% (w/v) saturated

formaldehyde solution).13. 0.02 μm filtered samples, formalin preserved, 2% (v/v)

final concentration.

218 Soares et al.

14. 2.0 mL clean and sterilized micro-centrifuge tubes.15. Pipettes suitable for 2–10 mL volumes.16. 5 and 10 mL clean, sterilized tips.17. 50 mL conical centrifuge tube.18. Glass Erlenmeyer filter flasks or multiple filter holder.19. Polystyrene Petri dish.20. Low-temperature dry-heat block with anodized aluminum

heat block.21. Non-fluorescent immersion oil for microscopy (refractive

index (n.d.) 1/4 = 1.516; Olympus).22. Epifluorescence microscope with high-pressure mercury

burner (HBO 100 W), 100× objective, 10× 10 ocu-lar grid, light filters as blue excitation (450–490 nmwide-bandpass) and green excitation (480–550 nm wide-bandpass). A minimum total magnification of 1000× isrequired, although 1250 or higher is preferred.

2.2. Bacteria 1. Polycarbonate membrane, black or stained with IrgalanBlack (1), 0.2 μm pore size, 2.5 mm diameter.

2. Glass 22 mm diameter filter holder with 15 mL funnel,Millipore-type.

3. Vacuum source, suitable for at least 25 cm Hg vacuum.4. Glass microscope slides, 25 × 75 mm, standard thickness,

frosted at one end.5. Glass coverslips (25 mm × 25 mm).6. Formaldehyde 37%, <0.2 μm filtered to a final concentration

of 2% (0.5 mL 37% formaldehyde per 10 mL sample).7. Acridine Orange stock solution 1 mg/mL; 0.2 μm filtered.

Storage cold and dark.8. Non-fluorescent immersion oil for microscopy (refractive

index (n.d.) 1/4 = 1.516).9. Epifluorescence microscope with high-pressure mercury

burner (HBO 100 W), 100× objective, ocular grid, lightfilters as blue excitation (450–490 nm wide-bandpass) andgreen excitation (480–550 nm wide-bandpass). A minimumtotal magnification of 1000× is required, although 1250×or higher is preferred.

2.3. Phytoplankton 1. Amber bottles.2. Lugol’s fixative: Dissolve 10 g I2 (pure iodine; toxic) and

20 g KI in 200 mL distilled water and 20 mL concentratedglacial acetic acid. Store in darkened bottle.

3. Sedimentation chambers of various volumes.


4. Inverted microscope with 40× objective, equipped with anocular grid.

2.4. Zooplankton 1. Glass bottle.2. Neutralized formalin solution (40% formaldehyde) to pro-

duce a final concentration of about 4%.3. Automatic volumetric pipette or a Hansen–Stempel pipette.4. Sedgwick-Rafter Counting Chamber.5. Microscope with 10–40× objectives, equipped with an ocu-

lar grid.

3. Methods

3.1. VirusQuantification

About 30 years ago, fluorescence microscopy became the stan-dard method for counting planktonic prokaryotic cells collectedon blackened polycarbonate filters. SYBR Green I and SYBR Goldare notable stains to enumerate microorganisms (2). The first oneis an extremely sensitive cyanine-based fluorescent dye that bindsto double-stranded DNA (dsDNA) and RNA. It is commonlyused for nucleic-acid gel staining due to its high sensitivity super-bright fluorescence and low background, and is considered a less-mutagenic alternative to ethidium bromide (3). SYBR Green I hasbeen used in both fluorescence microscopy and flow-cytometrycounts of viruses (Fig. 13.2a).

1. Rinse each tube (50 mL centrifuge tube) three times withthe samples. Fix samples with filtered formalin immediatelyafter collecting to 2% (v/v) final concentration. Use briefinversion of the sampling container to mix in the fixativeand then place it on ice. Slides need to be prepared within4 h; if it is not possible, samples should be stored at −80◦C(see Note 1).

2. Filter 2 mL of each sample with replicates. If 2 mL is toomuch, dilute the sample with 0.02 μm filtered samples pre-served with 2% (v/v) formalin. Work through the completeslide-preparation procedure to ensure the volume of samplesbeing filtered is appropriate for enumeration. It is impor-tant to know the exact sample volume filtered. Place samplesimmediately in ice.

3. Clean the filter tower with filtered ultra pure water andethanol. Moisten the 0.8 μm pore size, 25 mm diameterwith filtered H2O and put it onto the center of the filterholder. Put a 0.02 μm pore size, 25 mm diameter Anodiscfilter on top of the first one. Using a pipette to transfer thedesired volume of sample to the filter funnel, turn the vac-uum pump to B20 kPa or 20 cm Hg VAC.

220 Soares et al.

Fig. 13.2. Main planktonic groups imaged by light microscopy after staining with SYBR Green I (a, Virioplankton) andacridine orange (b, Bacterioplankton) and fixing with lugol’s solution (c–g, phytoplankton) and formalin (h–l, zooplankton).In (a), virus-like-particles are encircled; other particles are indicated by arrows. In b, bacterial cells are encircled andpicoplanktonic cells indicated by arrows, c cyanobacteria, d and e diatoms, f cryptomonads, g desmid, h and i rotifers,j copepod, k and l cladocerans. Bars, 10 μm (a); 10 μm (b); 30 μm (c); 5 μm (d); 10 μm (e–g); 25 μm (h, k);50 μm (i, l); 0.5 mm (j).


4. After the water has filtered through, leave the vacuum pumpon and carefully remove the clamp and filter funnel. Thesame 0.8 μm filter can be used for multiple samples. The0.02 μm filter needs to be dried before the stain process.While filters are drying, use a pipette and tip to dispense10 μL droplet of the working concentration SYBR GreenI reagent onto the middle of each Petri dishes labeled withthe number of samples. Once filters are dry, place each filteronto the 10 μm SBYR reagent dropped backside down ineach labeled Petri dishes. Microorganisms on the topside ofthe filter will be stained by the reagent, which will easily passthrough the filter from the underside to the top. Stain eachfilter for 18 min, keeping the stained filters in the dark.

5. Put the dried Anodisc filter (0.02 μm) onto the labeled glassmicroscope slide. To mount the filter, use a pipette and tipto dispense 27–30 μL of 0.1% (v/v) p-phenylenediamineanti-fade mounting medium onto a 25× 25 mm glasscoverslip.

6. To enumerate the virus, use a 100× fluorescence oil-immersion objective with immersion oil. Put a drop of oilonto the center of a coverslip and move the objective downto the drop of oil. SYBR Green I is excited with maximumat 488 nm. Using a 10× 10 ocular grid, count between 30and 40 boxes for a total of 10 fields by randomly moving thestage to a new position around the entire filter. The total for10 fields should be >200 viruses counted.

7. Calculate the total density (number of virus per mL) fromconversion factor: (NV × 100)/(n × RSF)/V, where NVis the average number of viruses per field, n is the numberof boxes (of 100 total) counted on the microscope eyepiecegrid per field, RSF is the grid-reticle scaling factor (repre-senting the ratio of the filtered area to the 10 × 10 gridvisible in the eyepiece) and V is the volume of the filteredsample (mL).

3.2. BacteriaEnumeration

Bacteria enumeration is performed under fluorescencemicroscopy equipped with specific light filters in accordancewith the staining used. Bacteria are stained on membranes usingfluorochromes. Acridine Orange and DAPI (4′,6′-diamidino-2-phenylindole) have been the most used stains for bacteriaenumeration by aquatic microbiologists (see Note 2). Theepifluorescence microscopy method was introduced by Hobbieet al. (2), which led to a better understanding of bacterioplanktonrole to the aquatic microbial food web. Fluorescence microscopyhas been used to determine bacterial biomass and bacteriaphylogenetic identification (see Note 3) in addition to bacteriacounting (see Note 4) (Fig. 13.2b).

222 Soares et al.

1. Fix samples immediately following sampling with free-particle formaldehyde (see Section 2.2, Item 6). Fixed sam-ples should be kept in the dark and slides should be preparedsoon to avoid cell degradation.

2. Moisten the filter unit with particle-free water; place the1.2 μm supporting filter and subsequently the 0.2 μm blackmembrane. The supporting filter can be used several times.Rinse the filter unit funnel with ultra pure water and place iton top of the filters and fasten it with the clamp.

3. Pour known volume of sample on 2.5 mm unit membranefilter and add Acridine Orange at a final staining concentra-tion of 0.01%. Wait for 1 min (3) and then gently filter. Referto Note 5 for sample volume adjustment.

4. Filter samples at a maximum vacuum of 100 mmHg to avoidcell damage until whole samples pass through the filter. Letthe suction continue and air flow through the filters whileremoving the funnel and let the filter dry a little. The filtershould be moist but not wet.

5. Place the filter on a glass slide using a small drop of immer-sion oil. Put on the central top of black membrane a smalldrop of immersion oil and mount with a clean coverslip,pressing down firmly in a folded paper towel until the oilmoves out of the edges of the coverslip. Store frozen andprotected from light.

6. Thoroughly rinse funnels with particle-free water betweensamples.

7. Count at least 10 grids randomly spread over the filter. Aminimum of 200 cells should be counted. This reduces thestandard deviation and provides a trusty mean (4). WhenAcridine Orange is used, bacteria should be counted on blueexcitation (U-MSWB2 450–490 nm wide-bandpass). Cellsshine in orange and green colors. Heterotrophic cells andautotrophic cells may be differed by using green excitationfilter (U-MSWG2 480–550 nm wide-bandpass), at whichonly autotrophy cells shine (5).

8. Calculate bacterial cell number from the following conver-sion factor: (n × A)/(V × a) where,n is the mean number of bacterial cells, A is the filter area,a is the area of the counting grid, V is the volume of sam-ple collected × volume of sample filtered/(volume of thecollected sample + volume of the fixative).

3.3. PhytoplanktonEnumeration

For phytoplankton enumeration, a whole water sample must becollected (unfiltered and unstrained) (Fig. 13.2c–g). Sedimen-tation is the preferred method for phytoplankton concentration,because it is nonselective (unlike filtration) and nondestructive(unlike filtration or centrifugation). For enumeration, several


alternative counting chambers can be used (Sedwick–Rafter,Lund, Palmer-Maloney). The commonly used Sedgewick–Rafterchamber is the most suitable for counting large algae at low con-centration, once high magnification objectives cannot be used.Here we discuss sedimentation and counting techniques using theinverted microscope (6) which is the better suited equipment forquantitative analysis of multispecies present in natural samples.Whenever possible, phytoplankton species should be examinedalive, particularly delicate species of flagellated algae.

1. Fix the samples in Lugol’s solution, at a final concentrationof 1:100 (see Note 6). Samples must be stored in dark glass-ware protected from light.

2. After mixing with the Lugol’s solution, gently add the sam-ple into a sedimentation chamber. Refer to Note 7 for sedi-mentation chamber volumes and care.

3. Fill the chamber to the top edge with sufficient excess to per-mit the water to “bead” upward above the edge. Slide thechamber cover cap across the top of the cylinder to removeany excess of water and to enclose the sample of exact vol-ume without entrapping any air bubbles. The time to ensurecomplete sedimentation of all organisms in hours should beat least three times the height in cm of the sedimentationchamber (7). Keep the chamber on a vibration and light-freearea.

4. Remove the extra volume and mount the chamber with anadequate coverslip.

5. Enumerate organisms using an inverted microscope. Beforecounting, the entire chamber must be analyzed to ensurethe homogeneous sedimentations of cells. Individual (cells,colonies, coenobium, filaments) are enumerated in randomnon-overlapping fields, previously defined (8). At least 100individuals of the dominant taxa must be counted (error<20%, p < 0.05) (9). When organisms lie across the marginareas, those along the upper and right edges (or lower andleft) must not be counted. Empty diatoms, dinoflagellates orother damaged cells are not counted.

6. Calculate the total density from the following conversionfactor: [field surface (μm2) × chamber length (mm) × num-ber of fields]/109. Cell volume can be calculated for eachspecies by comparing cellular morphology with solid geo-metric shapes most closely related (10).

3.4. Zooplankton Zooplankton samples often need to be concentrated in the fieldusing plankton net or a pump. The sampling success will largelydepend on the selection of a suitable gear; mesh size of nettingmaterial, time of collection, water depth of the study area andsampling strategy. After sampling, fixation should be carried out

224 Soares et al.

as early as possible preferentially within 5 min after the collectionto avoid damage to animal tissue by bacterial action and autolysis.The methodology described here is applied to large zooplanktongroups, such as rotifers (Fig. 13.2 h–i), microcrustaceans – cope-pods (Fig. 13.2j) and cladocerans (Fig. 13.2 k, l).

1. Depending on the used concentration, dilute the concen-trated sample of zooplankton to exactly 50, 100 or 200 mLwith tap water. Use a graduated cylinder.

2. Mix organisms thoroughly within the graduated cylinder(vortex mixer or magnetic stirrer) and immediately obtaina subsample. Work rapidly to minimize the effect of organ-ism settling.

3. Obtain a subsample of 1 mL with an automatic volumetricpipette. Use a wide-mouth pipette so as not to restrict uptakeof large zooplankton or a Hansen–Stempel pipette.

4. Add this subsample to Sedgwick–Rafter chamber whichholds exactly 1 mL.

5. Count all organisms in number of subsamples definedaccording to the concentration of organisms in the subsam-ples (2–5).

6. Calculate the total density from the following conversionfactor: ind. L−1 = average count per cell ± standard devi-ation of the mean × volume of sample (mL)/volume of lakewater filtered (L). The biovolume is estimated by measur-ing cells’ dimension and by approximation of a geometricshape (11–13). In the case of microcrustaceans (cladoceraand copepoda) is commonly used the value of dry weight,which is measured in an analytical microbalance (14, 15).

3.5. Identificationof Damaged Cells

In natural populations, an important fraction of bacteria and phy-toplankton cells are compromised or dying whenever lysis ratesare important. Indeed, recent works have demonstrated that lysisis an important process not only in oceans (16) but also inlakes (17). The identification of damaged cells requires appro-priate methods to reliably discriminate living from dead or com-promised cells in natural phytoplankton. A rapid procedure toidentify damaged cells can be done using tests for viability suchas the LIVE/DEAD BacLight Bacterial Viability kit (MolecularProbes) to estimate total counts of viable/damage of both bacte-ria (18) and cyanobacteria (19, 20). This kit utilizes a mixtureof SYTO R© 9 green-fluorescent nucleic-acid stain and the red-fluorescent nucleic-acid stain, propidium iodide. These stains dif-fer in their ability to penetrate into healthy cells. When both dyesare used together, SYTO 9 labels viable bacteria, while propidiumiodide penetrates into bacteria with damaged membranes. Meth-ods for viability evaluation of other phytoplanktonic organismsare also available (21).


4. Notes

1. If samples were stored at −80◦C, certify that the samples’temperature reach the room temperature (25◦C) before thefiltration.

2. The specific fluorescent probe DAPI (4′6′-diamidino-2-phenylindole) can also be used by the aquatic ecologist (22).When DAPI is used, the bacteria quantification is done onUV filters (U-MNU2, BP 330–385 nm, Olympus).

3. One common way to estimate bacterial biomass is to usea computer-assisted image analysis system. It consists ofusing a microscope of epifluorescence with a video cam-era, a computer equipped with video acquisition capabilitiesand appropriate software for image processing. In our case,we use the software Image Pro-Plus 5.0. Samples are pre-pared like described for enumeration of total bacteria andbiomass is determined by measure geometric formula. Theimage must be passing a following application of filter inorder to improve it, according to the following sequence(23): Gauss filter (kernel 5 × 5) followed by Laplace filter(kernel 5 × 5) and a median filter (rank 3), the latter runthree times. After calculating a three-dimensional param-eter volume, using the formula V = π/4 × W 2(L − W /3)(V= volume, L = Length, W = Width), the result is appliedin the formula (24): m = CV α, where, C = 120 fg C, V =volume of cell and a = 0.76.

4. Fluorescence in situ hybridization (FISH) with rRNA-targetprobes is another staining technique that allows phyloge-netic identification of bacteria in mixed assemblages (25)through fluorescence microscope. Such procedure is rou-tinely used in the analysis of bacterioplankton, althoughmust be necessary that other excitation filters besides blueand green to this approach.

5. The volume of the bacteria samples for staining must beadjusted in accordance to the bacterial abundance andthis can be determined by trial and error. One or twomilliliter has been the most usual sample amount for naturalsamples.

6. Lugol is a recommended fixative for phytoplankton. It actsnot only on the cell preservation but also accelerating thesedimentation process within the chamber, since it increasescells’ weight.

7. Sedimentation chambers are available in volumes of 1, 5, 10,25, 50 and 100 mL. The concentrated volume is related tothe sample turbidity and varies inversely to the amount of

226 Soares et al.

organisms. The sedimentation chambers and their coverslipsmust be cleaned thoroughly to avoid contamination withorganisms from previous samples.

References

1. Hobbie, J. E., Daley, R. J., Jasper, S. (1977)Use of nucleopore filters for counting bacte-ria by flurescence microscopy. Appl EnvironMicrob 49, 1225–1228.

2. Noble, R. T., Fuhrman, J. A. (1998) Useof SYBR Green I for rapid epifluorescencecounts of marine viruses and bacteria. AquatMicrob Ecol 14, 113–118.

3. Patel, A., Noble, R. T., Steele, J. A., Schwal-bach, M. S., Hewson, I., Furhman, A.(2007) Virus and prokaryote enumerationfrom planktonic aquatic environments by epi-fluorescence microscopy with SYBR Green I.Nat Protoc 2, 269–276.

4. Kirchman, D. J. (1993) Statistical analysisof direct counts of microbial abundance, in(Kemp, P. F., Sherr, B. F., Sherr, E. B.,Cole, J. J. eds.), Handbook of Methods inAquatic Microbial Ecology. Lewis Publishers,Boca Raton, FL, pp. 117–119.

5. Daley, R. J, Hobbie, J. E. (1975) Directcounts of aquatic bacteria by a modified epi-fluorescence technique. Limnol Oceanogr 20,875–882.

6. Utermöhl, H. (1958) Zur Vervolkomnungder quantitativen phytoplankton–methodik.Mitt Int Ver Theor Angew Limnol 9,1–38.

7. Margalef, R. (1969) Counting, in (Vollen-weider, R. A., Talling, J. F., Westlake, D.F. eds.). A Manual on Methods for Measur-ing Primary Production in Aquatic Environ-ments (I.B.P. Handbook 12). Blackwell Sci-ent. Publ., Oxford and Edinburgh, pp. 7–14.

8. Uhelinger, V. (1964) Étude statistiquedes methods de dénobrement planctonique.Arch Sci 17, 121–223.

9. Lund, J. W. H., Kipling, C., Lecren, E. D.(1958) The inverted microscope method ofestimating algal number and statistical basisof estimating by counting. Hydrobiologia 11,143–170.

10. Hillebrand, H., Dürselen, C. D., Kirschtel,D., Pollingher, U., Zohary, T. (1999) Bio-volume calculation for pelagic and benthicmicroalgae. J Phycol 35, 408–424.

11. Gates, M. A., Rogerson, A., Berger, J. (1982)Dry to wet biomass conversion constant forTetrahymena Elliot (Ciliophora, Protozoa).Oecologia 55, 145–148.

12. Ruttner-Kolisko, A. (1977) Suggestions forbiomass calculation of planktonic rotifers.Arch Hydrobiol 8, 71–76.

13. Pauli, H. R. (1989) A new method toestimate individual dry weights of rotifers.Hydrobiologia 186/197, 355–361.

14. Dumont, H. J, Van De Velde, I., Dumont, S.(1975) The dry weight estimate of biomassin a selection of cladocera, copepoda androtifera from the plankton, periphyton andbenthos of continental waters. Oecologia 19,75–97.

15. Manca, M., Comoli, P. (1999) Studies onzooplankton of Lago Paione Superior. J Lim-nol 58, 131–135.

16. Agustí, S., Sánchez, C. (2002) Cell viabilityin natural phytoplankton communities quan-tified by a membrane permeability probe.Limnol Oceanogr 47, 818–828.

17. Agustí, S., Alou, E., Hoyer, MV., Frazer, T.K., Canfield, D. E. (2006) Cell death in lakephytoplankton communities. Freshwater Biol51, 1496–1506.

18. Freese, H. M., Karsten, H. M., Schumann, R.(2006) Bacterial abundance, activity, and via-bility in the eutrophic river Warnow, North-east Germany. Microb Ecol 51, 117–127.

19. Dagnino, D., de Abreu Meireles, D., deAquino Almeida, J. C. (2006) Growth ofnutrient–replete microcystis PCC 7806 cul-tures is inhibited by an extracellular sig-nal produced by chlorotic cultures. EnvironMicrobiol 8, 30–36.

20. Lengke, M. F., Ravel, B., Fleet, M. E.,Wanger, G., Gordon, R. A. S., Southam,G. (2006) Mechanisms of gold bioaccumula-tion by filamentous cyanobacteria from Gold(III)-Chloride complex. Environ Sci Technol40, 6304–6309.

21. Schumann, R., Schiewer, U., Karsten, U.,Rieling, T. (2003) Viability of bacteria fromdifferent aquatic habitats. II. Cellular fluo-rescent markers for membrane integrity andmetabolic activity. Aquat Microb Ecol 32,137–150.

22. Poter, K. G, Feig, Y. S. (1980) The use ofDAPI for identifying and counting aquaticmicroflora. Limnol Oceanogr 25, 943–948.

23. Massana, R., Gasol, M. J., Bjørnsen, P. K.,Black-Burn, N., Hagström, A., Hietanen, S.,Hygum, B. H., Kuparinen, J., Pedrós-Alió,C. (1997) Measurement of bacterial size viaimage analysis of epifluorescence prepara-tions: description of an inexpensive systemand solutions to some of the most commonproblems. Scientia Marina 61, 397–407.


24. Norland, S. (1993) The relation betweenbiomass and volume of bacteria, in (Kemp,P. F., Sherr, B. F., Sherr, E. B., Cole, J.J. eds.), Handbook of Methods in AquaticMicrobial Ecology. Lewis, Boca Raton, FL,pp. 303–308.

25. Cottrell, M. T., Kirchman, D. L. (2000)Community composition of marine bac-terioplankton determined by 16 s RRNAgene clone libraries and fluorescence in situhybridization. Appl Environ Microbiol 66,5116–5122.

Chapter 14

Fluorescence Immunohistochemistry in Combination withDifferential Interference Contrast Microscopy for Studiesof Semi-ultrathin Specimens of Epoxy Resin-EmbeddedSamples

Shin-ichi Iwasaki and Hidekazu Aoyagi

Abstract

We have developed a technique, using a combination of immunofluorescent staining of semi-ultrathinsections of epoxy resin-embedded samples and the corresponding differential interference contrast (DIC)images obtained by light microscopy that provides detailed information about the immuno-localizationof histological and cellular structures. To demonstrate the effectiveness of our method, we examined theimmunofluorescence of immuno-stained keratin 13 (K13) and type III collagen (CIII) and the corre-sponding DIC images during the morphogenesis of filiform papillae on the rat tongue. Immunoreactivityspecific for K13 and CIII was detected on the lingual epithelium of juveniles on postnatal days 7 and 14(P7 and P14). The immunoreactivity specific for K13 was clearly located in the intermediate-layer cells ofthe interpapillary cell columns, while that specific for CIII was also distinct in the connective-tissue fibersbetween the lingual epithelium and the lingual muscle. The DIC images revealed the keratinization of thestratified squamous cells of the lingual epithelium and, also, myogenesis beneath the connective tissue. Inaddition, immunoreactivity specific for CIII was also recognizable in the endomysium and perimysiumaround the lingual muscle. Thus, our method demonstrated changes in patterns of immunoreactivity ofK13 and of CIII during the morphogenesis of the rat tongue.

Key words: Fluorescence immunohistochemistry, differential interference contrast (DIC), keratin13, type III collagen, lingual mucosa, morphogenesis.

1. Introduction

In our analyses of morphogenesis of the mammalian tongue, wehave found that it is impractical to use semi-ultrathin sectionsfor light microscopy after immunohistochemical staining, because


229

230 Iwasaki and Aoyagi

the histological structures on semi-ultrathin sections cannot bedistinguished by standard counter-staining. To solve this prob-lem, we have developed a technique that exploits a combinationof immunofluorescence staining of semi-ultrathin sections afterremoval of epoxy resin (1) and the corresponding differentialinterference contrast (DIC) images obtained by light microscopy.Haraguchi and Yokota (2) were the first to describe a similarmethod, including treatment of sections with 10% sodium ethox-ide to remove epoxy resin (3) and reaction with fluorescence-labeled second antibodies. In our method, we use 10% sodiummethoxide to remove epoxy resin from sections (1). There is nosignificant difference between the two treatments, but Haraguchiand Yokota (2) used fluorescence-labeled second antibodies whilewe used a combination of biotin-conjugated second antibody andstreptavidin fluorescence. Our preliminary experiments indicatedthat the latter system had greater sensitivity than the former inthe case of semi-ultrathin sections. Using our above-mentionedmethod, we are now easily able to detect immunofluorescence onsemi-ultrathin sections of epoxy resin-embedded specimens.

To demonstrate the utility of our method, we examinedthe tongues of juvenile rats. In prior unsuccessful experiments,counter-staining with hematoxylin and eosin of semi-ultrathinsections barely revealed any details of the histology and cell mor-phology. In the present study, we used DIC images to examine thehistology of the same specimens as those in which we had moni-tored the fluorescence of Alexa Fluor 488 and 633. By combiningimages, we were able to clearly define the histological location ofkeratin 13 (K13) and the extracellular distribution of type III col-lagen (CIII). Our method should be applicable to various kindsof tissue and cell of which only very small amounts are available.

2. Materials

2.1. Rat Tongues 1. Sprague–Dawley rats (SPF; Japan SLC, Hamamatsu, Japan)were used for all observations.

2. Tongues were removed from rats on P7 and P14 after theyhad been killed by an intraperitoneal overdose of sodiumpentobarbital (120 mg/kg body weight).

2.2. Preparationof Tissue

1. 4% formaldehyde titrated from paraformaldehyde in 0.1 Mphosphate buffer.

To prepare this solution: 25 mL of distilled water(DW) are heated to almost boiling pint. Two grams ofparaformaldehyde (fine granules) are added to the hot DWand are dissolved by gentle shaking. Upon addition of one

Fluorescence Immunohistochemistry with Differential Contrast Microscopy 231

to three drops of 1 N NaOH, the 2 g of paraformaldehydeare completely dissolved in DW. Finally, a small amountof DW is added to give a total volume of 25 mL. Thephosphate-buffered saline (PBS) is prepared in as 10× stocksolution that contains 1.37 M NaCl, 27 mM KCl, 100 mMNa2HPO4, 18 mM KH2PO4 (adjusted to pH 7.4 with HClif necessary) and stored at 4◦. A working solution is pre-pared by dilution of one part stock with four parts water. Mixthe solution of paraformaldehyde with 2× PBS just beforeuse. To prepare 10× stock solution of PBS: 1.37 M NaCl,27 mM KCl, 100 mM Na2HPO4, 18 mM KH2PO4 (adjustto pH 7.4 with HCl if necessary) and store at 4◦. Preparea working solution by dilution of one part stock with nineparts water.

2. An ascending ethanol series for dehydration of specimens(see Note 1)

3. Epoxy resin (Epon 812; TAAB, Berks, UK) for embeddingof specimens. Solution A: 62 mL Epon 812 and 100 mLDDSA. Solution B: 100 mL Epon 812 and 89 mL MNA.Mix equal volumes of solutions A and B, adding 1.5–1.8%DMP-30 for polymerization.

4. An ultramicrotome (MT-XL; RMC, Tucson, AZ, USA) witha diamond knife for semi-ultrathin sectioning.

5. Glass slides (Matsunami Glass Ind., Ltd., Osaka, Japan) formounting of specimens.

6. 10% sodium methoxide for removal of epoxy resin (1). Add20 g of sodium metal to 200 mL of absolute methanol ina 1000-mL flask because bubbling will occur for approxi-mately 20 min. Add methanol to a volume of 200 mL tocompensate for loss due to evaporation. After adding 200mL of benzene, add 50–100 mL of methanol until completefusion of methanol and benzene occurs.

7. A descending acetone series (see Note 2) for transfer of spec-imens to PBS.

2.3. PrimaryAntibodies

1. Mouse monoclonal antibodies specific for K13, which hadbeen purified from human esophagus, were purchased fromProgen Biotechnik GmbH (Heidelberg, Germany).

2. Rabbit polyclonal antibodies against CIII, which had beenpurified from rat skin, were purchased from Chemicon Inter-national (Temecula, CA, USA).

2.4.ImmunofluorescenceStaining of K13and CIII

1. 10 mM sodium citrate buffer (pH 6.0): prepare a 20× stocksolution with 20 mL of 200 mM citric acid, 80 mL of200 mM sodium citrate. Prepare the working solution bydilution of one part stock with nineteen parts water.


2. Phosphate-buffered saline (PBS; pH 7.4)3. Biotin-conjugated rabbit antibodies against mouse IgG, IgA

and IgM4. Biotin-conjugated antibodies raised in goat against rab-

bit IgG5. Streptavidin AlexaFluor 4886. Streptavidin AlexaFluor 6337. FluoroGuardTM antifade reagent8. Glass coverslips

2.5. Light Microscopy(Axioplan; Carl Zeiss)

1. All specimens were examined with an Axioplan fluorescencemicroscope (Carl Zeiss, Jena, Germany) that was equippedwith a mercury vapor lamp (HBO 103 W).

2. Images were digitized using Zeiss AxioVision 3.1 softwareand resolution of 1300 × 1030 pixels.

3. For detection of the fluorescence of Alexa Fluor 448 (peakexcitation, 495 nm; peak emission, 519 nm), we used lightfrom a mercury vapor lamp and the filter set for fluo-rescein isothiocyanate (FITC; excitation filter, BP485/20;beamsplitter, FT510; emission filter, BP515-565). Similarly,for Alexa Fluor 633 (peak excitation, 632 nm; peak emis-sion, 647 nm), we used the filter set for Cy3 (excita-tion filter, BP565/30; beamsplitter, FT585; emission filter,BP620/60) (see Note 3).

4. A combination of pixel sizes (450 μm/1300) × (357 μm/1030) and a 20× objective lens with NA (numerical aper-ture) = 0.5 was used for observations.

5. We examined the differential interference contrast (DIC)image of each specimen by light microscopy.

3. Methods

Using our technique, we were able to easily detect and local-ize immunofluorescence in the tongues of juvenile rats at earlypostnatal stages. The results obtained by the present methodwere almost identical to those obtained by confocal laser-scanningmicroscopy (4–11). In the present study, we used DIC imagesobtained by light microscopy to examine the same specimens asthose in which we monitored the fluorescence of Alexa Fluor 488


or 633. We were able to define the histological location of K13and CIII by combining the respective images.

3.1. Preparationof Lingual Tissue

1. Tongues, taken from juvenile rats on postnatal days 7 and 14(P7 and P14) were fixed in 4% formaldehyde, titrated fromparaformaldehyde, in 0.1 M PBS (pH 7.4) at 4◦ for 5 h.

2. After rinsing in 0.1 M PBS, all samples were dehydrated inan ascending ethanol series (see Note 1) and embedded inepoxy resin (Epon 812), which was allowed to polymerizeovernight at 60◦.

3. Epoxy resin-embedded samples were cut into 500-nm sec-tions with a diamond knife on an ultramicrotome.

4. The sections were mounted on glass slides and incubated in10% sodium methoxide for 2 min at room temperature toremove the epoxy resin (1).

5. After passage through an acetone series (see Note 2), sec-tions were finally transferred to PBS (pH 7.4).

3.2.ImmunofluorescenceStaining

1. After retrieval of antigens by heating in a microwave oven at500 W for 2 min in 10 mM sodium citrate buffer, pH 6.0(12), sections on slides were allowed to cool for 8 min andwere then transferred to PBS (pH 7.4) at room temperature.

2. Sections were then incubated with primary antibodiesovernight at 4◦. To determine the optimum working dilu-tions of preparations of antibodies, we tested dilutions from1:25 to 1:800.

3. After washing in PBS, sections were incubated with biotin-conjugated antibodies raised in rabbit against mouse IgG,IgA and IgM or with biotin-conjugated antibodies raised ingoat against rabbit IgG for 30 min at room temperature.

4. Sections were incubated with streptavidin-Alexa Fluor 488or 633 for 30 min at room temperature.

5. Sections were then mounted with FluoroGuardTM antifadereagent (Bio-Rad Laboratories) after washing in PBS. Eachspecimen was covered with a glass coverslip, which was fixedwith clear mail varnish.

6. The main steps for immunofluorescence staining are shownschematically in Fig. 14.1.

7. The specificity of immunoreactions was checked by prepa-ration of the following controls: a control without primaryantibodies; a control incubated with normal mouse seruminstead of primary antibodies and controls incubated withantibodies that had been incubated for 24 h at 4◦ with thecorresponding antigen at 10–100 μg/mL (see Note 4).


Fig. 14.1. Procedures for the retrieval and immunostaining of antigens.

3.3. Light Microscopy 1. We analyzed the localization of immunoreactivity spe-cific for K13 and CIII with an Axioplan fluorescencemicroscope.

2. We also examined DIC images that revealed the histologyand morphology of cells on the same semi-ultrathin sections.

3. We stacked the immunofluorescence images and the cor-responding DIC images by computer, as illustrated inFig. 14.2.

3.4. Detection ofImmunofluorescenceof K13 in Rat Tongueon P7

1. As judged from fluorescence in combination with DICimages, immunoreactivity specific for K13 was stronger inthe suprabasal cells of the narrow interpapillary cell columnson P7 than in those of the papillary cell columns, reflectingthe appearance of a keratinized layer in the papillary epithe-lium (see Note 5) (see Fig. 14.3a).

2. The immunoreactivity in immunopositive cells in the inter-papillary cell columns seemed to be more densely distributedin the cytoplasm than in the nuclei.


Fig. 14.2. Computerized stacking of an immunofluorescence image on the correspond-ing DIC image.

Fig. 14.3. Immunofluorescence of K3 in the rat tongue on P7 (a) and on P14 (b). E, epithelium; CT, connective tissue; M,lingual muscle; FP, filiform papillary cell column; IP, interpapillary cell column; KL, keratinized layer; arrows, connectivetissue papillae.

3. Immunoreactivity specific for K13 in the suprabasal cellsof the interpapillary cell columns was more densely dis-tributed than it was in the suprabasal cells of the papillarycell columns.

4. The corresponding DIC images revealed that the lingualepithelium was composed of stratified squamous cells and,in addition, that rounded rudiments of filiform papillae werearranged at equal intervals, for the most part.

5. The size of basal cells in the papillar and interpapillar regionswas almost same as that of cells in the dorsal epithelium ofthe tongue.


6. At this stage, the connective tissue was beginning to pen-etrate into the central part of each filiform papilla and,as a result, the epithelial–connective tissue border wasundulated.

3.5. Detection ofImmunofluorescenceof K13 in Rat Tongueon P14

1. The distribution of immunoreactivity specific for K13 injuveniles on P14 was extensive in the entire deep interme-diate layer of the interpapillary cell columns that containedsuprabasal cells (see Fig. 14.3b).

2. The immunoreactivity was stronger throughout the cyto-plasm of immunopositive cells than it was on P7 and nonewas evident in nuclei.

3. Immunoreactivity specific for K13 in the suprabasal cells ofthe papillary cell columns disappeared completely with thedevelopment of the thick keratinized layer of filiform papil-lae.

4. The corresponding DIC images revealed that the lingualepithelium was composed of stratified squamous cells onP14 and that rounded rudiments of filiform papillae werearranged at equal intervals, for the most part, just as theywere on P7.

5. The thickness of the dorsal epithelium of the tongue wasalmost same as or slightly greater than that on P7.

6. The cell composition of the epithelium and sizes of cells werenot very different from those on P7.

7. The penetration of connective tissue into the central partof each filiform papilla was clearer than that on P7 and theundulation of the epithelial–connective tissue border was,thus, more distinct.

3.6. Detection ofImmunofluorescenceof CIII in Rat Tongueon P7

1. As indicated by results for K13, DIC images revealed thatboth the morphogenesis of filiform papillae and the ker-atinization of the epithelium had already progressed con-siderably in the lingual epithelium of juveniles by P7 (seeFig. 14.4a).

2. Furthermore, the connective tissue papillae penetrateddeeply into the center of each papilla from the basal sideof the lingual epithelium.

3. The lingual muscle was fully developed, as revealed by DICimages.

4. Immunoreactivity specific for CIII (see Note 6) in tonguesof juveniles on P7 was very distinct within the connectivetissue between the epithelium and the muscle layer.

5. Immunoreactivity specific for CIII was clearly recognizablein the connective tissue papillae.


Fig. 14.4. Immunofluorescence of CIII in the rat tongue on P7 (a) and on P14 (b). E, epithelium; CT, connective tissue; M,lingual muscle; FP, filiform papillary cell column; IP, interpapillary cell column; KL, keratinized layer; arrows, connectivetissue papillae.

6. Immunoreactivity specific for CIII was also distinct in theendomysium and perimysium around the lingual muscle (seeNote 7).

3.7. Detection ofImmunofluorescenceof CIII in Rat Tongueon P14

1. In the lingual epithelium of juveniles on P14, DIC imagesrevealed that the filiform papillae were almost the same shapeas those in adults (see Fig. 14.4b).

2. The differences between the anterior and the posterior partsof filiform papillae and the interpapillar regions were easilydistinguishable and keratinization of the papillary epitheliumwas already very significant.

3. Furthermore, connective tissue papillae penetrated deeplyinto the center of each papilla.

4. The lingual muscle had matured completely and theendomysium and perimysium around the lingual musclewere also fully developed (see Note 7).

5. Immunoreactivity specific for CIII was significant and dis-tinct both within the connective tissue between the epithe-lium and the muscle layer and in the connective tissue papil-lae and it was also recognized in the endomysium and per-imysium around the lingual muscle.

3.8. ComparisonBetween LightMicroscopyand ConfocalLaser-ScanningMicroscopy AfterImmuno-stainingwith Fluorescence-LabeledAntibodies

1. There was no significant difference in terms of the detec-tion of immunofluorescence between light microscopy andconfocal laser-scanning electron microscopy

2. We found, for both light microscopy and confocal laser-scanning microscopy, that microwave heating of specimensjust before immunofluorescence staining was also useful forantigen retrieval in epoxy resin-embedded specimens, eventhough this treatment (4) was developed for antigen retrievalin formalin-fixed, paraffin-embedded specimens.


3. In our previous and present efforts, to clarify the localiza-tion of immunoreactivity in tissues and cells, we used a com-bination of immunofluorescence staining of semi-ultrathinsections and corresponding DIC images obtained by lightmicroscopy and by confocal laser-scanning microscopy.Using this technique, we were easily able to detect and local-ize immunofluorescence in the tongues of rat fetuses andjuveniles at embryonic and postnatal stages (4–7).

4. In some of our previous studies, however, we used confo-cal laser-scanning microscopy in the transmission mode toexamine the same specimens as those in which we had moni-tored fluorescence in an effort to reveal histological and cell-morphological features more distinctly than those revealedin DIC images (8–11).

5. By combining immunofluorescence images and the cor-responding images obtained by confocal laser-scanningmicroscopy in the transmission mode, we were able to definethe histological localization of K13 and CIII more clearlythan when we combined immunofluorescence images andthe corresponding DIC images.

6. Our method, using both light microscopy and confocallaser-scanning microscopy, should be applicable to variouskinds of tissue and cell of which only very small amounts areavailable.

4. Notes

1. Dehydrate specimens in ascending ethanol series (60, 70, 80,90, 99%; 10 min each) and then in absolute ethanol (twicefor 15 min each); transfer them to propylene oxide (twicefor 15 min each); and finally transfer them to the mixture ofpropylene oxide and epoxy resin (1:1,v/v).

2. The descending acetone series consists of absolute ace-tone (two times), 50% acetone and distilled-deionized water(DDW). Each specimen is passed through each solventquickly, for 30 s or so.

3. BP: Band-pass filter, FT: dichroic mirror.4. No immunolabeling of the lingual mucosa of juvenile rats

was seen in any of the negative controls.5. K4 and K13 are associated with differentiating suprabasal

cells of the nonkeratinized and parakeratinized oral epithe-lium (13–15).


6. CIII is found together with type I collagen in many stro-mal connective tissues (16). As Garrone et al. (17) stated,other types of collagen in addition to CI and CIII are lessabundant, although they are certainly extremely importantfor normal skin physiology.

7. The endomysium consists of fine collagenous and reticularfibers and is found between individual muscle fibers. Theperimysium consists of the deep layers of connective tissuethat invests bundles of muscle fibers.

Acknowledgments

The authors thank Professor Sumio Yoshie for his useful adviceand they also thank Drs. Hiroyuki Yokosuka and MasahikoKumakura for their skilled technical assistance. This work wassupported by Grants-in-Aid (NDUF-06-04 NDUF-06-21 andNDUF-07-01) from the Nippon Dental University.

References

1. Mayor, H. D., Hampton, J. C., Rosario, B.(1961) A simple method for removing theresin from epoxy-embedded tissue. J Cell Biol9, 909–910.

2. Haraguchi, C. M., Yokota, S. (2002)Immunofluorescence technique for 100-nm-thick semithin sections of Epon-embeddedtissues. Histochem Cell Biol 117, 81–85.

3. Litwin, J. A., Yokota, S., Hashimoto, T.,Fahimi, H. D. (1984) Light microscopicimmunocytochemical demonstration of per-oxisomal enzymes in Epon sections. Histo-chemistry 81, 15–22.

4. Iwasaki, S., Aoyagi, H., Yoshizawa, H.(2003) Immunohistochemical detection ofthe expression of keratin 14 in the lingualepithelium of rats during the morphogene-sis of filiform papillae. Arch Oral Biol 48,605–613.

5. Iwasaki, S., Aoyagi, H., Asami, T. (2006)Expression of keratin 18 in the peridermcells of the lingual epithelium of fetal rats:visualization by fluorescence immunohisto-chemistry and differential interference con-trast microscopy. Odontology 94, 64–68.

6. Iwasaki, S., Yoshizawa, H., Aoyagi, H.(2006) Immunohistochemical expression ofkeratin 13 and 14 in the lingual epithe-

lium of rats during the morphogenesisof filiform papillae. Arch Oral Biol 51,416–426.

7. Iwasaki, S., Aoyagi, H., Yoshizawa, H.(2007) Immunohistochemical detection ofepidermal growth factor and epidermalgrowth factor receptor in the lingual mucosaof rats during the morphogenesis of filiformpapillae. Acta Histochem 109, 37–44.

8. Iwasaki, S., Aoyagi, H. (2007) Expressionof keratin 14 in the basal cells of the lin-gual epithelium of mice during the mor-phogenesis of filiform papillae: visualizationby fluorescent immunostaining and confocallaser-scanning microscopy in the transmissionmode. Odontology 95, 61–65.

9. Asami, T., Aoyagi, H., Yoshizawa, H.,Wanichanon, C., Iwasaki, S. (2008)Immunohistochemical expression of type IIcollagen in the lingual mucosa of rats duringthe morphogenesis of the tongue. Arch OralBiol 53, 622–628.

10. Aoyagi, H., Asami, T., Yoshizawa, H.,Wanichanon, C., Iwasaki, S. (2008) Newlydeveloped technique for dual localizationof keratin 13 and 14 by fluorescenceimmunohistochemistry. Acta Histochem 110,324–332.


11. Iwasaki, S., Asami, T., Wanichanon, C.,Yoshizawa, H., Aoyagi, H. (2008) Immuno-histochemical analysis of type III collagenexpression in the lingual mucosa of rats dur-ing organogenesis of the tongue. Odontology96, 12–20.

12. Shi, S. R., Key, M. E., Karla, K. L. (1991)Antigen retrieval in formalin-fixed, paraffin-embedded tissues: an enhanced methodfor immunohistochemical staining based onmicrowave heating of tissue sections. J His-tochem Cytochem 39, 741–748.

13. Morgan, P. R., Leigh, I. M., Purkis, P. E.,Gardner, I. D., Van Muijen, G. N. P., Lane,E. B. (1987) Site variation in keratin expres-sion in human oral epithelia – an immuno-cytochemical study of individual keratins.Epithelia 1, 31–43.

14. Barrett, A. W., Selvarajah, S., Franey, S.,Wills, K. A., Berkovitz, B. K. B. (1998) Inter-species variations in oral epithelial cytokeratinexpression. J Anat 193, 185–193.

15. Heyden, A., Huitfeldt, H. S., Koppang, H.S., Thrane, P. S., Bryne, M., Brandtzaeg, P.(1992) Cytokeratins as epithelial differentia-tion markers in premalignant and malignantoral lesions. J Oral Pathol Med 21, 7–11.

16. Keene, D. R., Sakai, L. Y., Bächinger, H.P., Burgeson, R. E. (1987) Type III colla-gen can be present on banded collagen fibrilsregardless of fibril diameter. J Cell Biol 105,2393–2402.

17. Garrone, R., Lethias, C., Guellec, D.(1997) Distibution of minor collagens dur-ing skin development. Microsc Res Tech 38,407–412.

SUBJECT INDEX

A

Achromatic objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125Acridine orange . . . . . . . . . . . . . . . . . . . . . . . . . . .218, 220–222Acute infection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69, 75Adipose differentiation-related protein, ADRP,

adipophilin . . 152, 154, 158–160, 166, 168–169,173–174

Adiposome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149Aequorin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206Air-drying technique . . . . . . . . . . . . . . . . . . . . . . . . . 59, 61–62Amastigote . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69–70, 75Angiogenesis . . . . . . . . . . . 183–187, 189–193, 197, 199–201

tumor angiogenesis . . . . . . . . . . . 183–187, 189–190, 197Anna Morse’s solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34Aperture diaphragm. . . . . . . . . . . . . .103, 109–110, 115, 131Apochromatic objective . . . . . . . . . . . . . . . . . . . .122–123, 126Aquatic ecology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215–226Arachidonic acid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163–164Artifacts . . . . . . 4, 8–9, 20, 34, 46, 111, 113–114, 118, 125,

131–132, 134–136, 199, 213

B

Bacterioplankton . . . . . . . . . . . . . . . . . . . . . 216, 220–221, 225Bandpass filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116, 195BODIPY R© . . . . . . . 151, 154, 156–160, 166, 170, 173–174BrdU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7, 13–14

See also 5-bromo-2-deoxyuridineBright-field microscope . . . . . . . . . . . . . . . . . . . .100, 103–1045-Bromo-2-deoxyuridine . . . . . . . . . . . . . . . . . . . . . . . . . . 6, 13Buccal mucosa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55Buffers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8, 30, 139, 208

C

Ca2+-induced Ca2+ release or CICR . . . . . . . . . . . . . . . . 207Calcium spark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205–212Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51, 66, 164, 183–201Cardiac excitation-contraction or EC . . . . . . . . . . . . . . . . 206Cardiac myocyte . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .205–212Cardiomyocytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75, 77Cellular dysfunction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82Chagas’ disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69–70Chromatic aberration . . . . . . . . . . . . . . . . . . 95, 124–127, 129Chromosome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52–53, 62, 65

plant chromosome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53Condenser . . . 27, 54, 56, 101–103, 109, 117, 122, 134–135Confocal microscopy . . . . . . . . . 83, 102, 105, 109, 140, 142,

147, 206, 208–209, 212Contrast-enhancing method. . . . . . . . . . . . . . . . . . . . . . . . . .82Correction collar . . . . . . . . . . . . . . . . . . . . . . . . . . 126, 129–131Cremaster muscle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82–86

Crystal polyester resin . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22, 24Cutting-grinding . . . . . . . . . . . . . . . . . . . . . . 19–21, 24, 26, 28Cytogenetic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53, 65Cytometric . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53Cytospin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152, 155–157,

159, 166

D

DAPI or 4′, 6′-diamidino-2-phenylindole . . 116–118, 128,130, 165, 173, 221, 225

Demineralized teeth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20Demineralizer agent . . . . . . . . . . . . . . . . . . . 20–23, 22–23, 29Densitometric cytological analyses . . . . . . . . . . . . . . . . . 51–52Density calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57–58Dibenzolperoxide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6, 40, 71Dichromatic beam-splitter . . . . . . . . . . . . . . . . . . . . . . . . . . 120Dichromatic mirror . . . . . . . . . .102–103, 114, 117, 119–122DIC or DIC Nomarski . . . . . . . . . . 113, 127, 129, 131–136,

229–230, 232, 234–238Differential interference contrast . . . . . . . . . . . 127, 229–239

See also DIC or DIC NomarskiDigital camera . . . . . . . . . . . . . 15, 52, 65, 158, 168, 171, 198Digital image . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51–52, 210DNA quantitification . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51–67

nuclear and chromosomal DNA content . . . . . . . . . . . 52

E

EDTA . . . . . . . . . . . . . . . . . . . . . . . . . . 21, 23, 27, 29, 194, 197Eicosanoid

EicosaCell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163–180Eicosanoid lipid mediator . . . . . . . . . . . . . . . . . . . 163–180

Embeddingembedding problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40GMA embedding . . . . . . . . . . . . . . . .4, 10, 37–48, 71, 73large samples embedding . . . . . . . . . . . . . . . . . . . . . . . . . 38

Emission filter . . . . . . . . . . . . . . . . . . . 115–122, 195, 197, 232Emission spectrum . . . . . . . . . . . . . . . . . . . . 96, 100, 107, 207Endocervical swab. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55Endocytosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137–147Endosome recycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137Eosin . . . . . . . . . . 6, 9, 11–12, 21, 23, 29, 31, 34, 71, 74–75,

79, 230Eosinophil . . . . . . . . . . . . . . . . . . . . . . 154, 165, 172, 177–179Epifluorescence microscopy . . . . . . . . . . . . . . . . . . . . . . 82, 221

epi-illuminator . . . . . . . . . . . . . . . . . . . . . . . . . . . . .109, 113Epoxy resin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229–239Erythrosine-orange-toluidine blue . . . . . . . . . . . . 5, 9, 11–12Excitation balancer . . . . . . . . . . . . . . . . . . . . . . . . 113–114, 120Excitation filter . . . . 114–117, 120–122, 197, 222, 225, 232Exocytosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137–147Extracelluar matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

H. Chiarini-Garcia, R.C.N. Melo (eds.), Light Microscopy, Methods in Molecular Biology 689,DOI 10.1007/978-1-60761-950-5, c© Springer Science+Business Media, LLC 2011

241

242LIGHT MICROSCOPY

Subject Index

F

Feulgen reaction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .52, 56, 66Feulgen solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53Field diaphragm . . . . . . . . . . . . . . . . . . 57, 109–110, 113, 115Filter cube . . . . . . . . . . . . . 109–111, 114–116, 118–122, 132FITC . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82, 118, 140, 142, 232Fixation problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46Fixative

Bouin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3, 34–35buffered neutral formalin . . . . . . . . . . . . . . . . . . 21, 27, 33formalin-aceto-alcohol (FAA) . . . . . . . . . . . . . . 39, 46–47glutaraldehyde . . . . . . . . . . . . . . . . . . . . . . . . . . 4–5, 7–9, 16Karnovsky . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5, 7osmium . . . . . . . . . . . . . . . . . . . . 1, 58–159, 150, 153–155paraformaldehyde . . . . . . . . . . . 4–5, 39, 70, 77, 140, 142,

144, 150–151, 153, 155, 157, 159, 175,230–231, 233

volume of the fixative solution . . . . . . . . . . . . . . . . . . . . .27Flat field correction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110–111Fluo-4 AM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208–212Fluorescein . . . . . . . . . . . . . . . 82, 95, 98, 105, 111–112, 142,

158, 232Fluorescence efficiency. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97Fluorescence lifetime. . . . . . . . . . . . . . . . . . . . . . . . . . . .99–100Fluorescence quantum yield . . . . . . . . . . . . . . . . . . . . . 98, 156

See also Quantum efficiency (QE)Fluorescent probe . . . . . . . . . . . . . . . . . 83, 113, 158–159, 225Fluorite objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125–126Fluorochrome. . .82, 95–100, 102–103, 105, 107–108, 110,

114, 118–121, 125, 132, 165, 167, 172, 221Fluorophore . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83, 88, 199FM1-43 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137–147

G

Gelfoam R© . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .191–192Gemcitabine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189Germinated seed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55GFP. . . . . . . . . . . . . . . . . . . . 95, 183–187, 189–194, 196–201

See also Green fluorescent protein (GFP)Glass knife . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5, 9–10, 73Glioma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192–193Glycol methacrylate . . . . . . . . . . . . . . . . . . . . .3–17, 38, 71–72

2-hydroxyethyl methacrylate . . . . . . . . . . . . . . . . . . .38, 71See also GMA

GMAGMA embedding . . . . . . . . . . . . . . . .4, 10, 37–48, 71, 73GMA embedding disadvantages. . . . . . . . . . . . . . . . . . .38GMA infiltration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38GMA polymerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38hardener:GMA ratio . . . . . . . . . . . . . . . . . . . 40–42, 45–48See also Glycol methacrylate

Gomori’s trichrome . . . . . . . . . . . . . . . . . . . . 21–23, 29, 31–32Green fluorescent protein (GFP) . . . . . . . . . 82, 95, 183–201

H

Halogen lamp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113Hardener concentration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Harris’ hematoxylin . . . . . . . . . . . . . . . . . . . . . . . . 6, 23, 31, 71Heart . . . . . . . . . . . . . . . . 8, 69–70, 72, 75, 77, 187, 205–206,

209–210Heat filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113HEMA, see GMAHeparin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4, 8, 15, 53

High-resolution light microscopy . . . . . . . . 3–17, 19–35, 38,82, 84, 149–160, 212, 215–226, 229–230, 232,234, 237–238

Historesin, see GMAHuman nuclear DNA . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55, 59

I

ICM . . . . . . . . . . . . . . . . . . . . . . . 51–53, 55, 59, 61–64, 66–67See also Image cytometry

Image cytometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51–67Image fogging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113Imaging calcium sparks . . . . . . . . . . . . . . . . . . . . . . . . 205–213Immunohistochemistry . . . . . . . . . . . . . . . . . .9, 184, 229–239Infiltration

cooling improves infiltration . . . . . . . . . . . . . . . . . . . . . . 45full-strength infiltration solution . . . . . . . . . . . . . . . . . . 41intermediate infiltration solution or

pre-infiltration . . . . . . . . . . . . . . . . . . . . . 40–42, 45low temperature . . . . . . . . . . . . . 41, 45, 47, 145–146, 218poor infiltration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43

Inflammatory infiltrate . . . . . . . . . . . . . . . . . . . . . . . . 20, 70, 77Integrated optical density . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52Interference contrast microscopy . . . . . . . . . . . . . . . . 229–239

interference filter . . . 54, 57–58, 103, 114–115, 117, 119Intracellular calcium . . . . . . . . . . . . . . . . . . 205–207, 210–211Intravital microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81–89IOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52, 57, 63–65

See also Integrated optical density

K

Karyogram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65Keratin 13 or K13. . . . . . . . . . . . . . . .230–231, 233–236, 238Knee joint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82, 84–87Kohler illumination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134

L

Large specimens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41, 47Laser scanning confocal microscopy . . . . 206, 232, 237–238LED or light emitting diode . . . . . . . . . . 105, 108–109, 195,

197, 221Leukocyte recruitment . . . . . . . . . . . . . . . . . . . . . . . . . . . 81–89Leukocyte rolling. . . . . . . . . . . . . . . . . . . . . . . . . . . . .82, 86–87Leukocyte trafficking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81Light emitting diode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

See also LED or light emitting diodeLight microscopy . . . . . 3–17, 19–35, 38, 43, 52, 77–80, 82,

84–85, 89, 149–160, 212, 215–226, 229–230,232, 234, 237–238

Light source . . . . . . . . . . . . . 58, 102–110, 114–115, 121, 195Linearity test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58, 60, 66Lingual mucosa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238Lipid body . . . . 150, 152, 154, 156, 158–159, 173–174, 179Lipid droplet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149LPS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84–86, 88L-type Ca2+ channel or LCC . . . . . . . . . . . . . . . . . . 206–207

M

Mayer’s albumin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23, 29Mercury lamp . . . . . . . . . . . . . . . . . . . . . . . . . 87, 105–107, 197Metal halide lamp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105, 107Metaphase blocking solution . . . . . . . . . . . . . . . . . . . . . . . . . 56Microcirculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81–86, 89Microvascular disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

LIGHT MICROSCOPY

Subject Index 243

Molar extinction or molar absorption . . . . . . . . . . . . . 99–100Mordant solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12Morphometry . . . . . . . . . . . . . . . . . . . . . . . . . . 3–17, 20–21, 73

N

ND-GFP . . . . . . . . . . . . . . . . . . 186–187, 189, 193, 196, 201See also Nestin-drive green fluorescent protein or

ND-GFPNestin-drive green fluorescent protein or

ND-GFP . . . . . . . . . . . . . . . . . . . . . . . . . . 183–201Neurogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .192–193Neuromuscular junction . . . . . . . . . . . . . . . 138, 140–145, 147Neuronal activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137–138Neuronal communication . . . . . . . . . . . . . . . . . . . . . . . . . . . 138Neutral density . . . . . . . 54, 58, 103, 109–110, 112–114, 122Nile red . . . . . . . . . . . . . . . . . . . . . . . . . 151, 154–156, 158–160Nuclear genome size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .52Numeric aperture, (NA) . . . . . . . 54, 94, 100–102, 115, 123,

126–130, 209, 230, 232

O

OD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52, 57–60, 62–64See also Optical density

Odontoblast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30, 32Oil red O . . . . . . . . . . . . . . . . . . . . . . . 151, 154, 156, 158–159Optical density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .52Orthotopically-growing tumor . . . . . . . . . . . . . . . . . . . . . . 185Orthotopic breast cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185Orthotopic pancreas cancer . . . . . . . . . . . . . . . . . . . . . . . . . 185

P

Papillary epithelium . . . . . . . . . . . . . . . . . . . . . . . . . . . 234, 237Paraffin . . . . . . . . . . . . . . . . . . . . . . 3, 9–10, 12, 21, 29–31, 37,

72–74, 237embedding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21, 29, 73

Parasite nests . . . . . . . . . . . . . . . . . . . . . . . . . . 70, 73, 75–76, 78Parasitism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69–80PAS, see StainPerfusion. . . . . . . .8–9, 13, 15, 153, 155, 157, 172, 208–210Periodic acid-Schiff solution . . . . . . . . . . . . . . . . . . . . . 3, 6, 13Periodontal tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–35Periodontium . . . . . . . . . . . . . . . . . . . . . . . . . 19–21, 29, 32, 34Phase contrast . . . . . . . . . . . . . . 113, 129, 131–132, 167–168,

171–172, 212Phosphate buffered solution . . . . . . . . . . . . . . . . . . . . . . . . . . . 5Phosphorescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93, 96–97Photobleaching . . . . . . . . . . . . . . . 97, 99, 105, 143, 146, 213Phytoplankton . . . . . . . . . . . . . . . . . . .216–218, 220, 222–225Planck Rychol’s solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34Plankton

communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215–226Planktonic organism . . . . . . . . . . . . . . . . . . . 215, 217, 224

Plant chromosomal ICM . . . . . . . . . . . 55–56, 61–62, 64–65Plant sample

DNA content in plant . . . . . . . . . . . . . . . . . . . . . 52, 63–64large plant samples . . . . . . . . . . . . . . . . . . . . . . . . . . . 37–48plant nuclei . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53, 55, 62

Plant tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37–38, 46–47Plastic resin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4, 70, 72–73Ploidy determination . . . . . . . . . . . . . . . . . . . . . . . . . 55, 59–61Polymerization

polymerization of the sample core . . . . . . . . . . . . . . 44–45rapid polymerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47unsatisfactory polymerization . . . . . . . . . . . . . . . . . . . . . 47

Propidium iodide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224Putt’s eosin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23, 311-Pyrenedodecanoic acid (P-96) . . . . . . . 152, 154, 157–158

Q

Quantum efficiency (QE) . . . . . . . . . . . . . . . . . . . . . . . 98, 105See also Fluorescence quantum yield

Quenching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98–100, 146

R

Red fluorescent protein or RFP . . . . . . . . . . . . . . . . . . . . . . 184Refractive index . . . . . . . . . 54, 100, 102, 124–125, 127–130,

212, 218Resin infiltration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46Resin polymerization . . . . . . . . . . . . . . . . . . . . . . 16, 27, 40, 45Rhodamine . . . . . . . . . . . . . . . . . . . . . . . . . . 82, 84, 87, 89, 158Ryanodine receptor or RyRs . . . . . . . . . . . . . . . . . . . . . . . . . 206

S

Schiff ’s reagent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .54–56, 66Sedgewick-Rafter chamber . . . . . . . . . . . . . . . . . . . . . . . . . 223Semi-thin section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Semi-ultrathin section . . . . . . . . . . . . . . . . 229–231, 234, 238Silicone . . . . . . . . . . . . . . . . . . . . . . . . . . 21, 24–25, 31, 33, 139Spherical aberration . . . . . . . . . . . . . . . . . . 123–126, 128–130Squashing technique. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66Stability test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58–59Stain

BODIPY R© . . . . . . . . . . . . 151, 154, 156–160, 166, 170,173–174

eosin . . . . . . . 6, 9, 11–12, 21, 23, 29, 31, 34, 71, 74–75,79, 230

erythrosine-orange-toluidine . . . . . . . . . . . . . . 5, 9, 11–12Feulgen solution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53–54Gomori’s trichrome . . . . . . . . . . . . . 21, 23, 29, 31–32, 35Harris’ hematoxilin . . . . . . . . . . . . . . . . . . . . . 6, 23, 31, 71Nile red . . . . . . . . . . . . . . . . . . . . . 151, 154–156, 158–160Oil red O . . . . . . . . . . . . . . . . . . . . 151, 154, 156, 158–159PAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9, 11, 13

See also Periodic acid-Schiff solutionPutt’s eosin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23, 311-pyrenedodecanoic acid (P-96). . . . . . . . .152, 154, 157toluidine blue-basic fuchsin. . . . . . . . . . . . . . . . . . . .74–76toluidine blue-borate . . . . . . . . . . . . . . . . . . . . 5, 11–12, 72toluidine blue O . . . . . . . . . . . . . . . . . . . . . . . . . . . 5, 39, 72

Stokes shift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100, 120Styryl dye FM1-43 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137–147

See also FM1-43Synaptic vesicle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137–147Synovial tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85–86Synthetic rubber of white silicone . . . . . . . . . . . . . . . . . . . . . 21

T

Teeth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19–35Temozolomide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192–193Testes, testis

acrossomal system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3spermatogonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3–4, 13

Toluidine blue-basic fuchsin . . . . . . . . . . . . . . . . . . . . . . 74–76Toluidine blue-borate . . . . . . . . . . . . . . . . . . . . . . .5, 11–12, 72Toluidine blue O . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5, 39, 72Tongue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230, 232–238Transcellular emigration . . . . . . . . . . . . . . . . . . . . . . . . . . 82–83

244LIGHT MICROSCOPY

Subject Index

Trypanosoma cruzi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69–80T. cruzi nests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

Type III collagen or CIII . . . . . . . . . . . . . . . . . . . . . . . . . . . 230

U

Ultraviolet light . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94Undecalcified soft tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20Undemineralized samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20Uniformity test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

V

Vacuum pump. . . . . . . . . . . . . . . . . . . . . . . . . . 39, 46, 219, 221disposable syringe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

Virioplankton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215, 220

W

Wavelength . . . . . . . . . . 83, 94, 97–100, 103, 105, 108–109,113–117, 119–120, 124–125, 126, 173, 184,197, 199, 208

Working distance, (WD) . . . . . . . . . . . . . . . . . . 104, 127, 212

X

Xenograft . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184Xenon lamp . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105–107, 109

Z

Zooplankton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216–217, 219,223–224

Light Microscopy: Methods and Protocols

Documents

Transcript of Light Microscopy: Methods and Protocols