Handbook of Cryo-Preparation Methods For Electron Microscopy (Methods in Visualization) (Annie Cavalier, Daniele Spehner Etc.

HANDBOOK OF
Cryo-Preparation Methods
for Electron Microscopy
© 2009 by Taylor & Francis Group, LLC

Methods in Visualization
Series Editor: Gérard Morel
Genome Visualization by Classic Methods

in Light Microscopy
Jean-Marie Exbrayat
Handbook of Cryo-Preparation Methods for Electron Microscopy

Annie Cavalier, Danièle Spehner, and Bruno M. Humbel
Imaging of Nucleic Acids and Quantitation

in Photonic Microscopy
Xavier Ronot and Yves Usson
In Situ Hybridization in Electron Microscopy

Gérard Morel, Annie Cavalier, and Lynda Williams
In Situ Hybridization in Light Microscopy

Gérard Morel and Annie Cavalier
PCR/RT -PCR In Situ Light and Electron Microscopy

Gérard Morel and Mireille Raccurt
Visualization of Receptors: In Situ Applications

of Radioligand Binding
Emmanuel Moyse and Slavica M. Krantic

Methods in Visualization Series
HANDBOOK OF
Cryo-Preparation Methods
for Electron Microscopy
Edited by
Annie Cavalier
Danièle Spehner
Bruno M. Humbel

Painting on the cover by Annie Cavalier
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works

Printed in the United States of America on acid-free paper
10 9 8 7 6 5 4 3 2 1
International Standard Book Number-13: 978-0-8493-7227-8 (Hardcover)
This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been
made to publish reliable data and information, but the author and publisher cannot assume responsibility for the valid-
ity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright
holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this
form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may
rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or uti-
lized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopy-
ing, microfilming, and recording, or in any information storage or retrieval system, without written permission from the
publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://
www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923,
978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For orga-
nizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for
identification and explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data
Handbook of cryo-preparation methods for electron microscopy / editor, Annie Cavalier, Daniele
Spehner, and Bruno M. Humbel.
p. ; cm. -- (Methods in visualization)
Includes bibliographical references and index.
ISBN 978-0-8493-7227-8 (hardcover : alk. paper) 1. Cryobiology--Methodology. 2. Electron
microscopy--Methodology. I. Cavalier, Annie. II. Spehner, Daniele. III. Humbel, Bruno M. IV. Title.
V. Series.
[DNLM: 1. Cryopreservation--methods. 2. Microscopy, Electron--methods. QY 95 H236 2008]
QH324.9.C7H36 2008
570.28’25--dc22 2008016244
Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com

V
SERIES PREFACE
The second half of the 20th century was a period of intense exploration into the anatomy
of the cell. After the ground-breaking discovery of a whole new world of structural
complexity, the period between about 1950 and 1970 was characterized by the further
development of the electron microscope, and specimen preparation methods were
redefined so that cells could be studied at the organelle level, then that of single molecular
units. During the 1970s, correlating structure and function was the most productive area
of cell biology. Electron microscopy was applied to cytochemical reactions developed for
use with light microscopy, but also to histochemical reactions, immunolabeling and
autoradiography. The pioneers of the 1960s were already studying individual molecules
within cells, and by the 1970s these methods had gained wide acceptance.
Since the 1960s, the most effective method for exploring submicroscopic structures has
been ultramicrotomy. But the preparation steps carried out in conventional
ultramicrotomy, i.e., chemical fixation using osmium tetroxide, chemical solvents and
resin embedding, are harmful to structural conformation. In order to preserve the integrity
of molecules in sections, researchers such as Leduc,1,2 Pease3,4 and their co-workers
introduced water-miscible macromolecules as embedding material. This technique proved
effective in enzyme digestion5,6 and the immunochemical labeling of a small number of
antigens,7 and Bernhard’s group adapted it for use with sections of frozen aldehyde-fixed
tissue.8-10 Finally, Tokuyasu introduced inert cryoprotectants, such as sucrose, into
specimens in infusions before freezing so as to improve their plasticity, and the dry,
frozen sections were collected on droplets of saturated sucrose.11,12
For x-ray analytical studies, no suitable flotation fluid has yet been found onto which
frozen sections can be cut. Dry sectioning systems13,14 are therefore used, without
chemical fixation. Electron microscopy, while presenting the advantages associated with
frozen sections of unfixed and unembedded material, provides little ultrastructural detail,
due to the damage caused by the freezing process. “A very serious problem in any
morphological study of frozen tissue is the ice crystal damage that occurs during the
initial freezing and subsequent recrystallization.”13 Christensen14 found that rapid freezing
minimized the size of the ice crystals. A number of other ways of minimizing these
artifacts have been proposed. Chemical fixation or cryoprotection, for example, can
reduce them. But only very rapid freezing, using specifically designed equipment, will
result in satisfactory frozen preparations.
In 1981, Allakhverdov and Kuzminykh15 wrote: “Well-known advantages of specimen
preparation by cryomethods are accompanied by some disadvantages, resulting mainly
from the inadequate level of presently existing laboratories and instruments.” These days,
there are laboratories that specialize in cryomethods, and state-of-the-art equipment is
now available for the preparation and observation of frozen specimens.
Thin-film vitrification is a practical freezing technique that does not need very specialized
equipment, but its application is limited by the fact that the material to be observed has to
be thin enough for direct observation by transmission electron microscopy. Cryo-
ultramicrotomes and electron microscopes are now capable of dealing with larger and/or
thicker frozen samples, resulting in ever better observations and a growing range of
applications.

VI
For the moment, cryomicroscopy using water vitrification is the method of choice for
research in the field of structural biology, combined with x-ray analysis and particle
visualization (e.g., nanoparticles, macroparticles, viruses, filamentary structures).
Vitrification in cells and tissue can attain a depth of 10 µm, and it is now possible to
observe intact cells and tissue in their natural state. In the words of Dubochet16 — though
indeed Fernandez-Moran, in the early 1950s,17 might have said much the same thing —
“That is what cryo-EM of vitreous sections (CEMOVIS) does.”
Gérard Morel
Series Editor
1. Leduc, E.H. et al. The use of water-soluble glycol-methacrylate in ultrastructural

cytochemistry, J. Roy. Microscop. Soc., 81, 119, 1963.
2. Leduc, E.H. and Holt, S.J. Hydroxy-propyl methacrylate, a new water-miscible
embedding medium for electron microscopy, J. Cell Biol., 26, 137, 1965.
3. Pease, D.L. Eutectic ethylene glycol and pure propylene glycol as substitution
media for dehydratation of frozen tissue, J. Ultrastructural. Res., 21, 75, 1967.
4. Pease, D.L. and Peterson, R.G. Polymerisable glutharaldehyde-urea mixture as
polar, water-containing embedding media, J. Ultrastructural. Res., 41, 133, 1972.
5. Bernhard, W. and Granboulan, N. The fine structure of the cancer cell nucleus,
Exp. Cell Res. Suppl, 9, (19), 19, 1963.
6. Granboulan, N. and Granboulan, P. Ultrastructure cytochemistry of the
nucleolus. Demonstration of the chromatin inside the nucleolus, Exp. Cell Res., 34,
71, 1964.
7. Kawarai, Y. and Nakane, P.K. Localization of tissue antigens on the ultrathin
sections with peroxidase-labeled antibody method, J. Histochem. Cytochem., 18,
161, 1970.
8. Bernhard, W. and Leduc, E.H. Ultrathin frozen sections. I. Methods and
ultrastructural preservation, J. Cell Biol., 34, 757, 1967.
9. Bernhard, W. and Viron, A. Improved techniques for the preparation of ultrathin
frozen sections, J. Cell Biol., 49, 731, 1971.
10. Leduc, E.H. et al. Ultrathin frozen sections. II. Demonstration of enzymatic
activity, J. Cell Biol., 34, 773, 1967.
11. Painter, R.G. et al. Immunoferritin localization of intracellular antigens: The use of
ultracryotomy to obtain ultrathin sections suitable for direct immunoferritin
staining, Proc. Nat. Acad. Sci., USA, 70, 1649, 1973.
12. Tokuyasu, K.T. A technique for ultracryotomy of cell suspensions and tissues,
J. Cell Biol., 57, 551, 1973.
13. Appleton, T.C. A cryostat approach to ultrathin "dry" frozen sections for electron
microscopy: A morphological and x-ray analytical study, J. Microsc., 100, 49, 1974.
14. Christensen, A.K. Frozen thin sections of fresh tissue for electron microscopy, with
a description of pancreas and liver, J. Cell Biol., 51, 772, 1971.
15. Allakhverdov, B.L.and Kuzminykh, S.B. Some aspects of developing instruments
for cryomethods, Acta Histochem., 23, 75, 1981.
16. Al-Amoudi, A. et al. Cryo-electron microscopy of vitreous sections, EMBO J., 23,
3583, 2004.
17. Fernandez-Moran, H. Application of the freezing sectioning technique to the study
of cell structure with the electron microscope, Ark. Fys., 4, 471, 1952.

VII
ACKNOWLEDGMENTS
This handbook was initiated by Annie Cavalier and Danièle Spehner as part of a Centre
National de la Recherche Scientifique (CNRS) effort in France (GDR 2368) to promote
cryo-electron microscopy. Bruno Humbel, in Holland, enthusiastically joined the team.
We are indebted to the contributors who have made it come true. It was not such an easy
task explaining tricks of the trade, routinely carried out by each of them, but not so
readily put into words. Dr. Gerard Morel provided his guidance throughout and we are
very grateful to him. We also thank Profs. Daniel Boujard and Gilles Salbert who have
encouraged the project and ensured a propitious environment. Our thankfulness extends
to Dr. Patrick Schultz and Prof. Arie Verkleij who believed in the usefulness of such a
handbook and supported us during its elaboration. Bruno Humbel would also like to thank
the European Network of Excellence (NoE), FP6: “Three-Dimensional Electron
Microscopy” and the Dutch Cyttron Consortium, “A Window on the Molecular
Machinery of Life,” for support. We express our gratitude to Dr. Robert Drillien who not
only overviewed the use of the English language, but also shared his insight on the
contents of the book. This project has also benefited from the generous help of
Emmanuelle Guiot who kindly performed the layout of the entire handbook; her
participation was most valuable. Sandrine Pawlicki kindly constructed the book’s
temporary Web site for which we are thankful. Last but not least, we would like to thank
Christian Cavalier who offered us his home and his hospitality for a few hectic but very
enjoyable weeks during finalization of the manuscripts.
Annie Cavalier
Danièle Spehner
Bruno M. Humbel

IX
THE AUTHORS
Annie Cavalier, graduate of the University Claude Bernard-Lyon 1, is currently an

engineer in Biology in a CNRS laboratory at the University of Rennes I, France. Annie
Cavalier is a member of the French Society of Microscopy (SFµ). She was actively
involved in the organization of the 13th International Congress of Electron Microscopy
(Paris, July 1994), and the first Congress of the French Society of Microscopy (Rennes,
June 1996). She is secretary of the GUMP (Groupement des utilisateurs des microscopes
Philips-FEI; a unique association that brings together users of FEI-made electron
microscopes). Annie Cavalier has regularly given lectures and practical courses for
training in the field of cytology at the ultrastructural level. She has edited two technical
books and contributed to numerous scientific papers. Until recently, her research interests
have focused on the structural and functional characterization of water channels
(aquaporins) and glycerol facilitators. Her current research concerns the assembly and
dynamics of large macromolecular complexes by means of electron tomography.
Danièle Spehner obtained her Ph.D in Biology from the University Louis Pasteur in
Strasbourg, France, while working at the Virology Laboratory of the Medical School in
Strasbourg. After a few years at Transgene, a biotechnology company, where she
developed new vaccines based on poxvirus recombinants, she returned to fundamental
research at the French Institute of Health and Medical Research (INSERM). She is
currently an engineer at the Institute of Genetics and Molecular and Cellular Biology
(IGBMC) in Illkirch, France. Early in her career, she became interested in electron
microscopy to study poxviruses as such or recombinant poxviruses expressing foreign
proteins. She acquired a passion for cryo-electron microscopy in biology after meeting
Professor Hellmut Sitte at a workshop in Seefeld (Austria) and, later on, Bruno Humbel.
She transmitted her passion and her know-how through cryo or immuno workshops to the
Alsatian community and throughout France. Danièle’s current research focuses on the use
of electron tomography and cryo-electron tomography of vitreous cryo-sections
(Cemovis) to analyze poxvirus morphogenesis and the cell nucleus.
Bruno M. Humbel graduated in Biochemistry at the Federal Institute of Technology

(ETH), Zürich, Switzerland. He was privileged to do his Ph.D with Dr. Martin Müller in
the group of Prof. Dr. Hans Moor at the Institute of Cell Biology, ETH. After a postdoc
period of four years at the Max-Planck-Institute for Biochemistry in Munich, he joined
the lab for Electron Microscopy and Structure Analysis at Utrecht University in the group
of Prof. Dr. Arie J. Verkleij, where he currently holds the position of an associate
professor. Bruno Humbel’s main interest is to develop preparation methods for (electron)
microscopy, which not only allow a glimpse into life at high resolution, but also enable
identification and localization of the machinery of life within cells. The aim is to visualize
the living cell at low resolution and to zoom in to analyze its ultrastructure at high
resolution: The gateway to in situ biological nanostructures. Recently, his research
focuses on correlative microscopy to introduce the FIB/SEM technology into life sciences
and the application of a newly developed integrated laser electron microscope (ILEM).
He is teaching cryo-techniques and immunolabeling at Utrecht University and in different
workshops in Europe and Asia.

X
CONTRIBUTORS
AL-AMOUDI Ashraf CHRÉTIEN Denis

EMBL UMR CNRS 6026, Equipe TIPs
HEIDELBERG Campus de Beaulieu
Germany Université de Rennes 1
RENNES
France
ARNAL Isabelle CRUCIFIX Corinne

UMR CNRS 6026, Equipe TIPs Institut de Génétique et de Biologie
Campus de Beaulieu Moléculaire et Cellulaire
Université de Rennes 1 Inserm U596, CNRS UMR7104
RENNES Université Louis Pasteur de Strasbourg
France ILLKIRCH
France
BAUMEISTER Wolfgang DE CARLO Sacha

Max-Planck-Institut für Biochemie MCDB 347
Abteilung Strukturbiologie University of Colorado,
MARTINSRIED bei München BOULDER Colorado
Germany USA
BORDAT Christian DE HAAS Felix

INRA-CRJ FEI Company
NURELICE, Bât. 230 EINDHOVEN
Domaine de Vilvert The Netherlands
JOUY EN JOSAS
France
BOUCHER-MARQUIS Cédric DUBOCHET Jacques
University of Colorado Centre de Microscopie
Dept. of Molecular Cellular and Laboratoire d’Analyses Ultrastructurales
Developemental Biology, Porter Science, Université de Lausanne
BOULDER Colorado LAUSANNE
USA Switzerland
BRON Patrick EDELMANN Ludwig

UMR CNRS 6026, Equipe SDM Anatomie und Zellbiologie,
Campus de Beaulieu Universität des Saarlandes,
Université de Rennes 1 HOMBURG/Saar
RENNES Germany
France
CAVALIER Annie ELTSOV Mikhail

UMR CNRS 6026, Equipe SDM EMBL
Campus de Beaulieu HEIDELBERG
Université de Rennes 1 Germany
RENNES
France

XI
FREDERIK Peter M. KAECH Andres

Maastricht University Center for Microscopy and Image
EM Unit / Pathology Analysis at the University of Zurich
MAASTRICHT ZÜRICH
The Netherlands Switzerland
FUKAZAWA Yugo LEIS Andrew

Division of Cerebral Structure Max-Planck-Institut für Biochemie
National Institute for Physiological Abteilung Strukturbiologie
Sciences, Myodaiji MARTINSRIED bei München
OKAZAKI Germany
Japan
GEERTS Willie J. C. MALLETER Marine

Electron Microscopy & Structure Laboratoire de pharmacologie marine
Analysis Faculté des Sciences de Nantes
Cellular Architecture  Dynamics NANTES
Utrecht University France
UTRECHT
The Netherlands
GUERQUIN-KERN Jean Luc MASUGI-TOKITA Miwako

Institut Curie Recherche/INSERM U 759 Division of Cerebral Structure
Laboratoire Microscopie Ionique National Institute for Physiological
Bat.112 Centre Universitaire Sciences, Myodaiji
ORSAY OKAZAKI
France Japan
GRUSKA Manuela MÉSINI Philippe J.

Max-Planck-Institut für Biochemie Institut Charles Sadron-CNRS-ULP
Abteilung Strukturbiologie STRASBOURG
MARTINSRIED bei München France
Germany
HAGIWARA Akari MOREL Gérard

Division of Cerebral Structure UMR 5123 CNRS-
National Institute for Physiological Université Claude Bernard-Lyon 1
Sciences, Myodaiji VILLEURBANNE
OKAZAKI 444 France
Japan
HUMBEL Bruno M. PAPAI Gabor

Electron Microscopy & Structure Institut de Génétique et de Biologie
Analysis Moléculaire et Cellulaire
Cellular Architecture  Dynamics Inserm U596, CNRS UMR7104
Utrecht University Université Louis Pasteur de Strasbourg
UTRECHT ILLKIRCH
The Netherlands France

XII
SCHMUTZ Marc STUDER Daniel

Institut Charles Sadron-CNRS-ULP Institute für Anatomy
STRASBOURG University of Bern
France BERN
Switzerland
SCHULTZ Patrick TARUSAWA Etsuko
Institut de Génétique et de Biologie Division of Cerebral Structure
Moléculaire et Cellulaire National Institute for Physiological
Inserm U596, CNRS UMR7104 Sciences, Myodaiji
Université Louis Pasteur de Strasbourg OKAZAKI
ILLKIRCH Japan
France
SCHWARZ Heinz THOMAS Daniel
Max-Planck-Institut fuer UMR CNRS 6026, Equipe SDM
Entwicklungsbiologie Campus de Beaulieu
TÜBINGEN Université de Rennes 1
Germany RENNES
France
SHIGEMOTO Ryuichi VAN DONSELAAR Elly

Division of Cerebral Structure Electron Microscopy & Structure
National Institute for Physiological Analysis
Sciences, Myodaiji Cellular Architecture  Dynamics
OKAZAKI Utrecht University
Japan UTRECHT
The Netherlands
SITTE Hellmuth VANHECKE Dimitri
HOMBURG-SAAR Institute für Anatomy
Germany University of Bern
or SEEFELD in Tyrol BERN
Austria Switzerland
SPEHNER Danièle VERKLEIJ Arie J.
Institut de Génétique et de Biologie Electron Microscopy & Structure
Moléculaire et Cellulaire Analysis
Inserm U596, CNRS UMR7104 Cellular Architecture  Dynamics
Université Louis Pasteur de Strasbourg Utrecht University
ILLKIRCH UTRECHT
France The Netherlands
STIERHOF YorkDieter VONCK Janet
Zentrum für Molekularbiologie der Max-Planck-Institute of Biophysics
Pflanzen (ZMBP). Department of Structural Biology
Elektronenmikroskopie, Universität FRANKFURT am Main
Tübingen Germany
TÜBINGEN
Germany
STORMS Marc M.H. ZUBER Benoît

FEI Company MRC Laboratory of Molecular Biology
EINDHOVEN CAMBRIDGE
The Netherlands United Kingdom

XIII
TABLE OF CONTENTS
Introduction XVII
Abbreviations XIX
Chapter 1 1
Vitreous Water
Jacques Dubochet
Part I  Cryo-Fixation Methods
Chapter 2 17
Slam-Freezing, Metal-Mirror Freezing
Danièle Spehner and Ludwig Edelmann

Chapter 3 49
Plunge-Freezing (Holey Carbon Method)
Sacha De Carlo
Chapter 4 69
Controlled Vitrification
Peter M. Frederik, Felix de Haas and Marc M.H. Storms
Chapter 5 101
BAL-TEC HPM 010 High-Pressure Freezing Machine
Andres Kaech
Chapter 6 129
High-Pressure Freezing LEICA EMPACT
Dimitri Vanhecke and Daniel Studer

XIV
Part II  Cryo-Electron Microscopy
Chapter 7 159
Frozen-Hydrated Macromolecules for Structural Analysis
Corinne Crucifix, Gabor Papai and Patrick Schultz
Chapter 8 191
Two-Dimensional Crystals
Patrick Bron and Janet Vonck
Chapter 9 219
Cryo-Negative Staining
Sacha De Carlo
Chapter 10 237
Vitrification of Dynamic Microtubules
Isabelle Arnal, Marine Malleter and Denis Chrétien
Chapter 11 259
CEMOVIS: Cryo-Electron Microscopy of Vitreous Sections
Jacques Dubochet, Ashraf Al-Amoudi, Cédric Bouchet-Marquis,
Mikhail Eltsov and Benoît Zuber
Chapter 12 291
Cryo-Electron Tomography
Andrew Leis, Manuela Gruska and Wolfgang Baumeister
Part III  Low-Temperature Embedding
Chapter 13 319
Freeze-Substitution
Bruno M. Humbel
Chapter 14 343
Cryo-Fixation, Freeze-Substitution, Rehydration and Tokuyasu Cryo-
Sectioning
YorkDieter Stierhof, Elly van Donselaar, Heinz Schwarz, Bruno M.
Humbel
Chapter 15 367
Freeze-Drying and Embedding of Biological Material
Ludwig Edelmann

XV
Part IV  Freeze-Fracture and Metal Shadowing

Chapter 16 391
The Shadow of Hydrated Biological Specimens
Daniel Thomas
Chapter 17 411
Cryo-Fracture of Self-Assembled Systems in Organic Solvent
Marc Schmutz and Philippe J. Mésini
Part V  Analysis
Chapter 18 433
Progressive Lowering of Temperature for Immunolabeling and in situ
Hybridization
Annie Cavalier and Danièle Spehner
Chapter 19 467
Cryo-Sectioning According to Tokuyasu
Bruno M. Humbel and YorkDieter Stierhof
Chapter 20 499
Cryo-Preparation Procedures for Elemental Imaging by SIMS and
EFTEM
Christian Bordat and Jean-Luc Guerquin-Kern
Chapter 21 537
Correlative Light and Electron Microscopy
Heinz Schwarz and Bruno M. Humbel
Chapter 22 567
SDS-digested Freeze-Fracture Replica Labeling (SDS-FRL)
Yugo Fukazawa, Miwako Masugi-Tokita, Etsuko Tarusawa,
Akari Hagiwara and Ryuichi Shigemoto
Chapter 23 587
Immunolabeling of Ultrathin Sections with Enlarged 1 nm Gold or
QDots
YorkDieter Stierhof
Chapter 24 617
3-D Electron Tomography of Cells and Organelles
Willie J. C. Geerts, Bruno M. Humbel and Arie J. Verkleij
Final Considerations 651

Hellmuth Sitte
Glossary 657

XVII
INTRODUCTION
The purpose of this handbook is to provide guidance to newcomers to the field who wish
to learn and possibly apply the methods it describes. Of course, no biologist knows every
technique in his specialty and this also pertains to those using electron microscopy, so we
hope that our fellow colleagues will also find this book useful for one application or
another. The chapters have been written by experts in the field, in some instances the very
inventors of the method, who have taken care to explain the history behind the techniques
they describe and how they are most readily carried out today. Whenever possible, clear
step-by-step recipes are presented and the tools and ingredients of the methods, as well as
where they can be purchased, are listed. After a quick glance at the contents one will
realize that this handbook goes well beyond the preparation methods the title suggests and
into the realm of cryo-electron microscopy analysis itself, hence the numerous
illustrations from the contributors’ own laboratories.
Most biologists have their own idea of what cryo-electron microscopy is all about and this
notion may be strongly influenced by one’s own experience as well as the fashion of the
times. It has long been realized that freezing specimens is a nearly perfect way to
maintain biological samples in a state as close as possible to their native state. Over the
years many different methods have been devised to freeze samples and each of them has
been presented with advice on how they should be applied and to what kind of material.
The ultimate achievement in this field is certainly the ability to embed samples into a
somewhat mysterious form of ice known as vitreous water or crystal-free ice. Today,
there are numerous ways to take advantage of this technological feat. Whereas some
methods aim at perfect preservation for fine structural analysis, others use vitreous water
as an intermediate step for in situ localization of biological molecules and ions.
Sophisticated methods of resin embedding subsequent to removal of vitreous water, as
well as freeze-fracture, have become valuable tools for numerous applications as
illustrated in this handbook. Ice also turned out to be a convenient medium for embedding
and sectioning, as has been so exquisitely exemplified by the so-called Tokuyasu method,
and dealt with in several chapters herein. This and the so-called hybrid methods, freeze-
substitution and freeze-drying, have been found to be ideal intermediates to achieve
highly efficient immunolabeling, so critical for determining the molecular geography
inside tissues and cells. On the other hand, such methods have gained considerably from
parallel developments including new ways to stain samples and label them. This opus
would have been deficient if it had not also dealt with one of today’s most promising
developments in cryo-electron microscopy, namely cryo-electron tomography. Although
such tools would deserve an entire book in their own right, the chapters herein should
provide a useful overview of the topic.
The pace of cryo-electron microscopy has been accelerating in recent years and the
methods available are constantly evolving. It is our wish that this handbook be a useful
aid to the scientific community for current and future endeavors, and especially for
novices to show the beauty of cryo and encourage them to take up the challenge.

XIX
ABBREVIATIONS
2-D  Two-Dimension/two-Dimensional
3-D  Three-Dimension/three-Dimensional
ART  Algebraic Reconstruction Technique
CCD  Charge Coupled Device (camera)
CEMOVIS  Cryo-Electron Microscopy Of Vitreous
Sections
CET  Computerized electron tomography
CFD  Cryo-sorption Freeze-Drying system
CIA  Cutting-Induced Amorphous water
CMC  Critical Micelle Concentration
Cryo-ET  Cryo-Electron Tomography
Cryo-EM  Cryo-Electron Microscopy (of vitrified
specimens)
CS  Cryo-sectioning according to Tokuyasu
CTF  Contrast Transfer Function
EELS  Electron Energy-Loss Spectrum
EF  Energy Filtering
EFTEM  Energy Filtered Transmission Electron
Microscope
ESI  Electron Spectroscopic Imaging
ET  Electron Tomography
FD  Freeze-Drying
FEG  Field Emission Gun
FF  Freeze-Fracturing/Freeze-Fracture
FRL  Freeze-fracture Replica Labeling
FS  Freeze-Substitution
HAADF  High-Angle Annular Dark-Field
HC-PRO  Helper Component PROteinase
HPF  High-Pressure Freezer, High-Pressure
Freezing
LR  London Resin
LTE  Low-Temperature Embedding
MCS  Membrane Contact Side
MES  Morpholino Ethane Sulfonic acid
MMF  Metal Mirror Freezing
MS medium  Murashige and Skoog medium
PLT  Progressing Lowering of Temperature
Pt/C  Platinum used in conjunction with
carbon
QD  Quantum dot
RTS  Rapid transfer system
SIMS  Secondary-Ion Mass Spectrometry
SIRT  Simultaneous Iterative Reconstruction
Technique
WBP  Weighted Back-Projection
θ  Angle of shadowing

Vitreous Water 3
CONTENTS
1. INTRODUCTION ................................................................................................. 5
2. WATER IS REMARKABLE................................................................................ 5
2.1. Ice floats ....................................................................................................... 5
2.2. Water has a high specific heat ...................................................................... 6
3. VITREOUS WATER IS STRANGE.................................................................... 7
4. VITREOUS WATER MADE SIMPLE ............................................................... 9
4.1. Vitrification is a good preservation method.................................................. 9
4.2. What is vitrification and devitrification? ...................................................... 9
4.3. How can vitrification be achieved? ............................................................ 10
5. OBSERVED RESULTS ...................................................................................... 12
6. REFERENCES .................................................................................................... 14

Vitreous Water 5
1. INTRODUCTION
NASA would like to extend biological studies to other worlds and consequently makes
every effort to look for water on other planets. For us, on earth, aiming at a better
understanding of our earthly life, water is also of primary concern. It is the major
constituent of all living organisms; it is the medium in which life takes place and it
directly participates in most actions that make life work. Nevertheless, water is frequently
neglected and it remains poorly understood in several essential aspects. The systematic
exclusion of water as an object of investigation during the first 50 years of electron
microscopy certainly contributed to the bias against it. Times have changed, however; for
two decades now water has regained the central role it always had in nature. This book
tells the story of its changed fortunes.

Many electron microscopists know little about water and even less about the strange
properties of vitreous water. The present chapter aims at familiarizing them with the
substance with which they are working.
The literature on vitreous water is dangerously abundant and it is rather difficult for those
who want to gain an understanding of it from first principles.1-3 Most of our review of
1988 still holds.4 For pleasure, we recommend the Bibliography of Water by Philip Ball.5
A recent review by Angell6 sums up the present situation, gives abundant references to
the literature and, for the first time, incorporates the observations of cryo-electron
microscopy with the bulk of data from the water field.
In the following presentation, we aim to examine some fundamental questions about

water and vitreous water. This simple presentation — hopefully not simplistic — will
probably raise some specialists’ eyebrows. My excuse is my own incompetence. As a
long-time tourist of “Waterland” I hope that it may offer a useful understanding for
practitioners.
2. WATER IS REMARKABLE
Chapters on water traditionally start with a list of the unusual properties that make water
remarkable. We will consider only two of them: ice floats and liquid water has a high
specific heat. These two properties are easy to grasp and they are convenient proxies for
all the others.
2.1. Ice floats
When a room is full, a good way to gain space is to make order. The contrary holds for
water for which the disordered liquid form is 8 percent more dense than hexagonal ice,
the ordered form we are used to. How can this be? Very simply, it all depends on how
order is made. For example, in your home, you will use up a lot of space if you place only
one book per shelf. This is similar to what happens with water.

6 Handbook of Cryo-Preparation Methods for Electron Microscopy
In a plane, one circle can touch up to six identical nonoverlapping neighbours (see Figure
1.1A). In space, identical spheres can accommodate up to twelve nearest neighbours. In
hexagonal ice or cubic ice, each water molecule has only four next neighbours. Why so
few? Because of hydrogen bonds! Each water molecule offers two and accepts two in a
tetragonal symmetry (see Figure 1.1B).
The rule of four can be extended to the neighbours and to the neighbours of the
neighbours without limit. The result is ice. Two forms are possible: cubic ice (a form with
the same density frequently encountered in cryo-electron microscopy) when the bonds of
neighbouring tetrahedron are coplanar, or hexagonal ice in which tetrahedrons are rotated
by 60° with respect to each other, as is the case in Figure 1.1C. With only four instead of
twefve next neighbours, ice is full of holes. Hydrogen bonds can be seen as long sticks
holding the molecules at a distance rather than glue holding them tightly together.
Because of the large amount of free space between water molecules, other arrangements
involving bending and stretching of hydrogen bonds are possible. The result is several
other forms of crystalline ice.
In most cases, they are only stable under pressure. The hydrogen bond is not very strong.
It tends to break when temperature increases. Above 100°C at 1 atm, it cannot even hold
the water molecules together, and water vapour becomes the most stable form. At
intermediate temperatures, hydrogen bonds constantly break and reform, thus enabling
any molecule to fall down for an instant, into the hole, closer to its neighbour. The
structure is disordered, the density is higher; this is liquid water.
2.2. Water has a high specific heat
Everything that can shake or move does so at the molecular level. This is temperature.
Each of these things that can shake or move is called a degree of freedom. It is
remarkable that each degree of freedom contains on average the same amount of energy 
proportional to the temperature T:  = ½ kT, where the proportionality factor k is the
Boltzmann’s constant, 1.38.10-23 J/oC. Everything else remaining equal, warming means
increasing the amount of energy contained in each degree of freedom. The specific heat
— the amount of heat required to warm one gram by one degree Celsius — is therefore
directly proportional to the number of degrees of freedom n.
n is a remarkable number. It tells us how complicated a system is. It tells us how many
parameters must be introduced in any realistic model of the system. It tells us the
magnitude of the task for those who have the ambition to understand it. We shall try an
exercise. n is measured to be nine for water vapour. For each molecule, three degrees of
freedom are required to specify its location in space (X, Y, Z) and three others for its
orientation (3 Euler angles, three are left for internal vibrations of the molecule.
This makes sense! What about ice, which has n = 8? As compared to water vapour, we
expect the internal vibrations to be the same — that accounts for three degrees of freedom
— but the six involved in location and orientation are lost because they are now nearly
fixed in a crystal. So, where are the other five? Solid-state physics gives us the answer.
These five are the collective vibrations of the crystal lattice; the molecules are allowed to
move, but only in restricted cooperative movements. We come now to liquid water. What

Vitreous Water 7
do we expect? Internal vibrations remain; those are still three. Location and orientation:
Molecules are quite free in liquid water, but not so much as in gas; it will be some number
smaller than six, say, four. We are left with possible cooperative vibrations as in the
crystal, but water is not known for its crystalline sound; it goes “splash” not “cliiing”; so
instead of five, we can perhaps save three. And thus we have: n = 3 + 4 + 3 =10. But no,
this cannot be! The specific heat of liquid water is more than twice that of vapour or of
ice; n = 18. That is eight degrees of freedom more than our guess. We can add one or two
more degrees for positional freedom and better collective vibrations, but how to reach
eight? Where are the missing degrees of freedom? Apparently, nobody knows for sure.
They will probably have to be found in some more cooperative interactions resulting in a
much richer conformational space. Certainly, the years to come will bring a solution to
the problem of the missing degrees of freedom. I like to guess that the solution will have
interesting consequences in various fields where water is important — probably in
biology also. Certainly, it is wise to keep an eye on water.

3. VITREOUS WATER IS STRANGE
Vitrification of water is an old idea, loaded with hopes and dreams. Imagine cooling
water — or an organism — in such a way that it becomes immediately immobile. It is
frozen time, suspended life. A science was born out of these ideas and the state of
suspended animation got its name: the vitreous state.7 It turned out, however, that
vitrification was terribly difficult to achieve except in the case of concentrated solutions.
Even worse, the argument was made that vitrification of pure water is fundamentally
impossible. The discovery, at the beginning of the 1980s, that the vitreous state can be
obtained from liquid water8,9 was first greeted with scepticism, but because the procedure
is simple and reproducible, some way must be found to reconcile theory and practice. We
will explain why it is still a difficult task.
The phase diagram of water sets the scene (see Figure 1.2). There is one domain for the
liquid phase at the upper right, one domain for the vapour at the bottom, and one for ice
on the left. We note that the vapour and liquid domains are not completely separated. The
line goes to the so-called critical point and then stops. This is because there is not always
a difference between these two phases. Above the pressure and temperature of the critical
point, gas and liquid are all the same. On the contrary, ice is always separated from the
domains of liquid and vapour. There is a good reason for the strict separation of the solid
domain: The existence of a critical point between two phases means that it is possible to
go continuously from one to the other. This cannot exist between the liquid, which has no
symmetry, and hexagonal ice, which has hexagonal symmetry. An intermediate state with
“just a bit of symmetry” makes no sense.
At normal pressure, the most stable state is ice below 0°C, it is liquid between 0°C and
100°C and vapour is the normal state above 100°C. These are the conditions at which the
free energy G is minimal as illustrated in Figure 1.3 for the liquid (A) and for the vapour
region (B). This does not mean, however, that the liquid does not exist above 100°C; one
can still warm it above this temperature provided it does not cross the potential barrier
separating it from the vapour phase (see Figure 1.3 C). It is said to be in a metastable
state.

Upon further warming, the advantage of the vapour phase over the liquid phase becomes
greater, and the potential barrier separating them becomes smaller. The higher the
temperature, the more difficult it is to keep the superheated liquid from escaping as
vapour. Upon further warming, there is a point where the potential barrier becomes zero.
Above this temperature, the liquid state just does not exist anymore. The line depicting
this limit on the phase diagram is called the spinodal line (see Figure 1.2). An analogy
may help to clarify the concept. My pen on the desk is in a metastable state; the stable
state is when it lies on the floor. If the table is smaller, the pen can still be on the table
though it will fall more frequently. If the table is removed, the question whether the pen is
on the table or not ceases to make sense. Similarly, there is no question anymore about
the liquid state after the spinodal line. However, the region just before the spinodal line is
particularly interesting. Of course, it is difficult to explore it because the liquid takes
every opportunity to vaporize. In this region, all the properties of the liquid become
strange. A graph presenting the variation of any property of the superheated liquid always
shows that something special is happening when the temperature approaches the critical
line. The signature of the approaching catastrophe is characteristic.
What is the freezing temperature of water? This trap question calls for the wrong answer.
0°C is not the freezing temperature of water at normal pressure; it is the melting
temperature of ice! The freezing temperature of liquid water is somewhere below 0°C,
whenever the liquid finds its way toward the more stable state of ice. This is an important
difference because, in nature, it frequently happens that the temperature fluctuates around
0°C. It makes a lot of difference as to whether water freezes or not. The matter has been
studied in detail and one basic fact stands out: Water resists undercooling at low
temperatures pointing toward a spinodal line at around 45°C. The signature is the same
as for superheated water. It explains why attempts to vitrify pure water were unsuccessful
for so long, just like keeping the pen on the table does not work when there is no table.
But this does not explain the finding in 1980 that ultrarapid cooling can vitrify liquid
water! Perhaps, after all, there is no spinodal at 45°C and the strange behaviour of
undercooled water must be explained differently? Soon after, another finding was made,
complicating the situation even further. Under very high pressure (10,000 bar), and
cooled at liquid nitrogen temperature, hexagonal ice crumbled into a High-Density
Amorphous state (1.15 g/cm3 after restoration of normal pressure) poetically called HDA,
which is clearly different from LDA, the Low Density Amorphous state (0.94 g/cm3)
obtained by rapid cooling of liquid water.10 This could be explained by considering that
disorder can take many forms from low to high density amorphous, just as the disorder in
a room can correspond to any level of heaping. But as usual, water surprises us. When
HDA is rewarmed slowly, it suddenly changes into something resembling LDA. The
transformation cannot be ignored because the abrupt change of density results in large
internal tensions, which make the block fly apart in small fragments. It was also observed
by electron microscopy.11 Thus, it appears that amorphous water is disordered in at least
two mutually exclusive states. In HDA, the disorder is organised in some way and it
becomes unstable above 160°C, at which temperature the disorder must be organized in
a different way. The big trouble is that nobody knows what makes the difference.
Keep an eye on vitreous water, it could tell us important things about water!

Vitreous Water 9
4. VITREOUS WATER MADE SIMPLE
What are the consequences of the above consideration for us, cryo-electron
microscopists? For one thing, and as always in science, it tells us that we should remain
vigilant because we do not know what we are observing. It could be that our vitrified
specimens are fundamentally different from the original material. In them, water is
perhaps organised differently. The future will tell.
4.1. Vitrification is a good preservation method
The other thing we know is that vitrification is an excellent method for preserving
biological material. It makes it possible to observe samples without going through the
major transformations of chemical fixation, dehydration and staining. Even more
important, vitrification is, by itself, a remarkably conservative procedure. This is
demonstrated by a large body of evidence extending from cryo-preservation of embryos
for in vitro fertilization, to cryo-x-ray diffraction for best preservation of atomic structures
in delicate crystals. The quality and the coherence of the results obtained by cryo-electron
microscopy over the last 20 years also contribute to the credibility of the method. 
4.2. What is vitrification and devitrification?
In everyday life, and until we know more about vitreous water, cryo-electron
microscopists can resist becoming depressed by following Occam’s recipe, adopting the
simplest possible view of vitrification and of the vitreous state. In this view, vitrification
is the process by which the viscosity of the sample is increased to such a high value that
molecular movements become negligible before ice crystal formation has time to start.
Any subsequent change is blocked unless considerable forces are applied. Typically, such
forces may arise during cutting thin sections and during observation in the electron
microscope. For a normal experimental time scale, the vitreous state holds as long as the
temperature is well below 135°C. At this temperature, the movement of the water
molecules becomes significant enough so that crystallisation takes place within minutes.
This is the process of devitrification. It can take place at a slightly lower temperature if
one waits long enough or if it is accelerated by electron irradiation.
We note that ice formation by crystallisation from the cooling liquid is very different
from ice formation by devitrification. In the first case, it takes place in the liquid medium
where everything is mobile. As a consequence, the growing ice crystal, which only
incorporates water molecules, rejects all the other constituents of the solution, such as
salts, other solutes and macromolecules. Segregation takes place between pure water in
the ice crystals and all the dehydrated remains of the biological material. Ice formation by
devitrification, on the other hand, takes place under conditions of severely restricted
mobility — just enough to allow water molecules to rearrange into ice crystals. Under
these conditions, no large ramified hexagonal ice crystal is formed, but instead many
compact, sub-µm cubic crystals appear.

They are easily recognized from the characteristic circles of the powder diffraction
spectra. In spite of the fact that little is known about the devitrification process, it
probably takes place with minimum rearrangement of the biological material. One can
infer that the fine structure of a frozen biological specimen in which ice is formed by
devitrification is better preserved than when ice grows from the liquid phase.12 These
considerations are important for freeze-substitution (see Chapter 13), which is always
performed well above the devitrification temperature. Substitution, therefore, deals with
water in the form of crystalline ice. It probably makes a lot of difference whether this ice
was formed by devitrification of a vitrified specimen or whether it was crystallized during
cooling from the liquid phase. The best route is certainly to start with a vitrified sample.
4.3. How can vitrification be achieved?
In principle, the method is simple. It suffices to cool the specimen so rapidly that water
molecules are practically immobilized before an ice crystal starts to form. By chance,
nucleation, the beginning of ice crystal formation, is not an easy task. Even billions of
molecules take a while until, by chance, a group of them happens to be in the right
conformation. Once a crystal has started, the road is clear; it produces its own heat,
further amplifying its growth until all the available water is sequestrated. The question
we may address is, therefore: What is the cooling speed required for vitrification?

The answer is: Nobody knows! Fortunately we know quite well what are the important
parameters, and experience tells us how they can be combined to reach vitrification.
The easiest way is to increase cooling speed by reducing the size of the specimen.
Physics has found that cooling speed increases with the inverse square of the size of an
object immersed in a good cryogen. Calculation13 tells us that, one cm deep in a piece of
meat suddenly placed at liquid nitrogen temperature, the temperature drop is expected to
be in the oC/sec range; at 0.1 mm from the surface, it will reach ca. 104 oC /sec and for a
200 nm thin layer, it is deduced to reach the staggering value of 1011 oC /sec. It is a big
advantage for electron microscopists, and the green light towards cryo-electron
microscopy, that vitrification is relatively easy when the specimens are of a size that is
generally observed in a transmission electron microscope (TEM). Chapters 5 and 6
describe how it can be done.
In general, it is considered that vitrification of pure water is not possible for dimensions
in excess of some µm. Some µm are more than enough if the specimen is to be directly
observed in a transmission electron microscope. Macromolecules, viruses, many isolated
organelles and even whole cells, if they can be squeezed onto a thin layer, can be directly
vitrified in that way.14 Most eukaryotic cells, however, and all tissues are in a larger
range of dimensions. Vitrification at least in the 100 µm scale is needed.

Vitreous Water 11
The cooling speeds calculated above are theoretical maxima. They cannot be
significantly increased because the limiting factor is the thermal conductivity of water.
There are, however, two other parameters that can favour vitrification: the addition of a
cryoprotectant and high pressure. The former acts by reducing the capability of water
molecules to participate in crystal formation. For example, in a saturated sugar solution,
water molecules are so involved in their interaction with the sugar that they cannot
crystallize. Most soluble substances act as a cryoprotectant. The material inside a cell, the
cytoplasm, which typically corresponds to 15 to 30 percent dry weight, is a
cryoprotectant that nature provides freely. We take advantage of it.
High pressure is not so easily offered and its advantage must be paid for.15,16 The method
stems from the fact that water increases its volume upon freezing; therefore, great
pressure applied during cooling makes ice formation more difficult. The trouble is that
pressure has only a minor influence on the volume of water, and considerable pressure
must be applied in order to induce a significant effect. 2000 atmosphere is the optimal
pressure; it is more than that found in the deepest recesses of the ocean. Applying such a
pressure in milliseconds while the specimen is cooled at maximal speed is an engineering
challenge that may have some similarity with gun technology. Furthermore, it raises the
question as to whether this sudden high pressure may induce structural changes.
Certainly it does. The very small compressibility of biological matter suggests, however,
that the changes are small; this has been confirmed experimentally. For example, it is
surprisingly difficult to kill cells by applying high pressure for a short period of time,17
and reports on structural changes produced rapidly by high-pressure freezing concern
quite special systems or very minor transformations.18 Certainly, one should pay careful
attention to this problem; however, my personal guess is that the effect of cooling has
more important structural consequences than high pressure.
There are perhaps other means of helping to achieve vitrification. There are substances
that are said to be especially efficient cryoprotectants. Trehalose is one of them. Nature
has invented surprising ways to avoid — or to help — ice crystal formation. There are
efficient antifreeze proteins in plants and animals,5 and it has been claimed that some
other common proteins have similar properties (G. Prulière, personal communication).
People have also considered using high frequency electromagnetic waves to prevent ice
nucleation. These possibilities are certainly worth creative and careful investigation.
Summarizing, we note that vitrification of pure water or of any diluted aqueous solution
at normal pressure is easy for thicknesses compatible with direct observation in the
electron microscope (µm range). This can be increased by a factor of ten by taking
advantage of the natural cryoprotective capabilities of most biological specimens.
Another factor of ten is gained by resorting to high pressure.19 The 100 µm range for
practical bulk vitrification thus is within reach. Practicalities are described in Chapters 5
and 6.

5. OBSERVED RESULTS
 Figure on the Chapter’s title page  The three forms of solid water frequently
General view: The three forms of ice. encountered in electron microscopy and
their electron diffractogram. At the left is a
single, large, ramified hexagonal ice crystal
grown from a thin layer of water on a
supporting film. At upper right is a layer of
small cubic ice crystals grown from a thin
vitreous layer rewarmed at 135°C. The
electron diffractogram formed of concentric
rings is characteristic of the large number of
small crystals in the observed field. Note
also the dark regions in the two forms of
ice; they are Bragg reflexions taking place
where the crystal has exactly the correct
orientation. At the bottom is a layer of
amorphous water formed by low pressure
vapour deposition on a cold supporting
film. (From Dubochet et al. (1988),4 with
permission.)


 Figure 1.1 The water molecule.  Water molecules and their arrangement
in space. Centred on the oxygen atom, with
tetragonal symmetry, two strong covalent
bonds attach hydrogen atoms at short
distance (0.1 nm), whereas two weaker
hydrogen bonds are stretched at longer
distance (0.2 nm) toward other oxygen
atoms. (A) In a plane, a circle can have up
to 6 adjacent neighbours. (B) A water
molecule and its 4 bonds. (C) A pair of
water molecules attached as in hexagonal
ice. The shadowed plane marks 3 coplanar
bonds.
 Figure 1.2 Phase diagram of water.  Phase diagram of water. For an

explanation, see Section 3.
 Figure 1.3 Free energy and spinodal.  Stability of liquid (L) and vapour (V)
water close to the boiling point.
(A) The liquid is the most stable state
(below 100°C). (B) The vapour is the most
stable state (above 100°C). (C) After the
spinodal line, the liquid state exists no
longer.

Vitreous Water 13
Figure 1.1
Figure 1.2
Figure 1.3

6. REFERENCES
1. Eisenberg, D. and Kauzmann, W. The Structure and Properties of Water, Oxford

University Press, Oxford, 1969.
2. Franks, F. Properties of aqueous solutions at subzero temperatures, in Water: A
comprehensive treatise. Water and Aqueous Solutions at Subzero Temperatures,
Franks, F. ed., Plenum Press, New York, 1982, 215.
3. Debenedetti, P.G. Metastable Liquids, Princeton University Press, Princeton, 1996.
4. Dubochet, J. et al. Cryo-electron microscopy of vitrified specimens, Q. Rev.
Biophys., 21, 129, 1988.
5. Ball, P. H2O; A Biography of Water, Weidenfeld & Nicolson, London, 1999.
6. Angell, C. A. Amorphous water, Ann. Rev. Phys. Chem., 55, 559, 2004.
7. Luyet, B.J. and Gehenio, P.M. Life and Death at Low Temperature, Biodynamica,
Normandy, MO, 1940.
8. Mayer, E. and Brüggeller, P. Complete vitrification in pure liquid water and dilute
aqueous solutions, Nature, 288, 569, 1980.
9. Dubochet, J. and McDowall, A.W. Vitrification of pure water for electron
microscopy, J. Microsc., 124, RP3-RP4, 1981.
10. Mishima, O., Calvert, L.D., and Whalley, E. An apparently first-order transition
between two amorphous phases of ice induced by pressure, Nature, 314, 76, 1985.
11. Al-Amoudi, A., Dubochet, J., and Studer, D. Amorphous solid water produced by
cryosectioning of crystalline ice at 113K, J. Microsc., 207, 146, 2002.
12. Dubochet, J. et al., Freezing; facts and hypothesis, Scanning Microsc. Suppl., 5,
S11— S16, 1991.
13. Studer, D. et al. Vitrification of articular cartilage by high pressure freezing, J.
Microsc., 179, 321, 1995.
14. Garvalov, B.K. et al. Luminal particles within cellular microtubules, J. Cell Biol.,
174 (6), 759, 2006.
15. Moor, H. Theory and practice of high pressure freezing, Cryotechniques in
Biological Electron Microscopy, Steinbrecht, R.A. and Zierold, K. Springer, eds.,
Heidelberg, Germany, 175, 1987.
16. Studer, D., Michel, M., and Müller, M., High pressure freezing comes of age,
Scanning Microsc. 3, 253, 1989.
17. Sato, M. et al. Schizosaccharomyces pombe is more sensitive to pressure stress than
Saccharomyces cerevisiae, Cell Struct. Funct., 21, 167, 1996.
18. Leforestier, A. and Livolant, F. Cholesteric liquid crystalline DNA: A
comparative analysis of cryofixation methods, Biol. Cell, 71, 115, 1991.
19. Sartori, N., Richter, K., and Dubochet, J. Vitrification depth can be increased
more than 10 fold by high pressure freezing, J. Microsc., 172, 61, 1993.

Part I
Cryo-Fixation Methods

Slam-Freezing, Metal-Mirror Freezing 19
CONTENTS
GENERAL INTRODUCTION ...................................................................................... 21

1. PRINCIPLES OF SLAM-FREEZING .............................................................. 22
2. SUMMARY OF THE DIFFERENT STEPS ..................................................... 23
3. MATERIALS/PRODUCTS/SOLUTIONS ........................................................ 23
3.1. Materials ..................................................................................................... 23
3.2. Products ...................................................................................................... 26
3.3. Solutions ..................................................................................................... 26
4. PROTOCOLS ...................................................................................................... 26
4.1. Preparation of the Copper Block ................................................................ 26
4.2. Preparation of the Slammer ........................................................................ 28
4.3. Sample Preparation..................................................................................... 29
4.4. Cryo-Immobilization Procedure ................................................................. 30
4.5. Checking the Quality of Freezing............................................................... 34
4.6. Further Processing ...................................................................................... 35
5. ADVANTAGES/DISADVANTAGES................................................................ 39
5.1. Advantages ................................................................................................. 39
5.2. Disadvantages............................................................................................. 40
6. WHY AND WHEN TO USE A SPECIFIC METHOD .................................... 41
6.1. For Ultrastructural and Immunolabeling Experiments ............................... 41
6.2. To Study Dynamic Events .......................................................................... 41
6.3. At Liquid Helium Temperature .................................................................. 41
8. REFERENCES .................................................................................................... 46


GENERAL INTRODUCTION
Visualization inside a cell, a tissue or a whole organism is a fascinating challenge that has
exploded in the last century with the development of new microscopes at the light
microscopy or electron microscopy level. Light microscopy and, in particular, confocal
microscopy has its own limits due to the resolution (see Chapter 21). Electron
microscopy aims at acquiring ultrastructural information, which also implies ultra-
resolution, and therefore, requires the best preservation of samples. This goal was first
achieved by molecular structural biologists who preserve their samples in the native state
and keep them at low temperature in the cryo-electron microscope. However, only a few
organisms, such as viruses or bacteria, can be observed directly in their native state using
the bare grids method developed by J. Dubochet and M. Adrian (see Chapters 3, 7 and 9).
It has been noted by Bellare et al.3 that the environment (temperature, humidity and
chemical atmosphere) is essential for arresting the sample in its native state, and a
method was devised to achieve the appropriate conditions. This method was improved
for biological applications by P. Frederik,9 who has constructed a robot to plunge samples
in a humidity, and temperature-controlled atmosphere (see Chapter 4). Despite the fact
that electron microscopes are increasingly more powerful and that, for example, in
theory, the electron beam of a FEG 300 kV can go through a 1 µm thick specimen, it is
not possible to observe an entire cell or tissue. The largest organism that can be observed
in a frozen hydrated state are human platelets.10 Therefore, most biological samples that
one wants to observe in an electron microscope have to be sectioned after cryo-
immobilization. Sections can be cut directly after cryo-immobilization (see Chapter 11)
or after freeze-substitution or freeze-drying and resin embedding (see Chapters 13 and
15). Each method has its advantages and drawbacks, which are described in the chapters
mentioned.
The initial step of rapidly cooling to cryo-immobilize a sample is probably the most
critical one It is amazing to read that freezing techniques appeared very early, just before
the 20th century, as the best methods for structural preservation.1 However, “conventional
microscopy” involving chemical fixatives and dehydration of the specimen, thus
inducing various artifacts in samples, had more success than cryo-techniques. This is
perhaps due to the fact that cryo-techniques are difficult to work with and require specific
skills, experience and expensive apparatuses. Today, despite the fact that the apparatuses
are very expensive, cryo-techniques have gained in popularity due to the development of
new techniques, such as cryo-electron tomography (see Chapters 12 and 24).
Cryo-immobilization can be carried out on a simple impact freezing machine (a very

powerful machine unfortunately no longer on the market is Escaig’s slammer7), on a
double-jet freezing machine18 or by high-pressure freezing (see Chapters 5 and 6).
This chapter will concentrate on the slamming method also referred to as impact or metal
mirror freezing. For electron microscopy this method, which was originally described by
van Harreveld and Crowell,27 has been modified and improved by Heuser et al.,12
Escaig,8 Heath,11 and Phillips and Boyne.19 Despite the fact that the method is very easy
to perform and the apparatus is inexpensive, very few scientists use it. “The easiest way
is from time to time the most difficult way.” So, some tips and tricks included in this
chapter can make the slamming method an attractive technique for many applications.

1. PRINCIPLES OF SLAM-FREEZING
Slamming, the process of rapidly projecting  Structural preservation of a 1030 µm

cells or tissue onto a cooled metal block is thick region of the cell pellet or tissue.7
one of the fastest heat transfer methods.
 It is possible to slam large tissue
One can expect to get several micrometers
fragments of up to 2 cm in diameter (see
in depth of well-frozen material.
Chapter 20).
Principle of the method:23 A solid copper  The metal mirror (MM80 from Leica) is
block is cooled in a Dewar flask with liquid described in this chapter.
nitrogen (LN2). Then, using an insulated
 Other apparatuses exist, however most of
manipulator, the copper block is lifted just
them are no longer on the market.
far enough above the LN2 level that it still
remains in the cold atmosphere of
evaporated LN2, thus avoiding frosting.
Cells or tissue have to be prepared so that  This step influences the quality of
they remain in the best physiological state ultrastructural preservation
before slamming:
 Cells have to be carefully pelleted as is
commonly carried out when har-
vesting cultured cells.
 Tissue has to be excised and kept in an
appropriate physiological buffer.
 It is recommended to cut 0.5 to 2.0 mm
 thick slices using the tissue slicer developed
by Sitte et al.23 and previously described by
Stadie and Riggs25 included in the MM80
and CPC package (Leica).
Cells or tissue are rapidly frozen on the 
metal mirror block.

Further procedures:
 Freeze-substitution (FS)  See Chapter 13
 Freeze-drying (FD)  See Chapter 15
 Cryo-sectioning (CEMOVIS)  See Chapter 11
 Room temperature sectioning after FS
or FD

2. SUMMARY OF THE DIFFERENT STEPS
1. Preparation of the copper block
2. Preparation of the apparatus  LN2 is required.

 For safety procedures using liquid
nitrogen, see Sitte et al.24
3. Sample preparation
4. Slamming procedure
5. Checking vitrification
6. Further processing  Cryo-ultramicrotomy: CEMOVIS (see

Chapter 11).
 Freeze-substitution (see Chapter 13).
 Freeze-drying (see Chapter 15).
 Room temperature sectioning after
freeze-substitution or freeze-drying.
3. MATERIALS/PRODUCTS/SOLUTIONS
3.1. Materials
 Impact freezer  Homemade.23
 Some are “precious oldies” like the
helium Cryovacublock (Escaig) previously
sold by Reichert-Jung, Vienna, Austria, and
still used in some laboratories.15
 CPC (with an automatic LN2 refilling
system and microprocessor controlled)
from Leica, Microsystems, Vienna.
 This chapter will focus on use of the
MM80 from Leica (Figure 2.1).
Figure 2.1 MM80 apparatus

 Plastic sheets  Orange plastic spacer of a Leitz binder

0.3 mm thick. Small squares of about 3 mm
× 3 mm are cut.
 Plastic cups  Homemade.
A 2.4 mm diameter indentation can be
made with a punching device, which was
formerly used to make EM grids.
 The borders of the depression have to be
sharp (see Figure 2.2).
 Be careful: A belt punching device is not
applicable because the rim (A) will be
rounded. Consequently, the liquid droplet
does not form a meniscus, but flows over
and will not be frozen properly.
Figure 2.2 Image and sketch of the plastic

cup. The rim (A) has to be very sharp. The
liquid droplet (1.5 µL or less) stays in the
depression and forms a meniscus.


 Very dense polystyrene sheet  2 mm thick.
 Readily available in any laboratory or
can be purchased in a hobby shop.

 Spacer rings  #16860180; furnished with the impact
 freezer machine.
 For better results, fix the rings on
double-sided tape attached on a piece of
polystyrene. The best results are obtained
with 0.3 mm thick rings.
 Tweezers  Those commonly used for EM work,
 e.g., flat, short and antimagnetic (#5
Dumont & Fils, Switzerland).

 Insulated cryo-tweezers  The orange ones furnished with the
machine. (#16701955) Leica Microsystems,
Vienna, Austria.

 Heating plate or oven at 50°C  Useful to rapidly dry the tweezers,

copper block and plastic cups before each
slam.
Figure 2.3 A metal block (arrow) on filter

paper placed on a heating plate, after
slamming. Ice from condensed water vapor
is melted and evaporated. The two holes
visible on the side of the metal block are
useful to insert cryo-tweezers and carry the
metal mirror.
 Precision micropipette with very thin  Preferably the white Eppendorf

adapted tips 110 µL, #0030 073.363, Eppendorf AG,
2233 Hamburg, Germany.
 Binocular microscope  Cold light source.
 Specific material supplied with any of  Refer to the different supplier’s manuals.
the apparatus available will not be
described here:
 Spacer rings  #16860180 Leica.
 Specimen holder  #16701887 Leica.
 Liquid nitrogen Dewar  To refill the slamming apparatus.

 It is furnished with the CPC from Leica.
In the CPC, the temperature is micro-
processor controlled and the LN2 refills
automatically.
 Styrofoam boxes  To transfer the specimen after slamming
to one of the additional processing
machines.
 Cryo-ultramicrotome  For cryo-sectioning with an adapted
holder that clamps flat specimens,
# 16701880 Leica.
 Freeze-substitution apparatus  For freeze-substitution and low-
temperature resin embedding.
 Freeze-drying apparatus  For freeze-drying and low-temperature
resin embedding.
 Cryosorption freeze-drying system
(CFD), Leica.
 Room temperature ultramicrotome

3.2. Products
 Venol™  To clean and polish the copper block
before use.
 Any other copper cleaning product can
be tested and used.
 Liquid nitrogen
 Cotton hydrophilic wad 
 For cryo-ultramicrotomy  See Chapter 11 and A. Leforestier et al.15
 For freeze-substitution  See Chapter 13.
 For freeze-drying  See Chapter 15.
3.3. Solutions
 Fetal calf serum  Or any other concentrated serum protein.
 Water-free acetone  Any supplier, e.g., #00570; Fluka,
 Sigma-Aldrich Chemie GmbH, CH-9471
 Buchs SG, Switzerland.
 To rinse the copper block after
polishing.
 Methanol  M1775GA, Sigma-Aldrich: Fluka,
Buchs, Switzerland.
 Epon  LX112 Embedding kit, Ref. 5.21210,
Inland Europe.
4. PROTOCOLS
4.1. Preparation of the Copper Block

 Polish the copper block with Venol  This step is absolutely necessary.
Use a cotton hydrophilic wad (no  Paper napkins and Kimwipes slightly
paper napkins or other products, such scratch the copper block. The striations will
as Kimwipes) interfere with the slamming process after a
while.
 Make sure that all the Venol is removed
before using the copper block. Any trace of
the cleaning agent can interfere with the
specimen.
 It is not necessary to polish the copper
block after each freezing run, but only
when the sample has scarred the copper
block. In real life, this often happens and
the copper block has to be cleaned when
this occurs.

 An alternative to the polishing step is to

use gold-covered copper blocks that do not
need to be polished after each slam. It just
needs washing with acetone or ethanol
before each slamming.
Figure 2.4 Copper block.
 Rinse the block with acetone or

ethanol.
Figure 2.5 Copper block held with the

orange tweezers.
 Place the copper block in the apparatus

and let it cool.
Figure 2.6 The copper block is placed in

the apparatus. The two holes on the side are
made to pick up the copper block with the
orange tweezers to avoid scratching the
surface of the metal mirror.
 To test if the block has reached the  If the copper block is cold enough, the
working temperature and is ready for drop stays on the copper for a few moments
slamming, fill the chamber with liquid before evaporating.
nitrogen up to the brass plate (see  Has to be done just before slamming.
Figure 2.6, arrow). Cool down the  Wipe all traces of liquid nitrogen off the
orange tweezers by dipping them in copper block (when it is cold!) using a
the chamber and remove a drop of precooled, dry cotton wad.
liquid nitrogen (only one drop!) and  When using the CPC (Leica) instead of
place it rapidly on the copper block. the MM80, the temperature can be set and
controlled automatically.

4.2. Preparation of the Slammer

 Fill the MM80 with liquid nitrogen:  Here the metal mirror freezer MM80
from Leica will be described.
 In this case, the whole procedure has to
be started again: to heat, polish, pour liquid
nitrogen, etc.
 A beep signal is emitted when the
apparatus needs to be filled with liquid
 10 to 15 minutes are needed for the nitrogen. When the system has stabilized,
apparatus to sufficiently cool down. i.e., beeping does not occur every couple of
minutes, the apparatus is ready to start.
 When the system has stabilized, care  If this happens, cryo-immobilization is
must be taken to avoid pouring liquid not efficient. Under these circumstances,
nitrogen on the copper or gold-plated the warm specimen is projected toward the
copper block. liquid nitrogen covering the copper block.
The liquid nitrogen evaporates and
produces an insulated gas layer, which
prevents cryo-fixation.11
 After the light has been switched on, the
TF (TransFer of nitrogen vapor) process is
activated and will automatically start when
the slider is opened. When the TF process is
on, LN2 is evaporated and the cold nitrogen
vapor fills the chamber to keep the interior
dry and at 196°C. This process is essential
to prevent the sample from warming up and
subsequent ice crystal formation.
 Pressing the TF button will cause LN2
vapor formation even if the slider is closed
and the lights are off.
Figure 2.7 Control buttons of the MM80

(in color on the apparatus)
 On the right:
 Mains on button: green lamp
 Mains off button: red lamp
 On the left:
 White button corresponds to the
light and the TF
 Orange button corresponds to the
heating mode.
 Settings for the plastic cups:  Other sample holders can be used and
 Thickness = 0 are referred to in the material section.
 Force = 6  Thickness: Has to be adapted depending
 Speed = 6 on the specimen holder; a simple way to
adapt it is to measure the overall distance of
the mounted specimen holder squeezed
between the thumb and the index.

 Force: Choose the same value as speed.

 Speed: The best results are obtained with
short specimen contact (<5 ms-1) if the
specimen is mounted on a support of low
mass.
 Force and speed have to be adapted to
the specimen. It is preferable to try different
slamming conditions during an experiment
because the results can only be observed
under an EM. For example, keep the
thickness constant and make a series of
slamming from force speed 1 to 9.
Figure 2.8 Top view of the apparatus:

Buttons for force and speed are visible as
well as thickness.
4.3. Sample Preparation

 For freeze-substitution, isolated cells  Or the same serum that was used for the
are pelleted and the supernatant is cell culture.
removed. The cells are then  Serum protein serves as an extracellular
resuspended in a protein-rich medium, matrix for the entire procedure (cryo-
such as undiluted fetal calf serum. immobilization and freeze-substitution).
This extracellular matrix helps to keep the
cells together and acts as a cryoprotectant.
 Some cells should not be maintained at a
high density or pelleted in pure serum
protein because their physiological state can
change (e.g., human platelets are activated).
Figure 2.9 Activation of human platelets

pelleted in pure serum.
 For lyophilization, isolated cells are  They do not need an extracellular matrix
pelleted, the supernatant is removed because they stick together and are not lost
and the pellet is washed with growth during the freeze-drying procedure.
medium.

 Isolated cells to be further cryo-  The cells are maintained together with a
sectioned. cryoglue20 after slamming, especially when
using the Escaig’s or Heuser’s machines.
 The flattened cell pellet can be directly
mounted in the cryo-holder of the cryo-
ultramicrotome.
 Tissue  The use of the tissue slicer especially
developed by Sitte’s group is
recommended.
 Any other tissue chopper also can be
used.
 Slicing has to be rapidly done and the
tissue has to remain wet and preferably in
its physiological medium.
4.4. Cryo-Immobilization Procedure

 Place a thin but very dense polystyrene
layer on the supplied foam support
using double-sided tape as represented
in Figure 2.10.
Figure 2.10
 A: Represents the metal support (3 cm
in diameter) covered with 7 mm thick
foam (Leica ref 16701888).
 B: The very dense 2 mm thick
polystyrene is taped on the foam.
 C: Sample holder.
 D: Plastic cups.
This makes a three-layer sandwich.
 The plastic cups (up to four) are fixed  The Leica rings are fixed the same way.
on the polystyrene layer with double-
sided tape.
Figure 2.11 Four plastic carriers are taped

on the specimen holder.

 Fill the plastic cup. The total volume  Check how much liquid the plastic cup
makes a positive meniscus and the can contain before starting.
liquid stays in the cup. In general, the  The sample should make a small
volume is less then 1.5 µL. meniscus for good cryo-immobilization.
 Take care: No overflowing, otherwise
the sample will not be frozen properly.
 The rings provided with the apparatus
can also be used as described before.
Figure 2.12 Four homemade specimen
carriers are taped onto the polystyrene that
is fixed onto the sample holder and filled
with the cell suspension.
 The use of a very thin tip on a 10 µL
pipette is recommended.
 For a mixed population of small and big
cells, the big cells will move to the
meniscus more rapidly then the small ones.
If small cells are the more interesting ones,
they will never be cryo-immobilized
properly with this method (they will stay in
the not well-frozen part of the sample).
High pressure freezing is the best solution
in this case.
 The sample holder is placed upside
down onto the magnetic ejector. Wait
approximately one minute before
slamming to allow the cells to move
into the meniscus and to avoid freezing
of only ice or medium.
Figure 2.13 The sample is fixed upside-

down (arrow) on the “air-cushioned
injection system” developed by Sitte23 and
stays for one to two minutes.

Figure 2.14 By pressing the red button

(arrow) the sample is slammed onto the
metal mirror surface with the spring-loaded
injector rod.
Figure 2.15 The sample remains pressed

against the copper block for one minute.
 Very important to prevent heat transfer

to the frozen specimen and re-
crystallization.
 Again, wait for one minute.  The sample is not deformed because in
the initial phases of contact and rapid
cooling, pressure is built up slowly as the
23
 Maintain the sample holder with pre- damping air bag slowly empties.
cooled orange tweezers on the cooper  This is crucial to avoid recrystallization
block (with the left hand in Figure due to the bouncing phenomenon.
2.16) and detach it from the ejector by
simultaneously pushing on the tip of
the ejector and pulling out the ejector
(with the right hand in Figure 2.16).

Figure 2.16 Unloading the sample holder

from the ejector.
 The sample and the holder remain on

the copper block. With the precooled
orange tweezers, turn the entire holder
and place it on the copper block.
Figure 2.17 The sample transport unit

(arrow) and the orange tweezers are pre-
cooled.
Figure 2.18 The entire holder is placed on

the brass plate and the impact imprints are
checked.

Figure 2.19 The three well cryo-

immobilized samples are transferred to the
precooled transport unit using a precooled
tweezers.
Figure 2.20 The samples are then

transported in a Styrofoam box containing
LN2 where they are further checked. Arrow:
Plastic carrier to be checked.
 The sample can be ejected from the  This step is done under a binocular
cup using two clamped tweezers by microscope with a cold light source and in
pressing on both sides of the plastic LN2.
cup.
 For freeze-substitution or freeze-drying,
the sample can be cut in four pieces.
 This can be done easily by pressing once
with a fine tweezers in the center of the
sample.
4.5. Checking the Quality of Freezing

 On the copper block:
In general, the impact of the plastic  Slightly white and round traces indicate
cup is seen on the copper block as a that the sample contains some ice and is not
white imprint of frozen water vapor. well frozen: DISCARD IT.
The area surrounding the cup is white,  It is possible that only one or two
but the inner part has to be colorless samples are not well-frozen: Discard those
with no trace of condensation. only and not all four.
 The square, white imprints on the copper
block are made by the plastic cups; the
round imprints are from the sample droplet.

Figure 2.21 In the ideal configuration, all

the samples are “well-frozen.” One can
only see the imprints left by the carrier
itself on the metal block.
 The quality of the well-frozen samples  Since only 10 to 30 μm in thickness of

is verified under a binocular the specimen can be cryo-fixed the whole
microscope (see Figure 2.20). The sample will never be well-frozen with this
frozen sample is removed from the technique. The two phases are clearly
cup, positioned at an oblique angle on visible: the well preserved one (the smallest
a black background and viewed from in general) and the ice phase.
the side. The part that was in direct
contact with the metal mirror should  If only one milky phase is visible,
be colorless, whereas the opposite side discard the sample; it is not well-frozen.
is milky due to light scattering in ice.
 Checking under a cryo-electron  The ideal method!

microscope.  Using electron diffraction within the
cryo-electron microscope is the only
method to see if the sample is really
vitrified.
4.6. Further Processing

Cryo-ultramicrotomy (CEMOVIS)  See Chapter 11.
Freeze-substitution  See also Chapter 13.
The sample may be substituted as follows:

1. In methanol with 0.5% uranyl
acetate
 4872 hours at 90°C.  12 to 24 hours are sufficient if uranyl
 Go to 60°C: 1°C/hour. actetate is added to the substitution medium
 Remain at 60°C for 812 hours.
 Go to 45°C: 1°C/hour.

Embedding in Lowicryl HM20 45°C:  Care has to be taken using these very
(~ 2 hours for each step): toxic and allergenic resins. Please follow
 Methanol/Lowicryl: the manufacturer’s recommendations.
 2+1
 1+1
 1+2
 Pure Lowicryl
 Pure Lowicryl: overnight
Polymerization: 2448 hours with UV
light.
For embedding in Lowicryl K4M at

30°C, follow the same procedure as
for HM20.
2. In acetone with 2% osmium

tetroxide
 72 hours or more at 90°C.  12 to 24 hours are sufficient if osmium is
 Go to 60°C: 1°C/hour. added to the substitution medium.
 Go to 45°C : 1°C/hour.
 Wash in pure acetone: 3 times over
one hour.
 Raise the temperature to 0°C in
one hour.
Embedding in Epon:
 30% Epon in acetone for 3 hours at
0°C.
 70% Epon in acetone for 3 hours while
raising the temperature to 10°C.
 100% Epon in acetone for 3 hours
while raising the temperature to 25°C.
 100% Epon in acetone for 1 hour.
Polymerization:
 Increase the temperature from  Alternative: Place the specimen directly
25°C to 60°C over one hour. in the oven at 60°C.
 Heat polymerize at 60°C for 24
hours.

3. Membrane visualization28  Modification from Ludwig Edelmann.

 2% osmium tetroxide in pure acetone  Starting at -90°C.
for 24 hours.and linear temperature
increase from -90 to – 80°C.
 2% osmium tetroxide in pure acetone
and 2% H20 for 18 hours and linear
temperature increase from 80°C to
10°C.
 Pure acetone: 3 washes over one hour  During this step the container with the
while increasing the temperature to specimen should be closed or covered to
0°C. prevent contamination of the freeze-
substitution chamber with osmium vapors.
Embedding in Epon
 30% Epon in acetone for 1 hour at
0°C.
 70% Epon in acetone for 1 hour while
raising the temperature to 10°C.
 100% Epon in acetone for 1 hour
while raising the temperature to 25°C.
 100% Epon in acetone for 1 hour.
Polymerization:
 Increase the temperature from
25°C to 60°C over one hour.
 Heat polymerize at 60°C for 24
hours.
Freeze-drying  See Chapter 15.
Room temperature sectioning

Fix the block on the ultramicrotome
and check it.
Turn the block in a manner to cut  Except for SIMS (see Chapter 20) where
perpendicular to the slamming the sections are made parallel to the
direction. slamming surface.
 The two layers, seen in the frozen state,
are also clearly visible when checking the
block under the binocular of the
ultramicrotome. The one that is well
preserved has a brownish color, the other is
yellowish.
 Advantage: One can follow the gradient
from well-frozen to the ice-damaged phase
in the electron microscope.

Figure 2.22 Under the electron microscope,

the ice gradient ghost is clearly visible in
this section of a cryo-fixed/freeze-dried
sample (the arrow indicates the direction of
the gradient from the point of impact
toward the less well preserved area of the
sample).
 If the resin is too soft, several solutions  Polymerization can be continued with
exist: natural UV light; let the block stay on a
windowsill or in a homemade box for UV
polymerization (Carlemalm, Villiger,
Garavito and Acetarin in Lowicryl Letters
n°1 or in the Lowicryl user’s manual).
 It is recommended to cut Lowicryl with
a 35° knife.14
 If the blocks are still soft like the
consistency of chewing gum (due to the
high level of humidity in summer, for
example, ) they can be sectioned with the
oscillating diamond knife.26
Stabilization of trimmed Lowicryl

blocks by osmium vapours:6
 0.25 g osmium tetroxide or less is

deposited in an Eppendorf tube. The
Lowicryl block is fixed with Parafilm
on the top of the Eppendorf (the
trimmed pyramid facing the osmium
vapors).
 Wait for 2030 minutes and then
section.

 If the contrast obtained is not

satisfactory, K4M sections can be  This method of Roth et al.21 provides
embedded like Tokuyasu cryo-sections excellent results.
in:
 0.4 mL 2% uranyl acetate.  Contrast can also be increased in the
 1.8 mL 2% methylcellulose. electron microscope by reducing the
acceleration voltage and by choosing a
smaller objective diaphragm.
Other tips and tricks
 The specimen must not be rewarmed.

 If the specimen is warmed above
140°C, ice crystals will appear and
grow. This ice crystal damage can be
avoided when:
 The cover of the apparatus is always
closed when not in use.
 Tweezers are dry and precooled.
 Once prepared, the specimen is
maintained in liquid nitrogen.
5. ADVANTAGES/DISADVANTAGES
5.1. Advantages
 Cryo-immobilization:
 Quicker than chemical fixation.
 Ion gradients are maintained.
 Antigenicity of epitopes is not altered.  As it is the case with chemical fixatives.
16
 Retains more lipids than chemical  Maneta-Peyret et al.
fixation if followed by freeze-
substitution.
 No external mass is added. As happens with chemical cross-
linkers.17
 The process can be reversed with the  Useful for time-resolved slam freezing,
cryo-block falling onto the specimen. e.g., electrically stimulated muscle.5
 Less traumatic then high-pressure  After slamming cell suspensions, cells
freezing. do not float out of the carriers during
substitution as is commonly the case after
high-pressure freezing.13
 Material may be checked for proper  Poorly frozen samples are discarded.
freezing before further processing.  Gain in time.
 The apparatus is inexpensive.  Compared to the high-pressure freezing
machine.

 The unit is easily transportable.  From one lab to another.
 Impact freezing devices are available  A more sophisticated apparatus in which

commercially. the LN2 is automatically refilled and
temperature controlled is also available
(CPC from Leica).
 The only way to cryofix surface  Nearly as large as the cooling block
regions of quite large samples. (3 cm in diameter).
 However, the depth will never exceed
30 μm.
 Copper has a much higher thermal  The same amount of heat is transferred
conductivity than organic liquids. 10,000 times faster through copper than
through liquid propane.
 Physics related to the various freezing  See article by W. B. Bald.2
methods.
5.2. Disadvantages
 Major disadvantage: Only a small  Approximately 10 μm (30 μm in the best
portion of the sample will be properlycase): 10 times less then by high pressure
cryo-immobilized. freezing (see Chapters 5, 6).
 Cell mixtures (i.e., blood cells): Only
 They may not be the cells of interest.
 For this type of cell suspension, it is
the heaviest (largest) cells, which are
first in the meniscus, will be properly
preferable to use the high-pressure freezing
frozen. method.
 Intracellular ice crystal formation may
 The ice crystals grow and cellular
occur and damage cells. organelles are displaced.
 Utmost care has to be taken to keep the
specimen at the LN2 temperature.
 Bouncing phenomenon: If the  The quality of metal mirror freezing is
slamming apparatus is not well drastically reduced.
adjusted, the thermal contact between  Ice crystal formation.
the specimen and the metal block is
interrupted.
 There are significant limitations for  Only the superficial layer of the organ is
freezing organs and tissues. frozen and ice crystals disrupt the fine
structure of the cells underneath.
 Distortion and structural changes can  There is only one paper about structural
occur. changes after MM freezing in the literature,
but the authors concluded that freezing was
not optimal in their experiments.4

6. WHY AND WHEN TO USE A SPECIFIC METHOD
6.1. For Ultrastructural and Immunolabeling Experiments

 This method can be used for  Only 10 to 30 µm are well preserved.
ultrastructural experiments. At low  Sufficient for samples containing small
magnification, a gradient of ice crystal isolated cells or bacteria, limited for
formation can be observed. Only the dissected organs.
best preserved part has to be
investigated.
Figure 2.23 The cell nucleus is the most

sensitive organelle for ice crystal damage.
Here, the reticulate in the nucleus, indicates
insufficient cryo-fixation.

6.2. To Study Dynamic Events

 The method can be used for time lapse  Vesicle exocytosis.12
experiments.  For a review see Ryan and Knoll.22
 For special applications, the metal  Edelmann made a reverse slammer
mirror can be employed so as to apparatus to slam the copper block onto the
catapult onto tissue. heart or the muscle or any organ in situ.5
6.3. At Liquid Helium Temperature

 A heated debate is in the literature  The consensus appears to be that if a
concerning the most appropriate polished copper block is used, there is little
cryogen for best freezing of samples.2 or no gain to use liquid helium instead of
liquid nitrogen.
 The situation is different in the “Escaig”
machine.15 In this case, slamming is done
under vacuum to avoid any moisture
contamination and, for this application,
liquid helium appears to be the best
cryogen. However, this machine is no
longer on the market and only a few
machines are available throughout the
world.

7. OBSERVED RESULTS
 Figure on the Chapter’s title page  HeLa cells were cryofixed using the
MM80.
 Freeze-substitution was done in the
AFS using the protocol 3 for membrane
visualization described in this chapter.
 The sample was then embedded in
Epon.
 Observed results: Numerous nuclear
pores are visible around the nucleus. The
contrast of the membranes is enhanced.
 Figure 2.24  Human dendritic cells were cryofixed

using the MM80.
 Freeze-substitution was carried out in

the AFS using the protocol 1.
 The samples were then embedded in
Lowicryl K4M.
 Observed results: The nuclei, as well as
the nucleoli, are well preserved.
 Figure 2.25  Human dendritic cells were cryofixed

using the MM80.
 Freeze-drying was done in the CFD
from Leica, according to Ludwig
Edelmann (see Chapter 15).
Spurr’s resin.
 Observed results: The nucleus is well
preserved as are the cellular elements in
the cytoplasm.
 NOTE: The cytoplasm, as well as the
nucleus, is very dense indicating that more
cellular elements are preserved with this
method as compared with the freeze-
substitution method.
 N = Nucleus
 R = Endoplasmic reticulum


 Figure 2.26  HeLa cells were cryo-fixed using the

MM80.
 Freeze-substitution was carried out in
the AFS using the protocol C for
membrane visualization described in this
chapter.
Epon.
 Observed results: Two cells are
observed. The membranes are so enhanced
that the mitochondria appear as if they only
contain membranes. This suggests that the
“amplification of membrane contrast” may
have induced different artifacts.
 M = Mitochondria
 N = Nucleus


8. REFERENCES
1. Altmann, R. Die Elementarorganismen und ihre Beziehungen zu den Zellen. Veit

und Co., Leipzig, 1890.
2. Bald, W.B. The relative merits of various cooling methods J. Microsc., 140, 17,
1985.
3. Bellare, J.R. et al. Controlled environment vitrification system: An improved
sample preparation technique, J. Electron Microsc. Tech., 10, 87, 1988.
4. Bennett, P.M. Structural changes in samples cryofixed by contact with a cold metal
block, J. Microsc., 192, 259, 1998.
5. Edelmann, L. The contracting muscle: A challenge for freeze-substitution and low
temperature embedding, Scanning Microsc. Suppl., 3, 241, 1989.
6. Edelmann, L. and Ruf, A. Freeze-dried human leukocytes stabilized with uranyl
acetate during low temperature embedding or with OsO4 vapor after embedding,
Scanning Microsc. Suppl., 10, 295, 1996.
7. Escaig, J. New instruments which facilitate rapid freezing at 83K and 6K, J.
Microsc., 126, 221, 1982.
8. Escaig, J. Control of different parameters for optimal freezing conditions, in The
Science of Biological Specimen Preparation, Revel, J.-P., Bernard, T., and Haggis,
G.H., eds., SEM Inc, AMF O'Hare, IL, 1984, 117.
9. Frederik, P.M. and Hubert, D.H. Cryoelectron microscopy of liposomes, Methods
Enzymol., 391, 431, 2005.
10. Frederik, P.M. et al. Perspective and limitations of cryo-electron microscopy.
From model systems to biological specimens, J. Microsc., 161, 253, 1991.
11. Heath, I.B. A simple and inexpensive liquid helium cooled "slam freezing" device,
J. Microsc., 135, 75, 1984.
12. Heuser, J.E. et al. Synaptic vesicle exocytosis captured by quick freezing and
correlated with quantal transmitter release, J. Cell Biol., 81, 275, 1979.
13. Hohenberg, H., Mannweiler, K., and Müller, M. High-pressure freezing of cell
suspensions in cellulose capillary tubes, J. Microsc., 175, 34, 1994.
14. Jesior, J.C. Use of low-angle diamond knives leads to improved ultrastructural
preservation of ultrathin sections, Scanning Microsc. Suppl., 3, 147, 1989.
15. Leforestier, A., Dubochet, J., and Livolant, F. Bilayers of nucleosome core
particles, Biophys. J., 81, 2001, 2001.
16. Maneta-Peyret, L. et al. Immunocytochemistry of lipids: Chemical fixatives have
dramatic effects on the preservation of tissue lipids, Histochem. J., 31, 541, 1999.
17. Morgenstern, E. and Edelmann, L. Analysis of dynamic cell processes by rapid
freezing and freeze substitution, in Electron Microscopy of Subcellular Dynamics,
H, P., ed., CRC, Boca Raton, FL, 1989, 119.
18. Müller, M., Meister, N., and Moor, H. Freezing in a propane jet and its
application in freeze-fracturing, Mikroskopie, 36, 129, 1980.
19. Phillips, T.E. and Boyne, A.F. Liquid nitrogen-based quick freezing: Experiences
with bounce-free delivery of cholinergic nerve terminals to a metal surface, J.
Electron Microsc. Tech., 1, 9, 1984.
20. Richter, K. A cryoglue to mount vitreous biological specimens for
cryoultramicrotomy at 110K, J. Microsc, 173, 143, 1994.

21. Roth, J., Taatjes, D.J., and Tokuyasu, K.T. Contrasting of Lowicryl K4M thin
sections, Histochemistry, 95, 123, 1990.
22. Ryan, K.P. and Knoll, G. Time-resolved cryofixation methods for the study of
dynamic cellular events by electron microscopy: a review, Scanning Microsc., 8,
259, 1994.
23. Sitte, H., Edelmann, L., and Neumann, K. Cryofixation without pretreatment at
ambient pressure, in Cryotechniques in Biological Electron Microscopy,
Steinbrecht, R.A. and Zierold, K., eds., Springer-Verlag, Berlin, Germany, 1987, 87.
24. Sitte, H., Neumann, K., and Edelmann, L. Safety rules for cryopreparation, in
Cryotechniques in Biological Electron Microscopy, Steinbrecht, R.A. and Zierold,
K., eds., Springer-Verlag, Berlin, Germany, 1987, 285.
25. Stadie, W.C. and Riggs, B.C. Microtome for the preparation of tissue slices for
metabolic studies of surviving tissues in vitro, J. Biol. Chem., 154, 687, 1944.
26. Studer, D. and Gnaegi, H. Minimal compression of ultrathin sections with use of
an oscillating diamond knife, J. Microsc, 197, 94, 2000.
27. van Harreveld, A. and Crowell, J. Electron microscopy after rapid freezing on a
metal surface and substitution fixation, Anat. Rec., 149, 381, 1964.
28. Walther, P. and Ziegler, A. Freeze substitution of high-pressure frozen samples:
The visibility of biological membranes is improved when the subsitution medium
contains water, J. Microsc., 208, 2002, 2002.

Plunge-Freezing (Holey Carbon Method) 51
CONTENTS
1. PRINCIPLES OF PLUNGE-FREEZING ......................................................... 53

3. MATERIALS/SOLUTIONS............................................................................... 54
3.1. Materials ..................................................................................................... 54
3.2. Solutions ..................................................................................................... 54
4. PROTOCOLS ...................................................................................................... 55
4.1. Preparing Holey Carbon Grids ................................................................... 55
4.1.1. Making the plastic solution ........................................................... 55
4.1.2. Making “holey” plastic film.......................................................... 55
4.1.3. Choosing the plastic film .............................................................. 56
4.1.4. Covering grids with the plastic film.............................................. 56
4.1.5. Busting the holes........................................................................... 57
4.1.6. Finishing up .................................................................................. 57
4.2. Plunge-Freezing Grids................................................................................ 59
4.2.1. Glow-discharge treatment ............................................................. 59
4.2.2. Preparing the cryogen/vitrification device .................................... 60
4.2.3. Pipetting the sample on the freshly prepared grid......................... 64
4.2.4. Removing excess liquid ................................................................ 65
4.2.5. Vitrification................................................................................... 66
4.2.6. Specimen mounting and transfer................................................... 66
5.1. Advantages ................................................................................................. 67
5.2. Disadvantages............................................................................................. 68
6. REFERENCES .................................................................................................... 68



1. PRINCIPLES OF PLUNGE-FREEZING
Cryo-electron microscopy of frozen-hydrated macromolecules embedded in a thin layer

of vitreous water is nowadays a well-established method. It was developed more than
20 years ago at the European Molecular Biology Laboratory (EMBL) in Heidelberg, by
the pioneering group led by Jacques Dubochet.1 It is currently used in hundreds of labs
worldwide in order to study biological complexes in their near-native state3-5 (see Chapter
1).
The main advantage of cryo-EM versus air-drying negative staining is that the biological
object is fully embedded in its native environment, namely the vitrified buffer
surrounding it, thus its three-dimensional structure is fully preserved down to atomic
scale.
Typically, the frozen-hydrated specimen is observed suspended across the holes of a

carbon support. The first part of this chapter describes how to prepare the holey carbon
support grids. The second part describes how to apply the sample to the grid and the
vitrification process in order to obtain a thin layer of suspension, to be observed in low-
dose mode in the cryo-electron microscope.

1. Preparing holey carbon grids 

 Making the plastic solution
 Making “holey” plastic film
 Choosing the plastic film
 Covering grids with the plastic film
 Busting the holes
 Finishing up

2. Plunge-freezing grids 

 Glow-discharge treatment.

 Preparing the cryogen/vitrification

device.

 Pipetting the sample on the freshly

prepared grid.

 Removing excess liquid.

 Vitrification.

 Specimen mounting and transfer.

3. MATERIALS/SOLUTIONS
3.1. Materials
 Acetone chamber  With grate
 COLD aluminum block
 Copper grids  200 or 400 mesh
 EM-grade tweezers  Dumont biology N5; Ted Pella Inc.,
Redding, California
 Filter paper, wiping paper  E.g., Whatman grade 15
 Forceps
 Glass microscope slides  Clean
 Glass Petri dish
 Glass pipette with rubber bulb
 Glow-discharge apparatus
 Grid-boxes  For storage (Ted Pella Inc., Redding,
California)
 Humidity chamber  With humidity between 45 and 50%

 Humidity reader
 Parafilm
 Plastic container/liquid nitrogen
Dewar
 Razor blades
 Screwdriver
 Small plastic grid boxes  Ted Pella Inc., Redding, California
 Stop watch
 Vitrification device  Homemade or commercially available
 Vitrobot  Or other commercially available

plunging devices
 Water container  Bigger dimensions than filter paper

3.2. Solutions
 90% acetone
 Cellulose acetate solution (plastic)
 Ethane-35 gas bottle  With a pressure gauge and two security
valves
 90% ethanol
 99% ethyl acetate 
 25 L or 50 L liquid nitrogen Dewar  For grid storage
 Liquid nitrogen  4 L bottle

4. PROTOCOLS
4.1. Preparing Holey Carbon Grids

4.1.1. Making the plastic solution
To make the cellulose acetate solution,2 mix  As an example, if you use 0.026 g of the
the dry cellulose acetate-butyrate with ethyl dry cellulose acetate, you would need to
acetate to a final concentration of 0.15% add 17.3 mL of ethyl-acetate so that the
(w/v). solution would be 0.15% (w/v).
4.1.2. Making “holey” plastic film

To be performed in the humidity chamber if lab’s humidity is not within the 45% to 50%
range.
1. Remove aluminum block from

refrigerator.
2. Put it on ice to keep it cold and wipe

dry.
3. Wait for about 23 minutes before  2 to 3 minutes waiting time is to form
placing glass slide onto block. fine droplets (water condensation).
4. While waiting, clean a few (3 or 4) Use pure or 90% ethanol to clean

glass slides with ethanol. glass slides.
5. Mark on which side of each slide the  Mark slides with a permanent marker
plastic film is to be deposited. (a fine dot is sufficient).
6. Place the glass slide onto the 

aluminum block, designated side up,
for 10 seconds.
7. Remove slide from block. 
8. While tilting the slide at a 60angle, 

pipette a layer of plastic onto it.
9. Set the slide aside beveled and allow 

to dry. 

10. REPEAT steps 59 with no more 
than 3 slides before wiping the block 
dry and starting over again with step 
3. 

4.1.3. Choosing the plastic film

1. Hold the slides up to the light to make 
sure that the plastic has dried evenly  Discard slides that do not meet this
throughout. criterion.
2. Observe the slides in a phase contrast  Check to make sure that the entire length
microscope with an EM grid on top of of plastic has bubbles of the appropriate
it for reference. The bubbles should be size. If not, discard the section with poor
of homogenous shape and size. About bubble formation.
20 bubbles should fit across each side 
of each square (or more reasonably,
most of the bubbles should be of that
size, regardless of orientation).
4.1.4. Covering grids with the plastic film

1. Fill the water container to the top with
distilled water.
2. Scrape the edges of the plastic film to

facilitate floating off the film upon
entry into water.
3. Breathe on the slide and gently

introduce it into the water container,
designated side up, at a 30angle with
the water (the plastic should detach
from the slide and float on the water
surface).
4. Delicately arrange copper grids dull

side up on the plastic film (avoid the
edges of the plastic).
5. To retrieve the grids, overlay filter  The plastic film/grids should have
paper onto them. adhered to the underside of the filter paper.
 Before the filter paper has an
opportunity to slowly begin to sink into the
water, securely grab a portion without grids
under it and meticulously pull the filter
paper out of the water as if peeling the skin
off a fruit.
6. Dry flat for 34 hours (overnight is
even better) with the grids facing up
(the shiny side should be up with the
plastic over it).

4.1.5. Busting the holes

1. Fill the acetone chamber with
acetone, replace grate once this is
done.
2. Take one grid off the filter paper and

place it on a microscope slide; this is
the pretreated grid.
3. Place the filter paper with the grids in

the acetone chamber for one minute.
4. After removing the filter paper with

the grids from the chamber, transfer a
representative grid from the filter
paper to the microscope slide
containing the pretreated grid.
5. Compare the two grids: The  The plastic between the holes should be
representative grid should have quite thin. If it is not thin enough replace
thinner plastic between the the representative grid onto the filter paper
bubbles/holes (i.e., bigger holes in the and put the filter paper with grids into the
representative grid than the bubbles in acetone chamber for 30 seconds and then
the pretreated grid) compared to the compare again. Repeat if necessary.
pretreated grid.
Figure 3.1. Good homemade holey carbon

(the big holes have an approximate
diameter of 5 to 6 m).

4.1.6. Finishing up
1. Coat the grids with carbon using the The thickness of the carbon may vary
carbon discharge protocol (see depending on your own needs. Typically a
Harris.4). When done, you should see 10 to 20 nm layer is a good thickness for
a little bit of carbon deposited holey carbon, but thicker layers may work
underneath the grids (showing that the too, depending on the application and the
holes have popped). sample.

2. To dissolve the remaining plastic Wash the grids with pure ethyl-acetate
from the grids, remove the grids from or chloroform to completely remove the
the filter paper and put them onto the plastic.
grate in the acetone chamber, dull
side up, overnight.
3. Eventually, use a sputter-coater to

deposit a very thin layer of
gold/palladium (15 nm).
4. To make sure you removed the plastic
completely, wash the grids with ethyl-
acetate and rinse with water right
before use.
ALTERNATIVE METHOD
1. Use precleaned glass slides or clean Materials and chemicals for the
them with ethanol or methanol (90%) alternative method:
and Kimwipes (delicate task wipers)
beforehand.  Beaker 250 mL
2. In a small glass beaker, prepare a  Filter paper, wiping paper (e.g.,
0.5% Formvar solution in chloroform, Whatman grade 1)
 Forceps
e.g., 0.25 g in 50 mL chloroform (use
agitation + cover; it takes a while to  Glass microscope slides (clean)
dissolve).  Glass pipette with rubber bulb
 Glow-discharge apparatus
3. Add about 0.5 mL of 50% glycerol  Kimwipes wipers
with a Pasteur pipette on the surface  Optical microscope
of the Formvar/chloroform solution.  Pasteur pipette
You can adjust the volume of glycerol  Ultrasonicator with tip
added to get more or fewer holes
(higher ratio of glycerol/Formvar
solution gives smaller holes).  90% acetone
 99% chloroform
4. Use ultrasonic treatment to make an  90% ethanol and methanol
emulsion of glycerol droplets in the
 99% ethyl acetate
Formvar.  Formvar solution (plastic); Ladd
5. Dip an ultrasonicator tip in mix, Research Williston, Vermont
approximately 1 inch deep, sonicate 1  50% glycerol
minute at maximum power (50 W;  Water (distilled, nanopure, etc)
20-30 kHz). Quickly dip the glass
slides, one by one, vertically in this 
emulsion, blot the sides and let them 
dry vertically (you can use, for 
example, a big beaker with filter Vary the glycerol/Formvar ratio in the
paper at the bottom). You can prepare solution and sonication times to adjust hole
~ 20 slides each time. Make sure to number and size.
dip the glass slides immediately after
the ultrasonic treatment

6. Check the size of the holes with an

optical microscope. You can adjust
the number of glycerol drops and the
sonication time to obtain the hole
number and hole size you want.
7. When the slides are dry, cut the edges

of the holey membrane on one side
and float it on the surface of distilled
water. Cover with copper grids and
pick them up with a piece of filter
paper (or whatever works best). Let
them dry on filter paper. Soak the
entire sheet in methanol for
30 minutes to get rid of the glycerol
to form the holes, and then let them
dry.
8. Evaporate carbon onto the grids,

which are covered with the holey
plastic membrane (as described
before).
9. Remove the Formvar by soaking the
grids in chloroform right before use.
Let them dry on filter paper.
10. Eventually use a sputter-coater in

order to deposit a thin layer of
gold/palladium (15 nm).
11. Wash the grids with full-strength
ethyl-acetate (pure) and then with
water right before use.
4.2. Plunge-Freezing Grids

4.2.1. Glow-discharge treatment
 Glow-discharge treatment of the grids
right before use can help to obtain a good
spreading of the sample across the holes in
the carbon support. Apply right before use
because the charging effect of the carbon
film is only temporary (lasting time may
depend on the apparatus used). It also
cleans the carbon support surface. Glow-
discharging time may depend on the
apparatus you are currently using in your
lab. Please refer to the user’s manual of the
equipment for further operating instru-
ctions.

4.2.2. Preparing the cryogen/vitrification device

1. In the meantime, prepare the
vitrification system. Prepare the
material you are going to need on the
bench (see Figure 3.2), near the
vitrification device, as timing is
important in the next few steps.
Figure 3.2 Forceps, EM-grade tweezers,

Petri dish, Whatman filter paper, EM grids,
Parafilm and, of course, the sample are
displayed on the bench.
2. Make sure the Styrofoam box is placed 

on the vitrification device at the right 
place (see Figure 3.3). 



Figure 3.3 Placement of the Styrofoam

box.




3. In order to make your life easier, it is 

useful to have a direct light coming 
from behind, typically from a desk 
lamp (see Figure 3.4). 













Figure 3.4 Direct lightening coming from
behind is advisable.


4. If you are using a commercially 
available vitrification device (e.g., the
Vitrobot, (see Figure 3.5 and Chapter
4)), please refer to the user’s manual
for further instructions.
5. The small metal cup (typically

aluminum) that will contain the
cryogen should also be placed in the
box; keep the aluminum cup in the
center where the tweezers will fall
later (see Figure 3.6). Test before use.
Figure 3.5 Vitrobot vitrification device.

6. Prepare the gridbox and place it in the

area that will be filled with liquid
nitrogen (see Figure 3.7).
7. The plastic grid storage box should

also be placed underneath the liquid
nitrogen level; make sure it stays on
the bottom by using a metal holder.
Vitrified samples will be placed in this
storage box.
Figure 3.6 Small metal cup is placed in the

box.
8. Make sure you wear lab-certified

protection eyewear BEFORE you start
using ethane (see Figure 3.8).
9. Use your tweezers to pickup a grid

(see Figure 3.9).
Figure 3.7 Prepared grid box
10. Fill the whole box with liquid

nitrogen, do it slowly, no nitrogen
should splash into the aluminum cup
that will contain ethane (see Figure
3.10).

Figure 3.8 Protection eyewear.
11. Start filling the ethane cup (see

Figure 3.11) and wait for the liquid
ethane to reach the right temperature
(see Figure 3.12).
Figure 3.9 Tweezers for picking up a grid.
 The freezing point is a good way to

make sure the vitrification temperature is
reached (liquid nitrogen will be at about
192°C).
Figure 3.10 Box filled with liquid nitrogen
12. Typically, one would wait until the 

ethane starts to become solid and then
add some more ethane to the cup to fill
it to the top (see Figure 3.13).


By repeating this several times the
ethane remains liquid, at the right
temperature to vitrify biological samples.
Figure 3.11 Filling cup with ethane.
Figure 3.12 liquid ethane reaching the right

temperature
Figure 3.13 When ethane starts to become

solid, more is added to fill the cup.
4.2.3. Pipetting the sample on the freshly prepared grid

1. Typically 35 L of the sample 
solution are applied to the freshly 
glow-discharged EM grid (see Figure 
3.9). 

2. Mount the tweezers, holding the grid, 
in the vitrification device (see Figure 
3.14). 


 The grid should already be held with the

tweezers that will be mounted on the
vitrification apparatus.
 There is no need to wait further, as the

sample is not adsorbed to the carbon
support, but suspended in a droplet above
the support.
Figure 3.14 Tweezers mounted in

vitrification device.
4.2.4. Removing excess liquid

1. This is the key step. Use filter paper in  E.g., Whatman type I.
order to remove excess liquid of the
suspension on the grid. Typical
blotting time is 23 seconds at room
temperature and in a relatively dry
(< 60% relative humidity) environment
(see Figure 3.15)
2. The blotting time is also dependent on  The backlight should help one to see a
the concentration of the protein “halo” forming on the filter paper; release
solution and the presence of lipids or the plunging mechanism when the halo
detergents in the buffer. starts to disappear.
 If you are using a temperature and
humidity controlled chamber, blotting times
could be as long as 10 seconds.
 You can use blotting paper that has been

preexposed to water (or buffer) by dipping
or spraying (whichever is more conve-
nient).
Figure 3.15 Using filter paper to remove

excess liquid on the grid.


 You can also set up a homemade system

to blow humid air at 37°C right onto the
grid (typically from behind the grid) while
you apply the blotting paper to the front
side.
 Please refer to the lab equipment’s user’s
manual if you are using commercially
available vitrification devices.
4.2.5. Vitrification
The cryogen usually used for vitrification is
liquid ethane (in some labs liquid propane
is used). Immediately after blotting the grid
(previous step), the sample is plunged into
the cryogen (see Figure 3.16).
 Make sure the sample-releasing

mechanism works efficiently, as the cooling
process must be fast enough in order to
guarantee vitrification.
 Cubic and hexagonal ice will be

observed if the cooling process was too
slow, or if the sample was warmed up to a
temperature above 135°C.
Figure 3.16 Sample plunged into cryogen.
4.2.6. Specimen mounting and transfer

1. After plunging, keep the grid in liquid 
ethane and transfer it quickly to liquid
nitrogen, move it underneath the
nitrogen level to the storage area (see
Figure 3.17).

 The grid must always be kept at liquid
nitrogen temperature after vitrification.
Figure 3.17 Transferring the grid to the

storage area.


2. Use the precooled screwdriver to close 

the grid box firmly (see Figure 3.18).
 The grid can be stored in a Dewar under

liquid nitrogen, or directly mounted on the
cryo-specimen holder and, subsequently,
observed at liquid nitrogen (or liquid
helium) temperature in the cryo-electron
microscope.
Figure 3.18 Closing the grid box with a

screwdriver.


5.1. Advantages
 The electron microscope must be
 Preservation equipped with an anticontaminator in the
Unstained, frozen-hydrated biological column and a low-dose beam-deflection
macromolecules are prepared in a fully unit to avoid beam damage when the
hydrated environment, e.g., water or specimen is observed at liquid nitrogen
physiological buffer. Therefore, they temperature.
are observed in a near-native state. 
 The low-dose exposure for unstained,
frozen-hydrated biological specimens is
usually below 15 electrons/Å2.


 Artifact-free
Unstained, frozen-hydrated biological  Because the protein density is very
macromolecules, prepared and ob- close to that of vitreous water, the
served using the method described amplitude contrast arising from scattering
here are fully suspended in an aqueous is very low. Thus, object visibility in cryo-
environment. Provided that the EM is mainly obtained by defocusing
specimen is fully embedded in the (increasing phase contrast).
vitreous layer, flattening and stain-
related artifacts, often observed with  Defocusing alters the contrast transfer
conventional air-dried, negatively function (CTF) of the electron microscope,
stained samples, are thus avoided. usually leading to loss of resolution in the
 data.

5.2. Disadvantages
 Beam damage  A field-emission gun (FEG) electron
Unstained, frozen-hydrated biological source and computer-assisted CTF
complexes are highly sensitive to the correction are usually required to obtain
electron beam. Low-dose exposure medium to high resolution results.
techniques are required to avoid beam-
induced damage.

 Low signal-to-noise ratio (contrast)  FEG-equipped microscopes operating at
Because of the low-dose mode 200 to 300 kV are nowadays the standard in
required to image unstained protein many EM labs where single-particle cryo-
complexes, the final visibility of the EM and cryo-electron tomography are
object (more precisely signal-to-noise routinely employed.
ratio) is very low. 

6. REFERENCES
1. Adrian, M. et al. Cryo-electron microscopy of viruses. Nature, 308, 32, 1984.

2. Bradley, D.E. The preparation of specimen support films, in Techniques for
Electron Microscopy, Kay D., ed., Blackwell Scientific Publications, Oxford and
Edinburgh, UK, 57, 1965.
3. Dubochet, J. et al. Cryo-electron microscopy of vitrified specimens. Q. Rev.
Biophys., 21, 129, 1988.
4. Harris, J.R. Negative Staining and Cryoelectron Microscopy. BIOS Scientific
Publishers Limited, Oxford, UK, 1997.
5. Kasas, et al. Vitrification of cryoelectron microscopy specimens revealed by high
speed photographic imaging. J. Microsc., 211, 48, 2003.


Controlled Vitrification 71
CONTENTS
GENERAL INTRODUCTION ...................................................................................... 73

1. PRINCIPLES OF CONTROLLED VITRIFICATION ................................... 74
1.1. Effect of Environmental Conditions ........................................................... 74
1.2. Process Control prior to Vitrification ......................................................... 76
1.2.1. Thin film formation upon blotting ................................................ 76
1.2.2. Vitrification in melting ethane ...................................................... 77
1.2.3. Set up the temperature conditions for vitrification........................ 78
1.2.4. Set up humidity conditions for vitrification .................................. 78
1.2.5. Condensing ethane ........................................................................ 78
1.2.6. Checklist ....................................................................................... 79
3. MATERIALS/PRODUCTS/SAMPLES ............................................................ 82
3.1. Materials ..................................................................................................... 82
3.2. Products ...................................................................................................... 83
3.3. Samples ...................................................................................................... 84
4. PROTOCOLS ...................................................................................................... 84
5.1. Advantages ................................................................................................. 93
5.2. Disadvantages............................................................................................. 94
6. WHY AND WHEN TO USE A SPECIFIC METHOD .................................... 94
6.1. Preparation of a Suspension........................................................................ 94
6.2. Preparation of a Viscous Sample ................................................................ 95
6.2.1. Gels ............................................................................................... 95
6.2.2. Creams .......................................................................................... 95
6.3. Preparation of Cells .................................................................................... 96
6.3.1. Bacteria ......................................................................................... 96
6.3.2. Cells growing on an EM grid........................................................ 96
8. REFERENCES .................................................................................................... 99


Fundamental research within the scope of cell, structural biology and nanotechnology is
increasingly focusing on unraveling interactive biological and biochemical processes and
pathways at the macromolecular level. For this, high resolution transmission electron
microscopy (TEM) is indispensable. Of paramount importance is the three-dimensional
visualization of macromolecular structures and molecular machines in their native
hydrated state. Their physical fixation within ultra-thin vitrified ice layers is the crucial
starting point.
This chapter describes the essentials of controlled vitrification, a crucial step for cryo-
observation in particular and a starting point for various cryo-preparative methods in
general.
Cryo-observation of vitrified samples allows the ultrastructural study of macromolecules,
molecular assemblies and cells in their natural (= hydrated) environment (see Chapters 3,
7 – 10, 17). 3-D reconstructions can be obtained from vitrified specimens by applying
careful microscopy and data analysis. 3-D data are based on single particle analysis
(SPA; resolution better than 1 nm) or tomography (resolution better than 3 nm).
Controlled vitrification is the starting point for these 3-D studies and has the potential
advantage of accurate timing of the vitrification process (1 msec precision), thus enabling
time resolved studies.
Prior to vitrification, samples are vulnerable and sensitive to environmental conditions.
Control over humidity and temperature of the environment is the cornerstone of sample
preparation in a critical vitrification procedure,1,2 as will be outlined in this chapter.
Practical procedures will be illustrated using the Vitrobot™ as an instrument for
computer- controlled vitrification.
Figure 4.1: Modern cryo-electron

microscope: (FEI Titan) tailored for 3-D
cryo-EM by tomography and single particle
analysis.

1. PRINCIPLES OF CONTROLLED VITRIFICATION
1.1. Effect of Environmental Conditions

 For cryo-electron microscopy, samples  Thin samples are needed for
(macromolecules, molecular asse- vitrification: less than 100 nm thick for
mblies, virus particles, bacteria etc.) optimum resolution, less than a few
are suspended in buffer and cells micrometers for good vitrification (see
cultured on a flat support. Chapters 3, 7).
 A three microliter droplet is applied to

a hydrophilic grid (e.g., glow
discharged).
 A thin specimen is vulnerable to
 By blotting away most of the liquid, a
100 nm thick film is formed which environmental conditions as heat and mass
comprises about 1/5000 of the appliedexchange occur quickly because the surface
volume. to volume ratio is rather extreme
 The thin specimen that forms is shot  A high-entry velocity in ethane is
into melting ethane. required to optimize heat exchange to
vitrify the specimen. To obtain such
velocities over a short traveling distance
(about 7 cm) requires an acceleration of ca.
3 times the gravitational force.
 The vulnerability to the environment  The fragility at a nanometer scale
will be further explained below with contrasts with the robustness at the
its consequences for molecular and millimeter scale where the grid is subjected
cellular interactions. to blot forces and acceleration forces.
 Depending on the thickness of the  A thin specimen (typically around 0.1

specimen, the temperature decreases to μm), as used for cryo-EM, will thus attain
about the dew point in seconds or split its dew point temperature within 0.1 sec.
seconds. The temperature difference The temperature of a thin film in
thus established between specimen and equilibrium will be lower than the
environment depends on environ- temperature of the environment unless the
mental humidity. environment is saturated with water vapor.
Figure 4.2 Dew point effect and thickness.

Image shows the relation between time and
temperature with a thin specimen at 40oC
and 40% relative humidity (rH).
(Reproduced with permission.3)

 Depending on the temperature and the

environmental humidity, a temperature
gradient will be established. This results in
a heat influx toward the specimen and in
further evaporation of water. The
evaporation velocity is independent of the
thickness of the specimen. Evaporation is
calculated according to the kinetic theory of
gass and assumes evaporation from a free
spanning thin film (evaporation from two
flat surfaces).
Figure 4.3 Evaporation velocity versus

temperature and humidity. (Reproduced
with permission.3,4)
 Preparation at 20oC and 40% rH (typical

environmental conditions in our laboratory)
will result in the evaporation of water at a
rate of 40 nm/sec). A thin film of 100 nm
thus losing water will have an increasing
solute (and particle) concentration.
Liposomes, permeable to water, are
exposed to an increasing salt concentration
and will lose internal volume to maintain
the same osmotic pressure internally as
externally. The internal volume may
decrease by some 50% during “open air”
preparation of a thin film (concomitant with
a two-fold increase of the salt/solute
concentration). The loss of internal volume
is accompanied by a change in shape ―
spherical vesicles turn into concentric
vesicles connected through an “orifice” that
keeps the bilayer continuous.
Figure 4.4 Osmotic effects of lab

environment. Effect of evaporation on the
internal volume of vesicles. (Reproduced
with permission.3,4)

 DLVO theory (Derjaguin-Landau-  An increase in salt concentration has not

Verwey-Overbeek): Molecular dis- only osmotic effects, but also an effect on
tance related to attraction/repulsion. molecular interactions (DLVO theory). The
Relation depending on salt con- balance between electrostatic repulsion and
centration. Balance between van der attraction forces is modulated by the salt
Waals attraction and electrostatic concentration (and valence of the ions
repulsion. involved). For the study of macromolecular
assemblies, molecular machines and cells, it
is essential to maintain the “original” salt
concentration. This can only be achieved in
an environment saturated with a solvent
(e.g., water).
1.2. Process Control prior to Vitrification

 Although vitrification of a specimen is  The final result in this chain of events
a rather simple and straightforward depends on the weakest link and in the next
3step process (apply, blot, and paragraphs a concise description of the
plunge), every step requires careful principles behind the critical steps will be
attention. discussed.
1.2.1. Thin film formation upon blotting

 Preparing a thin vitrified object from  An applied volume of 3 µL would form
bulk material is an important step for a cylinder of about 0.5 mm on a diameter of
cryo-EM. The sample can be a 3 mm carrier grid and upon blotting the
solution/suspension (macromolecules, thickness is reduced to a thin film some 100
macromolecular assemblies, or nm thick.
particles) (see Chapters 3, 7 - 10, 17)
or small, thin cells (see Chapter 12).
A convenient (small) volume of liquid
is used to obtain a “representative”
thin film in which the object is  The reproducible production of a thin
suspended. film is thus an important processing step for
Producing the thin film in a controlled cryo-EM.
and, therefore, reproducible manner is
an asset of the Vitrobot™ technology.
 The blotting action is defined by time  An optimal thin film is formed when it is
and pressure and has to be optimized continuous over the whole grid and thick
for each specimen under investigation. enough to suspend the specimen in a
pseudomonolayer.
 A gentle and controlled movement of  The movement of the blot pads is
the blot pads toward the grid was symmetric and perpendicular to the grid
found to be important as well as the surface.
final pressure they exert onto the grid
in the stationary phase (maximum
contact/pressure).

 The liquid is removed from one side

by gently pressing the blotting paper.
The blot pad on the other side is used
as backing to prevent damage by
lateral or bending forces.
 Both the blot pressure and blotting  For viscous solutions, the blotting time
time can be varied and selected by the can be increased and the pressure should be
user interface. higher than for more liquid samples.
 Blotting is done with filter paper.  The absorbing properties of the filter
paper are important as they also influence
the blotting time and pressure.
 Other types and brands of paper than
those provided with the Vitrobot™ can be
used, but the optimum blotting parameters
have to be established.
 Note: The blotting paper may release
components, e.g., Ca2+, as contaminants
that can influence the structure of the
specimen.
1.2.2. Vitrification in melting ethane
 For vitrification of aqueous solutions,

a high freezing velocity (104oC/sec) is
the key and a coolant with high
thermal conductivity is necessary. In
addition, the specimen is plunged into
the coolant at a high velocity.
 Plunging blotted specimens into  Other hydrocarbons seem to be less

melting ethane (liquid between 88 suited, e.g., liquid methane (161 and
and 172°C) became the standard for 183°C). However, methane is liquid over
vitrification following the monumental a limited temperature range compared to
first observations on vitrified, aqueous liquid nitrogen (196 and 210°C) or liquid
solutions/suspensions by Dubochet at propane (42 and 187°C), which leaves
al.5,6 (see Chapters 1, 3, 7). residues on the specimen.
 Ethane residues may protect the  Most of the excess of liquid ethane
specimen during transfer, which originates from the liquid inbetween the
evaporates in the airlock and vacuum tweezers’ tips or from the fact that the grid
of the microscope. is removed from the ethane too abruptly.
 The excess liquid ethane on the grid

can be limited by performing two
manipulations:

1. Squeeze the tweezers’ tips together

while removing the grid from ethane.
2. Slowly lift the grid from the  The ethane will form a meniscus of the
solution through the ethane surface liquid from the solution to the grid that
while observing how the liquid film detaches perfectly when the movements are
detaches from the grid. done slowly.
1.2.3. Set up the temperature conditions for vitrification
 The Vitrobot is designed to vitrify  A Peltier element is chosen for heating

samples from temperatures between or cooling the instrument with appropriate
4oC and 60oC. accuracy.
 The temperature and humidity prior to  When the conditions chosen deviate
vitrification should be set using the from ambient conditions, time has to be
user interface of the Vitrobot. allotted for equilibration; 5 minutes at
25°C, 45 min at 4°C or 55°C are typical
 The bulk of the sample is normally at figures.
the same temperature as the chamber  The chamber of the Vitrobot could be
of the Vitrobo and, therefore, the used for equilibration or any incubator or
sample solutions should be thermally water bath.
equilibrated as well.
 Special care and attention should be
 At a high relative humidity, a slight
given to the thermal equilibration of
difference in temperature of the tweezers
the tweezers that are holding the and the environment may result in
specimen grid. condensation of water on the tweezers. The
condensed water may flow down on the
grid and dilute the solution.
 We strongly advise preheating the  Slightly overheating keeps you on the
tweezers before they enter the safe side.
chamber.
1.2.4. Set up humidity conditions for vitrification

 The ultrasonic humidifier has to be  Note that switching on the humidifier
filled with distilled water or has an effect on the temperature as
demineralized water following the evaporation of water extracts heat.
instructions of the manual. Although the software critically adjusts heat
input and water evaporation, it can take
some time to reach equilibrium temperature
(> 40°C) and humidity (> 95% rH).
1.2.5. Condensing ethane

 The ethane container should be filled  Note: Liquid ethane can cause severe
up to the rim to prevent precooling of burns and gaseous ethane may explode.
the sample in cold gas before it enters Always work in a fume hood. Follow the
the liquid ethane. safety rules of the manufacturer.

1.2.6. Checklist
 Control the temperature of ethane,  Temperature can be checked with a
which has to be about 172°C. thermocouple or by visual inspection. The
coexistence of liquid and solid ethane
indicates that ethane is at its melting point
(see Chapter 3, Figures 3.12, 3.13).
 Control (or load) all parameters in the  Are the set temperature and humidity
user interface of the Vitrobot. values already attained and stable?
 Check the tweezers of the Vitrobot.  Are they clean and at the same
temperature as the chamber?
 Sample and pipette should be ready for  Liquid nitrogen, cryo-storage box and all
application and eventually brought to necessary tools and utensils should be at
the same temperature as the chamber. hand.
 Replace blotting paper to ensure  The instrument keeps a record of the
reproducible working conditions. number of blotting actions and gives a
Upon every blotting action, the blot- message (“replace blot papers”) when a full
pads are turned to expose a fresh area turn is made.
of filter paper (16 blot actions to make  Blotting papers are mounted on the foam
a full turn). pads with a clamping ring around the
central hole.







Figure 4.5 Glow discharge grids.
 





Figure 4.6 Condense ethane (aluminum

spindle placed on top of the chamber).
Figure 4.7 Replace filter papers.
Figure 4.8 Mount the grids onto tweezers.

Figure 4.9 Set vitrification parameters.
Figure 4.10 Apply the sample to grid.
Figure 4.11 Blot the sample.
Figure 4.12 Vitrify the sample.

Figure 4.13 Put grids into storage

container.
Figure 4.14 Remove the storage container

and transfer the grid to a cryo-holder.
Images are taken at low temperature
(< 170oC) at low dose conditions (see
Chapter 7).
3. MATERIALS/PRODUCTS/SAMPLES
3.1. Materials
 Vitrobot  Automated device for reproducible
vitrification of thin films (FEI Company,
Eindhoven, The Netherlands).
 Cryo-holder and transfer system  The vitrified grid is mounted in a cryo-

holder and then inserted into a cryo-TEM.
 Alternatively, grids may be loaded into
the special cartridge of a Tecnai Polara EM.
 Electron microscope with low-dose  Cryo-samples are very sensitive to

imaging capabilities and low-dose radiation damage caused by electrons.
software  The Low Dose SW suite has been
developed to minimize the electron dose
during searching, focusing and image
acquisition of the specimen.

 Image acquisition  For single particle analysis (see Chapter

7), the Leginon SW suite allows automated
particle picking and acquisition in the most
optimal areas of vitrification.
 For cryo-tomography (see Chapter 12),
FEI’s Xplore 3D and Inspect 3D may be
used for acquisition and reconstruction of
low-dose tomograms.
 Image reconstruction  Various reconstruction packages exist
for particle averaging and/or tomographic
reconstruction analysis, e.g., Spider, eMan,
Imagic (see Chapters, 7, 12, 24).
3.2. Products
 Grids  Quantifoil R2/2 (Quantifoil GmbH, Jena,
Germany) or Lacey carbon film grids
(various suppliers) are used for vitrification.
 A typical mesh size is 300.
 Grids need to be glow-discharged in
order to make them hydrophilic.
 Grid box  Circular grid boxes that fit in the metal
grid box holder of the ethane container are
used. The grid boxes are provided with the
Vitrobot (FEI Company), but can be easily
self-made from a square grid box.
 Liquid nitrogen  Liquid nitrogen is used for precooling
the container and for condensation of the
coolant (e.g., propane or ethane).
 Ethane  Ethane with a purity > 99.9%.
 Liquid ethane enables a very fast cooling
rate (dT> 105K) in order to generate an
amorphous ice layer.
 Note: Liquid ethane can cause severe
burns and gaseous ethane may explode.
Always work in a fume hood. Follow the
safety rules of the manufacturer.
 Filter paper  Provided with the instrument, Schleicher
& Schuell 595 or Whatman/Schleicher &
Schuell 597; both Ø 55 mm.
 Any other filter paper may be used, but
has to be tested for its blotting properties.
 Homemade punch  Used to prepare blotting paper by
punching a hole in standard Ø 5 cm filter
papers.

3.3. Samples
 Various aqueous suspensions: Con-  E.g., proteins, protein complexes,
centration range may vary depending viruses, bacteria or soft matter chemical
on original concentration/density. substances, such as synthetic polymers.
 Isolated cells.  Cells in suspension or cultured as
monolayers on carbon coated grids.
4. PROTOCOLS
1. Glow discharging the grids  Glow discharge helps to spread the

applied solute and prevents the liquid film
from breaking into smaller (micro) droplets
after blotting.
 Upon glow discharge, a grid will remain
hydrophilic for several hours.
 In extreme cases, the material may
adhere to the support film instead of
spanning the holes in solution.
 Aging of the grids may reduce this
tendency.
Figure 4. 15 Bombardment of the grid with

ionized air.
The surface of the support film should be

hydrophilic. A droplet applied should
spread out and behave as one entity
transforming into a continuous thin film.
2. Filling the humidifier.

 Prior to activation of the Vitrobot, the
humidifier must be filled with 60 mL
distilled water through the plastic tube at
the bottom of the apparatus using a syringe
(see Figures 4.16, 4.17).
 NOTE 1: Be aware to fill the

humidifier beaker with sufficient distilled
water prior to enabling the ultrasonic
humidification.
 NOTE 2: The humidifier has to operate
on distilled water.
Figure 4.16 Filling the syringe.

 It is important that a “vacuum” is created

inside the syringe (by “pulling” the piston)
to remove the “air” from the humidifier.
 The humidifier is bayonet-attached to
the bottom of the climate chamber. By
turning and pulling, the humidifier can be
removed.
Figure 4.17 Filling the humidifier.
3. Starting up the Vitrobot.

 Before starting, make sure all cables and
wires are properly connected including the
pressurized air link.
 Activate the pressurized air switch and
make sure pressurized air flows into the
Vitrobot.
 The Vitrobot is switched on with the
hard lock switch on the backside of the
machine.
Figure 4.18. Switching on the Vitrobot.
 After activating the hard lock switch, the

embedded computer, with MS Windows®
operating system, automatically starts up.
The start-up page appears shortly followed
by the Vitrobot User Interface (see Figure
4.19).
4. Defining the Vitrobot user interface.
 The Vitrobot User Interface consists of  On the console screen, the temperature
two pages: The Console and the can be set to any value between 4° and
Options screen (see Figures 4.19, 60°C with the + and – buttons. The actual
4.20). A variety of vitrification temperature read-out is displayed in red
parameters can be set in both pages. (21.8).
 The humidity — displayed in blue (41.1)
— in the climate chamber can be set with
the + and – buttons.
 Enable the humidity switchbox to start
the evaporation (On/Off switch).
 The light in the climate chamber can be
switched on.
 A chronometer records the experimental
time. Once a specific time is set, the
chronometer counts down displaying a
counterclockwise movement.

 In the Controls box on the right, the

sequence of the complete vitrification
process (Place New Grid, Start Process,
etc.) or to Exit is displayed and operated.
 After 16 sequential blottings, the “reset
blot paper” button becomes red pointing out
that the blot papers need to be replaced.
 The Memo box on the left side of the
interface functions as an event logger. All
major actions and warnings are displayed.
Figure 4.19. The Console page.
 E.g., the time that the grid is dipped into

the vial (plunge time), the relaxation time
before blotting (wait time) and the level of
 In the Options screen, additional the liquid in the vial (in case of dipping).
vitrification parameters can be set.  Alternatively, manual application of a
drop of suspension onto the grid through
the right or left side entry port can be
selected.
 In the Miscellaneous box, the use of a
foot pedal switch (instead of the mouse)
and the possibility to switch off the
humidifier during manual application and
plunge-freezing can be selected.
Figure 4.20 The Options page.
 In addition, all essential freezing

parameters can be saved (Save) or loaded
 The semiautomatic grid transfer (i.e.,
(Load) depending on the type of sample or
the automatic movement of the grid
experiment.
from the liquid ethane/propane toward  If the “auto raise ethane lift” option is
the grid box in the liquid nitrogen checked, the coolant container will be lifted
atmosphere) is default activated, but toward the bottom of the climate chamber
can be deactivated by checking the simultaneously with the uplift of the
skip box. tweezers (this can also be done in separate
actions).
 Parameters that affect the blotting  Blot offset determines the force with
process are the number of blottings which the excess fluid is removed from the
(blot total), the time of each individual grid.
blot (blot time) and the position of the  Another factor of significance is the
grid between the blot pads (blot drain time, the time between blotting and
offset). plunge-freezing.

 One of the features in the Mark III  To activate this function, press ADD and
Vitrobot is the option to do repetitive define the application parameters for the
sample application onto the grid and first substance that is to be applied on the
subsequent blotting prior to plunge- grid.
freezing.
 In the Processes section, the Process ID
number, including subsequent parameters,
are displayed.
 By pressing ADD again, the application
parameters for the second substance can be
defined and displayed in Processes as
Process ID 2.
 Up to 20 application cycles can be added
in this way (1, 2, 3, 4 to 20) – Figure 4.23.
Figure 4.21 Repetitive Sample Application
Processes.
 INS allows for inserting application
parameters at a specific position in the
sequence of events that has been defined (1,
2, 3, 4 to 20).
By pressing DEL, selected application
parameters can be removed from the
sequence list.
5. Preparation of the climate chamber.
 Besides setting the proper parameter
conditions (see Section 4.4.), the LED
lights are switched on, the pneumatic
pressure control must be active and the
blotting papers attached to the blot pads by
using the white clipping rings (see Figure
4.23).
Figure 4.22 Mounting the filter papers.
6. Preparation of the climate chamber.

 A glow-discharged grid is attached to the
tweezers. Make sure that the black
clamping ring is fixed in such a way that
the grid does not fall off in vertical position.
Figure 4.23
Picking up the glow-discharged grid.

Figure 4.24 Mounting onto the connection

groove of the plunge axis.
 Mount the tweezers onto the  The tweezers with grid are subsequently
connection grove in the central axis. lifted into the climate chamber by selecting
To do this, first select the ‘Place New the Start Process button in the User
Grid’ button in the Vitrobot User Interface, or alternatively, use the foot pedal
Interface to put the central axis into the switch.
loading position.
7. Preparation and lifting of the coolant
container.
 Prior to setting the coolant container in  The outer ring of the container must be
the proper position for vitrification, it filled with liquid nitrogen. Cooling is a
needs to be precooled and filled with two-step process to be carried out in a fume
liquid ethane. hood; first, the peripheral reservoir will
attain liquid nitrogen temperature; then the
central part having a higher heat capacity
will cool down. Vigorous boiling
(“Leidenfrost effect”) followed by a “calm”
equilibrium indicates that the metal parts
have attained liquid nitrogen temperature.
 The central cup can be precooled with

liquid nitrogen before filling with ethane or
propane.
Figure 4.25 Cooling down the container.
 Note: Wait for complete evaporation of

the remaining liquid nitrogen in the central
part.
 To improve heat exchange/cooling of the
condensing ethane, an aluminum “spindle”
is placed on top of the cup. The spindle is a
temporary heat conductor between the
liquid nitrogen and the liquid ethane and
should be positioned during condensing
ethane and further cooling of ethane down
to 172oC.
Figure 4.26 Filling the container with
ethane.

 When the central cup is at liquid  The ethane container should be filled up
nitrogen temperature, ethane or to the brim to prevent precooling of the
propane can be condensed. With
sample in cold gas before it enters the
appropriate pressure reduction, a
gentle stream of gas is brought into liquid ethane.
contact with the cold metal surface
where it condenses into a liquid. When
enough ethane is condensed, the gas  Immediately after the appearance of the
flow can be adjusted (slightly halo of solid ethane, the metal spindle must
increased) to prevent clogging in the be removed.
feed line (visual inspection will then  The spindle may interfere by creating a
tell that solid ethane is formed). cold gas atmosphere (about –160oC) that
may precool the specimen before it enters
the liquid coolant.
 Thawing the frozen ethane between
spindle and ethane cup, by placing a bolt on
the spindle for 10 seconds, is a more careful
way to remove the spindle.
Figure 4.27 Removal of the spindle.
 When the ethane/propane container is

ready for vitrification, it is placed on the
platform ring under the Vitrobot. Then the
foot pedal switch or the Continue button in
the User Interface must be pressed in order
to raise the container toward the bottom of
the climate chamber.
Figure 4.28 Placing the container onto the

platform ring.
8. Sample application
 For manual application of the sample,
select Manual Application in the Process
parameters of the Options page.
Figure 4.29 Setting the application

parameters.

 The liquid sample can be manually

applied to the grid through the left-
and/or right-hand side entry port using
a pipette.
 When the foot pedal switch is pushed or
the mouse bottom (Continue to proceed) is
pressed, the tweezers slightly lower to
allow the application of suspension through
the side-entry port using a pipette.
 As a consequence, the suspension is
applied on one side of the grid only.
 The advantage of this method is that
only small volumes of the sample (typically
3 µL) are used.
Figure 4.30 Manual sample application.
9. Blotting and vitrification.

 Excess suspension must be removed  Select Continue in the user interface or
from the grid prior to plunge-freezing. use the foot pedal. This activates a slight
uptake of the grid toward the correct
position between the blot pads and a
subsequent blotting of the grid.
 After each blot, the blot pads undergo a
slight rotation to ensure a clean, new area of
filter paper for the next blotting.
Figure 4.31 The blotting process.
 The blotting procedure is immediately

followed by injection of the tweezers
with the grid into the liquid ethane or
propane. The only delay between
blotting and plunge-freezing is
determined by the drain time that can
be set in the Options page and the time
required for removing the shutter from
the hole in the climate chamber.
 The actual plunging mechanism is
mediated by pneumatics at the central axis
combined with the gravitational force.

 After plunge-freezing, both the liquid

coolant container and the tweezers with the
grid are automatically and simultaneously
lowered while keeping the grid in liquid
ethane. This prevents contamination and re-
warming of the freshly frozen sample.
10. Transfer of the vitrified sample.

 After vitrification, the frozen grid must  When the “semiautomated grid
be transferred into a storage box or transfer” is active, the grid is automatically
mounted into the cryo-holder. transferred from the liquid ethane into the
liquid nitrogen.
 The tweezers must be carefully

disconnected from the central axis prior to
positioning the grid into the grid box.
Figure 4.32 Removal of the tweezers.
 To make the grid transfer more

convenient, the coolant container may be
lifted from the support ring and positioned
next to the Vitrobot.
 The excess of liquid ethane on the grid
can be limited by squeezing the tweezers’
tips together and lifting the grid slowly
from the ethane surface while observing
how the liquid film detaches from the grid.
 The anticontamination ring, which floats
on the liquid nitrogen, creates a cold gas
atmosphere, which facilitates the transfer
and minimizes possible ice contamination
on and rewarming of the grid.
Figure 4.33 Transfer of the grid from the

liquid ethane into the liquid nitrogen.
 The outer ring contains a circular storage

grid box for four grids underneath a layer of
liquid nitrogen.

 After the grids have been transferred into

the grid box, the box is sealed with a
special screw and stored in a Dewar with
liquid nitrogen.
 The transfer of the grid box into the
Dewar should be done swiftly to minimize
the risk of rewarming.
Figure 4.34 Loading the grid into the grid

box.
11. Shutting down/switching off the

Vitrobot
 After removal of the tweezers, the
central axis moves into its parking position
(inside the climate chamber), the LED is
switched off and the user interface shuts
down.
 The computer may be shut down.
Figure 4.35 Switching off message.
 When pressing the Exit button in the

interface, two questions appear:
 Pair of tweezers removed? Confirm.
 Pair of tweezers not removed?

Confirm.
 After shutting down the Vitrobo PC,  NOTE: To prevent bacterial growth in
the remaining water in the humidifier the preheated water of the humidifier, it is
should be removed. Pull the metal ring advisable to drain the water supply at the
connector downward to disconnect the end of each working day.
electronic cable from the humidifier.

 Twist and pull the humidifier to unleash

the bayonet connection.
 Emptying the humidifier is performed in
two steps:
 Drain the humidifier to empty the
central reservoir.
 Reconnect the syringe to the plastic
tube at the bottom and remove the
water from the outside reservoir.
Figure 4.36 Emptying the humidifier.
5.1. Advantages
 Fully automated and reproducible  All essential vitrification parameters —
vitrification of aqueous suspensions. temperature, relative humidity, number of
blottings, blotting pressure and drain time
— can be programmed for each individual
application and stored.
 High sample throughput.
 High vitrification quality through  The Vitrobot provides a controlled
controlled environment. environment, preventing cooling of the
specimen and concentration of solutes due
to evaporation before freezing.
 These artifacts are inevitable when using
conventional freezing apparatus.
 The liquid coolant container with an  More efficient “after-freezing” handling.
anticontamination device.
 The transfer from the vitrification  More constant and high yield sample
medium into the liquid nitrogen has output.
been automated.
 Easy to Use.  Cryo-fixation has become easier with the
newly designed and software controlled
Vitrobot User Interface.

5.2. Disadvantages
 The Vitrobot is designed for  Monolayer cell cultures can also be
vitrification of aqueous suspensions vitrified.
only.  In special cases, material suspended in
organic solutes can be vitrified as well.
 Ethane, however, cannot be used as
coolant because it is a “lipid solvent” and it
often dissolves organic solutes even at low
temperatures.
 Decalin and acetone are examples of
organic solutes that can be vitrified upon
cooling in (the inert) liquid nitrogen, but
not all organic solutes will vitrify under
these conditions7 (see Chapter 17).
6.1. Preparation of a Suspension

 Use of the standard vitrification  Best structural preservation.
technique (see Section 4.2) that avoids  Most suitable to recover excellent
adsorption, staining and dehydration structural information upon image analysis.
 Most suitable to detect functional
artifacts.
conformational changes and to analyse
specimen dynamics.
Figure 4.37 Cowpea mosaic virus (CPMV)

embedded in a thin layer of plunge-frozen
water.
 Recommended starting conditions for

suspensions of macromolecules (2 mg/mL)
in aqueous buffers; blotting time of two
seconds, one single blotting operation at
100% humidity, a temperature of 21°C.
Setting blot offset at 0 mm is also a good
point to start.
 Depending on the observed results, the
parameters can be changed accordingly.
 In case the ice layer is too thick, change
the blotting time to three seconds.

 If ice is too thin or no sample is left,

change the blotting time to one second.
 Also, the blot offset, i.e., blotting
pressure on the grid, can be adjusted. Going
up in offset (positive offset values) will
leave more sample on the carbon film
because the pressure on the grid will be
lower. Going down (negative offset values)
will remove more of the sample.
Figure 4.38 Herpes virus capsids imaged in

a patch of vitreous ice.
6.2. Preparation of a Viscous Sample

6.2.1. Gels
 For a sample with higher viscosity than
water, longer blotting time (e.g., five
seconds) is advised.
 A sample containing long fibers like
DNA will have a gel-like behavior and
most often needs to be blotted longer.
Figure 4.39 Shampoo at 25,000×.
6.2.2. Creams
 For even more viscous material, an extra
number of blots may be helpful.
 Creams are difficult to prepare in a film
thin enough to be penetrated with electrons.
 Good results have been obtained with
one second blotting time and three blots.
 Sometimes it is useful to use 300 mesh
or 400 mesh bare grids to obtain a large
area of thin material in the center of a grid
square.
Figure 4.40 Handcream at 25 000×.

6.3. Preparation of Cells

6.3.1. Bacteria
 A suspension of Escherichia coli

bacteria can be vitrified under similar
conditions as a diluted suspension of
macromolecules.
Figure 4.41 Bacterium in a patch of a

frozen sample.
6.3.2. Cells growing on an EM grid

 Growing cells directly on the EM-grid
covered with a thin carbon film is a fast
way of observing cellular structures.
 This is, however, only possible in thin
parts of the cell, e.g., at its circumference or
in filopodiae.
Figure 4.42 Rat liver endothelial cell,

isolated and cultured on an EM grid8 and
frozen in liquid ethane.

7. OBSERVED RESULTS
 Single particle image reconstruction of a

set of 700 Cowpea mosaic virus (CPMV)
images.
 A typical 3-D map of 700 views has a
resolution of 2 nm. Clear structural features
can be observed.
Figure 4.43 CPMV from 700 views.
 Vitrified microtubules imaged using

low-dose mode.
Figure 4.44 Microtubules are reconstructed

by single particle reconstruction.

 T4-phage GP23 proteins form tubes of

different diameters. Tomograms of these
tubes provide more information on their
organization.9
Figure 4.45 Tubes from GP23 of T4-

phage.
 Polymer-layered silicate (PLS)

nanocomposites were subjected to plunge-
freezing and low-dose cryo-electron
tomography.
 The field of nanomaterials is a fast
developing area where the unique
properties of inorganic-layered silicates are
investigated.10
Figure 4.46 Laponite clay-enriched

nanoparticles by cryo-electron tomography
and reconstruction (see Chapter 12).

8. REFERENCES
1. Bellare, J.R. et al. Controlled environment vitrification system: An improved

sample preparation technique, J. Electron Microsc. Tech., 10, 87, 1988.
2. Talmon, Y. Electron beam radiation damage to organic and biological
cryospecimens, in Cryotechniques in Biological Electron Microscopy, Steinbrecht,
R.A. and Zierold, K., eds., Springer-Verlag, Berlin, Heidelberg, Germany, 1987, 64.
3. Frederik, P.M. and Hubert, D.H.W. Cryoelectron microscopy of liposomes,
Methods Enzymol., 391, 431, 2005.
4. Frederik, P.M. and Storms, M.M.H. Automated, robotic preparation of vitrified
samples for 2D and 3D cryo electron microscopy, Microsc. Today, November 2005,
32, 2005.
Biophys., 21, 129, 1988.
microscopy, J. Microsc., 124, RP3, 1981.
7. Butter, K. et al. Direct observation of dipolar chains in iron ferrofluids by
cryogenic electron microscopy, Nature Mat., 2, 88, 2003.
8. Braet, F. et al. The observation of intact hepatic endothelial cells by cryo-electron
microscopy, J. Microsc., 212, 175, 2003.
9. Kükrer Kaletas, B. et al. Structural analysis of self-assembled nanotubes of
bacteriophage T4 capsid protein gp23 by cryo electron microscopy and mass
spectrometry, Microsc. Microanal., 12, 658, 2006.
10. Negrete-Herrera, N. et al. Polymer/Laponite composite latexes: Particle
morphology, film microstructure, and properties, Macromol. Rapid Commun., 28,
1567–1573, 2007.


BAL-TEC HPM 010 High-Pressure Freezing Machine 103
CONTENTS
GENERAL INTRODUCTION 105

1. PRINCIPLES OF HIGH-PRESSURE FREEZING ....................................... 106
1.1. Fundamentals............................................................................................ 106
1.2. Basic Demands ......................................................................................... 106
1.3. BALTEC HPM 010................................................................................ 107
1.3.1. Introduction to the principle .......................................................... 107
1.3.2. Components and function.............................................................. 107
1.4. General Rules for Obtaining Good Results with High-Pressure Freezing 109
2. SUMMARY OF THE DIFFERENT STEPS ................................................... 110
3. MATERIALS/PRODUCTS .............................................................................. 111
3.1. Materials ................................................................................................... 111
3.2. Products .................................................................................................... 112
4. PROTOCOLS .................................................................................................... 113
4.1. General Tips and Hints ............................................................................. 113
4.2. Plant and Animal Tissue........................................................................... 114
4.2.1. Excision and sizing of plant tissue ................................................ 114
4.2.2. Excision and sizing of animal tissue ............................................. 114
4.2.3. Assembly of the specimen sandwich for subsequent freeze-
substitution (FS) ............................................................................ 115
4.2.4. Assembly of the specimen sandwich for subsequent freeze-
fracturing (FF), freeze-drying (FD) and cryo-sectioning (CS) (I) . 115
4.2.5. Assembly of the specimen sandwich for subsequent FF, FD
and CS (II)..................................................................................... 116
4.3. Suspensions .............................................................................................. 116
4.3.1. Preparation of cell suspensions or small organisms (bacteria, algae,
nematodes, etc.) for subsequent FS ............................................... 116
4.3.2. Preparation of cell suspensions, small organisms, liposomes and
emulsions for subsequent FF, FD and CS ..................................... 118
4.4. Monolayer Cell Cultures .......................................................................... 120
4.5. Loading, Freezing, Unloading and Storage of the Specimen Sandwich... 122
5. ADVANTAGES/DISADVANTAGES.............................................................. 124
5.1. Advantages ............................................................................................... 124
5.2. Disadvantages........................................................................................... 124

6. WHY AND WHEN TO USE A SPECIFIC METHOD.................................. 125

7. OBSERVED RESULTS.................................................................................... 126
8. REFERENCES .................................................................................................. 128

Optimum preservation of biological ultrastructure is achieved by cryo-immobilization,

provided that the biological material is not distorted by formation and growth of ice
crystals. At ambient pressure, this requires high cooling rates (> 10,000 K/s). Specimens
of up to 10 µm in thickness can be frozen without visible ice crystal damage and
without the use of intracellular antifreezing agents by using the propane-jet freezing
technique. Thicker samples can be adequately frozen only when the physical properties
of water are changed, e.g., using antifreezing agents or under high pressure (2100 bars).
The thickness of biological specimens to be adequately frozen by high-pressure freezing
is limited to about 200 µm due to the thermal properties of water.15
The high-pressure freezing technique was developed in the 1960s by Moor and Riehle at
the ETH Zurich.7,10 Two different approaches to implement high-pressure freezing were
described:
1. Pressure is applied to the specimen in a small container by a transmission fluid.

The pressurized sample is then cooled.
2. The specimen in the container is located in a pressure chamber. Liquid nitrogen at
high pressure is shot into the chamber where it completes pressurizing and
cooling sequentially (see Section 1, Principles of High-Pressure Freezing).
The first prototype was built using the first approach and patented by Balzers AG in
1968 (GB1230120). This approach showed that ice crystal formation and crystal size are
efficiently reduced by the effect of high pressure. It proved, however, to be unsuitable
for bulky, aqueous specimens such as animal and plant tissue. The scope of application
was limited by the physical properties and geometry of the container (e.g., required
thickness of specimen container to withstand pressure). This system was discontinued
by BAL-TEC (a new model of type I has been constructed by Studer, see Chapter 6) and
a high-pressure freezer according to the second approach was implemented. The new
machine provided superior results for all kinds of aqueous specimens. Today, it is
manufactured by ABRA-Fluid AG as the HPM 010 high-pressure freezing machine.
The vast majority of successful results so far published have been obtained using this
system.

1. PRINCIPLES OF HIGH-PRESSURE FREEZING
1.1. Fundamentals
Applying 2100 bars to the specimen prior to 
the freezing process reduces ice crystal
growth drastically and permits adequate
freezing of aqueous specimens up to
200 µm thickness.
Figure 5.1 The diagram4 shows the stable

state of water as a function of pressure and
temperature. Melting temperature and
homogeneous nucleation temperature reach
a minimum at ~ 2100 bars. Water can be
supercooled to 90°C at 2100 bars, turning
it into a very viscous liquid. I, II and III are
forms of hexagonal ice.
1.2. Basic Demands

Pressure buildup must be as rapid as  Pressure leads to changes in the
possible. Upon reaching 2100 bars, the physical/chemical equilibrium, e.g.,
cooling process should start and proceed as dissociation constant, enzymatic reaction.
quickly as possible to a temperature below A slow increase of the pressure and an
140°C where amorphous (vitreous) ice is extended maintenance of the high pressure
stable. would change the specimen unpredictably.
This must be avoided because the specimen
should be immobilized in a state as natural
as possible.

1.3. BALTEC HPM 010

1.3.1. Introduction to the principle
The specimen is enclosed and protected in a
small volume between two specimen
carriers and frozen in the high-pressure
chamber of the HPM 010. Liquid nitrogen
is used as a pressurizing and cooling  Now available from Boeckeler
medium. In order to coordinate pressure Instruments, Inc., Tucson, Arizona, USA or
increase and temperature decrease, a small ABRA-Fluid AG, Widnau, Switzerland.
volume of alcohol (room temperature) is www.abra-fluid.ch
injected into the pressure chamber so that  A new modified apparatus HPM 100 is
the specimen remains at room temperature available from Leica Microsystems,
during the pressure buildup period. Vienna, Austria.

1.3.2. Components and function

1. Oil pressure pump
2. Bladder-type pressure accumulator
3. Pressure valve (electromagnetic)
4. Low pressure line
5. Low pressure cylinder
6. Pressure piston
7. High-pressure cylinder
8. LN2 Dewar
9. Nonreturn valve


Figure 5.2 Pressure system and Dewar.
10. High-pressure line

11. Nonreturn valves
12. Specimen holder
13. Specimen pressure chamber
14. Quick fastening bolt
15. Specimen
16. LN2 entry lines
17. Pressure sensor
18. Temperature sensor
19. N2 exhaust with silencer
20. Outlet apertures
Figure 5.3 Object head and alcohol

container.

The bladder-type accumulator (2) is pre-

stressed with nitrogen gas to about 150
bars. This pressure is increased to 300 bars
using the oil pressure pump (1). After
starting a freezing cycle, a small volume of
alcohol is injected into the specimen
pressure chamber (13). Subsequently, the
300 bars are converted to high pressure
pushing liquid nitrogen from the high-
pressure cylinder (7) through the high-
pressure line (10) to the specimen pressure
chamber (13). The stream of the pressurized
liquid nitrogen compresses the alcohol
volume and forces it through the outlet
apertures (20) into the environment. The
pressure in the chamber containing the
specimen (15) increases to about 2100 bars
at room temperature before the stream of
pressurized liquid nitrogen cools the
specimen (15).
Figure 5.4 Complete system and

description of the process.
Pressure maintenance indicates the time of

the pressure exceeding 2100 bars (560 ms).
Temperature drops below 0°C only after
pressure has reached 2100 bars. The flow
rate of liquid nitrogen can be controlled by
the size of the outlet apertures — Higher
flow rate of liquid nitrogen = higher cooling
rate = shorter pressure maintenance.
Figure 5.5 The diagram indicates a typical

course of temperature and pressure during
the process.

1.4. General Rules for Obtaining Good Results with High-Pressure

Freezing
 Always use a “fresh” specimen of  It is absolutely crucial that the raw
which as much as possible is known material is of highest possible and known
(e.g., how it was prepared or grown). (defined) quality.
 Always use the smallest possible  The thinner a specimen, the better the
aqueous specimen surrounded by the freezing quality.
thinnest possible metal support.
The thickness of the metal support (carrier

membrane) has higher impact on the
cooling rate than the material from which it
is made.14
Figure 5.6 Effect of wall thickness and

material of the specimen carrier on the
cooling rate at the center of a 200 µm thick
specimen.
 Extracellular fluid surrounding the  Water in its liquid and solid (frozen)
specimen must be used to optimize state is a poor heat conductor. Excess water
heat transfer and prevent ice crystal (and later ice) surrounding the specimen
formation (transmission fluid or decreases the cooling rate and, therefore,
extracellular antifreezing agents). the quality of freezing.
Excess water and air in the specimen  Air acts as an insulator, which is
carriers and air in the specimen itself compressed during high-pressure freezing,
(e.g., plant leaves, antennae of insects) damaging the specimen mechanically.
need to be removed and substituted
with one of the following extracellular
fluids.
 Paraffin oils, e.g., 1-hexadecene.  1-Hexadecene is a universal trans-
mission fluid, which can be used for most
applications due to its favorable properties
such as low surface tension, good heat
transfer coefficient in liquid and solid state,
no osmotic effect, no miscibility with water.
 Dextran/BSA solutions.  Dextran and bovine serum albumin
(BSA) are used as antifreezing agents
usually at concentrations of 20% in a buffer
solution.
 Frozen dextran and BSA solutions are
said to have better cutting properties than 1-
hexadecene with regard to cryo-sectioning
(see Chapter 11).

NOTE: These solutions dry quickly at

ambient conditions leading to an increase
in the dextran / BSA concentration.
Preparation is preferably performed in a
humid environment (e.g., on a wet filter
paper or on agar).
NOTE: BSA dissolves cholesterol from
membranes (defatting).
 Polyvinylpyrrolidone (PVP) solution  Alternative antifreezing agents, e.g., for
specimens that are affected by 1-hexa-
 Hydroxyethyl starch solution decene (e.g., cheese).
 During freezing, the specimen  The specimen may be squeezed out of

sandwich must be kept firmly in place the carrier sandwich if it is not closed
by the sample holder. properly, e.g., suspensions. In addition,
alcohol may penetrate the specimen causing
damage.
1. Sample excision and sizing
 The sample is excised and sized to fit

into the cavity of the specimen carrier
(max. 2 mm diameter, 200 µm
thickness) using a biopsy gun,2 hand
microtome, vibratome, scalpel, razor
blade, etc.
2. Assembly of the specimen sandwich
 Cavity size is selected according to the

dimension of the specimen.
 The residual space between carrier
cavity and specimen must be filled
with an extracellular fluid (e.g.,
1-hexadecene).
Specimen assembly is completed with a
second carrier (e.g., flat carrier).
3. Loading of the specimen holder
 The complete specimen sandwich is

transferred to the specimen holder of
the high-pressure freezer.

4. Freezing of the specimen
 The holder is inserted into the

specimen pressure chamber of the
high-pressure freezer and locked.
Subsequently, the specimen is frozen
by a one-button operation.
5. Unloading of the specimen
 Immediately after the freezing process

the holder is pulled out and transferred
to liquid nitrogen in the unloading
device. The specimen sandwich is
removed from the holder.
6. Storage of the specimens
 The specimen in the carrier is stored in

liquid nitrogen until further
processing.
Figure 5.7 (1-6) Summary of the particular

steps.
3. MATERIALS/PRODUCTS
3.1. Materials
 Aluminum Type A specimen carrier Note: The BAL-TEC specimen carriers and
(100 µm/200 µm cavity) consumables are now supplied by Leica.
Figure 5.8 BAL-TEC LZ 02135 VN.
 Aluminum Type B specimen carrier

(flat, 300 µm cavity)
Figure 5.9 BAL-TEC LZ 02136 V.

 Aluminum Type C specimen carrier

(100 µm/200 µm cavity)
Figure 5.10 BAL-TEC LZ 02137 VN
 Gold specimen carrier, cylinder  BAL-TEC LZ 02125 VN

shaped indentation (bottom carrier)
(3 mm diameter, 0.8 mm thickness)
 Gold specimen carrier, dome-shaped  BAL-TEC LZ 02124 VN

indentation (top carrier) (3 mm
diameter, 0.8 mm thickness)
 Gold tubes (diameter: outside 300 µm,  BAL-TEC LH 01846 VN
inside 200 µm)
 Cellulose capillary tubes  BAL-TEC LH 01843 VN
 Promag biopsy gun (Set)  BAL-TEC BU 012 139 –T
 Tissue puncher 2 mm  BAL-TEC LH01847 KN
 Spacer rings 0.5 mm  BAL-TEC LH 01842 VN
 Athene G209G, BAL-TEC LZ 04884
 Gold grids, 8…10 µm thickness
KN
 Single hole EM grids, 50 µm thickness  Agar G2620C, BAL-TEC LZ 04885 KN
 EM grids, 400 mesh 12…15 µm  Pelco 1GC400, BAL-TEC LZ 04886
thickness KN
 Sapphire discs (50 µm thickness,  BAL-TEC LH 01845 VN
3 mm)
 Universal holder for freeze-fracturing  BAL-TEC LZ 04746 VN
or cryo-scanning electron microscope
(SEM).
3.2. Products
 1Hexadecene  Fluka 52278, BAL-TEC LZ
02079 KN
 Bovine serum albumin (BSA)  Sigma A4503
 Dextran  Sigma D4133
 Ethanol  Merck 100971

 Isopropanol  Merck 109634



4. PROTOCOLS
4.1. General Tips and Hints

1. Stereomicroscope  A stereomicroscope with low magni-
fication and a flat base plate is useful for
the preparation and handling of specimens.
2. Loading station
 A plastic Petri dish is helpful as a
loading station for the specimen holder. The
rim has the proper height to stabilize the tip
of the holder. The Petri dish must be pinned
down with a weight or tape to prevent
tilting during operation.
Figure 5.11 Simple loading station using a

plastic Petri dish.

3. Freezing cycle  Time between sample excision and

freezing should be as short as possible in
order to keep the specimen in its native
state. Working in pairs facilitates proc-
essing. Usually, one person prepares and
loads the specimen while a second person
operates the high-pressure freezer (freezing,
unloading, and drying of holder).

4. Specimen carriers  Always use clean specimen carriers.
Aluminum carriers are cleaned in detergent
solution and/or ethanol in an ultrasonic
bath.
NOTE: Do not use chloroform, acetone
or acids. These chemicals disintegrate the
aluminum carriers.

 Extracellular fluids NOTE: Remember that dextran or BSA
 solutions dry very quickly increasing the
BSA/dextran concentration. A Petri dish
with agar or a wet filter paper is helpful to
prevent drying of the solution and
specimens.



6. Preparation of the HPM 010 machine  The machine must be started up and
checked for faultless operation.
 Some test shots must be performed using
the temperature measuring holder. Pressure
maintenance and cooling time/maintenance
should meet specifications.
 Check the level of alcohol in the
container. A lack of alcohol will lead to
severe ice crystal damage.
NOTE: For subsequent freeze-
fracturing, isopropanol should be used
(melting point 88°C) instead of ethanol
(melting point 114°C).

4.2. Plant and Animal Tissue

4.2.1. Excision and sizing of plant tissue
Plant leaves not exceeding a thickness of  1-Hexadecene ideally is evacuated
200 µm can be excised using a 2 mm before use because air can dissolve it.
biopsy punch.  Plant tissue containing air needs to be
submitted to a slight vacuum in 1-
Plant leaves thicker than 200 µm or other hexadecene to remove the air (see Chapter
plant material must be sized and trimmed 6).
using a scalpel, razor blade or hand  The choice of extracellular fluid may
microtome to fit the cavity of the specimen vary depending on the scientist’s
carrier. preferences. A variety of different protocols
 are found in the literature.


4.2.2. Excision and sizing of animal tissue
Animal tissue is dissected and sized using a NOTE: The Promag biopsy gun system
scalpel, razor blade or biopsy gun. has proven to be very suitable.2 The
indentation of the smallest available biopsy
The time between excision of the tissue and needle (8 mm × 0.4 mm × 0.25 mm) is
the freezing process must be reduced to a larger than the space offered by the
minimum to prevent artifacts caused by the specimen carriers for high-pressure freezing
separation of the tissue from its natural (2 mm diameter, 0.2 mm thickness).
environment. However, the excised tissue is often smaller
(shorter) than the indentation of the needle,
The use of a biopsy gun allows fast excision e.g., depending on the size of the tissue and
and sizing of the tissue at the same time.2 its consistency. In many cases, the excised
Dissection of tissues using a scalpel or tissue “sausage” may be transferred directly
razor blade can be performed in a buffer to the specimen carrier containing 1-
solution dedicated for use with tissues or in hexadecene. The “sausage” must be cut into
1-hexadecene. shorter pieces if it is too long. Tissue can be
squeezed slightly without causing damage
due to its flexibility.

4.2.3. Assembly of the specimen sandwich for subsequent freeze-substitution (FS)
Figure 5.12 Assembly of the specimen

sandwich for subsequent FS. Left, bottom to
top: type A carrier with 1hexadecene (e.g.,
200 µm cavity), tissue, flat type B carrier
wetted with 1-hexadecene. Right, assembled
specimen sandwich.

For subsequent freeze-substitution (see NOTE: The cover carrier is dipped into
Chapter 13), the sandwich is usually 1-hexadecene before completing the
composed of a Type A carrier and a flat sandwich, regardless of the kind of extra-
cover Type B carrier. Whenever possible, cellular fluid used. This eases the removal
the 100 µm cavity is selected to carry the of the cover carrier after freezing.
specimen for superior freezing quality.
Punched leaf discs of 2 mm diameter fit NOTE: When using 1-hexadecene, crack
directly into the cavity of the Type A the frozen 1-hexadecene by tapping the
aluminum carrier (one side 100 µm cavity, surface with pointy tweezers. This ensures
other side 200 µm). that the freeze-substitution mixture
Tissues excised with a biopsy gun can be penetrates the specimen. This is especially
transferred directly from the needle into the important when the specimen is completely
cavity of the carrier. embedded in 1-hexadecene after freezing.1
After freezing the flat cover can be shifted
away and the sample stored in liquid NOTE: Excess 1-hexadecene does not
nitrogen until further processing. affect the freezing.
4.2.4. Assembly of the specimen sandwich for subsequent freeze-fracturing (FF),

freeze-drying (FD) and cryo-sectioning (CS) (I)

sandwich for subsequent FF, FD and CS.
Left (bottom to top): Type A carrier with
1hexadecene in 100 µm cavity facing up,
tissue, type A carrier with 1-hexadecene in
100 µm cavity facing down. Right,
assembled specimen sandwich.

For freeze-fracturing (cryo-SEM or replica, NOTE: Pure mechanical attachment is
see Chapter 22), freeze-drying (see Chapter essential if the specimen is to be “heated”
15) and cryo-sectioning (see Chapter 11), for sublimation after fracturing or for
the sandwich is composed differently than freeze-drying. Hence, the carrier
for FS. The specimen is enclosed between containing the specimen or the complete
two Type A carriers with the 100 µm sandwich must be mechanically attached
cavities facing each other. onto a special holder for freeze-fracturing
or onto a holder adapted for the cryo-
ultramicrotome
.

After freezing, the sandwich is either kept NOTE: High-pressure frozen aqueous
together for freeze-fracturing (sandwich material is very brittle. Often, the
fracturing) or opened to get access to the specimen is cracked or even lost during the
specimen with the knife in the cryo- fracturing process. This problem is reduced
ultramicrotome or freeze-fracturing device. by using the alternative approach described
For subsequent cutting of frozen hydrated below.
sections by cryo-sectioning, the specimen
can be also removed from the carrier and
glued (using a cryoglue5) to a stub for the
cryo-ultramicrotome.
4.2.5. Assembly of the specimen sandwich for subsequent FF, FD and CS (II)

Left, bottom to top: type C carrier with 1-
hexadecene in 100 µm cavity facing up,
tissue, type C carrier with 1-hexadecene in
100 µm cavity facing down. Right,


For FF, FD and CS, the sandwich can be NOTE: This approach is useful for
formed alternatively by using type C slit subsequent freeze-fracturing because the
carriers. The specimen is enclosed in the slit specimen is mechanically stabilized.
between two carriers with the 100 µm
cavities facing each other. Holes in the
type C carrier surface enable proper
alignment.
4.3. Suspensions
4.3.1. Preparation of cell suspensions or small organisms (bacteria, algae,
nematodes, etc.) for subsequent FS
  Cell suspensions or organisms normally
must be concentrated before freezing in
order to achieve a reasonable density for
EM investigation.

 Concentration by centrifugation  The suspension is centrifuged into a
viscous paste at a centrifugal force relevant
to the sensitivity of the specimen.
Subsequently, the paste is either placed
onto the specimen carrier directly or first
drawn into a cellulose capillary tube.

 Assembly of the specimen sandwich:

direct loading of specimen carriers.
sandwich for subsequent FS. Left (bottom
to top): Type A carrier with suspension in
100 µm/200 µm cavity facing up, flat type
B carrier with 1-hexadecene facing down.
Right, assembled specimen sandwich.
The specimen paste is transferred into the

100 µm (or 200 µm) cavity of a type A NOTE: It is important that no air is
specimen carrier with a spatula or included during transfer of the paste into
toothpick. The sandwich is completed with the carrier.
a flat type B carrier dipped previously into
1-hexadecene. NOTE: The cell suspension aggregate
usually stays together during freeze-
Centrifugation of a suspension can also be substitution and embedding. The
performed in a closed plastic pipette tip aggregate may break into a few pieces, but
(closed by heat). After centrifugation, the will not dislocate as single cells in the
pipette tip is opened with a razor blade and mixture.
the paste is squeezed directly into the
specimen carrier using the appropriate
pipette.
 Assembly of the specimen sandwich:

loading of specimen carriers using
cellulose capillary tubes.1

sandwich for subsequent FS. Left (bottom
to top): Type A carrier with 1hexadecene
with the 200 µm cavity facing up, filled
cellulose capillary tube, flat type B carrier
with 1hexadecene facing down. Right,
The viscous specimen paste is normally NOTE: Depending on the suspension, it

drawn into cellulose capillary tubes by is often difficult to close both ends.
capillary force. The tube may be attached to However, freezing is not affected and only
a pipette tip using wax or glue to increase a minor part of the suspension is lost during
suction with an attached pipette. This freeze-substitution or embedding. More-
approach is helpful for very thick and over, an open end supports the freeze-
viscous suspensions. substitution process. For this purpose, tubes
may be closed on one side only.

The filled tube is immediately immersed in NOTE: Dry, fresh cellulose tubes may be
1-hexadecene or another medium to prevent squeezed at one end. In that case, the ends
drying of the suspension. The tube is cut of the tube are cut off with sharp scissors
into small pieces (<2 mm) that fit in the before use.
specimen carrier. Cutting is best performed NOTE: Always prepare a fresh specimen
with a blunt scalpel blade. The tube is then tube for freezing. Do not leave the tubes in
cut and squeezed (closed) at the same time. 1-hexadecene for an extended time period
Tube pieces (1 to 4) are transferred into the prior to freezing.
200 µm cavity of a type A carrier dipped in
1-hexadecene. The sandwich is completed NOTE: Do not use cavities smaller than
with a flat type B carrier wetted with 150 µm. The tubes would be completely
1hexadecene. squeezed in a 100 µm cavity.
2. Alternative concentration method.
 Assembly of the specimen sandwich.

sandwich for subsequent FS (alternative).
Left ‘bottom to top): Type A carrier with
specimen, flat type B carrier with 1-
hexadecene facing down. Excess medium is
Suspensions consisting of cells or blotted off with filter paper. Right,
organisms may be prepared in the following assembled specimen sandwich.
way: Cells or organisms (e.g., nematodes,
cell aggregates) are transferred individually
or in groups onto the specimen carrier
containing the original medium. After
transfer, the excess medium is blotted off
with filter paper and the sandwich is
completed with a flat type B carrier dipped
into 1-hexadecene.
4.3.2. Preparation of cell suspensions, small organisms, liposomes and emulsions for
subsequent FF, FD and CS
Preparation of cell suspensions, small  Concentration of the specimen (except
organisms, liposomes and emulsions for liposomes, emulsions) is performed as
subsequent FF, FD and CS. described above.
 Assembly of the specimen sandwich

for “thin” suspensions.
sandwich for subsequent FF, FD. Left
(bottom to top): Scored flat type B carrier
with flat side up, spacer grid with
suspension, scored flat type B carrier facing
down. Right, assembled specimen
sandwich.

Figure 5.19 The flat side of the carriers

should be scored with a scalpel in order to
enhance the adhesion of the water/ice. By
using this method, the majority of the
sandwiches will adhere after high-pressure
freezing for subsequent freeze-fracturing.
Suspensions consisting of particles or NOTE: No 1-hexadecene is needed.

structures smaller than 10 µm (bacteria,
cells, liposomes) are frozen in a sandwich NOTE: This approach is not suited for
of two flat type B specimen carriers with a subsequent CS.
transmission electron microscope (TEM)
grid as a spacer (12 to 15 µm) in between.
The grid is simply plunged into the
suspension and clamped between the two
flat sides of the type B carriers.9

for “thick” and ”thin” suspensions.
Left (bottom to top): Type A carrier with
specimen in the 100 µm cavity facing up,
type A carrier with specimen in the 100 µm
cavity facing down. Right, assembled
specimen sandwich.
Suspensions consisting of particles or NOTE: Ensure that no air is trapped in

structures larger than 10 µm (cosmetic the carrier cavity. Depending on the
emulsions, cement solutions) are frozen in a sensitivity of the specimen, toothpicks or a
sandwich consisting of two type A pipette may be used for filling the carriers.
specimen carriers with the 100 µm cavities Both carriers are filled with the specimen
facing each other. before assembling the sandwich.
This approach may be applied to CS of any NOTE: No 1-hexadecene is needed.
type of suspension. In this case, the NOTE: Specimens may also be frozen
sandwich is opened and the carrier(s) with the gold specimen carriers for
containing the specimen is mounted on a subsequent FS.
corresponding holder for CS.

for “thin” suspensions for CS (altern-
ative). Figure 5.21 Assembly of the specimen
sandwich for subsequent CS. Left (bottom
to top): Stainless steel spacer ring 0.5 mm,
filled gold tube, stainless steel spacer ring
0.5 mm. Right, assembled specimen
sandwich.14

Alternatively, suspensions can be frozen in NOTE: The tube is cut into pieces of
gold tubes for subsequent CS. The gold about 5 mm length using a 200 µm thick
tubes have an inner diameter of 200 µm and wire and a scalpel. The wire with a
an outer diameter of 300 µm. The diameter corresponding to the inner
suspension is simply drawn into the tube by diameter of the tube (200 µm) must be
capillary forces. Thereafter, the tube is inserted into the tube prior to cutting in
clamped between two 0.5 mm spacer rings order to prevent squeezing of the gold tube.
for freezing.
NOTE: Sensitive structures of some
Turbulence in the pressure chamber of the suspensions (e.g., certain emulsions) may
high-pressure freezer can sweep the tube be damaged by the capillary forces of the
away. Therefore, the tip of the specimen tube.
holder is turned 45° clockwise: Primary
stream of liquid nitrogen will hit the small
face of the tip (rather than the tube) and,
therefore, reduce the shearing forces.
4.4. Monolayer Cell Cultures

Varying substrates have been developed for  Preparation of the Sapphire discs (50 µm
growing and processing monolayer cell thickness, 3 mm diameter): Sapphire discs
cultures by high-pressure freezing: Sapphire are thoroughly cleaned with sulfuric acid,
discs,13 Matrigel,12 PET discs,8 and Aclar rinsed with water and ethanol, and further
discs.3 scoured in a plasma cleaner. Subsequently,
the disc is coated with a carbon layer of
The assembly of the specimen sandwich is about 10 nm and a “2” is scratched into the
basically the same for all substrates (discs surface. Carbon is a good substrate for
of the corresponding material) except for various cell types and it creates a visual
the Matrigel approach where cells are contrast on the disc for easier manipulation.
grown on the specimen carrier itself. The asymmetric “2” helps to identify the
side with the cells.
The alternatives I to IV described here are 
sorted by increasing freezing quality.
(I).
sandwich, alternative I. Left (bottom to
top): Flat type B carrier with layer of 1-
hexadecene face up, Sapphire disc with
cells facing upwards, type A carrier with
100 µm cavity dipped in 1-hexadecene
facing cells. Right, assembled specimen
sandwich.
This assembly is the easiest one with NOTE: Excess medium is blotted off
respect to handling because it consists of with filter paper in order to reduce the
only three pieces. aqueous layer to be frozen.
NOTE: The layer of 1-hexadecene on the
flat carrier ensures good thermal contact
between disc and carrier surface.


(II).
sandwich, alternative II. Left (bottom to
top): Flat type B carrier with layer of 1-
hexadecene, Sapphire disc with cells facing
up, spacer (one hole grid with appropriate
thickness), flat type B carrier optionally
with a film of 1-hexadecene facing cells.
Right, assembled specimen sandwich.
NOTE: 1-Hexadecene is optional

between cells and carrier cover for
alternatives II and III because the
volume is reduced to a few µm by using
the spacer.

(III).
sandwich, alternative III. Left (bottom to
top): Spacer ring 0.5 mm, Sapphire disc
with cells facing up, spacer (one hole grid
with appropriate thickness), flat type B
carrier optionally with a film of 1-
hexadecene facing cells. Right, assembled
specimen sandwich.
Liquid nitrogen hits the support, e.g.,

Sapphire disc directly, which improves the
cooling rate and freezing quality.

(IV).
sandwich, alternative IV.6 Left (bottom to
top): Spacer ring 0.5 mm, Sapphire disc
with cells facing up, spacer (one hole grid
with appropriate thickness), and Sapphire
disc with cells facing down, spacer ring 0.4
mm. Right, assembled specimen sandwich.
Liquid nitrogen hits the support, e.g.,

Sapphire discs directly from both sides,
remarkably improving the cooling rate and
freezing quality. Additionally, two discs are
frozen at the same time.

4.5. Loading, Freezing, Unloading and Storage of the Specimen

Sandwich
 Transfer of the sandwich to the NOTE: The holder tip has to be dried
specimen holder for high-pressure after each freezing cycle with a hair
freezing. dryer.
1. The specimen sandwich is transferred to

the specimen holder for high-pressure
freezing.
2. The specimen sandwich is inserted into

the cavity of the specimen holder and then
closed.
3. Tighten the specimen holder. The

specimen holder is held in the left hand or
with tweezers and the holder bolt is turned
clockwise.
4. The specimen sandwich must be firmly

attached in the holder. The carriers should
not move after tightening. This can be
checked with tweezers.
Figure 5.26 (1-4) Loading of the specimen

holder.

 Freezing of the specimen.

5. Specimen holder is inserted in the
specimen pressure chamber of the high-
pressure freezer.
6. Specimen holder is secured with the

locking bolt and the freezing process is
started by pushing the corresponding
button.
Figure 5.27
(5-6) Freezing of the specimen.
 Unloading of the specimen after
freezing.
7. Immediately after freezing, the specimen
holder is removed from the specimen
pressure chamber and transferred to the
designated slit in the unloading box filled
with liquid nitrogen. The tip containing the
specimen is unscrewed by turning the
holder counter clockwise.
NOTE: The holder can be removed as
soon as the locking bolt can be retracted.
Transfer of the specimen holder to liquid
nitrogen should be performed within five
seconds after the freezing shot.
8. The specimen holder is opened in liquid

nitrogen using tweezers and the specimen
sandwich is pushed out from behind.
9. Example of a frozen specimen: cellulose

capillary tubes embedded in 1-hexadecene.
10. The specimen is stored in liquid

nitrogen until further processing, e.g., an
aluminum box.
Figure 5.28 (7-10) Unloading of the

specimen.
IMPORTANT: Always precool tweezers or

other tools prior to handling frozen
specimens.

5.1. Advantages
 Very good preservation of ultra-  Preservation of high-pressure frozen
structure without the use of (cryo-immobilized) specimens is superior
to
intracellular antifreezing agents or any conventional, chemical fixation
other chemical treatment. techniques.
 Processing of cryo-immobilized
specimens is no more time consuming than
the conventional room temperature
processing.
 Quality of freezing depends on the size
of the specimen and its composition, e.g.,
content of natural antifreezing agents.

 All kinds of specimens can be frozen  Handling is easy.
using the same set of carriers for all  No special tools are necessary.
subsequent cryo-preparation methods.
 High-pressure freezing technique  Pressure in the pressure chamber of the

using the HPM 010 allows the use of HPM 010 is the same throughout the whole
very thin-walled specimen carriers. chamber. This makes the use of thin-walled
Any type of specimen sandwich can be carriers possible without collapsing the
used, e.g., assembling a sandwich with sandwich.
spacer rings and 10 µm thick stainless  Freezing quality is remarkably enhanced
steel discs.11 by using thin-walled carriers or discs.
 Large specimen size (2 mm diameter).  Compared to other high-pressure

freezing systems, the HPM 010 (system II,
see General Introduction) allows the largest
specimens to be frozen.
5.2. Disadvantages
 Maximum specimen size feasible for  Sizing of the specimen to the required
high-pressure freezing (2 mm in dimensions is a very important step and
diameter and 200 µm in thickness) is may often be challenging and should be
small compared to conventional performed as fast as possible.
chemical fixation techniques.
 Although freezing of thicker specimens
is possible, freezing without visible ice
crystal damage for most biological material
is restricted to about 200 µm.

Desired subsequent cryotechnique High-pressure freezing methods


 Freeze-substitution (FS)
 Freeze-fracturing (FF)
 Freeze-drying (FD)
 Cryo-sectioning (CS)
Figure 5.29 Specimen carrier assembly in

relation to further processing.

7. OBSERVED RESULTS
 Figure 5.30  All specimens were high-pressure frozen

in their native state (no chemical pre-

treatment).
 Figure 5.30.A  Figures A, C, D, E:

Thin section of mouse kidney biopsy.
 Freeze-substituted in acetone with 2%
OsO4
 Embedded in Epon/Araldite
 Thin sectioned
 Figure 5.30.B  Stained with uranyl acetate 2% and
Cryo-SEM micrograph of freeze- Reynolds lead citrate
fractured mouse kidney biopsy.
 All images:
 Electron Microscopy Center Zurich
(EMEZ), ETH Zurich, Switzerland.
 Figure 5.30.C
Thin section of human lung fibroblast,
grown as monolayer on Sapphire disc.
 Figure 5.30.D
Thin section of Scenedesmus vacuolatus
alga.
 Figure 5.30.E
Thin section of Caenorhabditis elegans
frozen in a cellulose capillary tube.
 Figure 5.30.F
Freeze-fracture replica of a liposome
suspension.


8. REFERENCES

2. Hohenberg, H., Tobler, M., and Müller, M. High-pressure freezing of tissue
obtained by fine-needle biopsy, J. Microsc., 183, 133, 1996.
3. Jiménez, N. et al. Aclar discs: A versatile substrate for routine high-pressure
freezing of mammalian cell monolayers, J. Microsc., 221, 216, 2006.
4. Kanno, H., Speedy, R.J., and Angell, C.A. Supercooling of water to -92 degrees C
under pressure, Science, 189, 880, 1975.
5. Michel, M., Hillmann, T., and Müller, M. Cryosectioning of plant material frozen
at high pressure, J. Microsc., 163, 3, 1991.
6. Monaghan, P. et al. High-pressure freezing in the study of animal pathogens, J.
Microsc., 212, 62, 2003.
7. Moor, H. and Riehle, U. Snap-freezing under high pressure: A new fixation
technique for freeze-etching, Proc. Fourth Europ. Reg. Conf. Elect. Microsc., 2, 33,
1968.
8. Morphew, M.K. and Mcintosh, J.R. The use of filter membranes for high-pressure
freezing of cell monolayers, J. Microsc., 212, 21, 2003.
10. Riehle, U. Über die Vitrifizierung verdünnter wässriger Lösungen. Federal Institute
of Technology, Zurich, Switzerland, 1968.
11. Sawaguchi, A., Ide, S., and Suganuma, T. Application of 10-microm thin stainless
foil to a new assembly of the specimen carrier in high-pressure freezing, J. Electron
Microsc. (Tokyo), 54, 143, 2005.
12. Sawaguchi, A. et al. Direct attachment of cell suspensions to high-pressure freezing
specimen planchettes, J. Microsc., 212, 13, 2003.
13. Schwarb, P. Morphologische Grundlagen zur Zell-Zell Interaktion bei adulten
Herzmuskelzellen in Kultur. Federal Institute of Technology, Zurich, Switzerland,
1990.
14. Shimoni, E. and Müller, M. On optimizing high-pressure freezing: From heat
transfer theory to a new microbiopsy device, J. Microsc., 192, 236, 1998.
15. Studer, D. et al. Vitrification of articular cartilage by high-pressure freezing, J.
Microsc., 179, 321, 1995.

High-Pressure Freezing LEICA EMPACT 131
CONTENTS
CONTENTS................................................................................................................... 131
GENERAL INTRODUCTION .................................................................................... 133
1. PRINCIPLES OF HIGH-PRESSURE FREEZING ....................................... 135
Simultaneous Pressurizing and Cooling of Biological Samples .......................... 135
2. SUMMARY OF THE DIFFERENT HPF TOOLS......................................... 137
3.1. Materials ................................................................................................... 138
3.2. Products .................................................................................................... 143
4. PROTOCOLS .................................................................................................... 143
4.1. The Tube Holder System .......................................................................... 143
4.2. The Flat Carrier System............................................................................ 144
4.3. The Biopsy Carrier ................................................................................... 146
4.4. The Ring Carrier....................................................................................... 147
4.5. The Membrane Carrier ............................................................................. 148
4.6. The Live Cell Carrier................................................................................ 148
5.1. Advantages ............................................................................................... 149
5.2. Disadvantages........................................................................................... 150
6. WHY AND WHEN TO USE HIGH-PRESSURE FREEZING ..................... 150
7. OBSERVED RESULTS .................................................................................... 152
8. REFERENCES .................................................................................................. 156


High-pressure freezing is a means to  Vitrification: Samples that are cryo-

produce vitreous or well-frozen bulk fixed without ice crystal formation during
biological samples.6 Successful high- cooling (usually very high cooling rates
pressure freezing leads to improved are necessary). Vitrification can only be
ultrastructural preservation. proven by diffraction of cryo-sections
(electron or x-ray; see Chapters 1, 11).
 Well-frozen: Samples show no

detectable ice crystal-induced segregations
after embedding in polymers.
 Bulk samples have to be small. The

typical size is a disc of 1 to 3 mm in
diameter with a thickness of 200 m.
Figure 6.1 The Leica EMPACT high-

pressure freezer.
Bar = 10 cm
High pressure (2000 bars = 200 MPa) is a  High pressure lowers the freezing point
physical cryoprotectant that suppresses ice of water. At 204.8 MPa, the freezing point
nucleation and ice crystal growth to a is at a minimum of 251 K (22°C).
certain degree during cooling. The Applying more pressure increases the
cryoprotective influence can be quantified.8 freezing point again.
In practice, biological sample discs 200 m
thick (or less) and a few millimeters in  Sartori et al.8 showed that 10 times
diameter are well-frozen or vitrified.12 thicker samples can be vitrified when high-
pressure freezing is applied compared to
ambient pressure freezing, e.g., slam-
freezing and plunge-freezing.
The thinner the sample, the higher the  A 200 m-thick aqueous sample has in
cooling rates within the sample. its center a cooling rate of about
Slamfreezing (see Chapter 2) on a copper 5000 K/sec. This cooling rate is dictated
block may supply an almost infinitely high by sample size. To achieve optimal
superficial cooling rate;3 however, within cooling rates in the center of a 200 m-
the sample (depending on its thickness) this thick sample, the cooling rate applied at
cooling rate decreases rapidly. the surface has to be 10,000 K/sec or more.

High-pressure freezing machines can NOTE: Cooling rates higher than

generate superficial cooling rates in the 10,000 K/sec do not increase the cooling
range of 10,000 to 30,000 K/sec. Pressure rate in the center.10,12
and cold are simultaneously applied to the
sample.  The measured cooling rate (10,000 to
30,000 K/sec) depends more upon the
thermocouple than on the machine’s
performance. Cooling rate measurements
depend on each thermocouple. To have a
series of identical sensors is almost
impossible.
Figure 6.2 Calculations-based on the

properties of water show that in 200 m-
thick samples (shown only from surface to
centre, 100 m, the other half is exactly
mirrored) the highest cooling rates are
generated when the sample has reached a
temperature, of 253 K (20°C). At this
temperature water in the sample is still
liquid at 200 MPa.
A high-pressure frozen sample shows  Gaseous compartments are compressed

improved structural preservation when the to almost zero volume. The collapse of
sample does not contain gaseous gas-filled compartments can induce heavy
compartments and is well prepared prior to distortions in the sample. The aqueous
freezing. phase, however, is almost incompressible.

1. PRINCIPLES OF HIGH-PRESSURE FREEZING
Simultaneous Pressurizing and Cooling of Biological Samples

We consider all samples that are high-  Samples frozen under ambient
pressure frozen as aqueous samples conditions are in most cases heavily
because water is the main constituent segregated by ice crystal formation.
(around 80%) of living biological Cryoprotectants have to be added.
structures.
Pressure is an efficient “physical”  All statements made are based on

cryoprotectant. However, it can be harmful measured and calculated physical facts.12
when applied for too long a period of time Therefore it is always important to
(>100 ms7,9). Therefore, one has simul- correctly declare what one is doing. To
taneously to pressurize and cool the sample achieve maximum cooling rates in the
to vitrify or freeze it well. Best results are centre of a 200 µm-thick sample, a surface
achieved at a pressure of 2048 bars and cooling rate of 10,000 K/sec is sufficient;
cooling rates higher than 10,000 K/sec however, for optimal cooling rates in a
applied at the surface of a 200 µm-thick 100 µm-thick sample, about 50,000 K/sec
sample. are needed at the surface.
A correct high-pressure freezing cycle is  The total force generated by pressure

provided by the EMPACT. The sample is onto an object depends on the surface area
surrounded by a metal carrier (usually gold- of the object in contact with the pressure
plated copper able to withstand 2048 bars). source. The EMPACT is, on one hand,
This carrier is connected via a small tube to applying about 2000 bars onto a sample
a pressure system. Once the pressure is and, on the other hand, it cools the sample
applied, a double jet of liquid nitrogen from outside. For efficient cooling of the
cools the sample. The high-pressure and sample, the enveloping container should
cooling systems are separate from one have minimal mass. However, the container
another. has to withstand pressure and, therefore, the
mass cannot be reduced ad infinitum.
Figure 6.3 An optimal high-pressure

cooling cycle: The pressure slope reaches
2045 bars in 20 msec. The pressure
increase is nicely synchronized with the
temperature decrease. The cooling rate is
indicated as dT/dt: 23,650 K/sec.

F
S
Figure 6.4 The diagram illustrates how the
EMPACT is working. The inset shows a
cross section through a copper tube sealed
on both sides by cones.
 PF = Pressure fluid
 LN2 = Liquid nitrogen
The pressure is transmitted by a liquid,  The sample container is represented by

methyl cyclohexane. This solvent is liquid the copper tube system. The copper tube
in the temperature range of 150 to 370 K surrounded by the envelope and filled with
(123°C to +97°C), thus transmitting a cell suspension is introduced into the
pressure correctly over a wide range of EMPACT with the help of a manipulator.
temperatures. The high boiling point helps The tube is connected to the pressure
in that methyl cyclohexane does not system by pushing the copper tube end onto
evaporate quickly at room temperature. the cone-shaped end of a pressure tube
Furthermore, due to its hydrophobic (with pneumatic piston C). The opposite
properties, methyl cyclohexane does not end of the copper tube is sealed with a
mix with water. Because the interface stainless steel cone integrated into the
between sample and hydraulic fluid is small manipulator. When the sample is located in
(spatial as well as temporal), the impact of position and consequently sealed, piston D,
methyl cyclohexane is considered to have generating high pressure, is loaded with
no detrimental effects on the sample. In the compressed air. However, bar F stops
case of membrane carriers, no contact is piston D so that no pressure is applied to
made between sample and methyl the sample as of yet.
cyclohexane.
Technically, one can exchange methyl  Piston B jets the cryogen toward the
cyclohexane with any other solvent. sample; however, in the very first moment,
However, there is no experience available. the jet is deviated by shutter S, not hitting
the sample and, consequently, the sample is
not cooled at this step.
After a cycle time of 600 msec, the sample  Piston A coordinates the high-pressure
is released automatically into a reservoir freezing process. The connecting bar (F)
filled with liquid nitrogen. releases the pressure piston (D) applying
2048 bars pressure to the sample via the
pressure fluid (PF). The jets are
synchronized with the same rod to redirect
the liquid nitrogen onto the surface of the
sample carrier immediately after pres-
surization.

2. SUMMARY OF THE DIFFERENT HPF TOOLS
The handling of the EMPACT and When nothing else is indicated, the tools
EMPACT2/RTS is well described in the and small parts are supplied by Leica
manuals supplied by Leica. Their operation Microsystems (Vienna, Austria)
is supported step-by-step on the integrated
touch screen.Because a wide variety of
sample types exists and different follow-up
procedures — freeze-substitution (see
Chapter 13), freeze-fracturing (see  One system is based on hollow
cylindrical holders (the tube holder
Chapter 22), freeze-drying (see Chapter 15) system), the other six on disk-like carriers
and cryo-sectioning (see Chapter 11) — (the flat specimen system).
can be chosen, a variety of sample carriers
is available.

 The tube holder system.  The tube holder system allows very
efficient cooling of the sample, and is the
best suited carrier for cryo-sectioning,
CEMOVIS (Cryo-electron microscopy of
vitreous sections), if the sample allows it.
 The flat carrier.  The carrier is used to high-pressure
freeze samples that can easily be trimmed
to a disc of 200 m in thickness and a
diameter of 1.2 to 1.5 mm (depending on
the carrier) or for suspensions. It is
dedicated to freeze-substitution.
 The biopsy carrier.  Part of the microbiopsy system and
designed for biopsies of tissues. Dedicated
to freeze-substitution.
 The ring carrier.  It is originally a consumable of the
freeze-fracture system. It is suitable for
dense suspensions. Dedicated to freeze-
substitution.
 The membrane carrier.  Has the same function as the flat carrier;
however, it rules out that the biological
sample comes into contact with the
hydraulic fluid (methyl cyclohexane).
Dedicated to freeze-substitution.
 The live cell carrier.  Especially suitable for cell cultures
grown on sapphire discs. They can be
investigated in the light microscope and
then high-pressure frozen within five
seconds with the help of the rapid transfer
system (RTS) mounted on the EMPACT2.
Dedicated to freeze-substitution.
 The freeze-fracture carrier.  A platelet system suitable to perform
cryo-fixation prior to freeze-fracturing in
the machines available on the market.

3.1. Materials
 The high-pressure freezing system

Figure 6.5 Illustration of the Leica
EMPACT (left) and the Leica EMPACT2
with the rapid transfer system (right).
Apart from several small updates (Dewar
with drain, new software with USB
connection, bath led illumination), the most
prominent difference between the two
machines is the rapid transfer system
(RTS). The RTS allows specimen carrier
loading in less than five seconds, a pre-
requisite for correlative light and electron
microscopy (CLEM). The EMPACT is
described in Studer et al.11 and functioning
of EMPACT2 and RTS in Manninen et al.5
Bar = 20 cm
 The sample loading system  Dedicated to CEMOVIS samples and

 Envelop of tube holder system small samples in suspension (single cells,
invertebrates, embryos, etc.).
Figure 6.6 A copper tube carrier is inserted

into the tube holder (top), which in turn is
connected to the loading device (bottom).
Bar = 1 cm
 The flat specimen system  Dedicated to all other types of samples.

Five kinds of carriers exist, depending on
the type of sample. All five types fit in
the carrier holder (pod).
Figure 6.7 The carriers are locked into the

pod by means of a torque wrench.
Bar = 1 cm

 The carriers
 The tube holder system  Copper tube
 Length = 15.9 mm
 Outer diameter = 0.650 mm
 Inner diameter = 0.350 mm
Figure 6.8
The copper tubes are fixed in the tube
holder with the help of a recycling kit. The
tube is introduced into the holder and cones
are pushed into both ends to form funnels.
Bar = 2 mm
 The system is dedicated to vitrifying

suspensions for cryo-sectioning
(CEMOVIS, see Chapter 11).
 The carriers of the flat specimen  All carriers have a gold-coated copper
system body.
 Flat carrier
 Inner diameter = 1.25 to 1.5 mm
 Central opening diameter = 0.2 mm
 Inner wall height = 0.2 mm
 Carrier thickness = 0.5 mm
 For suspensions and easy to prepare
tissue samples (e.g., cartilage, leaves).
 Biopsy carrier
 Cavity length = 1.2 mm
 Cavity width = 0.3 mm
 For high-pressure freezing of biopsies
(tissue).
 Ring carrier
 Used as carriers above; not so easy to
prepare sample, but much easier to
handle sample in follow-up pro-
cedures.
Figure 6.9 Available carriers as mentioned

above.
Bar = 1 mm

 Membrane carriers
 Wall height = 0.2 mm or 0.1mm
 Carrier thickness = 0.26 mm or
0.16 mm
 Same use as flat carrier. However, the
pressure solvent is not in contact with
the sample.
 Live cell carriers, sapphire discs and

finder grid
Live cell Sapphire Finder

carriers disks grid
Outer 2.8 mm 1.4 mm 1.5 mm
diameter
Large inner 1.5 mm – –
diameter
Small inner 1.0 mm – –
diameter
Wall height 0.2 mm – –
Thickness 0.4 mm 0.05 mm 0.01 mm
Figure 6.10 Membrane carrier, live cell

carrier and sapphire discs (from top to
bottom).
Bar = 1 mm
 Dedicated for correlative light and

electron microscopy.
 Freeze-fracture carriers
Carriers Ring
Large outer diameter 2.8 mm 2.8 mm
Small outer diameter 2.0 mm –
Inner diameter 1.15 mm 1.2 mm
Wall height 0.2 mm –
Thickness 1.0 mm 0.4 mm
Figure 6.11 Carrier and ring are mounted

on top of one another and loaded into a pod.
These carriers provide a link between high-
pressure freezing with the EMPACT and
the freeze-fracture technique (BAL-TEC
freeze-fracture machine).
Bar = 1 mm

 The sampling systems

Sampling systems are tools that are
necessary to excise the biological
material from its environment (in vivo or
in vitro) and transfer it to an EMPACT-
compatible carrier.
 Piston (tube holder system). A piston  Stainless steel wires of the correct
is a thin wire (diameter 0.2 to 0.3 mm, diameter (0.25 mm) are available. The
less than the inner diameter of a likelihood of vitrifying the medium
copper tube), long enough to handle surrounding single cells can be improved
easily (4 to 5 cm). by adding 20% dextran.
 Cellulose tubes (tube holder system).  The cellulose tubes can be filled with
The cellulose tubes were introduced by samples (e.g., single cells, small
Hohenberg et al.4 The tubes fit in the invertebrates, embryos, etc.) using a suction
copper tubes and are porous, ensuring pressure pipette. Sealing a tube is easily
that the substitution medium reaches done by pressing with the back (i.e., the
opposite side of the blade) of a surgical
the sample inside.
blade.2
Figure 6.12 A sealed cellulose tube

containing nematodes is sticking out of a
copper tube.
Bar = 500 µm
 Commercial pipette system (ring  A volume of 2 to 5 L per carrier is
carrier, flat carrier, membrane carrier). sufficient. Use micropipettes (maximum
For the transfer of solutions, fluids and volume) 20 L with extra thin (white) tips.
suspensions.
 Punch (flat carrier, membrane carrier).

The punches are available in two sizes:
1.2 mm or 1.5 mm diameter.
 Sapphire discs (live cell carriers).

0.05 mm-thick sapphire discs
(diameter 1.4 mm) allow monolayer
culture. These discs withstand the
physical conditions of high-pressure
freezing and are good heat transducers.
 Microbiopsy system (biopsy carrier). Figure 6.13 The microbiopsy needle has a
notch in its central rod. By inserting the
needle into the tissue, the notch is filled
with tissue and excised upon releasing the
sharp hollow tube.
Bar = 1 mm

 The transfer station
Figure 6.14 The transfer station is the tool

for transferring the biological sample into
the biopsy platelet.
 Top: Transfer station with RTS
adapter installed for EMPACT2.
 Middle: The transfer station with
biopsy needle for EMPACT.
Bar = 2 cm
 Inset below: a higher magnification of

the transfer area. The needle is made
transparent for a better view.
Bar = 5 mm
The transfer station provides a situation  When an EMPACT2 (with RTS) is used,
that allows a swift transfer of the sample the loading of the platelet in the pod is done
into the biopsy carrier, the carrier into the automatically.
pod and the pod into the EMPACT.
 The rapid transfer system (RTS)
Figure 6.15 The rapid transfer system
(RTS) is an addition to the EMPACT2. It
speeds up the loading of the flat carrier
system by automation of a number of steps.
It was developed specifically for correlative
light and electron microscopy studies and
rapid biopsy work. The RTS system uses a
new type of pod, but it does not differ in
functionality from the pod described herein.
The adaptor provides the link between the
transfer station (see Figure 6.14) and the
high-pressure freezer.
Bars = 10 cm and 1 cm
 The adapter has a clamp diameter of 2.4
mm, stretchable to 3.0 mm. Hence, the
adapter can hold all types of flat carriers
(diameter 2.8 mm).

3.2. Products
 EMPACT, EMPACT2, EM RTS and  Leica Microsystems, Vienna, Austria
tools.
 Liquid Nitrogen (LN2). 15 L are
required to fill the EMPACT Dewar.
 Methyl cyclohexane as hydraulic fluid.  Merck (VWR International), Germany
Ord. no.: 806147
 1-Hexadecene.  Merck (VWR International), Germany
Ord. no.: 822064
 High molecular glycans, such as  Fluka, Switzerland, Ord. no.: 31390
dextran (4075 kDa).
 Biological samples depending on
investigators interest.
4. PROTOCOLS
4.1. The Tube Holder System

The tube holder system allows high-  The tube holder system is the most
pressure freezing of small organisms (cell elegant way to prepare samples for frozen
suspensions, Caenorhabditis elegans, etc.) hydrated sectioning and subsequent
as well as fluids (blood, milk). Addition of CEMOVIS1 see Chapter 11).
20% dextran (40 to 75 kDa) enables 
vitrification of the medium around the cells.
1. A copper tube is loaded into the tube  A copper tube is inserted into a holder
holder. and fixed into place with the loading jig
included with the EMPACT tool set. The
outer tips of the copper tube are funneled
by tightening the screw.
2. Prepare your sample. Place a 20 L  The hydrophobic Parafilm permits the

drop on a Parafilm sheet. formation of stable drops.
3. A thin wire used as a piston is inserted  Make new drops for every new run as
in the copper tube and directed through the small drops are very sensitive to drying
the tube into the suspension. The out. To ensure vitrification around the cells,
bottom end of the copper tube has to use 20% dextran (40 to 75 kDa).
be inserted into the suspension drop.
 Because the copper tube has a funnel-
4. The piston is pulled out at a constant shaped end, it is not difficult to insert the
speed of about 1 cm/sec. wire in the tube.
5. The tube is turned 180° (upside down),

i.e., the bottom end becomes the top
and the top becomes bottom. Repeat
Step 3 to 4.

Figure 6.16 (Left) Diagram of the

aspiration of a fluid into the copper tube
(encased by a tube holder) using a wire
(piston). (Right) Diagram of the tube holder
attached to the loading device prior to high-
pressure freezing.
Bars = 1 cm
6. The copper tube is now filled with  To prevent the loss of biological samples
suspension and can be loaded into the (e.g., due to their fluidity) during freeze-
loading device and high-pressure substitution, they can be enclosed in
frozen. cellulose tubes. The cellulose tubes were
7. After freezing, the tube is removed by introduced by Hohenberg et al.4 They are
cutting out the central 5 mm. This is thin walled, water permeable and have a
done with the tube cutter included in diameter less than the inner diameter of the
the EMPACT tools. Be aware that copper tubes. For additional sampling
only the sample in the central 5 mm of information on the cellulose tubes, consult
the tube is vitreous and will be further Claeys et al.2
processed.
4.2. The Flat Carrier System

1. A punch produces discs of a leaf or of  The flat carrier is intended as an all-
tissue sliced to 200 m thick sections. purpose carrier. Its use is explained using
leaves as an example.

Figure 6.17 Drawing of the excision of leaf

discs with a punch.
Bar (left) = 1 cm
Bar (right) = 1 mm
2. Gaseous reservoirs in the sample (e.g.,  1-Hexadecene was introduced by Studer

leaves or lung) must be replaced by an et al.13 in 1989.
inert, noncompressible fluid. 1-Hexa-
decene is a suitable and often used
inert fluid for replacing gaseous
volumes.4

Figure 6.18 Air evacuation

 The leaf samples are collected in a
5 mL syringe, which is half filled with
1-hexadecene.
 Air is removed from the syringe.
 The syringe is airtight closed at the
needle exit (Parafilm and finger) and
the piston is pulled. Gas bubbles will
appear around the biological sample.
Continue until no more gas bubbles
appear.
Bar = 2 cm
3. Prepare the transfer stage.
Figure 6.19 Install the pod (P) onto the

transfer stage (tool supplied with the
EMPACT). Insert a flat carrier into the fork
of the handle (H), making sure the cavity
points upward.
Bar = 2 cm
4. Introduce sample in carrier.

Figure 6.20 Diagram of the leaf disc in a
carrier. The leaf discs can now be
introduced into the flat carrier (one leaf disc
per carrier). A drop of 1-hexadecene can be
added to ensure all gas is removed from the
carrier cavity.
Bar = 1 mm
5. Once the sample is introduced into the  Move the handler to the end stop.
carrier, push the fork into the pod and  Close the pod with the torque wrench. A
lock it. Attach the pod to the loading torque of 30 N/m is applied.
device and freeze.  With the RTS, Step 5 is automatic.
Figure 6.21 Drawing of the pod attached to

the loading device, prior to high-pressure
freezing.
Bar = 1 cm

4.3. The Biopsy Carrier

1. Load the biopsy carrier into the  Only take biopsies of exposed organs.
loading fork and push the fork through
Do not stick the needle through the skin to
the pod to the end of the sampling take a biopsy of an organ.
bench. Now the biopsy needle notch  If notch and carrier cavity are not
and the carrier cavity should be aligned, adjust the needle position with the
aligned on top of each other. screw.
 The cavity in the biopsy carrier has the
2. Retract the outer tube of the biopsy same size as the biopsies obtained with the
gun. The gun is now “loaded.” Leica microbiopsy system. This allows fast
tissue sampling and little gaseous volume to
fill prior to high-pressure freezing (HPF).

Figure 6.22 The orientation of the cavity of
the biopsy carrier must be in line with the
notch of the biopsy needle.
Bar = 1 mm
3. Introduce the tip of needle in the organ

and take the biopsy.
Figure 6.23 The microbiopsy needle

 In essence, a solid rod surrounded by a
hollow tube. Both are sharp.
 The rod has a notch at its tip (0.3 mm
high × 0.6 mm wide × 1.2 mm long),
which will contain the biopsy.
 Upon inserting the rod in the organ of
interest, the notch is filled with tissue
due to its natural movement.
 Upon taking a biopsy, the outer tube
moves along the rod and cuts away the
content of the notch from the rest of
the organ.
Bar = 1 mm
4. Quickly dip the tip of the biopsy gun  Dipping in 1-hexadecene will prevent
(with the tissue) in 1-hexadecene. the sample from drying out.

5. Install the gun on the microbiopsy

transfer station and transfer the tissue
from the needle into the biopsy carrier
with the transfer tool.
Figure 6.24 Close-up (left) of the transfer

station with the pod, the biopsy carrier and
the transfer tool in place and ready to
receive a biopsy. The needle containing the
biopsy is not shown. (right) Once the
biopsy gun is installed on the transfer
station, the notch in the needle will be
positioned exactly above the cavity of the
biopsy carrier.
Bar = 1 mm
6. Proceed as explained in Step 4 to 5 of  The biopsy station ensures the alignment

the flat carrier. of the tools to one another (needle notch,
biopsy carrier, pod, transfer tool). (See
Vanhecke et al. 14 )
 This procedure takes less than
30 seconds from obtaining the biopsy until
it is frozen. With the EM RTS on the
EMPACT2, the procedure takes less time.
4.4. The Ring Carrier

The surface tension of water allows fluids  The main advantage over the copper
to be introduced into a bottomless carrier. tube is the possibility to observe the sample
The ring carrier can be used for cell under the microscope or stereoscope during
suspensions or fluids. The ring carriers are every step of the follow-up procedure. The
the same rings supplied for freeze- mammalian (rat) cell line Walker carcino
fracturing. sarcoma was grown in suspension and is
used as an example.
1. Install a ring carrier in the tip of the  Sample loading is still easy, although a
fork on the transfer stage. little more difficult than for flat carriers.
2. Prepare tools as in Step 3 in the flat  The advantage is that handling after
carrier system. freeze-substitution is easier.
3. Pipette a small volume of the  The theoretical cavity volume is

concentrated cell suspension or fluid in 0.226 L, but overfill (2 to 5 L) is needed
the central cavity of the carrier. to overcome the surface tension of water.
The excess liquid will be pushed into the
pressure tube of the pod during locking.

Figure 6.25 The filling of a ring carrier.
Bar = 1 mm
4. Proceed as explained in Step 4 of the

flat carrier system.
4.5. The Membrane Carrier

The membrane carrier is used in the same  The base of the carrier is thin and
way as explained for the flat carrier. flexible and, therefore, allows transmission
of the pressure.
The compression factor of water is:
5 × 10-6 cm2/N
At 2000 bars, this equals a volume change
of 0.01 (or 1%), which in 3-D is very small.
Assume a cube with an edge of 1000 m;
this edge is reduced to 996.6 m at
2000 bars.

4.6. The Live Cell Carrier

1. Cultivate cells on sapphire discs.  Sapphire discs are, after carbon coating,
well suited for cultivation of cells.
 Furthermore, sapphire discs are
transparent and very good thermal
2. Load a live cell carrier. conductors and perfect for high-pressure
freezing.
3. Introduce (in medium) a sapphire disc  Cultivated Madin-Darby canine kidney
with cells into the live cell carrier. (MDCK) cells are used as an example of
the live cell carrier functionality.
4. Clamp a finder grid.
5. Observe the cells in the light

microscope (e.g., confocal scanning
laser microscopy).
6. Proceed by using the RTS. All

necessary manipulations are performed
automatically.

Figure 6.26 Drawing of the loading of a

sapphire disc in a live cell carrier, allowing
observation in a light microscope.
Bar = 1 mm
The follow-up procedures are not  Freeze-substitution, freeze-fracturing

discussed in this chapter. and cryo-sectioning.

5.1. Advantages
 High-pressure freezing best preserves  Chemical fixation is comparatively slow
bulk biological samples. and during cross-linking destroys gradients
essential for life processes.
 Handling is easy and very safe.  All steps necessary to perform high-
pressure freezing are listed step-by-step on
the control panel of the EMPACT machine.
 Specimen preparation is easy with  Biopsies obtained with the microbiopsy
dedicated tools. system are frozen within 30 sec. Cell
cultures grown on sapphire discs and
sampled with the RTS are frozen within 5
sec.
 For all follow-up methods, suitable  All follow-up procedures are possible
sample carriers and holders are with the EMPACT frozen samples (cryo-
available. substitution, freeze-fracture, cryo-SEM).
The tube system is the most convenient for
cryo-sectioning.
 The EMPACT is small and easy to  In the case of a centrally organized
transport on a cart. electron microscopy unit, it is often easier
to fix the samples where they originated. It
takes little effort to move the EMPACT
around a campus.
 The EMPACT consumes only a small  15 liters of liquid nitrogen are sufficient
amount of liquid nitrogen. to run the EMPACT for at least 20 samples.
The remaining nitrogen can be used for, for
example, freeze-substitution.
 Data management  Information on temperature and pressure
changes is saved digitally on the storage
device of the machine. The curves can be
retrieved via a USB key and printed or
added to a protocol log.

5.2. Disadvantages
 All high-pressure freezing machines  This statement is valid for most
are unable to vitrify or adequately biological samples. The limit of 200 m
freeze native biological samples thickness is given by physics and not by
thicker than 200 m. any technical insufficiency.10,12 However,
eye lenses, for example, are enriched with
proteoglycan and this natural cryoprotectant
allows vitrification of these samples thicker
than 200 m.

 The success of the approach has to be  Water content and its state (bound vs.
experienced with every new sample. free water) is the most limiting factor for
Also this statement holds for all high- successful high-pressure freezing. There are
pressure freezing machines. “easy,” “difficult” and “impossible”
samples. What the outcome will be is
nearly impossible to predict. However,
most well prepared biological samples with
a thickness of 200 m or less are, in
general, segregation free (well-frozen).


6. WHY AND WHEN TO USE HIGH-PRESSURE FREEZING
The dream of all microscopists is to achieve  Optimal cryo-fixation is vitrification (see

optimal ultrastructural preservation of Chapter 1). A vitreous sample is, in theory,
biological samples in electron microscopy. preserved at the atomic level. A prerequisite
Up to now there is no better method is that the sample itself is not changed
available than cryo-fixation or cryo- during specimen preparation. This means
immobilization. Bulk samples in their no chemical prefixation, no addition of
native state have to be high-pressure frozen cryoprotectants. Specimen preparation prior
to obtain satisfactory (segregation free) to freezing must be gentle and as fast as
cryo-fixation. Therefore, high-pressure possible.
freezing is a prerequisite in achieving this.
Chemical fixation has to be avoided

whenever possible. The size of the sample
is a major limiting factor of high-pressure
freezing. With few exceptions, there is no
physical fixation available to date for
samples thicker than 200 m.

7. OBSERVED RESULTS
 Figure on the Chapter’s title page  Connective tissue of the endomysium of

a rat soleus muscle is shown.
 Figure 6.27 Brain tissue

 E = endothelium
Brain tissue obtained from an  M = mitochondrion
anaesthetized rat (Rattus norvegicus)  PoN = postsynaptic neuron
using the microbiopsy system (including  PrN = presynaptic neuron
the biopsy carrier). The tissue was freeze-  RBC = red blood cell
substituted and embedded in Epon.  S = synapse
(Samples prepared and images acquired
by Werner Graber, Institute of Anatomy, Bar = 1 µm
Bern.)
 Figure 6.28 Ivy leaf

 CP = chloroplast
1.2 mm-diameter ivy leaf discs were  DL = double lipid layer
excised from fresh ivy leaves (Hedera  M = mitochondrion
sp.) using a punch, evacuated in 1-hexa  G = granum
decene prior to cryo-fixation in a flat  N = nucleus
carrier. The tissue was freeze-substituted  NE = nuclear envelope
and embedded in Epon. (Samples  NP = nuclear pore
prepared and images acquired by Werner  PNS = perinuclear space
Graber, Institute of Anatomy, Bern.)  S = stroma
Bar = 250 nm
Bar (inset) = 100 nm


 Figure 6.29 Yeast  B = bubbling induced by the

electron beam
Cryo-section of a vitreous yeast sample.
The yeast cells (Saccharomyces  M = mitochondrion
cerevisiae) were vitrified using the
 RER = rough endoplasmatic
copper tube system.
reticulum
Bar = 1 m (left)
Bar = 500 nm (right)
Arrow = cutting direction
 Figure 6.30
Walker carcinosarcoma cells.  R = ribosome
The cells grew in vitro in suspension and  RER = rough endoplasmatic
were cryofixed in a ring carrier, followed reticulum
by freeze-substitution and Epon
 V = vesicle
embedding.
Bar = 1 m


8. REFERENCES
1. Al-Amoudi, A. et al. Cryo-electron microscopy of vitreous sections, Embo J., 23,

3583, 2004.
2. Claeys, M. et al. High-pressure freezing and freeze substitution of gravid
Caenorhabditis elegans (Nematoda: Rhabditida) for transmission electron
microscopy, Nematology, 6, 319, 2004.
3. Escaig, J. New instruments which facilitate rapid freezing at 83K and 6K,
J. Microsc., 126, 221, 1982.
5. Manninen, A. et al. Caveolin-1 is not essential for biosynthetic apical membrane
transport, Mol. Cell Biol., 25, 10087, 2005.
6. Moor, H. Theory and practice of high-pressure freezing, in Cryotechniques in
Biological Electron Microscopy, Steinbrecht, R.A. and Zierold, K., eds., Springer,
Berlin, Germany, 1987, 175.
7. Moor, H. and Hoechli, M. The influence of high-pressure freezing on living cells,
in Electron Microscopy: Proceedings of the 7th International Congress on Electron
Microscopy, Favard, P., ed, Grenoble, France, 1970, 445.
8. Sartori, N., Richter, K., and Dubochet, J. Vitrification depth can be increased
more than 10-fold by high-pressure freezing, J. Microsc., 172, 55, 1993.
9. Sato, N. et al. Ultrastructural effects of pressure stress to Saccharomyces cerevisiae
cells revealed by immunoelectron microscopy using frozen thin sectioning, in High-
Pressure Bioscience and Biotechnology, Hayashi, R. and Balny, C., eds., Elsevier
B.V., Amsterdam, The Netherlands, 1996, 109.
10. Shimoni, E. and Müller, M. On optimizing high-pressure freezing: from heat
transfer theory to a new microbiopsy device, J. Microsc., 192, 236, 1998.
11. Studer, D. et al. A new approach for cryo-fixation by high-pressure freezing,
J. Microsc., 203, 285, 2001.
12. Studer, D. et al. Vitrification of articular cartilage by high-pressure freezing,
J. Microsc., 179, 321, 1995.
13. Studer D., Michel, M., and Müller, M. High pressure freezing comes of age.
Scanning Microsc. Suppl. 3, 253, 1989.
14. Vanhecke, D. et al. A rapid microbiopsy system to improve the preservation of
biological samples prior to high-pressure freezing, J. Microsc., 212, 3, 2003.

Part II
Cryo-Electron Microscopy

Frozen Hydrated Macromolecules for Structural Analysis 161
CONTENTS

1. PRINCIPLES OF THE METHOD .................................................................. 164
1.1. Vitrification .............................................................................................. 164
1.2. Thin, Aqueous Water Layers.................................................................... 164
2.1. Standard Method ...................................................................................... 166
2.2. Reduce Surface Effects............................................................................. 168
3. MATERIALS/PRODUCTS/SOLUTIONS ...................................................... 169
3.1. Materials ................................................................................................... 169
3.2. Products .................................................................................................... 172
3.3. Solutions ................................................................................................... 172
4. PROTOCOLS .................................................................................................... 173
4.1. Preparation of Holey Carbon Films8 ......................................................... 173
4.2. Vitrification of a Thin Layer..................................................................... 174
4.3. Preparation of a Holey Carbon Film Covered with a Thin, Continuous
Carbon Foil ............................................................................................... 176
4.4. Adsorption of a Specimen on a Carbon Film............................................ 176
4.5. Adsorption of Poorly Concentrated Protein Samples9 .............................. 177
4.6. Chemical Cross-Linking........................................................................... 177
4.7. Preparation of a Sandwich9 ....................................................................... 178
4.8. Incubation with a Lipid Monolayer1,12,15,16 ............................................... 178
4.9. Specimen Transfer.................................................................................... 179
4.10. Low-Dose Imaging ................................................................................... 180
4.11. Image Analysis ......................................................................................... 181
4.12. Troubleshooting........................................................................................ 182
5.1. Standard Method ...................................................................................... 183
5.1.1. Advantages.................................................................................. 183
5.1.2. Disadvantages ............................................................................. 183
5.2. Adsorption on Carbon .............................................................................. 183
5.2.1. Advantages.................................................................................. 183
5.2.2. Disadvantages ............................................................................. 184
5.3. Glutaraldehyde Cross-Linking.................................................................. 184

5.3.1. Advantages ................................................................................. 184

5.3.2. Disadvantages............................................................................. 184
5.4. Interaction with Lipid Layers ................................................................... 184
5.4.1. Advantages ................................................................................. 184
5.4.2. Disadvantages............................................................................. 184
6.1. Stable Specimen Available at High Protein Concentration...................... 185
6.2. Stable Specimen Available at a Concentration Lower Than 0.3 mg/mL . 185
6.2.1. Adsorption on a carbon film (see Section 4.4) ........................... 185
6.2.2. Interaction with a lipid layer (see Section 4.8) ........................... 185
6.3. Unstable Specimen................................................................................... 185
6.3.1. Chemical stabilisation ................................................................ 186
6.3.2. Adsorption of the specimen ........................................................ 186
6.3.3. Use of lipid layers....................................................................... 186
8. REFERENCES .................................................................................................. 189


 Imaging of biological macromolecules

by electron microscopy has developed  Structure-function analysis.
into a major method for structural
biology during the last decade. This
approach aims at understanding the
mode of action of molecular
assemblies by correlating their
structure with their function. A
prerequisite for such structure-function  Keep the specimen hydrated in the
studies is to observe the molecules in a electron microscope.
state as close as possible to their  Keep the specimen at a very low
physiological state. temperature.
 A major obstacle in reaching this goal

is keeping the specimen hydrated in an  Needs a cold stage.
electron microscope that operates
under vacuum. Methods were  Needs an anticontaminator.
developed in the 1980s to preserve the
hydration of the specimen by vitrifying
a macromolecular suspension through
rapid freezing.10,13 The specimen is
observed by its own contrast in the
absence of stain so that the collected  A variety of methods adapted to fragile
structural information is comparable to specimen.
other structural methods, such as x-ray
crystallography. The direct observation
of a thin layer of suspension avoids
specimen adsorption onto a carbon
support and, thus, allows the analysis
of its dynamic behaviour.
 It was observed, however, that the

structure of some molecules can be
affected during the formation of the
thin layer of suspension, notably
through surface effects.5,6 This chapter
will describe the standard method used
by the authors to prepare thin films of
macromolecules and a special
emphasis will be given to variants of
the method that may help to reduce
surface effects.

1. PRINCIPLES OF THE METHOD
1.1. Vitrification
 Vitrification is the process by which a  Vitrification: A cryogenic immobi-
special form of solid water — vitreous lisation of the specimen in its hydrated
ice — is formed by rapid freezing, thus state.
preventing ice crystal formation.5 This
form of water is required for optimal
preservation and clear visibility of the
specimen. The high cooling rate of
about 10,000°C/sec is obtained by
plunging the specimen into ethane  Ideally, the molecules are randomly
slush (liquid ethane at the oriented and thus are observed through all
solidification temperature). directions.
Figure 7.1 Schematic representation of

randomly oriented particles within the
frozen hydrated film.
 The thickness of the water suspension

that can be cooled into a vitreous state
depends on the freezing velocity,11 on
the exact composition of the aqueous
solution and on the ambient pressure.
At atmospheric pressure and by using
the standard freezing devices a
thickness of about 200 nm of water
can be converted into a vitreous state.
This value is larger than the thickness
generally acceptable for high-
resolution imaging of particles that
should not exceed 100 nm.
1.2. Thin, Aqueous Water Layers

 During specimen preparation for cryo-
EM, the suspension has to be reduced
to a less than 100 nm thick layer,
which is generally performed by
blotting off the excess liquid with filter
paper.
 During the short interval between the

removal of the filter paper and the
actual vitrification, the suspension is in

a metastable state. Effects related to

dehydration are reported in Chapters 3
and 4 and will not be treated here.
 Just before vitrification, the specimen
has also the opportunity to interact
with the air/water interface, which may
generate various effects related to the
high surface tension of the interface.
1. Concentration
 Interaction with the surface may result
in concentrating the specimen. This is
likely to happen very often because at
a specimen concentration of 1 mg/mL,
the particles occupy only 0.1% of the  Because the particles are immobilised at
volume and few particles would be the surface, this effect can be used to
visible. remove undesirable agents, such as glycerol
 In some instances the surface can also or high salt concentrations, by rinsing the
repel the particles and this gives rise to grid with a drop of solution.
areas where particles are absent. This
phenomenon is generally observed in
the thinnest parts of the vitrified layer.
Figure 7.2 Locally concentrated particles at

the surface of the frozen hydrated film.
2. Orientation
 In several instances, it could be
observed that the particles are
preferentially oriented at the air–water
interface, a phenomenon not only due
to thinning of the film, but to the
preferred interaction of one face of the
particle with the interface.5
Figure 7.3 Partially oriented particles at the

surface of the frozen hydrated film.

3. Denaturation
 In extreme cases, the interaction of
the specimen with the air–water
interface can severely affect the
structure and the function of the
molecule.17 Proteins can be unfolded
to expose their hydrophobic residues
or domains at the interface. Particles
smaller than expected are then
observed.
Figure 7.4 Partial denaturation of fragile

particles at the air-water interface.
2.1. Standard Method


1. Place the liquid ethane container
and the grid box into the liquid
nitrogen container.




 
 
 
 
2. Lock the EM grid on tweezers and
fix the tweezers on the plunger.
3. Fill the liquid nitrogen container

and wait 10 to 15 min for the system
to equilibrate.

 


4. Liquefy the ethane by inserting
polyethylene tubing connected to an
ethane gas cylinder into the ethane
reservoir.



5. Wait for the ethane to solidify at the

rim of the container.




6. Place 4 µL of the sample on the grid.
7. Remove the excess liquid by blotting

with a filter paper.




8. Plunge the grid into the ethane
slush.
9. Store the grid in a grid storage

device in LN2.

 Figure 7.5 (1-9) Different steps of the
 vitrification procedure.


2.2. Reduce Surface Effects

 The structure of some macro-
molecules can be affected before
vitrification and during the short
exposure to surfaces. In this case, a
number of methods may be used to
reduce the effects the airwater
interface exerts on the molecules.
1. Adsorption methods
 Absorption to a solid surface such as a  Adsorption onto a thin carbon film that
plain carbon film, rather than to a has been previously glow-discharged.
“mobile” air–water interface may help
to reduce particle reorganisation. A  In case the sample concentration is very
deformation may still result from the low, the drop of solution can be left to
interaction with carbon, but it is likely interact with the carbon support for several
to be localised at the interaction hours.
interface.
 In the sandwich method (see Section
2.4.), the specimen is placed between two
carbon films, which may further limit
specimen rearrangements.
2. Cross-linking methods
 To prevent particle reorganisation, the  Use of glutaraldehyde stored at 20°C at
specimen may be stabilised by a final concentration of 0.05 to 0.1%.
chemical cross-linking with low
concentrations of glutaraldehyde.
3. Methods using amphiphilic mole-

cules
 To prevent the particles from surface
interaction, the air–water interface can
be covered with amphiphilic lipid  A detergent or a solution of liposomes
molecules that will instantly spread as may be mixed with the particles of interest.
a monolayer at the surface.1,12,15,16 The
properties of the hydrophilic part of
the lipid, such as its charge or the  The particles may be incubated prior to
presence of a specific recognition vitrification with a monolayer of lipids
function, will trigger the interaction of spread at the air–water interface.
the protein of interest. Thus, the
protein is concentrated at the
interface, but cannot be denatured
because the surface is covered by the
lipid. Eventually this interaction can
lead to a specific orientation of the
protein and to its 2-D crystallisation.

3.1. Materials
 Carbon evaporator
 Produce holey carbon films and plain

carbon films evaporated on mica. From Boc
Edwards model Auto 306.
A = Electrodes
B = Carbon rods
C = Shutter
D = Work plate
 The carbon evaporation device can be

replaced by a metal sputtering device in
case metal nanoparticles need to be
deposited onto the thick holey carbon film
as a defocalisation aid.
Figure 7.6 Carbon evaporator.

 Glow discharge apparatus

  Used to charge electrostatically the
 surface of the carbon films to favour
specimen adsorption.14
A = Vacuum gauge
B = Discharge current reading
C = Air inlet
D = Amylamine inlet
E = Vacuum chamber air
F = Vacuum chamber amylamine
G = Rotary pump
H = Vacuum selector
I = Discharge selector
J = Voltage setting
K = Current setting


Figure 7.7 Glow discharge apparatus





 Holey carbon grids  Used as a support for the thin layer of

 suspension to be vitrified.
A = EM grid bar
B = Hole
C = Thick carbon mesh
Figure 7.8 Holey carbon grid.



 Cryo-plunger  Used to vitrify a thin layer of
suspension.
A = Sliding shaft
B = Tuneable abutment to set the height
C = Electromagnet for automatic release
D = Fastening device to hold the tweezers
E = LN2 container
Figure 7.9 Cryo-plunger

 Nitrogen container  Used to liquefy the ethane and to vitrify
the sample.
A = Nitrogen container
B = Polyurethane foam isolating device
C = Ethane container
D = Grid storage device
Figure 7.10 Nitrogen container.

 Ethane container  To keep the ethane for a longer time in a

liquid state before solidification, the
Lausanne group5 designed a double-walled
container to prevent the inner ethane
container from being in direct contact with
LN2.
A = Outer container in contact with LN2

B = Inner ethane container
C = Isolating air layer
Figure 7.11 Ethane container.
 Cryo-box for grid storage  Storage device for four grids.
 Rotationally indexed for unambiguous

grid identification.
 Screw cover numbered for device

identification.
 Long term storage in liquid nitrogen.
Figure 7.12 Cryo-box for grid storage.

3.2. Products
 Filter paper
Round ashless filter paper, 9 cm in  Used for blotting. From Macherey-
diameter, cut into small sectors Nagel, Düren, Germany, grade. MN 640 W
 Bare Cu/Rh grids, 300 or 400 mesh  From EMS, Hatfield, PA, USA.
 Tweezers  Large size, cold-isolated tweezers to

manipulate the cold grid storage device.
 Small Dumont N5 tweezers for grid
manipulation.
 LN2 tank for storage of frozen grids
 Small plastic containers to pour LN2 Several are required to avoid

contamination with moisture.
 Ruby red mica sheets for carbon  From EMS, Hatfield, PA, USA, Ref
evaporation 71851.
 Carbon floating device  Used to prepare plain carbon films and
to sandwich the specimen between two
carbon films.
 Teflon® wells for incubation with  Wells 3 mm in diameter and 2.5 mm
lipids deep bored into a Teflon block.
 Hamilton syringe  Used to deposit lipids at the air–water
interface.
 Lipid storage 2 mL glass vials  Used to keep stock solution of lipids.
 Protection glasses  Used as a protection in case of ethane
spilling.
3.3. Solutions
 Ethane gas cylinder
 LN2 supply
 Glutaraldehyde  Keep frozen aliquots to avoid
polymerisation. From Sigma, Ref G-5882.
 Incubation with lipids
 Lipids: Preferably use oleyl C18:1  Lipids are available as powders or as
side chains, which combine good film solutions and can be ordered from most
stability and fluidity. Positive charges chemical companies. Lipids with
can be introduced by ammonium and derivatised polar groups may be found at
trimethyl ammonium groups, and Avanti Polar Lipids (Alabaster, Alabama,
negative charges can be introduced by USA).
phosphatidyl serine. To reduce the
amount of charges present in the
monolayer, oleyl alcohol can be used
as a dilution lipid.
 Ethane: Chloroform 1:1 solution  To prepare dilutions of the lipids.

 To produce grids covered with holey

carbon films
 7X®PF cleaning solution  From ICN Biomedicals, Aurora, Ohio,
 Pasteur pipette USA, Ref. 76-671-21
 Cellulose acetate  From Aldrich, Ref 18,095-5

 Benzalkonium chloride  From Sigma, Ref B-1383
 Ethyl acetate  From Aldrich, Ref 27,052-0
 Dioctyl sulfo succinate  From Sigma, Ref D-4422
4. PROTOCOLS
4.1. Preparation of Holey Carbon Films8

1. Place 10 glass slides in a 500 mL  The perfect cleaning of the glass slides is
beaker, cover with 7 × detergent and essential to obtaining a homogeneous
boil for 10 min. distribution of holes.
2. Wash extensively the slides with tap

water for 30 min and then with 5
beaker volumes of distilled water.
3. Render the glass slides hydrophobic by

immersion into Benzalkonium chloride
0.5% (w/v) in water for at least 30 min.
From now on, each slide is treated
individually.  The slide has to be perfectly dry when
withdrawn.
4. Dip each individual slide into a beaker

containing distilled water. Slowly
draw the slide out of the water by
holding it by one end with tweezers.  Pre cool a 10 mm-thick aluminium plate
for a few hours at 20°C.
5. Place the glass slide onto the pre-

cooled aluminium plate. Small water  The size of the holes will increase with
droplets will form on the cold glass the relative humidity and the length of time
plate. the slide is on the cold plate.
6. Cover the glass slide with a layer of  The slide is held inclined and the
cellulose acetate dissolved to 0.4% cellulose acetate solution is poured over the
(w/v) in ethyl acetate. slide with a Pasteur pipette.

7. Let the solvent evaporate.  At this stage, the slide can be observed
under a light microscope to inspect the
holes and to select the region of interest.
8. To transfer the holey cellulose film  To peel off the cellulose film more
onto bare grids, it has to be loosened easily, the dioctyl sulfosuccinate solution
from the glass slide through an can be heated to 80°C.
immersion into a dioctyl sulfo-
succinate solution 1% (w/v) in water.
9. The loosened film can now be floated

at the surface of a glass container in
which bare grids were immersed and  In case the film sticks at the border of
placed onto a filter paper positioned on the glass slide, the rim can be scratched
a metal support. with a scalpel.
10. The cellulose film is deposited onto  The best area covered with suitable holes
the grids by lowering the water level. can be marked and the grids can be
deposited only under this area.
11. After drying, the filter paper is placed  At this stage, gold nanoparticles used as
into the carbon evaporator and a 20 to a defocalisation aid can also be deposited
40 nm-thick carbon film is evaporated by evaporation.
on top of the cellulose film.
12. The cellulose film can be dissolved  To dissolve the cellulose film, the grids,
with ethyl acetate and the grids are still placed on the filter paper, are placed
ready for use. overnight onto several layers of filter paper
soaked with ethyl acetate in a glass Petri
dish.
4.2. Vitrification of a Thin Layer

1. Select 3 to 4 EM grids coated with a  The grid may be glow-discharged in air
holey carbon film. for 30 sec to render them hydrophilic. The
electrodes are separated by 4.5 cm; a
2. Place the ethane container and the grid voltage of 110 V is applied at a partial air
box into the LN2 container of the pressure of 0.2 mbar yielding a discharge
plunging device. current of 2.5 mA.
3. Fill the container with LN2.

4. Wait 10 min for temperature  In case of incubation with lipids, the

equilibration. lipid-protein film has to be transferred onto
a grid.
5. Refill with LN2.
6. Liquefy the ethane by flowing a stream  LN2 has to be dry. Take care that few ice
of gas through plastic tubing crystals accumulate in the LN2 containers.
connected to an ethane flask into the
ethane reservoir cooled to LN2
temperature.
7. Pick up the holey carbon-coated grid  Liquid ethane may spill and, therefore,
with the tweezers and fasten the the operator should wear protective glasses.
tweezers onto the plunger.
8. Wait for the ethane to solidify.  When the ethane starts to solidify, a
white solid rim appears around the
container.
9. Apply 4 µL of the sample onto the grid

or transfer the lipid-protein layer onto
a holey carbon film.
10. Remove the excess solution by  The filter paper can be applied only on
applying a filter paper onto the drop half of the grid. In this case, a gradient of
of solution. solution thickness will be obtained and at
least one area of the grid will have a
suitable thickness.
11. Remove the filter paper and

immediately plunge the grid into the
ethane slush.
12. Rapidly transfer the frozen hydrated  Any delay at this stage will result in
sample into the LN2 container and evaporation.
place it into the storage device.

4.3. Preparation of a Holey Carbon Film Covered with a Thin,

Continuous Carbon Foil
1. Evaporate a 50 to 100 Ǻ-thick carbon  A continuous carbon film is required for
film onto a freshly cleaved mica sheet. specimen adsorption. To reduce the signal
arising from the carbon film, it is
2. Place a filter paper onto a metal recommended to use a film as thin as
support immersed in a water- possible. Such a film is not stable enough
containing glass dish. over the grid bars and needs to be supported
by the thick holey film.
3. Place about 100 EM grids covered
with a holey carbon film onto the
immersed filter paper.
4. Dip the mica sheet into the liquid with

an angle of about 45° in order to float
the carbon film on the air–water
interface.
5. Lower the water level in order to

deposit the carbon film onto the holey
grids.
4.4. Adsorption of a Specimen on a Carbon Film

1. If required, the carbon-coated holey  If required, exchange the buffer by
film (see Section 4.3) can be glow placing the grid onto a drop of the new
discharged to render it hydrophilic. buffer deposited on a flexible laboratory
The grid is placed on an electrode into film (Parafilm).
a vacuum chamber pumped down to a
partial air pressure of 2 10-1 mbar. An
alternative discharge current of 2.5 mA
is applied for 30 sec to charge the grid.
2. Five µL of the sample at a protein

concentration of about 30 µg/mL is
applied onto the charged grid and left
for 45 sec to allow adsorption.
3. Proceed as described in Section 4.2 for

vitrification.

4.5. Adsorption of Poorly Concentrated Protein Samples9

1. Cut the carbon coated mica into 4 mm  This method can be used for protein
strips. concentrations lower than 5 µg/mL.
2. Deposit 75 µL of the protein sample

into a Teflon well 5 mm wide, 5 mm
long and 3 mm deep.
3. Float the carbon film from the mica

strip at the surface of the protein
sample placed in the Teflon well.
4. Leave the carbon in place for 12 hours

to allow specimen adsorption.
5. Plunge an EM grid coated with a holey

carbon film into the Teflon well and
pick up the floating carbon onto which
the specimen is adsorbed.

vitrification.
4.6. Chemical Cross-Linking

1. The sample is incubated for 2 min at  This method is used to stabilise the
room temperature with glutaraldehyde specimen independently of the subsequent
at a final concentration of 0.1 to preparation method.
0.05%.
2. The incubation time and glutar-

aldehyde concentration depends on
sample composition. It is strongly
recommended to screen the cross-
linking conditions by negative staining
EM to find the optimal conditions in
which the sample is stable and does
not form aggregates.

4.7. Preparation of a Sandwich9

1. Adsorb the specimen as described in  This method can be used to further
Section 4.4. immobilise the specimen by sandwiching it
between two carbon films.
2. Cut a carbon coated mica into 4 mm
strips.
3. Deposit 75 µL of buffer into a Teflon  This method can be combined with

well 5 mm wide, 5 mm long and 3 mm negative staining when the solution in the
deep. Teflon well is replaced by a 2% uranyl
acetate solution. The stained sample is dried
4. Float the carbon film from the mica for 2 min and plunged into LN2 before
strip at the surface of the buffer placed observation to keep a hydration shell
in the Teflon well. around the specimen.
5. Plunge the EM grid coated with a

holey carbon film on which the
specimen is adsorbed into the Teflon
well and pick up the floating carbon.

vitrification.
4.8. Incubation with a Lipid Monolayer1,3,12,15,16

1. Place the Teflon support into a Petri
dish cover placed upside down; add
5 mL of buffer into the Petri dish to
seal the incubation chamber once the
Petri dish base is placed upside down
on top of the chamber.
2. Place 10 µL of buffer into a Teflon

well 3 mm in diameter and 2.5 mm
deep.

3. Place 1 µL of lipids dissolved in a 1:1

chloroform/hexane solution at a
concentration of 0.5 mg/mL.
4. Add 5 µL of the sample. Different
concentrations need to be tested
between 30 and 500 µg/mL.
5. Incubate for various time periods
ranging between 1 and 24 hours.
6. Place an EM grid coated with a plain
or a holey carbon film on top of the
drop for 5 min.
7. Lift the EM grid from the drop. A
good transfer is witnessed by the
presence of a thin layer of solution on
the EM grid.
8. The transferred lipid-protein layer can
be negatively stained or processed for
cryo-EM as described in Section 4.2.
Figure 7.13 Illustration of the protein-lipid
layer incubation and transfer to the EM
grid.
4.9. Specimen Transfer

1. Cool down the anticontaminator and
the cryofork of the microscope.
2. Cool down the cryo-holder by pouring

LN2 into the Dewar and into the grid
mounting device.
A = Dewar of the cryo-holder

B = Mounting device
Figure 7.14 Grid mounting station.
3. Retrieve a cryo-box containing the

samples from its long-term storage
device and place it into the grid
mounting device.

A = Position of the grid

B = Place for the grid-box
C = Cryo-shield
D = Reservoir
Figure 7.15 Holder cryostation and grid

mounting device
4. Take out a grid from the cryo-box and

 Solid ethane may remain attached to the
mount it onto the cryo-holder. Close
grid at this stage and may cause problems
the cryo-shield over the sample to during transfer or EM observation. The
prevent ice contamination during the
solid ethane may be scratched off with the
transfer. tweezers before specimen mounting onto
5. Transfer the cryo-holder into the the cryo-holder.
electron microscope as recommended
by the manufacturer.
4.10. Low-Dose Imaging

1. Recording images at the lowest  Search mode: Select an area of interest
possible electron dose is required to (circle A) at low magnification (typically
limit the irradiation damage of the 5,000×) with the lowest possible
specimen. irradiation.
 Focus mode: Focus the objective lens at
high resolution (typically 40,000×), on a
neighbouring area (B), close enough to
have little changes in height and far enough
to avoid irradiation of the area of interest.
This is performed using an image shift
whose angle and amplitude can be selected.
 Record mode: Record the image at the
required magnification of the area of
interest.
Figure 7.16 Illustration of the principle of

low dose imaging.
2. Images can be recorded on film plates  To improve sensitivity, data may be

or on charged coupled device (CCD) recorded on SO163 films (Kodak) and
cameras. developed for 12 min with undiluted D19
developer.

4.11. Image Analysis

1. Image analysis of single molecules  Information on two commonly used
aims principally at improving the image analysis software suites can be
resolution of the noisy images and at found on:
determining a 3-D model of the http://www.wadsworth.org/spider_doc/spid
particles. This section is not meant to er/docs/spider.htmL
describe the image analysis protocols, and
but to get some starting directions.2- http://www.imagescience.de/
4,7,18
2. Digitise the images. Unless recorded  Most widely used scanners are flat bed
on CCD cameras, the silver halide (Nikon Coolscan 8000) or drum scanners
emulsions have to be digitised using a (Heidelberger Druck Maschinen,
high-resolution microdensitometer. Primescan D7100). They should provide a
linear response with at least 5000 dpi
3. Particle picking. An image repre- resolution.
senting a large field containing many  Several free displaying systems are
particles (see Figure 7.1) has to be available on line. Among them, EMAN can
displayed on a monitor and the be used for particle picking:
coordinates of the particles have to be http://blake.bcm.edu/EMAN/
determined.
4. Alignment. Particle images have to be  Cross correlation functions are generally

iteratively aligned in rotation and in used to align images.19
translation against references to be
placed in register.
5. Clustering. Similar images have to be  Multivariate statistical methods are used

grouped into classes according to their to classify images.
maximal resemblance.
6. Averaging. Aligned particle images  An image class corresponds generally to

within each class are averaged to a particular view of the particle.
improve the signal-to-noise-ratio.
7. 3-D reconstruction. The different

views of the particle have to be
combined to form a 3-D model of the
particle.
 An interactive molecular graphic
8. Visualisation. The 3-D model can be
program can be found under:
displayed and compared to other
http://www.cgl.ucsf.edu/chimera/
existing models.
 A protein structure data base is hosted
at: http://www.wwpdb.org/
 A macromolecular structure data base is
found under: http://www.ebi.ac.uk/msd/

4.12. Troubleshooting
1. No particles are observed
 Is particle concentration in the correct  The concentration should be between
range? 0.3 and 1 mg/mL.
 Particles smaller than 50 kDa are
difficult to spot at low dose.
 Is the contrast sufficient?  The presence of glycerol or sucrose will
increase the density of the media and will
reduce the contrast of the particles.
 Are the particles hidden in the thicker  Some particles do not like to be close to
part of the frozen hydrated layer? the surface and escape into the thick parts
of the vitreous water.
 Are the particles denatured?  In this case, smaller particles are
observed in the background indicating that
 Are the particles aggregated? the large assembly is dissociated.
2. Ice contamination
 The grid surface may be covered with
long filamentous aggregates that arise from
impurities present in the ethane. These
aggregates tend to be more numerous when
the grid is covered with a thick layer of
solid ethane that evaporates during transfer
into the microscope. Careful removal of the
excess ethane before mounting the grid
helps to avoid this problem.
 The grid may be contaminated with
hexagonal ice crystals that were present in
LN2 or may have condensed during grid
transfer. Working in a dry atmosphere,
heating up the cryo-holder between
successive transfers into the microscope
and possibly emptying and drying the LN2
container generally limits this problem.
 The grid surface may be covered by a
thin, almost continuous cubic ice layer
arising from the condensation of water
present in the microscope. This form of
contamination may appear like faint
discontinuous spots. It may reflect high
water pressure around the specimen either
because the anticontaminator is not
working properly or because poorly
adsorbed ice crystals are melting or
because there is a small leak in the air lock.
Figure 7.17 Illustration of ice con-

tamination.

5.1. Standard Method

5.1.1. Advantages
 No adsorption effects.  Use of holey carbon films.
 No staining effects. 
 Physiological ionic strength.  Beware of the effects of evaporation
after blotting.
 Fully hydrated structure.
 Particles observed through all orient-  Important for image analysis of single
tations. particles and 3-D model reconstruction.
5.1.2. Disadvantages
 Low intrinsic contrast.  Use of phase contrast and, thus, need to
correct for the contrast transfer function
(CTF) of the microscope to have a faithful
representation of the particle.
 High protein concentration.  0.3 to 1 mg/mL.
 Possible surface effects.  The structure of the specimen may be
affected.
 Particles pushed together may aggregate
or be too close to be separated.
5.2. Adsorption on Carbon

5.2.1. Advantages
 Adsorption concentrates the specimen.  Use plain carbon films.
 10 to 30 µg/mL; concentration range.
 Adsorption may reduce surface  The immobilisation of the particles
induced aggregation. reduces particle redistribution during thin
film formation.
 Carbon film gives a strong signal to 

determine the CTF of the microscope.
 Uniform particle distribution. 
 Particles in the same plane.  All particles have the same CTF.

 Preferred orientations.  Reduces the angular range for 3-D
reconstruction.
 Adsorption induced deformation.  The surface of the particle in contact
with the carbon may be altered.
 Carbon film introduces a granularity  Use of thin carbon films.
that may affect particle alignment.
5.3. Glutaraldehyde Cross-Linking

5.3.1. Advantages
 Stabilises specimen structure.  Prevents surface denaturation.
 May promote aggregation.  Adjust glutaraldehyde concentration.
 Stabilise lowest energy conformation.  Dynamic information is questionable.
5.4. Interaction with Lipid Layers

5.4.1. Advantages
 Covers the air–water interface.  Prevents surface denaturation.
 Mobile surface.  Promotes particle organisation.

 Two-dimensional crystallisation.
 Lipid–particle interaction.  The structure of the molecules may be
affected by this interaction.
 Possible orientation of the particles.  Tilting the specimen may help to record
all particle views.

6.1. Stable Specimen Available at High Protein Concentration

 Use of the standard vitrification  Best structural preservation
technique (see Section 4.2) that avoids  Most suitable to recover excellent
adsorption, staining and dehydration structural information upon image analysis.
 Most suitable to detect functional
artifacts.
conformational changes and to analyse
specimen dynamics.

6.2. Stable Specimen Available at a Concentration Lower Than
0.3 mg/mL
 In this case, an adsorption method is  For any method used, the interaction of
required to concentrate the specimen. the specimen with the support may alter the
specimen structure. However, high
resolution has been obtained with both
methods.
6.2.1. Adsorption on a carbon film (see Section 4.4)

 The specimen will interact with the  The surface properties of the carbon film
surface of the carbon film and will are generally hydrophobic and may be
concentrate at the interface. modified by glow-discharge.
6.2.2. Interaction with a lipid layer (see Section 4.8)

 The interaction of the specimen with  Lipids carrying an electrostatic charge or
the lipids can be tuned by using a specific recognition function, such as a
different lipid mixtures. ligand of the protein to be studied, can be
used.5
6.3. Unstable Specimen

 An unstable specimen will appear  Additional tests may be performed by
aggregated or dissociated when using buffers containing different ionic
processed using the standard strengths in order to find conditions where
the specimen might be stable.
vitrification method. Different
 The criteria for having little surface
strategies may be used to prevent denaturation is the structural homogeneity
specimen dissociation or aggregation. of the sample.

6.3.1. Chemical stabilisation

 Glutaraldehyde or other cross-linking  Be aware that the stabilised species may
agents can be used to stabilise the not correspond to the functional state of the
protein complex. complex.
6.3.2. Adsorption of the specimen

 Different adsorption methods can be
used that may help to prevent surface
denaturation. The sandwich method
may further reduce molecular
movements and aggregation prior to
vitrification.
6.3.3. Use of lipid layers

 Incubation with a spread lipid layer
prevents interaction of the specimen
with the air–water interface and
subsequent denaturation.

7. OBSERVED RESULTS
 Figure on the Chapter’s title page (see

Figure 7.19)
 Figure 7.18
Cryo-electron microscopy observation of a
frozen hydrated suspension of yeast RNA
polymerase at a concentration of
0.6 mg/mL in a buffer containing 10 mM
Tris/HCl pH 7.4 and 100 mM NaCl.
 Figure 7.19
Image analysis of the molecular images
shown on the title page. The top panel
shows several characteristic class averages
obtained upon several alignment/classifi-
cation iterations. Each class average
corresponds to a different view of the RNA
polymerase I molecule. The lower panel
shows the three-dimensional model
obtained by combining the different class
averages.

188 Handbook of Cryopreparation Methods for Electron Microscopy
 Figure 7.20
Panel of cryo-electron micrographs
showing different preservation conditions
and illustrating the effects of surface
denaturation. The upper left panel shows
well preserved RNA polymerase molecules
with an optimal distribution. The molecules
are represented as circles in the right panel.
The middle panel shows partially denatured
particles (represented as squares in the right
panel) that appear heterogeneous in size and
shape. The lower panel shows an
inhomogeneous distribution of the particles
that tend to stick to each other.
 Figure 7.21
Illustrations of yeast RNA polymerase I
molecules incubated with positively
charged lipids spread as a monolayer at the
air–water interface. The upper panel
represents a short incubation time (5 min)
and shows that the molecules (initial
concentration of 50 µg/mL) are
concentrated at the lipid/air interface. This
lipid/protein interaction favours or
stabilises RNA polymerase dimers that are
observed in large amounts. The middle
panel was obtained after one hour
incubation and shows the organisation of
the dimeric molecules into rows. Finally,
the lower panel shows an imperfectly
ordered, two-dimensional crystal formed at
the lipid interface.

8. REFERENCES
1. Bischler, N. et al. Specific interaction and two-dimensional crystallization of

histidine tagged yeast RNA polymerase I on nickel-chelating lipids, Biophys. J., 74,
1522, 1998.
2. Bischler, N. et al. Localization of the yeast RNA polymerase I-specific subunits,
Embo J., 21, 4136, 2002.
3. Crucifix, C., Uhring, M., and Schultz, P. Immobilization of biotinylated DNA on
2-D streptavidin crystals, J. Struct. Biol., 146, 441, 2004.
4. De Carlo, S. et al. Cryo-negative staining reveals conformational flexibility within
yeast RNA polymerase I, J. Mol. Biol., 329, 891, 2003.
Biophys., 21, 129, 1988.
6. Dubochet, J. et al. Emerging techniques: Cryo-electron microscopy of vitrified
biological specimens, Trends in Biochem. Sci., 10, 143, 1985.
7. Frank, J. Three-Dimensional Electron Microscopy of Macromolecular Assemblies:
Visualization of Biological Molecules in Their Native State, Oxford University
Press, Oxford, U.K.,2006.
8. Fukami, A. and Adachi, K. A new method of preparation of a self-perforated
micro plastic grid and its application, J. Electron Microsc. (Tokyo), 14, 112, 1965.
9. Golas, M.M. et al. Major conformational change in the complex SF3b upon
integration into the spliceosomal U11/U12 di-snRNP as revealed by electron
cryomicroscopy, Mol. Cell., 17, 869, 2005.
10. Jaffe, J.S. and Glaeser, R.M. Preparation of frozen-hydrated specimens for high
resolution electron microscopy, Ultramicroscopy, 13, 373, 1984.
11. Kasas, S. et al. Vitrification of cryoelectron microscopy specimens revealed by
high-speed photographic imaging, J. Microsc., 211, 48, 2003.
12. Lebeau, L. et al. Specifically designed lipid assemblies as tools for two-
dimensional crystallization of soluble biological macromolecules. Handbook of
Nonmedical Applications of Liposomes, Vol. 2, Barenholz, Y. and Lasic, D.D. eds.,
CRC Press, Boca Raton, FL, USA, 1996, 155.
13. Lepault, J., Booy, F.P., and Dubochet, J. Electron microscopy of frozen
biological suspensions, J. Microsc., 129, 89, 1983.
14. Oudet, P. et al. Electron microscopy of simian virus 40 minichromosomes,
Methods Enzymol., 170, 166, 1989.
15. Schultz, P., Bischler, N., and Lebeau, L. Two-dimensional crystallization of
soluble protein complexes, Methods Mol. Biol, 148, 557, 2001.
16. Schultz, P. et al. Three-dimensional model of yeast RNA polymerase I determined
by electron microscopy of two-dimensional crystals, Embo J., 12, 2601, 1993.
17. Tanford, C. Protein denaturation, Adv. Protein Chem., 23, 121, 1968.
18. van Heel, M. et al. Single-particle electron cryo-microscopy: Towards atomic
resolution, Q. Rev. Biophys., 33, 307, 2000.
19. van Heel, M., Schatz, M., and Orlova, E.V. Correlation functions revisited,
Ultramicroscopy, 46, 307, 1992.

TwoDimensional Crystals 193
CONTENTS

1.1. 2-D Crystallization of Membrane Proteins ............................................... 196
1.1.1. Naturally occurring crystals ........................................................ 196
1.1.2. 2-D crystallization by reconstitution........................................... 196
1.2. 2-D Crystallization of Soluble Proteins .................................................... 198
1.2.1. Lipid monolayer crystallization .................................................. 198
1.2.2. Surface crystallization................................................................. 199
1.3. Cryo-Preparation of 2-D Crystal Samples ................................................ 199
2.1. 2-D Crystallization by Membrane Fusion ................................................ 200
2.2. 2-D Crystallization by Reconstitution ...................................................... 200
2.2.1. Crystallization by dialysis........................................................... 200
2.2.2. Crystallization using BioBeads................................................... 201
2.3. Crystallization under a Lipid Monolayer.................................................. 202
2.4. Surface Crystallization ............................................................................. 203
3.1. Materials ................................................................................................... 204
3.2. Products .................................................................................................... 205
3.3. Solutions ................................................................................................... 205
4. PROTOCOLS .................................................................................................... 206
4.1. Membrane Fusion of Purple Membranes.................................................. 206
4.2. 2-D Crystallization by Dialysis ................................................................ 206
4.3. 2-D Crystallization using BioBeads ......................................................... 207
4.4. Preparation of Cryo-Samples by the Back-Injection Method ................... 208
4.5. Lipid Monolayer Crystallization............................................................... 209
5.1. Using Natural Crystals.............................................................................. 213
5.1.1. Advantages.................................................................................. 213
5.1.2. Disadvantages ............................................................................. 213
5.2. 2-D Crystallization by Dialysis ................................................................ 213
5.2.1. Advantages.................................................................................. 213
5.2.2. Disadvantages ............................................................................. 213

5.3. 2-D Crystallization Using BioBeads ........................................................ 214

5.3.1. Advantages ................................................................................. 214
5.3.2. Disadvantages............................................................................. 214
5.4. Lipid Monolayer Crystallization .............................................................. 214
5.4.1. Advantages ................................................................................. 214
5.4.2. Disadvantages............................................................................. 214
5.5.1. Advantages ................................................................................. 214
5.5.2. Disadvantages............................................................................. 214
6.1. 2-D Crystallization of Membrane Proteins............................................... 215
6.1.1. Natural 2-D crystals.................................................................... 215
6.1.2. Detergent removal by dialysis .................................................... 215
6.1.3. Detergent removal using BioBeads ............................................ 215
6.2. 2-D Crystallization of Soluble Proteins.................................................... 215
6.2.1. Lipid monolayer crystallization.................................................. 215
6.2.2. Surface crystallization ................................................................ 215
6.3. Specimen Preparation............................................................................... 215
8. REFERENCES .................................................................................................. 218


Two-dimensional (2-D) crystallization of

proteins is a powerful alternative to x-ray
crystallography when 3-D crystals of
sufficient quality cannot be grown or when
proteins are not suitable for this technique;
for instance, when a protein aggregates
above a specific concentration. The aim of
this technique is to force lateral interactions
between proteins in a plane and thereby to
induce the formation of 2-D crystals, as
shown in Figure 8.1.
Cryo-electron micrographs of 2-D crystals

have very low contrast. However, it is
possible to investigate the structure of
proteins by image analysis based on Fourier
methods because the molecules exist in
thousands of identical copies.
Developments in sample preparation,
imaging and image analysis now make it
possible to resolve protein structures from
intermediate (6 to 10 Å) (see Figure 8.25)
to high (2 to 4 Å) resolution from 2-D
crystals.6
Initially, electron crystallography was
mainly dedicated to the study of membrane
proteins.7 Over the last decades, new
methods have been developed to induce
2-D crystallization of both membrane and
soluble proteins.
The goal of this chapter is to present
commonly used methods to induce 2-D
crystallization of proteins and related cryo-
preparation techniques.
Figure 8.1
 A: HC-Pro, a soluble protein
expressed with a 6His-tag, can interact
with a lipid monolayer.
 B: When the pH is raised to 8.4,
lateral interactions between proteins
induce formation of a regular array.
Bar = 50 nm

2-D crystals can occur naturally or can be

produced by different approaches,
depending on the chemical properties of the

proteins. In this context, we will distinguish
between membrane and soluble proteins.

1.1. 2-D Crystallization of Membrane Proteins
1.1.1. Naturally occurring crystals
A few membrane proteins occur naturally in  Bacteriorhodopsin is a classic example.
2-D arrays. These protein arrays can be
purified and prepared for electron
microscopy (EM) directly. When the  Bacteriorhodopsin membrane patches
patches are too small, they can be merged can be fused using detergents (see Protocol,
into larger arrays of several µm in diameter. Section 4.1).


1.1.2. 2-D crystallization by reconstitution
The most general method for 2-D  The protein purification step is outside
crystallization of membrane proteins the scope of this book. We will suppose
involves mixing detergent-solubilized that the proteins are purified, solubilized in
proteins with suitable lipids, also in the presence of detergent and have a
detergent solution, and the removal of concentration of at least 1 mg/mL.
detergent that induces the reconstitution of 
the protein in a lipid bilayer (see Figure
8.2).

 mp = membrane protein
 d = detergent
Figure 8.2 Proteins are solubilized in

detergent.

Lipids used for reconstitution are prepared

in detergent (see Figure 8.3).
 l = lipid
 d = detergent
Figure 8.3 Lipids are solubilized in

detergent.
 Lipids commonly used are phos-
phatidyl choline (PC), dioleyl phos-
phatidyl choline (DOPC), dioleyl
phosphatidyl glycerol (DOPG),
dimyristoyl phosphatidyl choline
(DMPC). For a review see Walz and
Grigorieff.12
Proteins and lipids are mixed together (see
Figure 8.4).
 d = detergent
 l = lipid
Figure 8.4 Membrane proteins and lipids

are mixed together and incubated
overnight.
 The concentration of free detergent
has to be above the critical micelle
concentration (CMC).
Detergent is removed inducing formation of

a lipid bilayer (see Figure 8.5).
 d = detergent
 l = lipid
Figure 8.5 A lipid bilayer is reconstituted

by detergent removal.
 Proteins are usually inserted in an up
and down orientation in the mem-
brane.
The detergent can be either removed by
dialysis using a membrane that is permeable
to detergent monomers but not to micelles  A 10 kDa molecular weight cut-off is
(see Protocol, Section 4.2) or by absorption generally used.
onto polystyrene beads (see Protocol,
Section 4.3).

The dialysis can take days to weeks,  Detergent removal by BioBeads® is

depending mainly on the critical micelle much faster than by dialysis,10 which can
concentration (CMC) of the detergent. be an advantage when the CMC of the
Under appropriate conditions, regular arrays detergent is very low or when the protein is
will be formed. The most important very unstable.
parameters include the nature of the lipid (or  Determination of crystallization
lipid mixture), the lipid-to-protein ratio, the conditions is empirical. This means that all
temperature, pH, ionic strength and parameters should be tested individually
presence of additives. and in combination with others.
1.2. 2-D Crystallization of Soluble Proteins

1.2.1. Lipid monolayer crystallization
2-D crystallization can be induced under a  Proteins should have a concentration

monolayer of lipids (for a review, see Chiu above 1 mg/mL in a suitable buffer.
et al.3). The interactions between proteins

and lipids will force the proteins into a fixed
orientation and increase the concentration of 
proteins under the monolayer (see Figure
8.6).
 Interactions of proteins with the lipid

monolayer can be electrostatic or lipid-
specific, as for instance with a nickel
functionalized lipid.

Figure 8.6 The lipid monolayer is formed

at the air-buffer interface. Proteins bind
functionalized lipids.
 The lipid monolayer consists of
functionalized lipids and diluting
lipids, which play a role in the lateral
diffusion of the proteins.
 Dl = Diluting lipid
 Fl = Functionalized lipid
 P = Protein
 Protein with its interaction region with
the functionalized lipid.

1.2.2. Surface crystallization

During the 3-D crystallization of proteins,  Crystallization of the solubilized
physical phenomena involving specific H+ATPase from Neurospora crassa has
interactions between proteins occur at the been performed using this method.5
air-buffer interface. Thus, in a drop in
which 3-D crystals are formed, it is
interesting to observe what happens at the
surface of the drop. Under favorable
situations, 2-D crystallization can also
occur at the airbuffer interface.

1.3. Cryo-Preparation of 2-D Crystal Samples

2-D crystals are very large, but thin. For
cryo-preparation it is important to have the
crystals stretched out over the support film.
Flatness of the crystals on the support film  Depending on the size of the unit cell of
is essential for successful high-resolution the crystal and the expected resolution, a
data collection. deviation of as low as 1o over several µm
 can be tolerated.11

For cryo-microscopy, most 2-D crystals can  The carbon film usually wrinkles after
be deposited on carbon support films and freezing because the expansion coefficients
plunged into liquid ethane using a classical of copper1 and amorphous carbon do not
plunge-freezing device (see Chapters 3, 7). match. In practice, molybdenum is the grid
material of choice because it maintains the
film flatness best at liquid nitrogen (LN2)
temperature.
A plunge-freezing device in a chamber with  At a very high humidity level, a thin

controlled atmosphere (see Chapter 4) is layer of water is still preserved at the
very useful in some cases. surface of 2-D crystals.
2-D crystals can alternatively be  Glucose, trehalose or tannin are

cryoprotected by a sugar. This method, commonly used cryoprotectants.
which is most suitable for 2-D crystals of
membrane proteins, will be presented in
this chapter (see Protocol, Section 4.4).

2.1. 2-D Crystallization by Membrane Fusion

Membrane patches are mixed together in
the presence of detergents and incubated
for several weeks.
2.2. 2-D Crystallization by Reconstitution

2.2.1. Crystallization by dialysis
1. Mix detergent solubilized proteins,
lipids and detergent.
2. Incubate with gentle magnetic
stirring for 1 to 24 hours.
3. Transfer the mixture into a dialysis
device without forming air bubbles
on the dialysis membrane (see
Figures 8.7 and 8.8).
4. Put the dialysis device into a
detergent-free buffer (see Figure
8.9).
5. Incubate at fixed temperature with
gentle magnetic stirring of the
buffer.
Figure 8.7 6. Change buffer two or three times
during the first 72 hours.
7. Incubate several days.
8. Take a sample.
9. Deposit a few µL of solution onto a
carbon-coated grid (usually 300 or
400 mesh grids) and wait for
two minutes. The grid can be glow-
discharged before using it in order
to improve the interaction and
flatness of 2-D crystals with the
carbon film of the EM grid.
10. Blot the grid and freeze it in liquid
ethane using a plunge-freezing
device.
9&10 Alternatively, a cryo-sample can
be prepared with the back-injection
Figure 8.8 method (see Protocol, Section 4.4).
Figure 8.7 Dialysis device (Slide-A-

Lyzer®) from Pierce.
Figure 8.8 Bent glass capillary tube.

 P = Parafilm
 B = Buffer without detergent
 M = Magnetic bar
Figure 8.9 Overview of the dialysis device.
2.2.2. Crystallization using BioBeads

1. Mix proteins and lipids in detergent in
a microcentrifuge tube.
2. Incubate with gentle magnetic stirring
for 2 to 24 hours at 16°C.
3. Add wet BioBeads (see Figure 8.10A).
4. Incubate at 16°C with gentle magnetic
stirring.
5. To stop adsorption of hydrophobic
molecules to the BioBeads, remove the
Figure 8.10A sample and transfer it to a new tube
(see Figure 8.10B).
6. Freeze-thawing cycles can be used to
induce 2-D crystallization (see 2-D
crystallization of cytochrome b6f2).
The sample is plunged into liquid
nitrogen and thawed on ice. The cycle
is repeated at least three times.
Figure 8.10 BioBeads in detergent

removal.
 A: BioBeads are added to the solution.
 B: Detergent removal is stopped by
transferring solution into another tube.
Figure 8.10B

7. Deposit a few µL of solution onto a

carbon-coated grid and wait for two
minutes. Depending on sample
properties, the interaction and flatness
of 2-D crystals with the carbon
covering the grid can be improved by
using glow-discharged grids.
ethane using a plunge-freezing device.
7&8. Alternatively, a cryo-sample can be
prepared with the back-injection method
(see Protocol, Section 4.4).
2.3. Crystallization under a Lipid Monolayer
1. Wash a Teflon® crystallization plate

(see Figure 8.11) and place it in a Petri
dish that contains two layers of wet
filter paper.
2. Inject the buffer into the incubation
well. The surface of the well must be
perfectly plane. The volume of the well
is about 50 µL.
3. Deposit 0.5 µL of lipids at the surface
of the incubation well.
4. Incubate at least 4 to 12 hours at 16°C.
This temperature decreases the
dehydration phenomenon.
5. Inject 3 to 8 µL protein solution using
the injection well.
6. Protein incubation is performed under
very gentle magnetic stirring for 24 to
48 hours at 16°C.
Figure 8.11 The Teflon crystallization plate.
The technical features of this crystallization
plate are described in Levy et al.8
7. Deposit a carbon-coated grid at the
surface of the well and wait for
two minutes. The grid can be glow-
Figure 8.11 discharged before using it.

8. Pick up the grid and transfer it to a

plunge-freezing device.
ethane.
2.4. Surface Crystallization
1. Take a cell culture plate and fill the

wells with 1 mL of buffer.
2. Deposit 7 to 10 µL of the protein
solution onto a siliconized cover slide.
3. If 2-D crystallization occurs in the first
hours then a gold grid covered with a
carbon film is deposited onto the drop.
4. Apply grease around the crystallization
well for sealing.
5. Return the cover slide and seal it onto
the well.
6. Incubate the plate at 16°C. For the H+-
ATPase, 2-D crystals appear within the
first two hours of incubation.
7. Remove the cover slide, pick up the
grid with standard tweezers and place
it into a freezing device.
7’. When 2-D crystallization occurs
slowly (>24 hours), the EM grid is
only deposited onto the drop after
having removed the cover slide. Wait
for at least two hours at 16°C and then
pick up the grid with tweezers and
place it into a plunge-freezing device.
ethane.
Figure 8.12 The crystallization plate.

3.1. Materials
 A freezing device for cryo-samples.  See Chapters 3, 4, 7
 BioBeads.  SM2 (BioRad)

www.bio-rad.com
 Dialysis devices:  Any device that can dialyze volumes

between 10 to 100 µL.
 Bent glass capillary  Bent-glass capillary is homemade. A
 Dialysis button (Hampton Research) piece of dialysis membrane is fixed at the
 Slide-a-Lyzer (Pierce) short end with a rubber ring or a piece of
plastic tubing. These vessels are easy to
make, cheap and can be reused. The sticks
should be prepared and placed in buffer
several hours before filling to check for
leaks in the membrane.
www.hamptonresearch.com/
www.piercenet.com/
 Dialysis membrane.  10 kDa cut-off
 Fixed temperature incubating room.  16°C is the commonly used temperature.
 Refridgerator.  Storage of BioBeads, buffers, etc.
 Magnetic stirrers.  Small magnetic stirrers are used for

mixing proteins and lipids. Volumes
usually used are in the 50 to 100 µL range.
 2 mm × 2 mm magnetic stirrers are used
in Teflon wells.
 Teflon crystallization plate.  The depth of the well allows magnetic

stirring. The injection well is used for
protein injection.
 Drawing and technical features are
presented in Levy et al.8

3.2. Products
 Carbon-coated grids.  Commercially available. For example:
 Gold-coated grids.  Oxford Instruments
www.oxford-instruments.com
 Molybdenum grids.  E.M.S.
www.emsdiasum.com/microscopy/
For molybdenum grids, 300 mesh is the
smallest mesh size available
 Pacific Grid Tech
www.grid-tech.com.
 Diluting lipids.  Dioleyl phosphatidyl choline or dioleyl
phosphatidyl glycerol for instance
 DOGS-NTA lipid.  Commercially available from Avanti
Polar Lipids (www.avantilipids.com/)
 In chloroform/methanol (9:1) (v/v) at
0.5 mg/mL. Synthetic lipids with chelated
nickel that can interact with histidine-
tagged proteins.
 DTAC.  DTAC can be supplied by:
(N, N-dodecyl trimethyl ammonium www.sigmaaldrich.com
chloride).
 Ethane gas.  Extremely inflammable. Always use
ethane in well-ventilated areas.
 Liquid nitrogen.  May cause severe burns; wear protective
clothes and gloves for handling LN2;
always use LN2 in well-ventilated areas.
 Vacuum grease.  The grease is used to seal the cover slide
onto the well of the cell culture plate.
Can be supplied by:
www.hamptonresearch.com

3.3. Solutions
 Buffers
 100 mM potassium phosphate buffer  Used for fusion of purple membranes.
(pH 5.2) with octyl glucoside (6 mM)
and DTAC (200 µM).
 For cryoprotection
 Glucose  Usually 2% (w/v) aqueous solution
 Trehalose  A 4 to 8% (w/v) aqueous solution
 Tannin  0.5% (w/v) adjusted to pH 6.0 with KOH
and clarified by centrifugation.

4. PROTOCOLS
4.1. Membrane Fusion of Purple Membranes

1. Mix purple membranes in a suspension  Use 100 µL aliquots with slight
containing 3 mg/mL in 100 mM variations in detergent concentration.
potassium phosphate buffer (pH 5.2)
with octyl glucoside (6 mM) and  Adding 3 mM sodium azide will prevent
DTAC (200 µM). bacterial growth.
2. Incubate for several weeks at room

temperature.
3. Check for large sheets (10 µm or  2-D crystals of membrane proteins are
more) by EM. often cryoprotected by sugars before
freezing using the back injection method
4. Prepare crystals for cryo-electron (see Protocols, Section 4.4).
microscopy. 
4.2. 2-D Crystallization by Dialysis

1. Mix together in a 10 to 100 µL 
volume: 

 Membrane protein in detergent, about  Lipid/protein ratio: 0.1 to 10 (w/w). The
1 mg/mL. suitable ratio should be determined
 Lipid in detergent. empirically.
 Additional detergent, if necessary.  The final free detergent concentration
has to be above its CMC.
 10 to 15% glycerol, if desired.  Addition of glycerol often produces
 Add buffer to obtain the final volume. better crystals.


2. Let the mixture equilibrate at room
temperature for one hour.
3. Transfer the mixture to one of the  Use slow magnetic stirring.

following dialysis devices:
 Slide-A-Lyzers.  They have a large surface area, are
suitable for 100 µL volumes or more, are
very easy to use and a sample can be taken
during dialysis.
 Bent glass capillary.  The surface area is small, so dialysis is
slow.
 Dialysis button.  For very small sample volumes (down to
10 µL). They are not easy to handle. The
ratio of surface-to-volume is low.

4. Dialyze against detergent-free buffer  All dialysis installations are placed in 30

for several days or weeks. to 250 mL dialysis buffer and dialyzed at
defined temperature, most commonly at 20
5. Check for 2-D crystals by negative to 37°C.
staining.
6. Adjust crystallization parameters or  Dialyzing time can vary from two days
prepare crystals for cryo-electron to several weeks or months for detergents
microscopy. with a low CMC.
 Online Fourier transform of charged
coupled device (CCD) images can be used
to check if 2-D crystals are formed.
 See Figure 8.23.
 2-D crystals of membrane proteins are

often cryoprotected by sugars before
freezing using the back injection method
(see Protocol, Section 4.4).
 Crystal quality can sometimes be
improved by incubation at a higher
temperature (37oC) for a few hours.
4.3. 2-D Crystallization using BioBeads

1. Mix proteins with lipids as described 
(see Section 4.2, Step 1) in a 100 µL
volume.
2. Wash BioBeads thoroughly with  BioBeads can be prepared and then

methanol and then with water or stored at 4°C.
buffer.  To avoid bacterial contamination,
BioBeads need to be washed with filtered
3. Discard dry BioBeads. water before use.

4. Deposit a small amount of BioBeads  BioBeads must always remain wet.
onto a filter paper.
5. Weigh the desired amount of beads  Use a precision balance. The amount of
and add to protein/lipid mixture. BioBeads needed is in the mg range.
 The amount of BioBeads needed to
remove detergents is described in Rigaud et
al.10
 If detergent removal is too rapid, the
beads can be added in small aliquots every
hour.
6. When all the detergent is adsorbed to  Remove the sample from the bottom of
the beads, the sample has to be the tube with a pipette. The BioBeads
transferred to a new tube. remain in the tube.

7. Incubate the sample at 16°C.
8. Check for 2-D crystals by EM after

negative staining.
9. Adjust crystallization parameters or  2-D crystals of membrane proteins are

prepare crystals for cryo-electron often cryo-protected by sugars before
microscopy. freezing using the back-injection method
(see Protocol below).
4.4. Preparation of Cryo-Samples by the Back-Injection Method
1. 200 µL of sugar solution is put on a 

piece of Parafilm (see Figure 8.13).

Figure 8.13 Sugar solution onto Parafilm.
 Sugars are used to cryoprotect the
specimen.
 Glucose, tannin or trehalose can be
used.

2. A 3 mm × 3 mm piece of carbon film 
on mica is floated off and picked up on
a molybdenum grid (see Figure 8.14).
Figure 8.14 Carbon film is picked up on a

EM grid.

3. The grid is turned over so that the
carbon faces downwards (see Figure
8.15).
Figure 8.15 Turning over of the grid.

Some liquid remains on the grid (about
3 µL). Excess liquid can be removed from
the grid side with a pipette.
1 to 2 µL of the crystal suspension is
added to the grid side and mixed with
the sugar by pipetting up and down
(see Figure 8.16).

Figure 8.16 Deposit of crystal solution on

the grid.
 The sugar forms a thin layer protecting
the extra-membrane part of the
proteins.
 As the image contrast is very low,
defocused diffraction mode can be
used to identify 2-D crystals.
4. The grid is put on two sheets of filter

paper to blot for 5 to 10 seconds and
then frozen by plunging into liquid
nitrogen.
Figure 8.17 Deposit of the grid on filter

paper.
 2-D crystals of c-ring proteins are
prepared according to this method (see
Figures 8.23 to 8.25).
4.5. Lipid Monolayer Crystallization

A crystallization plate is placed in a Petri 
dish that contains two layers of wet filter
paper (see Figure 8.18).
.
 Wfp = Wet filter papers

 Tcp = Teflon crystallization plate
 Pd = Petri dish

Figure 8.18 Crystallization device.
 Volume of the wells is about 50 to
60µL.
 Wet filter paper is used to saturate the
atmosphere with water vapor and
prevent dehydration.

1. Prepare the Teflon plate. 
 Wash the plate with detergent, rinse 
with distilled water. 
 Sonicate the plate in ethanol two times 
for 15 minutes.
 Rinse with hot water at least  Hot water (>60°C) is used for removing
30 minutes. traces of ethanol.

2. Fill wells with buffer. 

 Place spherical stirrers into wells. 
 Inject the buffer. 
 Remove air bubbles by stirring. 
 Adjust volume to obtain a plane  The surface of the well is observed with
surface. a horizontal light.

3. Deposit the lipid monolayer.  Take lipids from freezer at least 10
 The lipid mixture contained in minutes before depositing the monolayer.
chloroform/methanol solution should  Lipid mixture is at 0.5 mg/mL.
be at room temperature.  In case of His-tagged proteins, the ligand
 With a very thin syringe, deposit lipid is DOGS-NTA-Ni.
0.5 µL lipid mixture that contains the  Diluting lipids usually used are DOPC,
ligand and the diluting lipid in the DOPG and DOPS.
crystallization plate.  DOGS-NTA-Ni/DOPC ratio between 1:1
 Put the plate in the Petri dish, cover it and 1:5.
and incubate at 16°C.  Lipids must be sealed with dry nitrogen
to avoid lipid oxidation.

4. Protein injection.
 Adjust volume with distilled water to
obtain a perfectly plane surface.
 Remove the necessary volume through
the injection well.
 Inject the protein solution through the
injection well.
 Check the planarity of the surface of
the well.
5. Incubate at 16°C with very slow

stirring.
6. Grids.  If the crystals are extremely fragile, use

 Use copper grids (300 to 400 mesh) gold grids covered with a carbon film and
covered with a thin carbon film or a deposit the grid on the monolayer for at
carbon-collodion film. least two hours (see Figures 8.26 and 8.27).
 If the grid slightly turns when it is
 Glow-discharge the grids. deposited on the surface of the monolayer,
 Check the planarity of the lipid this means that protein interactions with the
surface. lipid monolayer are weak.
 Deposit the grid on the monolayer,  If there is no liquid at the surface of the
carbon film in contact with the lipids. grid that was in contact with the monolayer,
 Wait for two minutes. then it means that very few proteins have
interacted with lipids.

7. Freeze.
 Take the grid with tweezers dedicated
for freezing.
 Mount the tweezers on the freezing
device.
 Remove excess liquid with a filter
paper and quickly plunge the grid into
liquid ethane.
 Transfer the grid into a cryo-holder  see Chapter 7.
and observe the grid in a cryo-electron
microscope.







Figure 8.19 The cell culture plate.
 Use the basic 24-well cell culture
plates.

1. Apply vacuum grease around the well. 
2. Fill with buffer solution (~ 1 mL) (see

Figure 8.20).
 G = Grease
 B = Buffer
Figure 8.20 Grease is applied around the

well.
 The grease should not be too liquid.

3. Rinse the cover slide with 70%  Handle the cover slip with standard
ethanol and dry. tweezers.
4. Add the protein solution containing  Be careful of a charging effect that

precipitant agents (~ 7 to 8 µL). repulses the drop.
5. Deposit a carbon-coated gold grid (see

Figure 8.21).

Figure 8.21. The grid is deposited on the

drop.
 The carbon face is in contact with the
drop.
 Gold grids are used to avoid oxidation.
6. Carefully turn the cover slip over (see

Figure 8.22).
Figure 8.22 The cover slide is turned over.

 The drop must stay in the center of the
cover slip.
7. Put the cover slide on the well and  For H+-ATPase, 2-D crystallization
press to seal it with grease. occurs within two hours. If crystallization
needs more time, do not put the EM grid on
8. Incubate at 4°C. the drop immediately. The grid should be
placed on the drop two hours before
9. Remove the coverslip with the grid freezing.
facing up.
10. Freeze the grid.
 Handle the grid with tweezers

dedicated for freezing.
 Mount the tweezers on the freezing
device.
 Remove the excess liquid with a filter
paper and quickly plunge the grid into
liquid ethane.
 Transfer the grid into a cryo-holder
and observe the grid in a cryo-electron
microscope.

5.1. Using Natural Crystals

5.1.1. Advantages
 Membrane proteins are in their natural  Proteins have never been extracted from
environment. the membrane.
  Membrane proteins often need to interact
with specific lipids to be functional.

 Possibility of merging crystal patches  Very large crystals.
(see Protocol, Section 4.1, 2-D crys-  Suitable for electron diffraction.
tallization by membrane fusion of
purple membranes).
 Natural 2-D crystals usually form very  The quality of the crystal is poor.
small arrays

 Contaminant proteins can be included  This perturbs the periodic arrangement
in arrays. in the array.

5.2. 2-D Crystallization by Dialysis

5.2.1. Advantages
 General method for membrane pro-  Can be tried with any protein.
teins 

 Crystallization parameters can be  Very highly ordered crystals can be
accurately controlled. formed.
 Detergent removal is slow.  Influence of detergent in 2-D crystal

formation can be investigated accurately.
 Detergent removal is slow.  Not suitable for sensitive proteins

 Detergent removal is not efficient for  Detergent concentration in dialysis
detergent with low CMC. buffer quickly reaches the CMC.



5.3. 2-D Crystallization Using BioBeads

5.3.1. Advantages
 Very fast detergent removal.  Suitable for sensitive membrane pro-
teins.
 Possibility to quantify the amount of  Very reproducible.
detergent to remove.
 Small amounts of detergent can be  Allows accurate control of the
added if detergent removal is too fast. detergent/protein/lipid ratio.
 Lipids and proteins can be adsorbed  The hydrophobic surface of BioBeads
onto BioBeads. can bind lipids and proteins after a while.
 Accurate weighing of BioBeads is  Beads dry very quickly.
difficult.
5.4. Lipid Monolayer Crystallization

5.4.1. Advantages
 Screening of crystallization conditions  Formation of the first 2-D arrays occurs
is very fast. within 6 to 12 hours.
 Requires very small amounts of  Usually between 2 to 5 µg protein/well.
protein.
 Only one crystallization trial per well.  The lipid monolayer is picked up with an
EM grid.


5.5.1. Advantages
 A very good alternative to obtain  Proton-ATPase from Neurospora crassa
structural data when 3-D crystals are is a good example.5
not well ordered.

 No structural data better than 8 Å has
been obtained.
 The method is not very general.

6.1. 2-D Crystallization of Membrane Proteins

6.1.1. Natural 2-D crystals
 Should be used when possible.
 Membrane fusion can be tested with  To increase size of 2-D crystals.
different detergents.

6.1.2. Detergent removal by dialysis
 The process is time consuming, but  The most general method.
usually gives large crystals suitable for
electron crystallographic studies.

6.1.3. Detergent removal using BioBeads
 Detergent removal by BioBeads  Useful with proteins solubilized in
should be used with unstable proteins Triton X100 or dodecyl maltoside.
or proteins solubilized with a detergent
of low CMC.

6.2. 2-D Crystallization of Soluble Proteins
6.2.1. Lipid monolayer crystallization
 The lipid monolayer approach is ded-  Proteins can be used whatever their size.
icated to proteins with a His-tag.  Often used with soluble proteins that
aggregate at low concentrations.
 The ligand lipid for His-tagged proteins
is DOGS-NTA-Ni.
 Charged lipids can be used to bind  His-tag is not necessary.
proteins to the lipid monolayer by  The charge distribution does not have to
electrostatic interactions. be homogeneous.
 Positively charged lipid: DOTAP.
 Negatively charged lipid: DOPG, DOPE.
6.2.2. Surface crystallization

 Surface crystallization should be tried  This method can be used for both
when small 3-D crystals appear in the soluble and detergent solubilized membrane
drop. proteins.
6.3. Specimen Preparation

 Sugar embedding often works well for  To avoid basic problems occurring
membrane proteins without large during vitrification.
extra-membranous domains.
 The back-injection method yields very
flat crystals for high-resolution data
collection.

7. OBSERVED RESULTS
 Figure on the Chapter’s title page  Image of negatively stained 2-D crystals
of HC-Pro.9
 Figure 8.23 2-D crystal of ATP-

synthase c-ring formed by detergent
dialysis (see Protocol, Section 4.2),
negatively stained with 1% uranyl acetate
diluted in water. The crystal is in a closed
tubular vesicle. The inset shows the
crystalline structure (superposition of the
two sides of the vesicle).
Size of the inset = 300 nm




 Figure 8.24 Fourier transform of an

image of a similar crystal, cryoprotected in
4.5 % trehalose and prepared by the back-
injection method (see Protocol, Section
4.4).
Figure 8.25 Projection map of this crystal

calculated at 5 Å resolution. Image pro-
cessing was performed using the MRC
package.4
Bar = 2 nm





 Figure 8.26 2-D crystal of HC-Pro
obtained by the lipid monolayer
crystallization method (see Protocol,
Section 4.5). A carbon-coated copper grid
was applied for two minutes on the lipid
monolayer and negatively stained with 1%
uranyl acetate diluted in water. The 2-D
crystals are extremely fragile and have
fragmented due to surface tension.





Bar = 100 nm






 Figure 8.27 A gold grid was used to

prepare HC-Pro 2-D crystals. Two hours
incubation of the grid on the drop prevents
breakage of the 2-D crystals, which are now
suitable for structural investigation. Gold is
used to prevent grid oxidation during
incubation.9
Bar = 100 nm
 

8. REFERENCES
1. Booy, F.P. and Pawley, J.B. Cryo-crinkling: What happens to carbon films on
copper grids at low temperature, Ultramicroscopy, 48, 273, 1993.
2. Bron, P. et al. The 9 Å projection structure of cytochrome b6f complex determined
by electron crystallography, J. Mol. Biol., 287, 117, 1999.
3. Chiu, W., Avila-Sakar, A.J., and Schmid, M.F. Electron crystallography of
macromolecular periodic arrays on phospholipid monolayers, Adv. Biophys., 34,
161, 1997.
4. Crowther, R.A., Henderson, R., and Smith, J.M. MRC image processing
programs, J. Struct. Biol., 116, 9, 1996.
5. Cyrklaff, M. et al. 2-D structure of the Neurospora crassa plasma membrane
ATPase as determined by electron cryomicroscopy, Embo J., 14, 1854, 1995.
6. Gonen, T. et al. Lipid-protein interactions in double-layered two-dimensional
AQP0 crystals, Nature, 438, 633, 2005.
7. Kühlbrandt, W. Two-dimendional crystallization of membrane proteins: A
practical guide, in Membrane Protein Purification and Crystallization, Hunte, C.,
Von Jagow, G., and Schagger, H., eds., Academic Press, NY, USA, 2003.
8. Levy, D. et al. Two-dimensional crystallization on lipid layer: A successful
approach for membrane proteins, J. Struct. Biol., 127, 44, 1999.
9. Plisson, C. et al. Structural characterization of HC-Pro, a plant virus
multifunctional protein, J. Biol. Chem., 278, 23753, 2003.
10. Rigaud, J.L. et al. Bio-Beads: An efficient strategy for two-dimensional
crystallization of membrane proteins, J. Struct. Biol., 118, 226, 1997.
11. Vonck, J. Parameters affecting specimen flatness of two-dimensional crystals for
electron crystallography, Ultramicroscopy, 85, 123, 2000.
12. Walz, T. and Grigorieff, N. Electron crystallography of two-dimensional crystals
of membrane proteins, J. Struct. Biol., 121, 142, 1998.

Cryo-Negative Staining 221
CONTENTS
1. PRINCIPLES OF CRYO-NEGATIVE STAINING....................................... 223

3. MATERIALS AND PREPARATION ............................................................. 224
3.1. Principles .................................................................................................. 224
3.2. Useful Stains............................................................................................. 224
3.3. Materials/Small Items/Products/Solutions................................................ 225
3.3.1. Materials ..................................................................................... 225
3.3.2. Small items ................................................................................. 225
3.3.3. Products ...................................................................................... 225
3.3.4. Solutions ..................................................................................... 225
4. PROTOCOL ...................................................................................................... 226
4.1. Prepare the Staining Solution ................................................................... 226
4.2. Pipet the Sample on the Grid .................................................................... 226
4.3. Prestaining Step ........................................................................................ 227
4.4. Staining (Dialysis) .................................................................................... 228
4.5. Sample mounting...................................................................................... 228
4.6. Vitrification .............................................................................................. 229
4.7. Transfer..................................................................................................... 230
5.1. Advantages ............................................................................................... 231
5.2. Disadvantages........................................................................................... 232
6. WHY AND WHEN TO USE A SPECIFIC METHOD .................................. 233
8. REFERENCES .................................................................................................. 236



1. PRINCIPLES OF CRYO-NEGATIVE STAINING
The only additional step in cryo-negative  The staining solution used is ammonium
staining compared to thin-film vitrification molybdate (see below).
(see Chapter 3) is a short time of contact
between the staining solution and the
sample.
What happens to the sample during the  When prepared with this staining
short time of contact? The biological technique, the resulting density of the
sample is maintained in a low-salt, vitrified ammonium molybdate is higher
physiological buffer. During this step, the than the density of the protein (without
ammonium molybdate mixes with the taking into account hydration water).That is
sample buffer. The difference in salt why the biological samples appear white in
concentration of the two solutions a dark background in the electron
equilibrates so that the biological particles micrographs of a cryo-negatively stained
become suspended in the high-salt medium preparation.
without being diluted in the staining
droplet.
Figure 9.1 Schematic view of the steps in

sample preparation. (Reprinted from Adrian
et al. 19981 with permission from Micron.)
1. Preparing the staining solution

2. Pipetting the sample on the grid
3. Prestaining step
4. Staining (dialysis)
5. Sample mounting
6. Vitrification
7. Transfer

3. MATERIALS AND PREPARATION
3.1. Principles
In the case of cryo-negative staining, the
staining solution is used to increase the
contrast and decrease the electron beam  Contrast is measured in terms of signal-
sensitivity of partially hydrated biological to-noise ratio.
samples.
Unlike conventional air-drying negative  The final state of hydration of the

staining, the samples are still in solution, sample depends on blotting time with filter
where the aqueous buffer surrounding the paper.
biological particles is replaced by a higher  The time taken for blotting depends on
density, heavy metal salt. the relative humidity of the surrounding
sample environment.
The staining solution, composed of ca.
1.2 M salt (saturated solution of ammonium
molybdate), has a higher density than the
protein. This results in a contrast inversion
as observed by electron microscopy.
Figure 9.2 The chaperonin GroEL

visualized by cryo-EM. (Left: 1.4 m
defocus) as compared to cryo-negative
staining (right: 0.5 m defocus).
(Reprinted from De Carlo et al. 20023 with
permission from Elsevier.)
Bar = 50nm
Between the staining and the vitrification  The gain in contrast strongly depends
step, the sample IS NOT DRIED AT ALL. on the thickness of the vitreous ice layer
and on the distribution of the stain
surrounding the biological particles.

3.2. Useful Stains
Ammonium molybdate so far is the only
heavy metal salt that can be used in a
reproducible way to successfully achieve
cryo-negative staining.
Other typical heavy metal salts often used  In some cases, a contrast increase
in conventional air-drying negative obtained after mixing a 1 to 2% PTA
staining, such as uranyl acetate, solution with the sample, prior to
phosphotungstic acid (PTA), do not vitrification, has been reported in the
produce a marked increase in contrast. literature.2

When used at 1 to 2% (w/v), no significant  Another technique also known as the

contrast increase is observed. When used at “sandwich method” uses uranyl formate
a higher concentration, the salt precipitates (0.5 to 2%) in addition to glycerol (>10 to
and no useful results can be obtained. 20%) prior to plunging in liquid nitrogen.4
 Ammonium molybdate tetrahydrate  Can be purchased from different

(H24Mo7N6O24.4H2O; MM = 1235.86) companies.
 Has an acidic pH (4.5) if not buffered.
 The volumetric density of the final

staining solution, after buffering with
sodium hydroxide (NaOH,) is 2.35
g/cm3.

3.3. Materials/Small Items/Products/Solutions
3.3.1. Materials
 Balance.
 Magnetic mixer and heater.  Used to dissolve 40g of NaOH 1 M in
100mL water.
 Pyrex container (100 mL).  For storage of the NaOH 10 M solution.
3.3.2. Small items

 Bulbs, pasteur pipettes.  Does not need to be sterile.
 Electron microscopy forceps.  Needs to be cleaned before and after
usage.
 EM grid.  Holey or continuous carbon already
mounted.
 Filter paper.  No need of specific type.
 Glass/plastic Petri dish.  Does not need to be sterile.
 Gloves, mask.  Indispensable (NaOH is highly
concentrated and ammonium molybdate is
toxic).
 Micro-pipettes.
 Parafilm foil.
3.3.3. Products
 Ammonium molybdate tetrahydrate.  From Fluka, Buchs, Switzerland or
Sigma, Saint Louis, Missouri, USA.
 NaOH  1 M stock solution.
3.3.4. Solutions
 Water.  Do not use deionized water.

4. PROTOCOL
4.1. Prepare the Staining Solution

1. Weigh 1.0 to 1.2 g of ammonium 
molybdate tetrahydrate.  Fluka, Sigma
2. Add 0.9 mL water.
3. Shake.  It is a saturated solution, some powder

will remain and form a slurry.
4. Add 0.1 mL NaOH 10 M.
5. Shake.







Figure 9.3 Keep the ammonium molybdate
solution in a vertical position (slurry on the
bottom) and use only the supernatant.

4.2. Pipet the Sample on the Grid

1. The EM grid can be previously glow-
discharged if needed.
2. The carbon film can be continuous or

holey depending on your needs.
3. Avoid plastic (e.g., Formvar) on grids. 

If still present, dissolve plastic with 
ethyl-acetate/chloroform prior to  See Chapter 3.
experiment. 

4. Typically use 3 to 5 µL of sample on 

the EM grid, let the sample adsorb for 
a 10 to 30 seconds if it is a continuous 
carbon film.







Figure 9.4 Make sure the sample droplet is
evenly distributed on the grid.




4.3. Prestaining Step
1. Prepare a Petri dish (plastic or glass) 
with cover.
2. Put a square piece of Parafilm paper in

the Petri dish.
3. Deposit a 80 to 100 L droplet of the

staining solution on parafilm paper.








Figure 9.5 Use a new stain droplet each
time.





4.4. Staining (Dialysis)

1. Place the grid on the staining solution  Place sample face down, toward the
droplet. stain.










Figure 9.6. Sample facing the stain droplet.

2. Remove the grid after 10 to 60 
seconds, depending on the sample 
sensitivity to highly concentrated 
salts. Longer times of contact allow
better stain distribution.








Figure 9.7 Make sure the grid does not
sink.

4.5. Sample mounting

Mount the specimen on the plunge-freezing 
apparatus as fast as you can to avoid further
water evaporation, or use commercially
available units that allow environment-
controlled vitrification.  E.g., Vitrobot, FEI Company, see
Chapter 4.



















Figure 9.8 Tweezers mounted on the
plunger.


4.6. Vitrification

1. Apply filter paper to remove excess
liquid.








Figure 9.9 Apply the filter paper directly
on the grid (one-sided blotting shown
here).


2. Remove filter paper and wait 1 to 3

seconds before plunging the sample in
the cryogen.










Figure 9.10 Wait 1 to 2 seconds before
releasing the plunger.








Figure 9.11 Vitrification occurs as the grid
falls gradually into liquid ethane
previously cooled at liquid nitrogen
temperature.
4.7. Transfer
1. Transfer the grid from liquid ethane to

liquid nitrogen.
2. Make the transfer as fast as possible.

Avoid warming up the sample.
Devitrification occurs at 135°C.








Figure 9.12 Transfer the grid gently to the
grid box.

3. Transfer the grid to the appropriate 

storage box. 










Figure 9.13 Make sure the grid box lid is
tightly closed before storage.


4. Transfer the sample to your cryo- 
workstation and prepare to transfer to 
your usual specimen holder for cryo- 
microscopy (cold stage). 

 Cryo-negative staining is a practical

method that combines negative
staining and electron cryo-
microscopy.5,6 The biological samples
are imaged with a high signal-to-noise
ratio (contrast) provided by the
surrounding heavy metal while
maintaining the advantages of the
frozen-hydrated state (for a
comparison, see Chapter 3).
5.1. Advantages
 The signal-to-noise ratio (contrast,  This technique is very useful to increase
object visibility) is dramatically the signal-to-noise ratio (SNR) of small
increased. biological complexes that are hard to
 visualize in standard cryo-EM because of
their relatively small size.

 The sensitivity of the biological  It has been shown that samples

samples exposed to the electron beam preserved in saturated ammonium
is decreased. molybdate can withstand a three-fold
increase in the exposure to the electron
beam (as demonstrated by imaging GroEL
at 1.5 nm resolution with a total electron
dose of ~ 3600 electrons/nm2; (see Figure
9.16 from De Carlo et al. 2002 3).
Figure 9.14 Fourier shell correlation (FSC)

plots. The FSC are represented with the
following labels: LowD (low-dose, 1000 e-
/nm2), HighD (high-dose, 3000 e-/nm2),
water (unstained particles), and Stain (cryo-
negatively stained particles). The corrected
3 FSC criterion is also represented “FSC
Crit”). (Reprinted from De Carlo et al.
20023 with permission from Elsevier.)
5.2. Disadvantages
 Partial evaporation.  Removal of excess buffer (partial

dehydration) is necessary to obtain a thin
film prior to vitrification.
 Sensitivity to salt.  Some macromolecular assemblies do not

withstand the high salt concentration and
may fall apart.

As mentioned in this chapter, the main advantage of the technique is an increased signal-
to-noise ratio when compared to unstained cryo-electron microscopy. Many enzymes of
fundamental importance in cellular processes are in the range of 200 kDa, and this
represents the limit of molecular cryo-EM in terms of particle size amenable to image
processing. With an increased signal-to-noise ratio provided by the ammonium
molybdate stain, we were able to reconstruct the 3-D model of a human transcription
factor whose size is only 120 kDa7 provided the sample can stand 0.8 M salt due to the
ammonium molybdate and does not fall apart; this is a method of choice to get better
contrast in cryo-EM. Ammonium molydbate is the only useful stain with this technique.
If you need to use uranyl acetate or uranyl formate, still widely and successfully used in
many EM labs, the cryo-negative staining sandwich method developed by Golas and
colleagues5 will be very useful, especially if the sample is stored in 20 to 30% glycerol
and glycerol is needed for complex stability.

7. OBSERVED RESULTS
 Figure 9.15  Tomato Bushy Stunt Virus, as

observed unstained by cryo-EM (left) or by

cryo-negative staining in a saturated
solution of ammonium molybdate (right).4
Bar = 50nm
(Reprinted with permission from M.A.
Hayat. Principles and Techniques of
Electron Microscopy: Biological Applica-
tions. 4th ed., Cambridge University Press,
Cambridge, U.K., 2000.)




 Figure 9.16  Cryo-electron micrographs of GroEL
 (a–c) Unstained vitrified native GroEL

solution.
 (d–f) Cryo-negatively stained prepar-
ation with saturated ammonium
molybdate.
Bar = 50 nm
The total cumulative dose is

 1000 electrons/nm2 in (a) and (d)
 2000 electrons/nm2 in (b) and (e)
 3000 electrons/nm2 in (c) and (f).3

(Reprinted from De Carlo et al. 2002,3 with
permission from Elsevier.)


8. REFERENCES
1. Adrian, M. et al. Cryo-Negative Staining, Micron, 29, 145, 1998.

2. Böttcher, C. et al. Structure of influenza haemagglutinin at neutral and at fusogenic
pH by electron cryo-microscopy, FEBS Letters, 463, 255, 1999.
3. De Carlo, S. et al. Cryo-negative staining reduces electron-beam sensitivity of
vitrified biological particles, J. Struct. Biol., 138, 216, 2002.
4. De Carlo, S. Cryo-negative staining: Advantages and applications for three-
dimensional electron microscopy of biological macromolecules. PhD thesis,
University of Lausanne, 2002.
5. Golas, M.M. et al. Molecular architecture of the multiprotein splicing factor SF3b.
Science, 300, 980, 2003.
6. Harris, J.R. Negative Staining and Cryoelectron Microscopy. BIOS Scientific
Publishers Limited, Oxford, UK, 1997.
7. Jawhari, A. et al. Structure and oligomeric state of human transcription factor

TFIIE. EMBO Rep., 7, 500, 2006.

Vitrification of Dynamic Microtubules 239
CONTENTS

1. PRINCIPLE OF THE METHOD .................................................................... 242
3. MATERIALS/PRODUCTS/SAFETY PRECAUTIONS................................ 244
3.1. Materials ................................................................................................... 244
3.2. Products .................................................................................................... 248
3.3. Safety Precautions .................................................................................... 249
4. PROTOCOL ...................................................................................................... 250
5. ADVANTAGES/DISADVANTAGES/POSSIBLE IMPROVEMENTS ....... 253
5.1. Advantages ............................................................................................... 253
5.2. Disadvantages........................................................................................... 254
5.3. Possible Improvements............................................................................. 254
6.1. Self-Assembly Versus Nucleated Assembly ............................................ 255
6.1.1. Microtubule self-assembly.......................................................... 255
6.1.2. Centrosome nucleated assembly ................................................. 255
6.1.3. Axoneme nucleated assembly..................................................... 255
6.2. Incubation in Tubes or on Grid................................................................. 255
6.2.1. Incubation in tubes...................................................................... 255
6.2.2. Incubation on grid ....................................................................... 255
8. REFERENCES .................................................................................................. 258



Microtubules are dynamic polymers built

from the  tubulin heterodimer and
associated proteins collectively called
MAPs (for microtubule-associated
proteins). Tubulin, from most organisms, is
soluble at 4°C and assembles into
microtubules when the temperature is
raised (e.g., to 37°C) in the presence of
guanosine-5´-triphosphate (GTP). There-
fore, a preparation of frozen-hydrated
microtubules in their dynamic state requires
that the temperature and humidity of the
specimen be regulated before freezing in
liquid ethane.
In this chapter, we describe a simple

environmental device that has been used to
study the structure of microtubule ends
during assembly and disassembly.1-4 This
device can also be used with any kind of
specimen that necessitates regulation of its
environmental conditions.
 CT = Centrosome
 MT = Microtubule
Figure 10.1 Microtubules assembled from

purified tubulin and nucleated by an
isolated centrosome, observed by video-
enhanced differential interference contrast
light microscopy.3 Microtubules displaying
dynamic instability are either growing
(white arrow) or shrinking (black arrow).

1. PRINCIPLE OF THE METHOD
The aim is to vitrify dynamic microtubules  Without regulation of humidity, the

in a state as close as possible to the one they solution evaporates, which induces an
had in solution before freezing. A small increase in the overall salt concentration.
aliquot of the suspension is deposited on a
holey carbon grid held by tweezers fixed to  The temperature drops as a consequence
a guillotine device over a small cup filled of evaporation.5 This can be easily
with liquid ethane at ~ 180°C. Most of the monitored using a thin thermocouple
specimen is removed using a filter paper so immersed in the specimen while it sits on
that the remaining suspension forms a thin the electron microscope grid.
layer of about 50 to 100 nm thickness
spanning the holes of the carbon film. The  To limit the extent of these artifacts, it is
grid is rapidly plunged into liquid ethane by necessary to regulate both the temperature
releasing the mobile arm of the guillotine and humidity around the specimen before
device, and is transferred to the cryo-stage freezing.
of the electron microscope or stored in
liquid nitrogen for further use.  Note that warming only the air around
the specimen without saturating the
For most structural studies where one wants atmosphere with water vapor will result
to have the overall structure of a only in increased evaporation and, thus,
macromolecule, the steps described above increased cooling of the specimen.
may appear sufficient. However, when one
wants to relate structural features to
dynamic events, such as conformational
changes or a polymerization process,
additional care should be taken to preserve
the environmental conditions during
specimen preparation, because its exposure
to the ambient atmosphere may induce
several significant artifacts.

1. Equilibrate the environmental device

to the desired temperature.
2. Incubate the specimen on a grid or in

a test tube.
3. Blot the specimen.
4. Freeze in liquid ethane.
5. Store in liquid nitrogen or observe in

the electron microscope.
Figure 10.2 Summary of the different steps

to prepare frozen samples under controlled
temperature and humidity conditions.

3. MATERIALS/PRODUCTS/SAFETY PRECAUTIONS
3.1. Materials
 
 Chemical Hood  Necessary, to reduce airflow and
evacuate gaseous ethane and nitrogen if
required. No electrical or electronic
apparatus that may produce sparks should
be placed in this area. An explosion-proof
hood is preferred.
A = Two-liter water flask warmed by the

heating mantle
B = Temperature controller
C = Guillotine device
D = Tubing output near specimen grid
E = Polystyrene box with small metallic
cup
F = Ethane bottle
Figure 10.3 Open environmental device.
 Two-liter round-bottom flask with two  Fill 2/3 with distilled water.
or three necks (see Figure 10.4D).

 Heating mantle  Should be equipped with a separate
temperature controller located outside the
hood (see Figure 10.7).
A = Tubes coming from the air compres-

sors
B = External temperature controller
probe
C = Thermometer to record the temper-
ature inside the flask
D = Two-liter round-bottom bottle flask
E = Tube between the water flask and
the reservoir
Figure 10.4 Source of humid and warm air.

 Thermometer  Used to record the temperature inside the
 flask (see Figure 10.4C).

 Tubing  Two pieces ~ 80 cm and 15 cm in

length, and ~ 1.5 cm internal diameter,
connected by a Y-shaped connector (see
Figures 10.4E and 10.5B).
 Reservoir  Used to collect condensed water from
the tubes. An Erlenmeyer flask may be
suitable.
A = Tube coming from the two-liter

water flask
B = Tube going to the guillotine device
C = Erlenmeyer flask
Figure 10.5 Reservoir to collect condensed

water.
 

 Plastic tube  A 10 × 2 cm tube placed at the output of
 the tubing system. A 15 mL Falcon tube cut
at its base will work just as well. A tiny
hole can be drilled at its extremity to install
a thin thermocouple to monitor the airflow
temperature (see Figure 10.6B).
A = Thermocouple plunged into the 4 L
sample
B = Thermocouple that records the
output temperature
C = Output tube
 Figure 10.6 Device output.


 
 Compressed air  Be careful of oil contamination when
using oil-lubricated air compressors; use air
filters if necessary. Alternatively, use a
nitrogen bottle (it might empty quickly) or
small compressors used to inject air in
aquariums (we use two of them). A
pressure regulator might be needed to finely
regulate the airflow.
A = Air compressors
B = External temperature regulation of
the heating mantle
Figure 10.7 Compressors and heating

 mantle.

 Guillotine device  This can be purchased from various

companies or be homemade. We use a
model that has been designed and built at
the EMBL workshop in Heidelberg.
Modifications of the environmental device
presented here might be necessary if other
types of guillotines are used.
A = Mobile arm
B = Gallows
C = Tweezers
Figure 10.8 Guillotine device.
 
 Thin thermocouple  Used to record the temperature inside the
specimen droplet. A 0.25 mm diameter
chromel-alumel thermocouple can be used.
A second thermocouple can be fixed on the
plastic tube to record the output
temperature (see Figure 10.6B).
A = Tweezers
B = Thermocouple
C = Four microliter sample on holey
carbon-coated grid.

Figure 10.9 Sample temperature recording.
 Tweezers  Several pairs of tweezers can be used in

a single study. Electronic tweezers covered
with plastic except at their tip may be used
to reduce water condensation and to
facilitate manipulation at liquid nitrogen
temperature. We use short tweezers (A) for
the plunging and long tweezers (B) for
manipulating the grids inside liquid
nitrogen.

A = Tweezers used for plunging
B = Tweezers used for manipulating
specimen grids inside liquid nitrogen
Figure 10.10 Tweezers of electronic
quality.

 Grids  Use 100 to 400 mesh-size grids covered
with holey carbon films. Holey carbon grids
can be purchased or made according to
published protocols (see Chapters 3, 7 or
Chrétien et al.5).
Figure 10.11 Holey carbon grid square
observed by light microscopy.

 Filter paper  Whatmann n° 1 or n° 4 depending on the

viscosity of the suspension. Small
rectangular pieces can be prepared in
advance (~ 1 × 3 cm) and folded at a right
angle to facilitate blotting.
Figure 10.12 Filter paper.


 Cone tips  Must be cut at the tip if the microtubules
become very long, and prewarmed if
needed.

 Metallic cup  Used to store liquid ethane. Must be
cooled by liquid nitrogen. Solidification of
the ethane may arise at liquid nitrogen
temperature. This can be reversed by
touching the solid ethane with a metallic
object (e.g., a scalpel handle).
A = Polystyrene box covered inside with
aluminum foil and filled with liquid
nitrogen
B = Small metallic cup filled with liquid
ethane
 Figure 10.13 Polystyrene box and metallic
cup.

 Polystyrene boxes (2)  One used to cool the ethane cup (see
Figure 10.13A) and another to transfer the
specimen grid inside small boxes under
liquid nitrogen. Can be covered inside with
aluminum foil to reduce liquid nitrogen
bubbling.

 Small electron microscope boxes  These can be prepared from rectangular
boxes commercially available. A round
piece is cut so as to include four grid
positions. A small Plexiglas lid is designed
that covers three positions and a thread is
drilled at their center to screw the lid.
A = Storage box
B = Box (~ 1.3 cm in diameter)
C = Plastic screw
D = Plexiglas lid

Figure 10.14 Storage box.

 Liquid nitrogen tank  Used to store the grid boxes. 50 mL
 Falcon tubes can be used to store several
round boxes (drill a hole in the cap to allow
air evacuation during cooling).

3.2. Products
 Ethane To cryofix the samples. Always have a
fire extinguisher at hand.

 Liquid nitrogen  To liquefy ethane. Keep free of ice
contamination.
 Tubulin  Tubulin is commercially available or can

be isolated from a variety of organisms and
cell types. The most convenient source of
tubulin is mammalian brain (e.g., calf or pig
brain) because this organ is rich in
microtubules that direct vesicular fluxes
within axons and dendrites. Purification of
brain tubulin relies on its ability to
polymerize into microtubules at 37°C and
to be soluble at 4°C. Several rounds of
polymerization at 37°C and depoly-
merization at 4°C followed by ultra-
centrifugation give rise to a fraction
enriched in tubulin and MAPs (2X or 3X
tubulin depending on the number of cycles).
Tubulin is globally negatively charged near
neutral pH and can be further separated
from positively charged MAPs by ion-
exchange (phosphocellulose) chroma-
tography.6
 Figure 10.15 Tubulin molecule.7
 Axonemes  Several protocols can be found in the

literature that allow purification of
axonemes from various organisms and cell
types, such as seaurchin sperm8 or the
protozoa Chlamydomonas. In contrast to
centrosomes, large quantities of axonemes
can be easily obtained at a high
concentration. The protocol involves
demembranation of the axonemes and
removal of the molecular motors (dynein)
that induce flagella or ciliary beating.
Figure 10.16 Axoneme observed by cryo-

EM (Chrétien D., unpublished observa-
tions).



 Centrosomes  Centrosomes can be purified from

different cell types, but a concentration of
~ 1.108/mL after dilution with the tubulin
sample should be reached to facilitate their
visualization by cryo-EM. A protocol for
the purification of centrosomes from the
KE-37 lymphoblastic cell line has been
described that gives rise to this
concentration range.9 The KE-37 cells grow
in suspension and display the special
feature of a high ratio of nuclear to
cytoplasmic volume, which provides
centrosome-enriched cytosolic fractions
after cell lysis. Centrosomes are further
purified on a discontinuous sucrose
gradient, but sucrose does not seem to
influence the assembly properties of the
microtubules nor does it cause major
problems with their visualization by cryo-
EM.
Figure 10.17 Centrosome observed by

 cryo-EM. (Reprinted with permission from
Elsevier.9)

3.3. Safety Precautions
 Ethane  Avoid placing any electrical apparatus
inside the hood. A 5 mL cup of liquid
ethane represents a volume of ~2 L gaseous
ethane at room temperature, and may
produce a large explosion. A fire
extinguisher must be placed close to the
hood where the guillotine device is
installed.

 Liquid nitrogen  Follow all safety instructions when
handling LN2. In particular, wear goggles
and trousers during specimen preparation.

 Tubulin  Tubulin is isolated from mammalian
brain, such as cow, so the risk of prions
being present in such samples cannot be
totally neglected. Be careful to avoid
piercing your fingers with tweezers that
have been in contact with the tubulin
preparations.

4. PROTOCOL
1. Environmental device preparation  Small variations in temperature (+/-1°C)

can be recorded. For tubulin, it is safer to
 Let the device equilibrate before work at an average temperature of 35°C to
starting an experiment avoid denaturation.
– 30 min The water flask must be heated above
 Check that the temperature inside the this temperature (e.g., 45°C) because
droplet (on a blank specimen) and cooling of the air will arise during its travel
beside it are identical. into the tubing system.
2. Preparation of self-assembled  Microtubule self-assembly requires two

microtubules main steps: Nucleation and elongation.
 Prepare tubulin at a concentration Nucleation is an energetically

above ~ 20 M in the presence of unfavorable step and requires a critical
GTP. concentration (~ 20 M) that must be
– 10 min at 4°C reached before elongation can proceed.

 Incubate in a test tube or directly on  Various buffers can be used to assemble
the grid (see below, Steps 5 and 5’). tubulin, but they share some common
– Variable time at 35 °C features. Microtubule assembly is efficient
at pHs just below neutrality (6.8 ~ 6.9).
Buffers below or above this pH range may
induce the formation of other polymeric
forms (sheets, ribbons, double-walled
microtubules). Magnesium ions and GTP
should be present unless other
polymerization-promoting agents are added
to the reaction mixture; such as the
antimitotic drug Taxol®. Calcium ions are
very efficient at depolymerizing micro-
tubules and ethylene glycol tetra-acidic acid
(EGTA) should always be present in the
assembly buffer (note that calcium can be
released from the filter papers used for
blotting the specimens). Some compounds
such as glycerol can also be used to
facilitate microtubule assembly, but these
can be difficult to use with cryo-EM of
vitrified samples because their density may
match that of proteins and, hence, decrease
the contrast present in the images.
3. Preparation of centrosome-nucleated  Centrosomes nucleate microtubules at
microtubules ~ 5 M tubulin. Most microtubules will
 Mix centrosome and tubulin solutions start to assemble synchronously, which
to get a final centrosome concentration facilitates observation of their extremities
of ~ 1.108/mL. by cryo-EM.

– 10 min at 4°C  The elongation rate is concentration

dependent (~ 0.2 mM-1min-1) and, thus,
this parameter can be adjusted to control
the length of the microtubules at a given
assembly time.
 Incubate in a test tube or directly on  Same protocol as with axonemes.
the grid (see below, Steps 5 and 5’).
– Variable time at 35°C
4. Holey carbon grids  See Chapters 3, 7 and Chrétien et al.5

 Choose grids with large holes  Grids with mesh sizes of 200 or 300 may
(~ 10 m) to observe microtubule be preferred.
asters.  This avoids droplet spreading on both
 Glow discharge lightly for a short time sides of the grid, which can be critical if
(light pink discharge). long incubation times are required.
– 10 to 30 sec 
5a. Incubation in tubes  Cut off cone tips to avoid breakage of

 Pipette a 4 L specimen and apply on the microtubules if they become too long.
the grid held by the guillotine device.  Preincubate cone tips in buffer at the
 desired temperature.
  The temperature inside the specimen can
be rapidly checked using the small

thermocouple if desired.
  Microtubule suspensions can become
 Blot and release the mobile arm of the very viscous if one works at high tubulin
guillotine device. concentrations or in the presence of
polymerization-promoting factors, such as
MAPs. A rapid filter paper, such as
Whatmann n° 4 (or equivalent), might be
needed to blot the specimen.

5b. Incubation on the grid  With some practice, time periods as
 Same steps as above. short as a few seconds after the beginning
of incubation can be reached.
 For diluted suspensions, such as
centrosome-tubulin mixtures, a slow-rate
filter paper (e.g., Whatmann n° 1) can be
used.
 In practice, time periods up to 3 min of
incubation at 35°C can be performed on the
grid. Beyond this time period, water tends
to condense onto the tip of the tweezers and
to fall inside the specimen, resulting in its
dilution. To overcome this effect, it may be
necessary to also warm the tweezers (see
Section 5.3).

5. Specimen storage  Studying a polymerization process

 Transfer specimen grids into small involves preparation of several grids at
round boxes (see Figure 10.14). different assembly times, and eventually in
different assembly conditions (e.g.,
different tubulin concentrations).
Hence, it is convenient to store the
specimen grids in a liquid nitrogen tank
until their observation by cryo-EM.
 Care should be taken to monitor the
liquid nitrogen level from time to time and
to avoid water contamination of the
nitrogen that may induce crystal formation
on the grid.

6. Electron microscope observation  Under these high defocus conditions, the
under low dose conditions microtubules are easily visualized and the
quality of the ice can be monitored
 Search specimen at low magnification (presence of ice contaminants, cubic or
(2000 to 3000 ×), with the beam hexagonal ice, ethane contamination (see
spread as much as possible and under Chapter 7).
high defocus (10 to 15 m). 
 Purified centrosomes without
microtubules or after short assembly times
may be difficult to find because they tend to
localize at the edges of the carbon holes
where the ice layer is thicker. This becomes
less critical after longer times of assembly
because the microtubules help to localize
the centrosomes inside the holes. Self-
assembled microtubules tend to span the
holes of the carbon, and visualization of
 their extremities is more difficult.
 Blank the electron beam from the  For those interested in fine structural
specimen. details (e.g., for three-dimensional
 Take pictures. reconstructions), a magnification of
~ 60,000 × may be required.
However, for those who wish to take

pictures of the whole microtubule asters
with a chance of visualizing several ends in
the same image, magnifications as low as
20,000 and ~ 1.5 m defocus can be used.



5. ADVANTAGES/DISADVANTAGES/POSSIBLE
IMPROVEMENTS
5.1. Advantages
 Open design  More sophisticated devices than the one

 presented here can be used to control the
environmental conditions around the
specimen when it sits on the electron
microscope grid. However, these devices
are closed and require an automatic system
to blot the specimen (see Chapter 4). These
are certainly very well suited for specimens
of constant viscosity, but seem to us more
difficult to use with a polymer such as
microtubules whose viscosity changes with
time and assembly conditions. We find that
it is easier to blot the specimen manually,
which provides — with a little practice — a
good feeling of the viscosity of the
suspension.

 Modularity  Can be easily modified to suit one’s
needs. We have used this device to observe
microtubules polymerized at 35°C,3 and
also near room temperature (~ 23°C) with a
very light heating of the water balloon.1
Under these conditions, the device avoids
evaporation and cooling of the specimen.
We found that formation of the thin layer of
suspension was facilitated and more
reproducible, probably because it remains
in equilibrium with the humid atmosphere
for a longer time than when exposed to
ambient atmosphere. We have also used a
similar device, but with a flow of cold air
generated by a reservoir installed into a
cooler.4 Addition of these two apparatuses
may be used to perform fast temperature
jumps.

 Economical

5.2. Disadvantages
 Water condensation on the tweezers.  The setup presented here can only be
used for short reaction times (i.e., up to
three minutes on the grid) because water
tends to condense on the tweezers and to
dilute the specimen.
 Specimen exposed to ambient  The specimen remains exposed to the

atmosphere during plunging. ambient atmosphere during its plunge into
liquid nitrogen. Although this is a very fast
process (a few tenths of a millisecond), we
cannot exclude some molecular
rearrangements before it is cryofixed.

 However, comparison of the structure
of the ends of growing microtubules with
those of cold-induced depolymerized ones
revealed that this device is well suited for
visualizing slow conformational changes
such as those involved in the assembly and
disassembly of microtubules.

5.3. Possible Improvements
 Regulation of the tweezers’ tem-  It may be advantageous to design a

perature. system that warms the tweezers at the same
temperature as that of the specimen to
avoid water condensation.

 Thermocouple.  It would be convenient to have special
tweezers whose tips would act as
thermocouples to monitor the temperature
of the specimen when it sits on the grid as
well as its temperature history during
plunging into liquid ethane.

 Automatic regulation of the tem-  We have tried to regulate the
perature. temperature of the heating mantle using a
temperature controller linked to the
thermocouple that records the droplet
temperature, but the inertia of the system is
such that this induced too large a
fluctuation of the output temperature.

6.1. Self-Assembly Versus Nucleated Assembly

6.1.1. Microtubule self-assembly
  Assembling microtubules in a test tube
is easy as long as the tubulin concentration
is above ~ 20 M (~ 2 mg/mL).

Ends may be difficult to observe, and
the polarity of the microtubules must be
determined to differentiate plus and minus
ends (see Chrétien et al.10).

6.1.2. Centrosome nucleated assembly
  All microtubules start to assemble at the
same time.
  Only one type of end is observed,
because the minus end is anchored at the
centrosome, while the plus end is growing
away from it.
  Centrosome preparation necessitates
two to three weeks of cell culture followed
by one day of purification.

6.1.3. Axoneme nucleated assembly
  Large quantities can be obtained in a
single day from sea urchin sperm.

  Although the plus end grows more
rapidly than the minus end, differentiating
them in the cryo-EM images may be
difficult.

6.2. Incubation in Tubes or on Grid
6.2.1. Incubation in tubes
  Easy to perform.
  Microtubule suspensions can become
very viscous (use fast-rate filter paper).
6.2.2. Incubation on grid
  Gives access to the initial reaction
times.

7. OBSERVED RESULTS


Figure 10.18 Microtubules nucleated by a

centrosome in the presence of purified
tubulin. Note the presence of the long
outwardly curved tubulin sheets at the ends
of growing microtubules (see Chrétien et
al.3).
 
 





Figure 10.19 Microtubules nucleated by an
axoneme in the presence of purified tubulin
(Chrétien, D., unpublished observations).
The arrows indicate the transition between
the axoneme (bottom left) and the
microtubules nucleated from it (upper
right).
 





Figure 10.20 Microtubules polymerized in
a cell-free extract from Xenopus eggs
(Chrétien et al.5). Filaments, vesicles and
ribosomes can also be visualized.
(Reprinted from Wade and Chrétien11 with
permission.)







Figure 10.21 Microtubules assembled in

the presence of the slowly hydrolysable
GTP analogue GMPCPP (guanylyl-()-
methylene-diphosphonate), and depolymer-
ized in the presence of calcium (see Müller-
Reichert et al.4 for details.)




Figure 10.22 Self-assembly of micro-

tubules in the presence of CLIP-170. Note
the presence of curved tubulin-CLIP-170
oligomers in the background and at the
extremity of the growing microtubules.
(Reprinted from Arnal et al.2 with
permission.)


8. REFERENCES
1. Arnal, I. et al. Structural transitions at microtubule ends correlate with their dynamic
properties in Xenopus egg extracts, J. Cell Biol., 149, 767, 2000.
2. Arnal, I. et al. CLIP-170/tubulin-curved oligomers coassemble at microtubule ends
and promote rescues. Curr. Biol., 14, 2086, 2004.
3. Chrétien, D. et al. Structure of microtubule ends: Two-dimensional sheets close into
tubes at variable rates. J. Cell Biol., 129, 1311, 1995.
4. Müller-Reichert, T. et al. Structural changes at microtubule ends accompanying
GTP-hydrolysis: Information from a slowly hydrolyzable analogue of GTP,
GMPCPP. Proc. Nat. Acad. Sci., 95, 3661, 1998.
5. Chrétien, D. et al. Lattice defects in microtubules: Protofilament numbers vary
within individual microtubules. J. Cell Biol., 117, 1031, 1992.
6. Ashford, A.J. et al. Preparation of tubulin from bovine brain, in Cell Biology: A
Laboratory Handbook, 2nd ed., Celis, J.S., eds., Academic Press, San Diego, CA,
USA, 2, 205, 1998.
7. Nogales, E. et al. Structure of the alpha beta tubulin dimer by electron
crystallography. Nature, 391, 199, 1998.
8. Detrich, H.W. et al. Purification, characterization, and assembly properties of
tubulin from unfertilized eggs of the sea urchin Strongylocentrotus purpuratus.
Biochem., 22, 2453, 1983.
9. Chrétien, D. et al. Reconstruction of the centrosome cycle from cryo-electron
micrographs. J. Struct. Biol., 120, 117, 1997.
10. Chrétien, D. et al. Determination of microtubule polarity by cryo-electron
microscopy. Structure, 4, 1031, 1996.
11. Wade, R.H. and D. Chrétien. Cryoelectron microscopy of microtubules. J. Struct.
Biol., 110, 1, 1993.

CEMOVIS, Cryo-Electron Microscopy of Vitreous Sections 261
CONTENTS

1. PRINCIPLES OF CEMOVIS........................................................................... 264
1.1. Cryo-Electron Microscopy ....................................................................... 264
1.2. Vitrification .............................................................................................. 264
1.2.1. General........................................................................................ 264
1.2.2. Cryoprotection ............................................................................ 265
1.2.3. High pressure .............................................................................. 266
1.3. Cryo-Sectioning........................................................................................ 267
1.3.1. Cutting dry .................................................................................. 267
1.3.2. Cutting vitreous material............................................................. 268
1.4. Observing in the Cryo-Electron Microscope ............................................ 272
1.4.1. Beam damage.............................................................................. 272
1.4.2. Focus........................................................................................... 274
1.4.3. Volume and not surface .............................................................. 274
3. INSTRUMENTS AND INSTALLATION/MATERIALS/CHEMICALS .... 275
3.1. Instruments and Installation...................................................................... 275
3.2. Materials ................................................................................................... 276
3.3. Chemicals ................................................................................................. 276
4. PROTOCOLS .................................................................................................... 277
4.1. Preparation of the Material ....................................................................... 277
4.1.1. Cell suspension ........................................................................... 277
4.1.2. Tissues ........................................................................................ 278
4.2. Vitrification .............................................................................................. 278
4.2.1. Cell suspension vitrified in a Leica EM-PACT high-pressure
freezer ......................................................................................... 278
4.2.2. Organotypic slice of rat brain in high-pressure freezer HPM 10
from BAL-TEC........................................................................... 279
4.3. Cryo-Sectioning........................................................................................ 279
4.3.1. Mounting the specimen in the cryo-ultramicrotome ................... 280
4.3.2. Trimming the specimen .............................................................. 280
4.3.3. Cutting sections........................................................................... 281
4.3.4. Flattening the section on the grid................................................ 282
4.4. Observing in the Cryo-Electron Microscope ............................................ 282
4.4.1. Transfer....................................................................................... 282
4.4.2. Evaluation of the specimen ......................................................... 283
4.4.3. Observation and image recording ............................................... 283

5.1. Advantages............................................................................................... 284
5.2. Disadvantages .......................................................................................... 284
6. WHY AND WHEN TO USE CEMOVIS ........................................................ 285
8. REFERENCES .................................................................................................. 288


The ultimate method… … with restrictions.

Can you imagine any better preparation  Water can only be vitrified in layers in
method than CEMOVIS ― cryo-electron the µm range (see Chapter 1).
microscopy of vitreous sections?  The vitrification depth is greater for
biological material, but it does not exceed
the 10 µm range for most cells and tissues.
Take a sample, cool it so rapidly that it is  High-pressure vitrification makes it

immobilised before it has time to possible to increase the depth of
crystallise, cut it into ultrathin sections and vitrification 10-fold, but it does not allow
observe, in a cryo-electron microscope, all vitrification of pure water or a diluted
the finest molecular details of the perfectly solution (physiological solution).
preserved specimen. You may even  In general, the extracellular medium
reconstruct the 3-D distribution of the must be cryoprotected. 20% dextran
material in the section by means of ensures that the entire sample has approx-
computerised electron tomography (CET) imately uniform vitrification properties.
(see Chapter 12). In addition, the time span  After vitrification has been achieved, the
from the sample to the image could be as temperature must remain below –135°C for
short as one hour rather than days. Yes, in all subsequent steps.
principle, you can’t dream of any thing
better than CEMOVIS.
The dream of CEMOVIS is old. Fernández-  Vitreous material is a liquid of high

Morán12 initiated the effort to make it come viscosity. Cutting-induced deformation is
true. Vitrification (see Chapter 1) and high- inherent to vitreous specimens. It can only
pressure freezing (see Chapters 5, 6) were be reduced by cutting conditions that
important milestones on the way toward its increase fractures, thus favouring crevasses.
realisation, which, it can be argued, is now Cutting artefacts are a serious problem.
successfully achieved.3,4 It remains to make  The quality of the knife is a critical
it generally available. This book may help. factor.
 It is not a trivial matter manipulating a
sample below 140°C without excessive
frost contamination.
Maybe, one day, CEMOVIS will replace  Beam damage is a critical factor.
conventional embedding and room Whereas parsimonious electron irradiation
temperature observation. For the moment, produces only marginal image improve-
however, it requires skill and patience. ment in conventional observations, an
Consequently, the domain of application of excessive electron dose rapidly causes
CEMOVIS is where conventional methods bubbling and total destruction of vitreous
fail, namely at high resolution. specimens. Even at lower doses, a specimen
undergoes significant modification.

The question is when to choose

CEMOVIS? Of course, this depends on the
nature of the observed structure ― rigid
bodies are better preserved by conventional
preparation methods than liquid structures.
In general, we suggest that CEMOVIS
should be considered for any observation in
the 10 nm range or below. The reader may
consult other general or recent articles.7-
11,14,16,19,21
1. PRINCIPLES OF CEMOVIS
1.1. Cryo-Electron Microscopy

Electrons only travel over micrometres in  The mean free path ― namely the
biological material and centimetres in air. distance between two scattering events in
Consequently, at room temperature, the biological matter ― is ca. 0.2 µm for
electron microscope column must be under 100 kV electrons. The thickness of a
high vacuum and water must be eliminated specimen should preferentially be much
from the specimen because it evaporates at smaller.
low pressure. Water can only be conserved  The evaporation rate is less than
in vacuum at a temperature where the 1 nm/sec at 100°C. It decreases 10-fold
evaporation rate becomes negligible. each time the temperature is reduced 10°C
Modern cryo-electron microscopes make thus leading to a negligible evaporation rate
such observations possible without loss of of 1 nm/hour at 135°C. It is negligible at
image quality and with minimum 196°C, the temperature of liquid nitrogen.
operational burden.
1.2. Vitrification
1.2.1. General
Under normal conditions, water freezes  Hexagonal crystals are always very large
upon cooling and ice formation is at the ultrastructural scale. In most cases,
associated with large-scale water they are only one or few per cell. However,
reorganisation. The induced rearrangement under favourable freezing conditions, an ice
of the biological material may be limited by crystal is not like a compact ice block, but it
careful freezing conditions, but it can never is ramified like a tree with its branches.
be completely avoided.

Cryo-EM became a fashionable method  Vitrification is not crystallisation with

when vitrification of water and diluted very small ice crystals. The vitreous and
aqueous solutions were found to be crystalline states are fundamentally
possible. different, nonoverlapping states. Electron
Vitrification ― namely making water solid diffraction unmistakably shows the
while keeping its amorphous state and difference.
avoiding any ice formation ― can be
achieved under some stringent cooling
conditions.
Three parameters are essential for  The cooling rate for vitrifying pure water
vitrification: cooling rate, cryoprotection, is very uncertain. It lies probably in the
and pressure. Vitrification of pure water at range from 106 to 1011°C/sec. For typical
normal pressure is only obtained with an cells and tissues cooled under high-
extremely high cooling rate. A slower pressure, the required cooling rate is
cooling rate is sufficient when a fraction of probably only in the 103 to 104°C/sec range.
the water molecules is mobilised by a In such a case, the immobilisation time can
cryoprotectant or when cooling is done be estimated to be in the 10 msec range.
under high pressure.
For pure water, vitrification can be  Thin film vitrification (see Chapters 3, 4,
achieved on a thin layer of up to ca. 1 µm 7) is very efficient because there is a direct
thickness. contact between the material to be vitrified
and the liquid cryogen.
The vitreous state is metastable. It can only  Pure vitreous water devitrifies within
exist at low temperatures. Ice forms when minutes at 135°C. Any operation with
the vitreous state is rewarmed. vitreous material must be performed below
this temperature.
1.2.2. Cryoprotection
The tendency of water to form ice decreases  The search for optimal cryoprotectants is
with solute concentration. It can even be a major goal of cryobiology, which has
practically suppressed with some solutes at stimulated considerable research during the
high concentration (2.4 M sucrose, gly- past 70 years. Much remains to be done,
cerol). In general, most solutes act as however, for the practice of CEMOVIS.
cryoprotectants. This also holds for the
material in a living cell.
The vitrification depth that can be reached  A vitrification depth in the 10 µm range
for typical cells and tissues by plunge- is not convenient for most eukaryotic cells
freezing or slam-freezing (see Chapter 2) is and tissues because it hardly corresponds to
typically in the 10 µm range. one cell layer. Good observation conditions
are thus limited to the surface layer that, in
many cases, is the region damaged during
excision.

Cryoprotection also increases the  This observation implies that freeze-

devitrification temperature of vitrified substitution, which takes place at 90°C,
samples. However, according to Lepault et never occurs in vitreous specimens, but
al.15 and our own experience, the always in samples where water is in a
devitrification temperature of typical crystalline form. It remains to be seen if
biological samples remains well below cubic ice is frequently the form of water
100°C. that freeze-substitution is dealing with.
In most tissues or cell suspensions, the  Experience has shown that a ca. 20%
extracellular medium is a poor sugar solution mimics the vitrification
cryoprotectant. It, therefore, has a tendency properties of typical cells. This condition
to crystallise even when vitrification could causes severe osmotic dehydration of the
be achieved in the cell. This is an cell. We have found that the high molecular
unacceptable situation because ice for- weight sugar-polymer dextran is a
mation in the surrounding of a cell causes convenient sugar substitute with a
major osmotic injuries. Therefore, it is negligible osmotic effect. The exact
important to add a cryoprotectant to the conditions for optimal cryoprotection must
extracellular medium in order to obtain be adjusted for every specimen.
homogeneous vitrification conditions.
1.2.3. High pressure

The density of hexagonal ice is ca.  2000 atmosphere may seem a very high
0.92 g/cm3 whereas it is 1 for pure water. In pressure. However, because the com-
other words, the volume of water increases pressibility of water is very low, the
upon freezing. Consequently, freezing can enthalpy change in the biological sample is
be retarded if high pressure is applied small unless some gas bubbles are left.
during cooling. The equilibrium melting  Using an even higher pressure is
temperature of ice can be reduced to as low detrimental because it favours the formation
as 22°C at 2000 bars. of a new form of ice with higher density.
Practice supports the rule of thumb that  There are no quantitative experimental
high pressure increases the vitrification data available on how the cooling speed
depth achievable in cells and tissues by a required for vitrification decreases with
factor of 10 from the 10 to the 100 µm increasing pressure. The vitrification depth
range. We frequently work with 300 µm- at normal and high pressure could,
thick specimens. however, be determined for a number of
comparable systems.
High-pressure freezing is performed in  In both cases, a conductive metal layer

commercial apparatus working according to of 100 to 200 µm separates the sample from
two different principles. In one case, the the direct action of the cryogen. The
well-protected sample is placed in a maximum cooling speed that can be
chamber where the cryogen is injected at reached at the surface of the sample is
very high pressure (see Chapter 5). In the always much smaller than what can be
other case, the sample is mounted in a reached by direct cooling. As a
small, protected volume (a capillary copper consequence, we have never been able to
tube or another sandwiched volume). High vitrify pure water or diluted aqueous

pressure is applied in the volume at the very solutions by high-pressure freezing. In our
instant it is cooled from outside with a experience, rare are the biological
cryogen jet at low (some bars) pressure (see specimens that can be vitrified with less
Chapter 6). than 10% added external cryoprotection.
We believe that claims of vitrification of
much more diluted aqueous solutions that
were previously published or announced
orally were erroneous.
Slam-freezing (see Chapter 2) is an  From the principle of the method, it can
interesting method, which should be studied be inferred that the cooling speed at the
more systematically. It does not provide the surface of the sample must be comparable
advantage of high pressure, but it enables to that obtained by plunge-freezing.
direct contact of the cryogen with the Consequently, one could guess that it would
sample; there is no intermediate protective be possible to vitrify, at least the surface, of
layer between the sample and the cooling water or a diluted solution layer. This has
material. Furthermore, it does not have the not yet been shown to be the case. The
disadvantage of high-pressure freezing, reason could be the retroactive effect of the
namely the effect of very high pressure heat generated by ice formation deeper in
applied during the cooling process. the sample.
1.3. Cryo-Sectioning
Any user of a sophisticated technology tends to forget the instrument that makes his work
possible. Electron microscopists depend on the electron microscope and the various
instruments, such as ultramicrotomes used for specimen preparation. Cryo-electron
microscopy became possible with the development of reliable electron microscopes
capable of dealing with low temperature specimens. CEMOVIS was made possible thanks
to the remarkable developments of cryo-ultramicrotomes and diamond knives.17
1.3.1. Cutting dry

A good modern cryo-ultramicrotome
provides an adequate working environment
for cutting vitreous sections. The
temperature is well controlled in the
working range of 140°C to 170°C. The
gas flow in the working area is strong
enough to repel frost, though it is smooth
enough not to disturb sections. The working
place is spacious, well illuminated and
observable with high-quality binoculars. In
addition, the cryo-ultramicrotome has kept
all the cutting properties, which makes it so
good for cutting conventional sections at
room temperature.

Cutting conventional plastic sections is  The required liquid probably does not
eased by the fact that it is possible to float exist because it must have mutually
them on water. A liquid performing the exclusive properties. It must remain a liquid
same function with vitreous sections has not at low temperature, i.e., have small
yet been found. cohesive energy and, at the same time, have
a high surface tension in order to force the
section to float and prevent the liquid from
wetting the specimen.
As a consequence, in addition to being very  The question of whether diamond or

sharp, a knife for cutting vitreous sections glass knives are better has not yet been
must offer a surface on which the section studied in detail, but for the moment the
can glide smoothly. A knife with a small best results have been obtained with
cutting angle is preferable as long as it does diamond knives.
not compromise the sharpness of the cutting For our purposes, and for the moment, the
edge. diamond knives of only one manufacturer
are adequate for producing good vitreous
sections (see Section 3.1.). The critical
point is how the section glides on the knife
surface.
Sections cut in the dry state are very  A device providing an adjustable ion
sensitive to electrostatic charges, which shower in the vicinity of the knife is an
make them fly away or stick to any surface. important aid for controlling the sections.
No thin section of good quality (feed of
50 nm) can be obtained using a diamond
knife without an ion shower.
1.3.2. Cutting vitreous material

Vitreous water is a liquid, but a liquid with  The basic features of the sectioning
very high viscosity. This means that, like process are probably quite well understood.
honey or butter, a section flows. It flows The consequences of the cutting process are
slowly if the acting force is small as is the so diverse and so remarkable that they
case during observation in the EM and it provide rich possibilities for the
flows rapidly when the force is large as is falsification of any of the theories that have
the case during cutting. Deformation, is been proposed. The current theory has
therefore, an inherent feature of vitreous undergone so many tests that it is probably
sections. globally correct, though it obviously needs
refinement.5

 The knife K hits the block at depth f

equal to the feed. It exerts a pressure N
perpendicular to the knife surface (if the
friction along the surface is negligible).
N can be decomposed into the compression
force C along the cutting direction and the
bending force B. The cutting energy is
concentrated in the grey region where most
of the deformation takes place,
transforming the advance ∆x in the block
into the displacement ∆x´ of the section on
the knife surface. The shear stress increases
along N from S to Q where there is no
material any more to counterbalance the
effect of the cutting force. The stress may
exceed the fracture limit of the material,
thus producing crevasses (dash line).
 A rich collection of cutting artefacts is
shown in Figure 11.14 A.
Figure 11.1 Forces and deformations

during cutting.
The enormous force density that is locally  All that glitters is not gold. The fact that
applied during cutting is amazingly a section is vitreous does not prove that the
illustrated by the effect of cutting-induced block was vitrified. Material that has
amorphous water (CIA) vitrification.1 It has undergone CIA, however, has very different
been shown that, under some uncommon cutting properties; in particular, it is more
conditions, crystalline ice is transformed
brittle. An experienced CEMOVISt will
into a form of amorphous water during
easily identify CIA by the unusual
cutting. This effect is known to take place
under a pressure of several thousand bars. mechanical properties of the material and
the fact that, in spite of vitrification, the
biological material has suffered from
crystallisation-induced segregation. In our
experience, CIA is not observed in thick
sections (feed ≥100 nm). Therefore, in case
of doubt, vitrification should be tested on
such sections.
The basic deformation that takes place  The compression factor C (length
during cutting is compression along the reduction along the cutting direc-
cutting direction compensated by a tion/original length) is generally between
proportional increase of the section 15 and 60%.
thickness.

 According to the above model, cutting-

induced deformation is remarkably
homogeneous, namely, it is the same
everywhere. A straight line in the block is
transformed into a straight line in the
section and a plane in a plane. The square
marked by the points O and P is
transformed in the parallelepiped O’P’.
The equations of the transformation are

 X´ = xcos  + ysin 
 Y´ = y t/f
in which  is the cutting angle, t is the
thickness of the section and f is the feed.
Figure 11.2 A square transformed into a

parallelepiped.
The ideal case of homogeneous  The basic goal of a CEMOVISt is to

deformation holds remarkably well as long obtain homogeneously compressed sections
as the material is homogeneous. Irregular with no other cutting artefacts, such as
deformations and fractures, however, are crevasses and chatter.
common due to discontinuity in the cutting
process or to the non homogeneity of the
material.
There are indications that an oscillating  The demonstrated advantage of

knife may considerably reduce com- oscillating for cutting soft material at room
pression.2 temperature has also been observed in
vitreous sections, but not yet on a
reproductive basis.
 The knife edge is very sharp, but not

perfectly sharp, thus causing an additional
deformation close to the surface of the
section. It seems that this effect does not
affect more than 1/5 of the thickness of the
thinnest sections.
 There are also places where the knife
edge is damaged (dashed line). At this place
the deformation is more severe. A furrow is
formed in the bottom face of the section
and a corresponding ridge on the block
face. These are knife marks.
 A more common source of knife marks
is particles adhering on the surface of the
knife.
Figure 11.3 Many knife marks due to
irregularities at the knife edge.

One may try to reduce deformation by  Crevasses are fractures generally arising
applying a higher force during a shorter at the top surface of the section. It is
time period. Unfortunately the cutting force common that a section becomes a field of
can only increase to the point where shear
crevasses. Crevasses are more severe in
stress produces a fracture in the material.
The result is the very common and nasty thick sections and at high cutting speed.
artefact of a crevasse. Using a small angle knife reduces them.
 Chatter is a wavy variation of the

section thickness along the cutting
direction. Contrary to what may be guessed
at first, chatter is not due to a vibration of
the knife relative to the block. It is rather a
variable compression due to irregular
friction on the knife surface. The section
moves on the knife like the string of a
violin on the bow.
According to the model described of the

sectioning process, a section of thickness t
is formed in absence of friction. Friction F
results in the cutting force R and the
increased section thickness t´.
Chatter is nasty because it is associated

with irregular compression and, thus, non
homogenous deformation. Chatter depends
primarily on the gliding properties of the
knife surface. It is reduced at high cutting
speed.
Figure 11.4 Friction on the knife surface

causes chatter.
Biological material itself is a source of  A good network of polymer, such as

irregular deformation. Membranes fre- condensed chromatin, is a favourable
quently resist deformation and are a source medium for cryo-sectioning. In this respect,
of fracture. Some regions are more brittle
the use of dextran, a sugar polymer, as a
than others and consequently more
crevasse-prone. An excellent trimming is cryoprotectant is advantageous, whereas
key to a tension-free block. other cryoprotectants, such as glycerol, are
detrimental.

1.4. Observing in the Cryo-Electron Microscope
In the electron microscope, a vitreous section differs from a conventional resin section in
several aspects. First of all, vitreous biological material is beam-sensitive in a
conspicuous way when the section bubbles before disappearing under the beam. Beam
damage starts, however, in a much more delicate way when the material of the section
flows under the effect of electrical forces generated by the electron beam. The second
major difference stems from the fact that, contrary to stained objects, imaging of native
biological material nearly exclusively relies on phase contrast, which strongly depends
on focus. As was learned from cryo-EM of vitrified suspensions, the choice of a correct
focus is of the utmost importance. Finally, contrary to post stained sections, typical for
conventional observations in which surface structures are emphasised by reinforced
staining, the structures in vitreous sections are equally visible over the entire thickness of
the section. Any image must be interpreted as originating from a 3-D object.
1.4.1. Beam damage
1. Bubbling
 For a dry specimen, beam damage is  There is a debate about whether beam
reduced 2 to 5-fold13 at liquid nitrogen damage is further significantly reduced at a
temperature. In the presence of water, temperature approaching that of liquid
molecular damage is similar, but other helium (4 K, minus 269oC). At this
effects are conspicuous. The most temperature, however, the greatly reduced
obvious is bubbling. electrical conductivity, and some other not
yet well understood transformations, make
it difficult to take full advantage of an
eventual considerable reduction of beam
damage. The future will tell if a very low
temperature is a useful avenue for cryo-EM.
 Bubbling takes place through  With 100 kV electrons, bubbling

excessive accumulation of gas typically starts between 2000 to
produced by electron beam-induced 6000 e/nm2. The onset depends on the
radio decomposition of biological nature of the organic material. Radio-
matter.9 Probably because pores sensitive substances, such as aliphatic
through which the gas could escape are hydrocarbons (lipids), bubble first, whereas
sealed with water, the accumulating aromatic-rich compounds (nucleic acids)
gas nucleates into bubbles. The are more resistant. For example, condensed
specimen seems to be boiling in ever chromatin resists longer to bubbling than
increasing bubbles, which finally the surrounding structures. Interestingly,
this effect is inverted after glutaraldehyde
collapse into a large hole marking the
fixation.
ultimate end of any observation.
Bubbling is not only a pain for the observer,
it can be used as a focusing aid.

 Obviously, any high-resolution  It is probable that a good part of the gas

structural observations cease when that produces bubbles is H2 because the
bubbling starts. It may happen, how- effect also exists at temperatures much
ever, that the differential onset of below the liquefaction temperature of other
bubbling provides valuable informa- candidates, such as O2, N2 or CO.
tion about the nature of material
present. Sometimes it also reveals
compartments that are not otherwise
recognisable.
2. Beam-induced deformation
 Because it is a high viscosity liquid,  The effect exists in thin vitrified films
the section undergoes global deform- but it is more prominent in vitreous sections
ation during irradiation, probably because they are not always well attached
because of the forces generated by non to a rigid support. This long-range deform-
balanced charge accumulations. ation is a severe problem in CET where the
information from a large number of
successive images must be combined into a
3-D model (see Chapter 12).
 Locally the material also is rearranged  The apparent curing of a badly crevassed
under the effect of the beam. In section by the electron beam is a contra-
particular, any sharp irregularities, productive beginner’s mistake. Fine
such as crevasses or knife marks, seem structural details, which were broken by
to be “ironed away” with irradiation. fracture, are rarely repaired when crevasses
Of course, such a procedure does not are ironed away. They leave characteristic
repair fine structures. discontinuities in any previously well-
defined structures, such as membranes or
filaments.
 There are also biological structures  The first visible effect of the electron
and particles, which aggregate under beam is to reinforce the compaction of
the effect of the beam. The effect has various regions, such as chromatin
been quantitatively documented for domains, chromatin fragments or other
domains of condensed chromatin, nucleoproteic structures. The effect may
which contract over a smaller surface also take place in other constituents, but it
during irradiation, thus producing an has not yet been quantitatively documented.
apparent density increase in these It could also be a beginner’s mistake to
domains.20 accept as bona fide compact bodies,
originally diffuse structures aggregated
under the effect of the electron beam. An
image free of knife marks should be
considered critically because it generally
means that the specimen has been
excessively irradiated.

1.4.2. Focus
 Amplitude contrast dominates images  Electrons intercepted in the objective
of conventionally stained sections. It aperture cause amplitude or area contrast.
reveals regions of different densities, These are mostly scattered by heavy atoms
such as stained versus unstained from the stain.
regions. It is most clearly visible when
the image is recorded close to focus.
 CEMOVIS demonstrates that the  Ribosomes are at the limit of visibility

density of native biological material is with pure amplitude contrast. It comes as a
remarkably constant. The material in surprise for many that even regions that are
the cell is relatively homogenously reputed to contain dense material, such as
distributed. Consequently, amplitude condensed chromatin, are only a few
contrast does not reveal much. There is percent above the rest of the nucleus.
little to be seen in a vitreous section
recorded at focus.
 Imaging vitreous specimens relies on  Phase contrast is due to a phase shift of

phase contrast. This mode of image the scattered beam induced by short
formation is extremely sensitive to distance density differences. It does not
small local density difference. The carry long-range information. The focus
sensitivity of phase contrast imaging is dependant dimension d, which is best
such that single unstained double- represented on the image, is
stranded DNA molecules floating in  dmax = (2∆z)1/2
water are perfectly visible.6 where  is the electron’s wave length
(3.8-3 nm at 100kV) and ∆z is the
Phase contrast is focus dependant.
defocus, whereas there is no information
Images must be recorded at a defocus
retrieval at:
corresponding to the expected dimen-  dmin = (∆z)1/2
sion of the explored structure. In For example: Best visibility of
general, several micrographs at chromatin’s granularity (10 nm):
different defocus are necessary in ∆z = 13µm
order to gain a realistic view of the closely packed dsDNA (2.5 nm):
object. ∆z = 0.8µm.
1.4.3. Volume and not surface

 In conventionally resin-embedded  Any image is expected to contain many
biological samples, post staining overlapping structures because the typical
reinforces features at the surface of the fine details that are observed (2 to 10 nm)
section. The latter is predominantly are much smaller than the section thickness
visible in the image. The situation is (rarely less than 50 nm).
different in vitreous sections of native
samples because the whole volume of
the section is equally represented on
the image. The image is strictly a two-
dimensional projection of a three-
dimensional object.21

 Conclusion 1: CEMOVIS is not  Conclusion 2: The combination of

limited by the lack of contrast of small, CEMOVIS and CET (see Chapters 12, 24)
unstained structures, but rather by the leading to a 3-D model of the material
plethora of overlapping information distribution in the vitreous section will
projected through the thickness of the overcome the limitation due to plethoric
section. information presented in any image.
1. Preparation of the material.
2. Vitrification.
3. Cryo-sectioning.
4. Observing in the cryo-electron

microscope.
3. INSTRUMENTS AND
INSTALLATION/MATERIALS/CHEMICALS
3.1. Instruments and Installation

 High-pressure freezer.  Leica EM PACT2; Leica, Vienna,
Austria: www.leica-microsystems.com
 BAL-TEC HPM 010, Balzers,
Liechtenstein. Now available from
Boeckeler Instruments, Inc., Tucson,
Arizona, USA or ABRA-Fluid AG,
Widnau, Switzerland: www.abra-fluid.ch
 A new modified apparatus HPM 100 is
available from Leica.
 Cryo-ultramicrotome.  Leica EM FC6 (2004); Leica FCS
Leica. (We have no experience with
instruments from other manufacturers.)
 Ionizer for cryo-chamber.  Static Line II, Diatome, Biel,
Switzerland:
http://www.diatome.ch/en/products/

 Diamond cutting and trimming knives;  Diatome, Biel, Switzerland:

oscillating knife. http://www.diatome.ch/en/products/
 In our hands and until 2005, vitreous
sections of comparable quality could
not be obtained with diamond knives
from other manufactures.
 The oscillating knife is promising, but
still in development.
 Controlled humidity and temperature  Necessary in the cryo-microtome room.
and control of O2 in air for safety.
3.2. Materials
 Cryo glue.  Mixture of 1 vol. ethanol in 3 vol.
isopropanol. Mixture ratios in the range
from 1:7 to 2:3 are used.
 1000 mesh grids with 50 nm carbon  Any producer.
film.
 Thinner grid, smaller holes.  Quantifoil: http://www.quantifoil.com/
 Manipulation aids for the  Provided with cryo-microtome and cryo-
ultramicrotome and tools for specimen specimen transfer system.
manipulation and storage
 Mounted eyelash  Own production: An eyelash is glued on
a wooden stick.
 Liquid nitrogen and long term
specimen storage place.
3.3. Chemicals
 Dextran.  Sigma-Aldrich, St. Louis, MO, USA.
 Ethanol and isopropanol for cryo glue.
 Methyl cyclohexane for Leica  Merck, Eurolab, VWR International,
EM PACT. Darmstadt, Germany.
 1-hexadecene for HPM 010.  Fluka, Buchs, Switzerland.

4. PROTOCOLS
4.1. Preparation of the Material
The CEMOVIS procedure depends on the material studied and the type of high-pressure
apparatus used for vitrifying the sample. In the following description, we consider two
characteristic samples. One is bacteria vitrified in a Leica EMPACT high-pressure
freezer. The other is an organotypic slice of rat brain vitrified in a BAL-TEC HPM 010
apparatus. In both cases, the external medium must be cryoprotected, but with as little
effect as possible on the cells.
4.1.1. Cell suspension
 Bacteria grown to observation

conditions in growth medium phosphate
buffered saline (PBS).
 Centrifugation
 1 min at 20,000 g
or
 15 min at 3000 g
Soft pellet gently re-suspended in
growth medium supplemented with
dextran at final concentration of 20%.
 Second centrifugation
 Pellet resuspended in ca. 5 × its
volume of medium supplemented with
20% dextran (minimal final volume:
30 µL, see Section 4.2.1).
 Dextran 40 kDa at 20% does not cross

the plasma membrane. It has little osmotic
effect (20 mOs).
A = 20% dextran in PBS
Figure 11.5 Cryo-preservation of bacteria.

4.1.2. Tissues
Tissues from the central nervous system are especially difficult to vitrify, probably
because they have a higher than average water content. Vitrification of the cells could
only be achieved after a slight osmotic dehydration of the tissue by 5% sucrose (total
osmolarity: 600 mOs).
 Organotypic slice of rat hypocampus in

experimental culture medium; 5 × 2 mm,
thickness: 200 µm.
Extraction of a 1 mm diameter slice
 The small slice is placed for five

minutes in the culture medium
supplemented with 20% dextran and 5%
sucrose. After this time, the tissue remains
functional.
Figure 11.6 Extraction of a tissue sample

for vitrification
4.2. Vitrification
4.2.1. Cell suspension vitrified in a Leica EM-PACT high-pressure freezer
 30 µL of bacteria suspension on a
Parafilm.
 Suck the suspension into the copper tube
with a piston until full. No air bubble must
remain in the tube.
 Dimensions of the copper tube:
 Length: 20 mm
 External diameter: 600 µm
 Internal diameter: 300 µm
The specimen is thus separated from the
cryogen by 150 µm of copper.
 High-pressure vitrification according to
instrument procedure (see Chapter 6).
 Pressure should not exceed 2000 bars;
higher pressure favours formation of high-
pressure ice.
 After cooling, the tube can be stored at
liquid nitrogen temperature or directly
mounted in the cryo-ultramicrotome.
Figure 11.7 Vitrification in EM-PACT
high-pressure freezer.

4.2.2. Organotypic slice of rat brain in high-pressure freezer HPM 10 from

BAL-TEC
 The excised slice in dextran and sucrose
is mounted in the HP 010 aluminium holder
with a minimum amount of liquid left. The
rest of the volume is filled with hexadecane
taking care to leave no air bubble; the
sandwich is closed.
 There are different holder sizes. The
specimen is typically separated from
the cryogen by 200 µm of aluminium.
 The rest of the procedure follows

immediately according to instrument
procedure (see Chapter 5).
Figure 11.8 Vitrification in HPM 010 high-

pressure freezer.
Extraction of the specimen from the holder.  With pins and scalpel at 140°C in the
microtome chamber.
The work in the cryo-ultramicrotome takes place below 140oC in the atmosphere of the
laboratory. Even though the instrument is carefully designed to produce a frost repelling
air stream around the working area, the presence of snowflakes may be a serious
problem, eventually leading to the deposition of a frost layer, thus impairing any useful
work. The cryo-ultramicrotome should be placed in a draft-free room with controlled
humidity and temperature. It must be securely ventilated and a safe level of oxygen must
be strictly controlled.
 Recommended conditions in the cryo-  Danger: Use of liquid nitrogen in a

ultramicrotome room area: 20°C and closed room is dangerous. One litre of
30% relative humidity. liquid nitrogen produces 700 litres of
nitrogen gas. Suffocation without
warning starts when 02 concentrations
are slightly reduced (below 19.5%) and it
rapidly leads to death.

4.3.1. Mounting the specimen in the cryo-ultramicrotome

 The copper tube from Leica EM PACT2
can be directly mounted in the specimen
holder.
 The fragment of material extracted from
the sandwich of the BAL-TEC HPM 010
freezer is mounted with cryoglue.18
1. Cryoglue is a mixture of ethanol and
2-propanol in a 2:3 ratio.
2. A drop of cryoglue is applied on the
specimen head at 140°C. The glue is
slightly soft at this temperature.
3. The specimen fragment is forced into
the glue and correctly oriented.
4. Temperature is reduced to 160°C.
The glue becomes rigid enough for
further operations.
Figure 11.9 Vitreous specimen mounted in
a cryo-ultramicrotome
4.3.2. Trimming the specimen

 The specimen surface is prepared flat
with a trimming knife. The whole section of
the EM PACT2 copper tube is sectioned
with ca. 0.05 µm (cryoglued and/or fragile
specimens) to 1 µm strokes (clamped
copper tubes). Typically, some 200 µm
must be removed from the tube; less from a
glued specimen.
 The first evaluation of the state of the
specimen can generally be made at this
stage. A well vitrified specimen seen with
top illumination is uniformly black in the
copper tube. A non vitrified region or
inhomogeneous material is shiny or milky.
Many bad specimens can already be
discarded at this stage.
 A square-based pyramid is an excellent
shape for final preparation of the specimen.
A 45° trimming knife must be used.
Suitable dimensions are 100 to 200 µm at
the base and 10 to 100 µm high.
A rectangular parallelepiped trimmed with
a 90° knife may offer advantages, but it can
be dangerously prone to fracture.
Figure 11.10 Vitreous specimen trimmed
in a cryo-ultramicrotome.

4.3.3. Cutting sections

 The conditions for obtaining good
sections are a well-prepared, homo-
geneously vitrified specimen, careful
trimming, and an excellent cutting knife.
The latter is of utmost importance. Besides
being sharp and clean, it must have good
gliding properties. The manufacturing
procedures are not known publicly.
Research on this matter should be
encouraged.
 In addition, the working room must be

reasonably free of snowflake contamination
and the forces induced by electric charges
must be neutralised. This is best done with
the ion shower incorporated into the cryo-
ultramicrotome. Operating conditions must
be adjusted according to the situation.
 Under good cutting conditions, the

sections appear as regular ribbons, which
neither stick to the surface of the knife nor
fly away. Adjust the ion shower by
changing the distance of the ion source to
the knife edge and the power of the ionizer.
 The ribbons are manipulated with the

eyelash. Under well-adjusted conditions,
the ribbon just sticks to the eyelash that can
be transferred directly onto the
prepositioned grid. Two to 10 ribbons can
be mounted on each grid.
 Some operators hold the growing ribbon

with the eyelash and pull it gently. This
longer ribbons can be obtained.
The incorporation of a micromanipulator
into future instruments could help in
democratising the method.
Figure 11.11 Ribbons of vitreous sections

are transferred from knife to grid.

4.3.4. Flattening the section on the grid

 The ribbons on the grid are flattened by
pressing gently with a flat surface. This
action is necessary to reduce the probability
of losing the sections during storage and
transfer and to improve the stability of the
section during irradiation. A clean, smooth
surface and tools, free of snow flakes, are
important. Experience dictates the force that
should be used for pressing. An
exaggerated force destroys the supporting
film.
The grids with the ribbons are stored in a

specimen box that can be closed and stored
for a long period of time in liquid nitrogen.
Figure 11.12 Ribbons of vitreous sections

are flattened on the grid.
4.4. Observing in the Cryo-Electron Microscope

4.4.1. Transfer
 Transfer takes place as for any vitrified  Under-liquid transfer minimises the risk
specimen (see Chapters 3, 4, 7–10, of contamination; above-liquid minimises
12). In the transfer station, the the risk of losing sections.
manipulation can be performed under
liquid nitrogen or above its surface.
 After transfer, the microscope must be  Poor vacuum may contaminate the
allowed time for vacuum restoration specimen with a layer of water condensed
and temperature stabilisation. from the gas phase in the form of
amorphous water or cubic ice.
 Non equilibrated temperature causes

drift of the specimen, reducing image
quality.

4.4.2. Evaluation of the specimen

 Overview  Low magnification observation of the
grid. Localisation and general appearance
of the ribbon; level of contamination by
snow.
 Test of vitrification by electron  No specimen should be studied without
diffraction. first checking the state of water. The test is
immediate and it prevents many mistakes.
 Determination of the thickness of the  This valuable measure can be made
sections rapidly and at low magnification. The
electron flux, shown in Figure 11.13, of the
section (Is) is compared to the flux where
there is no section (Io). The thickness is
given by the equation in which t is the
thickness for a specimen of unit density,
and L is the mean free mass thickness — a
parameter of the microscope, typically of
the order of 200 nm. Even if  is not known
precisely, the relative value of t is a
precious guide. (Practical details can be
found in Dubochet et al.10)
Figure 11.13
4.4.3. Observation and image recording

A competent cryo-electron microscopist is recognised by the fact that at any given
moment he knows the dose already received by the specimen.
 Calibration of the electron dose is  For a given illumination on the screen,
obtained from direct current the dose on the specimen increases with the
determination on the screen, from square of the magnification (a single
optical density of photographic film micrograph at 20,000 × produces the same
exposed under controlled conditions or dose as 100 micrographs at 2000 ×).
from charge coupled device (CCD)
calibration data.
 A normal illumination on the screen
(one second exposure) at 10’000
magnification typically corresponds to
ca. 100 electrons/nm2 in a 1 second
exposure.
 As compared to conventional  The quantitative values are given by the
specimens, images of vitreous sections equations in Section 1.4.2.
must be recorded at higher defocus.
The optimal defocus value depends on
the studied dimension.
 Computerised electron tomography,  Recording the long-tilt series needed for
CET CET requires a good section, adequate
hardware and software and expertise (see
Chapter 12).

5.1. Advantages
 Close to native specimen with all the  Is it really so, that apart from viscosity,
water immobilised in a liquid state. there is no difference between liquid and
vitreous water? It may be so in practice, but
fundamentally it is probably not the case
(see Chapter 1).
 No stain.  The image is a true representation of
material present in its natural state.
 No chemical fixation.  Chemical fixation is not the worst
among conventional artefacts. It produces
fine-scale aggregation.
 No dehydration.  Dehydration is arguably the most severe
artefact of a conventional preparation. It
acts at a molecular and at a cellular scale.
 No aggregation artefact.  The result of the two previous
advantages.
 High resolution.  Atomic preservation in the section; 2 nm
achievable in the image under favourable
orientation. In combination with CET, a
2 nm, 3-D map is a reasonable goal for the
years to come.
 Potentially very rapid.  Is less-than-an-hour EM possible?
Extraction of the sample and vitrification:
4 min; mounting, trimming and sectioning:
30 min; coffee: 5 min; observing, recording
and transferring micrographs to the
operating surgeon: 20 min.
5.2. Disadvantages
 Technically demanding (material and  Yes. But does any good work require
skill). less than top quality?
 Cutting artefacts.  Yes. But they can frequently be
 mastered.
 Low contrast.  No. This is not a limitation. Only the
limitation on signal-to-noise ratio is
important.
 Beam damage  Yes. But it can be minimised close to the
fundamental limit it imposes on signal-to-
noise ratio.
 Plethora of information within the  Yes. But it can be overcome by:
thickness of the section.  Cutting thinner sections
 CET

 Difficulty in placing fiducial marks for  Yes. But solutions are in progress by:
CET.  the adjunction of markers on the
vitreous section (see Chapter 12)
 fiducial-free tomographic alignment
 Not adequate for post immunolabel-  Yes. But staining and labelling can
ling. sometimes be performed before specimen
preparation and freeze-substitution is a
valuable alternative (see Chapter 13).
6. WHY AND WHEN TO USE CEMOVIS
CEMOVIS should be used whenever possible because it is the best method for preserving
whole cells and tissues in their native state. In particular, it should be used when “liquid”
structures are observed ― i.e., when the observed structure has no solid substrate or
scaffold ― and when high resolution is aimed for.
The contraindications for using CEMOVIS are: (1) the technical difficulty of the method
and (2) the beam sensitivity of the vitreous sections.
Neither would be fully relevant if freeze-substitution did not exist, but the latter method
provides a valuable alternative in many cases (see Chapter 13).
If freeze-substitution and CEMOVIS are compared, we note that both preparation

methods are identical up to the point where the specimens are vitrified. The roads then
diverge. In the first case, the sample is freeze-substituted before being cut and then
observed at room temperature, whereas with CEMOVIS the preparation procedure
continues at a low temperature.
Freeze-substitution, therefore, should be considered as the method of choice for rigid

specimens and when high resolution observation is not the major goal. It is then possible
to obtain relatively beam-resistant sections particularly suitable for CET (see Chapter 24)
allowing a resolution that compares favourably with conventional resin-embedded
specimens.
It remains, however, that any “liquid” structure, any feature that is partially or totally free
to float in the water medium of the native state, will never be faithfully represented in
dehydrated specimens. Experience with CEMOVIS shows that the effect of dehydration
has probably been grossly underestimated in structural biology and this is why
CEMOVIS should be used, whenever possible.

7. OBSERVED RESULTS
 Figure on the Chapter’s title page  CEMOVIS: This elegant axoneme —

probably a terminal portion — was
observed in an organotypic slice of rat
brain.21
See text for methodological details.
Bar = 100 nm
 Figure 11.14
 Three views with CEMOVIS.
A) Yeast cells in 20% dextran. Sectioning
artefacts. The progression of the knife
in the vitreous sample from top-left to
bottom-right has deformed the cells
into ellipses, produced knife-marks
(arrows), numerous crevasses and the
undulation of chatter (bottom left half).
B) Gram-positive bacteria (Enterococcus
faecalis). A good section.
C) Organotypic slice of rat brain.

 Axons (A)
 Presynaptic vesicles (V)
 Microtubules in the plane of the
section (µa) and perpendicular to it
(µb).
Bar = 1 µm
See text for details.


8. REFERENCES
1. Al-Amoudi, A. et al. Amorphous solid water produced by cryo-sectioning of

crystalline ice at 113K. J. Microsc., 207, 146, 2002.
2. Al-Amoudi, A. et al. An oscillating cryo-knife reduces cutting-induced deformation
of vitreous ultrathin sections. J. Microsc., 212, 26, 2003.
3. Al-Amoudi, A. et al. Cryo-electron microscopy of vitreous sections of native
biological cells and tissues. J. Struct. Biol., 148, 131, 2004.
4. Al-Amoudi, A. et al. Cryo-electron microscopy of vitreous sections. EMBO J., 23,
3583, 2004.
5. Al-Amoudi, A. et al. Cutting artefacts and cutting process in vitreous sections for
cryo-electron microscopy. J. Struct. Biol., 150, 109, 2005.
6. Bednar, J. et al. Determination of DNA persistence length by cryo-electron
microscopy. Separation of the static and dynamic contributions to the apparent
persistence length of DNA. J. Mol. Biol, 254, 579, 1995.
7. Bouchet-Marquis, C. et al. Cryoelectron microscopy of vitrified sections: A new
challenge for the analysis of functional nuclear architecture. Histochem. Cell. Biol.,
125, 43, 2006.
8. Bouchet-Marquis, C. et al. Visualization of cell microtubules in their native state.
Biol. Cell, 99, 45. 2007.
9. Dubochet, J. et al. Cryoelectron microscopy of vitrified specimens, in
Cryotechniques in Biological Electron Microscopy. Steinbrecht, R. A. and Zierold,
K., eds., Springer, Berlin, Germany, 1987, 114.
10. Dubochet, J. et al. Cryo-electron microscopy of vitrified specimens. Q. Rev.
Biophys., 21, 129, 1988.
11. Dubochet, J. et al. How to "read" a vitreous section, in Cryotechniques in
Biological Electron Microscopy, McIntosh, J.R., ed., Elsevier, Amsterdam, The
Netherlands, 2007, 386–403.
12. Fernández-Morán, H. Low-temperature preparation techniques for electron
microscopy of biological specimens based on rapid freezing with liquid helium II.
Ann. New York Acad. Sci., 85, 689, 1960.
13. International Study Group. Cryo-protection in electron microscopy. J. Microsc.,
141, 385, 1986.
14. Hsieh, C.-E. et al. Towards high-resolution three-dimensional imaging of native
mammalian tissue: Electron tomography of frozen-hydrated rat liver sections.
J. Struct. Biol., 153, 1, 2006.
15. Lepault, J. et al. Freezing of aqueous specimens: An x-ray diffraction study.
J. Microsc., 187, 158, 1997.
16. Matias, V.R.F. and Beveridge, T.J. Native cell wall organization shown by cryo-
electron microscopy confirms the existence of a periplasmic space in
Staphylococcus aureus. J. Bacteriol., 188, 1011, 2006.

17. Michel, M. et al. Diamonds are a cryo-sectioners best friend. J. Microsc., 166, 43,
1992.
18. Richter, K. A cryoglue to mount vitreous biological specimens for
cryoultramicrotomy at 110K. J. Microsc., 173, 143, 1994.
19. Sartori, N. and Salamin-Michel, L. Cryotransmission electron microscopy of thin
vitrified sections, in A Laboratory Handbook, Celis, E., ed., Academic Press,
London, U.K., 1994, 323.
20. Sartori Blanc, N. et al. Electron beam-induced changes in vitreous sections of
biological samples. J. Microsc., 192, 194, 1998.
21. Zuber, B. et al. The mammalian central nervous synaptic cleft contains a high
density of periodically organized complexes. Proc. Natl. Acad. Sci. USA, 102,
19192, 2005.

Cryo- Electron Tomography 293
CONTENTS

1. PRINCIPLES OF CRYO- ELECTRON TOMOGRAPHY........................... 296
1.1. Specimen Preservation ............................................................................. 296
1.2. Acquisition and Mutual Alignment of Projection Images ........................ 297
1.3. Reconstruction.......................................................................................... 299
1.3.1. Principles of three-dimensional reconstruction according to Radon29
1.3.2. Reconstruction algorithms .......................................................... 300
1.3.3. Distortions................................................................................... 300
1.4. Image Segmentation and Isosurface Rendering........................................ 301
1.4.1. Denoising .................................................................................... 302
1.4.2. Surface rendering........................................................................ 302
3. EQUIPMENT/PRODUCTS/REAGENTS....................................................... 305
3.1. Equipment................................................................................................. 305
3.2. Products .................................................................................................... 306
3.3. Reagents ................................................................................................... 306
4. PROTOCOLS .................................................................................................... 307
4.1. Preparation of Protein-Stabilised Colloidal Gold ..................................... 307
4.2. Application of Quantum Dots to the Surfaces of Vitreous Cryosections . 308
5.1. Advantages of Cryo- Electron Tomography............................................. 310
5.2. Technical Limitations in Cryo- Electron Tomography............................. 310
6.1. Vitrification Methods ............................................................................... 311
6.1.1. Plunge-freezing........................................................................... 311
6.1.2. High-pressure freezing................................................................ 312
6.2. Alignment and Reconstruction Methods .................................................. 312
6.2.1. Fiducial marker alignment .......................................................... 312
6.2.2. Cross-correlation alignment........................................................ 312
6.2.3. Choosing a reconstruction algorithm .......................................... 313
8. REFERENCES .................................................................................................. 316

Title page illustration: Courtesy of Dr. Stephan Nickell and with permission from Nature
Reviews Molecular Cell Biology.1

Cryo- electron tomography is a visualisation technique that aims to combine the

resolving power conferred by the short wavelength of electrons with optimal structural
preservation and three-dimensional image reconstruction.2 The resolution achieved
using this technique currently approaches four nanometres, which is sufficient to resolve
macromolecular complexes larger than ca. 400 kDa. The advantages of cryo- electron
tomography are obvious, even at this rather modest resolution: It provides the link
between high-resolution structural imaging performed on isolated macromolecules and
their imaging within the cellular context. The ultimate goal is to create a pseudoatomic
atlas of a cell, with high-resolution structures obtained by x-ray crystallography or
single particle electron microscopy being substituted for structures in tomographic
maps, creating synthetic tomograms that show the absolute positions and relationships
between macromolecules within the cell.1
To take full advantage of this emerging technology, steps must be taken to ensure that
the specimen remains stable under vacuum and during irradiation with electrons, but
also that it closely resembles its native state. These prerequisites are achieved by
converting the specimen into a specific frozen-hydrated (vitreous) state3 and by
carefully limiting the electron dose encountered by the specimen so as to preclude
radiation-induced damage. Vitrification and the low-dose acquisition procedure
distinguish cryo- electron tomography from electron tomography of chemically fixed
and stained, plastic-embedded material. This is not to suggest that embedding plastics
are not affected by the electron beam; on the contrary, they shrink markedly and are
typically pre-exposed until the initial, rapid shrinkage has stabilised. Finally, heavy
metal stains responsible for amplitude contrast in conventional electron microscopy are
also omitted such that the densities of the specimen are represented, rather than indirect
visualisation of stain. Contrast is enhanced by recording images at a specified
underfocus value, creating so-called phase contrast. The further from focus, the more
that resolution is sacrificed. Therefore, interpretation is hindered by the fact that signal
is difficult to discern from background noise.
Electron tomography is commonly performed using intermediate accelerating voltages,

typically 300 kV or 400 kV. The wavelength of a 300 kV electron is ca. 2 pm, i.e.
2 × 10-12 m. Cryo- electron tomography is commonly performed on macromolecules,
organelles or cells contained within thin films of ice created by blotting with filter paper
and shock-freezing in liquid ethane (see Chapter 3). With few exceptions, such
specimens can be captured without significant artefacts. When specimen thickness
exceeds more than ca. 500 nm, as is the case for many prokaryotes and nearly all
eukaryotes, vitreous slices must be obtained (see Chapter 11), analogous to sectioned
plastic material. This currently represents a major bottleneck for cryo- electron
tomography of vitrified biological specimens because the options for correcting cutting-
induced artefacts are severely hampered by the need to maintain vitreous specimens at
temperatures below 137°C to avoid a deleterious phase change.
This chapter is concerned with the workflow in cryo- electron tomography, from
preparing the specimen so that it can be frozen without the formation of ice “crystals” to
computational rendering of a three-dimensional volume that allows for detailed
interpretation.

1. PRINCIPLES OF CRYO- ELECTRON TOMOGRAPHY
1.1. Specimen Preservation

It is common knowledge that biological  Specimen preservation for cryo-
specimens typically comprise more than electron tomography is essentially identical
70% water. At first glance, this would seem to that used for cryo- electron microscopy.
to contradict the fundamental requirements A comprehensive description is provided
of an electron microscope: Water boils in Chapter 7.
spontaneously at the low pressures required
to allow electrons to travel unimpeded  There are two practical ways to achieve
through space. The most accurate vitrification: “plunge” freezing in liquid
preservation of molecular structure as well ethane (see Chapters 3, 4, 7) and high-
as specimen stability within the vacuum pressure freezing followed by vitreous
column of the microscope, thus, is achieved cryosectioning (see Chapters 5, 6, 11).
by rapid freezing, hence the prefix “cryo”,
meaning “cryogenic”. With a sufficiently  Cryo- electron tomography is sometimes
fast cooling rate, crystallisation of water is abbreviated to cryo- ET or CET. Although
inhibited, and the solid water formed is there is no such thing as a cryoelectron
amorphous (featureless). This is commonly (hence the space, as well as the hyphen to
described as vitreous or “glass-like” water. indicate the prefix), there is also no such
thing as a cryo- tomogram, nor does
electron microscopy refer to microscopy of
In addition to the frozen sample’s negligible electrons. The debate is rather pointless;
rate of evaporation in vacuum, cryo- names of techniques are invariably
fixation for electron tomography serves compromised by the attempt at being
additional purposes: concise.

1. Within the cell interior, biological  Every suspended “particle” has an
nanomachines (proteins and protein energy of approximately 1.5 kT, where k is
complexes) must be imaged within a the Boltzmann constant, and T is the
tiny fraction of a second so as not to be temperature in Kelvin (see Chapter 1): Of
blurred by Brownian or ATP- course, vitrification precludes observations
dependent motion if they are to be of dynamic events, placing emphasis on
amenable to unambiguous structural the need for novel correlative light and
recognition. electron microscopy approaches.
2. Three-dimensional structure is pre-  In fact, structural preservation resulting

served at least to the level that can from vitrification is superior to the
potentially be visualised by electrons practical limit of detection by electrons. It
(i.e. molecules). is sufficient for furnishing structures by
means of electron crystallography.

3. The native state is maintained without  After vitrification, image acquisition as

aggregation. well as any preparative steps (usually
limited to sectioning) must take place at
4. Improved resistance to irradiation with temperatures below 137°C, the devitri-
electrons. fication temperature.
1.2. Acquisition and Mutual Alignment of Projection Images

 Acquisition 
A tomogram is reconstructed from a series  Various acquisition schemes are
of two-dimensional projection images (see available in terms of constant exposure time
Figure 12.1). or variations to account for increasing
specimen thickness, e.g., linear exposure
and Saxton schemes.5



Figure 12.1 Acquisition of a series of two-
dimensional projection images from a
tilted specimen in the transmission electron
microscope. Simulated, selected projection
images of a macromolecular complex —
the 20S proteasome — are shown.

Imaging with electrons has a price: 
Biomolecules comprise mainly light atoms 
(C, H, O, N, P) and thus have little contrast 
to electrons. The most logical way to 
increase signal is to apply a larger electron  Or by using the artefact-prone, indirect
dose; however, biological specimens method: impregnating the specimen with
embedded in a thin film of vitreous ice heavy metal stains: Note that there have
tolerate only about 5000 electrons nm-2, been attempts to combine the virtues of
implying that the cumulative dose for a tilt- cryogenic preservation and heavy metal
series must be kept below this level and staining (see Chapter 9).
distributed between the individual 2-D
images, which in the case of cryo- ET
typically numbers 60 to 80. This is known
as “dose-fractionation.” Accordingly, each  The signal of individual projection
projection image has a very low overall images must be sufficient to ensure
contrast, and the final “signal” represents accurate tracking of the specimen during
the combination of signal from all acquisition.
projections. 


The completeness of information is

approximated by the Crowther criterion4
defined by the angular range sampled and a
finite number of equally spaced projection
images. The full angular range (180°) is
generally not applicable in electron  Actually, the Crowther criterion was
tomography because the grid holder described earlier by Bracewell and Riddle.6
geometry restricts the tilt range to ±70° and
perhaps, more importantly, a tilted
specimen with a flat (slab) geometry also
becomes prohibitively thick with respect to
the beam. For practical reasons (mitigation  Novel sample geometries and tilting
of beam damage and thus limiting the devices have the potential to solve both
electron dose and, to a lesser extent, problems.
speeding up acquisition), a compromise 
must be made: The number of images is
made constant and these are distributed
over the maximum possible angular range,
thereby defining the tilt increment.
An energy filter operating in “zero-loss”  Essentially chromatic aberrations.

mode removes contributions from inelas-  The mean free path for electrons in ice
tically scattered electrons. This diminishes is approximately 350 nm, implying that an
blurring, especially in the case of thicker energy filter will provide an advantage
specimens. even for a thin specimen that exceeds this
apparent thickness during tilting.
 Detectors
Digital recording devices were critical for
enabling automated data acquisition for
electron tomography. Because images are
acquired in real time, tracking and focus
should be taken into account for subsequent
images of a tilt series, leading to routine
data acquisition under low-dose conditions.
Pixel-based detectors, such as charged-
coupled device (CCD) cameras are rapidly
approaching the quality of silver halide
film, and single-electron detectors that do
not suffer from “leakage” of signal into
neighbouring pixels should provide the
single most important foreseeable
improvement in resolution for electron
tomography.1

 Alignment
Projection images must be aligned to a
common coordinate system to account for
rotations and unintentional changes in
focus. This is most easily achieved by
introducing an appropriate amount of 10 nm
colloidal gold particles into the sample prior
to vitrification. A minimum of three well-
spaced particles (and preferably five to
account for rotations) visible in all images
of a series is usually sufficient to perform a
satisfactory alignment. Quality control is
possible by analysing the root mean square
(RMS) errors and the trajectories (profiles)
of gold particles for a tilt series. Analysis of  An RMS error of less than two pixels is
marker profiles can help to eliminate an generally considered to be acceptable.
outlier (spurious marker) or to show if a
more serious problem exists, e.g.,
unexpected movement of the specimen.
1.3. Reconstruction
1.3.1. Principles of three-dimensional reconstruction according to Radon
The mathematical description for producing  Even for most German speakers with a
a three-dimensional reconstruction from a background in mathematics, the original
series of two-dimensional images was version is reputedly difficult to read. Two
described in 1917 by Radon.7,8 The English-language translations are available,
reconstruction is achieved by back- those by Parks9 and by Lohner.10
projecting the images in reciprocal or 
Fourier space (see Figure 12.2). A
weighting function (hence “weighted”
back-projection) is used to account for
variations in low- and high-frequency
information. Without the weighting factor,
reconstructions appear dull and featureless.

Figure 12.2. Computational synthesis of a

three-dimensional image volume using
back-projection. For clarity, only selected,
simulated projection images are shown.


1.3.2. Reconstruction algorithms

In addition to weighted back-projection, a
number of more sophisticated recon-
struction algorithms have been redis-
covered recently for electron tomography.
These include the iterative reconstruction
algorithms ART (algebraic reconstruction  Weighted back-projection will continue
technique) and SIRT (simultaneous to be applied for tomographic recon-
iterative reconstruction technique). Both struction because it is extremely rapid. If
begin with an initial back-projection the results of weighted back-projection look
followed by multiple cycles of image satisfactory, the user can start a more time-
comparison until a given stopping criterion consuming iterative reconstruction.
is fulfilled. ART compares one slice to
other slices, while SIRT compares each
slice to the entire volume.

1.3.3. Distortions
Distortions in electron tomograms arise  Even if holders did not restrict the
principally from the incomplete sampling of angular tilt range, progressive thickening of
information. The specimen holder restricts the specimen relative to the electron beam
the maximum tilt range to ±70° instead of during tilting to extreme angle prevents a
the full tilt range of ±90° required for full angular scan for most conventional
“isotropic” (balanced) resolution. This is sample geometries (plunge-frozen or cryo-
known as the “missing wedge” in reference sectioned materials have a “slab”
to its appearance in Fourier space. It gives geometry). For an untilted specimen with
rise to the perception that spherical objects thickness 500 nm, tilting to 70° results in an
are elongated. apparent thickness of well over 1 µm.
Acquisition of a complementary tomogram  Distortions in accurately representing the

obtained by tilting about an orthogonal axis 3-D information are covered in the
results in more complete and more isotropic following section.
resolution, the smaller “missing pyramid”.
A second potential source of distortions is

the fact that some features appear to be
continuous with one another due to
smearing, when in fact they might be
discontinuous. This is a consequence of the
finite tilt increment. More detailed
reconstructions would result from smaller
tilt increments.


1.4. Image Segmentation and Isosurface Rendering

Homo sapiens is a “visual-centric”
species,11 with a disproportional
dependence on the sense of sight, and a
unique ability to perceive three-dimensional
space and colour. A three-dimensional
volume that can be magnified, rotated and
viewed from any direction thus is
indispensable as a research tool.
The aim of segmentation and isosurface

rendering is the objective decomposition of
a complex, three-dimensional object into its  Post selection operations, such as
individual components for the purposes of Gaussian smoothing or the removal of
improved comprehension12 (see Figure outlying pixels, may improve the aesthetic
12.3). appearance of surface renderings, but
 possibly at the expense of accuracy and
objectivity.

 Interactive visualisation techniques —
manually rotating and zooming with a
computer or movies that animate aspects of
this process — offer the possibility to
examine a 3-D object to scale and without
distortion. Furthermore, human vision is
certainly capable of seeing potential
connections between objects, but to some
extent, it relies on motion for 3-D
perception.

Figure 12.3. Selected, surface-rendered
features of a cryo- electron tomogram
displayed as an offset above an orthoslice
(digital slice) of the tomogram. To aid in
segmentation, the tomogram was denoised
using nonlinear anisotropic diffusion (see
following section) .13
Although computer-based segmentation is 
in principle more objective, manual 
segmentation is often the only way to 
delineate some portions of a cryo- 
tomogram. This is because human per- 
ception is superior to that of available 
segmentation algorithms. An a priori 
knowledge of an object’s identity, e.g., an
actin filament, might be obtained by
independent means, such as diffraction
patterns indicating the characteristic
spacing of actin monomers; however, the
human mind is prone to biasing the
perception of whether two features are truly
interconnected.

1.4.1. Denoising
In cryo- electron tomography, it is common
practice to apply some form of denoising in 
order to facilitate segmentation.14 Because
denoising procedures typically reduce noise
at the expense of some signal, this practice
is not carried out prior to performing a
pattern recognition procedure, such as
template matching. Different levels of
denoising can be used to enhance the
delineation of various features having
specific shapes, e.g., microtubules. For this
purpose, a broad mask is applied for a
conservative segmentation that includes a
number of adjacent pixels on either side of
the object. The mask is then applied to the
non denoised data and thresholded to limit
bleeding of the selection into other features.
Where averaging of symmetrical or

repetitive structures is not applicable, as is
the case for the many uniquely shaped
(pleomorphic) cellular features, classical,
real-space digital filters, such as Gaussian
smoothing, can be useful. Filtering means
that the grey-scale value of a pixel is
replaced by a new value based on its
current value and the values of
neighbouring pixels. More advanced digital
filters, such as non-linear, anisotropic
diffusion have been applied recently.

1.4.2. Surface rendering
Surface rendering is used to improve the 
perception of the 3-D data. It is used to 
highlight, for example, the continuity of 
membranes, and the spatial relationships of 
macromolecular complexes to neigh- 
bouring organelles or molecules. A three- 
dimensional image volume necessitates 
creativity in image display — a 3-D 
volume cannot be portrayed on a flat 
medium such as this sheet of paper without 
causing distortions to either shape (lack of 
conformity) or size. This is most easily 
comprehended by examining 2-D 
representations of the surface of a spherical  The Earth is actually slightly elliptical,
object, such as the globe (see Figure 12.4). but the principle described here is still
Map projections are typically visualized valid.

geometrically as if they are transferred onto  “Conformal” refers to a projection for

a cylinder, cone or plane. Reliefs of the which all angles at each point are preserved.
Earth’s surface are described as conformal,  “Equal-area” projections show all areas
equal area, or neither — they cannot be in proportion to their true areas, but shapes
conformal and equal area. Cartographers become distorted.
have attempted to overcome this problem 
for centuries. All representations of this
kind are compromises, but they may be
useful for certain purposes. The Mercator
projection, only recently superseded for
navigation by satellite-based global
positioning systems (GPS), is extremely
useful for navigation because distortion is
constant along any parallel. In other words,
at any given point, east–west scale is the
same as the north–south scale.

 Despite its distortions, the Mercator

projection is the only map projection that
shows true bearings for navigation.
 Distortion map (Tissot Indicatrix) for the

Mercator projection. All circles are perfect
spheres, indicating that there is no angular
distortion, but the areas vary. The
consequence is that a land mass, such as
Greenland, appears to be the size of Africa,
when it is, in fact, only a fraction of the
size.


Figure 12.4 Distortions induced by
projecting a sphere onto a cylinder: the
Mercator projection.

Discontinuous projections (see Figure 12.5) 

are useful for emphasising certain features 
at the expense of others, e.g., land masses 
and oceans, and potentially restricting the  Selected references:
gross distortions to the regions that are not  John B. Garver Jr., New Perspective
emphasised. on the World, National Geographic
Magazine, 1988, pp. 911–913.
 John P. Snyder, Flattening The Earth –

2000 Years of Map Projections, The
University of Chicago Press, Chicago,
USA, 1993, pp. 214216.

Figure 12.5 Mollweide projection, which
biases the perception by emphasising
continuities between land masses.
1. Culture of cells in suspension or on

TEM grids.
2. Addition of colloidal gold to
suspension cells.
3. Vitrification by plunge freezing in
liquid ethane or using a high-
pressure freezing instrument.
4. Vitreous sectioning (if high-pressure
frozen).
 4a = Deposition of inorganic, colloidal
particles onto vitreous section
5. Transfer to electron microscope in
cryo- holder.
6. Mapping of specimen grids using
low magnification; determination of
suitable areas for tomography.
7. Basic transmission electron
microscope (TEM) alignments, set-
ting of eucentric height, setting of
tomography parameters (tracking
area and dose, focus area and dose,
exposure weightings).
Figure 12.6

3. EQUIPMENT/PRODUCTS/REAGENTS
3.1. Equipment
 Plunge-freezing apparatus  Leica Microsystems, Vienna, Austria, or
homemade; controlled temperature and
humidity using the Vitrobot, FEI Electron
Optics, Eindhoven, The Netherlands (see
Chapter 4).
 High-pressure freezer  Leica; BAL-TEC HPM 010 now
available from Boeckeler Instruments, Inc.,
Tucson, Arizona, USA; M. Wohlwend
GmbH, Sennwald, Switzerland (see
Chapters 5, 6).
 Grid boxes and opening tool  Self-made or commercial supplier.
 Self-closing forceps  Dumont #5, Dumont S.A., Switzerland,
preferably Teflon®-coated or insulated with
additional foam; more sophisticated and
costly forceps are available: Dumont
#545.
 Plasma device for glow-discharge  For rendering carbon-coated TEM grids
hydrophilic and for removing organic
contaminants, e.g., Harrick Plasma, Ithaca,
New York, USA.
 Liquid nitrogen Dewars for storage  For example, Taylor-Wharton, Husum,
Germany; Theodore, Alabama, USA; Air
Liquide S.A., Düsseldorf, Germany.
 Cryo- ultramicrotome  For example, Leica Ultracut UC6 + FC6
(see Chapter 11.)
 Diamond knives  For example, Diatome AG, Biel,
Switzerland; Drukker International B. V.,
Cuijk, The Netherlands; Delaware Diamond
Knives, Wilmington, Delaware, USA.
 Cryo- electron microscope  An energy filter is indispensable for
thicker specimens, where “thicker” refers to
the relationship between the mean free path
of electrons in ice and the maximum
specimen thickness encountered during
tilting.
 Cryo- tilt holder  For example, Gatan Model 626, Gatan,
Pleasanton, California, USA; dedicated
cryo-microscopes, such as the Polara and
the Krios, have built-in capabilities, FEI
Company, Eindhoven, The Netherlands.
 Light microscope and cell incubator  Quality control, correlative microscopy.

3.2. Products
 TEM grids  Various commercial suppliers; more
sophisticated carbon coatings from
Quantifoil Micro Tools GmbH, Jena,
Germany or C-flat, Protochips, Inc.,
Raleigh, North Carolina, USA.
 Tomography acquisition and  Commercial products, such as
reconstruction software Explore3D, FEI Electron Optics,
Eindhoven, The Netherlands or open-
source, dedicated tomography packages,
e.g., TOM Toolbox,
www.biochem.mpg.de/tom
 Segmentation and rendering software  Amira, Visage Imaging, inc., Karlsbad,
California, USA or Imaris, Bitplane AG,
Zürich, Switzerland; enhanced capabilities
are provided with 3DSMax or Maya,
Autodesk, Munich, Germany.

3.3. Reagents
 Liquid nitrogen  In Europe: Messer Griesheim, Linde,
Westfalen. Must be kept dry.
 Ethane gas  BOC gases, part of Linde AG, Munich,
Germany.
w
 20% and 40% ( /v) Dextran in  Neutral polysaccharide from
physiological buffer or growth Leuconostoc mesenteroides. Sigma D-4876,
medium Mr 100,000 to 200,000 dissolves quite
rapidly.
 Colloidal gold  Self-made as described in Section 4.1, or
e.g., Sigma G-1527, Sigma Chemicals, St.
Louis, Missouri, USA; note that the
commercial product contains sodium azide,
a potent respiratory poison. In either case, it
is preferable to conjugate with protein A or
bovine serum albumin to prevent auto-
aggregation. In the absence of NaN3, shelf-
life of protein-stabilised gold is limited to
several weeks at room temperature, and
dissociation may also limit the usefulness
of older preparations.
Safety information
Extra caution is required when using liquid
nitrogen (suffocation risk, cold injury) and
ethane (explosion risk). Read the Materials
Safety Data Sheets for all reagents before
use.
Disclaimer
The authors do not have financial interests
in any of the above-mentioned companies
or products.

4. PROTOCOLS
4.1. Preparation of Protein-Stabilised Colloidal Gold

It is common practice in cryo- electron 
tomography to add colloidal gold particles
to cell suspensions shortly prior to plunge
freezing (see Figure 12.7).



Figure 12.7 Projection image of a
bacterium embedded in vitreous ice. The
preparation contains colloidal gold particles
to enable subsequent alignment. For details,
see Section 7, Observed Results, Figure
12.3.
Bar = 200 nm

Colloidal gold with a size of 10 to 12 nm is 
most commonly prepared by the citrate
reduction method followed by conjugation
with either Protein A (from Staphylococcus
aureus) or albumin. Although there are
other ways of preparing colloidal gold, this
method is easy and reliable.
 Citrate reduction method, based on 

the 1926 Nobel Prize lecture of R. A. 
Zsigmondy, for 10 to 12 nm gold 
colloids.15 

1. Solution A: Add 1 mL 1% tetrachloro 
aurate to 80 mL distilled water in a 
strongly stirred beaker. 
2. Solution B: Dilute 4.5 mL 1% tri-  The ratio of gold to tannic acid governs
sodium citrate dihydrate to 20 mL with the resulting colloid size. 40 µL of tannic
distilled water and add 40 µL acid should give approximately 12 nm
1% tannic acid (e.g., Mallinckrodt colloids.
#1764).
3. Warm solutions A and B separately to  The resulting solution has a pH of about

60°C. Mix solutions A and B rapidly, 5.0.
which should become black and then
turn red after 30 minutes.

Advantage over commercial preparation: no 

sodium azide and, therefore no immediate
toxicity when added to cells.
 Stabilisation of gold colloids with 

bovine serum albumin (BSA)
1. Bring the pH of the gold colloid 

suspension to pH 6 using dilute base
(NaOH, KOH) while stirring.
2. Dissolve BSA at a concentration of  Protein A from Staphylococcus aureus

2 mg/mL in a pH 6, low-molarity can be used instead of BSA; the stock
buffer, e.g., 1 to 2 mM phosphate solution should be made at 1 mg/mL, and
buffer. Recheck the pH and adjust to 6 the amount to be added to the gold solution
if necessary. is about 5 µg/mL, i.e., about five times less
Protein A is needed than for BSA. The
exact amount to be added can be
determined by a microtitration assay.
Usually, coupling with Protein A is
followed by the addition of BSA to a
concentration of 0.2% to ensure maximum
stability.
3. Add 10 µL of the BSA solution 
per mL of gold suspension while 
stirring. Allow to stand for 10 minutes, 
and then adjust the pH depending on 
the intended application. 



4.2. Application of Quantum Dots to the Surfaces of Vitreous
Cryosections
This procedure was introduced by Masich  Analysis of marker alignment profiles
et al.16 It is based on Tokuyasu’s suggests that particles associated with
observation that isopentane is liquid at surface artefacts (crevasses) move during
150°C, and that commercially available the course of image acquisition, resulting in
inorganic semiconductor particles (lead poor alignment. It is possible to incorporate
sulfide quantum dots) are soluble in fiducial markers within suspensions of
toluene, which in turn is miscible with microorganisms prior to vitrification and
isopentane. An abbreviated version is sectioning; however, introduction of foreign
described here. particles to the cell interior is invasive
unless the aim is to study endocytosis.



The main advantage of this method is that it  Disadvantages: The method is based on
is universally applicable, i.e., independent adsorption of inorganic particles onto the
of specimen type (cryosections of section surface, the same region which
eukaryotic or prokaryotic cells) and thus suffers from the most artefacts and thus
non invasive. changes most rapidly during acquisition.
This is, of course, a fault of the section, not
Additional special equipment: steel or the marker deposition method. Unlike
aluminium blocks containing two wells, one colloidal gold, quantum dots are perfect
for the quantum dot suspension and one for crystals. This means that the contrast
liquid ethane. The well for the isopentane- changes abruptly for different tilt angles,
based quantum dot suspension should be making detection difficult for thicker
countersunk to form a funnel at the top of specimens containing large variations in
the well. contrast. A further problem is that this
particular type of quantum dot (PbS in
1. Cut vitreous sections and apply them toluene) is especially large (15 to 17 nm)
to TEM grids as usual. which is inaccurate for a tilt series taken
using high magnifications.
2. Cool the metal block using liquid
nitrogen and allow the nitrogen to
evaporate from the wells in the cryo-
ultramicrotome chamber.
3. Fill one well with ethane gas, which  The back of the chamber might be
will liquefy within the cooled well. significantly cooler than the setting
Immerse the block in liquid nitrogen. indicated on the microtome display panel.
The ethane will solidify to a white
mass. Transfer the block to the
chamber of the cryo-microtome. The  The ethane will liquefy quickly, but it
ethane will quickly become a liquid as does not evaporate very fast at this
the block warms to 150°C in the temperature; there is no need to rush the
microtome chamber. If the ethane fails experiment.
to liquefy, move the block to the
surface of the knife holder. Fill the
other well with a suspension of lead
sulfide quantum dots dissolved in
toluene and diluted 50:1 (v/v) with
isopentane.
4. Dip the grid containing the sections
into the quantum dot suspension, rinse
briefly in liquid ethane, and then blot
the excess ethane by wicking onto a
piece of filter paper in the chamber.
The ethane must be seen to have wet
the paper. Store the grids in liquid
nitrogen for microscopy.

5.1. Advantages of Cryo- Electron Tomography

 Intuitive, 3-D depiction.  Clear visualisation concerning the spatial
relationships of a macromolecular complex
to other macromolecules and organelles.
 Maintenance of specimen hydration.  Essential for maintaining the spatial
coordinates of “soluble” cytoplasmic
components.
 No staining.  Direct visualisation of native structure
rather than an uninterpretable combination
of stain (dominant) and specimen.
 No chemical fixation.  Physical fixation (cryo- fixation) serves
the purpose of immobilisation without the
aggregation artefacts; faster.
 Direct recognition of large complexes.  Indirect recognition by immunolabelling
is restricted to the section surface (2-D).
 Spatial resolution of 4 to 5 nm.  An advantage relative to light
microscopy.
 Safe.  Fewer and less hazardous reagents
reduce the potential for generating artefacts
in specimen and user.
5.2. Technical Limitations in Cryo- Electron Tomography

 Some specimens, e.g., brain tissue,  Vitrification is essentially an all-or-
may be difficult to vitrify.17 nothing phenomenon and depends on the
ratio of water to polymer.
 Limited contrast of specimen.  Reliance mainly on phase (cf. amplitude)
contrast; low signal-to-noise ratio makes
interpretation and segmentation difficult.
 Low tolerance to radiation damage.  Limit to the tolerable dose = limit to the
number of images = finite tilt increment =
incomplete information.
 “Working blind” — difficulty in
surveying the specimen without spending
the restricted dose on secondary tasks.
 Optimal alignment often possible only  Low signal-to-noise level means that
by means of fiducial markers. cross-correlation approaches are either
unreliable or less accurate; introduction of
markers should be done in a non invasive
manner.

 Limited tilt range, typically ± 60°.  Incomplete information, manifested as

distortions.
 Kinetic studies difficult.  Requirement to capture specific events,

and to relate tomograms to events
immediately prior to vitrification
(correlative microscopy); this is, of course,
a characteristic of all electron microscopy,
not just cryo- ET.
 Thinning of vitreous specimens is  Choices are few and imperfect; the

difficult. vitreous character of the specimen must not
be compromised.
 Dedicated cryo- stages/microscopes  The price for trying to maintain biology

necessary. within a vacuum column: “suspended
animation.”
6.1. Vitrification Methods

6.1.1. Plunge-freezing
 Depending largely on polymer  Specimens thicker than ca. 800 nm are
concentration and as a corollary, on not electron-transparent. The limiting factor
water content, plunge-freezing can be for tomography, therefore, is specimen
used to vitrify biological specimens thickness (cell + overlying ice) rather than
contained within thin films to a depth vitrification depth.
of 3 to 20 µm. Such specimens include
purified suspensions of macro-
molecules or suspensions of uni-
cellular microorganisms.
 
 Because the depth of vitrification  Physical ablation results in the
achieved by plunge-freezing is under- irreversible loss of a significant portion of a
utilised and bypasses the artefacts of specimen.
the cryo- sectioning process, it may be
possible in the future to take greater
advantage of the effective depth of
vitrification by means of specimen
thinning techniques, such as focussed
ion beam milling.
 

6.1.2. High-pressure freezing

 High-pressure freezing is the method  High-pressure freezing is the method of
of choice for specimens that are not choice for mammalian cells and tissue
electron transparent, i.e., > ca. 1 µm, biopsies.
with the implication that it can only be
used in conjunction with appropriate
sectioning techniques. The depth of
vitrification is at least 200 µm,
allowing vitrification of cells and
tissues that are not amenable to plunge
freezing.


6.2. Alignment and Reconstruction Methods
6.2.1. Fiducial marker alignment
 This method is generally used in  Incorporation of particles within the ice
combination with plunge-freezing (see and thus within the specimen volume
Section 5.1) for suspensions of small, should result in the most satisfactory results
unicellular organisms or macro- because multiple planes of the specimen are
molecules where the gold does not taken into account during alignment, and
penetrate the specimen and thus can be the markers are less likely to move as may
used for non invasive alignment. occur for surface-adsorbed particles.
 With a little more effort, the variant

method described for vitreous cryo-
sections in Section 4.2 is suitable for
any specimen because it is applied post
vitrification.

6.2.2. Cross-correlation alignment
 This method is generally used when  Cross-correlation or general feature-
fiducial marker alignment is not based alignment may be used in
possible. An example is the lack of combination with marker alignment,
sufficient gold particles in vitreous although this is generally not applicable for
cryosections, especially in the case of cryo- tomography; either marker alignment
eukaryotic cells where the extracellular is satisfactory or there are no markers
matrix may not be captured within the available. The “features” of unstained
imaged area. specimens are generally not obvious at all
tilt angles.

 Cross-correlation often fails at high tilt

angles because the signal-to-noise ratio
is insufficient to generate a reliable
cross-correlation peak.
6.2.3. Choosing a reconstruction algorithm

 Weighted back-projection (WBP) is a  WBP reconstructions are rich in contrast,
mathematically sound and compu- but this is partly a result of halo-like
tationally inexpensive means of brightness around dark objects. Depending
reconstructing an image volume. on once point of view, this is either
Iterative algorithms18 are compu- aesthetically pleasing or artificial.
tationally expensive, but may give
more realistic-looking reconstructions.
 SIRT is preferable to ART for cryo-  Computation time will eventually

electron tomography projections become irrelevant. WBP is a convenient
because it is less prone to fail when the way of seeing the reconstruction within a
images are dominated by noise. SIRT few moments.
reconstructions are not dramatically
different to those produced by
weighted back-projection, but the
images have a natural appearance in
that they lack artificial edge  If the goal is to perform template-
enhancements, and back-projection matching, it is essential to use a method,
“rays” are suppressed. which does not contain constraints, e.g.,
WBP.

7. OBSERVED RESULTS
 Figure on the Chapter’s title page  Three-dimensional reconstruction of a

hydrated biological specimen from a series
of two-dimensional projection images
(schematic).
 Figure 12.8 (see colour inserting  Multiscale imaging of hydrated

following page???.) mammalian cells.
A: Phase contrast light micrograph
B: Cryo- EM projection image
C: Segmented cryo-electron tomogram
from a plunge-frozen mouse
adenocarcinoma cell grown on an EM
specimen grid
Note that the central regions are
inaccessible unless sectioned.
Bar for A = 20 µm
Bar for B = 500 nm
 Additional colour information:

 Blue: Cell membrane
 Green: Microtubules
 Yellow: Transport vesicles
 Red: Macromolecular complexes
 Segmentation and surface rendering by
M. Gruska and L. Andrees, MPI for
Biochemistry.
 Cells were provided by W. G. Müller
and J. McNally, National Institutes of
Health.
 Figure 12.9  Low-dose projection image of a portion

of a bacterium, Magnetospirillum gryphis-
waldense, showing the 10 nm colloidal
gold particles used for alignment.
Magnetosomes are iron oxide particles
used for orientation according to the
Earth’s magnetic field. For details, see
Scheffel et al.19.
Bar = 200 nm


8. REFERENCES
1. Nickell, S. et al. A visual approach to proteomics, Nat. Rev. Mol. Cell Biol., 7, 225,
2006.
2. Baumeister, W. Mapping molecular landscapes inside cells, Biol. Chem., 385, 865,
2005.
3. Dubochet, J. and Sartori Blanc, N. The cell in absence of aggregation artifacts,
Micron, 32, 91, 2001.
4. Crowther, R.A., DeRosier, D.J., and Klug, A. The reconstruction of a three-
dimensional structure from its projections and its applications to electron
microscopy, P. Roy. Soc. Lond. A Mat., 317, 319, 1970.
5. Saxton, W.O., Baumeister, W., and Hahn, M. Three-dimensional reconstruction
of imperfect two-dimensional crystals, Ultramicroscopy, 13, 57, 1984.
6. Bracewell, R.N. and Riddle, A.C. Inversion of fan-beam scans in radio astronomy,
Astrophys. J., 150, 427, 1927.
7. Radon, J. Über die Bestimmung von Funktionen durch ihre Integralwerte längs
gewisser Mannigfaltikeiten. Berichte über die Verhandlung der Königlich
Sächsischen Gesellschaft der Wissenschaften zu Leibzig, in Mathematische
Physische Klasse, 1917, 262.
8. Radon, J. On the determination of functions from their integral values along certain
manifolds, IEEE Trans. Med. Imaging, MI-5, 170, 1986.
9. Parks, P.C. IEEE Trans. Med. Imaging, MI-5, 170, 1986.
10. Radon, J. On the determination of functions from their integrals along certain
manifolds, in The Radon Transform and Some of Its Applications, Deans, S.R., ed.,
Wiley-Interscience, New York, USA, 1983, Appendix A.
11. Anon Seeing is believing, Nat. Methods, 2, 889, 2005.
12. Frangakis, A.S. and Hegerl, R. Segmentation of three-dimensional electron
tomographic images, in Electron Tomography: Methods for Three-Dimensional
Visualization of Structures in the Cell, Frank, J., ed., Springer-Verlag, Heidelberg,
Germany, 2006, 353.
13. Frangakis, A.S., Stoschek, A., and Hegerl, R. Wavelet transform filtering and
nonlinear anisotropic diffusion assessed for signal reconstruction performance on
multidimensional biomedical data, IEEE Trans. Biomed. Eng., 48, 213, 2001.
14. Hegerl, R. and Frangakis, A.S. Denoising of electron tomograms, in Electron
Tomography: Methods for Three-Dimensional Visualization of Structures in the
Cell, Frank, J., ed., Springer-Verlag, Heidelberg, Germany, 2006, 331.
15. Slot, J.W. and Geuze, H.J. Sizing of protein A-colloidal gold probes for
immunoelectron microscopy, J. Cell Biol., 90, 533, 1981.
16. Masich, S. et al. A procedure to deposit fiducial markers on vitreous cryo-sections
for cellular tomography, J. Struct. Biol., 156, 461, 2006.
17. Zuber, B. et al. The mammalian central nervous synaptic cleft contains a high
density of periodically organized complexes, Proc. Natl. Acad. Sci. USA, 102,
19192, 2005.
18. Gilbert, P. Iterative methods for the three-dimensional reconstruction of an object
from projections, J. Theor. Biol., 36, 105, 1972.
19. Scheffel, A. et al. An acidic protein aligns magnetosomes along a filamentous
structure in magnetotactic bacteria, Nature, 440, 110, 2006.

Part III
Low Temperature Embedding

Freeze-Substitution 321
CONTENTS

1. PRINCIPLES OF FREEZE-SUBSTITUTION............................................... 325
3.1. Materials ................................................................................................... 327
3.2. Products .................................................................................................... 327
3.3. Solutions ................................................................................................... 329
3.3.1. Substitution media ...................................................................... 329
3.3.2. Embedding media ....................................................................... 330
4. PROTOCOLS .................................................................................................... 331
4.1. Sample Preparation................................................................................... 331
4.2. Preparation of Substitution Apparatus...................................................... 331
4.3. Freeze-Substitution................................................................................... 331
4.4. Preparation for Resin Embedding............................................................. 332
4.5. Low-Temperature Embedding in Methacrylates ...................................... 332
4.6. Low-Temperature Polymerisation ............................................................ 333
4.7. Curing of Methacrylates ........................................................................... 333
4.8. Embedding in LR White and Epoxy Resins ............................................. 333
5.1. Advantages ............................................................................................... 334
5.2. Disadvantages........................................................................................... 334
6.1. Choice of Substitution Medium................................................................ 335
6.2. Choice of Resin ........................................................................................ 335
6.3. Osmium Tetroxide and Low-Temperature Embedding ............................ 335
8. REFERENCES .................................................................................................. 340


Our life is confined to water. All cellular processes need water to work. As soon as water
is removed, not only do the processes cease and life stops, but also the cellular
ultrastructure is altered. Therefore, the most straightforward method to study cells and
cellular structure and architecture by microscopy is to visualise them in their water
environment. For light microscopy, live cell imaging has become a well-established
method. For higher resolution in an electron microscope, however, the vacuum required
to allow electrons travel does not allow visualisation of living cells in liquid water. The
closest approach is to embed the biological material in a thin layer of ice and image it in a
cryo-electron microscope (see Chapters 1, 11).1,2 This method has its limitations: A thin
sample is needed and ice or ice-embedded biology is very beam sensitive. Recently,
marvellous results were obtained by cryo-electron tomography (see Chapter 12)3,4 and
even non invasive, immunolabelling by template matching is within reach5 and with it,
imaging life and identification of the molecular machinery in situ.
Up to now this fantastic technique has been limited to very thin samples, which can be
cryo-fixed on a specimen grid by, e.g., plunge-freezing.3 A more useful method would be
to cryofix a thick specimen, a piece of tissue, and make thick cryo-sections. Cryo-fixation
of reasonably thick samples, (2 mm diameter, 100 to 300 μm thickness) can be achieved
by high-pressure freezing.6-8 From these samples ultrathin cryo-sections can be cut with a
cryo-ultramicrotome9, but not thick sections10 employed in electron tomography. Despite
the limitation of section thickness, electron tomography on ultrathin cryo-sections is very
useful to acquire high-resolution data (see Chapter 12), but cannot visualise the
distribution of larger entities in their cellular context. A solution for this problem is
looming on the horizon. With a new instrument, a focused ion beam (FIB) scanning
electron microscope, a semithin (ca. 300 nm) section can be cut with a gallium ion beam
under observation with an electron beam.11 The section can be lifted out and transferred
into a cryo-transmission electron microscope. This method is still emerging and, until this
new method is established for routine applications, we use a hybrid technique, freeze-
substitution, which can bridge the gap between cryo- and conventional electron
microscopy.
To ensure high quality preservation of biological samples, we choose cryo-fixation with

high-pressure freezing in combination with freeze-substitution and resin embedding.12-14
The freeze-substitution technique comprises dehydration of a cryo-fixed sample at 90oC
by dissolving the ice with an organic solvent and, if wished, chemical fixation during the
substitution process. The term freeze-substitution often also implies subsequent resin
embedding.

The methodology of freeze-substitution already has a long history. It was first described
by Simpson in 1941.15 Thereafter, it was further developed and improved.16-20 Today, it
has become a standard technique (for review, see12). Probably the most convincing data to
demonstrate the supremacy of cryo-fixation and freeze-substitution has been presented by
Studer et al.22 In chemically fixed nodules of soybean, the Rhyzobium bacteria have no
contact with the cell membranes of the nodules, whereas there is a tight connection when
visualised by cryo-fixation and freeze-substitution. In fact, this direct contact is a
prerequisite for their function because a mutant, which has no contact to the nodule
membranes, cannot convert atmospheric nitrogen into NH4+. Another great advantage of
the freeze-substitution technique is that it can be combined with on-section immunogold
labelling12,23 for localisation studies with reliable morphology.
Several modifications of the freeze substitution procedure have been devised to overcome
two major draw-backs: 1) the heavy metal salts used for contrasting penetrate poorly into
the section, and 2) in well preserved, i.e., optimally cryo-fixed cells, internal membranes
are difficult to visualise. These limitations have been addressed by changing the
substitution method. Three-dimensional contrast can be achieved by adding uranyl acetate
and osmium tetroxide to the substitution medium.12 Membrane contrast can also be
enhanced by adding contrasting agents,24 but more reliably by adding a small amount of
water to the substitution medium.25 The amount of water to be added is dependent on the
sample and varies between 1 and 5%.
In this chapter, I would like to present and discuss our recently developed protocols to
improve sample preparation by freeze-substitution, especially with respect to their
application in cellular electron tomography14,26-27 (also see Chapter 24).

1. PRINCIPLES OF FREEZE-SUBSTITUTION
The biological sample is cryofixed by  Optimum structural preservation, close

means of one of the methods suited for to the native state of the biological
cryo-fixation (see Part I, Cryo-Fixation specimen.
Methods), e.g., by high-pressure freezing.
The layer of optimum cryo-fixation is
dependent on the method used and on the
content of free water within the specimen.
For high-pressure freezing, the thickness of
well-frozen samples may vary between
100 μm to about 300 μm.
At about –90°C, the cryofixed samples are  At –90°C, the ice will be dissolved by
put into an organic substitution medium, the organic solvent, the sample is
based on, e.g., acetone, ethanol or meth- dehydrated. The solvent and the chemical
anol, containing chemical fixatives, such as fixatives will surround the biological
uranyl acetate, osmium tetroxide and structures. The dehydration time is
aldehydes. The samples are kept at –90°C dependent on the specimen, e.g., cells with
for 8 hours up to as much as 80 hours. thick cell walls, such as yeast or plant cells,
will need more time for the exchange of
water than cultured cells. In addition, slow
dehydration better preserves the natural
distribution of ions and reduces shrinkage
artefacts (see Chapter 15).
After dehydration, the temperature is raised  During the warming-up process the
to the desired embedding temperature, e.g., chemical fixatives start to react: uranyl
–40°C to –30°C for Lowicryl HM20 or acetate as soon as the negative charges of
K11M, but also for Epon and LR White. the nucleic acids and phosphate groups
become accessible, osmium tetroxide at
around –70°C28 and glutaraldehyde between
–40°C and –30°C.12
For certain protocols, the temperature is  To improve membrane contrast.24

raised up to 0°C for 1 hour before
embedding.
Embedding in the Lowicryls takes place  K4M at –30°C; HM20 from –30°C to
between –60°C and –30°C depending on –50°C; K11M to –60°C; and HM23 may be
the formulation of the resin. used down to –70°C.

The first dilutions of Epon and LR White  A 30% solution of Epon or LR White is
can also be infiltrated at around –30°C, but liquid at –30°C. Epon/acetone mixtures are
the further embedding steps are done at liquid at –30°C up to 60% Epon, however,
room temperature and curing at +50°C. not Epon/ethanol mixtures! In case of
LR White, mixtures with ethanol can and
should be used.
The methacrylate-embedded samples are
suited for on-section immunolabelling
studies, whereas the Epon-embedded
samples are best suited for morphological
analysis including electron tomography.
1. Cryo-fixation  Best preservation of the cellular and

molecular architecture by immediate arrest.
 Note, that during freeze-substitution the

dehydration and chemical fixation step are
inverted.
Figure 13.1 Schematic representation of

the freeze-substitution process.
2. Dehydration  During warming up, amorphous water

can crystallise into cubic ice; ice is the
3. Chemical fixation thermodynamically more stable form of
water. This process is also called de-
4. Resin infiltration and vitrification and in pure water de-
polymerisation vitrification takes place at around –135°C
(see Chapters 1, 11). At the temperature
5. Sectioning freeze-substitution starts, amorphous ice
should not exist anymore, this would imply
6. Immunolabelling that cubic ice has no influence on the
morphology at the level of resolution of
7. Analysis by light microscopy and biological samples.
electron microscopy

3.1. Materials
 Freeze-substitution apparatus. Insets  Self-made21
for the freeze-substitution apparatus Leica EM AFS or EM AFS2, Leica,
with small holes in the bottom to hold Microsystems, Vienna, Austria.
glass vials and microtubes. RMC FS-7500, Boeckeler Instruments Inc.
Tucson, Arizona, USA.
 Microtubes 500 μL and 1500 μL.  The screw and seal tubes are very well
suited for freeze-substitution (1.5 mL) tubes
#72.609; Sarstedt AG & Co, Nümbrecht,
Germany.
 The tubes used for methacrylate
embedding need to be absolutely air tight.
Here we use 0.5 mL Sarsteadt #72.699.
 Glass vials.  Used to prepare the solutions to be pre-
cooled in the substitution unit.
  Borosilicate glass counting vials
(20 mL) from Wheaton, Millville, New
Jersey, USA, with urea screw cap #986542
(Closure Liner: Metal Foil) or 986546
(Closure Liner: Poly-Seal Cone) or
Wheaton Liquid Scintillation Vials from
Electron Microscopy Sciences, Hatfield,
Pennsylvania, USA, #72632 or Perkin-
Elmer, Waltham, Massachusetts, USA,
#6000096.
 Perforated insets.  Leica plastic capsules D5 x H 15 mm
#16702738; Leica, Microsystems, Vienna,
Austria.
 Forceps.  For example, Dumont (#2a, #3, #4, #4N,
#7), Dumont & Fils, Switzerland.
 Styrofoam boxes.  Use different sizes for the preparation of
the sample and media.
 Ice.  Used for incubation at 0°C.
 UV lamp ~ 360 nm.  Used for low temperature embedding in
methacrylates.
 Oven.  Used for embedding in LR White or
epoxy resins.
3.2. Products
 Acetone.  Dry acetone (Merck #1.00299.0500,
Merck KGaA, Frankfurter Strasse 250,
D64293 Darmstadt, Germany): The open
bottle is stored under nitrogen gas in a
closed container above silica gel.

 Methanol.  Dry methanol (Merck #1.06012.0500),

the open bottle is stored under nitrogen gas
in a closed container above silica gel.
 Ethanol.  Dry ethanol (Merck #1.00990.0500), the
open bottle is stored under nitrogen gas in a
closed container above silica gel.
 Sodium hydroxide (NaOH).  Merck.
 Glutaraldehyde.  Glutaraldyde is available as a 10%
solution in water-free acetone (EMS
#16530), ethanol (EMS #6531) or methanol
(EMS #16532) or a 50% solution in water.
 Osmium tetroxide.  10 × 1 g ampoules #19110 or 10 × 0.1 g
ampoules #19134; EMS, Hatfield,
Pennsylvania, USA.
 Uranyl acetate.  EMS, Hatfield, Pennsylvania, USA or
SPI Supplies, West Chester, Pennsylvania,
USA.
 Double distilled water.  Quartz double distilled water is preferred
over Millipore filtered water, but both can
be used.
 Lowicryl resin:  The HM resins are hydrophobic,
 HM20, HM23, K4M, K11M whereas the KM resins have some
hydrophilic properties, e.g., EMS or
Polysciences Europe GmbH, Eppelheim,
Germany.
 There are premixed versions of the
Lowicryl resins. We, however, prefer to
mix the solutions ourselves to avoid aging
and prepolymerisation during storage.
 London Resins:  LR White is for heat polymerisation and
 LR White, LR Gold LR Gold may be UV polymerised down to
ca. 10°C. By changing the initiator, using
dibenzoyl peroxide, LR White is also suited
for low temperature embedding down to
about 5°C; EMS.
Caution: LR White does not always
contain the catalyst needed for
polymerisation. It can also happen that
mixtures of LR White and ethanol
polymerise prematurely. Therefore, it is
advisable to test every new batch before
use.
 Epon.  For example, #45359, Fluka, Sigma-
Aldrich Chemie GmbH, CH-9471 Buchs,
St. Gallen, Switzerland.

 Spurr’s resin.  Due to its toxicity, carcinogenicity and

the high vapour pressure, we prefer not to
use Spurr’s. In our hands, Epon embedding
works even with samples difficult to
infiltrate, such as yeast.
 #14300, EMS or #R1032, Agar
Scientific Ltd, Stansted, UK.

3.3. Solutions
3.3.1. Substitution media
 High contrast FS medium17  Uranyl acetate is prepared as a 20%
 3% (v/v) Glutaraldehyde (w/v) stock solution in methanol.
 1% (w/v) Osmium tetroxide
Equal amounts of 6% (v/v) glutaraldehyde,
0.5% (w/v) Uranyl acetate
in methanol 1% (w/v) uranyl acetate in methanol and 2%
(w/v) osmium tetroxide in methanol are
cooled above liquid nitrogen close to the
freezing point of the solutions. Then the
glutaraldehyde/uranyl acetate solution is
poured into the osmium tetroxide solution
and the container is well shaken. This
medium can be stored in Teflon® bottles
under liquid nitrogen. As long as the
solution has a greenish yellow colour, the
quality of the medium is good. Do not use it
anymore when it turns golden yellow.
Always keep close to the freezing point
when handling (Dr. M. Müller, personal
communication).
 Standard FS medium19,29  This is the classical freeze-substitution
 1 to 5% (w/v) osmium tetroxide medium.
in acetone  We usually use 2% osmium tetroxide.
 Membrane contrast medium  Either of the chemical fixatives or any

 0.5% (v/v) glutaraldehyde combination of them may be used.
 1% (w/v) osmium tetroxide  To enhance membrane contrast,
 0.2% (w/v) uranyl acetate depending on the sample, 1 to 5% water can
 x% (v/v) water25 be added.
in acetone
 Wild FS medium24  Used to enhance resolution of the

 0.25% glutaraldehyde (50% aqueous membrane bilayer.
stock solution)
 0.5% osmium tetroxide
in acetone

 Hawes FS medium30  In combination with rapid freeze-

a) morphology substitution and embedding in Lowicryl
 2% uranyl acetate in acetone (from a HM20 at 50oC.
20% methanol stock solution)
b) immunolabelling
 0.2% uranyl acetate in acetone (from
a 20% methanol stock solution)
 Pure solvent for FS20
 In combination with low-temperature
 Acetone, or embedding, this medium is suited for
 Ethanol, or immunolabelling of aldehyde-sensitive pro-
 Methanol. teins.
 Low concentration of aldehyde FS  Either of these chemical fixatives is

medium12 used.
 2% (w/v) formaldehyde, or  In combination with low-temperature
 0.05% (v/v) glutaraldehyde, or embedding, this medium is suited for
 0.5 % (w/v) uranyl acetate immunolabelling.
in methanol  Paraformaldehyde can be dissolved in
methanol by adding sodium hydroxide
pellets.
 Methanol may be replaced by acetone. In
this case, uranyl acetate can only be
dissolved at the maximum concentration of
0.2% (w/v).
 Epoxy fixation FS medium31  This medium is used for optimum
 20% Araldite/Epon embedding mix- preservation of the cellular ultrastructure.
The intracellular membranes are only
ture in acetone
visible in negative contrast.
 This medium is also suited for
immunolabelling.
3.3.2. Embedding media
 The embedding media are prepared
 Epon, Spurr’s, Lowicryl and London according to the manufacturer’s instruct-
Resin embedding tions.
 Araldite/Epon embedding mixture31  Dr. M. Müller, personal communication.
 49% (w/w) Araldite/Epon stock
solution:
 41% (w/w) Epoxy-812 substitute  Fluka #45345
 54% (w/w) Durcupan A/M epoxy  Fluka #D0291
resin
 5% Dibutyl phthalate  Fluka #80100
 49% (w/w) Hardener DDSA  Fluka #45346
 2% (w/w) Accelerator DMP-30  Fluka #45348

4. PROTOCOLS

 Sample the biological specimen as  Extreme care should be taken so that the
carefully as possible. natural environment is altered as little as
 Cryofix with one of the cryo-fixation possible (temperature, pH, osmolarity, etc.).
machines.
 For excision of tissue samples, it is
 The frozen samples may be stored
advised to use the biopsy systems of
under liquid nitrogen until use.
Hohenberg et al.32 (Transfer System; Leica
Microsystems, Vienna, Austria, see Chapter
6).
4.2. Preparation of Substitution Apparatus

 Fill the substitution apparatus with  Our standard settings are 8 hours at
liquid nitrogen. 90°C, 8 hours at 60°C and 8 hours at
 Choose the desired settings for the 30°C.17
different steps (time and temperature).
 If desired, e.g., for plant cells or fungi or
 Place all the holders needed for the
for more careful substitution to also retain
microtubes and the glass vials with the
ions (see Chapter 15), the initial step can be
substitution medium into the appara-
prolonged up to 80 hours.
tus.
 Start the cooling.  Recently Hawes30 et al. got very promising
results with much shorter time settings.
4.3. Freeze-Substitution
 Wait until the apparatus has  Caution: Often the temperature display of
equilibrated at 90°C. the substitution apparatus is not reliable. The
 Put the frozen samples into a 1.5 mL apparatus should be calibrated by measuring
microtube containing a perforated the actual temperature in the tubes.
inset.
 HPF samples and those sandwiched
 Add cold, 90°C substitution medium
between two copper plates (e.g., propane jet)
into the tube.
should be opened under liquid nitrogen.
 Start the program.
 When using 1-hexadecene in combin-ation
with HPF, it needs to be brushed off under
liquid nitrogen, otherwise the substitution
process could be hampered.33

 Alternatively, the medium is filled in the

microtubes and then frozen in liquid nitrogen.
The sample is put into the tube on top of the
frozen medium. The tubes are put into the
device and, during melting, the sample will
slowly sink into the substitution medium.
This method may not be useful for
substitution media containing water.
4.4. Preparation for Resin Embedding

 After the substitution process, the  At 30°C, osmium tetroxide and or uranyl
samples are kept at 30°C. acetate can be removed.
 The tubes can also be brought to 0°C in an
 They are washed with pure acetone, ice/water bath for one hour to increase
methanol or ethanol. membrane contrast before they are put back
to 30°C for resin infiltration.24
 For low-temperature embedding the
samples are kept at 30°C.  Note: Acetone may interfere with the
polymerisation of the methacrylates. In such
 Or they are brought to room tem- a case it is advised to wash with ethanol or
perature for embedding in LR White methanol before embedding, when acetone
or epoxy resin. was used as the substitution medium.
 When the epoxy fixation FS medium is

used, embedding is continued without
washing.
4.5. Low-Temperature Embedding in Methacrylates

 Embed in the methacrylates at 30°C,  We usually use 1/3, 2/3 resin/alcohol, then
or lower depending on the Lowicryl 100% resin 2 to 4 hours, 100% resin over-
formulation used, following the night and 100% resin 2 to 4 hours.
manufacturer’s instructions.
 For samples with thick cell walls, such as
yeast or plants, more and longer infiltration
steps may be used, e.g., 5 to 7 days in a series
of dilutions of ethanol and methacrylate
(1/100, 1/50, 1/10, 1/5, 1/2, 2/1, 5/1, and
6 changes of pure resin).23

4.6. Low-Temperature Polymerisation

 Methacrylates for low-temperature  Indirect illumination is important for more
embedding are polymerised for even polymerisation.12,20
24 hours in the substitution apparatus  The tubes should be cooled in an alcohol
at 30°C with indirect UV illumin- bath to guarantee optimum heat transfer and
ation. avoid bubbling during polymerisation.12,20
4.7. Curing of Methacrylates

 The low-temperature embedded  Fully polymerised K4M and HM20 have a
samples need to be cured at room yellowish and reddish colour, respectively.
temperature, either outdoors in the sun
or under a UV lamp in a fume hood.
4.8. Embedding in LR White and Epoxy Resins

 The samples are infiltrated with 30%  The first 30% step for LR White or epoxy
Epon in acetone or 30% LR White in may be done at 30°C, and then the sample is
ethanol at 30°C for 2 to 5 hours. warmed to room temperature in the closed
tube to avoid water condensation.
 Then the tubes are brought to room  Epon/acetone mixtures are liquid at 30°C
temperature and infiltration is up to 60% Epon.
continued as for routine embedding.
 Polymerisation takes place at 50°C for  For samples with thick cell walls, such as
at least 24 hours. yeast or plants, more and longer infiltration
steps may be used, e.g., 5 to 7 days in a series
of dilutions of acetone and Epon (1/100, 1/50,
1/10, 1/5, 1/2, 2/1, 5/1, and 6 changes of pure
resin).23
 Alternatively, Spurr’s low viscosity resin

may be used.34

5.1. Advantages
 The optimum preservation of the 
cellular architecture achieved by cryo-
fixation can be combined with resin
embedding.
 Methacrylate sections may be used for  Methacrylate sections can also be used
immunolabelling studies. for correlative light and electron
microscopy (see Chapter 21).
 Epon embedded samples are best  Epon is more stable in the electron beam
suited for reliable morphological than methacrylates.
studies.
 Thick Epon sections can easily be 

prepared and are suitable for electron

tomography (see Chapter 24).



5.2. Disadvantages
 Methacrylate sections are generally 
less sensitive for immunolabelling
studies than Tokuyasu cryo-sections
(see Chapter 19).
 In contrast to frozen-hydrated sections  Collapse of delicate, highly hydrated

(CEMOVIS, see Chapter 11,) the nanostructures.
biological structures are still exposed 
to organic solvents and chemical 
fixatives, which might alter the 
cellular architecture. 





6.1. Choice of Substitution Medium

 High contrast medium  These media are used for morphological
 Standard FS medium analysis and electron tomography.
 Membrane contrast medium  Used for morphological analysis and
 Wild FS medium electron tomography.
 Hawes FS medium (2% UA)
 Epoxy fixation FS medium  Used for morphological analysis and
electron tomography.
 With caution it can also be used for
immunolabelling.
 Hawes FS medium (0.2% UA)  In combination with methacrylate
 Low concentration of aldehyde FS embedding these media are best suited for
medium immunolabelling.
 Pure solvent for FS

6.2. Choice of Resin
 Freeze-substitution in pure solvent 
followed by low-temperature embed-
ding in methacrylate is used for
immunolabelling of aldehyde-sensitive
proteins.
 (Low-temperature) embedding in
methacrylate is used for immuno-
labelling.
 Embedding in epoxy resin is used for
morphological analysis, in particular
for electron tomography.

6.3. Osmium Tetroxide and Low-Temperature Embedding
 In contrast to the general belief,  As long as the substitution is stopped
biological samples fixed with osmium below or at 30°C and the temperature is
tetroxide during freeze-substitution not raised during embedding, the sample
can be embedded in methacrylates, has only a yellowish colour and does not
e.g., Lowicryl, at low temperature with get black. The UV light can easily penetrate
UV polymerisation. the sample and the resin polymerises.20
 Osmium tetroxide used during freeze-  At temperatures below 30°C, osmium
substitution affects antigenicity to a tetroxide reacts in the fixing manner as
much lesser extent than when used at described in the textbooks and does not get
higher temperatures. black,28 whereas at temperatures above
0°C, osmium tetroxide reacts more like a
protease with all the consequences
expected for protein epitopes.35,36

7. OBSERVED RESULTS
 Figure on the Chapter’s title page  Clamydomonas reinhardii cells were

pelleted and picked up with a copper grid
and cryofixed by plunging into liquid
propane.
Freeze-substitution was done in ethanol
containing 0.5% uranyl acetate and 0.1%
glutaraldehyde, 8 hours at –90°C; 8 hours
at 60°C; 8 hours at –35°C using a
homemade freeze-substitution apparatus.
The samples were embedded at 35°C in
Lowicryl HM20. (B.M. Humbel,
unpublished results, 1989.) Full width
corresponds to 2 µm.
 Figure 13.2  Escherichia coli cells grown at 37oC

were cryofixed by high-pressure freezing
(Leica EM HPF, Leica Microsystems,
Vienna, Austria; now M. Wohlwend,
Sennwald, Switzerland). Substitution was
done in acetone containing 2% osmium
tetroxide, 0.2% uranyl acetate and 1% H2O;
48 hours at 90°C; warmed with a slope of
2°C/h to 60°C, 8 hours at 60°C; warmed
with a slope of 2oC/h to 30°C, 8 hours at
30°C in a AFS substitution apparatus
(Leica EM AFS, Leica Microsystems,
Vienna, Austria). After one hour at 0°C,24
the E. coli cells were embedded in Epon.
(E. van Donselaar, B.M. Humbel, unpub-
lished results, 2006.)
Bar = 500 nm
 Figure 13.3  Saccharomyces cerevisiae cells were

grown to stationary phase. Then they were
cryofixed by high-pressure freezing (Leica
EM HPF, Leica Microsystems, Vienna,
Austria; now M. Wohlwend, Sennwald,
Switzerland) and freeze-substituted in
acetone containing 0.5% glutaraldehyde
and 0.2% uranyl acetate and low-
temperature embedded in Lowicryl HM20.
(E. van Donselaar, B.M. Humbel, unpub-
lished results, 2006.)
Bar = 500 nm


 Figure 13.4  Cartilage of mouse hip was cryofixed by

high-pressure freezing (Leica EM PACT,
Leica Microsystems, Vienna, Austria; see
Chapter 6) and freeze-substituted in
acetone containing 0.5% glutaraldehyde
and 0.2% uranyl acetate and 1% H2O; 48
hours at 90°C; warmed with a slope of
2°C/h to 60°C, 8 hours at 60°C; warmed
with a slope of 2°C/h to 30°C, 8 hours at
30°C in a AFS2 substitution unit using the
automatic reagent handling system (Leica
EM AFS2/Leica EM FSP, Leica
Microsystems, Vienna, Austria). The
sample was low-temperature embedded in
HM20. (E. Van Donselaar, J.W. Slot, B.M.
Humbel, unpublished results, 2005.)
Bar = 5 µm
 Figure 13.5  NRK cells grown on Aclar® disks were

cryofixed by high-pressure freezing (Leica
EM HPF, Leica Microsystems, Vienna,
Austria; now M. Wohlwend, Sennwald,
Switzerland) and freeze-substituted in
acetone with 2% osmium tetroxide;
48 hours at 90°C; warmed with a slope of
2oC/h to 60°C, 8 hours at 60°C; warmed
with a slope of 2°C/h to 30°C, 8 hours at
30°C in a AFS substitution apparatus
(Leica EM AFS, Leica Microsystems,
Vienna, Austria). After one hour at 0°C,24
the sample was embedded in Epon.37
(N. Jiménez, E. van Donselaar, B.M.
Humbel, unpublished results, 2005.)
Bar = 500 nm


8. REFERENCES
Biophys., 21, 129, 1988.
microscopy of biological specimens based on rapid freezing with liquid Helium II,
Ann. NY Acad. Sci., 85, 689, 1960.
3. Medalia, O. et al. Macromolecular architecture in eukaryotic cells visualized by
cryoelectron tomography, Science, 298, 1209, 2002.
4. Leis, A.P. et al. Cryo-electron tomography of biological specimens, IEEE Signal
Proc. Mag., 23, 95, 2006.
5. Frangakis, A.S. et al. Identification of macromolecular complexes in cryoelectron
tomograms of phantom cells, Proc. Nat. Acad. Sci. USA, 99, 14153, 2002.
6. Studer, D. et al. A new approach for cryofixation by high-pressure freezing, J.
Microsc., 203, 285, 2001.
7. Müller, M. and Moor, H. Cryofixation of thick specimens by high pressure
freezing, in Science of Biological Specimen Preparation 1983, Revel, J.P., Barnard,
T., and Haggis, G.H., eds., SEM Inc., AMF O'Hare, IL, USA, 1984, 131.
8. Riehle, U. Über die Vitrifizierung von verdünnter wässriger Lösungen. ETH Diss
Nr 4271. Federal Institute of Technology (ETH), Zürich, Switzerland, 1968.
9. Al-Amoudi, A., Norlen, L.P.O., and Dubochet, J. Cryo-electron microscopy of
vitreous sections of native biological cells and tissues, J. Struct. Biol., 148, 131,
2004.
10. Al-Amoudi, A., Studer, D., and Dubochet, J. Cutting artefacts and cutting process
in vitreous sections for cryo-electron microscopy, J. Struct. Biol., 150, 109, 2005.
11. Marko, M. et al. Focused-ion-beam thinning of frozen hydrated biological
specimens for cryoelectron microscopy, Nat. Methods, 4, 215, 2007.
12. Humbel, B.M. and Schwarz, H. Freeze-substitution for immunochemistry, in
Immuno-Gold Labeling in Cell Biology, Verkleij, A.J. and Leunissen, J.L.M., eds.,
13. Schwarz, H., Hohenberg, H., and Humbel, B.M. Freeze-substitution in virus
research: A preview., in Immunoelectron Microscopy in Virus Diagnosis and
Research, Hyatt, A.D. and Eaton, B.T., eds., CRC Press Inc., Boca Raton, FL, USA,
1993, 97.
14. Geerts, W.J.C. et al. Electron microscopy tomography and localization of proteins
and macromolecular complexes in cells, in Protein-Protein Interactions. A
Molecular Cloning Manual, Golemis, E.A. and Adams, P.D., eds., Cold Spring
Harbor Laboratories Press, NY, USA, 2006, 715.
15. Simpson, W.L. An experimental analysis of the Altmann technic of freezing-
drying, Anat. Rec., 80, 173, 1941.
Science, 129, 1284, 1959.
17. Müller, M., Marti, T., and Kriz, S. Improved structural preservation by freeze
substitution, in Proc. 7th Eur. Congr. Electron Microsc., Brederoo, P. and de
Priester, W., eds., The Hague, The Netherlands, 1980, 720.
18. Steinbrecht, R.A. and Müller, M. Freeze-substitution and freeze-drying, in
K., eds., Springer-Verlag, Berlin, Heidelberg, Germany, 1987, 149.

19. Van Harreveld, A., Crowell, J., and Malhotra, S.K. A study of extracellular
space in central nervous tissue by freeze-substitution, J. Cell Biol., 25, 117, 1965.
20. Humbel, B., Marti, T., and Müller, M. Improved structural preservation by
combining freeze substitution and low temperature embedding, Beitr.
Elektronenmikroskop. Direktabb. Oberfl., 16, 585, 1983.
21. Humbel, B. and Müller, M. Freeze substitution and low temperature embedding,
in The Science of Biological Specimen Preparation 1985, Müller, M., et al., eds.,
SEM Inc., AMF O'Hare, IL, USA, 1986, 175.
22. Studer, D., Hennecke, H., and Müller, M. High-pressure freezing of soybean
nodules leads to an improved preservation of ultrastructure, Planta, 188, 155, 1992.
23. Humbel, B.M. et al. In situ localization of β-glucans in the cell wall of
Schizosaccharomyces pombe, Yeast, 18, 433, 2001.
24. Wild, P. et al. Enhanced resolution of membranes in cultured cells by
cryoimmobilization and freeze-substitution, Microsc. Res. Tech., 53, 313, 2001.
The visibility of biological membranes is improved when the substitution medium
26. Marsh, B.J. et al. Organellar relationships in the Golgi region of the pancreatic beta
cell line, HIT-T15, visualized by high resolution electron tomography, Proc. Nat.
Acad. Sci. USA, 98, 2399, 2001.
27. Murk, J.-L.A.N. et al. 3-D Structure of multilaminar lysosomes in antigen
presenting cells reveals trapping of MHC II on the internal membranes, Traffic, 5,
936, 2004.
28. White, D.L. et al. The chemical nature of osmium tetroxide fixation and staining of
membranes by x-ray photoelectron spectroscopy, Biochim. Biophys. Acta, 436, 577,
1976.
29. Van Harreveld, A. and Crowell, J. Electron microscopy after rapid freezing on a
30. Hawes, P. et al. Rapid freeze-substitution preserves membranes in high-pressure
frozen tissue culture cells, J. Microsc., 226, 182, 2007.
31. Matsko, N. and Müller, M. Epoxy resin as fixative during freeze-substitution, J.
Struct. Biol., 152, 92, 2005.
32. Hohenberg, H., Tobler, M., and Müller, M. High-pressure freezing of tissue
obtained by fine-needle biopsy, J. Microsc., 183, 133, 1996.
34. Spurr, A.R. A low-viscosity epoxy resin embedding medium for electron
microscopy, J. Ultrastruct. Res., 26, 31, 1969.
35. Behrman, E.J. The chemistry of osmium tetroxide fixation, in The Science of
Biological Specimen Preparation 1983, Revel, J.P., Barnard, T., and Haggis, G.H.,
eds., SEM Inc., AMF O'Hare, IL, USA, 1984, 1.
36. Maupin, P. and Pollard, T.D. Actin filament destruction by osmium tetroxide, J.
Cell Biol., 77, 837, 1978.
37. Jiménez, N. et al. Aclar discs: A versatile substrate for routine high-pressure
freezing of mammalian cell monolayers, J. Microsc., 221, 216, 2006.

Cryo-Fixation, Freeze-Substitution, Rehydration and Tokuyasu Cryo-Sectioning 345
CONTENTS

1. PRINCIPLES OF THIS METHOD ................................................................. 348
1.1. Cryo-Fixation ........................................................................................... 348
1.2. Dehydration and Chemical Fixation ......................................................... 348
1.3. Rehydration and Postfixation ................................................................... 348
1.4. Tokuyasu Cryo-Sectioning ....................................................................... 348
1.5. Immunolabelling....................................................................................... 348
1.6. Staining and Methyl Cellulose/Moviol Embedding ................................. 349
3.1. Materials ................................................................................................... 350
3.2. Products .................................................................................................... 350
3.3. Solutions ................................................................................................... 352
4. PROTOCOLS .................................................................................................... 354
4.1. Cultured Mammalian Cells and Tissue, Bacteria and Yeast..................... 354
4.2. Plants, Nematodes, Drosophila................................................................. 355
4.3. Preparation for Tokuyasu Cryo-Sectioning .............................................. 356
5.1. Advantages ............................................................................................... 357
5.2. Disadvantages........................................................................................... 358
8. REFERENCES .................................................................................................. 364


Cryo-fixation has proven to be the best method for the preservation of cellular
ultrastructure.1-10 On the other hand, the open structures obtained with Tokuyasu cryo-
sections in most cases are ideally suitable for efficient immunolabelling.11-14 The
advantages of Tokuyasu cryo-sections are probably related to the absence of dehydration
so that molecules remain in their natural hydrophilic surrounding. Moreover, this method
usually involves weak fixation with low concentrations of aldehydes resulting in a partial
loss of proteins and an increase in the three-dimensional access of antibodies.
For many years, a method was sought that combines cryo-fixation with efficient
immunolabelling. Up until recently, only freeze-substitution and embedding in
methacrylate resins (see Chapter 13) or freeze-drying (see Chapter 15) could be used for
immuno-electron microscopy.15,16 Although cellular morphology is good or even
excellent using the latter methods, the section surface only provides a two-dimensional
access for antibodies, which makes it a difficult task to localise sparse antigens. In the
last years two groups, Slot and colleagues17 and Stierhof and Schwarz,18 have developed
a technique that combines high-pressure freezing with the Tokuyasu immunolabelling
technique. They applied this new method to tissue and cultured mammalian cells, but the
real power of this method is to successfully prepare difficult-to-chemically-fix material,
such as yeast, bacteria, plants, nematodes and insects.
In all cases, the cryo-fixation, freeze-substitution and rehydration method resulted in
good cellular ultrastructure and a high labelling efficiency, which is comparable to that of
the conventional Tokuyasu technique.













1. PRINCIPLES OF THIS METHOD
1.1. Cryo-Fixation
 The sample is cryofixed by high-  See Chapters 5, 6.
pressure freezing.
1.2. Dehydration and Chemical Fixation

 The cryofixed specimen is dehydrated  See Chapter 13.
and chemically fixed during freeze-  Uranyl acetate (UA), glutaraldehyde
substitution. (GA), and osmium tetroxide (OsO4) may be
used as fixatives.
1.3. Rehydration and Postfixation

 After freeze-substitution a rehydration  After completing the freeze-substitution
process is carried out on ice at about process, samples have to be rehydrated in
0oC to 4oC. order to enter the conventional Tokuyasu
cryo-sectioning procedure.
 During rehydration, samples are  Fixation during rehydration is necessary
additionally fixed with glutaraldehyde because the chemical fixation during
at 0oC to 4oC. freeze-substitution turned out to be
insufficient.17,18
1.4. Tokuyasu Cryo-Sectioning

 In the last step, the sample is  See Chapter 19.
impregnated with a cryoprotectant
(e.g., 2.3 M sucrose), frozen on a
specimen pin of the ultramicrotome
and sectioned at around 120oC to  For PVP-sucrose at around – 115oC.
140oC. The sections are picked up,
thawed, and mounted on grids.
1.5. Immunolabelling
 On-section immunogold labelling or  See Chapters 19, 21, 23.
immunofluorescence labelling.

1.6. Staining and Methyl Cellulose/Moviol Embedding

 See Chapters 19, 21, 23.
 UA staining and stabilisation of the
sections with methyl cellulose (EM)
or with Moviol (LM).
1. See Chapters 5, 6.
2. See Chapters 5, 6.
3. See Chapter 13.
5. See Chapter 19.
6. See Chapter 19.
7. See Chapters 19, 21, 23.
8. See Chapter 19, 21.



 Figure 14.1 Flow chart showing the
rehydration hybrid technique. The conven-
tional Tokuyasu cryo-section labelling
technique is in grey.



3.1. Materials
 Freeze-substitution apparatus with  Self-made.19
inset to hold glass vials and micro- Leica AFS or AFS2, Leica, Microsystems,
tubes. Vienna, Austria.
RMC FS-7500, Boeckeler Instruments Inc.
Tucson, Arizona, USA (no longer sold).
 Microtubes.  0.5 and 1.5 mL Sarstedt AG & Co,

Nümbrecht, Germany.
 Glass vials.  Used to prepare the solutions.
 Perforated insets.  Leica plastic capsules D5 × H 15 mm #

16702738; Leica, Microsystems, Vienna,
Austria.
 Forceps with insulation coating.  For example, Dumont e.g., (#2a, #3, #4,
#4N, #7), Dumont & fils, Switzerland.
 Rotating wheel.  Used for sucrose impregnation at 4oC.

 Stubs/pins.  Copper or aluminium rivets, used as
specimen holders.
3.2. Products
 Acetone.  Dry acetone (Merck #1.00299.0500,
Merck KGaA, Darmstadt, Germany); the
closed container above silica gel or
molecular sieve (3Å) is added to the
solution.
 Ethanol.  Dry ethanol (Merck #1.00990.0500); the

closed container above silica gel or a
solution.
 PIPES.  1,4-Piperazine-bis(ethanesulfonic acid).

 HEPES.  2-[4-(2-Hydroxyethyl)-1-piperazine] eth-

anesulfonic acid.
 Magnesium chloride.  MgCl2.

 EGTA.  Ethyleneglycol-bis(β-aminoethyl)-N,N,
N´,N´-tetracetic acid
 Sucrose.
 Gelatine.
 Glutaraldehyde (GA).  GA is available as a 10% solution in

water-free acetone (EMS # 16530), ethanol
(EMS # 16531) or methanol (EMS #
16532).
 50% or 25% solution in water (Sigma,
#G-5882; Sigma-Aldrich: Fluka, Buchs,
Switzerland) (keep frozen).
 8% solution, Polysciences, # 00216A.
 Glycine.
 Methanol.  Dry methanol (Merck # 1.06012.0500);
the open bottle is stored under nitrogen gas
in a closed container above silica gel or a
solution.
 Osmium tetroxide (OsO4).  EMS, Hatfield, Pennsylvenia, USA.
 Uranyl acetate (UA).  EMS, Hatfield, Pennsylvenia, USA or

SPI Supplies, West Chester, Pennsylvenia,,
USA.
 H2O.  Double distilled or MilliQ water.
 Polyvinylpyrrolidone (PVP, MW  Sigma PVP10-100G.

10,000).
 Sodium chloride.  NaCl.
 Disodium hydrogenphosphate.  Na2HPO4.
 Sodium dihydrogen phosphate.  NaH2 PO4.

 Silica gel.
 Molecular sieve (3 Å).
 Methyl cellulose.  Sigma M-6385.

3.3. Solutions
 PBS buffer (pH 7.4)  Phosphate buffered saline, a standard
 137 mM sodium chloride NaCl buffer.
 2.7 mM potassium chloride KCl  Note: Phosphate buffers are income-
 8.1 mM disodium hydrogenphosphate patible with uranyl acetate based substi-
Na2HPO4 tution media.
 1.5 mM sodium dihydrogen phosphate
NaH2 PO4
 PHEM buffer (pH 6.9)  Cytoskeleton buffer.20

 60 mM PIPES  Note: PHEM is compatible with uranyl
 25 mM HEPES acetate-based substitution media.
 10 mM EGTA
 2 mM MgCl2
 Membrane contrast medium17  Medium for bacteria, yeast, and cultured

 Acetone mammalian cells and tissue.
+ 15% (v/v) H2O21  Used to enhance membrane contrast,
depen-ding on the sample 1 to 5% water
can be added.
+ 0.10.2% (w/v) UA  Directly dissolved in dry acetone.
 Note: Uranyl ions can impair immuno-
labelling.
+ 0.5% (v/v) GA, or 12% (w/v) OsO4  GA from 10% stock in acetone.
 Membrane contrast medium18  Medium for plants, nematodes and

 Acetone Drosophila.
+ 24% (v/v) H2O21  Used to enhance membrane contrast,
depending on the sample, 2 to 4% water can
be added.
+ 0.10.2% (w/v) UA  From 20% UA stock in methanol.22
+ 0.5% (v/v) GA  From 10% stock in acetone.

+ 0.1% (w/v) OsO4  When OsO4 is omitted, we use 0.5%
UA.

 Postfixation.
 0.25% GA in H2O18
 Rehydration solution17
 95 to70%: Acetone in H2O + 0.5%
GA
 50 to 30%: Acetone in PHEM buffer  Note: To make a glutaraldehyde solution
+ 0.5% GA in 30% acetone in PHEM, 8% aqueous
 0.5% GA in PHEM buffer glutaraldehyde must be used, otherwise the
acetone-based GA precipitates.
 Rehydration solution18
 80 to 10%: Acetone in H2O + 0.5 to
0.25 % GA
 Then H2O + 0.25% GA
 10%12% (w/v) gelatine in PHEM.  Suspend in phosphate buffered saline

(PBS), warm up to about 60°C, stir until the
gelatine powder has dissolved.
 2.3 M sucrose in PHEM buffer.  Weigh 78.73 g sucrose in a 100 mL

volumetric flask. Add PHEM buffer pH:
7.4 while stirring and wait until the sucrose
has dissolved. Remove the stir bar and top
up with the buffer to the mark. The mixture
is stored frozen in 10 to 20 mL aliquots.
 Polyvinylpyrrolidone (PVP)-sucrose23  Infiltration time is longer compared to

 20% PVP pure sucrose.
 1.84 M sucrose  To make 100 mL, mix 20 g PVP and
in phosphate carbonate buffer 4 mL of 1.1 M disodium carbonate
(Na2CO3), in 0.1 M disodium hydrogen
phosphate (Na2HPO4) and 80 mL of 2.3 M
sucrose in 0.1 M Na2HPO4.
 Inactivation buffer for fixative (GA)  Weigh 38 mg glycine in 10 mL H2O.

50 mM glycine in H2O

4. PROTOCOLS
4.1. Cultured Mammalian Cells and Tissue, Bacteria and Yeast
1. Cryo-fixation  See Chapters 5, 6.

 Embedding of the cell pellet in 2%
gelatine before cryo-fixation may prevent
loss of cells during freeze-substitution.
2. Freeze-Substitution
 FS medium
 Acetone
 + 1 to 5% H2O  Addition of water may improve the
 + 0.1 to 0.2% UA visualisation of membrane bilayers.21
+ 0.5% GA 90°C, 48 h
60°C, 8 h  Increase with 2°C/h to 60°C.
Instead of GA, 1 to 2% 30°C, 8 h  Increase with 2°C/h to 30°C.
OsO4 can be used or
GA and OsO4
combined.
 4 rinses 30°C  This step is only done when the
 In freeze-substitution 4 x 10 min substitution medium contains UA.
solution without UA
 Warming up 0°C
 Incubation 0°C, 1 h
3. Rehydration and postfixation
 Rehydration 0°C  In case of OsO4 fixation, rehydration is
 95% acetone in H2O 10 min done with acetone only.
+ 0.5% GA
 90% acetone in H2O 0°C
+ 0.5% GA 10 min
+ 0.5% GA 10 min
+ 0.5% GA 10 min
 50% acetone in 0°C
PHEM buffer 10 min
+ 0.5% GA
 30% acetone in 0°C Note: 8% aqueous glutaraldehyde must be
PHEM buffer 10 min used!
+ 0.5% GA

 PHEM buffer 0°C, 10 min

+ 0.5% GA
+ 0.5% GA
 Preparation for Tokuyasu cryo-sec-  See Section 4.3.

tioning.
4.2. Plants, Nematodes, Drosophila

1. Cryo-fixation  For details, see Chapters 5, 6.
2. Freeze-Substitution
 FS medium
Acetone
 + 2% H2O  Water is required for improved
 + 0.1% OsO4 visualisation of the structure of membrane
 + 0.1-0.2% UA bilayers (plants, Drosophila; for nematodes,
 + 0.5% GA 4% H2O may be better).18
90°C, 48 to 72 h
 When OsO4 is omitted, we use 0.5%
UA.
 Washing
Acetone
 + 2% H2O 4 to 5 × 30 to 45 min
 + 0.5% GA 35°C
 Warming up 20°C, 10 to 60 min
 Warming up 0°C
3. Rehydration and postfixation

 Rehydration  Note: If a buffer is used instead of H2O,
 80% Acetone in H2O 5 to 10 min it should be compatible with UA, e.g.,
+ 0.5% GA 0°C PHEM buffer, to prevent precipitation of
 50% Acetone in H2O uranyl ions.
+ 0.38% GA 5 to 10 min
0°C

 25% Acetone in H2O 5 to 10 min

+ 0.25% GA 0°C
 10% Acetone in H2O 5 to 10 min
+ 0.25% GA 0°C
 H2O + 0.25% GA 5 min
0°C
 Postfixation  Necessary for stabilisation of ultra-

 H2O + 0.25% GA 30 to 90 min structure
0°C  Postfixation time can vary, depending on
sample type and size, fixation cocktail
during freeze-substitution and on the
antigen to be localised.18
 Inactivation of residual fixative mole-  Optional.

cules (GA)
 H2O 5 min, 0°C
 50 mM glycine in 2 × 10 min
H2O 0°C
 Preparation for Tokuyasu cryo-  See Section 4.3.

sectioning
4.3. Preparation for Tokuyasu Cryo-Sectioning
1. Gelatine embedding  To improve sectioning properties.
 Cell pellets and cells grown on filters  The gelatine density should match the
or beads17,24 are embedded in gelatine sample density; we usually use 10 to 12%.
in PHEM buffer and the gelatine is
 The pellet should be loose. In dense
solidified on ice.
pellets, the cells are closely packed and
 Small blocks of not more than 1 mm3 individual cells are difficult to photograph.
are cut with razor blades from the Often dense packing also impairs proper
gelatine-embedded cells. sectioning.
2. Sucrose impregnation  To prevent freezing artefacts and

improve sectioning properties.

 The small blocks of tissue or gelatine  After successful impregnation, the

embedded cells are put into a blocks float in the solution or sink to the
microtube containing 2.3 M sucrose in bottom. Under no circumstances should
PHEM buffer. The tube is mounted on they swim.
a rotating wheel and the tubes are  Infiltration in several steps, e.g., 0.7 M,
slowly rotated. 1.4 M, 2.3 M sucrose, may reduce possible
 Impregnation is done for 4 to 20 hours osmotic effects, such as vacuole collapse
in a cold room at 4°C. and shrinkage of highly vacuolated plant
tissue.
 Alternatively, PVP-sucrose may be used
if gelatine embedding was omitted. PVP
does not permeate cells and may reduce
differences in matrix densities that can
impair the sectioning process.23
3. Mounting the sample  For details, see Chapter 19.

 The gelatine or tissue blocks are
mounted on stubs, frozen and inserted
in the specimen holder of the cryo-
ultramicrotome.
4. Tokuyasu cryo-sectioning.  For details, see Chapter 19.
5. Immunolabelling.  For details, see Chapters 21, 23.
6. Staining and methyl cellulose or  For details, see Chapters 19, 21, 23.
Moviol embedding.
5.1. Advantages
 The rehydration hybrid technique  See Chapters 2, 5, 6, 13, 19.
combines the excellent preservation of  This method can also be used for in situ
cellular ultrastructure by cryo-fixation hybridisation (see Chapter 18).
with the high labelling efficiency of
thawed cryo-section labelling accord-
ing to Tokuyasu.
 Chemical fixation during freeze-  Suitable for fixation-sensitive antigens.

substitution causes fewer artefacts than
when carried out at temperatures 
above 0°C. 

 Osmium tetroxide and uranyl acetate  The chemical reaction of osmium

may be used as primary fixatives tetroxide at low temperature is different
during freeze-substitution than at ambient temperature. At low
temperatures, it works as a fixative and
displays little if any proteolytic activity
(See Chapter 13).
5.2. Disadvantages
 During freeze-substitution, the spec-  In contrast to the original Tokuyasu
imen is exposed to inorganic solvents. technique.
 Limited sample size.  To prevent ice crystal damage during

cryo-fixation.
 More time consuming when compared  Due to the additional freeze-substitution

to the original Tokuyasu procedure. and rehydration steps.
 Requires freezing apparatus (e.g., a

HPF). 

 Possible artefacts due to rehydration 
(e.g., extraction). 

 Sometimes inhomogeneous appear-
ance of cytoplasm.
1. If cellular ultrastructure cannot be  For example, in the case of bacteria,

preserved when using chemical yeast, plants, nematodes and insects.
fixation needed for the Tokuyasu cryo-
sectioning technique.
2. If conventional fixation at ambient  Due to inactivation of antigens by
temperature is not useful. fixation at ambient temperature.
 Due to artefacts caused by slow fixative
penetration (due to large sample size or
dense cell walls, cuticles, etc.).
3. If cryo-fixation is necessary but resin  For example, in case of sparse antigens.

section labelling does not work.

7. OBSERVED RESULTS
 Figure on the Chapter’s title page  Immunogold localisation of the vacuolar

ATPase in a thawed cryo-section of an
Arabidopsis thaliana pollen grain after
high-pressure freezing, freeze-substitution,
rehydration and further processing for
Tokuyasu cryo-section labelling.
 Figure 14.2  Mouse hip cartilage was cryo-fixed by

high-pressure freezing. Freeze-substitution
and rehydration was done as described in
Section 4.1. The substitution medium
contained 0.5% glutaraldehyde, 0.2%
uranyl acetate and 1% H2O. The rehydrated
sample was further processed for Tokuyasu
cryo-sectioning (see Section 4.3.) and the
sections labelled for COPII.
Bar = 500 nm
 Figure 14.3  Saccharomyces cerevisiae was cryo-

fixed by high-pressure freezing. Freeze-
substitution and rehydration was done as
described in Section 4.1. The substitution
medium contained 2% osmium tetroxide,
0.2% uranyl acetate and 5% H2O in
acetone. The rehydrated sample was further
processed for Tokuyasu cryo-sectioning
(see Section 4.3.).
Bar = 500 nm


 Figure 14.4  Immunogold labelling of a thawed cryo-

section of an Arabidopsis thaliana embryo
after high-pressure freezing, freeze-
substitution, rehydration and further
processing for Tokuyasu cryo-section
labelling.
Freeze-substitution was carried out in
acetone containing 0.5% glutaraldehyde,
0.2% uranyl acetate, 0.1% osmium
tetroxide and 2% H2O, rehydration was
done in the presence of 0.5%
glutaraldehyde, and postfixation with
0.25% glutaraldehyde in H2O was carried
out for 60 min at 0°C.
Ultrathin thawed cryo-sections were
labelled with rabbit antibodies and silver-
enhanced goat antirabbit IgG coupled to
Nanogold.
 DV = Dense vesicle
 G = Golgi stack
 SV = Storage vacuole
Bar = 250 nm
 Figure 14.5
 Immunogold labelling of a thawed cryo-
section of the nematode Pristionchus
pacificus, after high-pressure freezing,
freeze-substitution, rehydration and
Tokuyasu cryo-section labelling. Freeze-
substitution was carried out in acetone
containing 0.5% glutaraldehyde, 0.2%
uranyl acetate, 0.1% osmium tetroxide and
4% H2O. Rehydration was done in the
presence of 0.5% glutaraldehyde, and post-
fixation with 0.25% glutaraldehyde in H2O
was carried out for 90 min at 0°C.
A microtubule bundle close to the pharynx
region was labelled with mouse antitubulin
antibodies and silver-enhanced goat anti-
mouse IgG coupled to Nanogold.®
 N = Nucleus
Bar = 1 µm


8. REFERENCES
1. Dubochet, J. et al. Cryo-electron microscopy of vitrified biological specimens,

Trends Biochem. Sci., 6, 143, 1985.
Ann. NY Acad. Sci., 85, 689, 1960.
3. Knoll, G., Braun, C., and Plattner, H. Quenched flow analysis of exocytosis in
Paramecium cells: Time course, changes in membrane structure, and calcium
requirements revealed after rapid mixing and rapid freezing of intact cells, J. Cell
Biol., 113, 1295, 1991.
4. McDonald, K.L. High pressure freezing for preservation of high resolution fine
structure and antigenicity for immunolabelling, Methods Mol. Biol., 117, 77, 1999.
5. Menco, B.P.M. A survey of ultra-rapid cryofixation methods with particular
emphasis on applications to freeze-fracturing, freeze-etching, and freeze-
substitution, J. Electron Microsc. Tech., 4, 177, 1986.
6. Moor, H. Theory and practice of high pressure freezing, in Cryotechniques in
Biological Electron Micoscopy, Steinbrecht, R.A. and Zierold, K., eds., Springer-
Verlag, Berlin, Heidelberg, Germany, 1987, 175.
7. Müller, M. The integrating power of cryofixation-based electron microscopy in
biology, Acta Microsc., 1, 37, 1992.
9. Studer, D., Hennecke, H., and Müller, M. High-pressure freezing of soybean
nodules leads to an improved preservation of ultrastructure, Planta, 188, 155, 1992.
10. Szczesny, P.J., Walther, P., and Müller, M. Light damage in rod outer segments:
The effects of fixation on ultrastructural alterations, Curr. Eye Res., 15, 807, 1996.
11. Liou, W., Geuze, H.J., and Slot, J.W. Improving structural integrity of
cryosections for immunogold labeling, Histochem. Cell Biol., 106, 41, 1996.
12. Tokuyasu, K.T. A technique for ultracryotomy of cell suspensions and tissues, J.
Cell Biol., 57, 551, 1973.
13. Tokuyasu, K.T. A study of positive staining of ultrathin frozen sections, J.
Ultrastruct. Res., 63, 287, 1978.
14. Tokuyasu, K.T. Cryosections for immunohistochemistry, J. Electron Microsc., 35,
1977, 1986.
research: A preview., in Immunoelectron Microscopy in Virus Diagnosis and
1993, 97.
17. Van Donselaar, E. et al. Immunogold labeling of cryo-sections from high-pressure
frozen cells, Traffic, 8, 471, 2007.
18. Ripper, D., Schwarz, H., and Stierhof, Y.-D. Cryo-section immunolabelling of
difficult to preserve specimens: Advantages of cryofixation, freeze-substitution and
rehydration, Biol. Cell, 100, 109, 2008.

19. Humbel, B. and Müller, M. Freeze substitution and low temperature embedding,
in The Science of Biological Specimen Preparation 1985, Müller, M., et al., eds.,
SEM Inc., AMF O'Hare, IL, USA, 1986, 175.
20. Schliwa, M., van Blerkom, J., and Porter, K.R. Stabilization of the cytoplasmic
ground substance in detergent-opened cells and a structural and biochemical
analysis of its composition, Proc. Nat. Acad. Sci. USA, 78, 4329, 1981.
The visibility of biological membranes is improved when the substitution medium
22. Hawes, P. et al. Rapid freeze-substitution preserves membranes in high-pressure
frozen tissue culture cells, J. Microsc., 226, 182, 2007.
23. Tokuyasu, K.T. Use of poly(vinylpyrrolidone) and poly(vinyl alcohol) for
cryoultramicrotomy, Histochem. J., 21, 163, 1989.
24. Zeuschner, D. et al. Immuno-electron tomography of ER exit sites reveals the
existence of free COPII-coated transport carriers, Nat. Cell Biol., 8, 377, 2006.

Freeze-Drying and Embedding of Biological Material 369
CONTENTS

1.1. Check Freeze-Drying Apparatus CFD...................................................... 371
1.2. Cryo-Fixation ........................................................................................... 371
1.3. Preparation of Small Samples on the Object Table .................................. 371
1.4. Automatic Freeze-Drying (FD) ................................................................ 372
1.5. Optional: Osmication of Freeze-Dried Specimens ................................... 372
1.6. Infiltration with Resin............................................................................... 373
1.7. Polymerization.......................................................................................... 374
1.7.1. UV polymerization at low temperature....................................... 374
1.7.2. Heat polymerization.................................................................... 374
1.8. Ultramicrotomy and Analysis of Ultrathin Sections................................. 374
3.1. Materials ................................................................................................... 375
3.2. Products .................................................................................................... 375
3.3. Solutions ................................................................................................... 376
4. PROTOCOLS .................................................................................................... 376
5.1. Advantages ............................................................................................... 379
5.2. Disadvantages........................................................................................... 379
8. REFERENCES .................................................................................................. 387


Freeze-drying (FD) is a technique in which the frozen water of a cryofixed specimen is

removed by sublimation at low temperature in a vacuum chamber. Any contact with an
organic solvent is avoided during the dehydration process. The dry specimen may be resin
embedded at low or high temperature and used for conventional thin sectioning at room
temperature. The sections may be used for electron microscopic detection of
ultrastructural features of native biological material that are not preserved in specimens
obtained by other embedding and sectioning techniques. Note that the quality of structure
preservation after freeze-drying is highly dependent on the quality of cryo-fixation and on
the physiological state of the biological material. Even mildly chemically fixed (dead)
cells are differently freeze-dried than undisturbed living cells.5 The technique as
performed with the Leica EM cryo-sorption freeze-drying (CFD) System is presented.
1.1. Check Freeze-Drying Apparatus CFD

Prepare CFD according to the instructions  This step requires about two hours and
of the manufacturer. It is placed in the neck can be done one day before starting freeze-
of a Dewar flask filled with liquid nitrogen drying.
(LN2) and evacuated to a pressure of at least
10-3 mbar.
1.2. Cryo-Fixation
Cryo-fix biological material, e.g., by  See Chapters 2-6 in part I, Cryo-Fixation
metal-mirror freezing (MMF) or by high- Methods
pressure freezing (HPF).
1.3. Preparation of Small Samples on the Object Table

Under the control of a stereo light  At least one side of a specimen should
microscope, prepare and transfer small not exceed 0.3 mm.
frozen samples to the cold object table (OT)  Small samples are important for
of the CFD apparatus. minimizing freeze-drying times, for
uniform drying at low temperatures and for
easy resin infiltration.

 Small specimens (S) are prepared on a

metal plate (MP) using a precooled scalpel
(SC) and transferred into small metal (MC)
or plastic containers (PC).
 Plastic containers (PC) are used if

drying, infiltration and UV polymerization
are carried out without any further manual
transfer of specimens until final embedding.
Figure 15.1 Object table (OT) cooled by

liquid nitrogen in a Styrofoam box (SB).
1.4. Automatic Freeze-Drying (FD)

1 Vent the vacuum chamber of the CFD  The open vacuum chamber is always
by pure nitrogen gas (GN2) (TF button flooded with GN2.
of the CFD) and open the chamber.  During transfer, water molecules from
2 Transfer the object table into the CFD the air may condense at the cold surface of
and fix it with a built-in cold finger the object table. This condensed water will
(cold trap). sublimate during freeze-drying.
3 Close the vacuum chamber and
evacuate it.  The vacuum chamber of the CFD is
4 Start automatic temperature controlled evacuated in two steps:
freeze-drying. 1. To a pressure of about 15 mbar by
means of a membrane pump and
2. To a pressure of about 10-4 mbar by a
built-in cryo-sorption pump.
The extremely low water vapor pressure
necessary for successful freeze-drying at
low temperature is provided by built-in
cold traps.2,8
1.5. Optional: Osmication of Freeze-Dried Specimens

Osmication is performed outside the CFD  Osmicated specimens are usually not
under a fume hood: suited for low-temperature embedding
(LTE) in Lowicryl because of insufficient
1. Warm up object table a few degrees penetration of UV light through black
above room temperature (RT) inside
specimens. Osmication of freeze-dried
the vacuum chamber. specimens may be performed before
2. Vent vacuum chamber, remove metal embedding in epoxy resins.
plate (MP) (see Figure 15.1), close
object table with a glass plate (GP) and
transfer object table to a fume hood
(see Figure 15.2).

3. Open object table, place a small  Osmication directly after freeze-drying

container with an osmium tetroxide is not necessary to obtain excellent
(OsO4) crystal on the table and close it
structure preservation (See Section 1.8. and
for some time, e.g., our hour. 7, Observed Results).
 The temperature of the specimens
4. Remove the container with OsO4 and
must be higher than RT in Step 3 in order
put drops of pure resin, e.g., Spurr’s
to avoid adsorption of water to the
resin, into container with osmicated
hygroscopic freeze-dried specimens.
specimens.
COS = Container with an OsO4 crystal

GP = Glass plate
OT = Object table
Figure 15.2 Object table (OT) tightly

closed with a glass plate (GP).
1.6. Infiltration with Resin

1. Vent and open vacuum chamber,  The resin or a component of the resin
remove metal plate (MP) (see Figure with which infiltration starts must be liquid
15.1) from object table (OT) inside the at the given object table temperature. For
chamber. example, pure Lowicryl HM 20 can be
used at an object table temperature of –
2. By using a syringe with a metal
40oC. When final embedding is carried out
needle, introduce small amounts of
with Spurr’s resin or EPON, the
embedding medium, e.g., 0.1 mL, into
infiltration may start at an object table
the containers with the dried
temperature of –20oC with a 1:1 DER-ERL
specimens.
mixture (two components of Spurr’s resin
3. Vacuum infiltration of small freeze- or a 1:1 propylenoxide/Epon mixture,
dried specimens is not recommended respectively).
for most biological material.5
 By introducing 0.1 mL of embedding
medium, the liquid will be precooled during
flow through the needle of the syringe
inside the chamber, which is constantly
flooded with cold GN2.

1.7. Polymerization
1.7.1. UV polymerization at low temperature
1. Completely fill plastic containers (PC)  Plastic containers are used for LTE;
on the object table inside the CFD metal containers are used for heat
chamber with resin (e.g., Lowicryl polymerization (see Section 1.7.2).
HM 20).
 Different Lowicryls can be used for
2. Close object table by a UV translucent embedding and UV polymerization at low
glass plate (GP), see Figure 15.2. temperature. (See Chapter 13, Figures
15.3.9 and 15.3.10, and References 2,5,6)
3. Place a UV lamp above the glass plate
thereby closing the vacuum chamber  A Lowicryl K11M/HM20 mixture,
(not evacuated). supplemented with 0.3% uranyl acetate,
may be used to stabilize the specimen
4. After a few hours the infiltration of the during low temperature embedding, (see
specimens may be complete and the Section 4, protocol P10; Reference 6 and
UV lamp is switched on. Section 7, Figure 15.3.10).
1.7.2. Heat polymerization

Warm up object table with metal  Transfer of the small specimens has to
container a few degrees above RT (see be carried out under the control of a stereo
Section 1.5.) and remove it from the light microscope.
CFD.
1. Transfer small metal container into  The embedding moulds should have a
larger containers with pure resin. glossy bottom in order to see the very small
specimens in the polymerized blocks under
2. After sufficient time of infiltration (a a stereo light microscope. A small piece of
few hours) transfer specimens into aluminum foil may be placed into
embedding moulds with fresh pure embedding moulds with a rough surface.
resin and polymerize at about 60°C. The foil may be removed after
polymerization.
1.8. Ultramicrotomy and Analysis of Ultrathin Sections

See Section 7, Observed Results.  For analysis of mobile ions, such as
sodium Na+ and potassium K+, prepare
dry-cut sections (see Section 7, Figure
15.3.2 and References 2-4).
 Lowicryl embedded specimens may be

very labile and difficult to cut (see Section
7, Figure 15.3.9 and References 5,6). In this
case the preparation can be stabilized by
exposing the entire polymerized blocks to
OsO4 vapor for, e.g., 30 min (see Section 4,
protocol P9).

1. Cryo-fixation. 
2. Transfer of small frozen samples on

cold object table.
3. Transfer of cold object table into

freeze-drying apparatus CFD and
automatic temperature controlled
freeze-drying (FD).
4. Infiltration of freeze-dried speci-

mens with resin.
5. Polymerization.

3.1. Materials
 Equipment for cryo-fixation  See Chapters in ACryo-Fixation
Methods.
 Freeze-drying equipment, e.g., Leica  Other freeze-drying units suited for long-
EM CFD (Leica Microsystems, term freeze-drying at low temperatures
Vienna, Austria) with UV lamp. down to 140ºC are no longer7
commercially available.
 Homemade freeze-drying equipments,
including adapted freeze-etch units, are
mentioned in Reference 10.
 Stereo light microscope 
 LN2 Dewar 
 Oven for heat polymerization 
 Scalpel 
 Tweezers 
3.2. Products
 Liquid nitrogen (LN2).
 Resins for electron microscopy:  Spurr’s resin is mostly used due to its
 Epoxy resins, e.g., Epon, Araldite, low viscosity that enables easy infiltration.
Spurr’s resin. For resins used, see Section 7, Observed
Results.
 Acrylic resins, e.g., Lowicryl HM 20,
HM23, K4M, K11M

 Uranyl acetate. 
 Osmium tetroxide.  OsO4.


3.3. Solutions
 Spurr’s resin.  The embedding media are prepared
 Epon. according to the manufacturer’s instruct-
tions.
 Lowicryl  See Chapter 13.
 HM20
 K11M
4. PROTOCOLS
A general protocol for freeze-drying (FD)  These parameters are, e.g.:

of biological material cannot be given  Quality of cryo-fixation
because of the many parameters that  Heat transfer between specimen and
determine the freeze-drying process which object table in the vacuum chamber
are different for different biological  Size of specimen
material in different physiological states.  Degree of hydration of the specimen
 Distribution and structure of macro-
molecules
 Distribution of ions
 Interaction of water with macro-
In general, long freeze-drying times molecules and ions.
between 100ºC and 80ºC and long 
freeze-drying times between 80ºC and
50ºC are essential in order to avoid
severe shrinkage artefacts.2
Freeze-drying Protocols (P1P8) used for  The water of chemically fixed biological
the different biological materials shown material is less attracted by cellular proteins
under Observed Results are listed below. than the water of living cells. Therefore,
freeze-drying of chemically fixed material
Note: Infiltration with resin (see Section is faster than freeze-drying of native
1.6) as a rule starts at lower temperatures material.5
than the final temperatures of the freeze-
drying protocols. Specimens embedded in
Lowicryls were infiltrated at 25°C for one
day and polymerized for one day at 25°C.
Protocol P1. FD
 Striated frog muscle, (see Section 7,
 3 days at 80ºC Figures 15.3.1, 15.3.2, 15.3.4) freeze-dried
 6 days at 60ºC in a homemade, cryo-sorption freeze-drying
unit.

Protocol P2. FD
 Temperature increase 1ºC h-1 from  Rat kidney, see Section 7, Figure
125ºC to 50ºC (75 h) 15.3.5 AB.
 90 h at 50ºC
 1ºC h-1 from 50ºC to +25ºC (25 h)
 24 h at +25ºC
 Total time: 214 h ~ 9 days
Protocol P3. FD
 Temperature increase 0.2ºC h-1 from  Rat liver, see Section 7, Figure 15.3.6
90ºC to 30ºC (300 h) and the figure on the Chapter’s title page.
 Temperature increase 1ºC h-1 from
20ºC to 10ºC (20 h)
 10 h at 10ºC
Protocol P4. FD
 Temperature increase 25ºC h-1 from  Dendritic cells, see Section 7, Figure
140ºC to 90ºC (2 h) 15.3.7 and Reference 9.
 Temperature increase 0.4ºC h-1 from
90ºC to 50ºC (100 h)
50ºC to 0ºC (50 h)
 24 h 0°C
Protocol P5. FD
 Temperature increase 50ºC h-1 from  Human platelets, see Section 7, Figure
140ºC to 90ºC (1 h) 15.3.8 and Reference 5.
 Temperature increase 0.27ºC h-1 from
90ºC to 50ºC (150 h)
50ºC to 0ºC (50 h)
 24 h 0°C
 Total time: 225 h ~ 9.5 days

Protocol P6. FD
 24 h 100ºC  Human lymphocytes, see Section 7,

 Temperature increase 0.9ºC h-1 from Figure 15.3.9 and Reference 6.
100ºC to 10ºC (100 h)
 10 h 10°C  After freeze-drying and LTE in Lowicryl
 Total time: 134 h ~ 5.6 days HM20, the polymerized block has been
stabilized by OsO4 vapor, see P9.
Protocol P7. FD
 Temperature increase 10ºC h-1 from  Jurkat cells, see Section 7, Figure
160ºC to 90ºC (7 h) 15.3.10 and Reference 5. During embed-
 32 h 90ºC, temperature increase ding in Lowicryl, the freeze-dried cells
0.3ºC h-1 from 90ºC to 0ºC (300 h) have been stabilized by uranyl acetate, see
 48 h 0°C P10.
Protocol P8. FD
 Temperature increase 1ºC h-1 from  This freeze-drying procedure has been
140ºC to 70ºC (7 h) called Molecular Distillation Drying.7
-1
 Temperature increase 10ºC h from After freeze-drying and exposure to OsO4,
70ºC to +25ºC (9.5 h) vapor, specimens were embedded in
 48 h at +25°C Spurr’s resin. The rather short period of
 Total time: 127.5 h ~ 5.3 days freeze-drying at low temperature has been
criticised.5,8
Protocol P9. Osmication after freeze-

drying and LTE in Lowicryl
 Place a polymerized block for about  As a rule the block is first sectioned until
30 min into a 0.5 mL Eppendorf the embedded biological material appears at
capsule together with a small amount the surface. After exposure to OsO4 vapor,
(< 0.1 g) of OsO4 crystals. the block appears dark brown and turns
black after a few hours without further
exposure.6
Protocol P10. Fixation with uranyl

acetate during LTE
 Dissolve overnight 0.4% (w/v) uranyl  Uranyl acetate could not be dissolved in
acetate in Lowicryl K11M. pure HM20.
 Mix 3 parts of this solution with 1 part  See Section 7, Figure 15.3.10 and
of pure HM 20. Reference 6.
 Use this mixture with 0.3% uranyl
acetate for complete LTE.

5.1. Advantages
 No chemical fixatives are necessary.  Exceptions are described.
 No organic solvent during
dehydration.
 Only minute amounts of embedding 

media are needed.
 Retention of mobile macromolecules  See References 2-5.

and mobile ions (Na+, K+).
 Preservation of selective ion adsorp-  See Section 7, Figure 15.3.4 and

tion to cellular proteins. Reference 4.
 Preservation of cellular ultrastructure  Precondition: excellent cryo-fixation, see

and molecular conformation of Section 7, Figures 15.3.6, 15.3.10 and
proteins. Reference5
 Preservation of antigenicity.  See References 5,9.
5.2. Disadvantages
 Highly dependent on quality of cryo-  Not necessarily a disadvantage.
fixation.
 Handling of very small specimens. 
 Time consuming.  One to two weeks for small samples.
 Sophisticated freeze-drying equipment  The apparatus may be available through

is necessary. a special order to Leica or possibly other
companies or homemade.

 A general protocol for every kind of  Optimization of protocols is required.
specimen is not available.

The advantages of the freeze-drying 

technique for electron microscopic studies 
are given in Section 5.1 and exemplified by 
Observed Results (see Section 7). From the 
results obtained so far, it appears that the 
interactions between the main components 
of living cells (namely water, proteins and 
ions) are best preserved during and after 
proper freeze-drying because the biological 
material can be dehydrated by sublimation 
without changing the life-like conformation 
of proteins as it takes place during freeze- 
substitution with an organic solvent.  See Detailed Discussion in Reference 5.

As a consequence, mobile ions, such as 
potassium K+, remain associated at or near  See Discussions in References3-5
their original protein sites and layers of
water molecules remain associated at
cytoplasmic protein sites thereby preventing
condensation artifacts. This is concluded
from the findings:
1. That even after long freeze-drying at
rather high temperatures, a con-
siderable amount of water remains in
cryo-fixed native biological material.5  See Section 7, Figures 15.3.6, 15.3.8.
2. That the proteins of the cytoplasm are
very homogeneously distributed. 
The discovery of the unknown ultrastructure 

of cytoplasmic proteins of living cells and 
of their interactions with water and ions is a  See Reference 5.
demanding goal that may be approached by 
using and further developing the freeze-
drying technique for electron microscopy.

7. OBSERVED RESULTS
 Figure on the Chapter’s title page  See Figure 15.3.6.
 Figure 15.3.1 Chemically unfixed frog  The muscle has been freeze-dried for
Sartorius muscle, cryofixed on a LN2 three days at 80˚C followed by six days at
cooled copper block, freeze-dried, 60˚C (see Section 4, Protocol P1.) Shorter
embedded in Spurr’s resin, and section drying times produced severe structural
stained with uranyl acetate and lead artifacts.2
citrate. (From Reference 2, reprinted
with permission.)
 Figure 15.3.2 Dry-cut sections of a  Cellular K+ has been reversibly replaced

thallium Tl+-loaded (A), and of normal with Tl+ in living muscle. The Tl+ contrast
potassium K+-containing (B), cryo- in (A) represents adsorption sites of K+ onto
fixed, freeze-dried and embedded cellular proteins. For confirmation of this
muscle as in Figure 15.3.1. (From claim by independent microanalytical
Reference 1, reprinted with permis- methods, see References 1,2,4.
sion.)
 Figure 15.3.3 Cryo-section (0.1 µm  Severe shrinkage even of ultrathin cryo-

thick) after freeze-drying for 33 h at sections of chemically unfixed biological
100˚C. Cryo-transfer of the dry cold material during freeze-drying can only be
section to a Zeiss EM 902. (From avoided by long drying times (e.g., 33
Reference 8, reprinted with permis- hours) at low temperature.2,8
sion.)
 Figure 15.3.4 Muscle sections stained  In (A) the electron dense Cs+ is adsorbed
with a solution containing 100 mM with a high selectivity to certain cell
lithium chloride LiCl and 10 mM proteins. Selective adsorption of alkali
caesium chloride CsCl. (A) after metal ions like K+ or Cs+ to cellular
freeze-drying and embedding in proteins can only be demonstrated in
Spurr’s resin (see Figure 15.3.1.), (B) freeze-dried preparations, but not in
after glutaraldehyde fixation and chemically fixed preparations (no staining
embedding. (From Reference 3, re- of section (B) nor in protein solutions4).
printed with permission.)

 Figure 15.3.5 Rat kidney, cryo-fixed  Without osmication, the membranes
(Leica MM80), freeze-dried and appear in negative contrast. They are not
embedded in Spurr’s resin. For freeze- visible at the low magnification (A). Black
drying times, see Section 4, membranes are shown in (B).
Protocol P2. Sections stained with
uranyl acetate and lead citrate. (A) No
chemical fixation, N: nucleus; (B) with
osmium tetroxide vapor fixation after
drying (see Protocol P9),
M: mitochondria. (From Reference 5,
reprinted with permission).



 Figure 15.3.6 Rat liver, cryo-fixed  The section was taken from the area of
(Leica MM80), freeze-dried and best cryo-fixation, at a distance less than
embedded in Spurr’s resin (no 0.5 µm from the freezing plane. Membranes
chemical fixative). Section stained are clearly seen in negative contrast.
with uranyl acetate and lead citrate.
For freeze-drying times, see Section 4
(Protocol P3). aER: agranular
endoplasmic reticulum, rER: rough
endoplasmic reticulum, M: mito-
chondrion, P: peroxisome. The area
around the asterisk (*) in the mito-
chondrion (M) is shown at higher
magnification in the inset: Membranes
are oriented perpendicular to the plane
of the section. The two cristae
membranes (CM) are thicker than the
outer and the inner membrane (MM)
of the mitochondrion. (From
Reference 5, reprinted with permis-
sion.)
 Figure 15.3.7 Highly magnified MHC  Note satisfying immunolabeling of

class II compartment from a human freeze-dried specimens embedded in Spurr’s
immature dendritic cell, cryofixed resin.5,8
(Leica MM80), freeze-dried, see
Section 4 (Protocol P4), embedded in
Spurr’s resin. Immunolabeling (10 nm
gold) for the MHC class II protein.
(from Reference 9, reprinted with
permission.)
 Figure 15.3.8 Human platelet, cryo-  The platelet stems from an area distant
fixed (Leica MM80), freeze-dried (see from the freezing plane less than 5 µm.
Section 4 (Protocol P5) and resin Membranes appear in well-defined negative
embedded in Spurr’s resin. contrast (as in Figure 15.3.6). The very
Immunolabeling (10 nm gold) for CLP homogeneous aspect of the cytoplasm is
36, a PDZ-LIM domain protein typical for excellent cryo-fixation.
associated with α-actinin. G: secretory
storage organelles, DTS: dense tubular
system = agranular endoplasmic retic-
ulum. (From Reference 5, reprinted
with permission.)


 Figure 15.3.9 Ultrathin section of  Compression and extraction artifacts

human lymphocytes after freeze- may occur during wet cutting of well-fixed,
drying; see Section 4 (Protocol P6) and freeze-dried and Lowicryl embedded
low temperature embedding in specimens. After exposure of the
Lowicryl HM20 at 25°C. Poly- polymerized block for 30 min to OsO4
merized block exposed to OsO4 vapor vapor, the whole resinbiological material
(see Protocol P9). Uranyl acetate and complex is stabilized and the ultrastructure
lead citrate staining. (From Reference is well preserved.6
6, reprinted with permission.)
 Figure 15.3.10 Jurkat cells, high-  The freeze-dried cells are stabilized
pressure frozen in cellulose during infiltration and UV polymerization
microcapillaries, freeze-dried, see of the Lowicryl mixture by supplemented
Section 4 (Protocol P7) and resin uranyl acetate.
embedded at 25°C in a K11M/HM20
Lowicryl mixture, supplemented with
0.3% uranyl acetate (see Protocol
P10). Uranyl acetate and lead citrate
staining. The asterisk (*) represents the
microcapillary that was used. (From
Reference 5, reprinted with permis-
sion).

8. REFERENCES
1. Edelmann, L. Subcellular distribution of potassium in striated muscles, Scan.

Electron Microsc., 875, 1984.
2. Edelmann, L. Freeze-dried embedded specimens for biological microanalysis, Scan.
Electron Microsc., 1337, 1986.
3. Edelmann, L. Two opposing theories of the cell: Experimental testing by
cryomethods and electron microscopy, in The Science of Biological Specimen
Preparation, Müller, M., et al., eds., Scanning Electron Microscopy., Inc., AMF
O'Hare, Chicago, IL, USA, 1986, 33.
4. Edelmann, L. Basic biological research with the striated muscle by using
cryotechniques and electron microscopy, Physiol. Chem. Phys. Med. NMR, 33, 1,
2001.
5. Edelmann, L. Freeze-dried and resin-embedded biological material is well suited for
ultrastructure research, J. Microsc., 207, 5, 2002.
6. Edelmann, L. and Ruf, A. Freeze-dried human leukocytes stabilized with uranyl
acetate during low temperature embedding or with OsO4 vapor after embedding,
Scan. Microsc. Suppl., 10, 295, 1996.
7. Linner, J.G. et al. A new technique for removal of amorphous phase tissue water
without ice crystal damage: A preparative method for ultrastructural analysis and
immunoelectron microscopy, J. Histochem. Cytochem., 34, 1123, 1986.
8. Sitte, H. et al. A new versatile system for freeze-substitution, freeze-drying and low
temperature embedding of biological specimens, Scan. Microsc. Suppl., 8, 47, 1994.
9. Spehner, D. et al. Embedding in Spurr's resin is a good choice for immunolabelling
after freeze drying as shown with chemically unfixed dendritic cells, J. Microsc.,
207, 1, 2002.
10. Steinbrecht, R. and Müller, M. Freeze substitution and freeze drying, in
Cryotechniques in Biological Electron Microscopy, Steinbrecht, R. and Zierold, K.,
eds., Springer, Berlin, Germany 1987, 149.


Part IV
Freeze-Fracture and Metal

Shadowing

The Shadow of Hydrated Biological Specimens 393
CONTENTS

1. PRINCIPLE OF THE METHOD .................................................................... 396
1.1. Freeze-Drying and Metal Shadowing ....................................................... 396
1.2. Freeze-Fracture and Freeze-Etching......................................................... 398
3.1. Materials ................................................................................................... 400
3.2. Products .................................................................................................... 402
4. PROTOCOLS .................................................................................................... 403
5.1. Advantages ............................................................................................... 407
5.2. Disadvantages........................................................................................... 407
6.1.1. Visualization of fine surface details............................................ 408
6.1.2. Information on small molecules.................................................. 408
6.2.1. Freeze-fracture ............................................................................ 408
6.2.2. Freeze-etching............................................................................. 408
8. REFERENCES .................................................................................................. 410


Freeze-drying followed by unidirectional

shadowing is a technique that enables one
to directly observe the surface topography
of biological objects. The contrast
enhancement is limited to the surface
exposed to the metal while the contrast
contribution from non surface exposed
areas is negligible. The appearance of a
shadow is caused by the metal deposited
under a fixed azimuthal and elevation
angle. This provides images with a quasi 3-
D appearance.
Figure 16.1 Isolated membrane of an

epithelial cell, observed after freeze-drying
and shadowing.
Bar = 200 nm
Freeze-fracture and freeze-etching are two

related techniques. They allow electron
microscopic investigation of cell surfaces
and biological membranes in the frozen
state. The fracturing of the frozen object
produces cross-sectional views of the
structure, but also views inside fractured
membranes and surface views of
organelles. Moor and Mühlethaler,1 in
1963, were the first to fully describe the
yeast cell using the freeze-etching
technique.
Figure 16.2 Yeast spheroplast. Total view

of a freeze-fractured cell after high-pressure
freezing.
Bar = 500 nm

1. PRINCIPLE OF THE METHOD
Shadowing or shadow casting means  Platinum (Pt) is mostly used in

evaporating atoms of a heavy metal from a conjunction with carbon (Pt/C).
point source at an oblique angle to the
specimen.2-6 Metal piles up on surfaces that
face the source, but surfaces facing away
from the source are shielded and receive no
metal. This is the “shadow” that appears  Inverting the contrast will give
white under the electron beam. Given the contoured bright objects illuminated from
angle of shadowing θ and the length of the the side by light and projecting a dark
shadow l, the height of the object h, which shadow.
produces the shadow, can be determined 
simply from the formula:
h = l * tan θ.









Figure 16.3 Isolated viral particles after
shadowing. For θ = 45°, h = l = 27 nm.
Bar = 100 nm



1.1. Freeze-Drying and Metal Shadowing
The goal is to freeze a hydrated suspension 
of particles (virus, membrane, macro-
molecular complexes), then to freezedry
the sample under vacuum and finally to
shadow the surface of the specimen by
metal evaporation.
A sample suspension is applied to carbon-
coated grids and adsorbed. The grids are
washed, blotted to remove excess water,
quick frozen in liquid nitrogen and then
transferred into a freeze-drying/metal
shadowing unit (i.e., a freeze-fracture
apparatus). In the apparatus, samples are
freeze-dried for two hours at 80°C.

Freeze-dried specimens are then shadowed 

with platinum–carbon (Pt/C) at a given 
elevation angle (commonly 45°), followed 
by evaporation of a layer of carbon at an 
elevation angle of 90° to prevent possible 
metal grain redistribution. 






Figure 16.4 Freeze-dried ferritin molecules

unidirectionally shadowed by Pt/C. θ = 45°.
Bar = 50 nm

















Figure 16.5 Freeze-dried ferritin molecules
rotationally shadowed by Pt/C. θ = 45°.
Bar = 50 nm




1.2. Freeze-Fracture and Freeze-Etching

In the freeze-fracture technique, the  A method of specimen preparation for
specimen is frozen and broken into two electron microscopy, in which a replica is
pieces to expose fracture faces. The surface made from a sample that has been rapidly
is immediately imprinted by making a frozen and then fractured along natural
shadow replica. In the freeze-etching planes of weakness, mostly through the
variant, the specimen surface is allowed to hydrophobic part of membranes9 to reveal
freeze-dry slightly in order to remove the its internal structure.
ice surface.7-9 
Bulk biological samples are mounted
between copper plates then frozen either by
quick plunging in liquid propane or by
high-pressure freezing.
For freeze-fracture, the preparations are
immediately shadowed at an elevation
angle of 45° with 2 nm of platinum–carbon
(Pt/C) and backed with a 10 to 20 nm-thick
layer of carbon.
For freeze-etching, after fracturing, the
specimen temperature is raised to 100°C
for about 2 min for etching. The specimen
is then shadowed as above.
After thawing, the replicas are cleaned of
the remaining sample, rinsed in distilled
water and mounted on copper grids.
Figure 16.6 Freeze-fractured microvilli

from a midgut epithelial cell of an insect.
Pt/C, θ = 45°.
Bar = 100 nm


1. Cool the specimen stage of the
freeze-fracture machine to 160°C.
2. Apply a 5 µL drop of suspension on
a collodion carbon-coated grid.
3. Wash with a volatile buffer and/or
distilled water.
4. Blot with filter paper.
5. Freeze in liquid nitrogen.
6. Introduce the specimen into the
high vacuum chamber of the
freeze-fracture machine.
7. Freeze-dry for two hours by
raising the specimen stage to
80°C.
8. Evaporate Pt/C at a given elevation
angle.
9. Evaporate carbon to stabilize the
Pt/C layer at an angle of 90oC.
10. Warm up the specimen in vacuo to
room temperature.

1. Cool the specimen stage of the
freeze-fracture machine to 160°C.
2. Mount the bulk biological sample

between two copper plates.
3. Freeze the sandwich by quick

plunging into liquid propane or
better, if available, by high-
pressure freezing.
4. Introduce the specimen into the

high vacuum chamber of the
freeze-fracture machine.

5. Fracture the specimen:

The fracture can be done by using a
knife or by breaking the sandwich
into two pieces, depending on the
equipment available.
6. Optional step of etching:

Raise the temperature of the
specimen stage to 100°C so as to
allow sublimation of surface ice for 2
min.
7. Evaporate 2 nm of Pt/C at a given

elevation angle (45°).
8. Evaporate at 90° a thick layer

(20 nm) of carbon to stabilize the
Pt/C replica.
9. Thaw the specimen and clean the

replica by digesting the tissue.
Wash the replica in distilled water.
10. Collect the replicas on uncoated

grids.
3.1. Materials
 Equipment for rapid freezing.  See Part I, Cryo-Fixation Methods

 Freeze-fracture/-etching unit.
Figure 16.7 JFD-9000 Freeze-Etching

Equipment from JEOL, Tachikawa, Tokyo,
Japan.


Figure 16.8 Cryofract 190 from

ReichertJung, (now Leica Microsystems,
 Vienna, Austria).

Figure 16.9 BAF 060 from BALTEC,

Balzers, Liechtenstein, (now Leica Micro-
systems, Vienna, Austria).
 .
 
 Evaporating material
Figure 16.10 Carbon tip (left), platinum-

carbon tip (right) and ring-shaped filament,
components of the electron beam
evaporation gun10 of the Jeol JFD-9000.

 Specimen holders
 For unidirectional or rotary shadow  See Figure 16.11.
casting (Jeol) with five magnetic slots
for holding five nickel grids.
 For freeze-fracture or freeze-etching  See Figure 16.12.
(left: Cryofract, right: Jeol).
Figure 16.11 Specimen holder.
 Left: The holder can receive four

sandwiches made with two copper
plates. The sandwiches are broken by
opening the holder.
 Right: The holder can receive one
disk, whose surface has been scratched
for better adherence of the object.

Figure 16.12 Specimen holders.
 Dewar vessels
 Filter paper
 Nickel or copper grids, 300 to 400
mesh
 Platinum loop
 Porcelain spot plates  The depth of the well (6 to 8 mm) should
allow manipulation of replicas.
 Tweezers
 Wide bore pipette
3.2. Products
 2% (w/v) Ammonium acetate
 Carbon
 2% (w/v) Collodion in amyl acetate
 30% (w/v) Chromium oxide
 Distilled water
 Household bleach (10 to 12% sodium
hypochlorite)
 Liquid nitrogen
 Pt/C
 Sample suspension for freeze-drying/
metal shadowing
 70% (v/v) Sulphuric acid

4. PROTOCOLS

1. Cool down the specimen stage of the
freeze-fracture apparatus to approx-
imately 160°C.
2. Grids 
 Nickel or copper grids, 300 to 400 
mesh, covered with a carbon film or 
a carbon collodion film. 
 Glow discharge the grids for 30 sec  To create positive charges.
in low pressure air (2 × 10-1 Torr).
3. Sample suspension
 Spot on the grids 5 to 10 µL of the  It is helpful to examine the sample by
sample suspension at approximately negative staining to determine the right
0.1 to 0.5 mg/mL for 1 min. concentration and time needed for contact
of the sample with the grid.
4. Wash
 3 to 5 times with distilled water or  This step is crucial to avoid salt
with a volatile solvent like precipitation during freeze-drying. If
ammonium acetate followed by buffering is necessary, a freshly made
water. solution of 2% ammonium acetate is most
often used.
5. Blot
 Remove excess liquid with filter  Leave a very thin layer of water.
paper.
6. Freeze
 Quickly plunge the grid into liquid  Liquid ethane and liquid propane may
nitrogen. also be used because they have better
 Mount the grid on the specimen cryogenic properties than liquid nitrogen.
holder maintained at liquid nitrogen
temper-ature.
7. Dry  The freeze-drying process can be

 Introduce the specimen into the high followed by monitoring the pressure of the
vacuum chamber of the freeze- chamber. During water sublimation, the
fracture/freeze-etching device. pressure increases progressively to reach a
 Raise the temperature of the plateau, which corresponds to the freeze-
temperature controlled specimen drying state of the specimen. A simple
stage to 80°C and maintain the alternative is to examine the grid through a
specimen at this temperature for binocular; the specimen is ready when ice is
about two hours. no longer seen.

8. Evaporate Pt/C  The grain size (and ultimate resolution)

of the deposited metal depends on the
 Turn on the high voltage (HV) temperature of the specimen prior to
switch and set the voltage to 3000 V. evaporation. Cooling down the specimen to
100°C or lower will reduce the grain size.5
 Select the Pt/C source.
 Tilt the specimen stage to 45°.  Rotary shadowing is recommended for

studying small objects like macromolecules
(see Figure 16.5). The specimen is mounted
on a revolving holder, tilted under a chosen
angle and rotated during Pt/C evaporation.
 For small particles, DNA, filamentous
proteins, use a low angle (5 to 30°).
 Raise gradually the emission current

to 40 mA.
 Monitor the thickness of the metal  If the device for quartz crystal thickness
with the quartz crystal thickness measurement is not available or is not
monitor. When the evaporation is working properly, the alternative to obtain
complete, return emission to 0 mA. reproducible results is to calibrate the metal
deposition by timing for a given current
emission (i.e., 30 sec for 40 mA emission).
 Return the specimen stage to its
horizontal position.
9. Evaporate C
 Select the C source.  See above.
 Gradually raise the emission current

to 40 mA.
 Monitor the thickness of the C.

When the evaporation is complete,
return emission to 0 mA.
10. Warm up the specimen in vacuo to  Good results are obtained for a Pt/C
room temperature. thickness of 1 nm and a C thickness of
10 nm.


1. Cool down the specimen stage of the 
freeze-fracture apparatus to approx- 
imately 160°C. 

2. Freeze  Freezing is a crucial step for the success
 Mount the bulk biological specimen of the experiment. These methods are
between two copper or aluminum presented and fully documented in Part I,
plates. Cryo-Fixation Methods, of this book (see
Chapters 26).
 Freeze the sandwich by high-pressure
freezing (see Chapters 5, 6).  Instead of making a sandwich, one can
or use one copper disk with a scratched
 Drop or plunge the object into liquid surface for better adherence of the object. A
propane. small droplet is placed on the disc and then
dropped into liquid propane.

 Introduce the specimen into the high
vacuum chamber of the freeze-
fracture/freeze-etching device.
 Mount the object on the precooled  Wait for a brief period until the
specimen stage. temperature is stable.
3. Fracture  Freeze-fracture refers to experiments

 Freeze-fracture: Strike the sample where ice removal is negligible; it is
mounted on one copper disk with the performed at low temperature (usually
cold knife or break the sandwich into 150°C), and Pt/C evaporation is started
two pieces. Then immediately before breaking the specimen.
evaporate the metal (see Step 4).
or
 Freeze-etching: After the fracture  The cold knife held above the specimen
process, the sample stage is warmed is often used as a cold trap. It traps the
up to 100°C, while a cold trap water, which sublimates out of the fracture
(below 150°C) is held over the surface and prevents its redeposition onto
sample. The etching process lasts the fracture surface.
usually between one and two minutes.
4. Evaporate
 Turn on the HV switch and set the
voltage to 3000 V.
 Select the Pt/C source.
 Tilt the specimen stage to 45°.  Most freeze-fracture methods use
 Gradually raise the emission current to unidirectional shadowing at a nominal
40 mA. angle of 45°. However, fine details can be
enhanced by using a lower angle of, e.g., 10
to 30°.

 Rotary shadowing is also an alternative

to reveal fine details in replicas. The
specimen is mounted on a revolving holder,
tilted under a chosen angle and rotated
during Pt/C evaporation.
 Monitor the thickness of the metal  A 2 nm-thick layer of Pt/C provides

with the quartz crystal thickness good results.
monitor. When the evaporation is
complete, return emission to 0 mA.
 Return the specimen stage to its

horizontal position.
 Select the C source.  A 20 nm backing layer of carbon will

strengthen the replica for the cleaning
 Gradually raise the emission current to procedure.
40 mA.
 Monitor the thickness of the C. When

the evaporation is complete, return
emission to 0 mA.
5. Clean  The thin replica needs to be lifted off the

 Remove the specimen from the freeze- sample before it can be viewed in the
fracture device and leave it in distilled electron microscope.
water for thawing.
 Float the replicas on sulphuric acid  The digesting solution should remove
(70%, 12 hours). the sample without harming the replica.
Bleach, sulphuric acid, chromium oxide are
usually used, but there is no universal
recipe. This step can last from hours to
days, depending on the tissue analyzed.
 Transfer to a bath of distilled water for  Transfer from one liquid to the next with
washing (five minutes). a platinum loop or wide-bore pipette.
 Transfer to a bath of household bleach  The major problem is the breakage of

(2 to 12 hours). the replica due to swelling and shrinkage of
 Transfer to bath of distilled water the specimen in the cleaning solutions. A
(20 minutes). device has been described for washing
replicas11 or methods for strengthening the
replica12 have been developed.
 Repeat the previous step until the
bleach is replaced with distilled water.
 Collect the replicas on uncoated 100
to 300 mesh copper grids.

5.1. Advantages
 Shadowing is interesting to examine: 
 Fine order surface details, such as 
membrane proteins, which cannot be
observed in any other way.3,5,6,13
 Small objects of relatively low
electron density because their contour
may not be clearly defined otherwise.
 Freeze-fracture and freeze-etching are 
straightforward methods to obtain fine
structural details within the cell.
Specific details are revealed, based on
the fracturing behavior of each
individual tissue sample.
 In freeze-fracture, the preferential
fracture plane goes through the lipid
bilayer.14 This is widely used for
studying distribution and localization
of integral membrane proteins.15-17

5.2. Disadvantages
 The grain size of Pt/C (2 nm) limits the  The alloy tungsten–tantalum gives finer
resolution. grains than Pt/C (1 nm) but is more difficult
to evaporate.13
 Freeze-drying can induce shrinkage of 
the material.6
 Freeze-etching on model systems has  To reduce these artifacts, rapid freezing
demonstrated that cryo-fixation can procedures and high-pressure freezing
introduce artifacts as a result of slow methods have been proposed (see Chapters
freezing like segregation of dispersed 26).
material from the solvent or of
intramembrane particles in biological
membranes.18
 Fracturing proceeds at its own will. It  The preferential fracture plane goes
14
is not possible to choose a particular through the lipid layer of the membranes.
path for the fracture plane to follow.
 Freeze-fracture apparatuses are expen-
sive.


 Shadowing is of great interest to

examine:
 Small objects of relatively low
electron density because their contour

may not be clearly defined otherwise.
6.1.1. Visualization of fine surface details

 Protruding membrane proteins.
 See Figure 16.1.

 One-sided information on tubular 
crystals to determine the pitch of a 
helical structure. 

 
6.1.2. Information on small molecules
 Rotary shadowing can give informa-  See Figure 16.5.
tion on the oligomeric state of small
proteins. 
 DNA etc. 



 Freeze-fracture and freezeetching are 
straightforward methods to obtain fine
structural details within the cell.
6.2.1. Freeze-fracture
 Freeze-fracture provides a high mag- 
nification view of the physical
relationships between membrane
components (see also Chapter 22).

6.2.2. Freeze-etching
 Freeze-etching is useful to study 
relationships between cytoskeletal
elements within the cell.


7. OBSERVED RESULTS
 Figure on the Chapter’s title page

Isolated membrane of an epithelial cell,  Only one side of the membrane is
observed after freeze-drying and shad- observed. Membrane proteins forming
owing. small aggregates are clearly seen (Pt/C,
θ= 45°, unidirectional shadowing).
 Figure 16.1  See General Introduction.

Both sides of the membrane are
Isolated membrane of an epithelial cell, observed, showing that the membrane has
observed after freeze-drying and shad- an asymmetrical structure (Pt/C, θ = 45°,
owing. unidirectional shadowing).

 Figure 16.2  See General Introduction.

 Nucleus, vacuole-like structures and
mitochondria are clearly observed (Pt/C,
Yeast spheroplast. Full view of a freeze-
θ = 45°).
fractured cell after high-pressure freezing.
 Figure 16.3  See Section 1.

 The metal accumulates on the dark side
while the shadow on the other side appears
Isolated viral particles after freeze-drying
white (Pt/C, θ = 45°).
and shadowing. 
 Figure 16.4  See Section 1.1.

 Pt/C, θ = 45°
Ferritin molecules freeze-dried and uni-
directionally shadowed.

 The metal accumulates around the
Ferritin molecules freeze-dried and spherical molecule and, in this case is
rotationally shadowed. structurally more informative than
unidirectionally shadowing Pt/C, θ = 45°.

Freeze-fractured microvilli from a midgut  Pt/C, θ = 45°
epithelial cell of an insect. Intramembrane
particles (IMP) form a regular network
within the microvilli membranes, with a
constant 8 nm repeat unit.


8. REFERENCES
1. Moor, H. and Mühlethaler, K. Fine structure in frozen-etched yeast cells. J. Cell
Biol., 17, 609, 1963.
2. Kistler, J., Aebi, U. and Kellenberger, E. Freeze drying and shadowing a two-
dimensional periodic specimen, J. Ultrastruct. Res., 59, 76, 1977.
3. Studer, D., Moor, H. and Gross. H. Single bacteriorhodopsin molecules revealed
on both surfaces of freeze-dried and heavy metal-decorated purple membranes, J.
Cell Biol., 90, 153, 1981.
4. Fowler, W.E. and Aebi, U. Preparation of single molecules and supramolecular
complexes for high-resolution metal shadowing, J. Ultrastruct. Res., 83, 319, 1983.
5. Gross, H., et al. High resolution metal replication quantified by image processing.
6. Gross, H. High resolution metal replication of freeze-dried specimens. In
Cryotechniques in Biological Electron Micoscopy, Steinbrecht, R. A. and Zierold,
K. eds., Springer-Verlag, Berlin, Heidelberg; Germany, 1987, 203.
7. Steere, R.L. Electron microscopy of structural detail in frozen biological specimens.
J. Biophys. Biochem. Cytol., 3, 45, 1957.
8. Moor, H. et al. A new freezing-ultramicrotome. J. Biophys. Biochem. Cytol., 10,
1961.
9. Branton, D. Fracture faces of frozen membranes. Proc. Nat. Acad. Sci. USA., 55,
1048 1966.
10. Moor, H. Evaporation and electron guns. In Freeze-Etching Techniques and
Applications, Benedetti, E.L. and Favard, P., eds., Société Française de Microscopie
Electronique, Paris, France,1973, 27.
11. Hohenberg, H. and Mannweiler, K. Semiautomatic washing device for
simultaneous cleaning of surface replicas under identical conditions. Mikroskopie
(Wien), 36, 145, 1980.
12. Stolinski, C., Gabriel, G. and Martin, B. Reinforcement and protection with
polystyrene of freeze fracture replicas during thawing and digestion, J. Microsc.,
132, 149, 1983.
13. Bachmann, L., et al. Decoration and shadowing of freeze-etched catalase crystals.
14. Branton, D. The fracture process of freeze-etching. In Freeze-Etching Techniques
and Applications, Benedetti, E.L. and Favard, P., eds., Société Française de
Microscopie Electronique: Paris, France, 1973, 107.
15. Hohenberg, et al. Plasma membrane antigens detected by replica techniques. In
Science of Biological Specimen Preparation 1985, Müller, M., Becker R.P., Boyde,
A. and Wolosewick, J.J. eds., SEM Inc. AMF O'Hare, IL, USA, 235.
16. Fujimoto, K. SDS-digested freeze-fracture replica labeling electron microscopy to
study the two-dimensional distribution of integral membrane proteins and
phospholipids in biomembranes: Practical procedure, interpretation and application.
Histochem. Cell. Biol., 107, 87. 1997.
17. Bron, P., et al. Oligomerization state of MIP proteins expressed in Xenopus oocytes
as revealed by freeze-fracture electron microscopy analysis. J. Struct. Biol., 128,
287, 1999.
18. Costello, M.J. and Gulik-Krzywicki, T. Correlated x-ray diffraction and freeze-
fracture studies on membrane model systems: Perturbation induced by freeze-
fracture preparative procedures, Biochim. Biophys. Acta, 455, 412, 1976. 

Cryo-Fracture of Self Assembled Systems in Organic Solvents 413
CONTENTS

1.1. Preparation of the Sample, Preservation of the Internal Structure and
Environment by Rapid Freezing ............................................................... 418
1.2. Transfer and Fracture under Vacuum and at Low Temperature: 170°C. 419
1.3. Shadowing and Replication ...................................................................... 419
1.4. Dissolving the Material and Washing the Replicas .................................. 420
1.5. Recovery of the Replicas and Observation............................................... 420
3.1. Materials ................................................................................................... 421
3.2. Products .................................................................................................... 421
4. PROTOCOL ...................................................................................................... 422
5.1. Advantages ............................................................................................... 424
5.2. Disadvantages........................................................................................... 424
6. VITREOUS ORGANIC SOLVENT MADE SIMPLE ................................... 425
8. REFERENCES .................................................................................................. 429

A number of difficulties in carrying out electron microscopic studies are due to the
physico-chemical state of the sample. For “hard materials”, such as ceramics and metal
alloys, the structure of the sample depends very little on their environment and many
techniques can be used routinely to study their internal organization (ion milling,
polishing, focused ion beam (FIB), etc.). The difficulties increase when the samples to be
characterized have a structure that highly depends on their solvent content. This is, in fact,
the case for most biological samples. Many chapters in this book examine how
microscopists have overcome these problems in the field of biology. Biological objects
can be observed in dilute conditions, in thin films, by cryo-TEM. There are a number of
instances where dilution is undesirable because it may destroy the architecture of the
material or it does not allow observation in the native state. Thin films are also inadequate
for large objects such as large liposomes or tissues. Therefore, cryo-fracturing was
developed by several groups9,10,22 to allow the investigation of large samples (tissue
pieces or whole cells) or lipid phases,11 where the concentrations are the same as under
physiological conditions (see also Chapter 16). These technical advances were achieved
exclusively in aqueous medium. However, in organic solvents, microscopy has not yet
reached this state-of-the-art. This sharp contrast is unfortunate because there is a growing
demand from material scientists to examine objects in organic solvents. In the last
decades, material science has exploited the concept of self-assembly to fabricate new
materials.13,26 Self-assembly is the property that some molecules possess to associate
spontaneously with themselves or with other molecules to form larger structures. In the
field of chemistry, molecules have a limited size, typically at most 1 nm. But, the self-
assembled objects they can form have dimensions as large as 10 to 1000 nm. The unique
properties of self-assembled materials emerge from those large aggregates and not from
the single molecules. Therefore, microscopic techniques with a resolution of a few nm are
necessary to evaluate the shape and the structure of the aggregates. Among these
materials, one can find liquid crystals and noncovalent polymers. This section will be
devoted only to the case of organogelators because they illustrate most of the difficulties
that can arise during the structural analysis of self-associated systems in organic solvents.
Organogelators23,25 are a growing class of compounds that are able to gel organic solvents
at low concentrations (typically a few percent per weight). This property is due to their
ability to form self-assembled structures with a high aspect ratio (fibrils, platelets,
ribbons, nanotubes, etc.). These structures form a 3-D network that is responsible for the
visco-elastic properties of the materials. The self-assembly proceeds through non-
covalent bonds, such as H-bonds, dipole or - interactions and Van der Waals forces.
The weakness of these interactions is expressed at the macroscopic level by the thermo-
reversibility of the gels; when they are heated, a sharp transition to a sol phase is
observed, and the gel forms again upon cooling.
The rheological properties of organogels explain their use in many industrial applications,
mainly as thickeners or rheomodifiers; they are used in coatings, in cosmetics or
lubricants. The formation of a network with a high specific area enables their use as
nucleating agents for semicrystalline polymers.19 Besides these widespread applications,
recent work has shown that they can be used to template inorganic or organic compounds
to form tubes or mesoporous materials.16,21,24 In the medical field, biocompatible gelators
have been used as drug delivery systems3,14,15 or supports for tissue growth.8,12,20

The interest in these compounds is also triggered by more fundamental questions. For
instance, it is still hard to establish a clear relationship between the chemical structure of
their gelators and the ability to form gels. The analysis of the shape and the dimensions of
the self-assemblies are necessary to explore this relationship. Moreover, it is interesting to
relate the morphology of the self-assembled fibrils to the rheological properties of the gels
or to the conditions of their formation (cooling rate during gelation, concentration, etc.).
Each of these questions requires an efficient tool to systematically measure the size of the
self-assemblies. At this scale, only a few techniques are available to investigate the shape
of the objects. Among them are small angle neutron or x-ray scattering techniques but
these measurements often require beam time on large facilities. Electron microscopy is
precious in this regard because it can provide a well-established method to study the
structure of individual self-assemblies.
The cryo-fracture technique of preparation and observation of organogels was chosen

because of its specificity. We describe below some of the properties of organogels that
determined this choice.
1. The structure of the self-assembled objects strongly depends on the solvent. The
formation of fibrils and gels occurs in certain solvents only. In other organic
solvents, the gelators may just form crystals or may remain soluble. Some of the
behavior can be rationalized easily with basic chemical concepts. For instance, H-
donor or H-acceptor solvents can form H-bands with the gelators, competing with
the intermolecular bonds of the self-assemblies, which results in a solubilization of
the compounds. But very often it is hard to rationalize why certain solvents favor the
formation of gels. The shape of the fibrils also varies with the solvent in which they
are formed.
2. The structure of the self-assembly depends on its concentration. Self-assembly is
governed by the following chemical equilibrium between single molecules, on the
left and associated ones on the right:
Figure 17.1
As a consequence of the mass action law, the equilibrium is expected to shift toward the
formation of single molecules upon dilution of the mixture. The phase diagram has been
established for many organogelators. An example is given in the figure below. It can be
easily observed that for a given temperature, there is a concentration C*, namely the gel
concentration or minimal gel concentration, below which, the gel can no longer form.

Figure 17.2 Phase diagram of gelators. For each temperature (here 25°C) the gel
concentration C* is the concentration below which the gel does not form.
The phase diagrams can be more complex, and for instance, the shape of the self-
assemblies can vary with the concentration. The sol phase can be constituted of soluble
self-assemblies, but their shape may differ from the ones in the gel.
Both the properties described above result in severe limitations for the choice of the
techniques used to prepare the sample — the solvent must be preserved in the samples
and the concentration must be kept constant. For instance, when the observation is carried
out on a solution that has been deposited on a carbon grid, blotted and shadowed, self-
assembled objects can be observed, but their relevance to the objects in the real gel is
always questionable. Either the concentration is kept constant during the preparation of
the sample, and then, what is observed are the self-assemblies that are actually present in
the sol, but not in the gel. It is more probable that the solvent evaporates during the
preparation, but in this case, the final composition is not known. In this regard, it must be
reminded that many organic solvents are more volatile then water. For instance, the vapor
pressure at 25°C is 24 mmHg for water, 300 for ethyl acetate, 309 for cyclohexane, and
623 for chloroform. The evaporation of the solvent is a real concern when handling
organogels.
Other difficulties arise from the rheological properties themselves. The samples studied
are gels. Their high viscosity forbids any spreading in thin films. Because of their high
solvent content, they are soft materials and, hence, difficult to cut into ultrathin sections
suitable for direct observation. Cryo-ultramicrotomy would be necessary, but it needs to
be implemented for organic solvents first.
These reasons have stimulated the implementation of the well-know technique of cryo-
fracture to study the ultrastructure of organogels. The technique is based on rapid freezing
of the sample to prevent crystallization of the solvent and to obtain it in a vitreous state.
The benefits of the process that has been widely exploited in biology are also ideal to
study the case of the organogelators because the solvent environment of the objects is
preserved and the concentration is fixed. Rapid freezing is also expected to prevent any
reorganization of the molecules within the objects, and thus allows observation of the
self-assemblies at equilibrium and in their native state. Talmon has written that “freeze-
fracture replica, regrettably a dying art, should be revived and used more for imaging
high-viscosity systems.”5

This chapter will present how to perform freeze-fracture experiments in organic solvents
to study self-assembled or high-viscosity systems. We will present here some protocols
that help in overcoming many of the obstacles associated with organic media. A number
of papers illustrate the success of the method.1,2,4,6,7,17,18 Herein we will deliberately show
some pictures resulting from negative experiments or careless protocols as well. These
examples may help to identify problems in the case of poor results. However, the
proposed protocols have been improved to make cryo-fracture a routine technique.
1.1. Preparation of the Sample, Preservation of the Internal Structure

and Environment by Rapid Freezing
 The choice of the solvents in which the  Some solvents tend to crystallize upon
self-assembly will be performed is the rapid freezing, whereas others become
first important step. The propensity of amorphous. Unlike cryo-fracture in aqueous
organic solvents to be vitrified is medium, no cryoprotectant (like glycerol in
variable. water, for instance) can be added.
 So far, no rule has been established  From our experience, we can empirically
that predicts how easily a solvent can distinguish three kinds of solvents. (1)
be vitrified. Good solvents that seldom crystallize.
These solvents are dichloromethane,4
toluene,17,18 cyclohexane2,6,7 acetonitrile,
pyridine,4 tetrachloroethane, mixtures of
solvents. (2) Borderline solvents that can be
vitrified, but often crystallize in the same
sample. Obtaining clean pictures with the
latter is still possible although difficult.
This is the case of nonane,1 or tetralin. (3)
Poor solvents that exclusively give rise to
crystals. No other structures can be
observed after fracture. This is the case of
dodecane or heptane.
 As discussed in the introduction,  For example, replacing dodecane by

changing the solvent may change the nonane or cyclohexane
structure of the self-assembly.
However, it might sometimes be wise
to switch to a solvent with close
chemical properties, but with a
different behavior in the freezing
process.

 The second step is the rapid freezing  The rate of freezing must be as fast as
of the sample between two copper possible so the solvent will be amorphous
cups. The goal of this high rate is two- (104 K/s for water). In the case of
fold: preserve the structure and obtain organogelators, care must be taken so that
a vitreous solvent. The size of the the gel does not undergo a phase transition
sample is chosen so that it is as small or additional texturation due to a shear
as possible. distortion while preparing the sandwiches.
1.2. Transfer and Fracture under Vacuum and at Low Temperature:

170°C
 After transferring the samples into the  At low temperature (–170°C) and at high
apparatus by taking care that the vacuum (between 10-8 to 10-7 mbar).
holder is always in a liquid nitrogen
bath, the fracturing can be performed.
 The fracture will occur along the plane  It is well known that for lipid bilayers it
that offers least resistance to the occurs mainly between the aliphatic chains.
applied forces.
Etching (optional step)

 In some cases, etching is performed.  First, an experiment is carried out
without etching and the replicas are
observed. If the structures appear embedded
in the solvent, a second set of experiments
has to be done with an etching step.
 In this case, after fracturing, the  The optimal time of etching cannot be
temperature is raised to the known in advance, but has to be determined
temperature of sublimation of the for each sample by several assays.
solvent (for toluene, it is 90°C) and
left for a few minutes. The temperature
is then lowered to 170°C and the
shadowing step is performed.
1.3. Shadowing and Replication

 The replication process should  First, a metal layer is evaporated at an
faithfully reproduce the 3-D surface oblique angle (45°, in general) to provide
revealed by the fracturing. Replication contrast and then reinforced by a supporting
is a two-step process. layer of carbon (C) at an angle of 90°.

 Shadowing is often performed using Pt/C is the most used as it is easier to

an electron gun and two metals are evaporate, but Ta/W provides a finer grain
most suitable for this purpose: allowing a better resolution. The thickness
Platinum/carbon (Pt/C) or of the deposited metal layer is 2 nm and the
tungsten/tantalum (W/Ta). The metal C layer 20 nm, measured with a quartz
layer will cover the inner structures of balance.
the gel revealed by the fracturing
process.
1.4. Dissolving the Material and Washing the Replicas

 At this point, it is important to choose a
solvent in which the sample is highly
soluble. This choice is determined by
chemical considerations. For instance, a
solvent with H-donor or H-acceptor groups
can disassemble molecules associated
through H-bands. Very often chloroform is
the most suitable solvent. In some
instances, heating the solvent has been
shown to better dissolve the remaining
sample. If heating is necessary, take care to
perform all the steps under a fume hood.
1.5. Recovery of the Replicas and Observation

 The floating replicas are picked up  If the replicas break down into small
with naked 400 mesh copper grids. pieces, higher mesh-size grids or grids
Finally the grid is air-dried and covered with a supporting carbon film can
observed under standard transmission alternatively be used.
electron microscope (TEM) condi-
tions.
1. Preparation of sandwiches.
2. Rapid freezing of sandwiches.
3. Transfer into the cryo-fracture

apparatus.
4. Fracture under high vacuum and

low temperature.

5. Optional etching.
6. Evaporation of the metal (Pt/C or

TaW).
7. Evaporation of the carbon layer.
8. Entry of air and increasing tem-

perature.
9. Washing the replicas.
10. Catching the replicas on copper

grids with or without a carbon
support film.
11. Observation in the TEM.
3.1. Materials
 Cryo-fracture apparatus  Balzers, Cressington, or homemade
apparatus (see Chapters 5, 6).
 Binocular microscope
 High-pressure freezing apparatus 
 Electron microscope
 Ultrasonic bath
 Fast freezing device  Plunge-freeze: “Guillotine” (see
Chapters 3, 4, 7).
3.2. Products
 Solvents for washing and cleaning  Technical grade.
 Product for cleaning copper
 400 mesh copper grids

4. PROTOCOL
1. Carefully clean the cups supplied with

the cryo-fracture apparatus (either
dilute HCl or a common copper
cleaning household product) and
sonicate 10 minutes in acetone), dry.
Figure 17.3 Two different cups used
2. Deposit a drop or a portion of the

substance under investigation on one
cup and gently cover with a second
one.
Figure 17.4 Closing the sandwich
3. Close the sandwich while taking care

not to leave any substance on the sides.
Figure 17.5 Not good
Figure 17.6 Good
4. Hold firmly the sandwich with a pair

of tweezers.
5. Plunge the sandwich very quickly into  The plunging can be done either with a
the cryogen (either liquid ethane if the freezing device ,manually or though a high-
solvent is not miscible in it or liquid pressure freezing machine (see Chapters 3,
nitrogen. 4, 7).

6. Position the sandwich tightly into the

object holder of the cryo-fracture  The entire process is carried out in liquid
apparatus maintained in liquid nitrogen.
nitrogen.
Figure 17.7 “Escaig”-like object holder in

the closed state at room temperature.
7. Introduce into the cryo-fracturing

apparatus
Figure 17.8 Our cryo-fracture apparatus

was developed and built by Dr. J.-C. Homo.
 Wait until vacuum has reached the  The higher the vacuum the smaller the
highest possible level (about 10-8 size of the metal grains deposited on the
mbar). specimen.
8. Begin evaporation of metal and as  By evaporating first and then fracturing,

soon as the quartz crystal thin film it is possible to avoid the readsorption of
monitor detects a deposit, begin the dirt on the surface that appears after
fracture step. If an etching step is fracturing (dirt may be residual solvent
necessary, it will be performed before vapors in the apparatus or substances of
metal evaporation. other origin).
Figure 17.9 Holder in the open state with

one fractured sandwich.

9. After depositing about 2 nm Pt/C or  The carbon layer improves the physical
Ta/W measured by the thickness strength of the replica and, therefore, the
detector, a 20 nm carbon layer is latter can be more readily handled (see
deposited at a 90° angle. Chapter 16).
10. The holder is removed and heated

up to room temperature under a dry
nitrogen stream.
11. The cup is put into a recipient

containing a solvent in which the  The choice of solubilizing solvent
initial product is totally soluble. directly influences the final result.
12. The replicas are gently recovered  If necessary, grids with a smaller mesh
onto 400-mesh naked copper grids. size or with a carbon supporting film can be
used. This is mainly helpful if the replicas
have broken into small pieces.
5.1. Advantages
 Preservation of the native 
environment (solvent is maintained).
 Enables structural studies on “large”  Greater than 150 nm.
objects or concentrated solutions.
 Preservation of the three-dimensional
organization.
 Enables access to the internal and
external structure of vesicular
systems.
5.2. Disadvantages

 Requires sophisticated equipment.
 Not all solvents are suitable for such 
analysis (crystallization may be a
problem).
 Time consuming. 
 About 2 nm.
 Low resolution.
 Unsuitable for small objects.  Less than 2 nm if ordered, otherwise
10 nm.

6. VITREOUS ORGANIC SOLVENT MADE SIMPLE
The method works in most situations provided it is possible to find a solvent in which the
substance being investigated is highly soluble. Chloroform has been found to be one of
the most useful solvents in our experience.
7. OBSERVED RESULTS
Figure on Chapter’s title page  Nanotubes of self-assembled diamides

3 forming a gel in cyclohexane.
F F
B  Figure 17.10
N N
 Chemical structure of the molecule (1)
HN O 

O O 
C16H33O
N N
OC16H33 
H H 
C16H33O OC16H33 
OC16H33 OC16H33  



 Figure 17.11
 Poor freezing: In gels of 1 in nonane,
solvent crystals have formed. They
appear as terraces corresponding to
the molecular layers and present
generally a faceted shape.















 Figure 17.12 Good preservation of
structures in nonane.
 Cryo-fracture of a gel formed by self-
assembly of borondipyrromethenes in
nonane. These molecules associate as
small fibers 20 nm wide and up to 500
nm in length (arrows) and form a
thermotropic liquid crystal. The gel
also presents some interesting
fluorescence properties.1
R1 

O O
 Figure 17.13 Chemical structure of the
diamide gelators (compounds 2 to 5)
visualized in figures 17.14 to 17.20.
O O Compound R1 R2 R3
R2 R2 2 C11H23 C5H10 C7H15
O 3 C10H21 C5H10 C7H13
O
4 C10H21 C10H21 C11H23
HN NH 5 C20H41 C5H10 C7H15
R3 R3





 Figure 17.14
 Poor freezing of a gel of the diamide

gelator 2 in cyclohexane, solvent
crystals appear as angular structures
(arrows).









 Figure 17.15 Same sample as in Figure
17.14 at a higher magnification. Self-
assembled fibers of the diamide 2 (arrows)
are seen lying between the solvent crystals
(star).





 Figure 17.16 Poorly washed sample
(gel of 3 in toluene, 2% (w/v)). The black
structures (arrow) overlying the replica are
the remains of the sample. Underneath, the
real structures are visible as straight
platelets (arrowheads).









 Figure 17.17 Replica of a gel of
diamide 4 in toluene 2% (w/v). The
structures are bathing in the solvent and,
therefore, difficult to visualize.












 Figure 17.18 Same sample as in Figure
17.17 but after freeze-etching at –90°C for
three minutes. The structures are clearly
visible as flat ribbons.

 Figure 17.19 Replica of a gel of 5 in

cyclohexane 2% (w/v) after freeze-drying.
Excessive freeze-drying has entailed
collapse of the structures, which appear as
dense fibers (arrows).
 Figure 17.20 Same sample as above but

this time well preserved.

8. REFERENCES
1. Camerel, F. et al. Highly luminescent gels and mesogens based on elaborated

borondipyrromethenes, J. Am. Chem. Soc., 128, 4548, 2006.
2. Camerel, F. et al. Tuning the thermotropic and lyotropic properties of liquid-
crystalline terpyridine ligands, Eur. Chem. J., 12, 4261, 2006.
3. Couffin-Hoarau, A.-C. et al. In situ-forming pharmaceutical organogels based on
the self-assembly of L-alanine derivatives, Pharm. Res., 21, 454, 2004.
4. Cuccia, L.A. et al. Encoded helical self-organization and self-assembly into helical
fibers of an oligoheterocyclic pyridine-pyridazine molecular strand, Angew. Chem.,
Int. Ed. Engl., 39, 233, 2000.
5. Danino, D. and Talmon, Y. Direct-imaging and freeze-fracture cryotransmission
electron microscopy of molecular gels, in Molecular Gels. Materials with Self-
Assembled Fibrillar Network, Weiss Richard, G. and Terech, P., eds., Springer,
Dordrecht, The Netherlands, 2006, 253.
6. Diaz, N. et al. Self-assembled nanotubes in organic solvents, Macromol. Symp.,
241, 68, 2006.
7. Diaz, N. et al. Self-assembled diamide nanotubes in organic solvents, Angew.
Chem., Int. Ed. Engl., 44, 3260, 2005.
8. Ellis-Behnke, R.G. et al. Nano neuro knitting: Peptide nanofiber scaffold for brain
repair and axon regeneration with functional return of vision, Proc. Nat. Acad. Sci.
USA, 103, 5054, 2006.
9. Escaig, J. and Nicolas, G. Cryofractures of biological material performed at very
low temperatures in ultravacuum, C. R. Acad. Sci. (Paris), 283, 1245, 1976.
10. Gross, H., Bas, E., and Moor, H. Freeze-fracturing in ultrahigh vacuum at -196
degrees C, J. Cell Biol., 76, 712, 1978.
11. Gulik-Krzywicki, T. Electron microscopy of cryofixed biological specimens, Biol.
Cell, 80, 161, 1994.
12. Holmes, T.C. et al. Extensive neurite outgrowth and active synapse formation on
self-assembling peptide scaffolds, Proc. Nat. Acad. Sci. USA, 97, 6728, 2000.
13. Lehn, J.M. Supramolecular chemistry, Science, 260, 1762, 1993.
14. Murdan, S. Organogels in drug delivery, Expert Opin. Drug Deliv., 2, 489, 2005.
15. Murdan, S., Gregoriadis, G., and Florence, A.T. Sorbitan mono-
stearate/polysorbate 20 organogels containing niosomes: A delivery vehicle for
antigens?, Eur. J. Pharm. Sci., 8, 177, 1999.
16. Ono, Y. et al. Organic gels are useful as a template for the preparation of hollow
fiber silica, Chem. Commun., 1477, 1998.
17. Schmidt, R. et al. New bisamides gelators: Relationship between chemical
structure and fiber morphology, Tetrahedron Lett., 44, 3171, 2003.
18. Schmidt, R. et al. Organogelation properties of a series of oligoamides, Langmuir,
18, 5668, 2002.
19. Shepard, T.A. et al. Self-organization and polyolefin nucleation efficacy of
1,3:2,4-di-p-methylbenzylidene sorbitol, J. Polym. Sci. Part B: Polym. Phys., 35,
2617, 1997.
20. Silva, G.A. et al. Selective differentiation of neural progenitor cells by high-Epitope
density nanofibers, Science, 303, 1352, 2004.

21. Simon, F.X. et al. Formation of helical mesopores in organic polymer matrices, J.
Am. Chem. Soc., 129, 3788, 2007.
22. Steere, R.L. Electron microscopy of structural detail in frozen biological
specimens, J. Biophys. Biochem. Cytol., 3, 45, 1957.
23. Terech, P. and Weiss, R.G. Low molecular mass gelators of organic liquids and
the properties of their gels, Chem. Rev., 97, 3133, 1997.
24. van Bommel, K.J.C., Friggeri, A., and Shinkai, S. Organic templates for the
generation of inorganic materials, Angew. Chem., Int. Ed. Engl., 42, 980, 2003.
25. Weiss, R.G. and Terech, P. Molecular Gels. Materials with Self-Assembled
Fibrillar Network. Springer, Dordrecht, The Netherlands, 2006.
26. Whitesides, G.M., Mathias, J.P., and Seto, C.T. Molecular self-assembly and
nanochemistry: A chemical strategy for the synthesis of nanostructures, Science,
254, 1312, 1991.


Part V
Analysis

Progressive Lowering of Temperature PLT 435
CONTENTS

1. PRINCIPLES OF PROGRESSIVE LOWERING OF TEMPERATURE ... 438
1.1. PLT Dogma .............................................................................................. 438
1.2. General Outline of the PLT Method ......................................................... 438
3.1. Materials ................................................................................................... 442
3.2. Products .................................................................................................... 443
3.3. Solutions ................................................................................................... 443
4. PROTOCOLS .................................................................................................... 446
4.1. Chemical Fixation of the Sample of Interest ............................................ 446
4.1.1. Tissue ............................................................................................ 446
4.1.2. Cell suspensions ............................................................................ 447
4.1.3. Cell monolayers............................................................................. 447
4.1.4. Adherent cells grown on sapphire disks as monolayers ................ 448
4.1.5. Polarized cell monolayers ............................................................. 448
4.1.6. Precious samples ........................................................................... 449
4.2. Preparation of the Low-Temperature Embedding Apparatus ................... 450
4.3. Programming the Apparatus According to the Resin Chosen .................. 450
4.3.1. For Lowicryl K4M ........................................................................ 450
4.3.2. For Lowicryl HM20 ...................................................................... 451
4.3.3. For AFS2....................................................................................... 451
4.4. Dehydration .............................................................................................. 452
4.4.1. Postfixation ................................................................................... 452
4.4.2. Dehydration................................................................................... 452
4.5. Embedding................................................................................................ 453
4.6. Polymerization.......................................................................................... 454
4.7. Further Processing .................................................................................... 454
4.7.1. Room temperature sectioning........................................................ 454
4.7.2. Immunolabeling ............................................................................ 455
4.7.3. In situ hybridization....................................................................... 455
5.1. Advantages ............................................................................................... 458
5.2. Disadvantages........................................................................................... 458


6.1. With Regards to the Tokuyasu Immunostaining Method......................... 459
6.2. Compared to Cryo-Fixation ..................................................................... 459
6.3. New Technique ........................................................................................ 459
8. REFERENCES .................................................................................................. 464


The progressive lowering of temperature (PLT) procedure was introduced in the 1950s
when it was common to use acrylic resins. These methacrylate-based resins, however,
were unstable under the electron beam, unreliable and difficult to section. Therefore, they
disappeared giving way to epoxy resins until the 1980s when they became the resin of
choice for immunocytochemistry or in situ hybridization. In fact, Kellenberger et al.14 re-
introduced the acrylic resins, now commonly called Lowicryls, a combination of the
name of the company Lowi Werke and the term “acryl.” Their formulation was also
improved. The advantages of these new resins are multiple: they have low viscosity,
therefore, they infiltrate rapidly into tissues; they are fluid at low temperatures, some of
them are miscible with water; they are hydrophilic when polymerized so that they have no
nonspecific attraction for immunoreagents; and, in addition, they have good electron
beam stability. The Lowicryl resins were not only developed for the low temperature
embedding procedure, but also to enable optimal ultrastructural preservation and to allow
observation of unstained samples in scanning transmission electron microscopy (STEM).
Indeed, they give low scattering ratios due to their low density as compared to the higher
density of tissue.3 The contrast (ratio of elastically and inelastically scattered electrons
known as the Z contrast) is provided by the tissue itself.5 However, the ability to view
biological samples in STEM without contrasting agents is not applicable to transmission
electron microscopy (TEM). The contrast required to view cellular structures embedded
in Lowicryl K4M can be obtained by conventional methods or enhanced using the method
described by Roth et al.24
A variety of acrylic resins have been developed and are known under the following
trademarks:
 LR White and LR Gold (London Resin, UK)
 Lowicryl K4M, HM20, K11M and HM23 (Lowi Werke, Germany)
 Bioacryl (Unicryl – British BioCell International, UK)
In the meantime, colloidal gold was developed and the perspective of fine localization of
macromolecules at the ultrastructural level12 required improved methods of specimen
preparation. It was then clear that fixation, dehydration and resin embedding altered the
specimen, therefore, the low temperature and freezing methodologies developed in the
1950s were considered as better alternatives to improve our understanding of
ultrastructure and develop the science of immunocytochemistry. The PLT method, the
freeze-substitution method (see Chapter 13) and the Tokuyasu method (see Chapter 19)
are the three main procedures used today for immunocytochemistry. A new emerging
technique based on cryo-fixation, rehydration and Tokuyasu cryo-sectioning (see Chapter
14) also appears promising.

1. PRINCIPLES OF PROGRESSIVE LOWERING OF

TEMPERATURE
1.1. PLT Dogma

 In the PLT method, the temperature of
the specimen is gradually decreased as  Macromolecules in the specimen are
water is removed from it by increasing immobilized in an “inert” manner.
the concentration of organic solvents  Causes less conformational changes to
so as to minimize the denaturing the tertiary structure of protein molecules.
effects of solvents on proteins. This  Reduces lipid extraction
method originally developed by  The final temperature depends on the
Fernandez-Moran6 was reintroduced resin chosen.
by Carlemalm et al; in the 1980s.4
 The samples are then impregnated
with increasing concentrations of
acrylic resin in solvent at the low
temperature most suitable for the resin
chosen. Finally, the sample is floating
in a 100% acrylic resin solution.
 The resin is then polymerized by UV  Inhibition of the polymerization step can

irradiation, which activates a occur if osmium has been used as a post-
photoinitiator mixed into it. fixative or if a colored pigment is present in
the sample (e.g., intense color of the
erythrocytes in blood vessels. Use
perfusion fixation that will wash out the red
blood cells).
 The resin is maintained at a low
temperature in a cooled ethanol bath  But there is no injection of nitrogen gas
during polymerization, thus only a into the specimen chamber during
minimal temperature increase can be polymerization.
measured during polymerization.7
1.2. General Outline of the PLT Method

 The sample is chemically fixed.  The sample should be fixed as soon as
the cells are collected or the organ
extracted from an animal.
 Perfusion fixation is preferred.
 Mild fixation is preferred because it
does not interfere with antigenicity.
 The overall structure is preserved.
 The sample is dehydrated.  Water is progressively replaced by the
organic solvent in a series of steps.
 The solvent environment should remain
as “water like” or as polar as possible
during the entire embedding procedure.

Based on this idea numerous attempts to

develop water-soluble or water-compatible
embedding media were made without
success.
 Ethanol is commonly used.
 Acetone is unsuitable for the PLT
technique as 70% acetone has a freezing
point of 27°C and PLT requires the 70%
step to be at 35°C.
 The solvent concentration has to be
carefully chosen such that the temperature
is above the freezing point of the organic
solvent concentration used in the previous
step (see Graph 1 in the Lowicryl booklet
supplied with the resin).
 One must ensure that enough time is
allotted to allow the interior of the sample
to equilibrate with the dehydrating agent at
each step.
 The resin has to be soluble in the solvent
used. K4M and HM20 are soluble in most
dehydrating agents. However, ethylene
glycol and dimetyl formamide are not
miscible with HM20.
 The temperature is lowered during the  Until the polymerization temperature of
dehydration procedure. the resin chosen is attained.
 Lowicryls are the only low temperature
embedding resins (below 30°C).
 Lowicryl embedding can be performed  Different formulations of Lowicryls are
at temperatures between 70°C to available. We prefer K4M, which
30°C depending on the type of resin. polymerizes down to 30°C or HM20
down to 45°C.
 K4M was largely used because of its
good properties in on-section colloidal gold
labeling.23
 HM20 was compared to K4M for
immunodetection22 and was found to be
easier to section and yield labeling
comparable to K4M.
 HM23 and K11M which polymerize at
even lower temperatures are more
dedicated to cryo-substitution.1
 Other resins exist and have their  LR white (see Chapter 13).
advantages and disadvantages.  LR gold (see Chapter 13).
 Unicryl/Bioacryl.
 Polymerization is initiated by benzoin  In case of the use of Lowicryl.
methylether and long wave UV  First by indirect UV light.
irradiation (~ 360 nm, a wavelength  For further hardening the block is
that is not efficiently absorbed by exposed to direct UV light at room tem-
proteins or nucleic acids) at low perature for three days.
temperature.

1. Chemical fixation of the sample of

interest
1.1. Tissue
 Fixation by immersion
 Sample size:
 ~ 0.5 mm31 mm3  Samples not larger then 0.5 mm3 are
recommended.
 Fixation by perfusion  This is the method of choice.
 Only possible on animals.
1.2. Cell suspensions

 Add a two times concentrated fixative
to the medium
 Pellet the cells
 Fix again
1.3. Adherent cell monolayers grown in

culture flasks
 Fix the cells
 Scrap the monolayer with a rubber
policeman
 Centrifuge and discard the supernatant
 Add fresh fixative over the pellet
1.4. Adherent cell monolayers grown

on sapphire discs
1.5. Polarized cell monolayers
1.6. Precious samples  Samples from human biopsies (e.g.,

bone marrow)
 Samples from a Bio Safety Lab 3 when
very few cells can be obtained.
 A = Tissue fixation
 B = Cell suspension
 C = Cell monolayer
 D = Sapphire disc
Figure 18.1

2. Preparation of the low temperature  The Leica AFS25 is used in the authors’
embedding apparatus laboratories and presented in this chapter
(see Figure 18.4).
3. Programming the apparatus

depending on the resin chosen and
on the apparatus
 For Lowicryl K4M

 For Lowicryl HM20
 For the AFS2
4. Dehydration
5. Embedding depending on the sam-

ple and purpose of the work
 Flat embedding  Mostly for tissue, useful to orientate the
specimen.
 Microtube embedding  For all purposes. Dehydration is done in
Sarstedt tubes or in 1.5 mL microtubes.
Embedding can be done in 0.5 mL Sarstedt
microtubes, which can be directly fixed in
the ultramicrotome after polymerization.
 Leica molds  Dehydration and embedding can be done

in the same mold.
 Sapphire discs  Dehydration and embedding are

performed using the specific tools (Leica
SD FS Unit).
 Embedding after protein internal-  Step 1: A protein or antibody coupled to

ization colloidal gold is internalized.
 Step 2: Immunogold labeling on sec-
tions of the PLT embedded sample.
6. Polymerization
7. Further processing
 Room temperature sectioning

 Immunolabeling  See Chapter 23.
 In situ hybridization  See Section 4.7.3.

3.1. Materials
 Apparatus for PLT  Homemade.27
 Leica EM AFS or EM AFS2, Leica,
Microsystems, Vienna, Austria.
 RMC FS-7500, Boeckeler Instruments
Inc., Tucson, Arizona, USA (no longer
sold).
 Centrifuge  Used to pellet cells.
 Chemical hood  Used to prepare solutions.
 Resins are very toxic and allergenic.
Read the technical note before using them.
 Adapted gloves and chemical face  Caution: Methacrylates are aller-genic;
mask please refer to the safety information
provided with the product.
 Gelatin capsules  Agar Scientific or any other EM material
supplier. Size 00.
 Leica plastic capsules  #16702738, perforated tubes used for
freeze-substitution.
 Glass bottles  Adapted to the chamber of the apparatus.
 Some are furnished with the #16702737
Leica apparatus.
 Rubber policeman  To scrape the cells off culture dishes.
 Microcentrifuge tubes  Sarstedt # 16702765.
 To store ice.
 Styrofoam boxes
 for example, #16708826 or #16708864
Leica.
 Special waste containers  To recycle leftover solutions (should be
performed by professionally qualified
personnel). Polymerized resins are still
toxic.
 Sapphire disc freeze-substitution unit  SD FS is a special tool for substitution
and embedding of sapphire discs. Leica
 Sarstedt tubes for the AFS2  Sarstedt.
 #16707150 Leica.
 Microvettes CB300  Sarstedt #16 440.
 To keep the low melting point agarose
 Stove
and the instruments at 40°C.
 Water bath  To maintain the low melting point
agarose fluid.

3.2. Products
 Cacodylic acid sodium salt trihydrate  # 8.20670.0250, Merck.
 Gelatin  To embed sparse cells, very small
biopsies or very precious samples.
 EMS #16560, Hatfield, Pennsylvania,
USA or SPI Supplies, West Chester,
Pennsylvania, USA or Merck 1.0470.
 Uranyl acetate  EMS:
http://www.emsdiasum.com/microscopy/asp
/a.aspx
 Low melting point agarose  Same use as gelatin (EMS #1670B).
 Caution: Normal agarose may destroy
the diamond knife.
 Gold colloids  Ultra-small colloidal gold and specific
secondary antibodies: Aurion, The
Netherlands, British BioCell International,
U.K.
 Silver enhancement  SE-EM, Aurion, The Netherlands.
 Methyl cellulose  25 centipoises.
 Osmium tetroxide  10 x 0.25 g, # R 1016, Agar.
3.3. Solutions
 Ethanol  #02860, Fluka.
 30%, 50%, 75%, 100%  The best method is to prepare ethanol
solutions with an alcoholmeter.
 Due to volume changes in mixtures of
ethanol and water, dilutions are incorrect.
 Fixatives
 2 to 4% (w/v) formaldehyde (FA) and  Cell culture: 2.5% FA + 0.1% GA.
less then 0.5% (v/v) glutaraldehyde  Cell suspension: 5% FA + 0.2% GA.
(GA) in 0.1 M cacodylate buffer;  Fixative preparation:
pH 7.2  Dissolve 2.5 or 5 g paraformaldehyde
in 45 mL H2O: 2 hours at 65°C.
 Warm up to about 65C.
 Stir vigorously for some time.
 Add 20 to 80 L of NaOH 1 M, wait
for 10 minutes.
 If the paraformaldehyde has not
dissolved into formaldehyde, repeat
the NaOH step. Thereafter, add
glutaraldehyde if desired and adjust to
50 mL with H2O.
 Then add 50 mL double strength
buffer and check the final pH.
 Note: Can be stored frozen in aliquots.

 Glutaraldehyde (GA)  Glutaraldyde 25% solution in water

(EMS #16410).
 Toxic by inhalation. Use under a fume
hood.
 Uranyl acetate solutions (UA)
2%, 5% Aqueous uranyl acetate  In double-distilled water. Kept at 4°C in
the dark.
 0.5% Uranyl oxalate, pH 7.4  Prepared as described in Chapter 19.
 Buffers
 100 mM Phosphate buffer, pH 7.4  Na2HPO4 disodium hydrogen phosphate,

Preparation of 1 M stock solution: NaH2PO4 sodium dihydrogen phosphate.
 Dissolve 14.19 g Na2HPO4  Solution A.
(MW 141.96) in 100 mL double-
distilled water
 Dissolve 13.8 g NaH2PO4, H2O (MW
138.00) in 100mL double-distilled  solution B.
water  Add 19 mL of solution A to 81 mL of
 The pH should be 7.4, according to the solution B before use.
phosphate sodium buffer chart  dilute 10-fold to make 100 mM.
 0.1 M Cacodylate buffer; pH 7.4  Cacodylic acid sodium salt trihydrate

Preparation of 0.1 M Cacodylate Na(CH3)2AsO2, 3H2O (MW 214.05).
buffer:
 Dissolve 21.4 g in 1 L double-
distilled water  Check and adjust the pH to 7.2 to 7.4.
 Add 1mM MgCl2 from a 1 M stock
solution
 Add 1mM CaCl2 from a 1 M stock  The osmolarity should be ~ 300
solution mosmoles.
 Add 2% sucrose.  Storage at 4°C or frozen in aliquots.
 Lowicryl resin:  EMS or Polysciences Europe GmbH,
Eppelheim, Germany.
 Ready to use K4M  K4M Monostep (EMS#14335).
 Is hydrophilic.
 Preparation of polar K4M:
 Cross-linker A: 2.70 g or 3.6 mL  Increase the amount to obtain harder
blocks.
 Initiator C: 0.10 g  Benzoin methyl ether for UV
polymerization.
 Monomer B: 17.30 g or 25 mL  Gives the polar property to the resin.
 Procedure:
Mix cross linker and monomer in  Bubble a small stream of dry nitrogen in
a vial and take care not to the mixture through a Pasteur pipette.
introduce any air into the resin  Oxygen inhibits polymerization of any
mixture. The initiator is then acryl resin.
added and mixed the same way.

 Ready to use HM20  HM20 Monostep (EMS#14345).

 Is hydrophobic.
 Preparation of apolar HM20:
 Cross-linker D: 2.98 g or 4 mL
 Initiator C: 0.10 g  Benzoin methyl ether for UV poly-
merization.
 Monomer E: 17.02 g or 25 mL  Gives the apolar property to the resin.
 In Situ Hybridization solutions

 SSC buffer 20X:
 175.3 g sodium chloride
(MW 58.44)
 88.2 g sodium citrate
(MW 294.10)
 Fill up to 1 L sterile double-  Adjust the pH to 7.0 with sodium
distilled water hydroxide 10 M.
 Phosphate/NaCl buffer:
 1 M Phosphate buffer  14.6 g Sodium chloride (MW 58.44) fill
 5 M Sodium chloride up to 50 mL with phosphate buffer.
 1 M Tris/HCl buffer (MW 121.16)  C4H11NO3 .
 20 mM Tris/HCl–300 mM NaCl
buffer
 6 mL 5 M Sodium chloride
 2 mL 1 M Tris
 Fill up to 100 mL with sterile
double-distilled water  Adjust the pH to 7.6 with HCl.
 Blocking buffer
 1% BSA in
Phosphate/NaCl buffer
 Washing buffer
 20 mM Tris/HCl–300 mM NaCl
buffer or 2X SSC
 Hybridization buffer:  The hybridization buffer can be stored at

 2X SSC buffer 20°C with or without the probe.
 30% deionized formamide  Roche 11 814 320 001.
 2X Denhardt’s solution
 250 µg/mL yeast tRNA  Roche 10 109 495 001.
in a 100 µL final volume

4. PROTOCOLS
4.1. Chemical Fixation of the Sample of Interest

 Choose the fixative so that:  Extraction of cellular components is
reduced.
 No denaturation is observed.
 No alteration of epitopes occurs because
immunolabeling is the purpose of the PLT.
 No alteration of DNA or RNA for in situ
hybridization (ISH) occurs.
 The volume and the shape of the sample
are preserved.
 A mild fixation is preferred  Formaldehyde enters organs or cell

usually 2 to 4% FA supplemented suspensions slowly.
with less then 0.5% GA  Preserves antigenicity.
 The overall morphology is not well
preserved.
 Take care: The reaction is reversible
during the washing steps. The sample
should be kept in a 1/10 dilution of the
initial concentration if it is not immediately
treated.19
 Glutaraldehyde at a low concentration is
added to preserve the ultrastructure.11
 To preserve membrane structures, a  Osmium tetroxide cannot be used
postfixation with neutral uranyl acetate because it interferes with antigen reactivity.
is recommended.
4.1.1. Tissue
 Fixation by immersion  The sample size should not exceed
 2 to 4% FA + < 0.5% GA 0.5 mm3.
 Fixation by perfusion  Method of choice: Only possible on

 2 to 4% FA + < 0.5% GA animals.
 Depends on the organ and the purpose of
the investigation. Some tissues (e.g., lung)
must be “washed” with 9‰ NaCl at 37°C
before perfusion to prevent cells from
plugging capillaries. Very soft or poorly
structured tissues, such as spleen or bone
marrow, are difficult to recover after
perfusion, although it is not impossible to
do it.

4.1.2. Cell suspensions

1. Add an equal volume of  Freshly prepared FA.
5% FA supplemented
with 0.2 % GA in 0.1 M 30 min
cacodylate buffer. at 4°C
2. Pellet the cells 5 min

2500 rpm
3. Transfer the cells in a

microtube and further fix
them for half an hour with 30 min
2.5% FA supplemented with at 4°C
0.1% GA.
4. Wash the cells with 0.1 M cacodylate  Pellet the cells 1 min at 7000 rpm.
buffer.
5. If not further processed immediately,

the pellet should be resuspended in
cacodylate buffer containing 0.25%
FA.
4.1.3. Cell monolayers

1. The culture medium is removed.  Avoid the use of trypsin-EDTA, which
complexes the divalent ions and modifies
cell membranes.
 Chemical detachment also leads to gross
changes of the morphology.
2. Add enough 2.5% FA
supplemented with 0.1% 30 min
GA to cover the monolayer. at 4°C
3. Scrap the cells with a rubber
policeman.
4. Pellet the cells. 5 min
2500 rpm
5. Resuspend the pellet in
the same fixative and 30 min
transfer it into a micro- at 4°C
tube.
6. The cells are then washed with 0.1 M
cacodylate buffer.
FA.

4.1.4. Adherent cells grown on sapphire disks as monolayers
1. The cells are fixed with  Some authors (see section 4.4 in Chapter
2.5% FA supplemented 30 min - 1 h 5) coat the sapphire discs with carbon
with 0.1 % of GA. at 4°C before growing cells.
2. Wash the cells with 3 × 10 min

0.1 M cacodylate buffer, at 4°C
pH 7.4.

FA.
4.1.5. Polarized cell monolayers
1. The cells grown in a Petri  Such as Caco cells.

dish are fixed with 2.5% FA 1 h  The objective is often to embed a single
supplemented with 0.1% at 4°C cell layer and to section it transversally to
GA. or less the culture.
 Cheaper compared to sapphire discs.
2. The monolayer in the Petri dish is cut

into small ribbons with a razor blade.
3. With a thin rubber policeman, a ribbon

of cells is removed and placed on a
slide.
4. A thin layer of low-melting point

agarose (2% in 1 M cacodylate buffer)  Let it solidify on ice.
is poured over the ribbon.
5. The sample is hereafter treated as a  The embedding can be done:

cell pellet.  To further section the cell layer (see
Figure 18.2, Steps 5, 6).
 To further section the cell layer
transversally (see Figure 18.2, Step 7).

 1 = Ribbons on the Petri dish.

 2 = Lift off the ribbon with a rubber
policeman and place it on a slide.
 3 = Pour agarose on the sample and
let it solidify.
 4 = Cut several pieces of the sample.
 5 = PLT and embedding in flat
molds (AFS2).
 6 = Embedding in flat molds (AFS).
 7 = Dehydration and embedding (for
further transverse sectioning) in a
Leica tube (AFS).
Figure 18.2 The different steps to embed a

monolayer of polarized cells.
 Caco cells have been agarose embedded

as described above. After PLT and
Lowicryl K4M flat embedding, sections
were cut through the cell monolayer. Only
one cell layer can be observed and the
apical and basal sites are clearly visible
under the electron microscope.
Figure 18.3 Observed results of above

steps
4.1.6. Precious samples

 Some very precious samples, such as
human biopsies, are easily damaged  The low melting point agarose is kept in
and/or lost during the whole a water bath at 40°C after dissolution by
procedure. Therefore, after fixation in boiling.
2.5% FA supplemented with 0.1%  Pipettes and microtubes are kept at 40°C
GA: in the water bath or in a stove.
 Small biopsies are enwrapped in low-  Let it solidify on ice.

melting point agarose.  Most of the agarose surrounding the
sample is removed with a razor blade.
 Isolated cells are pelleted in low-  An easy way to pellet the cells is to use
melting point agarose. Microvette (Sarstedt).
 Solidified cell pellets are cut into small
cubes.

4.2. Preparation of the Low-Temperature Embedding Apparatus

 Fill with liquid nitrogen  See Figure 18.4
Figure 18.4 Leica AFS.
4.3. Programming the Apparatus According to the Resin Chosen

4.3.1. For Lowicryl K4M
 Settings for the AFS Settings are made according to Carlemalm
et al.4 and Whitehouse et al.27
 The different steps are important: Do not
forget any of them!
 When in 30% ethanol, the temperature
should be 0°C or it is better to set a slope
from 0°C to 2°C or 4°C.
 When in 50% ethanol, the temperature
should progressively decrease from 0°C to
20°C.
 It should also progressively decrease
from 20°C to the next temperature: 25°C
(K4M) or 45°C (HM20). This is possible
with the AFS2 where two successive slopes
can be selected. While not possible in the
AFS1, a short step (T2) is inserted before
the next slope.
 The two steps in 100% ethanol are also
important.
Figure 18.5 Control unit of the AFS.
 T indicates the temperature chosen and  The temperature can vary from one
S the slope or rate in decrease from apparatus to another by ± 5°C. Check the
one temperature to another. The temperature the first time you use it.
temperatures are set in decimals: half  An easy test is to put some K4M
an hour is 0.5 hour, 0.1 hour is 6 min. Lowicryl in the apparatus at 30°C. If the
viscosity of the Lowicryl has increased, the
temperature is too low.

The program should be set at:

 T1 : 0 C; 0.5h  S2: 5 C; 1h
 S1 : 20 C; 1h  T3: 25 C; 48h
 T2 : 20 C; 0.2h  Another way is to put some ethanol 30%
 S2 : 10 C; 1h in the AFS chamber at 0°C. The ethanol
 T3 : 30 C; 48h should stay fluid. If this is not the case,
your AFS is too cold. Set the program as
above.
4.3.2. For Lowicryl HM20

 Settings
 T1 : 0 C; 0.5h  Same comments as above.
 S1 : 20 C; 1h
 T2 : 20 C; 0.2h
 S2 : 25 C; 1h
 T3 : 45 C; 48h
4.3.3. For AFS2

 Settings
An example of settings is shown in Table  When equipped with the freeze
18.1. substitution processor (FSP), the AFS2
becomes a real robot: dehydration,
embedding and UV polymerization are
done automatically.
STEP T° T° End Slope Time Reagent Content Transfer Agitation UV Pause

Start °C °C/h
01 0 0 0 0.05 Ethanol 30% stay off
02 0 15 30 0.30 Ethanol 50% mix on
03 15 30 30 0.30 Ethanol 75% mix on
04 30 35 10 0.30 Ethanol 100% Exch/fill on
05 35 35 0 0.30 Ethanol 100% Exch/fill on
06 35 35 0 1:00 HM20 30% mix on
07 35 35 0 1:00 HM20 60% mix on
08 35 35 0 1:00 HM20 100% Exch/fill on
09 35 35 0 15:00 HM20 100% Exch/fill on
10 35 35 0 30:00 HM20 100% stay off X
11 35 20 13.8 4:00 HM20 100% stay off X
Table 18.1 Settings of the Different Steps

4.4. Dehydration
4.4.1. Postfixation
 0.5% Neutral uranyl acetate  30 minutes on ice (not more!).
 Neutral uranyl acetate is prepared in a
Veronal-based Mickaelis buffer.
 Note: Because Veronal is very toxic, a
special authorization to manipulate this
product is required. Therefore, we will not
give the recipe in this chapter.
 The uranyl oxalate prepared for the
Tokuyasu method should be preferred (see
Chapter 19).


4.4.2. Dehydration
Dehydration is done using ethanol at  The solvent must never freeze. The
different concentrations. temperature of the chamber should always
be well above the freezing point of the
solvent to be used.
Method:
 For embedding in Lowicryl K4M:
 30% Ethanol: 30 min, 0°C  Cool the pipette, discard the medium and
 50% Ethanol: 1h, 0°C to 20°C add precooled 30% ethanol.
 75% Ethanol: 1h, 20°C to 30°C At each step the next ethanol solution is
 100% Ethanol: 1h, at 30°C precooled in the AFS chamber.
 100% Ethanol: 1h, at 30°C
 For embedding in Lowicryl HM20:

 30% Ethanol: 30 min, 0°C  Same comments as above.
 50% Ethanol: 1h, 0°C to 20°C
 75% Ethanol: 1h, 20°C to 45°C
 100% Ethanol: 1h, at 45°C
 100% Ethanol: 1h, at 45°C
 Use of Lowicryls (follow the supplier’s

E Temperature °C T instructions).
Vol K4M HM20 K11M HM23 min  E = Ethanol
%
 T = Time
30 0 0 0 0 30
50 20 20 20 20 60
70 35 50 50 50 60
95 35 50 60 60 60 Table 18.2 Typical dehydration scheme for
100 35 50 60 80 60 PLT. (Carlemalm, personal communication,
100 35 50 60 80 60 see Reference 10.)

4.5. Embedding
Embedding depending on the sample and the purpose of the work.
1. Procedure 
 EthanolLowicryl: (1:1) (v/v)  For one hour
 EthanolLowicryl: (1:2) (v/v)  For one hour
 Pure Lowicryl  For one hour
 Pure Lowicryl  Overnight
 Pure Lowicryl  For one hour before inserting in molds
and embedding.
2. Flat embedding 
 After dehydration, the sample is  Mostly for tissue.
placed in the adequate holder.  The block can be orientated.
Embedding is done as described  Caution: Flat embedding cannot be done
before. in parallel with Leica mold embedding.
 In the AFS2, flat embedding single-use
molds are furnished.
3. Eppendorf embedding  Convenient technique in the AFS but

only with the tightly sealing Sarstedt
microtubes.
 Works well for cryo-substitution and
Epon embedding.
 In the AFS2, Eppendorf embedding is
replaced by “Sarstedt” tube embedding.
4. Leica capsule embedding  Convenient because dehydration and

embedding can be done in the same
capsule.
 In the AFS2 “Leica” capsules are
replaced by very convenient single-use
molds.

5. Cell culture on sapphire discs  Special tools exist; just follow the
supplier’s manual.

6. Embedding after immunostaining  It is possible to follow internalization of
a protein coupled to colloidal gold and to
perform an immunostaining on sections.15



 Procedure 
 In this experiment, the heat shock
protein Hsp-70 coupled to 10 nm colloidal
gold particles was traced in human
dendritic cells.
The cells were then embedded in K4M

Lowicryl using the PLT method. They were
further immunolabeled with anti-MHC
class II or anti-CD63 antibodies that were
revealed by secondary antibodies coupled
to Ultra-small gold. Ultra-small gold
particles were amplified by silver
enhancement.
Figure 18.6 A: Hsp-70 internalization and

silver enhancement. The gold particles are
localized in the so-called MHC class II
compartments.
B: CD63 (open arrow) co-localizes with
Hsp-70 as all the gold particles were silver
enhanced.
4.6. Polymerization
 Polymerization of the resin is
triggered by an activator sensitive to
UV light, which is incorporated in the
resin mixture. The reaction is very  Polymerization 24 to 48 hours under UV
exothermic and controlled by the low light at the embedding temperature.
temperature.
4.7. Further Processing

4.7.1. Room temperature sectioning
 Trouble shooting.  See Chapter 2.

4.7.2. Immunolabeling
 See Chapter 23.
4.7.3. In situ hybridization

1. Principle
The in situ hybridization (ISH) technique is  Use labeled probes (oligonucleotides,
based on the pairing reaction between the cDNA or cRNA). Optimal conditions of in
nucleic acid sequences which are under situ hybridization have been described.17,21
study in a biological material and the  It is also possible to simultaneously
complementary labeled probes of a detect a nucleic acid sequence and an
hybridization solution. Hybridization at the antigen.16
LM level was first described by John et
al.13 and Pardue and Gall.20 It was
developed afterwards for EM cryo-sections
or resin embedded specimens by Morel et
al.18
 Three major types of probes are  They have to be labeled before use. The
commonly used: label depends on the probe chosen.17
 cDNA  Stable probe to be denatured before use.
 Several different oligonucleotides can be
 oligonucleotide(s) used simultaneously.
 Single-stranded probe. The hybrids
 cRNA obtained with these probes are more stable
but these probes are more fragile.
 Three types of antigenic markers  The revelation process depends on the
exist: marker chosen.
 Digoxigenin  Plant molecule that does not exist in

animal specimens (less background). Do
not use it for plant material.
 Fluorescein  In principle, could be observed at the
LM level.
 Biotin  This is the smallest molecule available.
 Endogenous biotin (vitamin H) often
creates a background.
 Three methods are available:  Each has its advantages and drawbacks.
 ISH pre-embedding  Hybridization is done on fixed material

before embedding. Sensitive method, but
time consuming.
 ISH on resin-embedded sections  Easiest method, but is the least sensitive.
After PLT embedding, the sensitivity is
enhanced.
 ISH on cryo-sections according to  Most sensitive method, but a cryo-
Tokuyasu. ultramicrotome is required.

 A general procedure involves the

different steps:
 Labeling the probe  Labeled oligonucleotide probes can be
purchased commercially.
 Tissue preparation and sectioning  As described in this chapter.
 Pretreatment  To render the target accessible.
 Hybridization  To form the complexes between nucleic
acid and probe.
 Washing  Washing out of nonspecific hybrid-
ization label.
 Detection  Immunocytological reaction using
colloidal gold particles.
 Observation  Under an electron microscope
2. Protocol
ISH is carried out on ultrathin Lowicryl  Easiest method.
K4M or HM20 sections (100 nm
thickness) using nucleic acid probes.
 Labeling the probe

 Oligonucleotide  100 pmoles/mL  Labeling a short oligonucleotide
probe of hybridization (30 mers with a GC at the 3´end) is carried
buffer out by the addition at the 3´ end of labeled
nucleotides using an enzyme: terminal
transferase (TdT).
 The labeled nucleotides are carriers of
antigenic molecules.17
 Mixture solution  cDNA should be denatured.
 Add the labeled probe (i.e.,  Oligonucleotides and cRNA probes do
oligonucleotide probe) in the not have to be denatured.
hybridization buffer and vortex  But cRNA should be hydrolyzed to
 Centrifuge produce sequences of around 300 nuc-
 Chill rapidly on ice leotides for the best results.
 Hybridization
 Place the grid on the  30 µL/grid  The volume can be reduced to a few µL.
mixture solution
 Incubation 3 h  Humid chamber.
at RT
 Washing
 2X SSC 3 ×10 min  Necessary.
at RT
 1X SSC 2 × 5 min  Optional.

 Detection  Indirect immunocytological reaction.

 Stabilizing the structures  Immunodetection is carried out on drops
 4% FA in 2X SSC 5 min of 40 to 50 µL.
 Washing
 2X SSC 5 min
 Blocking nonspecific 1030 min  Necessary

sites in blocking buffer
 Incubate the first 60 min  See “Labeling the probe.” above.

antibody antiprobe  40 µL/grid  Raised in species X.
label diluted 1:50 in  Humid chamber.
phosphate/NaCl
buffer
 Washing
 Phosphate/NaCl 2 × 5 min
buffer
 20 mM Tris/300 mM 2 × 5 min
NaCl buffer
 Incubate colloidal 60 min  Antispecies secondary antibodies: IgG or

gold immunoglobulin  20 µL/grid Fab fragments (5mU/mL).
conjugate diluted in  Detection of the complex.
20 mM Tris/300 mM  Humid chamber.
NaCl buffer
 Washing
 TrisHCl/NaCl buffer 2 × 5 min
 2X SSC 5 min
 Complex fixation
 2.5% GA in 2X SSC 5 min
 Staining
 2% Aqueous uranyl 25 min  Uranyl oxalate can be successfully used
acetate (see Chapter 19).
in darkness
 Distilled water 5 × 4 min  Finish washing in a jet of double
distilled water from a wash bottle.
 Observation  See Section 7, Figure 18.8.8

5.1. Advantages
 Fast and easy method to localize  Can localize proteins or protein com-
proteins in situ. plexes inside a cell or an organ.
 Can localize heterologous proteins, e.g.,
expressed with a recombinant virus.
 Can only localize the protein or over-
expressed protein, biochemical experiments
have to be performed to see if the protein is
functional.
 Fast and easy method to localize a  It can only localize; other experiments
gene. have to be performed to see if the gene is
active or not.
 If DNA is destroyed by DNase
treatment, only RNA can be detected. In
this case, the messenger RNA detected
reflects gene activity.
 Lowicryl resin mixtures have the  K4M tolerates 5% water.

advantage of keeping some hydration
in the specimen.
 Lowicryl resin K4M and HM20 can be  75% K4M and 25% HM20, for example.
mixed together to combine their
advantages.
5.2. Disadvantages
 If the gene of the protein to be  Intracellular proteins involved in
localized is poorly expressed, labeling trafficking, such as those localized in rafts
becomes difficult or even impossible. or scarce proteins.
 Protein or nucleic acid localization can
be tedious. In such situations, the Tokuyasu
method is preferred because it is more
sensitive.

6.1. With Regards to the Tokuyasu Immunostaining Method

 Low-cost apparatus  Can be homemade.
 No cryo-sections  An ordinary ultramicrotome can do the
job.
 No experience in cryo-sectioning is
needed.
 Block of resin  Can be sectioned years after it was
prepared.
 Easy storage, no need of a special Dewar
flask and liquid nitrogen.
 Good ultrastructural preservation  In our hands with the ready-to-use
resins, polymerization is reproducible and
the overall morphology comparable to
sections of Epon-embedded specimens.
 Antigenic sites preserved  Perfect immunolabeling if the antigen is
abundant.
 In situ hybridization  PLT is the easiest method to start with
the ISH technique.
6.2. Compared to Cryo-Fixation

 HPF followed by cryo-substitution  Expensive apparatus.
 Time consuming procedure.
 Very little material to be sectioned.
 Specific “how to do” is needed.
 Slam-freezing followed by cryo-sub-  Less material available because the
stitution entire sample is not well preserved.
 Time consuming.
6.3. New Technique

 High-pressure freezing, freeze-  See Chapter 14.
substitution fixation, rehydration,  Takes advantage of cryo-fixation and the
Tokuyasu cryo-sectioning, and classical Tokuyasu method.
immunolabeling  To be considered in case of problematic
samples.


7. OBSERVED RESULTS
 Figure on the Chapter’s title page

1. Description  Immunolocalization of a membrane
protein in a prokaryote.
2. Method  PLT.
3. Sample  Mycoplasma.
4. Fixation  4% FA + 1% GA, 30 minutes at 4°C.
5. Embedding  Lowicryl K4M, 30°C.
6. Visualization for  Goat antirabbit coupled to 10 nm
immunocytochemistry (ICC) colloidal gold.
7. Comments  The colloidal gold particles are found
only on the plasma membrane.
 Figure 18.7
1. Description  Cellular and subcellular distribution of
AQP8 in the gastrointestinal tract.
2. Method  PLT.
3. Sample  Rat hepatocytes.
4. Fixation  4% FA + 0.01% GA, 2 hours at 4°C.
5. Embedding  Lowicryl K4M, 35°C.
6. Visualization for ICC  Goat antirabbit coupled to 10nm
colloidal gold.
7. Magnification  40,000 ×; insets, 130,000 ×.
8. Comments  APQ8 in the gastrointestinal tract of the

rat may be involved in the (1) absorption of
water in small and large intestine, (2)
secretion of canalicular bile and (3)
secretion of pancreatic juice. (Bc) bile
canaliculus, (cyt) cytoplasm. (From
Calamita, et al.2, with permission from
Elsevier.)


 Figure 18.8
1. Description  Co-localization by immunolabeling of
nucleocapsid-like particles in cells infected
with a vaccinia virus recombinant encoding
the measles virus nucleoprotein and
phosphoprotein. (From Spehner, et al. 26,
with permission from Elsevier.)
2. Method  PLT.
3. Sample  Hamster BHK21cells.
4. Fixation  2.5% FA + 0.1% GA, one hour at 4°C.
5. Embedding  Lowicryl K4M at 25°C.
6. Visualization for ICC  Specific antibody: Mouse antimeasles
nucleoprotein; secondary antibody: Goat
antimouse coupled to 5 nm colloidal gold.
 Specific antibody: Rabbit antimeasles
phosphoprotein; secondary antibody: Goat
antirabbit coupled to 10 nm colloidal gold.
7. Magnification  ~ 120,000 ×
8. Comments  Some vaccinia virus particles are visible
in the top right of the picture. The electron
dense mass corresponds to the accum-
ulation of nucleocapsid-like structures.
These structures are labeled with antiNP (5
nm) and antiP (10 nm) antibodies.
 Figure 18.9
1. Description  In situ hybridization.9
Simultaneous detection of the three receptor
subtype mRNAs in distal (A) and proximal
tubules (B).
2. Method  PLT.
3. Tissue  Kidney cortex.
4. Fixation  4% FA for 2 hours at 4°C.
5. Embedding  Lowicryl K4M at 35°C.
6. Visualization for  Goat antimouse coupled to 10 nm
immunocytochemistry colloidal gold particles.
7. Scale bar  0.5 µm.
8. Comments  The 10 nm gold particles were localized
in the cytoplasmic matrix near the nucleus
(N), but not in mitochondria (m). (From
Grandclement and Morel8, with permission
from Portland Press.)


8. REFERENCES
1. Acetarin, J.D., Carlemalm, E., and Villiger, W. Developments of new Lowicryl

resins for embedding biological specimens at even lower temperatures, J. Microsc.,
143, 81, 1986.
2. Calamita, G. et al. Expression and immunolocalization of the aquaporin-8 water
channel in rat gastrointestinal tract, Eur. J. Cell Biol., 80, 711, 2001.
3. Carlemalm, E., Colliex, C., and Kellenberger, E. Contrast formation in electron
microscopy of biological material, in Advances in Electronics and Electron
Physics,, 1985, 269.
4. Carlemalm, E., Garavito, R.M., and Villiger, W. Resin development for electron
microscopy and an analysis of embedding at low temperature, J. Microsc., 126, 123,
1982.
5. Carlemalm, E. and Kellenberger, E. The reproducible observation of unstained
embedded cellular material in thin sections: Visualisation of an integral membrane
protein by a new mode of imaging for STEM, Embo J., 1, 63, 1982.
6. Fernandez-Moran, H. The fine structure of vertebrate and invertebrate
photoreceptors as revealed by low temperature microscopy, in The Structure of the
Eye, Smelser, G.K., ed., Academic Press, New York, NY, USA 1961, 521.
7. Glauert, A.M. and Young, R.D. The control of temperature during polymerization
of Lowicryl K4M: There is a low-temperature embedding method, J. Microsc., 154,
101, 1989.
8. Grandclement, B. and Morel, G. Ultrastructural characterization of atrial
natriuretic peptide receptors (ANP-R) mRNA expression in rat kidney cortex, Biol.
Cell, 90, 213, 1998.
9. Grandclement, B., Ronsin, B., and Morel, G. The three subtypes of atrial
natriuretic peptide (ANP) receptors are expressed in the rat adrenal gland, Biol. Cell,
89, 29, 1997.
10. Griffiths, G. Fine Structure Immuno-Cytochemistry. Springer-Verlag, Berlin,
Heidelberg, New York, 1993.
11. Hayat, M.A. Glutaraldehyde: Role in electron microscopy, Micron Microsc. Act, 1,
115, 1986.
12. Hobot, J.A. Lowicryls and low temperature embedding for colloidal gold methods,
in Colloidal Gold: Principles, Methods and Applications, Hayat, M.A., ed.,
Academic Press, San Diego, CA, USA, 1989, 75.
13. John, H.A., Birnstiel, M.L., and Jones, K.W. RNA-DNA hybrids at the
cytological level, Nature, 223, 582, 1969.
14. Kellenberger, E. et al. Low Denaturation Embedding for Electron Microscopy of
Thin Sections. Chemische Werke Lowi G.m.b.H., West Germany, 1980.
15. Lipsker, D. et al. Heat shock proteins 70 and 60 share common receptors which are
expressed on human monocyte-derived but not epidermal dendritic cells, Eur. J.
Immunol., 32, 322, 2002.
16. Morel, G. In situ hybridization on ultrathin frozen section, in Hybridization
Techniques for Electron Microscopy, Morel, G., ed., CRC Press, Boca Raton, FL,
USA, 1993, 163.
17. Morel, G., Cavalier, A., and Williams, L. In Situ Hybridization in Electron
Microscopy, ed. Morel, G., CRC Press, Boca Raton, FL, USA, 2001.

18. Morel, G., Dubois, P., and Gossard, F. Ultrastructural detection of messenger
RNA coding for growth hormone in the rat anterior pituitary by in situ
hybridization, C. R. Acad. Sci. III, 302, 479, 1986.
19. Newman, J.R. and Hobot, J.A. Resin Microscopy and On-Section
Immunocytochemistry. Springer-Verlag, Berlin, Germany, 1993.
20. Pardue, M.L. and Gall, J.G. Molecular hybridization of radioactive DNA to the
DNA of cytological preparations, Proc. Nat. Acad. Sci. USA, 64, 600, 1969.
21. Puvion-Dutilleul, F. and Puvion, E. Non-isotopic electron microscope in situ
hybridization for studying the functional sub-compartmentalization of the cell
nucleus, Histochem. Cell Biol., 106, 59, 1996.
22. Robertson, D. et al. An appraisal of low-temperature embedding by progressive
lowering of temperature into Lowicryl HM20 for immunocytochemical studies, J.
Microsc., 168, 85, 1992.
23. Roth, J. et al. Enhancement of structural preservation and immunocytochemical
staining in low temperature embedded pancreatic tissue, J. Histochem. Cytochem.,
29, 663, 1981.
24. Roth, J., Taatjes, D.J., and Tokuyasu, K.T. Contrasting of Lowicryl K4M thin
sections, Histochemistry, 95, 123, 1990.
25. Sitte, H. et al. A new versatile system for freeze-substitution, freeze-drying and low
temperature embedding of biological specimens, Scan. Microsc. Suppl., 8, 41, 1994.
26. Spehner, D., Drillien, R., and Howley, P.M. The assembly of the measles virus
nucleoprotein into nucleocapsid-like particles is modulated by the phosphoprotein,
Virology, 232, 260, 1997.
27. Whitehouse, R.L. et al. Immunolabelling of bacteriophage lambda receptor protein
(LamB) on thin sections of E. coli embedded in Lowicryl, Biol. Cell, 51, 389, 1984.

Cryo-Sectioning according to Tokuyasu 469
CONTENTS

1. PRINCIPLES OF CRYO-SECTIONING ACCORDING TO TOKUYASU472
3.1. Materials ................................................................................................... 475
3.2. Products .................................................................................................... 476
3.3. Solutions ................................................................................................... 477
4. PROTOCOLS .................................................................................................... 480
4.1. Preparation of Animal Cells and Tissue ................................................... 480
4.2. Preparation of Microorganisms ................................................................ 481
4.3. Preparation of Plant Tissue....................................................................... 483
4.4. Cryo-Sectioning........................................................................................ 484
4.5. Staining and Methyl Cellulose Embedding .............................................. 489
4.6. Cleaning Diamond Cryo-Knives .............................................................. 489
5.1. Advantages ............................................................................................... 490
5.2. Disadvantages........................................................................................... 491
6. WHY AND WHEN TO USE CRYO-SECTIONING ACCORDING TO
TOKUYASU ....................................................................................................... 491
8. REFERENCES .................................................................................................. 496


The desire to visualise biological ultrastructure in its aqueous surrounding is more than
50 years old. One of the great pioneers of cryo-electron microscopy, Humberto
Fernández-Morán,1 described the first cryo-ultramicrotome in 1952.2 In the succeeding
two decades, Wilhelm Bernhard,3 Christensen,4 Dollop and Sitte,5 and Sevéus6 and
Barnard constructed the first functional cryo-microtomes, but it was not until the early
1970s that the technique had its breakthrough, initiated and propagated by Kiyoteru
Tokuyasu.7-9 He made cryo-ultramicrotomy of chemically fixed and cryo-protected
samples a standard method for sensitive immunogold labelling.10
Since then, the technique was further improved and brought to perfection by the group of
Jan Willem Slot and Hans Geuze.11-13 This process of perfection would not have been
possible without the new design and development of high-precision, reliable and easy to
use cryo-ultramicrotomes.14 In general, today, Tokuyasu cryo-sectioning is as easily done
as resin sectioning with the same quality. Even serial sectioning and sectioning parallel
to the substrate are no longer impossible tasks.12,15 Today, the Tokuyasu cryo-sectioning
method has become a standard method in combination with highly efficient immunogold
labelling.12
Unfortunately, not all specimens are equally well suited for this method. Efficient and fast
chemical fixation is impeded in samples containing air or large vacuoles (leaves), cell
walls (plants, fungi) or cuticles, hydrophobic surfaces (insects, nematodes); and cryo-
sectioning is more difficult with samples that exhibit differing internal stiffness, e.g., due
to extracellular cell walls or cuticle-like barriers in plant, arthropod, and nematode tissues.
Therefore, in plant as well as in nematode and arthropod research, there are only a few
labs with a limited number of examples of successful application of the Tokuyasu cryo-
sectioning method.16-18 Especially in these cases, a novel technique, combining cryo-
fixation and ultrathin cryo-section labelling according to Tokuyasu offers a number of
potential advantages19 (see Chapter 14).
Despite the drawbacks for a few biological specimens, cryo-sectioning according to

Tokuyasu and described in this chapter, is the method of choice for most
immunolabelling applications.

1. PRINCIPLES OF CRYO-SECTIONING ACCORDING TO

TOKUYASU
 The biological material is chemically  Chemical fixatives cannot penetrate

fixed with low concentrations of deeply into dense tissue.
aldehydes.  Mammalian tissue is best preserved after
a perfusion fixation. If perfusion fixation is
not possible, the samples are cut in small
1 mm3 pieces in the presence of fixatives.
 Tissue culture cells are fixed in situ and
then scraped from the culture wells with a
rubber policemen. They can also be used
for flat-embedding according to a special
protocol.15

 The suspended cells are collected by  Caution: For scraped tissue culture
centrifugation in 1.5 mL microtubes cells, e.g., human foreskin fibroblasts
and washed several times in buffer. (HFF), 1% gelatine or 1% bovine serum
albumin (BSA) should be added to the
washing buffer to avoid sticking of the cell
layers to the tube walls.15
 Cells are pelleted into 10 to 12%  Tissue blocks are also infiltrated with
gelatine, which is solidified on ice, 12% gelatine to fill empty spaces. Large
and then small cubes, not more than 1 differences in matrix densities have a
mm3, are cut. negative influence on sectioning.
 The cubes of tissue or cells containing  Sucrose or sucrose/PVP impregnation is
gelatine are infiltrated with 2.1 to 2.3 important to avoid ice crystal damage
M sucrose or a mixture of polyvinyl during freezing and to make sure that the
pyrrolidone (PVP) and sucrose on a ice has good sectioning properties7,9 (see
slowly rotating (4 rpm) spinning wheel also Chapter 11).
at 4°C. Impregnation is done for 2 to
24 hours until the blocks do not float
to the surface anymore.
 If needed, the blocks can be cut  The smaller the block the less trimming
smaller and mounted on pins that can is needed.
be fit into the holder of the microtome.  Due to the high sucrose concentration,
Remove the excess of sucrose solution the sample will vitrify independently of the
with a filter paper. Then the sample is way freezing is carried out.
frozen either in liquid nitrogen or in
the sectioning chamber of the cryo-
ultramicrotome.

 Semithin and ultrathin sections are cut  The optimum temperature has to be
with either glass or diamonds knives at found experimentally. It is mainly depen-
temperatures between 80°C to dent on the desired thickness of the
140°C. sections, but also on the texture of the
sample.
 During trimming and sectioning an  The sectioning process causes elec-

ioniser is directed toward the knife trostatic charging, which interferes with the
edge. The intensity is adjusted so that gliding properties of the sections on the
the sections glide from the knife edge knife surface. Sticking to the surface can
without touching the surface, but do cause severe compressions (see also
not fly away. Chapter 11)!
 The sections are retrieved from the  The exact timing of gently touching the
knife surface with a loop containing a cryo-sections with the surface of the
drop of pick-up solution and trans- freezing drop is important for the structural
ferred to room temperature on a integrity of the cryo-section. It has to be
Formvar- or Pioloform-carbon-coated found experimentally.
electron microscopy grid or on
coverslips or silane-coated glass slides
for light microscopy applications.
 At this step immunolabelling can be  See Chapter 21, 23.

performed.
 Then the sections are stained with  Methyl cellulose prevents drying
uranyl acetate and embedded in artefacts and irregular shrinkage of the
methyl cellulose. different cell organelles.
 After drying, the sections are ready for

analysis by electron microscopy.
 For light microscopy, coverslips with  See Chapter 21.

sections are mounted on a slide in a
drop of mounting medium containing
an antifading reagent.

1. Chemical prefixation
 Alternative: cryo-fixation (see Chap-
ters 5, 6, 14)
2. Gelatine impregnation (optional)
3. Cryo-protection
4. Mounting/freezing (a,b)
5. Trimming (c: Top view on specimen

holder (stub) with frozen sample)
6. Semithin cryo-sectioning for light

microscopy (LM) (d)
7. Ultra-thin cryo-sectioning for EM

(d)
8. Section retrieval (e)
9. Transfer to grid (EM) or coverslip

(LM) (f,g)
10. Immunolabelling (h)
11. Staining and methyl cellulose

embedding (ultrathin cryo-sections)
(i) or
 Moviol embedding (semithin cryo-
sections)
Figure 19.1 (a-i) Outline of the cryo-

sectioning and immunolabelling procedure
of thin, thawed cryo-sections according to
Tokuyasu.

3.1. Materials
 Cryo-ultramicrotome  Leica EM UCT/EM FCS or EM
UC6/EM FC6, Leica Microsystems, Vienna
Austria.
 Knife maker  Leica Knifemaker, Leica Microsystems,
Vienna Austria. Please check the quality of
the knife maker carefully before buying!
 Glass knives  Made from glass strips: Gumhag shi 6.4
x 25/30, Glass Ultra Micro Co., Stockholm,
Sweden; Glass strips (several sizes), Leica
 For breaking glass knives (see Reference
20, pp 145152).
 Diamond cryotrim  Diamond trimming blades with a
trimming angle of 20° and 45°
(cryotrim 20, cryotrim 45; Diatome AG,
Bienne, Switzerland).
 Diamond cryo-knives  There are special cryo-knives available
with different cutting angles, 25°, 35°, 45°,
e.g., Diatome AG, Bienne, Switzerland and
Element Six, Cuijk, The Netherlands.
 Forceps  For example, Dumont #2a, #3, #4, #4N,
#7, Dumont & Fils, Switzerland.
 Loop  Homemade loops from remanium wire
or perfect loop, EMS, Fort Washington,
Pennsylvania, USA.
 Rubber policemen/cell scraper or cell  Costar #3010, Corning B.V.; Corning
lifter New York, USA, Schiphol-Rijk, The
Netherlands.
 Costar #3008, Corning B. V.
 Table centrifuge  Swing-out centrifuge, e.g., Rotima 35,
Andreas Hettich GmbH&Co. KG.,
Tuttlingen, Germany.
 A centrifuge with six different swing-out
elements allows for more combinatorial
freedom.
 Microfuge  Centrifuge 5415C, Gerätebau Eppendorf
GmbH, Enelsdorf, Germany.
 Microtubes  0.5 and 1.5 mL Eppendorf-Netheler-
Hinz GmbH, Hamburg, Germany or
Sarstedt AG & Co, Nümbrecht, Germany.
 Razor blades  Double-edged razor blades are best
suited for trimming the blocks.

 Filter paper  Coffee filter paper or Whatman 50,

Whatman International Ltd, Maidstone,
England.
 H2O  Double distilled or MilliQ water.
 Coverslips  For polylysine coating, see Chapter 21.
 Glass slides  For immunolabelling, they may be
silane-coated.
 Pioloform- or Formvar-coated and  Copper, nickel or gold grids, 50 to
carbon-covered grids. 400 mesh: e.g., Stork Veco B. V., Eerbeek,
The Netherlands; or Gilder Grids,
Grantham, England.
 For coating, see Reference 21.
3.2. Products
 Gelatine  From the supermarket, e.g., Dr. Oetker.

 BSA, bovine serum albumin,  A-7906, Sigma-Aldrich: Fluka, Buchs,
Fraction V Switzerland.
 Uranyl acetate  Stain for electron microscopy. EMS,
(UO2(CH3COO)2 × 2H2O) Fort Washington, Pennsylvania, USA or
USA.
 Methyl cellulose  25 centipoises, Sigma M-6385.
 Fluorescent secondary antibodies  Against the species in which the primary
antibody was produced.
 Provided by several companies, e.g.,
Jackson Immunoresearch Europe Ltd,
Suffolk, U. K.
 Gold-labelled secondary antibodies  Against the species in which the primary
Dianova: Hamburg, Germany; Aurion,
Wageningen, The Netherlands; British
BioCell International (BBI) Ltd, Cardiff,
U.K. For 1 nm gold markers, see Chapter
23.
 Gold-labelled protein A  Binds preferentially to the Fc region of
rabbit, pig, human and mouse (subclasses
IgG2a and IgG2b) antibodies (for details see
Reference 20 pp 320323).
 Can easily be homemade22,23 or
purchased from Dr. G. Posthuma, Depart-
ment of Cell Biology, Utrecht Medical
Centre, Utrecht, The Netherlands.
Dianova, Aurion, and BBI, see above..

 Paraformaldehyde
 Disodium hydrogenphosphate  Na2HPO4.
 Polyvinyl pyrrolidone (PVP, MW  Sigma PVP10-100G.
10,000)
 Sodium carbonate  Na2CO3.
 Sucrose  C12H22O11.
 Sodium chloride  NaCl.
 Potassium chloride  KCl.
 Glycine
 PIPES  1,4-Piperazine-bis(ethanesulfonic acid).
 HEPES  2-[4-(2-Hydroxyethyl)-1-piperazine] eth-
anesulfonic acid.
 Magnesium chloride  MgCl2.
 Magnesium sulphate  MgSO4.
 Potassium chloride  KCl.
 EGTA  Ethyleneglycol- bis(β-aminoethyl)-N,N,
N’,N’-tetracetic acid.
 MES  2-(N-Morpholino)-ethanesulfonic acid.
 Oxalic acid
 Potassium dihydrogen phosphate  KH2PO4.
 Glutaraldehyde (aqueous, e.g., 8%)  For example, Sigma #G-5882.
 8% solution, Polysciences, # 00216A.
 Keep at 4oC.
 Sodium metaperiodate  NaIO4; Merck #106597.
 Periodic acid  H5IO6; Merck #10052.
 Ethanol
 Triton X-100  Nonionic detergent, e.g., Fluka; Buchs,
Switzerland.
 Citric acid monohydrate  Merck 231211.
3.3. Solutions
 137 mM sodium chloride NaCl buffer.
 2.7 mM potassium chloride KCl
 8.1 mM disodium hydrogenphosphate
Na2HPO4
 1.5 mM sodium dihydrogen phosphate
NaH2PO4
 PB buffer (200 mM, pH 7.4)  Phosphate buffer, a standard buffer.
 162 mM disodium hydrogen-  Make two stock solutions of 200 mM:
phosphate Na2H PO4  Buffer A: 27.6 g NaH2PO4 x 1H2O in
 38 mM sodium dihydrogen phosphate 1 L H2O
NH2 PO4  Buffer B: 35.6 g Na2HPO4, 2H2O in
1 L H2 O

 Add 19 mL buffer A to 81 mL buffer

B to get 200 mM phosphate buffer; pH
7.4. Check pH and if necessary add
more of the appropriate buffer to get
pH 7.4.
 PHEM buffer (pH 6.9)  Cytoskeleton buffer.24

 60 mM PIPES
 25 mM HEPES
 10 mM EGTA
 2 mM MgCl2
 Citrate buffer  For low pH 1.55.0.

 200 mM citric acid monohydrate  Used as a primary fixative for difficult to
fix organisms, e.g., Dictyostelium
discoideum.
 Buffers for plants  Often relatively low salt buffers.

 Phosphate buffer
100 mM potassium phosphate buffer;
pH 7.025
 MTSB (microtubule stabilising  Cytoskeleton buffer.

buffer)26
 15 g PIPES (50 mM)
 1.9 g EGTA (5 mM)
 0.6 g MgSO4 (5 mM)
 2,5 g KOH (solid) in 1 L H2O
 pH 6.97.0
 MS medium (pH 5.8)  Plant culture medium.27

Including micro and macro elements  Duchefa Biochemie B.V., Haarlem, The
(0.5 ×). Netherlands.
 Dissolve 2151.04 mg medium in 1 L
H2O.
 Fixatives  For a 16% stock solution, suspend 16 g

 2 to 8% (w/v) formaldehyde paraformaldehyde in 45 mL H2O, warm to
about 60C (au bain Marie) and stir
 2% (w/v) formaldehyde and 0.20.02% vigorously for some time. Then add 20 to
(v/v) glutaraldehyde 80 l 1 M NaOH and wait for about
 In PB, PHEM, or 200 mM 10 minutes. If the paraformaldehyde has not
HEPES; pH 7.2 or dissolved into formaldehyde, repeat the
 In MTSB; pH 6.97.0 or PB; NaOH step. Cool down on ice, filter and
pH 7.0 adjust volume with H2O to 50 mL.
If desired, glutaraldehyde is added to the
working dilution of formaldehyde in buffer
(HEPES pH 7.2, PB, PHEM, MTSB),
check the final pH.
 Note: Fixatives can be stored frozen in
aliquots.

 2.5 parts 8% formaldehyde in H2O  This fixative can be useful for difficult to
 1.5 parts of a saturated solution of fix samples like yeast (see Figure 19.13) or
picric acid in H2O Dictyostelium discoideum.28
 5.0 parts citrate buffer  Note: The cytoplasm often gets a
 Add H2O to make 10 parts grainy appearance in the presence of
 Adjust pH to the pH of the growth picric acid.
medium.
 1 to 12% (w/v) gelatine in PBS  Suspend in PBS, warm to about 60°C,

stir until the gelatine powder has dissolved.
Cool to 37oC, add sodium azide, aliquot (ca
2 mL) and store at 4oC.
 2.3 M sucrose in PBS or PHEM  Weigh 78.73 g sucrose in a 100 mL

buffer volumetric flask. Add PBS or PHEM buffer
pH 7.4 while stirring and wait until the
sucrose has dissolved. Remove the stir bar
and top up with the buffer to the 100 mL
mark. The mixture is stored at 4oC or frozen
in 10 to 20 mL aliquots.
 Polyvinyl pyrrolidone (PVP)-  Infiltration time is longer compared to

sucrose9 pure sucrose.
 20% PVP  To make 100 mL, mix 20 g PVP and
 1.84 M sucrose 4 mL of 1.1 M disodium carbonate
in phosphate carbonate buffer (Na2CO3), in 0.1 M disodium hydrogen
phosphate (Na2HPO4) and 80 mL of 2.3 M
sucrose in 0.1 M Na2HPO4.
 Pick-up solution7,12
 2.3 M sucrose or  Reduces mechanical damage induced by
 1% (w/v) methyl cellulose, 1.15 M spreading of sections during thawing on the
sucrose in PHEM buffer surface of the 2.3 M sucrose drop.
 Alternatively, one can use a mixture of
one part sucrose in PBS and one part PVP-
sucrose to reduce spreading upon thawing.
 Staining solution for transmission
electron microscopy (TEM)  To avoid vesiculation of cellular
 Neutral uranyl acetate (uranyl oxalate) membranes.8
2% (w/v) uranyl acetate in 0.15 M  Stir a 0.3 M oxalic acid solution in H2O
oxalic acid; pH 7.4 vigorously and add slowly the same volume
of a 4% (w/v) aqueous uranyl acetate
solution. Adjust pH to 7.4 with 10%
ammonium hydroxide.8
w
 2% ( /v) aqueous uranyl acetate  Make 4% to add to the methyl cellulose
for embedding.

 Staining solution (LM)  See Chapter 21.

 Toluidine Blue O
 Methyl cellulose  Dissolve methyl cellulose in prewarmed

 2% (w/v) methyl cellulose in H2O H2O and stir in the cold for several hours.
(25 centipoises)10 Centrifuge the solution for 60 min at
100,000 g. Store in the refrigerator.
 Embedding solution (TEM)

 Methyl cellulose + uranyl acetate
 9 parts methyl cellulose  More uranyl acetate can be added to
 1 part 4% (w/v) aqueous uranyl increase contrast.
acetate29
 Embedding solution (LM)  See Chapter 21.
 Moviol
4. PROTOCOLS
4.1. Preparation of Animal Cells and Tissue
1. Chemical fixation of culture cells

 Cells are grown in culture dishes or  Routinely, we use two fixatives: One
culture flasks to about 70% with formaldehyde only and one that
confluency. contains formaldehyde and glutaraldehyde.
 Preferentially, growth medium is
refreshed one day before fixation
 Cells in culture are fixed for 1 to 2  Fixative is made in phosphate, PHEM or
hours at room temperature by adding HEPES buffer. PHEM and HEPES buffers
equal volumes of double-strength have higher buffering capacity and
fixatives to the growth medium. generally result in better structural
 Then the cells are scraped in the preservation.20
presence of 1% gelatine in buffer15 to  For adherent cells, the medium may be
reduce damage and loss of cells due to replaced by the fixative solution and the
adherence to plastic or glass surfaces cells are fixed for 1 to 2 hours at room
or to the walls of microtubes. temperature.
 Suspension cells are directly fixed
with the same volume of double-
strength fixative.
 The cells are collected in a centrifuge  1000 rpm in a microcentrifuge.
tube and pelleted at about 800 to1200
g.
 Then, they are washed three times
with buffer containing 1% gelatine.

2. Chemical fixation of tissue

 If possible, the tissue is fixed by  Perfusion fixation is preferred because
perfusion fixation. tissue is very sensitive to the postmortem
 Otherwise, the tissue is sampled, alterations and damage by excision.
submersed into the fixative and cut
into small pieces of about 1 mm3.
 The tissue pieces are fixed for 1 to 2
hours at room temperature.
 Thereafter, they are washed three
times with buffer.
3. Gelatine impregnation
 The cells are pelleted into gelatine in  The gelatine density should match the
PBS or PHEM buffer and the gelatine sample density; we usually use 10 to 12%.
is solidified on ice.  The pellet should be loose. In dense
 Small blocks of not more than 1 mm3 pellets, the cells are closely packed and
are cut from the gelatine-embedded individual cells are difficult to photograph.
cells. The plasma membrane labelling may also
be reduced.
 The pieces of tissue are also im-  Tissue blocks are infiltrated with
pregnated with 12% gelatine for 30 gelatine to fill spaces, e.g., capillaries, with
minutes at 37oC on a rotating wheel. protein. Large differences in matrix
Before gelling, the excess gelatine is densities can also impair the sectioning
drained off with a piece of filter paper process.
or tissue.
4.2. Preparation of Microorganisms

1. Chemical fixation of bacteria
 Bacteria are grown in growth medium  Note that the fixative solution should
or on agar plates to the desired mimic the culture conditions as closely as
density. possible, therefore, fix with the pH of the
 An equal volume of double-strength growth medium at the growing temperature.
fixative is added to the growth For acid or basic pH, appropriate buffers
medium. Cells grown on agar plates have to be chosen.
are collected with a spatula or a  Adding equal volumes of double-
bacteria loop and suspended in single- strength fixatives to the growth medium is
strength fixative. the preferred initial fixation.
 After about 15 to 30 minutes the cells

are pelleted at 800 to 1200 g and
resuspended in fresh single strength
fixative.
 Fixation is continued for 1 to 2 hours
overall.
 The fixed cells can be stored in 1%
formaldehyde at 4oC.

2. Chemical fixation of yeast cells and

other fungi
 Yeast cells (fungi) are grown in  For yeast cells it seems to be even more
culture medium or on agar plates to important that the fixative solution mimics
the desired density. the culture conditions. If disturbed by a
 An equal volume of double-strength change of environment or even by
fixative is added to the growth centrifugation, yeast cells can close their
medium (see above) and stirred well. membrane pores and the fixation process
Cells grown on agar plates are will be delayed. As a result, yeast cells are
collected with a spatula or a bacteria not well fixed and can lose a large amount
loop and suspended in single-strength of cellular content.
fixative.
 After about 15 to 30 minutes the cells  4000 rpm in a Hettich swing-out
are pelleted at about 2500 × g and centrifuge.
resuspended in fresh single-strength  After initial fixation, the pH should be
fixative at pH 7.2. raised to 7.2 because at low pH aldehydes
do not fix very well.20
 Fixation is continued for 1 to 2 hours
overall.
 After the cells are washed in buffer,
then fixed cells can be stored at 4°C or
even shipped in 1% formaldehyde.
 Now the cells are washed three times  The treatment with periodate30 (De
in buffer, e.g., PHEM, and then Maziere, A.M., personal communication)
incubated in 1% periodic acid in buffer reduces detaching of the cell wall during
for onr hour at ambient temperature. sectioning. Frequently, the cell wall nicely
Alternatively, they are treated with 1% lines the plasma membrane in contrast to
(w/v) sodium metaperiodate for one standard fixation methods where the cell
hour at 4°C. wall detaches and lies on the section like
twisted bicycle tires.
 Under these conditions, the oxidation
process is very limited and the glucane
chains in the cell wall can still be labelled
with specific antibodies31 (see Figure
19.13).
 The cells are pelleted in 12% gelatine
in PBS or PHEM buffer and the
gelatine is solidified on ice.
 Small cubes or prisms of not more
than 1 mm3 are cut from the gelatine-
embedded cells.

4.3. Preparation of Plant Tissue

1. Chemical fixation of tissue without
enclosed air
 Seedlings grown on agar are  For example, meristem tissue like root
transferred to single strength fixative tips.
 Seedlings treated with reagents, e.g.,  Samples are fixed either with 4% or 8%
dissolved in half-strength (0.5 ×) MS formaldehyde or with a mixture of
medium27 (pH 5.8) are fixed by formaldehyde and glutaraldehyde, in a
adding the same amount of double relatively low ionic strength buffer
strength fixative in PB, or MTSB. (100 mM PB, MTSB).
 After a few minutes, the seedlings are  Highly vacuolated cells (with high
transferred to fresh single-strength internal pressure) may collapse.
fixative at ~ pH 7.0.  In order to enhance wetting of samples
and to facilitate fixative diffusion into
cells/tissues covered by hydrophobic
surfaces (e.g., ovules, anthers/pollen), the
fixation buffer can be supplemented with
low concentrations of detergent (e.g.,
0.01% Triton X-100).
Caution: Detergent treatment increases
extraction!
 In case of detaching of the cell wall
during sectioning and section transfer,
sodium periodate or periodic acid treatment
may help (see Section 4.2.2).
2. Chemical fixation of tissue con-
taining air
 The tissue of interest is transferred to  Samples are fixed either with 4% or 8%
single strength fixative and is cut in formaldehyde or with a mixture of
smaller pieces (1 to 2 × 1 to 2 mm). formaldehyde and glutaraldehyde, in a
relatively low ionic strength buffer (PB,
MTSB).
 Degassing during fixation (pressure  Removes air and improves fixative
should not be lower than 200 to 300 diffusion.
mbar to prevent boiling of buffer).
 Tissue pieces are embedded in 12%  Optional, may improve sectioning
gelatine. properties (see 3. in Section 4.1).
 May be done in several steps, e.g., 2%,
4%, 8%, and 12% gelatine, to reduce
deleterious osmotic effects, such as vacuole
collapse and shrinkage of highly vacuolated
plant tissue.

1. Sucrose impregnation
 The small blocks of tissue or gelatine  After successful impregnation, the
embedded cells are put into a blocks float in the solution or sink to the
microtube containing 2.3 M sucrose. bottom. Under no circumstances should
The tube is mounted on a rotating they swim.
wheel and the tubes are rotated head  Infiltration in several steps, e.g., 0.7 M,
over. 1.4 M, 2.3 M, may reduce possible osmotic
 Impregnation is done for 4 to 20 hours effects, such as vacuole collapse and
in a cold room at 4oC. shrinkage of highly vacuolated plant tissue.
 Alternatively, PVP-sucrose may be used,
if gelatine embedding was omitted. PVP
does not permeate cells and may reduce
differences in matrix densities, which can
impair the sectioning process.9
2. Mounting the sample
 The gelatine or tissue blocks are cut  The smaller the block, the shorter the
smaller if needed (in cubes or prisms) time of trimming.
and mounted on the stubs of the cryo-  Please remember that the thinner the
ultramicrotome. section has to be cut, the smaller the length
 Drain off the sucrose solution from the of the edge: 600 μm edge for 200 nm
surface of the block, but retain a section; 400 μm for 100 nm; 300 μm for 80
conical-shaped portion around the nm and 200 μm for 60 nm, 150 μm for 40
bottom of the block at the stub. Then nm and 100 μm for 30 nm (H. Gnägi,
the block is frozen on the stub either Diatome AG, personal communication).
by immersing it in liquid nitrogen or in  The conical-shaped drop stabilises the
the cold atmosphere of the block during sectioning.
ultramicrotome.  Excess sucrose can cause cracking and
the block may break off.
 Before mounting, the surface of the stub
is scratched, then the stubs are cleaned with
acetone in a sonicator to improve
attachment of the sample.
 In the case of PVP-sucrose, samples
have to be processed very rapidly to prevent
evaporation.
 In contrast to cryo-fixation, here slower
freezing is preferred to prevent fast changes
in material shrinkage during the freezing
process. Unequal shrinkage of the metal
and the sample can lead to breaks at the
stub — block interface and loss of the
sample. The high sucrose content
guarantees vitrification even at low freezing
rates.
Figure 19.2 Mounting and freezing.

3. Trimming
 The temperature of the cryo-
ultramicrotome is set to about 80oC
to 100oC.
 The tip of the ioniser is placed close to  To avoid sticking of the sections to the
the blade and set to full power. surface of the trimming tool.
 The blocks are trimmed to the desired  With the glass knife, the angle of the
size either with a glass knife or the pyramid will be 90o; with the cryotrim 20;
diamond cryotrim. the angle will be 70°; and with cryotrim 45,
 Trimming can be done at high speed it will be 45°. The smaller the angle the
with a feed of a few hundred nano- more stable the resulting pyramid, but the
metres. faster the sections grow in size.
Figure 19.3 Trimming with the cryotrim

diamond knife. Top view on the frozen
gelatine block mounted in the pin/stub.
4. Semithin sections
ultramicrotome is set to about 100°C,
for PVP-sucrose to about 80°C.
 The tip of the ioniser is placed close to  Semithin sections are very useful for
the blade and set to full power. correlative light and electron microscopy
 The knife, glass knife or cryo-diamond (see Chapter 21), but also for screening new
knife slowly approaches the face of the antibodies and orientation in the tissue.
pyramid.
 The reflection of the knife edge mirrored
in the clean surface of the pyramid shows
whether the knife is parallel to the pyramid
and indicates the distance from it.
 The sections are cut with a nominal  During sectioning, the strength of ioniser
feed of 200 to 400 nm at a speed of is adapted by the power, but also by
about 1 mm/sec. changing the distance to the knife edge so
 With an eyelash mounted on a wooden that the sections float above the knife
stick, some sections are brought surface but do not fly away. When the
together. sections stick to the surface, it is advisable
to use a fresh edge of the knife.

 A drop of pick-up solution is mounted  Caution: The ioniser has to be

in a loop and held into the cryo- switched off during section retrieval.
chamber close to the sections. At the  Electrical contact between the ioniser
moment a drop of 2.3 M sucrose and the microtome can damage the circuits
freezes, indicated by a thin trail of fog, of the cryo-ultramicrotome!
the sections are touched and will stick  The ioniser also changes the freezing
to the drop. With a mixture of behaviour of the pick-up solution and
sucrose/methylcellulose the sections impairs section pickup.
are picked up before the drop has  When the drop is to close to the sections,
completely frozen and turned white. they may jump to the drop by electrostatic
interactions.
Figure 19.4 Semithin cryo-sectioning and

retrieval or pickup of the sections. The
sections are arranged on the knife with an
eyelash.
 The loop is brought to room

temperature and thawed. Depending
on the pick-up solution the sections
spread more or less evenly on the
surface of the melting drop.
 Then the sections are placed on a
coverslip or glass slide.
Figure 19.5 Transfer of thawed cryo-

sections to ambient temperature and
mounting on a coverslip, which is placed
upside down on a drop of buffer for
subsequent immunolabelling. Alternatively,
coverslips can be handled as described in
Chapter 21.
 The sections can be treated with

common histological stains, e.g.,
Toluidine blue; or for immuno-
fluorescence studies and correlative  Observe preferably with bright-field or
microscopy, (see Chapter 21). phase contrast illumination.

Figure 19.6 Immunolabelling for LM in a

moist chamber on Parafilm (see Chapter 21,
23).
5. Ultrathin sections
ultramicrotome is set between 120°C
and 140°C, for PVP-sucrose to about
110 to 115°C.
 The tip of the ioniser is placed close to  The reflection of the knife edge mirrored
the blade and set to full power. in the clean surface of the pyramid shows
 The knife, glass knife or cryo-diamond whether the knife is parallel to the pyramid
knife should slowly approach the face and indicates the distance from it.
of the pyramid.
 The sections are cut with a nominal  The Diatome diamond knives come with
feed of 50 to 100 nm at a speed of an indication of the preferential cutting
about 1 mm/sec. speed. The optimum cutting speed,
however, is dependent on the material, the
temperature, the thickness of the section,
the quality of the knife and has to be
determined empirically.
 A drop of pick-up solution is mounted  Caution: The ioniser has to be switched

in a loop and the sections are picked off during section retrieval!
up as described above.  Caution: The methyl cellulose/sucrose
solution freezes much faster than sucrose
alone.
 The loop is brought to room  In 2.3 M sucrose, the sections spread
temperature and thawed. Depending considerably, a phenomenon that can lead
on the pickup solution, the sections to severe changes of the ultrastructure. In
spread more or less evenly on the the methyl cellulose/sucrose solution,12 the
surface of the melting drop. ultrastructure is of much superior quality.
 The methyl cellulose/sucrose solution
should be used with perfect sections only,
compressions and folding of the section
will remain.

 The moment of pickup is important, too

early may lead to folding of the sections
and too late may mechanically destroy the
sections or they may not attach at all.
Figure 19.7 Picking up/retrieval of ultra-

thin cryo-sections, after removing them
from the knife edge with the help of an
eyelash.
 Then the sections are placed on a  Copper grids can be used for
Formvar- or Pioloform-carbon-coated morphological analysis or for immuno-
electron microscopy grid. Now the labelling procedures, which do not exceed
sections are ready for immunogold one day. Copper grids can also be used for
labelling (see Chapter 23) and/or long-term storage of sections.32
embedding into methyl cellulose for  For overnight incubation and silver-
inspection by electron microscopy. enhancement (see Chapter 23), inert
material, such as nickel or gold, is
preferred.
 For plastic coating and carbon covering
of grids, see Reference 21.
Figure 19.8 Transfer of thawed cryo-

sections to a Formvar/Pioloform and
carbon-coated grid and incubation of the
grid (upside down) on a buffer drop for
subsequent immunolabelling.
Figure 19.9 Immunolabelling in a moist

chamber on Parafilm. Transfer of grids can
be done with forceps or a loop.

4.5. Staining and Methyl Cellulose Embedding

 After thawing or immunolabelling,  Here the full embedding procedure is
respectively, the sections are carefully
given. Depending on the application and the
washed on several drops of H2O. intensity of the contrast in a given
 Then they are stained for 10 minutes application, some of the steps can be
on uranyl oxalate at pH 7.4. omitted: staining with aqueous uranyl
acetate only or directly embedding in
 The sections are quickly washed over uranyl acetate containing methyl cellulose.
two drops of H2O. Other combinations are also valid.
 Stained for two minutes on aqueous  The main purpose of the neutral uranyl
uranyl acetate. oxalate is to protect the morphology of
delicate membrane structures.8
 Incubated for five minutes on a drop of  The methyl cellulose embedding is done
methyl cellulose or methyl cellulose on ice because the solution is less viscous
with uranyl acetate. in the cold.
 The grids are removed individually
from the drops with a loop.
 The excess methyl cellulose is drained
away with a wet piece of tissue paper
or filter paper, such that a very thin
film remains.
 Then the grids are dried for about
30 minutes before they are carefully
removed from the loop with forceps.
Figure 19.10. Uptake of grids in loop (left)
and removal of excess methyl cellulose
from the grid, placed in a loop, using a
piece of filter paper (right) (the distance
between grid rim and loop is larger then in
reality).
 After silver-enhancement the grids

should be stored in dry air, under nitrogen
gas or under vacuum because the silver
layer is not stable in a humid atmosphere
(see Chapter 23).
4.6. Cleaning Diamond Cryo-Knives

 Remove knife from the cryo-chamber  Caution: A diamond knife is very
before heating up. delicate! In order not to damage your
 Flush it with running tap water until diamond knife, please refer to the
the water does not freeze anymore on manufacture’s instructions prior to
the knife. cleaning. This is the method used routinely
in our lab by experienced personnel;
however, we do not take any liability!

 The tip of the polystyrene rod is cut to  The rod is available from Diatome and
an angle of about 90° using a razor supplied with new diamond knives.
blade.  Caution: The razor blade must be
 The sectioned polystyrene tip is dipped cleaned with white spirits before use. The
in 50% ethanol/water. remaining oil layer destroys the coating of
 Gently run the rod tip across the the diamond knife (H. Gnägi, Diatome,
cutting edge of the diamond knife personal communication).
without applying lateral pressure.  Caution: Do not touch the glue (if
present) between knife and knife holder.
Fig. 19.11 Moving the polystyrene rod tip

along the knife edge.
 When sections or debris have dried  Caution: Refer to the user manual of
onto the knife edge, place the knife in manufacturer.
H2O overnight before cleaning.
5.1. Advantages
 Highly efficient immunolabelling  Especially suited for low-copy number
antigens.
 The sample is only partly dehydrated  20 to 30% water remains in the sample.
and the proteins remain in their natural  Suited for antigens sensitive to solvents
aqueous environment. (ethanol) and resin components.
 The topology of the thawed sections  Partial penetration of antibodies and
enables better access to the antibodies. ultra-small gold markers into the section is
possible (see Chapter 23).
 Excellent definition of intracellular  See Reference 12 and 20.
membrane systems.

5.2. Disadvantages
 The biological material is chemically  Aldehyde fixation is selective.
fixed by aldehydes.
 An undefined number of soluble  Usually only a weak aldehyde fixation is
proteins may be lost. used.
 Suboptimum preservation of the  Alterations induced by chemical fixation,
cellular ultrastructure. mainly in membrane architecture33-35 (see
Part I, Cryo-Fixation Methods; and Chap-
ters 11, 13).
 The contrast in the cytoplasm is  Stronger fixation leads to less loss of
dependent on the fixation. proteins. Therefore, there is very little
difference in contrast, especially in the
smaller cytoplasmic structures, e.g.,
ribosomes and cytoskeletal components
cannot be differentiated. As an additional
note, stronger fixation also prevents
antibodies from penetrating the cryo-
sections, which may result in reduced
labelling efficiency.36
6. WHY AND WHEN TO USE CRYO-SECTIONING

ACCORDING TO TOKUYASU
 Cryo-sectioning according to  The proteins remain in their aqueous

Tokuyasu is the most sensitive environment.
preparation method for on-section  Suited for antigens sensitive to ethanol
immunolabelling. or methacrylates.
 Especially suited for low-copy number
antigens.
 The thawed sections have a rough
surface topology exposing more antigens
than resin sections.
 Antibodies can access antigens in the
sections of weakly fixed material.37
 The method is very fast.  If necessary, within one day a sample
can be prepared, sectioned, immunolabelled
and analysed.
 It is best suited for locating proteins in  It excels in membrane contrast.
correlation to cellular compartments.

7. OBSERVED RESULTS
 Figure on the Chapter’s title page  A light and electron microscopy image
of rat liver labelled for catalase (correlative
microscopy).
Rat liver was perfusion-fixed with 2%
formaldehyde and 0.2% glutaraldehyde in
phosphate buffer. Cryo-sections of about
300 and 80 nm thickness were cut in a cryo-
ultramicrotome and labelled with a rabbit
anticatalase antibody. For fluorescence light
microscopy (upper left corner), the semi-
thin sections were detected with a
secondary FITC-conjugated antibody
(bright dots) and stained with DAPI to
visualise the DNA (grey contours). For
electron microscopy labelling (lower right
corner), the primary antibody was detected
with a goat antirabbit antibody coupled to
ultra-small gold particles (Aurion). The
particles were silver-enhanced with
RGENT SE-EM (see Chapter 23) ca. 40
min at 23°C, see below.
 Figure 19.12  Electron micrograph of a cryo-section of

rat liver labelled for catalase, see above.
Bar = 500 nm
 Figure 19.13  Yeast cells were fixed in 2% FA, 15%

saturated picric in 100 mM citrate buffer pH
4.0 at 30°C for about 15 min. Fixation was
continued in 2% FA and 0.02% GA in 200
mM HEPES, pH 7.4 for 1 h, then the cells
were incubated in 1% NaIO4 in water for 1
h at 4°C. Finally, 100 nm-thick cryo-
sections were labelled for 1-6 β-glucan and
detected with 10 nm gold particles.31,38
This result confirms that the oxidation by
iodate did not dramatically change the
chemical structure of the glucan chains.
Bar = 200 nm


 Figure 19.14  Arabidopsis root tip cell in mitosis.

Seedling root tips were fixed with 4%
(30 min) and 8% FA (90 min) in MTSB,
pH 7.0. Root tips were infiltrated with PVP-
sucrose, mounted on a pin, frozen in LN2,
and sections of about 100 nm were cut. A
monoclonal mouse antimyc antibody was
used to label the myc-tagged version of the
cytokinesis specific syntaxin KNOLLE. The
primary antibody was detected with a
secondary antibody conjugated to
™
Nanogold . Nanogold was silver-enhanced
with HQ Silver™ for 8.5 min (see Chapter
23). KNOLLE is necessary for fusion of
Golgi-derived vesicles with the newly
formed cell plate (CP) membranes between
the two daughter cells. KNOLLE can be
also detected on the trans-Golgi network
(TGN).
 Figure 19.15  Arabidopsis seedling root tips were
processed as described above. The GFP-
tagged Sialyltransferase, which localizes to
trans-Golgi cisternae and TGN was labelled
with a rabbit anti-GFP antibody and
detected with a secondary antibody coupled
to ultra-small gold. The ultra-small gold
colloids were silver-enhanced with R-
GENT SE-EM for 45 min (see Chapter 23).
 Figure 19.16  Arabidopsis seedling root tips were
processed as described above. Clathrin
coats on TGN vesicles were detected with
rabbit anticlathrin antibodies and secondary
antibodies conjugated to Nanogold.
Nanogold was silver-enhanced with HQ
Silver for 8.5 min (see Chapter 23). The gap
between the trans-Golgi stack and TGN
vesicles is a stretching artefact caused by
the use of sucrose alone for section
retrieval. MC-sucrose mixture is less
harmful to the thawing cryo-section.
 G = Golgi
 M = mitochondrion
 TGN = trans-Golgi network
Bars = 250 nm


8. REFERENCES

Ann. N. Y. Acad. Sci., 85, 689, 1960.
2. Fernández-Morán, H. Application of the ultrathin freezing-sectioning technique to
the study of cell structures with the electron microscope, Arkiv Fysik, 4, 471, 1952.
3. Bernhard, W. Ultramicrotomie à basse température, Ann. Biol., 4, 5, 1965.
4. Christensen, A.K. Frozen thin sections of fresh tissue for electron microscopy with
a description of pancreas and liver, J. Cell Biol., 51, 1971.
5. Dollhopf, F. and Sitte, H. Die Shandon-Reichert-Kühleinrichtung FC-150 zum
Herstellen von Ultradünn- und Feinschnitten bei extrem niedrigen Temperaturen. I.
Gerätetechnik, Mikroskopie, 25, 15, 1969.
6. Sevéus, L. Preparation of biological material for x-ray microanalysis of diffusible
elements. I. Rapid freezing of biological tissue in nitrogen slush and preparation of
ultrathin frozen sections in the absence of trough liquid., J. Microsc., 112, 269,
1978.
7. Tokuyasu, K.T. A technique for ultracryotomy of cell suspensions and tissues,
J. Cell Biol., 57, 551, 1973.
8. Tokuyasu, K.T. A study of positive staining of ultrathin frozen sections,
J. Ultrastruct. Res., 63, 287, 1978.
9. Tokuyasu, K.T. Use of poly(vinylpyrrolidone) and poly(vinyl alcohol) for cryo
ultramicrotomy, Histochem. J., 21, 163, 1989.
10. Tokuyasu, K.T. Immunocytochemistry on ultrathin frozen sections, Histochem. J.,
12, 381, 1980.
11. Geuze, H.J. et al. Use of colloidal gold particles in double-labelling
immunoelectron microscopy of ultrathin frozen tissue sections, J. Cell Biol., 89,
653, 1981.
12. Liou, W., Geuze, H.J., and Slot, J.W. Improving structural integrity of
cryosections for immunogold labelling, Histochem. Cell Biol., 106, 41, 1996.
13. Slot, J.W. and Geuze, H.J. A new method of preparing gold probes for multiple-
labelling cytochemistry, Eur. J. Cell Biol., 38, 87, 1985.
14. Sitte, H. Process of ultrathin sectioning, in Science of Biological Specimen
Preparation 1983, Revel, J.P., Barnard, T., and Haggis, G.H., eds., SEM Inc., AMF
O'Hare, IL, USA, 1984, 97.
15. Oorschot, V. et al. A novel flat-embedding method to prepare ultrathin
cryosections from cultured cells in their in situ orientation, J. Histochem.
Cytochem., 50, 1067, 2002.
16. González-Melendi, P. et al. New in situ approaches to study the induction of pollen
embryogenesis in Capsicum annuum L, Eur. J. Cell Biol., 69, 373, 1996.
17. Dettmer, J. et al. Vacuolar H+-ATPase activity is required for endocytic and
secretory trafficking in Arabidopsis, Plant Cell, 18, 715, 2006.
18. Pimpl, P. et al. In situ localization and in vitro induction of plant COPI-coated
vesicles, Plant Cell, 12, 2219, 2000.
19. Van Donselaar, E. et al. Immunogold labelling of cryo-sections from high-pressure
20. Griffiths, G. Fine Structure Immunocytochemistry. Springer-Verlag, Berlin,
Heidelberg, Germany, 1993.

21. Hayat, M.A. Principles and Techniques of Electron Microscopy. Biological

Applications. 4th ed., Cambridge University Press, Cambridge, UK,2000.
22. Faulk, W.P. and Taylor, G.M. An immunocolloid method for the electron
microscope, Immunochemistry, 8, 1081, 1971.
23. Slot, J.W. and Geuze, H.J. Sizing of protein A-colloidal gold probes for
immunoelectron microscopy, J. Cell Biol., 90, 533, 1981.
25. Hinz, G. et al. Vacuolar storage proteins and the putative vacuolar sorting receptor
BP-80 exit the Golgi apparatus of developing pea cotyledons in different transport
vesicles, Plant Cell, 11, 1509, 1999.
26. Goodbody, K.C. and Lloyd, C.W. Immunofluorescence techniques for analysis of
the cytoskeleton, in Plant Cell Biology. A Practical Approach, Harris, N. and
Oparka, K.J., eds., IRL Press, Oxford, UK, 1994, 221.
27. Murashige, T. and Skoog, F. A revised medium for rapid growth and bio assays
with tobacco tissue cultures, Physiologia Plantarum, 15, 473, 1962.
28. Humbel, B.M. and Biegelmann, E. A preparation protocol for postembedding
immunoelectron microscopy of Dictyostelium discoideum cells with monoclonal
antibodies, Scan. Microsc., 6, 817, 1992.
29. Griffiths, G. Selective contrast for electron microscopy using thawed frozen
sections and immunocytochemistry, in Science of Biological Specimen Preparation,
1983, Revel, J.P., Barnard, T., and Haggis, G.H., eds., SEM Inc., AMF O'Hare, IL,
USA, 1984, 153
30. Kärgel, E. et al. Candida maltosa NADPH-cytochrome P450 reductase: Cloning of
a full-length cDNA, heterologous expression in Saccharomyces cerevisiae and
function of the N-terminal region for membrane anchoring and proliferation of the
endoplasmic reticulum, Yeast, 12, 333, 1996.
31. Müller, W.H. et al. Immuno-electron microscopy in yeast cell research, Recent
Res. Devel. Mol. Microbiol., 1, 119, 2002.
32. Griffith, J.M. and Posthuma, G. A reliable and convenient method to store
ultrathin thawed cryosections prior to immunolabelling, J. Histochem. Cytochem.,
50, 57, 2002.
33. Van Harreveld, A., Crowell, J., and Malhotra, S.K. A study of extracellular
space in central nervous tissue by freeze-substitution, J. Cell Biol., 25, 117, 1965.
34. Murk, J.L.A.N. et al. Influence of aldehyde fixation on the morphology of
endosomes and lysosomes: Quatitative analysis and electron tomography, J.
Microsc., 212, 81, 2003.
36. Stierhof, Y.-D., Schwarz, H., and Frank, H. Transverse sectioning of plastic-
embedded immunolabeled cryosections: Morphology and permeability to protein A-
colloidal gold complexes, J. Ultrastruct. Mol. Struct. Res., 97, 187, 1986.
37. Stierhof, Y.-D. et al. Yield of immunolabel compared to resin sections and thawed
cryosections, in Colloidal Gold: Principles, Methods, and Applications, Hayat,
M.A., ed., Academic Press, Inc., San Diego, CA, USA, 1991, 87.
38. Humbel, B.M. et al. In situ localization of β-glucans in the cell wall of
Schizosaccharomyces pombe, Yeast, 18, 433, 2001.

Cryo-Preparation Procedures for Elemental Imaging by SIMS and EFTEM 501
CONTENTS

1. PRINCIPLES OF THE METHODS ................................................................ 504
1.1. Energy Filtered Transmission Electron Microscopy (EFTEM)................ 504
1.2. Secondary Ion Mass Spectroscopy (SIMS) .............................................. 505
2.1. Chemical Procedure.................................................................................. 506
2.2. Cryo-Procedures ....................................................................................... 507
3.1. Materials ................................................................................................... 509
3.2. Products .................................................................................................... 511
3.3. Solutions ................................................................................................... 512
4. PROTOCOLS .................................................................................................... 512
4.1. Sampling................................................................................................... 512
4.2. Chemical Procedure.................................................................................. 513
4.3. Cryo-Fixation ........................................................................................... 514
4.3.1. Slamming .................................................................................... 514
4.3.2. High-pressure freezing................................................................ 515
4.4. Dehydration of Cryo-Fixed Samples ........................................................ 516
4.4.1. Freeze-substitution...................................................................... 516
4.4.2. Freeze-drying .............................................................................. 516
4.5. Resin Embedding...................................................................................... 517
4.5.1. Low viscosity epoxy resin mixture ............................................. 517
4.5.2. Lowicryl HM® ............................................................................ 518
4.6. Sectioning ................................................................................................. 519
4.6.1. Sectioning on fluid...................................................................... 519
4.6.2. Dry sectioning............................................................................. 520
4.7. Elemental Analysis ................................................................................... 520
4.7.1. EFTEM ....................................................................................... 520
4.7.2. SIMS ........................................................................................... 521
5.1. Advantages ............................................................................................... 522
5.1.1. Chemical procedure .................................................................... 522
5.1.2. Cryo-fixation............................................................................... 522
5.1.3. Dehydration ................................................................................ 523

5.1.4. Resin embedding ........................................................................ 523

5.1.5. Sectioning................................................................................... 524
5.1.6. Elemental analysis ...................................................................... 524
5.2. Disadvantages .......................................................................................... 525
5.2.1. Chemical procedure.................................................................... 525
5.2.2. Cryo-fixation .............................................................................. 525
5.2.3. Dehydration ................................................................................ 526
5.2.4. Resin embedding ........................................................................ 526
5.2.5. Sectioning................................................................................... 527
5.2.6. Analysis ...................................................................................... 527
6.1. Chemical Procedure ................................................................................. 528
6.2. Cryo-Fixation ........................................................................................... 528
6.2.1. Cryo-fixation by slamming......................................................... 528
6.2.2. Cryo-fixation by high-pressure freezing..................................... 528
6.3. Dehydration.............................................................................................. 529
6.3.1. Dehydration by freeze-substitution............................................. 529
6.3.2. Dehydration by freeze-drying..................................................... 529
6.4. Resin Embedding ..................................................................................... 529
6.4.1. Embedding in Spurr’s resin ........................................................ 529
6.4.2. Embedding in Lowicryl HM23................................................... 530
6.5. Sectioning................................................................................................. 530
6.5.1. Sectioning on fluid ..................................................................... 530
6.5.2. Dry-sectioning ............................................................................ 530
6.6. Analysis.................................................................................................... 531
6.6.1. EFTEM....................................................................................... 531
6.6.2. SIMS........................................................................................... 531
8. REFERENCES .................................................................................................. 536



Identification, localisation and quanti-

fication of intracellular chemical elements

are important questions in many areas of

biological research, especially in

pharmaco-toxicology to understand mech-

anisms by which drugs interfere with living

processes.

Nowadays some analytical techniques, such EFTEM = Energy Filtered Trans-
as EFTEM3,4 or SIMS imaging8,9 allow mission Electron Microscopy
mapping of most of the chemical elements SIMS = Secondary Ion Mass Spec-
present in samples. trometry
During such analysis, samples are subjected

to drastic conditions (vacuum, bomb-
Hydrated samples are not com-
ardment energy, etc.) that prevent obser-
patible with the high vacuum conditions
vation of living cells. Therefore, all
of the analysis chamber.
samples should be prepared prior to any
analysis. First, biological processes should
be immobilized so that the samples stay as
close as possible to the living state. The
samples are then dehydrated and generally
embedded in a plastic resin.
When preserving samples, the preparation
methods should first and foremost
minimize artefacts, particularly delocal-
isation of the molecule of interest at the
level of resolution of the imaging
technique, which is in a nanometric range
for analytical methods using electron
microscopy.

1. PRINCIPLES OF THE METHODS
1.1. Energy Filtered Transmission Electron Microscopy (EFTEM)

 For more details see Reference 7.
 In a transmission electron microscope
(TEM) when incident electrons from the
electron beam go into a sample, they
interact with the constituent atoms. These
interactions depend on the atomic structure
of the sample. Some of these electrons are
scattered. When inelastic collisions with the
electron shell of the sample’s atoms occur,
the primary electrons undergo a deviation of
their trajectories and a loss of energy, which
is specific for every element present in the
sample. This phenomenon is the basic
principle of electron energy-loss
spectroscopy (EELS).
E = Energy of incident electrons

E = Energy lost during inelastic
scattering
Figure 20.1 Electron scattering by a single

Chemical information can also be atom.
determined by other analytical techniques,
which are secondary events of inelastic  Such as auger or x-ray emissions.
collisions of electron beam and sample
atoms.
 Characterization of energy losses of
scattered electrons is obtained by using
different kinds of dispersive energy filters
placed either in or below the column of the
microscope.
S = Sample
P = Alpha filter in the column
M = Electrostatic mirror
E = Energy of incident electrons
E = Energy lost during inelastic
scattering
Figure 20.2 Example of an in column filter

(Alpha filter) in an energy filtered
transmission electron microscope
(EFTEM).

1.2. Secondary Ion Mass Spectroscopy (SIMS)

 When a solid sample is bombarded by
primary ions of a few keV in energy,
molecular fragmentation occurs and
particles are emitted from the surface. The
portion of sputtered material that leaves the
sample surface as ions (secondary ions) is
accelerated into a mass spectrometer where
it is separated by its mass-to-charge ratio,
and detected. Secondary ion emission
supplies information about either the
elemental and isotopic composition of the
surface uppermost atomic layers by
dynamic SIMS5 or the molecular
composition by static SIMS (TOF-SIMS).2
SIMS is the most sensitive surface analysis
technique and can be applied to any type of
material that can be maintained under
vacuum.
PI = Primary ion beam
S = Sample
SI = Sputtered secondary ions
Figure 20.3. Primary ion bombardment

produces a collision cascade and sputter
elements from the sample surface.
 Elemental or isotopic distribution

imaging can be obtained by rastering the
primary ion probe upon the surface.17
D = Detectors
ES = Electrostatic sector
M = Magnet
MS = Mass spectrometer
PI = Primary ions
IS = Primary ions source
S = Sample
SI = Secondary ions
1, 2, 3, 4, 5 = Images of the chemical
element distribution observed from the data
measured on detectors d1, d2, d3, d4 and
d5.
Figure 20.4 SIMS-instrument (NanoSIMS)

scheme.

Because living samples cannot be observed directly by EFTEM or SIMS, preparation

steps are essential and specific strategies must be adapted to each problem and each kind
of sample.
2.1. Chemical Procedure

 Numerous procedures are described in
all classical laboratory manuals. For more
details, see Reference 10.
1. Chemical fixation  Glutaraldehyde or a mixture of
glutaraldehyde / formaldehyde.
2. Dehydration by solvent  Water is usually removed by passing the

specimen through solutions of ascending
concentrations of organic solvents (ethanol,
acetone): 50 to 100%.
3. Resin infiltration  In epoxy resin by passing the specimen

through solutions of ascending concen-
trations of resin mixture without
accelerator (50%, 75% and pure resin
twice) followed by resin mixture with
accelerator.
4. Polymerization by heat at 65°C  At 65°C in resin mixture with

accelerator 12 to 24 hours depending on the
resin.
5. Sectioning at room temperature and A = Embedded specimen

collecting sections by floating B = Diamond knife
C = Sections
D = Water or glycerol boat
E = Section on a glass slide for light
microscopy
F = Section on a grid for EFTEM
analysis
G = Section on stainless steel or silicon
holder for SIMS analysis
Figure 20.5 Sectioning by floating on fluid.
6. EFTEM and/or SIMS analysis  Sections 40 to 50 nm thick for EFTEM,

adjacent 200 nm thick sections for SIMS.

2.2. Cryo-Procedures
1. Cryo-fixation
Cryo-fixation of specimens can be carried  Quick freezing is the best way for rapid
out in a number of ways, but we have immobilisation of any biological material.
chosen to use either slamming or high- Depending on the experimental conditions,
pressure freezing: the freezing process can lead to formation
of different ice states going from vitreous
 Slamming (see Chapter 2). to hexagonal ice. Hexagonal crystals alter
drastically cellular ultrastructure by solute
segregation and could redistribute
erratically some intracellular elements.
A = Injector rod with pneumatic

damping system
B = Specimen holder: metal plate, foam
rubber and double-sided adhesive tape
C = Moist sample fixed downward onto
the adhesive tape
D = LN2-cooled, gold-coated copper
mirror
Figure 20.6 Scheme of specimen slamming

onto cooled metal mirror.
 High-pressure freezing
(see Chapter 6).
A = Flat specimen holder (perforated

steel envelope)
B = Flat specimen carrier (gold-coated
copper)
C = Direct liquid nitrogen jets
D = Thin tube transmitting pressure P
(about 2000 bar) to specimen holder
Figure 20.7 Schematic drawing of the

high-pressure principle according to Studer.

2. Dehydration of cryo-fixed samples
 Freeze-substitution12,13  Dry acetone with or without osmic acid

(see Chapter 13). at 90°C for about three days; ice crystal
formation can occur at 90°C without
apparent damage to the sample at the
nanometre scale.
 Freeze-drying6  Freeze-drying process starts around

(see Chapter 15). 100°C.
3. Embedding in a resin  Fluid epoxy resin or Lowicryl® HM23.
4. Polymerization at 65°C or by UV  The aim of embedding is to fill the

light at 70°C "water space" in the biological specimen by
fluid resin that, after polymerization, will
harden the sample. The resin acts also as a
homogenous matrix for the analytical
beam.
5. Sectioning as in Section 4.6.1. or dry

sectioning
A = Embedded specimen
B = Diamond knife
C = Sections
E = Ionizer (to neutralize electrostatic
charges on the dry section by an ion
emission)
F = Section on a glass slide for light
microscopy
G = Section on a grid for EFTEM
analysis
H = Section on stainless steel or silicon
holder for SIMS analysis
Figure 20.8 Dry sectioning.
6. EFTEM and/or SIMS analysis

3.1. Materials
 Refridegerator  Storage of fixatives and embedding
products used in chemical procedure.
 Magnetic stirrer  To mix chemicals and resin mixtures.
 Oven 65°C  To dry flasks, moulds and for heat
polymerization of epoxy resins.
 Horizontal self pressurized LN2  50 litres: intermediate LN2 storage inside
container with a roller base laboratory.
 A little LN2 container with handle  4 to 5 litres to gently fill the freeze
devices.
 Slam-freezing device  Commercially available. For example:
 EM MM80E, or EM CPC, Leica
 RMC MF-7000, Boeckeler Instru-
ments Inc. Tucson, Arizona,USA.
 High-pressure freezing device  Commercially available. For example:
 HPM 010, BAL-TEC product. Now
available from Boeckeler.
  A new modified apparatus HPM
100 is available from Leica
 EM PACT or more recently
 EM PACT 2 + RTS device, Leica
 HPF Compact 01, Engineering Office
M. Wohlwend GmbH, Sennwald,
Switzerland.
 Freeze-substitution device  Commercially available. For example:
 EM AFS or more recently
 EM AFS2, Leica
 RMC FS-7500
 Freeze-drying device  Commercially available. For example:
 EM CFD, Leica (upon request)
 MED/020 GBF, Leica.
 Ultramicrotome  Commercially available. For example:
 EM UC6, Leica
 Power Tome X and XL, RMC.
 Diamond knifes  Diamond knives are available from
various manufacturers (e.g., Diatome AG,
Bienne, Switzerland or Element Six, Cuijk,
The Netherlands).
 Diamond knife with a boat  Sectioning by floating on fluid.
 knife angle: 35°  Biological tissues, heterogeneous (in
term of hardness) blocks (i.g., non
decalcified spongy bone).

 knife angle: 45°  Recommended for uniformly hard

samples.
 Cryoimmuno diamond knife  Dry sectioning thick sections up to

 knife angle: 35° 200 nm.
 Ionizer  To neutralize electrostatic charges on the

dry section to facilitate manipulations.
 Commercially available. For example,
Static Line II by Diatome Ltd.
 Grid storage box with 50 to 100
alphanumeric labelled compartments
 Uncoated hexagonal nickel or gold  600 mesh to support sections in EFTEM

specimen grids analysis.
 250 mesh to support sections in SIMS
analysis.
 Carbon-coated, square specimen grids
with alphabetic identification
 Small items:
 Sapphire disks  1.4 mm diameter, 50 µm thick.
Commercially available: Leica Product.
 Thermanox® film  Nalge NUNC Products, Rochester, NY,
USA.
 Scalpels, scissors, fine forceps, biopsy  To dissect and prepare samples.
needles
 Glass vials, flasks and beakers  25 mL and 60 mL for storage and
mixing the different resins kept in a dry
oven (60°C).
 Magnet bars  Used to mix resin solutions.
 Disposable plastic Pasteur pipettes  3 and 1.5 mL.
 Watch-glass  Used to transfer samples under resin if
necessary.
 Embedding flat moulds
 Embedding gelatine capsules  In any microscopy product catalogue,
e.g., Pelco International, Clovis, California,
USA.
 Supports with holes  Used to maintain gelatine capsules
vertically during resin polymerization.
 Small section pickup loop  Used to pickup sections obtained by
flotation.
 Gloves: latex and nitrile  Important: Some components for
embedding resins can be very toxic.

3.2. Products
 Cacodylic acid sodium salt trihydrate  Toxic by inhalation and if swallowed.

 50% Glutaraldehyde  Toxic by inhalation. Use under a fume
hood.
 Formaldehyde in powder  Toxic by inhalation. Use under a fume
hood.
 2% OsO4 in water  Extremely toxic for eyes, skin, lung,
kidneys at very low concentrations. Use
under a fume hood.
 Acetone GR dry  Maximum 0.01% water.
 Available from various manufacturers
(e.g., Fluka Chemie GmbH, Switzerland).
 Calcium chloride in powder
 Spurr’s epoxy resin mixture as  Viscosity of the complete mixture: about
reported by Spurr:18 60 cP at 25°C, polymerization at 65°C.
 ERL® 4206 or VCD (vinyl cyclo-  Harmful if swallowed, eye and skin
hexene dioxyde): diepoxide irritant, may cause allergic skin reaction,
causes cancer in laboratory animals.
 Production of ERL®4206 is discontinued
and is actually replaced by ERL®4221.
®
 DER 736 (Diglycidyl ether of  Eye and skin irritant.
polypropyleneglycol): softener
 NSA (Nonenyl succinic anhydride):  Harmful if swallowed, eye and skin
hardener irritant.
 DMAE (Dimethylaminoethanol):  Eye, skin and mucous membrane
accelerator irritant, may injure the central nervous
system.
 Use and prepare Spurr’s mixture
under a fume hood.
 Lowicryl® HM23 hydrophobic, non-  See Reference 1.
polar resin  Low viscosity at –80°C.
 Monomer G: a mixture of Ethyl  Methacrylates may cause eczema and
methacrylate and n-Butyl metha- allergic reactions: wear nitrile and latex
crylate gloves.
 Cross-linker:
1,3 Butanediol dimethacrylate
 Photo-initiator J  BDA: Benzyl dimethyl acetal.
 For polymerization at –80°C or –70°C
by UV radiation (360 nm).
 LN2: Cryogen  May cause severe burns; wear
protective clothes, glasses and gloves for
handling LN2. Always use LN2 in well-
ventilated areas.

3.3. Solutions
 Phosphate buffer 0.15 M according to
Sörensen (see Reference 10)  Di Na/mono K, pH 7.4
 0.2 M Cacodylate buffer  pH 7.4
 1% Glutaraldehyde in 0.15 M
phosphate buffer  For chemical fixation of cultured cells.
 1.6% Formaldehyde, 3%
glutaraldehyde, 0.05% CaCl2 in 0.1 M
cacodylate buffer  For chemical fixation of samples.
 1% OsO4 in buffer or small crystals  When post fixation is required.
 Increasing concentrations of ethanol
from 50% to 100%  Chemical procedure.
 Spurr’s mixture, manufacturer’s
standard medium:
 VCD: 10.0 g
 DER 736: 6.0 g
 Polymerization at 65°C for 8 hours
 NSA: 26.0 g
minimum.
 DMAE: 0.4 g
 Lowicryl mixture (HM23) for  UV polymerization at –70°C: at least
infiltration and UV polymerization two days, if necessary complete by UV
below 70°C: polymerization for one day to three days at
 Cross-linker: 1.10 g 50°C.
 Monomer G: 18.90 g
 Initiator J: 0.15 g
 Lowicryl now also exists as pre-
mixed ready to use solutions.
4. PROTOCOLS
4.1. Sampling
Because in vivo and in vitro observations by
EFTEM and SIMS are so far impossible, the
first step in any analysis consists in collecting
samples. This step is crucial and should be
performed while minimizing cellular stress  Cellular stress induces drastic changes
that could modify molecular distributions. in free ion distribution.
Biological materials come from:

 Cell or microorganism cultures that are  Cells cultured on organic support
homogeneous and simple models and that films (Thermanox™ type) or on mineral
come in adherent or suspended form. disks minimize stressing steps.
 Biopsies (plant or animal).  Reflect more tissue heterogeneity.

 Useful for pharmacological or trace
element studies.


1. Fixation
1.1. Cultured cells on support

 1% Glutaraldehyde in 0.15 M phosphate  First replace culture medium by warm
buffer, pH 7.4 fixative mixture (35°C) for 5 min, then
change for cold (4°C) fixative mixture
for 20 min.
 Wash buffer 3 × 10 min
at 4°C
 Postfixation: 30 min
1% OsO4 in buffer at RT
at RT
1.2. Biopsy samples

 1.6% Formaldehyde, 4-5 h
3% glutaraldehyde in at 4°C
0.1 M Cacodylate
buffer, 0.05 % CaCl2,
pH 7.4
0.1 M Cacodylate at 4°C
buffer, 0.2 M Sucrose,
0.03 M CaCl2, pH 7.4
 Postfixation: 45 min
1% OsO4 in 0.1 M at RT
Cacodylate buffer
at RT
2. Dehydration
 Ascending concentrations of organic  Usually ethanol or acetone.
solvents:
 30%, 50%, 70%, 95% 510 min/each
at 4°C
 100% twice 10 min  For cultured cells dehydration time
RT can be reduced.
3. Resin infiltration
 Ethanol + Spurr’s 60 min
resin: 1/1 at RT
 Ethanol + Spurr’s 60 min
resin : 1/3 at RT
 Spurr’s 112 h (2 baths)  For cultured cells infiltration time can
100% at RT be reduced.
4. Polymerization 10 h
at 65°C

5. Sectioning  For more details see Section 4.6.

 EFTEM: Usually section thickness varies  The thinner the sections, the easier the
between 40 to 50 nm. analysis.
 SIMS: Typical sections are 100 to 500  Below 100 nm, the sections are too
nm-thick. thin to resist ion bombardment for a long
period of analysis, while over 500 nm the
electrostatic charge of the sample could
perturb the analysis.
4.3. Cryo-Fixation
4.3.1. Slamming
Cryo-fixation using EM MM80 E, Leica.  For more details, see Chapter 2.

 For cultured cells on Thermanox™
type film or sapphire disks or thin
biopsies.
 Parameter settings:
 Force: 11
 Speed: 11
 Thickness: 4
 Depending on specific studies, cell
stimulation can be carried out at this
step.
 Attach the fresh sample onto the  See Figure 20.9 One to three samples
specimen holder and impact it onto the can be attached onto the holder and cryo-
precooled metal block. fixed at the same time. Just before fixation
excess culture medium should be gently
removed with a piece of filter paper.
 After freezing, the specimens are
stored in LN2 or directly transferred in
LN2 to the freeze drying device.
 Scheme of process. A = Cultured cells:
 On Thermanox® type film (3 × 3 mm)
 On sapphire discs (diameter 1.4 mm,
50 µm thick) or small biopsies (1mm3)
coated with 1 µl culture medium, 20%
foetal calf serum: (omit if stimulation
will be done later).
B = Stimulation if necessary (if studies of
dynamic processes)
C = Gently remove excess medium with
filter paper
D = Sample fixed downward onto the
adhesive tape
E = Slamming onto the LN2 cooled metal
mirror
Figure 20.9 Scheme of the slamming
process.

4.3.2. High-pressure freezing

High-pressure freezing using the EM PACT  For more details, see Chapter 6.
apparatus from Leica with specimen  Note: In this part, terms describing small
carriers for: tools come from the EM PACT’s Leica
 catalogue.

1. Cultured cells on sapphire disks  1.4 mm diameter, 50 µm thick for
sapphire.
2. Thin and regular samples  For example, one-day-old rat calvarium
or plant leaf cut with a biopsy needle
(diameter 1 mm).
A = View from the top of an empty

carrier
B = Sectional drawing of a carrier with a
sample (cells cultured on a sapphire disk)
in culture medium.
Figure 20.10 Scheme of a flat specimen

carrier for sapphire disks.
 Adherent cells are cultured on the  To avoid handling the disk before
sapphire disk. If necessary, the cavity fixation, cell cultures are set up directly
of the specimen carrier is filled with in the specimen carrier.
appropriate culture medium.
 Samples are punched with the biopsy  If cell stimulation is needed proceed at
needle and pushed directly in the this step.
cavity of the specimen carrier pre-
filled with 0.6 µL of culture medium.
 Once the flat specimen carrier is filled  The outer ring of the specimen carrier
with sample and medium, it is should be dry; if necessary, gently remove
sandwiched in the flat specimen holder medium with filter paper.
using a torque wrench.  Applying a force of 20 N/cm2.
 Total time for sampling and filling the
carrier should be minimized and be below
30 seconds.
 The specimen holder is mounted on  Two persons are needed for this
the rod and transferred in the EM operation, one to prepare the specimen
PACT and freezing is automatically carrier (filling, specimen stimulation,
carried out. positioning, etc.) and the other to fix it on
the rod and freeze it.

 The HPF device operates up to  It takes about 15 ms to freeze down from

2000 bars with a cooling rate of about 30°C to 190°C with a pressure near
7000 to 10000°C/s. 2000 bars.
 After freezing, the specimen carriers  After use some of the flat specimen
are separated from the holder by carriers are deformed. Samples from
unscrewing under LN2 with the torque damaged carriers should be rejected to
wrench and left in LN2 or directly ensure the quality of freezing.
transferred in LN2 to the freeze-
substitution device.
4.4. Dehydration of Cryo-Fixed Samples

4.4.1. Freeze-substitution
 Frozen samples are transferred in  For handling of the AFS, see Chapter 18.
liquid nitrogen into a freeze-  AFS2 now in Leica catalogue.
substitution device (AFS, Leica).

 Ice is removed by substitution using  Commercial dry acetone (maximum
pure dry acetone (90°C, three days). 0.01% water). Avoid molecular sieves that
can carry impurities altering the analyses.
4.4.2. Freeze-drying
 Freeze-drying in the cryo-sorption  For details, see Chapter 15.
freeze-drying (CFD) apparatus accor- 
ding to Sitte.16 
 A = Chamber filled with cold LN2 gas
B = Object table heating (regulation of
temperature against LN2 temperature)
C = Samples
D = Cold trap wings
E = Translucent glass
F = Connection for pumping with valve
V1 = Valve for communication of mole-
cular sieve with chamber A
V2 = Valve connection with cold LN2 gas
(breaking vacuum)
MS = Molecular sieve (zeolite)
Figure 20.11 Schematic cross-section of

freeze-drying chamber.



 Before freeze-drying, the molecular  At last 12 hours at 90°C, valves F and

sieve should be regenerated by V1 open, V2 closed and glass E in position.
heating.
 Fill the FD device with 30 L LN2.  Allow 40 L for filling the machine.
 Precool the chamber to 150°C.
 Transfer cryo-fixed samples in the pre-  From freeze-slamming or high-pressure
cooled object table (190°C). freezing experiment.
 Place the cold trap wings over the
samples.
 Primary vacuum is obtained in the
chamber of the CFD by a membrane  18 mbars.
pump.
 Secondary vacuum is obtained by
cryosorption pumping (zeolite).  2 × 10-3 mbars.
 Freeze-drying process is achieved by
increasing temperature slowly up to  To better preserve molecular structure
0°C. and biological functions, freeze-drying can
 Dehydration is then stopped by be stopped earlier (60°C or 10°C).
breaking vacuum with LN2 gas entry.
4.5. Resin Embedding

4.5.1. Low viscosity epoxy resin mixture
We use the Spurr formulation18 (see
Section 3.2.) with a low viscosity, which
facilitates infiltration of densely packed  The viscosity of the resin mixture is
tissues. lower than 60 cP.
 Freeze-dried samples
 When the dehydration process is
stopped (at 10°C or 0°C) (see Section
4.4.2) chamber A (see Figure 20.11)  Freeze-drying process stopped at
can be opened. atmospheric pressure.
 Pure Spurr’s resin is dropped on
samples and infiltration is carried out
with a weak vacuum (about 500 mbar)
while the temperature is increased to  Infiltration time depends on sample.
20°C (1°C/hour).
 Resin infiltration is then completed at  Embedding medium is renewed.
room temperature in small dry flasks
outside the CFD apparatus.
 Place infiltrated samples in flat moulds  Previously dried in an oven at 65°C.
or in gelatine capsules.
 Process polymerization in an oven.  At 65°C for 10 hours.

 Freeze-substituted samples
 After freeze-substitution, resin infil-
tration is carried out in ascending  50, 75, 100% ERL.
concentrations of ERL in acetone for
about 24 hours from –40°C to 10°C.
 Infiltration is completed in Spurr’s  Infiltration time depends on sample.
mixture for one or two days at 20°C.
 Process polymerization in an oven.  At 65°C for 10 hours.
4.5.2. Lowicryl HM®

Lowicryl HM resins are nonpolar and
hydrophobic. They are methacrylate-based  Lowicryl HM20 (can be used at 45°C).
and remain fluid at low temperature.1  Lowicryl HM23 (can be used at 80°C).

Polymerization by UV light takes place at  Polymerization with Lowicryl HM20 can
low temperatures. be carried out at 45°C.
 Polymerization with Lowicryl HM23 can
We use a Lowicryl HM23 mixture: be carried out at 80°C.
 Freeze-dried samples
 When freeze-drying is achieved (see
Section 4.4.2), infiltration of resin is
carried out at 60°C by adding a few
drops of resin mixture onto the
samples in the LN2 cold gas-vented
CFD chamber.

 Polymerise at 60°C, 48 hours.  UV polymerization.
 If necessary, complete the polymer-
ization for 24 to 72 hours at 50°C.
 Freeze-substituted samples
 When freeze-substitution is achieved
(see Section 4.4.1), infiltration of resin
is performed:
 50% resin in acetone at 90°C,
three hours
 100% resin twice 30 hours at
80°C.
 Polymerize at 70°C, 48 hours.  Using UV.

 If necessary, complete the polymer-  If resin blocks are too soft or if a liquid
ization for 24 to 72 hours at 50°C. film remains at the surface.

4.6. Sectioning
Sections are cut with an ultramicrotome at 
low speed (1 mm/sec) or manually using a 
diamond knife (angle: 35° or 45°). 

4.6.1. Sectioning on fluid
Usually, sections are collected on a water 
boat. Glycerol may be used to minimize
element losses (see Section 2.1. and Figure
20.5 for general procedure).
 For EFTEM 
 EFTEM analysis requires ultrathin
slices (thinner than 100 nm) to avoid  Ultrathin sections: 40 to 50 nm thick.
multiple electron scattering through
the section that alters the transmitted
signal.
 Sections are collected on fluid and
recovered on hexagonal 400 to 600  Dry carefully the grids on a filter paper
mesh nickel or gold grids to remove fluid before storing.
 Sections floating on water are picked
up quickly. Only one or two sections  Only one or two sections can be cut and
are transferred on one grid without picked up in a short time (about 10
any coating film. seconds).
 If sections are collected on anhydrous
glycerol, work in a dry room.  To reduce glycerol hydration.
 Grids are stored in a grid box until  Hygrometry < 50%.
observation. 
 For SIMS
 Section thickness can vary from  Less than 100 nm sections are too thin to
100 nm to 500 nm and sections are resist ion bombardment for a long period of
made using a histological diamond analysis, while in the case of over 500 nm
knife (angle: 45°). sections electrostatic charges of the sample
could perturb the analysis.
 See Figure 20.5 (G).
 Sections are floated on water and put  Standard stainless steel or silicon holders
on a warm holder with a small section for SIMS analysis have a 10 mm diameter.
pickup loop.
 If TEM imaging is performed on the
same section for ultrastructural  Section thickness can be reduced to 80 to
identification, then sections can be 100 nm.
directly put on 250 mesh carbon-
coated grids.
 Sections are dried for one hour on a
hot plate and then stored in a vacuum  To keep the sections dry.
oven at 50°C or directly in a storage
chamber of the SIMS apparatus for
impending observations.

4.6.2. Dry sectioning

 Dry sectioning is recommended for  See Section 2.2. and Figure 20.8.
examination of highly diffusible
elements and it is performed manually
at room temperature using a 45°
immuno cryo-diamond knife.
 Dry sections for EFTEM are thicker
than sections obtained on fluid (up to  Ion emission neutralizes electrostatic
60 nm). charges on dry sections and avoids folding
 Static Line II (ionizer) should be until sections are put on the holder.
operated during the entire sectioning 
procedure.
 Sections are grabbed with fine
tweezers and put on prewarmed
holders and stored at 50°C.
4.7. Elemental Analysis

4.7.1. EFTEM
 After locking a grid with sample
sections in the column of the electron  Element specific signals (ionization
microscope, element specific signals edge) are superimposed on a background
are identified by energy loss of the that has to be removed from data to obtain
transmitted electrons that interact with the net distribution of each element of
the sample. interest.
 Data can be acquired on:

 Spectrum mode
 EELS:
(electron energy loss spectroscopy).  Acquisition is relatively time consuming
The energy loss of the imaged for just one spectrum (about two minutes)
electrons on one spot (10 nm and damages the sample locally.
diameter) is scanned in a large
window (100 eV) and the electron 
intensity is recorded and processed.
 PEELS:
(parallel electron energy loss spec-  Intensity profile of the image is recorded
troscopy). and processed to identify elements present
Beyond the spectrometer, the electron locally in the sample.
beam is dispersed according to the  The time required for acquisition of the
energy of the electrons. The spectrum spectrum is much faster (<1 sec).
is visible on the image plane as a line
with varying brightness and can be
imaged via the camera.

 Image mode
 ESI:  Image signal contains specific element
(electron spectroscopic imaging) information buried in unspecific back-
A short series of images is acquired ground. Therefore, the signal of each pixel
around a specific energy edge of the has to be processed to extract the useful
element of interest. information and to map the distribution of
the specific element.
 Acquisition of images and processing for
each element takes typically less than one
 Image EELS: minute.
(combination of spectrum and images)
A series of images (up to 100)
obtained at different energy-loss  A complete spectrum is associated to
values (typically with an energy each pixel. Therefore, after processing, the
increment of 1 eV) are processed to chemical distribution of various elements
either spectra or element distribution can be easily mapped.
images.
4.7.2. SIMS
 Sample sections are placed and  A good vacuum is essential for efficient
maintained under vacuum for several transmission of secondary ion beams.
hours in the SIMS apparatus.
 Implantation of the sample surface  In dynamic SIMS, implantation is

with primary ion beam. required so that secondary ion current
reaches a steady state.
 Setting the microprobe and detectors
for analysis.
 The probe is rastered over the area of  In dynamic SIMS each secondary ion
interest for analysis. Emitted species is detected within specifically tuned
secondary ion beam goes then through detectors.
the mass spectrometer where it is  Typical acquisition time varies from 5 to
analysed. 30 minutes.

 Data processing.  Mapping of selected chemical species
and correlation with light or TEM images
for ultrastructural characterisation.

5.1. Advantages
5.1.1. Chemical procedure
 A well-established method with
various types of samples.  A lot of references.
 Quick and easy to do.  Specimen preparation takes about two
days.
 Quality of fixation is reproducible.  All specimens in an experiment are
identical in quality.
5.1.2. Cryo-fixation
 All biological processes and bio-
chemical reactions are quickly stopped 
(in a few milliseconds).
1. Cryo-fixation by slamming
 No specific sampling is required.
 Can be used with every kind of biopsy.
 Method is quick and easy to perform.
 Time between sampling and freezing  Typically less than 10 seconds with
can be very short. cultured cells.
 Quality of fixation is quite satisfactory.  Usually, according to the resolution of
the analytical method, a thickness up to 10
µm is usable for EFTEM studies and up to
20 µm for SIMS analyses (see. References
15 and 16).
2. Cryo-fixation by high-pressure
freezing
 HP-freezing is very efficient  See Reference 19.
(temperature and pressure are well
controlled).
 With specimen carriers (see Figure
20.10), high quality freezing can be
obtained on thicker samples than with
slamming.  Up to 200 µm.
 The copper tube holder is very suitable

for suspensions of cells, bacteria,
virus, etc.
 New developments in the HPF device  The new EMPACT2 + Rapid Transfer
simplify the sampling procedures System (RTS) in the Leica apparatus reduce
(shorter and easier to do). the interval to about three seconds between
stimulation and the freezing of the material.

5.1.3. Dehydration
Cryo-dehydration procedures tend to 
minimize or to suppress element losses
and/or delocalization during ice removal
from cryofixed samples.
1. Dehydration by freeze-substitution
 Procedure is very easy.
 Apparatuses are commercially avail-
able and widely used.
 Requires minimum handling of  Especially with the new AFS2 (Leica)
samples during freeze-substitution. using the FSP (Freeze-Substitution Pro-
cessor).
 Some commercial devices (such as the
EM AFS) include extraction of toxic
fumes directly from the cryo-
substitution chamber.  Safer to use.
2. Dehydration by freeze-drying
 The best way to minimize loss or  See Reference 6.
redistribution of elements (better
than cryo-substitution).
 No handling of samples during
cryosorption.
5.1.4. Resin embedding

Embedded specimens can be stored for
unlimited periods of time. Sections can be
handled and observed at room temperature.
1. Embedding in Spurr’s resin

 Good resistance to analytical beams
(electrons or ions).
 When samples are freeze-dried prior to
embedding in pure resin, there is very
little loss or delocalization of labile
ions or molecules.
 Easy sectioning of Spurr’s embedded
samples by floating on fluid.
2. Embedding in Lowicryl HM 23
resin
 Mixture fluid down to –80°C
 Polymerization at –80°C  Under UV light.
 Enhanced preservation of protein and
membrane structures.

5.1.5. Sectioning
1. Sectioning on fluid
The cutting process generates section compression and subsequent charging phenomena.
Floating sections on a fluid allows surface tensions to stretch the sections and restore
them to initial dimensions.
 Easy to perform
 Ultrathin and regular sections (40 to  Ultrathin and regular sections are
50 nm thick) can be obtained. required for ETEM analysis.
 Glycerol (anhydrous solvent) can
reduce element loss during sectioning  Work in dry atmosphere (hygrometry
(but it should remain dry). lower than 50%) with fresh pure glycerol.
2. Dry sectioning
 Avoids element losses during  When very soluble elements are
sectioning. analysed.
 Dry sectioning is easier with Lowicryl  Ionizer should be used.
HM23 resin than with Spurr’s resin.
5.1.6. Elemental analysis

1. EFTEM
 Particularly well adapted for light  Atomic number (Z): 3 to 30
element analysis.  However, some elements with Z > 30
can be analysed.
 Most significant and detectable
elements in biological samples are N,
O, Mg, P, S, Ca and Fe, etc.
 Some exogenous elements can be  Applications in trace element and
detected (Al, Ni, In, Pt, etc.). pharmacological studies (see Reference 11).
 High local resolution of element distri-  In biological samples, elements can be
bution localized with a resolution of about 5 nm.
 High sensitivity  With respect to x-ray microanalysis.

 Adapted for elements concentrated in
very small volumes.
 Easy correlation between ultrastruc-  An ultrastructural image is acquired in
tural information and chemical the same area as the analytical data.
mapping
 Possible correlation with other  Particularly with methods using the same
analytical methods type of fixed sample.

2. SIMS
 High sensitivity in elemental  Slightly less in molecular analysis with
analysis. static SIMS (TOF-SIMS).
 Most significant and detectable  CN, P, S are useful for structural
elements in biological samples are characterisation.
CN, O, P, S, Br, I, Pt, Ca, Fe, etc.  I, Br, Pt can be used in molecules as
tracers for pharmacological studies.
 Trace element analysis.  Particularly for elements with high
electron affinity (Pt, Se, Ni, Ag, etc.).
 The only technique that can  Isotope analysis makes it possible to
differentiate between isotopes of the study intracellular and/or intercellular
same chemical element. metabolic pathways by use of stable
isotope-labelled tracer molecules.
 Subcellular resolution.  Particularly in dynamic SIMS.
 Possible correlation with other  Particularly useful, for example, with

methods. TEM to correlate chemical mapping with
subcellular structures.
5.2. Disadvantages
5.2.1. Chemical procedure
 Fixation occurs progressively from  Loss of membrane functions precede
outer cell to inner cell. inner fixation of cells; free water and ion
displacements occur across the cell
membrane.
 Fixation and dehydration solutions  Small molecules and free ions are
facilitate diffusion and/or precipi- subjected to diffusion.
tation of cellular material.
 Severe extraction cell  For instance, solvents used for

of some
constituents. dehydration and resin infiltration extract
lipids.
 Deformation of surface membranes
and organelle volume.14
5.2.2. Cryo-fixation
 Cryo-fixation is restricted to small 
samples.
 Cryoprotectant agents that improve the
quality of freezing cannot be used for
analytical purposes.
 Parameters to control high-quality
freezing are not easily determined and
conditions have to be defined and
adapted for each new sample.

1. Cryo-fixation by slamming
 For analytical purposes, slamming is  HPF is better for this kind of sample.
less adapted for fixing biological
suspensions than for biopsies and
adherent cell cultures.
2. Cryo-fixation by high-pressure
freezing
 Specific care is required concerning
sampling.
 Sampling of some tissues is very  For example, muscular fibre or plant
difficult. root.
 Some tissues cannot be cryofixed by  Lung tissue, for instance.
HPF.
 Due to the tiny size of the holder,  Inner diameter of carriers for EM PACT
sample handling is not easy and takes is only 1.5 mm.
at least 10 to 15 seconds to place the
sample and fill properly the carrier
prior to freezing.
5.2.3. Dehydration
1. Dehydration by freeze-substitution
 Samples are in contact with organic  Risk of loss or at least redistribution of
solvents for a long period of time up to soluble elements.
the embedding step.
2. Dehydration by freeze-drying
 Very time-consuming process.  More than 8 days are needed for
dehydration because a slow rate of increase
in the temperature is required (typically
0.4°C/hour).
 The cryosorption chamber is not really  Osmium is useful for ultrastructural
adapted for treating samples with observations by EFTEM.
osmium vapours.
 Some samples are difficult to  Particularly for plant tissues.
dehydrate properly.
5.2.4. Resin embedding

1. Embedding in Spurr’s resin
 Only fluid at room temperature.  Viscous below 0°C.
 Polymerised by heat.  65°C.
2. Embedding in Lowicryl HM23 resin

 Poor resistance to analytical beams  For analytical purposes, sections cannot
(electrons or ions). be protected by support film.

5.2.5. Sectioning
1. Sectioning on fluid
 Soluble elements from sample can be  Sections have to be quickly picked up
removed or displaced during section (10 seconds maximum) to minimize
floating. element delocalization.
 Glycerol modifies the resin.  Temporary softening of the resin
 Glycerol is a hydrophilic absorbent for  Sectioning should be carried out in a
atmospheric water. room where hygrometry is lower than 50%.
 Potential contamination of the sample

sections by glycerol.
2. Dry sectioning
 Dry sectioning of epoxy resin blocks  Almost impossible to obtain 50 to 70
is tricky. nm-thick dry sections for EFTEM.
 Typical dry sections are thicker than
100 nm.
 Laying down dry sections flat on a  Thickness variation alters EFTEM

TEM grid is very difficult. analytical data.
It is easier to put dry sections flat on a
prewarmed SIMS holder.
5.2.6. Analysis
1. EFTEM
 Complicated method for mapping a
limited number of elements in each
experiment.
 Analysis is restricted to a limited area. Analysis is limited to a region of a cell.
 No molecular information.
 Data are related to the global con-  Free and bound elements cannot be
centration of an element. distinguished.
 The evaporation process can occur
under the electron beam for some
elements.  Na, K, Cl.
 Although the carbon signal is intense,  Carbon is extensively present in
its biological usefulness is limited. embedding medium.
 K mapping is very difficult to obtain.  K signal is masked by the strong carbon
signal.
2. SIMS
 Only fixed samples can be analyzed.
 In dynamic SIMS, organic molecules
can only be visualized through specific
exogenous atoms or isotopes labelling.
 The data provide the global con-  Free and bound elements cannot be
centration of an element. distinguished.
 Identification of subcellular structures
is difficult and usually requires
correlation of TEM imaging.


 Why:  For more than 50 years, a number of
 A lot of well-established procedures
articles and books have been written
for various kinds of samples. concerning problems in processing
 specimens from both animal and plant
domains.

 When:  For example, to analyse components
 For structural TEM, EFTEM and strongly bound to cellular structures and/or
SIMS analysis when other methods that cannot be easily removed by solvents
are not required or failed. (insoluble crystals, clusters such as ferritin,
 etc.).


6.2. Cryo-Fixation
6.2.1. Cryo-fixation by slamming
 Why: 
 Quick and easy procedure that can be
used with biopsies of almost any type
of biological tissue.
 When:
 Using cultured cells.  Particularly on Thermanox® type film.
 Nature of tissue does not allow HPF  For instance, lung tissue cannot be fixed
fixation. by the HPF method.
 Structure of tissue is not compatible  Muscle or lung tissue is difficult to
with an easy sampling for HPF system. sample for an HPF carrier.
 Fixation of relatively large specimens  Up to 25 mm2.
is required.
 Studying dynamic processes.
6.2.2. Cryo-fixation by high-pressure freezing

 Why: 
 HP freezing is the only method to
properly freeze biological samples up
to 200 µm in depth.
 When:
 Organ biopsies, cultured cells on  Tools have been created specifically:
sapphire disks or biological biopsy needles, carriers.
suspensions. 


6.3. Dehydration
6.3.1. Dehydration by freeze-substitution
 Why: 
 Easy procedure that can be used with
cryofixed biopsies of almost any type
of biological tissue.
 When: 
 Analysis is restricted to components
strongly bound to cellular structures
and/or that cannot be easily removed
by solvents.
 Dehydration will take too much time  For example, freeze-drying of plant
by freeze-drying. tissues can require more than three weeks.


6.3.2. Dehydration by freeze-drying
 Why: 
 To avoid use of any solvent prior to
analysis.
 When: 
 Anytime molecular redistribution can 
potentially occur with use of solvent.  To minimize chemical localisation
artefacts.


6.4. Resin Embedding

6.4.1. Embedding in Spurr’s resin
 Why:
 Low viscosity (about 60 cP) at room  Infiltration of sample is possible without
temperature. using any solvent.
 Ultrathin sectioning is very easy by
floating sections on water.  Particularly useful for EFTEM analysis.
 Resin withstands electron or ion

bombardment.

 When: 
 Samples of even thin thicknesses are  EFTEM analysis requires ultrathin
required. sections: 40 to 50 nm-thick.

 A good resistance of section to 
electron bombardment is required for  For example, samples are subjected to
specific analysis. severe irradiation during EELS analysis.


6.4.2. Embedding in Lowicryl HM23
 Why:
 Embedding is achieved at low  Down to 80°C.
temperatures. 
 When: 
 Antigenic properties of samples have
to be preserved for ultrastructural
characterisation by immunolabelling to
complete analytical studies.
 Dry sectioning is needed.  It is easier to perform dry-sectioning on
Lowicryl than on Spurr’s resin.


6.5. Sectioning
6.5.1. Sectioning on fluid
 Why:
 Thin and regular sections are easily  40 to 50 nm-thick for EFTEM analysis.
collected on water.
 When: 
 Standard use.  However, sections have to be collected
quickly to avoid potential molecular
redistribution.

6.5.2. Dry-sectioning
 Why: 
 To avoid contact of samples with any
solvent.
 When: 
 Analysing the distribution of labile 
components. 


6.6. Analysis
6.6.1. EFTEM
 Why: 
 Elemental analysis at the subcellular
level.
 When: 
 High spatial resolution in elemental  About 2 to 5 nm.
analysis is needed.
 To map clusters of a few hundred
atoms.

6.6.2. SIMS
 Why: 
 It is the only technique that can 
differentiate between isotopes of the  It is possible, for instance, to
same chemical element. differentiate in the same mapping 12C and
13
 To analyse trace elements or drug C or 14N and 15N to analyse the cellular
targeting. distribution of a 15N or 13C labelled
molecule.
 Highly sensitive analytical technique. 


 When: 
 Mapping distribution of specific, 
exogenous atoms or isotope labelled  Lateral resolution of about 100 nm.
molecules at a cellular level. 














7. OBSERVED RESULTS
 Figure on the Chapter’s title page Left: Cryo-fixed cultured osteoblasts.

Top left: Energy filtered image by EFTEM.
Bottom left: Calcium mapping in the same
region by ESI. Experimental EELS
spectrum for calcium recorded from the
same area is superimposed.
Right: SIMS imaging of cultured
osteoblasts.
Top right: Phosphorus mapping.
Bottom right: Sulphur mapping.
 Figure 20.12, A-D: EFTEM data
(see colour insert following page )
1. Description  Calcium distribution in quiescent
cultured osteoblasts (left) and in osteoblasts
perfused with 5 µM ionomycin (right).
2. Method  EFTEM analysis.
3. Tissue  Cell culture.
4. Fixation  Cryo-fixation by slamming.
5. Embedding  Spurr’s resin.
6. Visualisation  EELS and ESI with an energy filtered
transmission electron microscope.
7. Bar  0.5 µm.
8. Comments  Calcium mapping is in grey spotted (in
red, see the colour insert), superimposed on
an energy filtered image obtained at 250
eV (inverted contrast without staining).
 White arrows: mitochondria with dense
or clear matrix, E.R.: endoplasmic
reticulum.
 A and C: Data for unstimulated cells.
Dense mitochondria exhibit a weak
calcium signal. In experimental EELS
spectrum (purple in C, see colour
insert) recorded in circle in A, Ca-L2,3
signal (344 eV) is below the limit of
detection.
 B and D: Data for cells stimulated by
5 µM ionomycin. Dense mitochondria
exhibit a strong calcium signal and
endoplasmic reticulum an identifiable
weak calcium signal. In experimental
EELS spectrum (green in D, see
colour insert) recorded in circle in B,
Ca-L2,3 signal is clearly visible.
 C-K predominant carbon signal in C
and D.
 For more information, see Reference 4.


















 Figure 20.12, E- H: SIMS images

(see colour insert following page)
1. Description  Ultrastructural distribution of an

iodinated drug within B16 melanoma cells
from a mouse lung colony.
2. Method  SIMS analysis.
3. Tissue  Small sample of lung.
4. Fixation  Cryo-fixation by slamming.
5. Embedding  Spurr’s resin.
6. Visualisation  SIMS images (carried out with a
CAMECA NanoSIMS 50 microprobe).
7. Bar  Field of view: 25µm.
8. Comments  Iodobenzamide (BZA) was injected
intravenously (2.6 µmoles) in tumour
bearing mice six hours before sacrifice.
 E: Elemental distribution of CN- ions
(m = 26). Melanosomes containing the
melanin grains rich in carbon and
nitrogen appear as bright spots.
 F: Elemental distribution of S- ions (m
= 32). Part of the melanin biopolymer
(pheomelanin) contains sulphur.
 G: Elemental distribution of I- ions (m
= 127). Iodine is representative of the
drug distribution and appears to be
specifically located within melano-
somes.
 H: Merge (CN-: blue, S-: green, I-:
red). Co-localisation of the observed
ions leads to colour variations. It
appears that melanosomes containing
mainly eumelanin (arrow “a”) without
the sulphur element bind BZA
(magenta colour corresponding to
merge blue (CN-) and red (I-) while
melanosomes richer in pheomelanin
(arrow “b”) do not bind BZA and
appear in cyan colour (merging of CN-
in blue and S- in green). Arrow “c”
corresponds to white melanosomes
(merging of blue, green and red)
containing a balanced mixture of
pheomelanin and eumelanin (for
colour indications, see colour insert).
For more information, see Reference
9.


8. REFERENCES
1. Acetarin, J.D. et al. Developments of new Lowicryl resins for embedding

biological specimens at even lower temperatures, J. Microsc., 143, 81, 1986.
2. Benninghoven, A. Surface investigation of solids by the statical method of
secondary ion mass spectroscopy (SIMS), Surf. Sci., 35, 427, 1973.
3. Bordat, C. et al. Calcium distribution in high-pressure frozen bone cells by electron
energy loss spectroscopy and electron spectroscopic imaging, Histochem. Cell Biol.,
109, 167, 1998.
4. Bordat, C. et al. Direct visualization of intracellular calcium in rat osteoblasts by
energy-filtering transmission electron microscopy, Histochem. Cell Biol., 121:31,
2004.
5. Castaing R. and Slodzian G., Microanalyse par émission ionique secondaire,
J. Microsc., 1, 395, 1962.
6. Edelmann, L. Freeze-dried and resin-embedded biological material is well suited
for ultrastructure research, J. Microsc., 207, 1, 5, 2002.
7. Egerton, R.F. Electron Energy Loss Spectroscopy in the Electron Microscope. 2nd
ed., Plenum Press, New York,USA, 1989.
8. Guerquin-Kern J.L. et al. Progress in analytical imaging of the cell by dynamic
secondary ion mass spectroscopy (SIMS microscopy), BBA, 1724, 228, 2005.
9. Guerquin-Kern, J.L. et al. Ultra-structural cell distribution of the melanoma
marker iodobenzamide: Improved potentiality of SIMS imaging in life sciences,
Biomed. Eng. Online, 3:10, 2004.
Applications. 4th ed., Cambridge University Press, Cambridge, UK, 2000.
11. Jeanguillaume, C. Electron energy loss spectroscopy and biology, Scan. Microsc.,
2, 437,1987
12. Kellenberger, E. The potential of cryo-fixation and freeze substitution:
Observations and theoretical considerations, J. Microsc., 161, 2, 183, 1991.
13. Nicolas, G. Advantages of fast-freeze fixation followed by freeze-substitution for
the preservation of cell integrity, J. Electr. Microsc. Tech., 18 (4), 395, 1991.
14. Murk, J.L. et al. Influence of aldehyde fixation on the morphology of endosomes
and lysosomes: Quantitative analysis and electron tomography. J. Microsc., 212, 1,
81, 2003.
15. Reipert, S. et al. Cryo-fixation of epithelial cells grown on sapphire coverslips by
impact freezing, J. Microsc., 209, 2, 76, 2003.
16. Sitte, H. et al. Cryo-fixation without pre-treatment at ambient pressure, in
Cryotechniques in Biological Electron Microscopy, R.A. Steinbrecht and K.
Zierold, eds., Springer-Verlag, Berlin, Germany, 87, 1987.
17. Slodzian, G. et al. Scanning secondary ion analytical microscopy with parallel
detection, Biol. Cell, 74, 43,1992.
18. Spurr, A.R. A low viscosity epoxy-resin embedding medium for electron
microscopy, J. Ultrastruct. Res., 26, 31, 1969.
19. Studer, D. et al. A new approach for cryo-fixation by high-pressure freezing,
J. Microsc., 203, 3, 285, 2001.

Correlative Light and Electron Microscopy 539
CONTENTS

1.1. Preparation................................................................................................ 543
1.2. Imaging..................................................................................................... 543
3.1. Materials ................................................................................................... 545
3.2. Products .................................................................................................... 546
3.3. Solutions ................................................................................................... 547
4. PROTOCOLS .................................................................................................... 549
4.1. Protocols for Materials Needed for On-Section Immunolabelling ........... 549
4.1.1. Poly-L-lysine coating of coverslips ............................................ 549
4.1.2. Mounting solution for fluorescence microscopy......................... 549
4.2. Sample Preparation................................................................................... 550
4.3. Toluidine Blue O Staining ........................................................................ 551
4.4. Protocol for On-Section Immunolabelling ............................................... 551
5.1. Advantages ............................................................................................... 556
5.2. Disadvantages........................................................................................... 556
6.1. Specimen Preparation ............................................................................... 556
6.2. Marker Systems ........................................................................................ 557
6.3. Light Microscopy ..................................................................................... 558
6.4. Staining for Electron Microscopy............................................................. 559
8. REFERENCES .................................................................................................. 564


On the search of the holy grail to image cellular ultrastructure in its native state, electron
microscopists have established a wealth of preparation techniques, which result in
excellent preservation of cellular components.1-11 Cryo-preparation techniques form this
groundwork as dealt with in depth throughout this book.
In light microscopy, recent years have resulted in major developments in:
1. Imaging: Confocal microscopy12-15 and deconvolution algorithms.16
2. Detection: A large variety of fluorochromes with a wide range of excitation and

emission spectra as well as improved photostability.
3. Live imaging: Green fluorescent protein (GFP), their variants17 and alternatives
(ReAsh, Lumio Technology).18 These improvements have almost exclusively been in
the area of fluorescence microscopy. Dark-field imaging techniques, such as epi-
fluorescence microscopy have improved contrast, which results in a high signal-to-
noise ratio.
Nevertheless, in light microscopy, little attention has been paid to preparation methods
that remain crude and harsh, live imaging excluded. In fluorescence microscopy the
deterioration of cellular ultrastructure is invisible and, therefore, neglected. It should,
however, be kept in mind that in the dark-field mode, particles smaller than the
resolution of a light microscope can be detected, in the same way the eye can see
microscopical particles of dust in a beam of sunlight.
We suggest combining the advantages of the developments in (fluorescence) light

microscopy with the knowledge of high resolution preparation techniques in electron
microscopy. The basic concept is to inspect serial (ultra) thin sections of the very same
sample block, stained and immunolabelled, in both light and electron microscopes.
Histological stains are used for orientation in the tissue; the labelling used under the light
microscope identifies the structure of interest, which is further analysed at high
resolution in the electron microscope
The high-resolution objectives of light microscopes have inevitably a very low depth-of-
field, much smaller than the thickness of even one cell layer. This results in out-of-focus
information, hence, blurring the image and reducing the final resolution. The principle of
confocal (fluorescence) light microscopy or the application of deconvolution algorithms
partially overcomes this drawback by generating three-dimensional stacks of in-focus
optical sections.

However, we would like to encourage the use of very thin (50 to 500 nm) sections for
histological analysis by conventional bright-field, phase contrast or interference contrast
light microscopy and localisation of cellular components by wide-field epifluorescence
microscopy.19-21 Physical sectioning produces much higher z-resolution data than can be
obtained with confocal light microscopy of thick specimens. The section is thinner than
the depth-of-field of the high-resolution objectives and, therefore, all structures imaged
are perfectly in focus.
In addition, the high-quality preparation methods used in electron microscopy guarantee

the preservation of high-resolution structural information. A further advantage is the fact
that a subsequent ultrathin section, containing the same information as studied by light
microscopy, can be analysed in the electron microscope. In the electron microscope, the
label can directly and unambiguously be assigned to a defined cellular structure. Thus
this approach allows correlative light and electron microscopy investigations: firstly,
easy orientation in the tissue; secondly, wide-field evaluation of labelled cells; and
thirdly, high-resolution identification of the labelled substructures.21,22-25
Now, the described concept of correlative microscopy is being applied for cryo-electron
tomography. The GFP-labelled material, whole mount or CEMOVIS cryo-sections (see
Chapters 11, 12) is cryofixed and mounted on a special cryostage attached to a light
microscope. The position of the labelled structures are identified on a finder grid and
recorded. Then the grid is transferred to the cryo-electron microscope and tomograms of
the respective structure at high resolution are recorded (see Chapter 12).26
This concept for correlative light and electron microscopy is, of course, not limited to
samples prepared by cryo-fixation, but can also be applied to chemically fixed samples
processed at room temperature or, even better, to progressive lowering the temperature
methods27 (PLT, see Chapter 18) or to cryo-sectioning according to Tokuyasu28 (see
Chapter 19). The Tokuyasu technique often results in more efficient labelling than resin-
based techniques.

1.1. Preparation
 The biological specimen is, whenever  Excellent structural preservation close to
possible, cryofixed, freeze-substituted the native state.
and low-temperature embedded into  For morphology epoxy is the resin of
methacrylate (or epoxy) resins. choice.
 For higher label efficiency, however,
methacrylate resins are preferred.
 Alternatively, the biological specimen  Highly efficient immunolabelling.

is chemically fixed (see Chapter 19) or
even better prepared by the emerging
cryo-fixation/rehydration method29
(see Chapter 14) followed by
Tokuyasu cryo-sectioning.
 After chemical fixation the sample  This is a good compromise when cryo-
may also be embedded into fixation or the Tokuyasu cryo-section
methacrylates by standard procedures method cannot be applied.
or preferentially by the progressive
lowering of the temperature (PLT)
method27,30 (see Chapter 18).
 Ultrathin serial sections are cut and in  Observation of the same structure by
sequence mounted for histology, light and electron microscopy.
immunofluorescence and immunogold
staining.
1.2. Imaging
 The sections for light microscopy are  High z-resolution.
even thinner than the depth-of-field of
the high numerical aperture objectives.
Further, on resin sections the label is
restricted to the section surface.
 Correlation of light and electron  Easy orientation and high-resolution

microscopy. data.

Figure 21.1 Flow diagram of the different preparation steps.

 The sample is either embedded in methacrylate (or epoxy) for resin sections or
prepared for Tokuyasu cryo-sections.
 Subsequent sections are cut and collected for the different applications, one 500 nm for
Toluidine blue staining and a few 50 to 100 nm for immunofluorescence and
immunogold labelling.
 Because these sections are thin and collected subsequently, the same organelles may be
examined by immunofluorescence and electron microscopy.
 If ultra-small gold particles are used for labelling, a silver enhancement step needs to
be performed (see Chapter 23).

3.1. Materials
 Epifluorescence microscope
 (Oil) immersion objectives with the
highest numerical aperture
 High quality fluorescence filter  Band pass filter dedicated to the specific
fluorochromes:
 Chroma Technology Corp., Bellows
Falls, Vermont, USA.
 Omega Optical Inc, Brattleboro,
Vermont, USA.
 In Europe, available from AHF
Analysentechnik, Tübingen, Germany.
 Forceps  For ewample, Dumont #2a, #3, #4, #4N,
#7, Dumont & Fils, Switzerland.
 Loop  Self-made loops from platinum wire or
perfect loop, EMS, Fort Washington,
Pennsylvania, USA.
 Round glass coverslips, 12 mm in
diameter
 Scratching diamond  To mark the position of the sections on
the microscope slide.
 Moist chamber
 Copper or nickel grids for electron  In general, copper grids can be used for
microscopy, 50 to 200 mesh. The immunolabelling.
grids are preferentially covered with For long-term incubation (overnight) or if
carbon-coated Formvar or Pioloform the copper grids have no or a perforated
support film.31 support film or if they are used in
combination with azide, however, they will
be oxidised.
 Copper grids may impair silver
enhancement (see Chapter 23).
 Nickel grids are magnetic, therefore
demagnetised forceps have to be used.
 In addition, nickel grids may influence
the astigmatism in the electron microscope,
resulting in blurred images close to the grid
bars.
 Because of their magnetic properties,
nickel grids are essential in combination
with the automatic labelling apparatus from
Leica (IGL, Leica Microsystems, Vienna).
 Mercury lamps  Can be, 50, 100 or 200 W depending on
the fluorescence microscope used and the
manufacturer.

 Precentred mercury lamps  For optimal illumination in fluorescence

microscopy Osram HXPTM R 120W/45C,
Osram GmbH, Munich, Germany.
 Liquid light guide  For optimal illumination in fluorescence
microscopy, EXFO Life Sciences Group,
Ontario, Canada.
3.2. Products
 Toluidine blue O [C.I.52040]  Serva 36693.02; Fluka 89640; Sigma T
0394.
 Mowiol  4-88 Hoechst; Calbiochem 475904.
 DABCO, 1,4-Diazabicyclo[2.2.2.]-  Antifading agent.
octane
 PPD, p-Phenylene diamine  Antifading agent.
 DAPI  4´,6-diamidine-2-phenylindole; for DNA
staining.
 Alternatively, Hoechst 33342 and
Hoechst 33258 can be used for DNA
staining.
 Propidium iodide  3,8-Diamino-5-(3-diethylaminopropyl)-
6-phenanthridinium iodide methiodide to
stain DNA and RNA.
 Gelatine
 BSA, bovine serum albumin,
Fraction V
 Coldwater fish skin gelatine  Sigma G 7765.
 Skim milk powder
 BSA-c™  Acetylated BSA; Aurion, Wageningen,
The Netherlands.
 Poly-L-lysine  Sigma P 1274; MW 70,000-150,000 Da.
 0.1% ready-to-use solution, Sigma
P 8920.
 Primary antibodies  Against target protein.
 Fluorescent secondary antibodies  Against the species in which the primary
 Gold-labelled secondary antibodies  Against the species in which the primary
antibody was produced (for details, see
Chapter 23).
 Gold-labelled protein A  Binds preferentially to the Fc region of
rabbit, pig or human antibodies (for details,
see Chapter 23).

 Mouse IgG2a, IgG2b and IgG3 can also

bind to protein A under the standard
labelling conditions (pH 7.4). For mouse
IgG1, however, the pH must be higher,
~ 8.5.
 Can easily be homemade or purchased
from Dr. G. Posthuma, Department of Cell
Biology, Utrecht Medical Centre, Utrecht,
The Netherlands.
 (biotinylated) Lectins  Detection of specific sugar moieties.
Lectins are also successfully applied on
epoxy sections; Vector Laboratories,
Burlingame, California, USA.
 Uranyl acetate  Stain for electron microscopy.
USA.
 Lead citrate  Stain for electron microscopy.
 Methyl cellulose  Tokuyasu cryo-section embedding.28
 Sodium tetraborate  Stable alkaline buffer solution; Merck
106308.

3.3. Solutions
 137 mM NaCl buffer.
 2.7 mM KCl
 8.1 mM Na2HPO4
 1.5 mM NaH2PO4
 TRIS (pH 7.4)  Label buffer; at pH 8.5 this buffer may

 150 mM NaCl be used to detect mouse IgG1 antibodies
 50 mM TRIS with protein A gold complexes.
 pH adjusted with HCl
 PHEM buffer (pH 6.9)  Cytoskeleton buffer.32 This buffer can be

 60 mM PIPES helpful to suppress persistent background
 25 mM HEPES (E. van Donselaar, personal communi-
 10 mM EGTA cation).
 2 mM MgCl2

 Blocking solution
 Blocking protein dissolved in PBS or  0.045% (w/v) coldwater fish skin gelatine
TRIS. Every lab uses its own recipe and 0.5% (w/v) BSA, or
for blocking solutions. In distinct 0.2% (w/v) gelatine, 0.5% (w/v) BSA,
cases, specific blocking proteins, to Fraction V
be found empirically, are required. 0.5%1% (w/v) coldwater fish skin
We routinely use 0.045% (w/v) gelatine33
coldwater fish skin gelatine and 0.5% 1% (w/v) skim milk powder
(w/v) BSA in PBS. 1% BSA
 Centrifuge prior to use. 0.1% BSA-c (BSA-c should not be used
in combination with gelatine!)
 Toluidine blue O  Histological stain for orientation in the

 Dissolve 0.5% (w/v) Toluidine blue O light microscope.
in an aqueous solution of 1% sodium
tetraborate, pH 9.2.
 Filter and store as aliquots, e.g., in
microtubes.
 Centrifuge prior to use.
 Mounting solution for fluorescence  Preparation, see Section 4.1.2.

microscopy  Mounting solutions are also commer-
cially available (e.g., Vectashield, Vector
Laboratories or ProLong and SlowFade,
Molecular Probes, Invitrogene).
 Note: Vectashield is incompatible with
Cy fluorochromes and DAPI staining.
 Uranyl acetate
 2% (w/v) in double distilled water  See Chapter 19.
 2% (w/v) in 1.5 M oxalate pH 7.028
 Lead citrate34
 0.04 g lead citrate
 10 mL H2O
 100 μL 10 M NaOH
 Poly-L-lysine stock solution  Alternatively, the 0.1% ready-to-use

 Dilute 2 mg/mL (w/v) in double solution, Sigma P 8920 may be used.
distilled water.
 Methyl cellulose  To embed Tokuyasu cryo-sections,28 see

Chapter 19.

4. PROTOCOLS
4.1. Protocols for Materials Needed for On-Section Immunolabelling

4.1.1. Poly-L-lysine coating of coverslips
1. Clean round glass coverslips with  For cleaning, 1% (v/v) hydrochloric acid
ethanol. in 70% (v/v) ethanol or soap may also be
used.
2. Rinse with double distilled water.
3. Dry coverslips separately on a filter

paper in a large dish.
4. Place a 30 to 50 µL drop of poly-L-

lysine solution in the centre of the
coverslip.
5. Allow the coating to settle for at least

30 minutes in a moist chamber.
6. Rinse briefly with double distilled  Coverslips can be coated in large batches
water and dry face up on a filter paper and stored indefinitely in dust-free dishes at
for one hour at 60°C or overnight at room temperature.
ambient temperature.
4.1.2. Mounting solution for fluorescence microscopy

1. Dissolve 5 g Mowiol in 20 mL buffer
(100 mM TRIS/HCl; pH 8.0) and stir
for 16 hours.
2. Add 10 mL glycerol and stir again for

16 hours.
3. Remove undissolved Mowiol by  Aliquots can be stored at –20oC.

centrifugation.
4. Add 20 to 50 mg/mL DABCO or  DABCO and PPD are antifading agents

1 mg/mL p-phenylenediamine (PPD) and reduce photobleaching of the
to the final Mowiol solution. fluorescence signal.
 DABCO is dissolved preferentially at
60°C.
 PPD is more efficient in preventing
bleaching, but turns a brown colour within
one or two days, resulting in orange
background fluorescence. DABCO-
containing solutions, however, can be
stored for months in a freezer.

For our approach of correlative light and electron microscopy, any electron microscopic
preparation technique suited for thin sectioning can be used. Note that in the following
protocol all types of sections are treated in exactly the same way.
1. The specimen is cryofixed, freeze-  For optimal preservation of the cellular

substituted and low-temperature ultrastructure (see Chapter 13).
embedded into a methacrylate.  Alternatively, chemically fixed samples
are conventionally dehydrated or dehy-
drated by PLT before embedding into a
methacrylate (see Chapter 18).
2. Sample is chemically fixed and  In most cases, the label efficiency is

prepared for Tokuyasu cryo- higher on Tokuyasu cryo-sections than on
sectioning. resin sections (see Chapters 14, 19).
3. Sections are cut on a (cryo-) ultra-  The 500 nm section will be used for
microtome: one 500 nm section to be histological staining, e.g., toluidine blue for
mounted on a coverslip and a few 50– orientation.
100 nm sections to be mounted  The ultrathin sections on coverslips will
alternatively on coverslips and be used for fluorescence labelling and those
EM grids. on EM grids for gold labelling.
4. Resin sections are picked up from the

water trough with a loop and
transferred onto the coverslip or EM
grid. Cryo-sections are picked up from
the dry knife edge with a loop
containing a drop of, e.g., 2.3 M
sucrose solution and after thawing  See Chapter 19, for alternative picking
transferred onto the coverslip or EM up solutions
grid.
5. Touch the loop edge with a pointed  The dried resin sections can be stored for
piece of filter paper to remove the years without loss of antigenicity.
water droplet.
6. The section will attach to the surface.
7. Cryo-sections are picked up from the  Cryo-sections, without removing the

dry knife edge with a loop containing a drop of the transfer solution, can be stored
drop of sucrose solution. up to one year at 4oC.
 See Chapter 19.

4.3. Toluidine Blue O Staining

1. 500 nm thick resin sections are put on
a droplet of water on a poly-L-lysine
coverslip.
2. They are dried on a hot plate (60 to

80oC) and immediately incubated with
toluidine blue for 30 to 60 sec.
3. The sections are destained using a  Overstained sections can be recovered

stream of (tap) water. by destaining in ethanol.
4. For high-resolution light microscopy  Do not use embedding media containing

and long-term storage, the stained apolar solvents as they will destain the
sections can be embedded in a thin sections.
layer of Epon under a coverslip.
4.4. Protocol for On-Section Immunolabelling

1. The coverslips are put section side up  Note: For on-section labelling of resin
on a piece of parafilm, mounted in the embedded material, the same rules as for
moist chamber. solid-phase immunoadsorption apply. The
reaction is restricted to the surface,
therefore, not diffusion-dependent. This
implies that the time for the washing steps
is not critical.
Figure 21.2 Rehydration of sections

mounted on 12 mm glass coverslips with
buffer.

2. For EM, drops of buffer are put on a

piece of parafilm and the grids on the
drop sections side down.
3. The sections are incubated with PBS  Resin sections are rehydrated and the
for 10 minutes. sucrose drop on cryo-sections is removed.
4. Then they are incubated 2 × 10  The blocking proteins cover non-

minutes with blocking buffer. specific binding sites and reduce back-
ground labelling due to charge or
hydrophobic interactions.
 We routinely use 0.045% (w/v) cold
water fish skin gelatine and 0.5% (w/v) BSA
in PBS.
5. Incubation with the primary antibody  For an unknown antibody, it is

diluted in blocking buffer for preferable to first make serial dilutions in
30 minutes to 2 hours at room temp- blocking buffer: e.g., 1/10, 1/30, 1/100,
erature. 1/300, 1/1000, etc., to evaluate the
concentration at which a clear signal is
visible without background.
 As a rule of thumb, the appropriate
concentration of reacting antibody is in the
range of 0.1 to 1 μg/mL.
 To save precious primary antibodies, the
grids may be put on the very same drop on
the coverslip.
Figure 21.3. The coverslips with the

sections facing upward and the grids with
the sections facing downward are incubated
in the same drop of primary antibody.
6. Wash six times for about two minutes

with blocking buffer.

 The drops on the coverslip are removed

with a pipette and a fresh drop is
immediately added. Take care that the
sections never dry.
Figures 21.4 and 21.5 The drop of buffer

or antibodies is aspirated with a pipette tip
connected to a vacuum flask hooked to a
vacuum pump.
Figure 21.6 A fresh drop of buffer or

antibody is immediately added on the
coverslip to prevent drying of the section.

 The grids are lifted off the drop with

forceps, blotted with a tissue or filter paper
and immediately transferred on a fresh drop
of blocking buffer.
Figure 21.7 During transfer from one drop

to the other, the remaining drop of liquid on
the grid is drained off by a short touch to a
piece of tissue. Take care that the grids do
not dry.
Figure 21.8 Coverslips (left) and grids

during the label process.
7. Incubation with the secondary  For light microscopy, fluorescent

antibody diluted in blocking buffer for antibodies are used.
one hour at room temperature.  For electron microscopy, gold
conjugated secondary antibodies or protein
A gold is used.
8. Wash two times about two minutes

with blocking buffer.
9. Wash four times about two minutes

with PBS.
10. For light microscopy, DNA may  DAPI and PI penetrate methacrylate
be counterstained with 0.1 to 0.4 sections and, therefore, the signal intensity
μg/mL DAPI or any nucleic acid may reflects the thickness of the section. They
be stained with 0.1 to 0.4 μg/mL do not react with epoxy resin embedded
propidium iodide. nucleic acids.

11. For electron microscopy, the  For light microscopy, the label should
sections on the grids are fixed with 1% not be fixed, as glutaraldehyde at this
glutaraldehyde in PBS. concentration may cause autofluorescence!
12. Wash four times with double  All salt ions, especially Cl- and PO43-,
distilled water. have to be removed as they will interfere
with silver enhancement and uranyl
staining.
13. Silver enhancement, if ultra-small  See Chapter 23.

gold particles were used as a marker.
14. Wash extensively with double

distilled water.
15. The fluorescently labelled sections  Use a very small droplet of mounting
are mounted with Mowiol. medium. The medium must not be
squeezed out from the coverslips. Excess
Mowiol will pollute the front lens of the
microscope objectives.
16. The sections for electron
microscopy are counter-stained with
uranyl acetate and if desired with lead
citrate.
17. Tokuyasu cryo-sections are stained  Methyl cellulose must always be kept in
with uranyl acetate and embedded in the cold because it dissolves better in the
methyl cellulose. cold.
 The embedding is done on ice (see
Chapter 19).
18. The fluorescently labelled prepar-  DAPI or PI counter stains may be lost
ations can be stored for months in a during storage and PPD-containing media
refrigerator or freezer. turn brown. For restoration, the coverslips
can be removed after submerging in PBS
and thereafter again stained and mounted.
 Even additional immunolabelling or
“section recycling” is possible (bound
antibodies can be released as described for
immunoadsorption techniques, e.g., with
4 M magnesium chloride, MgCl2).

5.1. Advantages
 Excellent structural preservation in  Application of cryo-electron microscopy
light and electron microscopy. techniques.
 Ease of orientation in stained  Large overview.
histological tissue sections.
 Fast overview on labelled structures.  Which cells are labelled, what is the
level of expression, evaluation of labelling
conditions (label efficiency, background)?
 High z-resolution in light microscopy.
 Correlation of label and cellular
ultrastructure.
5.2. Disadvantages
 No live imaging possible. 
 Dedicated instrumentation for prepar- 
ation is needed.
 The preparation is rather time con-  Standard preparation for light
suming. microscopy is comparably faster.
 Limited access to antigens.  Only the antigens on the section surface

are accessible to antibodies.


 Cryofixed, freeze-substituted and  See Chapter 13.
resin-embedded samples in general
have a better morphology.
 Resin-embedded samples are easy to
store and antigenicity is retained for
many years.
 Resin sections are easier to handle.  Large, undistorted sections are easier to
obtain than with Tokuyasu’s cryo-
sectioning.

 The cellular morphology is preserved  The epoxy resin covalently binds to the
much better by epoxy embedding; biological structures, masking protein
however, epoxy sections are less epitopes.
suited for immunolabelling of protein
epitopes.
 Carbohydrate chains, such as sugar  Sugar moieties can be labelled with
moieties, nucleic acids, biotin and specific (biotinylated) lectins.
digoxigenin, are much less affected.  Biotin and digoxigenin can be used as
tags in combination with epoxy embedding.
 Methacrylate-embedded samples are  Methacrylates do not covalently bind to
better suited for immunolabelling but biological structures, therefore, the cellular
the cellular morphology is inferior. morphology suffers, but the protein
epitopes are more accessible.
 Cryo-sections of chemically fixed  During dehydration for resin embedding
samples are optimal for immuno- the tertiary structure of proteins may be
labelling. changed resulting in reduced or even loss of
antigenicity.
 The proteins are not denatured by
organic solvents and the matrix of the
section is less dense allowing easier access
for antibodies (see Chapter 19).
 Cryofixed, freeze-substituted rehydra-  See Chapter 14.

ted samples further prepared for
Tokuyasu cryo-sectioning may com-
bine good structural preservation and
antigenicity.29
6.2. Marker Systems

 Fluorescent labels are preferred to  Fluorescence microscopy is a dark-field
coloured enzyme reaction products. imaging mode that excels in contrast
compared to bright-field imaging,
therefore, the labelling is more intensive.
 In fluorescence microscopy, double or  There is a wealth of fluorochromes
multiple labelling is easy. available. In contrast to gold particles, the
fluorochromes do not affect the binding
properties of the antibodies.
 Protein A gold is used in combination  At neutral pH, protein A has a very low
with primary antibodies from rabbits, binding affinity to mouse or rat monoclonal
pig or man. antibodies. It is, however, successfully used
with a bridging antibody produced in rabbit
or pig.
 At pH 8.0 or higher, protein A can bind
to mouse IgG1 antibodies. Here it is
absolutely necessary to fix the protein A to
the antibodies with glutaraldehyde before
the sections are stained with aqueous
(acidic!) uranyl acetate.

 Protein A gold complexes generally  This is only true for homemade

have a higher label density and do not protein A complexes or those supplied by
aggregate. the UMC Utrecht.
 Gold-tagged secondary antibodies are
used in combination with all other
primary antibodies.
 For electron microscopy, single  Valid controls for multiple labelling are
labelling is preferred to double or even laborious and often difficult to design,
multiple labelling. while the extra information is minimal.
 We prefer labelling of subsequent serial
sections for different antigens.
6.3. Light Microscopy

 Even illumination of the field of view  Centre the light source meticulously.
without a light gradient is a pre-  A simple and more efficient approach is
requisite. using precentred mercury lamps (Osram
HXPTM R 120W/45C) coupled to the
microscope via a liquid light guide. The
system is available as X-Cite 120 from
EXFO Life Sciences Group, Ontario,
Canada.
 Optimal alignment of the microscope  Remember, even an epifluorescence
is essential. microscope must be aligned according to
Köhler.
 For fluorescence microscopy, oil  To collect as much signal as possible, a
immersion objectives with a high large opening angle of the objective is
numerical aperture (NA) must be used essential.
for all magnifications.  The medium between the glass coverslip
and the objective should have the same
refraction index as glass to avoid loss of
light by reflection at the glass surfaces.
 The deflected light can also cause
background noise because it does not
contain any object information.
 Though phase contrast objectives
prolong the exposure time (by 10 to 15%)
they are still suitable for fluorescence
imaging. They offer the advantage of
correlative phase contrast imaging for
better orientation, e.g., phase contrast
image overlaid with fluorescence signal.
http://www.micro.magnet.fsu.edu/optics/ind
ex.html

 Sections may be searched with a low  The edges of the sections are high-
magnification phase contrast objective lighted.
fitted with a wrong phase ring (e.g.,
Ph 3) to obtain pseudo dark-field
illumination.
 Sections may be found using DAPI or  The DAPI and PI signals in general are
PI staining with a low magnification very bright.
objective.
 As a general hint to find the focal  Under these conditions, the objective is
plane of the object, close the focussed in the plane of the sections.
illumination diaphragm and focus the  This operation requires an optimally
rim. aligned microscope.
 Close the illumination diaphragm so  Only the light directly emitting from the
that only the field of imaging is object contains image information.
illuminated.
 Further closing results in an even  Light outside the field of view only
better signal-to-noise ratio. increases the background noise.
6.4. Staining for Electron Microscopy

 The gold-labelled resin sections are  The contrast of the gold label, especially
stained with aqueous uranyl acetate of smaller gold particles, e.g., 5 nm, is quite
and then examined. Only if the weak and is easily overlooked on heavily
contrast is too weak, a second staining stained sections.
step with lead citrate follows.
 Alcoholic uranyl acetate solution may  The remaining blocking proteins may be
cause dirty grids. precipitated by the alcohol.
 There are several methods to stain and
embed Tokuyasu cryo-sections:
a) Staining for 10 minutes with neutral  See Chapter 19.
uranyl oxalate followed by 5 minutes
with acid uranyl acetate before
methylcellulose embedding.
b) Embedding in methyl cellulose
containing 0.2 to 0.4% uranyl
acetate.

7. OBSERVED RESULTS
 Figure on Chapter’s title page  Localization of β-catenin in the adherens

junctions, but also in the Z-line of guinea
pig heart muscle.23
 Figure 21.9  A 500 nm-thick section of Lowicryl-

embedded rat duodenum is stained with
Toluidine blue. The bright field mode offers
a good overview and orientation within the
tissue.
 Multiple exposure of a rat

 Figure 21.10 (see colour insert
duodenum section stained with actin
following page.) (FITC), β-catenin (Cy3) and nucleic
acids
 Actine = FITC, green

 β-catenin = Cy3, orange
 Nucleic acids = PI, red
Bar = 20 µm


 Figure 21.11  At low magnification the cellular

morphology can directly be related to the
light microscopy images, the cell–cell
interphase in this case (see Figure 21.10).
 The same area at an intermediate

magnification.
 Step-by-step zooming in on the position

located by fluorescence microscopy the
labelled structures can be analysed at high
resolution.
 Inset: β-Catenin is predominately located

at the cell membrane between adjacent cells
(arrows).
Note that there is no β-catenin in the
desmosomes.
Top bar = 20 µm
Middle bar = 5 µm
Bottom bar = 1 µm


8. REFERENCES

Biophy., 21, 129, 1988.
microscopy of biological specimens based on rapid freezing with liquid helium II,
Ann. NY. Acad. Sci., 85, 689, 1960.
3. Gilkey, J.C. and Staehelin, L.A. Advances in ultrarapid freezing for the
preservation of cellular ultrastructure, J. Electr. Microsc. Tech., 3, 177, 1986.
Immuno-Gold Labelling in Cell Biology, Verkleij, A.J. and Leunissen, J.L.M., eds.,
5. Kaneko, Y. and Walther, P. Comparison of ultrastructure of germinating pea
leaves prepared by high-pressure freezing-freeze substitution and conventional
chemical fixation, J. Electr. Microsc., 44, 104, 1995.
Struct. Biol., 152, 92, 2005.
7. Müller, M., Marti, T., and Kriz, S. Improved structural preservation by freeze
substitution, in Proc. 7th Eur. Congr. Electron Microsc., Brederoo, P. and de
Priester, W., eds., The Hague, The Netherlands, 1980, 720.
8. Müller, M. and Moor, H. Cryo-fixation of thick specimens by high pressure
freezing, in Science of Biological Specimen Preparation 1983, Revel, J.P., Barnard,
T., and Haggis, G.H., eds., SEM Inc., AMF O'Hare, IL, USA, 1984, 131.
research: a preview, in Immunoelectron Microscopy in Virus Diagnosis and
1993, 97.
10. Steinbrecht, R.A. and Müller, M. Freeze-substitution and freeze-drying in
K., eds.. Springer-Verlag, Berlin, Heidelberg, Germany, 1987, 149.
11. Van Harreveld, A. and Crowell, J. Electron microscopy after rapid freezing on a
12. Brakenhoff, G.J. et al. 3-dimensional imaging of biological structures by high
resolution confocal scanning laser microscopy, Scan. Microsc., 2, 33 1988.
13. Petran, M. et al. Tandem scanning light microscopy, in Science of Biological
Specimen Preparation 1985, Müller, M., et al., eds., SEM Inc., AMF O'Hare, IL,
USA, 1986, 85.
14. White, J.G., Amos, W.B., and Fordham, M. An evaluation of confocal versus
conventional imaging of biological structures by fluorescence light microscopy, J.
Cell Biol., 105, 41, 1987.
15. Wilson, T. Scanning optical microscopy, Scanning, 7, 79, 1985.
16. DeBiasio, R. et al. Five-parameter fluorescence imaging: wound healing of living
Swiss 3T3 cells, J. Cell Biol., 105, 1613, 1987.
17. Tsien, R.Y. The green fluorescent protein, Annu. Rev. Biochem., 67, 509, 1998.
18. Gaietta, G. et al. Multicolor and electron microscopic imaging of connexin
trafficking, Science, 296, 503, 2002.

19. Schwarz, H. Immunolabelling of ultrathin resin sections for fluorescence and

electron microscopy, in Electron Microscopy 1994, ICEM 13, Jouffrey, B. and
Coliex, C., eds., Les Editions de Physique, Les Ulis, France, 1994, 255.
20. Schwarz, H. Correlative immunolabelling of ultrathin resin sections for light and
electron microscopy, in Electron Microscopy 1998, ICEM 14, Calderón Benavides,
H.A., et al., eds., Institute of Physics Publishing, Bristol, PA, USA, 1998, 865.
21. Tonning, A. et al. Hormonal regulation of mummy is needed for apical
extracellular matrix formation and epithelial morphogenesis in Drosophila,
Development, 133, 331, 2005.
22. Bierkamp, C. et al. Desmosomal localization of β-catenin in the skin of
plakoglobin null-mutant mice, Development, 126, 371, 1999.
23. Kurth, T. et al. Fine structural immunocytochemistry of catenins in amphibian and
mammalian muscle, Cell Tissue Res., 286, 1, 1996.
24. Nica, G. et al. Eya1 is required for lineage-specific differentiation, but not for cell
survival in the zebrafish adenohypophysis, Develop. Biol., 292, 189, 2006.
25. Schwarz, H., Müller-Schmid, A., and Hoffmann, W. Ultrastructural localization
of ependymins in the endomeninx of the brain of the rainbow trout: Possible
association with collagen fibrils of the extracellular matrix, Cell Tissue Res., 273,
417, 1993.
26. Sartori, A. et al. Correlative microscopy: Bridging the gap between fluorescence
light-microscopy and cryo-electron tomography, J. Struct. Biol., 160, 135, 2007.
27. Carlemalm, E., Garavito, R.M., and Villiger, W. Resin development for electron
microscopy and an analysis of embedding at low temperature, J. Microsc., 126, 123,
1982.
28. Tokuyasu, K.T. Immunocytochemistry on ultrathin frozen sections, Histochem. J.,
12, 381, 1980.
29. Van Donselaar, E. et al. Immunogold labelling of cryo-sections from high-pressure
30. Acetarin, J.D., Carlemalm, E., and Villiger, W. Developments of new Lowicryl
resins for embedding biological specimens at even lower temperatures, J. Microsc.,
143, 81, 1986.
Applications. 4th ed., Cambridge University Press, Cambridge, UK, 2000.
33. Birrell, G.B., Hedberg, K.K., and Griffith, O.H. Pitfalls of immunogold labelling:
Analysis by light microscopy, transmission electron microscopy, and photoelectron
microscopy, J. Histochem. Cytochem., 35, 843, 1987.
34. Venable, J.H. and Coggeshall, R. A simplified lead citrate stain for use in electron
microscopy, J. Cell Biol., 25, 407, 1965.


SDS-Digested Freeze-Fracture Replica Labeling (SDS-FRL) 569
CONTENTS

1.1. Fixation of Tissue (optional) .................................................................... 571
1.2. High-Pressure Freezing of Specimen ....................................................... 572
1.3. Freeze-Fracture Replication...................................................................... 573
1.4. SDS-Treatment of Replicated Materials................................................... 574
1.5. Immunodetection of Target Molecules with Gold Particles ..................... 574
3.1. Materials ................................................................................................... 576
3.2. Products .................................................................................................... 577
3.3. Solutions ................................................................................................... 577
4. PROTOCOL ...................................................................................................... 578
5.1. Advantages ............................................................................................... 581
5.2. Disadvantages........................................................................................... 582
6.1. Preparation of Tissue ................................................................................ 582
6.1.1. Fixation ....................................................................................... 582
6.1.2. Unfixed tissue ............................................................................. 583
6.2. Replication................................................................................................ 583
6.2.1. Fracturing and shadowing temperature....................................... 583
6.2.2. Shadowing .................................................................................. 583
6.2.3. Grid-mapping of the replicated tissue ......................................... 583
6.3. Tissue Removal ........................................................................................ 583
8. REFERENCES .................................................................................................. 586


One of the questions to be answered in the post-genomic era, at a time when most
proteins constituting living organisms have been identified, is what particular protein
species and amount of each species is expressed in a particular cell type and in their
subcellular domains. Such information is indispensable for understanding mechanisms
involved in living cells. New techniques for molecular localization that combine
immunofluorescent labeling or expression of fluorescent proteins with laser microscopy
have been developed and have successfully revealed new aspects of protein logistics and
its dynamic regulation in living organisms. However, for subcellular localization that
requires nano-scale resolution, conventional immunoelectron microscopy (pre- and post-
embedding immunolabeling) has been widely used. Although these techniques, especially
with immunogold particles, have provided precise details of molecular localization in
different subcellular domains, reliable quantification of immunoreactivity has often been
hampered by low sensitivity and limited accessibility of antibodies to target molecules
buried in tissues. In 1995, Fujimoto developed an epoch-making technique for
localization of plasma membrane molecules, termed sodium dodecyl sulfate (SDS)-
digested freeze-fracture replica labeling (SDS-FRL), by which molecules on the plasma
membrane can be directly approached by antibodies and visualized in a two-dimensional
manner with high sensitivity and high resolution.1,2 In collaboration with Fujimoto, we
have applied this technique to brain tissue with some modifications and proved its high
sensitivity and quantitative capability in visualizing molecules in the plasma membrane.3,4
Here, we introduce our current protocol of SDS-FRL, which could be further modified for
various purposes and other tissues to bring out the full potential of this powerful
technique.
1.1. Fixation of Tissue (Optional)

For fixation of brain tissue, animals are 
transcardially perfused with formaldehyde
(0.5 to 4%). Use of formaldehyde without
glutaraldehyde (GA) allows better removal
of tissue debris from replica membranes  Fixation of tissue with formaldehyde
and denaturation of membrane-attached (FA)
proteins by SDS treatment. Thereby the
molecules present in the plasma membrane
can be studied quantitatively.

Figure 22.1 SDS treatment resulted in

better dissociation of proteins in FA-fixed
brain tissue than in GA-fixed tissue.
A = Coomassie brilliant blue (CBB) stain.
B = Western blot immunodetection of
AMPA-type glutamate receptor subunit
GluR2.
 Lane 1: unfixed brain tissue
 Lane 2: 4% FA fixed tissue
 Lane 3: 4% FA plus 0.05% GA fixed
brain tissue
 Numbers indicate molecular weight in
kDa.
Note that a larger proportion of proteins and

GluR2 immuno-reactivity are detected at
high molecular weight ranges in GA-fixed
brain tissue.
It seems likely that fixation of tissue
improves reproducibility of labeling inten-
sities from one experiment to another.
1.2. High-Pressure Freezing of Specimen

For tissue freezing, chemically prefixed or 
unfixed brain slices between 90 to
150 micrometer thick are cut with a
vibrating microtome and are rapidly frozen  Consistent production of the same
with a high-pressure freezing machine (see quality of frozen tissues is necessary for
Chapters 5, 6), which produces a series of quantitative analysis.
uniformly frozen tissues in a consistent
manner. This is important for quantitative
analysis of protein localization.
Homogeneous freezing throughout the  Homogeneous freezing within the slice

whole slice is also important because is also necessary for quantitative analysis.
random fracturing through the slice occurs
when using a double replica specimen table.

1.3. Freeze-Fracture Replication

High-pressure frozen tissues are transferred 
into a freeze-fracture replica machine, and
the lipid bilayer of the plasma membranes
is split into two pieces at the hydrophobic
interfaces by fracturing. Proteins in the
plasma membrane are allocated in either of
two membranes, exoplasmic or proto-
plasmic membranes, and molecules in these
membranes, including lipids and proteins,
are then immobilized with a thin layer (< 5  Deposition of the first carbon layer onto
nm thick) of carbon deposit. This material exposed faces of plasma membrane halves
is further coated with a 2 nm-thick ensures immobilization of membrane
platinum/carbon layer for shadowing the proteins and phospholipids on the replica.
membrane faces and then strengthened with
a 15 nm-thick carbon deposit.




















Figure 22.2 The first carbon layer on the
exposed faces retains molecules at the
surface. The plasma membrane in frozen
tissue is frequently fractured at its
hydrophobic interface (B) and exposed
faces of the membrane half are coated with
evaporated carbon (gray) and platinum
(black) (C).



1.4. SDS-Treatment of Replicated Materials

After preparation of the replicas, they are  SDS treatment exposes molecules
treated with SDS, which dissolves lipids immobilized on replicas.
and proteins that are not immobilized in the 
replica membrane.
Figure 22.3 SDS treatment of replicated

tissue dissociates molecules that are not
immobilized in the replica membrane.
SDS treatment also denatures immobilized

molecules and unfolds secondary and
tertiary structures of the protein, and
thereby facilitates detection of epitopes in  SDS denaturation of molecules on
target molecules by immunolabeling. plasma membrane.
1.5. Immunodetection of Target Molecules with Gold Particles

SDS-treated replicas are subjected to
immunodetection where specific primary
antibodies for target molecules are bound
with immunogold-coupled secondary  Quantification of immunogold particles
antibodies, and then investigated with a under EM.
transmission electron microscope (TEM).
Quantification of immunoreactivity at a
nanometer scale resolution can be achieved.
Figure 22.4 Localization of target

molecules is visualized with immunogold
particles, which coincident with platinum/
carbon (Pt/C) replica images of membrane
faces.

1: Brain slice preparation

 Perfusion of animal with fixative
(optional)
A) Slicing
B) Trimming regions of interest
C) Cryoprotection in freezing medium
2: Freezing
 Mounting trimmed slices within two
specimen carriers with a spacer made
of double-sided adhesive tape (a)
 Freezing by high-pressure freezing
machine
 Storage of frozen sample in liquid
nitrogen
3: Freeze-fracture replication
 Equilibration of sample to target
temperature
 Fracturing of sample
 Replication of exposed surfaces with
C-Pt/C-C
4: Removal of tissue
A) Transferring replica with tissue into
SDS solution
B) Incubation at high temperature
5: Immunolabeling
 Removal of SDS
 Blocking
 Labeling
 Mounting the labeled replica onto
electron microscopy (EM) grid
6: LM observation of replicas
 Identification of tissue architecture in
replicas by light microscopy (LM)
(stereo or upright microscope)
7: EM observation of replicas
 Investigation of immunoreactivities.
Figure 22.5 Summary of the different

steps.

3.1. Materials
 Vibrating microtome  For preparation of slices

 High-pressure freezing machine  For rapid freezing of tissue
 HPM010, BAL-TEC, Balzers, Princi-
pality of Liechtenstein. Now available from
Boeckeler Instruments, Inc., Tucson,
Arizona, USA or ABRA-Fluid AG,
Widnau, Switzerland: www.abra-fluid.ch
 Specimen carriers with or without a  BAL-TEC, 4.6 mm diameter, 0.6 mm

ring-shaped, double-sided adhesive thickness, LZ 02127 VN. Now available
tape (1.5 mm diameter hole, 150 µm from Leica.
thickness or less)
 Cell locator  For storage of frozen samples
 Freeze-fracture replica machine  For freeze-fracture and replication of

specimens
 BAL-TEC, BAF060 Now available from
Leica.
 Double replica specimen table  BAL-TEC. Now available from Leica.
 Stereo microscope  For trimming of slices, placing trimmed

tissues into freezing carriers, and immuno-
labeling of replicas
 Hybridization oven or autoclave  For SDS treatment of replicated samples
 Sealable glass vial (5 mL)
 A glass rod with a platinum wire or a  For replica handling

glass rod with a fine tip
 Shaker with flat bottom  For immunolabeling
 Porcelain spot dish
 EM specimen grids coated with  For observation under the EM

Pioloform film

3.2. Products
 Liquid nitrogen  For storage and handling of frozen
samples and for operation of HPM010 and
BAF060
 Primary antibodies  For immunolabeling
 Secondary antibodies coupled to 5, 10  British Biocell International, Cardiff and

or 15 nm colloidal gold Amersham, Biosciences, Buckinghamshire,
U.K.
3.3. Solutions
 Fixation
 Saline or phosphate-buffered saline  Higher concentrations of formaldehyde
(PBS) generally result in fracturing of plasma
 0.5 to 4% formaldehyde (FA), 15% membranes into smaller pieces, making
saturated picric acid (PA) solution in interpretation of images difficult.
0.1 M PB.
 Freezing medium
 30% glycerol or 30% sucrose in  Stepwise treatment of tissues with 10,
0.1 M PB (pH 7.4) 20, 30% glycerol or sucrose solutions may
work better in terms of preservation of
morphology.
 SDS solution
 2.5% SDS, 20% sucrose in 15 mM  Solution at pH 6.8 can also be used.
Tris-HCl (pH 8.3)  Sucrose can be replaced with 10%
glycerol.
 Buffers for immunolabeling

 Washing buffer
 0.05% BSA  Addition of BSA in buffers dramatically
 0.1% Tween-20 reduces stickiness of the replica to
 0.05% NaN3 glassware and platinum wire and facilitates
in 50 mM Tris buffered saline the isolation of intact (large) replicas after
(TBS, pH 7.6) immunolabeling.
 Blocking buffer
 5% BSA
 0.1% Tween-20
 0.05% NaN3
in TBS
 Primary antibody solution
 1% BSA
 0.1% Tween-20
 0.05% NaN3
in TBS

4. PROTOCOL

Standard protocol for SDS-FRL
 Dissected tissues must be kept cold
(4 °C) throughout Step 1.
1. Preparation of tissue slices  Glass rods or paint brushes can be used
 After brief cardiac perfusion with to handle the tissue slices.
saline, animals are fixed with 0.1 M PB  Unfixed slices can also be used.
containing 2% FA and 15% saturated  Preparation of fixative solution and
PA, and the brain is sliced at 150 m perfusion of animals must be done under a
thickness with a vibrating microtome chemical hood to avoid inhalation of
in 0.1 M PB. formaldehyde.
 FA ranging in concentration between 0.5
to 4% is used.
 Thickness of the slice should be
optimized by each user so that it fits into
the space between specimen carriers (see
below).
 Use of stereo microscope is recom-
mended for the following steps.
 Trimming of slice  Region of interest in the slice is trimmed

out with a razor blade to fit into the hole of
the specimen carrier. To prevent tissue
damage, trimming is carried out on a 1%
agarose gel in 0.1 M PB.
 Cryoprotection  The trimmed slices are immersed into

freezing medium for at least three hours.
 The order of trimming and cryo-
protection is reversible.
2. Rapid freezing of the trimmed slices 

 Mounting trimmed slices onto  In the freezing media, the slice is placed
specimen carriers into a hole of double-sided tape attached to
a specimen carrier.
The carrier with the slice is transferred onto
filter paper and the excess solution in the
hole is removed with a brush.
The carrier is covered with another
specimen carrier without the tape so that
the slice is sandwiched between the two
carriers.
Figure 22.6 Tissue slice mounted on the
specimen carrier.

 Freezing and storage of specimens  The room must be ventilated to avoid

oxygen shortage. Wear cryo-protection
gloves when handling chilled equipment.
 The carrier pair is rapidly frozen with
the high-pressure freezing machine and
stored in liquid nitrogen until use.
3. Freeze-fracture replication
 Installation of frozen samples into a  The frozen sample inserted into a slot of
replica machine. a double replica specimen table in liquid
nitrogen is transferred into the freeze-
fracture machine.
 Freeze-fracturing of samples  Fracture temperature at 110 to 120°C
is recommended for ease of interpretation
of the images as fracturing at lower
temperature results in smaller membrane
fragments in the replicas.
 Replication of exposed tissue surface  Indispensable. To have homogeneous
immobilization of surface molecules, a thin
layer of carbon coating as the first layer is
recommended. Conditions for each layer
are as follows:
 First layer, 5 nm carbon from
overhead with sample rotation at 30
rpm
 Second layer, 2 nm platinum/carbon at
60° angle without sample rotation
 Third layer, 15 nm carbon from
overhead with sample rotation
 Grid-mapping of the replicated tissue  Optional. For better orientation, the
replicas are attached to EM finder grids
before tissue removal and nuclear stain
images on the EM grid are recorded by
light microscopy.
To keep replicas in their original shape
throughout the labeling procedures, they
are stabilized with Lexan resin (see Section
6.2.3).
4. Tissue removal from replica mem-  Replicas with tissues are transferred into
brane by SDS treatment glass vials containing 0.75 mL of the SDS
solution.
 These samples are incubated at +80°C
for 18 hours with continuous shaking.

5. Immunolabeling  All of the following steps are carried out

at room temperature unless otherwise
noted.
 All of the following steps are carried out
under mild shaking conditions.
 Wash
 SDS solution 10 min  Optional.
 Wash buffer 3 × 10 min
 Blocking non specific sites  Indispensable !

 Blocking buffer 1 hour
 Primary antibody reaction 

 Primary antibody Overnight  30 L drops of the solution are prepared
solution at 15°C on a Parafilm sheet and placed in a humid
chamber. Blocked replicas are transferred
onto these drops. The primary antibody
concentration is between 1 and 5 g/mL in
the primary antibody solution.
 Wash
 Blocking non specific sites  For reduction of background labeling,

 Blocking buffer 30 min but not indispensable.
 Secondary antibody reaction  The secondary antibody is diluted at

 Secondary antibody Overnight 0.5 ~ 1 g/mL in the blocking buffer. For
in blocking buffer multiple labeling, several secondary
antibodies coupled with different sizes of
colloidal gold are mixed. In general, those
with smaller gold show higher labeling
intensity and those with > 20 nm gold give
very low labeling intensity.
 Wash
 Wash
 Double distilled 1 × 3 min  In solution without BSA, the tissue side
water of replicas tends to stick onto glassware
and platinum wire leading to damaged
replicas. Handle them on the nontissue
side.

 Mounting replica membranes onto the  Replicas should be kept on the water
EM grids surface and be unfolded with platinum wire
under a stereo microscope.
 At this step it is important to have large,
intact replica membranes in which
information about tissue architecture can be
retrieved.
 Practice is required in handling replicas
in solution.
6. LM observation
 Stereomicroscopy  Stereomicroscopic images of replicas
with transparent illumination and/or
reflecting illumination are useful to obtain
information about tissue architecture in the
replicas.
7. EM observation  For investigation of immunoreactivity

and morphology of plasma membrane
faces.
 For interpretation of images, please refer
to original papers listed in the reference
section.
5.1. Advantages
 Two-dimensional visualization of  Instead of reconstructing serial ultrathin
protein localization in the plasma sections, SDS-FRL can directly visualize
membrane at EM level two-dimensional distribution of membrane
molecules. Localization of the immunogold
particles coincident with Pt/C replica
images of membrane faces are obtained by
TEM.

 Quantitative analysis of protein  SDS treatment dissociates proteins from
localization target proteins in the protein complex and
denatures their higher-order structure,
resulting in an equal detectability by
specific antibodies. Visualization of
immunoreactivity with antibodies coupled
to colloidal gold particles enables
quantitative analysis of protein localization.

 Highly sensitive detection and clear-  In the conventional pre-embedding

cut distinction of the pre- or post- immunolabeling method, molecules in
synaptically localized molecules synaptic sites are often undetectable
because of poor penetration of antibodies
into the densely packed protein matrix of
the specialized synaptic membranes. On the
other hand, in the post-embedding
immunolabeling method, the deviation of
immunogold particles (up to 30 nm) from
the antigen site makes it difficult to
distinguish pre- and post-synaptic location.

 Wide applicability of specific  SDS-FRL uses similar antigen–antibody
antibodies reaction conditions to those used for the
Western blot immunodetection. Therefore,
a large proportion of antibodies usable in
the latter method may also be applicable to
this technique.
5.2. Disadvantages
 Difficult to identify cellular or sub-  Information about fine structure in the
cellular origin of each replicated vicinity of replicated membranes is lost.
membrane
 Difficult to produce replicas com-  Fracture occurs in a random fashion

prising specific cellular components within the frozen tissues.
within a frozen tissue.


6.1. Preparation of Tissue

6.1.1. Fixation
 0.5% FA + 15% saturated PA in 0.1 M  For more synaptic profiles in the replica
PB membrane

 4.0% FA + 15% saturated PA in 0.1 M  For finer structure of the plasma
PB membrane and intramembrane particles
(IMPs)

6.1.2. Unfixed tissue

 Artificial cerebrospinal fluid in which  For some target molecules that show low
sodium chloride is substituted with immunoreactivity in the fixed tissue.
0.8% sucrose.



6.2. Replication
6.2.1. Fracturing and shadowing temperature
 Fracture and replication at 130°C or  For finer structure of the plasma
below. membrane and IMPs.

6.2.2. Shadowing
 Rotary shadow for the second layer,  For finer structure of the IMPs.
2.5 nm platinum/carbon from 25° with 
sample rotation. 
 
6.2.3. Grid-mapping of the replicated tissue

 Attach an EM grid to an entire replica  To prevent replica membranes from
membrane with Lexan resin before breaking into small pieces during tissue
tissue removal. removal and immunolabeling procedures.
 For detailed mapping of the cyto-

architecture within replicas. Detailed
procedure is described in References 3,5.
6.3. Tissue Removal

 Incubations
 105°C for 10 min by autoclave  When the hybridization oven is not
available.

 60°C for one day and 37°C overnight  For detection of membrane-associated
proteins.3

7. OBSERVED RESULTS
 Figure on Chapter’s title page  See Figure 22.7.
 Figure 22.7
1. Description  Co-localization of AMPA- and NMDA-

type glutamate receptors in the postsynaptic
specializations in the plasma membrane of
the dentate gyrus granule cell. The post-
synaptic specialization was identified by a
cluster of IMPs on the exoplasmic fracture
face.
2. Method  SDS-FRL
3. Tissue  Adult rat hippocampus
4. Fixation  2% FA, 15% saturated PA in 0.1 M PB
5. Fracture temperature  120°C
6. Coating layers  5 nm C, 2 nm Pt/C from 60°, 15 nm C
7. SDS treatment  80°C for 18 hours
8. Antibodies  Primary antibodies:

 Rabbit anti-GluR1-4 antibody was
kindly donated by Prof. E. Molnar at
University of Bristol U.K. (3.3 g/L).
 Mouse anti-NMDA receptor 1
antibody (Chemicon, 3.6 g/L).
 Secondary antibodies:
 Antirabbit and antimouse IgG coupled
with 5 and 10 nm colloidal gold
particles, respectively (0.67 g/mL)
9. Bar  200 nm


8. REFERENCES
1. Fujimoto, K. Freeze-fracture replica electron microscopy combined with SDS

digestion for cytochemical labeling of integral membrane proteins. J. Cell Science,
108, 3443, 1995.
2. Fujimoto, K. SDS-digested freeze-fracture replica labeling electron microscopy to
study the two-dimensional distributioin of integral membrane proteins and
phospholipids in biomembranes: Practical procedure, interpretation and application.
Histochem. Cell Biol., 107, 87, 1997.
3. Hagiwara, A. et al. Differential distribution of release-related proteins in the
hippocampal CA3 area as revealed by freeze-fracture replica labeling. J. Comp.
Neurol., 489, 195, 2005.
4. Tanaka, J. et al. Number and density of AMPA receptors in single synapses in
immature cerebellum. J. Neurosci., 25, 799, 2005.
5. Rash, J.E. et al. Grid-mapped freeze fracture: Correlative confocal laser scanning
microscopy and freeze-fracture electron microscopy of preselected cells in tissue
slices, in Rapid Freezing, Freeze Fracture, and Deep Etching, Severs, N.J. and
Shotton, D.M., eds., Wiley-Liss, New York, 1995, 127.

Immunolabeling of Ultra-thin Sections with Enlarged 1 nm Gold or QDots 589
CONTENTS

1.1. Small Markers .......................................................................................... 592
1.1.1. Small colloidal gold markers ...................................................... 592
1.1.2. NANOGOLD markers ................................................................ 593
1.1.3. Quantum dot (QD) markers ........................................................ 593
1.2. Silver Enhancement.................................................................................. 594
1.3. Gold Enhancement ................................................................................... 595
3.1. Materials ................................................................................................... 597
3.2. Products .................................................................................................... 598
3.3. Solutions ................................................................................................... 599
4. PROTOCOL ...................................................................................................... 602
4.1. Labeling with Small Markers ................................................................... 602
4.2. Appendix 1: Repeated Enhancement/Enhancement after Staining........... 605
4.3. Appendix 2: Double Labeling .................................................................. 606
4.4. Appendix 3: Inactivation and Blocking Solutions .................................... 606
5.1. Advantages of Small Markers .................................................................. 608
5.2. Disadvantages of Small Markers .............................................................. 608
5.3. Advantages of Enhancement Procedures.................................................. 609
5.4. Disadvantages of Enhancement Procedures ............................................. 609
6.1. Small Gold and QD Markers .................................................................... 610
6.1.1. NANOGOLD.............................................................................. 611
6.1.2. Ultra-small colloidal gold ........................................................... 611
6.1.3. Small QDs................................................................................... 611
6.2. Silver Enhancement Techniques............................................................... 611
6.2.1. HQ SILVER................................................................................ 611
6.2.2. R-GENT SE-EM......................................................................... 612
6.2.3. Danscher solution ....................................................................... 612

6.3. Gold Enhancement Techniques................................................................ 612

6.3.1. GoldEnhance-EM ....................................................................... 612
6.3.2. Published recipes ........................................................................ 612
6.4. Silver Stabilization/Gold Toning.............................................................. 612
8. REFERENCES .................................................................................................. 615


Very small (1 nm) gold markers, such as ultra-small (colloidal) gold (Aurion) and the
gold compound NANOGOLD™ (Nanoprobes), have become more important in electron
microscopic immunocytochemistry due to considerably improved sensitivity (see Figure
on Chapter’s title page).1,11,22 The main reasons for improved sensitivity are less steric
hindrance and reduced electrostatic repulsion (the colloidal gold surface is negatively
charged). These markers penetrate into permeabilized cells and tissues more readily than
larger gold colloids and, in ultrathin resin and cryo-section labeling experiments, the
density of the label is often considerably higher when compared to colloid sizes of 4 nm
or larger. In contrast to permeabilized samples, ultrathin resin sections cannot be
penetrated by gold markers whatever the marker size used20 (see Figure 23.1). Also in the
case of well preserved thawed cryo-sections obtained according to Tokuyasu,25 gold
labeling is restricted mainly to the section’s surface, but (local) low matrix density, either
due to inherent specimen properties or section damage, may allow some penetration,
especially of 1 nm gold markers24 (see Figure 23.2).
Due to their small size, the electron density of 1 nm gold particles is low and their
visualization in a conventional transmission electron microscope (TEM) is difficult.22
This also holds true for small semiconductor nanocrystals, so-called quantum dot (QDs,
e.g., QDOT® (Invitrogen) markers that exhibit low intrinsic contrast.6 Therefore, these
markers have to be enlarged, either by deposition of metallic silver (so-called silver
enhancement1,4,5,21,22) or gold (so-called gold enhancement26). Both enhancement
techniques can be a source of problems, e.g., due to the low stability of the silver
layer.1,18,21,22
The aim of this chapter is to help to choose the most useful small marker and
enhancement method for a specific application and to cope with the typical practical
problems associated with the use of very small markers and the application of
enhancement techniques.
Figure 23.1 Cross-section through a gold

labeled ultrathin resin section. Due to the
high matrix density (polymerized resin,
hatched area) luminal (white stars) and
cytoplasmic (black) epitopes of vacuolar
membrane antigens are only accessible for
immunoreagents at the resin section surface
(spheres, gold marker). Antibodies are not
visualized.

Figure 23.2 Cross-sections through gold-

labeled, ultrathin Tokuyasu cryo-sections.
 (a) In the case of well-preserved
sections, gold markers are
preferentially bound to the cryo-
section surface. Section damage
(arrow) and local low matrix density in
the vacuolar lumen allow some
penetration of immunoreagents into
the thawed cryo-section.
 (b) In the case of weakly fixed and,
therefore, extracted sections, anti-
bodies and especially 1 nm gold
markers can permeate the thawed
cryo-section to a certain extent due to
its low matrix density.
1.1. Small Markers

The preparation of 1 nm gold markers and  For example, unconjugated 1 nm gold
their conjugation to ligands such as colloids are less stable than larger ones.
antibodies, are considerably more difficult  Several recipes (for review see
when compared to larger ones.1 Therefore, Reference 1).
commercial products are preferred. 
1.1.1. Small colloidal gold markers

The smallest gold colloids in use have sizes  Few commercial suppliers (Aurion
between 0.8 nm and ~ 3 nm.1,14,17,21-23 The (Ultra-small); British Biocell International
colloid surface is hydrophobic and (2 nm gold); Amersham/GE Healthcare
negatively charged (due to adsorbed ions). (AuroProbe One).
The actual particle size is larger due to an  Relatively large size distribution, e.g.,
additional hydration shell. Antibodies or ultra-small gold (Aurion).
antibody fragments are noncovalently (but  Gold colloids have a rigid and thick coat
tightly) bound via hydrophobic and of water dipoles due to their negative
electrostatic interactions. surface charge.

Figure 23.3 Fab (antibody) fragment non-
covalently bound to 1 nm colloidal gold
(black). The Fab fragment is 5 to 6 nm in
length. Antigen-binding site (bright ends),
IgGs may have more than one colloid disulphide bridge (white rectangle) (drawn
bound. to scale).

1.1.2. NANOGOLD markers

NANOGOLD (Nanoprobes) is a gold  One commercial supplier.
compound with a gold core diameter of 1 to
1.4 nm and including the organic shell,  There are also double labeled
2.7 nm.10,11 Antibodies or antibody NANOGOLD-antibody (fragment) conju-
fragments are covalently linked to the gates, which are additionally linked to
organic shell via a hinge region thiol (one fluorochromes like fluorescein or Alexa
gold particle per Fab fragment or IgG). dyes 488 and 594.
 
Figure 23.4 Fab fragment covalently linked

to the organic shell (grey) of NANOGOLD
(gold core, black). Antigen-binding site
(bright ends), disulphide bridges (white
rectangles) (drawn to scale).
Undecagold (Nanoprobes) is even smaller

(gold core 0.8 nm, organic shell 2 nm in  Undecagold can be used to label-specific
diameter),11 but cannot be satisfactorily sites on a protein for single particle
enhanced by silver deposition. analysis.
1.1.3. Quantum dot (QD) markers

QDs are fluorescent semiconductor  Several commercial suppliers of
nanocrystals made of substances, such as antibody-conjugated nanocrystals, e.g.,
cadmium selenide as the core, which is QDOTs (Molecular Probes/Invitrogen), or
surrounded by a core shell. QDs offer EviTags (Evident Technologies/Antibodies
unique optical properties like high quantum Inc.).
yield, resistance to photobleaching and  Important for immunofluorescence
size-tunable emission spectra.6 Core and microscopy.
core shell exhibit moderate electron  The relatively small, green fluorescent
density6 (see Figure 23.9 a-c). QDOT 525 (Molecular Probes) has an
electron dense spherical core/core shell of
about 3 to 5 nm in diameter. The whole
particle is 11 to 14 nm in diameter
including the organic layer and bound
(Fab`)2 fragments (see Figure 23.9 a-c).

Figure 23.5 Model of QDOT 525
conjugate: The core (black) is surrounded
by a core shell (grey) and an organic layer
(black ring), which is covalently linked to
several (Fab`)2 fragments. Antigen-binding
sites (bright ends), disulphide bridges
(white rectangles) (drawn to scale).


 There are QDs conjugated to secondary

antibodies, which may be smaller (e.g.,
EviTag Adirondack Green), but it is not
known whether they can be silver-enhanced
or not.
1.2. Silver Enhancement

Gold colloids and organo-gold clusters as  Silver enhancement for immunogold
well as QDs can act as nuclei, comparable electron microscopy was introduced in
to silver nuclei, in photographic 1983.15
development. In the presence of a reducing
agent, the gold surface acts as a catalyst for  The enhancer solution contains a silver
the reduction of silver ions to metallic salt, a reducing agent, a buffer and often the
silver. The metallic silver deposited on the so-called protective colloid gum arabic,
gold surface itself serves as nuclei for which improves the enhancer’s
further reduction of silver ions to metallic performance: it makes the reaction more
silver, resulting in a growing silver layer.3,4 efficient and reproducible.1,3,4,21-23

Figure 23.6 Principle of silver

enhancement. The gold surface serves as a
catalyst for the reduction of silver ions, e.g.,
by hydroquinone.
Whereas silver enhancement of larger gold

particles is relatively unproblematic, it
turned out that enhancement of 1 nm gold is
associated with a number of problems, such
as inefficient enhancement, uneven growth  As a consequence, not all enhancers are
and increased auto-nucleation due to suitable for 1 nm gold markers and
prolonged enhancement time.1,21-23 QDOTs.1,21-23
A general drawback of silver enhancement
is the instability of the silver layer caused  OsO4 dissolves the silver layer; electron
by oxidation, which requires additional beam and air humidity cause dis-
treatment or special storage conditions for location/loss of silver (see Figure 23.14).
protection.1,18

We will describe the use of two commercial  Nickel grids should be used because
silver enhancers and one published method: nickel does not interfere with silver
enhancement, in contrast to copper.
 HQ SILVER (Nanoprobes) contains  See Figure 23.10a, b.

the protective colloid gum arabic. Its
pH is near-neutral, the viscosity is
high.
 R-GENT SE-EM (Aurion) is a low  See Figure 23.10c.

viscosity enhancer with near-neutral
pH.
 The Danscher method3,4,23 is an  See Figure 23.11.

example of a published recipe of high  Two other published recipes deserve
efficiency, which is suited for all small mentioning: The first one,16 a fast-working
markers mentioned. The enhancer enhancer, can be used for copper grids. The
contains the protective colloid gum second one7 causes less structural damage
arabic. The pH is low, the viscosity on whole mount samples or Tokuyasu cryo-
high. sections than the Danscher method,
probably due to a near-neutral pH.
1.3. Gold Enhancement

In principle, the deposition of gold instead  A process similar to silver enhancement.
of silver offers potential advantages: it  Higher particle density (improved
confers stability to the enhanced particle, contrast).
thereby making it resistant to oxida-  Improved backscatter detection in the
tion.1,18,26 scanning electron microscope (SEM).
 Caused by OsO4 treatment, beam
damage, and air humidity.

GoldEnhance™-EM (Nanoprobes) is the  In our lab, gold markers bound to resin
only commercial product available (Near sections can be readily enhanced (see
neutral pH, low viscosity, less sensitive to Figure 23.12), but treatment of labeled
buffer salts).26 cryo-sections results in strong background.

Gold chloride treatment has been described  In our lab, gold enhancement following
as being able to enlarge small colloidal published procedures leads to loss of gold
gold markers (for review see References 1 rather than to gold particle growth1,18 (see
and 18). Figure 23.13).


Ultra(cryo)microtomy and mounting of

sections:
 Mounting ultrathin cryo-sections
(swimming on the surface of the pick-
up drop; see Chapter 19) or resin
sections on plastic- and carbon-
coated13 grids (Figure 23.7, top and
middle).
 Transfer of grids (upside down) to
drops of buffer placed on a
hydrophobic film in a dish covered by
a glass plate (moist chamber) (Figure
23.7, middle and bottom).
 Optional: Cryo- and resin sections
mounted on grids can be stored prior
to immunolabeling.8
Figure 23.7 Mounting of sections on grids

(top, middle) and transfer of grids to drops
of buffer (middle, bottom) for immuno-
labeling.
Figure 23.8 Labeling of grids in moist

chamber. Grids can be transferred with a
forceps or a loop (see also Chapter 21).
1: Inactivation of residual fixative

molecules (mainly aldehyde groups)
(optional)
2: Blocking unspecific protein binding
sites
3, 4: Antibody labeling
5, 6: Marker incubation
7, 8: Fixation
10, 11: Silver or gold enhancement
12, 13: Stabilization of silver layer
(optional) (gold toning)
14: Staining (heavy metals)
15, 16: Drying (resin sections) or methyl
cellulose embedding (cryo-sections)

3.1. Materials
 Vortex  For antibodies, markers, and enhancers;
e.g., Vortex-Genie 2, Scientific Industries
Inc, Bohemia, New York., USA.
 Centrifuge  For all solutions used; e.g., Centrifuge
#5415D, Eppendorf-Netheler-Hinz GmbH,
Hamburg, Germany.
 Forceps (antimagnetic)  Grid transfer (nickel grids!); e.g., # 5/45
or #7, Dumont & Fils, Switzerland.
 Petri dish (glass)  For lead staining.
 Dish (stainless steel) with glass plate  Moist chamber for labeling and
enhancement.
 Hydrophobic film  For incubating grids on drops; Parafilm
“M”, Pechiney Plastic Packaging, Chicago,
USA.
 Loops (stainless steel, 3 mm in  Self-made loops from platinum wire for
diameter) attached, e.g., to pipette tips grid transfer and for drying cryo-sections.
 Tissue paper/wipes  For generating humid air; e.g.,
Kleenex®, Kimberly-Clark Corp, USA.
 Pioloform or Formvar coated and  For supporting sections (100 to 200
carbon-covered nickel grids for silver mesh nickel grids), e.g., Stork Veco B.V,
enhancement, copper or nickel grids Eerbeek, The Netherlands; or Gilder Grids,
for gold enhancement Grantham, England.
 For plastic and carbon covering of
grids, see Reference 13.
 Pipettes (glass or plastic)  For placing drops.
 Pipette tips  For dilution, placing drops, etc.
Eppendorf-Netheler-Hinz GmbH, Ham-
burg, Germany.
 Desiccator (made of glass or plastic)  For storing silver-enhanced sections in
dry air, nitrogen gas, or vacuum.
 Silica gel-desiccant  For dry storage of silver-enhanced
sections.
 Black cardboard  For dark chamber for light sensitive
reactions (e.g., silver enhancement).
 Filter paper  For drying grids, usually Whatman 541,
Whatman International Ltd, Maidstone,
England.
 Water bottle (containing double  For jet wash of grids.
distilled water)
 H2O  Water, double distilled (on quartz) or
Milli Q.
 Microtubes (0.5 mL, 1.5 mL, 2 mL)  For mixing solutions, for storing frozen
aliquots; Eppendorf-Netheler-Hinz GmbH,
Hamburg, Germany or Sarstedt AG & Co,
Nümbrecht, Germany.

3.2. Products
 Ammonium hydroxide (NH4OH)  For example, Merck Eurolab, VWR
International; Darmstadt, Germany.
 Bovine serum albumin (BSA)  Sigma-Aldrich, Fluka, Buchs,
fraction V Switzerland.
 Bovine serum albumin, acetylated  Aurion, Wageningen, The Netherlands.
(BSA-c)
 Citric acid (C6H8O7 x H2O)  Merck.
 Cold water fish gelatin  Aurion, Sigma.
 Disodium hydrogenphosphate  Merck.
(Na2HPO4)
 Ethanol  Merck.
 Fetal / new born calf serum  Gibco, Invitrogen, Eugene, Oregon,
USA.
 Gelatin  Merck.
 Glutaraldehyde (aqueous, 25%)  Sigma #G-5882 (keep frozen).
 Glycine  Merck.
 Gold chloride  See tetrachloroauric acid trihydrate.
 Gum arabic  Merck.
 HEPES  Merck.
 Hydroquinone  Merck.
 L-(+)-lactic acid  Sigma #L-7771 (Caution: light
sensitive).
 Lead (II) citrate trihydrate  Fluka, Buchs, Switzerland; or Agar
(Pb3(C6H5O7)2 x 3H2O) Scientific.
 MES (C6H13NO4S)  Serva 29834, Heidelberg, Germany.
 Methyl cellulose  Sigma M-6385.
 Milk powder (non fat, dried)  Regular non fat milk powder.
 NANOGOLD™ conjugates  Nanoprobes (Stony Brook,New York,
USA).
 Ultra-small gold and NANOGOLD
conjugates can be stored at 20°C after
mixing equal parts of marker and 99%
glycerol.
 Nonimmune serum / normal serum  Invitrogen.
 N-propyl-gallate  Merck.
 Ovalbumin (albumin from egg white)  Sigma.
 Oxalic acid (C2H2O4 x 2H2O)  Merck.
 Potassium chloride (KCl)  Merck.
 Potassium dihydrogen phosphate  Merck.
KH2 PO4
 Quantum dot (QDOT®) conjugates  Molecular Probes, Invitrogen, Eugene,
Oregon, USA.
 Sodium borohydride (NaBH4)  Merck.
 Sodium chloride (NaCl)  Merck.
 Sodium hydroxide (NaOH) pellets  Merck.

 Sodium thiosulfate (Na2O3S2)  Merck.

 Tetrachloroauric(III)acid trihydrate  Merck #101582.
(AuCl4H x 3H2O)
 Trisodium citrate dehydrate  Merck.
(C6H5Na3O7 x 2H2O)
 Tween-20®  Sigma, Serva.
 Ultra-small gold conjugates  Aurion.
 Uranyl acetate  Stain for electron microscopy. EMS,
(UO2(CH3COO)2 x 2H2O) Fort Washington, Pennsylvania, USA or
USA.
 Uranyl oxalate  See Staining solutions.

3.3. Solutions
1. R-GENT SE-EM Activator, initiator,  Aurion, Wageningen, The Netherlands.
enhancer  Shelf life ~ 18 months from date of
production.
 Preparation of developer:  Light insensitive.
 Put 40 drops of activator (1385 µL)
into the developer bottle  The concentrated initiator can be stored
 Add 1 (36 µL) drop initiator to the in a freezer.
developer bottle  Shelf life of developer is about one
 Mix well on a vortex month.
 Use nickel grids.
 Preparation of final enhancer:
 Mix 20 drops (870 µL) of enhancer
with 1 drop (34 µL) of developer
 Mix well on a vortex
2. HQ SILVER  Nanoprobes, Stony Brook, New York,
 Moderator, activator, initiator USA
 Equal amounts of moderator, activator,  Solutions should be kept frozen in small
and then initiator are thoroughly mixed aliquots.
before use  Caution: Light sensitive.
 Shelf life ~ 30 months.
 Use nickel grids.
3.Danscher (acidic) silver enhancer3,23
 Stock solutions:
 Gum arabic: 33% (w/v) in H2O  Stir for one day, centrifuge (30,000 g,
 Hydroquinone: 0.57 g in 10 mL H2O 2 h, 4°C).
(5.7%, 0.52 M)
 Citrate buffer: 2.55 g citric acid  Gum arabic, citrate buffer, and
(23.5%, 1.5 M) + 2.35 g trisodium hydroquinone can be premixed and stored
citrate dehydrate (23.5%, 0.5 M), add in a freezer. Silver lactate has to be stored
H2O to make 10 mL (pH 3.53.8) separately in a freezer.
 Silver lactate: 7.3 g in 10 mL H2O  Caution: Light sensitive!
(0.73 %, 37 mM)  Caution: Silver lactate salt may age with
time and change properties. If in doubt, buy
a new package.

 Preparation of final enhancer:

 Mix 6 parts gum arabic stock with 1
part citrate buffer stock and 1.5 parts  Use nickel grids.
hydroquinone stock. Add 1.5 parts  Caution: Moderate light sensitivity after
silver lactate stock and mix mixing.
thoroughly.
4. Fast working silver enhancer16  Suitable for copper grids.

 See above, replace citrate buffer by  Caution: Light sensitive!
0.2 M HEPES buffer, pH to 6.8 with
NaOH.
5. Neutral silver enhancer7  Reduced ultrastructure damage

 Stock solutions: (probably due to near-neutral pH).
 0.5 M MES adjusted to pH 6.15 with
NaOH
 Gum arabic: 33% (w/v) in H2O
 N-propyl gallate: dissolve 10 mg NPG
in 0.25 mL ethanol, then add 4.75 mL
of H2O
 Silver lactate: 36 mg in 5 mL H2O
 Preparation of final enhancer:  Caution: Light sensitive!

 2 parts MES stock, 5 parts gum arabic
stock, 1.5 parts N-propyl gallate stock,  Use nickel grids.
and 1.5 parts silver lactate stock.
 The first three components are mixed
30 min before use, the N-propyl  Caution: Light sensitive (dark room)!
gallate stock is added 2 min before
use.
6. GoldEnhance™-EM  Nanoprobes, Stony Brook, New York,

 Mix equal amounts of enhancer, USA.
activator, initiator, and buffer.  Shelf life ~ 36 months.
 Mix enhancer and activator, wait 5 to  Use copper or nickel grids.
10 min, then add initiator and buffer.  For numerous modifications, see
www.nanoprobes.com.
7.PBS buffer (pH 7.4)  Standard buffer.

 137 mM sodium chloride (NaCl)
 2.7 mM potassium chloride (KCl)
 8.1 mM disodium hydrogen-phosphate
(Na2HPO4)
 1.5 mM potassium dihydrogen
phosphate (KH2 PO4)

8. Staining solutions  Stains for resin and cryo-sections.

 Uranyl oxalate or neutral uranyl  Stir a 0.3 M oxalic acid solution
acetate:25 vigorously and add slowly the same volume
 2% (w/v) uranyl acetate in 0.15 M of a 4% (w/v) aqueous uranyl acetate
oxalic acid, pH 7.4 solution. Adjust pH to 7.4 with 10%
ammonium hydroxide.25
 Store in refrigerator.
 4% (w/v) aqueous uranyl acetate  Store in refrigerator.

 1% (w/v) aqueous uranyl acetate  Store in refrigerator.
9. Methyl cellulose  Embedding of cryo-sections.

 2% (w/v) methyl cellulose in H2O  Centrifuge at 100,000 g at 4°C.
(25 centipoises).  Store in refrigerator.
 Methylcellulose–uranyl acetate (UA)  See also Chapter 19.
solution:  Stronger contrast and darker structures
 3% UA in H2O with higher UA concentration. Too high
 2% aqueous methyl cellulose UA concentration causes membrane
 Mix 1 (1.5) part UA with 9 (8.5) parts blebbing.25
methyl cellulose.
10. Lead citrate staining  Stain for resin sections.
 Recipe  See Chapter 21.
11. Gold chloride  See gold toning.
12. Gold toning  Protection of silver layer.

 0.05% aqueous gold chloride (diluted  Store in darkness (at 4°C).
from 1 to 2% stock solution of
tetrachloroauric acid trihydrate)
13. Fixative  Stabilization of antigen-antibody-marker

 0.5 to 1% aqueous glutaraldehyde complex prior to enhancement.
14. Blocking buffer  See Appendix 3 for alternative reagents.

 0.5% BSA plus 0.5% nonfat milk
powder in PBS
15. Glycine-PBS  Inactivation of residual reactive

 50 mM glycine in phosphate buffer aldehyde groups.
saline (PBS)
16. Silver fixer, e.g.,  To stop silver enhancement.16

 250 mM sodium thiosulfate in 20 mM
HEPES (pH 7.4)

4. PROTOCOL
4.1. Labeling with Small Markers

Labeling with small markers  Indirect labeling.
 All of the following steps are carried out
at room temperature (20 to 22°C) in a
moist chamber.
 The entire immunolabeling and silver
enhancement of sections mounted on grids
is carried out on 50 to 100 L drops of
solutions placed on a hydrophobic film (in a
dish covered by a glass plate) (see Chapter
21).
 All solutions should be centrifuged
before use (13,000 g/2 to 3 min) to pellet
aggregates and debris and to remove
bubbles.
Ultrathin resin/cryo-sections collected on  Nickel does not interfere with silver

Pioloform and carbon-coated13 nickel grids enhancement, in contrast to copper.
(e.g., 100 or 200 mesh)  Copper and nickel grids can be used for
gold enhancement.
 Formvar or collodium coating is also
possible.13

1. Inactivation of residual fixative  Mainly aldehyde groups (see Section
molecules 4.3) Optional.
 Glycine (50 mM) in 510 min  See Section 4.4.
PBS
2. Blocking unspecific protein-binding

sites
 Blocking buffer 2 × 710 min  See Section 4.4, for different recipes.
0.5% BSA, 0.5%
milk powder in PBS
3. Primary antibody incubation 

 Antibody diluted in 10 µL/grid  Detection of antigens.
blocking buffer 3060 min  Final (specific) IgG concentration 1 to
5 µg/mL.
 Moist chamber.
 For double labeling, see Section 4.2.
4. Wash
 Blocking buffer 6 × 35 min  Removal of unbound antibodies.

5. Marker incubation  Detection of bound primary antibodies.

 Marker conjugate 10 µL/grid  NANOGOLD conjugates: 1:50 to 1:100.
diluted in blocking  Ultra-small gold conjugates 1:20 to 1:50.
buffer 3060 min  QDOT conjugates: 1:20.
 Moist chamber.
6. Wash  Removal of unbound markers.
 Blocking buffer 2 × 35 min
 PBS 4 × 35 min
7. Fixation  To stabilize antigen-antibody-marker

 0.5% glutaraldehyde 5 min complexes during subsequent treatment.
in PBS
8. Wash  To remove ions that could interfere with
 H2O 5 × 23 min silver enhancement (e.g., chloride).
9. Drying (only resin sections)  Optional.
10. Silver (a-c) or gold (d)  To make small markers visible.

enhancement  Caution: Speed of enhancement
depends on temperature, higher temper-
atures will speed up deposition of silver or
gold.
 Enhancement of larger gold is less
problematic. All the silver enhancers
mentioned as well as GoldEnhance-EM can
be used.
Caution: Enhancement time is shorter due
to larger colloid size.
10a. HQ SILVER  Three solutions.

 Incubation 50100
µL/grid  Caution: Light sensitive.
 NANOGOLD in darkness
 Ultra-small gold 89 min
 QDOT 525 56 min  (Inefficient enhancement).
~ 5 min
10b. R-GENT SE-EM  Four solutions.
 Incubation 50100  Light insensitive.
 NANOGOLD µL/grid  Not recommended, no enhancement.
 Ultra-small gold
 QDOT 525 4050 min  Not recommended, inefficient enhance-
ment.
10c. Danscher solution  Four solutions (or two solutions when

stored frozen).
µL/grid  Moderate light sensitivity.
 NANOGOLD in darkness
 Ultra-small gold 3545 min
 QDOT 525 3035 min  Best enhancer for QDOT 525.
30 min

10d. GoldEnhance-EM  Four solutions.

mL/grid  Low light sensitivity.
 NANOGOLD (in darkness)
 Ultra-small gold 3060 sec
 QDOT 525 3060 sec  Not recommended, inefficient enhance-
ment.
11. Wash  On drops, to stop reaction (and remove

 H2O 5 × 4 min gum arabic).
 R-GENT SE-EM: 5 × 2 min is sufficient.
 For more rigorous halt of reaction, if
desired: fixation with 250 mM sodium
thiosulfate, 20 mM HEPES, pH 7.4
(5 min).16
12. Stabilization of silver layer by gold  Optional, to prevent redistribution or

toning loss of silver due to oxidation (not
necessary for gold enhancement).
 0.05% gold chloride 50 µL/grid  Freshly diluted from stock solution.

15 min
13. Wash  To remove ions that could interfere with

 H2O 5 × 2 min staining, e.g., phosphate.
14. Staining
14a. Resin sections

 1% aqueous uranyl 20 µL/grid  See Step 15b.
acetate (UA) 15 min  Depending on desired contrast.
 Wash
 H2O 2 × 5 sec  Final washing in a jet of H2O (wash
bottle).
 Drying  Optional, e.g., to check labeling in

TEM.
 Lead citrate solution, 20 µL/grid

in Petri dish with 15 sec3 min  Depending on desired contrast.
NaOH pellets  Keep Petri dish closed to prevent contact
with CO2.
 Wash
 H2O 2 × 1 sec  Final washing in a jet of H2O from a
wash bottle.
 Drying

14b. Cryo-sections
 Uranyl oxalate 20 µL/grid  Optional.
 Or 1% UA 10 min  We have never observed deleterious
effects of UA or neutral uranyl oxalate on
the silver layer during resin or cryo-section
staining or in preembedding labeling
experiments when samples were treated
with aqueous UA for one hour before
dehydration. However, silver-enhanced
gold in epoxy sections (if samples were
enhanced before embedding) is sensitive to
oxidation (electron beam, air humidity).
 Wash
 H2O 2 × 1 sec
15. Embedding (only cryo-sections)

a. Infiltration with 0.3 50 µL/grid  Numerous recipes for staining and
to 0.5% UA in 10 min embedding, see Chapter 19.
methyl cellulose on ice  Facilitates infiltration with methyl
cellulose.
b Picking up grids with a loop and  Dry methyl cellulose film should display
removal of excess methyl cellulose by an interference color between gold and
touching the loop at an angle of 45° to blue, see Chapter 19.
90°C to a filter paper (section side
down).
16. Air-drying  In loop; after drying, remove grids

carefully (see Chapter 19).
17. Storage (only for nonstabilized

silver)
 In dry air, under nitrogen gas or under
vacuum (in desiccator).  To prevent redistribution or loss of silver
due to oxidation.
4.2. Appendix 1: Repeated Enhancement/Enhancement after Staining

 Repeated enhancement is possible  Except for the area illuminated by the
21,22
even after inspection in the TEM electron beam.
(labeled) Lowicryl sections that were
already stained with aqueous UA and
Pb citrate can be silver-enhanced.  After washing the sections on several
drops of H2O.

4.3. Appendix 2: Double Labeling

1. Simultaneous or sequential labeling  Simultaneous labeling: Two specific
and simultaneous enhancement: After antibodies, and later two markers are
completing the labeling procedure applied simultaneously.
(sequential or simultaneous labeling of
two antigens), both markers can be  Sequential labeling: In a first round, the
enhanced in one enhancement step. first specific antibody is applied, followed
by the corresponding marker. In a second
round, the second specific antibody is
applied, followed by the corresponding
marker. A fixation step (e.g., 0.5%
glutaraldehyde) after the first labeling
round is recommended to stabilize the
antigen-antibody-marker complex, provided
the “second” antigen tolerates this
treatment. Sections should be washed and
blocked again before labeling the “second”
antigen.
 Sequential enhancement: Only possible

if the “second” antigen tolerates the
enhancement procedure.
2. Sequential labeling and sequential  It is possible to use 1 nm gold markers

enhancement: Sections can be twice (only for sequential
enhanced twice; after labeling the first labeling/enhancement19).
antigen and after labeling the second  Caution: Sizes of silver-enhanced
antigen. This means that the first gold particles are not uniform, so care has to be
marker is silver enhanced twice. taken to ensure that there is no overlap in
particle size distribution curves of the two
enhanced markers.
 Caution: Gold chloride treatment results
in partial disintegration of the silver layer
and, therefore, is not compatible with
double labeling or quantification studies.
4.4. Appendix 3: Inactivation and Blocking Solutions

Undesirable interactions between antibodies
and specimen/resin can be a serious
problem. Primary antibodies and secondary  Such antibodies can be eliminated, e.g.,
antibodies can contain antibodies that by immunoaffinity purification12 (column,
recognize epitopes, which differ from the beads, dot blot) or by absorption on
epitope under investigation. specimens lacking the antigen under
investigation.
Non-specific background can derive from
reaction of antibodies with residual fixative
molecules, mainly aldehyde groups.

From a theoretical point of view, fixative  Useful inactivation strategies:

inactivation should be done directly after Aldehyde groups can be inactivated by
completing fixation. Most probably, after amino acids like glycine or lysine (50 mM)
resin embedding, but also after cryo- and molecules like ammonium chloride
sectioning, there are no reactive fixative (NH4Cl; 50 mM).
sites left.
Nonspecific background can also derive

from hydrophobic interactions between
sample and antibodies, and from ionic
interactions, e.g., from positively charged  Useful protein blocking strategies:
specimen compounds, such as histone Gelatin (0.2 to 0.5%), liquid cold water fish
proteins in the nucleus, starch grains in skin gelatin (0.5% or 0.005%), ovalbumin
plant cells, or molecules like polycations (0.5%), bovine serum albumin (BSA)
and collagen. In contrast, negatively (0.5%), fetal or newborn calf serum (0.5%),
charged areas (phospholipids, aldehyde- nonimmune serum (0.5 to 5%), especially
fixed proteins) can repulse negatively from species used for secondary antibody
charged antibodies, thereby reducing label production), and nonfat dry skim-milk are
density. the most common blocking agents.9
In general, blocking should be done not BSA (negatively charged) is often used
only before but also during antibody and together with (cold water fish) gelatin or
marker incubation because, with time, nonfat dry skimmilk powder. In our lab, it
blocking molecules can also detach from turned out that milk powder is the strongest
the specimen. For a detailed discussion, see blocking agent.
Reference 9 and Web site of Aurion Acetylated (linearized) BSA-c (0.1%, used
(www.aurion.nl). during antibody and marker incubation)
displays increased negative charge and
hydrophobicity (www.aurion.nl). It is
thought to mimic gold surface properties,
therefore, it competes with gold colloids
during binding to unspecific sites.
Caution: Do not mix BSA-c with gelatin as
gelatin exhibits a high tendency (low gold
number) to adsorb to gold surfaces.
Electrostatic interactions can be also
influenced by buffers displaying increased
ionic strength, e.g., 0.5 M NaCl or KCl.
Hydrophobic interactions can be also

influenced by detergents added to the
antibody solution, e.g., 0.05% Tween-20,
and are also used to reduce background. It
should be kept in mind that detergents also
extract lipids/cytoplasmic material and
enhance wettability (i.e., accessibility of
antigens, e.g., in entire mount labeling
experiments).

5.1. Advantages of Small Markers

 Often considerably higher label  Necessary to detect scarce antigens.
density and efficiency when compared  Most probably, fluorochrome-coupled
to larger gold markers. IgGs are even more sensitive (for LM).
 NANOGOLD seems to be the most  Due to less electrostatic repulsion and
sensitive gold marker available for smaller hydration shell when compared to
labeling resin sections, cryo-sections colloidal gold (see above).
or permeabilized specimens.
 NANOGOLD (with bound fluoro-  Useful for permeabilized samples, but
chrome) and fluorescent QDs can be also for resin and cryo-sections, which can
used for correlative light/electron be labeled in parallel for LM and TEM.
microscopy.
 NANOGOLD conjugation kits avail-  Covalent conjugation to thiols or
able. primary amines of antibodies.
 Custom labeling service for colloidal
ultra-small gold and NANOGOLD.
5.2. Disadvantages of Small Markers

 Low intrinsic contrast, requires  Enhancement is a source of additional
additional enhancement step. problems.
 Only a few commercial suppliers,  Use test sample with known label
quality of conjugate should be checked density; check every batch for aggregates,
before use. label efficiency, unspecific background.
 Risk of increased background.  For example, due to higher sensitivity
(see Section 4.3).
 After resin labeling, unenhanced  Enhancement should be performed
NANOGOLD particles are unstable. directly after the immunolabeling pro-
cedure.
 NANOGOLD and Qdots are degraded  Silver or gold enhancement should be
by some chemicals and high done directly after immunolabeling.
temperatures.
 No protein A-1 nm gold conjugate.  Due to inactivation of protein A by
coupling.
 Quantitation is difficult.  Due to difficulty in obtaining uniformly
sized enhanced particles and unknown
stochiometry.
 Double/multiple labeling is difficult,  Due to difficulty in obtaining uniformly
but possible. sized enhanced particles (see Section 4).
 QDs are relatively large.  Due to the “coating” of the nanocrystal
core necessary for stabilization and
coupling to antibody (fragments).
 Reduced penetration requires strong
permeabilization for whole mount labeling.

5.3. Advantages of Enhancement Procedures

 Small sensitive markers with low 
intrinsic contrast can be visualized in 
LM, conventional TEM, STEM, and 
SEM.21,22 
 Final marker size can be freely chosen  Suitable for overview and detailed
by varying enhancement time. images.

 Gold-enhanced particles are stable  In contrast to silver-enhanced particles.
against oxidation.  For example, by OsO4, beam damage, air
humidity.

 Gold chloride treatment (gold toning)  Enhanced gold markers are no longer
stabilizes silver-enhanced particles. sensitive to oxidation.

5.4. Disadvantages of Enhancement Procedures

 The use of nickel grids is strongly  Exception: The recipe published by Lah
16
recommended for silver enhancement. et al is compatible with copper grids.

 Often long enhancement times.  Especially true for the most efficient
enhancers.

 Not all commercial enhancers are  See 10, Section 4.1.
useful for all markers.
 Large size distribution of enlarged  Double (multiple) labeling difficult but

gold and QDs. possible (see Section 4.2).
 Quantitation difficult (unknown
stochiometry).

 Silver and gold enhancement products  Quality of enhancer should be checked
can be a source of problems. before use, e.g., by using a test sample
(sections with known label density); check
on silver precipitates (unspecific back-
ground), enhancement efficiency, enhance-
ment time.

 Silver is sensitive to OsO4 treatment  Requires gold toning before OsO4
(oxidation). treatment, otherwise OsO4 treatment should
be reduced in time and concentration (e.g.,
0.1%; 15 min) or omitted; or gold has to be
“over” enhanced.2

 Silver layer is sensitive to oxidation  Requires gold chloride treatment or

when sections have been exposed to special storage conditions for long-term
electron beam and/or air humidity storage (e.g., in nitrogen gas or in vacuum).
(storage) (see Figure 23.12).
 Some enhancers are harmful to  Use enhancer with neutral pH or with

ultrastructure (especially important for short enhancement time (R-GENT SE-EM,
cryo-section, preembedding, and HQ SILVER, or see references 2, 7, 16.).
whole mount labeling).2,7,16
 Gold enhancement (GoldEnhance-EM)  Use test sample with known label

may be problematic with respect to density; check on enhancement time,
enhancement efficiency, reproduce- enhancement efficiency and gold
bility, and background. precipitates (unspecific background).
Published enlargement procedures  Gold markers become even smaller18
using gold chloride do not work in our (see Figure 23.13).
lab.1,18  The silver layer decomposes in several
smaller silver colloids18 (see Figure 23.13).
 Silver-enhanced gold is not compatible  Silver-enhanced gold emits fluorescent
with fluorescence markers in light light when illuminated with a Hg lamp.21,22
microscopy.
6.1. Small Gold and QD Markers

 High sensitivity markers should be  Reduced steric hindrance, reduced
used for low copy number antigens. electric repulsion, improved penetration
properties.
 All markers require routine tests for
quality.
 Useful for resin section, cryo-section,  Limited suitability for double or multiple
and preembedding/whole mount label- labeling due to large size distribution.
ing.
 All markers require additional  Susceptible to enhancement artifacts.
enhancement for LM, and convent-
ional TEM and SEM.
 Suitable for high resolution studies as 
well as for overview images. 



6.1.1. NANOGOLD
 NANOGOLD is probably the most  For low copy number antigens.
sensitive gold marker available.  Not all enhancers work with
NANOGOLD.
 Best “prerequisites” for penetration  Especially for cryo-section and pre-
into sample. embedding/whole mount labeling.
 Gold/fluorochrome conjugate.  For correlative light/electron microscopy
6.1.2. Ultra-small colloidal gold

 Sensitivity slightly lower compared to  For low copy number antigens.
NANOGOLD markers.
 Penetration properties slightly poorer  Relevant to cryo-section and pre-
when compared to NANOGOLD embedding/whole mount labeling.
markers.

6.1.3. Small QDs
 Sensitivity between 4 to 6 nm colloidal  For low copy number antigens.
gold and 1 nm gold markers  For cryo-section and resin section
labeling, limited suitability for whole
mount labeling.
 Fluorescent marker of low electron  For correlative light/electron micros-
density, which can be silver-enhanced. copy.

6.2. Silver Enhancement Techniques

 Necessary to visualize small markers  All enhancers require routine tests for
in LM, conventional TEM and SEM. quality.
 Nickel grids should be used (for TEM).
 Suitable for overview and detailed  Caution: Silver is sensitive to oxidation;
images. stabilization useful prior to OsO4 treatment
and for long-term storage (see Section 6.4).
6.2.1. HQ SILVER
 Suitable for all gold and QDOT  Tailored specifically for NANOGOLD
markers and all labeling techniques.  Slightly reduced efficiency for
QDOT 525.
 Low/no interference with ultrastruc-  Important for cryo-section and pre-

ture preservation. embedding/whole mount labeling.

6.2.2. R-GENT SE-EM

 Suitable for all colloidal gold markers  Tailored specifically for ultra-small gold.
(all labeling techniques).  Not suitable for NANOGOLD and
QDOT 525.
 No interference with ultrastructure  Important for cryo-section and pre-
preservation. embedding/whole mount labeling.
6.2.3. Danscher solution

 Suitable for all gold and QDOT  Easy to prepare.
markers and all labeling techniques.  Relatively cheap.
 Premixed aliquots can be stored frozen.
 May be harmful to ultrastructure in cryo-
sections and preembedding/whole mount
labeling experiments.
6.3. Gold Enhancement Techniques

 Gold enhancement results in stable  Gold-enhanced particles are not sensitive
gold particles. either to OsO4 treatment, electron beam or
air humidity.
6.3.1. GoldEnhance-EM
 Suitable for all gold markers.  Limited reproducibility?
 Slightly reduced efficiency?
 Not suitable for QDOT markers.
 Strong background on labeled cryo-
sections.
6.3.2. Published recipes

 Not recommended.  Treatment with gold chloride is not
suitable for enlargement of small gold
markers.
6.4. Silver Stabilization/Gold Toning

 Recommended for stabilization of  Gold chloride-treated particles are no
silver-enhanced gold particles for long longer sensitive to oxidation.
term storage and against the action of  Caution: Gold chloride treatment
OsO4. slightly reduces the size of silver-enhanced
 Alternatively, storage in dry air, under particles and leads to their disintegration
nitrogen gas or vacuum (e.g., in (see Figure 23.15).
desiccator).

7. OBSERVED RESULTS
 Figure on Chapter’s title page  Centriole labeling in ultrathin cryo-

sections of U2OS cells fixed with 8% FA.
C-Nap1 was detected at the proximal end
with specific rabbit antibodies and protein
A-15 nm gold (left) or silver-enhanced
NANOGOLD coupled to goat anti-rabbit
F(ab)2 fragments (right). Bar = 250 nm
Ultrathin resin sections of Escherichia.
coli cells embedded in methacrylate 
Lowicryl HM20 were incubated with 
rabbit antibodies raised against the outer 
membrane protein OmpA. In this 
overproducing strain, OmpA is mainly 
found in the periplasmic space. Markers 
used were protein A-6 nm colloidal gold, 
ultra-small (colloidal) gold, NANOGOLD, 
or QDOT 525. UA and lead citrate 
staining were omitted. 

 Figure 23.9 QDOT 525F(ab`)2
marker:
(a) Without staining.
(b) Negatively stained with uranyl acetate.
Bar = 50 nm
QDOT 525 marker after immuno-

labeling of E. coli sections:
(c) Without enhancement.
(d) For comparison, labeling with protein
A-6 nm gold marker is shown.
Bar = 100 nm

 Figure 23.10 HQ SILVER enhancer
applied to:
(a) NANOGOLD marker (8 min).

(b) Ultra-small gold marker (5 min).
(c) R-GENT SE-EM enhancer applied to
ultra-small gold marker (40 min).
Bar = 100 nm


 Figure 23.11 Danscher enhancer

applied to:
(a) NANOGOLD marker (35 min).

(b) Ultra-small gold marker (30 min).
(c) QDOT 525 marker (30 min).
Bar = 100 nm
 Figure 23.12 GoldEnhance-EM

enhancer applied to NANOGOLD marker
(30 sec).
Bar = 100 nm
 Caution: High background on
immunolabeled ultrathin cryo-sections
 Figure 23.13 Gold chloride treatment

reduces gold size rather than leading to
enhancement:
(a) Untreated control.
(b) Protein A-6 nm gold treated with 0.05%
gold chloride for 30 min.
Bar = 100 nm
 Figure 23.14 The silver layer is

sensitive to oxidation, e.g., during storage
of labeled sections:
(a) Image was taken directly after silver
enhancement.
(b) Image was taken several weeks after
silver enhancement (same section, stored
in humid ambient air).
Bar = 100 nm
 Figure 23.15 Gold chloride treatment

(0.05%, 10 min, 20°C) stabilizes silver but
leads to partial disintegration of the silver
layer:
(a) Untreated silver-enhanced control.
(b) After gold chloride treatment, the silver
layer decomposes into several smaller
silver particles.
Bar = 100 nm

8. REFERENCES
1. Baschong W. and Stierhof, Y.-D. Preparation, use, and enlargement of ultrasmall

gold particles in immunoelectron microscopy. Microsc. Res. Techn., 42, 66, 1998.
2. Burry, R.W. Pre-embedding immunocytochemistry with silver-enhanced small
gold particles, in Immunogold-Silver Staining, Hayat, M.A., ed., CRC Press, Boca
Raton, FL, USA, 1995, 217.
3. Danscher, G. Histochemical demonstration of heavy metals: A revised version of
the sulphide silver method suitable for both light and electron microscopy.
Histochem., 71, 1, 1981.
4. Danscher, G. et al. Trends in autometallographic silver amplification of colloidal
gold particles, in Immunogold-Silver Staining, Hayat M.A., ed., CRC Press, Boca
Raton, FL, USA, 1995, 11.
5. Dettmer, J. et al. Vacuolar H+-ATPase activity is required for endocytic and
secretory trafficking in Arabidopsis. Plant Cell, 18, 715, 2006.
6. Giepmans, B.N.G. et al. Correlated light and electron microscopic imaging of
multiple endogenous proteins using quantum dots. Nat. Methods, 2, 743, 2005.
7. Gilerovitch, H.G. et al. The use of electron microscopic immunocytochemistry
with silver-enhanced 1.4 nm gold particles to localize GAD in the cerebellar nuclei.
J. Histochem. Cytochem., 43, 337, 1995.
8. Griffith, J.M. and Posthuma, G. A reliable and convenient method to store
ultrathin thawed cryo-sections prior to immunolabelling. J. Microsc., 212, 81, 2003.
9. Griffiths, G. Fine Structure Immunocytochemistry, Springer-Verlag, Berlin,
Germany, 1993.
10. Hainfeld, J.F. and Furuya, F.R. A 1.4-nm cluster covalently attached to antibodies
improves immunolabeling. J. Histochem. Cytochem., 40, 177, 1992.
11. Hainfeld, J.F., and Furuya, F.R. Silver-enhancement of Nanogold and
undecagold, in Immunogold-Silver Staining, Hayat, M.A., ed., CRC Press, Boca
12. Harlow, E. and Lane, D. Using Antibodies, Cold Spring Harbor Laboratory Press,
New York, 1999.
13. Hayat, M.A. Principles and Techniques of Electron Microscopy, Cambridge
University Press, Cambridge, UK, 2000.
14. Hermann, R., Schwarz, H., and Müller, M. High precision immunoscanning
electron microscopy using Fab fragments coupled to ultra-small colloidal gold. J.
Struct. Biol., 107, 38, 1991.
15. Holgate, C. et al. Immunogold-silverstaining: New method of immunostaining with
enhanced sensitivity. J. Histochem. Cytochem., 31, 938, 1983.
16. Lah, J.J., Hayes, D.M., and Burry, R.W. A neutral pH silver development method
for the visualization of 1-nm gold particles in pre-embedding electron microscopic
immunocytochemistry. J. Histochem. Cytochem., 38, 503, 1990.
17. Leunissen, J.L.M. and Van De Plas, P. Ultrasmall gold probes and cryo-
ultramicrotomy, in Immuno-Gold Electron Microscopy in Virus Diagnosis and
Research, Hyatt A.D. and Eaton B.T., eds., CRC Press Inc., Boca Raton, FL, USA,
1993, 327.
18. Pohl, K. and Stierhof, Y.-D. Action of gold chloride (“gold toning“) on silver-
enhanced 1 nm gold markers. Microsc. Res. Techn., 42, 59, 1998.

19. Sibon, O.C.M. et al. Ultrastructural localization of epidermal growth factor (EGF)-
receptor transcripts in the cell nucleus using pre-embedding in situ-hybridization in
combination with ultra-small gold probes and silver enhancement. J. Histochem.,
101, 223, 1994.
20. Stierhof, Y.-D. et al. Yield of immunolabel compared to resin sections and thawed
cryosections, in Colloidal Gold: Principles, Methods, and Applications. Hayat,
M.A., ed., Academic Press Inc., San Diego, IL, USA, 1991, 87.
21. Stierhof, Y.-D. et al. Direct visualization and silver enhancement of ultra-small
antibody-bound gold particles on immunolabeled ultrathin resin sections. Scan.
Microsc., 6, 1009, 1992.
22. Stierhof, Y.-D. et al. Use of TEM, SEM, and STEM in imaging 1-nm colloidal
gold particles, in Immunogold-Silver Staining. Hayat, M.A., ed., CRC Press, Boca
23. Stierhof, Y.-D., Humbel, B.M., and Schwarz, H., Suitability of different silver
enhancement methods applied to 1 nm colloidal gold particles: An immunoelectron
microscopic study. J. Electr. Microsc. Tech., 17, 336, 1991.
24. Stierhof, Y.-D. and Schwarz, H. Labeling properties of sucrose-infiltrated
cryosections. Scan. Microsc. Suppl., 3, 35, 1989.
25. Tokuyasu, K.T. A study of positive staining of ultrathin frozen sections. J.
Ultrastruct. Res., 63, 287, 1978.
26. Weipoltshammer, K. et al. Signal enhancement at the electron microscopic level
using Nanogold and gold-based autometallography. Histochem. Cell. Biol., 114,
489, 2000.

























3-D Electron Tomography of Cells and Organelles 619
CONTENTS

1. PRINCIPLES OF ELECTRON TOMOGRAPHY......................................... 622
3. INSTRUMENTAL REQUIREMENTS ........................................................... 625
3.1. The Transmission Electron Microscope ................................................... 625
3.2. The Charge Coupled Device (CCD) Camera............................................ 627
3.3. High-Tilt Holders ..................................................................................... 628
3.4. Software.................................................................................................... 628
3.4.1. Data acquisition software............................................................ 629
3.4.2. 3-D data reconstruction software ................................................ 629
4. PROTOCOLS .................................................................................................... 631
4.1. Specimen Preparation ............................................................................... 631
4.2. Data Acquisition ....................................................................................... 633
4.3. Alignment, Reconstruction and Modeling................................................ 638
4.4. Short Protocol for Electron Tomography ................................................. 639
5.1. Advantages of 3-D Electron Tomography................................................ 641
5.2. Disadvantages of 3-D Electron Tomography ........................................... 642
6.1. Why Electron Tomography?..................................................................... 644
6.2. Why Electron Tomography on Resin Sections? ....................................... 644
8. REFERENCES .................................................................................................. 648



One of the techniques in electron microscopy that evolved rapidly in the past decade is
transmission electron tomography. The method was first applied in 1970;1 however, only
with increased computer power and once several of the technical challenges of the
method were worked out,2 did electron tomography became a valuable and powerful tool.
In the meantime, many electron microscopists realized that conventional electron
micrographs merely were two-dimensional projections of three-dimensional structures
and, in fact, an oversimplification of the actual structures in a cell. During the last few
years, methods were developed to overcome some of the remaining limitations of electron
tomography and focused on developing methods that provide a better understanding of
the actual processes within cells, between tissues and in organs.
The use of 3-D electron tomography in life sciences can be separated into several major
fields of application, which have different requirements with respect to specimen
preparation, the microscope and subsequent image processing. A possible distinction can
be made between 1) cryo-electron tomography, 2) cellular tomography, 3) scanning
transmission electron microscopy (STEM) tomography, and 4) energy filtering (EF)
tomography. They all have their advantages and disadvantages.
Cryo-electron tomography aims at imaging cryofixed and unstained material with low-
dose transmission electron microscope (TEM)3 (see Chapter 12). Cryo-electron
tomography is based on optimal specimen preservation, but is restricted to 3-D imaging
of macromolecular complexes (>200 kDa), isolated cell organelles (< 500 nm), thin areas
of cells (< 500 nm) or thin sections of cryo-fixed material.4,5 To date, there are no
possibilities for specific labeling in cryo-electron tomography, but localization of large
proteins with a characteristic and distinguishable shape can be mapped in their cellular
context by template matching.6
Cellular tomography aims at reconstructing relatively large, unique structures by

preferentially using high-pressure freezing (see Chapters 5, 6) and freeze-substitution (see
Chapter 13) for specimen preparation.7-13 Often, several tomograms are combined to build
3-D reconstructions of even larger areas.14 Because of the dehydration and staining step,
the specimen may be less well preserved than frozen-hydrated samples (see Chapters 1,
11). The method has, however, the advantage that large samples of cells and tissue can be
processed easily. Furthermore, cellular components can be localized by specific labeling
and the sections are more stable in the electron beam.
Cellular tomography can also be applied to chemically fixed material and even to
Tokuyasu cryo-sections for combination with immunogold labeling; however, one has to
realize that structural artifacts are introduced.12
STEM tomography is a relatively new type of tomography.15-18 The technique has been
shown to be very powerful for reconstructing highly diffracting specimens (e.g.,
crystalline material) that are not suitable for TEM imaging. When using a high-angle
angular dark-field detector (HAADF), HAADF-STEM imaging collects images for which
the contrast is more sensitive to differences in atomic weight. Therefore, for stained
sections and for detecting ultra-small (gold) labels,19,20 STEM tomography has potential
applications.18 To date, however, HAADF-STEM tomography is not often used in the life
sciences.

Electron tomography using energy filtering16 is an approach for imaging very thick
sections (up to 1 µm) or cryo-sections (see Chapter 12). Often zero-loss filtering is used
for cellular tomography. In addition, with sufficient electron dose, it may be possible to
acquire 3-D images of areas containing a high concentration of a particular element, e.g.,
calcium.21
In this chapter, we will introduce the basic procedures and exemplify the methods that are
used in electron microscopy facilities to obtain 3-D information on unique subcellular
structures with nanometer-scale resolution and overview the continuing development of
automated data acquisition programs as well as the improved user interface procedures.
1. PRINCIPLES OF ELECTRON TOMOGRAPHY
 Electron tomography is a method to

computationally generate a three-
dimensional (3-D) digital volume (the
tomogram) from multiple 2-D
projection images of the 3-D structure.
 There are two assumptions that have
to be taken into consideration when
electron tomography is performed.
1 The first assumption is that the  For this reason, stained resin-embedded
specimen does not change (shape, specimens have to be preirradiated,
density) during the recording of the tilt ensuring that the amount of specimen
series. shrinkage during data collection is
minimal.22
2 The second assumption is that the  This assumption does not hold for highly
image is a true projection of the diffracting specimens (for those specimens
densities inside the specimen STEM tomography is the method of
(projection requirement). choice).
 These assumptions set a limit to the  The use of energy filtering helps to meet
thickness of the specimen that can be the projection requirement.
observed. For very thick specimens
(multiple), electron scattering events
within the specimen can become too
dominant.
 For optimal results, the object must be  This is usually between 65o and +65o
projected under an as wide as possible with an increment of 1o, resulting in a tilt
range of viewing directions. series of 131 images.

 The images need to be aligned and the  There are several procedures to compute
reconstruction (called tomogram) can the 3-D reconstruction. Currently, the most
be computed. widely used approach is resolution-
weighted back projection (WBP).7 In
addition to this, there are other methods like
the simultaneous iterative reconstruction
technique (SIRT)23), the algebraic
reconstruction technique (ART) or a dual-
axis iterative algorithm from Tong et
Midgley24
 Once the tomogram is computed,

complex structures in the interior of
the specimen can be viewed in thin
digital slices (often as thin as 3 to
10 nm), much thinner than can be  This means that in a 350 nm-thick
accomplished with physical sectioning section with a recorded tilt series of 141
(50 to 80 nm). The obtainable images (70 to +70o), a resolution of about
resolution can be calculated by the 7.5 nm can be achieved.
following rule of thumb: three times  3 × 350/141 = 7.5 nm
section thickness divided by the
number of images.
 Furthermore, in the 3-D digital  In this way, the structure of interest can
volume, the object can be sampled then be analyzed and modeled.
with image processing tools, e.g.,
segmented in any direction one wishes
without being hampered by
information from overlying structures
in the region of interest.
 With the current type of specimen  New approaches to achieve full, 180o
holders, it is not possible to tilt the tilt, angles are under investigation.25
specimen to angles higher than 70o; at
higher tilt angles the material holding
the specimen blocks the field of view.
 Due to the limitations of the angular  The beads will not be completely round,
tilt range, there is a lack of but will appear elongated in Z-directions. In
information at higher tilt angles. In the the Fourier transform of the tomogram, the
tomogram, this missing information is missing information can be seen as a
clearly visible in the shape of missing wedge of information corres-
reconstructed high-density particles ponding to the missing angular tilt range
(e.g., gold beads). (see Section 5.2, Figure 24.18).
 The missing wedge can be reduced to  The improvement in resolution

a missing pyramid by rotating the grid (especially in Z-direction) is very sig-
over 90o, recording a second tilt series nificant and results in a more isotropic
of the same area and combining the resolution throughout the tomogram.
two tomograms into a dual-axis tilt
reconstruction.26,27







A: The thick resin section with the
structure of interest.
B: The projection images of the thick

section. After recording the projection,
images need to be aligned with respect to
each other.
C: Individual slices from the computed

tomogram.














D: Segmentation and annotation of the

optical slices.
E: 3-D model of the structure.
Figure 24.1 (A-E) Schematic represent-

ation of the basic goal of performing
electron tomography, namely the
reconstruction of 3-D information from an
object. (Sketch courtesy of Misjaël N.
Lebbink.)
3. INSTRUMENTAL REQUIREMENTS
3.1. The Transmission Electron Microscope

 To acquire 3-D information, electron  Keep in mind that at a 60o tilt angle the
tomography is performed on relatively apparent thickness of a 500 nm-thick
thick specimens. specimen has doubled (1000 nm), and at tilt
 angles of 70o tripled (1500 nm).
 For imaging, sections of up to 500 nm  Increase in specimen thickness results in
thickness and a TEM with an increased scattering inside the specimen
acceleration voltage of at least 200 kV creating inelastic scattered electrons, which
are necessary. leads to blurring and reduced image
contrast.

  In many cases an imaging energy filter

will be very helpful in generating high-
contrast images. The images are taken at
zero-loss, filtering out the inelastic
scattered electrons that lost energy within
the specimen.
 The goniometer should allow accurate  A goniometer that can tilt under cryo
tilting of the specimen holder over a conditions is required for cryo-electron
large angular range. microscopy (see Chapter 12).
 For room temperature applications,
dedicated tomography holders are available
that allow tilting at high angles, sometimes
in combination with dual-axis tilting.
 Sophisticated acquisition packages  Many of the automated tracking and

(commercial and academic) are under focusing procedures are based on methods
constant development. They allow the introduced in the early 1990s.31-33
automated tracking, focusing and  Recording of a tilt series (131 images)
recording of the area of interest. can be performed within 45 minutes.
 The stability of the goniometer and the
high-tilt holders allow holder
calibration procedures28-30 to predict
the movement of the holder at
eucentric height in the microscope.
Figure 24.2 Tecnai-20 with LaB6 filament

(FEI Company, Eindhoven, The Nether-
lands) at Utrecht University. This dedicated
microscope is used solely for recording tilt
series for 3-D electron tomography. The
microscope is equipped with the automated
data acquisition software Xplore 3-D (FEI
Company, Eindhoven, The Netherlands).

3.2. The Charge Coupled Device (CCD) Camera

 Before the advent of digital image  The size of the individual pixels on the
recording, one of the most laborious CCD chip determines the final resolution
and time consuming tasks of doing (sampling distance) and the sensitivity of
tomography was the processing and the camera.
digitization of recorded tilt series from  The characteristics of the scintillator
film. During the last decade, digital (type of material, thickness) in combination
data acquisition using CCD cameras with the noise characteristic of the CCD
has become the most common type of chip determine the sensitivity of the camera
data recording. The incoming electrons system. 
are converted to photons on a 
scintillator and then transmitted to the
CCD chip either via fiber coupling or
via a lens-coupled camera system.
 For cellular tomography on high  Currently, CCD cameras for

contrast specimens it might be better to transmission electron microscopy will have
use a camera with smaller pixels and between 10242 and 40962 pixels.
imperfect noise characteristics,  The increasing amount of pixels leads to
whereas for cryo-electron tomography, large data sets close to 1 Gbyte in size.
with beam sensitive specimens and  Especially for iterative reconstruction
therefore low beam intensities, it is algorithms this may lead to challenges with
better to have a camera with larger respect to computation and data storage.
pixels and very good noise char-
acteristics. The quality of the digital
cameras is improving continuously.
Figure 24.3 The TemCam-F214 CCD-

camera (TVIPS GmbH, Gauting, Germany)
that records the projection images at the
Tecnai-20 LaB6 microscope at the electron
microscopy department of Utrecht Univer-
sity.

3.3. High-Tilt Holders

 The specimen grid holders must be  Causing shadowing on the specimen.
able to tilt to high-tilt angles without  These high-tilting angles are necessary to
obstructing the electron beam. The minimize the size of the missing wedge.
higher the achievable tilt angle the  By combining the two tomograms, the
less loss in Fourier space and the missing wedge can be reduced to a missing
smaller the so-called missing wedge
artifact. pyramid. In our experience, it is more
 A second approach to reduce the size advantageous to record two tilt series from
of the missing wedge is to rotate the 60 to +60o and to combine them in one
grid by 90o and record a second tilt tomogram instead of recording one single-
series of the same specimen area.26,27 tilt series from 70 till +70o.

 To achieve these high-tilt angles, the  Currently, we use the Fischione 2020
material of the holder near the two rotational tilt holder (Fischione Instruments,
sides of the grid is made as thin as Pittsburgh, Philadelphia, USA) for this
possible to have a relatively large purpose.
field of view, even at high-tilt angles  They allow tilting angles up to +/80o.
(see Figure 24.4). Until recently, the
in-plane rotation of the grid over 90o
for recording the second tilt series
had to be done manually. To facilitate
the process of recording dual-axis tilt
series dedicated specimen holders are
now available, which allow in-plane
rotation of the grid even when the
holder is inserted into the electron
microscope.
Figure 24.4 The Gatan high-tilt holder

(Gatan U.K., Abingdon, U.K.) with a
clamping device for easy and firm clamping
of the grid.

3.4. Software
 Not only hardware (see above) but  Three different types concerning the
also software for automated data Three different steps:
acquisition, data handling and data
analysis is necessary. Several
software packages are available for  Acquisition
these purposes. Some are
commercially available (meaning  Reconstruction
expensive and without source code)
and some are academic (meaning  Analysis and modeling
much cheaper or even free and with 
source code).

3.4.1. Data acquisition software

 The main goal of the data acquisition  Some of them are commercial packages
software is to accurately collect a tilt (e.g., Xplore 3-D, FEI Company,
series of digital images from a fixed Eindhoven, The Netherlands; EMmenu,
location on the specimen. Irregu- TVIPS GmbH, Gauting, Germany;
larities in stage movement (x, y) and TEMography™, JEOL System Technology
focusing (z) during the recording of Co, LTD, Tokyo, Japan).
the tilt series should be corrected.  others have an academic origin and are
Many data acquisition software freely available (e.g., TOM toolbox, Max-
packages are available for image Planck-Institute for Biochemistry, Martins-
acquisition. ried, Germany; Serial EM, University of
Colorado, Boulder, Colorado, USA; and
Priism/IVE, UCSF, San Francisco,
California, USA).
 The commercially available packages  Currently, in our department the Xplore

are stable and supported by a 3-D package and the TOM toolbox are used
company but do not allow users to for data acquisition at the electron
make adjustments in the software. microscopes.
The packages available through
academia are more flexible and under
constant development by the user
community.
3.4.2. 3-D data reconstruction software

 The 3-D reconstruction software  In most cases, it takes less then an hour
performs two tasks: alignment and to compute a dual-axis tilt tomogram.
reconstruction. Currently, we use the
IMOD package34 from University of
Colorado (Boulder) to align the
recorded projection images using
fiducial markers. The tomograms are
computed by means of the resolution-
weighted back projection method.
 For those occasions when the

resolution-weighted back projection is
less optimal (e.g., for very low
contrast images), we use the ART or
the SIRT modules available in the
Xplore 3-D package. The resolution-  The advantage of a SIRT reconstruction
weighted back projection algorithm algorithm is that the low frequency
implemented in IMOD is relatively information is better resolved than with
fast. resolution-weighted back projection.

 The ART and SIRT reconstruction  There is, however, a considerable

algorithms have the disadvantage that reduction in ART and SIRT calculation
they are computationally intensive and times possible by using the computational
still lack the ability to generate dual- power of graphical processing units or
axis tilt tomograms. graphic cards on a standard desktop
computer.
Figure 24.5 Two images of the user

interface from the data reconstruction
software package IMOD.34 This is one of
the few packages that allow the
reconstruction of double-tilt tomograms.

4. PROTOCOLS











Figure 24.6 Specimen preparation step scheme used for cellular tomography of cells and
tissue.

 The quality of the tomogram

produced with electron tomography
completely depends on the quality of
the sample in the microscope. The
cellular structures under investigation
must be prepared in a way that they
represent a cellular state that is as
close as possible to the native state. In
order to achieve this, several cryo-
fixation protocols were developed.

 Specimen preparation protocols are  For specimens prepared by freeze-
under constant development and substitution, several protocols have been
alternatives for traditional chemical developed that include staining of cellular
fixation are used to overcome structures with heavy metals, like uranyl
problems, such as displacement of the ions, lead ions or osmium tetroxide36 (see
cellular proteins, conformational Chapter 13). For this type of specimen, the
changes and denaturation of proteins. contrast mechanism is mainly scattering
Also the shrinkage of tissues and cells contrast.
during the fixation step12,35 and the
loss of lipid content36 are important
issues in developing new fixation
protocols (see Matsko and Müller37 for
a recent report).
 In our laboratory, the cryo-fixation  In our approach, section thickness

method of high-pressure freezing normally ranges between 250 nm to
(HPF) followed by freeze-substitution 500 nm.
(FS) and resin embedding is often used
(see Geerts et al8 for a detailed
description). Briefly, the cryofixed
cellular materials are embedded in a
resin and thick sections are cut with an
ultra-microtome using diamond knifes.
The thick sections are collected on
Formvar (or Pioloform)/carbon-coated
copper grids.
 In most of our studies on cellular  In cases where specific immunogold

arrangement, e.g., membrane contin- labeling for the localization of particular
uities between organelles or connec- proteins is performed, the specific gold
tions between cellular compartments, label can sometimes be used also as a
6, 10 or 15 nm gold particles are fiducial marker for the alignment step.
added randomly to both sides of the
section. These gold beads function as
fiducial markers during the alignment
step that precedes the generation of
the tomogram.

 G = Golgi complex
 ER = Endoplasmic reticulum
 M = Mitochondria
Figure 24.7 Overview of a plasma cell in a

300 nm-thick section. The plasma cell was
prepared by high-pressure freezing
followed by freeze-substitution and
embedding in Epon.

 This procedure is especially suited to
study membranes and their relationship.
Figure 24.8 Detail of the previous plasma

cell showing an extensive endoplasmic
reticulum (ER), a large Golgi complex (G)
in which cisternae and the trans Golgi
network (TGN), with numerous small
vesicular structures, can be clearly
recognized.

4.2. Data Acquisition

 The first prerequisite is a well aligned  In our department we use a Tecnai 20
TEM with an accelerating voltage of at with a LaB6 filament equipped with a Tietz
least 200 kV. The microscope Temcam F214 CCD camera with 2048 ×
alignment procedures are not described 2048 square pixels.
because they are outside the scope of
this chapter.
 Next, it is required that the CCD  For the installation and proper
camera is properly installed and that calibration of the CCD camera, we refer
the reference images (bias and gain readers to the manuals provided by the
reference images) are correctly manufacturer of the electron microscope
recorded to produce high-quality and/or the camera.
digital images.

 When one acquires a manual tilt series,

it becomes clear that the repetitive
actions needed to keep the area of
interest centered and focused are time
consuming. Area selection, focusing
and tracking has to be done by intense
interaction between the user and the
computer program controlling the
microscope. Once digital imaging
became a commodity on most electron
microscopes, automation was
developed that replaced the human
operations to center and focus.
 In 2001, it was noted that imper-

fections of the tilting movement of the
goniometer (see Figure 24.9) were
very reproducible.28
Figure 24.9 The holder calibration curves

for the Gatan high-tilt holder that can be
used for automatic recording of data sets.
For visualization, purposes the thickness of
the calibration line is adjusted in Adobe
®
 Photoshop .
 Calibration curves were used to  Today automated tomography data

predict image and focus shifts and to acquisition software packages are available
compensate for them automatically on a variety of TEMs.
during data acquisition. Variations
and improvements on this basic idea
were realized and further improved in
several packages (UCSF package
2004,29 TOM toolbox, Martinsried,
2005, SerialEM.30)


 An important concept in automated  Theoretically, the radius of the circular

data acquisition is the “optimized movement is then minimal.
position” concept. From a theoretical
point of view, any point on the  By setting the specimen at eucentric
specimen undergoes a circular height, the position of the specimen within
motion, with the center of the circle the goniometer is changed until the height
lying on the tilt axis (a line in of the specimen and the tilt axis are in the
horizontal direction determined by the same plane.
goniometer) and the radius of the
circle being the distance between the
point and the tilt axis.

 However, in practice there still will be  The distance between these two axes, the
movement visible when the misalignment, can be several m.
goniometer is tilted. This is due to the
fact that the position of the tilt axis
(horizontal) does not intercept with
the optical axis of the microscope
(vertical) that defines the area that is
imaged on the center of the CCD
image.

 The major effects of this mis-  However, focus changes and
alignment are a focus change and translational shifts can be corrected by
translational shift when the specimen adapting the objective lens and the image
is tilted. shift controls of the microscope.

 Fortunately, automated correction of  The purpose of these calibrations is to
the misalignment is carried out by give a physical meaning to the pixel size of
adjusting the optical axis of the the CCD camera (the magnification) and to
microscope (vertical) to coincide with calibrate the image shift, stage shift and
the tilt axis of the goniometer focus shift controls
(horizontal). To carry this alignment
through, a number of basic calibra-
tions need to be performed.

 Apart from these general calibrations  Used to predict the image shift and focus
there are two tomography-specific changes that are needed during tilt-series
calibrations, those of the tilt axis acquisition to keep the area of interest in
(measuring the optimized-position the center of the image.
shift — the distance between the  These calibrations have to be checked
optical axis and the tilt axis) and once a month.
measuring the remaining focus
changes and translational shifts of the
specimen when it is tilted.

 When these prerequisites are fulfilled,

recording a tilt series is a straight
forward procedure that can be
performed within an hour.
 As stated before, in our laboratory we  For optimal alignment, we need 10 to 15

image specimens prepared with cryo- gold particles per image.
fixation and freeze-substitution. We  The required dilution of the gold particle
use stained resin-embedded sections solution can be checked by applying a
of 250 to 500 nm thickness, often droplet of a particular dilution on bare grids
decorated with fiducial markers (gold and examining them in the microscope at
beads in the size range of 5 to 15 nm) the specific magnification that will be used
on both sides of the specimen. for recording the tilt series
 The specimen grid is tightly clamped  This is absolutely necessary for

into a high-tilt holder (single-tilt or recording proper tilt series.
rotation holder) and carefully inserted
into the goniometer of the
microscope.
 Next, the area of interest has to be set  Eucentricity has to be checked every
at eucentric height. time before recording a new tilt series,
either by hand or with the data acquisition
software.
 Before recording the tilt series, it also
has to be checked (by tilting the
holder) on how far the area of interest
can be tilted without any obstructions
from the grid bars or the holder itself.
 Next, the intensity of the beam on the  This is necessary to prevent over-
specimen is checked at the zero illumination of the area of interest during
degree tilt angle. the recording of the tilt series.
 Keep in mind that the light intensity on
the CCD screen depends on the tilt angle of
the sample; at high-tilt angles the section
thickness increases considerably resulting
in much weaker image intensity on the
CCD chip.
 Finally, the actual recording of the

projection images can start. In
general, it takes about 45 minutes to
record a complete series of projection
images.

Figure 24.10 User interface of the FEI

automated data acquisition software where
data acquisition parameters can be adjusted.
Figure 24.11 User interface showing the

tracking and focusing images and the
corresponding cross-correlation images that
are used for automated focusing and
tracking of the area of interest.
Figure 24.12 The user interface that allows

visualization of the recorded tilt series. The
scale bar, the tilt angle and the tilt axes are
viewed and stored as tiff images.
Figure 24.13 The goniometer of the

Tecnai-20 (FEI Company).

4.3. Alignment, Reconstruction and Modeling

 The series of projection images (tilt  For the reconstruction process, several
series) are stored for subsequent academic and commercial packages are
computation of the tomogram. available.
 The resolution-weighted back project-  WBP is a fast reconstruction procedure.

tion algorithm,23 which is also
implemented in the IMOD package is
a widely used method for computing
the tomogram. For a more detailed
description of this package, we refer
to Kremer et al.34
 In short, the individual projection

images of the recorded tilt series show
small image shifts from one
projection image to the next. The first
alignment step in IMOD reduces the
image shifts from one image to the
next by means of a cross correlation-
based alignment of the individual
images.
 In the second alignment step, fiducial  For a more elaborate description we

markers present in the projection refer to the excellent manual and the Web
images (of the gold particles) are site of the University of Colorado
selected and used for a more accurate (http://bio3d.colorado.edu/imod/).
alignment of the projection images.
Then the tomogram can be computed
using the resolution-weighted back
projection method. Also available in
IMOD is the possibility to generate
dual-axis tilt tomograms. After the
tomogram is calculated, the analysis
of the recorded data set can start.
 In the IMOD package, several options  So far, all the steps in the acquisition
are available for modeling. There is a process can be performed without specific
contour drawing mode (most often knowledge of the object under study.
used), a thresholding tool for  However, for interpretation of the
automatic selection and drawing of tomogram and modeling, specific
contours and contour volume knowledge of the object is absolutely
rendering options. necessary.
 In many studies, using the contour mode,
the extent to which membranous com-
partments are connected to other
membranous compartments is investigated,
both in the exocytic as well as in the
endocytic pathways.38,39

 An illustrative example is the connection

between the ER and the newly originating
peroxisomes.40
 The contour drawing mode is, however,
prone to the interpretation by the
investigator. Therefore, efforts are now
being undertaken to find computational
methods for automated pattern recognition
by means of thresholding and template
matching by computers.18,41
Figure 24.14 Projection image of a

300 nm-thick plastic-embedded section
with randomly applied fiducial gold
markers (10 nm), which can be seen as
dense black dots all over the section.
4.4. Short Protocol for Electron Tomography

1. Before starting data acquisition check  Also the TEM basic calibrations and the
the general microscope conditions: TEM tomography calibrations must have
Beam tilt pivot points, tomo rotation been carried out.
center, beam tilt calibrations, image
(beam) calibration, image shift pivot
points, image shift for the magnifica-
tion range.
2. Check the CCD dark image and gain  For the dark image correction, close the
correction. The fully corrected image screen and record the dark image.
should appear homogenously grey.  For the gain correction, remove the
specimen and illuminate the CCD evenly
with a number of counts not too different
from the working conditions later.
3. Load the sample and locate the area of
interest.
4. Adjust eucentric height manually.

5. Focus.
6. Center the condenser aperture.
7. Adjust the gun conditions.
8. Align rotation center.
9. Correct objective astigmatism and

focus image.
Figure 24.15 User interface of Xplore 3-D

for checking light intensity and
astigmatism.
10. Check for the maximum tilt angles

without blocking the area of interest.
11. Adjust the CCD settings for search,

focus, exposure and tracking.
12. Set the parameters for recording a tilt

series.
Figure 24.16 User interface of Xplore 3-D

for setting the data acquisition parameters
for automated recording and storage of the
tilt series.
13. Provide a unique name for storage of

the image series.
14. Check if the focusing and tracking

steps are performed well.
15. Check the recorded tilt series in

visualization mode.

16. Align the individual projection

images with IMOD.
17. Calculate the tomogram with the

weighted back projection method in
IMOD.
18. Analyze the tomogram.
19. Perform the contouring and/or surface  Amira is available from Mercury
rendering in IMOD or Amira. Computer Systems SAS, 33708 Merignac
Cedex, France.
5.1. Advantages of 3-D Electron Tomography

 The produced dataset contains  To achieve the best possible Z-resolution
projection images of the same in transmission electron microscopy
position under different angles. studies, sections need to be as thin as
possible. In the recorded projection image,
however, structures obscure each other and
the achieved Z-resolution is never better
than the section thickness.
 The third dimension in the section is  By the use of the back projection
restored. method. In this way a “virtual” 3-D block is
created and individual images can be
retrieved from this block and used for
analysis.
 The Z-resolution obtained is  In general, the Z-resolution of a

determined by the number of recorded tomogram is approximately 10 times better
projection images and the section then in a single projection image.
thickness.
 The 3-D digital volume: The object

can be sampled with image processing
tools, e.g., segmented in any direction
one wishes without being hampered
by information from overlying
structures in the region of interest.

Figure 24.17
A. The information in a projection
image of a 300 nm-thick plastic
section displays low Z-resolution and
is obscured by overlying structures
and, therefore, difficult to interpret.
 Note: The increased resolution in all

directions and the amount of detail present
in the slice.
B. Single optical slice extracted from

the 3-D reconstruction of the thick
section (300 nm) shown above.
N = Nucleus
ER = Endoplasmic reticulum with
ribosomes on the membranes
G = Golgi
S = Spindle
5.2. Disadvantages of 3-D Electron Tomography

 The current types of specimen holders  Due to this limitation of the angular tilt
prevent tilting the specimen to angles range, there is a lack of information at the
higher than +/70o. higher tilt angles.
 In the Fourier transform of the

tomogram, the missing information
can be seen as a missing wedge of
information corresponding to missing
angular tilt range.

Figure 24.18 Schematic drawing of the

missing wedge information in Fourier
space. The gray area contains information
that is not present in the reconstruction due
to the limited tilt angles within an electron
microscope.
 Have to combine two perpendicular  The improvement in Z-direction is

tomograms into a dual-axis tilt significant and results in a more isotropic
reconstruction.26,27 resolution throughout the tomogram.
 No real 3-D labeling is possible.  Many EM studies are performed on

resin-embedded samples. This usually
provides good morphological quality, but
prevents specific localization studies
because the penetration of antibodies into
the section is not possible. So immuno-
labeling can be performed only at the
surface of the resin-embedded section.
Figure 24.19 Comparison of the same area

(X/Z) in a single and double-tilt tomogram.
In the single-tilt tomogram (top), a clear
distortion/elongation in the Z direction is
visible. The double-tilt tomogram (bottom)
shows improved Z-resolution.

6.1. Why Electron Tomography?

 With transmission electron microscopy  This makes it difficult to give unam-
of thin sections (40 to 70 nm), unique biguous answers to certain questions, like,
subcellular structures are studied and for instance, is there a connection between
high X, Y resolution can be obtained. organelles (direct open contact or MCS
However, many fine details (especially (membrane contact sides) or between sub-
in Z-direction) are obscured by over- compartments of an organelle (direct
and underlying structures in the connections between cisternae of a Golgi
section. system)?
 3-D transmission electron tomography
can provide unmistakable answers to
these questions with a resolution
especially in Z that is 10 times better
then obtainable with the classical
methods.
 By performing 3-D tomography, it  Resolution with light microscopy is at
was possible to gain new insight and best, 200 nm.
generate new hypotheses for many
biological processes, such as the
generation of the peroxisome,9 COP II
localization,42 connections between
cisternae in the Golgi39,43 transport in
the endo- and exocytic pathways,
nuclear organization and many more.
6.2. Why Electron Tomography on Resin Sections?

 Performing tomography on resin
sections has a number of advantages.
 First, the specimen preparation steps  Many EM labs can perform the basic
are relatively simple. steps of fixing (either chemical or the more
advanced high-pressure freezing followed
by freeze-substitution), resin embedding,
and sectioning and contrasting of sections.
 Second, the presence of abundant
amounts of heavy metals (uranium,
osmium, lead) in the section results in
a high signal-to-noise ratio (good
contrast), making it easy to recognize
structures and focus on the specimen.

 Third, it is even possible to determine  When enough specific gold labels are
protein localization on metacrylate- present on the section surface, they can be
embedded material. used for alignment of the projection images
and tomograms were specific labeling on
top of the section can be studied and
analyzed.
 Fourth, the presence of resin provides  Initial volume shrinkage can be up to

rigidity to the specimen. After the 30%.
initial shrinkage of the resin under the
electron beam, the section is rather
stable allowing the recording of a
large number of projection images
with small incremental values. This
results in a resolution that is  In cryo-EM, the total dose must be
impossible to obtain with cryo- divided over all the recorded images. This
tomographic methods. The total elec- results in fewer images with a much lower
tron dose that can be used on the signal-to-noise ratio (see Chapter 12).
specimen is much higher than in cryo-
tomography.
 Fifth, it is easy to reduce the so-called  The two computed tomograms can be
missing wedge artifact into a missing combined into one double-tilt tomogram.
pyramid by rotating the grid by 90o The resolution (especially in Z-direction)
and recording a second tilt series of the benefits considerably from this. Rotating
same area of interest. The stability of can be done in the microscope (rotation
the specimen and the unique nature of high-tilt holder) or outside the microscope.
the object allow the microscopist to
find the region of interest after rotating
the grid.

7. OBSERVED RESULTS
 Figure 24.20  The whole process of doing 3-D electron

tomography is exemplified on a cellular
organelle (a multivesicular lysosome).
 A-C: Representative images of the

first stack of projection images
recorded from 65 to +65o with an
increment of 1o Shown are the
projection images at 60o (A), 0o (B)
and +60o (C).
 D-E: Images of the stack perpen-

dicular to the first one from the same
area at 60o (D), 0o (E) and +60o (F).
 G: The combined tomogram. Because

it is constructed from two
independently constructed tomograms,
there is a region that contains single-
tilt information (upper right part) and
one that contains double-tilt
information (lower left part).
 H-I: Two representative optical slices

from the double-tilt tomogram
illustrating the increased resolution of
the tomogram compared to the zero tilt
images in B and E.
 Figure 24.21 (see colour insert  A: Optical slice with some contours of
following page ) the different suborganellar structures in
the multivesicular lysosome. The
contouring was done with the IMOD
software package.
 B: A surface rendered model that was
made from the contouring of the
multivesicular lysosome. For demon-
stration, only a few of the subcellular
structures are visualized.


8. REFERENCES
1. Crowther, R.A., DeRosier, D.J., and Klug, A. The reconstruction of a three-

dimensional structure from its projections and its applications to electron
microscopy, P. Roy. Soc. Lond. A Mat., 317, 319, 1970.
2. Koster, A.J. et al. Perspectives of molecular and cellular electron tomography,
J. Struct. Biol., 120, 276, 1997.
3. Medalia, O. et al. Macromolecular architecture in eukaryotic cells visualized by
cryoelectron tomography, Science, 298, 1209, 2002.
4. Al-Amoudi, A., Norlen, L.P.O., and Dubochet, J. Cryo-electron microscopy of
vitreous sections of native biological cells and tissues, J. Struct. Biol., 148, 131,
2004.
5. Leis, A.P. et al. Cryo-electron tomography of biological specimens, IEEE Signal
Proc. Mag., 23, 95, 2006.
6. Frangakis, A.S. and Förster, F. Computational exploration of structural
information from cryo-electron tomograms, Curr. Opin. Stuct. Biol., 14, 325, 2004.
7. Frank, J. Electron Tomography: Three-Dimensional Imaging with the
Transmission Electron Microscope. Plenum Press, New York, London, 1992.
8. Geerts, W.J.C. et al. Electron microscopy tomography and localization of proteins
and macromolecular complexes in cells, in Protein-Protein Interactions. A
Molecular Cloning Manual, Golemis, E.A. and Adams, P.D., eds,. Cold Spring
Harbor Laboratories Press, New York, USA, 2006, 715.
9. Geuze, H.J. et al. Involvement of the endoplasmic reticulum in peroxisome
formation, Mol. Biol. Cell, 14, 2900, 2003.
10. Horowitz, R.A. et al. The three-dimensional architecture of chromatin in situ:
Electron tomography reveals fibers composed of a continuously variable zig-zag
nucleosomal ribbon, J. Cell Biol., 125, 1, 1994.
11. Marsh, B.J. et al. Direct continuities between cisternae at different levels of the
Golgi complex in glucose-stimulated mouse islet beta cells, Proc. Nat. Acad. Sci.
USA, 101, 5565, 2004.
12. Murk, J.L.A.N. et al. Influence of aldehyde fixation on the morphology of
endosomes and lysosomes: Quatitative analysis and electron tomography, J.
Microsc., 212, 81, 2003.
13. Tomova, C. et al. New comprehension of the apicoplast of Sarcocystis by
transmission electron tomography, Biol. Cell, 98, 535, 2006.
14. Marsh, B.J. Reconstructing mammalian membrane architecture by large area
cellular tomography, Methods Cell Biol., 79, 193, 2007.
15. Midgley, P.A. et al. Probing the spatial distribution and morphology of supported
nanoparticles using rutherford-scattered electron imaging, Angew. Chem. Int. Ed.
Engl., 41, 3804, 2002.
16. Midgley, P. and Weyland, M. 3D electron microscopy in the physical sciences:
The development of Z-contrast and EFTEM tomography, Ultramicroscopy, 96, 413,
2003.
17. Ziese, U. et al. Three-dimensional localization of ultrasmall immuno-gold labels by
HAADF-STEM tomography, J. Struct. Biol., 138, 58, 2002.
18. Yakushevska, A. et al. Comparison of TEM and STEM tomography in cell
biology, J. Struct. Biol , 159, 381, 2007.
19. Otten, M.T. et al. High-angle annular dark-field STEM imaging of immunogold
labels, Scanning, 14, 282, 1992.
20. Stierhof, Y.-D. et al. Use of TEM, SEM, and STEM in imaging 1-nm colloidal
gold particles, in Immunogold-Silver Staining: Principles, Methods, and
Applications, Hayat, M.A., ed., CRC Press, Boca Raton, FL, USA, 1995, 97.
21. Leapman, R.D. Detecting single atoms of calcium and iron in biological structures
by electron energy-loss spectrum-imaging, J. Microsc., 210, 5, 2003.

22. Luther, P.K. Sample shrinkage and radiation damage, in Electron Tomography.
Three-Dimensional Imaging with the Transmission Electron Microscope, Frank, J.,
ed., Plenum Press, New York, London, 1992, 39.
23. Gilbert, P. Iterative methods for the three-dimensional reconstruction of an object
from projections, J. Theor. Biol., 36, 105, 1972.
24. Tong, J. and Midgley, P. A novel dual-axis reconstruction algorithm for electron
tomography, J. Physics: Conf. Ser., 26, 33, 2006.
25. Kamino, T. et al. Application of a FIB–STEM system for 3D observation of a
resin-embedded yeast cell, J. Electr. Microsc., 53, 563, 2004.
26. Penczek, P. et al. Double-tilt electron tomography, Ultramicroscopy, 60, 393,
1995.
27. Mastronarde, D.N. Dual-axis tomography: An approach with alignment methods
that preserve resolution, J. Struct. Biol., 120, 343, 1997.
28. Ziese, U. et al. Automated high-throughput electron tomography by pre-calibration
of image shifts, J. Microsc., 205, 187, 2002.
29. Zheng, Q.S. et al. An improved strategy for automated electron microscopic
tomography, J. Struct. Biol., 147, 91, 2004.
30. Mastronarde, D.N. Automated electron microscope tomography using robust
prediction of specimen movements, J. Struct. Biol., 152, 36, 2005.
31. Dierksen, K. et al. Three-dimensional structure of lipid vesicles embedded in
vitreous ice and investigated by automated electron tomography, Biophys. J., 68,
1416, 1995.
32. Koster, A.J. et al. Automated microscopy for electron tomography,
33. Fung, J.C. et al. Toward fully automated high-resolution electron tomography,
J. Struct. Biol., 116, 181, 1996.
34. Kremer, J.R., Mastronarde, D.N., and McIntosh, J.R. Computer visualization of
three-dimensional image data using IMOD, J. Struct. Biol., 116, 71, 1996.
Struct. Biol., 152, 92, 2005.
38. Murk, J.L.A.N. et al. Endosomal compartmentalization in three dimensions:
Implications for membrane fusion, Proc. Nat. Acad. Sci. USA, 100, 13332, 2003.
39. Trucco, A. et al. Secretory traffic triggers the formation of tubular continuities
across Golgi sub-compartments, Nat. Cell Biol., 6, 1071, 2004.
40. Murk, J.L.A.N. 3-D analysis of endosomes, lysosomes and peroxisomes. PhD
thesis, Utrecht University, Utrecht, The Netherlands, 2004.
41. Lebbink, M.N. et al. Template matching as a tool for annotation of tomograms of
stained biological structures, J. Struct. Biol., 158, 327, 2007.
42. Zeuschner, D. et al. Immuno-electron tomography of ER exit sites reveals the
existence of free COPII-coated transport carriers, Nat. Cell Biol., 8, 377, 2006.
43. Polishchuk, R.S. and Mironov, A.A. Structural aspects of Golgi function, Cell.
Mol. Life Sci., 61, 146, 2004.

FINAL CONSIDERATIONS
The publication of this handbook is a wonderful satisfaction for a scientist engaged
throughout his long life in the development of instrumentation and methodologies for
research in biological ultrastructure. It is proof that some dreams may become reality and
that “cryo” is still in a dynamic and fruitful state. twenty years ago, Steinbrecht and
Zierold10 edited the first compilation in this field. Their book already demonstrated the
incredible advances since the early 1950s. At that time, high resolution transmission
electron microscopes were available for research.3,7 The first chapters of this book
document the current state of the art. It’s a story of success. A dream became reality; the
dream to look directly by microscopic observation inside the basic structures of life. The
way to do this starts with the proper vitrification of macromolecules, cells in suspension
or tissues. Both instrumentation and methodology for freezing, post- and prefreezing steps
of such preparations have now reached a state of perfection, which makes it almost an
easy task and at the very least a feasible one.
Progress in cryo-technology had an important side effect for science — the tiny water
molecule was brought again into the forefront. It was astonishing that most scientists
— even biologists — for a long period of time had no real interest in this simple
compound, albeit with a complex behaviour, made of two hydrogen and one oxygen
atom. Really astonishing, because it was obvious and well known that water molecules
represent a major component of our huge universe. Astonishing also because more than
half of our earth’s surface is covered by oceans. Astonishing with respect to the
community of biologists because any student knows that cells, plants and animals,
including man, contain more water than other components. Nevertheless, water was
ignored successfully. It was hampering structural studies in the light microscope and had
to be removed completely to allow wax (paraffin) embedding. A good penetration of the
dehydrating agent was only possible after poisoning the cells to make their membranes
leaky and penetration possible. Those were useful methods, but also a solid basis for
plenty of artefacts.
Concerning ultrastructural research, it was obvious that proper “vitrification” (see
Chapter 1) within milliseconds was the only way to stabilise all components of the
sensitive, mixed plasmatic phases, including also small molecules and rapidly moving
ions. But there was a general scepticism by the experts in physical chemistry as to
whether real vitrification of pure water and diluted aqueous solutions was actually
possible. In 1980/81, two independent teams led by Brüggeller and Dubochet
demonstrated this possibility.1,2 This was a breakthrough of great importance that opened
the door to answering a lot of questions. If water has no time to aggregate into crystals,
then all other components must also be held in their original position. Proper
vitrification, therefore, justifies the considerable effort carried out to develop cryo-
ultramicrotomy and perform diffraction analysis of ultrathin cryo-sections in the
“frozen-hydrated state.” That now really makes sense.9

Most progress discussed above was achieved against all odds. Similar obstacles were
encountered in the development of the electron microscope. Gabor, as a member of the
team around Knoll at the Techniche Hochschule Berlin, wrote the most impressive and
competent report about the early days of this development.3 I cite in the original language:
“Niemand konnte anhand der damals verfügbaren Unterlagen vorausahnen, dass
das Elektronenmikroskop so erfolgreich sein würde, wie es sich später erwies”. He
continued, “Wir wollen im Voraus die Moral aus dieser Geschichte ziehen: Selbst in
der Wissenschaft ist es oft besser, Mut zu besitzen, als gescheit zu sein”. Translated
into English: “At that time, nobody could according to the available data predict that the
electron microscope would be such a success as it turned out later on” and “We draw in
advance the conclusion from this story: even in science, it is often better to be courageous
than to be very intelligent.” I noticed with much pleasure the cynical comments of all the
honourable old professors in scientific journals at the time. They widely discussed this
useless crazy new instrument. These comments would be worth a historical study — as an
example for the old slogan, “errare humanum est.”
I would like to recall a few other examples of such errors:

a. Ultramicrotomy: No one believed it would be possible to reproducibly cut ultrathin
sections in the nanometer-range suited for the EM. Neither precision mechanics nor
knife materials nor fixatives nor embedding media seemed to be achievable.
Nevertheless, among others, Porter and Blum6 did it.
b. High-pressure freezing: Moor and Riehle5 were successful. I cite. Moor4: “The
interest of the audience was not overwhelming because everybody thought that this
approach is over sophisticated and unnecessary” – what an error! High pressure
freezing provided the first pictures of well-frozen ultrastructures and a world of new
information.
c. Finally, my own experience in cryo-ultramicrotomy:9 Nobody wanted to believe that
sectioning at 180°C would be successful. A group of smart, experienced cryo-
technologists from industry visited our prototype FC4 and stated that such a set-up
with direct LN2 cooling and an open top could never work (it was already running
and not so badly). One of these fellows stated that we probably had manipulated the
temperature indicators. He wet his fingertip and felt if the knife holder was really
cold (it was approximately 150°C). I was silent and smiling when he yelled because
he suffered a severe burn.
I noticed something else, all these “crazy developments, which never would work” were
achieved by rather young scientists, often students. The first EM was built by a “young
crew” around the also youthful assistant. Knoll. His team was made up only of students of
the TH Berlin. Jacques Dubochet2 was young as well as his team at the EMBO,
Heidelberg. Nevertheless, they achieved and documented proper vitrification of pure
water. Porter and Blum6 were young fellows when they built (at the Rockefeller Institute
in New York), the first ultramicrotome, which was clearly suited for routine work. That
was the Volkswagen in ultramicrotomy and served generations of scientists over decades,
an astonishingly simple apparatus. Moor and Riehle5 as young scientists also did an
excellent job at the ETH Zürich. I started as a young student in the third semester on an
EM and built my first ultramicrotome.8 I wanted, together with my brother Peter (also a
student), to have a look into cells. In general, youngsters have the courage necessary to
enter into risky projects and have the drive needed for success. The slogan:“Doing

the Impossible” fits to their kind. One should be careful with this precocious resource
and try to keep these youngsters instead of firing them after some years of successful
engagement. The United States tenure track system seems a good solution, which we lack
in Europe.
I want to make some further comments on “cryo” analyses (and on expensive research in
general): “Enthusiasm cannot substitute for Money.” This means that efficient cryo-
work in many —if not most — sectors needs some money. Electron microscopes
(equipped with a cryo-stage, low dose, possibly filter lenses), cryo-ultramicrotomes (with
cryo-diamonds, antistatic devices, etc.), a high-pressure freezer, cryotransfer units,
automated freeze-substitution and freeze-drying units are not available for nothing. The
same applies to running costs. This is a threat for universities and grant-giving
organisations. Money is rare — especially in Europe and the Southern Countries. This is a
severe problem for cryo.
Our colleagues in the US and Canada are in a splendid situation. They also complain that
the budget decreases every year. But a true comparison reduces my compassion
immediately to nothing. Certainly there are some low-cost set-ups like home-built
plunging systems working with a rubber belt like David’s catapult to kill Goliath. But
such instruments are rather the exception. There are some examples in which lack of
money may provoke inventions that end in a Nobel Prize award. The one most often
cited is about the Austrian analytical chemist Fritz Pregl who earned the Nobel Prize for
the foundation of quantitative organic microanalysis in 1923. It was said that he was
forced to this discipline simply by lack of money. I do not believe this. The same rumour
concerns the discipline of theoretical physics in Göttingen (Germany). Maybe, but I want
to warn governments from reducing budgets of universities and research organisations to
increase the number of Nobel Prizes that such an inverse relationship will certainly not
work. It would also be dangerous to split the old fashioned universitas litterarum into
expensive and inexpensive disciplines. Then cryo will come to an end together with other
expensive disciplines.
Markl, former president of the big German research organisation (Deutsche

Forschungsgemeinschaft) stated in an official meeting about the financial situation, that
we must avoid supporting, for financial reasons, more low-cost projects. Then we would
provoke the degeneration of research into “Mickey Mouse research.” I guess personally
that plenty of Mickey Mouses are already at work in our scientific community. These are
scientists who calculate exactly the probability of success and the time needed. Only
100% success rates and short periods of time are wanted. They should leave the university
and change over to a commercial employment. In my opinion, they will not do this and
also will not be wanted in this alternative area — sorry. But, I again warn governments.
There is a pitfall, a kind of sin. The “splitting of money into fractions of “useful”
(applied) and “useless” (basic) research” is a horrible mistake. Useful research is most
often light-weight research that serves commerce and governments. Useless research is
heavy-weight research, which has no immediate visible and predictable influence on
commerce and, therefore, income taxes for the government. The claim is that this useful
research increases the welfare of all citizens. Most citizens feel that this is a striking
argument. The truth is just the contrary.

The economic success in the United States resulted to a great extent from the generous
support of the so-called “useless” basic research. Only such basic research opens
completely unexpected new areas for commerce, industry and other avenues in a country.
Internationally, well-known scientists made very clear statements and warned the
Austrian Government. I cite two colleagues working in Vienna in prominent positions. (1)
Freissmuth, pharmacologist and director of a big Institute of the Medical University in
Vienna, stated, “You cannot save (basic) research.” (2) Penninger, pathologist and
director of the Institute of Molecular Biotechnology (Austrian Academy of Sciences) just
recently stated in an official discussion of the broadcast ORF1, “If the government saves
money on basic research it will change our university into a museum.” This is
perfectly valid for cryo-work. With an old EM, a historical ultramicrotome and an old
evaporation unit, you may impress plenty of visitors. But you are unable to carry out
state-of-the-art cryo-work. You degenerate into a “Mickey Mouse,” according to Markl.
Since I wish all young colleagues a satisfying and fruitful career in this fascinating field, I
have made some comments concerning a sufficient amount of money for this type of
research. I repeat: “Enthusiasm in most situations cannot substitute for money.” A strong
financial basis is of key importance for success besides courage, intelligence and the
always needed drive. In this sense, I wish all fellow workers, especially the young ones,
“Good luck in the fascinating cryo-work.”
REFERENCES
1. Brüggeler, P. and Mayer, E. Complete vitrification of pure liquid water and

diluted aqueous solutions, Nature (London), 288, 569, 1980.
microscopy, J. Microsc. (Oxford), 124, RP3, 1981.
3. Gabor, D. Die Entwicklungsgeschichte des Elektronenmikroskopes, Elektrotechn.
Zschr., 78, 522, 1957.
4. Moor, H. Theory and practice of high pressure freezing, in Cryotechniques in
Biological Electron Microscopy, Steinbrecht, R.A. and Zierold, K., eds., Springer
Verlag, Berlin, Heidelberg, Germany, 1987, 175.
5. Moor, H. and Riehle, U. Snap-freezing under high pressure: A new fixation
technique for freeze-etching, in Proc. 4th Europ. Reg. Conf. Electron Microsc.,
Bocciarelli, S., ed., 1968, 33.
6. Porter, K.R. and Blum, J. A study in microtomy for electron microscopy, Anat.
Rec., 117, 685, 1953.
7. Ruska, E. Die frühe Entwicklung der Elektronenlinsen und der
Elektronenmikroskopie, Acta historica Leopoldina Nr. 12. Deutsche Akademie der
Naturforscher, Halle/Salle, 1979.
8. Sitte, H. Ein einfaches Ultramikrotom für hochauflösende elektronen-
mikroskopische Untersuchungen, Mikroskopie (Wien), 10, 365, 1955.
9. Sitte, H. Advanced instrumentation and methodology related to cryo-
ultramicrotomy: A review, Scan. Microsc. Suppl., 10, 387, 1996.
10. Steinbrecht, R.A. and Zierold, K. Cryotechniques in Biological Electron
Microscopy. Springer Verlag, Berlin, Heidelberg, Germany, 1987.

Glossary 659
A
 Algebraic reconstruction technique  A reconstruction algorithm that
(ART) compares the differences between the
reprojections of a tomogram and the
measured data, one image at a time, and
corrects the volume iteratively until a
given stopping criterion is fulfilled.
 Amorphous ice  Solidified water devoid of crystals, also
called vitreous ice.
 Analytical probes  Electron beam or ion beam that interacts
with specimens are analytical probes.
 Antibody  Immunoglobulins that recognize
specific antigens.
 Antifreezing agent  A substance that is added to water to
lower its melting point.
B
 Bar  Measure of pressure:
1 Bar = 105 Pa = 752.5 Torr = 0.99 atm.
 Basic calibrations  Measured curves describing the
movement of a high tilt holder in x, y and z
during a tilt range.
 Blocking buffer  Used to prevent undesired binding of
antibodies to biological structures and resin
(usually interactions of electrostatic and
hydrophobic nature).
 Blotting  Removal of excess fluid from the grid
prior to vitrification in liquid ethane.
 Boltzmann constant, k  The physical constant relating temp-
erature to energy; experimentally
determined as 1.38 × 10 to 23
joules/Kelvin.
C
 Celsius  Anders Celsius 1701–1744, Swedish
astronomer, devised the Celsius (oC)
temperature scale (Dictionary of Science
and Technology, Academic Press, San
Diego, CA, 1992).
 CEMOVIS  Technique allowing observation of
sections of fully frozen-hydrated samples
under an electron beam (in a cryo-electron
microscope).

 Charge-coupled device (CCD)  An image sensor comprising an

integrated circuit and a linked array of
radiation-sensitive capacitors.
 Chatter  One of the major cutting artifacts. It is a
periodic variation of section thickness at
the µm scale. It may be due to vibration of
the knife in respect of the specimen. In
vitreous cryo-sections, it is generally due
to variation of friction of the section on the
knife surface.
 Chemical mapping  Construction of 2– or 3–dimensional
maps of the chemical composition of
specimens for one or several elements.
 Critical micelle concentration  The CMC corresponds to the minimum
(CMC) concentration of detergent at which
micelles form. The CMC is very sensitive
to temperature and polarity of the medium.
 Colloidal gold marker  Colloidal gold bound to antibody
(fragments) via hydrophobic and
electrostatic interactions.
 Contouring  Manual drawing of contour lines in
slices of a tomogram.
 Contrast transfer function  Function associated to the electron
magnetic lens and which describes the
proportion of signal that is transferred by
the imaging system for each spatial
frequency. The CTF is an oscillating
function that varies with defocus value and
the spherical aberration of the objective
lens.
 Conventional  A useful term to designate anything
except the present subject. In this book,
conventional electron microscopy is non-
cryo-electron microscopy. More specifi-
cally, a conventional preparation method
means chemical fixation and resin
embedding.
 Correlative light microscopy (LM)  Combination of observing and imaging
and electron microscopy (EM) with light microscopy before detailed
analysis at high resolution with electron
microscopy.
 Crevasse  One of the major cutting artifacts
induced by cutting stress in vitreous
samples.
 Crowther criterion  A simple formula that defines the
theoretical resolution of a tomogram only
in terms of the total number of projection
images and the tilt range, which by
definition also defines the tilt increment.

Glossary 661
 Cryo-fixation or  Solidification of a biological specimen

cryo-immobilization by cooling with the aim of minimal
displacement of its components.
 Cryogen  A liquid that boils at extremely low
temperature, e.g., liquid nitrogen (boiling
point 196°C). Liquid nitrogen is a pri-
mary cryogen. Other cryogens, such as
liquid ethane, are referred to as secondary
cryogens because they are cooled by liquid
nitrogen.
 Cryogenic  Pertaining to preservation and/or
maintenance of molecules, cells or organ-
isms at extremely low temperatures,
typically in liquid nitrogen.
 Cryo-section according to Tokuyasu  Ultrathin thawed frozen section.
Biological samples are weakly fixed with
aldehydes, protected against freezing
damage by infiltration with sucrose, frozen
in liquid nitrogen and sectioned at 90°C
to 120°C by cryo-ultramicrotomy
(usually used for immunolabeling.
 Cryo-sectioning  Production of frozen-hydrated sections
for observation in a cryo-electron
microscope (see CEMOVIS).
D
 Dehydration  Free water from the fixed sample must
be removed when water immiscible resins
are used for embedding.
 Dehydration of cryofixed samples
 Freeze drying through the sublimation of ice
(lyophilisation). Starting temperature of
procedure: 100°C.
 Freeze-substitution  Dehydration of cryofixed samples by
using an organic solvent carried out below
90°C.
 Depth-of-field  The distance in object space over which
the objective focused on the specimen can
provide adequate definition or clarity
(Dictionary of Science and Technology,
Academic Press, San Diego, CA, 1992).
 Detergent  Amphiphilic molecules used to
solubilize membrane proteins.
 Devitrification  Crystallization of water by warming a
vitrified sample. For pure water, the
devitrification temperature is ca. 135°C.
It is higher when the solute concentration
is high.

 Dialysis  Mechanism allowing removal of some

components from a solution.
 DOGS-NTA-Ni  Synthetic lipid formed by a
hydrophobic domain and a hydrophilic
domain containing chelated nickel.
 Dose fractionation  The process of distributing the total
electron dose that can be tolerated by a
specimen between the individual pro-
jection images (cryo-electron tomography).
E
 EELS  Among other structural information, the
electron energy-loss spectrum contains
chemical information about the specimen
in the form of specific edges (with specific
shape and position in energy) super-
imposed on a decreasing background.
 EFTEM  Energy filtered transmission electron
microscopy can be used for the
measurement of the elemental composition
of irradiated specimens. This analysis is
possible by the use of an imaging energy
filter integrated into the microscope
column or by using a prism spectrometer
below the final screen.
 Elvanol/Mowiol  Semisolid embedding medium for light
microscopy.
 Energy filtered image  See ESI.
 Epi-fluorescence microscope  The illumination light of the object is
using the same light path through the
objective as the emission signal.
 Epon  Epoxy resin.
 Epoxy resin  Chemically: polyaryl esters of glycerol
with terminal epoxy groups and hydroxyl
groups spaced along the chain. With
addition of cross-linking agents, they are
converted into an inert solid.
 ESI  Generates images with electrons of
selected energy loss. These data concern
only element-specific electrons = imaging
for one element.

Glossary 663
 Etching  Sublimation of a solvent for a short

period of time.
 Ethane  In a liquid form, ethane is the ideal
vitrification coolant.
F
 Fiducial marker (or “fiducial,”  An exogenous, inorganic colloidal
slang) particle, usually 10 nm colloidal gold,
introduced into or onto a specimen for the
purposes of facilitating subsequent
alignment of projection images.
 Fixation  The goals of fixation are to preserve the
structure of samples with minimum
alterations from the living state.
 Fluorochrome  A chemical component, which upon
illumination radiates light at a longer wave
length.
 Formaldehyde  Weak fixative that has little influence
on immunolabeling efficiency.
 Fourier space  Synonymous of reciprocal space; the
“space” defining the Fourier transform of
an object, i.e., its decomposition into a
continuous spectrum of its component
frequencies, as opposed to Euclidean or
“real” space, where positions are defined in
terms of an x, y, z coordinate system.
 Freeze-drying  A technique by which the frozen water
of a cryo-fixed specimen is removed by
sublimation at low temperature in a
vacuum chamber.
 Freeze-etching  After breaking the specimen and before
evaporating metal, the specimen undergoes
a process of etching, i.e., ice sublimation,
for a short period of time.
 Freeze-fracturing  To break a frozen specimen into pieces
and immediately make a heavy metal
replica of the fractured plane.

 Freeze-substitution  Dehydration method by replacing water

(in solid state) for a (fluid) solvent (mostly
methanol, ethanol or acetone). After freeze
substitution, the sample can be warmed up
to room temperature (i.e., the process to
dehydrate and chemically fix specimens at
temperatures between 90°C and 30°C).
G
 Gaussian denoising  Denoising of a signal with the
assumption that the noise has a Gaussian
distribution, or in other words, that it is
“white noise.”
 Glow discharge  Procedure by which ionized atoms are
deposited onto a carbon support to modify
its surface properties. The carbon-coated
grids are deposited onto an electrode that is
used to ionize residual gas under vacuum.
 Glutaraldehyde  Aliphatic dialdehyde, an efficient cross-
linking fixative. It may considerably
reduce immunolabeling efficiency.
 Gold enhancement  Similar to silver–enhancement. Small
gold particles increase size by gold atoms,
instead of silver atoms, deposited onto the
surface.
 Gold toning  Treatment of silver–enhanced structures
with gold chloride. Originally used to
improve contrast in LM and to
cover/enlarge silver enhanced gold.
 Green fluorescent protein  Protein from the jelly fish Aequorea
victoria. GFP is a marker of gene
expression and protein targeting. It is used
for live cell imaging.
H
 H+-ATPase  An ion pump that actively transports
hydrogen ions across lipid bilayers in
exchange for ATP.
 HC-Pro  The helper component proteinase is
encoded by plant virus of the genus

Glossary 665
Potyvirus. HC-Pro is involved in different

steps of the viral cycle, aphid transmission,
replication, and virus cell-to-cell
movement and is a suppressor of post-
transcriptional gene silencing.
 High-pressure freezing  Rapid cooling of a sample at
204.8 MPa.
I
 Immunoadsorption  Purification of antibodies by adsorption
on the specific antigen often bound to a
sepharose column.
 Indirect labeling  The primary antibody is not coupled to
a marker molecule. The secondary
antibody, which is used to detect bound
primary antibodies, is coupled to a marker
molecule.
K
 Kelvin  Baron William Thomson Kelvin 1824-
1907), British physicist and
mathematician, devised the Kelvin (K)
temperature scale (Dictionary of Science
and Technology, Academic Press, San
Diego, CA, 1992). The international
standard unit of temperature where 0°
Kelvin is equivalent to 273°C.
L
 Lipid monolayer  Layer of lipids formed at the interface
buffer-air.
 Low-dose  Microscope operation mode with a
beam-deflection unit that allows the
precise determination of the total exposure
dose on the specimen during data
collection. Typically between 10 to 20
electrons/Å2.

 Lowicryl resin  A mixture of acrylate-metacrylate resins

with very low viscosity at low temp-
eratures (up to 80°C).
 Low-temperature embedding  Embedding in methacrylates, typically

Lowicryls, at temperatures (below 0°C) by
UV irradiation.
 LR White, LR Gold  London Resin, brand name for
methacrylates.
M
 MAPs  Proteins that interact with microtubules
in the cell, regulating their dynamic
behavior or allowing them to build
complex organelles, such as axonemes and
centrosomes.
 Marker  Necessary to visualize the antigen-
antibody reaction by light microscopy
(fluorescent molecules and enzymes) or
electron microscopy (usually gold
particles).
 Methyl cellulose  Embedding medium for ultrathin
Tokuyasu cryo-sections.
 Microtubules  Polymers of the tubulin molecule
associated with proteins (MAPs) in the
cell, and involved in various functions,
such as cell division through the mitotic
spindle, vesicular traffic through the action
of molecular motors, cell compart-
mentalization, or cell motility through the
motion of cilia and flagella.
 Missing pyramid  Term used to describe the region of
Fourier space that remains unsampled in a
dual-axis tilting experiment, defined by
orthogonal tilt axes. By combining two
perpendicular recorded tomograms of the
same region, the missing wedge can be
reduced to a missing pyramid.
 Missing wedge  Missing information in Fourier space,
due to limited tilt angles. Results in
anisotropic resolution in X,Y,Z.
 Molecular distillation drying  A special method of freeze-drying.

Glossary 667
N
 NANOGOLD marker  Commercially available gold compound
that can be covalently bound to antibodies
and other molecules.
 Neurospora crassa  A type of red bread mold of the phylum
Ascomycota.
 Numerical aperture  Measure for the resolution power of an
objective:
 NA = n × sin α
 n = Refraction index of the medium
between sample and objective lens
 α = Half opening angle of the
objective.
 The correlation with the resolution d
is:
 d = 0.61λ/NA
 λ = The wave length of the
electromagnetic wave.
O
 Optimized position  Theoretical positions in a perfectly
adjusted specimen stage where the distance
between the optical axis of the microscope
and the tilt axis of the specimen is zero.
 Osmication  A method to stain and fix biological
specimens with the vapors from osmium
tetroxide crystals or with an osmium
tetroxide solution.
 Osmium tetroxide  Strong fixative, but also contrasting
agent especially for membranes. It may
degrade proteins and, therefore, has a
strong effect on immunolabeling
efficiency.
P
 Phosphotungstic acid  Heavy metal chemical used for negative
staining.
 Photobleaching  Loss of intensity of the fluorescence
signal during imaging. Fluorochromes are
affected by radical oxygen species
produced upon illumination.

 Primary antibody  Antibody specific to the antigen/epitope

under investigation (usually raised in rabbit
or mouse).
 Protective colloid  Slows down silver enhancement and
makes the reaction more efficient and
reproducible.
 Protein A  In EM, often used instead of secondary
antibody. Binds to a number of IgGs from
different species.
Q
 Quantum dot (QD) marker  Fluorescent semiconductor nanocrystal
covalently bound to antibodies. Most QDs
exhibit relatively low electron density, but
can be silver-enhanced.
R
 Replica  The metal imprint or cast of the surface
of an object.
 Resin embedding  Dry specimens are fragile and porous
and must be permeated with a fluid resin
that can polymerized resin for conservation
and sectioning.
 Resin polymerization  Resin can be polymerized by heat or by
UV radiation with addition of accelerating
agents: cross-linking and catalysts.
 Rotary shadowing  The specimen is tilted under a chosen
angle and rotated while shadowing.
S
 Secondary antibody  Antibody (usually raised in goat)
coupled to a marker molecule that detects
the bound primary antibody.
 Segmentation  The delineation of the features of a
complex, three-dimensional image, either
manually using a mouse or drawing tablet,
or automatically by means of appropriate,
e.g., edge-detection, algorithms.

Glossary 669
 Shadowing  To evaporate a heavy metal from a

point source at an oblique angle onto a
specimen surface in order to produce a cast
of the specimen.
 Silver enhancement  Process similar to photographic
development. In the presence of a reducing
agent, the gold surface acts as a catalyst for
the reduction of silver ions to metallic
silver. The metallic silver deposits on the
gold surface, resulting in a growing silver
layer.
 Silver stabilization  Protection of the silver layer that is
sensitive to oxidation (OsO4, electron
beam, air humidity), by gold chloride
treatment.
 SIMS (SIMS imaging)  SIMS is a mass spectrometric technique
using a primary ion beam to analyze the
specimen. Chemical and isotopic micro-
analyses can be carried out by focusing this
primary beam to get a microprobe
(diameter under 100 nm). SIMS imaging is
obtained by scanning such a microprobe
over the surface of interest of the
specimen.
 In TOF-SIMS, the primary ion beam is
pulsed and a whole mass spectrum can be
analyzed. This imaging method gives
molecular information, but is less sensitive
than dynamic SIMS imaging.
 Simultaneous iterative  A reconstruction algorithm similar to
reconstruction technique (SIRT) ART (see above) where the reconstructed
volume is updated only after all corrections
have been performed.
 Slam-freezing  Cryo-fixation technique where the
surface of a biological sample is cooled
down by the polished surface of a metal
block, which, in turn, is cooled by the
cryogen.
 Spurr’s resin  One of the most fluid resins used in
electron microscopy (about 60 cP at am-
bient temperature for fresh mixture). Easily
sectioned and resistant under the analytical
probes.

T
 Tomogram  Computed 3-D volume reconstruction
of a specimen by using as many projection
images as possible (usually between 121
and 141 images).
 Tomography  The process of obtaining a three-
dimensional image volume (tomogram)
from a series of two-dimensional images,
represented either by x-y slices, (e.g.,
confocal microscopy) or projections, (e.g.,
x-ray tomography or electron tomography).
 Torr  Old unit for measuring vacuum still
frequently used in electron microscopy.
Derived from an Italian physicist and
mathematician, Evangelista Torricelli
(1608–1647), who invented the mercury
barometer and was the first to create
vacuum. 1 Torr = 1.33 × 102 Pa or N/m2
(SI unit for pressure).
 Transmission fluid  A fluid used to mediate transfer of heat
and pressure during high-pressure freezing
without interacting with cellular
specimens.
 Triton X100  A nonionic surfactant that has a
hydrophilic polyethylene oxide group and
a hydrocarbon or hydrophobic group.
U
 Uranyl acetate  Fixative and contrasting agent. The
divalent uranyl cation forms salt bridges
with negatively charged groups, e.g.,
phosphate groups, hence, stabilizing
membranes.
V
 Vitreous  From the Latin, literally “glass-like”,
often used interchangeably with “amor-
phous.”

Glossary 671
 Vitrification  The formation of a non-crystalline

(vitreous) solid state.
 Weighted back-projection (WBP)  Synonymous of filtered back-

projection; the standard method of
reprojecting aligned, two-dimensional
projection images into a three-dimensional
volume; the weighting factor is used to
account for differences in low- and high-
frequency information in Fourier space.
 Well-freezing/well-frozen  Term to describe the state of solid water

in a biological object. A sample is said to
be well frozen if ice crystal ramifications
formed cannot be seen with the electron
microscope.
 Whole mount labeling  Immunolabeling of large samples that
have been permeabilized and extracted
(e.g., by detergent or solvent treatment)
prior to antibody and marker incubation to
allow their penetration into the sample.

Handbook of Cryo-Preparation Methods For Electron Microscopy (Methods in Visualization) (Annie Cavalier, Daniele Spehner Etc.

Uploaded by

Copyright:

Available Formats

Handbook of Cryo-Preparation Methods For Electron Microscopy (Methods in Visualization) (Annie Cavalier, Daniele Spehner Etc.

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Handbook of Cryo-Preparation Methods For Electron Microscopy (Methods in Visualization) (Annie Cavalier, Daniele Spehner Etc.

Uploaded by

Copyright:

Available Formats

What are some techniques discussed for preparing samples?

What are some techniques discussed for preparing samples?

What are some imaging techniques mentioned?

What are some imaging techniques mentioned?

HANDBOOK OF

© 2009 by Taylor & Francis Group, LLC

Genome Visualization by Classic Methods